Crojfe 2 2017 kb

B. Talbot et al.

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

8

338

2

2017

Croat. j. for. eng. 38(2017)2

Editorial

50 Years of FORMEC International Network Respected readers, distinguished colleagues, dear friends, This issue of Croatian Journal of Forest Engineering (CROJFE) is dedicated to a special and very important anniversary. In the year 1966, in the former Czechoslovak Republic, the first conference entitled »Forestry Mechanization« was held, and ever since 1994 these conferences have been organized under the acronym FORMEC (Forestry Mechanization). This year, in Romania, Brasov, the FORMEC Conference celebrates a 50-year jubilee. Through the following pages, closer history will be presented, as well as the current situation and possibly the near future of the international FORMEC Network. Mutual co-operation between the FORMEC International Network and CROJFE journal will also be presented, as well as the current thematic issue of CROJFE, one of the world-leading journals exclusively specialized in articles dealing with forestry engineering.

The initiators of the idea of organizing »Forestry Mechanization« conferences were distinguished professors: Stevan Bojanin (Croatia), Franz Hafner (Austria), Jozsef Kaldy (Hungary), Amer Krivec (Slovenia), Gabor Pankotai (Hungary), Ernst Pestal (Austria), Hans Bruno Platzer (Germany), Eugen Rónay (Slovakia) and others. In the beginning, the official language of the conference was German. Scientists from the countries of Central and Eastern Europe, whose field of scientific research was related to forestry engineering, could have used Forestry Mechanization network as a platform for sharing ideas, knowledge and research, all in order to overcome differences in the technological development and the level of mechanization of timber harvesting operations. In the beginning, the number of participants was between 20 and 40, and the conference host (organizer) covered all the costs of organization except the participants’ travel expenses.

At the Conference held in 1999 in Zalesina (Croatia), the introduction of English, as a second official conference language, was intensively discussed, with Croat. j. for. eng. 38(2017)2

the main aim of popularizing FORMEC among younger colleagues, scientists, teachers and operational foresters, all involved in forestry engineering. The idea was accepted after a lengthy discussion, but it took a while to get visible results of this great initiative. In 2003, FORMEC Conference was officially transformed into an international network of forestry engineering specialists. The president elected was, and still is, Professor Karl Stampfer, PhD. (BOKU, Vienna). For fifteen years, together with his BOKU team and a strong support of leading scientists and teachers in the field of forestry engineering, primarily from Europe but also beyond, step by step, he has been building, developing, and modernizing the FORMEC Network by adapting it to new circumstances (https://www.formec.org/).

Today’s FORMEC conferences are the largest annual event in the field of forestry engineering for all scientists, teachers and operational foresters involved in forest engineering. These are respectable events from the standpoint of forestry itself. The participants, now ranging between 150 and 200 (in Dubrovnik, Croatia, in the year 2012 even around 250), come not only from almost all European countries but also from all around the world. Every fourth year, FORMEC is organized in Austria, in combination with AUSTROFOMA – a vast presentation of forest machines in real working conditions. In 2019, the FORMEC Conference will be organized for the first time together with the annual COFE (Council of Forest Engineering) meeting (Vienna, Austria and Sopron, Hungary).

In 2015, an international scientific conference, entitled »Forest Engineering – Current Situation and Future Challenges« (CROJFE 2015), was held in Zagreb and Zalesina (Croatia). The Conference was held on the occasion of the tenth anniversary of the CROJFE journal and forty-eight years of publishing New Forestry Mechanization (Forestry Mechanization) journal.

Namely, in 1976, ten years after the first »Forestry Mechanization« Conference, a journal entitled »Forestry Mechanization«, a scientific and professional

151

Editorial (151–153)

journal dealing with a specialized field of forestry i.e. forestry engineering, was published in Croatia.

This journal played a very important and irreplaceable role, especially in the period of the most intensive mechanization of forestry operations in the Republic of Croatia in the seventies and eighties of the last century, primarily in the field of timber harvesting. Research results of domestic and foreign scientists and domestic forestry practitioners were presented, development of new forestry machinery of Croatian producers were presented and promoted, and in some way the introduction of different types of forestry machinery and vehicles was encouraged, directed and publicly followed, adapted to biodiversity of Croatian stand and habitat conditions. Guidelines were also given for a sustainable management in Croatian forests.

The first twenty years of publishing Forestry Mechanization journal is probably one of the best and most persistent examples of excellent cooperation between forestry science and practice, not only in Croatia but also in Europe. At the very end of the twentieth century, due to some objective and understandable reasons, but also some subjective and difficult-to-understand reasons, the problems in regular publishing of Forestry Mechanization journal occurred. Therefore, at the meeting of the editorial board and publisher of the journal at the end of 2004, it was concluded that publishing needed to be quicker (and more regular), and decision was made to launch two new journals in the year 2005: Croatian Journal of Forest Engineering (CROJFE) and New Forestry Mechanization (NMŠ), both successors of Forestry Mechanization journal.

Over the past 12 years of continuous development and improvement of CROJFE, there have been a lot of events that should be highlighted and which, each in their own way, have given a positive feedback to the journal itself. However, one of the most important ones – at the end of 2010, was a first two-year contract on cooperation between CROJFE and the FORMEC International Network, only to be followed by two three-year cooperation contracts, whereby FORMEC International Network became the co-publisher of the journal. Very good cooperation over the past seven years, where there is still space for improvement, is certainly one of the reasons why CROJFE is now one of the world’s leading journals in the field of forestry engineering.

During an informal conversation between several professors, including CROJFE’s Editor-in-Chief and

152

President of FORMEC Network, during the Danube boat trip in Linz (FORMEC 2015 Conference), the idea of CROJFE’s thematic issue arose as a tribute to the 50th anniversary of FORMEC. After conducting consultations with colleagues in both the management and the scientific committee of FORMEC Network and having obtained the consents of the editorial board of CROJFE, the implementation of this idea started.

Invitations with framework themes of invitatory papers, which should collectively cover the whole area (or at least most of it) of forestry engineering, were sent to the addresses of 25 selected potential authors (counting on publishing about 15 papers).

Finally, after the review process was carried out, 14 papers have been published in CROJFE’s thematic issue. Some selected authors have not been able to respond due to various obligations, some authors could not comply with the deadlines required, and some of the papers were rejected or required a major revision during the review process and will be published in one of the following issues of the journal.

In the upcoming years, there will be plenty of challenges for both CROJFE journal and FORMEC Network, but with mutual cooperation, difficulties should be easily overcome. Some of the upcoming questions, regarding not only FORMEC Network or CROJFE journal, but the entire forest engineering community/society are:  positioning of forest engineering as an important and well known component in forestry and life sciences as well  higher quality of undergraduate, graduate, master and doctoral studies in all country members of FORMEC Network  inclusion of international study programs (universities from several countries) of forest engineering or in combination with other fields of forestry (for example silviculture, forest management, etc.) in English  higher quality of scientific and research works, methods, procedures and equipments in all country members of FORMEC Network  equalization of standard procedure for gaining doctoral degree in forest engineering in all country members of FORMEC Network  enhancement of FORMEC conferences in terms of: (1) receiving papers/presentations/posters of higher quality and of the same level, (2) introducing new forms of personal communications (workshops, seminars, etc.) to discuss the Croat. j. for. eng. 38(2017)2

Editorial (151–153)

most important topics, (3) increasing the number of participants, (4) higher share of young colleagues (both Postdocs and PhD. students), (5) more female PhD students  equalization of quality of FORMEC conferences (in all components), which should be met by any future organization committee  even better cooperation between FORMEC Network and CROJFE journal, not only for their benefit, but in the interest of the entire forest engineering community/society  higher quality and efficiency of review process in CROJFE, as well as expansion of reviewers’ database with world known scientists and experts in forest engineering

Croat. j. for. eng. 38(2017)2

 retaining regularity of printing CROJFE, quality of technical editing, as well as its financial security and independence  higher quality of published papers (as a result of enhanced review process and higher number of uploaded papers), and increase of the Impact Factor from the current 1.415 to 2.0.

Last, but not least, we would like to congratulate 50 successful years of FORMEC International Network to all members who have contributed in some way to its development and progress, wishing many more such anniversaries to come.

Tibor Pentek, Karl Stampfer, Ivica Papa, Mario Šporčić, Željko Tomašić

153

Subject review

Forest Road Network and Transportation Engineering – State and Perspectives Hans Rudolf Heinimann Abstract The paper reviews traditional and computer-assisted road network layout approaches and brings them together in an overall stream of development. It results in the main finding that changes in the representation of the road network layout problem triggered major scientific advancements. A systematic, 2D transport geometry representation emerged in the 1870s and led to the mathematical derivation of optimal road spacing. The representation of road network and harvest layout problem as a mathematical graph and the solution of the corresponding linear programming problem, triggered a representational shift in the early 1970s. The broad availability of digital elevation models DEMs at the beginning of the 1990s was another representational innovation, enabling an automatic road route layout on the terrain DEM surface.

The most recent shift consisted of systems to semi-automatically, concurrently laying out harvest/transport-network problems on DEMs in the mid-1990s. The review identifies challenges for future research, among which the extension of the concurrent harvest/road-network layout systems for multi-objective functions is the first importance. Considering that scientific advancement is mostly going along with changes in problem representations, research should explore improved representations for lattice type terrain representation, among which triangular irregular network (TIN) meshes seem to be the first interest. Additional paths for improvements are the integration of road network planning with detailed road engineering, the refinement of optimization problems formulations, and the cross-national adaptation of road network planning courses to operations-research-based approaches.

Keywords: road network, network layout, road spacing, road density, computer-assisted network layout

1. Introduction An efficient road network has been the backbone of forestry, the design of which is based on fundamental principles. Those principles have been changing in time due to scientific advancements and due to the evolution in both off-road and on-road transportation technology. The underlying scientific concepts cover a broad range. Computer-assisted systems that are automatically generating the concurrent layout of road and harvesting systems represent the edge of development (Epstein et al. 2006, Epstein et al. 2001), with a scope on plantation forestry conditions. On the other hand, traditional expert approaches, which are relying on the skills and experience of the designers, are still in use and are still part of training programs for forest operations specialists. Although, according to Google Scholar, there have been about 70 scientific publications on forest road network planning since 1960, there Croat. j. for. eng. 38(2017)2

is no comprehensive review of the state of knowledge in the field. Traditional text and handbooks (Dietz et al. 1984, Hafner 1971, Kuonen 1983, Wenger 1984) documented expert-based approaches, while review articles (Church et al. 1998, Epstein et al. 2007) summarized the state of knowledge for mathematical approaches stemming on operations research techniques.

The goal of the present contribution is to bring computer-assisted and traditional road network layout approaches together and to sketch an overall stream of development. In particular, it aims to: ďƒž bring out the concepts and methods for the different conceptual approaches in time ďƒž identify discontinuities, at which major progress occurred.

The following reflections will focus on forest road networks, thus neglecting other types of low-volume road networks. Additionally, they will cover the time

155

H. R. Heinimann

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

span from the emergence of scientific literature on forest road networks by the end of the 18th century until today. The article will first sketch the early developments up to World War I, and second, analyse the transport geometry approach that led to the concept of optimal road spacing. Third, it will review approaches that were using operations research tools that developed further into semi-automatic computeraided road network layout methods. Finally, we will discuss computer-aided approaches to concurrently layout road network and harvesting systems.

2. Early Developments The establishment of the French state Corps of Bridges and Roads (Corps des Ingénieurs des Ponts et Chaussées) in 1716 was the nucleus for enlightened, modern engineering sciences (Belhoste 1989). Before, engineering knowledge was tacit without formal knowledge generation and dissemination, and without formal education and training systems. Modern engineering sciences grew out of a military initiative, the foundation of the Corps of Fortifications (Corps des Ingénieurs des Fortifications) in 1691. French military ambitions – the establishment of a powerful fleet and the construction of fortifications – required an improved timber supply system. A French marine engineer, Duhamel du Monceau, was a mastermind of the wood supply improvement programme, which he documented in a series of books, among which »Transport, Conservation and Forces of Timber« (Duhamel Du Monceau and Prévost 1767) documented forest transport practices of those times. Waterborne transport with vessels or by rafting was the backbone of transportation, in particular for long distances. Land transportation was mainly limited to charcoal and firewood to the nearest centres of usage (villages, cities), and large -sized »carpenter timber« could not be transported with carriages or carts, because their carrying capacity was limited to about two metric tons. Although France was the cradle of engineering sciences, there was no theory on land transportation until the 1860s (Picon 2016). About 60 years after Duhamel’s book, a German textbook »Handbook on Timber Transport and Floating« appeared (Jägerschmid 1827), describing the transportation practices at that time.

There were two road standards, one for firewood and one for roundwood. The author did not give technical specifications for the layout of road networks, but for vertical alignment and cross-section design, which are essential for detailed road engineering. Since biomechanical power (humans, draught animals) was the only energy source for locomotion, gradients had to

156

be limited to about 8%. Both Duhamel and Jägerschmid, took a purely technical point of view, thus totally neglecting economic aspects, which can be explained by the fact that mercantilism was the dominating economic paradigm that was only slowly replaced by classical economics.

In 1842, von Thünen wrote a seminal book »The Isolated State about Agriculture and Political Economy …« (Thünen 1842), which is somewhat the cradle of location theory and transportation economics.

Thünen investigated how different types of land uses arranged spatially around a city that had absolutely no economic exchange with other economic entities.

For that kind of assumptions, various types of landuses arrange in concentric rings. He investigated where the production of timber should be located, concluding that this would happen in a second circle.

The outer limit was at about 8 miles, beyond which the transportation was economically infeasible because transportation cost became higher than timber prizes.

One horse and carriage unit was able to transport about two m³ of firewood and could do a distance of about 20 to 30 km a day, which illustrates how demanding land transportation was at that time.

Around 1840, a discontinuity in the development of transportation technology occurred with the appearance of large-scale railway networks. This development cut ground transportation costs tremendously, triggering the exchange of goods over long distances. Around 1860 the theory of transport network layout emerged concurrently in Germany and France (Lalanne 1863, Launhardt 1872), providing quantitative methods for the design of transport networks. Both authors introduced a conceptual design phase – nowadays called »architecture definition process« (Walden et al. 2015), aiming to find the best possible topological network arrangement. Urban transportation started with the centres of traffic demand and then looked for the best possible connection.

Whereas the French approach was purely geometric, Launhardt’s approach was quantitative. His conceptual design phase, which he called »commercial trace« (Launhardt 1872), totally neglected terrain conditions, aiming to layout the best possible network on a twodimensional plane. His process started with the identification of »transport locations« (cities, villages, production plants) and their specific traffic inflows and outflows. The problem was then to find a network that connects all the transport locations at minimal transportation cost. He developed a procedure to optimally locate intersection points, considering the traffic flow and the geometry of the adjacent points. Overall, Launhardt’s method resulted in a type of Steiner tree, Croat. j. for. eng. 38(2017)2

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

connecting all the »transport locations« to the network. The succeeding detailed road route engineering phase – now called »design definition process« (Walden et al. 2015), which Launhardt called the »technical trace«, aims to locate road routes on the terrain and to define the vertical alignment of the road centreline in such a way that it considers the technical constraints and results in a smooth horizontal and vertical alignment. It is not clear how this seminal work affected the theory of forest road network planning.

However, some key conceptual design ideas, such as the definition of »transport locations« (»fixed points«, »control points«) and the search for a network topology can still be found in later textbooks (Dietz et al.

1984, Hafner 1956, Kuonen 1983, Wenger 1984).

The book »Forest Road Construction and Its Preliminary Work« (Schuberg 1873) is, to our knowledge, the first textbook with a comprehensive treatment of forest road networks. The author recognized that transportation planning in forestry is not a point connection problem, but a problem of how to make a whole area accessible as even as possible. Schuberg introduced principles of forest transport geometry, quantifying the relationship between road spacing and average skidding distance ASD, although his analytical solutions might not be correct, yielding an ASD of 5/24 road spacing, which is however quite close to the correct solution of 1/4 road spacing. He suggested lattice-type road networks for gentle terrain with an average road spacing of up to 700 m, if there is a secondary network of skidding roads. In hilly terrain, the terrain constraints the route location, resulting in network types that were called »contour-type networks«. Following Launhardt’s logic, those considerations defined the conceptual layout of forestry road networks. The detailed road network layout had to stem on reliable contour maps, and compass-based methods to layout routes with a regular grade by stepping with a constant distance from one contour to the next were already known. Once a good enough solution was found on the map, the design was transferred to route locations in the terrain with the use of surveying instruments. It is not clear how Schuberg’s methods spread because other textbooks were focusing on road engineering and road construction (Stoetzer 1877, Stoetzer and Hausrath 1913). In North America, road network and transportation engineering started to become formalized around World War I (Greulich 2002), amplified by the need for economic rationalization and the understanding of logging cost. World War I triggered another discontinuity in development with a boost of motor-vehicle-based transport systems after the 1920s, which resulted in quantitative insights that will be discuss below.

Croat. j. for. eng. 38(2017)2

H. R. Heinimann

3. Optimum Road Spacing/Optimum Road Density The emergence of industrial engineering as a scientific discipline in the aftermath of Taylor’s seminal work on time studies (Taylor 1895) resulted in an improved understanding of production cost and productivity. It had a stimulating effect on scientific studies in forest operations (Ashe 1916, Braniff 1912), leading to an increasing number of time and cost studies. By the end of the 1930s, the tractor started to appear for skidding operations, enhancing the alternatives to design forest harvesting systems tremendously. An operational study reflected this new variability (Matthews 1939), coming to a conclusion that – depending on road conditions – off-road transportation cost per unit of volume and unit of transportation distance were about 6 to 9 times higher than on-road cost. The figures led to the insight that increasing the share of on-road transportation, while decreasing the proportion of off-road transportation, must lower the total cost up to a minimum, beyond which the total cost would raise again. In his seminal paper (Matthews 1939), Matthews raised the question at what road spacing this minimum would occur. He developed a transport geometry model that was based on the following assumptions:  terrain conditions of the forest area are flat, homogeneous, and consequently, road building cost are constant, while off-road transportation costs depend only on the distance  road network layout follows a pattern of parallel roads with equal road spacing, and there are no links between those routes  road network is built at one point in time; there is no sequencing across several time periods  Off-road transportation takes place on the shortest path between the loading point in the stand and the landing point on the road  forest stand conditions are homogeneous for the whole area  there is one forest management strategy that applies to any part of the area under consideration.

Matthews’ mathematical formulation (Matthews 1939) is not straightforward, and some relationships, such as between road length and road spacing for a unit area, appeared implicitly in his formulation. This is why a more comprehensible formulation follows below. Fig. 1 illustrates the basic transport geometry model, characterized by road spacing (sr) and road length (L). Assuming that the area covered by sr and L equals the unit area Au, road spacing sr equals the unit

157

H. R. Heinimann

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

Below, any formula will be based on consistent length and area dimensions. Matthew represented the cost of road construction per unit of volume as (4):

Cr =

Where: Cr cr sr V

cr

sr × V

(4)

road cost, EUR.m-3 road construction cost, EUR.m-1 road spacing, m harvesting volume, m3.m-2

Writing the skidding cost per unit of volume results in (5):

Cs =

Where: Cs cs sr

Fig. 1 Transport geometry for a system of parallel roads. L = lengths of road segment, sr = road spacing area Au divided by the length L of the road segment (1). Equation [1] makes it possible to express both sr and L, respectively, as a function of sr.

From L × sr = Au it follows sr =

Au

L

(1)

Where: L length of road segment, m sr road spacing, m Au unit area, m2 Au equals 1 (length-unit)2 if length and area units are consistent. It has been a tradition, in particular in Europe, to use non-consistent units, meters for length and hectares for the area. If Au is expressed in hectares (2), unit consistency can be achieved by multiplying with a conversion factor (10,000 m2 per hectare), which results in (2).

 10 ,000 × m2  L × sr = Au  ha ×  ha  

(2)

Where: L length of road segment, m sr road spacing, m Au unit area, ha Solving (2) for road spacing sr yields (3), expressing road spacing in meters:

158

sr =

10 ,000 × Au  ha 10 ,000 × m2  ×   L ha m 

(3)

sr × c s

4

(5)

skidding cost, EUR.m-3 variable skidding cost, EUR.m-1.m-3 road spacing, m

Summing (4) and (5) up yields (6):

Ctot = Cr + Cs =

sr × c s c + r 4 sr × V

(6)

The derivative of (6) with respect to sr yields (7), which is a quadratic equation:

C’ tot =

cs c − r 4 sr2 × V

(7)

Finally, setting C’tot = 0 and solving (7) for sr yields the Matthews formula (8), which explains the optimal road spacing with three variables: skidding cost cs, road construction cost cr, and harvesting volume V.

sr,opt =

cr × 4

cs × V

(8)

Since Matthew used imperial units, acres and feet, his mathematical constant in (8) equals 0.33 instead of 4. Matthew did not use calculus to derive (8); he made use of the fact that total cost becomes minimal if road cost Cr and skidding cost Cs are equal. He used different data sets that were characteristic for the southern US to calculate optimal road spacing, arriving at spacing ranges between 1700 and 3800 feet, which equals 520 m and 1160 m, respectively.

Whereas road spacing is a network design parameter, road density is a »dual« parameter that equals the inverse of road spacing. Road density is the ratio between the length L of the road segment and the unit area Au (Fig. 1, (9)).

Croat. j. for. eng. 38(2017)2

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

RD =

L

Au

H. R. Heinimann

(9)

Multiplying road density (9) with road spacing (1) yields the primary relationship (10), yielding that road spacing sr is the inverse of road density RD and vice versa.

L Au × = 1 Au L

RD × sr =

(10)

Replacing sr with 1/RD and solving it for RD, yields the equation for the optimal road density (11), which is the inverse of (8).

RDopt

c ×V = s

cr × 4

(11)

Different countries developed different traditions.

Whereas the road spacing point of view has been dominating in the US, the road density view has been dominating in Europe. As equations (8) and (11) demonstrate, there is a duality between the two viewpoints, and the implications for road network layout planning are the same.

A Swiss investigation derived a mathematical approach to identify the optimal road spacing (Soom 1950, 1952) for both industrial timber and fire wood extraction. Although the author seemed not to have been aware of Matthews’ work, the basic formalism is the same. He relaxed the assumption that extraction costs are strictly linear to the extraction distance by introducing a quadratic extraction cost function. The analysis resulted in optimal road spacing of about 500 m, which was significantly larger than the then customary best practices.

More or less at the same time, Ulf Sundberg started his work on transportation economics at the Department of Operational Efficiency at the Royal College of Forestry in Sweden. In a study (Sundberg 1953), he developed generalized cost functions for both road construction and off-road transportation, using the basic relationships of (4) and (5). He relaxed the assumptions of a purely parallel road network and the shortest off-road transportation path by introducing correction factors, which he investigated for different geometrical extraction patterns. He then extended (8) with those correction factors to derive the optimal road spacing. His work triggered some follow-up studies, such as a study to identify the optimal road standard and the spacing of roads on networks with primary and secondary roads (Larsson 1959). Another study (Segebaden 1964) investigated how the assumptions of purely parallel roads and the shortest off-road transportation path could be relaxed and introduced Croat. j. for. eng. 38(2017)2

Fig. 2 Transport geometry for a lattice-type road network, sr = road spacing, ASDeff = effective yarding distance into the determination of the optimal road spacing and the optimal road density, respectively.

Fig. 2 illustrates the relaxed transport geometry for a lattice-type road network. Assuming that there are two systems of parallel roads, being perpendicular to each other, a comparison with the basic model of Fig. 1 yields that the road network layout affects the average skidding distance ASD, which goes into the skidding cost function (5) as the road spacing sr divided by 4.

Considering that road spacing and road density are reciprocal, the average skidding distance equals the reciprocal of 4 times the road density (12). The unit area in Fig. 2 equals 4sr2, while the road lines per unit area are 4sr. Introducing those terms into (12) yields the average skidding distance, sr/4, which we call theoretical average skidding distance because ASDtheor is predicted by the basic model.

ASDtheor =

s 4 × sr2 1 = = r 4 × RD1 4 × 4 × sr 4

(12)

Fig. 2 illustrates how the extraction pattern looks like for a lattice-type network. Assuming that logs are moving on the shortest path to the nearest road results in a pattern of 8 triangles per unit area, and the average skidding distance for a single triangle equals the distance from its centroid to the nearest road, which is one third of the road spacing sr (13).

ASDeff =

sr

3

(13)

If we calculate the ratio between the effective and theoretical average skidding distance (14), we get a factor of 1.33, which is characteristic for triangular, rectangular, and hexagonal networks (Segebaden 1964). Matern, a mathematician, analysed a Poisson

159

H. R. Heinimann

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

field type network for Segebaden, demonstrating that the corresponding network correction factor cnet equals exactly 2.0.

cnet

ASDeff s ×4 1 = = r =1 ASDtheor sr × 3 3

(14)

If we take the inverse of the network correction factor cnet, we get a metric for the road network layout efficiency enet, which was introduced based on empirical investigations (Backmund 1966).

enet =

ASDtheor sr × 3 3 = = ASDeff sr × 4 4

(15)

Segebaden introduced the network correction factor cnet into the skidding cost term of the total cost function (6). Additionally, he proposed an off-road transport correction factor coffr that considers that logs do not move on the shortest path to the nearest road. Considering both, cnet and coffr, the overall cost function results in a generalized cost function (16).

ctot =

sr × cs × cnet × coffr c + r 4 sr × V

(16)

Road network developments are long-term investments, which require a capital budgeting view, balancing cost and revenue flow over the whole project lifecycle. The formula for optimal road spacing (8) and road density (11), respectively, do not consider this long-term investment aspect. The equivalent annual cost (EAC) approach is a straightforward method to calculate the cost of owning and operating an asset over its entire lifespan. Assuming that the road network has a lifespan of 50 years and that a conservative value for the interest rate is 2%, an annuity factor of 3.18% is obtained. If the interest that has to be paid for the capital is neglected, the annuity factors become 2.0%, resulting in a difference in road construction cost of 35% and illustrating the importance of capital budgeting aspects. This dynamic view was first introduced in Great Britain (Gayson 1958), where a type of annuity and annual maintenance factors were introduced, expressing the maintenance costs as the ratio of the investment cost. To make the overall cost function consistent, the harvesting volume V, measured in volume units per area unit, has to be converted into an annual harvesting flow volume Va, which equals the mean annual increment under steady-state assumptions.

Introducing these capital budgeting aspects into the total cost function (16), calculating the derivative with respect to road spacing sr, equalling it to 0, and solving it for sr, the equation for the generalized optimal road spacing (17) is obtained.

160

sr,opt =

cr × ( a + m ) × 4

cs × cnet × coffr × V a

(17)

Where: sr road spacing, m cr road construction cost, EUR.m-1 a annuity factor, a-1 m maintenance factor, a-1 cs variable skidding cost, EUR.m-1.m-3 cnet network correction factor coffr off-road correction factor Va harvesting volume flow, m3.m-2.a-1 (17) gives the basic insight into the main factors affecting the economic efficiency of road network layout. The ratio of road construction to skidding cost (cr/cs) is crucial, and inefficient skidding technology calls for lower road spacing. Inefficient road network geometries, expressed by a high network correction factor cnet, have a similar effect. A Poisson field layout, yielding a network correction factor of 2.0 (see Segebaden), reduces the road spacing to about 70% compared to a network with parallel roads. Whereas a risky capital budgeting policy (low interest rates) results in relatively high road spacing, low-risk budgeting policies (high interest rates) increase road spacing.

Finally, the management intensity – expressed by the harvesting flow volume – is affecting road spacing.

Under a plantation forestry regime with an annual flow of about 30 m³ per hectare, road spacing is half of that under traditional forestry regimes within an annual flow of about 7.5 m³ per hectare. Although those insights are not new, they cannot be taken for granted.

RDopt =

cs × cnet × coffr × V a

cr × ( a + m ) × 4

(18)

[18] is the generalized equation to determine the optimal road density, which is the dual problem to optimal road spacing. In 1963, the so-called »Joint Committee« of ILO/FAO/ECE organized a symposium on the planning of forest communication networks that took place in Geneva, Switzerland. Most of the forest operations engineering specialists of that time participated, sharing their knowledge of road network developed from different perspectives. Sundberg delivered a paper on road network economics, reviewing the state-of-the-art of that time (Sundberg 1963). He emphasized that both the network correction cnet and the off-road transport correction coffr factors multiply with the theoretical average skidding distance, resulting in higher effective than theoretical distances. He also explained how this should be considered in the calculation of optimal road spacing and roads density, Croat. j. for. eng. 38(2017)2

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

respectively. Silversides gave an overview on the influence of logging methods on the road network layout (Silversides 1963). He emphasized that Matthews book (Matthews 1942) still build the basis of the North American approach to estimate optimal road spacing, thus neglecting some of the newer Swedish achievements. He requested that future work on road spacing should distinguish two cases:  systems where the logs are moving on the shortest path to the nearest road  systems where the logs are moving to specific transhipment points (landings) that are located in intervals.

In the seminal book, »Cost Control in the Logging Industry« (Matthews 1942), the author proposed a procedure for the simultaneous determination of road and landing spacing. Assuming that the logs are moving on a radial axis to the landing, the mean yarding distance does no longer equal the distance from the centroid of an area to the nearest road. This called for a more precise estimate of the average skidding distance, which appeared in a report (Suddarth and Herrick 1964). Later, Peters developed an approach that is slightly more precise than Matthews approximation, but quite challenging to solve. Interestingly, the effect of non-efficient road network layout, represented by Segebaden’s network correction factor cnet, was never considered. This means that all the concurrent landing/road spacing approaches assumed the transport geometry of parallel roads (Fig. 1). The traditional geometric layout of roads and landings assumed that the landings are located on access perpendicular to the roads. Bryer relaxed this assumption by shifting the landing locations on every second road (Bryer 1983). The analysis showed that shifting reduces the average skidding distance by about 5% to 9%, if the ratio of landing spacing to road spacing is between 1.5 and 2.0.

After the 1970s, contributions to the optimal economic layout of road networks became marginal.

Whereas some work refined the cost function by including travel cost of workers, the opportunity cost of growth loss due to aisle clearing and other effects (Abegg 1978), most of the work made adaptations to local contexts. There were also contributions that are hardly defendable, such as the consideration of overhead costs in the calculation of optimal road spacing (Thompson 1992). If one formulates a total cost function and finds the derivative for road spacing, only the variables will remain that directly dependent on road spacing. It is not appropriate to formulate an overall cost function, in which overhead costs are directly dependent on road spacing.

Croat. j. for. eng. 38(2017)2

H. R. Heinimann

The transportation geometry models (Fig. 1, Fig. 2) were all two-dimensional. In order to study how slope affects the road network layout or how the total cost function for a combination of road and off-road transportation increases with slope, a 3D-model must be used. Heinimann developed an approach to differentiate skidder and cable-yarder-based road network concepts on steep slopes (Heinimann 1998). He found that at 30% slope, the skidder-based outperformed the yarder-based total cost function. At 50% slope, the situation was inverse, and the yarder-based outperformed the skidder-based function. This means that there must be a slope at which the minima of both costs functions are equal, which is the threshold, at which a cable-based system is more efficient than a yarder-based one. If context-specific data are available, this approach can be calibrated for different road network concepts. Previously, the discrimination of road network concepts was mainly based on rules of thumb.

4. OR-Tool-Supported Road Network Layout A high-level panel on decision-making and problem-solving concluded that the way in which problems are presented affects the quality of the solutions that will be found (Simon et al. 1986). In the 1950s, the shift from the pre- to the post-computer era took place, empowering humans to solve problems that were not tractable before. As a consequence, the field of linear programming started to spread into commercial applications, providing methods to identify the optimal solution for a system represented by a set of linear relationships and constrained by linear inequalities or equalities. It seems obvious that this quantitative field of knowledge started to trigger

Fig. 3 Conceptual model to integrate silvicultural activities and road construction over multiple time periods, adapted from (Weintraub and Navon 1976)

161

H. R. Heinimann

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

novel approaches to forest management and engineering. In 1973, a symposium on »Planning and Decision-Making as Applied to Forest Harvesting« took place in Oregon (O’leary 1972) and it paved the way to operations research techniques in forest engineering. Two contributions addressed the use of ORtechniques for road network layout planning (Kirby 1972, Mandt 1972). Fig. 3 illustrates the problem representation, based on which a mathematical program was formulated.

The representation (Fig. 3) relaxes the assumption that the forest cover is homogeneous for the whole area of interest. The area is divided up into similar units (stands, harvesting units, timber stratum), each of which belongs to a specific age class and is managed according to a specific silvicultural regime. The representation further relaxes the assumption that all the roads are built at one point in time, allowing road segments to be constructed in different time periods.

The conceptual model (Fig. 3) assumes that each forest unit has one access node (N1–N8), from which the whole unit can be managed. It further defines a set of road segments between the nodes (S03 to S67), the combination of which allows the timber to flow from the source nodes (N1 to N7) to the sink node (N0). Whereas Kirby discussed problem formulation options (Kirby 1972), Mandt introduced the analysis of road networks from a network analysis perspective (Mandt 1972), drawing on the then state-of-the-art in the field (Ford and Fulkerson 1962). A seminal paper (Weintraub And Navon 1976) described the problem formulation and solution procedure for the Fig. 3 planning problem. The authors defined sets of (1) access nodes I, (2) road segments J, (3) time periods T, and (4) timber classes K. The problem is to allocate all harvest units (i1...in) and the required road segments (j1...jk) to the time periods (t1...tl), such that the discounted (revenues – cost) become maximal. While it is not too difficult to transfer this goal into an objective function, the formulation of constraints is much more tricky. The authors found a formulation that could be solved on a mainframe computer in about 2 1/2 minutes (Weintraub And Navon 1976). To keep the solution feasible, the road network had to be restricted to a small number of paths, generated as shortest paths between access and sink nodes (Kirby 1986). Using the classical transhipment problem formulation, large-scale network problems with many links and with many road construction and reconstruction projects could be formulated and solved (Kirby et al.

1979). This model developed into the integrated resource planning model (IRPM) by the beginning of the 1980s, and it became operational for use by the

162

U.S. Forest Service. The model extended the Weintraub-Navon model with traffic flow sets, defining the traffic capacity for a given period and with a set of land-use alternatives instead of timber classes. The solution for the model was feasible for relatively small problem size only, requiring heuristic solution procedures, based on a sequence of the linear program run (Kirby 1986). The authors were aware that such a heuristic procedure would not yield optimal plans except by coincidence.

A review paper on locational issues in forest management discussed the state of models to concurrently locate and schedule both harvest and road network layout (Church et al. 1998). It slightly re-formulates the problem as to model »the scheduling/location which harvest units will be cut in each period, meets adjacency restrictions and ensures that no unit is harvested without the completed road route that can reach the unit«. The authors also discussed the variety of solution techniques used to solve this type of problems, such as dual ascent, Lagrangian heuristics, MonteCarlo integer programming, simulated annealing, and tabu search. A follow-up paper (Murray 1998) provided a formal mathematical specification for this multiple target access problem (MTAP), emphasizing that a promising exact solution approach be based on the Lagrangian relaxation with branch and bound method.

The integrated formulation of harvest/road-network layout and scheduling made it possible to assess harvesting costs comprehensively, travel cost and road construction and maintenance cost, and to balance the scaling of harvest and roads building activities. However, it still requires the planner to predefine:  access nodes  exact location of road segments between access points in the terrain  detailed layout of harvesting operations beyond each access node.

5. Computer-Aided Road Network Layout Planning Contour maps built – since their emergence in the Renaissance – the backbone for any spatial planning activity, such as road network layout or harvest planning. By the end of the 1950s, a master thesis at MIT proposed digital terrain models (DTM) as an alternative to contour maps (Miller 1958). A digital terrain model approximates a part of the continuous terrain surface with a large number of selected discrete points with known XYZ coordinates in an arbitrary data coCroat. j. for. eng. 38(2017)2

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

H. R. Heinimann

Fig. 4 Automatic road route location on a digital terrain model. From (Liu and Sessions 1993), adapted. SP Steiner points; filled circles: mandatory source and sink nodes ordinate field (Miller 1958), stored electronically. Early applications appeared in the areas of highway earthwork analysis and of highway location (Laflamme 1959). Data acquisition was expensive and laborious, letting the new technology only slowly grow into the broad application. Forest engineering applications appeared in the 1970s, when Burke developed the DTM to automatically extract skyline road profiles (Burke 1974). The situation changed by the beginning of the 1990s, when topographic agencies of many countries started to produce and distribute DEMs systematically.

Fig. 4 illustrates a road layout example on a digital terrain model (Liu And Sessions 1993). Mathematically, a graph is a set of nodes together with a set of edges, each edge associated with two nodes. From a flow perspective, a graph is defined by inflow-, transhipment- and outflow-nodes, all of which have to be identified for the layout of the road network. Traditionally, the planner had to determine those nodes, which were called »control points«, »fixed points«, or »access nodes«. Fig. 4 shows four access and one exit node. After the definition of the nodes, the planner has to identify a pattern of connections between the nodes, such that all »inflow-nodes« are connected to one or more »outflow-nodes«. Traditionally, the solution deCroat. j. for. eng. 38(2017)2

pended on the skills and experience of the planner, resulting mostly in sub-optimal solutions, compared to the optimum.

Graph theory knows the concept of the minimum spanning (Weisstein online-b) tree to connect all the nodes with the minimum possible total link weight, which can be cost or any other metric. It is easy to see that the automatic solution presented in Fig. 4 is a type of minimum spanning tree. The insight that access nodes could be connected by a minimum spanning tree appeared already in the 1960s in the literature (Kanzaki 1966). The solution in Fig. 4 is a Steiner Minimum Tree (Weisstein online-c), which introduced additional »Steiner points« to improve the minimum spanning tree solution in the best possible way. There are two algorithms to solve the minimum spanning tree problem, Prim’s (Prim 1957) and Kruskal’s (Kruskal 1956). While there is no exact solution for Steiner tree problem, good approximations are available (Robins and Zelikovsky 2000).

So far, we discussed some options on how to automatically identify a solution and connect nodes into a directed graph. However, the issue of how to automatically layout a road between two nodes on a digital elevation model still needs a solution. A digital elevation model is a discretization of a continuous,

163

H. R. Heinimann

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

Fig. 5 Link patterns for the automatic generation of road routes on a digital elevation model, following (Stückelberger et al. 2007) real world terrain, consisting of grid cells and grid points. The guiding idea is to define possible links from each grid point to its adjacent grid points and then to search for a set of serial links that connect the two road network points at minimum cost while maintaining feasibility. The most simple link patterns from one point to its adjacent points are the »Von Neumann« and the »Moore« neighbourhoods (see Fig. 5, left). The solution presented in Fig. 4 is based on a »Moore« neighbourhood, and it results in zigzagging. To our knowledge, a study conducted in Japan pioneered the use of »Moore« neighbourhoods for automatic road network layout (Kobayashi 1984). Liu and Sessions evaluated the feasibility of each link for maximum grade and calculated its construction, transport and maintenance cost, whereas construction cost was adjusted to the side slope. The example in Fig. 4 had 25×25 cells and resulted in about 3300 feasible links. Although the problem size is relatively small, the exact solution for the related mixed integer problem was too time-consuming. This is why a heuristic algorithm was used to solve the problem (Sessions 1987).

The use of an 8-link pattern (Fig. 5, left) results in a quite course discretization of the solution space. Experience in finite element representation showed that both the granularity and geometry matter for the quality of the solution. The eight link-pattern is a low granularity discretization of the solution space that

164

tends to result in a chain of consecutive straight lines without any curve or switchback constraints. The two concerns triggered the investigation of alternative link-patterns (see Fig. 5, right) and of horizontal alignment restrictions (Stückelberger et al. 2007). It demonstrated that the link pattern specification heavily influences road network locations and alignments.

The main result was that the 24 link-pattern model that penalizes switchbacks yields good solutions for slope gradient of up to 30%. Steep terrain requires both the refined link model (e.g. 48 links per node, Fig. 5, right) and the introduction of horizontal curvature constraints. The introduction of curvature constraints increases the size of the graph representation by a factor of 256, resulting in a substantial increase in computing times. The authors found that the 8-link zigzag is not always able to identify road segments between 2 points in steep terrain, whereas the 48-link pattern always did so. They found that cost lower by about 30% for the 48-link model in steep terrain, and by about 10% for a constraint 8-link model in moderate terrain (Stückelberger et al. 2007), both compared with the unconstraint 8-pattern. Another investigation developed a model to estimate the spatial variability of road construction cost for a specific area of interest, based on geotechnical information and parametric cost modelling as used in the construction industry (Stückelberger et al. 2006a). Road network layouts based on the assumption of route-indepenCroat. j. for. eng. 38(2017)2

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

dent construction cost resulted in a 10% shorter overall road length but in an increase in road construction cost of about 20% (Stückelberger et al. 2006a) compared to route-dependent cost assumptions. The study further demonstrated that cost-estimating procedures that consider only slope gradient are still resulting in a 20% lower total construction cost compared to the route-independent cost alternative. Based on the work presented above, computer-aided engineering approaches for the layout of forest road networks under difficult terrain conditions reached some maturity, based on which future tools and solutions can be built.

6. Computer-Aided, Concurrent Harvest Road-Network Layout Planning The computer-aided, concurrent solution of harvest/road-network layout problems has to combine two NP-hard problems, the plant (harvest) location and a road location problem, which requires a representation in a huge network with hundreds of thousands of nodes and millions of edges (Epstein et al.

2001). A general solution, based on the conceptual model of Fig. 3, was presented in the 1970s (Weintraub and Navon 1976), but the harvest layout activities required to harvest and extract the timber to the input nodes, which were termed access nodes, were not explicitly considered. This calls for the spatially explicit harvest layout model, which was presented in a seminal work in the 1970s (Dykstra And Riggs 1977). Such a design model has to be spatially explicit, dividing the area of interest in a grid of quadratic cells that are usually positioned on a digital elevation model. The problem is then to delineate ground-based and cable-based extraction areas, to locate transhipment points (landings) for tower yarders and skidders in such a way that the majority of cells are accessible, while the cost is minimum (Church et al. 1998). For a given road network, transhipment points have to be located on roads themselves. The concurrent harvest/road-network layout looks additionally to connect all the transhipment points to the exit points.

Weintraub and Epstein, Chilean operations researchers, supported by Bren and John Sessions, did the seminal work to develop the methods and tools to solve the concurrent harvest/road-network layout problem (Church et al. 1998). In 1993, a Chilean state agency, Fondef, started to fund the development of basic operations research tools for the Chilean forest industry. Solving the concurrent layout problem requires the following decisions to be modelled: Croat. j. for. eng. 38(2017)2

H. R. Heinimann

 which areas to harvest by ground-based and cable-based systems  where to locate the transshipment points (landings)  what area to allocate to each cable system setting  what roads to build  what volume of timber to harvest and to transport (Epstein et al. 1999).

To support those decisions with tools, the evaluation of the best-known systems, PLANS and PLANZ (Cossens 1992, Twito et al. 1987), did not perform satisfactorily, triggering the development of PLANEX to solve the concurrent layout problem. Modelling the problem required topographic information of the harvesting area, including timber inventories at an appropriate spatial resolution. PLANEX was designed to interact with a GIS system.

Another paper (Epstein et al. 2006) explains the model formulation and solution in more detail. The modelling philosophy followed a mixed-integer linear programming approach, thus developing a large -sized network design model that uses information stored in cells with a size of 10×10 m. The link pattern (Fig. 5) consists of 16 links, which is close to the 24-link pattern proposed earlier. The approach also considers horizontal alignment constraints, in particular by considering minimum turn radii. The model minimizes the cost of road construction, machine installation, harvesting and transportation and consists of a vast number of sets, parameters and variables. The typical problem involves about 75,000 timber cells, 400,000 potential road segments, and about 300 transhipment points for cable systems and 5000 for ground-based systems. The algorithm to solve the problem is similar to a heuristic to identify Steiner Minimum Trees (Weisstein online-c) and produces solutions within an error of about 3.5% compared to exact solutions gained with a commercial solver.

PLANEX has been in use in about 8 Chilean forest companies since the mid-1990s and resulted in savings of 15% to 20% of the operating cost (Epstein et al.

2006). Overall, the use of PLANEX and other operations research tools resulted in annual savings of about US$20 million for the two largest Chilean companies, Bosques Aurauco and Forestal Celco. It is interesting to observe that traditional Central European forestry countries have rarely been using this type of sophisticated systems, thus giving away part of the operational margin and losing competitiveness.

Although PLANEX has been the most widely used system, there are alternatives (Chung et al. 2004, Chung et al. 2008, Meignan et al. 2012). PLANEX has

165

H. R. Heinimann

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

Fig. 6 Comparison of a new exact solution procedure (a) with Epstein’s greedy heruistic (b). (a) saves about 7% of the cost of (b). From (Bont et al. 2015). ©2011 swisstopo (JD100042, JA100120) been using a greedy heuristic (Weisstein online-a), and there have been efforts to improve the solution process.

A Lagrangian relaxation, decomposing the problem into its two building blocks, resulted in a moderately better approach (Vera et al. 2003). A meta-heuristic, tabu search, which was tailored to the problem led to significantly faster solution times compared to the exact solution with a standard solver (Legües et al. 2007).

However, on average, the obtained solutions perform slightly lower than optimal solutions.

Considering that PLANEX was designed for clearcut harvesting systems and is based on a greedy heuristics, Bont et al. (2015) developed a model for the concurrent harvest/road-network layout problem, which was based on a simpler model formulation.

They excluded the road transportation cost as a variable by introducing a set cover constrained formulation, which delineated a problem that was much smaller and, therefore, easier to solve than previous formulations. Fig. 6 compares the new, exact solution procedure (a) with Epstein’s greedy heuristic (b). The solution for optimality outperforms the greedy heuristic by about 7% regarding cost and results in a different spatial layout. Although the solution performed well, there are still opportunities for improvement.

The study demonstrated that the process on primary

166

assigning candidate nodes formed the fitness of the solution as well. At present, the upper feasible limit for a successful solution is less than 1000 ha.

In the early 1980s, the adverse consequences of forest road construction and use started receiving public attention. Studies on how to assess the environmental impacts of forest roads appeared, following the environmental impact assessment philosophy, which was mainly based on expert appraisals. In parallel, the design of forest road networks still followed a monoobjective, cost-optimal approach, which is even true for PLANEX (Epstein et al. 2006, Epstein et al. 2001), the most advanced system to semi-automatically design concurrent harvest/road-network layout. A study (Stückelberger et al. 2006b) developed an approach to map the spatial variability of several objective functions, such as lifecycle cost, ecologically effects and cable yarder landing attractiveness. The weighted sum of objective functions method was used, systematically varying the weights to gain the pareto frontier.

The cost-optimal solution is a benchmark, against which eco-optimal solutions were compared. Spatially explicit, eco-optimal solution for bird habitat protection (Caipercaillie) resulted in a length increase of 60% and in an increase of cost of about 80%. Another eco-optimal solution for marshland protection Croat. j. for. eng. 38(2017)2

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

led to an increase of cost and road length of about 30%.

Those results illustrated that the minimization of environmental impacts is context-dependent and that there is no overall »silver bullet« solution for the environmental impact problem. Multi-objective optimization approaches improve our understanding of trade-offs, which is crucial for the expert-stakeholder dialogue. Another study (Bont 2012) investigated multi-objective layout problems for mountain protection forests, aiming to concurrently optimize the cost, protection against natural hazards, and residual stand damage. Protection against natural hazards (snow creeping, flow avalanches) requires that cable roads and slope direction be minimized, whereas minimum residual stand damages require uphill logging. Taking the cost-optimal solution as a reference, the »slopeline-optimal« solution resulted in 7% increase in cost, reducing cable roads in slope line directions to zero.

H. R. Heinimann

The »uphill-yarding-optimal« solution showed a 22% growth in cost by reducing the slope-line impact to zero. To our knowledge, this investigation was the first to semi-automatically generate a concurrent harvest/road-network layout for multiple objectives.

7. Discussion and Perspectives The present review aimed (1) to bring out the concepts and methods for different road network layout approaches in time, and (2) to identify discontinuities, at which major progress occurred. Our review yielded a set of road network layout methods, covering a path of development that started with rules-of-thumb approaches in the 18th century, and that emerged into semi-automatic mathematical optimization approaches by the beginning of the 21st century. Table 1 shows the characteristics of the different layout approaches

Table 1 Features characterizing different road network approaches. Each vertical sequence of points specifies the profile of a design approach, whereas the shadings indicate major scientific advancements Representation and design assumptions

Early approaches

●

Terrain

2D, perfect plane 2D, contour maps

CAE Harvest Optimum Optimal entry CAE network road-network road spacing point access optimization layout

●

●

● ●

3D, digital elevation models DEMs

Forest cover

Continuous, uniform conditions

●

● ●

Discrete spatial harvesting units

● ●

Discrete harvesting units, represented by a set of grid cells (e.g. 10 m x 10 m)

Road network layout design

Expert cognition and techniques

●

Expert cognition and experience + mathematical derivation of the key design parameter: road spacing

● ●

Expert cognition and techniques to identify road segments between entry points

●

Automatic identification of road segments between entry points

Harvesting layout design

Mathematical identification of an optimal network No harvest layout Manual layout of one entry point for each harvesting unit Automatic delineation of harvesting technologies (ground- or cable-borne) and generation of a large set of entry points

Croat. j. for. eng. 38(2017)2

●

●

●

●

●

●

●

●

●

167

H. R. Heinimann

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

with feature profiles, covering aspects of terrain, forest cover, road network layout and harvesting design.

Early approaches (Table 1) started to get formalized with the appearance of textbooks in the 1870s.

They recommended lattice-type networks in flat terrain, the spacing of which relied on rules of thumb.

In steep terrain, the definition of a set of access points, which had to be connected by a road network, was an important layout design strategy, although the location of those access points (control points) was not based on scientific evidence, but on very general rules of thumb. The effectivity and effectiveness of those early approaches heavily depended on the skills and expertise of the planning expert. Additionally, road network layout did not take into account the variability of the forest cover and the layout of harvesting units.

In a follow-up phase, the design approaches got scientifically more mature with the introduction of quantitative tools to estimate the optimal road spacing (Table 1). Looking at models of a purely parallel or a Manhattan grid layout, it is easy to see that there is a single design parameter that defines those layouts, road spacing. Starting with seminal work in 1939 (Matthews 1939), a research stream emerged, aiming to mathematically identify the optimal road spacing, which is a dual property to optimal road density. Whereas optimal road density has been mainly used to formulate policies as to what level of accessibility should be achieved on regional or national scales, optimal road spacing has been a design parameter for the spatial layout of specific forest road networks. Although there are tools available to estimate the optimal road spacing, the design of road network layouts for specific terrain units resulted in a small set of network alternatives, of which the most suitable alternative has to be selected. Textbooks on forest road network design that have been widely used, such as (Dietz et al. 1984, Hafner 1971, Kuonen 1983, Wenger 1984) all relied on the manual layout of control points and road spacing as the main network design parameter.

The third type of network design approach – optimal entry point access (Table 1) – emerged in the 1970s, when a new representation of forest road network planning problems appeared. Whereas previous methods neglected the variability of the forest cover and the harvesting layout, the new approach identified harvesting units to be cut in different time periods. Each harvesting unit was characterized by an entry point, to which logs were expected to move by various off-road transport technologies, such as skidding or cable yarder. Once those entry points

168

were identified, a set of road segments was located, each connecting a pair of entry points. Experts had to do two tasks: locate the entry points and road segments. The search for the minimum tree that connects all the entry points and the sequencing of road construction that considers harvesting activities taking place in different time periods was based on a mathematical optimization formulation, which resulted in a near-optimal solution.

The U.S. Forest Service refined this approach, which became widely applied as the so-called Integrated Resource-Planning Model (IRPM) in the 1980s.

The formal approach is nowadays known as the Multiple Target Access Problem (MTAP), for which there are exact solutions. This new, operations-researchbased road network planning approach triggered a bifurcation in forest road network design methods.

Whereas regions with considerable forestry tradition, such as Europe, stayed with traditional forest road network planning methods (control point and road spacing led expert layout), the North American forest road community moved to OR-based approaches, which started to be widely used, particularly by the U.S. Forest Service. Weintraub, Chilean, contributed to the development of OR-based road network design methods at an early stage of his career in the U.S., from where the methodologies spread to Chile.

The fourth type of road network design methods – CAE network optimization (Table 1) – appeared in the early 1990s in the U.S., triggered by the broad availability of digital elevation models (DEMs). It built upon the multiple target access problem, which required the planner to identify harvest units, their entry points and road segments between pairs of entry points manually. The representation of the terrain surface as a 3D-grid (see Fig. 4) made it possible to automate the layout of road segments between entry points, and the identification of harvesting units and the location of entry points were the only design task that a planner had to do. From a computational point of view, a shortest path algorithm has to be used to identify the set of the shortest path between all pairs of entry points, whereas minimal spanning tree algorithms provide optimal connection of the entry points to the exit points. The location of control points (entry points) has been the first step in road network design since the 1870s, and the computer-aided-engineering approach fully automated and optimized all the remaining design steps, resulting directly in an optimal solution.

The fifth type of road network design methods – concurrent harvest/road network layout (Table 1) – stems from the DEM-based CAE network optimizaCroat. j. for. eng. 38(2017)2

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

tion method by integrating the harvest layout and concurrently solving the harvest/road network layout problem. Funded by a Chilean state agency, a team of Chilean and U.S. researchers developed PLANEX, a system that semi-automatically generates harvest/ road layouts for specific forest areas that are nearoptimal. Eight major Chilean forest companies have been using PLANEX, which resulted in significant cost savings within the industry. While this methodology was designed for clear-cutting regimes in a plantation forestry context, it was an obstacle for the transfer to areas with close-to-nature forestry schemes. Around 2005, research emerged, aiming to improve the efficiency and accuracy of the automatic road segmentation on a DEM, to adapt the problem formulation to close-to-nature and continuous cover silvicltural regimes, and to solve the problem exactly for optimal results. However, more recent network design methods have mainly been used for case studies. The reasons for this limited use are the high level of road network development and the fact that the road network specialists lack skills for using quantitative, OR-based methods.

The feature profiles in Table 1 show the major scientific advancements in forest road network layout in the last 150 years. The first advancement was the emergence of the theory of optimal road spacing/density, which matured between 1940 in 1960 allowing a scientifically informed layout of road networks. The second advancement, maturing in the 1970s and 1980s, represented the road layout problem as a timber flow problem from entry nodes (harvest units) to exit nodes for a set of time periods, which became known as Multiple Target Access Problem (MTAP).

While the location of entry/exit nodes and harvest volumes had to be located and estimated by the planner, a mathematical optimization program yielded the best possible connections of entry to exit nodes.

The third advancement, triggered by the broad availability of Digital Elevation Models (DEMs), developed methods to automatically identify the minimum cost road path between any pair of entry points, providing detailed information to determine the optimal minimum-cost connection between entry points and exit points. The fourth and most recent advancement integrated two spatial layout problems – road network and harvest design – representing harvesting units as a set of grid cells, from each of which logs have to »flow« on the minimum cost path to the exit nodes. This latest development reconciled different technologies, such as geographical information, digital terrain and mathematical optimization technology Croat. j. for. eng. 38(2017)2

H. R. Heinimann

that require operations research and engineering knowledge.

Any review of past developments raises the question of how the observed trends could continue into the future. Considering that operations-researchbased road network layout approaches have been mainly in use under clear-cut and plantation management regimes, the OR-based and the traditional forest road layout community have been unlinked. Whereas the road layout optimization community is exchanging its knowledge in the operations research publications, the traditional forest road layout community still relies on forest publications. The separation also affected education and training, and there are still forest operations courses around the world that are mainly focusing on the traditional road spacing approaches relying on OR-based methods only marginally or even neglecting them. This calls for an adaptation of road network layout courses and related training programs for forest road specialists across countries, or even across continents. The further development of all OR-based road network layout methods for multi-objective settings is the second path for improvement, which allows for the identification of pareto frontiers that quantify very present trade-offs between efficiency (cost minimization) and environmental impact objectives. An improved representation of the terrain surface is a third path for improvement that could stem from finite element theory. The approaches reviewed above are all based on a regular, lattice -type discretization of the terrain surface. Finite element theory provides a broad experience of how to best represent the surface with a mesh (Lo 2015), and triangular irregular networks (TIN) are a class of surface meshes that are superior to lattice-type meshes. Their advantage is that the granularity of basic elements (triangles) increases in areas with high variation in height, and decreases in regions with low variation. We hypothesize that moving from lattice -type DEMs to TINs or even more sophisticated meshes (Lo 2015) will further improve the quality of the solutions. The integration of road network layout with detailed road engineering is a fourth path for improvement. Road network layout identifies corridors, within which the future centreline of road will be located. The availability of Digital Elevation Models (DEMs) that are derived from airborne light detection and ranging (LIDAR) systems are offering new opportunities. The availability of unmanned aerial vehicles (UAVs) and low-mass LIDAR sensors with masses of less than 5 kg brought significantly down the cost for the development of high-resolution digital elevation models (Favorskaya and Jain 2017). A future

169

H. R. Heinimann

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

scenario could be as follows: based on an optimal or near-optimal road network layout, a UAV-borne LIDAR sensor could scan all road corridors, producing geo-referenced point cloud, out of which a high-resolution DEM could be extracted. Detailed road engineering starts with the location of a traverse, which is defined by intersection points, which is followed by the definition of the vertical alignment. Traditionally, those two activities had to be done manually, and only about 10 years ago scholars proposed a semi-automatic procedure with the capability to concurrently optimize horizontal and vertical alignments, aiming to minimize construction and maintenance cost (Aruga 2005). Integration of road network and detailed road engineering is expected to further reduce construction, maintenance and transportation cost. The refinement of optimization techniques is the fifth path for improvement. Even one of the most sophisticated systems, PLANEX, is based on a heuristic solution technique, which is providing near-optimal, but not optimal solutions. The increasing performance of computers and the power of optimizers to solve large mixed-integer problems to optimality has been a trend that will continue. However, we should not forget that smart model formulation and the tuning of integer programming algorithms have a considerable potential to reduce solving time and to make problems tractable, resulting in feasible or near-optimal solutions (Klotz and Newman 2013). Set cover problem formulations are a promising approach to decrease the problem size and to achieve optimal or near-optimal solutions (Bont et al. 2015). Finally, it is still unclear how the scientific developments from the early transport geometry approaches (Launhardt 1872, Schuberg 1873) spread and evolved into the traditional layout theory, represented by traditional textbooks (Dietz et al. 1984, Hafner 1971, Kuonen 1983, Wenger 1984).

Belhoste, B., 1989: Les Origines de l‘École polytechnique.

Des anciennes écoles d‘ingénieurs à l‘École centrale des Travaux publics. Histoire de l’éducation 42(1): 13–53.

8. References

Duhamel Du Monceau, H.-L., Prévost, B.L., 1767: Du transport, de la conservation et de la force des bois ou l‘on trouvera des moyens d‘attendrir les bois, de leur donner diverses courbures, sur-tout pour la construction des vaisseaux; et de former des pieces d‘assemblage pour suppléer au défaut des pieces simples faisant la conclusion du Traité complet des bois et des forets. Paris. L.F. Delatour, 556: XXVII p.

Abegg, B., 1978: Die Schaetzung der optimalen Dichte von Waldstrassen in traktorbefahrbarem Gelaende. Mitteilungen, Eidgenoessische Anstalt fuer das forstliche Versuchswesen 52(2): 99–213.

Aruga, K., 2005: Tabu search optimization of horizontal and vertical alignments of forest roads. Journal of forest research 10(4): 275–284.

Bont, L., 2012: Spatially explicit optimization of forest harvest and transportation system layout under steep slope conditions. ETH. Accessed [Dec-28-2016]. [http://dx.doi. org/10.3929/ethz-a-007558027].

Bont, L.G., Heinimann, H.R., Church, R.L., 2015: Concurrent optimization of harvesting and road network layouts under steep terrain. Annals of Operations Research 232(1): 41–64.

Braniff, E.A., 1912. Scientific Management and the Lumber Business. A Possible Field for Foresters. Forestry Quarterly 10(1): 9–14.

Bryer, J.B., 1983: The Effects of a Geometric Redefinition of the Classical Road and Landing Spacing Model Through Shifting. Forest Science 29(3): 670–674.

Burke, D., 1974: Skyline logging profiles from a digital terrain model. In 1974 Skyline Logging Symposium, Univ. of Washington Seattle, WA, 52–55 p.

Chung, W., Sessions, J., Heinimann, H.R., 2004: An application of a heuristic network algorithm to cable logging layout design. International journal of forest engineering 15(1): 11–24.

Chung, W., Stückelberger, J., Aruga, K., Cundy, T.W., 2008: Forest road network design using a trade-off analysis between skidding and road construction costs. Canadian journal of forest research 38(3): 439–448.

Church, R.L., Murray, A.T., Weintraub, A., 1998: Locational issues in forest management. Location Science 6(1–4): 137– 53.

Cossens, P., 1992: Planning and control of short-term log allocation in New Zealand. In Integrated Decision-Making in Planning and Control of Forest Operations, Christchurch.

University of Canterbury, 46–56 p.

Dietz, P., Knigge, W., Löffler, H., 1984: Walderschliessung ein Lehrbuch für Studium und Praxis unter besonderer Berücksichtigung des Waldwegebaus. Hamburg etc. Parey, 426 p.

Ashe, W.W., 1916: Cost of Logging Large and Small Timber.

Forestry Quarterly 14(3): 441–452.

Dykstra, D.P., Riggs, J.L., 1977: An application of facilities location theory to the design of forest harvesting areas. AIIE Transactions 9(3): 270–277.

Backmund, V.F., 1966: Kennzahlen für den Grad der Erschließung von Forstbetrieben durch autofahrbare Wege.

Forstwissenschaftliches Centralblatt 85(11): 342–354.

Epstein, R., Morales, R., Seron, J., Weintraub, A., 1999: Use of OR systems in the Chilean forest industries. Interfaces 29(1): 7–29.

170

Croat. j. for. eng. 38(2017)2

Forest Road Network and Transportation Engineering – State and Perspectives (155–173) Epstein, R., Rönnqvist, M., Weintraub, A., 2007: Forest transportation. In Handbook of operations research in natural resources. Springer, 391–403 p.

Epstein, R., Weintraub, A., Sapunar, P., Nieto, E., Sessions, J.B., Sessions, J., Bustamante, F., Musante, H., 2006: A combinatorial heuristic approach for solving real-size machinery location and road design problems in forestry planning.

Operations Research 54(6): 1017–1027.

Epstein, R., Weintraub, A., Sessions, J., Sessions, B., Sapunar, P., Nieto, E., Bustamante, F., Musante, H., 2001: PLANEX: A system to identify landing locations and access. In International Mountain Logging and 11th Pacific Northwest Skyline Symposium, eds. Schiess, P., Krogstad, F., Seattle.

University of Washingthon, College of Forest Resources, 190–193 p.

Favorskaya, M.N., Jain, L.C., 2017: Overview of LiDAR Technologies and Equipment for Land Cover Scanning. In Handbook on Advances in Remote Sensing and Geographic Information Systems, Favorskaya, M.N., Jain, L.C., Editors. Springer International Publishing, 19–68 p.

Ford, L.R., Fulkerson, D.R., 1962: Flows in networks. Princeton, New Jersey. Princeton University Press, 194 p.

Gayson, A.J., 1958: Road density in Extraction Planning.

Empire Forestry Review 37(3): 305–310.

Greulich, F., 2002: Transportation networks in forest harvesting: early development of the theory. In International Seminar on New Roles of Plantation Forestry Requiring Appropriate Tending and Harvesting Operations, ed. Yoshimura, T., Tokyo. Japanese Forest Engineering Society, 57–65 p.

Hafner, F., 1956: Forstlicher Strassen- und Wegebau, Wien.

Fromme, 399 p.

Hafner, F., 1971: Forstlicher Strassen- und Wegebau, 2., neu bearbeitete Auflage Ed. Wien. Oesterreichischer Agrarverlag, 360 p.

Heinimann, H.R., 1998: A computer model to differentiate skidder and cable-yarder based road network concepts on steep slopes. Journal of Forest Research 3(1): 1–9.

Jägerschmid, K.F.V., 1827: Handbuch für den Holztransport und das Flosswesen zum Gebrauche für Forstmänner und Holzhändler, und für solche die es werden wollen. Erster Band, enthaltend die Fällung, Zurichtung und den Transport des Holzes zu Land. Karlsruhe. Verlag der Ch. Fr. Müller‘schen Hofbuchhandlung, 391 p.

Kanzaki, K., 1966: A planning of the forest road network by the theory of graphs. Journal of the Japanese Forestry Society 48(10): 365–371.

Kirby, M., 1972: A Resource-Allocation Procedure for Planning and Scheduling. In Planning and Decisionmaking as Applied to Forest Harvesting, J. O‘leary, J., Corvallis, Oregon. School of Forestry, Oregon State Universtiy, 75–83.

Kirby, M., 1986: Simultaneous Planning of Wildland Management and Transportation Alternatives. TIMS Studies in the Management Sciences 21: 371–387.

Croat. j. for. eng. 38(2017)2

H. R. Heinimann

Kirby, M., Wong, P., Cox, W., 1979: Optimization of Rural Road Networks – An Application of the Timber Transshipment Model. In Roads of Rural America, eds. Bunker, A.R., Hutchinson, T.Q., U.S. Department of Agriculture. Economics, Statistics and Cooperative Service, 17–26 p.

Klotz, E., Newman, A.M., 2013: Practical guidelines for solving difficult mixed integer linear programs. Surveys in Operations Research and Management Science 18(1): 18–32.

Kobayashi, H., 1984: Planning system for road-route locations in mountainous forests. Journal of the Japanese Forestry Society 66(8): 313–319.

Kruskal, J.B., 1956: On the shortest spanning subtree of a graph and the traveling salesman problem. Proceedings of the American Mathematical society 7(1): 48–50.

Kuonen, V., 1983: Wald- und Güterstrassen Planung – Projektierung – Bau. Pfaffhausen. Eigenverlag, 743 p.

Laflamme, R.A., 1959: The application of the digital terrain model principle to the problem of highway location. Department of Civil and Sanitary Engineering, Massachusetts Institute of Technology, MSc. Lalanne, L., 1863: Essai d’une théorie des réseaux de chemins de fer, fondée sur l’observation des faits et sur les lois primordiales qui président au groupement des populations.

Comptes rendus des séances de l’Académie des Sciences 57(2): 206–210.

Larsson, G., 1959: Studies on Forest Road Planning. Kungl.

Tekniska Hoegskolans Handlingar (Transactions of the Royal Institute of Technology) 147: 1–136.

Launhardt, W., 1872: Kommerzielle Tracierung der Verkehrswege. Hannover, 32 p.

Legües, A.D., Ferland, J.A., Ribeiro, C.C., Vera, J.R., Weintraub, A., 2007: A tabu search approach for solving a difficult forest harvesting machine location problem. European Journal of Operational Research 179(3): 788–805.

Liu, K., Sessions, J., 1993: Preliminary planning of road systems using digital terrain models. Journal of Forest Engineering 4(2): 27–32.

Mandt, C.I., 1972: Network Analysis in Transportation Planning. In Planning and Decisionmaking as Applied to Forest Harvesting, ed. O‘leary, J., Corvallis, Oregon. School of Forestry, Oregon State Universtiy, 95–102.

Matthews, D.M., 1939: The Use of Unit Cost Data in Estimating Logging Costs and Planning Logging Operations. Journal of Forestry 37(10): 783–787.

Matthews, D.M., 1942: Cost control in the logging industry.

American forestry series. New York. McGraw-Hill, 374 p.

Meignan, D., Frayret, J.-M., Pesant, G., Blouin, M., 2012: A heuristic approach to automated forest road location. Canadian Journal of Forest Research 42(12): 2130–2141.

Miller, C.L., 1958: The theory and application of the digital terrain model. Department of Civil and Sanitary Engineering, Massachusetts Institute of Technology, MSc. 107 p.

171

H. R. Heinimann

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

Murray, A.T., 1998: Route planning for harvest site access.

Canadian Journal of Forest Research 28(7): 1084–1087.

of various link patterns and road design elements. Canadian journal of forest research 37(11): 2281–2298.

O‘leary, J., 1972: Planning and Decisionmaking as Applied to Forest Harvesting. In 160. Corvallis, Oregon. School of Forestry, Oregon State Universtiy.

Stückelberger, J.A., Heinimann, H.R., Burlet, E.C., 2006a: Modeling spatial variability in the life-cycle costs of lowvolume forest roads. European Journal of Forest Research 125(4): 377–390.

Picon, A., 2016: French theory of road network layout. Personal Communication. July-21-2016. email.

Prim, R.C., 1957: Shortest connection networks and some generalizations. Bell system technical journal 36(6): 1389– 1401.

Robins, G., Zelikovsky, A., 2000: Improved Steiner tree approximation in graphs. In Proceedings of the eleventh annual ACM-SIAM symposium on Discrete algorithms. Society for Industrial and Applied Mathematics, 770–779.

Schuberg, K., 1873: Der Waldwegbau und seine Vorarbeiten.

Zweiter Band. Die Bauarbeiten, Kostenüberschläge und der Gesamtbau im wirthschaftlichen Betriebe. Berlin. Julius Springer, 575 p.

Segebaden, G.V., 1964: Studies of Cross-Country Transport Distances and Road Net Extension. Studia Forestalia Suecica 14: 1–70.

Sessions, J., 1987: A heuristic algorithm for the solution of the variable and fixed cost transportation problem. In The 1985 Symposium on System Analysis in Forest Resources.

Univ. of Georgia, Athens. University of Georgia, 324–336 p.

Stückelberger, J.A., Heinimann, H.R., Chung, W., Ulber, M., 2006b: Automatic road-network planning for multiple objectives. In The 29th Council on Forest Engineering. Working Globally – Sharing Forest Engineering Challenges and Technologies Around the World, ed. Chung, W., Coeur d’Alene.

University of Montana, 233–248 p.

Suddarth, S.K., Herrick, A.M., 1964: Average skidding distance for theoretical analysis of logging costs. Forest Service Bulletin, USDA (789): 52 p.

Sundberg, U., 1953: Studier i skogsbrukets transporter.

Svenska Skogsvårdsföreningens Tidskrift (Journal of the Swedish Forestry Society) 51(1): 15–72.

Sundberg, U., 1963: The Economic Road Standard, Road Spacing and Related Questions. Some Views on the Theory of Planning a Forest Road Network in Non-Alpine Conditions. In Proceedings, Symposium on the Planning of Forest Communication Networks (Roads and Cables), Genève.

Joint Committee on Forest Working Techniques and Training of Forest Workers, 231–247 p.

Silversides, C.R., 1963: The Technological Structure (Layout) of a Transport Network as Influenced by Various Logging Methods. In Proceedings, Symposium on the Planning of Forest Communication Networks (Roads and Cables), Genève. Joint Committee on Forest Working Techniques and Training of Forest Workers, 299–308 p.

Taylor, F.W., 1895: A piece-rate system being a step toward partial solution of the labor problem. Transactions of the American Society of Mechanical Engineers 16(647): 865–903.

Simon, H.A., Dantzig, G.B., Hogarth, R., Plott, C.R., Raiffa, H., Schelling, T.C., Shepsle, K.A., Thaler, R., Tversky, A., Winter, S., 1986: Report of the Research Briefing Panel on Decision Making and Problem Solving. In Research Briefings, E., Committee On Science, And Public Policy, Editor.

National Academy Press: Washington, DC, 17–36 p.

Thünen, J.H., 1842: Der isolirte Staat in Beziehung auf Landwirthschaft und Nationalökonomie, oder Untersuchungen über den Einfluss, den die Getreidepreise, der Reichthum des Bodens und die Abgaben auf den Ackerbau ausüben.

Rostock, G.B., Leopold’s Universitäts-Buchhandlung, 3 Theile.

Soom, E., 1950: Rückeaufwand und Wegeabstand. Schweizerische Zeitschrift fuer Forstwesen 100: 546–560.

Twito, R.H., Reutebuch, S.E. Mcgaughey, R.J., Mann, C.N.; 1987: Preliminary logging analysis system (PLANS): overview. USDA Forest Service. Portland, Oregon, Technical Report.

SOOM, E. 1952. Rückafwand und Wegabestand beim Rücken von Brennholz. Schweizerische Zeitschrift für Forstwesen 102(8).

Stoetzer, H., 1877: Waldwegebaukunde ein Handbuch für Praktiker. Frankfurt am Main. Sauerländer’s, J.D., Verlag, 170 p.

Stoetzer, H., Hausrath, H., 1913: Waldwegebaukunde nebst Darstellung der wichtigsten sonstigen Holztransportanlagen ein Handbuch für Praktiker und Leitfaden für den Unterricht, 5., bearb. von Hans Hausrath Ed. Frankfurt, A., Sauerländer, M.J.D., 251 p.

Stückelberger, J., Heinimann, H.R., Chung, W., 2007: Improved road network design models with the consideration

172

Thompson, M.A., 1992: Considering overhead costs in road and landing spacing models. Journal of Forest Engineering 3(2): 13–19.

Vera, J.R., Weintraub, A., Koenig, M., Bravo, G., Guignard, M., Barahona, F., 2003: A lagrangian relaxation approach for a machinery location problem in forest harvesting. Pesquisa Operacional 23(1): 111–128.

Walden, D.D., Roedler, G., Forsberg, K., Hamelin, R., Shortell, T., 2015: INCOSE Systems Engineering Handbook: A Guide for System Life Cycle Processes and Activities, 4th Ed. Hoboken. Wiley, 290 p.

Weintraub, A., Navon, D., 1976: A forest management planning model integrating silvicultural and transportation activities. Management Science 22(12): 1299–1309.

Croat. j. for. eng. 38(2017)2

Forest Road Network and Transportation Engineering – State and Perspectives (155–173)

H. R. Heinimann

Weisstein, E.W., online-a: Greedy Algorithm. In Wolfram Web Ressources. Accessed [Dec-28-2016]. [http://mathworld.wolfram.com/GreedyAlgorithm.html]

Weisstein, E.W., online-c: Steiner Tree. In Wolfram Web Ressources. Accessed [Dec-28-2016]. [http://mathworld.wolfram.com/SteinerTree.html].

Weisstein, E.W., online-b: Minimum Spanning Tree. In Wolfram Web Ressources. Accessed [Dec-28-2016]. [http://mathworld.wolfram.com/MinimumSpanningTree.html].

Wenger, K.F., 1984: Forestry handbook, Second Ed. New York etc. Wiley, 1335 p.

Authors’ address:

Received: January 17, 2017 Accepted: June 14, 2017 Croat. j. for. eng. 38(2017)2

Prof. Hans Rudolf Heinimann, DTech Sc. e-mail: hans.heinimann@frs.ethz.ch Programme Director, Future Resilient Systems Singapore-ETH Centre 1 CREATE Way #06-01 CREATE Tower Singapore 138602

173

Subject review

GIS Applications in Forest Operations and Road Network Planning: an Overview over the Last Two Decades Stefano Grigolato, Omar Mologni, Raffaele Cavalli Abstract A systematic literature review was settled to investigate the application of GIS in terms of methods, complexity and accuracy to support decision on forestry operations and forest road network planning. A comprehensive search for relevant studies was performed to retrieve as many relevant international scientific publications dealing with forestry operations and forest road network planning in the period 1996–2015.

The analysis was based on the development of a systematic literature review comprising three steps:  implementation of the database searches by well-defined search terms  identification of all the publications meeting the requirements of the search terms by abstract  choice of the most relevant publications analysis of the contents.

In this review, »GIS and forest operations« includes all the descriptors dealing with GIS applied to support forest operations decision and analysis, while »GIS and forest roads« includes all the papers dealing with the analysis, management and planning of forest road or forest road networks. A total of 372 references and 82 publications were selected for the analysis as they were clearly in conformity with the review topics (GIS applications in forest operations and road network planning).

The analysis showed that GIS has also been applied successfully and unambiguously to harvesting and transportation engineering in forest operations management. Further to the prevailing use concerning applications to support tactical planning, a significant number of recent publications have turned successfully to GIS applied at operational level. Again, despite the prevailing use concerning applications to support tactical planning, a significant number of recent publications have also turned successfully to GIS applied at operational level with the topics of Forest Operations Management in terms of optimization, productivity and safety analysis.

By considering the recent evolution and improvement of GIS technology and the increasing availability of spatial data, as well their improvement in quality and resolution, the application of GIS in forest harvesting and transportation engineering as well as in forest operations management will expand in the near future.

Keywords: systematic review, GIS, forest operations, forest road network, spatial analysis, IUFRO

1. Introduction Nowadays the application of logical and numerical modelling and statistical methods to spatial data is almost routine thanks to the Geographical Information Croat. j. for. eng. 38(2017)2

System – GIS (Burrough and McDonnell 1998). The application of GIS for spatial analysis is well known.

GIS has the primary means of storing, viewing and analysing spatial data as well integrating a range of different spatial data and information. Spatial infor-

175

S. Grigolato et al.

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

mation can be obtained from map digitizing, field data collection (generally by Global Navigation Satellite System, GNSS), aerial photogrammetry, remote sensing and/or Computer Aid Design (CAD) (Malczewski 1999).

GIS started to support environmental decisions in the middle of the seventies (Worral 1991). One of the first GIS applications in natural resources dealt with planning and management of a forest recreation area in western Maryland (Becker 1976). Again, first GIS applications on harvesting operations focused on strategic harvesting planning to minimise site disturbance (Reisinger and Davis 1986) and on harvesting visual impact assessment in landscape design (Dunningham and Thompson 1989).

Over the last decades, the use of GIS in forestry has been increasingly expanding and widely applied in many fields in forest engineering at professional level as well as at academic and research level (Li et al.

2007). GIS application in forest engineering has also generated considerable recent research interest by highlighting different analytical approaches (Bettinger et al. 2009).

Typically, the GIS covers the range from the application of essential functions of spatial analysis to the application and development of statistical modelling and mathematical modelling (Longley et al. 2011). The choice of the analytical approach depends on the aim of the GIS or spatial analysis. To support decisions on the identification of the most suitable harvesting systems or on the evaluation of forest accessibility, morphology (largely slope, roughness analysis, etc.) and distance (largely straight line or Euclidean distance and cost distance) parameters are the most common spatial analysis applications (Reisinger and Davis 1986, Pentek et al. 2008, Cavalli and Grigolato 2010, Hayati et al. 2012, Dupire et al. 2015, Mologni et al.

2016, Sitzia et al. 2016). Further, spatial analysis essential functions (as Euclidean distance or cost path calculation) integrated in a multi-criteria model can be considered an advanced spatial analysis approach that can be used to support decisions on harvesting planning by introducing diverse criteria for selecting the most suitable harvesting techniques (Kuhmaier and Stampfer 2010, Synek and Klimánek 2014) or the energy wood supply management (Kühmaier et al.

2014), to define forest road maintenance priorities (Pellegrini et al. 2013) as well as new forest road alignment (Babapour et al. 2014, Hayati et al. 2013).

The development of complex mathematical modelling in GIS environment or comparable spatial analysis tool can enable forest engineers to design unequivocally a forest road network layout (Stückelberger et

176

al. 2007, Najafi et al. 2008, Bont et. al. 2015), to optimize cable road layout (Bont and Heinimann 2012), to define the best skidding trail according to tree position (Sterenczak and Moskalik 2014), to support the decision on forest road upgrading (Karlsson et al. 2006), as well as to understand environmental impact generated by forest operations (Proto et al. 2016).

It is currently excessively expensive to collect information on the relevant forest characteristics (as terrain morphology and/or forest growth parameters) by means of field inventories (Maltamo et al. 2014). Light Detection and Ranging (LiDAR) technologies system, aircraft-mounted (Airborne Laser Scanner – ALS) or ground-based (Terrestrial Laser Scanner, TLS), is a cost-efficient means to obtain data with high spatial resolution and high positional accuracy (Akay et al.

2009, Pirotti et al. 2012). GIS and LiDAR thus seem to be a prominent integration for most forest engineering topics e.g. the use of high resolution terrain and stand information to support forest road network planning and designing or to analyse forest operations and forest machine performance. The increasing accuracy of spatial information and a suitable spatial resolution on forest cover and terrain has resulted in the integration of GNSS and GIS technologies (McDaniel et al. 2012).

Accurate analyses, based mainly on essential functions of spatial analysis, are proposed for extracting forestyield maps by combining accurate spatial data and harvester information (Olivera and Visser 2016), as well for proving forest machine efficiency (Alam et al.

2012, Strandgard et al. 2014) or for identifying factors, such as terrain and volume distribution, machine production efficiency and time element identification (Pellegrini et al. 2013, Grigolato et al. 2016, Macrì et al.

2016, Olivera et al. 2016).

Again, at small scale of application, essential spatial analysis commonly used in GIS environments has been used to evaluate spatial distribution of soil disturbance by forest machines by integrating high resolution data, such as micro-DEM derived from light detection and ranging (LiDAR) technologies (Koreň et al. 2015) or photogrammetry methods (Haas et al.

2016, Pierzchała et al. 2014, Pierzchała et al. 2016).

Clearly, in the last years, the forest engineering community has shown great interest in GIS and spatial analysis. As a consequence, a systematic review is currently a prerequisite to highlight the importance of the subject and help to understand the main questions that have driven this interest. Further, the same result can also be useful to encourage future interest and development.

A key motivation for this review is to identify the progress of the last two decades that can contribute to Croat. j. for. eng. 38(2017)2

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

improving the knowledge and the application of GIS in the traditional forest engineering topics and in particular in forest operations and road network planning. The three main questions are: what have been the main spatial analysis techniques applied? What has been the complexity of the applied methods? What has been the accuracy of the data?

2. Materials and methods 2.1 Systematic review To synthesize and discuss different approaches, issues and findings of GIS application in forest operations and forest road network planning, a review of the existing literature was conducted. A systematic review is the procedure of identifying and evaluating multiple studies on a topic using a clearly defined methodology (Wolf et al. 2016). An evidence-based approach to scoping reviews (Khalil et al. 2016) was thus adopted by:  defining and refining research search terms  identifying databases and search engines  querying database using the search terms  creating and applying the inclusion and exclusion criteria filters  verifying the representation of the sub-selections.

A comprehensive search for relevant studies was thus performed to retrieve as many relevant scientific publications dealing with GIS application in forest operations and forest road network planning in the period 1996–2015 (20 years). The papers were gathered up in September 2016.

The systematic literature review thus comprised three main steps. In Step 1, search terms and their various combinations were defined to seek for articles relevant from the perspective of GIS and Forest Operations and Forest Road Network. In this case, the literature search was achieved for peer-reviewed articles and conference proceedings within the Scopus (www.scopus.com) database.

Scopus Database is one of the largest multidisciplinary peer-reviewed literature database also including cited references for citation searching. For citation analysis, Scopus database also offers a higher coverage than other peer-reviewed literature databases. The main constraint of the Scopus database is the limitation in covering articles published before 1995 (Falagas et al. 2007). In this study, this limitation will not influence the results as the search for relevant studies was fixed to retrieve scientific publications in the period 1996–2015.

Croat. j. for. eng. 38(2017)2

S. Grigolato et al.

The search terms were so defined and combined using Boolean operators (AND, OR) and wild-cards representing any group of characters, including no character. The search string for GIS application in forest operations and forest road network planning was thus compiled as follows: TITLE-ABS-KEY (»GIS« OR »G.I.S.« OR »Geographic information system«) AND TITLE-ABS-KEY (»Forest operation*« OR »forest mechanisation« OR »forest road*« OR »forest road network*« OR »Skid* trail*« OR »logging« OR »cable« OR »yarder«) AND TITLE-ABS-KEY (»Forest*«) AND (PUBYEAR > 1995 AND PUBYEAR < 2016) AND TITLE-ABS-KEY (»forest*«).

The additional inclusion criteria were then defined: language: »English«; subject area: »Forestry, Agriculture and Environmental Science« or »Engineering«; document type: »article«; source type: »Journal« or »Conference proceeding«.

In Step 2, the abstracts of the articles were examined to identify the papers meeting the main requirements of the research. In the subsequent Step 3, the whole contents of the articles selected in Step 2 were studied to identify their main research approach and vision.

To better understand the rising in number of publications in »GIS and Forest operation/Forest roads«, the number of papers identified by the systematic search was also compared with the total number of publications (also indexed into Scopus database) identified by the search terms: TITLE-ABS-KEY (»GIS« OR »G.I.S.« OR »Geographic information system«).

2.2 Data organization and inclusion criteria Each identified reference in Step 1 was imported into Mendeley Desktop (Mendeley® Ltd.) library and thus re-organised into an external spreadsheet in the form of a database. The database consisted of three parts:  mandatory fields about the reference  categorization of the subject according to the IUFRO 3.00.00 Division and its unit subjects  fields on main GIS data type and quality as well on GIS analysis.

The following mandatory field were thus compiled in order to fully describe the reference as well its current impact: »Authors«, »Article title«, »Year«, »Journal/conference title«, »Number of citations«, »Type«, »Language« and »Abstract«.

For the purpose of the analysis, the main subjects were associated with the designation of the eight Units

177

S. Grigolato et al.

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

of the IUFRO 3.00.00 Division »Forest operations engineering and management« as follows: »Harvesting and transportation engineering«, »Stand establishment and treatment«, »Forest ergonomics«, »Forest operations management«, »Forest operations ecology«, »Forest operations in mountainous conditions«, »Forest operations in the tropics« and »Small-scale forestry«. »Harvesting and transportation engineering« was sub-divided into three sub-subjects »Road networks and transportation«, »Road engineering and management« and »Harvesting and processing systems«. »Forest operations management« was also sub-divided into two sub-subjects »Operations systems analysis and modelling« and »Supply chain management«.

When the papers could not be clearly identified under the Unit of the IUFRO 3.00.00 Division, an additional field annotated them to the most appropriate IUFRO Division or eventually excluded them when they were not relevant for the subject review.

Step 2 consisted of selecting relevant documents by application of inclusion criteria. Inclusion criteria were first applied to the document title, then to the abstract and in the final phase, to the whole document. Each paper was thus screened according to the following four categories:  paper with the subject clearly in conformity with the review topics (GIS applications in forest operations and road network planning)  paper with the subject partially in conformity with the review subjects (GIS application concerns mainly forest planning, forest management and/or forest landscape and less forest operations and road network planning)  subject slightly conforming to the review subjects (GIS applications are not the main aims of the paper and slightly related to forest operations and road network planning)  paper not related to GIS application neither to forest operations or forest road network planning.

In Step 3, the whole contents of the articles selected as paper with the subject strictly conforming to the review topics (GIS applications in forest operation and road network planning) were analysed.

The database was thus complied by including the following mandatory fields:  planning level (strategic level, tactical level, operational level)  main type of input data (raster data, vector data, aerial photograph; satellite image; miscellany)

178

 limiting spatial data resolution in case of image and raster data (greater than 10 m; between 10 m to 1 m; smaller than 1 m)  complexity of GIS analysis (static analysis; dynamic analysis)  outputs:  cartographic models (temporally static, combined spatial)  spatial-temporal models (dynamics in space and time, time-driven processes)  network models (modelling flow and accumulation over a road network).

3. Results and discussions 3.1 Scientific literature and time distribution In presenting the findings of the literature research, a simple chronology of the publications gathered in the first step is presented. In line with the broad focus of the review, peer reviewed publications on GIS applications in forest operations and road network planning have steadily risen over the past two decades with a rapid increase (Fig. 1), particularly since 2005. The identification process (Step 1) gathered 372 studies for the period 1996–2015. The average number of publications per year was 18.6 when calculated for the entire period 1996–2015. The average number of publications resulted statistically different

Fig. 1 Publications on »GIS and Forest operations/Forest roads« gathered from Scopus database in the period 1996–2015 Croat. j. for. eng. 38(2017)2

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

S. Grigolato et al.

ences both accounting for 56% of the papers. About 13% (37) of the papers were published in just three journals: Croatian Journal of Forest Engineering, Forest Ecology and Managements and Biomass Bioenergy.

In terms of impact, 69.6% of the papers were also cited at least one time during the same period, with a maximum of 406 citations for a paper (Curran et al.

2004) focusing on logging in tropical forest and only partially concerning the use of GIS to support forest operations or forest road network analysis. In total, the 372 papers generated 5299 citations in 20 years with an average of 265 citations on average per year and an average of 14.2 citations per publication.

Most common keywords were: GIS (included »Geographic Information System«, »G.I.S«) with 7% and »Forest road« (including »Forest road«, »Forest road network«) with 5%.

3.2 Application of inclusion criteria Fig. 2 Peer reviewed publications on »GIS and Forest operations/ Forest roads« and on »GIS« and »Forestry« during the period 1996– 2015 normalized for the total number of peer reviewed publications focusing on »GIS« published during the same period (Paired t-test, α=0.05) for the first decade 1996–2005 (9.8 publications per year) if compared with the second decade 2006–2015 (27.3 publications per year).

Scientific publications in »GIS and Forest operation/Forest roads« thus seem to be constant in time (Fig. 2) when compared to the total number of publications focusing on GIS application and GIS development published during the same period. This is confirmed by the analysis of the period 1996–2015 divided in two sub-periods corresponding to two decades 1996–2005 and 2006–2015, in which the percentage of publications on »GIS and Forest operations/Forest roads« related to the total publications on GIS application and development shows that the results are comparable (0.31% in the period 1996–2005 and 0.33% in the period 2006–2015) (Paired t-test, α=0.05). In contrast, the yearly »GIS« publications related to »forestry« in general (search terms: TITLE-ABS-KEY (»GIS« OR »G.I.S.« OR »Geographic information system«) and TITLE-ABS-KEY (»Forest*«) seem to increase with time (Fig. 2).

The 372 collected papers have been published in 132 different scientific journals or conference proceeding (6.4% of the retrieved publications) and written by 156 different first authors from 60 different countries.

The papers covered a broad range of scientific disciplines (20), of which the most represented were Environmental Science and Agricultural and Biological SciCroat. j. for. eng. 38(2017)2

The application of inclusion criteria (Step 2) first identified 33 publications not strictly related to GIS

Fig. 3 Application of inclusion criteria process to select publications relevant to the review analysis

179

S. Grigolato et al.

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

application nor to forest operations or forest road network planning.

Secondly, other 173 publications were excluded from the analysis because they only partially conformed to the review subjects (GIS application concerns mainly forest planning, forest management and/ or forest landscape and less forest operations and road network planning) or just slightly conformed to the review subjects (GIS applications are not the main aim of the paper and slightly related to forest operations and road network planning).

The remaining 166 unique peer reviewed publications were analysed in detail to identify the most relevant and appropriate publications for the reviews. As a final point, 82 unique peer reviewed publications proved to be eligible for the detailed review analysis.

In order to clearly determine the review decision process, Fig. 3 summarized, in the form of a flow chart, the logic of the inclusion criteria process used in this review.

3.3 Relevant publications 3.3.1 Subjects distribution Across the entire period covered, the number of relevant peer reviewed publications with the subject clearly in conformity with the review topics (GIS applications in forest operations and road network planning) resulted in 82 unique papers after the application of the criteria section.

The average number of publications per year was 4.05, when calculated for the entire period 1996–2015.

The average is statistically different (Paired t-test, α=0.05) when it is calculated separately for the first decade 1996–2005 (1.1 publications per year) and for the second decade 2006–2015 (7.0 publications per year).

The 82 unique publications were published into 53 different scientific journals or conference proceedings (7.3% of the selected publications) written by 67 different first authors.

In terms of impact, 70.7% of the publications were cited at least one time across the entire period covered.

Within the same period, 82 publications generated 874 citations in 20 years with an average of 43.7 citations per year in total and an average of 10.6 citations per publication.

About 58% (48) of the relevant peer reviewed publications were strongly related to forest road and/or forest road planning subject, while 39% (31) were assigned to subject of forest operations. The remaining 2% (2) consisted of publications focusing mainly on

180

assessing hydrologic integration of extensive loggingroad network with the stream network (Wemple et al.

1996, Murphy et al. 2007) as crucial information to improve forest operations planning, road layout and construction, culvert location and size and off-road drive. Consequently, the publications were associated to a third subject, entitled »water control«.

Peer reviewed publications related to »Harvesting and transportation engineering« accounted for 60% (49), while publications related specifically to »Forest operations management« accounted for the remaining 40% (33).

The analysis of each reference related to »Harvesting and transportation engineering« through the tree sub-subjects »Road networks and transportation«, »Road engineering and management« and »Harvesting and processing systems« reported that »Road networks and transportation« was the most frequent description accounting for 36 (73%) out of 49 publications, while »Road engineering and management« and »Harvesting and processing systems« accounted, respectively, for 11 (22%) and for 2 (4%) out of 49 publications (Table 1). The high number of publications on »Road networks and transportation« also highlighted that this sub-subject was very attractive for the researchers, as it was absolutely the most recurring with 44 publications out of 82 publications (44% of the total). It highlighted the strong interest of the researchers in designing efficient forest road networks and wood transportation systems within the investigated period.

Within the subject »Forest operations management«, »Operations systems analysis and modelling« overlooked the sub-subject »Supply chain management«. In total »Operations systems analysis and modelling« accounted for 24 (73%) publications out of 33, while »Supply chain management« accounted for 9 (11%) publications. It should also be emphasised that »Operations systems analysis and modelling« was the second most frequently occurring sub-subject among all the selected publications.

3.3.2 Planning level In forestry, planning and decision making is traditionally performed at strategic, tactical or operational level depending on the time scale to which they are applied (D’Amours et al. 2008). From the selected publications, the GIS application in forest operations and road network planning seemed to support all the planning levels. It should be noted that almost 50% of the publications reported GIS application to support tactical planning.

In the case of »Road networks and transportation«, GIS was extensively used to support tactical analysis Croat. j. for. eng. 38(2017)2

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

Table 1 Selected publications in the period 1996–2015 broken down according to their main subjects Subjects and sub-subjects

n.

% subject

% total

Harvesting and transportation engineering Road networks and transportation

36

73.5

43.9

Road engineering and management

11

22.4

13.4

Harvesting and processing systems

2

4.1

2.4

Sub-total

49

100

–

Forest operations management Operations systems analysis and modelling

24

72.7

29.3

Supply chain management

9

27.3

11.0

Sub-total

33

100

–

Total

82

–

100

on the existing forest road network and to provide useful information for the decision-makers on further development of road network (Table 2). For example, Pentek et al. (2005) propose one of the first approaches to analyse the quantity and quality of the existing primary forest road network in different forest management units in Croatia to support forest manager in allocating efficiently the resources to specific forest areas. Again Krč and Beguš (2013) provided a GIS model to support tactical decision on identifying inaccessible forest areas by density of forest roads. In this case, the model is based on the analysis of distances between the existing network of public and forest roads and inaccessible forest areas. At tactical level,

S. Grigolato et al.

GIS also focused on road maintenance (Karlsson et al.

2006, Pellegrini et al. 2013, Talebi et al. 2015).

Some examples focused on the use of GIS to support decisions on solving issues related to the increment of volume of heavy trucks and the related rise of costs for road maintenance (Millot et al. 2006, Dowdle and Douglas 2007, Grigolato et al. 2013). Different applications were also related to transportation and logistics used for energy wood supply chains (Ranta 2005, Kanzian et al. 2009, Emer at al. 2011, Friso et al.

2011, Röser et al. 2011, Tahvanainen and Anttila 2011, Zambelli et al. 2012). They were also used for roundwood logistics (Gerasimov et al. 2008) and to support analysis on speed and fuel consumption (Holzleitner et al. 2010, Holzleitner et al. 2011, Sosa et al. 2015).

GIS used to assist in tactical planning and scheduling adapted to forest operations issues was also investigated. Kuhmaier and Stampfer (2010) provided a Decisions Support System (DSS) based on GIS and specifically designed to support timber harvesting decisions by comparing harvesting systems considering stakeholders interests and environmental circumstances. Synek and Klimánek (2014) proposed the use of GIS for multi-criteria evaluation of environmentally friendly skidding technologies.

Despite the prevailing use of applications to support tactical planning, a significant number of recent publications have also turned to the possibility of applying GIS successfully at operational level. In particular, the GIS applications focused on forest operations management in terms of optimization, productivity and safety analysis (Alam et al. 2012, Sterenczak and Moskalik 2014, Hiesl et al. 2015), as well as on forest road networks and transportation in

Table 2 Selected publications in the period 1996–2015 broken down according to their main subjects related to planning level Subjects and sub-subjects

Strategic, n. (%)

Tactical, n. (%)

Operational, n. (%)

Total, n. (%)

Harvesting and transportation engineering Road networks and transportation

5 (13.9)

23 (63.9)

8 (22.2)

36 (100)

Road engineering and management

–

4 (36.4)

7 (63.6)

11 (100)

Harvesting and processing systems

1 (50.0)

–

1 (50.0)

2 (100)

Sub-total

6 (12.2)

27 (55.1)

16 (32.7)

49 (100)

Forest operations management Operations systems analysis and modelling

6 (25.0)

10 (41.7)

8 (33.3)

24 (100)

Supply chain management

5 (55.6)

3 (33.3)

1 (11.1)

9 (100)

Sub-total

11 (33.3)

13 (39.4)

9 (27.3)

33 (100)

Total

45 (73.2)

21 (9.7)

16 (17.1)

82 (100)

Croat. j. for. eng. 38(2017)2

181

S. Grigolato et al.

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

terms of single road design and in terms of decision support (Bont and Heinimann 2012, Contreras et al.

2012, Ciesa et al. 2014, Craven and Wing 2014).

3.3.3 GIS approach, analysis and outputs GIS analyses are commonly based on process models. These can be static, such as defining the roughness index by a Digital Elevation Model (DEM), or dynamic in time and space as with a wildfire simulation model. The identified publication reported only GIS static model. According to Table 3, in the topics »Harvesting and transportation engineering« and »Forest operations management«, GIS model is not adequately applied in approaching the analysis by dynamic reiteration process through time and space.

As a result, simple GIS simulations, just based on changing model inputs or parameters to obtain different outputs in time and space, are more common.

Again in the in the period 1996–2015, GIS models and derived applications were mainly based on deterministic variables and functions and rarely included random/stochastic variables as found in Dean (1997) and in Najafi and Richards (2013).

Most of the described and applied models were cartographic models (about 2.05 publications per year for the entire covered period and about 3.7 publications per year in the period 2006–2015) based on numerical operations combining temporally static data with spatial data (Table 3). Most of them were mainly spatial elaborations aimed at calculating the distance from the forest road network and terrain slope only based on vector analysis (Pentek et al. 2010, Pičman et

al. 2011, Hayati et al. 2012), either integrated raster or vector analysis (Kluender et al. 1998, Najafi et al. 2008, Pentek et al. 2008, Cavalli et al. 2011, Emer et al. 2011, Nakahata et al. 2014) to provide harvesting and extraction map indicating the most suitable technology or consideration about the efficiency in terms of layout and location of the primary and/or secondary forest road network.

A consistent number of publications (almost 11%) showed interest in combining GIS tools and multicriteria decision analysis to provide Decision Support System to improve forest road planning and maintenance (Abdi et al. 2009, Norizah and Hasmadi 2012, Çalişkan 2013, Pellegrini et al. 2013, Tampekis et al.

2015) and to evaluate the most suitable harvesting system (Mihelič and Krč 2009, Kuhmaier and Stampfer 2010, Enache et al. 2013).

Examples of publications, where GIS was used in conjunction with equations involving different variables and constraints, were proposed by Contreras and Chung (2007) to determine the optimal landing location for ground-based timber harvesting by minimising total skidding and spur road costs, or by Epstein et al. (2006) to find a design of the road network that minimises the cost of installation and operation of harvest machinery, road construction, and timber transport. Again Stückelberger et al. (2006) provided a GIS model integrating different mathematical procedures to estimate accurately the spatial variability of road life-cycle costs, based on terrain surface properties as well as geological properties of the subsoil. Later, Stückelberger et al. (2007) proposed a mathematical

Table 3 Selected publications in the period 1996–2015 broken down according to their main subjects in relation to the type of GIS analysis and type of model output Model Cartographic, n. (%)

Spatial-temporal, n. (%)

Network, n. (%)

Total, n. (%)

Harvesting and transportation engineering Road networks and transportation

19 (52.8)

13 (36.1)

4 (11.1)

36 (100)

Road engineering and management

6 (54.5)

3 (27.3)

2 (18.2)

11 (100)

Harvesting and processing systems

–

1 (50.0)

1 (50.0)

2 (100)

25 (51.0)

17 (34.7)

7 (14.3)

49 (100)

Sub-total

Forest operations management Operations systems analysis and modelling

14 (58.3)

8 (33.3)

2 (8.3)

24 (100)

Supply chain management

2 (22.2)

–

7 (77.8)

9 (100)

Sub-total

20 (60.6)

4 (12.1)

9 (27.3)

33 (100)

Total

60 (73.2)

8 (12.1)

14 (17.1)

82 (100)

182

Croat. j. for. eng. 38(2017)2

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

graph model to be integrated into GIS to derive a road network that is optimal in terms of its construction costs.

3.3.4 Data structure and data Only 5% of the publications did not explicitly define the type of data used in the elaboration. 33% of the publications used exclusively raster data, while 20% used only vector data, while 43% used both types of data.

In the case of raster data, 16% of the publications reported a cell size larger than 10 m and only in 8% of cases equal to or lower than 1 m. However, the growing availability of LiDAR data is contributing to increase the use of high accuracy Digital Elevation Model, as well as the use of Digital Canopy Model for pinpointing tree crowns and stems. Heinimann and Breschan (2012) explored a LiDAR-data-based approach to improve the sourcing of stands to be harvested by integrating the Airborne Laser Scanner information into a GIS system and by characterizing the tree attributes that were required for stand bucking optimization. Again, Sterenczak and Moskalik (2014) reported that combining the high-accurate Digital Elevation Model and Digital Canopy Model with GIS is appropriate for determining the optimal or near-optimal locations of forest skid trails. Recently, Strandgard et al. (2014) have shown the possibility to analyse the impact of slope on productivity of a self-levelling processor by inputting into a GIS the data obtained from the GNSS system of the machine and a Digital Elevation Model derived from LiDAR data.

4. Conclusions It is well recognized that considerable scope exists for the application of GIS technology to aid planning, design and management of forestry. The literature review conducted for this study showed that GIS has also been applied successfully and unambiguously to harvesting and transportation engineering as well as to forest operations management.

High accuracy 3D model dataset of terrain and canopy by Aerial Laser Scanner or Terrestrial Laser Scanner, as well as photogrammetric technology, are nowadays available at lower cost than in the past. The perspective to use 2.5D or 3D spatial analysis is nowadays a reality in forestry (Vauhkonen et al. 2014) and in forest engineering applications (Alam et al. 2012, Strandgard et al. 2014, Koreň et al. 2015).

Besides the advantage of the availability and use of high resolution data, a prominent future development of GIS applications in forest operations and road netCroat. j. for. eng. 38(2017)2

S. Grigolato et al.

work planning could be a fully integrated 4D spatial dataset and Virtual Reality (VR) system (Roßmann et al. 2013 and Roßmann et al. 2016). In fact, the evolution of GIS techniques and the improvement of data resolution, in conjunction with the skill of researcher to integrate complex mathematical model and VR, could provide new development opportunities for the forest engineering community.

5. References Abdi, E., Majnounian, B., Darvishsefat, A., Mashayekhi, Z., Sessions, J., 2009: A GIS-MCE based model for forest road planning. Journal of Forest Science 55(4): 171–176.

Akay, A.E., Oguz, H., Karas, I.R., Aruga, K., 2009: Using LiDAR technology in forestry activities. Environmental Monitoring and Assessment 151(1–4): 117–125.

Alam, M.M., Strandgard, M.N., Brown, M.W., Fox, J.C., 2012: Improving the productivity of mechanised harvesting systems using remote sensing. Australian Forestry 75(4): 238– 245.

Babapour, R., Naghdi, R., Salehi, A., Ghajar, I., 2014: A decision support system for allocation of mountain forest roads based on ground stability. Arabian Journal for Science and Engineering 39(1): 199–205.

Becker, R.H., 1976: Computer visualizations as aids to forest recreation management. Journal of Forestry 74(12): 820–823.

Bettinger, P., Boston, K., Siry, J.P., Grebner, D.L., 2009: Forest management and planning. Academic Press, New York, 331 p.

Bont, L.G., Heinimann, H.R., Church, R.L., 2015: Concurrent optimization of harvesting and road network layouts under steep terrain. Annals of Operations Research 232(1): 41–64.

Bont, L., Heinimann, H.R., 2012: Optimum geometric layout of a single cable road. European Journal of Forest Research 131(5): 1439–1448.

Burrough, P.A., McDonnell, R., 1998: Principles of geographical information systems. Oxford university press, 333 p. Çalişkan, E., 2013: Planning of forest road network and analysis in mountainous area. Life Science Journal 10(2): 2456– 2465.

Cavalli, R., Grigolato, S., 2010: Influence of characteristics and extension of a forest road network on the supply cost of forest woodchips. Journal of Forest Research 15(3): 202–209.

Cavalli, R., Grigolato, S., Pellegrini, M., 2011: Determination of the forest road network influence on the supply chain for firewood production by Discrete Event Simulation. Journal of Agricultural Engineering 42(4): 41–48.

Ciesa, M., Grigolato, S., Cavalli, R., 2014: Analysis on vehicle and walking speeds of search and rescue ground crews in mountainous areas. Journal of Outdoor Recreation and Tourism 5–6: 48–57.

Contreras, M., Aracena, P., Chung, W., 2012: Improving Accuracy in Earthwork Volume Estimation for Proposed Forest

183

S. Grigolato et al.

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

Roads Using a High-Resolution Digital Elevation Model.

Croatian Journal of Forest Engineering 33(1): 125–142.

Contreras, M., Chung, W., 2007: A computer approach to finding an optimal log landing location and analyzing influencing factors for ground-based timber harvesting. Canadian Journal of Forest Research 37(2): 276–292.

Craven, M., Wing, M.G., 2014: Applying airborne LiDAR for forested road geomatics. Scandinavian Journal of Forest Research 29(2): 174–182.

Curran, L.M., Trigg, S.N., McDonald, A.K., Astiani, D., Hardiono, Y.M., Siregar, P., Caniago, I., Kasischke, E., 2004: Lowland forest loss in protected areas of Indonesian Borneo. Science 303(5660): 1000–1003.

D’Amours, S., Rönnqvist, M., Weintraub, A., 2008: Using operational research for supply chain planning in the forest products industry. INFOR: information systems and operational research 46(4): 265–281.

Dean, D.J., 1997: Finding optimal routes for networks of harvest site access roads using GIS-based techniques. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 27(1): 11–22.

Dowdle, L., Douglas, R., 2007: Log truck transportation on public roads in New Zealand: regional network analysis with geographic information systems. Transportation Research Record 1989(1): 34–40.

Dunningham, A., Thompson, S., 1989: Use of geographic information systems in New Zealand forestry applications.

Commonwealth Forestry Review 68(3): 203–213.

Dupire, S., Bourrier, F., Monnet, J.-M., Berger, F., 2015: Sylvaccess – A model for automatically mapping accessibility to forests. Revue Forestière Française 67(2): 11–126.

Emer, B., Grigolato, S., Lubello, D., Cavalli, R., 2011: Comparison of biomass feedstock supply and demand in Northeast Italy. Biomass Bioenergy 35(8): 3309–3317.

Enache, A., Kühmaier, M., Stampfer, K., Ciobanu, V.D., 2013: An integrative decision support tool for assessing forest road options in a mountainous region in Romania. Croatian Journal of Forest Engineering 34 (1): 43–60.

Epstein, R., Weintraub, A, Sapunar, P., Nieto, E., Sessions, J.B., Sessions, J., Musante, H., 2006: A Combinatorial Heuristic Approach for Solving Real-Size Machinery Location and Road Design Problems in Forestry Planning. Operations Research 54(6): 1017–1027.

Falagas, M.E., Pitsouni, E.I., Malietzis, G.A., Pappas, G., 2007.

Comparison of PubMed, Scopus, Web of Science, and Google Scholar: strengths and weaknesses. The FASEB Journal 22(2): 338–342.

Friso, D., Grigolato, S., Cavalli, R., 2011: Energetic and exergetic analysis of steam production for the extraction of coniferous essential oils. Biomass Bioenergy 35(9): 4045–4056.

Gerasimov, Y., Sokolov, A., Karjalainen, T., 2008: GIS-Based decision-support program for planning and analyzing shortwood transport in russia. Croatian Journal of Forest Engineering 29(2): 163–175.

184

Grigolato, S., Pellegrini, M., Cavalli, R., 2013: Temporal analysis of the traffic loads on forest road networks. iForest – Biogeosciences and Forestry 6(4): 255–261.

Grigolato, S., Panizza, S., Pellegrini, M., Ackerman, P., Cavalli, R., 2015: Light-lift helicopter logging operations in the Italian Alps: a preliminary study based on GNSS and a video camera system. Forest Science and Technology 12(2): 88–97.

Haas, J., Hagge Ellhöft, K., Schack-Kirchner, H., Lang, F., 2016: Using photogrammetry to assess rutting caused by a forwarder—A comparison of different tires and bogie tracks.

Soil and Tillage Research 163: 14–20.

Hayati, E., Majnounian, B., Abdi, E., 2012: Qualitative evaluation and optimization of forest road network to minimize total costs and environmental impacts. iForest – Biogeosciences and Forestry 5(3): 121–125.

Hayati, E., Majnounian, B., Abdi, E., Sessions, J., Makhdoum, M., 2013: An expert-based approach to forest road network planning by combining Delphi and spatial multi-criteria evaluation. Environmental Monitoring and Assessment 185(2): 1767–1776.

Heinimann, H.R., Breschan, J., 2012: Pre-harvest assessment based on LiDAR Data. Croatian Journal of Forest Engineering 33(2): 169–180.

Hiesl, P., Waring, T.M., Benjamin, J.G., 2015: The effect of hardwood component on grapple skidder and stroke delimber idle time and productivity – An agent based model. Computers and Electronics in Agriculture 118: 270–280.

Holzleitner, F., Kanzian, C., Stampfer, K., 2011: Analyzing time and fuel consumption in road transport of round wood with an onboard fleet manager. European Journal of Forest Research 130(2): 293–301.

Kanzian, C., Holzleitner, F., Stampfer, K., Ashton, S., 2009: Regional energy wood logistic – optimizing local fuel supply.

Silva Fennica 43(1): 113–128.

Karlsson, J., Rönnqvist, M., Frisk, M., 2006: RoadOpt: a decision support system for road upgrading in forestry. Scandinavian Journal of Forest Research 21(S7): 5–15.

Khalil, H., Peters, M., Godfrey, C.M., Mcinerney, P., Soares, C.B., Parker, D., 2016: An evidence-based approach to scoping reviews. Worldviews on Evidence-Based Nursing 13(2): 118–123.

Kluender, R., Weih, R., Corrigan, M., Pickett, J., 1998: The use of a geographic information system in harvest landing location for ground-based skidding operations. Forest Products Journal 50(3): 87–92.

Koreň, M., Slančík, M., Suchomel, J., Dubina, J., 2015: Use of terrestrial laser scanning to evaluate the spatial distribution of soil disturbance by skidding operations. iForest – Biogeosciences and Forestry 8(3): 386–393.

Krč, J., Beguš, J., 2013: Planning forest opening with forest roads. Croatian Journal of Forest Engineering 34(2): 217–228.

Kühmaier, M., Kanzian, C., Stampfer, K., 2014: Identification of potential energy wood terminal locations using a spatial multicriteria decision analysis. Biomass and Bioenergy 66: 337–347.

Croat. j. for. eng. 38(2017)2

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

S. Grigolato et al.

Kühmaier, M., Stampfer, K., 2010: Development of a multiattribute spatial decision support system in selecting timber harvesting systems. Croatian Journal of Forest Engineering 31(2): 75–88.

Norizah, K., Mohd Hasmadi, I., 2012: Developing priorities and ranking for suitable forest road: Allocation using Analytic Hierarchy Process (AHP) in peninsular Malaysia. Sains Malaysiana 41(10): 1177–1185.

Li, R., Bettinger, P., Danskin, S., Hayashi, R., 2007: A historical perspective on the use of GIS and remote sensing in natural resource management, as viewed through papers published in north american forestry journals from 1976 to 2005. Cartographica 42(2): 165–178.

Olivera, A., Visser, R., 2016: Development of forest-yield maps generated from Global Navigation Satellite System (GNSS)-enabled harvester StanForD files: preliminary concepts. New Zealand Journal of Forestry Science 46(1): 1–10.

Longley, P., Goodchild, M.F., Maguire, D.J., Rhind, D.W., 2011: Geographic information science and system. 4th Edition.

New York: John Wiley Sons, Ltd. 496 p.

Malczewski, J., 1999: GIS and multicriteria decision analysis.

New York: John Wiley Sons, Ltd. 408 p.

Vauhkonen, J., Maltamo, M., McRoberts, R.E., Næsset, E., 2014: Introduction to forestry applications of airborne laser scanning. In Maltamo, M., Næsset, E., Vauhkonen, J., Editors.

Forestry applications of airborne laser scanning: concepts and case studies. Dordrecht: Springer, 1–16.

Macrì, G., Zimbalatti, G., Russo, D., Proto, A.R., 2016: Measuring the mobility parameters of tree-length forwarding systems using GPS technology in the Italian Apennines.

Agronomy Research 14(3): 836–845.

McDaniel, M.W., Nishihata, T., Brooks, C.A., Salesses, P., Iagnemma, K., 2012: Terrain classification and identification of tree stems using ground-based LiDAR. Journal of Field Robotics 29(6): 891–910.

Mihelič, M., Krč, J., 2009: Analysis of inclusion of wood forwarding into a skidding model. Croatian Journal of Forest Engineering 30(2): 113–125.

Millot, M., Torre, F., Labbe, S., 2006: An evaluation of timber flows over the forest road system using a geographic information system. Application to road maintenance and logging cost assessment. Revue Forestiere Francaise 58(2): 125–140.

Mologni, O., Grigolato, S., Cavalli, R., 2016: Harvesting systems for steep terrain in the Italian Alps: state of the art and future prospects. Contemporary Engineering Sciences 9(25): 1229–1242.

Murphy, P.N.C., Ogilvie, J., Castonguay, M., Meng, F.R., Arp, P.A., 2007: Verifying calculated flow accumulation patterns of mapped and unmapped forest streams by culvert location.

Forestry Chronicle 83(2): 198–206.

Olivera, A., Visser, R., Acuna, M., Morgenroth, J., 2016: Automatic GNSS-enabled harvester data collection as a tool to evaluate factors affecting harvester productivity in a Eucalyptus spp. harvesting operation in Uruguay. International Journal of Forest Engineering 27(1): 15–28.

Pellegrini, M., Grigolato, S., Cavalli, R., 2013: Spatial multicriteria decision process to define maintenance priorities of forest road network: An application in the italian alpine region. Croatian Journal of Forest Engineering 34(1): 31–42.

Pellegrini, M., Ackerman, P., Cavalli, R., 2013: On-board computing in forest machinery as a tool to improve skidding operations in South African softwood sawtimber operations.

Southern Forests: a Journal of Forest Science 75(2): 89–96.

Pentek, T., Nevečerel, H., Dasović, K., Poršinsky, T., Šušnjar, M., Potočnik, I., 2010: Analysis of secondary relative openness in hilly areas as a basis for selection of winch rope length.

Sumarski list 134(5–6): 241–248.

Pentek, T., Nevečerel, H., Poršinsky, T., Pičman, D., Lepoglavec, K., Potočnik, I., 2008: Methodology for development of secondary forest traffic infrastructure cadastre. Croatian Journal of Forest Engineering 29 (1): 75–83.

Pentek, T., Pičman, D., Potočnik, I., Dvorščak, P., Nevečerel, H. 2005: Analysis of an existing forest road network. Croatian Journal of Forest Engineering 26(1): 39–50.

Pičman, D., Pentek, T., Nevečerel, H., Papa, I., Lepoglavec, K., 2011: Possibilities of application of relative openness in secondary forest opening of slope forests in Croatia. Croatian Journal of Forest Engineering 32(1): 417–427.

Pierzchała, M., Talbot, B., Astrup, R., 2014: Estimating soil displacement from timber extraction trails in steep terrain: application of an unmanned aircraft for 3D modelling. Forests 5(6): 1212–1223.

Pierzchała, M., Talbot, B., Astrup, R., 2016: Measuring wheel ruts with close-range photogrammetry. Forestry 89(4): 383– 391.

Najafi, A., Richards, E.W., 2013: Designing a forest road network using mixed integer programming. Croatian Journal of Forest Engineering 34(1): 17–30.

Pirotti, F., Grigolato, S., Lingua, E., Sitzia, T., Tarolli, P., 2012: Laser Scanner Applications in Forest and Environmental Sciences. Italian Journal of Remote Sensing 44(1): 109–123.

Najafi, A., Sobhani, H., Saeed, A., Makhdom, M., Mohajer, M.M., 2008: Planning and assessment of alternative forest road and skidding networks. Croatian Journal of Forest Engineering 29(1): 63–73.

Proto, A.R., Grigolato, S., Mologni, O., Macrì, G., Zimbalatti, G., Cavalli, R., 2016: Modelling noise propagation generated by forest operations: a case study in Southern Italy. Procedia – Social and Behavioral Sciences 223: 841–848.

Nakahata, C., Uemura, R., Saito, M., Kanetsuki, K., Aruga, K., 2014: Estimating harvest costs and projecting quantities of logging residues for small-scale forestry in Nasushiobara, Tochigi Prefecture, Japan. Journal of Forestry Research 25(4): 965–974.

Ranta, T., 2005: Logging residues from regeneration fellings for biofuel production-a GIS-based availability analysis in Finland. Biomass and Bioenergy 28(2): 171–182.

Croat. j. for. eng. 38(2017)2

Reisinger, T.W., Davis, C.J., 1986: A map-based decision support system for evaluating terrain and planning timber har-

185

S. Grigolato et al.

GIS Applications in Forest Operations and Road Network Planning: an Overview ... (175–186)

vests. Transactions of the ASAE-American Society 29(5): 1199–1203.

Röser, D., Sikanen, L., Asikainen, A., Parikka, H., Väätäinen, K., 2011: Productivity and cost of mechanized energy wood harvesting in Northern Scotland. Biomass and Bioenergy 35(11): 4570–4580.

Roßmann, J., Hoppen, M., Bücken, A., 2013: Semantic world modelling and data management in a 4D forest simulation and information system. ISPRS – International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XL-2/W2, 65–72.

Stückelberger, J., Heinimann, H.R., Chung, W., 2007: Improved road network design models with the consideration of various link patterns and road design elements. Canadian Journal of Forest Research 37(11): 2281–2298.

Synek, M., Klimánek, M., 2014: Proposal of using GIS for multi-criteria evaluation of environmentally friendly use of skidding technologies in forestry. Journal of Forest Science 60(2): 51–60.

Tahvanainen, T., Anttila, P., 2011: Supply chain cost analysis of long-distance transportation of energy wood in Finland.

Biomass and Bioenergy 35(8): 3360–3375.

Roßmann, J., Bücken, A., Hoppen, M., Priggemeyer, M., 2016.

Integrating Virtual Reality, Motion Simulation and a 4D GIS. Research in Urbanism Series 4(1): 25–42.

Talebi, M., Majnounian, B., Abdi, E., Tehrani, F.B., 2015: Developing a GIS database for forest road management in Arasbaran forest, Iran. Forest Science and Technology 11(1): 27–35.

Sitzia, T., Campagnaro, T., Grigolato, S., 2016: Ecological risk and accessibility analysis to assess the impact of roads under Habitats Directive. Journal of Environmental Planning and Management 59(12): 2251–2271.

Tampekis, S., Sakellariou, S., Samara, F., Sfougaris, A., Jaeger, D., Christopoulou, O., 2015: Mapping the optimal forest road network based on the multicriteria evaluation technique: the case study of Mediterranean Island of Thassos in Greece. Environmental Monitoring and Assessment 187(11): 1–17.

Sosa, A., McDonnell, K., Devlin, G., 2015: Analysing performance characteristics of biomass haulage in ireland for bioenergy markets with GPS, GIS and fuel diagnostic tools.

Energies 8(10): 12004–12019.

Sterenczak, K., Moskalik, T., 2014: Use of LIDAR-based digital terrain model and single tree segmentation data for optimal forest skid trail network. iForest – Biogeosciences and Forestry 8(5): 661–667.

Strandgard, M., Alam, M., Mitchell, R., 2014: Impact of slope on productivity of a self-levelling processor. Croatian Journal of Forest Engineering 35(2): 193–200.

Stückelberger, J.A., Heinimann, H.R., Burlet, E.C., 2006: Modeling spatial variability in the life-cycle costs of low-volume forest roads. European Journal of Forest Research 125(4): 377–390.

Wemple, B.C., Jones, J.A., Grant, G.E., 1996: Channel Network Extension By Logging Roads in Two Basins, Western Cascades, Oregon. Journal of the American Water Resources Association 32(6): 1195–1207.

Wolf, C., Klein, D., Weber-Blaschke, G., Richter, K., 2016: Systematic review and meta-analysis of Life Cycle Assessments for wood energy services. Journal of Industrial Ecology 20(4): 743–763.

Worral, L., 1991: Spatial analysis and spatial policy using Geographic Information Systems. London: Belhaven press.

Zambelli, P., Lora, C., Spinelli, R., Tattoni, C., Vitti, A., Zatelli, P., Ciolli, M., 2012: A GIS decision support system for regional forest management to assess biomass availability for renewable energy production. Environmental Modelling and Software 38: 203–213.

Authors’ addresses:

Received: January 01, 2017 Accepted: April 07, 2017

186

Assoc. prof. Stefano Grigolato, PhD. * e-mail: stefano.grigolato@unipd.it Omar Mologni, MSc. e-mail: omar.mologni@phd.unipd.it Prof. Raffaele Cavalli, PhD. e-mail: raffaele.cavalli@unipd.it University of Padova Department of Land, Environment, Agriculture and Forestry Viale dell’Università 16 35020 Legnaro ITALY * Corresponding author Croat. j. for. eng. 38(2017)2

Subject review

Challenges in Road Construction and Timber Harvesting in Japan Hideo Sakai Abstract Japan Islands fundamentally consisted of the accretionary wedges. The stratum of accretionary wedges inclined, and it made dip slope and opposite slope. Granite magma intruded among accretionary wedges, and later granite itself became decomposed soil. Much annual precipitation of nearly 2000 mm, caused by monsoons, attacks forest roads, so innovative technologies for road construction are required. From ancient days, numerous circular slips and deepseated landslides occurred. For the inclined stratum at Shimanto belt composed of sedimentary rocks, deep excavation for preparing structural road foundation was invented. After deep excavation until the depth that ensured the road width, about 30 cm thick blocks of compacted subsoil were piled up. It is difficult to make high cutting slope on smooth soil area, and the retaining wall with log structure is effective both for cutting and fill slopes at spur roads.

Underground water that comes out through crushing zones is often troublesome. When crashing zone and dip slope overlap, the valley head is unstable causing landslide. This makes this area inadequate for road construction. Some forest road retaining technologies have been introduced recently in Japan. When crossing the crushing zone by simple structures, cage or L-shaped steel retaining wall is effective because the weight of stones and rocks in the structure presses down the road foundation with high water permeability. It is important to adopt the most appropriate road construction method in accordance with the soil, geology and terrain conditions to provide sustainable forestry.

Keywords: accretionary wedge, crashing zone, dip slope, reinforced soil wall, rolled grade drainage

1. Introduction Planted forests have been increased since 1950s in Japan, and they have reached 10 million ha, which is equivalent to a quarter of the land area. About half of these forests are more than fifty years old, and they can be harvested any time by final cutting. However, timber demand in Japan is decreasing year by year to 70 million cubic meters, while domestic timber supply is only 30% in spite of the potential of 100% based on the statistics of the annual growth increment (Japanese Forestry Agency 2015). There are so many reasons for this unbalance of supply and demand, and these are primarily the low market price of timber and difficult terrain. From the view point of forest engineering, enhancement and upgrading of road network are still necessary for solving the problem. Simultaneously, new forestry mechanization must be established both for timber and fuel biomass production. These infraCroat. j. for. eng. 38(2017)2

structure developments can strengthen the management of logging contractors and create employment of forestry workers in mountainous areas.

In this review article, challenges in road construction related to timber harvesting are introduced and discussed from technological and economic point of view for future applications.

2. Geological background of road construction in Japan Originally, Japan Islands were formed by accretionary wedges from ocean plate at the east end of the Asian continent until 65 million years ago (Taira 1990).

At that time, Japan belonged to the Asian continent.

Accretionary wedges were named after the stratum attached to the continent when the ocean plate sank under the continent plate. Japan Islands fundamentally

187

H. Sakai

Challenges in Road Construction and Timber Harvesting in Japan (187â&#x20AC;&#x201C;195)

consisted of the accretionary wedges. The stratum of accretionary wedges inclined, therefore, terrains were classified into either dip or opposite slope. A dip slope is a slope in the same direction of the inclined stratum, which generally makes filling difficult. The dip slope on a mountain side is generally gentle having long and smooth river systems (Suzuki 2000), but landslides and collapse of the cutting slope frequently occur, so that roads on a dip slope require higher maintenance costs.

A dip slope is often covered with thick colluvial soil, and such accumulated soil is prone to slide after a heavy rain. On the contrary, a road planned on the opposite slope, with short and parallel waterfalls (Suzuki 2000), is winding in short sections and may take much time to construct, but is stable once constructed.

Then, the western part of Japan, the eastern part except Hokkaido Island, and the western and eastern parts of Hokkaido left the Asian continent until 65 million years ago, and combined and formed Japan Islands and Hokkaido Island as they are today.

In these processes of ancient times, debris flow of deposit under sea and on the ground shaved accretionary wedges, and granite magma intruded among accretionary wedges in the interior. It brought heat denaturation, and later granite itself became decomposed soil. Volcanic ash also covered thickly all over the Japan Islands.

In the western part from Kyushu Island to Kanto district, and in the southern seashore that is newer, there are geological belts named Shinmato belt, Chichibu belt, Sambagawa belt, Ryouke belt, which suffered heat denaturation by granite magma, granite mountains, Sangun belt, and Hida belt where heat denaturation still occurs (Saito 1992). Among Sambagawa and Ryouke belts, median tectonic line lies from Kyushu Island to Kanto district. Geology of the eastern part of Honshu Island is very complicated because it was under the sea for a long time and because there were a lot of volcanoes, which erupted a large quantity of volcanic ash.

In addition to the above complicated geological history, much annual precipitation of nearly 2000 mm, caused by monsoons, attacks forests and forest roads, so high and innovative technologies are required for road construction (Yuasa and Sakai 2012), while growing trees and providing rice cultivation with water from mountainous forest area.

Fig. 1 Shimanto belt showing dip slope and opposite slope ones faced to south or Pacific Ocean is opposite slope (Fig. 1). Numerous circular slips and deep-seated landslides occurred from ancient days especially on the slopes faced to north. It is still risky to construct roads on the dip slope. Therefore, downhill yarding by skyline system is indispensable. The possible skyline systems are conventional yarding systems, eg.

Endless Tyler system (Samset 1985). Recently, a domestic interlocked three drums tower yarder, which could operate both downhill and uphill, was developed (Fig. 2).

For the inclined stratum in Shimanto belt, composed of sedimentary rocks, deep excavation for preparing structural foundation of the road was developed (Fig. 3). After deep excavation that ensured the depth required for the road width, about 30 cm thick blocks of compacted subsoil were piled up. This method is suitable for gravelly soil, which is the main soil

3. Road construction in Shimanto belt Shimanto belt is the latest belt and shows typical features of accretionary wedges. The slopes faced to north or Japan Sea are generally dip slope, and the

188

Fig. 2 Interlocked three drums tower yarder which can operate both downhill and uphill Croat. j. for. eng. 38(2017)2

Challenges in Road Construction and Timber Harvesting in Japan (187â&#x20AC;&#x201C;195)

Fig. 3 Road construction method by deep excavation for preparing structural foundation

H. Sakai

Fig. 5 Small forwarder

light trucks, which are typical of the forestry and agriculture in Japan (see Fig. 8 below). They are called operation roads in Japan because of their function of forestry operation (Sakai 2004).

4. Retaining wall with log structure for Chichibu belt and decomposed granite area

Fig. 4 Switch curve combining the function of both switch back and hair pin curve type in Shimanto belt, while a small portion of clay serves as bondage. Surface soil of natural slope was kept during construction and used for the surface of the filling slope for providing early green surface because it contained buried seeds. Fig. 3 shows the section constructed eight years ago when the photo was taken.

It is easy to understand why the boundary of the road surface and cutting slope was sharp. The surface of the filling slope was quickly green due to the surface soil.

New technology named switch curve was also developed combining the function of both switch back and hair pin curve at steep slopes (Fig. 4). Road bed was made overhang in front of the curve to secure the space of landing using the soil of cutting slope immediately after the curve turning. These innovative roads cannot be used for logging trucks, but are useful for small scale forestry with small forwarders (Fig. 5) and Croat. j. for. eng. 38(2017)2

Chichibu belt is the north neighbor of Shimanto belt. This belt collapsed from Sambagawa belt and deposited under the sea, so that the geologic stratum was not formed and the soil was smooth. Big land slide sometimes occurred by heavy precipitation, because smooth soil around rocks and stones fluidized by much rain, and once the collapse occurred, it continued from the bottom to the top of slopes like domino (Fig. 6).

Fig. 6 Large-scaled land slide like domino caused by heavy precipitation in Chichibu belt

189

H. Sakai

Challenges in Road Construction and Timber Harvesting in Japan (187â&#x20AC;&#x201C;195)

Fig. 7 Retaining wall with log structure both at cutting and fill slopes in Chichibu belt

Fig. 9 Structure of retaining wall with log structure (Original: Oohashi 2001)

In such smooth soil area, it is difficult to make high cutting slope on steep slopes, and it is, therefore, necessary to make the retaining wall with log structure for providing operation roads (Figs. 7 and 8). It can be used for cutting slope and for stretching road width at filling by forming shoulder supported by the retaining wall of log structure at the bottom of fill slope (Fig. 9) (Oohashi 2001). This method was developed by Keizaburou Oohashi about sixty years ago in his forest located in decomposed granite area (Fig. 10). Fig. 10 shows the road section 50 years after the prescription of the retaining wall with log structure. Decomposed granite is sandy and it is difficult to make road bed only by earth work (Fig. 11).

Slope collapses of shallow depth have often occurred caused by recent torrential rain at decomposed granite area. Retaining wall with log structure is effective not only for decomposed granite but also for de-

posit of debris flow (Fig. 12), because its soil is mixed with soil and stones as is the case in Chichibu belt.

The logs used for making the retaining wall with log structure were first used for the frame with the aim of distributing the loads. However, these logs began to decay. It was reported that the Youngâ&#x20AC;&#x2122;s modulus of logs buried in the road would be zero within 30 years for sugi (Cryptomeria japonica) and 40 years for hinoki (Chamaecyparis obtusa) (Aizawa 2011). It is supposed that the internal and adhesive friction among logs and compacted soil makes the road bed stronger due to repeated traffic and soil and stones supplied during maintenance regardless of the decay of logs.

Fig. 8 Road constructed by retaining wall with log structure with narrow width of 2.5 m and a light truck

Fig. 10 Road by retaining wall with log structure at decomposed granite area

190

5. Black soil A great deal of forest land in Japan is covered with volcanic ashes or black soil originating from volcanic

Croat. j. for. eng. 38(2017)2

Challenges in Road Construction and Timber Harvesting in Japan (187â&#x20AC;&#x201C;195)

Fig. 11 Road construction with expanded mesh metal wall at decomposed granite area

H. Sakai

Fig. 13 Good sample of thick gravel on volcanic black soil

6. Crashing zone In addition to the soft soil and steep slopes, underground water that comes out through crushing zones is often troublesome. Crushing zones were made by frictions among bedrocks by frequent crustal movements and/or earthquakes since ancient days.

Fig. 12 Retaining wall with log structure is effective for deposit of debris flow soil. It is difficult to compact them, and they turn into mud after rain. The Shimonto method suitable for gravelly soil cannot be applied. Indeed, thick gravel is effective for volcanic black soil (Fig. 13) (Kishida and Abiko 1973), but it is not always easy to prepare the gravel.

Another method is to remove them, or to make reversal road from subsoil (Ryan et al. 2004) according to the thickness of the surface soil. However, the reversal road is not always strong enough for fully loaded trucks and crawler carriers. Logging companies sometimes cover the road with branches (Fig. 14).

On soft soils, it is recommended to separate the road bed and introduced gravel by geotextile (Keller and Sherar 2003). The use of an articulated forwarder with rubber crawler is also effective (Fig. 15), because it does not form ruts as there is no skid steering and due to large contact area and low ground pressure. It is also important for route location to find a stable natural bench on the slope in a volcanic area (Fig. 16).

Croat. j. for. eng. 38(2017)2

Fig. 14 A road covered with branches on volcanic black soil

191

H. Sakai

Challenges in Road Construction and Timber Harvesting in Japan (187â&#x20AC;&#x201C;195)

Fig. 15 Articulated forwarder with rubber crawler

usually not tall, but trees at the ridge of the crushing zones are high because of sufficient water.

Although crashing zones sometimes include water in the crashed rocks within the clay, which are also formed by numerous frictions, crashing zones are usually a way of water. Small-sized crushing zone is not so dangerous for road construction if the road crosses the zone at the right angle. However, it is costly to provide drainage using culverts or fords. If the route is parallel to a crashing zone, the road will be destroyed at many sections after heavy rain. Therefore, when crossing crushing zones with the planned route, it is important to take into consideration the drainage methods.

However, when the crashing zone and dip slope overlap, its valley head and the beginning of land slide are unstable, and it is not only impossible but also dangerous to construct roads on such terrain (Fig. 18). It was pointed out that the concave slope at the valley head was a dangerous area to construct roads (Oohashi, 2011; 2012). As shown in Fig. 18, the crack will

Fig. 16 Nick point, that is natural bench, is stable on the slope in a volcanic area Crashing zones can often be found by irregular profile of the ridge line and an isolated high tree at the ridge (Fig. 17) (Oohashi 2012). Trees at the ridge are

Fig. 18 Filling and constructing the retaining wall is impossible at the ridge when crashing zone and dip slope overlap

Fig. 17 Crashing zones can be found by irregular profile of the ridge line and a high tree at the ridge

Fig. 19 Underwater is sometimes hidden under thick volcanic ash in such concaved slope and this can result in the collapse of the filling

192

Croat. j. for. eng. 38(2017)2

Challenges in Road Construction and Timber Harvesting in Japan (187â&#x20AC;&#x201C;195)

H. Sakai

Fig. 20 Old alluvial fan area where old rivers are covered by sediment and/or deposit of debris flow and drainage facilities are required

Fig. 21 A reinforced soil wall, TK wall, applied at the site where the height of the filling exceeded 2 m

be caused at the road side of the filling by intensive compaction because of unstable natural slope and its weathered gravel soil. The maintenance is difficult, and in the worst case, the road will collapse by penetrated rain water flow in the future. It is also difficult to construct the retaining wall or reinforced soil wall because the basement is unstable.

Another dangerous case is when the crashing zone is hidden under thick volcanic ash. The filling will collapse due to underwater (Fig. 19). This is also applied to old alluvial fan area, where the rivers are covered by sediment and/or deposit of debris flow. When constructing roads in such areas along a contour line, fords and drainage facilities are required (Fig. 20).

(Takahashi et al. 2015a) (Fig. 21). TK wall structure is the combination of L-shaped outer frame and the wall by diagonal brace frame, using thinned logs. They were wound by polyethylene geotextiles. The geotextile in the filling makes the resistance force of shearing strong, and can be integrated into the soil resulting in the stabilization of road bed. The thinned logs in the soil will decay soon, but the durability of the structure will be maintained by its self-strength and a few tensile elongations of geotextile.

L-shaped steel retaining wall technology, the socalled L-shaped mesh wall, is easy to construct, and it is efficiently used for underground water drainage (Nippon Steel & Sumikin Metal Products Co., Ltd., 2011). The structure is composed of L-shaped stiffening steel mesh and horizontal steel mesh, which reinforce the tensile strength and prevent the deformation of the wall. It was not designed as reinforced soil wall, but the effect of the reinforced soil wall in stabilizing the bearing capacity of the filling was soon recognized (Takahashi et al. 2015b). When crossing a crushing zone by simple structures, L-shaped mesh wall or cage filled with stones is effective because the weight of stones and rocks in the structure presses down the road foundation with high water permeability (Fig. 22). The difference between the L-shaped mesh wall and the cage is that the latter is easily deformed when using soil on site whereas the former can be pressed by rolling without deformation.

It is necessary to use road width efficiently especially at the outside of curves by strengthening the filling. These technologies are effective and provide environmentally friendly landscape with the vegetation grown from seeds contained in the surface.

7. Road retaining technologies In addition to the main forest road, which has the function similar to the public road in a region in Japan, spur roads, constructed mainly for the purpose of forestry use, are required to be strong and to incur low maintenance costs. Road structures are fundamentally made only by earthwork, where retaining structures can only be used for unavoidable reasons such as terrain, geology, and soil. Reinforced soil wall can save the volume of earthwork on steep slopes because of nearly perpendicular slope gradient of the wall and narrow clearing width of roadway (Tatsuoka 2005).

The height of the filling and cutting slopes can be lower by adjusting formation of the road. Some forest road retaining technologies have been introduced recently in Japan (Yoshida et al. 2016).

A reinforced soil wall made by using thinned logs and geotextile, the so-called TK wall, was developed Croat. j. for. eng. 38(2017)2

193

H. Sakai

Challenges in Road Construction and Timber Harvesting in Japan (187â&#x20AC;&#x201C;195)

Fig. 22 Retaining wall whose surface was covered by vegetation, L-shaped steel retaining wall, set on three cages stacked on top of each other when crossing a crushing zone

Fig. 24 Excavator with bucket, which can be rotated, and a felling knife

8. Rolled grade

could be induced and distributed not to the valley but to the ridge, where it is usually dry (Oohashi 2011). On the contrary, when crossing the ridge, the level of the road should be lowered because surface water could be drained to the dry ridge. When crossing the valley, the level of the road should be raised in accordance with the topography and require minimum earthwork (Keler and Sherar 2003).

Drainage of road surface is important and indispensable. Rubber belt diversion is used in Japan. The belt diversion is also used on low-volume roads to divert water off the road surface in United States (Bloser et. al 2012). It consists of a piece of used conveyor belt bolted to treated lumber and buried in the road.

Rolled grade without belt diversion is also attempted in Japan. It is effective along contour line (Fig. 23), but dangerous on sloping road when going down because of unexpected slip under wet condition. It is always necessary to provide distributed drainage. Oohashi emphasized that the level of the road should be raised when crossing valleys, except streams with stable water, because water from the upper slope

9. Construction machinery At the construction site, excavators with bucket, which can be rotated, are used for soft soil. Some of them are equipped with a felling and bucking knife, and the operator can continue both felling trees on the route and doing earthwork without getting out of the cabin (Fig. 24).

10. Conclusions

Fig. 23 Rolled grade whose sloping at lower sections is used for drainage

194

It is important to locate a stout and stable road network in advance to provide sustainable forestry. Forest road is constructed on slopes, so that the inclination of slopes directly affects the road width requiring minimum earthwork, which is related to the vehicle sizes and operation systems. High road density may be effective for harvesting operations, but will increase the cost of road construction and maintenance, on the other hand. Therefore, appropriate harvesting systems should be selected from the view point of forest civil engineering. For example, in case of slopes of more than 35 degrees, cable logging and low density of stout forest road for trucks are available. Especially in Japan, tower yarders, which can operate both uphill and Croat. j. for. eng. 38(2017)2

Challenges in Road Construction and Timber Harvesting in Japan (187–195)

downhill, are needed because of the existing dip slopes.

In case of flat land of less than 15 degrees, a combination of the main road for trucks and skidding roads for CTL systems are available. In case of gentle and steep slopes ranging between 15 and 35 degrees, both vehicle systems and cable systems are available, and the choice depends on geology, terrain, soil, and forest management system. The existing forest roads tend to suffer unexpected situations such as heavy rain and spring water, and sometimes they are exposed to erosion after being finalized. Therefore, road maintenance, especially related to drainage system and its cost, is important to keep these structures effective. By adopting the most appropriate road maintenance method in terms of the soil, geology and terrain conditions and using the road repeatedly for a long time, the investment effect will be maximum and the construction cost will be negligible during long use. The early design and formation are important especially for controlling curves and longitudinal grade and for avoiding the overlap of the crashing zone and dip slope.

In the nearest future, new technology such as geotextile and reinforced soil wall should be developed and introduced. These challenges of road construction must take into account CO2 reduction.

11. References Aizawa, H., Sakai, H., Ando, N., Soma, T., Sakurai, R., 2011: Aging degradation of wood structure of operation road. J. Jpn. For. Eng. Soc. 26(4): 215–220.

Bloser, S.M., Creamer, D., Dimas, S., Scheetz B., Ziegler, T., 2012: Environmentally Sensitive Road Maintenance Practices for Dirt and Gravel Roads. USDA Forest Service. 126 p. http://www.fs.fed.us/t-d/pubs/pdf/11771802.pdf (Referred 25/11/2016) Japanese Forestry Agency, 2015: Annual Report on Forest and Forestry in Japan, Fiscal Year 2014. Japanese Forestry Agency, Tokyo. 206 p. (In Japanese).

Keller, G., Sherar, J., 2003: Low-Volume Roads Engineering.

US Agency for International Development, 158 p.

H. Sakai

Kishida, S., Abiko, Y., 1973: Road Engineering. Coronasha, Tokyo, 389 p. (In Japanese).

Nippon Steel & Sumikin Metal Products Co., Ltd, 2011: Lshaped Mesh Wall. Catalog No. L532. 4 p. (In Japanese).

Oohashi, K., 2001: All of Road Construction. Zenkoku-Ringyou-Kairyou-Fukyu-Kyokai, Tokyo. 159 p. (In Japanese).

Oohashi, K., 2011: Forestry Operation Road – Road Network Planning and Route Location. Zenkoku-Ringyou-KairyouFukyu-Kyokai, Tokyo, 121 p. (In Japanese).

Oohashi, K., 2012: How to Observe Mountain and Forest.

Zenkoku-Ringyou-Kairyou-Fukyu-Kyokai. Tokyo, 136 p. (In Japanese).

Ryan, T., Phillips, H., Ramsay, J., Dempsey, J., 2004: Forest Road Manual – Guidelines for the Design, Construction and Management of Forest Roads. COFORD. Dublin, 156 p.

Saito, Y., 1992: Understanding Early Life of Japan Islands.

Iwanami Shoten. Tokyo, 153 p. (In Japanese).

Sakai, H., 2004: History and importance of »operation roads« in Japan. Науковий Βісник: 354–360.

Samset, I., 1985: Winch and Cable Systems. Martinus Nijhoff/ Dr W. Junk Publishers. Dordrecht, 539 p.

Suzuki, R., 2000: Introduction to Map Reading for Civil Engineers. Terrace Hills and Mountainous. Kokonshoin. Tokyo, 942 p. (In Japanese).

Taira, A., 1990: Birth of Japan Islands. Iwanami Shoten. Tokyo, 226 p. (In Japanese).

Takahashi, H., Minami, Y., Yoshida, M., Sakai, H., 2015a: Relationships between reinforced soli wall and bearing capacity of forest road and management. Chubu Shinrin Kenkyu 63: 99–102. (In Japanese).

Takahashi, H., Sakai, H., Usuda, H., Watada, T., Furukawa, K., 2015b: Constructing L-shaped steel retaining wall and its bearing capacity. J. Jpn. For. Eng. Soc. 30(2): 79–84. (In Japanese with English Summary).

Tatsuoka, F., supervision. 2005: All of New Reinforced Soil Wall. Sougou Doboku Kenkyusho, 414 p. (In Japanese).

Yoshida, M., Takahashi, H., Sakai, H., 2016: Analysis of alternative forest road retaining technologies on difficult slopes in Japan. Eur. J. Forest Eng. 2(2): 61–66.

Yuasa, I., Sakai, H., 2012: Rule of Practice for Road Construction. Zenkoku-Ringyou-Kairyou-Fukyu-Kyokai. Tokyo, 217 p. (In Japanese).

Authors’ address:

Received: November 30, 2016 Accepted: June 13, 2017 Croat. j. for. eng. 38(2017)2

Prof. Hideo Sakai, PhD. e-mail: sakaih@fr.a.u-tokyo.ac.jp The University of Tokyo Graduate School of Agricultural and Life Sciences Department of Forest Sciences Bunkyo-ku Yayoi 1-1-1 113-8657 Tokyo JAPAN

195

Subject review

Transfer System to Adapt Timber Harvesting Operations to Local Conditions JĂśrn Erler Abstract Sustainable management requires equal consideration of economic, ecological and social criteria. The science looks back on more than 50 years, in which different multi-criteria decisionmaking models have been developed and refined. They are well suited for the solution of complex tasks, but are dependent on case studies und limited to them. Due to this high complexity, it is not possible to transfer the results of such case studies into practice. It is, therefore, necessary to prepare a transfer model that gives the opportunity to the practice to translate the scientific findings into their local multi-criteria decisions. Such a transfer model should provide a fixed basic structure, with which the complexity is reliably depicted. However, it should be open for individual additions and adaptations in order to adapt to the locality. In the process of finding the action options, it should support the user to enlarge the search space as far as possible. The criteria and attributes should be largely fostered by scientific if-then rules, in order to meet the transfer task. In this context, uncertainties, risks and side effects must be pointed out. In the selection of the scales, in contrast, it is recommended to dispense with scientific objectivity in favor of simpler applicability in practice. On the basis of these demands, a model is developed for finding optimal wood harvesting methods. In phases Develop â&#x20AC;&#x201C; Assess â&#x20AC;&#x201C; Evaluate, the user is guided through the decision-making process. Initially, he is commissioned to develop concrete action options for his individual task and to predict their results.

After that, he has to check six criteria with attributes and assess the options. Several methodological concepts are offered for the final evaluation. The model has proven its worth in various teaching environments. Therefore, it is recommended to develop it into an online tool for a wider target group as a continuing education module.

Keywords: harvesting, decision making, assessment

1. Need for a transfer model in decision-making questions Since the 1960s, it has become increasingly accepted that, for a sustainable management, several criteria have to be considered at the same time. The consequence is that the decision-maker needs a suitable multi-criteria decision-making methodology.

Jischa (1998) describes the path taken by the debate on sustainable development and identifies as milestones the book Silent Springs by Carson in 1962, Limits of Growth by Meadows in 1968, Global 2000 by Barney in 1978, Brundtland-Report in 1987 and Agenda 21 at UNCED in 1992. Parallel to this political process, both in technology development and in science, efforts were made to integrate ecological consequencCroat. j. for. eng. 38(2017)2

es and social needs into decision-making and to raise them at the same level as the business objectives. Kangas et al. (2008) and Diaz-Balteiro et al. (2016) give a systematic overview of the development of scientific decision-making methods. A shift in focus is apparent here: First of all, the focus was on methods which, using the linear algebra, sought to develop classical optimization functions and to calculate the shortest paths, minima or similar optima. As important representatives of this group, the authors name the Goal Programming, TOPSIS (Technique for Order Preferences by Similarity to Ideal Solutions), and the Point Method.

In the 1970s, discrete methods emerged, in which the finest option was selected from a finite set of options. The ELECTRE (Eliminating and Choice Trans-

197

J. Erler

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197–208)

lating Algorithm) and PROMETHEE (Preference Ranking Organization Method for Enrichment Evaluation) are mainly concerned with the gradual containment of the decision by exclusion of options.

Saaty uses another approach, the AHP (Analytic Hierarchy Process), to assess all options based on a set of criteria, and it is only in the end that he decides which option is best in the sum of the assessment criteria (Saaty 1977, Corrente et al. 2013). The central problem of these methods is to find as many objective judgments as possible and to exclude subjective influences. AHP solves this problem by looking for the better option in paired comparisons and by calculating the overall value using the eigenvalue method (Vaidya and Kumar 2006). According to Diaz-Balteiro et al.

(2016), in addition to AHP, the Development Analysis (DEA) and the MAUT method (Multi Utility Method) belong to this group of methods.

A major disadvantage of these methods is that they originally cannot deal with uncertainties or risks.

Therefore, there are many approaches to expand them, including the application of fuzzy methodology (Kangas et al. 2008).

Although the development of these methods has been intensively pursued since then, no standard has been developed. Since, in specific case studies, the basic idea of the AHP method is very suitable for resolving decision-making problems and, for example, considering that it can integrate expert groups into the set-up of pair comparisons to solve the subjectivity problem, this method is most likely to become a scientific standard.

However, all these decision-making methods have a problem: They only work in a specific case study, which defines all, the framework conditions as well as the options for action and their probable impact, and the objectives of the decision. However, when trying to generalize the results, these methods reach insurmountable limits:  A key issue is that the decision-making structures, as well as the selection of the criteria used in the case study, are subject to changing interests and are tailored to the local framework conditions in the case study. As soon as other structures or criteria are important in another case, a new case study has to be carried out, since the results of the given study cannot be transferred  Likewise, the options for action can hardly be generalized. As soon as in another case only one single deviating action option occurs, the pair comparisons of the case study lose their basis.

198

However, since technical processes are subject to rapid change, results based on case studies lose their relevance with every technical change  To exclude prejudice and to uncover undiscovered connections, scientific investigations are based on extensive data collected in a case study.

These data are not limited to the cause-and-effect relationships that can be determined by the natural sciences, but also include the assessments and the objectives of the case study. Thus, their results are only valid for this case study and are not transferable  As soon as the decision-making method tries to simulate risks and uncertainties and to consider the question of whether the decision-maker is more risk-averse, neutral or risk-tolerant, the results assume such a high complexity that a transfer to real questions falls into the background  Almost all methods are based on the same principle that the different criteria are scaled with a uniform assessment scale. However, the finding of a common scale that meets the criterion of scientific objectivity presents itself as a challenge. In particular, when »soft facts« are involved, for which there are hardly reliable data, the methods reach their limits.

If these assessments are combined, it is obvious that the transfer of decisions obtained in case studies cannot be generalized. Without wishing to challenge its scientific value, it should be noted that its value is severely restricted in practice.

The practice, however, expects the scientists to point out the solution of their problems. It does not want to be fed with singular results, which are no longer transferable with increasing complexity. Because this lack of transferability is system-dependent and cannot be resolved, the scientists are invited to develop a methodology that enables the practitioner to answer his own questions on the basis of findings obtained in scientific case studies:  First, it will be examined what characteristics a suitable transfer model should have  Subsequently, a model for the application to forestry decisions will be developed and presented  It will then be discussed whether this particular model meets the requirements of a transfer model  To conclude, in a short future, the tasks to be undertaken by scientists should be determined.

Croat. j. for. eng. 38(2017)2

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197–208)

2. Characteristics of a suitable transfer model Beforehand a note of warning should be sounded: From the requirement to overcome the limited transferability of scientific working methods, it is imperative that any model that arises here must deviate from scientific standards! The one who goes on this path inevitably leaves the quality standards of scientific work.

In doing so, he is exposed to the criticism – often violating – by his colleagues. The task, however, of creating a transfer instrument, which can be understood and used in practice, is fatally tied to this disadvantage. If, however, it were possible to build a bridge to practice in this way, it would seem worth all the efforts. Who else, if not the scientists, could build such a bridge?

Further to this basic view, the individual requirements will be discussed:

2.1 Specification of a decision-making structure and criteria The decisions that are made here are usually so complex that they must first be given a structure and the appropriate criteria for making the decision have to be selected. This requires not only a certain methodological knowledge, but also a great deal of time, which the practitioner can hardly muster. This is why it is necessary to offer frequently recurring basic structures and also to specify the most important criteria.

Such a skeletal decision-making model should be a middle way between standardization on the one hand and adaptability to individual needs and conditions on the other.

The AHP model can be used here as a sponsor: One important reason for its success is that it offers a very simple, easily adaptable structure, which is limited to three levels. On the lower level, the options of action are to be assessed; on the upper lever, the values are to be satisfied, and the criteria for judging the options are in between (Saaty 1977). This structure can easily be captured and transferred to any application.

When it comes to the question of the appropriate criteria, it becomes more difficult. Inexperienced model builders tend to take as many criteria as possible in order not to overlook anything. However, they run the risk of overshooting the model and making the decision impossible.

On the other hand, Saaty pointed out the psychological research results of G. Miller (1956) in the development of AHP, according to which the number of criteria should be no less than 5 and no more than 9 (»seven plus or minus two«, Saaty in 1990). Given such a limited number of criteria, it is the responsibility of Croat. j. for. eng. 38(2017)2

J. Erler

the model maker to select them very thoroughly, and to ensure that they are independent from one another and that they completely cover the set objectives.

Therefore, it is unpreventable that these criteria remain relatively abstract (e.g. protection of the soil).

The user must, therefore, be given the opportunity to search for suitable attributes, which in his case allow a more concrete assessment of the respective criterion.

In the example considered, it would be useful to take the specific ground pressure in kilopascal (kPa) as a criterion. However, it often appears necessary to define several explanatory attributes for one single criterion, which can be used either as an alternative or as a complement to one another. In our example, the decision-maker could consider the number of crossings as important in addition to ground pressure. For more than one attribute, he must specify how the assessment of the criterion is made up of the values of the different attributes.

2.2 Help in finding options for action In my experience, many practitioners tend to focus on very few favorites at the beginning of the decisionmaking process, thus severely restricting the search space. As a result, some options for action, which would be particularly well suited under one criterion, are simply overlooked.

The opposite should, therefore, be the case: first to collect as widely as possible different options for action and to evaluate them in a later step, because it is always better to leave an option after negative assessment fall back than not to take it up from the very start.

A transfer model should, therefore, offer a heuristic that helps the user to creatively generate as many different action options as possible.

2.3 Intuition where no data are available Since the practitioner does not have the opportunity to collect extensive data before his decision, he has to resort to the available information. While this sacrifices a fundamental property of scientific work, the transfer model must be able to partially engage with intuitive assessments.

However, in order not to open the floodgates to any subjective assessment, a middle way is required between data-related rules and intuitive application: Wherever there is sufficient evidence, which can be generalized via if-then rules, the scientist should also offer it to the practical user. However, where such a generalization is not possible, it should be up to the decisionmaker to follow his own intuition and assessment.

A suitable transfer model should, therefore, have a comprehensive set of decision-making rules that give

199

J. Erler

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197–208)

the user security. It has similarities to a decision support system, but it should allow him to deviate from those rules, where it is more appropriate for his or her individual decision-making situation due to locally based experience and knowledge, even if it cannot be subjected to any scientific review.

2.4 Dealing with uncertainty and risk Decision-making models tend to conceal uncertainties and risks behind figures and pretend to assume a higher degree of certainty. To compensate for this disadvantage, various methods have been proposed by scientists that can be used to describe both the factual uncertainties and risks, as well as their individual handling of the decision-makers (Kangas et al. 2008). However, they have further increased the already high complexity of the models. Thus, they are diametrically opposed to the effort to simplify and unify models for practical applicability.

Nevertheless, it would be wrong to ignore uncertainties and risks. Wherever possible, the user should be aware of uncertainties and risks so that he can take them into account when making his decision. Within the framework of the if-then rules, they should be communicated openly so that the user can only choose those options in which he is sufficiently secure. In this way, one leaves the judgment of the decision-maker’s intuition: after all, he is the one who has to make the decision and to take responsibility for it.

2.5 Choice of an appropriate assessment scale As already shown, a central problem of optimization is to compare the scores of different criteria.

Therefore, most of the scientific methods mentioned are looking for a uniform scale on which all criteria can be represented (exceptions such as outranking methods, see Kangas et al. 2008).

As such, the economist likes to resort to monetary units because they have indisputable advantages:  Everyone understands them intuitively  They make an abstract assessment without the user becoming aware of this abstraction  In addition to the criterion-related assessment, they also meet the balancing between the criteria, so that the sum directly contains a final judgment.

However, these advantages are counterbalanced:  Money values can be easily found only if the criterion has a real market value  All scientific attempts to assess non-market criteria with monetary values are critical; the assessment depends, in particular, on the method by which it was calculated

200

 As soon as attempts are made to present the impairment of any absolute value (e.g. the probability of a fatal accident) with monetary values, errors or even morally questionable statements can hardly be avoided  The obvious temptation to simply exclude such soft facts from the evaluation leads to ignoring relevant criteria.

Therefore, monetary scales are not recommended here.

We have seen that the AHP method solves this problem by making pair comparisons and thus automatically developing a uniform internal scale. However, this method has considerable disadvantages, too: On the one hand, it only works in a closed system. As soon as another option is added or the viewing space is changed in some way, the pair comparison must begin at least in parts from the beginning. Secondly, the pair comparison grows very quickly to a very timeconsuming task, so that the practitioner may be forced to keep the search space as small as possible in the sense of the time economy, which is counterproductive, as shown.

Another solution is to deliberately allow the subjectivity of the decision-maker and leave it to the judgment with a scale that corresponds, for example, to grading in the school, see value analysis by Zangemeister (1973). This seems to be justifiable if the decision-maker only has to justify his decision to himself.

However, if he were to justify his decisions to a third party, it would be necessary to disclose and discuss these subjective judgments.

It is, therefore, best to adopt a decision-making rule that does without a uniform scale. Then one could use a suitable scale for each criterion, whose values depend entirely on the question and which can be easily estimated by any expert.

3. A decision-making model for forest technology On the basis of these demands a sketch of a model will be given that fits to the strategic development of harvesting methods in forestry.

3.1 Balance between prebuilt structure and flexibility In section 2.1, the transfer model was to have balance between the specification of a structure on the one hand and the freedom to adapt this structure to the diversity of technical situations. What is meant by diversity in this case?

Croat. j. for. eng. 38(2017)2

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197–208)

 These are the framework conditions like geographical attributes of the ground, biological details of the woodland like species, age, dimensions, mixture, climate and the actual weather  These also are socio-economic conditions like wage-level, education of the workers, infrastructure, etc.  Finally, for the decision-maker, the objectives of his company have the status of a framework, too. For him a decision can only be made when he fulfills the priorities and restrictions of his own normative system.

If all of these conditions and limitations were integrated into our model, it would go beyond the scope of a transparent transfer system. Therefore, a model should be developed that presets a skeleton of basic structures that can be filled by the local knowledge of the user.

The model should be a bit more restrictive in terms of handling the assessment of the criteria, which are

J. Erler

represented by if-then rules (Weiss 2007). Here, the complex knowledge of experts should be collected and represented by a strong tool of rules, which are presented to the user. Normally, he will accept these general assessments. However, in some cases, when he has a more precise local knowledge or possesses some insights that differ from the main stream, he seeks for the chance to alter the assessment for his own to familiarize the model. In both cases the advantage is that all relevant criteria can be integrated, while the user may concentrate on those criteria that seem to be the most interesting for him.

Saaty (1977) has predefined a simple structure of options, criteria, and objectives that should be adapted to our transfer model. In case of forest technology, it can be translated to (Fig. 1):  First step »design of technical options«: It is the task of the user to look for technical options that seem to be adaptable under the local conditions.

For each option, the functional results of the

Fig. 1 Basic structure of the decision model for forest technology with the steps design, assessment and evaluation Croat. j. for. eng. 38(2017)2

201

J. Erler

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197–208)

work, like time consumption, productivity, quality of products etc., have to be forecasted  Second step »assessment of the options«: The options will be assessed by a set of important criteria. In order to prevent shortcomings, a fixed set of criteria should be proposed. The user will get the chance to underlay these criteria by fitting attributes by himself. If one option fails, it will not reach the third step because it is dropped out of the decision making process  Third step »final evaluation of the options«: Under those options that reach this third step, the decision-maker is free to elect that option that seems to fulfill the individual needs of his company in a best way. This part of the decision will never be taken over by any anonymous rules, because it is the expression of the responsibility of the decision-maker.

Following up, the steps will be described more intensively. As an example, a simple decision making situation will be carried out in parallel (data and assessments are fictive and have no relevance to real situations, see Fig. 3 and 5).

3.2 Design of different options On the first step, the decision-maker looks for technical options, or in our case: He has to find harvesting methods that could be able to reach the technical goal. Under these options, he will elect the best one. This shows the relevance of this first step. The search-area defines the quality of the decision. If the decision maker is lazy – and in practice this is often the case – and feels happy with one or two alternatives, the likelihood to find a really better solution is limited.

Experienced decision-makers suggest to look for more than 5 options, which should be as diverse as possible. Among them, there may also be some options, that seem to be unproductive, too expensive, old fashioned and so on. Though it seems to be inefficient to deal with options that will never have a chance to be the best, the advantage is that they enlarge the field of vision. A good method to enlarge the number of options is to look for different levels of mechanization.

And there is also a special option that is called the zero-option: i.e. to do nothing. More often than expected, the zero-option had the best assessment because it saved money where the positive effects of any activity never justified the negative side effects or risks.

202

Talking about options concerning the harvesting operations, the following should be specified: harvesting operations are all operational steps from the living tree to the storage place alongside the forest road. This includes the processing steps like felling, delimbing, topping, bucking and in some cases chipping. However, it also includes the transportation steps like hauling and skidding or forwarding. In most cases, the sequence of the processing steps is fixed. So is, of course, the sequence of hauling from the original stand of the tree to the skid road and the skidding to the forest road. However, the combination of processing steps and transport steps can be altered in several ways:  For example, in traditional forestry the tree is felled, delimbed and topped by a chain saw.

Then the tree length can be hauled by cable to the skid road and skidded to the forest road; in this case the first processing is finished at the stump site, followed by the extraction as second sub process. In a third sub process, the tree length is crosscut into pieces at the landing.  A harvester, however, first fells the tree, then moves this full tree to the skid road, where it starts delimbing and crosscutting and finally topping. In a second sub process, a forwarder loads the logs and forwards them to the forest road.

Normally, the decision-maker is creative enough to develop some functional alternatives. This task can be supported by a functiogram that has been developed

by Dög and Erler (2009). On a two-dimensional table, the most important functions of harvesting operations

Fig. 2 Functiogram of a harvesting method; here a fully mechanized cut-to-length method is presented with harvester and forwarder on assessable ground and permanent skid roads Croat. j. for. eng. 38(2017)2

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197â&#x20AC;&#x201C;208)

J. Erler

Fig. 3 Functiograms of four harvesting options: a) Semi-mechanized cut-to-length method with horse and tractor; b) highly mechanized tree length method with skidder; c) fully mechanized cut-to-length method with harvester and forwarder (see also Fig. 2); d) highly mechanized cut-to-length method with motor-manual pre-cutting outside the crane zone, harvester and forwarder; not illustrated here: e) the zero-option, which means no harvesting at all are specified, the processing functions top to down and the transport functions from left to right, building together a net with sides (functions) and nodes (states).

Each sub process can be depicted as a combination of procedural steps from the starting node to the end node. Between two sub processes, which can operate independently, a bubble marks a buffer. As soon as the decision maker decides to use a specific machine, he can define the sub process that this machine is able to run. While barely the total harvesting needs are covered by only one process (except the use of a harwarder or biomass harvester), the fixation of one sub process limits the choice of the other sub processes. This can be shown graphically by means of the functiogram (Fig. 2). Filling up the functiogram, other options can be generated that the decision-maker probably had not in mind before.

3.3 Assessment of options by a set of criteria To understand the assessment process, it is important to recognize that the options are not assessed as Croat. j. for. eng. 38(2017)2

Fig. 4 Complete system of six partial objectives with typical criteria such but rather their effects that are likely to be reached, as well as the expected efficiency of work. In some cases, if we are lucky, this information can be

203

J. Erler

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197–208)

taken from our own experience. However, in all other cases, a forecast that is based on fictional assumptions must be made:  Adaption: In a case that this harvesting method can be observed at a fair or anywhere else, we can try to assess whether the foreign results can be transferred to our own conditions and – if not – what will be the difference  Prolongation: In a case that the unknown method is very near to a known method and only small elements are different, the effect can be estimated on the basis of the effects of the old method  Analytical approach: If both possibilities fail, we can try to estimate some effects in analogy to other methods, where similar tools or machine elements (type of wheels and axles, harvesting head, etc.) are used.

In case of harvesting operations, the decision-maker will first forecast the quantity, quality, and value of the products (effectiveness). On the other hand, he will try to estimate the costs and procedural risks (efficiency). In the skeleton, these assessments of effectiveness and efficiency are foreseen as two independent criteria that have to be assessed in any case. The transfer system can offer some basic estimation for productivity and the costs depending on technical details (volume per tree, skidding distance, etc.), but they should be proved and altered by the decision maker, if needed.

However, according to the idea of sustainable development, the ecological and social aspects have to be regarded at the same hierarchical rank. The suggestion at this point sounds a bit schematic, but helps to find a systematic one that can be easily understood and communicated: To make the same difference for ecology and social aspects like in economy. This means as follows:  Ecological effectiveness asks for risks and side effects on soil, water, stand, biodiversity, and productivity. To forecast these effects, there are a lot of well-known correlations. The problem is normally to condense them to a common assessment of the partial objective of ecological effectiveness that should be called »ecological compatibility«  Ecological efficiency asks for the sparing use of natural resources like energy. In our case, the most important resource that we disturb for a long time and, therefore, »use« for our harvesting operations is the soil under the driving machines. An important attribute to measure ecological efficiency can, therefore, be the percentage

204

of forest soil that is covered by skid roads; with 20 m distance, it can reach 20% of the productive soil, in case of no driving restrictions it can easily reach 40% and more  Social effectiveness is reflected in the effects towards the society and answers to their societal needs. They involve recreation, cultural demands and the hope of employment. In parallel to the ecological partial objective, it can be called »social compatibility«  Social efficiency asks for the sparing use of social resources – a strange concept. Social resources involve all human beings who are working in the process. Sparing does not mean to rationalize them in order to lower the costs (that would be economy), but to lower the damage and danger for them. In other words: Social efficiency asks for the standard demands of ergonomics and can be condensed in a partial objective called »humanity«.

In most countries, there are some legal restrictions that have to be regarded in the decision making process. In Germany, for example, clear cuts are normally not allowed, machines have to stay only on skid roads and special hydraulic liquids have to be used.

In addition to these restrictions, the forest owner can decide to respect specific restrictions, which is called a voluntary self-limitation (Faber 2008). Well known are the forest certification systems like PEFC and FSC, which – depending on national rules – set standards for a specific behavior. In Germany, for example, they provide certain distances of skid roads and for wheel ruts after harvesting operations. If the decision-maker fears that the risk is high to exceed such a limit, he is well advised to drop the option completely.

3.4 Uncertainty and risk Many of these assessments contain open questions and risks, which must be considered by the decisionmaker. This is particularly true of the risks and sideeffects of forestry actions on the environment. The transfer system should, therefore, take particular account of the environmental risks and provide the user with as much information as possible.

A well-fitting tool, developed by Grüll, is in use in the North East of Germany (see Erler and Grüll 2008).

It assesses the sensitivity of the forest soil as well as the productivity of the forest site and condenses this information to a 5×5 diagram that is called technogram. Each field of the diagram represents the sensitivity against the traffic of forest machines (as indicator Croat. j. for. eng. 38(2017)2

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197â&#x20AC;&#x201C;208)

J. Erler

Fig. 5 Optimal option depending on the preference pattern of the company, shown for efficiency, effectiveness and overall assessment (sm: semi-mechanized; hm: highly mechanized; fm: fully mechanized; ctl: cut-to-length method; tl: tree length method; zero-opt: no harvesting at all) of ecological compatibility) and the productivity value of the stand (as indirect indicator for eco-efficiency).

In a comparable scheme, the impact of each technical option can be expressed by a 5Ă&#x2014;5 diagram called ecogram, by which the compatibility for this very site is assessed. A comparable tool would support the user of the transfer model in assessing the ecological compatibility and efficiency and herewith in lowering the risk of decision-making.

3.5 Scales and evaluation Up to this point, we have not made any determination regarding the scales. In Chapter 2.5, we have spoken only against the use of monetary values and pair comparisons.

As long as a relative evaluation of all criteria is attempted, however, a uniform scale is irreplaceable.

Croat. j. for. eng. 38(2017)2

Despite all the doubts, it has proved useful in this case to use a scale, which is based on evaluation in school teaching because the associated values and their relations are commonly known and balanced. In some cases, it is necessary to provide an explanatory guidance, if the assessments are not logically accessible.

As soon as all options are assessed under all criteria, the evaluation can follow. It seems to be obvious to follow a hierarchical structure that brings the partial objectives to a ranking order. First, all options are assessed under the focus of the most important criterion.

When different assessments are found for all options, the best option will be selected and the decision process is finished. Only when two or more options are similar under this first criterion, they (and only they) are assessed by the next important criterion. The advantage of this lexicographical way of decision-mak-

205

J. Erler

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197–208)

ing is that it easily leads to a decision. However, the disadvantage is obvious: it is only acceptable when all options are in principle elective near to each other.

Another solution to be applied in scientific research is to weight the criteria. However, this requires a very intensive focus on target systems and is, therefore, not suitable for practical applications.

Therefore, we suggest to use the decision making process: different decision structures are calculated to show which options were the best under which structure. A basic structure, in which the relationships between economy, ecology and social needs are altered in steps of one third, has proven very good. Between 1–0–0 and 0–1–0 and 0–0–1, there are intermediates like 0.66–0.33–0 and 0.33–0.33–0.33 and so on with 10 combinations in all. For each combination, the best option is shown with the effect that the decision-maker can see which consequence a certain preference concept would have (Fig. 5). This helps him to get a feeling for the technical consequences of his preferences.

 For each criterion, one or more attributes are selected for which natural scales are created  For each scale, a minimum condition is defined that must be met by each option  If an option in one attribute does not meet this minimum condition, you can try to improve it and re-evaluate it under all criteria; if this is not successful, the option must be excluded from further selection  As soon as all options have met all the minimum conditions, they can be regarded as usable without exception  Now, there is nothing to select a single criterion as decisive. This is usually the economic efficiency.

This method has the advantage of being very similar to the intuitive approach of many decision-makers. It is, therefore, subjectively perceived as meaningful and accepted.

Here is an example:  Concerning the efficiency: Under given assessments, highly and fully mechanized cut-tolength methods (c and d) are best, when ecology has a lower value, otherwise the zero-option e) is the best  Concerning effectiveness: With the presetting, the semi-mechanized cut-to-length method a) is optimal; only under maximum ecological focus, it will be lapped by the zero-option e)  Concerning efficiency and effectiveness together with balanced weighting, three options are competing: Under social focus the highly mechanized cut to length method d) is optimal, while under ecological focus, the zero-option wins; with economic focus or balanced weights, the semi-mechanized cut-to-length method a) is the best.

In chapter 2.5, the scales based on the natural assumptions of the criterion to be assessed have been described as optimal. They can thus be understood by any professional user and applied without contradiction. In the above example, the specific ground pressure would be in kPa, or an ordinal scale derived from it (light – moderate – heavy). With this type of scale, it is possible to provide decision-making, which is specialized in the exclusion of inappropriate options and reduces step by step the set of options.

In our transfer model, this could look like this:  Firstly, as many options as possible are required, all of which fulfill the functional objective

206

4. Discussion The presented model is not seen to be an alternative to the scientific model approaches that are used to evaluate multi-criteria decision-making situations. It was made clear that this model, in contrast, should be a transfer model that should bridge the gap between scientific knowledge and practical decision-making needs.

As this model is neither right nor wrong, it can only prove that it stands the test in reality:  It should be teachable between people so that they can understand and use it  It needs to be open for participation in a way that is open for the individual knowledge of the user and ready to be integrated into rules  It has to be adaptable to local conditions and variability over time  It must be operational in a sense that it can be controlled with clear results.

4.1 Readiness for teaching In study situations at university, this model has been taught numerous times. It could be proved that students are able to adapt it to specific problems after a short time with high transparence and quality. The same model could also be used with success in international context. In doctorate theses, it has been adapted to conditions in Greece (Dimou 2002), Brazil (Saraiva da Rocha 2011), and Iran (Badraghi 2013). For Croat. j. for. eng. 38(2017)2

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197–208)

students of tropical forestry, it serves as a basic pattern to explain the diversity and different technical solutions in global forestry with different regions and cultures.

4.2 Readiness for participation This model gives to the user the chance to introduce its own experiences and local knowledge. The combination of easy and fixed skeleton structure on one hand, and complex assessment on the other hand allows individual adaptations without lacking the complexity that is necessary for an overall assessment.

However, this flexibility has its shadows, too. In contrast to any standardized models, a comparison between different users and different companies is not possible. In each case, it is essential to look at all adjustments at all levels. So, this model is not a tool to normalize the decision results, it is rather a tool to ease the decision making process in a complex surrounding.

4.3 Readiness for communication Though the model is very complex, its structure is simple and near to reality. This helps people to adapt it to local conditions and needs without profound knowledge of decision-making structures. By its transparency, it is a valuable tool in the leadership process and may even help to find the legal liability.

4.4 Readiness for operational control A public forest service in Germany took this model to modify the forms for working advices with the consequence that finally the character of the advices changed to a real contract with clear objectives and competences (Erler 2009). The experiences of the last 10 years have shown that this tool is in practical use and gets positive feedback.

5. Future outlook The model has been in use for several years and has proven its benefits. It is mature enough to be made available to a larger group of users. It has been available as an online tool, which can be used for the training of forestry staff, service entrepreneurs and forest owners.

To this end, however, considerable efforts still need to be made:  The tool itself has to be prepared and adapted to the possibilities and limits of E-learning  Criteria and attributes must be filled with up-todate scientific findings and must be stored with appropriate scales Croat. j. for. eng. 38(2017)2

J. Erler

 All the calculations required for the entire decision-making process must be carried out and made transparent by the online system, without limiting the intuitive use.

6. References Badraghi, N., 2013: Productivity, Cost and Environmental Damage of Four Logging Methods in Forestry of Northern Iran. Doctoral dissertation, TU Dresden, Germany, 77 p.

Corrente, S., Greco, S., Slowinski, R., 2013: Multiple Criteria Hierarchy Process with ELECTRE and PROMETHEE. Omega 41(5): 820–846.

Diaz-Balteiro, L., Gonzalez-Panchon, J., Romero, C., 2016: Measuring systems sustainability with multi-criteria methods: A critical review. European Journal of Operation Research 258(2): 607–616.

Dög, M., Erler, J., 2009: Funktiogramme für Holzernteverfahren – komplex und trotzdem gut veständlich. Forsttechnische Informationen 61(9–10): 14–17 (in German).

Dimou, V., 2002: Multivariante Bewertung von Holzernteverfahren für die Waldarbeit in Nord-Ost-Griechenland. Doctoral dissertation, TU Dresden, Germany, 219 p. (in German).

Erler, J., 2009: Die teilautonome Arbeitsgruppe (WAG) bei Bundesforst – WAG-Broschüre. (B. Bundesforst, Hrsg.), 43 p. (in German).

Erler, J., Grüll, M., 2008: Standortgerechte Holzernteverfahren – ein Beitrag zur Harmonisierung von biologischer und technischer Produktion. FTI 3–4: 36–40 (in German).

Faber, R., 2008: Freiwillige Selbstbeschränkung bei forsttechnischen Handlungen im Wald und Möglichkeiten der öffentlichen und privaten Honorierung – dargestellt am Beispiel des Kommunalwaldes in Nordrhein Westfalen. Doctoral dissertation, TU Dresden, Germany, 246 p. (in German).

Jischa, M., 1998: Sustainable Development and Technology Assessment. Cehm. Eng. Technol. 21(8): 629–636.

Kangas, A., Kangas, J., Kurttila, M., 2008: Decision Support for Forest Management. Springer, Berlin, Germany, 222 p.

Miller, G.A., 1956: The magical number of seven, plus or minus two: some limits on our capacity for processing information. Psychological review 63(2): 81–97.

Saaty, T.L., 1977: A Scaling Method for Priorities in Hierarchical Structures. Journal of Mathematical Psychology 15(3): 234–281.

Saaty, T.L., 1990: How to make a decision: the analytic hierarchy process. European journal of operational research 48(1):9–26.

Saraiva da Rocha, E., 2011: Valuation of the logs skidding operation impacts in the wood crop process in upland primary forests of the Brazilian Amazonia. Doctoral dissertation, TU Dresden, Germany, 97 p.

207

J. Erler

Transfer System to Adapt Timber Harvesting Operations to Local Conditions (197–208)

Vaidya, O.S., Kumar, S., 2006: Analytic hierarchy process: An overview of applications. European Journal of operational research 169(1): 1–29.

Weiss, M., 2007: Entwicklung eines flächenbezogenen Entscheidungsunterstützungssystems für Holzernteverfahren

unter Berücksichtigung der Wertvorstellungen des Waldbesitzers. TUDPress Dresden, Germany, 114 p. (in German).

Zangemeister, C., 1973: Nutzwertanalyse in der Systemtechnik. Wittmannsche Buchhandlung, München Germany, 370 p. (in German).

Author’s address:

Received: November 8, 2016.

Accepted: February 17, 2017.

208

Prof. Jörn Erler, PhD. e-mail: erler@forst.tu-dresden.de TU Dresden Dresdner Straße 24 01737 Tharandt Germany Croat. j. for. eng. 38(2017)2

Subject review

Cable Yarding in North America and New Zealand: A Review of Developments and Practices Rien Visser, Hunter Harrill Abstract Cable yarders have been an integral part of harvesting timber on steep terrain for over 150 years.

They have developed from basic labour intensive steam powered winch operations to sophisticated and automated mechanised systems. While European yarder development has focused on relatively small but highly mobile machines operating with standing skyline configurations, the North American and Southern Hemisphere developments have tended towards larger, taller and more powerful machines capable of higher daily production. Two dominant North American brands, Madill and Thunderbird, produced over 3000 yarders and many of their machines continue to work today. Often working with 4 or 5 drums, they were able to develop and utilise an expansive range of rigging configurations to suit different extraction needs.

Modern developments continue to focus on increasing production capability and cost-effectiveness suited to clear-cut plantation forestry. With safety becoming more paramount in terms of a licence to operate, a strong preference is given to fully mechanised systems. By definition, these are yarders with rigging systems that support grapple carriages, extracting timber that has been mechanically felled on steep slopes. While mechanical grapple carriages have long been combined with swing yarder systems, the further development of a motorised grapple carriage allows tower yarders to operate without choker-setters. Ergonomic improvements for the operator, long established in European machinery, are being integrated including cab design with greatly improved visibility and partially automated electric over hydraulic control systems. Logic would suggest that, over time, yarder developments will combine the strength and robustness of North American design and the finesse and automation of European design. However, recent machine sales in North America and New Zealand continue to show a clear difference with the preference of large swing yarders capable of running fully mechanised extraction systems.

Keywords: cable logging, extraction efficiency, system development, rigging configurations, ergonomics

1. Overview of yarder developments Using cables to extract felled stems emerged as a common practice around the turn of the 20th century.

It became known as Âťcable loggingÂŤ and was a preferred method of extraction on steep slopes (Studier and Binkley 1974). While the use of rope logging practices date back centuries in Europe and Asia, modern cable yarding practices were developed in the late 19th century with the advent of steam powered engines like the Dolbeer Steam donkey in 1881 in Eureka, California (City of Eureka, 2010). The machinery used has improved over the years from the early steam powCroat. j. for. eng. 38(2017)2

ered winch sets to current yarders with highly-sophisticated diesel powered engines, water-cooled brakes, interlocking drums and electric controls (Mann 1976, Samset 1985, Sessions 1991). However, the challenges of productive cable logging remain the same; to get lift on the logs to provide partial or full suspension in order to avoid ground objects and reduce the friction, and thus the pulling force required (Pestal 1961, Lysons and Mann 1967, Falk 1981, Conway 1982).

In addition to being able to operate on slopes, where it is not physically feasible to operate groundbased systems, cable yarding can also be preferred on

209

R. Visser and H. Harrill

Cable Yarding in North America and New Zealand: A Review of Developments ... (209–217)

intermediate slopes due to its reduced environmental impacts because of partial or full suspension of logs, and not needing to drive an extraction machine into the forest, reducing soil disturbance (Liley 1983, McMahon 1995). Despite its considerable developments and wide use, cable logging is expensive and is inherently more complex than either tractor or skidder logging (Kirk and Sullman 2001, Visser 2014). It has a high incidence of accidents to workers and is generally less productive than ground-based methods of harvesting timber (Slappendel et al. 1993, Spinelli et al. 2015). The alternatives for the extraction on steep slopes are the helicopters but they are not often preferred due to their high rate of fuel consumption and expensive operating costs (Horcher and Visser 2011).

There are many different cable logging systems that have been developed over the years in different regions of the world. They are composed of different machines, rigging systems and operating methods available: each with their own requirements and capabilities. This culminates in a very large number of different combinations, whereby texts and manuals provide a good overview (e.g. Studier and Binkley 1974, Larson 1978, Liley 1983). Standardised terminology can be found in Stokes et al. (1989). While early cable yarders used a spar tree or independent tower to keep the rigging off the ground and provide lift to the trees to bring them onto on the landing, modern yarders are almost exclusively integrated tower yarders. »Integrated« refers to the winch set and the tower being on the same base machine.

A clear difference between North America and Europe in the design of the integrated yarders began in the 1950s. The North American systems were designed to operate in larger forest areas (initially native, but now predominantly in either second growth or plantations), so the emphasis was on larger machines that were able to harvest efficiently from a single purpose built landing location (Fig. 1). To accommodate modern high production extraction, processing and loading out operations, these landings are large and a recent survey showed the average yarder landing in New Zealand to be 3900 m3 (Visser et al. 2011). The infrequent yarder moves most often involved a relatively short distance shift to the next landing only a few hundred meters away. As such, most found favour with a strong and large track base that provided the ability to relocate on poor infrastructure. In comparison, the European design focus was based around smaller scale harvests with a need to relocate often, and required to be completed from existing forest roads (Heinimann et al. 2000). These smaller machines had relatively low payload capability and hence the need to design for speed of operation and relocation.

210

Fig. 1 A typical PNW or NZ yarder set-up, with a Madill 172 working a larger clear-cut operation from a purpose built landing

2. Madill Yarders as an Example A few yarder manufacturers based in the Pacific Northwest (PNW) produced the majority of cable yarders that have been, and are still, operating in PNW and New Zealand; Madill, Skagit and Washington built between 1500–2000 yarders each, Berger around 1000–1200, and Thunderbird 350–500 (pers com Cole).

Madill is considered one of the most successful and enduring manufacturers of cable yarders in the PNW and led many developments along the way. By the end of the 1990s, they had merged with other manufacturers like Cypress, Thunderbird, Skagit, Berger and Washington, but still retained their respected Madill name.

Early developments of diesel powered yarders revolved around integrating the tower to the chassis to reduce the dependence on time to rig-up a spar tree for desired lift. The idea was to make them more mobile by adding a tracked or wheeled undercarriage and the earliest recognised example of this was the Lidgerwood (Samset 1985). The concept was largely successful as indicated by sales and popularity of models such as the Madill 009 of which there were approximately 900 machines produced (Table 1). By the 1960s, yarders were becoming more commonly employed and they were produced by several manufacturers. Water cooled brakes added smoothness to the rigging movement and controls of a yarder that were far superior to the band brakes. The new types of brakes allowed for faster speeds, higher tensions and longer brake life, all of which helped improve productivity and justify larger and costlier yarders (Rice 1974). Interlock became another important development in winch sets, allowing drums to turn at the same Croat. j. for. eng. 38(2017)2

Cable Yarding in North America and New Zealand: A Review of Developments ... (209–217)

R. Visser and H. Harrill

speed or torque, and improved the effectiveness of running skyline systems, particularly the use of grapples on swing yarders (Carson and Jorgenson 1974).

Originally, interlock was mechanical but transition to hydraulic functions started, which allowed for energy recovery from braking and smoother control of functions (Hartsough et al. 1987).

The Madills of the 90s were a mixed design product, closer to the Cypress machine, and some Cypress machines (6280, 7280) were painted orange and called Madills. In the late 90s, the Cypress/Madill joint venture purchased Thunderbird, which had the rights to Skagit designs, and the last Thunderbird 255B machines also took on the orange and white colours.

Table 1 An overview of cable yarding developments at Madill (after Heavy Equipment Forum, 2016)

3. Aging Yarder Workforce, the New Zealand Case

Model

Number built

Basic Description 1960s

Madill 009

900

2 drum highlead machine, with strawline.

Usually 90’ tower height

Madill 046

56

3 drum slackline yarder, with straw and taglines. Usually 90’ tower height

Madill 052

20

3 drums as 046, but running skyline. Huge, heavy. Usually 90’ tower also

»Yarding Crane«

3

Swinging grapple yarder. Huge. Prototype for the 044 1970s

Madill 071

235

50’ Tower, slackline yarder mounted on Terex, tank, or rubber Madill SP

Madill 044

136

Large swing grapple yarder. Smaller, lighter version of the »Yarding Crane«

Madill 084

3

Bigger version yet of the 044. Only 3 were built. Just too big and heavy 1980s

Madill 121

3

Lightweight and mobile swing yarder, prototype for the 122

Madill 122

47

New and improved 121, very popular and still in use today. Weighs 50 tons

Madill 123

24

Larger, basically bigger version of the 122.

Weighs about 70 tons

Madill 144

19

Bigger still version of the 123. Basically huge, swing grapple yarder

Madill 171

29

Bigger version of the 071, with 70 tower, bigger drums, etc. in big tank 1990s

Madill 120

31

Swing Yarder. Grapple or dropline machine, all hydraulic, 45 tons

Madill 124

56

Swing Yarder. Same as above, much larger, all hydraulic

Madill 172

34

70’ slackline machine, hydraulically controlled mobile standing tower

Croat. j. for. eng. 38(2017)2

Cable logging, as practiced in New Zealand, differs in several respects from how it is practiced elsewhere.

The reasons are various, but the nature of Pinus radiata, the value of the wood recovered, features of New Zealand’s terrain and climate, and the reliance on plantation forestry, are all factors (Liley 1983). However, New Zealand shares many similarities in cable logging operations with the PNW due to the two regions long history of interactions and trade. When diesel yarders became well established in North America for harvesting such a resource, New Zealand started importing the machines to harvest their plantation forests of the 1950s and became a catalyst for future collaboration. For example, New Zealand joined the Pacific Logging Congress and Setup the Logging Industry Research Association (LIRA) to get up to speed with the technological developments of the machines and find how to most effectively use them (Ellegard 2016). The two regions embarked on a journey that saw a few decades of focused and collaborative research and training efforts; like the Forest Engineering Institute (FEI) of the late 1980s. The financial downturns at the end of the 20th century put an end to a prosperous period of cable logging developments for both regions, until a resurgence in the last decade.

Clear evidence of New Zealand’s reliance on yarder developments in the PNW can be found in a comprehensive yarder survey of working machines in 2012 (Visser 2013). There were 68 different »models« of yarders recorded in the survey and 85% were designed and built in the PNW. The survey indicated that 305 yarders were operating, compared to 214 recorded in the 2002 survey and only 82 yarders in 1985 (Finnegan and Faircloth 2002). Of a total of 2012, 67% were tower yarders and 30% were swing yarders and the remaining 3% were identified as excavator-based yarders.

The Thunderbird TMY 70 tower yarder is the most common model yarder in New Zealand with 31 machines (Table 2).

By identifying the make and model and ownership, the survey was also able to determine that approximately 130 yarders were introduced into the

211

R. Visser and H. Harrill

Cable Yarding in North America and New Zealand: A Review of Developments ... (209–217)

Table 2 Ten most popular yarder models operating in NZ as reported in the 2012 yarder survey (from Visser 2013) Make/Model

No. in operation

Thunderbird TMY 70

31

Madill 071

26

Madill 123

18

Madill 124

17

Madill 171

17

Brightwater BE70LT

13

Madill 009

13

Brightwater BE 85

10

Thunderbird TSY 255

10

Madill 122

9

workforce and that around 40 yarders were decommissioned in the last 10 years. The majority of additions have used machines purchased from the PNW and either put into service directly or reconditioned in New Zealand. With the closure of most North American yarder manufacturers, the reduced availability of used machines on the market and the expected increase in cable harvesting volume, it is unlikely that this trend can continue for much longer. Therefore, New Zealand will likely see an increase in new yarders entering the yarder population over the next few years.

In the survey, carriage information was recorded for only 213 yarders. Of this total, 129 (61%) did not have access to any sort of carriage. Of the 84 yarders that used a carriage in the survey, 17 were simple mechanical grapple carriages (20%), mainly associated with swing yarder machines. While European yarder systems have predominantly used standing skyline configurations, a survey of rigging configurations (Harrill and Visser 2012) showed that cable yarding operations in New Zealand rely heavily on North Bend, running skyline, highlead and gravity return (shotgun) configurations. These configurations, which do not employ a carriage, provide for higher payloads by sharing the payload with the ground (Fig. 2). The disadvantage is that they require greater mainline pull, higher fuel use per unit volume and create more ground-disturbance.

By 2010, New Zealand had already begun to make another transition of harvesting increasing volumes from forests on steep terrain in even greater proportions (Harrill 2014). The last decade has seen many

212

Fig. 2 North Bend rigging configuration, with the mainline going through a fall-block before being attached to the »carriage«. When the mainline is pulled, the stems being pulled balance forward motion with lift so that the payload nearly always remains in contact with the ground new developments in steep slope harvesting and cable logging technology in New Zealand (Raymond 2012, Visser et al. 2014), some of which have been commercialized and are now being exported back to North America. At the same time the adoption of cable-assisted machines and the aging fleet of yarders in North America is stimulating new developments there.

4. Yarder studies – productivity and safety System productivity has been extensively researched in logging operations, as increasing productivity typically results in lower logging rate costs ($/ton or $/m3) (Cavalli 2011, Visser 2014). An example of studies that provided insight and understanding into production potential of various logging systems and rigging configurations was known as the Pansy Basin Studies carried out in the Pacific Northwest. Production rates and costs for cable, balloon and helicopter yarding systems in old growth stands were established (Dykstra 1975) with a follow up study on the same systems in thinned and clearcut young growth forests, including their delays (Dykstra 1976a and 1976b). There were other research projects carried out at the time such as: running skyline production using a mechanical slack pulling carriage (Mann and Pyles 1988); production of a manual slack pulling carriage in thinned stands (Sinner 1973); comparison of skyline carriages for small wood harvesting (Balcom 1983); production of pendulum balloon logging (Ammeson 1984); production costs and optimal line spacing of Croat. j. for. eng. 38(2017)2

Cable Yarding in North America and New Zealand: A Review of Developments ... (209â&#x20AC;&#x201C;217)

running skyline and standing skyline systems using slack pulling carriages (Rutherford 1996).

Other studies quantified systems production rates, and even compared production rates of different systems and equipment side by side (Kellogg 1987, Forrester 1995) or over the same terrain and stand conditions such as: comparison of Washington 88 and Madill 009 (Bell 1985); cycle time comparison of Timbermaster and Wilhaul yarders (Douglass 1979); shift level comparisons between Ecologger, Bellis, Lotus, and Thunderbird yarders in down-hill logging (Evanson and Kimberly 1992); and a case study of a mobile Madill 009 in mature radiata pine (Murphy 1977).

These studies and many other yarder trials, carried out

by LIRA/LIRO between 1973â&#x20AC;&#x201C;1991, have been summarized in a book by Harper (1992). Some have investigated different rigging systems and their productivities such as: alternative rigging variations for the North Bend configuration to improve productivity by improving control and reducing required line shifts (Bennett and McConchie 1995); and a system evaluation of a Madill 071 using North Bend, Shotgun, Slackline and mechanical slack pulling carriage configurations (McConchie 1987). Stampfer et al. (2010) showed the benefits of using radio-controlled chokers. Visser and Stampfer (1998) showed the increased productivity associated with mechanised felling and processing on subsequent yarder extraction. More recently, Harrill (2014) investigated the productivity and skyline tensions of different rigging configurations in a series of case studies. Cavalli (2012) found that the last 10 years of research by forest engineers interested in cable logging was mainly (45%) directed towards efficiency.

According to Cavalli, in the near future, efficiency will continue to be the topic in cable logging research, and efforts in optimization, including computer automation and control of machinery, will aid this focus on efficiency.

Many guide books on cable logging safety and best practices have been produced over the years to educate workers and to reduce accidents. Notably, the Yarding and Loading Handbook by OR-OSHA (1993) and revised (2008) were built on the Cable Yarding Systems Handbook by WorkSafeBC (2006) and subsequent versions. Similar guides exist in New Zealand, like the Approved Code of Practice by the (MBIE 2012) and the Best Practice Guidelines by (FITEC 2000). Unfortunately, worker fatalities occur in the same ways as they did 40 years ago (OR-OSHA 2008). Improving our knowledge of forces and tensions involved with complex cable logging systems, as well as a better understanding of control over the extraction process, can help improve safety. Slappendel and others (1993) inCroat. j. for. eng. 38(2017)2

R. Visser and H. Harrill

vestigated factors affecting work related injury in forest workers in New Zealand. Hartsough (1993) studied the use of remote tension monitors and the benefits they provide to safety. Physical demands of steep terrain workers were quantified by Kirk and Parker (1994), who later investigated heart rate and strain of choker setters (Kirk and Sullman 2001).

Yarder tower collapses became a concern prompting two studies by Fraser (1996) and Fraser and Bennett (1996) on collapses and their potential causes. The New Zealand accident reporting scheme was established to combat increasing rates of accidents (Sullman et al. 1999). Bentley et al. (2002) outlined how the accident reporting scheme data could be used to identify priority areas for ergonomics, safety and health research attention. Social acceptability towards safety has led to many logging contractors and forestry companies taking a more proactive approach to fully mechanising cable logging operations via cable-assisted felling and grapple yarding.

5. Fully mechanised systems Mechanisation of cable logging operations offers both safety and productivity benefits (Amishev 2012).

Most cable logging operations require the use of skilled workers (e.g. fallers and choker-setters) to get the trees on the ground and connect wire ropes to the stems for extraction to a roadside or landing (Kirk and Parker 1994, Harrill and Visser 2012). The task of choker-setting is not only physically demanding but also poses a high risk to workers being struck by logs. Forestry is classified as one of the most dangerous jobs in

Fig. 3 T-Mar Log Champ 650 shown above, like Madill 124, is a very large (70+ tonne) swing yarder designed for larger scale mechanised clear-cut operations

213

R. Visser and H. Harrill

Cable Yarding in North America and New Zealand: A Review of Developments ... (209–217)

New Zealand and the tasks of tree felling and chokersetting have been identified as the most common to serious harm accidents and deaths (Raymond 2014).

Grapple carriages have been around since the 1960s and most are a mechanical type, where the yarder’s wire ropes are used to open and or close the grapple. Using a mechanical type grapple carriage requires a running skyline system, which is most effectively operated with a more modern and expensive interlocked swing yarder (Carson and Jorgenson 1974, Hartsough et al. 1983, Harrill and Visser 2012). New yarders commissioned in the PNW and New Zealand in recent years have mostly been of this type (Fig. 3).

One way to develop the use of grapple carriages onto other yarders was the powered grapple carriage.

Powered grapple carriages require a minimum of two wire ropes to be deployed and can operate on a live skyline system as they have an internal power source which opens, closes and sometimes rotates the grapple. Early developments of motorised grapple carriages started in the PNW in the late 60s when labour costs became unfavourable. Some of the first motorised grapple carriages weighed as much as four tonnes and commonly had problems with mechanical reliability, to a point where they never gained wide spread acceptance and almost vanished from the industry until a recent resurgence. The Forestry Falcon Claw (FFC) is a New Zealand built motorised grapple carriage that has gained market popularity since its introduction in 2012. Another example is the Alpine carriage (with hydraulic accumulator), which is lighter weight (Evanson 2014). Both carriages followed two earlier versions produced by Eagle Carriage & Machine Co. in the USA and numerous other attempts, less commercially successful, by different PNW enterprises. The nature of how these carriages operate compared to mechanical grapples differs and has been the focus of many studies (McFadzean and Visser 2013, Evanson 2014, Nuske 2014). McFadzean and Visser (2013) found that »feeding« grapple carriages (i.e. handing bunches of stems to the carriage with an excavator) could nearly double the productivity of yarder extraction. This practice is becoming more common in New Zealand and worldwide and it has been facilitated by cable-assisted machine use.

6. Integrating New Technologies One disadvantage of the grapple carriage extraction system is that, due to visibility limitations, the operator can have difficulty in picking up stems. Traditionally this limitation is overcome by using a »spotter«. Spotters place themselves along the harvest cor-

214

ridor, where they are able to see the grapple and the stems to be extracted, and provide feedback to the yarder operators in terms of carriage control commands through a walkie-talkie. Although using a spotter is effective, it does not meet the objective of fully mechanising the system.

Two types of camera systems have been developed to provide the yarder operator with good visual information to assist in the grappling phase. The first is a camera system that is either directly integrated into the carriage such as used on the Falcon Forestry Claw, or towed behind the carriage on a rider block which can then be used on a mechanical grapple carriage as well. The second system places the camera in the cutover (Evanson 2013). While this camera also sends the video image back into the yarder cab, the operator can remotely control both the direction and zoom of the lens so as to focus on the immediate area below the carriage.

New yarders incorporate features designed to optimize both the system controls as well as the ergonomics to reduce operator fatigue. Cabs and systems are now being increasingly retrofitted to older machines with features such as adjustable seats with electronic controls simplified to two joy-sticks. Most utilise Programmable Logic Controllers (PLC) or CANBUS for electric over air or electric over hydraulic controls, which can be programmed to optimise the winch performance by selecting a rigging configuration or mode (e.g. grapple, carriage or scab) (Visser et al. 2014). Other features offered by these new computer-operated control systems are: simplicity of operation including more precise control of machine functions, audible alarms and indicators for distance and tension, ease of maintenance with machine diagnostics displayed on a screen in the cab and troubleshooting using a visual Test Points function.

The computer controlled systems also provide the opportunity for autonomous control of certain processes (e.g. carriage outhaul), which have become standard for many of the European manufacturers.

Another example of feedback software is the ACDAT (active data) system, an on board data storage computer system produced by Active Equipment, which is standard with the company’s new yarders or can be retrofitted into older machines (Active Equipment 2016). This system is a one screen, multiple application computer that has four key functions: GPS tracking of choker-setters, live time tension display, modelling of the terrain and operations data recording. For the choker-setter tracking system, the chokersetters wear a special GPS unit that is synced each morning with the yarder. This information is stored by day and data is retained on the computer for a month.

Croat. j. for. eng. 38(2017)2

Cable Yarding in North America and New Zealand: A Review of Developments ... (209–217)

R. Visser and H. Harrill

Information categories include: skyline tension, mainline length, skyline length and engine voltage to name a few. There is also the option for the yarder operator to manually enter haul statistics (e.g. number of pieces extracted for each turn) (Visser et al. 2000).

Increasing the operating range of ground-based systems has been significantly aided by the fast development of winch-assist (also known as cable-assist or tethered) harvesting system (Heinimann 1999, Visser and Stampfer 2015). These systems are now being used to operate on slopes up to 100% (45 degrees), and while they potentially complement the cable yarding systems by providing a safe and effective felling system, in many cases the extraction systems are also winch-assist. This will result in a decline for cable yarding services over time. The full extent of this development has not yet reached its potential, but it will challenge the cable yarders for providing cost-effective extraction services.

Bell, G., 1985: A comparative study of Madill 009 and Washington 88 log haulers. University of Canterbury, Christchurch, New Zealand, 67 p.

7. Conclusion

Cole, E., 2016: Personal communication. Former Madill sales person and cable logging industry historian.

Cable yarders continue to be the backbone of timber extraction on steep terrain. This paper has provided an overview of yarder developments in North America and New Zealand, and identified differences with European developments. While many of these differences can be explained by the forest management requirements, many of the preferences simply reflect regional development. With increases in the cost of labour and fuel, and increasing global market competition, there will be increased focus on operational efficiency. Reduction in energy expenditure and fuel consumption, as well as automated controls for improved safety and worker satisfaction will increase the interest in modern, mainly European designed, yarders. However, in the interim, the continued focus in the PNW and New Zealand will continue to be on robust high production yarders.

8. References Active Equipment, 2016: Active products. Retrieved from www.activeequipment.co.nz/products. Accessed in December 2016.

Amishev, D., 2012: Mechanisation on Steep Slopes in New Zealand. Harvesting Technical Note HTN04-07, Future Forests Research Limited, Rotorua, New Zealand, 8 p.

Bennett, D., McConchie, M.S., 1995: Rigging options for the North Bend system. Logging Industry Research Organisation (LIRO), LIRO Report 20–20, 8 p.

Bentley, T.A., Parker, R., Ashby, L., Moore, D., Tappin, D., 2002: The role of the New Zealand forest industry injury surveillance system in a strategic ergonomics, safety and health research programme. Applied Ergonomics 33(5): 395–403.

Carson, W.W., Jorgenson, J.E., 1974: Understanding interlock yarders. USDA Forest Service, Pacific Northwest Forest Range and Experiment Station. Portland, Oregon, USA. Research Note, PNW-221, 13 p.

Cavalli, R., 2012: Prospects of Research on Cable Logging in Forest Engineering Community. Croatian Journal of Forest Engineering 33(2): 339–356.

City of Eureka. 2010. California History. Available at: http:// www.ci.eureka.ca.gov/civica/filebank/blobdload. asp?BlobID=4459. Accessed December 20, 2010.

Conway, S., 1982: Logging Practices – Principles of Timber Harvesting Systems, Revised Edition. Published by Miller Freeman Publications, San Francisco, California, USA, 432 p.

Dykstra, D.P., 1975: Production rates and costs for cable, balloon, and helicopter yarding systems in old-growth Douglasfir. Forest Research Laboratory, Oregon State University, USA. Research Bulletin 1, 57 p.

Dykstra, D.P., 1976a: Production rates and costs for yarding by cable, balloon, and helicopter compared for clearcuttings and partial cuttings. Forest Research Laboratory, Oregon State University, USA, 44 p.

Dykstra, D.P., 1976b: Yarding delays for advanced logging systems. Forest Research Laboratory, Oregon State University, USA. Research Paper 33, 11 p.

Ellgard, J., 2016: The man from LIRA. New Zealand Logger 2016(3): 34–45.

Evanson, A.W., Kimberley, M.O., 1992: An analysis of shiftlevel date from six cable logging operations. FRI Bulletin 174, Forest Research Institute, Rotorua, New Zealand, 24 p.

Evanson, T, 2013: Hauler vision system: testing of Cut-over camera. Future Forests Research Ltd. (FFR) HTN 05-03, 6 p.

Evanson, T., 2014: Alpine grapple carriage – from prototype to production. Future Forests Research Ltd. (FFR) HTN 06-09, 6 p.

Ammeson, J.E., 1984: A production study of the pendulum balloon logging system. Oregon State University, Corvallis, Oregon, USA, 78 p.

Falk, G.D., 1981: Predicting the payload capability of cable logging systems including the effect of partial suspension.

USDA Forest Service. RP-479. Northeast For. Exp. Stn., Broomall, Pennsylvania, USA, 29 p.

Balcom, J.C., 1983: Skyline carriages for small-wood harvesting. Oregon State University, Corvallis, Oregon, USA, 98 p.

Finnegan, D., Faircloth J., 2002: New Zealand Cable Yarder Survey. Unpublished Report, Rotorua, New Zealand, 9 p.

Croat. j. for. eng. 38(2017)2

215

R. Visser and H. Harrill

Cable Yarding in North America and New Zealand: A Review of Developments ... (209–217)

FITEC, 2000: Best Practice Guidelines for Cable Logging. Forest Industry Training and Educational Council, Auckland, New Zealand, 138 p.

Under Mountainous Conditions with special attention to Ergonomics, Accessibility and Environmental Protection, Harbein, P.R. of China, 9 p.

Forrester, P.D., 1995: Evaluation of three cable-yarding systems working in a coastal old-growth forest. Technical report – Forest Engineering Research Institute of Canada 112, 17 p.

Kirk, P.M., Sullman, M., 2001: Heart rate strain in cable hauler choker setters in New Zealand logging operations. Applied Ergonomics 32(4): 389–398.

Fraser, D., Bennett, D., 1996: Potential causes of tower collapses. LIRO Report 21(6), Logging Industry Research Association, 5 p.

Larson, R.S., 1978: Compendium of major cable logging systems. Published by Interforest AB, Lidingo, Stockholm, Sweden, 112 p.

Fraser, D., 1996: Hauler tower collapses. New Zealand Logging Research Organisation Report 21(7): 6 p.

Liley, W.B., 1983: Cable logging handbook. New Zealand Logging Industry Research Association. Rotorua, New Zealand, 103 p.

Harper, C., 1992: Summary of New Zealand cable logging production studies. Logging Industry Research Organisation (LIRO), Rotorua, New Zealand, 155 p.

Harrill, H., Visser, R., 2012: Matching Rigging Configurations to Harvesting Conditions. Harvesting Technical Note HTN04-06, Future Forests Research Limited, Rotorua, New Zealand, 6 p.

Harrill, H., 2014: Improving Cable Logging Operations for New Zealand’s Steep Terrain Forest Plantations. A thesis submitted in partial fulfilment of the requirements for the degree of Doctorate of Philosophy in Forest Engineering, School of Forestry, University of Canterbury, New Zealand, 242 p.

Hartsough, B.R., Miles, J.A., Darling, G.W., 1987: Running skyline analyses: consideration of yarder characteristics. Forest Products Journal 37(1): 51–55.

Hartsough, B., 1993: Benefits of remote tension monitoring.

Logging Industry Research Organisation. LIRO Report 18– 23, 13 p.

Heavy Equipment Forums 2016: S. Madill Blacksmith Founded in 1911 in Nanaimo BC (thread). Available at: http://www. heavyequipmentforums.com/threads/s-­madill-­blacksmithfounded-in-1911-in-nanaimo-bc.16539/ [Accessed January 9 2016] Heinimann, H.R., 1999: Ground-based harvesting technologies for steep slopes. In Proc., International Mountain Logging and 10th Pacific Northwest Skyline Symposium March 28 – April 1, eds. J. Sessions and W. Chung, Corvallis, Oregon, USA, 19 p.

Heinimann, H., Stampfer K., Loschek, J., Caminada, L., 2000: Perspectives on Central European Cable Yarding Systems.

The International Mountain Logging and 11th Pacific Northwest Skyline Symposium, Schiess, P., and F. Krogstad (eds.), College of Forest Resources, University of Washington, Seattle, Washington, USA, 268–279.

Lysons, H.H., Mann, C.N., 1967: Skyline Tension and Deflection Handbook. RP-PNW-39. USDA Forest Service. Portland, Oregon, USA, 39 p.

Mann, C.N., 1976: Why running skylines and interlock yarders. Proceedings from the Skyline Logging Symposium, December 1976, Vancouver, British Columbia, Canada, 17–26.

Mann, J.W., Pyles, M.R., 1988: Cable logging mechanics research at the Oregon State University. International Mountain Logging and Pacific Northwest Skyline Symposium, December 12–16, Portland, Oregon, USA, 1–6.

MBIE, 2012: Approved Code of Practice for Safety and Health in Forest Operations. Ministry of Business Innovation and Employment, Wellington, New Zealand, 134 p.

McConchie, M.S., 1987: Logging system evaluation: Madill 071/MSP system/Northbend/Shotgun system. Case studies 1, 2, 3 and 5. Forest Research Institute Project Records 1556, 1643, 1704, 205 p., (unpublished).

McFadzean, S., Visser, R., 2013: Falcon Forestry Claw Grapple: Productivity and Ergonomics Harvesting. Technical Note HTN05-06, Future Forests Research Limited, Rotorua, New Zealand, 7 p.

McMahon, S., 1995: Cable logging disturbance on an unstable slope: A case study. New Zealand Logging Industry Research Organization Report 20(12): 9 p.

Murphy, G., 1977: Cable logging in mature radiata pine: a case study of a mobile Madill operation. In Economics of silviculture report 103, New Zealand Forest Research Institute, 76 p. (unpublished).

Nuske, S., 2014: A comparative study of mechanised cable harvesting systems in New Zealand. Bachelor dissertation, School of Forestry, University of Canterbury, Christchurch, New Zealand, 47 p.

Horcher, A., Visser, R., 2011: Using on-board GPS to identify training needs of helicopter pilots. Croatian Journal of Forest Engineering 32(2): 481–488.

OR-OSHA, 1993: Yarding and Loading Handbook. Oregon Occupational Safety and Health Division, Salem, Oregon, USA, 184 p.

Kellogg, L., 1987: A cable hauling trial with the Madill 071 – using three different rigging systems. LIRA Report, Logging Industry Research Association 12(11): 8 p.

OR-OSHA, 2008: Division 7 forest activities code. Services, D.o.C.a.B. (ed.). Oregon Occupational Safety and Health Division, Salem, Oregon, USA, 28 p.

Kirk, P.M., Parker, R.J., 1994: Physical demands of steep terrain forestry work in New Zealand. Proceedings of the International IUFRO/NEFU/FAO Seminar on Forest Operations

Pestal, E., 1961: Seilbahne und Seilkrane für Holz und Material Transport. Published by Verlag Georg Fromme & Co., Horn, Austria, 473 p. [in German]

216

Croat. j. for. eng. 38(2017)2

Cable Yarding in North America and New Zealand: A Review of Developments ... (209–217)

R. Visser and H. Harrill

Raymond, K., 2012: Innovation to Increase Profitability of Steep Terrain Harvesting in New Zealand. Proceedings of NZ Institute of Forestry Conference, 1–4 July, Christchurch, New Zealand, 15 p.

Stokes, B.J., Ashmore, C., Rawkins, C.L., Sirois, D.L., 1989: Glossary of terms used in timber harvesting and forest engineering. General Technical Report SO-73. USDA Forest Service, Southern Forest Experiment Station, USA, 33 p.

Raymond, K., 2014: Safety benefits of harvesting research programme. New Zealand Journal of Forestry 59(3): 9–13.

Studier, D.D., Binkley, V.W., 1974: Cable logging systems.

USDA For. Serv., Div. of Timber Management, Portland, Oregon, USA, 205 p.

Rice, F.M., 1974: Increasing yarder production with water cooled brakes. Loggers Handbook. Vol. 34. Pacific Logging Congress. Portland, Oregon, USA, 3 p.

Rutherford, D.A., 1996: Productivity, costs, and optimal spacing of skyline corridors of two cable yarding systems in partial cutting of second-growth forests of coastal British Columbia. University of British Columbia, Vancouver, British Columbia, Canada, 100 p.

Samset, I., 1985: Winch and Cable Systems, Construction Work and Forest Operations. Nijhof and Junk Publishers, Dordrecht, The Netherlands, 533 p.

Sessions, J., 1991: Understanding the mechanical and operational characteristics of heavy equipment used in forest engineering. Oregon State University Forest Engineering Institute course notes, Oregon, USA, 43 p.

Sinner, H.U., 1973: Simulating skyline yarding in thinning young forests. Oregon State University, Corvallis, Oregon, USA, 90 p.

Slappendel, C., Laird, I., Kawachi, I., Marshall, S., Cryer, C., 1993: Factors affecting work-related injury among forestry workers: A review. Journal of safety research 24(1): 19–32.

Spinelli, R., Visser, R., Thees, O., Sauter, U-H., Krajnc, N., Riond, C., Magagnotti, N., 2015: Cable Logging Contract Rates in the Alps: the Effect of Regional Variability and Technical Constraints. Croatian Journal of Forest Engineering 36(2): 179–187.

Stampfer, K., Lietner T., Visser, R., 2010: Efficiency and ergonomic benefits of using radio controlled chokers in cable yarding. Croatian Journal of Forest Engineering 31(1): 1–9.

Sullman, M.J., Kirk, P.M., Parker, R.J., Gaskin, J.E., 1999: New Zealand logging industry accident reporting scheme: Focus for a human factors research programme. Journal of safety research 30(2): 123–131.

Visser, R., Stampfer, K., 1998: Cable extraction of harvester felled thinnings: An Austrian case study. Journal of Forest Engineering 9(1): 39–46.

Visser, R., Evanson, T., Palmer W., 2000: Operational Monitoring for Evaluating Work Force Performance. Proceedings of the Council on Forest Engineering Annual Meeting, Technologies for the New Millennium. Kelowna, Canada, 4 p.

Visser, R., Spinelli, R., Magagnotti, N., 2011: Landing characteristics for harvesting operations in New Zealand. International Journal of Forest Engineering 22(2): 23–27.

Visser, R., 2013: Survey of Yarders Used in New Zealand.

Harvesting Technical Note HTN06-03, Future Forests Research Limited, Rotorua, New Zealand, 4 p.

Visser, R., 2014: Harvesting cost and Productivity Benchmarking: 2013 Update. Harvesting Technical Note HTN06-06, Future Forests Research Limited, Rotorua, New Zealand, 5 p.

Visser, R., Raymond, K., Harrill, H., 2014: Mechanising Steep Terrain Harvesting Operations. New Zealand Journal of Forestry 59(3): 3–8.

Visser, R., Stampfer, K., 2015: Expanding Ground-based Harvesting onto Steep Terrain: A Review. Croatian Journal of Forest Engineering 36(2): 321–331.

WorkSafeBC, 2006: Cable Yarding Systems Handbook. Workers’ Compensation Board of British Columbia. Vancouver, British Columbia, Canadam, 204 p.

Authors‘ addresses:

Received: March 20, 2017 Accepted: June 14, 2017 Croat. j. for. eng. 38(2017)2

Prof. Rien Visser, PhD. * e-mail: rien.visser@canterbury.ac.nz Hunter Harrill, PhD. e-mail: hunter.harrill@canterbury.ac.nz Senior Research Assistant University of Canterbury College of Engineering Private Bag 4800 Christchurch NEW ZEALAND * Corresponding author

217

Subject review

Trends and Perspectives in Coppice Harvesting Raffaele Spinelli, Natascia Magagnotti, Janine Schweier Abstract Coppice management is applied to many species, in many countries and in many ways, so that several harvesting techniques have been developed depending on specific local conditions.

However, all techniques designed for handling coppice stands must be suitable for coping with small stem size and stump crowding, and often with steep and generally difficult terrain.

Traditional harvesting systems are labor intensive because they usually include motor-manual felling and processing into one-meter lengths at the stump site, and manual loading of the short logs onto pack animals or tractors. Thus, in industrialized countries, these systems are no longer viable and they are being replaced with mechanized cut-to-length and whole-tree harvesting, depending on site conditions. Mechanization dramatically improves worker safety, and compensates for the reduced availability of rural labor, with their propensity to perform heavy and low-paying jobs. Much progress has already been made, with the massive introduction of modern harvesters, forwarders and tower yarders in coppice harvesting operations. The presence of multiple stems on the same stump offers a serious challenge to the introduction of mechanized felling to coppice harvesting operations, because stump crowding hinders felling head movements. However, new machines have been designed that can handle coppice stumps.

Further research should address the relationship between stump damage and regeneration vigor, in order to define new standards for cut quality. Silvicultural practice may need adapting to the new harvesting technology and to the products required by the modern bio-economy.

Keywords: felling, extraction, productivity, logging, mechanization, biomass, management

1. Introduction Coppicing is a traditional silvicultural system whereby stand regeneration after cut is obtained from the re-sprouting of cut stumps, rather than from the establishment of new trees from seed. As a consequence, this system is only suited to those species that can sprout new shoots from their stumps after cutting.

Such capacity is typical of some hardwood species, if the interval between cuts does not exceed 50â&#x20AC;&#x201C;60 years.

To keep the re-sprouting ability, frequent cutting is needed and the application of coppice management requires relatively short rotations.

Coppice management is extremely efficient, because it offers the benefits of simplified care, prompt regeneration and short waiting time. On the other hand, coppice management has some important limitations, and especially the exclusion of softwood species and the relatively small size of assortments proCroat. j. for. eng. 38(2017)2

duced, which is the obvious consequence of the short rotations characterizing coppice management.

For these reasons, wood from coppice forests is likely used as energy and as industrial wood, typically firewood and pulpwood, although coppice forests are also a main source of posts, tool handles and fencing materials (Buckley 1992). Historically, coppice forests are associated with rural communities and represent the ideal complement to conventional agricultural systems. Coppice woodland was widespread all over Europe until recent times, when industrialization transformed both the economy and the landscape of many regions (Coppini and Hermanin 2007). In the post-war years, traditional coppice systems have suffered from the competition of oil and plastic, which have resulted in a decreasing interest towards the active management of traditional coppice stands (HĂŠdl et al. 2010). Therefore, large areas of coppice forests are not managed any longer.

219

R. Spinelli et al.

Trends and Perspectives in Coppice Harvesting (219–230)

However, in the last years new applications of the coppice concept have been developed, specifically designed for industrial use and/or for a changing agriculture. For the sake of clarity, the authors of this paper have decided to distinguish among three broad types of coppice stands, as follows (Table 1).

This review focuses only on the harvesting of conventional coppice forests, because a comprehensive analysis of all three types could be too long for just one paper, and potentially confusing. Besides, the surface covered with SRF and SRC is still relatively small in Europe, especially if compared with that covered by conventional coppice. In fact, SRF covers about half million hectares concentrated in the Iberian Peninsula, where SRF eucalyptus plays a major economical role. On the other hand, SRC is unlikely to cover more than 25,000 hectares in Europe, and it does not have any significant impact on the European economy yet. In contrast, the importance of conventional coppice is vastly larger, and it dwarves those of both SRF and SRC. The total surface of conventional coppice in the EU and its neighbors is estimated to over 26 million hectares (Table 2), which is 50 times larger than the surface of SRF and 1000 times larger than that of SRC.

Table 1 Main types of coppice stands Coppice definition

Conventional

SRF

SRC

Species type

Quercus sp.

Fagus sylvatica L. Ostrya carpinifolia L. Etc.

Populus sp.

Eucalyptus sp.

Acacia sp.

Salix sp.

Populus sp.

Eucalyptus sp.

Rotation years

15–40

5–15

2–5

Firewood

Pulpwood

Chips

Industrial and small-scale forestry

Industrial forestry

Industrial agriculture

Forest

Forest

Agricultural

Product type Economy domain Harvest technology

Conventional coppice. This is established with indigenous hardwood species (oaks, chestnut, beech, hornbeam etc.) and occasionally exotic ones (Robinia). It is harvested on 15–40 years rotations for a large variety of products, and it is managed within the framework of a rural economy according to local traditional practice. It is generally harvested with forestry equipment, small-scale or industrial.

Short rotation forestry (SRF). Stands are established with exotic fast-growing species (eucalypt, acacia) and harvested on 5–15 years rotations for the production of industrial feedstock (generally pulpwood). SRF developed within the framework of a large-scale industrial economy and it is often geared to supply large industrial plants. SRF stands are often (but not exclusively) managed as coppice, and they occasionally undergo shoot reduction treatments (thinning). Stands are generally harvested with industrial forestry equipment, and occasionally with small-scale forestry equipment.

Short rotation coppice (SRC). Stands are established on ex-arable land with genetically-improved fast-growing species, indigenous (willow, poplar) or exotic (eucalypt, robinia). They are harvested on 2–5 years rotations for the production of industrial feedstock (generally energy biomass), and managed within the framework of small-scale or industrial agriculture. So far, SRC represents a niche sector and it is generally harvested with modified agricultural equipment.

220

Table 2 Coppice forests in the EU and its neighbors Country

Mi, ha

France

6.8

Turkey

5.7

Italy

3.3

Spain

3.0

Bulgaria

1.8

Greece

1.6

Serbia and Montenegro

1.4

Bosnia and Herzegovina

0.8

Republic of Macedonia

0.6

Croatia

0.5

Hungary

0.5

Albania

0.4

Romania

0.3

TOTAL

26.7

Note: the list includes only the Countries with at least 100,000 ha of coppice.

Coppice is present in many other European countries than reported in the table (extracted from Nicolescu et al. 2015)

Conventional coppice forests represent a very large biomass resource, or a very serious landscape management problem if no productive use can be made of their potential, because abandoned coppice forests may degrade and become very susceptible to pests and forest fires. However, these immense reserves of Croat. j. for. eng. 38(2017)2

Trends and Perspectives in Coppice Harvesting (219–230)

R. Spinelli et al.

woody biomass may represent the ideal solution to matching the large demand for biomass feedstock generated by a rapidly growing bio-economy (Matula et al. 2012). Biomass users need huge amounts of lowquality wood at short intervals, which is what coppice was designed to offer in the first place (Jansen and Kuiper 2004). While new short-rotation plantations are being established on ex-arable land, existing conventional coppice forests might be simply recruited into the new economy as an even larger source of raw material, thus being returned to active and profitable management when the demand for traditional coppice products is dwindling (Hédl et al. 2010).

The goal of this paper is to produce a general review of existing literature about the harvesting of conventional coppice stands, with the intent of:  building a general framework of available technologies and techniques  providing general productivity benchmarks that may serve as a base reference  describing current trends and future perspectives.

This paper does not have the ambition to include every single study appeared in the past, or to describe all possible techniques or to offer a comprehensive coverage of all aspects of coppice harvesting – and especially site impacts and human factors. However, the general picture drawn in this paper may represent a viable background for framing existing and new studies.

thinning (conversion) generates losses (Motta et al.

2015). That is true for coppice as well as for high forest (Petty and Kärhä 2011). In the past decades, conversion was often subsidized with public grants, in an attempt to drive heavily anthropized ecosystems towards more natural forms, which was especially attractive at a time when energy wood was being phased out (Stajic et al. 2009). Today, a new appreciation of the cultural and ecological value of coppice stands has combined with the growing demand for wood biomass in causing a general reconsideration of the past emphasis on coppice conversion (Urbinati et al. 2015).

Coppice management implies short rotations, and that has a strong effect on product type. Stems are cut before they can get very large, and they are best suited for conversion into small-size assortments. Mean stem size varies most often between 0.05 and 0.25 m3, and it is smallest for oak and largest for chestnut, regardless of treatment type (clearcut or thinning). In general, coppice harvesting yields very limited amounts of timber, which is obtained from the standards released in the previous harvest.

2. Silviculture and products The traditional management of conventional coppice forests is quite simple, and it is based on clearcutting at the end of rotation. Standards are often released, with a density ranging between 50 and 100 trees per hectare, depending on the species. No other interventions are needed. If coppice management is no longer desirable, then the over-mature stand is thinned by removing approximately 40% of the standing volume. This intervention is expected to favor conversion into high forest, and it is followed by additional thinning treatments until the mature transitional forest is ready for regeneration felling.

The final harvest of a mature coppice stand commonly yields between 90 and over 200 m3 ha-1, depending on species, age and site productivity. The harvest obtained from thinning (conversion) over-mature coppice is more variable and depends on how old is the stand, but it generally varies from 40 to 200 m3 ha-1. As a general rule, clear-cutting accrues profits, whereas Croat. j. for. eng. 38(2017)2

3. Traditional harvesting systems In former times, manual work was dominant and it made sense to reduce cut stems to such a size that could be easily handled manually as early as possible, if that would not degrade assortment value. Firewood was cut into one-meter lengths at the stump site, before loading it on pack animals (Fig. 1) for extraction and transportation (Carette 2003, Lepper and Frere 1988). With minimum adjustments, animal extraction remained in use until a few years ago in countries such as Italy and France (Baldini and Spinelli 1989). Although there is no recent bibliography on the subject, anecdotal evidence points at its widespread current

Fig. 1 Extraction by pack-mules in a mixed oak coppice

221

R. Spinelli et al.

Trends and Perspectives in Coppice Harvesting (219–230)

use in the Balkans and Greece (Gallis 2004), with significant survivals in southern Italy as well (Civitarese et al. 2006). The only modern adaptations to this ancestral system are the introduction of chainsaws for felling and processing, and of trucks for transportation, so that animal work is limited to extraction (Piegai et al. 1980).

Small stem size, uncomfortable working position and the need for turning all stems into one-meter lengths combine to determine a very low productivity of motor-manual felling and processing, which is reported in the range between 0.3 (Piegai 2005) and 1.4 m3 (Picchio et al. 2009) per scheduled machine hour (SMH) and operator. Manual bunching of one-meter logs contributes to such low productivity figures. Significant manual inputs are also required for leading the animals to the loading site, loading them with ca.

200 kg of firewood each, and unloading the product once back at the roadside landing (Spinelli et al. 2016a).

The typical team comprises 2 operators and between 6 and 12 mules or horses. Extraction distances commonly vary from 200 to 800 m, depending on slope gradient and the direction of extraction: shorter distances can be sustained if extraction proceeds uphill, on steep slopes. The productivity of such team may range from 1.3 to 1.8 m3 SMH-1, depending on the number of animals and on work conditions (Baldini and Spinelli 1989).

The hourly cost of animals is relatively low, and it has been estimated to less than 8 € hour-1, excluding the driver (Magagnotti and Spinelli 2011). However, if operators were paid the 25 € SMH-1 rate characterizing modern logging operations in industrialized European countries, this system would be too expensive to run. In that case, the cost would range between 18 and 80 € m-3 for motor-manual felling and processing, and between 70 and 100 € m-3 for extraction. Even the lowest cost combination (90 € m-3) would be higher than the cost paid at the landing for quality firewood in the same countries, which is reported to be around 75 € m-3 (Spinelli et al. 2014). Obviously, this system is still competitive where labor cost is much lower than in industrialized economies, or where irregular underpaid labor is introduced, which is a growing phenomenon in many regions and represents the bane of law-abiding regular loggers (Pettenella and Secco 2004).

Once solved the problem of labor cost, mule logging offers several advantages, because it can be deployed in rough terrain and does not require opening new skid trails, which may incur additional cost and impact (Magagnotti and Spinelli 2011). More in general, animal logging allows a dramatic reduction of site impacts compared to other logging methods

(Shresta et al. 2008, Spinelli et al. 2010a), especially when a dense residual stand is present, as is the case with coppice conversions. In fact, the anachronistic survival of animal logging in modern countries such as Italy is partly due to the subsidies released in the recent past for coppice conversion, which allowed bearing the cost of otherwise unsustainable work methods. However, the main threat to the survival of animal logging in industrialized countries is not financial viability, but the inconvenience of constant care.

Animals must be attended to on a daily bases, and they cannot be parked in a barn and forgotten when the logging season is over.

222

4. Attempts at modernizing traditional harvesting systems For a short while, chutes were the rage, and they were purchased in significant numbers especially by cooperatives and public administrations (Piegai 1985).

While chutes could be easily stored for extended periods with no further care, log sliding turned out to require larger manual inputs than animal logging (Baldini 1987). Obviously, a cooperative or a public administration cannot offset labor cost by hiring irregular workers, and therefore chutes remained in some use as long as subsidies were available, disappearing with the end of public support.

All the above explains the search for a mechanical surrogate of the traditional mules, already started in the late 1980s (Baldini and Spinelli 1990). Over time, various micro-tractors have been designed and tested (Gallis 2004, Magagnotti et al. 2012) but none has ever obtained commercial success. Eventually, pack-mules have been replaced with the so called pack-tractor, i.e. a farm tractor equipped with front and rear bins capable of containing ca. 3 tonnes of one-meter logs (Piegai and Quilghini 1993). The bins are normally mounted on hydraulic lifts, so that they can be lowered to the ground for easier manual loading (Fabiano 2006). This solution is quite crude and it does stress the tractor frame, so that much anecdotal evidence is available about tractors splitting in half at the clutch flange. However, simple solutions often stick, and so it is for this artless method, which offers the benefits of minimum investment and specialization. The limits of the method are represented by extraction distance and terrain roughness. Small payload size prevents efficient use on distances longer than a few hundred meters, while the limited mobility of an encumbered farm tractor requires relatively easy terrain, or a good network of skid trails. Productivity is higher than reported for mule teams, and it Croat. j. for. eng. 38(2017)2

Trends and Perspectives in Coppice Harvesting (219–230)

R. Spinelli et al.

varies from 2 to 4 m3 SMH-1 with a crew of two (Piegai 2005, Verani and Sperandio 2003).

it includes about the same task sequence observed for the traditional system, and namely: felling and processing at the stump site, and forwarding of short logs to the roadside over the forest floor. The main difference is that all tasks are performed by machines, so that »no man is on the ground, no hand touches the wood«.

For this reason, the system must be adapted by increasing log length, because one-meter logs are too short for efficient mechanical handling. When CTL is introduced to coppice harvesting, log length is generally increased to 2 or even 3 m (Fig. 2).

The presence of multiple stems on the same stump offers a serious challenge to the mechanized felling of coppice, because stump crowding hinders head movements, and can be handled by very compact units only (Labelle et al. 2016). The ideal head for harvesting coppice is short (Zinkevičius et al. 2012), has two mobile knives only (Spinelli et al. 2002), and does not close its rollers in a triangular configuration (Moscatelli et al. 2010). That is the case of AFM 60, Kesla Foresteri RH20, SIFOR 350 or UTC CTL40 HW, just to mention some of the heads that have been successfully tested for hardwood harvesting (Martin et al. 1996, Spinelli et al. 2002, Suchomel et al. 2012).

Regardless of machine choice, operator skills play a major role when applying CTL harvesting to coppice stands (McEwan et al. 2016).

The productivity of a modern harvester deployed in conventional coppice operations may vary from 2 (Forestry Commission 2011) to almost 10 (Spinelli et al. 2010c) m3 SMH-1, depending on stem size and operator proficiency. The productivity of the forwarder commonly ranges between 5 (Grulois et al. 1996) and 10 (Spinelli et al. 2014) m3 SMH-1, depending on machine model and extraction distance. Assuming an hourly rate of 120 € for the harvester and 80 € for the forwarder, the harvesting and extraction cost would vary from 20 to 50 € m-3, which is within the price bracket of industrial wood users and much cheaper than the cost incurred for motor-manual work. Extraction can also be performed with forestry-fitted farm tractors, which allows reducing investment cost but results in lower payload and productivity (Spinelli et al. 2004).

Introduction of CTL harvesting is easier in the presence of industrial users, who can better support a sustained work flow. That is why CTL harvesting was first introduced to commercial coppice operations in central France, where the abundant chestnut resource was supplied to large particle board factories (Martin et al. 1996). It is only much later that the Italians (Spinelli et al. 2010b) and the Germans (Suchomel et al. 2011) followed suit.

It is very important to remember that all these developments are the consequence of specific silvicultural trends, and especially the strong drive towards coppice conversion. The maneuverability constraints imposed by selection thinning have systematically favored such methods as mule extraction, sliding in chutes and pack-tractors. There would be no reason to use relatively small size boxes, if the circulation of a proper tractor-and-trailer unit was not hindered by a dense residual stand, without suitable openings for machine traffic. In fact, dedicated forwarding trailers are used in clearcuts, and offer better performance than boxes, even when manual loading is applied (Spinelli and Baldini 1992). Over 30 years of experience with suboptimal working method should motivate a general revision of the traditional relationships between silviculture and operation management, and lead to re-assessing the past emphasis on coppice conversion. Low operational efficiency is acceptable as long as grant money is available to cover unsustainable harvesting costs. When such grants are no longer on the table, then re-thinking the whole strategy is the only alternative to the end of active management, which would also configure as the end of coppice.

5. Mechanized cut-to-length harvesting Mechanized cut-to-length (CTL) harvesting is based on the introduction of the classic harvester-forwarder combination (Kellogg et al. 1993). While representing a radical technological innovation, CTL harvesting is not a revolutionary system change, because

Fig. 2 Mechanized cut-to-length harvesting in a chestnut coppice Croat. j. for. eng. 38(2017)2

223

R. Spinelli et al.

Trends and Perspectives in Coppice Harvesting (219–230)

6. Whole-tree harvesting and tree-length harvesting Whole-tree harvesting (WTH) consists of felling trees and extracting them whole to the landing, where they are processed into commercial assortments (Stokes et al. 1989). WTH offers the advantage of simplified in-forest handling and is first documented in the US (Kammenga 1983). This basic scheme has proven to be so effective that it has remained virtually unchanged and appreciated until our days (Mitchell and Gallagher 2007). The main advantage of this system is to postpone processing to the landing, where it can be mechanized if terrain constraints make the stand inaccessible to harvesters (Adebayo et al. 2007). Even if no harvester is available, WTH moves motor-manual processing to a better worksite, where operation is more comfortable and productive (Spinelli et al. 2009).

As processing is moved to the landing, stump site work is simplified, which results in a relatively high productivity. Motor-manual directional felling may proceed at a pace between 1 (Bajić and Danilović 2004) and 4 (Spinelli and Magagnotti 2007) m3 SMH-1 operator-1. If the terrain is accessible to mechanical equipment, then feller-bunchers (Fig. 3) can be introduced and productivity will increase dramatically, reaching values between 4 (Spinelli et al. 2007) and over 8 (Schweier et al. 2015) m3 SMH-1. In fact, the main operational benefit of mechanized felling is not only the increased productivity, but rather the better presentation of felled trees, which are gathered in bunches and aligned towards the skidding tracks, so that extraction productivity receives a dramatic boost. Studies about the skidding of whole coppice trees report a wide range of productivity figures, which go from less than 3 m3 SMH-1 for skidding with a forestry-fitted farm

Fig. 3 Felling coppice with accumulating shears

224

Fig. 4 A light tower yarder routinely used in coppice operation tractor (Cantiani and Spinelli 1996), to 5 (Currò and Verani 1984) or even 8 (Canga et al. 2014) m3 SMH-1 when a dedicated skidder is used.

On steep terrain, cable yarding (Fig. 4) is the costeffective alternative to building an extensive network of skidding trails, and results in a much lighter site impact compared to ground-based logging (Bolding et al. 2011, Spinelli et al. 2010). Productivity is somewhat lower than in ground-based operations, and varies from 3 (Currò and Verani 1986, Verani et al.

2008) to 7 (Spinelli et al. 2014) m3 SMH-1. However, the main difference is the crew size, which increases to 3 or occasionally 4 workers, whereas only 1 or 2 workers are required for a skidder. Furthermore, yarder set up and dismantle are time consuming, and they may add 20–25% to the actual extraction time (Spinelli et al. 2016b).

Once at the landing, whole trees are converted into conventional assortments (i.e. firewood, pulpwood, etc.), or thrown straight into a chipper. Whole-tree chipping offers the benefits of increased product recovery, simplified processing and higher productivity (Herrick 1982). Whole-tree chipping was tested early on in the Italian coppice stands (Baldini 1973), at about the same time as it appeared in the US (Koch 1973) and well before it was introduced to softwood thinning.

Since then, whole-tree chipping has plaid a minor but steady role in coppice operations (Spinelli and Hartsough 2001), with the main purpose of supplying particle-board factories and some of the early chipfuelled boilers. Today, a booming demand for biomass chips has created the conditions for a rapid expansion of whole-tree chipping, which has become very popular in many regions. The efficiency gains obtained with whole-tree chipping often lead to turning into chips Croat. j. for. eng. 38(2017)2

Trends and Perspectives in Coppice Harvesting (219â&#x20AC;&#x201C;230)

those stems that could yield quality firewood, despite the higher price fetched by the latter assortment. An additional advantage of chip production is in the type of customer, because chips are generally delivered to industrial customers that absorb large quantities and offer better solvency, whereas firewood is sold to a large number of individual buyers, which complicates all administrative matters, including negotiation, billing and payment collection.

Despite its many advantages, WTH must be considered with some caution because of the risk for soil nutrient depletion (Helmisaari et al. 2011), which may result from removing nutrient-rich branch material (Lamers et al. 2013). Furthermore, taking branches to the landing may cause significant slash accumulation and disposal problems, if no market is available for them. In those cases, trees can be delimbed and topped before extraction, but not cut to length. That allows reducing inefficient stump-site work compared to traditional short wood harvesting, while increasing onsite biomass retention to mitigate possible adverse effects (Mika and Keeton 2013). This work system is known as tree-length harvesting (TLH) and is widely used to avoid the accumulation of residues at spaceconstrained landings (Westbrook et al. 2007). Substitution of TLH determines a large (>50%) increase of stump-site work compared to WTH, whereas landing work is only slightly reduced. Decreased work efficiency leads to a general increase of logging cost, which has been estimated at 10â&#x20AC;&#x201C;15% over WTH (Putnam 1983, Spinelli et al. 2016b).

7. Cutting technology and coppice regeneration One of the main obstacles when trying to introduce mechanized cutting to coppice operations is represented by the absolute need to prevent stump damage, in order to guarantee prompt regeneration. All cuts should be clean and as near to the ground as possible.

Unfortunately, mechanical felling can seldom guarantee that these requirements are met, and therefore forest managers often forbid mechanized felling in their coppice forests and prefer incurring the higher cost of motor-manual felling.

Harvesting machines equipped with shears used in coppice forests are not a favored option because they do produce taller stumps than obtained with chainsaws or disc saws under the same conditions (Schweier et al. 2015, Spinelli et al. 2007). That depends on a number of factors, and especially on their working mechanism, which requires engulfing the stem within the full arc described by the closing blades.

Croat. j. for. eng. 38(2017)2

R. Spinelli et al.

Fig. 5 Tall chestnut stumps after cutting with a harvester That might be difficult to achieve when too close to the ground and near the insertion of the stems on the stump. Therefore, operators tend to move the cutting point higher up, where the shear can wrap around the stem, thus leaving tall stumps (Fig. 5).

Furthermore, shears may also cause significant stump damage (De Souza et al. 2016), which is generally explained by high compression stress (McNeel and Czerepinski 1987). Cracks and stump pull may be observed on a large proportion of the stumps cut with a shear, and their incidence varies between 20% (Spinelli et al. 2014b) and 70% (Schweier et al. 2015).

In contrast, disc saws may produce very low cuts if the operator is skilled, even lower than could be produced with a chainsaw (Han and Renzie 2005, Hall and Han 2006). Stump damage levels are also lower for disc saws than for shears (Schweier et al.

2015). The use of a disc saw generally results in improved cutting quality, which should relieve most concerns. The main obstacle to the introduction of disc saws is the excessively large size (and cost) of most machines currently available on the market.

With few exceptions (Delasaux et al. 2009), commercial disc saw models weigh over 2 tonnes and are installed on expensive dedicated prime movers, or on very large excavators. On the other hand, chainsaw type felling heads are vulnerable to contact with soil and to frequent chain derail, the latter being generally caused by crowded stumps.

To conclude, shears represent the cheapest and most effective solution to mechanized felling in coppice stands, despite the lower cut quality (Chakroun et al. 2016). If shears are deployed, then motor-manual post-harvest stump trimming is a viable solution to excessive cutting height and felling-related stump

225

R. Spinelli et al.

Trends and Perspectives in Coppice Harvesting (219â&#x20AC;&#x201C;230)

damage, despite the additional cost and value loss derived from such practice (Martin et al. 1996).

In fact, there is very little scientific evidence about the effect of cut height and stump damage on stump mortality and re-sprouting vigor (Piskoric 1963, Roth and Hepting 1943). Increased stump mortality seems to be associated with the most severe damage type only (De Souza et al. 2016, Ducrey and Turrel 1992, Spinelli et al. 2017), which is relatively rare. None of the studies that have compared manual and mechanical cutting have found any significant differences in stump mortality or resprouting vigor (Crist et al. 1983, Ducrey and Turrel 1992, Giudici and Zingg 2005, Pyttel et al. 2013, Spinelli et al. 2016c). If at all, cutting with shears seems to prompt the emission of a larger number of shoots than when cutting with a saw (Cabanettes and Pagès 1986, HytÜnen 1994, De Souza et al.

2016, Spinelli et al. 2017). In fact, resprouting vigor seems directly related to stump size, rather than to cut quality (Johnson 1975, Ducrey and Turrel 1992, McDonald and Powell 1983, Souza et al. 2016).

active management impossible, unless subsidies are released.

In fact, financial viability is not the main issue when decisions on coppice management strategies are taken. Manual work is associated with the highest accident risk and accident severity, and it accounts for most of the fatal accidents recorded in forest operations (Albizu et al. 2013). Previous studies have shown that the introduction of mechanized felling may reduce accident rates by a factor 4 (Bell 2002), and therefore replacing manual felling with mechanized felling is a strategic ethical requirement, not just a financial goal. Furthermore, mechanization is the only solution for the continued management of forest areas, in the face of a declining availability of qualified forest workers (Tsioras 2012).

Such crucial issues must be solved, if coppice management has to be rescued from its slow decline. In the absence of new public grants for cautious coppice management, the alternative is often no management at all. However, coppice is one of the few silvicultural models that depend on active management: there are no widespread natural ecosystems that are based on coppice regeneration. Thus, the end of management would be the end of coppice at all. That would be sadly ironic, since the moment is most favorable for a revival of coppice management. Coppice forests may be entering a new season, where they are reinstated to their important economical role because they are present, productive and efficient. However, coppice forest will enjoy the benefits of the modern bio-economy only if coppice management is modernized. For this reason, it is important to facilitate the transition of coppice management from a part-time rural activity to a modern industrial business. Mechanization is the obvious solution, because it compensates for the reduced availability of rural labor, with their propensity to perform heavy and low-paying jobs. For this reason, a compromise must be found between ideal practice and the operational limits of mechanization. Much progress has already been made, but the introduction of mechanized operations still encounters great resistance. That might be mitigated by a better knowledge about the effects of mechanized harvesting on coppice forests, which can only derive from dedicated research. Similarly, research may help developing new low-impact technology solutions, when these are needed.

8. Conclusions: a new season for coppice Coppice management is applied to many species, in many countries and in many ways, so that it may be difficult to describe a single example epitomizing the typical coppice forest and its management. And yet, all coppice stands present two common elements that have a strong impact on operational choices, namely: small stem size and stump crowding. Therefore, all the many solutions devised for coppice harvesting will reflect a variety of local conditions, but they will invariably contain some measures to cope with such common elements.

Small stem size affects the type of products that can be obtained from coppice stands, while limiting work productivity. At the same time, small stem size may favor mechanization and multi-tree handling, which are the main strategies to push down harvesting cost when low-wage labor is no longer available. In such event, stem crowding represents a major technical obstacle, because it hinders mechanized felling and may result in excessive cut height. New small-size disc saws are appearing, which may contribute to solving this problem.

If mechanization is the goal, then silviculture should be adapted to favor it whenever possible. All interventions should offer large enough removals (>80 m3 ha-1) and should allow machine access through the opening of roughly rectilinear paths, about 4 m wide.

The systematic application of light selection thinning is a main obstacle to mechanization and it can make

226

Acknowledgements Much of the bibliographic material used in this study has been collected and analyzed within the scope of European COST Action FP1301 Eurocoppice.

Croat. j. for. eng. 38(2017)2

Trends and Perspectives in Coppice Harvesting (219–230)

9. References Adebayo, A., Han, H., Johnson, L., 2007: Productivity and cost of cut-to-length and whole-tree harvesting in a mixedconifer stand. Forest Products Journal 57(6): 59–69.

Albizu, P.M., Tolosana-Esteban, E., Roman-Jordan, E., 2013: Safety and health in forest harvesting operations. Diagnosis and preventive actions. A review. Forestry Systems 22(3): 392–400.

Bajić, V., Danilović, M., 2004: Optimizacija tehnologije prorednih seča i izdaničkim sestojinama bukve na področju crnog vrha. Glasnik šumarskog fakulteta 89: 41–51.

Baldini, S., 1987: Prove di utilizzazioni meccanizzate nelle conversioni. In: AA.VV: 1987 La conversione dei boschi cedui in alto fusto: stato attuale delle ricerche. UNIF, Viterbo, 84 p.

Baldini, S., 1973: Relazione sulla utilizzazione sperimentale di bosco ceduo nella FD di Cecina. Cellulosa e Carta 6: 37–51.

Baldini, S., Spinelli, R., 1990: Miniskidders in Italy. Small Scale Forestry 1: 23–27.

Baldini, S., Spinelli, R., 1989: Utilizzazione di un bosco ceduo matricinato con esbosco effettuato da animali. Monti e Boschi 2(89): 39–43.

Bell, J., 2002: Changes in logging injury rates associated with use of feller-bunchers in West Virginia. Journal of Safety Research 33(4): 462–471.

Bolding, C., Aust, W., 2011: Potential Soil Erosion following Skyline Yarding versus Tracked Skidding on Bladed Skid Trails in the Appalachian Region of Virginia. Southern Journal of Applied Forestry 35(3): 131–135.

Buckley, G.P., 1992: Ecology and management of coppice woodlands. Chapman & Hall, London.

Cabanettes, A., Pagès, L., 1986: Effet des techniques de coupe sur la hauteur des cépées dans un taillis de châtaignier (Castanea sativa Mill.). Canadian Journal of Forest Research 16(6): 1278–1282.

Canga, E., Fanjul, A., Sánchez-García, S., Alonso-Graña, M., Majada, J., 2014: Replacement of steel cable with synthetic rope in mountain logging operations in Castanea sativa Mill. coppice stands. Forest Systems 23(3): 461–469.

Cantiani, P., Spinelli, R., 1996: Conversion to high forest of Turkey oak coppices. Annali dell’Istituto Sperimentale per la Selvicoltura 27: 191–200.

Carette, J., 2003: La mulasserie, ses origines, ses pratiques.

Ethnozootechnie 72: 7–12.

Chakroun, M., Bouvet, A., Ruch, P., Montagny, X., 2016: Performance of two shear heads for harvesting biomass in hardwood stands in France. Biomass and Bioenergy 91: 227–233.

Civitarese, V., Pignatti, G., Verani, S., Sperandio, G., 2006: Planning wood extraction in a forest coppice. Forest@ 3(3): 367–375.

Crist, J., Mattson, J., Winsauer, S., 1983: Effect of severing method and stump height on coppice growth. In: Intensive Plantation Culture: 12 Years Research (Hansen, E.A., ed.), USDA Forest Service Gen. Tech. Rep. NC-91: 58–63.

Croat. j. for. eng. 38(2017)2

R. Spinelli et al.

Coppini, M., Hermanin, L., 2007: Restoration of selective beech coppices: A case study in the Apennines (Italy). Forest Ecology and Management 249(1): 18–27.

Currò, P., Verani, S., 1984: Tempi di lavoro e rendimenti di esbosco in un ceduo di cerro con Timberjack 225. CSAF Quaderni di Ricerca 4, 6 p.

Currò, P., Verani, S., 1986: Prove di concentramento del legname di un ceduo di cerro con due tipi di gru a cavo. CSAF Quaderni di Ricerca 10, 11 p.

Delasaux, M.J., Hartsough, B.R., Spinelli, R., Magagnotti, N., 2009: Small parcel fuel reduction with a low-investment, high-mobility operation. Western Journal of Applied Forestry 24(4): 205–213.

De Souza, D., Gallagher, T., Mitchell, D., McDonald, T., Smidt, M., 2016: Determining the effects of felling method and season of year on the regeneration of short rotation coppice.

International Journal of Forest Engineering 27(1): 53–65.

Ducrey, M., Turrel, M., 1992: Influence of cutting methods and dates on stump sprouting in Holm oak (Quercus ilex L.) coppice. Annales de Sciences Forestières 49(5): 449–464.

Forestry Commission – Forest Research Technical Development, 2011: Tractor Based Mechanised Harvesting in Sweet Chestnut Coppice. Project Report FCPR040, 18 p.

Gallis, C., 2004: Comparative cost estimation for forwarding small sized beech wood with horses and mini-skidder in northern Greece. Forest Products Journal 54(11): 84–90.

Grulois, S., Cassotti, P., Julien, C., Perinot, C., 1996: Productivity of harvesting operations in coppice forests in the Mediterranean region: situation in France. Annali Istituto Sperimentale di Selvicoltura 27: 183–190.

Han, H.S., Renzie, C., 2005: Effect of ground slope, stump diameter and species on stump height for feller-buncher and chainsaw felling. International Journal of Forest Engineering 16(2): 81–88.

Hall, R., Han, H.S., 2006: Improvements in value recovery through low stump heights: Mechanized versus manual felling. Southern Journal of Applied Forestry 21(1): 33–38.

Hédl, R., Kopecky, M., Komàrek, J., 2010: Half a century of succession in a temperate oakwood: from species-rich community to mesic forest. Diversity and Distributions 16(2): 267–276.

Helmisaari, H., Hanssen, K., Jacobson, S., Kukkola, M., Luiro, J., Saarsalmi, A., Tamminen, P., Tveite, B., 2011: Logging residue removal after thinning in Nordic boreal forests: Longterm impact on tree growth. Forest Ecology and Management 261(11): 1919–1927.

Herrick, O.W., 1982: Estimating benefits from whole-tree chipping as a logging innovation in Northern US Forests.

Forest Products Journal 32(11–12): 57–60.

Hytönen, J., 1994: Effect of cutting season, stump height and harvest damage on coppicing and biomass production of willow and birch. Biomass and Bioenergy 6(5): 349–357.

Jansen, P., Kuiper, L., 2004: Double green energy from traditional coppice stands in the Netherlands. Biomass and Bioenergy 26(4): 401–402.

227

R. Spinelli et al.

Trends and Perspectives in Coppice Harvesting (219–230)

Johnson, P.S., 1975: Growth and structural development of red oak sprout clumps. Forest Science 21(4): 413–418.

Motta, R., Berretti, R., Meloni, F., Nosenzo, A., Terzuolo P.G., Vacchiano, G., 2015: Past, present and future of the coppice silvicultural system in the Italian north-west. In: Book of Abstracts from the Conference »Coppice forests: past, present and future«, 9–11 April, Brno, Czech Republic, 74 p.

Kammenga, J., 1983: Whole-tree utilization system for thinning young Douglas-fir. Journal of Forestry 81(4): 220–224.

Kellogg, L., Bettinger, P., Studier, D., 1993: Terminology of ground-based mechanized logging in the pacific northwest.

Research Contribution 1. Forest Research Laboratory, Oregon State University, Corvallis, OR, USA. 12 p.

Koch, P., 1973: Whole-tree utilization of southern pine advanced by developments in mechanical conversion. Forest Products Journal 23(10): 30–33.

Labelle, E., Soucy, M., Cyr, A., Pelletier, G., 2016: Effect of tree form on the productivity of a cut-to-length harvester in a hardwood-dominated stand. Croatian Journal of Forest Engineering 37(1): 175–183.

Lamers, P., Thiffault, E., Paré, D., Junginger, M., 2013: Feedstock specific environmental risk levels related to biomass extraction for energy from boreal and temperate forests. Biomass and Bioenergy 55: 212–226.

Lepper, F., Frere, S., 1988: Trajan’s column. A new edition of Cichorius plates. Introduction, Commentary and Notes. Alan Sutton Publishing, Gloucester, UK. P. 339 p.

Magagnotti, N., Spinelli, R., 2011: Financial and energy cost of low-impact wood extraction in environmentally sensitive areas. Ecological Engineering 37(4): 601–606.

Magagnotti, N., Pari, L., Spinelli, R., 2012: Re-engineering firewood extraction in traditional Mediterranean coppice stands. Ecological Engineering 38(1): 45–50.

Martin, P., Lapeyre, D., Douchet, O., Restoy, G., Guegand, G., 1996: Récolte mécanisée des taillis en bois ronds (Mechanized harvesting of roundwood from coppice forests). AFOCEL Fiche Information-Forêt 540(4): 6 p.

Matula, R., Svátek, M., Kůrová, J., Úradníček, L., Kadavý, J., Kneifl, M., 2012: The sprouting ability of the main tree species in Central European coppices: implications for coppice restoration. European Journal of Forest Research 131(5): 1501– 1511.

MacDonald, J., Powell, G., 1983: Relationships between stump sprouting and parent-tree diameter in sugar maple in the 1st year following clear-cutting. Canadian Journal of Forest Research 13(3): 390–394.

McNeel, J., Czerepinski, F., 1987: Effect of felling head design on shear-related damage on Southern yellow pine. Southern Journal of Applied Forestry 11(1): 3–6.

Mika, A., Keeton, W., 2013: Factors contributing to carbon fluxes from bioenergy harvests in the U.S. Northeast: an analysis using field data. Global Change Biology and Bioenergy 5(3): 290–305.

Mitchell, D., Gallagher, T., 2007: Chipping Whole Trees for Fuel Chips: A Production Study. Southern Journal of Applied Forestry 31(4): 176–180.

Moscatelli, M., Pettenella, D., Spinelli, R., 2007: Produttività e costi della lavorazione meccanizzata dei cedui di castagno in ambiente appenninico. Forest@ 4(1): 51–62.

228

Nicolescu, V., Pyttel, P., Bartlett, D., (Eds.) 2015: Evolution and perspectives of coppice forests in European countries and South Africa. University of Transilvania, Braşov, Romania, 57 p.

Pettenella, D., Secco, L., 2004: L’organizzazione economica delle imprese di utilizzazione boschiva. L’Italia Forestale e Montana 6: 533–546.

Petty, A., Kärhä, K., 2011: Effects of subsidies on the profitability of energy wood production of wood chips from early thinnings in Finland. Forest Policy and Management 13(7): 575– 581.

Picchio, R., Maesano, M., Savelli, S., Marchi, E., 2009: Productivity and energy balance in conversion of a Quercus cerris L. coppice stand into high forest in Central Italy. Croatian Journal of Forest Engineering 30(1): 15–26.

Piegai, F., Quilghini, G., 1993: Esbosco a soma con trattore.

Monti e Boschi 2: 36–44.

Piegai, F., 2005: Tagli di utilizzazione e di avviamento nei cedui quercini (Clearcutting and conversion of oak coppice).

Sherwood – Foreste e Alberi Oggi 117: 5–8.

Piegai, F., 1985: Impiego delle risine in polietilene per l’avvallamento. Monti e Boschi 36(3): 45–47.

Piegai, F., Uzielli, L., Hippoliti, G., 1980: Diradamento geometrico a strisce in un ceduo di cerro: prove comparative fra sei sistemi di lavoro con vari mezzi di esbosco. Cellulosa e carta 31(3): 3–23.

Piskoric, O., 1963: The dynamics of height increment of coppice shoots of Evergreen oak. Sumarski List 87(3–4): 122–133.

Putnam, N., 1983: A comparison of productivity for whole tree, tree length, and log length skyline thinning in 35 year old Douglas-fir stands of western Oregon. Master Thesis.

Oregon State University, Corvallis, OR., 113 p.

Pyttel, P., Fischer, U., Suchomel, C., Gärtner, S., Bauhus, J., 2013: The effect of harvesting on stump mortality and resprouting in aged oak coppice forests. Forest Ecology and Management 289: 18–27.

Roth, E., Hepting, G., 1943: Origin and development of the oak stump sprouts as affecting their likelihood to decay. Journal of Forestry 41(1): 27–36.

Schweier, J., Spinelli, R., Magagnotti, N., Becker, G., 2015: Mechanized coppice harvesting with new small-scale fellerbunchers: results from harvesting trials with newly manufactured felling heads in Italy. Biomass and Bioenergy 72: 85–94.

Shresta, S., Lanford, B., Rummer, R., Dubois, M., 2008: Soil disturbances from horse/mule logging operations coupled with machines in the southern United States. International Journal of Forest Engineering 19(1): 17–23.

Croat. j. for. eng. 38(2017)2

Trends and Perspectives in Coppice Harvesting (219–230)

R. Spinelli et al.

Spinelli, R., Magagnotti, N., 2007: Biomassa dai boschi di neoformazione: casi di studio in Friuli-Venezia Giulia. Sherwood – Foreste e Alberi Oggi 135: 45–49.

Spinelli, R., Owende, P., Ward, S., Tornero, M., 2004: Comparison of short-wood forwarding systems used in Iberia.

Silva Fennica 38(1): 85–94.

Spinelli, R., Cacot, E., Mihelic, M., Nestorovski, L., Mederski, P., Tolosana, E., 2016a: Techniques and productivity of coppice harvesting operations in Europe: a meta analysis of available data. Annals of Forest Science 73(4): 1125–1139.

Spinelli, R., Owende, P., Ward, S., 2002: Productivity and cost of CTL harvesting of Eucalyptus globulus stands using excavator-based harvesters. Forest Products Journal 52(1): 67–77.

Spinelli, R., Magagnotti, N., Aminti, G., De Francesco, F., Lombardini, C., 2016b: The effect of harvesting method on biomass retention and operational efficiency in low-value mountain forests. European Journal of Forest Research 135(4): 755–764.

Spinelli, R., Pari, L., Aminti, G., Magagnotti, N., Giovannelli, A., 2016c: Mortality, re-sprouting vigor and physiology of coppice stumps after mechanized cutting. Annals of Forest Science 74(1): 5.

Spinelli, R., Ebone, A., Gianella, M., 2014a: Biomass production from traditional coppice management in northern Italy.

Biomass and Bioenergy 62: 68–73.

Spinelli, R., Brown, M., Giles, R., Huxtable, D., Laina Relaño, R., Magagnotti, N., 2014b: Harvesting alternatives for mallee agroforestry plantations in Western Australia. Agroforestry Systems 88(3): 479–487.

Spinelli, R., Magagnotti, N., Nati, C., 2010a: Benchmarking the impact of traditional small-scale logging systems used in Mediterranean forestry. Forest Ecology and Management 260(11): 1997–2001.

Spinelli, R., Magagnotti, N., Picchi, G., 2010b: Deploying Mechanized Cut-to-Length Technology in Italy: Fleet Size, Annual Usage, and Costs. International Journal of Forest Engineering 21(2): 23–31.

Spinelli, R., Hartsough, B.R., Magagnotti, N., 2010c: Productivity standards for harvesters and processors in Italy. Forest Products Journal 60(3): 226–235.

Spinelli, R., Magagnotti, N., Nati, C., 2009: Options for the mechanized processing of hardwood trees in Mediterranean forests. International Journal of Forest Engineering 20(1): 39–44.

Spinelli, R., Cuchet, E., Roux, P., 2007: A new feller-buncher for harvesting energy wood: Results from a European test programme. Biomass and Bioenergy 31(4): 205–210.

Croat. j. for. eng. 38(2017)2

Spinelli, R., Baldini, S., 1992: Utilizzazione di un ceduo quercino in stazione pianeggiante (Harvesting oak coppice in flat terrain). Cellulosa e Carta 43(1): 33–41.

Staijc, B., Zlatanov, T., Velichov, I., Dubravac, T., Trajkov, P., 2009: Past and recent coppice forest management in some regions of southeastern Europe. Silva Balcanica 10(1): 9–19.

Stokes, B., Ashmore, C., Rawlins, C., Sirois, D., 1989: Glossary of terms used in timber harvesting and forest engineering. General Technical Report SO-73. USDA, Forest Service, Southern Forest Experimental Station, New Orleans, LA, USA, 33 p.

Suchomel, C., Spinelli, R., Magagnotti, N., 2012: Productivity of processing hardwoods from coppice forests. Croatian Journal of Forest Engineering 33(1): 39–47.

Suchomel, C., Becker, G., Pyttel, P., 2011: Fully Mechanized Harvesting in Aged Oak Coppice Stands. Forest Products Journal 61(4): 290–296.

Tsioras, P., 2012: Status and Job Satisfaction of Greek Forest Workers. Small-scale Forestry 11(1): 1–14.

Urbinati, C., Iorio, G., Agnoloni, S., Garbarino, M., Vitali, A., 2015: Beech forests in Central Apennines: adaptive management and functions in transition. In: Book of Abstracts from the Conference »Coppice forests: past, present and future«, 9–11 April, Brno, Czech Republic, 38 p.

Verani, S., Nati, C., Spinelli, R., Nocentini, L., 2008: Meccanizzazione avanzata in bosco ceduo. Sherwood – Foreste e Alberi Oggi 144: 41–46.

Westbrook, M., Greene, D., Izlar, R., 2007: Utilizing forest biomass by adding a small chipper to a tree-length southern pine harvesting operation. Southern Journal of Applied Forestry 31(4): 165–169.

Zinkevičius, R., Steponavičius, D., Vitunskas, D., Čingas, G., 2012: Comparison of harvester and motor-manual logging in intermediate cuttings of deciduous stands. Turkish Journal of Agriculture and Forestry 36(5): 591–600.

229

R. Spinelli et al.

Trends and Perspectives in Coppice Harvesting (219–230)

Authors’ addresses: Raffaele Spinelli, PhD. * e-mail: spinelli@ivalsa.cnr.it Natascia Magagnotti, PhD. e-mail: magagnotti@ivalsa.cnr.it CNR – IVALSA Via Madonna del Piano 10 I-50019 Sesto Fiorentino (FI) ITALY

Received: September 15, 2016 Accepted: April 25, 2017

230

Janine Schweier, PhD. e-mail: janine.schweier@foresteng.uni-freiburg.de Albert-Ludwigs-Universität Freiburg Werthmannstraße 6 D-79085 Freiburg GERMANY * Corresponding author Croat. j. for. eng. 38(2017)2

Subject review

Timber Harvesting Methods in Eastern European Countries: a Review Tadeusz Moskalik, Stelian Alexandru Borz, Jiri Dvořák, Michal Ferencik, Sotir Glushkov, Peeter Muiste, Andis Lazdiņš, Oleg Styranivsky Abstract The social and economic changes that began over 25 years ago in post-communist Eastern Europe and the countries of the former Soviet Union also affected the forestry sector. Forested areas were privatised in many countries, and timber harvesting, also in state-owned forests, was contracted out to private sector logging companies. An analysis was conducted of the following countries: Belarus, Bulgaria, Czech Republic, Estonia, Latvia, Lithuania, Poland, Romania, Slovakia, and Ukraine. The basic parameters of forestry, like the characteristics of forest resources, the volume of harvested timber and logging processes used, were given for each country. Special attention was paid to the methods of timber harvesting. The main findings of the study are that various methods are used in Eastern Europe depending on site and forest conditions. In some countries, especially the wealthier ones, a dynamic increase in the cut-to-length method is observed, with the use of harvesters and forwarders.

Keywords: forests privatisation, cut-to-length method, tree-length method, timber harvesting costs, work productivity

1. Introduction Forest utilization, including timber harvesting, has been part of human life since the dawn of time.

Throughout this period, however, the ways humans have impacted the forest have been constantly changing. This applies particularly to the period after World War II, when various types of technologies have been introduced on a large scale in forestry to achieve partial or full mechanization of work.

The systemic transformation that followed the collapse of communist regimes in the early 1990s in European countries of the so-called Eastern Bloc affected virtually all areas of the economy, including forestry (Zälïtis 2015).

The changes that shaped the new political and economic systems, as well as the social changes, had their own characteristics and their own pace in each country of Eastern Europe. This was conditioned by many different factors, including geographic location and the degree of dependence on the Soviet Union.

The concept »Eastern European countries« is not completely unequivocal. Generally, it is used to define European countries with common cultural and hisCroat. j. for. eng. 38(2017)2

torical roots. Various classifications are used for this area (according to the United Nations Group of Experts on Geographical Names, United Nations Statistics Division and political classification during the Cold War). Since the 1980s, the concept of Central and Eastern European (CEE) countries has been more commonly used.

All of the countries share a history of having been socialist countries between 1948–1990, when private forests were nationalized or used by the state. More than 20 years ago, these countries began transitioning from communist regimes with centrally planned economies and one-party political systems to democratic rule and market economies. Today, most of them (except Belarus and Ukraine) are full members of the European Union, meeting all its requirements and conforming to its policy developments. These changes in the political system also stimulated new phenomena, which changed the forestry sectors of these countries: restitution of forest land, privatisation of forest industries, formation of a liberalized timber market, an increased level of timber exports, and new models of forest management, i.e. private businesses, logging companies (Sarvašová et al. 2015, Weiss et al. 2011).

231

T. Moskalik et al.

Timber Harvesting Methods in Eastern European Countries: a Review (231–241)

In terms of the forest economy, the breakthrough of the 20th and 21st centuries, especially in Europe, is the attempt to manage forests with continuous and sustainable development. This concept coincides with the contemporary multifunctional forestry model. It should be noted that forest utilization and harvesting, as the main sector of forestry, comprises the most important part of the multifunctional forest economy (Moskalik 2004).

The aim of this study is to present the state of forestry in the countries of Eastern Europe, with particular emphasis on forest utilization, in the context of socio-economic changes that have occurred in the last 25 years. This also reflects the diversity of natural conditions affecting the structure of the timber and the techniques and technologies used for its harvesting.

2. Methods On the basis of their similar history, ten Eastern European countries were selected for the analysis of aspects of forest utilization: Belarus, Bulgaria, the Czech Republic, Estonia, Latvia, Lithuania, Poland, Romania, Slovakia, and Ukraine.

In order to achieve the intended aim, it was necessary to analyse a large amount of data on, among others, the characteristics of forest resources in a given country (its size, indicator of forest cover, ownership structure, species structure), the volume of harvested timber and logging processes used. While obtaining information on general indicators of individual countries is relatively easy (though the indicators provided

by FAOSTAT, EUROSTAT and Statistical Yearbooks are often inconsistent), it is difficult to obtain current data on the technologies used in logging. These data are often not available in the literature, or dated, due to the rapid pace of changes primarily caused by the mechanization of processes. For this reason, the survey method addressed to experts, mainly the authors of this paper, on the forest utilization in a given country was used.

The survey included questions about the degree of mechanization of harvesting and skidding, with special emphasis on providing the number of working harvesters and forwarders, technological processes used and costs of timber harvesting and extraction.

Nine methods of obtaining shortwood timber were distinguished, assortments under 6 m in length, and five methods of obtaining tree-length timber (over 6 m).

These methods differ in the degree of mechanization of work and distance of skid trails. Respondents were asked to indicate the extent to which methods were used in their country: very often, often, rarely or never.

232

The exact characteristics of the processes are presented in Table 3. The descriptions in the table also includ explanations of the abbreviations used for the machine systems listed. A short analysis of the state of timber harvesting for each individual country was also made.

3. Results 3.1 Characteristics of forest resources The forest resources of Eastern European countries are relatively diverse in terms of size. As shown in Table 1, the largest areas of forest among those analysed are found in three countries: Ukraine, Poland, and Belarus. Significantly smaller areas, at a level of 1.9–2.5 million ha, are found in Slovakia, the Czech Republic and the Baltic countries (Adamczyk at al.

2015, Ambrušová at al. 2015, Jarský at al. 2015).

An important indicator is the forest cover, showing the proportion of the forested area to the total area of the country. In this respect, the best situation is in Estonia and Latvia (50%). The lowest proportion of the forest cover, 16.7%, is in Ukraine. It should be noted, however, that these resources are very heterogeneous, so the distribution of forests across countries is often uneven. In Ukraine, for example, the forest cover varies from 3.7% in the Zaporozhye region to 51.4% in Transcarpathia (Teder at al. 2015a, 2015b, Vilkriste and Zālīte 2015). Most of the countries have more forest cover than the average in Europe of 32.2% (excluding Russia) (EUROSTAT 2016). A very positive feature of Table 1 Forest area and cover in the studied Eastern European countries

Country

Forest area 1000 ha

Change in Forest area available to Forest cover in forest cover from 1990 to 2015, % supply wood, 2015, % %

Belarus

8633.50

75.0

42.5

10.7

Bulgaria

3774.70

57.9

37.4

16.9

Czech Republic

2597.18

86.3

34.5

1.5

Estonia

2231.95

89.3

52.7

1.3

Latvia

3356.00

93.9

54.0

5.9

Lithuania

2180.00

88.3

34.8

12.3

Poland

9197.90

87.7

29.4

6.2

Romania

6520.00

67.4

29.8

7.2

Slovakia

1941.52

92.0

41.0

0.7

Ukraine

9657.00

54.1

16.7

4.4

Croat. j. for. eng. 38(2017)2

Timber Harvesting Methods in Eastern European Countries: a Review (231â&#x20AC;&#x201C;241)

all countries, in contrast to the global trend, is the new afforestation that has been underway for the last 25 years. The greatest progress in this respect was recorded in Bulgaria, Lithuania, and Belarus (Mizaraite at al.

2015, Stoyanov at al. 2015).

The domination of state forest ownership, state capital goods and a centrally planned economy characterised Eastern European countries until 1990 (Teder et al. 2015a). One of the most important factors influencing the current state of the forestry sector and ownership structure in these countries was the restitution of land rights that were lost during the communist regime. This process started in the 1990s and faced many problems. New, the so-called Âťnon-stateÂŤ owners (a term that includes individual owners, cooperatives, private companies, churches, environmental groups and municipalities) lacked sufficient knowledge about how to manage their forests to achieve financial and ecological sustainability. Properties returned to private individuals were often too small for viable independent management and highly fragmented in their location. New forest owners also lacked financial capital, technological know-how, and the necessary equipment and tools (Kocel 2010).

Fig. 1 presents the structure of forest ownership. It shows that the process of reprivatisation has not yet been completed, as there are still areas of forest whose ownership has not been settled in Estonia, Lithuania, and Slovakia. To date, forests have not been returned

Fig. 1 Forest ownership structure in Eastern European countries under analysis Croat. j. for. eng. 38(2017)2

T. Moskalik et al.

to private ownership in three countries: Belarus, Ukraine, and Poland. According to the forest policies presented, their reprivatisation is not foreseen in the coming years. In the future of course, the policy system can change over and privatisation can be implemented more efficiently (Adamczyk at al. 2015).

A very important feature of Eastern European forests is their economic function as a raw material for the wood products industry. The structure of obtained assortments largely depends on the specific conditions of the stands. Undoubtedly, one of the most important factors in this regard is the share of each tree species. Fig. 2 shows a very large variation among the countries. In the case of Poland and Belarus, pine stands prevail; in the Baltic countries, there is far more spruce, birch or alder. In the southern part of the region, with mostly upland and mountainous forests (Bulgaria, Romania, Slovakia), there are mixed deciduous forests dominated by beech trees and various oak species (Nichiforel at al. 2015).

In the past, the region was dominated by native forests. However, due to intensive logging in the late 19th and 20th century, native forests were replaced with spruce monocultures in some countries. Today these forests are often destroyed by the bark beetle.

3.2 Roundwood production and trade Wood is harvested within final felling, thinning, sanitary and other cuttings. A limit of timber harvest-

Fig. 2 Tree species structure in Eastern European countries under analysis

233

T. Moskalik et al.

Timber Harvesting Methods in Eastern European Countries: a Review (231–241)

Table 2 Roundwood production and the timber trade in Eastern European countries under analysis (FAO 2016) Production 1000 m3

Production per capita m3/person

Export, % of production

Import, % of production

Belarus

19,550

2.10

11.1

0.1

Bulgaria

5570

0.76

9.7

0.6

Czech Republic

15,476

1.47

31. 9

15.8

Estonia

12,600

9.57

21.9

1.8

Latvia

12,597

6.29

30.4

10.3

7351

2.49

23.2

4.6

Poland

40,565

1.05

6.6

6.5

Romania

15,068

0.75

2.2

6.7

Slovakia

9417

1.73

31.1

8.2

Ukraine

18,300

0.43

18.8

0.1

Country

Lithuania

ing within the final felling is provided by allowable cuts, which should be approved taking into account the principles of continuity and sustainability of the use of forest resources.

Table 2 shows the volume of roundwood production in the analysed countries. These values are varied, amounting from 5.5 million m3 (Bulgaria) to just over 40 million m3 in Poland. Of course, the determining factors here are the size of the forest area of the country, the type of forest management and the age and species structure of the stands.

Large differences are observed in production per capita. The largest number of timber per person is harvested in the two Baltic countries – Estonia and Latvia, where the ratio is at the level of 9.6 and 6.3 m3, respectively. Less than 1 m3 per person is obtained in Ukraine, Bulgaria, and Romania.

Countries exporting significant quantities of roundwood are the Czech Republic, Slovakia, and Latvia.

The share of exports, compared to the amount harvested, is just over 30%. The smallest amount of timber is exported from Romania (2.2%). The Czech Republic and Latvia also import wood, in amounts of 15.8% and 10.3%, respectively. There is practically no import of wood in Ukraine and Belarus. This is because of the prices offered on the European wood market. For example, the average amount of the price for pine ranges from 50 €/m3 (Ukraine, Belarus) to 83 €/m3 in Romania (Fordaq 2014). The prices can be much higher, reaching up to 500 €/m3, when most valuable wood is under a special offer (Zastocki at al. 2012, 2015).

234

3.3. Timber harvesting processes and work productivity Different methods of logging are used in the countries of Eastern Europe. Their selection depends on site conditions, silvicultural treatments, species composition, tree sizes, stand density and the economic condition of each country. The degree of mechanization of work also differs. According to Asikainen et al. (2009), the proportion of mechanization varies greatly among European countries. The percentage is close to 100% in the Nordic countries, United Kingdom and Ireland, and notably smaller in Eastern Europe.

Currently, most European countries use two methods of harvesting wood: the tree-length method (TL) and cut-to-length method (CTL). These methods refer to the form in which wood is delivered to the road. In the TL method, trees are felled, delimbed and topped in the cut-over or bucked. In this analysis, the minimum length of timber was 6 m. Delimbing and crosscutting are done at the stump. Trees are mainly skidded to roadside by using skidders or agricultural tractors equipped with winches. In the CTL method, also called the shortwood method, trees are felled, delimbed and bucked to various assortments directly at the stump. Harvesting can be fully mechanized or motor-manual. Off-road transport is usually done by forwarders or agricultural tractors equipped with selfloading trailers.

Table 3 shows the most common timber harvesting methods applied in each analysed country. Certain trends in particular regions are clearly visible. In the Baltic countries, a significant proportion of timber is harvested using the CTL method by harvesters and forwarders. A clear increase in this type of machinery has also taken place in Poland, the Czech Republic, and Belarus. Long wood is still harvested in considerable quantities in all countries but not the Baltic ones.

In Ukraine, tree-length timber is extracted, but processed into lengths of 2–4 m at the landing located directly by haul roads. Horse skidding is used primarily in Bulgaria, Romania and Slovakia. Cable yarding is used in all countries, where the forests are located in mountainous areas.

Important aspects influencing the effectiveness of the processes used are the condition and age of the technical equipment. Malinen at al. (2016) conducted interesting research in this area, which showed that among the machines being used in Europe, the oldest harvesters were in Eastern Europe (8.2 years). The average age of forwarders was 9.9 years. In comparison, the average age of harvesters and forwarders in Nordic countries is 6.5 and 6.1 years, respectively.

Croat. j. for. eng. 38(2017)2

Timber Harvesting Methods in Eastern European Countries: a Review (231–241)

T. Moskalik et al.

C–H–F

Ukraine

Slovakia

Romania

Poland

Lithuania

Latvia

Fully mechanized

Estonia

20

H–F

Transported wood

Bulgaria

Mechanization degree

Belarus

Harvesting method

Average distance between skid rails m

Czech Republic

Table 3 Timber harvesting methods applied in each country under analysis

>20

C–F

20

C–TT

20

H–CC

20

C–H–CC

>20

C–O–H–F

>30

C–OT

20

C–O–F/TT

>20

H–TW/S

20

C–TW/S

>20

C–CC

>40

C–O–TW/S

>20

C–O

>20

Highly mechanized

Short wood <6m

Partially mechanized

Highly mechanized

Long wood >6m

Partially mechanized

Very often

Frequency of usage:

often

rarely

Explanation of abbreviations: H – Harvester; F – Forwarder; C – Chainsaw; O – Horse; OT – Horse with a trailer; S – Skidder; TW – Agricultural tractor with a winch; TT – Agricultural tractor with a trailer; CC – Cable crane

Table 4 Work productivity of timber harvesting in different forest conditions depending on the level of mechanization Work productivity, m3/h Cutting category

Volume of cut trees m3

Technological operations (Felling-delimbing-bucking)

Extraction 300 m

Tree-length method (TL) Chainsaw

Cut-to-length method (CTL) Harvester

Tree-length method (TL) Skidder

Cut-to-length method (CTL) Forwarder

Early thinning

0.06–0.08

0.5–1.1

4.4–5.6

4.2–5.3

4.6–5.4

Late thinning

0.18–0.32

0.7–1.5

8.6–16.1

6.3–14.2

8.8–15.1

Clear cut

0.32–0.57

1.8–4.3

19.8–32.1

13.7–16.3

16.5–17.9

The significantly high capital investment in CTL machines requires high productivity and high annual use of machines. Factors affecting productivity include machine size and properties, equipment configuration, stand and site conditions, forest management Croat. j. for. eng. 38(2017)2

objectives, and operator capabilities. The effects of these factors have been studied widely over the last 25 years (Malinen at al. 2016, Mederski at al. 2016a, Moskalik 2004, Nurminen et al. 2006, Oikari et al. 2010, Stankić at al. 2012).

235

T. Moskalik et al.

Timber Harvesting Methods in Eastern European Countries: a Review (231–241)

Table 4 presents the average work productivity of various operational logging processes, depending on cutting category, method applied, and chosen machine. The productivity can, of course, increase or decrease considerably as a result of a slight change in any of these factors.

Individual data on timber harvesting for each analysed country is presented below.

3.3.1 Belarus Tree-length (TL) and cut-to-length (CTL) methods are applied in Belarus to harvest wood. These methods differ in relation to the technology used. The typical TL system employs chainsaws for felling and delimbing, and a cable skidder for extraction; the typical CTL system employs chainsaws for felling, delimbing and cross-cutting, and a forwarder for extraction. The fully mechanized »harvester and forwarder« CTL system is becoming a common practice in Belarus (Gerasimov and Karjalainen 2010).

In the last 10–15 years, the Belarus Ministry of Forestry has upgraded its enterprises, moving in the direction of the mechanized harvesting (Baginsky 2015, Fedorenczik at al. 2013). The forestry enterprises of Belarus have about 160 harvesters, and in 2016, another 72 Vimek harvesters and 52 forwarders for thinning will be supplied (BelTA 2016).

In 2014, the amount of timber harvested by machines was 41%. The prognosis is that this number will increase to 80–85% by 2030. It is also planned that by this year, about 30% of harvested timber will be cut by external companies. The cost of wood harvesting using the government’s resources is 2.7–3.45 €/m³, while this cost is 10% lower with external contractors.

3.3.2 Bulgaria Bulgaria’s forest areas are divided among six stateowned logging companies that hire private firms to perform the harvesting. Cutting and delimbing are done with chain saws. There are only single harvesters, forwarders and cable cranes (about 10 skylines are still in use). There is no accurate statistics about wood extraction, but about 60% of timber is extracted mainly by horses and mules; oxen are rarely used; 40% of the wood is extracted by skidders, agricultural tractors and military trucks. The average logging cost paid by the state forest enterprises to private logging companies, which deliver wood to the roadside was 15.10 €/m³ at the end of 2015.

3.3.3 Czech Republic The tree-length method dominates in the Czech Republic (71%), with the use of chainsaws for felling

236

and delimbing, as well as skidders and agricultural tractors fitted with winches. National cable yarding system »Larix« is used in mountainous areas. About 120,000 m3 of timber is extracted in this way. 29% is harvested in a fully mechanized manner with the use of harvesters and forwarders. It is estimated that there are about 500 harvesters and 850 forwarders (MZe 2015). The cost of felling and extracting timber is 17.1 and 8.3 €/m³, respectively.

Municipal forests and private owners take an individual approach to logging operations. They either do it by themselves or through outsourced services – depending on the economic efficiency.

3.3.4 Estonia Wood is harvested by private companies in Estonia. The distinctive feature of the 1990s was the rapid growth of harvesting volumes and the transition from the tree-length method to the cut-to-length method.

The share of mechanized harvesting also started to grow. In 1995 there were only approx. 20 modern forwarders and 10 harvesters in Estonian forests (Muiste at al. 2006). Today, mechanized harvesting dominates.

Over 95% of final fellings are made by harvesters (up to 100% in state forests), as are over 80% of thinning operations.

The estimated total number of machines of cut-tolength technology is: 220–250 harvesters, 300–350 forwarders and 1000 agricultural tractors with self-loading trailers.

3.3.5 Latvia Timber is harvested by private companies. Several thousand companies declared forest operations as one of their business activities; however, most of the felling operations are carried out by less than 50 companies.

There is no accurate statistics about harvesting methods, but it is estimated that about 70% of wood is cut by using the fully mechanized CTL method; 30% (mostly for thinning and low valued deciduous stands) is cut by chainsaws. The distribution and number of forest machines (estimated) is: 312 harvesters, 1024 forwarders (some of them are agriculture tractors with trailers, which cannot be distinguished in the statistics), and 281 skidders.

The average cost in 2015 was 16.54 €/m³ for final felling, including harvesting, off-road transport and delivery to customer (CSB 2015). For specific operations, the costs were 5.70, 4.94, and 5.90 €/m³ respectively; the average cost of thinning was 21.6 €/m³ (9.39, 6.14, and 6.07 €/m³, respectively). Harvesting costs are mostly affected by the type of felling (thinning or final Croat. j. for. eng. 38(2017)2

Timber Harvesting Methods in Eastern European Countries: a Review (231–241)

felling) and forwarding conditions (soil bearing capacity and forwarding distance). Species do not have a significant impact on harvesting costs.

3.3.6 Lithuania Private logging companies dominate in Lithuanian forests. Contractors harvested 90% of timber procured in state forest enterprises. Chain saws, forwarders or agricultural tractors with self-loading trailers are used in the highly mechanized harvesting of the CTL method in Lithuania. Recently, timber harvesting by chainsaws is being replaced with harvesters (Mizaras at al.

2008).

In accordance with the data provided by the Register of Tractors, self-propelled agricultural machines and their trailers, 47 harvesters were registered in Lithuania in 2008 (Steponavičius and Zinkevičius 2010). It is estimated that nowadays there are about 90 harvesters and 170 forwarders operating in Lithuanian forests.

3.3.7 Poland Despite the fact that most forests belong to the state, a private forestry services sector has developed in Poland since the beginning of the 1990s. At this time, about 98% of the work associated with harvesting and extracting timber is carried out by such firms. The treelength method continues to dominate in mature stands, with the use of chain saws for cutting trees and skidders or agricultural tractors for wood extraction.

The number of operational skidders is estimated to about 1500 machines. In younger stands, the CTL method is partially used, as well as agricultural tractors with self-loading trailers.

In 2004, there were about 15 harvesters operating in Poland. The level of mechanization in forestry was then relatively low (Moskalik 2004). In recent years, rapid changes related to forest operations have been observed in Poland. A growing number of harvesters, an increased volume of harvested timber and a larger proportion of broadleaved species are considered the most important. There were 368 harvesters reported in the survey in early 2014, although at the end of 2015, this number rose to 530 machines (Mederski et al.

2016b). With such a number of machines, the level of mechanization with the harvester-forwarder system should be estimated to about 20%.

On average, last year the cost of wood harvesting was 6.5 €/m³ for cutting-delimbing-bucking and 5.65 €/m³ for extraction. Special tender auctions are organized each year in every forest district and there is one cost of logging, which is not dependent on the cutting category, tree species or extraction distance.

Croat. j. for. eng. 38(2017)2

T. Moskalik et al.

3.3.8 Romania Romanian timber harvesting operations still rely heavily on the use of motor-manual tree felling and processing followed by skidding, carried out mostly

by Romanian-produced wheeled winch skidders (Borz 2015, Sbera 2007, Sbera 2012). About 98% of the work is done with chainsaws and only 2% by fellerbunchers and harvesters. In 2012, Romania had about 35 harvesters and forwarders (Sbera 2012), There are some reasons for this particular situation.

While the Romanian Forest Code states that preference should be given to cable yarding in mountainous and hilly forests to protect the soil, there is a lack of cable yarding operators at the national level (Oprea 2008). The transition to a market economy left the Romanian timber harvesting industry with a serious lack of qualified personnel for cable yarding operations.

The number of existing cable yarders was estimated to about 135 in 2012 (Sbera 2012).

In 96% of cases, wood extraction is done using Romanian and foreign tractors, including forwarders, and only in 4% of cases by other equipment, such as cable yarders.

The most often used harvesting methods are the tree-length and cut-to-length methods, with intermediary adaptations depending on the equipment used and operational conditions. The tree-length method is usually implemented in mountainous and hilly regions when extracting the wood by skidders. The cutto-length method is used in lowland forest areas when procuring firewood directly from the stump. It is also used in mountainous forests in the process of aligning the cable yarding capabilities to the size of extracted wood or to the spatial limitations of the cable yarding corridors (Borz at al. 2015). The full tree method is forbidden by law (MO 2011) in order to limit the environmental impact of logging.

3.3.9 Slovakia In Slovakia, wood is mostly harvested by subcontractors. There is only one state-owned enterprise, which harvests and transports timber – Forestry Mechanization of the Forests of the Slovak Republic, a state enterprise. This enterprise harvests about 5% of the annual cut. It owns 3 harvester-forwarder units, and 20 cable cranes.

The exact number of mechanized units used in the harvesting process and their exploitation in the annual harvest of timber are not known today because of the lack of data. Chainsaw work is still most commonly used in felling operations – about 95% of annual felling is done this way. The rest is performed by the three state-owned harvesters and private ones (the

237

T. Moskalik et al.

Timber Harvesting Methods in Eastern European Countries: a Review (231–241)

actual number is unknown, but most probably it is less than 10).

The estimated percentages of individual mechanized means of extraction, based on the latest data available from 2006 (Green Report 2007) were: horses (8%), cable cranes (8%), agricultural tractors with winches (37%), skidders (45%), and forwarders (2%).

The average costs of individual operations of timber harvesting are as follows: felling – 10.79 €/m3, transportation from the stand to the forest landing (mainly by skidding) – 7.89 €/m3, log bucking – 2.52 €/m3.

3.3.10 Ukraine To perform certain types of work, including logging, state forest enterprises use their own workers or private contractors that have the appropriate licence.

In 2015, about 83% of logging operations were performed with own technical equipment.

Practically, all harvested timber is obtained using the chainsaw. Only a few harvesters are encountered (9 machines in 2015). Two methods of harvesting dominate: in lowlands – chainsaw and agricultural tractor with a trailer/winch or skidder; in the Carpathians – chainsaw-cable system-skidder). Wood is transported to the customer mainly as logs of 2–4 m in length (82–87%).

4. Discussion One of the primary functions of sustainable forest management, among others, is the broadly understood concept of the forest utilization, which also includes logging. Logging carried out in accordance with the rational planning of silviculture and forestry work, taking into account the protective functions of forests and socio-economic needs, is an activity that helps in forming stable and sustainable ecosystems.

However, this field has been undergoing profound transformation in recent years. In practice, it must strive for profound harmony in reconciling environmental requirements, ergonomics and work safety and the appropriate effectiveness of performed tasks, using specialized techniques and technologies (Paschalis and Moskalik 2000).

In Eastern European countries, at the beginning of the economic transformation, when the private forestry services sector was being established, the equipment acquired was of low technical quality. Enterprises owned mainly chainsaws and small numbers of archetypal forest machines, such as tractors with grapples or winches; they rarely had forwarders. Additionally, some of the companies kept horses, especially for use in mountainous regions.

238

Increasing competition in the forestry services set off a search for new technological solutions to decrease labour costs and increase work efficiency.

Many entrepreneurs invested in specialised machinery for timber harvesting despite growing financial difficulties and other challenges. Some forest service companies were able to obtain subsides from EU funds (Kocel 2010).

With each year, also in the analysed region, the level of mechanization in logging is increasing, but the variation between countries is great due to the availability of personnel, fear of unemployment and the rate of investment capability. The rate of investment for mechanization also depends on the regional availability of skilled workers. The complexity of operating high-end forest machines demands a long training period for operators before the person-machine unit can reach its full productivity (Asikainen et al. 2011).

In terms of this issue, there is very much to be done in Eastern European countries.

Mechanized forest logging processes using harvesters and forwarders have vastly increased in some parts of the region, especially in the Baltic countries.

60% of Eastern European harvesters are mainly small and medium class machines. In terms of forwarders, the most popular size class is final felling forwarders (Malinen at al. 2016). In the remaining countries, chainsaws and agricultural tractors with trailers are still used to a great extent to harvest shortwood.

Introducing modern technologies is linked to the need to gain access to tree stands (Stereńczak and Moskalik 2015, Pentek at al. 2008). However, the area of strip roads should not exceed 20% of the stand area.

This limitation mainly affects thinning operations and generally no regulations exist for the final felling. Considering the average width of a strip road (4 m), the standard distance between strip roads is 20 m.

5. Conclusions The social and economic changes, which began in the countries of Eastern Europe in the early 1990s, also led to the restitution or privatisation of forests. Of the countries under analysis, only Belarus, Ukraine and Poland have not introduced such changes. It should be noted that, except for Belarus and Ukraine, the analysed countries are currently members of the European Union.

All of the analysed countries have public companies that manage national forests. These companies entrust most of the field work of harvesting and extraction to private contractors. This work is carried out Croat. j. for. eng. 38(2017)2

Timber Harvesting Methods in Eastern European Countries: a Review (231–241)

by the state only in Ukraine. The owners of private forests harvest timber primarily with their own equipment.

In recent years, we have seen a dynamic growth in the use of the cut-to-length method using a harvester and forwarder to obtain wood. This applies especially to the Baltic countries. The leader in this respect is Estonia, where over 95% of final felling and over 80% of thinning operations are performed by harvesters. The lowest level of mechanization of logging processes is seen in Bulgaria, Romania, Slovakia and Ukraine.

The tree-length method is still dominant in all countries of Eastern Europe, with the exception of the Baltic countries. This method is based, in most cases, on using chainsaws and skidders or agricultural tractors with self-loading trailers. In mountainous regions, cable yarding systems are also used.

While the total productivity of the work methods used to obtain wood is comparable to the results obtained in other EU countries, unit costs, particularly with the less mechanized technologies, are at a lower level. This is mainly due to the availability of relatively cheaper labour.

6. References Adamczyk, W., Jodłowski, K., Socha, J., 2015: Forest land ownership change in Poland. COST Action FP1201 FACESMAP Country Report. European Forest Institute CentralEast and South-East European Regional Office, Vienna, 27 p. Accessed November 20, 2016. Available at: http://facesmap.boku.ac.at/library/countryreports Ambrušová, L., Dobšinská, Z., Sarvašová, Z., Hricová, Z., Šálka, J., 2015: Forest land ownership change in Slovakia.

COST Action FP1201 FACESMAP Country Report. European Forest Institute Central-East and South-East European Regional Office, Vienna, 35 p. Accessed November 20, 2016.

Available at: http://facesmap.boku.ac.at/library/countryreports Asikainen, A., Leskinen, L., Pasanen, K., Väätäinen, K., Anttila, P., Tahvanainen, T., 2009: The current state and the future of the forest machinery sector (in Finnish). Working Papers of the Finnish Forest Research Institute 125, 48 p.

Asikainen, A., Anttila, P., Verkerk, H., Diaz, O., Röser, D., 2011: Development of forest machinery and labour in the EU in 2010-2030. Austro/FORMEC: Pushing the boundaries with research and innovation in forest engineering. Graz, Austria. Accessed October 12, 2016. Available at: http://formec.boku.ac.at/images/proceedings/2011/formec2011_paper_asikainen_etal.pdf Baginsky, V.F., 2015: Problems and prospects of the organization and felling in Belarusian forests. The Proceedings of the St. Petersburg Forestry Research Institute 12(3): 44–54.

Croat. j. for. eng. 38(2017)2

T. Moskalik et al.

BelTa, 2016: Swedish company to sell timber harvesters to Belarus. Belarusian Telegraph Agency. Accessed November 12, 2016. Available at: http://eng.belta.by/economics/view/ swedish-company-to-sell-timber-harvesters-to-belarus-91749-2016/ Borz, S.A., 2015: A review of the Romanian and international practices in skidding operations. XIV World Forestry Congress, Durban, South Africa, 7–11 September. Accessed October 18, 2016. Available at: http://foris.fao.org/wfc2015/api/ file/552ba2919e00c2f116f8e1fc/contents/5129927e-582d-45578194-c5e47d5225aa.pdf Borz, S.A., Ignea, G., Popa, B., Spârchez, G., Iordache, E., 2015: Estimating time consumption and productivity of roundwood skidding in group shelterwood system – a case study in a broadleaved mixed stand located in reduced accessibility conditions. Croatian Journal of Forest Engineering 36(1): 137–146.

CSB, 2015: Central Statistical Bureau of Latvia.

EUROSTAT, 2016: Forestry statistics. Accessed December 8, 2016. Available at: http://ec.europa.eu/eurostat/statisticsexplained/index.php/Forestry_statistics FAO, 2016: Forestry Statistics. Accessed November 18, 2016.

Available at: http://www.fao.org/forestry/46203/en/ Fedorenczik, A.S., German, A.A., Protas, P.A., 2013: Forest machines AMKODOR. Belarusian State Technological University. Minsk, 239 p.

Fordaq, 2014: Accessed November 18, 2016. Available at: http://www.ihb.de/wood/news/Roundwood_prices_Central_Eastern_Europe_39821 Gerasimov, Y., Karjalainen T., 2010: Atlas of the forest sector in Belarus. Working Papers of the Finnish Forest Research Institute 170. Accessed November 20, 2016. Available at: http://www.metla.fi/julkaisut/workingpapers/2010/ mwp170.htm Green Report, 2007: Forestry sector. Accessed November 21, 2016. Available at: http://www.mpsr.sk/en/index. php?navID=17&id=18 Jarský, V., Hrib, M., Riedl, M., Dudík, D., Ventrubová, K., Šišák, L., 2015: Forest land ownership change in Czech Republic. COST Action FP1201 FACESMAP Country Report, European Forest Institute Central-East and South-East European Regional Office, Vienna, 48 p. Accessed December 2, 2016. Available at: http://facesmap.boku.ac.at/library/countryreports Kocel, J., 2010: Development of the forestry services sector in Poland. Folia Forestalia Polonica 52(1): 44–53.

Malinen, J., Laitila, J., Väätäinen, K., Viitamäki, K., 2016: Variation in age, annual usage and resale price of cut-tolength machinery in different regions of Europe. International Journal of Forest Engineering 27(2): 95–102.

Mederski, P.S., Bembenek, M., Karaszewski, Z., Łącka, A., Szczepańska-Álvarez, A., Rosińska, M., 2016a: Estimating and modelling harvester productivity in pine stands of different ages, densities and thinning intensities. Croatian Journal of Forest Engineering 37(1): 27–36.

Mederski, P.S., Karaszewski, Z., Rosińska, M., Bembenek, M., 2016b: Dynamics of harvester fleet change in Poland and

239

T. Moskalik et al.

Timber Harvesting Methods in Eastern European Countries: a Review (231–241)

factors determining machine occurrence (in Polish). Sylwan 160(10): 795–804.

Mizaraite, D., Mizaras, S., 2015: Forest land ownership change in Lithuania. COST Action FP1201 FACESMAP Country Report, European Forest Institute Central-East and South-East European Regional Office, Vienna, 35 p. Accessed November 20, 2016. Available at: http://facesmap. boku.ac.at/library/countryreports Mizaras, S., Sadauskiené, L., Mizaraité, D., 2008: Productivity of harvesting machines and costs of mechanized wood harvesting: Lithuanian case study. Baltic Forestry 14(2): 155–162.

MO, 2011: Order No. 1540 for the approval of instructions regarding the temporal extent, periods and modes of timber harvesting and transportation, 3rd of June. Ministry of Environment and Forests.

Moskalik, T., 2004: Model of fully mechanized timber harvesting in sustainable Polish forestry (in Polish). Wydawnictwo SGGW Warszawa, 134 p.

Muiste, P., Kurvits, V., Mitt, R., Teder, M., Kakko, T., 2006: Forest harvesting in Estonia during the transition period.

Metsanduslikud uurimused (Forestry studies) 45: 164–171.

MZe, 2015: Report on Forest Management of the Czech Republic in 2014 (in Czech). Ministerstvo zemědělství České Republiky. Praha, 108 p.

Nichiforel, L., Bouriaud, L., Dragoi, M., Dorondel, S., Măntescu, L., Terpe, H., 2015: Forest land ownership change in Romania. COST Action FP1201 FACESMAP Country Report, European Forest Institute Central-East and South-East European Regional Office, Vienna, 49 p. Accessed November 20, 2016. Available at: http://facesmap.boku.ac.at/library/ countryreports Nurminen, T., Korpunen, H., Uusitalo, J., 2006: Time consumption analysis of the mechanized cut-to-length harvesting system. Silva Fennica 40: 335–363.

Oikari, M., Kärhä, K., Palander, T., Pajuoja, H., Ovaskainen, H., 2010: Analyzing the views of wood harvesting professionals related to the approaches for increasing the cost-efficiency of wood harvesting from young stands. Silva Fennica 44: 481–495.

Oprea, I., 2008: Timber harvesting technology (in Romanian). Transilvania University Press, Braşov, 237 p.

Paschalis, P., Moskalik, T., 2000: Stand and development of timber harvesting technology at the turn of the century (in German). 34. Internationales Symposium Mechanisierung der Waldarbeit – Formec. Wydawnictwo SGGW Warszawa, 207 p.

Pentek, T., Nevečerel, H., Poršinsky, T., Pičman, D., Lepoglavec, K., Potočnik, I., 2008: Methodology for development of secondary forest traffic infrastructure cadastre. Croatian Journal of Forest Engineering 9(1): 75–83.

Sarvašová, Z., Zivojinovic, I., Weiss, G., Dobšinská, Z., Drăgoi, M., János, G., Jarský, V., Mizaraite, D., Põllumäe, P., Šálka, J., Schiberna, E., Šišák, L., Wolfslehner, B., Zälïte, Z., Zälïtis, T., 2015: Forest Owners Associations in the Central

240

and Eastern European Region. Small-scale Forestry 14(2): 217–232.

Sbera, I., 2007: Forest Resources and the market potential in Romania (in Romanian). Meridiane Forestiere 2: 3–7.

Sbera, I., 2012: Adpoting ecological strategies for timber harvesting (in Romanian). Revista Pădurilor 127(4): 24–26.

Stankić, I., Poršinsky, T., Tomašić, Ž., Frntic, M., 2012: Productivity models for operational planning of timber forwarding in Croatia. Croatian Journal of Forest Engineering 33(1): 6–78.

Steponavičius, D., Zinkevičius, R., 2010: The study of the logging methods prevailed in Lithuania and other countries of Central Europe. EJPAU 13(1) #01. Accessed October 20, 2016. Available at: http://www.ejpau.media.pl/volume13/issue1/art-01.html Stereńczak, K., Moskalik, T., 2015: Use of LIDAR-based digital terrain model and single tree segmentation data for optimal forest skid trail network. iForest 8(5): 661–667.

Stoyanov, N., Kitchoukov, E., Stoyanova, M., Sokolovska, M., 2015: Forest land ownership change in Bulgaria. COST Action FP1201 FACESMAP Country Report, European Forest Institute Central-East and South-East European Regional Office, Vienna, 65 p. Accessed November 20, 2016. Available at: http://facesmap.boku.ac.at/library/countryreports Teder, M., Mizaraite, D., Mizaras, S., Nonić, D., Nedelković, J., Sarvašová, Z., Vilkriste, L., Zälïte, Z., Weiss G., 2015a: Structural changes of state forest management organisations in Estonia, Latria, Serbia and Slovakia since 1990. Baltic Forestry 21(2): 326–339.

Teder, M., Põllumäe, P., Korjus, H., 2015b: Forest land ownership change in Estonia. COST Action FP1201 FACESMAP Country Report, European Forest Institute Central-East and South- East European Regional Office, Vienna, 30 p. Accessed November 20, 2016. Available at: http://facesmap. boku.ac.at/library/countryreports Weiss, G., Tykka, S., Nichiforel, L., Dobšinská, Z., Sarvašová, Z., Mizaraite, D., Nedelkovic, J. 2011: Innovation and sustainability in forestry in Central and Eastern Europe: challenges and perspectives (SUSICEE). Final report. Draft June.

Unpublished. European Forest Institute, Joensu, Finland.

Vilkriste, L., Zālīte, Z., 2015: Forest land ownership change in Latvia. COST Action FP1201 FACESMAP Country Report, European Forest Institute Central-East and South-East European Regional Office, Vienna, 54 p. Access November 20, 2016. Available at: http://facesmap.boku.ac.at/library/countryreports Zastocki, D., Dobosz, L., Moskalik, T., Sadowski, J., 2012: Submission sale of valuable wood on the example of the Krosno Regional Directorate of the State Forests. Sylwan 156(7): 483–493.

Zastocki, D., Moskalik, T., Sadowski, J., 2015: Evaluation of submission as a form of sales of supreme quality timber.

Sylwan 159(9): 707–713.

Zälïtis, T., 2015: Forest owners associations in the Central and Eastern European region. Small-scale Forestry 14(2): 217–232.

Croat. j. for. eng. 38(2017)2

Timber Harvesting Methods in Eastern European Countries: a Review (231–241)

T. Moskalik et al.

Authors’ addresses:

Received: January 09, 2017 Accepted: June 07, 2017 Croat. j. for. eng. 38(2017)2

Assoc. prof. Tadeusz Moskalik, PhD. * e-mail: tadeusz.moskalik@wl.sggw.pl Warsaw University of Life Sciences – SGGW Faculty of Forestry Department of Forest Utilization Nowoursynowska 159 02 776 Warsaw POLAND Assoc. prof. Stelian Alexandru Borz, PhD. e-mail: stelian.borz@unitbv.ro Transilvania University of Braşov Faculty of Forestry Department of Forest Engineering Şirul Beethoven No. 1 500 123 Braşov ROMANIA Assoc. prof. Jiri Dvořák, PhD. e-mail: dvorakj@fld.czu.cz Czech University of Life Sciences Prague Faculty of Forestry and Wood Sciences Kamycka 1176 165 21 Prague 6 – Suchdol CZECH REPUBLIC Assist. prof. Michal Ferencik, PhD. e-mail: ferencik@tuzvo.sk Technical University in Zvolen Faculty of Forestry Department of Forest Harvesting, Logistics and Ameliorations T. G. Masaryka 24 960 53 Zvolen SLOVAKIA Assoc. prof. Sotir Glushkov, PhD. e-mail: sotirgluschkov@abv.bg Forest Research Institute Kl. Ohridski Blvd 132 1756 Sofia BULGARIA Prof. Peeter Muiste, PhD. e-mail: Peeter.Muiste@emu.ee Estonian University of Life Sciences Institute of Forestry and Rural Engineering Department of Forest Industry Kreutwaldi 5 51014 Tartu ESTONIA Andis Lazdiņš, PhD., Senior researcher e-mail: andis.lazdins@silava.lv Latvian State Forest Research Institute »Silava« Rigas 111 2169 Salaspils LATVIA Assoc. prof. Oleg Styranivsky, PhD. e-mail: styranivsky@ukr.net Ukrainian National Forestry University Gen. Chuprynka 103 79057 Lviv UKRAINE * Corresponding author

241

Subject review

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review of Technological Innovation Ola Lindroos, Pedro La Hera, Carola Häggström Abstract Timber harvesting operations vary greatly around the world, as do the adaptations of technology to the complex, locally variable conditions. Similarly, technological innovations occur as a response to a large number of different situations. This review examines the three main drivers considered to generate substantial technological change in mechanized timber harvesting: 1) availability of new technology, 2) demand for new products and 3) introduction of new regulations. The main focus is on Nordic cut-to-length harvesting using a harvester and forwarder, partly due to its advanced level of technology and partly due to the authors’ backgrounds. Examining new technology, progress towards increased automation is highlighted with examples of entry-level products that provide computer-assisted motion control and semiautomation. Examples of unmanned machines and other high-level automation are also presented. Innovations in the field of bioenergy harvesting are presented as examples of advances addressing the demand for new products. Thus, illustrations span from harvesting of tree parts other than stemwood, to how such harvesting and transportation can be integrated into the traditional stemwood harvest. The impact of new regulations on technological innovation is demonstrated with advances aimed at reducing soil damage. Examples range from technical solutions for reducing soil pressure, to walking, flying and even climbing machines. Some predictions are given as to when certain advances can be expected to become reality. However, even though the main drivers are likely to change timber harvesting with new products and new rules, they will probably do so through a continued adaptation of technology to local needs.

Keywords: mechanization, automation, technological change, harvester, forwarder, CTL, logging

1. Introduction Timber harvesting operations vary greatly around the world. Current practices adapt to complex, locally variable conditions in, for example, geo-physical conditions (terrain), management regimes, tree properties, climate, ownership structure, industrial infrastructures, labor availability and capacity, and societal rules for acceptable practices. As most harvesting operations are mechanized to some extent, we had three choices to cover the proposed scope for this invited paper: 1) attempting to cover all developments for all kinds of harvesting operations, 2) to focus only on a limited set of harvesting operations, or 3) to find a way to cover the scope in a generalized manner. The first two alternatives would, however, easily end up as a Croat. j. for. eng. 38(2017)2

list of current technological advances and related predictions, such as those found in, for example, Hellström et al. (2009) and Vanclay (2011), and have a rather local or limited scope (e.g. Warkotsch 1990, Gellerstedt and Dahlin 1999, Guimer 1999, Harstela 1999). Lists of technological advances are naturally interesting, but become rapidly obsolete. Therefore, it was considered more relevant to address the mechanisms behind the progress of changes in the technology used for timber harvesting. However, that is obviously a great challenge, particularly in the limited space of an article format. The aim of the following paper is to highlight the driving forces that result in development of logging operations. A simplified framework for technological innovation is used to highlight the major general driving forces, for which

243

O. Lindroos et al.

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

examples of various interesting technical advances are presented. It should, however, be noted that examples have been selected based on their relation to the innovation drivers, and not on the authors’ opinions of their potential (good or bad) viability.

2. Background 2.1 Definitions and limitations In this paper, the focus will be on the technological part of timber harvesting. Technology is a well-used, but ill-defined, term in the sense that it also encompasses the know-how and tools to solve a practical task (e.g. Berry and Taggart 1994). Technological innovation spans across various research fields. Therefore, there is a variety of definitions, approaches and conceptual models depending on different viewpoints (e.g. Porter 1985, Garcia and Calantone 2002, Crossan and Apaydin 2009). Within this paper, we merely skim the surface of this wealth of research on technological innovation, and are aware of the simplifications that come with such an approach. Moreover, we will mainly focus on the technological part of timber harvesting development. Other work has examined labor, environmental and organizational aspects as important areas for improvement and development of mechanized forestry work (Silversides and Sundberg 1988, Heinimann 2007, Vancay 2011, Häggström and Lindroos 2016).

For the sake of clarity, we will use a simplified frame­work to highlight drivers for technological innovation. Nevertheless, there may be several drivers that independently or interactively result in a given innovation (Trott 2008). However, we have no intention to provide a full classification of drivers. Moreover, we will focus on timber harvesting in general, without addressing drivers for a given innovation.

We are aware that this kind of work will always be biased by the perceptions, values and expectations of the authors. Thus, there will be a significant focus on the Nordic CTL harvesting system with a harvester and forwarder, partly due to the fact that the authors are based in Sweden. However, the topic was deliberately chosen, since the Nordic CTL system is the most technologically advanced in the world, and thereby provides many good examples for future progress.

It should be noted that we will mainly use the term »mechanization« for technological innovations in timber harvesting operations.

2.2 Timber harvesting At a high level, logging operations are part of a production system in which raw material is converted

244

Fig. 1 Simplistic conceptual model of the production system of timber harvesting, with a conversion process designed and adapted to the local physical environment and rules into products. Machines and labor are used for the production work, which is carried out whilst being affected by the local environment and complying with defined rules (Fig. 1). Appropriate labor standards have to be met, and the production system has to be profitable, on a macro scale (forest industry) as well as on a micro scale (individual firm). In other words, forest operations should be carried out in a way that is bio-physically effective, economically efficient, individually compatible, environmentally sound and institutionally acceptable (Heinimann 2007). This is common to other production systems, such as agriculture, mining and various kinds of factory-based manufacture. However, forest harvesting is different because, for instance, the work is done outdoors, in rough terrains and in remote areas.

Tree harvesting can be divided into five distinct work elements:  i – accessing/reaching the tree  ii – felling the tree  iii – debranching the tree  iv – cross-cutting the stem/tree  v – transporting the stem/log/tree to a roadside landing.

All five elements have to be carried out to enable delivery of roundwood logs to industry, but in what order and where they are carried out differ greatly. In fact, order and location determines what harvesting method is used. For instance, in cut-to length (CTL) harvesting, elements i – iv are carried out in the forest with the trees being felled, debranched and bucked into saw-logs and pulpwood lengths according to industry demands and quality features, before off-road transportation to the roadside. In full tree harvesting, tree felling is often followed by transportation to the roadside. The equipment used for harvesting defines Croat. j. for. eng. 38(2017)2

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

the harvesting system. Various harvesting systems can be used to carry out a given harvesting method. The harvesting method CTL, for instance, can be achieved using a manual harvesting system (hand tools and animals), as well as using a fully mechanized harvesting system consisting of a harvester and forwarder (»Nordic CTL«).

Although there are many technical and transactional processes required to enable a full harvesting operation (see e.g. Heinimann 2007), our scope will be limited to tree conversion and off-road transportation.

2.3 Mechanization and automation Mechanization here implies the use of machinery to replace human or animal labor. In order to understand the future of mechanized timber harvesting, it is necessary to understand the work involved, the mechanization process, and the drivers for mechanizing the work.

The mechanization process has a long history in forestry. From the 20th century, timber harvesting has progressed from being entirely manual and animalpowered to being fully mechanized and partly automated (Silversides 1997). This progress has been described as having six phases, from hand tools to feedback-controlled machines (Silversides 1997).

However, with recent progress in automation technology, as observed in the fields of robotics, artificial intelligence and control systems, there are reasons to consider additional phases. In engineering, the degree to which a given task is automated is known as the Level of Automation (LOA). LOA serves to explain the ability of an algorithm to carry out a given automatic function, and how much human involvement there is in the process. Although the definition of LOA varies slightly, it is quite similar in most fields involving au-

O. Lindroos et al.

tomation technology, such as robotics, artificial intelligence, and automatic control. Currently, there is a summary of five levels to educate the wider community about the step-wise progression of automation. A comprehensive definition of these levels is presented in IEEE (2000). For the sake of simplicity, we have provided an overview of LOA definitions in Table 1, together with examples related to a field familiar to most people: the automotive industry.

According to the LOA specifications, a chain-saw operator in combination with a farm tractor would be a harvesting system with an automation level of 0. In contrast, a harvester-forwarder combination would have an automation level of 2, if relying on automated functions such as automatic bucking and computerassisted crane control.

The LOA enables comparisons of the state-of-theart of automation in different industries. From the examples in Table 1, it can also be seen that automation level 3 is the highest level of automation currently available in modern engineering. Industries using equipment with an automation level of 3 include the automotive, robotics, and aerospace industries, where systems equipped with advanced artificial intelligence and embedded hardware are able to compete against human skills. While these autonomous systems are good under certain conditions (conditional automation), they are not good at everything, especially tasks such as learning, making maps, easily identifying objects, or other basic human abilities needed to accomplish more advanced operations. Recent developments in the areas of automation involve efforts to improve technology to an automation level above 3, but such developments will take years to reach maturity. Examples of research problems include learning from demonstrations, understanding spoken lan-

Table 1 Definitions and examples of Levels of Automation (LOA) Level

Description

Human involvement

Example

0

Operator only

A human operator carries out all tasks

–

1

Operator assistance

Basic simplified control functions

A human operator carries out all tasks, but receives computer support simplifying some actions.

Some examples include automatic transmission, cruise control, or anti-sliding control

2

Partial automation

Function-specific automation

Vehicles performing automatic self-parking, or automatic braking to avoid collisions

3

Conditional automation Limited self-driving automation

A vehicle trained to drive in a city, but under constant supervision of a person. The ability to reason outside a given set of conditions is limited

4

High automation

Fully automated for a defined use

A vehicle trained to drive on its own, and not requiring supervision from a person, but will request help when a situation not covered in its database arises

5

Driverless

Fully automated for all situations

A vehicle driving on its own, not requiring any supervision, as it is able to make its own decisions and learn from its surroundings

Croat. j. for. eng. 38(2017)2

245

O. Lindroos et al.

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

guage, wireless network communication, and quickly identifying objects in images. Experts believe that an automation level of 4 will be achieved no earlier than the year 2025. Level 5, however, takes us into the more distant future, where autonomous systems may work by themselves in all situations, without any human supervision.

To describe it in words without using LOA nomenclature, a forestry machine must be capable of advanced localization and decision-making to achieve higher levels of automation. For instance, it should be capable of understanding where it is located, and the status and location of its parts. It should also understand the surrounding environment, and how the work objects (trees/stems/logs) are placed within it, their qualitative features, etc. Consequently, it should possess the computing ability to decide how to carry out the work, whether it is harvesting or transporting logs. In other words, an intelligent machine has to possess all the basic human operator abilities through sensing and computing. Before reaching full automation, semiautomated solutions and increased decision support can be expected first (Westerberg 2014, Hellström et al. 2009).

2.4 Technology innovation Technology innovation in forestry has been described as following paradigm shifts (Heinimann 2007) and discontinuous evolution (Samset 1966), analogous to Schumpeter’s (1942) process of »creative destruction«. This can be understood in the context of harvesting operations, locally or over larger areas, progressing and maturing in alternating leaps of evolution. It can also be seen as adaptations to various stimuli that force current operations to become new types of operations. Irrespective of which perspective is taken, it can be concluded that there would be no progress without some kind of driver for change. It is also understood that technological change is inevitable (Schumpeter 1942, Porter 1985), so it is just a matter of when and what drivers cause the change. Even though mechanization is applied within individual firms, in this paper the drivers are mainly addressed generally and at a forest industry level.

Technology innovations in general are also referred to as technology change, technology shift and technology development (e.g. Porter 1985, Tongur and Engwall 2014). In its most simplistic form, the process of technology innovation can be conceptualized as a linear process in which either a novel device or method is offered to the market (technology push) or market needs trigger innovations (market pull). Drivers for the innovation process can be either internal (e.g.

246

available knowledge) or external (market opportunities or imposed regulations) (Crossan and Apaydin 2010). Hence, technology push corresponds well to the internal driver, whereas market pull corresponds to the two external drivers. Naturally, innovation processes are far more complex than described here (Trott 2008, Crossan and Apaydin 2010), but the simple approach is useful for structuring current technological advances. Thus, for the categorization purpose of this selective review, the following main drivers of harvesting mechanization will be used:  availability of new technology (new technology)  new needs of forest-based products (new products)  need for changes in current operations (new rules).

Below, we briefly describe the aspects (»triggers«) of the production system that are considered to trigger innovation processes, as shown in Fig. 1. It is understood that it will not be possible to provide a complete list of all possible triggers here, or their interactions with the main drivers. Thus, also the categorization should be seen as a simplification, for the sake of clarity.

Production costs, labor and technology can be seen mainly as triggers of the main driver of new technology (i.e. to improve current operations), whereas product value applies to the market’s need for new products (i.e. to change operations to (also) produce new products). Last but not least, rules, labor and, to some extent, environment are triggers of the need for new operations (i.e. need to make the same products in a different way).

2.4.1 Production costs and product value The need to decrease costs and/or increase product value is an essential driver of mechanization (e.g. Porter 1985). However, this economic aspect is funded in the economic system of constant growth, with expectations of steadily increasing production costs (e.g. salaries) but without a corresponding increase in product prices. Hence, the economic drivers of mechanization would be less obvious without the growthbased economy. Competitiveness might then be achieved in other ways. However, since there is no apparent viable alternative to a growth-based economy, economic performance can be expected to continue to be a highly influential driver of timber harvesting mechanization.

The value of the products determines the acceptable production costs. Thus, with high-value timber, expensive harvesting systems such as heli-logging are feasible, whereas stands with low-value trees may not even be possible to harvest profitably. Cheaper is natCroat. j. for. eng. 38(2017)2

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

urally better, but high-value products enable larger profit margins and thereby other options for harvesting. Thus, the harvesting of certain products might be dependent on product price, and/or enabled through advances that decrease harvesting costs. Bioenergy harvesting is an example of that, with a profitability that is highly dependent on energy prices.

In financial value creation, there are two distinct results, depending on the factors limiting the production (Sundberg and Silversides 1988). With unlimited forest resources, forest operations are limited by other shortages, for example a shortage of labor, capital or markets. Then the focus is to maximize the profit per production unit (e.g. per machine), and so only harvest the high-value trees, leaving the low-value ones. There is room to expand operations, and development of new machine systems might enable the harvesting of unused forest resource. Historically, various production shortages have vanished, and eventually the forest resource has become the limiting factor. Western Europe is an example of this kind of transition (Sundberg and Silversides 1988). With limited forest resources, the operation revolves around maximizing the profit per area of forest. Measures to increase forest production are implemented, and as much forest as possible is harvested, using all profitable types of trees. With a limited amount to harvest, this implies that harvesting turns into a more-or-less steady state, with a limited opportunity to expand harvesting operations.

2.4.2 Labor Protection of workers from harsh environments is an important trigger in the mechanization process.

Labor-triggered mechanization involves the improvement of the work environment for health, safety and comfort reasons, but can also be economically-motivated since it expands the possible conditions that allow work (Häggström and Lindroos 2016). Thus, heated cabins and artificial light enable operations to take place in cold and dark conditions. Moreover, work from within a machine cab is safer than motormanual felling with only a helmet for protection from falling trees. Indeed, mechanization has been shown to substantially improve work safety in logging operations (Axelsson 1998).

While operators in general, and expert operators in particular, are becoming difficult to recruit (Bernasconi and Schroff 2011, Baker and Greene 2008), a shortage of qualified labor highlights the need for usable, user-friendly machines in the future (Häggström and Lindroos 2016), both to enable a larger part of the available workforce to operate machines and to shorten the lengthy time needed to become proficient in the operation of, for instance, harvesters (Purfürst 2010).

Croat. j. for. eng. 38(2017)2

O. Lindroos et al.

More productive machines, as well as some automation initiatives, are also a means to address the labor shortage, since they can enable a single operator to harvest larger quantities.

It is traditionally understood that machines enable operators to work faster, longer and with more strength. However, there are areas where the operator’s abilities limit the operations. To operate a harvester or a forwarder efficiently involves considerable cognitive work (Häggström et al. 2015), and often over very long work shifts. One example is the complex coordination required to seamlessly issue joystick commands resulting in motions of the crane and vehicle. Precise control of the many crane links and the harvester head usually requires a series of expertly coordinated movements that can prove tiring over time. Hence, computer-based assistance could improve performance and reduce operator strain.

Legislation of labor health and safety is also an important trigger for technological innovations, with vibration and noise reduction laws being typical examples (Andersson 1988).

2.4.3 Technology Technological advances in society present an abundance of possible applications for forestry. However, with limited numbers of machines sold annually (compared to agricultural and construction machines, for example), forest machine manufacturing is a tough business with scarce resources available for product development. Ironically, for a given size of harvest, even fewer new machines will be needed the more productive they are. Nevertheless, there is no shortage of technological advancements within forestry and related fields. With an active forestry industry, the question is not only whether things could be done differently but whether a change would be beneficial regarding costs and other important aspects. In fact, most technological innovations do not result in a change in operations (c.f. Porter 1985).

Put simply, there are two ways to cope with the challenges of low production numbers of forest machines: To produce highly specialized, advanced and very expensive machines (e.g. harvester and forwarder), or general, basic and rather cheap machines (cf.

Gellerstedt and Dahlin 1999). Both are able to do the job, but might differ in usefulness, labor competence, safety, product quality and cost-efficiency under various conditions. To some extent, the choice might be the result of differences in the machine capabilities to handle the specific trees and the industrial requirements (e.g. CTL versus full tree). Furthermore, the choice is also likely to be substantially influenced by

247

O. Lindroos et al.

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

whether the machine production is a limiting factor or not.

Some of the challenges in forest operations correspond to military and space research and development, in terms of developing robust machines capable of navigating rough terrain. However, military and space-oriented research and development receives substantially higher investments. Thus, forest mechanization is more likely to benefit from military and space innovations than the other way round.

2.4.4 Environment Development is largely triggered by the challenging environment machines have to cope with during work. If the forest operations are carried out in an intense, agricultural-like fashion, the demands are somewhat similar to agricultural machines, in terms of the potential to alter the area of operation. On the other hand, in »close-to-nature« forestry, forestry machines should ideally manage to operate in rough environments without changing the environment to needs and without damaging that environment. Thus, technical development aims to construct machines capable of navigating rough, soft and steep terrain, while simultaneously being able to handle the trees they are processing (Billingsley et al. 2008). This requires very robust and, possibly, very advanced machines (cf.

2.4.3). Climate change might change local environmental conditions, which might trigger technological innovation. However, given the multitude of existing machine systems adapted to various local conditions, it might also only trigger a change to other existing technology.

2.4.5 Rules Rules define the framework that dictates how forest operations are permitted to be carried out, and derive from laws, regulations and certification schemes as well as informal rules resulting from areas such as landowner objectives and social values. Such rules vary geographically and over time. Radical changes of rules could force forestry to either adapt, or to shut down.

Thus, machines must be able to cope with the given operating conditions, and also avoid unacceptable effects on the workforce and the environment. Labor health and safety rules have been important triggers for technological innovations in mechanized harvesting (Andersson 1988). With a continuing focus on environmental concerns (Ollikainen 2014), environmental rules are gaining in importance as a trigger for technology innovation, with the aim of better meeting the rules on avoidance of environmental damage caused by mechanized harvesting. Minimizing the

248

damage to soil is probably the most common requirement (Cambi et al. 2015). This is challenging even when leaving the machines out of the picture, since the weight of the harvested trees alone is several hundred tonnes per hectare. To transport such loads on natural soil without causing damage is naturally challenging, encouraging small (i.e. light) loads and careful driving. Economic considerations, on the other hand, call for large loads and high speeds. However, the same considerations also imply the avoidance of soil damage, since driving on soft ground reduces speed and increases fuel consumption. Additionally, a machine that becomes stuck in the mud results in both severe time losses and possible machine damage.

3. Current technological innovations As demonstrated by the many »triggers« listed above, together with those not mentioned, it is naturally difficult to single out one that will be the main source of future developments. This is especially true since there is such variability in forest operations worldwide, with variation in expectation of future developments. However, based on current trends, it is considered that the three specified main drivers, either individually or in combination, are currently responsible for producing significant advances in timber harvesting. Below, we present examples of various technical progress that can be seen as responses to the main drivers.

3.1 New technology – automation The use of the term new technology is debatable when applied to automation, since the interest in automated forest operations developed soon after the first mechanization. Examples of this interest are, for instance, the IUFRO Div. 3 symposium on »Forest Harvesting Mechanization and Automation« in 1974 (Silversides 1974), and a Swedish workshop on »Automation and Remote Controlling of Forest Machinery« in 1983 (Uusijärvi 1985). More than a decade later, ideas to produce fully automated, but supervised, logging systems were described (Hallonborg 1997). More recent publications have summarized the state-of-the art and the possible ways ahead (e.g. Hellström et al. 2009, Parker et al. 2016). Indeed, over time there have been plenty of innovative projects that have attempted to automate forest operations. So far, however, few have successfully reached the market.

3.1.1 From automation level 0 to level 2 Among the forest machines being operated conventionally, Nordic harvesters are the most advanced Croat. j. for. eng. 38(2017)2

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

ones. Nevertheless, a harvester still requires almost complete operator input. For instance, the operator has to control the many crane links and the harvester head precisely using a series of expertly coordinated movements. Computer-assistance is available for bucking, in the form of an automated decision support system that suggests value-maximizing log lengths and assortments. Nonetheless, mechanized harvesting, even with a harvester, can be considered to be at an automation level of 0. However, there are efforts to introduce LOA 1, mainly by providing computer-assistance for motion control. Over recent years, several entry-level products with automation level 1 technology have appeared on the market, such as:  Cranes equipped with motion sensors, providing entry-level products that use improved motion control software (Cranab 2015)  Basic boom-tip control, where the operator receives computer support to carry out expertly coordinated end-effector movements with less effort (John Deere 2013)  Reduced crane vibrations, making the operation of the crane more comfortable (John Deere 2013, La Hera and Ortiz Morales 2015)  Active suspension, improving the ride quality over uneven terrain (Ponsse 2017)  Hydraulic valves equipped with digital electronics, providing entry-level products that use improved software for dynamic motion control of the machine (Mathworks 2016, Danfoss 2015).

Among the examples listed above, the concept of boom-tip control has long-been anticipated. Finally, in 2013, John Deere became the first forestry machine manufacturer to produce smooth and intelligent boom control (SBC&IBC) systems for forwarders. By now, John Deere IBC system has expanded towards harvesters as well. At the same time, Cranab released their »Cranab Intelligent System« (CIS), a system comprising of sensors integrated in the cranes. Simultaneously, different producers of hydraulic valves have released products involving sensors and computers, resulting in a technology known as »intelligent valve«.

This combination of sensors in cranes and intelligent hydraulics provides sufficient technology for more machine manufacturers to develop their own automated crane functions. All these examples are entrylevel solutions, opening the door to automation. Various concepts for automated crane functions have been tested and/or implemented in test beds (Ortiz Morales 2015, Hansson and Servin 2010). For instance, the number of different boom tip-control algorithms that can be implemented on a machine is huge, because Croat. j. for. eng. 38(2017)2

O. Lindroos et al.

these algorithms respond to selectable optimization options such as minimum kinetic energy control, minimum potential energy control, failure recovery, strength optimization and fuel consumption (La Hera 2011, Westerberg 2014, Ortiz Morales 2015).

The five examples listed above show how current developments are starting to consider the hardware requirements and initial software needed for automation. However, transitioning towards this technology will not be easy, because developing software and redesigning all hydraulics and embedded electronics for forestry machines will be challenging, particularly when trying to make a profit in this process. Therefore, entering the world of automation level 1 will be a difficult step, and we expect that it will take the forest industry at least 15 years to complete it. In those years, however, improvements can be expected in control performance, particularly precision boom movements using motion sensors and operator-assistance software. Creating smarter machine movement will rely on libraries containing specific automated functions, many of which have been demonstrated by scientists over the past few years (Ortiz Morales et al. 2014, La Hera and Ortiz Morales 2015). However, the operator will still be an essential part in the correct use of these features, and many difficult movements will still be carried out manually.

At the later stages, operators are expected to take advantage of advanced computer vision systems in this emerging human-machine partnership. This is likely to enable new ways of controlling the machine, with the operator choosing from actions suggested by the computer (Fig. 2). Consequently, expertly coordinated automatic movements will harvest and collect trees, dramatically increasing productivity and reducing operator fatigue. At this point, the industry will have reached automation level 2. The operator will coordinate the tasks of the machines and, by then, technology will enable the possibility of operating machines with many cranes, because cranes will be able to operate autonomously for short periods (Figs. 3 and 4)

Fig. 2 Augmented reality will be used in future machines, presenting the possibility to select trees by, for instance, pointing to their location (Photo courtesy of Luu et al. 2016)

249

O. Lindroos et al.

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

Fig. 3 Technology will enable a single operator to control many cranes simultaneously, because cranes will operate autonomously for short periods (Photo courtesy of Mellberg 2013)

Fig. 5 Automated machines might eventually begin to have bioinspired designs, to improve the efficiency of off-road navigation and reduce soil damage (Photo courtesy of Ludwign 2016)

(Ersson et al. 2013). Nonetheless, most planning tasks will be carried out by people, who will also carry out tasks manually in very difficult situations. For both of these cases, the user interfaces will become simpler, because many unnecessary buttons and joysticks will be replaced by software algorithms. On the other hand, an interface for controlling several cranes will also add complexity.

better movement capabilities, better power sources, and use dynamic motion control, all of which will contribute to the overall energy efficiency. Later still, machines may begin to have bio-inspired designs (Fig. 5), because designs of this kind could improve the efficiency of off-road navigation and reduce soil damage (Winkler et al. 2015).

3.1.2 Automation level 3 and beyond From automation level 2, it will be possible to rethink fundamentally how machines are designed. This might enable further increases in work and fuel efficiency. Surpassing automation level 3 will produce machines that do not necessarily need to be manned (Fig. 4). Thus, designing smaller and lighter machines will become possible. Having machines without an on-board operator will remove the need for comfortable, ergonomic cabs. Therefore, machines will be cheaper to manufacture, and most of the costs will come from the hardware, software, number of cranes, and power source. Machines of this kind will have

This technological progress will enable a re-structuring of timber harvesting operations, because it will present the opportunity to run forest operations with practically no people in the forests. In essence, these technological advances will enable the complete automation of the forest operation. Operators will initially be located in a command center nearby the machines, but eventually they will be moved far away, close to cities. Hence, progress is expected to follow developments in, for instance, the mining industry, and in harbor and airport management.

Past and current forest machine developments have indeed considered many of the scenarios mentioned above. For instance, machines without cabs have been described (Bergqvist et al. 2006, Konrad 2017), machines with efficient (hybrid electric) power sources have been designed (Elforest 2017), and initial ideas for bio-inspired designs were presented two decades ago (Billingsly et al. 2008).

3.1.3 Challenges

Fig. 4 Having machines without people will remove the need for comfortable, ergonomic cabs; Automation might also enable several cranes to operate on the same machine (Photo courtesy of Leijon 2016)

250

The challenges presented in achieving automation levels 0 to 2 relate to integration of sensor technology and development of control systems, to control machine movements efficiently. The research into teleoperated forestry vehicles (Milne et al. 2013, Westerberg and Shiriaev 2013, Bergkvist et al. 2006) and unmanned self-navigating vehicles (Ringdahl et al.

2011, Hellström et al. 2009, Vestlund and Hellström 2006) have highlighted the challenges in making sensors perceive and understand the structure of »natural« forest land. Moreover, development is needed to enCroat. j. for. eng. 38(2017)2

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

able automatic detection of qualitative features of the trees and logs. Such capability will be required to enable automated decisions on which tree to harvest, as well as to enable automated value-optimized bucking.

Automated machines must be proven safe to humans and animals residing near the operating site.

Thus, safety issues might delay implementation. On the other hand, automation can also be prompted by operator safety, as exemplified by the use of tele-operated ground-based machines on steep terrain (Milne et al. 2013). Increased automation might also influence an operator’s capacity to interact with automated operations. With increasing automation, operators receive less on-the-job training in manual procedures, thus reducing their knowledge and, specifically, their skilled expertise. Insufficient operator knowledge and ability to override the automation, when necessary, could lead to significant effects on both safety and productivity (Amalberti and Deblon 1992).

Other problems to be solved before successfully implementing teleoperation are the problems of information presentation and visibility. For instance, the viewing angle and abstraction level have been shown to affect operator performance (Westerberg and Shiriaev 2013). When introducing two cranes, they will be positioned in new ways that might restrict the operator’s line of view. If not carefully designed, this may result in musculoskeletal injury and accidents if the operator has to alter their position to see properly (Eger et al. 2010, Thomas et al. 1994, Hansson 1990).

Moreover, it may have a negative effect on operator performance (Häggström and Lindroos 2016).

3.2 New products – bioenergy harvesting When an operation is expected to produce new products, it may be influenced by the adaptations required to produce the new product. The products from timber harvesting are traditionally roundwood of various lengths, with the production system being able to meet the industrial need for specific dimensions and quality features. However, operations might have to change to meet the requirements of new industries that may be interested in chemical content and not the dimension or structure of the wood (Ollikainen 2014). It is still too early to predict how such new products might influence silviculture and harvesting. Therefore, the focus will be on another »new« product – to use forest biomass for energy. Burning wood is not new but, nevertheless, it has received renewed attention recently (Björheden 2006, Hakkila 2006). The drive to replace fossil fuels introduced a desire to use increased amounts of forest resources for energy. However, the forest industry had no unused Croat. j. for. eng. 38(2017)2

O. Lindroos et al.

surplus to redirect apart from residuals from conventional forest products, which were already being substantially used in energy production. Thus, the focus turned to the use of hitherto unused parts and types of trees, to avoid competition between traditional products (that naturally could be burnt) and bioenergy assortments (Helmisaari et al. 2014). Even though this new feedstock was introduced in response to the oil crises during the 1970s and the expected fiber shortage, it never became part of the product range that conventional mechanized harvesting was adapted for (Björheden 2006). The recently renewed interest in forest-based bioenergy has resulted in substantial recent research. In fact, this bioenergy-oriented effort has most likely formed the majority of forest engineering research in the new millennium, and has contributed to maintaining, or even increasing, the number of people active in forest engineering research.

Some examples of areas investigated are machines and methods for harvesting of stumps (e.g. Spinnelli et al. 2005, Lindroos et al. 2010a, Berg et al. 2012), branches and tops (also known as logging residues or slash) (Wolf et al. 2014) and small trees (Jundén et al.

2013, Bergström and Di Fulvio 2014, Hanzelka et al.

2016). Interest has also increased in biomass production from the border between agricultural land and forestry, in the form of short-rotation woody crops for energy purposes. How such plantations should be harvested has sparked interest in both new use of traditional forest and agricultural machines, as well as the development of new machines (Spinelli et al.

2012b, Ehlert and Pecenka 2013).

The bulkiness of the material, relative to roundwood, is a challenge especially for transportation, since it gives low payloads. Since payment is given per energy unit in the material, and energy content is related to (dry) mass, low payloads are related to low profit per transport round. This has been addressed by various means that have tried to increase payload, mainly by various attempts to densify the material (e.g. Lindroos et al. 2010b, Bergström et al. 2010, Wolfsmayr and Rauch 2014, Wästerlund and Öhlund 2014, Nuutinen and Björheden 2016, Manzone 2016).

In addition to harvesting technologies, it should be mentioned that interest has also been shown in how to process the material into sizes and qualities suitable for combustion (e.g. Spinelli et al. 2012a, Eriksson et al. 2013, Anerud et al. 2016, Nuutinen et al. 2016) as well as in new analytical methods, aiming to find, define and measure the new products (Routa et al. 2015, Fridh et al. 2014, 2017).

An important aspect of research is to integrate the products i.e. how to combine the harvesting of round-

251

O. Lindroos et al.

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

wood and bio-energy assortments (e.g. Berg et al. 2014, Joelsson et al. 2016).

Besides operational aspects, the effects of bioenergy harvesting on site productivity, ecology and climate change have been subject to a good deal of research (e.g. Magnusson 2016, Bouget et al. 2012, Achat et al. 2015, Egnell et al. 2015).

The dependency on energy price and availability of industry residues have resulted in a current downturn in the harvesting of forest biomass in the Nordic countries. Nevertheless, we expect that energy-related and other new products will result in a substantial change in current timber harvesting within the next 10 to 20 years.

3.3 New rules – avoidance of soil damage Although rules differ substantially geographically, the general trends indicate a continuous increase in environmental consideration necessary during timber harvesting, and especially with regards to the avoidance of soil damage.

Over time, there has been plenty of development aimed at minimizing and, ultimately, avoiding negative impact on soil. However, financial as well as practical restrictions have limited the success for most development projects with this as their main driver. The approaches applied can be split into at least three separate groups: those trying to minimize driving by improved planning, reinforcing the soil and altering the machine usage. The first two are covered only briefly here, whereas the latter is addressed in more depth.

3.3.1 Improved planning Soil damage is likely to be reduced if the operator could be guided into making better choices of where to drive, and how often. Decision support systems, such as the development of LOA 1, may play an important role in this process. Algorithms to extract the best (driest) route are under development (Mohtashami et al. 2012, Flisberg et al. 2007), but this is a complex problem if trying to produce an optimum route for the full operation of, for example, forwarding a stand while simultaneously considering both environmental and economic aspects. However, progress in remote sensing as well as sensor technology is rapidly advancing the frontier of inputs to such planning systems (Lideskog et al. 2015, Ågren et al. 2014, Pohjankukka et al. 2016).

3.3.2 Reinforcing the soil Soil damage from machine traffic can be reduced by applying various materials to the soil surface (Cambi et

252

al. 2015). Most commonly, the logging residues from CTL harvesting (branches and tops) are collected onto the strip-road, and the created »brush-mat« increases the carrying capacity of the soil. The thicker the layer, the better the capacity, but with an energy-based demand for logging residues, there is a trade-off between usage for energy and soil damage avoidance. Other materials and structures have also been tested as soil reinforcement. However, they have to be transported to the site, laid out, and possibly also removed. Thus, if found successful in preventing soil damage, they have often been found to be too costly to use, and particularly in comparison to the use of logging residues.

3.3.3 Alternations of machines and mode of transport Machines can be modified in several ways to reduce soil damage, and the various aspects of such damage. One approach is to reduce the pressure of the machine on the soil (Cambi et al. 2015). The bearing capacity varies between soil types, as well as within soil types over time (due to variables such as weather conditions). However, the less pressure applied on soils, the less damage is caused. Since the pressure is the result of the mass distribution on the area in contact with the soil, both those aspects can be altered to achieve pressure reduction. Increased area for wheeled machines can be achieved by, for instance, the use of additional wheels (Ala-Ilomäki 2011) or bogie tracks (Edlund et al. 2013a, 2013b). Tracked machines are another option, and they also tend to be more suitable for working on steep terrain (Visser and Stampfer 2015).

Lower mass can be achieved by the use of small machines. However, small machines and the normally related small loads, result in more journeys for a given volume of product. Thus, there is a trade-off between the load carried on a single journey of the vehicle, and the total load of all journeys required to move all the products harvested (Cambi et al. 2015, Solgi et al. 2016).

The mass of a machine can also be reduced by the use of lightweight materials, to achieve a good relationship between the machine’s laden weight and its load capacity (i.e. a high load index). However, recent developments have produced machines with lower load indexes than before. On the other hand, the heavier, more robust machines are also more durable (Nordfjell et al. 2010). To equip the main machine with a trailer (Lindroos and Wästerlund 2014, Manzone 2015) is an option for increasing the load index and reducing the soil impact.

Another approach is to address the ground-based mode of transport. Here we can distinguish between new ground-based solutions, and those not groundbased. Among the ground-based solutions, there are Croat. j. for. eng. 38(2017)2

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

O. Lindroos et al.

4. Discussion

Fig. 6 The Portalharvester, with its two tripod legs and sliding cab (Photo courtesy of Christian Knobloch) some walking machines designed for timber harvesting, such as the PlusTech Ltd. (now John Deere Ltd.) harvester of the 1990s (Billingsly et al. 2008) and the recent Portalharvester (Fig. 6) (Anon. 2013, Erler 2013).

The benefits of walking machines, compared to wheeled and tracked machines, include the improved negotiation of certain obstacles and terrains, although such machines have limitations in terms of complexity, fuel consumption, etc. (Billingsly et al. 2008). The benefit from a soil damage perspective is that only soil compression points are created and not continuous tracks. Thus, avoidance of tracks prevents the risk of blocking off roots and water from certain areas by walls of compacted soil.

Aerial logging is another option, with several conventional systems available, such as cable yarding (Lindroos and Cavalli 2016) and heli-logging (Bigsby and Ling 2013). Balloons were suggested until the 1970s (Peters 1973), whereas the recent advances in unmanned aerial vehicles (UAVs) suggest usage in forestry for various monitoring purposes (Torresan 2016). However, given the large loads needed to be carried when harvesting or extracting trees, current UAV technology is unlikely to be used for such purposes, at least in the near future.

Tree-based transportation is a solution that lies between ground and aerial transportation. Indeed, the tree-to-tree moving robot developed in New Zealand was inspired by how monkeys move (Parker et al.

2016). As with aerial systems, it would avoid soil damage and would not be influenced by how rough or steep the terrain is. However, as with UAVs, the work that can be carried out by climbing machines is probably limited in relation to harvesting purposes. To develop a tree-to-tree moving machine capable of tree felling might be feasible. However, the weight of logs that could be carried while climbing is probably limited.

Croat. j. for. eng. 38(2017)2

As can be clearly seen by this limited selection of ongoing development related to mechanized harvesting, there is no shortage of innovation. There is also a great variation in innovation focus, which is to be expected since current timber harvesting practices are a complex mixture of adaptation to complex, locally variable conditions. Future development will be influenced by the necessity for local adaptation, and there is no »perfect solution« in sight (besides some very futuristic scenarios as described below). Thus, in this paper, we have not tried to cover the full range of timber harvesting scenarios. Instead, we have attempted to provide an understanding of the drivers of development. We have chosen a rather simplistic approach, and focused on what we perceive are the main drivers that will lead to substantial change in the conversion process (Fig. 1). Naturally, the innovation process is far more complex (see, for instance, Crossan and Apaydin 2010) and, depending on the point of view, there are other ways to categorize the involved drivers. For instance, Guimier (1999) chose to define another set of drivers, some of which are what we have called triggers. We have also chosen to have a very narrow scope, with the focus on machine development, even though we acknowledge the complex network-like structure required to run modern harvesting operations efficiently (e.g. Heinimann 2007). Simple models facilitate understanding of complex systems, but require that the simplifications are duly handled when attempting to turn the understanding into action. The limitations of the study allows for clarity at the expense of coverage. Nevertheless, the chosen scope can be considered useful for highlighting how innovation is the result of various drivers, among which some are responses to external needs to adapt current operations, whereas others originate from the internal requirement to improve operations constantly.

Following this line, new technology constantly emerges, and can be perceived as being pushed into the timber harvesting operations that already work (more or less) as desired. Possible advances might be intriguingly fascinating, but do not necessarily originate from a well-defined need that requires substantial change. UAV development is one such current example: the technology exists and now it is being investigated for possible forestry usage (Torresan 2016). Another example is the recent concern over increasingly stagnating productivity in Nordic harvesting operations that has triggered a new development focus.

However, it is easy to desire increased productivity and profit, but virtually impossible to achieve it without a well-defined idea about a method of doing so.

253

O. Lindroos et al.

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

With a general, but ill-defined, need for efficiency improvements, it is easy to wait impatiently for rapid advances. However, there is no need for a risky search for change if the actual need is small. Thus, it is important to analyze thoroughly whether or not innovations will render actual and important improvements (e.g.

Lindroos et al. 2015). What problems are solved, and what might be created? Otherwise, innovative technical solutions lacking operational viability might be supported (c.f. Lindroos 2012, Ringdahl et al. 2012) at the expense of more relevant development projects, and might even become burdens for the entrepreneurs that start using them.

New technology will slowly but surely change current timber harvesting operations. Until we see substantial advances in automation, however, there will be a limited effect on the conversion process (Fig. 1).

The process will be the same, but with different machines and slightly differently structured operations.

However, with unmanned machines, there will be a substantial change, of a magnitude similar to when powered tools and machines were brought into the process. However, automation will advance slowly in forestry due to the challenges specified above and, also, because it is a response to the general need for improvement of ongoing, functional operations.

As emphasized above, slowly decreasing profit margins have not proven to be a reliable driver for substantial and fast changes to current operations. In contrast, this and other »small« drivers are likely to result in slow change. That is not to say that such development is bad in any way, but it might fail to meet expectations.

With a well-defined problem, as with new products and new rules, the needs for change are more obvious. The old operations should be adapted to accommodate new conditions, to provide the new desired products to make more profit and to meet new rules or close the business down. The challenge is then to find the most appropriate changes.

Energy wood has (again) complemented the product mix of pulpwood and sawn wood, and has thereby substantially contributed to recent efforts in technical development. Other, less conventional, products can be expected in the form of new usage of trees, and in alternative/complementary products such as biochemicals (Ollikainen, 2014). In fact, the CTL system might not prove to be the most efficient one when the number of products starts to increase, due to the logistic challenges of handling numerous (and possibly differing sizes of) products (cf. Harstela 1999). Instead, it might prove more efficient to extract trees to a central point (terminal or log yard), where the various products are created, collected and distributed. If it is

254

relevant to collect small and rather unusual materials, it is likely to be done efficiently at sizeable facilities. It is reasonable to suppose that CTL might not be efficient in such a supply system, but that will depend on the price relationship between traditional and new products, as well as on how well the extraction of new products will fit the current CTL system. As an example of how new products might influence operations, it is noted that in the Nordics, where logging residues are used for energy, the tops and branches have to be collected separately from the logs (i.e. an adaption). With full tree harvesting on the other hand, both logs and logging residues end up by road-side even if just aiming for the logs.

A completely new product may be in the form of eco-system services, in the sense that future forestry is likely to have the responsibility of creating, balancing and maintaining various kinds of such services. The concept is far from operationalized, but it is likely that there will be a trade-off between various eco-system services. It is also likely that new business models will be developed, in order to form eco-system services into a product that is paid for when being produced (or charged for when being consumed). Thus, this might require forest operations to produce other ecosystem services rather than supplying forest biomaterials. The fact that timber harvesting commonly integrates restoration and creation of social and ecological forest environments (e.g. Gustafsson et al. 2012) is an indication of how this might proceed.

Irrespective of the size of jumps in technological advances, those expected over the next few decades will most likely be seen as fine tuning of current timber harvesting operations. However, to stretch this futureoriented prose a little, two truly drastic advancements that would alter operations substantially will be mentioned. To defy gravity and to be able to teleport would alter the laws of physics that currently define and limit timber harvesting operations.

If machines, trees and logs could be handled without the effect of gravity (i.e. to have them fly in new ways), substantial advances in transportation-related work could be expected. Most of such advances are described in the section about avoiding soil damage.

The even more tantalizing step would be the possibility of teleportation. With that, trees could be disintegrated in the forest, teleported to a desired location and materialized into a desired shape. Thus, trees would be the raw material, and the teleportation would be the transportation and possibly also the conversion process. Teleportation would naturally be a paradigm shifter for mankind, in so many more aspects than enabling new timber harvesting operations.

Croat. j. for. eng. 38(2017)2

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

O. Lindroos et al.

Teleportation is not likely to happen for many decades or even centuries, but there have been some intriguing advances, although on a scale substantially smaller than timber (Pirandola et al. 2015).

Berg, S., Bergström, D., Nordfjell, T., 2014: Simulating conventional and integrated stump-and round-wood harvesting systems: a comparison of productivity and costs. International Journal of Forest Engineering 25(2): 138–155.

As with all studies that aim to predict the future, this study has some strengths and many weaknesses.

We can only present our best educated guesses and speculations, from our limited viewpoint. However, although we are aware that there might be a multitude of other ways of viewing the here-and-now and the possible future, we have tried to provide a somewhat general view of the advances in already highly mechanized operations. We have great hopes for the advances over the coming decades. Moreover, we are curious to see what changes will arrive and what will be their driving forces – and how far from our predictions they will be.

Bergström, D., Nordfjell, T., Bergsten, U., 2010: Compression processing and load compression of young Scots pine and birch trees in thinnings for bioenergy. International Journal of Forest Engineering 21(1): 31–39.

5. References Achat, D.L., Deleuze, C., Landmann, G., Pousse, N., Ranger, J., Augusto, L., 2015: Quantifying consequences of removing harvesting residues on forest soils and tree growth– A metaanalysis. Forest Ecology and Management 348: 124–141. Ågren, A.M., Lidberg, W., Strömgren, M., Ogilvie, J., Arp, P.A., 2014: Evaluating digital terrain indices for soil wetness mapping – a Swedish case study. Hydrology and Earth System Sciences 18(9): 3623–3634.

Ala-Ilomäki, J., Högnäs, T., Lamminen, S., Sirén, M., 2011: Equipping a conventional wheeled forwarder for peatland operations. International Journal of Forest Engineering 22(1): 7–13.

Amalberti, R., Deblon, F., 1992: Cognitive Modeling of Fighter Aircraft Process-Control – a Step Towards an Intelligent on-Board Assistance System. International Journal of ManMachine Studies 36(5): 639–671.

Anerud, E., von Hofsten, H., Eliasson, L., 2016: An alternative supply system for stump biomass – coarse grinding combined with sieving of the produced hog fuel. International Journal of Forest Engineering 27(3): 109–114.

Bergkvist, I., Nordén, B., Lundström, H., 2006. Besten med två virkeskurirer – studier av prestation och bränsleförbrukning (The Beast with two shuttles – studies of performance and fuel consumption). Arbetsrapport; no. 616, Skogforsk. Uppsala, Sweden. (In Swedish)

Bergström, D., Di Fulvio, F., 2014: Evaluation of a novel prototype harvester head in early fuel-wood thinnings. International Journal of Forest Engineering 25(2): 156–170.

Bernasconi, A., Schroff, U., 2011: Professions and Training in Forestry. Results of an Inquiry in Europe and northern America. Bern: Federal Office for the Environment. Retrieved from www.umwelt-schweiz.ch/ud-1052-e Berry, M.M., Taggart, J.H., 1994: Managing technology and innovation: a review. R&D Management 24(4): 341–353.

Bigsby, H., Ling, P., 2013: Long-term productivity of helicopter logging in Sarawak. International Journal of Forest Engineering 24(1): 24–30.

Billingsley, J., Visala, A., Dunn, M., 2008: Robotics in agriculture and forestry. In: Springer handbook of robotics, Springer Berlin Heidelberg, 1065–1077 p.

Björheden, R., 2006: Drivers behind the development of forest energy in Sweden. Biomass and Bioenergy 30(4): 289–295.

Bouget, C., Lassauce, A., Jonsell, M., 2012: Effects of fuelwood harvesting on biodiversity – a review focused on the situation in Europe. Canadian Journal of Forest Research 42(8): 1421– 1432.

Cambi, M., Certini, G., Neri, F., Marchi, E., 2015: The impact of heavy traffic on forest soils: a review. Forest Ecology and Management 338: 124–138.

Cranab, 2015: Forwarder cranes with world leading technology. www.cranab.se/site_specific/.../05/cran_fc_2015-eng-lr. pdf. Retrieved 2017-01-24.

Anon., 2013: Vollmechanisiertes Fällen und Aufarbeiten mit Portalharvester; Rücken mit Flachlandseilkran (Fully mechanized harvesting with the portalharvester, transporting with flat-land cable crane). AFZ-DerWald 18/2013: 50–59.

Crossan, M.M., Apaydin, M., 2010: A multi‐dimensional framework of organizational innovation: A systematic review of the literature. Journal of management studies 47(6): 1154– 1191.

Axelsson, S.-Å., 1998: The mechanization of logging operations in Sweden and its effect on occupational safety and health. International Journal of Forest Engineering 9(2): 25– 31.

Danfoss, 2015: Robust and efficient in harsh environments. http://www.danfoss.com/technicalarticles/cf/robust-and-efficient-in-harsh-environments/?ref=17179879857#/ Retrieved 2017-01-24.

Baker, S.A., Greene, D.W., 2008: Changes in Georgia’s Logging Workforce, 1987–2007. Southern Journal of Applied Forestry 32(2): 60–68.

Konrad, 2017: Ground Carriage PULLY. Retrieved from http://www.forsttechnik.at/en/products/ground-carriagepully.html. Retrieved 2017-01-24.

Berg, S., Bergström, D., Athanassiadis, D., Nordfjell, T., 2012: Torque required to twist and cut loose Scots pine stumps.

Scandinavian Journal of Forest Research 27(8): 724–733.

Edlund, J., Keramati, E., Servin, M., 2013a: A long-tracked bogie design for forestry machines on soft and rough terrain.

Journal of terramechanics 50(2): 73–83.

Croat. j. for. eng. 38(2017)2

255

O. Lindroos et al.

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

Edlund, J., Bergsten, U., Arvidsson, H., 2013b: A forest machine bogie with a bearing capacity dependent contact area: acceleration and angular orientation when passing obstacles and drawbar pull force and free rolling resistance on firm ground. Silva Fennica 47(3): article id 1017.

Ehlert, D., Pecenka, R., 2013: Harvesters for short rotation coppice: current status and new solutions. International Journal of Forest Engineering 24(3): 170–182.

Eger, T., Godwin, A.A., Henry, D.J., Grenier, S.G., Callaghan, J., Demerchant, A., 2010: Biomechanics and evaluating the workplace. Why vehicle design matters: exploring the link between line-of-sight, driving posture and risk factors for injury. Work 35(1): 27–37.

Egnell, G., Jurevics, A., Peichl, M., 2015: Negative effects of stem and stump harvest and deep soil cultivation on the soil carbon and nitrogen pools are mitigated by enhanced tree growth. Forest Ecology and Management 338: 57–67.

Elforest, 2017: Elforest Technology, About us. http://elforest. se/eng/home-2.html. Retrieved 2017-01-24.

Eriksson, G., Bergström, D., Nordfjell, T., 2013: The state of the art in woody biomass comminution and sorting in Northern Europe. International Journal of Forest Engineering 24(3): 194–215.

Erler., J., 2013: Portalharvester und Flachlandseilkran (Portalharvester and Flat-land cable crane). Forst & Technik 9: 18–21.

Ersson, B.T., Jundén, L., Bergsten, U., Servin, M., 2013: Simulated productivity of one-and two-armed tree planting machines. Silva Fennica 47(2): article id 958.

Flisberg, P., Forsberg, M., Rönnqvist, M., 2007: Optimization based planning tools for routing of forwarders at harvest areas. Canadian Journal of Forest Research 37(11): 2153–2163.

Fridh, L., Volpé, S., Eliasson, L., 2014: An accurate and fast method for moisture content determination. International Journal of Forest Engineering 25(3): 222–228.

Fridh, L., Volpé, S., Eliasson, L., 2017: A NIR machine for moisture content measurements of forest biomass in frozen and unfrozen conditions. International Journal of Forest Engineering 28(1): 42–46.

Garcia, R., Calantone, R., 2002: A critical look at technological innovation typology and innovativeness terminology: a literature review. Journal of product innovation management 19(2): 110–132.

Gellerstedt, S., Dahlin, B., 1999: Cut-to-length: The next decade. Journal of Forest Engineering 10(2): 17–24.

Guimier, D.Y., 1999: Canadian forestry operations in the next century. Journal of Forest Engineering 10(1): 7–12.

Gustafsson, L., Baker, S.C., Bauhus, J., Beese, W.J., Brodie, A., Kouki, J., Lindenmayer, D.B., Lõhmus, A., Martínez Pastur, G., Messier, C., Neyland, M., Palik, B., Sverdrup-Thygeson, A., Volney, W.J.A., Wayne, A., Franklin, J.F., 2012: Retention forestry to maintain multifunctional forests: a world perspective. BioScience 62(7): 633–645.

Hakkila, P., 2006: Factors driving the development of forest energy in Finland. Biomass and Bioenergy 30(4): 281–288.

256

Hallonborg, U., 1997: Ingen man på maskinen – en förarlös vision (No man in the machine – a driverless vision). Arbetsrapport; no. 399, Skogforsk. Uppsala, Sweden. (In Swedish).

Hansson, A., Servin, M., 2010: Semi-autonomous shared control of large-scale manipulator arms. Control Engineering Practice 18(9): 1069–1076.

Hanzelka, N.C., Bolding, M.C., Sullivan, J., Barrett, S.M., 2016: Productivity and costs of utilizing small-diameter stems in a biomass-only harvest. International Journal of Forest Engineering 27(1): 43–52.

Harstela, P., 1999: The future of timber harvesting in Finland.

Journal of Forest Engineering 10(2): 33–36.

Hellström, T., Lärkeryd, P., Nordfjell, T., Ringdahl, O., 2009: Autonomous Forest Vehicles: Historic, envisioned, and stateof-the-art. International Journal of Forest Engineering 20(1): 31–38.

Helmisaari, H.S., Kaarakka, L., Olsson, B.A., 2014: Increased utilization of different tree parts for energy purposes in the Nordic countries. Scandinavian Journal of Forest Research 29(4): 312–322.

Heinimann, H.R., 2007: Forest operations engineering and management – the ways behind and ahead of a scientific discipline. Croatian journal of forest engineering 28(1): 107–121.

Häggström, C., Englund, M., Lindroos, O., 2015: Examining the gaze behaviors of harvester operators: an eye-tracking study. International Journal of Forest Engineering 26(2): 96– 113.

Häggström, C., Lindroos, O., 2016: Human, technology, organization and environment – a human factors perspective on performance in forest harvesting. International Journal of Forest Engineering 27(2): 67–78.

IEEE, 2000: A Model for Types and Levels of Human Interaction with Automation. IEEE transactions on systems, man, and cybernetics: A publication of the IEEE Systems, Man, and Cybernetics Society. New York, NY: Institute of Electrical and Electronics Engineers 30(3).

Joelsson, J., Di Fulvio, F., De La Fuente, T., Bergström, D., Athanassiadis, D., 2016: Integrated supply of stemwood and residual biomass to forest-based biorefineries. International Journal of Forest Engineering 27(2): 115–138.

John Deere, 2013: Smooth Boom Control / Intelligent Boom Control. http://www.deere.com/en_US/docs/forestry/SBC_ IBC_FastFact_FNL.pdf Retrieved 2017-01-24.

Jundén, L., Bergström, D., Servin, M., Bergsten, U., 2013: Simulation of boom-corridor thinning using a double-crane system and different levels of automation. International Journal of Forest Engineering 24(1): 16–23.

La Hera, P., 2011: Underactuated Mechanical Systems: Contributions to trajectory planning, analysis, and control. Doctoral dissertation thesis. Umeå University, Sweden.

La Hera, P., Ortiz Morales, D., 2015: Model-based development of control systems for forestry cranes. Journal of Control Science and Engineering: 15 p., article ID 256951.

Croat. j. for. eng. 38(2017)2

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258) Leijon, M., 2016: Forestry machines in 2050. Umeå Institute of design. Umeå University. Sweden.

Lideskog, H., Karlberg, M., Bergsten, U., 2015: Development of a Research Vehicle Platform to Improve Productivity and Value-extraction in Forestry. Procedia CIRP 38: 68–73.

Lindroos, O., 2012: Evaluation of technical and organizational approaches for directly loading logs in mechanized CTL harvesting. Forest Science 58(4): 326–341.

Lindroos, O., Wästerlund, I., 2014: Theoretical Potentials of Forwarder Trailers with and without Axle Load Restrictions.

Croatian Journal of Forest Engineering 35(2): 211–219.

Lindroos, O., Cavalli, R., 2016: Cable yarding productivity models: a systematic review over the period 2000–2011. International Journal of Forest Engineering 27(2): 79–94.

Lindroos, O., Henningsson, M., Athanassiadis, D., Nordfjell, T., 2010a: Forces required to vertically uproot tree stumps.

Silva Fennica 44(4): 681–694.

Lindroos, O., Pettersson, M., Johansson, P., Nordfjell, T., 2010b: Productivity of a prototype truck-mounted logging residue bundler and a road-side bundling system. Silva Fennica 44(3): 547–559.

Lindroos, O., Ringdahl, O., La Hera, P., Hohnloser, P., Hellström, T., 2015: Estimating the position of the harvester head – a key step towards the precision forestry of the future?

Croatian Journal of Forest Engineering 36(2): 147–164.

Luu, T., Sabir, E., Sundelin, L., 2016: Forestry in 2020. Operating the Harvester Crane. Umeå Institute of design. Umeå University. Sweden.

Ludwign, 2016: Forestry in 2050. Autonomous Harvester.

Umeå Institute of design. Umeå University. Sweden.

Magnusson, T.P., 2016: Long-term effects of stump harvest on mercury and general chemistry in discharge water. Scandinavian Journal of Forest Research 32(3): 230–238.

Manzone, M., 2015: Performance of an electronic control system for hydraulically driven forestry tandem trailers. Biosystems Engineering 130: 106–110.

Manzone, M., 2016: A bundler prototype for forestry and agricultural residue management for energy production. International Journal of Forest Engineering 27(2): 103–108.

Mathworks, 2016: INCOVA Designs Intelligent Valve-Control System for a 20-Ton Excavator. https://www.mathworks.com/ company/user_stories/incova-designs-intelligent-valve-control-system-for-a-20-ton-excavator.html Retrieved 2017-0124.

O. Lindroos et al.

Nordfjell, T., Björheden, R., Thor, M., Wästerlund, I., 2010: Changes in technical performance, mechanical availability and prices of machines used in forest operations in Sweden from 1985 to 2010. Scandinavian Journal of Forest Research 25(4): 382–389.

Nuutinen, Y., Petty, A., Bergström, D., Rytkönen, M., Di Fulvio, F., Tiihonen, I., Lauren, A., Dahlin, B., 2016: Quality and productivity in comminution of small-diameter tree bundles.

International Journal of Forest Engineering 27(3): 179–187.

Nuutinen, Y., Björheden, R., 2016: Productivity and work processes of small-tree bundler Fixteri FX15a in energy wood harvesting from early pine dominated thinnings. International Journal of Forest Engineering 27(1): 29–42.

Ollikainen, M., 2014: Forestry in bioeconomy – smart green growth for the humankind. Scandinavian Journal of Forest Research 29(4): 360–366.

Ortiz Morales, D., 2015: Virtual Holonomic Constraints: from academic to industrial applications. Doctoral dissertation thesis. Umeå University, Sweden.

Ortiz Morales, D., Westerberg, S., La Hera, P.X., Mettin, U., Freidovich, L., Shiriaev, A.S., 2014: Increasing the Level of Automation in the Forestry Logging Process with Crane Trajectory Planning and Control. Journal of Field Robotics 31(3): 343–363.

Parker, R., Karen, B., Clinton, P.W., 2016: Robotics in forestry.

New Zealand Journal of Forestry 60(4): 8–14.

Peters, P.A., 1973: Advanced Logging System-II: Balloon Logging: A Look at Current Operating Systems. Journal of Forestry 71(9): 577–579.

Pirandola, S., Eisert, J., Weedbrook, C., Furusawa, A., Braunstein, S.L., 2015: Advances in quantum teleportation. Nature photonics 9(10): 641–652.

Pohjankukka, J., Riihimäki, H., Nevalainen, P., Pahikkala, T., Ala-Ilomäki, J., Hyvönen, E., Varjo, J., Heikkonen, J., 2016: Predictability of boreal forest soil bearing capacity by machine learning. Journal of Terramechanics 68: 1–8.

Porter, M.E., 1985: Technology and competitive advantage.

Journal of business strategy 5(3): 60–78.

Ponsse, 2017: Ponsse Active Frame. http://www.ponsse.com/ se/produkter/skotare/activeframe. Retrieved 2017-01-24.

Purfürst, T., 2010: Learning curves of harvester operators.

Croatian Journal of Forest Engineering 31(2): 89–97.

Mellberg, A., 2013: Extension–Operator Environment for Forest Harvesters. Master thesis. School of Engineering, Jönköping, Sweden.

Ringdahl, O., Lindroos, O., Hellström, T., Bergström, D., Athanassiadis, D., Nordfjell, T., 2011: Path tracking in forest terrain by an autonomous forwarder. Scandinavian Journal of Forest Research 26(4): 350–359.

Milne, B., Chen, X., Hann, C., Parker, R., 2013: Robotisation of forestry harvesting in New Zealand – An overview. In 2013 10th IEEE International Conference on Control and Automation (ICCA): 1609–1614.

Ringdahl, O., Hellström, T., Lindroos, O., 2012: Potentials of possible machine systems for directly loading logs in cut-tolength harvesting. Canadian Journal of Forest Research 42(5): 970–985.

Mohtashami, S., Bergkvist, I., Löfgren, B., Berg, S., 2012: A GIS approach to analyzing off-road transportation: a case study in Sweden. Croatian Journal of Forest Engineering 33(2): 275–284.

Routa, J., Kolström, M., Ruotsalainen, J., Sikanen, L., 2015: Precision measurement of forest harvesting residue moisture change and dry matter losses by constant weight monitoring.

International Journal of Forest Engineering 26(1): 71–83.

Croat. j. for. eng. 38(2017)2

257

O. Lindroos et al.

Drivers of Advances in Mechanized Timber Harvesting – a Selective Review ... (243–258)

Samset, I., 1966: Loven om den sprangvise vikling (The law of discontinuous evolution). Norsk Skogbruk 12: 737–741. (In Norwegian) Schumpeter, J., 1942: Capitalism, socialism and democracy.

Harper and Brothers. New York.

Uusijärvi, R., 1985: Automation and remote controlling of forest machinery. Report from NSR workshop in Sweden, April 1983. Wood Technology Report No. 70, Swedish Institute for Wood Technology Research. Stockholm, Sweden. (In Swedish with English summary).

Silversides, C.R. (ed.), 1974: Proceedings of the IUFRO Division 3 symposium »Forest harvesting mechanization and automation. Proceedings«. 25 Sept – 5 October, Ottawa, Thunder Bay and Sault Ste. Marie, Canada. Canadian Forestry Service, Dept. of the Environment. Ottawa, Ontario, Canada.

Warkotsch, P.W., 1990: Timber Harvesting in South Africa – Past, Present, Future. South African Forestry Journal 153(1): 55–63.

Silversides, C.R., Sundberg, U., 1988: Operational Efficiency in Forestry. Volume 1: Analysis. Dordrecht, The Netherlands: Kluwer Academic Publishers.

Silversides, C.R., 1997: Broadaxe to flying shear: the mechanization of forest harvesting east of the Rockies. Transformation Series no. 6, National Museum of Science and Technology. Ottawa, Canada.

Solgi, A., Naghdi, R., Labelle, E.R., Tsioras, P.A., Nikooy, M., 2016: Effect of varying machine ground pressure and traffic frequency on the physical properties of clay loam soils located in mountainous forests. International Journal of Forest Engineering 27(3): 161–168.

Spinelli, R., Nati, C., Magagnotti, N., 2005: Harvesting and transport of root biomass from fast-growing poplar plantations. Silva Fennica 39(4): 539–548.

Spinelli, R., Cavallo, E., Facello, A., 2012a: A new comminution device for high-quality chip production. Fuel processing technology 99: 69–74.

Spinelli, R., Schweier, J., De Francesco, F., 2012b: Harvesting techniques for non-industrial biomass plantations. Biosystems engineering 113(4): 319–324.

Trott, P., 2008: Innovation management and new product development. 4th ed. Pearson Education Ltd, Harlow, England.

Torresan, C., Berton, A., Carotenuto, F., Di Gennaro, S.F., Gioli, B., Matese, A., Miglietta, F., Vagnoli, C., Zaldei, A., Wallace, L., 2016: Forestry applications of UAVs in Europe: a review. International Journal of Remote Sensing 38(8–10): 2427–2447.

Vanclay, J.K., 2011: Future harvest: what might forest harvesting entail 25 years hence? Scandinavian Journal of Forest Research 26(2): 183–186.

Westerberg, S., 2014: Semi-automating forestry machines: motion planning, system integration, and human-machine interaction. Doctoral dissertation thesis. Umeå University, Sweden.

Westerberg, S., Shiriaev, A., 2013: Virtual environment-based teleoperation of forestry machines: Designing future interaction methods. Journal of Human-Robot Interaction 2(3) :84– 110.

Vestlund, K., Hellström, T., 2006: Requirements and system design for a robot performing selective cleaning in young forest stands. Journal of terramechanics 43(4): 505–525.

Winkler, A.W., Mastalli, C., Havoutis, I., Focchi, M., Caldwell, D.G., Semini, C., 2015: Planning and execution of dynamic whole-body locomotion for a hydraulic quadruped on challenging terrain. In: 2015 IEEE International Conference on Robotics and Automation (ICRA), 5148–5154.

Visser, R., Stampfer, K., 2015: Expanding ground-based harvesting onto steep terrain: a review. Croatian Journal of Forest Engineering 36(2): 321–331.

Wolf, D.P., Meek, P., Jones, T.A., Cormier, D., Caspersen, J.P., 2014: Forest harvest residue recovery in semi-mechanized hardwood selection operations. International Journal of Forest Engineering 25(3): 229–237.

Wolfsmayr, U.J., Rauch, P., 2014: The primary forest fuel supply chain: a literature review. Biomass and Bioenergy 60: 203–221.

Wästerlund, I., Öhlund, A., 2014: Productivity Study of WoodPac Bundling of Logging Residues and Small Stems.

Croatian Journal of Forest Engineering 35(2): 139–151.

Authors’ addresses:

Received: January 24, 2017 Accepted: June 9, 2017

258

Prof. Ola Lindroos, PhD. * e-mail: ola.lindroos@slu.se Pedro La Hera, PhD. e-mail: xavier.lahera@slu.se Carola Häggström, PhD. e-mail: carola.haggstrom@slu.se Department of Forest Biomaterials and Technology Swedish University of Agricultural Sciences SE-901 83 Umeå SWEDEN * Corresponding author Croat. j. for. eng. 38(2017)2

Subject review

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland Rolf Björheden Abstract The role of the forest sector in Finland and in Sweden is the starting point for a case study presenting motifs for forest bioenergy in the two countries. Forest bioenergy, evolving in symbiosis with the forest industry, has become important. The successful development builds on piggy-backing conventional forestry, rather than on parallel supply systems.

After thirty years, forest biomass has become the largest energy source in the two countries, contributing almost 1/5 of the energy needs. For developed countries, Sweden and Finland have leading positions in the use of forest fuel, and in related technologies and methods. However, progress has not been simple and drivers for the development have changed over time.

The 1970s »oil crises« put initial focus on energy security and on reducing the dependence on imported fuels. Later, other motifs have become fundamental. Sustainability aspects – especially mitigating climate change – have emerged as key arguments. Fuels from sustainably managed forests cause minor, if any, emissions of carbon dioxide.

The facts that wood-based fuels create rural jobs and improve the trade balance have been ancillary motifs, and the increased net sale of forest products that follows on fuel production will increase the cash flow of the forest owner. However, due to low compensation and high costs compared to the traditional forest products, from the forest owners’ perspective, the economic motifs for forest fuel harvesting are not decisive.

For economic use of biomass, heat sinks are important. Combined heat and power, e.g. for district heating plus electricity to the grid or for industrial process heat and power are profitable options. Further refinement is possible but its potential to increase profitability seems limited.

Keywords: bioenergy, forest biomass, forest harvesting residues, climate change, sustainability

1. Introduction Bioenergy from forest biomass represents one of the oldest ways to utilize forest biomass. It is a prerequisite for man’s existence in the boreal parts of the world (James 1989). Also for the large-scale protoindustrial manufacture and for initial stages of the industrial manufacture, wood played an important role both as a material and as a fuel (Sundberg et al. 1994).

Despite this, for a long time, utilization of forest biomass for energy attracted only minimal attention. It was considered a primitive, dirty, inefficient and even wasteful way to produce energy that could easily, Croat. j. for. eng. 38(2017)2

cleanly and cost-efficiently be replaced by modern and more qualified sources of energy such as electricity from different sources or from fossil fuels, such as coal, oil and gas. In Nordic countries, efforts to modernize the use of forest biomass for energy were done only under the threat of isolation through disturbances to international trade (Fig. 1).

For a number of reasons, this negative view on forest biomass as a fuel changed (Silveira 2001, Hakkila 2006, Björheden 2006). Considerable efforts have been made to reintroduce forest biomass for energy in a large scale. Some of the globally most successful ex-

259

R. Björheden

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland (259–268)

Fig. 1 In times of disturbances to international trade and threatening isolation, such as the great wars, efforts were made to modernize and intensify the use of forest biomass for energy in the Nordic countries. The picture shows chunking of coniferous branchwood with the manually driven smallwood chunker »Ursus« in a south Swedish forest at the time of the First World War (from Nilsson 2016) amples of the modernized, large-scale rebirth of forest biomass as a fuel and a source of energy can be found in Sweden and Finland, which have also managed to combine a strong traditional forestry sector with an extensive use of forest biomass for energy.

Forests, the utilization of forests for various goods and services and the functions of the forest sector vary strongly. Some of the reasons behind the large differences depend on differences in ecosystem dynamics between forests of different natural regions such as boreal vs tropical forests (Kuusela 1990, Rudel and Roper 1996), some depend on differences of the forest sector and its demand for raw materials, while other explanations must be attributed to differences in traditions and public/social views on forestry (Parrotta et al. 2006). Universal trends concerning forestry and forest biomass utilization, e. g. for energy, can be identified but, as a result of the heterogeneity of forestry, it is normally necessary to present cases. In this review paper, the universal drivers behind the strong expansion of forest fuel use will be illustrated through case descriptions from Sweden and Finland. The relations between forest energy on the one side and forestry, forest industries and bio-economy on the other will be discussed.

2. Discussion 2.1 Forestry and agricultural sectors and proportion of renewable energy in the EU The EU strongly promotes the use of renewable energy sources as important means both for reducing the dependence on foreign energy imports, and for

Fig. 2 In 2015 Sweden and Finland, together with 9 other member states, had overshot their goals for the share of renewables in the gross energy consumption 2020 according to RED (Eurostat 2017a). Countries that have prematurely managed to meet the goals often have strong forest sectors

260

Croat. j. for. eng. 38(2017)2

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland (259–268)

mitigating climate change. In the Renewable Energy Directive 2009/28/EC, RED, (European Council 2009), binding targets are set for all EU member states. By 2020, the Union will reach a 20% share of energy from renewable sources. By 2017, the EU realized a 16.7% share of energy from renewable sources with eleven member states already achieving their goals. The goal for 2020 as well as the current share per member state is shown in Fig. 2.

Biomass, and especially forest biomass, has been identified as an important feedstock for production of renewable fuels and energy. To investigate if a strong forest sector may delimit the growth of renewable energy, a regression was made with the use of renewable energy per capita in each EU member state as the dependent variable and the annual production of roundwood, also per capita, as the independent variable.

The per capita production of roundwood was highly significant (P<0.0001) and explained 71.4% (adjusted r2) of the variation in achieved renewable energy share between countries. The hypothesis that a strong forest sector would be an obstacle to bioenergy through competition for biomass was falsified.

A similar regression with »utilised agricultural land per capita« as the independent variable could not explain anything of the variation. A P value over 0.6 further showed that the relative area of agricultural land does not show a strong correlation to the utilization of biomass for energy. All data for the two regressions were accessed on the Eurostat homepage (Eurostat 2017b).

As Fig. 2 shows, the ambition and implementation of RED varies greatly between the member states. The Union is above the target trajectory but some countries are clearly falling behind. Countries with a strong forestry sector and strong national regulation (e.g. Sweden and Finland) have been more successful in reaching their targets than others. Sweden reached its record high 49% renewable energy target eight years ahead of 2020 and was also the first member state to reach the 10 percent renewable transport fuels target. Probable explanations are the implementation of stable, strong national incentives such as a high carbon dioxide tax, an efficient certificate system for renewable electricity and tax deductions in the transport sector.

2.2 The role of forests and forestry in the national economy Both Swedish and Finnish economies are diversified, but the contributions of the respective forest sectors are very important since they produce a significant net surplus in foreign trade, representing as much as 15 per cent of the value of exports in Sweden and Croat. j. for. eng. 38(2017)2

R. Björheden

almost 40 per cent of the Finnish export earnings. Forest harvesting amounts to around 160 million m3, from an annual yield of just under 250 million m3. More than half the land area of the two countries is covered by forests (Anon. 2013, Anon. 2014).

The unusually important position of the forest sector can be explained by history and by the natural geography of the countries. Boreal and sub-boreal softwood forests are, by far, the most common types of ecosystem. Historically, the early industry in both Sweden and Finland was formed largely around raw materials and fuel from the forests. In both countries, the industrial revolution was »powered by wood« and much of the current industry has close ties to the forest sector (Björheden 2006, Hakkila 2006). In short, both countries have a history of high forest use and build their welfare to a large extent on the forest sector. Public acceptance of forestry is traditionally high, and over 60 per cent of the forests are managed by private, nonindustrial forest owners as a traditional part of family farms (Anon. 2013, Anon. 2014).

Traditionally, most Swedes and Finns are used to, and accept forestry as a sustainable industry. This fact is of great importance to the efforts in intensifying the use of forest biomass. Ongoing urbanization, however, has led to a growing alienation and an increasing proportion of people questioning the sustainability of intensified forest harvesting. This has become an important issue for Nordic forestry to address (Richardson et al. 2003).

2.3 Initial motifs: secure and cost efficient energy sources Although both Sweden and Finland have a strong tradition of large scale use of wood for energy, after the Second World War wood fuels were rapidly and almost completely replaced by fossil fuels. In Finland, the cheap oil of the 1960s displaced wood fuels. In the early 1970s, only 14 per cent of the Finnish energy supply was based on wood. In the same period, over 70 per cent of the Swedish energy supply was based on imports of petroleum, mainly from the Arab states (Doherty et al. 2002). Therefore, it was a severe shock to the national economies when, in 1973, the Arab States, proclaimed an oil embargo (Helby 1997).

Cheap, abundant oil was no longer guaranteed. In 1979, the oil crisis deepened as oil production fell because of the Iranian revolution. OPEC did not increase production to compensate. The Iraqi invasion of Iran in 1980 made matters worse and it became clear that alternatives had to be found to ensure supply of energy with less susceptibility to turmoil in other parts of the world. An Energy Research Programme aiming

261

R. Björheden

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland (259–268)

at gradual relinquishment of oil/fossil fuels was launched in 1975. It became the second largest sectoral research programme in Sweden. The pursuit of sustainability was identified in the long-term energy policy goal: to base the energy system on »durable, preferably domestic, sources of energy« (Haegermark 2001).

Sweden decided to launch an extensive nuclear energy program, planned to develop more of its potential hydropower, but also looked to forest biomass to decrease vulnerability and dependence on imported oil. The research program »Whole Tree Utilization« (1974 to 1977), which aimed to alleviate a potential shortage of pulpwood through use of fibre from small diameter wood and stumps in conventional industrial processes, was complemented with the goal to investigate if dependency on imported fossil fuels could also be relieved (Anon. 1977). There was an intense public debate challenging the decision to build on nuclear power. In 1979 there was a partial nuclear meltdown at the Three Mile Island Nuclear Generating Station, in Pennsylvania, USA. When, subsequently, in 1980 a Swedish referendum was held to decide the future of nuclear power, distrust prevailed. Nuclear power should be a parenthesis and Sweden should develop its renewable sources of energy and save energy to the extent that the nuclear reactors could be phased out. Public scepticism was strengthened when, in 1986, the Soviet Chernobyl accident triggered downfall of radioactive contaminants in large areas of Sweden (Haegermark 2001) and again by the Fukushima accident in Japan in 2011. So far, three of twelve reactors have been shut down and another three reactors are planned to be closed shortly. A green movement of »river rescuers« halted development of the remaining unharnessed rivers (Doherty et al. 2002).

Bioenergy was left as the only sizeable alternative, much to the discontent of leading industrialists who feared that rising energy costs would decrease the competitiveness of Swedish industry (Vedung 2001).

In Sweden, with almost two billion SEK, the Oil Replacement Fund 1980–1987 financed rebuilding oil burners to alternative fuels, mainly wood chips, and infrastructural subsidies, e.g. for investment in terminals to simplify the supply of wood for energy purposes (Hillring 1998, Hillring et al. 2001). The key political driver for bioenergy – to replace imported fossil fuels – was strengthened by the public wish to move away from nuclear power. Improved trade balance and increased earnings in rural areas were ancillary arguments for bioenergy in Sweden (Silveira 2001).

Driven by the same concerns of vulnerability and rising oil prices the Finnish government began to pro-

262

mote wood energy in the mid-1970s, also with complementary goals concerning rural employment and intensified thinning of young forests. However, in contrast to the Swedish development, when oil prices fell and availability was again ensured, »Finnish promotion of wood for energy more or less terminated« (Hakkila 2006). In contrast to Sweden, Finland assumed a reserved, or at least, cautious stance towards forest bioenergy.

2.4 Sustainability and climate change become powerful international drivers The Finnish scepticism to forest bioenergy changed radically when in the early 1990s Finland signed the UN Climate Convention in Rio de Janeiro. By this time, sustainability and greenhouse gas management to reduce the extent of climate change, had become the main drivers for bioenergy (Anon. 2012). All EU member states must reduce greenhouse gas emissions and Finland and Sweden are faced with extraordinarily ambitious goals. A Finnish national Action Plan for Renewable Energy Sources was adapted (Anon 2000b).

In Sweden, a milestone decision came in 1991, when carbon dioxide emissions from fossil fuels were taxed, making bioenergy an economic choice. In Finland, bioenergy was promoted through specific subsidies, especially focussing on supporting small-tree fuel from young stands (Aguilar 2014). This explains why the fractions of additionally harvested forest biomass differ strongly between the two countries with a higher proportion of small trees specifically mobilised by the Finnish KEMERA support (Tanttu and Sirén 2004) and a larger share of logging residues, integrated with conventional harvesting, being the most economical in Sweden (Petty and Kärhä 2011).

2.5 The supply of forest biomass for energy conversion The ten years of development that had been »lost« in Finland, compared to Sweden, was quickly caught up – forestry conditions are very similar, so technology and methods developed in Sweden could easily be used in Finland. Within a few years, the two countries complemented each other and prospered from joint research cooperation and mutual development.

During the 1990s, the technologies and methods for harvesting and converting forest biomass for energy matured in the two countries. Biofuels became an important part of the energy budget of both countries (Anon. 2000a, Hakkila 2006). The build-up of district heating provided an excellent heat sink for large parts of the year and through deliberate concentration on Combined Heat and Power production almost one Croat. j. for. eng. 38(2017)2

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland (259–268)

R. Björheden

Fig. 3 Proportional wood flows in Sweden 2012. The additional biomass in primary forest fuels such as logging residues, small trees and stumps only contribute to around 5% extra biomass for energy, while the sum of energy fractions of roundwood represent 42% of the annual roundwood harvest (Anon. 2014) third of the bioenergy was produced in the form of high value electricity. The current use of primary forest fuel in Sweden and Finland corresponds to around 10 million m3 biomass solid or some 20 TWh (Dìaz-Yáñez et al. 2013), which is more than double that of any other country. Even so, as shown below, the main source of forest biomass for energy, in both counties, consists of by-products from the industrial processing of sawlogs and pulpwood, an annual flow corresponding to 70–75 million m3 biomass solid or 140–150 TWh (cf. Fig. 3).

The supply of primary forest biomass available for direct energy conversion is ultimately determined by the appraisal of the alternative uses. In Sweden, the supply of forest biomass available for energy purposes has been investigated many times, with different results (Björheden and Fick 2014). The reasons for the differences are not missing or incorrect data on standing inventory and forest growth but that the investigations have approached the question with different conditions and restrictions. In principle, all forest biomass can be used for energy production, but this is usually not seen as an economically viable alternative.

The average net export revenue for lumber or pulp is 3–4 times higher per volume unit of unbarked coniferous roundwood (Hakkila 2006) and up to 16 times higher if the product is paper (Björheden 2006) than if the wood is used for the production of energy to replace imported oil. As large a share of the annual felling as possible is consequently used as raw material for the forest industries. Less than 10 per cent of the annually harvested roundwood – normally of low quality or unwanted tree species – will be used directly as fuel, mainly in farms and rural private homes (Anon. 2013, Anon. 2014).

Croat. j. for. eng. 38(2017)2

Thus, the forest industry actual need for wood is normally deducted from the tally of available quantities. The requirement of sustainable production is another common restriction. The latter requirement imposes the felling level to be lower than net forest growth, and that less fertile sites are partly exempt from the removal of forest biomass in addition to stemwood. Finally, technical and economic impediments are considered, i. e. forest areas that are too small or distant are excluded as are forest sites where harvesting conditions are too technically difficult to allow economic extraction of additional biomass.

In countries with developed and internationally competitive forest industries, like Sweden and Finland, this entails that the supply of forest biomass available for energy conversion will depend on and closely follow the felling levels induced by the demand for raw materials by the conventional forest industries. So far, thus, primary forest bioenergy has been retrieved only from harvestable fractions that are not demanded by the conventional forest industry. In principle, this is likely to prevail, unless energy prices rocket or, alternatively, forest biomass become very cheap.

In Sweden, the average net felling (excluding wood left in the forest) for the last five years amounts to 84 million m3 stem volume ob. Annual felling results in approximately 125 million m3 of solid biomass if also branches, foliage, tops, small trees and harvestable parts of the stump-root system are tallied. The felling level corresponds to 73 per cent of the annual growth, i.e. 27 per cent of the annual growth contributes to build-up of the standing inventory. The relative usage of forest biomass increment in Sweden is shown in Fig. 4.

263

R. Björheden

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland (259–268)

Fig. 4 Relative usage of forest biomass increment in Sweden (2009–2014). The procurement of primary forest biomass for energy conversion may be tripled without increasing the level of felling. The build-up of inventory corresponds to 27% of the annual increment. In principle, most of this volume could also be realized as feedstock

2.6 The role of conventional forest industries in the forest-energy value chain The logging operations carried out to supply the conventional forest industry with roundwood of different qualities also mobilizes forest biomass suitable for bioenergy. As shown in Fig. 5, each harvested m3 of industrial roundwood yields another 0.3 m3 (=0.6 MWh) of biomass in the form of fuel fractions such as low grade roundwood, small trees, slash and stumps. In Sweden, the large-scale procurement of wood for the conventional industry will thus contribute effectively to making wood fuel resources available at low costs.

This fuel feedstock is a consequence product only to be burdened with their induced incremental costs.

The forest industry’s most important contribution to the value chain of energy is, however, not that it makes available biomass that is poorly suited to the current industrial production. The main contribution is, instead, that by-products from industrial processing will become available in large volumes. Bark, sawdust, breakage and black liquor is used almost entirely for energy production, turning a potential waste into a valuable resource. In fact, forest industries are forerunners in substitution of fossil fuels – the Swedish forest industry uses some 50 TWh of forest bioenergy/ year, corresponding to 25 million m3 of solid biomass.

As shown in Fig. 6, between 45 and 50 per cent of the biomass in industrial wood becomes available for en-

264

Fig. 5 After deducting volumes for ecological and techno-economic restrictions, each m3 of harvested roundwood releases a potential primary fuel feedstock with an energy content of at least 0.6 MWh. The figure shows the Swedish usage level 2013 of these fuel fractions. The harvest of primary fuel may be tripled ergy conversion. In comparison to this, the Swedish extraction of primary fuels is modest (Nilsson 2006, Thorsén and Björheden 2010).

Fig. 6 In Sweden, after industrial processing, 45 per cent of the biomass of a m3 sob of roundwood is turned into fuel feedstock by-products, representing 0.9 MWh. These by-products are practically completely used. Thus, in 2013, more than five times more energy was gained from by-products of industrial wood, than from the harvest of primary forest bioenergy fractions Croat. j. for. eng. 38(2017)2

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland (259–268)

Today, fuel with a forest origin is the single most important energy source in the Swedish energy balance. This large-scale use of forest biomass for energy purposes was initially developed to reduce dependence on oil. With time, however, climate issues have become increasingly important.

From a climate perspective, durable products are preferable. And all products made from biomass can be converted into energy when their useful life is over.

The old debate question »pulp, saw or burn?« should be replaced by the exclamation »pulp, saw and burn!«.

Reuse of biomass-based products, with energy conversion as the final step, is a way to maximize the benefits and value of the Swedish forest from a combined climate-and-economic perspective (Joelsson and Gustavsson 2012).

2.7 Technological development for improved procurement of forest biomass for energy The development of technologies and methods to enable and streamline the harvest of additional biomass for energy became necessary. Apart from such roundwood that is not demanded by industry because of decay or other defects, the primary fuel fractions (logging residues, small sized trees and harvestable stumps) are bulky, difficult to handle, heterogeneous, wet and often contaminated and represent a low value. The development of technologies and methods to enable harvest of primary fuel feedstock reflect these problems but also include attempts to reduce the amount of mineral nutrients removed from the forest – especially in the branches and tops.

As mentioned, the first Swedish development efforts were made as an appendix to the research programme »Whole Tree Utilization«, (Anon. 1977). It has been followed by several more dedicated programs, in Sweden normally funded by the Energy Authority (Nilsson and Lönner 1999). Richardson et al. (2001, 2003) provide a brief review of the early technical development. Later, programmes funded jointly by the government and forestry made significant contributions to development and systems evaluations. Hakkila (2004), Thorsén and Björheden (2010) and Iwarsson Wide and Björheden (2016) offer broad reviews of this recent development, which are summarized in the following section.

With a few exceptions, the technology proven in Finland and Sweden, builds on the idea of piggy-backing on conventional forestry, using almost the same machines as in conventional harvest, with minor modifications to address the difficulties of the primary assortments. Examples of modifications are residue grapples simplifying loading of residues and decreasing the Croat. j. for. eng. 38(2017)2

R. Björheden

risk of contamination, detachable extra wide loading racks on the forwarder to allow full loads of the bulky fuel fractions and accumulating felling and harvesting heads for small sized trees, addressing the problem of very small piece sizes, an invention that has become standard also for harvesting of small sized pulpwood.

The most important specialised equipment for additional forest biomass harvest is tractor or truck mounted technology for comminution – chippers for clean fuels and crushers for contaminated fuels such as stump-wood. The main reasons that decentralised comminution has become a viable solution are that, in addition to producing a ready fuel, it will significantly reduce the bulk density of the material, decreasing transport costs, simplify the subsequent handling.

Several bundling, baling and compaction devices that have been developed have offered the same level of compaction and simplified handling but have not been able to provide economic feasibility to any large extent. Also, designated stump harvesters have been developed. They have become rather common in Finland but are presently used only for experimental purposes in Sweden.

2.8 Conditions for value creation in the forest-energy value chain One of the causes for the modest introduction of highly specialized equipment for harvest of residual forest biomass is the delimited scope for value creation in the forest-energy value chain, illustrated by Fig. 7.

There are several reasons for this. Forest products have mainly been used to generate heat, which is the simplest form of energy, with the lowest value. Another reason is that fuel feedstock, which mainly consists of by-products from logging and industrial processes, is very heterogeneous, difficult to handle and bulky. A troublesome seasonality of demand makes it necessary to store the biomass over longer periods. This adds significant costs to production. However, the main reason for the difficulties to increase the profitability of forest-based energy is the very low overall energy price established in the post-war period.

This fact forms an impediment for refinement of forest biomass into more attractive forms of biofuels as syn-gas based FT diesel, DME, methanol, hydrogen, etc. (Bengtsson 2012). There are well known and researched technologies for such refinement, but the processes are costly and need to be run in large scale to be viable. This, on the other hand, induces diseconomies of scale as the supply area and the following transport costs increase. The economic incentives for forest owners or the traditional industry to venture into large scale facilities for solid-to-gas or solid-to-

265

R. Björheden

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland (259–268)

Fig. 7 Solid wood fuels (by-products, primary chips and recycled wood) have a low real-price development compared to more qualified energy carriers as electricity or fossil liquid fuel. Current prices in SEK per MWh liquid refineries for the primary fuel fractions will thus be limited.

In spite of the difficulties mentioned in the previous sections, the achievements of forest based bioenergy are impressive in both Sweden and Finland. According to the Renewable Energy Directive (2009/28/EC), 49% of the consumed energy in Sweden should be based on renewable sources by 2020. This target was reached already in 2012 and has since been surpassed (Kühmaier et al. 2017). Also Finland is overshooting its trajectory, and renewable energy in Finland grew to 38.7% of t energy consumption in 2014, achieving a second place (joint with Latvia) in terms of renewable share of energy consumption, behind Sweden in first position with a 52.6% share (Eurostat 2016) The Swedish climate policy targets are, however, even more ambitious, aiming at net zero GHG emissions by 2050 and »fossil free« road transports by 2030 as pointed out by Cintas et al. (2017), who even mention a possibility of reaching »negative emissions« to increase the allowable GHG emissions for the rest of the world. It seems unlikely that these goals will be fulfilled without powerful and sustained control or incentive systems.

3. Conclusions Most governments are working hard to maintain and develop welfare. Abundant availability of energy is the basis for achieving this objective. The already

266

existing combined heat and power production, as a simple and straightforward addition to heat production, will continue to thrive (and the need for process heat opens this possibility also for countries in temperate zones). However, it is not likely that any major investments will be seen in high end biofuels from additional forest biomass, until game changing events occur, such as e. g. much higher fossil energy prices, highly efficient enzymatic cellulose technologies or strong international subsidies/fees favouring biofuel.

The EU member states show varying success in fulfilling the RED goals for renewable energy, seemingly coupled to the national incentive systems. On the EU level, the general incentives are weaker. The Union does not have a common carbon taxation system and the ETS returns very low prices on carbon dioxide emissions. Also, globally, the agreements on climate gases have been relative failures (Cooke 2012), including fairly straightforward initiatives such as the international CO2-emmission rights trade (Zacher 2015).

Together, such failures and weaknesses mean that the incentive to invest is very limited. The ILUC directive (EU) 2015/1513 (Indirect Land Use Change) introduced in 2015 will, if anything, slow down the development of renewable fuels for the transport sector.

4. References Aguilar, F.X. (Ed.), 2014: Wood energy in developed economies. Resource management, economics and policy. Routledge Earthscan Series, Taylor and Francis 2014, 338 p.

Croat. j. for. eng. 38(2017)2

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland (259–268) Anon., 1977: Project whole tree utilisation. Final summary report. Swedish Royal College of Forestry.

Anon., 2000a: Resultatredovisning av forskning och utveckling inom energiområdet. (Results of R&D in the energy sector), ER 16:2000, Swedish Energy Authority, 103 p.

Anon., 2000b: Finland’s Action plan for renewable energy sources. Publication 1/2000. Finnish Ministry of Trade and Industry.

Anon., 2012: En nationell strategisk forskningsagenda för den skogsbaserade näringen i Sverige. (A national strategic research agenda for the forest-based industries in Sweden) Skogsindustrierna 2012, 59 p. Available at: http://www.nrasweden.se/sites/nra-sweden.se/files/nra-dokumentationhela_0.pdf Anon., 2013: Statistical yearbook of forestry 2013, Metla and Official Statistics Finland, 416 p.

Anon., 2014: Swedish Statistical Yearbook of Forestry 2014, Swedish Forest Agency and Official Statistics Sweden, 406 p.

Bengtsson, S., 2012: Final report – CHRISGAS (Clean Hydrogen-rich Synthesis Gas), EC-CORDIS, 22 p. Available at: http://cordis.europa.eu/publication/rcn/11476_en.html Björheden, R., 2006: Drivers behind the development of forest energy in Sweden In: Richardson, J., (Ed.): Sustainable production systems for bioenergy: impacts on forest resources and utilization of wood for energy. Proceedings of the third annual workshop of Task 31, Flagstaff, Arizona, USA, October 2003. Biomass and Bioenergy 30(4): 289–295.

Cintas, O., Berndes, G., Hansson, J., Poudel, B.C., Bergh, J., Börjesson, P., Egnell, G., Lundmark, T., Nordin, A., 2017: The potential role of forest management in Swedish scenarios towards climate neutrality by mid-century. Forest Ecology and Management 383: 73–84.

Cooke, S., 2012: Why UN climate agreements fail. Global Research, December 03. Available at: http://www.globalresearch.ca/why-un-climate-agreements-fail/5313881 Dìaz-Yáñez, O., Mola-Yudego, B., Anttila, P., Röser, D., Asikainen, A., 2013: Forest chips for energy in Europe: Current procurement methods and potentials. Elsevier publishers: Renewable and Sustainable Energy Reviews 21: 562–571.

Doherty, S.J., Nilsson, P.O., Odum, H.T., 2002: Emergy evaluation of forest production and industries in Sweden. Swedish University of Agricultural Sciences, Department of Bioenergy, Report No 1. Uppsala, Sweden.

European council, 2009: Directive 2009/28/EC of the European Parliament and of the Council of 23 April 2009 on the promotion of the use of energy from renewable sources and amending and subsequently repealing Directives 2001/77/EC and 2003/30/EC. http://www.nezeh.eu/assets/media/fckuploads/file/Legislation/RED_23April2009.pdf. Accessed April 04, 2017.

European council, 2015: Directive 2015/1513/EC of the European Parliament and of the Council of 9 September 2015 amending Directive 98/70/EC relating to the quality of petrol and diesel fuels and amending Directive 2009/28/EC on the Croat. j. for. eng. 38(2017)2

R. Björheden

promotion of the use of energy from renewable sources. http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CE LEX:32015L1513&from=en. Accessed April 05, 2017.

Eurostat, 2017a: News Release 43/2017 Renewable energy in the EU. http://ec.europa.eu/eurostat/documents/2995521/ 7905983/8-14032017-BP-EN.pdf/af8b4671-fb2a-477b-b7cfd9a28cb8beea. Accessed April 04, 2017.

Eurostat, 2017b: Statistics on annually harvested roundwood, utilised agricultural area and population. http://ec.europa. eu/eurostat/data/database. Accessed April 04, 2017.

Haegermark, H., 2001: Priorities of energy research in Sweden. In: Silveira, S. (ed.): Building sustainable energy systems – Swedish Experiences. Swedish National Energy Administration: 163–196.

Hakkila, P., 2006: Factors driving the development of forest energy in Finland. In: Richardson, J., (Ed.): Sustainable production systems for bioenergy: impacts on forest resources and utilization of wood for energy. Proceedings of the third annual workshop of Task 31, Flagstaff, Arizona, USA, October 2003. Biomass and Bioenergy 30(4): 281–288.

Hakkila, P., 2004: Developing technology for large-scale production of forest chips Wood Energy Technology Programme 1999–2003. Technology Programme Report 6/2004, Final Report, 99 p.

Helby, P., 1997: Energi som säkerhetsfråga. (Energy as an issue of security). In: Jarvas, G., (Ed.): 2000-talets stora utmaningar – Aktuella resurs- och miljöproblem i ett konfliktperspektiv (Current resource and environmental problem from a conflict perspective – great challenges of the 21st century).

Stockholm: SNS Förlag: 88–130.

Hillring, B., 1998: National strategies for stimulating the use of bioenergy: Policy instruments in Sweden. Biomass and Bioenergy 14(5/6): 425–437.

Hillring, B., Ling, E., Blad, B., 2001: The potential and utilisation of biomass. In: Silveira, S., (Ed.): Building sustainable energy systems – Swedish Experiences. Swedish National Energy Administration.

Iwarsson W.M., Björheden, R., (Eds.), 2016: Forest energy for a sustainable future – composite report from the R&D programme Efficient Forest Fuel Supply Systems 2011–2015, 129 p.

James, S.R., 1989: Hominid use of fire in the lower and middle Pleistocene: A review of the evidence. Current Anthropology, University of Chicago Press. 30(1): 1–26.

Joelsson, J., Gustavsson, L., 2012: Swedish biomass strategies to reduce CO2 emission and oil use in an EU context. Energy 43(1): 448–468.

Junginger, M., Faaij A., Björheden, R., 2004: Technological learning and cost reductions in woodfuel supply chains. In Proceedings of the 2nd World Conference on Biomass for Energy, Industry and Climate Protection. Rome, Italy.

Kuusela, K., 1990: The dynamics of boreal coniferous forests.

SiTRA 112, Helsinki, Finland, 172 p.

267

R. Björheden

Development of Bioenergy from Forest Biomass – a Case Study of Sweden and Finland (259–268)

Kühmaier, M., Spinelli, R., Visser, R., Devlin, G., Eliasson, L., Laitila, J., Laina, R., Wide, M.I., Egnell, G., 2017: An international review of the most productive and cost effective forest biomass recovery technologies and supply chains. Renewable and Sustainable Energy Reviews 74: 145–158.

Nilsson, B., 2016: Extraction of logging residues for bioenergy – effects of operational methods on fuel quality and biomass losses in the forest. Linnaeus University Dissertations No 270/2016, 200 p.

Nilsson, P.-O., Lönner, G. (Eds.), 1999: Energi från skogen.

(Energy from the forest). SLU kontakt 9, Uppsala, Sweden.

Nilsson, P.-O., 2006: Biomassaflöden i svensk skogsnäring 2004 (Biomass flows in Swedish forestry 2004), Swedish Forest Agency, Rapport 2006, 23 p.

Parrotta, J., Agnoletti, M., Johann, E. (Eds), 2006: Cultural heritage and sustainable forest management: The role of traditional knowledge. Proceedings of a IUFRO conference held in Florence, Italy, 8–11 June, 547 p.

Petty, A., Kärhä, K., 2011: Effects of subsidies on the profitability of energy wood production of wood chips from early thinnings in Finland. Journal of Forest Policy and Economics 13(7): 575–581.

Richardson, J., Björheden, R., Hakkila, P., Lowe, A.T., Smith, C.T. (eds.), 2001: Bioenergy from sustainable forestry: guiding principles and practice. Kluwer Academic Publishers, The Netherlands, 348 p.

Richardson, J., Smith, T., Björheden, R., Lowe, A. (Guest eds.), 2003: Principles and practice of forestry and bioenergy in

densely-populated regions. Proceedings of the IEA Bioenergy Task 31 workshop, Garderen, The Netherlands, 16–21 September. Biomass and Bioenergy 24: 4–5.

Rudel, T., Roper, J., 1996: Regional patterns and historical trends in tropical deforestation, 1976–1990: A qualitative comparative analysis. Ambio 25(3): 160–166.

Silveira, S. (ed.), 2001: Building sustainable energy systems – Swedish experiences. The Swedish National Energy Administration, 552 p.

Sundberg, U., Lindegren, J., Odum, H.T., Doherty, S.J., 1994: Skogens användning och roll under det svenska stormaktsväldet. (Utilisation and role of forests in Sweden as a great power). The Royal Swedish Academy of Agriculture and Forestry, Stockholm.

Tanttu, V., Sirén, M., 2004: Co-operation and integration in wood energy production. International Journal of Forest Engineering 15(2): 85–94.

Thorsén, Å., Björheden, R. (Eds.), 2010: Efficient forest fuel supply systems – composite report from a four-year R&D programme 2007–2010. Skogforsk, Sweden, 113 p.

Vedung, E., 2001: The politics of Swedish energy policies. In: Silveira, S. (ed.): Building sustainable energy systems – Swedish Experiences. The Swedish National Energy Administration: 95–130.

Zacher, S., 2015: The world’s worst market failure: Greenhouse gas emissions. The Gate, Chicago. http://uchicagogate. com/2015/06/01/the-worlds-worst-market-failure-greenhouse-gas-emissions. Accessed April 05, 2017.

Authors’ address:

Received: January 10, 2017 Accepted: May 10, 2017

268

Prof. Rolf Björheden, PhD. e-mail: rolf.bjorheden@skogforsk.se The Forestry Research Institute of Sweden – Skogforsk Uppsala Science Park SE-751 83 Uppsala SWEDEN Croat. j. for. eng. 38(2017)2

Subject review

Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten Years (2007–2017) – Part One: Harvesting and Storage Gernot Erber, Martin Kühmaier Abstract Forest fuel is a renewable source with the potential to substitute fossil fuels in several application fields, such as reducing greenhouse gas emissions and supporting rural areas by fostering income and jobs. Contribution margins in fuelwood supply are small and ensuring supply chain efficiency plays a crucial role in delivering high quality products at competitive costs.

This paper provides a review of research trends related to this issue in the field of harvesting and storing fuelwood and the impact of recent technology during the last decade.

Whereas the basic suitability of supply chains and machines was research’s main priority 20 years ago, the focus shifted to improving the efficiency of machines appropriate for harvesting fuelwood during recent years. Significant increase of productivity could be achieved by introducing fuelwood harvesting heads for processing whole trees to bunches convenient for forwarding (»multi-tree handling«), and adapted working techniques (»boom-corridor thinning«) were developed. Development of compaction measures for bulky raw materials, like logging residues, applied during or before processing and forwarding, culminated in dedicated machines (»bundle-harvester«). Improving the final product quality by appropriate storage practices was emphasised. The phrase »moisture content management« voiced the urgent need for prompt monitoring of fuelwood drying behaviour, which was answered by research in the form of meteorological data based drying models. Among the treatments applied before and during storage, covering has evolved as the most suitable measure. Furthermore, research advocated for greater attention to dry matter losses during storage and the development of basic energy density.

Keywords: fuelwood, supply chains, biomass, research trends, harvesting, storage, multipletree handling, meteorological drying models

1. Introduction Energy from biomass is considered a major contributor to current international climate change mitigation and energy security. By 2030, the European Union (EU) is aiming to produce 27% of its primary energy from renewable sources (COM/2014/015). Forest biomass is a key renewable energy source with the potential to meet these long-term renewable energy targets. However, the potential is not exploited due to low efficiency and high costs in the supply chain, reCroat. j. for. eng. 38(2017)2

sulting in low or even negative contribution margins (Ghaffariyan et al. 2017). For this reason, efficiency improvement and cost reduction remain top priority topics of research in this field.

Recently, several overviews of research on technologies and procedures suitable for increasing the efficiency in harvesting and storage of fuelwood have been compiled. While Stampfer and Kanzian (2006) outlined the challenges and opportunities prevalent in mountainous regions, Routa et al. (2013) investigated the driving forces behind current technical solutions

269

G. Erber and M. Kühmaier Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten ... (269–278)

of forest energy procurement systems in Finland and Sweden, ending with some perspectives on possible future developments. Lately, Ghaffariyan et al. (2017) provided an extensive overview of best practice examples of state-of-the art forest biomass harvesting technologies and supply chains used in North America, Europe and the Southern Hemisphere. Focusing on the harvesting process, Bergström and Di Fulvio (2014a) estimated the effect of future harvesting and handling technologies on the cost and energy efficiency of supply chains for young, dense thinnings.

Nevertheless, there is no paper available which comprehensively covers relevant and promising research trends in the field of forest fuel supply chains, also due to an apparent conflict between subject extent and limitations in manuscript length. As part one of a series of two, this paper aims to cover the first two steps in the forest fuel supply chain, namely harvesting and storage, while the second will cover the steps comminution and transport.

In detail, the objective of this paper is to provide an overview of research papers in forest biomass supply in Europe considering the years 2007 to 2017, covering the »harvesting« and »storage« steps in the forest fuel supply chain and to analyse the impact of the presented technologies and work methods on the biomass supply chain, especially in terms of economy. The papers will be classified according to key supply processes and further into research trends. Finally, future research needs will be identified to fuel both industrial and academic development.

2. Material and Methods In order to access and collect the papers relevant to this review, an extensive online literature search was conducted. »Scopus«, »Web of Science« and »Google Scholar« are the most used search engines. A combination of the key words »harvesting«, »extracting«, »storage«, »fuelwood«, »energy wood« and »supply chain« was applied so that at least one word from each of the search terms (logical OR operator) or multiple terms should appear (logical AND operator) either in the title or the abstract of the paper. As the focus of this paper is the development of harvesting and storage in Europe during the last decade, except for some highly relevant papers from other continents, only highly relevant studies executed or demonstrated in Europe were included in the analysis.

The search resulted in a gross list of 99 papers, of which 32 were not relevant for the focus of this paper, leaving 67 papers for the review. These papers were classified by and summarized on supply process level

270

(»harvesting« and »storage«). These were further split into seven research trend groups, each containing novel knowledge on a certain subject gained during the last decade.

3. Results 3.1 Harvesting Throughout the last 25 years, the research on mechanized harvesting of fuelwood underwent continuous change, steadily triggered by improvements of existing and development of new technology. Trying to figure out the most promising ones, Bergström and Di Fulvio (2014a) report that, out of a gross list of 14, five systems were considered likely to improve harvesting of fuelwood from young dense thinnings. Consequently, this section deals with these findings as well as other relevant subjects of research in this field during the period under study.

3.2 Multi-tree handling Accumulating harvesting heads or »fuelwood harvesting heads« enable felling of more than one tree per cycle and collecting these by means of accumulating arms, a feature consequently termed »multi-tree handling«. Equipped with feeding rollers and delimbing knives, these heads process and buck tree bundles to length. Cutting elements include disk saws, saw bars and shear blades, latter available in elliptical, guillotine or scissor configuration. Guillotine cutting devices are advantageous in young stands, as, due to the absence of bar and chain, small trees cannot slip between them (Erber et al. 2016a).

Spinelli et al. (2007) tested two models of Timberjack feller-buncher heads, capable of accumulating up to six trees. Productivity depended mostly on tree mass and ranged from four to eight green tonnes per effective hour of work in thinnings. However, the cutting quality of shear heads was not satisfying in the coppice treatments in Italy and France. For a forwarder equipped with a feller-buncher head (Moipu 400E), Rottensteiner et al. (2008) found that tree volume and forwarding distance significantly influenced system productivity. Ovaskainen et al. (2008) came to a similar conclusion, adding harvesting intensity to their explanation. Unexpectedly, a ceiling to the positive effect of accumulation in terms of time required per tree was discovered at above ten trees per cycle. Belbo (2010) points out that utilisation of the theoretical accumulation capacity is limited by hampered manoeuvrability of a bunch of trees and poor visibility in dense, young stands, an issue not relevant in clear cutting operations (Jylhä and Bergström 2016).

Croat. j. for. eng. 38(2017)2

Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten ... (269–278)

The introduction of feed rollers and delimbing knives in fuelwood harvesting heads enables integrated harvesting of delimbed, cut-to-length bundles of fuelwood and pulpwood. Laitila and Väätäinen (2013) investigated harvesting with a Naarva EF28 fuelwood head mounted on an excavator-based harvester. It showed that the efficient use of the multi-tree handling-capability by the operator could increase the productivity to a level similar to harvesting machinery dedicated to forestry. The same head, mounted on a conventional harvester, was studied while harvesting hornbeam undergrowth from below a mature stand of broadleaves and conifers (Erber et al. 2016a). Extensive employment of (>70% of the cycles) multi-tree handling resulted in 27% longer harvesting cycles than in single tree harvesting. However, the volume extracted per cycle increased by 33%, which hence caused the overall productivity to rise by 5% compared to single tree harvesting, resulting in a cost decrease of 2.6 € per m3. The average time per tree decreased by 59% compared to single tree handling.

Hardwood conditions noticeably reduced (15%) the head’s maximum felling diameter. Laitila et al. (2016) point out that crooked and forked trees present a serious obstacle to multi-tree handling in combination with delimbing and bucking. Delimbing can further be a drawback regarding removal per hectare. However, it can be beneficial and cost-efficient in stands where whole tree harvesting is precluded due to nutrient loss or other ecological reasons (Laitila et al. 2010).

Petty and Kärhä (2014) state that heads with multi-tree handling, delimbing and bucking capability are only cost-efficient in early thinnings with an average diameter at breast height (DBH) below 15 cm. Di Fulvio et al. (2011) showed that additional harvest of fuelwood results in 200% more volume compared to pulpwoodonly harvesting in this DBH range and that it is crucial in receiving a net income.

In thinning operations, plenty time-consuming, non-linear movements are required during felling and bunching. If »boom-corridor thinning«, a variant where trees between strip roads are harvested in 1 m wide corridors with a length corresponding to the reach of the crane, higher productivity can be achieved.

Compared to conventional thinning from below, Bergström et al. (2010) observed that boom-corridor thinning brought a significant productivity increase (16%), while the time consumption for crane movement between trees decreased by 17%. In a simulation performed by Sängstuvall et al. (2012), selective multipletree handling increased productivity by 20% to 46% compared to single tree handling, and boom-corridor thinning resulted in a further 41% increase compared Croat. j. for. eng. 38(2017)2

G. Erber and M. Kühmaier

to selective multiple-tree handling. Triggered by these results, Bergström et al. (2012) decided to investigate a novel harvesting head, which could fell multiple trees of up to 8 cm DBH in a row while moving the crane at a speed 1.3 m s-1.

Forwarding is significantly improved by multi-tree handling, as the harvested volume is concentrated in a smaller number of locations, consequently increasing grapple loads and speeding up forwarding (Erber et al. 2016a, Laitila et al. 2016). However, the concentration effect declines with increasing forwarding distance, because the share of loading time decreases in the total time consumption (Laitila et al. 2007). Schweier et al. (2015) reported substantially better load presentation and, therefore, drastically enhanced forwarding after multi-tree handling, while the costs did not differ from motor-manual harvesting in coppice harvesting.

3.3 Compressing and bundling The concept of bundling small, whole trees, logging residues and pulpwood, was developed to enhance forwarding and transport economics. During the last ten years, substantial progress has been made in this field.

Jylhä and Laitila (2007) published the first study on a bundle-harvester (Fixteri). A rotating, semi-automatically operating bundling unit, mounted on the rear end of a harwarder base machine, was combined with an accumulating felling head. Uniformly sized bundles, 2.6 m long, 0.5 m3 (pulpwood) and 0.3 m3 (fuelwood) large bundles were formed with a productivity of 2.8–3.7 m3 per effective hour of work. If technical issues, namely the underpowered base machine and inefficient felling and feeding units were solved, this machine would have potential, especially in smalldiameter (7–10 cm) stands (Kärhä et al. 2011). However, to achieve cost-efficiency, the productivity would have to rise to 4.6 m3 and 8.7 m3 per effective hour depending on the tree diameter (7–13 cm). An improved version (Fixteri II) displayed a productivity increase of 38% to 77%, which was primarily attributed to a new accumulating felling head, which enabled smooth feeding of the bundling unit. Multipletree handling (2.9 trees per grapple during more than 80% of the crane cycles) was more efficient compared to the previous machine (1.3 trees per grapple during 19% of the cycles), increasing the volume fed into the bundling unit per cycle. Improved hydraulic capacity led to an increased share of total effective working time covered by simultaneous work elements (26% to 36%) (Nuutinen et al. (2011). The Fixteri FX15a, the third version of the Fixteri had a productivity range

271

G. Erber and M. Kühmaier Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten ... (269–278)

from 9.7 m3 (average tree volume of 0.027 m3) to 13.8 m3 (average tree volume of 0.084 m3) per productive machine hour. Multiple-tree handling was improved again (2.1 to 4.3 trees per cycle). However, utilization of the bundling capacity remained low (11% to 57%), and the felling unit was accordingly considered to remain the system bottleneck (Nuutinen and Björheden (2016).

Laitila et al. (2013) found that conventional terrain bundling of logging residues is the most cost-competitive option for forwarding distances greater than 200 m, while forwarding with successive bundling at the roadside is only beneficial for short forwarding distances (<100 m) and for a removal of >50 m3 per ha.

Spinelli et al. (2012) concluded that under mountainous conditions operational efficiency of a truckmounted bundler could only be achieved if either productivity increased by 30% or capital cost could be decreased by one third. The authors concluded that a forwarder-mounted unit would have been more efficient, as the truck-mounted bundlers operation area was limited to the roadside and entailed frequent relocation.

Manzone (2016) presented a small, prototype bundler for woody forestry and agriculture residues, especially suitable for small businesses and individual farmers. Up to 30 bundles of 18 kg to 20 kg could be produced per hour. At least under Italian pricing schemes, cost-efficient operation is possible.

Bergström et al. (2010) presented two different concepts for compressing fuelwood loads. The first concept involved breaking and flattening branches and removing foliage and fine branches from tree bunches during processing by a dedicated unit mounted on the boom-tip of a conventional harvester. Bulk density and net energy density of fresh, small Scots pine trees increased by 40% and 80% (5 cm to 8 cm DBH) and by 160% (12 cm to 15 cm DBH), while the bundle diameter decreased by 26% and 40%, respectively. However, these effects could not be observed for trees stored for as long as 10 months, and mass loss during compressing was similar to that of stored trees. The second concept involved a device for hydraulically compressing the trees on the load-bed of a forwarder by moving the stakes from the outward to the inward position. Through compression, the bulk density of birch whole trees decreased by more than 30% and a significant increase in utilized load capacity share was achieved (75% vs. 55% to 60%). A prototype head (MAMA), equipped with a feed-roller system and capable of compression-processing of whole tree bunches was compared to a conventional Bracke C16 felling head in early fuelwood thinnings by Bergström and di

272

Fulvio (2014b). Its use increased bunch bulk density by 47% to 70%. Compression-processing related mass loss (12% to 14%) was equalized by significantly increased forwarding productivity (12%).

3.4 Storage Fuelwood storage and drying of fuelwood have been important research subjects for at least half a century, and comprehensive knowledge has been gathered during this period. Nevertheless, research on this topic is carried out unrelentingly, nowadays focusing also on the downstream effects of fuelwood drying, such as transport economics. Consequentially, this section deals with novel knowledge in this field during the period under study.

3.4.1 Storage season and duration It is common knowledge that spring and summer are the periods most suitable for drying of fuelwood.

Contrary, autumn and winter exhibit significantly reduced drying rates or even rewetting. Badal et al.

(2015) confirmed that the particularly suitable drying regime in spring results in lower moisture content of fuelwood at the time of chipping compared to material stored during other periods. Klepac et al. (2008) found that the drying rate of Pinus taeda trees stored in summer was about 50% higher (0.36% per day) than in autumn and winter (0.22% and 0.23%). Brand et al.

(2010) found that the optimal storage duration for fuelwood was four to six months, starting in spring. Longer storage into autumn resulted in rewetting and decrease of net calorific value. Gautam et al. (2012 and 2013) witnessed decrease of logging residues moisture content for two successive drying seasons, but not during a third.

3.4.2 Pile size and shape, position in the pile and fuelwood size The size and shape of a fuelwood pile determine its drying surface and susceptibility to ventilation.

Gautaum et al. (2012) report that, while half-sphere (»beehive pile«) shaped piles are suitable for longterm softwood storage, half-ellipsoid (»windrow pile«) shaped ones are best for short-term softwood storage and hardwood storage in general. This effect is attributed to the greater surface to volume ratio of the windrow piles and decreased resistance to airflow due to their limited width. The more pronounced branching of hardwoods leads to less compaction and more void space, which further improves drying.

Ichihara et al. (2010) report that Japanese cedar logs dried similarly regardless of piling in either triangle or quadrilateral shape. Pile size was the determining factor for the drying performance of Loblolly pine Croat. j. for. eng. 38(2017)2

Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten ... (269–278)

whole trees. Those stored for 70 days in skidder-sized bundles dried significantly better (final moisture content of 23.8% to 29.2%) than trees in a large pile (28.6% to 48.9%) (Klepac et al. 2014). Fuelwood stored in piles displays a decreasing moisture content gradient from the inside to the outside (Erber et al. 2012, Ichihara et al. 2010, Kofman and Kent 2009a, Klepac et al. 2014, Röser et al. 2011). Röser et al. (2011) and Kofman and Kent (2009b) point out that this gradient is not likely in covered logwood piles. However, Manzone (2015) could not observe a gradient in uncovered piles under Italian conditions either. Piece size impacts fuelwood drying, as larger (thicker and/or longer) logs dry significantly slower than smaller ones (Kim and Murphy 2013, Visser et al. 2014, Elber 2007, Bown and Lassere 2015).

3.4.3 Covering, debarking, splitting, bundling and pre-drying in the stand The effect of covering in fuelwood drying is strongly dependent on the climate of the storage area. In Nordic countries, it is regularly applied to prevent rewetting during winter and thaw. Röser et al. (2011) observed that covering was beneficial for coniferous and deciduous logwood in the wet climate of Scotland and Finland, while it had no effect under Italian conditions. Nurmi and Hillebrand (2007) quantify the effect of covering at a 3% to 6% lower moisture content after storage from winter to autumn. Most of the effect manifested during thaw. Filbakk et al. (2011a), Nurmi (2014), Röser et al. (2011) and Visser et al. (2014) consider covering summer-dried fuelwood crucial in maintaining the moisture content decrease gained until then. A positive effect of covering was witnessed for Sug logging residues (Yukio and Tomohiro 2011) and coniferous and deciduous logwood (Elber 2007). Kofman and Kent (2009a and 2009b) point out that covering is far more important when storing fuelwood in the forest.

Debarking breaks the barrier that a tree bark constitutes to evaporation of moisture. Röser et al. (2011) report a significant, positive effect, when the piles are covered. Bown and Lassere (2015) observed that debarked Eucalyptus logs dried 8% faster than those in bark. Contrarily, Nurmi and Lehtimäki (2011) could not detect this effect after partial or strip debarking by a modified single grip harvester head. Splitting of large logs significantly increases drying rates during summer, when the piles are covered (Visser et al.

2014).

While bundling is primarily carried out for logging residues and small trees to enhance manipulation, it can also have a positive effect on the drying performance through decreased susceptibility to penetration Croat. j. for. eng. 38(2017)2

G. Erber and M. Kühmaier

by snow and rain (Petterson and Nordfjell 2007). However, this behaviour could not be observed by Afzal et al. (2010) and Filbakk et al. (2011b).

Pre-drying in the stand is a common practice for logging residues in Nordic countries. Three weeks of pre-drying in the stand can reduce the moisture content by more than 20% (Petterson and Nordfjell 2007), an effect similarly observed by Cutshall et al. (2010).

Yet, Nilsson et al. (2013 and 2015) witnessed that, at the end of summer, the moisture content differed only slightly between logging residues stored in-stand and those piled fresh at roadside. Similar results were observed by previous studies (e.g. Nurmi and Hillebrand 2001). However, the moisture content in Nurmi and Hillebrand (2001) started to developed differently after the start of autumn when the residues in the stand took up significantly more moisture than the piled ones.

3.4.4 Dry matter losses Dry matter losses during fuelwood supply can be caused either by microbial activity (commonly fungal attacks) or spillage of material during handling and storage (Pettersson and Nordfjell 2007). Recent studies confirm that dry matter losses related to microbial activity during storage of logwood (1% and 4% per year; Erber et al. 2012, 2016b and 2017) were considerably lower than those of whole trees and logging residues (0% to 24% during storage periods between 1.2 to 20 months; Routa et al. 2015). Residues pre-dried in small heaps in the stand did not lose dry matter, while extensive dry matter losses have been observed for logging residues piled at the roadside immediately after harvesting, which is considered to result from poor ventilation and, therefore, increased biological activity in the large piles. Bundling of logging residues is considered advantageous regarding dry matter losses (Eriksson and Gustavsson 2010, Filbakk et al. 2011b).

However, Petterson and Nordfjell (2007) point out that a rather large share of logging residues dry matter is lost along the fuelwood supply chain, either during bundling or by purposely not recovering logging residues probably contaminated by soil. Nilsson et al.

(2015) revealed that only 50% to 60% of the logging residues at the harvesting site arrive at the energy conversion plant.

Nurmi (2014) advocates the use of volumetric energy density as the criteria for assessing the effect of drying. In his study, both Scots pine and Downy birch whole trees exhibited a reduction of moisture content.

Nonetheless, due to extensive dry matter losses (8.5% to 14.1%), the volumetric energy density of Downy birch dropped considerably (3.4% to 9.6%), even though the moisture content had also plummeted by

273

G. Erber and M. Kühmaier Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten ... (269–278)

10.3% to 15.5% simultaneously. On the contrary, the volumetric energy density of Scots pine ascended by 0.8% to 16.5% during the same period and by 4.8% to 17.6% after 17 months.

3.4.5 Meteorological data based drying models Increased storage levels and thus higher capital and financial costs are a consequence of keeping fuelwood piles in storage too long to ensure that they are sufficiently dry, thus exposing them to increased dry matter losses (Acuna et al. 2012, Sosa et al. 2015). To ensure economic efficiency of fuelwood supply, forest managers must apply systems for tracking the drying performance of their piles (Gautam et al. 2012). Yet, keeping track of pile moisture content by physical sampling at intervals is time consuming, costly and error-prone. Modelling the development of the moisture content of piles is considerably more economical and reliable than »educated guesses« (Erber et al.

2016b). Appropriate meteorological data based models have been developed recently and are considered a valuable tool for resource allocation in fuelwood supply (Routa et al. 2016).

There are models for differing combinations of species, material type and treatment (e.g. covering and splitting). Most are dedicated to logwood (Bown and Lassere 2015, Erber et al. 2012 and 2014, 2016b and 2017, Kim and Murphy 2013, Murphy et al. 2012, Raitila et al. 2015, Visser et al. 2014), while whole tree (Filbakk et al. 2011a) and logging residue (Routa et al.

2016, Filbakk et al. (2011b) are less frequent. Treatments include covering (Murphy et al. 2012, Raitila et al. 2015, Filbakk et al. 2011a) and splitting (Visser et al.

2014).

The joint target variable is either the alteration of moisture content during a defined period or, in reverse, the period required to reach a defined moisture content. Explaining variables are highly diverse. However, meteorological variables are commonly included. Some models (Erber et al. 2014, Kim and Murphy 2013, Murphy et al. 2012, Raitila et al. 2015, Routa et al. 2016) employ a combination of cumulative precipitation and cumulative evapotranspiration (ET0), a parameter derived according to the FAO PenmanMonteith method. Most models estimate on a daily basis (Bown and Lassere 2015, Erber et al. 2012, 2014 and 2016b, Filbakk et al. 2011a and 2011b, Raitila et al.

2015, Routa et al. 2016). Some models (Visser et al.

2014, Kim and Murphy 2013, Murphy et al. 2012) opted for a weekly basis or a 10-day estimation period.

Erber et al. (2017) concluded that the modelling period, apart from the 10-minute basis, does not affect the modelling accuracy to a large degree.

274

All models comply with the ±5% limit for the target accuracy, which is suggested by Erber et al. (2014) and Routa et al. (2016) and derived from discussions with fuelwood suppliers. Erber et al. (2012 and 2014) could show that exceeding a model valid range resulted in an utter overestimation of the drying performance.

Nevertheless, validation against data from field studies is crucial to assess model accuracy (Routa et al.

2016).

Several approaches for scientifically monitoring a fuelwood pile drying performance are available. The »classic« approach is sampling by chainsaw (Filbakk et al. 2011a and 2011b) or weighing single stems or parts of the pile in intervals (Bown and Lassere 2015, Kim and Murphy 2013, Raitila et al. 2015, Visser et al.

2014). Yet, each variant is either associated with extensive workload or frequent disturbance of the experimental design. A more sophisticated approach, termed »continuous weighing approach« avoids these drawbacks (Murphy et al. 2012, Erber et al. 2012, 2014, 2016b and 2017, Raitila et al. 2015, Routa et al. 2016). Its basic principle is to track the drying performance of a fuelwood pile through the alteration of its weight. For this reason, the pile is put onto a metal frame, reminiscent of racks on logwood trucks, which rests on load cells.

By cumulating weight alterations and starting from the physically sampled initial moisture content, the actual moisture content can be estimated. Along with the weight data, meteorological data, necessary for modelling, is recorded (Erber et al. 2012, 2014, 2016 and 2017, Murphy et al. 2012, Routa et al. 2015 and 2016b). Nevertheless, this method does have drawbacks of its own, as extensive dry matter losses and snow cover can jeopardize the estimation.

4. Discussion and Conclusion Increasing the efficiency of fuelwood supply is a topic of research as old as fuelwood supply itself.

Naturally, the focus and trends have undergone changes over time, interacting with emerging technologies and milestones in improvement.

As Stampfer and Kanzian (2006) identified, small tree dimensions are the challenging factor in fuelwood harvesting. A decade later, the challenge remains the same. There is still plenty of room for improvement, even though significant improvement has been made.

Accumulating harvesting heads (multi-tree handling) and dedicated working techniques (boom-corridor thinning) significantly enhance fuelwood harvesting and the most sophisticated heads are equipped with feed rollers and delimbing knives, which enables the production of delimbed, cut-to-length bundles of fuCroat. j. for. eng. 38(2017)2

Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten ... (269–278)

elwood, convenient for forwarding. The suitability of these heads is perfectly illustrated by Belbo (2010), who could show that the degree of accumulation increases with decreasing tree size. However, the utilization of the theoretical accumulation capacity is limited by hampered manoeuvrability of a head full of trees and limited visibility in dense stands. What is more, the operator’s skills remain the deciding factor for the productivity of generally highly sophisticated and costly procurement machinery. For this reason, special attention shall be paid to usability when developing new technological solutions (Routa et al. 2013). Multitree handling capacity and dedicated harvesting techniques increased the productivity of harvesting by 2% to 17%, depending on the study conditions, in comparison to a single-tree handling. However, as stated

by Laitila et al. (2007), it is the forwarding productivity that benefits most (+9% to 17%) from multi-tree handling.

Bundling logging residues and small trees with dedicated machines has been considered as a feasible option to increase the load density for transportation and to increase productivity when chipping for more than a decade (Johansson et al. 2006, Kärhä and Vartiamäki 2006, Ranta and Rinne 2006, Cuchet et al.

2004). However, the effects of bundling have not been found to be as significant as expected. Bundling itself proved to work well. However, the preceding processes remain the bottlenecks in bundle-harvesting.

Even though improvements have been achieved, they severely hamper cost-efficient operation. However, bundle-harvesting is still considered a feasible option for small-diameter wood (Nuutinen et al. 2016) in Nordic countries. On the contrary, in mountainous regions, bundling has been ruled out as a feasible option due to space limitations on the forest roads.

Fuelwood storage and drying have been the subject of research for a long time, as indicated by early publications, as the one of Byram (1940). During the last decade, the mechanisms of storage and drying have been investigated in detail and several treatments to facilitate fuelwood drying have been tested. Covering is feasible, but the extent of its effect depends on the climatic conditions at the storage location. Meteorological data based drying models are a recent research trend, although there has already been work on this topic as early as in the eighties (Stokes et al. 1987 and 1993). However, modern ICT has opened a whole new dimension for these models. Integrated into fuelwood procurement systems, models enable day-to-day monitoring of fuelwood pile drying performance and present a major step in improving forest fuel quality and logistics management. However, more research is reCroat. j. for. eng. 38(2017)2

G. Erber and M. Kühmaier

quired to better understand degradation processes during drying.

In fuelwood harvesting, semi-automated and remotely controlled harvesting machines, also equipped with devices for enhancing vision in dense young stands, can be considered trends for the far future. In the short term, developments will certainly focus on incremental improvements of existing technology (combined harvesting of saw logs, pulpwood and fuelwood), ergonomic innovations and widespread introduction of innovative working techniques (boomcorridor thinning). Linked to the operation of winch-supported fully mechanized harvesting systems in the transition area to cable yarding-only terrain, fuelwood harvesting heads working range will expand, too.

The future of bundling is less certain. With the insolvency of Fixteri Oy, the main proponent and driver of bundling has disappeared. However, as the bundling unit has never been the system bottleneck, a revival of this technology could be fuelled by more productive harvesting heads. In terms of handling and transport economics, bundling can still be considered a feasible option. Contrarily, fuelwood harvesting heads capable of compressing have to be considered a pioneering technology, both in terms of compaction and avoidance of excessive nutrient removal.

In fuelwood storage, research can be expected to focus on improving existing and developing new, more sophisticated drying models, which are capable of addressing degradation effects more accurately. As these models are fuelled by modern ICT and readily available high-resolution meteorological data, they will be introduced in forestry with less reluctance.

Moisture content management is not limited to fuelwood. Sawmills show increasing interest in meteorological models for scheduling the point of transport and sawing. However, the latest calls in research indicate a shift of focus from energetic towards material use, a circumstance that researchers in this field will surely have to consider. Either way, moisture content management will be a topic in the future, regardless of the exact purpose.

5. References Acuna, M., Anttila, P., Sikanen, L., Prinz, R., Asikainen, A., 2012: Predicting and controlling moisture content to optimise Forest biomass logisitcs. Croatian Journal of Forest Engineering 33(2): 225–238.

Afzal, M., Bedane, A., Sokhansanj, S., Mahmood, W., 2010: Storage of comminuted and uncomminuted forest biomass and its effect on fuel quality. BioResources 5(1): 55–69.

275

G. Erber and M. Kühmaier Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten ... (269–278) Badal, T., Kšica, J., Vala, V., Kupčák, V., 2015: The influence of the average monthly temperature and precipitation on cumulative moisture, calorific value and ash of energy chips made from logging residues. Zpravy Lesnickeho Vyzkumu 60(4): 299–308.

Belbo H., 2010: Comparison of two working methods for small tree harvesting with a multi tree felling head mounted on farm tractor. Silva Fennica 44(3): 453–464.

Bergström, D., Di Fulvio, F., 2014a: Comparison of the cost and energy efficiencies of present and future biomass supply systems for young dense forests. Scandinavian Journal of Forest Research 29(8): 793–812.

Bergström, D., Di Fulvio, F., 2014b: Evaluation of a novel prototype harvester head in early fuel-wood thinnings. International Journal of Forest Engineering 25(2): 156–170.

Bergström, D., Bergsten, U., Nordfjell, T., 2010: Comparison of boom-corridor thinning and thinning from below harvesting methods in young dense Scots pine stands. Silva Fennica 44(4): 669–679.

Bergström, D., Bergsten, U., Hörnlund, T., Nordfjell, T., 2012: Continuous felling of small diameter trees in boom-corridors with a prototype felling head. Scandinavian Journal of Forest Research 27(5): 474–480.

Bergström, D., Nordfjell, T., Bergsten, U., 2010: Compression processing and load compression of young Scots pine and birch trees in thinnings for bioenergy. International Journal of Forest Engineering 21(1): 31–39.

Bown, H., Lasserre, J-P., 2015: An air-drying model for piled logs of Eucalyptus globulus and Eucalyptus nitens in Chile.

New Zealand Journal of Forestry Science 45(1): 1–9.

Brand, M., Bolzon de Muñiz, G., Quirino, W., Brito, J., 2011: Storage as a tool to improve wood fuel quality. Biomass and Bioenergy 35(7): 2581–2588.

Brand, M., de Muñiz, G., Quirino, W., Brito, J., 2010: Influence of storage time on the quality of biomass for energy production in humid subtropical regions. Cerne 16(4): 531–537.

Byram, G.M., 1940: Sun and wind and fuel moisture. Journal of Forestry 38(8): 639–640.

Com/2014/015: Communication from the commission to the European parliament, the council, the European economic and social committee and the committee of the regions. A policy framework for climate and energy in the period from 2020 to 2030.

Cuchet, E., Roux, P., Spinelli, R., 2004: Performance of a logging residue bundler in the temperate forests of France. Biomass and Bioenergy 27(1): 31–39.

Cutshall, J., Greene, D.W., Baker, S., 2013: Transpirational drying effects on energy and ash content from whole-tree southern pine plantation chipping operations. Southern Journal of Applied Forestry 37(3): 133–139.

Di Fulvio, F., Kroon, A., Bergström, D., Nordfjell, T., 2011: Comparison of energy-wood and pulpwood thinning systems in young birch stands. Scandinavian Journal of Forest Research 26(4): 339–349.

276

Elber, U., 2007: Feuchtegehalt-Änderungen des Waldfrischholzes bei Lagerung im Wald (Moisture content alteration of fuelwood during storage in the forest). Bern (Switzerland): Swiss Federal Office of Energy, 31 p.

Erber, G., Holzleitner, F., Kastner, M., Stampfer, K., 2016a: Effect of multi-tree handling and tree-size on harvester performance in small-diameter hardwood thinnings. Silva Fennica 50(1): 1–17.

Erber., G., Holzleitner, F., Kastner, M., Stampfer, K., 2017: Impact of different time interval bases on the accuracy of meteorological data based drying models for oak (Quercus L.) logs stored in piles for energy purposes. Croatian Journal of Forest Engineering 38(1): 1–9.

Erber, G., Kanzian, C., Stampfer, K., 2012: Predicting moisture content in a pine logwood pile for energy purposes. Silva Fennica 46(4): 555–567.

Erber, G., Kanzian, C., Stampfer, K., 2016b: Modelling natural drying of European beech (Fagus sylvatica L.) logs for energy based on meteorological data. Scandinavian Journal of Forest Research 31(3): 294–301.

Erber, G., Routa, J., Kolström, M., Kanzian, C., Sikanen, L., Stampfer, K., 2014: Comparing two different approaches in modeling small diameter energy wood drying in logwood piles. Croatian Journal of Forest Engineering 35(1): 15–22.

Eriksson, L., Gustavsson, L., 2010: Comparative analysis of wood chips and bundles – Costs, carbon dioxide emissions, dry-matter losses and allergic reactions. Biomass and Bioenergy 34(1): 82–90.

Filbakk, T., Hoibo, O., Nurmi, J., 2011a: Modelling natural drying efficiency in covered and uncovered piles of whole broadleaf trees for energy use. Biomass and Bioenergy 35(1): 454–463.

Filbakk, T., Hoibo, O., Dibdiakova, J., Nurmi, J., 2011b: Modelling moisture content and dry matter loss during storage of logging residues for energy. Scandinavian Journal of Forest Research 26(3): 267–277.

Gautam, S., Pulkki, R., Shahi, C., Leitch, M., 2012: Fuel quality changes in full tree logging residue during storage in roadside slash piles in Northwestern Ontario. Biomass and Bioenergy 42: 43–50.

Gautam, S., Pulkki, R., Shahi, C., Leitch, M., 2013: Quality assessment of cut-to-length logging residues for bioenergy production in Northwestern Ontario. International Journal of Forest Engineering 24(1): 53–59.

Ghaffariyan, M.R., Brown, M., Acuna, M., Sessions, J., Gallagher, T., Kühmaier, M., Spinelli, R., Visser, R., Devlin, G., Eliasson, L., Laitila, J., Laina, R., Iwarsson Wide, M., Egnell, G., 2017: An international review of the most productive and cost effective forest biomass recovery technologies and supply chains. Renewable and Sustainable Energy Reviews 74: 145–158.

Ichihara, T., Takano, S., Yamasaki, T., Masaoka, H., Itai, T., Noji, K., Matsuoka, Y., Kobatake, A., Suzuki, Y., Fujiwara, S., 2010: Transpirational drying of stacked logging residue logs for wood fuel chips. Nihon Ringakkai Shi/Journal of the Japanese Forestry Society 92(4): 191–199.

Croat. j. for. eng. 38(2017)2

Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten ... (269–278)

G. Erber and M. Kühmaier

Johansson, J., Liss, J., Gullberg, T., Bjorheden, R., 2006: Transport and handling of forest energy bundles – advantages and problems. Biomass and Bioenergy 30(4): 334–341.

Manzone, M., 2016: A bundler prototype for forestry and agricultural residue management for energy production. International Journal of Forest Engineering 27(2): 103–108.

Jylhä, P., Bergström, D., 2016: Productivity of harvesting dense birch stands for bioenergy. Biomass and Bioenergy 88: 142–151.

Murphy, G., Kent, T., Kofman, P., 2012: Modeling air drying of Sitka spruce (Picea sitchensis) biomass in off-forest storage yards in Ireland. Forest Products Journal 62(6): 443–449.

Jylhä, P., Laitila, J., 2007: Energy wood and pulpwood harvesting from young stands using a prototype whole-tree bundler. Silva Fennica 41(4): 763–779.

Kärhä, K., Vartiamäki, T., 2006: Productivity and costs of slash bundling in Nordic conditions. Biomass and Bioenergy 30(12): 1043–1052.

Kärhä, K., Jylhä, P., Laitila, J., 2011: Integrated procurement of pulpwood and energy wood from early thinnings using whole-tree bundling. Biomass and Bioenergy 35(8): 3389– 3396.

Kim, D-W., Murphy, G., 2013: Forecasting air-drying rates of small Douglas-fir and hybrid poplar stacked logs in Oregon, USA. International Journal of Forest Engineering 24(2): 137– 147.

Klepac, J., Mitchell, D., Thompson, J., 2014: The effect of pile size on moisture content of loblolly pine while field drying.

In: Proceedings of the 37th Council on Forest Engineering Annual Meeting at Moline, Illinois, 9 p.

Klepac, J., Rummer, B., Seixas, F., 2008: Seasonal effect on moisture loss of loblolly pine. In: Proceedings of the 31st Council on Forest Engineering Annual Meeting at Charleston, South Carolina, 9 p.

Kofman, P.D., Kent, T., 2009a: Forest storage and seasoning of conifer and broadleaf whole trees. COFORD connects – Harvesting/Transportation 18, 4 p.

Kofman, P.D., Kent, T., 2009b: Long term storage and seasoning of conifer energy wood. COFORD connects – Harvesting/ Transportation 20, 4 p.

Nilsson, B., Blom, T., Thörnqvist, T., 2013: The influence of two different handling methods on the moisture content and composition of logging residues. Biomass and Bioenergy 52: 34–42.

Nilsson, B., Nilsson, D., Thörnqvist, T., 2015: Distributions and losses of logging residues at clear-felled areas during extraction for bioenergy: Comparing dried- and fresh-stacked method. Forests 6(11): 4212–4227.

Nurmi, J., Hillebrand, K., 2007: Storage alternatives affect fuelwood properties of Norway spruce logging residues.

New Zealand Journal of Forestry Science 31(3): 289–297.

Nurmi, J., Hillebrand, K., 2007: The characteristics of wholetree fuel stocks from silvicultural cleanings and thinnings.

Biomass and Bioenergy 31(6): 381–392.

Nurmi, J., Lehtimäki, J., 2011: Debarking and drying of downy birch (Betula pubescens) and Scots pine (Pinus sylvestris) fuelwood in conjunction with multi-tree harvesting. Biomass and Bioenergy 35(8): 3376–3382.

Nurmi, J., 2014: Changes in volumetric energy densities during storage of whole-tree feed stocks from silvicultural thinnings. Biomass and Bioenergy 61: 114–120.

Nuutinen, Y., Björheden, R., 2016: Productivity and work processes of small-tree bundler Fixteri FX15a in energy wood harvesting from early pine dominated thinnings. International Journal of Forest Engineering 27(1): 29–42.

Laitila, J., Väätäinen, K., 2013: The cutting productivity of the excavator-based harvester in integrated harvesting of pulpwood and energy wood. Baltic Forestry 19(2): 289–300.

Nuutinen, Y., Kärhä, K., Laitila, J., Jylhä, P., Keskinen, S., 2011: Productivity of whole-tree bundler in energy wood and pulpwood harvesting from early thinnings. Scandinavian Journal of Forest Research 26(4): 329–338.

Laitila, J., Asikainen, A., Nuutinen, Y., 2007: Forwarding of whole trees after manual and mechanized felling bunching in pre-commercial thinnings. International Journal of Forest Engineering 18(2): 29–39.

Nuutinen, Y., Petty, A., Bergström, D., Rytkönen, M., Di Fulvio, F., Tiihonen, I., Lauren, A., Dahlin, B., 2016: Quality and productivity in comminution of small-diameter tree bundles.

International Journal of Forest Engineering 27(3): 179–187.

Laitila, J., Heikkilä, J., Anttila, P., 2010: Harvesting alternatives, accumulation and procurement cost of small-diameter thinning wood for fuel in Central Finland. Silva Fennica 44(3): 465–480.

Laitila, J., Kilponen, M., Nuutinen, Y., 2013: Productivity and cost-efficiency of bundling logging residues at roadside landing. Croatian Journal of Forest Engineering 34(2): 175–187.

Laitila, J., Niemistö, P., Väätäinen, K., 2016: Productivity of multi-tree cutting in thinnings and clear cuttings of young downy birch (Betula pubescens) dominated stands in the integrated harvesting of pulpwood and energy wood. Baltic Forestry 22(1): 116–131.

Manzone, M., 2015: Energy and moisture losses during poplar and black locust logwood storage. Fuel Processing Technology 138: 194–201.

Croat. j. for. eng. 38(2017)2

Ovaskaenen, H., Palander, T., Jauhlaenen, M., Lehtemäki, J., Tekkanen, L., Nurmi, J., 2008: Productivity of energywood harvesting chain in different stand conditions of early thinnings. Baltic Forestry 14(2): 149–154.

Pettersson, M., Nordfjell, T., 2007: Fuel quality changes during seasonal storage of compacted logging residues and young trees. Biomass and Bioenergy 31(11–12): 782–792.

Petty, A., Kärhä, K., 2014: Productivity and cost evaluations of energy-wood and pulpwood harvesting systems in first thinnings. International Journal of Forest Engineering 25(1): 37–50.

Raitila, J., Heiskanen, V-P., Routa, J., Kolström, M., Sikanen, L., 2015: Comparison of moisture prediction models for stacked fuelwood. BioEnergy Research 8(4): 1896–1905.

277

G. Erber and M. Kühmaier Research Trends in European Forest Fuel Supply Chains: a Review of the Last Ten ... (269–278) Ranta, T., Rinne, S., 2006: The profitability of transporting uncomminuted raw materials in Finland. Biomass and Bioenergy 30(3): 231–237.

Röser, D., Mola-Yudego, B., Sikanen, L., Prinz, R., Gritten, D., Emer, B., Väätäinen, K., Erkkilä, A., 2011: Natural drying treatments during seasonal storage of wood for bioenergy in different European locations. Biomass and Bioenergy 35(10): 4238–4247.

Rottensteiner, C., Affenzeller, G., Stampfer, K., 2008: Evaluation of the feller-buncher moipu 400E for energy wood harvesting. Croatian Journal of Forest Engineering 29(2): 117– 128.

Routa, J., Asikainen, A., Björheden, R., Laitila, J., Röser, D., 2013: Forest energy procurement: State of the art in Finland and Sweden. Wiley Interdisciplinary Reviews: Energy and Environment 2(6): 602–613.

factured felling heads in Italy. Biomass and Bioenergy 72: 85–94.

Sosa, A., Acuna, M., McDonnell, K., Devlin, G., 2015: Managing the moisture content of wood biomass for the optimisation of Ireland’s transport supply strategy to bioenergy markets and competing industries. Energy 86: 354–368.

Spinelli, R., Cuchet, E., Roux, P., 2007: A new feller-buncher for harvesting energy wood: Results from a European test programme. Biomass and Bioenergy 31(4): 205–210.

Spinelli, R., Magagnotti, N., Picchi, G., 2012: A supply chain evaluation of slash bundling under the conditions of mountain forestry. Biomass and Bioenergy 36: 339–345.

Stampfer, K., Kanzian, C., 2006: Current state and development possibilities of wood chip supply chains in Austria.

Croatian Journal of Forest Engineering 27(2): 135–145.

Routa, J., Kolström, M., Ruotsalainen, J., Sikanen, L., 2015: Precision measurement of forest harvesting residue moisture change and dry matter losses by constant weight monitoring.

International Journal of Forest Engineering 26(1): 71–83.

Stokes, B.J., McDonald, T.P., Kelley, T., 1993: Transpirational drying and costs for transporting wood biomass: a preliminary review. In: Proceedings of IEA/BA Task IX, Activity 6: Transport and Handling; Aberdeen University, UK: 76–91.

Routa, J., Kolström, M., Ruotsalainen, J., Sikanen, L., 2016: Validation of prediction models for estimating the moisture content of logging residues during storage. Biomass and Bioenergy 94: 85–93.

Stokes, B.J., Watson, W.F., Miller, D.E., 1987: Transpirational drying of energywood. ASAE paper No. 87–1530. St. Joseph, Michigan: American Society of Agricultural Engineers, 13 p.

Sängstuvall, L., Bergström, D., Lämås, T., Nordfjell, T., 2012: Simulation of harvester productivity in selective and boomcorridor thinning of young forests. Scandinavian Journal of Forest Research 27(1): 56–73.

Schweier, J., Spinelli, R., Magagnotti, N., Becker, G., 2015: Mechanized coppice harvesting with new small-scale fellerbunchers: Results from harvesting trials with newly manu-

Visser, R., Berkett, H., Spinelli, R., 2014: Determining the effect of storage conditions on the natural drying of radiata pine logs for energy use. New Zealand Journal of Forestry Science 44(1): 1–8.

Yukio, T., Tomohiro, G., 2011: Natural-drying process of Sugi logging residues in a field under different seasons and conditions of waterproof coverage. Nihon Ringakkai Shi/Journal of the Japanese Forestry Society 93(6): 262–269.

Authors’ addresses:

Received: March 7, 2017 Accepted: June 2, 2017

278

Gernot Erber, PhD. * e-mail: gernot.erber@boku.ac.at Kühmaier Martin, PhD. e-mail: martin.kuehmaier@boku.ac.at University of Natural Resources and Life Sciences Peter Jordan Strasse 82 1190 Wien AUSTRIA * Corresponding author Croat. j. for. eng. 38(2017)2

Subject review

Timber and Biomass Transport Optimization: A Review of Planning Issues, Solution Techniques and Decision Support Tools Mauricio Acuna Abstract The transport of timber and biomass represents a significant proportion of the operational cost for the forest industry worldwide. This calls for optimization solutions so that companies can organize their transport operations and allocate resources efficiently, and reduce the impact of transport operations on the environment. This paper presents an extensive overview of the transport and biomass optimization problem in the forest industry. It provides a detailed description of mathematical solutions, including linear programming models and algorithms, to solve complex transportation planning problems involving annual, monthly and daily decisions. Also, the paper presents and describes two decision support tools, MCPLAN and FastTRUCK, which have been implemented to assist transport planners to optimize the flows of timber and biomass from the forest to mills and energy plants, and to schedule and route the trucks efficiently between these supply and demand points.

Keywords: timber and biomass transport optimization, transport operations, wood flows, linear programming, log-truck scheduling, simulated annealing, heuristics

1. Introduction Transportation from forestry harvest areas to mills costs the forestry industry millions of dollars annually, accounting for up to half of the operational costs in forestry supply chains (McDonald et al. 2001, Acuna 2011, Audy et al. 2012). Given the level of spending on transportation costs, even small increases in efficiency can reduce costs substantially (Rรถnnqvist et al. 1998, Palmgren 2001, Palmgren et al. 2004). It is, therefore, important to optimize deliveries from the forest to the mill, and organize the scheduling of trucks efficiently as they have a high initial investment cost in their purchase, and high operational cost over their daily operations and ownership lifetime. Operations Research (OR) models can significantly improve decisions, as has been shown in real experiences, in particular for daily scheduling. Improvements in technology and communications drive the development of systems based on OR algorithms, which support transportation decisions (Weintraub et al. 2007).

In forestry, several modes of transportation are used: truck, train, ship and water (Epstein et al. 2007).

Croat. j. for. eng. 38(2017)2

As timber and biomass are mainly carried by trucks for parts of the transportation, the focus of this paper is on the transport of roundwood (logs delivered to pulp mills and sawmills) and biomass (chipped or packed) by trucks. It is also assumed that a forest company organizes a fleet of trucks using centralized scheduling and dispatching, with trucks belonging to one or several haulage contractors. This approach has been proposed not only to reduce costs and improve the tactical coordination of trucks and loads but also to reduce the impact of transport operations on the environment resulting from reduced emissions, traffic congestions, and noise and dust on forest roads (Gronalt and Hirsch 2007, Haridass 2009, Acuna et al. 2011).

From a planning perspective, the paper focuses on problems arising in annual, monthly and daily time periods, describing the planning problems/issues, and presenting mathematical solutions and decision support tools to optimize the transport of timber and biomass. The document is organized as follows: Section 2 describes the harvest and transport planning in the timber and biomass industry including major decisions at each planning level; Section 3 describes the

279

M. Acuna

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279–290)

timber and biomass transportation problem including basic mathematical formulations; Section 4 presents a review of some mathematical and algorithmic solution methods for the timber and biomass transport optimisation problem; Section 5 describes two decision support systems developed by the author of this paper, including a description of the tools and mathematical solutions implemented. The final section offers some concluding remarks.

2. Harvest and transport planning in timber and biomass industry Forest management planning is usually performed using a top-down hierarchical approach, where decisions regarding harvest and transport, are made at different levels and planning horizons: strategical, tactical, and short-term operational (Palmgren 2001, Rönnqvist 2003, D’Amours et al. 2008). There is an increased level of desegregation on the spatial and non-spatial information used from the strategic to the operational plans, and decisions made at higher levels become input constraints to lower levels plans (Bettinger et al. 2008). Church (2007) points out that this hierarchical approach is adopted due to the complexity associated with capturing all elements and implementing forest planning decisions and constraints in a single model.

The strategic level often describes broad-scale planning decisions over an extended period, which can range from several years (e.g. plantations in the Southern Hemisphere) to several decades (e.g. natural forests in the Northern Hemisphere). Regarding transport and logistics, the plan includes decisions related to facility building (e.g. location and opening/closing of mills, stockyards, and fuel terminals), inventory location and levels, procurement strategies (e.g. push, pull), transport technology, resources and investment (e.g. road and rail networks, transport equipment, and port infrastructure), and supply chain network design (D’Amours et al. 2008).

In the next level of the hierarchical forest planning approach, tactical planning often is linked to longterm strategic planning described above, and shortterm operational planning. Decisions at a tactical level span between 1 and 30 years, embracing the spatial allocation and scheduling of forest resources to forest operations. Connections with operational planning have impacts on transport and harvest operations involving annual, monthly and daily decisions such as timber and biomass flows, truck scheduling and dispatching, and allocation of cutting patterns and logging crews to harvest coupes. Transport decisions at a

280

tactical level include allocation of harvest units and catchment areas to customers, allocation of transport resources including trucks and drivers, and fleet management. Other typical tactical decisions concern allocating customers to mills and defining the necessary distribution capacity (D’Amours et al. 2008).

The last level to be mentioned is short-term operational planning, which involves monthly and daily decisions. Operational decisions in transportation deal with truck scheduling and dispatching, to optimally establish a set of routes and schedule that a fleet of trucks has to perform to deliver timber (log products) from wood pickup locations (harvest areas) to customer sites (mills). Scheduling means to plan the entire fleet’s route in advance (typically one day ahead), listing every pickup and delivery of timber products. An estimation of the stocks of logs at the pickup locations and demand at the customer sites is required to come up with a good plan. In practice, it is quite difficult to run a log truck fleet according to a pre-arranged schedule. Unexpected delays, breakdowns, and queues mean some logs are delivered late or not at all, all of which disturb the planned schedule. In such cases, it is essential to have a dispatcher on hand to repair the schedule in cooperation with truck drivers and transport coordinators (Palmgren 2001).

3. Timber and biomass transportation problem In this chapter, the focus is on two transportation problems, one that comprises annual and monthly decisions (timber and biomass flow, and backhauling), and another that involves daily decisions (truck scheduling).

3.1 Timber and biomass flow and backhauling The timber and biomass flow problem balances demand at consumption centers against supply from the forests, whereas backhauling extends the network flow problem where routes are included to explicitly consider efficient routing (Epstein et al. 2007).

Fig. 1 Network structure for timber and biomass flow problem Croat. j. for. eng. 38(2017)2

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279–290)

M. Acuna

Fig. 1 presents the network structure of the wood flow problem, assuming two generic products, roundwood and biomass. A forest site can supply just roundwood or both products, whereas a destination point can demand one or both products. In the latter case, the mill (e.g. a pulp mill) and the energy plant are part of the same industrial complex and have the same geographic location.

Assuming multiple products (e.g. roundwood and biomass), and multiple supply and demand points, the optimization flow model can be stated as:

∑ ∑ ∑ Cijk Xijk

(1)

∑ Xijk ≤ Sik ,∀ i ∈I, k ∈K

(2)

∑ Xijk = Dik ,∀j ∈J, k ∈K

(3)

Xijk ≥ 0,∀i ∈ I , j ∈ J , k ∈K

(4)

Min =

i∈I j∈J k ∈K

subject to

j∈J

i∈I

In this mathematical formulation, the index sets I, J, and K represent the set of supply points (harvest areas), demand points (wood mill or energy plant), and products (roundwood or biomass), respectively.

The flow of product k from supply point i to demand point j is represented by the decision variable Xijk.

The objective function minimizes total wood and biomass flow costs, where Cijk represents the flow costs of product k, between supply point i and demand point j. Constraint (1) ensures that the flow of products does not exceed the capacity of the supply points. Constraint (2) represents that the demand at the consumption centers (e.g. mill and energy plant) is equal to the supply from forest areas, whereas constraint (3) establishes the non-negativity of decision variable Xijk. The formulation is an extension of the transportation problem (Taja 2016), and corresponds to a linear programming (LP) model, which can be solved quickly and efficiently using any commercial LP solver. This basic flow model is the basis for other more complex models, for example, when only a subset of the supply points meets the demand of one of the mills or energy plants. Also, the model can contain temporary (e.g. monthly) decisions, with additional constraints that ensure that the demand for roundwood and biomass is satisfied in each planning period.

As shown by Epstein (2007) and Sosa et al. (2015), sometimes these flow models can be used to decide the catchment areas and products allocated to each Croat. j. for. eng. 38(2017)2

Fig. 2 Flow in two directions (left) and a backhauling tour (right) industry monthly or weekly. These areas represent supply points that deliver roundwood and biomass to one or more customers.

In the timber and biomass problem, it is assumed that a truck runs loaded from a wood pickup point to a customer (demand) point and empty in the other direction (left side of Fig. 2). Thus, the distance traveled loaded is just half of the total distance traveled.

This efficiency metrics is referred as to the running loaded percentage, which in this example is equal to 50%. It can be higher if routes involving several loaded trips are performed, that is, backhauling, which reduces the unloaded distance run by the truck (right side of Fig. 2).

Back-haulage flows may dramatically reduce trans­ port costs, with savings between 2% and 20% being reported in different studies (Carlsson and Rönnqvist 1998, Forsberg et al. 2005). As indicated by Epstein et al. (2007), the possibilities for backhauling are dependent on the type of transportation and the geographical distribution of mills and harvest areas and requires wood flows going in opposite directions.

3.2 Truck scheduling The daily transportation problem to be solved considers a heterogeneous fleet of trucks (T) that need to be optimally allocated to different transport tasks. The main objective of this optimal truck allocation is to minimize total unloaded travel times (TU) and waiting times (WT) at delivery points (wood pickup locations), assuming that a central transportation system provides schedules for trucks that maximize their utilization.

An essential input to solve the truck scheduling mathematical model corresponds to the daily transport tasks. These tasks are predefined with the assistance of a wood flow optimization model as the one described in the previous section or a short-term harvest model. A task corresponds to a truckload or transport order that must be performed to obtain a feasible solution.

281

M. Acuna

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279–290)

The problem considers the following attributes associated with transport tasks, trucks, and tours; Tasks:  a transport task starts at a wood pickup location and ends at a customer site (mill)  each transport task has to be fulfilled  each task is characterized by a wood pickup location (forest coupe or log landing), destination (mill), product, loading time, travel loaded, unloading time, and time window at the wood pickup location  some tasks can only be performed by certain trucks. Some trucks have limited or no access to certain wood pickup points, or they are not able to carry some wood products (logs of different specifications).

Trucks:  a heterogeneous fleet of logging trucks of the same capacity is assumed  trucks are based at different depots. One specific depot can be the daily starting point of one or several trucks  trucks are unloaded at the customer sites based on a first-in, first-out servicing. A truck follows a queue if there is one at the wood pickup point.

To maximize truck utilization, waiting times of trucks in queue is not wanted and must be minimized or eliminated completely  each truck has a maximum shift time that must be met. The shift time includes travel empty, travel loaded, waiting and loading times at wood pickup points, and unloading times at customers.

Tours:  a truck always starts its daily tour empty at the corresponding depot  after leaving the depot, a truck visits a wood pickup location where it is fully loaded. Neither partial loads nor transport between different wood pickup points is allowed  after delivering and unloading the load at the customer, the truck can continue the tour by performing another transport task (provided that remaining loads are to be carried from that particular wood pickup point to any mill) or returning to its depot  during one daily tour, a truck can visit one or more wood pickup points or customers, which in turn, can be visited by several trucks throughout the day  a tour must satisfy constraints regarding shift time and time windows at depots and pickup points.

282

Fig. 3 Small truck scheduling problem A small example (Fig. 3) illustrates the problem described above, which includes 11 transport tasks, 2 truck depots, 7 wood pickup points, 3 customer sites, and two products. Table 1 presents details of the tasks including wood pickup point, customer site, and product. It is worth noting that two or more transport tasks can share the same wood pickup point (e.g. tasks 2 and 11), the same customer site (e.g. tasks 9, 10, and 11), or the same wood pickup point and customer site (tasks 7 and 8). Also, a customer can demand just a single product (e.g. customers 1 and 2), or two different products (e.g. customer 3). Each product is defined by a different log length combination.

The scheduling planning problem described above is referred in the literature as to the timber transport vehicle routing problem (TTRVP) (Gronalt and Hirsch 2007). Multiple depots can be added to the problem as seen in Fig. 3, and the new formulation is known as the Multiple Depot Vehicle Routing Problem with Pickup and Delivery, and Time Windows (MDVRPPDTW) (Oberscheider et al. 2013). Palmgren (2001) Table 1 Description of transport tasks for the small truck scheduling problem shown in Fig. 4 Task

Wood pickup location, W

Customer site, C

Product

1

1

2

1

2

2

2

1

3

3

2

1

4

4

2

1

5

5

3

2

6

6

3

2

7 and 8

7

3

1

9

7

1

2

10

1

1

2

11

2

1

2

Croat. j. for. eng. 38(2017)2

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279â&#x20AC;&#x201C;290)

points out that the major differences that set the log truck scheduling problem apart from the VRPTW and PDPTW are the number of coupling constraints combinations and the possibility of visiting the same location several times during the same day. Thus, a truck can drive between several pickup and delivery points during its entire daily route. Palmgren (2001) also points out that in the truck scheduling problem a customer can be visited several times during the day until its demand is satisfied. Other specific features of the truck scheduling problem include the specific type of products being demanded by the mills, and the possibility that not all trucks can serve all pickup or delivery nodes.

One mathematical formulation for the above problem is presented by Acuna and Sessions (2015). The mathematical model corresponds to an adaptation of the formulation presented by Gronalt and Hirsch (2007) and Oberscheider et al. (2013), and it is formulated as a standard MDVRPPDTW problem.

4. Solution methods for timber and biomass transport optimization problem Several solution approaches have been proposed to solve transport problems arising in annual, monthly and daily time periods (for a comprehensive review see Audy et al. 2012, Acuna and Sessions 2014, and Devlin et al. 2016).

In Sweden, Skogforsk developed a system called FlowOpt (Forsberg et al. 2005), which integrates a Geospatial Information System (GIS) with a database and uses a heuristic approach based on a Tabu Search algorithm. A major goal of the system is to facilitate the planning process of managers in forest companies.

Shen and Sessions (1989) proposed a network-based method to generate a daily truck schedule that meets a mill delivery program with multiple time windows.

Rey et al. (2009) propose a column generation method in which each column corresponds to one feasible route for a truck. These models are based on generalized set partitioning models or general column-based Mixed Integer Programming (MIP) models. Several authors have used metaheuristics to solve the truck scheduling problem. Gronalt and Hirsch (2007) propose different strategies of the Tabu Search (TS) heuristic, while Rummukainen et al. (2009) propose a method that includes a TS heuristic to create routes and a full truckload-size request, in combination with a MIP model to allocate truckloads to demand sites.

Also, McDonald et al. (2010) proposed a simulated annealing (SA) method to generate a daily truck schedule to deliver a set of requests.

Croat. j. for. eng. 38(2017)2

M. Acuna

Previous work has considered the problem of optimized scheduling of log trucks and thus some truck scheduling and dispatching systems for commercial operations have been developed. In Chile, for example, a computerized system called ASICAM has been in use since 1990. It uses simulation with an embedded heuristic to produce a complete trucking schedule for one day of operations for more than 100 trucks (Epstein et al. 2007). The implementation of ASICAM in real operations has led to reductions in costs between 10% and 20% (Weintraub et al. 1996). Similar systems can be found in other countries. In Finland, a system called EPO/KUORMA based on the Tabu Search algorithm was developed to deal with all stages of planning, from strategic to operational (Palmgren et al.

2004). Forest companies have reported savings of about 5% in wood transport costs after the implementation of the system. In Sweden, Skogforsk developed a system called FlowOpt (Forsberg et al. 2005), which integrates Geospatial Information Systems (GIS) with a database and uses a heuristic approach based on a Tabu Search algorithm. In Canada, a system named MaxTour was developed to compute the potential in back-haulage tours within the volume of one or several types of products, based on an adaptation of the well-known savings heuristic of Clarke and Wright (Marier et al. 2007). Several other systems and tools have been developed to solve transport problems involving tactical and daily decisions. These include, among others, the Trimble Forest Logistics and Optimization (FLO) system (USA), the Asset Forestry WSX Logistics system (New Zealand), the DESCARTES Route Planner (Canada), the SEON Compass Logistics system (formerly ESRI ArcLogistics), the TRUCK SCHEDULER system (Canada), and the ORTEC routing optimization system (USA). A more detailed description of these systems is given in Audy et al. (2012).

5. Decision support tools to optimize timber and biomass transport This chapter presents two decision support tools developed by the author of the paper, one that comprises annual and monthly decisions (MCPLAN), and another that involves daily truck scheduling decisions (FastTRUCK).

5.1 MCPLAN MCPLAN is a tool developed and designed to optimize timber and biomass supply chains (Acuna et al.

2017). It provides spatial and temporal solutions, which include volumes to be harvested per period and supply point, drying times for roundwood and bio-

283

M. Acuna

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279–290)

mass, and flows from supply to demand points. It builds on and extends a previous tool named BIOPLAN, which has been used to investigate alternative biomass supply chains in Europe and Australia (Acuna et al. 2012a, Acuna et al. 2012b). MCPLAN is a tool that runs a linear programming model implemented with Visual BasicTM macros and solved with the What’s bestTM solver add-in for MS ExcelTM (Fig. 4).

MCPLAN uses moisture content (MC) curves as the driving factor for the optimization of supply chain costs, and the optimization tool is typically used to investigate the effect of MC on storage, chipping and transportation costs of roundwood and biomass delivered to mills and energy plants under different MC, operational, and drying scenarios. Geographical Information Systems (GIS) are used to get some of the inputs of MCPLAN. Some of these parameters include the availability of roundwood and biomass per supply area, the geographical location of supply and demand points, and the average transportation distances from supply areas to demand points. Centroids from supply areas are used to calculate the distance between supply and demand points (Sosa et al. 2015).

A basic formulation of the optimization model implemented in MCPLAN considers two products (roundwood and biomass), multiple supply points and demand destinations for both products (pulp mill and energy plant in the same location). Decisions on how much volume of roundwood and residues to be harvested and collected are made on a monthly basis (24 periods). Roundwood and residues are stacked at the roadside to reduce their transportation costs to the pulp mill and energy plant, which results from moving materials with a reduced MC. In MCPLAN, storage of these materials at the roadside is allowed for up to 24 months, but the user can modify this nominal

period that best suits the problem under consideration. The optimal drying period is provided by the optimal solution of the linear programming model and does not exceed the maximum nominal drying period established in the model’s formulation. It is important to mention that MCPLAN can be easily extended to include roundwood for sawmills, as well as dry matter content and volumetric losses, and losses resulting from fire, pests, etc.

It is also assumed that the residues can be stacked at the roadside for an equal or shorter period than the roundwood. After the storage period, the roundwood is transported to the pulp mill where it is chipped using a static chipper, whereas the residues are chipped at the roadside with a mobile chipper and then transported to the energy plant. Both roundwood and chips from residues are consumed during the same month in which they arrive at the energy plant and, therefore, there are no costs associated with the storage of roundwood at the pulp mill or chips from residues at the energy plant.

Also, it is assumed that the pulp mill and energy plant demand a monthly volume of roundwood and chips from residues during the production year (Year 2 for modeling purposes). However, both roundwood and residues may be harvested and stacked at the roadside for drying as from period 1 of Year 1. Thus, the optimal solution specifies for each supply area, when and how much volume of roundwood to harvest, and for how long to stack the roundwood before being delivered to the pulp mill. Likewise, it specifies how much volume of residues to collect and stack for later chipping at the roadside and delivery to the energy plant.

The model displays the results in a series of matrices including among others:  solid volume and tonnes of roundwood to be harvested and residues to be collected in each supply area and period  drying times for roundwood and residues in each supply area. These are generated from drying models that include season, geographical location, local conditions, tree species, and tree dimensions, among others  loose volume of chips from roundwood produced at the pulp mill  loose volume of chips from residues produced at the roadside in each supply area and period

Fig. 4 Graphical user interface of MCPLAN in MS Excel

284

 volume of roundwood and chips from residues delivered to the pulp mill and energy plant from each supply area Croat. j. for. eng. 38(2017)2

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279–290)

M. Acuna

ed in Equation 5, whereas the constraints are listed in Equations 6 to 12.

The objective function presented in Equation 5 minimizes total supply chain costs including the following components:

 number of truckloads delivered to the pulp mill and energy plant  energy content of chips produced from residues arriving at the energy plant  harvesting, forwarding, chipping, storage, and transportation costs.

 transport of roundwood and chips from residues  harvesting and extraction of roundwood and residues to roadside  drying of roundwood and residues at the roadside

The sets, parameters, and variables used in the mathematical model implemented in MCPLAN area are listed in Table 2. The objective function is present-

Table 2 Sets, parameters, and variables used in MCPLAN’s model formulation Term

Definition

Set i, j = periods

Periods, i ÎI = {1... 24}, j ÎJ = {13... 24}

S

Supply areas, s ÎS Parameters

BF

Biomass factor – ratio between solid volume of residues and solid volume of roundwood

LVF

Loose volume factor – loose volume of chips produced from 1 m3 solid of residues

SCs

Solid volume availability of roundwood in supply area s (m3s)

RWDj

Demand for roundwood in production period j (m3s)

ECij

Energy content of chips produced in period j from residues stacked at the roadside from period i (energy unit per loose volume of chips, e.g. MWh/m3l)

EDj

Demand of energy plant in period j (energy unit, MWh)

EXTECj

Energy from external suppliers to be purchased in period j (energy unit, e.g. MWh)

MCRWij

Moisture content of roundwood stacked at the roadside from period i to period j (%)

MCREij

Moisture content of residues stacked at the roadside from period i to period j (%)

MinMCRW, MinMCRE

Minimum moisture content of roundwood and residues delivered to pulp mill and energy plant, respectively (%)

MaxMCRW, MaxMCRE

Maximum moisture content of roundwood and residues delivered to pulp mill and energy plant, respectively (%)

CTRWijs, CTWijs

Transport cost for roundwood and chips from residues stacked at the roadside in supply area s and period i and delivered to the pulp mill and energy plant in period j, respectively ($/m3s for roundwood and $/m3l for chips from residues)

CHRWis, CHREis

Harvest and extraction cost for roundwood and residues produced in period i and supply area s ($/m3s)

CDRWij, CDREij

Drying cost for roundwood and residues stacked at the roadside from period i to period j ($/m3s)

CCHRWij, CCHREij

Chipping cost for roundwood and residues stacked at the roadside from period i to period j ($/m3s for roundwood and $/m3l for residues)

CEC

Energy cost purchased to external suppliers ($/MWh) Variables

Xijs

Decision variable. Solid volume of roundwood harvested in period i, and dried at the roadside until period j in supply area s (m3s)

Yijs

Xijs × BF Solid volume of residues harvested in period i, and dried at the roadside until period j in supply area s (m3s)

Y 'ijs

Yijs × LVF Loose volume of chips produced from residues harvested in period i, and dried at the roadside until period j in supply area s (m3l)

Croat. j. for. eng. 38(2017)2

285

M. Acuna

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279–290)

 chipping costs of roundwood (at the mill), and residues (at the roadside)  purchase of energy units delivered by external providers to the energy plant.

Also, the objective function includes a few assumptions: 1. Collection and extraction of roundwood and residues to roadside occur at the time of harvest; this cost varies by supply area and depends on the MC of roundwood and residues, 2. Drying cost only depends on the length of storage of roundwood and residues at the roadside, 3. Chipping cost depends on the MC of roundwood and residues. The assumption here is that chipping of materials with a reduced MC increase blades tear-out, which consequently, increases their maintenance and replacement costs.

Min =

∑ ∑ ∑ Xijs × ((CTRWijs + CHRWis + CDRWij + CCHRWij ) + i

+

s

∑ ∑ ∑Yijs × (CHREis + CDREij + C CHREij ) + ∑ ∑ ∑Y ’ijs × CTWijs + i

+

j

j

s

i

j

s

∑ EXTEC j × CEC

(5)

j

Equation 6 specifies that, in each supply area, the residues can be stacked at the roadside for an equal or shorter period than the roundwood. The equation also ensures that the solid volume of residues stacked at the roadside does not exceed the availability of residues in the supply area.

Xijs × BF ≥

∑Yijs ∀i ∈I ,s ∈S

(6)

j≤i

Equation 7 ensures that the capacity of roundwood in each supply area is not exceeded.

∑ ∑ Xijs ≤ SCs ∀s ∈S

i

(7)

j

Equation 8 ensures that the demand for roundwood at the pulp mill is met in each period.

∑ ∑ Xijs = RWDj ∀ j ∈J

i

(8)

s

Equation 9 ensures that the demand for energy at the energy plant is met in each period. The energy supply includes deliveries from harvest areas and the energy purchased from external providers.

∑ ∑(Y ’ ijs × ECij ) + EXTEC j = EDj ∀j ∈J

i

(9)

s

Equations 10 and 11 are optional and ensure that the roundwood and chips from residues are delivered with a certain MC range to the pulp mill and energy plant, respectively.

∑ ∑ Xijs × MinMCRW ≤ ∑ ∑ Xijs × MCRWij ≤ ∑ ∑ Xijs × MaxMCRW ∀j ∈J

(10)

∑ ∑Y ’ijs × MinMCRE ≤ ∑ ∑Y ’ijs × MCREij ≤ ∑ ∑Y ’ijs × MaxMCRE ∀j ∈J

(11)

i

s

i

i

s

i

s

s

i

i

s

s

Equation 12 establishes the non-negativity of decision variable Xijk.

Xijk ≥ 0,∀i ∈ I , j ∈ J ,k ∈K

(12)

5.2 FastTRUCK FastTRUCK is a software tool developed to assist transport and forest managers in solving the daily truck scheduling problem. FastTRUCK uses a standard Simulated Annealing (SA) procedure, which is encoded in the C++ programming language, using an object-oriented design. The tool has been implemented in the Qt programming framework to provide users with a friendly graphical interface (Fig. 5) so that they can easily import input data from MS ExcelTM, review and modify parameters, display numerical and graphical solutions, and export output data back to MS ExcelTM.

286

Croat. j. for. eng. 38(2017)2

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279â&#x20AC;&#x201C;290)

M. Acuna

Fig. 5 Graphical user interface of FastTRUCK scheduler The use of FastTRUCK allows forest companies to reduce costs in their daily transport operations, and its integration with flow optimization tools such as MCPLAN (which provides the transport tasks that are used as an input by FastTRUCK), provide solutions that help pinpoint inefficiencies across the planning spectrum which involves annual, monthly and daily transport decisions. Also, the tool can be used to run scenarios that allow transport planners to design better haulage contracts as well as optimize truck design and fleet configuration. During dispatching, the tool can also assist truck dispatchers in making better realtime decisions, in particular, when unforeseen events such as breakdowns, changes in mill demand, and harvest site and mill closures occur during the daily transport operations.

FastTRUCK includes SA heuristics as an optimization engine in combination with a deterministic discrete

event simulation to emulate the movement of trucks throughout the day. In each iteration, the optimization engine allocates trucks to transport tasks, minimizing the number of trucks required and the waiting times at origins and destinations while meeting demand at mills. Also, the system checks the availability and capacity of the trucks and their current location, reviewing the current operational constraints, such as shift length, closing times of harvest sites and mills, and the tasks that can be performed by each truck in the fleet.

The tool requires several input parameters to perform the simulations and come up with optimal solutions. These parameters need to be collected daily and include data about trucks, supply points and demand points, transport tasks, travel distances and travel times (Table 3). The daily transport tasks to be performed must be known in advance; each of them is comprised of a wood pickup point, a customer site,

Table 3 Inputs required by FastTRUCK Trucks

Supply points

Demand points

Tasks

Travel distance/time

1 Truck configuration

1 Number of active supply points

1 Number of active demand points

1 Supply point

1 Depot to coupe

2 Operational and fixed costs

2 Starting and ending working time

2 Starting and ending working time

2 Demand point

2 Coupe to customer

3 Maximum legal payload

3 Loading time

3 Unloading time

3 Product type

3 Customer to coupe

4 Working hours per day

Croat. j. for. eng. 38(2017)2

4 Customer to depot

287

M. Acuna

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279–290)

Fig. 6 Basic version of the Simulated Annealing algorithm implemented in FastTRUCK and the product to be delivered (in case multiple product combinations are produced by the logging crews at the wood pickup point).

In the simulation system, the events for the trucks (departures, arrivals, loading, unloading) are derived from known routes with their associated travel distances and times which are used as inputs. Usually, a forestry company will specify the roads to be traveled by each haulage contractor. Also, in allocating trucks to transport tasks, the planners must know which trucks can perform which tasks, as some of the trucks configurations cannot access some of the wood pick up points due to their size or weight.

288

Four basic types of trips are performed by the trucks:  an empty trip from the truck’s depot to a wood pickup location  a loaded trip from a wood pickup location to a customer site  an empty trip from a customer site back to a wood pickup location  an empty trip from a customer site to the truck’s depot at the end the day.

On completion of any of these trips, the distance and time traveled between the origin and destination locations are calculated, and the system updates the Croat. j. for. eng. 38(2017)2

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279–290)

M. Acuna

total distance, time and cost accordingly. The first trip to a wood pickup location must be completed before or at the latest starting point of the loader’s time window. If during the arrival to a wood pickup point or customer site there is another truck being loaded or unloaded, the truck must wait until the loading or unloading equipment becomes available. This waiting time is added to the total time and cost for the truck.

As a task is performed, the system updates the total number of tasks being completed to that point and compares them with the total number of loads planned for the day. Performing all the transport tasks planned for the day is critical to get a minimum cost solution as a penalty is applied for every task that is planned but not performed.

Once a run is completed, FastTRUCK provides a report containing critical statistics at a fleet and truck level. These include among others, optimal fleet size, total and unit cost, daily volume moved from wood pickup points to customer sites, average truck utilization, and average waiting time. In addition to performance metrics, the model generates a work schedule for the trucks which can be used as a guide by dispatchers to control the allocation of trucks to tasks during the day. In the case of unforeseen events, they can manually adjust the solution to consider these factors.

This optimized plan is the least-cost way of routing trucks to satisfy the demand for products and typically provide solutions that lead to substantial reductions in costs resulting from a more efficient way to route the trucks, maximizing their utilization and avoiding delays at pickup points and customer sites.

Fig. 6 shows a reduced version of the SA algorithm implemented in FastTRUCK, which allows obtaining near-optimal solutions in a very short time (usually in less than 10 minutes). More details of the design and implementation of the algorithm are found in Acuna and Sessions (2014).

Given that the transport of timber and biomass represents a significant proportion of the operational cost for the forest industry, there is a big opportunity for the implementation of Operations Research (OR) solutions so that companies can organize their transport operations and allocate resources efficiently, and reduce the impact of transport operations on the environment. Moreover, acquiring technical knowledge and developing skills for the implementation of OR techniques will enable the forest industry, the forest operations research and teaching community, and our students, to develop innovative solutions that will help solve increasingly complex problems arising in the management of forest supply chains.

6. Conclusions

Audy, J., D’Amours, S., Rönnqvist, M., 2012: Planning methods and decision support systems in vehicle routing problems for timber transportation: a review. Quebec, Canada, Interuniversity Research Centre on Enterprise Networks, Logistics and Transportation: 38–45.

This paper has presented an extensive overview of the transport and biomass optimization problem in the forest industry, including a detailed description of mathematical solutions, including linear programming models and algorithms, to solve complex transportation planning problems involving annual, monthly and daily decisions. Also, the paper presents and describes two decision support tools, MCPLAN and FastTRUCK, which have been implemented to assist transport planners from the Australian forest industry to optimize the flows of timber and biomass from the forest to mills and energy plants, and the efficient routing of trucks that transport those products.

Croat. j. for. eng. 38(2017)2

7. References Acuna, M., Brown, M., Mirowski, L., 2011: Improving forestry transport efficiency through truck schedule optimization: a case study and software tool for the Australian industry. Austro and FORMEC, October 9–12, Graz and Rein, Austria.

Acuna, M., Anttila, P., Sikanen, L., Prinz, R., Asikanen, A., 2012a: Predicting and controlling moisture content to optimise forest biomass logistics. Croatian Journal of Forest Engineering 33(2): 225–238.

Acuna, M., Mirowski, L., Ghaffariyan, M.R., Brown, M., 2012b: Optimising transport efficiency and costs in Australian wood chipping operations. Biomass and Bioenergy 46: 291–300.

Acuna, M., Sessions, J., 2014: A simulated annealing algorithm to solve the log-truck scheduling problem. In: Simulated Annealing: Strategies, Potential Uses and Advantages.

NOVA Science, 191–220.

Acuna, M., Canga, E., Sánchez-García, S., 2017. Aplicación de la herramienta de optimización logística MCPLAN para la planificación del transporte de madera rolliza y biomasa: Un caso de estudio en Asturias. (In Spanish). In: 7 Congreso Forestal Español. 26–30 Junio. Plasencia, Cáceres, Extremadura, España.

Bettinger, P., Siry, J., Boston, K., Grebner, D.L., 2008: Forest Management and Planning. Academic Press, 331 p.

Carlsson, D., Rönnqvist, M., 1998: Tactical planning of forestry transportation with respect to backhauling, Report LiTH-MAI-R-1998–18, Linköping University, Sweden.

Devlin, G., Sosa, A., Acuna M., 2016: Solving the woody supply chain for Ireland’s expanding biomass sector: a case study. In Biomass supply chains for bioenergy and biorefining. Holm-Nielsen, J.B., Ehimens, E.A. (eds.). Woodhead publishing: 333–355.

289

M. Acuna

Timber and Biomass Transport Optimization: A Review of Planning Issues ... (279–290)

Church, R.L., 2007. Tactical-level forest management models.

Handbook on Operations Research in Natural Resources.

Weintraub, A., Romero, C., Bjørndal, T., Epstein, R. (eds), Springer, New York, 343–363.

D’Amours, S., Rönnqvist, M., Weintraub, A., 2008: Using operational research for supply chain planning in the forest products industry. INFOR 46(4): 265–281.

Epstein, R., Karlsson, J., Rönnqvist, M., Weintraub, A., 2007: Forest transportation. In Handbook on Operations Research in Natural Resources. Weintraub, A., Romero, C., Bjørndal, T., Epstein, R. (eds), Springer, New York, 391–403.

Forsberg, M., Frisk, M., Rönnqvist, M., 2005: FlowOpt – a decision support tool for strategic and tactical transportation planning in forestry. International Journal of Forest Engineering 16(2): 101–114.

Gronalt, M., Hirsch, P., 2007: Log-truck scheduling with a tabu search strategy. In: Doerner, K.F., Gendreau, M., Greistorfer, P., Gutjahr, W.J., Hartl, R.F., Reimann, M. (Eds.). Metaheuristics – Progress in Complex Systems Optimization.

New York: Springer, 65–88.

Haridass, K., 2009: Optimization and scheduling of a pooled log transport system. Master of Science thesis. Auburn University, USA, 92 p.

Kanzian, C., Holzleitner, F., Stampfer, K., Ashton, S., 2009: Regional Energy Wood Logistics – Optimizing Local Fuel Supply. Silva Fennica 43(1): 113–128.

Marier, P., Audy, J.-F., Gingras, C., D’Amours, S., 2007: Collaborative wood transportation with the Virtual Transportation Manager. In: Blanchet, P., (Eds.). International Scientific Conference on Hardwood Processing, September 24–26, Quebec City, Canada. Quebec: FPInnovations-Forintek, 191– 198.

Oberscheider, M., Zazgornik, J., Henriksen, C. B., Gronalt, M., Hirsch, P., 2013: Minimizing driving times and greenhouse gas emissions in timber transport with a near-exact solution approach. Scandinavian Journal of Forest Research 28(5): 493–506.

Palmgren, M., 2001. Optimisation methods for log truck scheduling. Theses No. 880. LiU-TEK-LIC-2001:17, Linköping University, Sweden.

Palmgren, M., Rönnqvist, M., Varbrand, P., 2004: A nearexact method for solving the log-truck scheduling problem.

International Transactions of Operations Research 11(4): 447–464.

Rey, P.A., Muñoz, J.A., Weintraub, A., 2009: A column generation model for truck routing in the Chilean forest industry.

Information Systems and Operational Research 47(3): 215– 221.

Rönnqvist, M., Sahlin, H., Carlsson, D., 1998: Operative planning and dispatching of forestry transportation. Research paper LiTH-MAT-R-1998-18, Linköping University, Sweden.

Rönnqvist, M., 2003: Optimization in forestry. Mathematical Programming 9(1–2): 267–284.

Rummukainen, H., Kinnari, T., Laakso, M., 2009: Optimization of wood transportation. In: Madetoja E., Niskanen H., Hämäläinen J. (Eds.). Papermaking Research Symposium, Kuopio, Finland, June 1–4, Kuopio: University of Kuopio.

Shen, Z., Sessions, J., 1989: Log truck scheduling by network programming. Forest Products Journal 39(10): 47–50.

Sosa, A., Acuna, M., McDonnell, K., Devlin, G., 2015: Managing the moisture content of wood biomass for the optimisation of Ireland’s transport supply strategy to bioenergy markets and competing industries. Energy 86: 354–368.

McDonald, T., Taylor, S., Valenzuela, J., 2001: Potential for shared log transport services. In: Wang, J., Wolford, M., McNeel, J., (Eds.). 24th Annual COFE Meeting, July 15–19 Snowshoe, USA. Corvallis: Council on Forest Engineering, 115– 120.

Taja, H.A., 2016: Operations Research: An Introduction. 10th Edition. Pearson, 848 p.

McDonald, T., Haridass, K., Valenzuela, J., 2010: Mileage saving from optimization of coordinated trucking. In: 33rd Annual COFE Meeting. June 6–9 Auburn, USA.

Weintraub, A., Romero, C., Bjørndal, T., Epstein, R., 2007: Handbook of Operation Research in Natural Resources.

Springer, New York.

Weintraub, A., Epstein, R., Morales, R., Serón, J., Traverso, P., 1996: A truck scheduling system improves efficiency in the forest industries. Interfaces 26(4): 1–12.

Authors’ address:

Received: December 1, 2016 Accepted: May 24, 2017

290

Mauricio Acuna, PhD. e-mail: macuna@usc.edu.au AFORA – University of Sunshine Coast Private bag 12 Hobart, TAS AUSTRALIA Croat. j. for. eng. 38(2017)2

Subject review

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals from 2005–2016 Igor Potočnik, Anton Poje Abstract The occupational safety and health change through time due to technological and social development. It is an obligation of scientific research to impartially and critically examine these changes and propose measures to reduce negative impacts on people. Since the forestry as an industry sector follows general changes, we tried to establish the situation in the field of occupational safety and health in forestry by reviewing the studies in the period 2005–2016. The review included studies, the results of which have been published particularly in scientific journals relating to the field of forestry and ergonomics with an impact factor. The findings show that the number of published articles in the field of occupational safety and health and ergonomics increases. Studies were mostly limited to only three continents, namely Europe and North and South America, and 26 countries in total. The majority of research was conducted in Canada, Brazil and Sweden. The largest number of research relates to traditional technologies of harvesting (chainsaw and skidder), whereas the Nordic states prevail in terms of modern, mechanized technologies. The study shows that international and intercontinental cooperation of researchers must be further stimulated in the field of research and education. It has been identified that there is a lack of studies addressing the issue of biomass production, forest road and skid trail construction, and some new technologies. There is a deficiency of cognitive studies, studies of workers’ burnout and comprehensive studies of ergonomics and productivity in the field of ergonomics. The uniform statistics of recording accidents will provide the research to be conducted in all forestry operations and enable the preparation of efficient preventive measures.

Keywords: forestry, ergonomics, safety, review

1. Introduction The aim of each society and individual must be to preserve work as a basic right and need of every human being, as through work men achieve fulfillment at a personal and social level. However, the work should have no negative impacts on workers’ health and it should provide them a dignified life in old age.

Within this context, the ergonomics as a science and a profession adjusts work to workers so as to reduce difficulty and harmfulness of work by providing adequate work efficiency, which is reflected in a reduced level of illnesses or injuries at work. Thus, a commitment of the ergonomics society is to provide ergonomic knowledge and tools in a form useful for decisionCroat. j. for. eng. 38(2017)2

makers who require assistance (Brewer and Hsiang 2002).

Regardless of introducing modern methods of harvesting and production, the forest work is considered one of the most dangerous industrial activities with a high share of lethal accidents and injuries (EU-OSHA 2008, Adams et al. 2014). Due to natural environment, heavy loads and frequent use of manual tools and machines, workers are exposed to physical, physiological and environmental factors that result in various illnesses, related in particular to muscles, skeleton, nerves and vascular system and impairment of hearing (Gallis 2006, Bovenzi 2008, Fonseca et al. 2015).

With the development of mechanical harvesting, the

291

I. Potočnik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

work transferred from outdoors to a cabin thus reducing physical difficulty of work and exposure to most environmental risk factors. An increased human/machine interaction (Burman and Löfgren 2007) and the change of work from »doing to thinking« (Hollnagel and Woods 2005) caused the focal points of accident to change (Axelsson 1998) and new injuries (Repetitive strain injuries – Axelsson and Pontén 1990) as well as new, cognitive risk factors to occur. As a consequence, it appears that, in mechanical harvesting, the traditional ergonomic paradigm »less is better« should be replaced by »more can be better«, since reduced physical activity also has harmful impacts on health (Straker and Mathiassen 2009). To comprehensively improve the situation of occupational safety and health in the field of forestry, the rooted culture of »can do« must be replaced by »can do safely«, which requires changes in the entire chain of wood production (Adams et al. 2014).

Notwithstanding the mechanized wood production expanding to the fields where the work was executed in a traditional way, e.g. using chainsaw and cable skidder, traditional harvesting technologies remain applicable in difficult working conditions (e.g. co-natural, mostly deciduous forests) and private small-scale forests. Work technology used, out-of-date work equipment, non-use of personal protection equipment and lack of knowledge and experience cause poorer occupational safety and health of private forest owners than is the case with professional workers (EU-OSHA 2008).

Constant social and technological changes have also an impact on forest use and importance of its functions, i.e. on forest production and consequently occupational safety and health. For this reason, the aim of this study is to analyze ergonomics and occupational safety in the field of forestry in the past twelve years and in particular identify the focal points of research in terms of technology, risk factors and characteristic fields. On the basis of the results, potential deficiencies/ uncovered fields of the previous research were defined and guidelines for further research given.

(Safety Science). To select articles from journals, the key words »ergonomics« and »safety« were used for forestry journals, while in other journals the word »forestry« was applied. It was additionally required that studies be partially or entirely conducted in the field of forestry, forest workers, forest machines or tools. The review thus excluded marginal fields of forestry, e.g. arboriculture and fire protection, and the works not performed in forests (e.g. ornamental nursery). After examining titles and abstracts, the final selection included 136 full-text articles, i.e. 76 in the field of ergonomics and 60 in the field of occupational safety.

The ergonomics articles were then classified according to their contents and risk factors in 16 fields, which are defined in the European ergonomic and safety guidelines for forest machines (ErgoWood 2006) and 8 additional fields. In terms of the latter, the field »Independent data« needs to be highlighted. It included articles that used ergonomic indicators only as basic data for other purposes, e.g. establishing the impact of piece rate wages on health and safety (Johansson et al. 2010) or similarity of processes (Leszczyñski 2010).

In addition to articles directly addressing a specific factor, the articles with an indirect impact on the level of workers’ exposure were also included in the classification. Thus, for example, coupling forces exerted by fellers during wood harvesting (Malinowska-

2. Methods The study addressed the period from 2005 to December 2016. It included journals in the field of forestry and ergonomics from the Web of Science collection, which had an impact factor (Fig. 1). According to our experience and the purpose and content of the journals, the review additionally included two journals, one in the field of forestry (International Journal of Forest Engineering) and one in the field of occupational safety

292

Fig. 1 Flow diagram of articles selection Croat. j. for. eng. 38(2017)2

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

Borowska et al. 2011) or characteristics of a seat (Ji et al. 2015) were included in »Vibrations«. If a study dealt with several fields, the article was also classified in several fields.

Similar to ergonomics articles, the articles relating to occupational safety were divided according to their content into 5 fields. A special field »Integrated safety« included articles where the results of research directly affected the level of occupational safety and health.

Thus, for example, this field included articles where tools and methods of directional felling were developed (Noll and Lyons 2010) or which established the suitability of mobile anchors in cable skidding (Leshchinsky et al. 2016).

All articles were additionally divided according to the year of publication, countries and continents where the research has been conducted. If the location was not clear from the article, it was attributed to the country of the first author of research. In a similar way, all articles were also divided according to the respective technology or phase of work within the technological process of forest exploitation. Under the selected technology or phase of work, articles dealing with tools, accessories or working methods of the selected technology or phase of work were also included. For example, under the section »Chainsaw«, in addition to articles that measured the exposure of work with a chainsaw (e.g. Horvat et al. 2005), the articles that addressed the issue of antivibration gloves (e.g. Goglia et al. 2008b), coupling forces exerted by fellers (Malinowska-Borowska et al. 2011) or the impact of using forestry equipment on hearing loss (Fonseca et al. 2015) were also included. If several technologies were included in a study, the article was also classified in several technologies.

3. Results 3.1 Temporal and spatial dynamics of publishing The number of articles related to safety and ergonomics gradually increased in the analyzed period (Fig. 2). If we include this article, which will be published in March 2017, in 2016, the average number of articles published in the last six years increased by 34% (from 9.6 to 13.0 articles). Two thirds of all articles were published in forestry journals.

Ergonomics researches included 18 countries on five continents (Fig. 3). The vast majority of researches (95%) were conducted on three continents, i.e. Europe (58%), South America (20%) and North America (17%).

The largest number of studies according to individual countries was conducted in Brazil (15) and Canada Croat. j. for. eng. 38(2017)2

I. Potočnik and A. Poje

Fig. 2 Number of articles published in the period 2005–2016 (11). Sweden, Croatia, Italy and Austria, with more than 6 studies, stand out in Europe in terms of the number of studies.

Compared to ergonomics research, the studies related to safety were more widespread, since they were conducted on seven continents and in 20 countries (Fig. 3). In this case as well, approximately 75 % were performed in Europe (50%), and North America (27%), i.e. most of them in the USA, Canada and Sweden.

3.2 Technologies and risk factors The largest share of ergonomics articles (Fig. 4) addressed harvesting (35%) and skidding (38%) technology, followed by planting, care and protection of forests (pre-harvesting operations, 12%), while the smallest share dealt with biomass production technology (6%) and transport (1%). In terms of individual machines, the articles mostly analyzed the work with chainsaw, skidders and agricultural tractors, and forwarders and harvesters, which in total represents 59% of all forestry machines included in the articles. In terms of technology, the fully mechanized cut-tolength (CTL – harvester + forwarder) and whole-tree (WT – feller buncher + grapple skidder) harvesting technologies were addressed in 24% or 13% of articles, respectively.

The results (Fig. 4) also show differences in technologies considered according to continents. Thus, for example, studies of harvesting with chainsaw and skidding with skidders occur on all five continents, studies of technologies used in pre-harvesting operations only

293

I. PotoÄ?nik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291â&#x20AC;&#x201C;310)

Fig. 3 Number of forestry ergonomics (up) and safety (down) articles published in journals with impact factor per world countries in the period 2005â&#x20AC;&#x201C;2016 in North and South America, while the majority of studies of CTL technology (69%) and cable skidding (67%) were conducted in Europe. There are also large differences within European countries. Thus, for example, the studies of CTL technologies (83%) prevail in the Nordic countries, while the studies of work with chainsaw (100%) and skidding with skidders are more prevalent in other nine European countries.

In the analyzed period, more than a half (51%) of safety-related studies addressed the general situation

294

of occupational health in the field of forestry, in forestry as an industrial activity, in individual companies or in the work of forest workers (Fig. 5). In terms of the number, felling with the chain saw (25%) stands out among individual forestry operations. Two most frequently represented fields of research are present on seven and five continents, respectively. Except for wood transport, studies of all other operations took place in Europe.

From the analysis of articles according to risk factors, it is established (Fig. 6) that more studies adCroat. j. for. eng. 38(2017)2

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

I. Potočnik and A. Poje

Fig. 4 Representation of forestry machines in ergonomics articles published in journals with impact factor per continents in the period 2005–2016 dressed physical and physiological risk factors, which limit the use of technologies due to unsuitable dimensions and physiological fitness of workers, than environmental risk factors (29%), such as noise, vibrations, etc. In terms of individual risk factors included in the studies, three stand out most prominently: vibrations, cardiovascular load and working positions, which together represent more than a half (51%) of all studied factors. However, no study included four risk factors

referring to manuals and instructions, maintenance, maintenance index and biological agents.

Differences in individual risk factors are also evident in the classification according to continents. Thus, for example, most studies of physical risk factors were conducted in South America, where physical characteristics of workers were compared with dimensions of machines, establishing spine load through body positions and psychophysical workload of workers.

Fig. 5 Representation of forestry operations in safety articles published in journals with impact factor per continents in the period 2005–2016 Croat. j. for. eng. 38(2017)2

295

I. Potočnik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

Fig. 6 Representation of risk factors in ergonomics articles published in journals with impact factor per continents in the period 2005–2016 On the contrary, the majority of studies in Europe addressed environmental factors, while in terms of individual factors, the studies of vibrations, cardiovascular load and musculoskeletal disorders prevailed.

The studies of most frequently addressed risk factors are distributed differently according to technologies. Thus, the studies relating to musculoskeletal disorders were most frequently established in connection with the CTL harvesting technology, while the studies of loads due to vibrations and mental and

physical load were most frequently established in relation to traditional wood production (chainsaw + skidders).

In terms of occupational safety, the largest number of studies dealt with the prevention of accidents at work, since the studies of preventive measures and integrated safety present almost a half (48%) of all studies. In terms of their scope, only the studies in Europe and North America cover all five fields considered (Fig. 7). Except for the CTL harvesting technology,

Fig. 7 Representation of risk factors in safety articles published in journals with impact factor per continents in the period 2005–2016

296

Croat. j. for. eng. 38(2017)2

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

accidents at work were analyzed in all operations, while the studies of lethal accidents were only included in general forestry studies.

3.3 Overview of main study results by forest operations 3.3.1 General 3.3.1.1 Ergonomics Previous studies have established that the number of ergonomics studies in certain fields, such as skidding with cable cranes, is increasing (Cavalli 2012).

This can result from an increasingly wider scope of studies dealing with wood production that reflects the sensitivity of the modern society to the sustainability of human and environmental resources (Košir et al.

2015). In this context, the establishment of ergonomic suitability of work may be the main or side product of broader research. According to previous research, the forestry ergonomics will have to face the challenges, e.g. dissemination of the existing knowledge of classic ergonomics, adjustments of standards to specific local conditions and workers, development of cognitive ergonomics, adjustments of organization of work and people to (exceedingly) fast developing technologies and production processes (Heinimann 2007).

Ergonomics risk factors may also occur in research as impact factors. Thus, by using environmental risk factors and unfavorable body postures, the needed reduction of norms in harvesting was established (de Souza et al. 2015). Similarly, in addition to productivity factors, ergonomics factors in harvesting were used to estimate the benefits of four working operations (Leszczyñski 2010). The work ability index is negatively related to some physical characteristics of a worker, i.e. age and weight, and work experience (Landekić et al. 2013).

3.3.1.2 Safety In the world of high technological development and internationalization of forestry companies, it is a common concern to provide a healthy, safe and physically acceptable work for all (Rickards 2008).

The frequency of all lethal accidents and lethal accidents of young workers in the industrial sector of agriculture, forestry and fishing is often higher than in other sectors (Cohen et al. 2006, Ehsani et al. 2013).

With the rate of 63 accidents/100,000 employees, it may represent 10% of all accidents (Suchomel et al.

2013). The same applies to non-lethal accidents, where the frequency of accidents in individual countries may reach the rate of 98.5 accidents/1000 employees (Suchomel et al. 2013).

Croat. j. for. eng. 38(2017)2

I. Potočnik and A. Poje

The frequency of accidents in forestry companies, with rates between 14.8 and 97.0 accidents/1000 employees or 54.2 accidents/million cubic meters of wood, is often higher than in other sectors (Landekić 2010, Grzywiński et al. 2013, Tsioras et al. 2014, Laschi et al. 2016). The severity of accidents ranged between 18.2 and 48.6 days/accident (Grzywiński et al.

2013, Tsioras et al. 2014, Laschi et al. 2016) and was higher in simultaneous injuries of several body parts (Tsioras et al. 2014) and in older workers (Laschi et al. 2016). The number of lethal accidents of professional forest workers is decreasing and it is even fivetimes lower than that of non-professionals. The frequency of lethal accidents is the lowest in the Nordic countries, while the highest rate may even reach up to 9.5 accidents/million cubic meters of timber (Klun and Medved 2007).

The largest number of accidents of forest workers occurs at the beginning of a working week, supposedly due to lower attention of workers (Tsioras et al.

2011, Laschi et al. 2016). Regarding the parts of the body, the injuries of arms and legs are the most frequent, falls prevail in terms of causes, while contusions represent the most frequent type of injuries (Potočnik et al. 2009, Tsioras et al. 2011, Grzywiński et al. 2013, Enez et al. 2014, Tsioras et al. 2014, Laschi et al. 2016). The probability of accidents is increasing with the use of chainsaw, hookaroon, number of breaks during work, workday duration (Enez et al.

2014), while the risk also increases due to weather conditions, e.g. rain (Suchomel and Belanová 2009).

In addition to insufficient education and training of workers, the causes, such as disturbances of biological rhythm, increased difficulty of work due to changes in patterns of working time, quick shifting of workplaces, language barriers, job insecurity, occupational stress, works with sub-contractors, deterioration of general welfare of workers, have negative impacts on workers’ health and may result in accidents (Papadopoulos et al. 2010, Mylek and Schirmer 2015). Personal and organizational factors are most important causes of lethal accidents in harvesting, while the following causes are the most important individual factors: positioning in danger zones, carelessness, inappropriate behavior and unsuitable selection of workers (Melemez 2015).

The frequency of accidents can be reduced by implementing programs, such as Forest Protection program (e.g. Sustainable Forestry Initiative, Chapman and Husberg 2008), indirectly with the system for environmental management – ISO 14001 (Ackerknecht et al. 2005), proper corporate risk management (Albizu-Urionabarrenetxea et al. 2013) and certification of

297

I. Potočnik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

contractors (Martinić et al. 2011). The frequency of accidents may be reduced with an unconditional commitment to the safety culture at all organizational levels. The safety culture may be enforced through proper education, vocational training, selection of workers and sufficient motivation (Martinić et al. 2007, Albizu Urionabarrenetxea et al. 2010, Tsioras 2010, Melemez 2015). Education and training must be designed to be sufficiently extensive (Tsioras 2012), adjusted to the characteristics of individual groups of workers (Poje et al. 2016) and must primarily reduce the risk of accidents – and then prevent the occurrence of chronic diseases (Montorselli et al. 2010). A special attention must be paid to workers who suffered several injuries, since the cause may lay in recidivism or repeated violations of safe work procedures (Laschi et al. 2016). Since the work in forests is often performed in groups, special focus must be placed on the selection of group members and methods of introducing new members (Burt et al. 2008, Burt et al. 2009).

The data on accidents collected through general forms are usually not useful to design preventive measures (Bentley et al. 2005, Robert et al. 2015), and should therefore be adjusted to individual fields and workplaces (Cohen et al. 2006, Poje et al. 2016). Underreporting (Papadopoulos et al. 2010) and presenteeism (Wilmsen et al. 2015) also affect the statistics of accidents.

3.3.2 Pre-harvesting operations 3.3.2.1 Ergonomics The majority of ergonomics studies relating to planting, care and protection of forests addressed physical and physiological risk factors. Thus, the loads on lumbar spine due to unfavorable body postures exceed permissible values in transporting saplings (Alves et al. 2006), fertilization and planting (Da Silva et al.

2007, Silva et al. 2008, Vosniak et al. 2010, De Britto et al. 2014), and manual mowing (de Oliveira et al. 2014).

The loads on wrists in planting also pose a potential risk of musculoskeletal disorders and repetitive strain injuries (Da Silva et al. 2007, Denbeigh et al. 2013). Cardiovascular loads exceed permissible values (40%) in manual mowing (de Oliveira et al. 2014), brush cutting (Toupin et al. 2007), planting (Da Silva et al. 2007, Silva et al. 2008), weeding with knapsack sprayers (Sasaki et al. 2014) and manual pruning (Nutto et al. 2013).

Contrary to the expected, more demanding working conditions in brush cutting reduce cardiovascular loads. This means that workers with piece rate wages compensate their loss of income in more favorable working conditions (Toupin et al. 2007). Since, according to the majority of the previous research, the piece

298

rate payment has a negative impact on occupational safety and health (Johansson et al. 2010), a changed, variable method of payment for work (Toupin et al.

2007) is proposed to reduce the impact of more difficult working conditions on psychophysical workload.

The second, general measure reducing the exposure of workers to other risk factors is balancing the difficulty of work with active breaks (Gallis 2013).

Two studies addressing the issue of oxygen consumption, maximum aerobic capacity and heart rate of workers in planting and brush cutting are also very significant for further research. The studies established that oxygen consumption and maximum aerobic capacity can be established precisely enough by measuring heart rate and relations between oxygen consumption and heart rate established in laboratory. This means that there is no need to measure oxygen consumption on-site (Dubé et al. 2015). However, to obtain the right estimation of oxygen consumption, the thermal component must be eliminated from the heart rate (Dubé et al. 2016).

Environmental factors were studied only in using various knapsack sprayers for weeding, pest management or combating diseases (Sasaki et al. 2014). The study thus established that some types of knapsack sprayers exceed permitted thresholds of exposure to noise and whole body vibrations (WBV), and that their operation is very demanding in terms of energy consumption.

3.3.2.2 Safety The study addressing the machine stability (tractor-subsoiler) for soil preparation established that the maximum transverse slope of terrain must not exceed 23.7⁰ (Pereira et al. 2011). During pre-harvesting operations, the frequency and severity of accidents is significantly lower than during harvesting and skidding (Potočnik et al. 2009).

3.3.3 Harvesting 3.3.3.1 Ergonomics Contrary to pre-harvesting operations, the share of the studies of harvesting with chainsaw addressing physical and physiological risk factors was lower than the share of environmental risk factors. Thus, unfavorable body postures detrimental to health during harvesting with chainsaw (Barbosa et al. 2014) result in musculoskeletal disorders that mostly affect lower back, arms and wrists. The share of fellers with disorders increases with their age and working experience (Grzywiński et al. 2016).

Cardiovascular loads in working with manual saw or chainsaw heavily exceed the permitted values (40%) Croat. j. for. eng. 38(2017)2

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

and increase with the work productivity (Silayo et al.

2010). The loads on fellers are higher during work in younger than in older stands, and higher when working with a processor than with a skidder (Leszczyński and Stańczykiewicz 2015).

The firmness of grip of the chainsaw handle, affecting the transfer of HA vibrations, depends on worker’s experience, work operations and wood hardness.

Thus, coupling forces exerted by fellers are higher with less experienced workers, they are higher in felling and cross-cutting than in limbing, and higher with tree species of higher wood hardness (MalinowskaBorowska et al. 2011, Malinowska-Borowska et al.

2012, Malinowska-Borowska and Zieliński 2013).

When working with chainsaw in regeneration stands, loads on hands and arms caused by vibrations (HAV) exceeded the action value of daily exposures (Goglia et al. 2012a) defined in the European and, due to harmonization, also national legislations of the European countries (Goglia et al. 2012b). The HAV exposures, when using Kasper safety bar, are not different compared to the use of conventional bar (Rottensteiner and Stampfer 2013). However, HAV differs between tree species and is higher with tree species of higher wood density (Rottensteiner et al. 2012). One of the possible measures to reduce the harmful effect of HAV on workers is the use of anti-vibration gloves, which must comply with international standards (Goglia et al. 2008a). The studies show that there are significant differences between the types of anti-vibration gloves in terms of their insulation effectiveness (Goglia et al.

2008b).

High exposures of fellers to noise cause hearing impairments, in particular if the personal protection equipment is not used (Fonseca et al. 2015). On the contrary, the exposure of fellers to fir wood dust (1.29 mg/m3) does not exceed limit values of daily exposure (Horvat et al. 2005).

Dimensions of machines must be adjusted to physical characteristics of an operator to provide a safe and healthy work. Thus, the largest deficiency in mechanized harvesting with feller-bunchers is its limited possibility of adjusting the seat, cab access, controls and working postures (Fernandes et al. 2009, Fernandes et al. 2011). In addition to physical risk factors in mechanized harvesting with feller-bunchers, the WBV loads also significantly exceed the action value of daily exposure (Almeida et al. 2015).

A long-term exposure of operators of mobile forestry machines to WBV increases the risk of musculoskeletal disorders and reduces work productivity (Jack and Oliver 2008, Village et al. 2012). Thus, the Croat. j. for. eng. 38(2017)2

I. Potočnik and A. Poje

consequences of a relatively safe work with harvesters are manifested in musculoskeletal and nerve disorders. Pains most often occur in the area of the neck, lower back and shoulders (Hanse and Winkel 2008, Rehn et al. 2009, Silva et al. 2014). Neck pains may thus occur due to long periods of static muscle operation at low intensity and short time of muscle rest (Østensvik et al. 2008a, Østensvik et al. 2009). The occurrence of musculoskeletal disorders depends on the organization of work (Østensvik et al. 2008b). It decreases with the level of work control and job rotation (Hanse and Winkel 2008), while it increases with the duration of employment in the forestry sector (Silva et al. 2014).

The negative impact of exposure to WBV on the occurrence of musculoskeletal disorders in mechanized CTL harvesting and skidding was not proved (Rehn et al. 2009).

High incidence of musculoskeletal disorders of harvester operators and other operators intensely engaged in operating machines and cranes requires an ergonomic optimization of the levers. Thus, the results show that by using levers with short handles and when working at a higher gain, the efficiency of work increases, while the physical load decreases or remains the same as when using levers with long handles (Huysmans et al. 2006).

The study of cognitive abilities of harvester operators, which is also the only study in this field included in the research, showed that the following characteristics of an operator are needed for a skilful and productive work: comprehensive perception, wide use of memory functions, non-verbal deduction, spatial perception, coordination, concentration and motivation.

In terms of work efficiency, none of the aforementioned abilities is superior; it is only important to have these abilities (Ovaskainen and Heikkilä 2007). Information collected in the eye-tracking study (Häggström et al. 2015) is among the most important data to provide an efficient machine management, possibility to automate procedures and develop decision support systems.

Since the work with harvesters and forwarders is often done also at night and under artificial light, it is recommended to use xenon lights which, compared to halogen lights, have a higher colour temperature and thus improve peripheral vision. Xenon lights are also more energy efficient, since one 35W xenon light provides the same illumination as do three 70W halogen lights (Poom et al. 2007).

Exposure of harvester operators to noise is lower than in wood extracting with forwarders and does not exceed low action values of daily exposure (Messingerová et al. 2005).

299

I. Potočnik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

3.3.3.2 Safety During harvesting, accidents most often occur when felling trees (Laschi et al. 2016), the main reason being an interaction with parts of trees (Potočnik et al.

2009). The risk of accident is also different for different methods of felling. Thus, for example, the felling of individual trees poses a higher risk than clear-felling (Neyland et al. 2012). The risk of accident of professional workers is due to their underestimating the danger and a higher tempo of work during regular felling than during salvage felling (Poje et al. 2016).

Fatigue and dehydration cause a higher frequency of accidents in summer months and in late morning hours (Bentley et al. 2005).

Unforeseen interactions with a falling tree or branches and equipment are usually the main causes of accidents during felling performed by private forest owners (Neely and Wilhelmson 2006). Additionally, poor or wrong working technique, insufficient training, deficient or no personal protection equipment result in an increased risk of lethal and non-lethal accidents (Neely and Wilhelmson 2006, Häggqvist et al.

2010, Lindroos and Burström 2010, Brzózko 2016).

To reduce the risk of accidents during harvesting, fellers must be properly trained, and the contents and emphasis of trainings must be different for professional and non-professional fellers (Poje et al. 2016), beginners and experienced workers (Bentley et al.

2005), and adapted to their linguistic comprehension (O’Neal et al. 2007). The level of harvesting mechanization and control must be increased and employee turnover reduced (Bell and Grushecky 2006). Fellers must be physically and mentally fit to adapt to changing working conditions, because even experienced workers cannot anticipate all dangers (Bentley et al.

2005, DeMille and Lyons 2016). Immediately before the harvesting starts, a felling plan must be prepared in compliance with working conditions hazards (Lyons and Demille 2015). Efforts to improve the safety during felling are also made by developing new methods and auxiliary tools for directional tree felling (Lindroos et al. 2007, Noll and Lyons 2010, Lyons and Noll 2011, Lyons and Ewart 2012, Lyons 2015), since the studies show that the positioning of workers in the danger zone is one of the major factors causing lethal accidents (Melemez 2015). It is possible to reduce the risk of a kickback during cross-cutting with chainsaw by applying caution according to types and state of wood, right selection of chainsaw power and the length of the bar, the selection and maintenance of chain and its tension, and efficient chain brake (Dąbrowski 2012, Dąbrowski 2015). An innovative method to improve the ability of feller foremen is a

300

computer game Felling Safety Game (Yovi and Yamada 2015).

When introducing mechanized felling, which reduces the risk of accidents (Bell and Grushecky 2006), physical capacities and demographic characteristics of operators must be taken into account (e.g. age, Ferrari et al. 2012). However, not only proper skills, abilities, work techniques and training, but also all interactions between workers, technology, organization and environment must be comprehensively taken into account to provide an efficient and safe work (Häggström and Lindroos 2016).

3.3.4 Wood extraction 3.3.4.1 Ergonomics Operations with a high share of manual work, e.g. chainsaw felling and skidding with mules and agricultural tractors, cause musculoskeletal disorders and, mainly due to pains in the arms and wrists, lower back and neck, prevent the works to be executed (Gallis 2006). Since WBV is one of the possible causes for musculoskeletal disorders, the studies established that WBV in wood skidding still exceeds the permitted thresholds and that there have been no significant changes in the respective field in the last two decades (Cation et al. 2008, Jack et al. 2010, Pandur et al. 2013, Almeida et al. 2015). Exposure to VBW, in addition to working conditions (Cation et al. 2008, Jack et al. 2010), also depends on physical characteristics of an operator, e.g. his weight and characteristics of the seat (Ji et al. 2015, Ji et al. 2017). To provide a safe and healthy work with skidders, the areas, such as technical safety devices, accessing devices, cabin design, lighting devices, handling and safety requirements for designing and operating winches (Beuk et al. 2007), are important in particular.

The first phase of skidding, transporting of wood from tree-stumps to the skid trail (machine) is done with a winch. Due to high forces caused by friction between log and surface, all integral parts, such as wire rope and stop-end connections, must be properly designed to provide minimum safety requirements for their operation (Hartter and Garland 2006).

The replacement of metal wire ropes with synthetic ones has reduced the force needed to extend a rope, but failed to reduce the expected cardiovascular workload (Ottaviani et al. 2011, Magagnotti and Spinelli 2012). The same also applies to the slack-pulling device, where its application only increased the work efficiency (Spinelli et al. 2015). According to authors, the main reason for the unchanged cardiovascular workload is that the movement of workers, in particular the older ones, is physically the most demanding Croat. j. for. eng. 38(2017)2

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

task in the forest (Aalmo et al. 2016). Secondary causes, such as lighter means of work and tools, only temper the issue of excessive workload. To reduce the workload, several measures must be implemented simultaneously (Magagnotti and Spinelli 2012). On the contrary, the application of an auxiliary winch in collecting the wood increased the efficiency and reduced the cardiovascular workload as a lower number of workers was required (Magagnotti et al. 2016).

Although muscoloskeletal disorders are also experienced by forwarder operators, in particular pains in the lower back, neck, shoulders and the upper back (Phairah et al. 2016), these disorders are experienced more frequently by harvester operators (Silva et al.

2014). This can be the result of a less intense operation of the control lever and consequently shorter static loads of muscles (Østensvik et al. 2008a, Østensvik et al. 2009). Positions of parts of the body determined by the cabin design are not necessarily the most suitable, since they are not adjusted to physical characteristics of operators (Fontana and Seixas 2007).

When operating forwarders, the variability and exposure to WBV are the highest during movements (Rehn et al. 2005, Häggström et al. 2016). Factors that affect the variability are the operator, machine model and terrain type (Rehn et al. 2005). Different types of grabs have no impact on the exposure to WBV, but have nevertheless an impact on the damage to other trees in the stand (Häggström et al. 2016). Although the exposure of forwarder operators to noise does not exceed permissible loads, it is higher than in felling with harvesters due to faster and more frequent movements (Messingerová et al. 2005).

The study, which was based on the interviews with operators, showed that some types of cable cranes do not meet ergonomic requirements. This applies in particular to cable cranes without a cabin (Penna et al.

2011). By using radio controlled but heavier chokers, compared to the standard ones, the productivity increases, and consequently also the workload (Stampfer et al. 2010).

The most complex ergonomics studies of forestry machines and systems were conducted in North-West Russia (Gerasimov and Sokolov 2009, Gerasimov and Sokolov 2014). By using more than a hundred measured parameters and parameters obtained from the interviews with operators, the studies evaluate their suitability in terms of ergonomics. The findings show that the fully mechanized harvesting is the most suitable in terms of ergonomics and that harvester and forwarder are the most suitable combination of machines in CTL harvesting, while in terms of tree methods this applies to feller-buncher and grapple skidder.

Croat. j. for. eng. 38(2017)2

I. Potočnik and A. Poje

The least suitable technologies are those that include feller, skidder operator and choker setter. Visibility and working postures stand out as the worst fields of ergonomics.

3.3.4.2 Safety The largest number of accidents during cable yarding occurs due to breaking of spar and anchor trees.

On average, the accident severity is lower than during harvesting (Potočnik et al. 2009) and manual skidding or skidding with skidders, while the established frequency of accidents is 36 accidents/million cubic meters of extracted wood (Tsioras et al. 2011). Instead of trees, mobile machines can be used as end supports in cable yarding, whereby their mass and state of the ground are important for a safe operation of the entire system (Leshchinsky et al. 2015). The planning of cable yarding system by using theoretical models and software provides its rational and safe implementation (Bont and Heinimann 2012, Dupire et al. 2016).

3.3.5 Biomass production 3.3.5.1 Ergonomics The production of firewood with a firewood processor is at least 25% more efficient and physically less demanding when using a combination of circular saw and hydraulic log splitter. The cardiovascular workload does not exceed the limit in either case (Lindroos 2008). In addition to a lower workload, the health hazard due to unfavorable body positions during firewood production with semi-mechanized or fullymechanized systems is lower (Spinelli et al. 2017).

Measured WBV loads during the production of wood chips are higher in truck-mounted chippers, while the noise load is higher in tractor-trailer chippers. WBV exposure is higher in the production of chips from hardwood than from softwood (Rottensteiner et al. 2013). A daily exposure to noise exceeds the lower action value of daily exposure in case of higher utilization of machines (Poje et al. 2015), while some types of truck-mounted chippers can exceed the action value of daily exposure to WBV (Rottensteiner et al. 2013). WBV loads on operators of slash grapple during the collection of harvesting residues exceed the action value of daily exposure (Almeida et al. 2015).

3.3.5.2 Safety During the production of firewood, most accidents occur when using splitters, with finger injuries being the most frequent (Lindroos et al. 2008). The causes of accidents are not in compliance with the safe work procedures, improper level of machine safety, use of gloves and operation of a machine with more than one worker (Lindqvist and Nilsson 2011).

301

I. Potočnik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

3.3.6 Wood transport

publications increased in the period concerned. In addition to the necessity of publishing and increasing the number of IF journals, the cause may also be in an accelerated technological development of machines (eg. harwarders and autonomous forest machines, Hellström et al. 2006, Ovaskainen and Heikkilä 2007, Ringdahl et al. 2012) or their individual parts (eg. chocker, winch, grapple, seat, Stampfer et al. 2010, Häggström et al. 2016, Magagnotti et al. 2016, Ji et al. 2017) and in changes of work processes (eg. tree-topping, Huber and Stampfer 2015). The major driving force of occupational safety and health development may be linked with the development of human values, where health and environmental protection are considered as basic values; general (exceedingly) fast development of technology which, inter alia, calls for a change of standardized methods of work (Brewer and Hsiang 2002); and economic differences and crises that promote changes in work technologies (Archibugi et al. 2013).

By classifying the studies according to technologies, it was established that the majority of studies were still conducted on »traditional« technologies (chainsaw + skidder). The result makes sense considering that the majority of wood is still produced by using these technologies on a global scale, and that the work in this case is more dangerous (Potočnik et al. 2009, Tsioras et al.

2011). On the other hand, it seems that there has been no transfer of knowledge, for example, on traditional ergonomics, as one of the challenges in 2007 (Heinimann 2007). The reason for this failure may be in linguistic barriers, inaccessibility of older literature, and still insufficient international, and in particular, as the results indicate, intercontinental cooperation.

In terms of participation in studies, three fields of ergonomics stand out, namely: vibrations, cardiovascular workloads and working postures. It is obvious that the exposure to HA and WB vibrations still remains the central problem of forestry, since both in harvesting and extraction, regardless of the technology used, the vibrations exceed action values of permissible daily exposure. Similar also applies to unfavorable body positions during work, which together with other factors result in musculoskeletal disorders. If the studies of exposure to vibrations and noise are compared in terms of their number, obviously the number of the latter is significantly lower. The reason may be ascribed to a lower number of machines that cause overloads and to the fact that protection measures are more efficient and relatively simpler. Thus, for example, a regular use of personal protection equipment and noise insulated machine cabins may significantly reduce the exposure, without affecting the work productivity. On the contrary, the measures for reducing

3.3.6.1 Safety The review of previous studies provided only one article addressing wood transport. Authors established that the safety of wood transport with trucks was lower, although the number of accidents due to mechanical faults and alcohol decreased. From 1991 to 2003, the frequency of accidents increased from 11 to 19 accidents per 1 million of transported wood (Greene et al. 2007).

4. Discussion The greatest drawback of this research is that its scope is limited only to journals with impact factor. This means that, in addition to journals not included in the Web of Science collection, all grey literature was also left out, although it definitively includes important ergonomics project reports, such as Vibrisks, Ergowood and Comfor (EC 2016). Not observing this part of literature may definitely have an impact on our result, in particular if considering its purpose. Thus, for example, the projects are mostly focused on establishing guidelines and may cover very narrow, specific fields or areas. On the contrary, the publishing in journals with impact factor (IF journals) often depends on the popularity, originality and generality of their content.

Despite the aforementioned drawbacks, the research is believed to have succeeded in showing the focal points of studies in the field of occupational safety and health in forestry in the period 2005–2016. The complexity of results of these studies published in IF journals ensures to some extent their quality and current relevance in a broader research community. In addition to assessing the current state and potentially uncovering fields of interests, the research results represent a referential point for all similar future research and also the identification of trends in the development of occupational safety and ergonomics in forestry.

The research established that studies are mostly limited only to three continents. If the above mentioned issue of the scope of research is taken into account, the result of the spatial distribution of studies depends on actual capacities, such as research funding, tradition, education and staff potential, and also on the needs of researchers to be published in IF journals. The needs to be published are greater in countries where publications are crucial for an academic or research career, and directly or indirectly related to the funding of program groups and research (Rijcke et al. 2016).

Similar to Cavalli (2012) in its review paper on cable yarding, this research established that the number of

302

Croat. j. for. eng. 38(2017)2

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

the exposures to vibrations, psychophysical workload and unfavorable body positions often decrease the work productivity (Gallis 2013). To efficiently reduce exposure to vibrations without significantly affecting the productivity, it is necessary to replace technologies, e.g. chainsaw with harvester or skidder with forwarder, which however is not always possible because of working conditions (Mihelič and Krč 2009). The second option is to change the operators behavior (Tiemessen et al. 2007), which must be correlated with the method of payment (Johansson et al. 2010).

A great share of manual work in forestry occurring in all phases of wood production, and prevailing in pre-harvesting tasks, traditional harvesting and firewood production, causes a high frequency of studies addressing cardiovascular workloads. Cardiovascular loads frequently exceed permitted limits. Due to relative simplicity, the measurement of cardiovascular loads is frequently included in broader, complex work studies (Košir et al. 2015).

On the other hand, there are no or very few studies in some fields. Thus, no studies referring to manuals and instructions or maintenance were detected, although the studies indicate that the maintenance of machine, for example in mechanized harvesting, is one of the main reasons for the dissatisfaction of operators (Walker et al. 2005) and poses the highest risk of accident (Väyrynen 1982). Similar applies to the exposure to biological substances, although working in natural environment poses a constant health hazard as a consequence of operating with flora and fauna.

One of the reasons for the aforementioned fields not being included in the research is the above mentioned scope of research. Other reasons for not including these field may be an a priori underestimation of the risk factor significance (e.g. maintenance), inaccessibility of data (e.g. maintenance index) and blind trust (e.g. to producers).

Similar to ergonomics studies, also the studies addressing the occupational safety differentiate in their representation of individual forestry operations and content of research. The main reason may lie in the nature of work, risks of accident and accessibility of data. Thus, most studies have been conducted in forestry in general, which may be ascribed to the nature of work of forest workers, who take part in several work phases of wood production (e.g. harvesting and extracting) simultaneously, which is in particular true for working in groups (Poje and Potočnik 2007). In terms of individual operations, most studies deal with chainsaw felling, i.e. the working operation with the highest accident risk (Potočnik et al. 2009). Since nonlethal accidents are over 100-times more frequent than Croat. j. for. eng. 38(2017)2

I. Potočnik and A. Poje

the lethal ones, the accidents at work were addressed almost in all technologies concerned, while the lethal accidents were dealt with only in general.

5. Conclusions and recommendations According to our knowledge, this research is the first attempt to review the articles published in journals with impact factor in the field of occupational safety and ergonomics in forestry. Notwithstanding the drawbacks of the respective research, we are convinced that it succeeded in showing the research situation in terms of time dynamics, spatial scope and classification according to technologies and risk factors.

On the basis of the results, we thus assume that the number of publications in journals with the impact factor will also increase in the future, namely due to an increasing number of forestry and ergonomics journals with the impact factor, uniformity of the evaluation of research performance and also development of technologies and measurement techniques. By increasing the number of publications, the studies will also increase their spatial scope, but only under the condition that the number of researchers be increased and international cooperation and education improved.

The up-coming technological development of the existing technologies and automation and robotization of the forest work will require more emphasis on ergonomics research of cognitive risk factors. The robotization of work will even enhance the issue of mental overloads, which is already acute during current CTL harvesting (Berger 2003), and through the extension of a work day increase the possibility of ergonomic traps (Synwoldt and Gellerstedt 2003). At the same time, a new paradigm »more can be better« will have to be considered, since the studies establish that too low physical activity can have a negative impact on the health of workers (Straker and Mathiassen 2009). Development of new technologies, adjustment of work organization to technological progress (Heinimann 2007) and escalation of competition between manual and mechanized wood production will still require a continuous control of occupational safety and physical, physiological and environmental risk factors. It is expected that new requirements, primarily intended to protect the environment (e.g. reduction of emissions, use of alkylate fuels), will also have a positive impact on the health hazard (Neri et al. 2016).

Due to the technological development of measurement techniques and international projects, it is expected that measurements of ergonomics factors will be more often included in broader, more complex studies of the impact of wood production technologies on the efficiency, environment and people.

303

I. Potočnik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

Further studies need to be directed in supplementing this research and improving the knowledge on individual technologies and fields that have been identified as insufficiently explored. To complete the overall picture, the research must also include the articles in the field of medicine and grey literature, as well as a study for the period before 2005. No analyses of lethal accidents according to individual phases of wood production have been found in the studies of occupational safety. There are also no risk analyses that would be based on a comparison between injured and non-injured workers (or workers with one or multiple injuries) or hazardous or non-hazardous working conditions.

This would provide the answers to the question: »Which of the factors, such as environmental, organizational or personal, increase the accident probability?«.

In terms of technologies and fields, additional studies are needed in particular in the fields of wood transport, biomass production, construction and maintenance of roads and skid trails, and new technologies, such as battery powered tools. According to individual fields of ergonomics, there is an insufficient number of: cognitive studies in all work phases of wood production (Häggström 2010); studies of burnout workers due to overload that additionally increases the risk of accident (Ahola et al. 2013); comprehensive ergonomics studies similar to Gerasimov and Sokolov (2014) and studies of productivity that would take into account all ergonomics requirements for safe and healthy work in setting the standards. To successfully reduce accident risks, it is first necessary to prepare and harmonize the statistics of forestry data collections. Only this will provide the collection of suitable information and in-depth analyses to prepare preventive measures for individual fields. In terms of the objective of this research, the impact and loads of forest production and wood industry (e.g. noise, dust) on wildlife and urban centers still remains completely uncovered. Due to high environmental awareness, the care for healthy natural habitats is one of the fields of future studies.

6. References Aalmo, G.O., Magagnotti, N., Spinelli, R., 2016: Forest workers and steep terrain winching: The impact of environmental and anthropometric parameters on performance. Croatian Journal of Forest Engineering 37(1): 97–105.

Ackerknecht, C., Bassaber, C., Reyes, M., Miranda, H., 2005: Environmental certification systems and impacts of their implementation on occupational health and safety in Chilean forest companies. New Zealand Journal of Forestry Science 35(2–3): 153–165.

304

Adams, G., Armstrong, H., Cosman, M., 2014: Independent forestry safety review – An agenda for change in the forestry sector. http://fica.org.nz/wp-content/uploads/2014/10/IFSRReportSummary-Web.pdf, 12 p.

Ahola, K., Salminen, S., Toppinen-Tanner, S., Koskinen, A., Vaananen, A., 2013: Occupational burnout and severe injuries: an eight-year prospective cohort study among Finnish forest industry workers. Journal of Occupational Health 55(6): 450–457.

Albizu-Urionabarrenetxea, P.M., Tolosana-Esteban, E., Roman-Jordan, E., 2013: Safety and health in forest harvesting operations. Diagnosis and preventive actions. A review. Forest Systems 22(3): 392–400. doi:10.5424/fs/2013223-02714 Albizu-Urionabarrenetxea, P.M., Tolosana Esteban, E., Ulecia Zaldívar, J., Fernández Carretero, M., 2010: Diagnosis of safety in forest harvesting operations from company records, official databases and land surveys. Improvement action plan. Forest Systems 19(2): 221–233.

Almeida, S.F., Abrahão, R.F., Andrade Tereso, M.J., 2015: Evaluation of occupational exposure to whole body vibration in forest harvesting machines. Cerne 21(1): 1–8. doi:10.1590/ 01047760201521011446 Alves, J.U., De Souza, A.P., Minette, L.J., Gomes, J.M., Da Silva, K.R., Marçal, M.A., Da Silva, E.P., 2006: Biomechanical evaluation of activities of Eucalyptus ssp. seedling production. Revista Arvore 30(3): 331–335. doi:10.1016/j.desal.2006.03.078 Archibugi, D., Filippetti, A., Frenz, M., 2013: The impact of the economic crisis on innovation: Evidence from Europe.

Technological Forecasting and Social Change 80(7): 1247– 1260. doi:http://dx.doi.org/10.1016/j.techfore.2013.05.005 Axelsson, S.A., 1998: The mechanization of logging operations in Sweden and its effect on occupational safety and health. Journal of Forest Engineering 9(2): 25–31.

Axelsson, S.A., Pontén, B., 1990: New ergonomic problems in mechanized logging operations. International Journal of Industrial Ergonomics 5(3): 267–273. doi:http://dx.doi. org/10.1016/0169-8141(90)90062-7 Barbosa, R.P., Fiedler, N.C., do Carmo, F.C.A., Minette, L.J., Silva, E.N., 2014: Analysis of posture in semi-mechanized forest harvesting in steep areas. Revista Arvore 38(4): 733–738 doi:0.1590/S0100-67622014000400016 Bell, J.L., Grushecky, S.T., 2006: Evaluating the effectiveness of a logger safety training program. Journal of Safety Research 37(1): 53–61. doi:10.1016/j.jsr.2005.10.019 Bentley, T.A., Parker, R.J., Ashby, L., 2005: Understanding felling safety in the New Zealand forest industry. Applied Ergonomics 36(2): 165–175. doi:10.1016/j.apergo.2004.10.009 Berger, C., 2003: Mental stress on harvester operators. In: Limbeck-Lilienau, Steinmüller, Stampfer (eds) Proceedings of the Austro2003 meeting: High Tech Forest Operations for Mountainous Terrain. Schlaegl, Austria, 10 p.

Beuk, D., Tomašić, Ž., Horvat, D., 2007: Status and development of forest harvesting mechanisation in Croatian state forestry. Croatian Journal of Forest Engineering 28(1): 63–82.

Croat. j. for. eng. 38(2017)2

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310) Bont, L., Heinimann, H.R., 2012: Optimum geometric layout of a single cable road. European Journal of Forest Research 131(5): 1439–1448. doi:10.1007/s10342-012-0612-y

I. Potočnik and A. Poje

statures at planting and fertilizing forest activities. Scientia Forestalis/Forest Sciences 42(102): 191–196.

Bovenzi, M., 2008: A follow up study of vascular disorders in vibration-exposed forestry workers. Int Arch Occup Environ Health 81(4): 401–408. doi:10.1007/s00420-007-0225-9

de Oliveira, F.M., Lopes, E.S., Rodrigues, C.K., 2014: Evaluation of the physical work load and biomechanical of workers at manual and semi-mechanized mowing. Cerne 20(3): 419– 425. doi:10.1590/01047760201420031431

Brewer, J.D., Hsiang, S.M., 2002: The ‘ergonomics paradigm’: Foundations, challenges and future directions. Theoretical Issues in Ergonomics Science 3(3): 285–305. doi:10.1080/14639220110114681

de Souza, A.P., Dutra, R.B.C., Minette, L.J., Marzano, F.L.C., Schettino, S., 2015: Production targets for workers in forest harvesting. Revista Arvore 39(4): 713–722. doi:10.1590/010067622015000400014

Brzózko, J., 2016: Work safety at timber harvesting in private forests in Poland. Sylwan 160(7): 591–596.

DeMille, G.J., Lyons, K., 2016: Unexpected events when manually falling trees in coastal British Columbia. Forest Science 62(4): 433–439. doi:10.5849/forsci.15-164

Burman, L., Löfgren, B., 2007: Human-machine interaction improvements of forest machines. Skogforsk – The Forestry Research Institute of Sweden. http://www.nordiskergonomi. org/nes2007/CD_NES_2007/papers/A26_Burman.pdf. Accessed December 26, 2016.

Burt, C.D.B., Chmiel, N., Hayes, P., 2009: Implications of turnover and trust for safety attitudes and behaviour in work teams. Safety Science 47(7): 1002–1006. doi:10.1016/j. ssci.2008.11.001 Burt, C.D.B., Sepie, B., McFadden, G., 2008: The development of a considerate and responsible safety attitude in work teams. Safety Science 46(1): 79–91. doi:10.1016/j. ssci.2006.10.005 Cation, S., Jack, R., Oliver, M., Dickey, J.P., Lee-Shee, N.K., 2008: Six degree of freedom whole-body vibration during forestry skidder operations. International Journal of Industrial Ergonomics 38(9–10): 739–757. doi:10.1016/j.ergon.2007.10.003

Denbeigh, K., Slot, T.R., Dumas, G.A., 2013: Wrist postures and forces in tree planters during three tree unloading conditions. Ergonomics 56(10): 1599–1607. doi:10.1080/00140139.2 013.824615 Dubé, P.A., Imbeau, D., Dubeau, D., Auger, I., Leone, M., 2015: Prediction of work metabolism from heart rate measurements in forest work: some practical methodological issues. Ergonomics 58(12): 2040–2056. doi:10.1080/00140139.2 015.1044920 Dubé, P.A., Imbeau, D., Dubeau, D., Lebel, L., Kolus, A., 2016: Removing the thermal component from heart rate provides an accurate VO2 estimation in forest work. Applied Ergonomics 54: 148–157. doi:10.1016/j.apergo.2015.12.005 Dupire, S., Bourrier, F., Berger, F., 2016: Predicting load path and tensile forces during cable yarding operations on steep terrain. Journal of Forest Research 21(1): 1–14. doi:10.1007/ s10310-015-0503-4

Cavalli, R., 2012: Prospects of research on cable logging in forest engineering community. Croatian Journal of Forest Engineering 33(2): 339–356.

EC, 2016: CORDIS – Communitiy research and development information service. http://cordis.europa.eu/guidance/home_ en.html. Accessed December 25, 2016.

Chapman, L.J., Husberg, B., 2008: Agriculture, Forestry, and Fishing Sector. Journal of Safety Research 39(2): 171–173. doi:10.1016/j.jsr.2008.02.008

Ehsani, J.P., McNeilly, B., Ibrahim, J.E., Ozanne-Smith, J., 2013: Work-related fatal injury among young persons in Australia, July 2000 – June 2007. Safety Science 57: 14–18. doi:10.1016/j.ssci.2013.01.012

Cohen, M.A., Clark, R.E., Silverstein, B., Sjostrom, T., Spielholz, P., 2006: Work-related deaths in Washington State, 1998–2002. Journal of Safety Research 37(3): 307–319. doi:10.1016/j.jsr.2006.02.007 Da Silva, E.P., Minette. L.J., De Souza, A.P., 2007: Evaluation an ergonomic of the activity of semi mechanized pit for the planting of eucalyptus. Scientia Forestalis/Forest Sciences 35(76): 77–83.

Dąbrowski, A., 2015: Kickback risk of portable chainsaws while cutting wood of different properties: Laboratory tests and deductions. International Journal of Occupational Safety and Ergonomics 21(4): 512–523. doi:10.1080/10803548.2015.1 095547

Enez, K., Topbas, M., Acar, H.H., 2014: An evaluation of the occupational accidents among logging workers within the boundaries of trabzon forestry directorate, Turkey. International Journal of Industrial Ergonomics 44(5): 621–628. doi:10.1016/j.ergon.2014.07.002 ErgoWood, 2006: European ergonomic and safety guidelines for forest machines. Swedish University of Agricultural Sciences, Uppsala, Sweden, 100 p.

EU-OSHA, 2008: Occupational safety and health in Europe’s forestry industry. European agency for safety and health at work. https://osha.europa.eu/en/publications/e-facts/efact29/ view. Accessed December 28, 2016.

Dąbrowski, A., 2012: Reducing kickback of portable combustion chain saws and related injury risks: Laboratory tests and deductions. International Journal of Occupational Safety and Ergonomics 18(3): 399–417.

Fernandes, H.C., Brito, A.B.D., Minette, L.J., Leite, D.M., Leite, E.D.S., 2011: Application of ergonomic index indices in the evaluation of the cabin of a forest tractor »Feller-Buncher«. Scientia Forestalis/Forest Sciences 39(90): 273–281.

De Britto, P.C., Da Silva, E.P., Lopes, E., De Laat, E.F., Fiedler, N.C., 2014: Biomechanical evaluation in workers of different

Fernandes, H.C., De Brito, A.B., Santos, N.T., Minette, L.J., Rinaldi, E.P.C.N., 2009: Anthropometric analyses of a group

Croat. j. for. eng. 38(2017)2

305

I. Potočnik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

of Brazilian feller-buncher operators. Scientia Forestalis/Forest Sciences 37(81): 17–25.

vian Journal of Forest Research 25(suppl. 9): 25–32. doi:10.10 80/02827581.2010.506784

Ferrari, E., Spinelli, R., Cavallo, E., Magagnotti, N., 2012: Attitudes towards mechanized Cut-to-Length technology among logging contractors in Northern Italy. Scandinavian Journal of Forest Research 27(8): 800–806. doi:10.1080/02827 581.2012.693192

Häggström, C., 2010: Human factors in forest harvester operation. Arbetsrapport NR-728-2010. Skogforsk, Uppsala, 26 p.

Fonseca, A., Aghazadeh, F., de Hoop, C., Ikuma, L., Al-Qaisi, S., 2015: Effect of noise emitted by forestry equipment on workers’ hearing capacity. International Journal of Industrial Ergonomics 46: 105–112. doi:10.1016/j.ergon.2014.05.001 Fontana, G., Seixas, F., 2007: Ergonomic evaluation of the workstation of forwarder and skidder models. Revista Arvore 31(1): 71–81. doi:10.1590/S0100-67622007000100009 Gallis, C., 2006: Work-related prevalence of musculoskeletal symptoms among Greek forest workers. International Journal of Industrial Ergonomics 36(8): 731–736. doi:10.1016/j. ergon.2006.05.007 Gallis, C., 2013: Increasing productivity and controlling of work fatigue in forest operations by using prescribed active pauses: A selective review. Croatian Journal of Forest Engineering 34(1): 103–112.

Gerasimov, Y., Sokolov, A., 2009: Ergonomic characterization of harvesting work in Karelia. Croatian Journal of Forest Engineering 30(2): 159–170.

Gerasimov, Y., Sokolov, A., 2014: Ergonomic evaluation and comparison of wood harvesting systems in Northwest Russia. Applied Ergonomics 45(2): 318–338. doi:10.1016/j.apergo.2013.04.018 Goglia, V., Suchomel, J., Žgela, J., Dukić, I., 2012a: The effectiveness of forest pre-commercial thinning in the context of Directive 2002/44/EC. Šumarski List 136(9–10): 471–478.

Goglia, V., Suchomel, J., Žgela, J., Dukić, I., 2012b: Forestry workers’ expossure to vibration in the context of Directive 2002/44/EC. Šumarski List 136(5–6): 283–289.

Goglia, V., Žgela, J., Dukić, I., 2008a: The effectiveness of antivibration gloves: Part I. Šumarski List 132(3–4): 115–119.

Goglia, V., Žgela, J., Dukić, I., 2008b: The effectiveness of anti-vibration gloves: Part II. Šumarski List 132(5–6): 239–244.

Greene, W.D., Baker, S.A., Lowrimore, T., 2007: Analysis of Log Hauling Vehicle Accidents in the State of Georgia, USA, 1988–2004. International Journal of Forest Engineering 18(2): 52–57. doi:10.1080/14942119.2007.10702550 Grzywiński, W., Sawa, L., Nowik, A., Nowicki, G., 2013: Structure of work accidents in the Regional Directorate of the State Forests in Szczecinek in the years 1990–2009. Sylwan 157(6): 403–411.

Grzywiński, W., Wandycz, A., Tomczak, A., Jelonek, T., 2016: The prevalence of self-reported musculoskeletal symptoms among loggers in Poland. International Journal of Industrial Ergonomics 52: 12–17. doi:10.1016/j.ergon.2015.07.003 Häggqvist, P., Lejon, S.B., Lidestav, G., 2010: Forest days as an educational method in Swedish family forestry. Scandina-

306

Häggström, C., Englund, M., Lindroos, O., 2015: Examining the gaze behaviors of harvester operators: an eye-tracking study. International Journal of Forest Engineering 26(2): 96– 113. doi:10.1080/14942119.2015.1075793 Häggström, C., Lindroos, O., 2016: Human, technology, organization and environment – a human factors perspective on performance in forest harvesting. International Journal of Forest Engineering 27(2): 67–78. doi:10.1080/14942119.2016.1 170495 Häggström, C., Öhman, M., Burström, L., Nordfjell, T., Lindroos, O., 2016: Vibration exposure in forwarder work: Effects of work element and grapple type. Croatian Journal of Forest Engineering 37(1): 107–118.

Hanse, J.J., Winkel, J., 2008: Work organisation constructs and ergonomic outcomes among European forest machine operators. Ergonomics 51(7): 968–981. doi:10.1080/00140130801961893 Hartter, J., Garland, J., 2006: Synthetic Rope End Connections for Use in Timber Harvesting. International Journal of Forest Engineering 17(1): 39–51. doi:10.1080/14942119.2006.1070252 8 Heinimann, H.R., 2007: Forest operations engineering and management – The ways behind and ahead of a scientific discipline. Croatian Journal of Forest Engineering 28(1): 107–121.

Hellström, T., Johansson, T., Ringdahl, O., 2006: Development of an Autonomous Forest Machine for Path Tracking. In: Corke, P., Sukkariah, S., (eds) Field and Service Robotics: Results of the 5th International Conference. Springer Berlin Heidelberg, Berlin, Heidelberg, p. 603–614. doi:10.1007/978-3540-33453-8_50 Hollnagel, E., Woods, D.D., 2005: Joint cognitive systems: foundations of cognitive systems engineering. Taylor & Francis, Boca Raton, USA, 240 p.

Horvat, D., Čavlović, A., Zečić, Ž., Šušnjar, M., Bešlić, I., Madunić-Zečić V., 2005: Research of fir-wood dust concentration in the working environment of cutters. Croatian Journal of Forest Engineering 26(2): 85–90.

Huysmans, M.A., de Looze, M.P., Hoozemans, M.J.M., van der Beek, A.J., van Dieën, J.H., 2006: The effect of joystick handle size and gain at two levels of required precision on performance and physical load on crane operators. Ergonomics 49(11): 1021–1035. doi:10.1080/00140130500424102 Jack, R.J., Oliver, M., 2008: A Review of Factors Influencing Whole-Body Vibration Injuries in Forestry Mobile Machine Operators. International Journal of Forest Engineering 19(1): 51–65. doi:10.1080/14942119.2008.10702560 Jack, R.J., Oliver, M., Dickey, J.P., Cation, S., Hayward, G., Lee-Shee, N., 2010: Six-degree-of-freedom whole-body vibration exposure levels during routine skidder operations. Ergonomics 53(5): 696–715. doi:10.1080/00140130903581631 Croat. j. for. eng. 38(2017)2

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310) Ji, X., Eger, T.R., Dickey, J.P., 2015: Development of a seat selection algorithm to match industrial seats with specific forestry vibration exposures. International Journal of Forest Engineering 26(1): 48–59. doi:10.1080/14942119.2015.1007631 Ji, X., Eger, T.R., Dickey, J.P., 2017: Evaluation of the vibration attenuation properties of an air-inflated cushion with two different heavy machinery seats in multi-axis vibration environments including jolts. Applied Ergonomics 59: 293–301. doi:10.1016/j.apergo.2016.06.011 Johansson, B., Rask, K., Stenberg, M., 2010: Piece rates and their effects on health and safety – A literature review. Applied Ergonomics 41(4): 607–614. doi:10.1016/j.apergo.2009.12.020 Klun, J., Medved, M., 2007: Fatal accidents in forestry in some European countries. Croatian Journal of Forest Engineering 28(1): 55–62.

Košir, B., Magagnotti, N., Spinelli, R., 2015: The role of work studies in forest engineering: status and perspectives. International Journal of Forest Engineering 26(3): 160–170. doi:10 .1080/14942119.2015.1111043

I. Potočnik and A. Poje

ysis and Prevention 42(6): 1729–1735. doi:10.1016/j. aap.2010.04.013 Lindroos, O., Gullberg, T., Nordfjell, T., 2007: Torques from Manual Tools for Directional Tree Felling. International Journal of Forest Engineering 18(2): 40–51. doi:10.1080/14942119 .2007.10702549 Lyons, C.K., 2015: New development of a remotely operated falling wedge. Forestry Chronicle 91(2): 176–181. doi:10.5558/ tfc2015-028 Lyons, C.K., Demille, G., 2015: Management requiring conditions when manually falling trees. Forestry Chronicle 91(3): 299–305. doi:10.5558/tfc2015-051 Lyons, C.K., Noll, F., 2011: Optimizing a novel method for manual tree falling. Forestry Chronicle 87(4): 537–541. doi:10.5558/tfc2011-052 Lyons, K., Ewart, J., 2012: The wood duck: A new tree falling tool. Western Journal of Applied Forestry 27(3): 137–142. doi:10.5849/wjaf.11-026

Landekić, M., 2010: Organizational culture and occupational safety in the Croatian forestry sector. Šumarski List 134(11– 12): 613–622.

Magagnotti, N., Aalmo, G.O., Brown, M., Spinelli, R., 2016: A new device for reducing winching cost and worker effort in steep terrain operations. Scandinavian Journal of Forest Research 31(6): 602–610. doi:10.1080/02827581.2015.1133845

Landekić, M., Martinić, I., Bakarić, M., Šporčić, M., 2013: Work ability index of forestry machine operators and some ergonomic aspects of their work. Croatian Journal of Forest Engineering 34(2): 241–254.

Magagnotti, N., Spinelli, R., 2012: Replacing Steel Cable with Synthetic Rope to Reduce Operator Workload During Log Winching Operations. Small-Scale Forestry 11(2): 223–236. doi:10.1007/s11842-011-9180-0

Laschi, A., Marchi, E., Foderi, C., Neri, F., 2016: Identifying causes, dynamics and consequences of work accidents in forest operations in an alpine context. Safety Science 89: 28–35. doi:10.1016/j.ssci.2016.05.017

Malinowska-Borowska, J., Harazin, B., Zieliński, G., 2011: The influence of wood hardness and logging operation on coupling forces exerted by lumberjacks during wood harvesting. International Journal of Industrial Ergonomics 41(5): 546–550. doi:10.1016/j.ergon.2011.06.001

Leshchinsky, B., Sessions, J., Wimer, J., 2015: Analytical design for mobile anchor systems. International Journal of Forest Engineering 26(1): 10–23. doi:10.1080/14942119.2015.1023 014 Leszczyñski, K., 2010: Different evaluations of motor-manual wood harvesting processes on the basis of conjoint analysis. Croatian Journal of Forest Engineering 32(2): 165–172.

Leszczyński, K., Stańczykiewicz, A., 2015: Workload analysis in logging technology employing a processor aggregated with a farm tractor. Forest Systems 24(2): 1–8. doi:10.5424/ fs/2015242-06607 Lindqvist, A., Nilsson, O., 2011: Hand injury from powered wood splitters: Machine safety, patterns of use and injury events. International Journal of Occupational Safety and Ergonomics 17(2): 175–186.

Lindroos, O., 2008: The Effects of increased mechanization on time consumption in small-scale firewood processing.

Silva Fennica 42(5): 791–805.

Lindroos, O., Aspman, E.W., Lidestav, G., Neely, G., 2008: Accidents in family forestry’s firewood production. Accident Analysis and Prevention 40(3): 877–886. doi:10.1016/j. aap.2007.10.002 Lindroos, O., Burström, L., 2010: Accident rates and types among self-employed private forest owners. Accident AnalCroat. j. for. eng. 38(2017)2

Malinowska-Borowska, J., Harazin, B., Zieliński, G., 2012: Measuring coupling forces woodcutters exert on saws in real working conditions. International Journal of Occupational Safety and Ergonomics 18(1): 77–83.

Malinowska-Borowska, J., Zieliński, G., 2013: Coupling forces exerted on chain saws by inexperienced tree fellers. International Journal of Industrial Ergonomics 43(4): 283–287. doi:10.1016/j.ergon.2013.04.006 Martinić, I., Landekić, M., Šporčić, M., Lovrić, M., 2011: Forestry at the eu’s doorstep – how much are we ready in the area of occupational safety in forestry? Croatian Journal of Forest Engineering 32(1): 431–441.

Martinić, I., Vondra, V., Šporčić, M., 2007: Development of a new concept for improvement of forest techniques in Croatia – Areas of possible contributions. Croatian Journal of Forest Engineering 28(1): 47–54.

Melemez, K., 2015: Risk factor analysis of fatal forest harvesting accidents: A case study in Turkey. Safety Science 79: 369–378. doi:10.1016/j.ssci.2015.07.004 Messingerová, V., Martinusová, L., Slančík, M., 2005: Ergonomic parameters of the work of integrated technologies at timber harvesting. Croatian Journal of Forest Engineering 26(2): 79–84.

307

I. Potočnik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

Mihelič, M., Krč, J., 2009: Analysis of inclusion of wood forwarding into a skidding model. Croatian Journal of Forest Engineering 30(2): 113–125.

Montorselli, N.B., Lombardin, C., Magagnotti, N., Marchi, E., Neri, F., Picchi, G., Spinelli, R., 2010: Relating safety, productivity and company type for motor-manual logging operations in the Italian Alps. Accident Analysis and Prevention 42(6): 2013–2017. doi:10.1016/j.aap.2010.06.011 Mylek, M.R., Schirmer, J., 2015: Beyond physical health and safety: Supporting the wellbeing of workers employed in the forest industry. Forestry 88(4): 391–406. doi:10.1093/forestry/ cpv011 Neely, G., Wilhelmson, E., 2006: Self-reported incidents, accidents, and use of protective gear among small-scale forestry workers in Sweden. Safety Science 44(8): 723–732. doi:10.1016/j.ssci.2006.03.002 Neri, F., Foderi, C., Laschi, A., Fabiano, F., Cambi, M., Sciarra, G., Aprea, M.C., Cenni, A., Marchi, E., 2016: Determining exhaust fumes exposure in chainsaw operations. Environmental pollution 218: 1162–1169. doi:10.1016/j. envpol.2016.08.070 Neyland, M., Hickey, J., Read, S.M., 2012: A synthesis of outcomes from the Warra Silvicultural Systems Trial, Tasmania: Safety, timber production, economics, biodiversity, silviculture and social acceptability. Australian Forestry 75(3): 147– 162.

Noll, F., Lyons, C.K., 2010: A novel method for manually falling trees. Forestry Chronicle 86(5) :608–613.

Nutto, L., Malinovski, R.A., Brunsmeier, M., Schumacher Sant’Anna, F., 2013: Ergonomic aspects and productivity of different pruning tools for a first pruning lift of Eucalyptus grandis hill ex maiden. Silva Fennica 47(4): 1–14. doi:10.14214/ sf.1026 O’Neal, B., Shaffer, R., Rummer, R., 2007: Assessing the safety training needs of Spanish-speaking workers in the southeastern logging industry. Southern Journal of Applied Forestry 31(3): 124–128. Østensvik, T., Nilsen, P., Veiersted, K.B., 2008a: Muscle Activity Patterns in the Neck and Upper Extremities Among Machine Operators in Different Forest Vehicles. International Journal of Forest Engineering 19(2): 11–20. doi:10.1080/1494 2119.2008.10702563 Østensvik, T., Veiersted, K.B., Cuchet, E., Nilsen, P., Hanse, J.J., Carlzon, C., Winkel, J., 2008b: A search for risk factors of upper extremity disorders among forest machine operators: A comparison between France and Norway. International Journal of Industrial Ergonomics 38(11–12): 1017–1027. doi:10.1016/j.ergon.2008.01.016 Østensvik, T., Veiersted, K.B., Nilsen, P., 2009: Association between numbers of long periods with sustained low-level trapezius muscle activity and neck pain. Ergonomics 52(12): 1556–1567. doi:10.1080/00140130903199889 Ottaviani, G., Talbot, B., Nitteberg, M., Stampfer, K., 2011: Workload Benefits of using a synthetic rope strawline in cable

308

yarder rigging in Norway. Croatian Journal of Forest Engineering 32(2): 561–569.

Ovaskainen, H., Heikkilä, M., 2007: Visuospatial cognitive abilities in cut-to-length single-grip timber harvester work.

International Journal of Industrial Ergonomics 37(9–10): 771–780. doi:10.1016/j.ergon.2007.06.004 Pandur, Z., Horvat, D., Šušnjar, M., Zorić, M. 2013: Possibility of determination of daily exposure to vibration of skidder drivers using fleet manager system. Croatian Journal of Forest Engineering 34(2): 305–310.

Papadopoulos, G., Georgiadou, P., Papazoglou, C., Michaliou, K., 2010: Occupational and public health and safety in a changing work environment: An integrated approach for risk assessment and prevention. Safety Science 48(8): 943–949. doi:10.1016/j.ssci.2009.11.002 Penna, E.S., Machado, C.C., de Souza, A.P., Silva, E., da Silva, E.N., 2011: Ergonomic evaluation of skyline models used in forest extraction. Revista Arvore 35(3): 565–571. doi:10.1590/ S0100-67622011000300019 Pereira, D.P., Fiedler, N.C., De Souza Lima, J.S., De Oliveira Bauer, M., Rezende, A.V., Missiaggia, A.A., Simão, J.B.P., 2011: Lateral stability limits of farm tractors for forest plantations in steep areas. Scientia Forestalis/Forest Sciences 92: 433–439.

Phairah, K., Brink, M., Chirwa, P., Todd, A., 2016: Operator work-related musculoskeletal disorders during forwarding operations in South Africa: an ergonomic assessment. Southern Forests 78(1): 1–9. doi:10.2989/20702620.2015.1126781 Poje, A., Potočnik, I., 2007: Influence of working conditions on overlapping of cutting and ground skidding in group work. Croatian Journal of Forest Engineering 28(2): 157–167.

Poje, A., Potočnik, I., Košir, B., Krč, J., 2016: Cutting patterns as a predictor of the odds of accident among professional fellers. Safety Science 89: 158–166. doi:10.1016/j. ssci.2016.06.011 Poje, A., Spinelli, R., Magagnotti, N., Mihelic, M., 2015: Exposure to noise in wood chipping operations under the conditions of agro-forestry. International Journal of Industrial Ergonomics 50: 151–157. doi:10.1016/j.ergon.2015.08.006 Poom, L., Löfroth, C., Nordén, B., Thor, M., 2007: Testing Human Visual Detection with Xenon and Halogen Lamps as Used on Forest Machines. International Journal of Forest Engineering 18(2): 9–14. doi:10.1080/14942119.2007.10702545 Potočnik, I., Pentek, T., Poje, A., 2009: Severity analysis of accidents in forest operations. Croatian Journal of Forest Engineering 30(2): 171–184.

Rehn, B., Lundström, R., Nilsson, L., Liljelind, I., Järvholm, B., 2005: Variation in exposure to whole-body vibration for operators of forwarder vehicles – Aspects on measurement strategies and prevention. International Journal of Industrial Ergonomics 35(9): 831–842. doi:10.1016/j.ergon.2005.03.001 Rehn, B., Nilsson, T., Lundström, R., Hagberg, M., Burström, L., 2009: Neck pain combined with arm pain among professional drivers of forest machines and the association with Croat. j. for. eng. 38(2017)2

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310) whole-body vibration exposure. Ergonomics 52(10): 1240– 1247. doi:10.1080/00140130902939889 Rickards, J., 2008: The human factor in forest operations: Engineering for health and safety. Forestry Chronicle 84(4): 539–542.

Rijcke, S.D., Wouters, P.F., Rushforth, A.D., Franssen, T.P., Hammarfelt, B., 2016: Evaluation practices and effects of indicator use – a literature review. Research Evaluation 25(2): 161–169. doi:10.1093/reseval/rvv038 Ringdahl, O., Hellström, T., Lindroos, O., 2012: Potentials of possible machine systems for directly loading logs in cut-tolength harvesting. Canadian Journal of Forest Research 42(5): 970–985. doi:10.1139/x2012-036 Robert, K., Elisabeth, Q., Josef, B., 2015: Analysis of occupational accidents with agricultural machinery in the period 2008–2010 in Austria. Safety Science 72: 319–328. doi:10.1016/j. ssci.2014.10.004 Rottensteiner, C., Stampfer, K., 2013: Evaluation of operator vibration exposure to chainsaws equipped with a Kesper safety bar. Scandinavian Journal of Forest Research 28(2): 193–200. doi:10.1080/02827581.2012.706636 Rottensteiner, C., Tsioras, P., Neumayer, H., Stampfer, K., 2013: Vibration and noise assessment of tractor-trailer and truck-mounted chippers. Silva Fennica 47(5): 14 p. doi:10.14214/sf.984 Rottensteiner, C., Tsioras, P., Stampfer, K., 2012: Wood density impact on hand-arm vibration. Croatian Journal of Forest Engineering 33(2): 303–312.

Sasaki, R.S., Furtado Júnior, M.R., Leite, E.S., Souza, A.P., Teixeira, M.M., Fernandes, H.C., 2014: Ergonomic evaluation of knapsack sprayers used in forestry. Revista Arvore 38(2): 331–337. doi:10.1590/S0100-67622014000200013 Silayo, D.S.A., Kiparu, S.S., Mauya, E.W., Shemwetta, D.T.K., 2010: Working conditions and productivity under private and public logging companies in Tanzania. Croatian Journal of Forest Engineering 31(1): 65–74.

Silva, E.P., Minette, L.J., Sanches, A.L.P., de Souza, A.P., Silva, F.L., Mafra, S.C.T., 2014: Prevalence of musculoskeletal symptoms in forest harvesting machine operators. Revista Arvore 38(4): 739–745. doi:10.1590/S0100-67622014000400017 Silva, E.P.D., De Souza, A.P., Minette, L.J., Baeta, F.D.C., Vieira, E.H.A.N.F., 2008: Biomechanics evaluation of manual timber removal work in mountainous areas. Scientia Forestalis/Forest Sciences 36(79):231–235.

Spinelli, R., Aalmo, G.O., Magagnotti, N., 2015: The effect of a slack-pulling device in reducing operator physiological workload during log winching operations. Ergonomics 58(5): 781–790. doi:10.1080/00140139.2014.983184 Spinelli, R., Aminti, G., De Francesco, F., 2017: Postural risk assessment of mechanised firewood processing. Ergonomics 60(3): 375–383. doi:10.1080/00140139.2016.1172738 Stampfer, K., Leitner, T., Visser, R., 2010: Efficiency and ergonomic benefits of using radio controlled chokers in cable yarding. Croatian Journal of Forest Engineering 31(1): 1–9.

Croat. j. for. eng. 38(2017)2

I. Potočnik and A. Poje

Straker, L., Mathiassen, S.E., 2009: Increased physical work loads in modern work – A necessity for better health and performance? Ergonomics 52(10): 1215–1225. doi:10.1080/00140130903039101 Suchomel, J., Belanová, K., 2009: Influence of selected meteorological phenomena on work injury frequency in timber harvesting process. Croatian Journal of Forest Engineering 30(2): 185–191.

Suchomel, J., Belanová, K., Vlčková, M., 2013: Analysis of work accidents in selected activities in Slovakia, Czech Republic and Austria. Croatian Journal of Forest Engineering 34(2): 311–320.

Synwoldt, U., Gellerstedt, S., 2003: Ergonomic initiatives for machine operators by the Swedish logging industry. Applied Ergonomics 34(2): 149–156. doi:10.1016/S0003-6870(03)00006-1 Tiemessen, I.J., Hulshof, C.T.J., Frings-Dresen, M.H.W., 2007: An overview of strategies to reduce whole-body vibration exposure on drivers: A systematic review. International Journal of Industrial Ergonomics 37(3): 245–256. doi:10.1016/j. ergon.2006.10.021 Toupin, D., LeBel, L., Dubeau, D., Imbeau, D., Bouthillier, L., 2007: Measuring the productivity and physical workload of brushcutters within the context of a production-based pay system. Forest Policy and Economics 9(8): 104–1055. doi:10.1016/j.forpol.2006.10.001 Tsioras, P.A., 2010: Perspectives of the forest workers in Greece. IForest 3(5): 118–123. doi:10.3832/ifor0547-003 Tsioras, P.A., 2012: Status and Job Satisfaction of Greek Forest Workers. Small-Scale Forestry 11(1): 1–14. doi:10.1007/s11842011-9164-0 Tsioras, P.A., Rottensteiner, C., Stampfer, K., 2011: Analysis of accidents during cable yarding operations in Austria 19982008. Croatian Journal of Forest Engineering 32(2): 549–560.

Tsioras, P.A., Rottensteiner, C., Stampfer, K., 2014: Wood harvesting accidents in the Austrian State Forest Enterprise 2000–2009. Safety Science 62: 400–408. doi:10.1016/j. ssci.2013.09.016 Väyrynen, S., 1982: Occupational accidents in the maintenance of heavy forest machinery. Journal of Occupational Accidents 4(2-4): 175. doi:http://dx.doi.org/10.1016/03766349(82)90022-0 Village, J., Trask, C., Chow, Y., Morrison, J.B., Koehoorn, M., Teschke, K., 2012: Assessing whole body vibration exposure for use in epidemiological studies of back injuries: Measurements, observations and self-reports. Ergonomics 55(4): 415–424. doi:10.1080/00140139.2011.643243 Vosniak, J., Da Silva Lopes, E., Fiedler, N.C., Alves, R.T., Venâncio, D.L., 2010: Demanded physical effort and posture in semi mechanical hole-digging activity at forestry plantation. Scientia Forestalis/Forest Sciences 33(88): 589–598.

Walker, M., Tobisch, R., Weise, G., 2005: The machine operator current opinions and the future demands on technical ergonomics in forest machines. Institutionen för skogens produkter och marknader. Sveriges lantbruksuniversitet, Uppsala, 72 p.

309

I. Potočnik and A. Poje

Forestry Ergonomics and Occupational Safety in High Ranking Scientific Journals ... (291–310)

Wilmsen, C., Bush, D., Barton-Antonio, D., 2015: Working in the shadows: Safety and health in forestry services in southern Oregon. Journal of Forestry 113(3): 315–324. doi:10.5849/ jof.13-076

Yovi, E.Y., Yamada, Y., 2015: Strategy to disseminate occupational safety and health information to forestry workers: The felling safety game. Journal of Tropical Forest Science 27(2): 213–221.

Authors’ addresses:

Received: January 11, 2017 Accepted: February 2, 2017

310

Prof. Igor Potočnik, PhD. e-mail: Igor.Potocnik@bf.uni-lj.si Assist. prof. Anton Poje, PhD. * e-mail: Anton.Poje@bf.uni-lj.si University of Ljubljana Biotechnical Faculty Večna pot 83 1000 Ljubljana SLOVENIA * Corresponding author Croat. j. for. eng. 38(2017)2

Subject review

LCA Studies in Forestry – Stagnation or Progress?

Andreja Đuka, Dinko Vusić, Dubravko Horvat, Marijan Šušnjar, Zdravko Pandur, Ivica Papa Abstract Today, LCA is one of the leading and most used tools for environmental management, but the application of LCA in forestry is still in an initial phase. Due to a high amount of different wood products which can be produced in forestry sector, production of raw material itself is not included enough in the whole LCA process. Raw wood products and biomass used to be widely declared as »carbon neutral« and renewable, but production steps have a significant influence on the environmental impact depending on machinery used, opening forest with new roads, management type (clear-cut, even-aged management or selective cut), etc. This paper gives a review of LCA studies in forestry based on three segments:  harvesting operations  biomass for energy  road construction and maintenance.

Keywords: life cycle assessment, energy consumption, GHG emission, carbon neutral, wood products.

1. Introduction Forestry is a traditional supplier for various industries in terms of renewable raw materials, household fuel wood and increasingly for biofuels. Mechanized harvesting systems increased productivity, improved conditions for forest workers and decreased the demand for manpower in forest operations (Holtzscher and Lanford 1997), but have also increased fuel and oil requirements (Athanassiadis 2000, Berg and Karjalainen 2003), which contributed to higher GHG (Green House Gases) emissions (Berg 1997, Athanassiadis 2000). The development of environmentally friendly technologies, which are essentially based on utilization of renewable resources, is still happening at a slow pace, which makes them not-so-cheap replacements of the current fossil fuel technologies and processes and delays the achievement of sustainable development (Perić et al. 2016).

Carbon dioxide is a dominant greenhouse gas and its increasing levels, together with other greenhouse gases (i.e. nitrous oxide, methane, chlorofluorocarbons Croat. j. for. eng. 38(2017)2

and tropospheric ozone), may have contributed to the increase in atmospheric temperatures between 0.3 and 0.6 ºC since the late 1800s (Nowak and Crane 2002).

Increased atmospheric CO2 is mostly attributable to fossil fuel combustion (about 80–85%) and deforestation worldwide (Schneider 1989, Hamburg et al. 1997).

Atmospheric carbon is estimated to be increasing by approximately 2600 million tons annually (Sedjo 1989) and its present concentration is the highest in the last 650,000 years (Petit et al. 1999, Siegnethaler et al. 2005).

Trees represent a sink for CO2 by fixing carbon during photosynthesis and storing excess carbon as biomass and net long-term CO2 source/sink dynamics of forests change through time as trees grow, die, and decay. In addition, human influences on forests (e.g. management) can further affect CO2 source/sink dynamics of forests through such factors as fossil fuel emissions and harvesting/utilization of biomass (Nowak and Crane 2002). Forest ecosystems cover about 4.1 billion hectares globally (Dixon and Wisniewski 1995) and through forest vegetation and soils about 1240 Pg of carbon is stocked (Dixon et al. 1994). Out of the total

311

A. Đuka et al.

LCA Studies in Forestry – Stagnation or Progress? (311–326)

Fig. 1 Data categories by Guinée et al. (2002) terrestrial carbon stock in forest biomes, 37% is in low latitude forests, 14% in mid-latitudes and 49% in high latitudes. The above-ground plant carbon stock increases with decreasing latitude from tundra to tropical rainforest (Fisher 1995). Old-growth managed forests stock more carbon as opposed to young fast-growing forests and their conversion to young-fast growing forests will not decrease atmospheric carbon dioxide (Harmon et al. 1990). Increase in carbon stock of forest soils can be achieved through forest management including site preparation, fire management, afforestation, species management/selection, use of fertilizers and soil amendments (Lal 2005).

International Organization for Standardization – ISO (2006) defines LCA as a method used for quantification and improvement of possible impacts associated with products by:  improvement of environmental performance  design or redesign of manufacturing process  selecting and quantifying environmental indicators  establishing environmental soundness for ecolabels for products.

Before defining a system of unit processes, system boundaries should be defined between the product system (Guinée et al. 2006) as a part of the physical environment and the environment (Fig. 1). Authors continue that forestry can be regarded as a part of socio-economic system, but timber extracted from a natural forest will have to be regarded as a critical re-

312

source taken from the environment. Likewise, landfill managed without any control measures should be regarded as part of the environment.

Life Cycle Assessment is one of the leading and most used tools for environmental management (Curran 2016, Finnveden et al. 2009). It provides a systematic, holistic and multidisciplinary approach in quantification of environmental burdens and their potential impacts over the whole life cycle of a product, process or activity. Its scope is the entire life cycle of a product, from the extraction of raw materials, through to manufacturing, use, and end of life. Data from life cycle inventories (LCI) of forest operations provide the forest industry with the input required for assessing its products (Berg and Lindholm 2006). Since LCA came into wider application during the 1990s, efforts have been made to make progress with LCI in relation to forest operations with sufficient relevance and quality (Richter 1995, Schweinle 1999, Heinimann 1999, Knechtle 1999). From a production context point of view, LCA is a suitable tool to assess wood supply systems, because it was designed for product systems (ISO 2006). The idea of LCA was to obtain or provide product information from which the consumer would choose between several alternatives considering differences in environmental effects of the product. This information may be provided by industry, environmental or consumer organizations or by the public sector (Guinée et al. 1992), but also from the scientific community whose objective is to provide environmental soundness.

Croat. j. for. eng. 38(2017)2

LCA Studies in Forestry – Stagnation or Progress? (311–326)

Various databases (European Life Cycle Database – ELCD, U.S. Life Cycle Inventory Database, Ecoinvent Database, Sustainable Product Information Network for the Environment – SPINE, etc.) have been developed to allow communication between different software tools used for practicing LCA such as SimaPro (developed by PRé Consultants), Umberto (developed by IFU Hamburg and IFEU Heidelberg), TEAM (developed by Ecobalance), GaBi (developed

by Department of Life Cycle Engineering of the Chair of Building Physics at the University of Stuttgart and PE International GmbH), POLCAGE (developed by De La Salle University, Philippines, and University of Portsmouth, UK) and GEMIS (developed by Öko-Institut) (Perić et al. 2016).

Land use and forestry aspects of LCA are a complicated issue because of the dynamic nature of forests and long-term production period, which usually corresponds to the rotation period. Modeling carbon, nutrients and energy flows offers a solution that incorporates forestry operations and forest growth in lifecycle inventory without using specific indicators (Wessman et al. 2003). The same authors continue that modeling regarding carbon and nitrogen is usual, while other nutrient flows in the forest are usually ignored. They suggest landscape-related indicators for achieving biodiversity. Environmental system has to be a part of the analysis, characterized by input flows such as CO2, solar energy, mineral resources and land both occupied and transformed. Inventory analysis consists of mapping the structure and functions of the product system, usually in the form of a process flow diagram that is the basis for the following modeling of materials, energy, emission and waste flows (Heinimann 2012). LCA studies in forestry, however, have a wider context than the ones dealing with machine emissions and fuel consumption and report values of CO2 and other GHG emissions relative to energy consumed (Cosola et al. 2016).

As Ecoinvent database (Wernet et al. 2016, Frischknecht et al. 2005) highlights, in most cases the production of materials and services creates a mix of burdens and credits to the environment. When the score is positive, like in most cases, the net effect is the damage to the environment. However, in some cases, the score is negative, indicating that the credits are larger than the burdens. While searching through Ecoinvent multi-product activity datasets that form the basis for all other system models with terms such as: timber, roundwood, oak, spruce, fir or beech, it can be concluded that most of the data regarding production of timber (log production, softwood forestry, debarking at forest road, etc.) are based on literature reviews; Croat. j. for. eng. 38(2017)2

A. Đuka et al.

time aspect (growth of trees) is not included, and very often there are no data related to to/from environment aspect. Timber production is usually referred to as »motor-manual«, without further specifying vehicles used for primary transport and without including primary and secondary forest road infrastructure. Due to high amount of different wood products that can be produced in forestry sector, it seems that the production of raw material itself is not included enough or is even neglected in the whole LCA process. This was also highlighted by Frühwald (1995), Heinimann (2012), Klein et al. (2015), who conclude that inventory analysis is the heart of LCA, taking a considerable amount of time and being extremely data intensive and that it is not properly connected to forestry itself.

Raw wood products and biomass used to be widely declared as »carbon neutral« and renewable, but production steps have a significant influence on the environmental impact (Zah et al. 2007, Miner and Gaudreault 2013, Klein et al. 2015) depending on machinery used, opening forests with new roads, management type (clear-cut, even-aged management or selective cut), etc.

SimaPro 8.2.3 inventory, which includes the following databases: ecoinvent v3, Agri-footprint, US LCI, ELCD, EU and Danish Input Output, Industry data v.2 and Swiss Input Output, contains the term »wood« in many processes related to construction materials (doors, windows), carbon content biogenic materials, paper + board industries, and as a separate entry, »wood« can be found in a vast number of products from wood chips, raw cork, sawnwood, pulpwood, cleft timber, sawlogs, etc. However, roundwood, a starting point for many of these products, can be found in 11 inventory processes, where eight of them refer to azobe, eucalyptus, meranti and parana pine and other three to roundwood itself. Datasets on roundwood consider rough estimation of used machinery in European forestry and the associated occupation impact, but do not include wood burning emissions, land transformation and occupation. Datasets of the above mentioned species are more detailed and include harvesting and extraction operations as well as fuel used for forest road construction, but do not include land use of forest roads and gravel, nor logging impacts on further vegetation and environmental impacts of postharvest processes (potential forest degradation/deforestation) as well as forest road area that is not included in the land use. Majority of these datasets are valid for one specific company and region, so the uncertainty of their further use is rather high and they cannot be assumed to be the standard case. What is even more interesting, not to say ironic, SimaPro 7.1 tutorial

313

A. Đuka et al.

LCA Studies in Forestry – Stagnation or Progress? (311–326)

(Goedkoop et al. 2007) is actually based on example »production of planks from a tree«, where tree felling is process 1, and saw mill procedures is process 2.

Even though the idea of LCA methodology goes back to 1960s and 1970s (Ayers and Kneese 1969, Hall et al.

1979, Odum et al. 1977), it is still not widely used nor accepted in the forestry community and if it is performed, it is often a »truncated LCA« where environmental burdens of machines and forest road infrastructure are neglected, which results in an underestimation of environmental impacts of forest product systems (Heinimann 2012).

assembly of vehicles additionally consumes energy in the amount of 11 MJ/kg for tractors, 9.1 MJ/kg for harvesters, 6.3 MJ/kg for plough, etc. Heller et al. (2003) recorded similar results, where the calculation for agriculture tractor amounted to 26.04 MJ/kg of consumed energy. Engel et al. (2012) provide in their paper an analysis of the raw materials used in the forestry equipment and energy needed for the production of each material. According to their analysis, based on the vehicles mass, Pandur et al. (2015) calculated the total energy consumed in production of materials used for forwarder Valmet 840.2, forwarder Valmet 860.4 and agricultural tractor John Deere 8430, which amounts to 26.79 MJ/kg, 26.79 MJ/kg and 26.56 MJ/kg, respectively. Athanassiadis et al. (2002) estimated the energy used in the production of forwarders to be related to the machine mass, namely 66.4 MJ/kg.

Karjalainen and Asikainen (1996) state that greenhouse gas emissions caused by machinery used in silvicultural and stand preparatory operations, wood harvesting, and timber transportation in Finland were 424.2 Gg carbon dioxide, 10.6 Mg nitrous oxide, 3.5 Gg carbon monoxide, 31.5 Mg methane, 5.6 Gg nitrogen oxide, and 0.7 Gg non-methane volatile organic compounds. Silvicultural and stand preparatory operations accounted for 8% of the total emission, cutting of timber for 13%, primary transport for 18%, secondary (long-distance) transportation for 57% and transportation of machinery for 4%.

Berg (1997) uses LCA techniques in assessing the environmental loads imposed by different types of felling (clear cutting and shelterwood cutting), different level of mechanization (motor-manual felling with chainsaws and mechanized logging with harvesters), timber extraction by forwarders and conveyance of people, machinery and materials to and from the site in northern and southern part of Sweden. Forwarding was not separated from felling. The emissions in shelterwood cutting were 10% higher than in clear cutting and forwarding. The emissions were 20–25% higher in shelterwood management system and it can be expected that in selective forests, energy inputs will be even higher. According to the author, motor-manual felling had lower emissions per cubic meter than mechanized felling and even heavy deployment of resources for transporting personnel to and from work would not be sufficient to balance that difference.

Since, shelterwood and clear cutting were performed in different types of stands and terrain, figures presented here cannot be used for straight comparison of felling systems.

Athanassiadis (2000) estimated a combined fuel and oil energy use for harvesting and forwarding of

It is predictable that future LCA studies will focus on reducing the uncertainties of the current key issues such as: inclusion in the assessment of indirect effects of land-use-changes and their amortization over time, estimation of bioenergy impacts on biodiversity, better determination of fertilizer induced N emissions, and others (Cherubini and Strømman 2011). Authors continue that standardization in GHG balance accounting (also called carbon footprint) of products is particularly perceived as urgent by policy makers, and the methodological standards provided by consultants and stakeholders try to address this need.

This paper gives a review of scientific literature that used life cycle assessment (LCA) methodology or its parts to estimate sustainability and recycling values and environmental impacts of forestry operations, with the focus on three areas of interest: harvesting operations, biomass for energy and forest road construction and maintenance.

2. LCA studies in harvesting operations Combustion engines have been the backbone of forest machinery and the quality of the combustion process is crucial for all subsequent results. Machines consume resources through maintenance, which should also be considered in the analysis process. The materials of which a machine is manufactured embody environmental burdens that have to be considered to fulfill their »cradle to grave« requirement (Heinimann 2012).

Klvač et al. (2003) state that energy used in the manufacture and maintenance of machinery contributes to the overall energy use of the system and must be included in any LCA of machines. During calculation of energy embodied in forest machines and vehicles, Pandur et al. (2015) assumed it to be 66 MJ/kg.

Börjesson (1996) states that the energy required for the production of material embedded in vehicles amounts to an average of 24 MJ/kg, while manufacturing and

314

Croat. j. for. eng. 38(2017)2

LCA Studies in Forestry – Stagnation or Progress? (311–326)

A. Đuka et al.

Table 1 Studies on energy use in forest operations (Lindholm 2006, 2010 Energy use, MJ/m3

Silviculture and logging

Secondary transport

Total

Germany, saw logs, spruce (transport distance 50 km) (Schweinle 1996)

135

92

227

Switzerland, mechanized logging (Knechtle 1997, 1999)

91

–

–

Switzerland, motor-manual logging (Knechtle 1997, 1999)

111

–

–

Germany, (transport distance 50 km) (Wegner 1994)

62

125

187

Norway (hybrid LCA 3 scenarios from best to worst depending on transport distance) (Michelsen et al. 2008)

–

–

Best 48 Average 162 Worst 390

116+155=271

124

395

12+136

223

370

Spain (Gonzáles-Garcia et al. 2009) Sweden (Gonzáles-Garcia et al. 2009)

82 MJ/m3, but this did not include the energy used during the production of oils. The energy consumed during production is reported as ca 4.5 MJ/l for diesel fuel and 15.6 MJ/l for biodiesel (gained from rapeseed).

Klvač et al. (2003) calculated total energy input per unit of wood production (m3) from the fuel and oil consumption and the average mass of machines and replacement materials. The mean energy input was 66.7 MJ/m3 for harvesters and 52.7 MJ/m3 for forwarders, thereby giving a total system energy requirement of ca 120 MJ/m3 (with fuel accounting for approximately 82% of the total energy use) in Ireland.

Pandur et al. (2015) also calculated total energy inputs for chainsaws, forwarders and forest tractor assemblies, which were: 1) chainsaws 17.46 MJ/m3 (felling and processing logs) and 31.92 MJ/m3 (felling and processing of one-meter-long firewood), 2) 65.81 MJ/m3 for forwarders and 3) 59.72 MJ/m3 for forest tractor assemblies. For input parameters, they used fuel and oil consumption and energy embodied in machines and spare parts (tires, chains, sprockets and guide bars of chainsaws).

Berg and Lindholm (2005) differentiate seedling production, silviculture, logging and secondary transport to identify the most significant process in terms of energy inputs and output of timber and emissions.

The authors state that half of the energy used per cubic meter in Swedish forestry is provided for secondary transport from forests to industries. Enhancing payload per distance, removing return unloaded trips, improving forest roads (road width, curvature and better surfacing as well as »soft« driving) would improve the current situation. The type of cutting operation (final felling or thinning) had greater influence on energy input per volume of timber than geographical area of operations. Final felling consumes less energy Croat. j. for. eng. 38(2017)2

(30 MJ/m3) than thinning (48 MJ/m3). The energy consum­ed for forwarding timber to forest roads in final felling was 22–27 MJ/m3 s.u.b. and in thinning 31–34 MJ/m3 s.u.b. (solid under bark). The energy consumed in silviculture operations was 11 MJ and in seedling production 8 MJ. In conclusion, in Sweden during one year all forest operations produced 15 kg/ CO2-equiv./m3, which is a small amount (0.3 Tg C a–1) compared to national emissions from fossil fuels that amounted to 18.9 Tg C a–1.

Lindholm (2006, 2010) states that according to several European forestry studies (Tab. 1), the energy used in silviculture and logging ranges from less than 60 MJ/m3 of timber up to 270 MJ/m3. These findings have been corroborated by the studies of Schweinle and Thoroe (2001), who also considered road building and provide estimates of 170–270 MJ/tonne of dry wood (70–120 MJ/m3). Secondary haulage accounts for 90 to 223 MJ, raising total energy use to a level of 180–395 MJ/m3. However, energy use has been shown to be higher in exceptionally difficult terrain conditions (Wegner 1994), in long-distance haulage of pulpwood (Gonzáles-Garcia et al. 2009; Michelsen et al. 2008) and when silviculture is highly mechanized and the use of chemicals is high (Gonzáles-Garcia et al. 2009).

Table 2 Energy consumption for lorries of different gross weight.

The energy values are based on the lower heating values of diesel fuels (HD=42.8 MJ/kg diesel) Transport service

Diesel energy consumption kg/tkm

Final energy consumption MJ/tkm

Lorry 16t

0.072

3.08

Lorry 28t

0.05

2.14

Lorry 40t

0.036

1.54

315

A. Đuka et al.

Spielmann and Scholz (2005) compare LCA of different transporting vehicles in terms of payload used in Switzerland, depending on fuel (kg/tkm) and total energy consumption (MJ/tkm), which is given in Table 2.

The same authors presented the data on CO2 and NOX emissions during truck transport with 50% and 100% transport utility i.e. full return trip, and concluded that trucks with 100% full return trips produce 25–30% lower emissions of CO2 and NOX. Pandur et al. (2015) state that energy consumption during timber transport by forest truck assemblies (with a mounted crane) – FTA, at the distance of 53 km, is 199.3 MJ/t of fresh wood. The reason of higher values lies in the fact that loading and unloading of timber with crane is not separated from the driving itself.

In the year 1996, the Croatian state company »Hrvatske šume« Ltd. owned 259 FTAs and participated in total long distance timber transport with a share of 85%. Fuel consumption in all operations necessary for the production of 1 m3 of wood was 6.96 l/m3, and fuel consumption in timber truck transport was 2.33 L/m3 or 33.4% of total fuel consumed (Sever and Horvat 1996).

Karjalainen and Asikainen (1996) report that fuel consumption in Finland is 56 l/100 km, while the emission of greenhouse gases (CO2, CH4 and N2O) is 0.03 kg/m3km. According to Svenson (2011) fuel consumption in Sweden is 28 l/100 km, and according to Klvač et al. (2013) in the Czech Republic fuel consumption amounts to 2.19 l/m3 and 67.4 l/100 km.

Klvač et al. (2003) state that in the overall energy audit of mechanized wood harvesting systems in Ireland, fuel consumption was the most significant item (82%), followed by oils (7%) and machine repairs and replacement (11%). Pandur et al. (2015) point out that the total energy consumption in all the operations necessary for the production of 1 m3 of wood in lowland forests is 634 MJ/m3, of which fuel amounts to 86%, which is similar to the findings of Klvač et al. (2003).

Of all operations necessary for the production of 1 m3 of wood, energy consumption in timber truck transport amounts to 31% of the total energy consumption (Pandur et al. 2015).

Athanassiadis (2000) states that, during harvesting operations, the type of fuel and oil used by machinery, depending on their origin i.e. whether they are mineral or bio-produced products, significantly affects the environment. The author concludes that the production of RME (rapeseed methyl ester) generates high amounts of CO2 and NOX emissions as expected from mineral diesel fuel, and vice versa – in combustion, mineral diesel fuel emissions of HC and CO compounds prevail.

316

LCA Studies in Forestry – Stagnation or Progress? (311–326)

3. LCA studies on biomass for energy Lignocellulosic biomass was the first and, for many centuries, the main source of energy. With the development of civilization, a major shift towards the use of technical properties of wood occurred, but the role of wood in energy production has remained significant (Vusić and Đuka 2015). In recent years, increasing environmental concerns have resulted in policy measures, strongly shifting the focus in energy production towards sustainable sources of energy. In this respect, forest industry is expected to play a significant role due to the fact that among all the available alternative energy sources (hydro, solar, wind, etc.), biomass is the only carbon based sustainable option (Khan et al.

2009) and, therefore, it can effectively be transformed into different energy carriers (heat, electricity and fuel for transportation) making it the most desirable option for the replacement of fossil fuels.

Different techniques and approaches have been used to assess the environmental effects and energy balance of biomass production and use for energy.

Earlier research relied mostly on energy analysis, quantifying consumed energy and CO2 or GHG emissions in the production system, while recent studies favor LCA and include a wider range of environmental impacts (Djomo et al. 2011). Klein et al. 2015 state that the first tangible LCAs for the European forestry and wood products sector appeared in the 1990s, with the aim to scientifically analyze the impacts arising from nonrenewable inputs into a system.

LCA biomass studies are usually designed either as stand-alone assessments (describing the production system and presenting environmental impacts) or as comparative LCA studies (opposing the environmental impacts of the bioenergy system to the environmental impacts of alternative energy systems, either other renewable or fossil ones) (Djomo et al. 2011).

Cherubini and Strømman (2011) state that LCA can be carried out using different methods based on the purpose of the study, and make a distinction between attributional and consequential LCA. The first describes the environmentally relevant flows to and from a lifecycle (and its sub-systems), while the latter describes how environmentally relevant flows will change in response to possible decisions (Finnveden et al. 2009).

Although the attributional method is the most used in LCA, in LCA of bioenergy systems the consequential method is broadly applied for comparing the environmental impacts with those of a fossil reference system (Cherubini and Strømman 2011).

Klein et al. (2015) identified a total number of 28 studies where LCAs for forest production were at least Croat. j. for. eng. 38(2017)2

LCA Studies in Forestry – Stagnation or Progress? (311–326)

one of the main study objectives and supported the statement by Heinimann (2012) that, although LCAs have already been discussed in the forestry sector for 20 years, there is still little information based on scientific research. They name two reasons for this situation. One is the fact that, in many cases, forest production is not the main study objective, while the products of forest production frequently are (e.g. fuel chips or pellets), and environmental impacts of the previous forestry processes are only deduced from literature or calculated starting from the latest stage of the forest product chain (e.g. with the collection of wood residues or chipping), and thereby neglecting important processes of forest production. The other is the general opinion that the respective processes have only minor environmental impacts, and that providing wood for material or energetic purposes is nearly carbon-neutral (Miner and Gaudreault 2013).

Klein et al. (2015) distinguish two central questions related to climate change and forestry; the influence of forest management (and land use change) on carbon stocks of forests and harvested wood products, and GHG-emissions caused by forestry processes mainly originated from non-renewable inputs like fossil fuels or construction material for machineries.

As stated by Cherubini and Strømman (2011), bioenergy systems generally ensure GHG emission savings when compared to conventional fossil reference systems; net GHG emissions from generation of a unit of electricity from biomass are usually 5–10% of those from fossil fuel-based electricity generation (Cherubini et al. 2009, Bhat and Prakash 2009). This ratio will be even lower, if biomass is produced with low energy input (or derived from residue streams), converted efficiently, ideally in CHP (Combined Heat and Power) applications, and if the fossil fuel reference use is inefficient and based on a carbon-intensive fuel such as coal (Cherubini and Strømman 2011).

Klein et al. (2015) state that, considering that all removed biomass from sustainably managed forests will be sequestered again in the future (Helin et al.

2013), and based on the overall opinion that the provision of wood as raw material does not cause high GHG (Green House Gases) emissions, wood and wood products are commonly claimed as »carbon neutral«.

They question the »absolute carbon neutrality« of raw wood products, by reporting the results of 28 LCA studies of forestry production (14.3 kg CO2-equiv. per m3 o.b. (over bark) mean GWP (Global Warming Potential) from site preparation to forest road, adding 6.3–67.1 kg CO2-equiv. per m3 o.b. for transport processes and on average 20.5 kg CO2-equiv. per m3 o.b. for chipping processes. They suggest that raw wood products Croat. j. for. eng. 38(2017)2

A. Đuka et al.

should be described as »low emission raw materials«, if long-term in situ carbon losses by changed forest management or negative direct or indirect effects of land use change (LUC – Land Use Change, iLUC – indirect Land Use Change) can be excluded (Klein et al. 2015). In support to their report, the GHG-emissions, even in the worst case of 28 analyzed literature sources, are still low (9%) compared to the respective carbon content of the harvested wood (the range of C-emitted/C-stored in wood is 0.008–0.09 from forest to plant gate or consumer).

Djomo et al. (2011) synthesized 26 studies on energy and GHG balance of bioenergy production from poplar and willow published between 1990 and 2009.

Results reported on energy ratios varied from 13–79 for the cradle-to-farm gate and 3–16 for cradle-to-plant assessments, and the intensity of GHG emissions ranged between 0.6 and 10.6 g CO2-equiv. per MJ (39–132 g CO2-equiv per kWh). Although the substantial variation of reported values (caused by different system boundaries and methodological assumptions in reviewed studies) is evident, the review revealed a general consensus that short rotation coppice (SRC) willow yielded 14.1–85.9 times more energy per unit of fossil energy input compared to coal, and that GHG emissions were 9–161 times lower than those of coal (Djomo et al. 2011).

In their research of SRWC (Short Rotation Woody Crop) willow for energy, Heller et al. (2003) stressed the importance of analyzing the whole rotation period with the focus on redistributing the environmental burdens of establishing the plantation over each cutting cycle. They reported the production of 55 units of biomass energy per unit of fossil energy consumed over the biomass crop life cycle of 23 years. The research concluded that inorganic nitrogen fertilizer inputs have a strong influence on overall system performance, accounting for 37% of the non-renewable fossil energy input into the system and that net energy ratio varies from 58 to below 40 as a function of fertilizer application rate. Heller et al. (2003) also suggested substituting inorganic N fertilizer with sewage sludge biosolids, claiming that this practice could increase the net energy ratio of the willow biomass crop production system by more than 40%. They report net greenhouse gas emissions of 0.68 g CO2 per MJ of biomass produced and point out that, for reasonable biomass transportation distance and energy conversion efficiencies, generating electricity from willow biomass crops could produce 11 units of electricity per unit of consumed fossil energy. The same authors conclude that in biomass truck transport (40 t total weight), energy consumption was 188.9 MJ/t of dry matter on an

317

A. Đuka et al.

LCA Studies in Forestry – Stagnation or Progress? (311–326)

average distance of 96 km, while Pandur et al. (2015) state that energy consumption in hauler truck transport of wood chips with the moisture content of 35% on an average distance of 50 km was 77.35 MJ/t.

Pandur et al. (2015) calculated EROI (energy returned on energy invested) for wood chips from shelterwood cuttings of lowland oak forests. The following parameters were included in the calculation: energy invested for manufacturing all vehicles, machines and tools used in harvesting operations, road building and maintenance, fuel and lubricant consumption, energy invested in manufacturing of components (spare parts) such as: tires, chains, guidebars, drive spockets, etc. and energy invested for production of pesticides used in forestry.

Börjesson (1996) estimates that total energy consump­ tion during biomass transport by truck is 1.4 MJ/tkm, while for adapted farm tractor energy consumption doubles to 2.9 MJ/tkm. Energy required for biomass transport by railroad is 0.7 MJ/tkm, twice less than by truck, while by water transport it is 0.23 MJ/tkm – six times less energy than required by truck transport, which is, by the way, the most common timber transport in Sweden, Austria, Denmark, Finland, Norway, Germany, Slovenia, Italy, Ireland and Croatia (Schwaiger and Zimmer 2001, Beuk et al. 2007). The largest direct energy input i.e. fuel consumption ranges from 72.4% for adapted farm tractor to 97.1% for railway, while the remaining energy is needed for building infrastructure traffic networks and manufacturing and transporting vehicles.

Lindholm et al. (2010) investigated stumps and logging residues as raw material for energy generation, modeled seven different procurement chains of forest energy in Sweden (variations in geographical location, technology employed and resource use), and calculated their environmental performance from a Life Cycle Assessment (LCA) perspective. They reported the energy output/input ratio of chips from residues and stumps in the range of 21–48, and the greenhouse gas emissions from 1.5–3.5 g CO2-equiv. per MJ chips.

Results presented in the study by Lindholm et al.

(2010) confirmed the conclusions of previous research (Näslund-Eriksson and Gustavsson 2008) that transportation of forest fuel dominates the primary energy use, and that the use of primary energy in transporting forest products varies across different parts of Sweden (Berg and Lindholm 2005) due to different transportation distances as a result of different procurement chain organization. The results for the bundle forest energy supply system show that bundling process has the second highest energy use and environmental impact, but due to the fact that the forest energy systems based on

bundles rely on immature technologies, they have the potential to be improved (Lindholm et al. 2010). The primary energy use and environmental impact of the comminution of forest fuel, as the central feature of the forest energy supply chain (Hakkila 2004), strongly depends on the technology used, diesel driven vs. electrical driven (Lindholm et al. 2010), again depending on the design of the procurement chain.

Yoshioka et al. (2005) analyzed the energy balance and the carbon dioxide (CO2) emission of logging residues from Japanese conventional forestry as alternative energy resources over the entire life cycle of the residues using the method of a life cycle inventory (LCI). They calculated the ratio of energy output to input to be 5.69 and concluded that the production system they researched could be feasible as an energy production system. Comparing the CO2 emission per MWhe (1 MWhe=2.6136 MWh) of the biomass-fired power generation plant (61.8 kg CO2/MWhe) with that of coal-fired power generation plants in Japan (960 kg CO2/MWhe), the reduction in the amount of CO2 emission that would result from replacing coal with biomass for power generation could be as much as 3.0 million dry-t/year (Yoshioka et al. 2005).

According to Klein et al. (2015), system boundaries are crucial to identify all relevant processes for a specific LCA. They suggest that the forest system should start with site preparation processes and end at least at the forest road, including all relevant primary and secondary processes of the entire forest product chain (from cradle-to-forest road), and if in some cases, emissions do not appear (for example, if planting processes are not required because natural regeneration occurs), energy balance of this process should be set to zero (Klein et al. 2015). On the other hand, Lindholm et al. (2010), in the study of fuel chip production, set the system boundary starting in the forest after final felling (and including lifting of stumps by harvesters and forwarding stumps and logging residues) and ending when wood chips have been comminuted and delivered to the energy plant. Yoshioka et al. (2005) consider bioenergy as a by-product of conventional forestry, and in this sense set the bioenergy system boundary starting with comminuting logging residues at the landing of the logging site by a mobile chipper accrediting all environmental impacts up to this point to forestry. Similar to Yoshioka et al. (2005), Johnson et al. (2012) in the research of the first thinning by fulltree method, state that the primary products should bear the environmental burdens of the stand management activities because the whole tree is delivered to the landing as part of the primary product harvest.

There is no allocation of cost, fuel, nor any correspond-

318

Croat. j. for. eng. 38(2017)2

LCA Studies in Forestry – Stagnation or Progress? (311–326)

ing environmental burdens required to deliver the tops and limbs to the landing. Those are carried by the primary product.

It is evident that system boundaries are affected by the raw material characteristics and the place where they are produced/located. This is especially important when analyzing wood energy products, because raw material for their production can be regarded either as waste or product depending on the market situation and cost effectiveness of available harvesting systems. The issue of product/by-product/waste definition was identified by Berg (2001) and its strong influence to allocation procedures was discussed.

Allocation in LCA is carried out to attribute shares of the total environmental impact on different products of a system (Cherubini and Strømman 2011). The allocation of environmental burdens is needed if a process causes several outputs or products (Klein et al.

2015). Allocation concept is extremely important for bioenergy systems, which are usually characterized by multiple products and have a large influence on final results (Cherubini and Strømman 2011).

The functional unit is the unit to which all LCA results of a system are referred to and, therefore, its clear definition is essential (Klein et al. 2015). Cherubini and Strømman (2011), in their literature analysis, identify four types of functional units: input unit related (mass or energy unit, where the results are independent of conversion processes and type of endproducts and in studies aimed at comparing the best uses for a given biomass feedstock); output unit related (unit of heat or power produced or km of transportation service is usually selected by studies aiming at comparing the provision of a given service from different feedstocks); unit of land (hectare of land needed to produce the biomass feedstock as the first parameter to take into account when biomass is produced from dedicated energy crops); and year (used in studies characterized by multiple final products, since it allows avoiding an allocation step). Klein et al.

(2015) argue that calculating the impacts only on a hectare or annual base without any product-based unit would not be helpful, due to the fact that the raw wood product is usually the base for different final products, and its inherent ecological impacts represent just a part of all impacts. Therefore, they suggest that, as a default, results should be referred to 1 m3 o.b. as the most common functional unit in forestry. They also state that, in addition to the default functional unit, information about the moisture content and wood density should be given in order to be able to calculate additional functional units like 1 t biomass o.d. (oven dry), 1 t of carbon, 1 MJ (lower heating value), or 1 ha, Croat. j. for. eng. 38(2017)2

A. Đuka et al.

depending on subsequent use of the wood. Moisture content is not important only for calculating conversion efficiency but also for understanding results of the transportation processes. Lindholm et al. (2010) take the calculations one step further accounting for dry matter losses and the ash content of harvested stumps and logging residues as parameters affecting the mass balance of the systems. It can be concluded that the functional unit depends on the goal of the study and on further use of the raw wood and that, as a consequence, different study objectives result in different functional units, which in some cases causes difficulties in making quantitative comparisons (Klein et al. 2015).

Cherubini and Strømman (2011) state that, in the light of the future expected competition for fertile land, one of the important research questions will be that of efficient land use (bioenergy vs. carbon sequestration). They predict that future LCA studies will focus on reducing the uncertainties of these current key issues (inclusion in the assessment of indirect LUC effects and their amortization over time, estimation of bioenergy impacts on biodiversity, better determination of fertilizer induced N emissions, and others).

LCA studies are crucial to understand and quantify environmental impacts and to avoid possible negative effects of increasing wood use as energy source (Klein et al. 2015). The use of different input data, functional units, allocation methods, reference systems and other assumptions complicates comparisons of LCA bioenergy studies (Cherubini and Strømman 2011). Some authors recognized that different accuracy levels and reliability of the input parameters have a strong influence on the final results, and therefore tried to solve this problem by applying sensitivity analyses, modeling different productivity levels (Johnson et al. 2012), energy requirements (Lindholm et al.

2010), or biomass-fired power generation plant parameters (Yoshioka et al. 2005). When analyzing 28 different literature sources of LCA in the forestry sector, Klein et al. (2015) concluded that the results of the GWP varied considerably between studies, depending on the processes included and decisive assumptions (like productivity rates and fuel consumption of machineries), but also stated that, compared with the carbon stored in wood, the GWP actually varies on a low scale.

4. LCA studies in forest road construction and maintenance Forest traffic infrastructure gives access to forests and forest land and, therefore, it is today an essential

319

A. Đuka et al.

LCA Studies in Forestry – Stagnation or Progress? (311–326)

Table 3 Environmental loadings caused by road construction and maintenance (Mroueh et al. 2000) Environmental loadings

Construction

Maintenance

CO2, kg/km

263,000 – 562,000

33,900

SO2, kg/km

280 – 610

4.1

NOX, kg/km

2600 – 3800

140

CO, kg/km

600 – 1100

20

Violatile organic compounds (VOC), kg/km

550 – 980

210

63,000 – 100,000

18,200

790,000 – 1,470,000

183,300

Fuel consumption, l/km Energy consumption, kWh/km

part of intensive forest management (Šikić et al. 1989, Potočnik 1996, Gucinski 2001, Loeffler et al. 2009, Stampfer 2010, Whittaker et al. 2011, Bosner et al. 2012, Pentek and Poršinsky 2012, Sokolović and Bajrić 2013, Papa et al. 2015). Enache and Stampfer (2014) state that significance of forest traffic infrastructure as environmental burden is actually two-sided, because in forests with poor accessibility, the environmental footprint of forest operations is significant due to long timber extraction distances. Improving the environmental performance of forest operations requires a well-developed forest infrastructure, specifically the density and quality of roads. Even though forest traffic infrastructure and long-distance transport have a share of about 60% in the overall environmental burden of timber procurement process, environmental performance of silviculture operations, timber harvesting and transport are extensively addressed in the literature, while forest roads are kept aside from the analyzed system boundaries except a few recent studies (Berg and Karjalainen 2003, Whittaker et al. 2011, Bosner et al. 2012, Heinimann 2012).

Karjalainen and Asikainenen (1996), in their extensive study made in Finland, conclude that the highest GHG emissions in silvicultural and forest improvement work were caused by building of permanent forest roads. Building of one kilometer of permanent forest road requires nearly 47 h of work with an excavator, 4.25 h with a bulldozer, 6.8 h with a loader, and 24 h driving materials with a truck to complete the upper structure of the road, giving a total fuel consumption of 1236.2 l km-1. Respectively, GHG for building one kilometer of permanent forest road in Finland is 3290.74 kg CO2, 0.0826 kg N2O, 27.50 kg CO, 0.2374 kg CH4, 39.648 kg NOx and 5.646 NMVOC. Mroueh et al. (2000), in a study of life cycle assessment of road construction, analyze numerous factors (environmental loadings) and divide them into five

320

categories: 1) resource use, 2) effluents to soil and waters, 3) emissions to air, 4) wastes, and 5) other loadings. According to the study of Häkinnen and Mäkelä (1996), the same authors estimate the environmental burdens that arise during maintenance and repair of roads in Finland in the period of 50 years. The frequency of repairs is determined by a preset strategy (Tab. 3).

Heinimann and Maeda-Inaba (2003) developed a model that evaluates environmental burden of forest road construction based on an input-output model of the underlying process network. This approach enabled the study of the influence of 6 road construction parameters: 1) roadbed width, 2) cut slope, 3) fill slope, 4) thickness of base course, 5) thickness of surface course, and 6) transport distance of base course materials. The entire analysis was based on the following average values of forest road parameters in hillymountainous parts of Switzerland: 1) roadbed width of 4.2 m, 2) cut slope angle of 1:1, 3) fill slope angle of 4:5, 4) thickness of the base course of 0.3 m, 5) thickness of the surface course of 0.08 m, and 6) transport distance for base course materials of 10 kilometers. Authors concluded as follows: On moderate slopes of up to 40%, construction of one meter of forest road consumes about 350 MJ of energy, while emitting about 20 kg of greenhouse gases; Energy consumption is equivalent to the heating value of about 10 l of diesel fuel per meter of road length, and about 10 kg of wood mass that has to be grown to sequestrate the amount of emitted greenhouse gases; Transport distance of base course materials is the most sensitive factor of influence. Compared to on-site preparation of aggregates, a 50-kilometer transport increases energy consumption by a factor of about five; Croat. j. for. eng. 38(2017)2

LCA Studies in Forestry – Stagnation or Progress? (311–326)

A. Đuka et al.

Table 4 Breakdown of energy requirements and GHG emissions for forest road construction and maintenance (Whittaker et al. 2011) Stage

Energy requirement MJ/km

Emissions kg CO2/km

kg CH4/km

kg N2O/km

kg CO2 eq/km

Road construction – diesel fuel Loading roadstone

48,867.96

3,376.73

0.925

0.026

3,407.59

Haulage

99,545.84

6,878.53

1.883

0.053

6,941.38

Spreading roadstone

25,338.94

1,750.90

0.479

0.013

1,766.90

Grading

2,714.89

187.60

0.051

0.001

189.31

Rolling

1,680.64

116.13

0.032

0.001

117.19

Material inputs Roadstone (blasted)

127,509.70

7,605.12

32,075

39,175

20,081.00

Roadstone (crushed)

51,657.34

3,416.85

5,226

5,369

5,147.44

Machine manufacture Excavator

9,003.75

668.96

0.982

0.039

705.21

Haulage

25,725.00

1,911.31

2.806

0.112

2,014.87

Bulldozer

9,261.00

688.07

1.010

0.040

725.35

Grader

1,194.38

88.74

0.130

0.005

93.55

Roller

422.63

31.40

0.046

0.002

33.10

Machine maintenance Excavator

180.08

13.38

0.020

0.001

14.10

Haulage

514.50

38.23

0.056

0.002

40.30

Bulldozer

185.22

13.76

0.020

0.001

14.51

Grader

23.89

1.77

0.003

0.000

1.87

Roller

8.45

0.63

0.001

0.000

0.66

Total

403,834.19

26,788.09

45.745

44.841

41,294.33

Slope is the second important factor that shows a nonlinear influence on energy consumption and greenhouse gas emissions. Increasing slope to about 50% doubles energy consumption and greenhouse gas emissions, while a slope of 70% almost triples them; Roadbed width is the third important factor of influence. Energy consumption doubles when the roadbed width is increased from 4.2 m to 6.2 m.

The above stated results and conclusions were later confirmed by Heinimann (2012), who reported that during construction and maintenance of forest roads, embodied energy rates of 315 MJ m-1 to 735 MJ m-1 depend on the side slopes and CO2 emission rates between 19 and 47 kg m-1 In the mountains of the United States, Loeffler et al. (2009) study the actual excavation of road paths in Croat. j. for. eng. 38(2017)2

»extreme terrain conditions« with regard to energy consumption and CO2 emissions. Similarly to Heinimann and Maeda-Inaba (2003), authors estimate that diesel fuel required for roads constructed on slopes of up to 50%, while using a cut-fill construction method, was 1400 l/km, with emitting 3777.59 kg of CO2/km. On slopes of more than 50%, by using a full bench road construction method, between 7680 and 18,800 l/km diesel fuel was consumed and between 20,974.06 and 51,504.86 kg of CO2/km was emitted. It is evident that fuel consumption and CO2 emissions were 5.5 times greater on slopes of more than 50%.

Whittaker et al. (2011) state that forest road construction is a highly energy-intensive operation, where operations such as grading, rolling and hauling stone requires approximately 4.7 l diesel for 1 m

321

A. Ä?uka et al.

of road and, in total, road construction requires 404 GJ and emits 41 t CO2-equiv. km-1 road. Detailed analysis of energy requirements and GHG emissions are given in Table 4.

It should be mentioned that it is even more difficult to find understanding of road maintenance operations as environmental burden in scientific studies, such data usually being ignored due to lack of databases or due to its overall complexity. This is confirmed by Schwaiger and Zimmer (2001), who collected data from 11 European countries regarding LCA of forestry and forest products, Berg and Karjalainen (2003) who analyzed emissions during harvesting operations in Finland and Sweden, Loeffler et al. (2009), all without including estimates of fuel consumption or emissions for road reconstruction, grading and maintenance. In the following year, Whittaker et al. (2010) emphasize that the actual extent and frequency of forest road maintenance should be further investigated in terms of environmental burdens. Whittaker et al. (2011) state that road maintenance operations are less energy intensive due to the smaller quantities of aggregate used per km, and fewer machinery operations. 102 GJ and 9000 kg CO2-equiv. are required to maintain 1 km of road.

Authors, further divide forest roads into two groups depending on the necessity of road maintenance: Type A roads, which are maintained once a year, Type B roads, which are maintained before each harvesting operation.

Furthermore, authors state that over the full forest rotation period, road maintenance requirements exceed those of the original road construction. In the study area, where road density of type A roads was 0.008 km/ha and type B was 0.007 km/ha, over a 50year forest rotation period with six felling periods, original road construction required 120 MJ ha a-1 of energy and emitted 8.0 kg ha a-1 CO2-equiv., while 1912.2 MJ ha a-1 of energy is required and 129.9 kg ha a-1 CO2-equiv. is emitted during forest road maintenance operations.

In forestry, environmental impact studies usually exclude forest transport infrastructure impact, which is correlated, according to Treloar et al. (2004), to road construction, maintenance and use, due to its high complexity, where a complete LCA of forest roads is difficult and time consuming, and it depends on the system boundaries and on the number of inputs in the process analysis. Treloar et. al (2004) and Sharrard (2007) state that a hybrid based process and inputoutput based LCA approach is recommendable for estimating project specific environmental impacts of forest roads.

322

LCA Studies in Forestry â&#x20AC;&#x201C; Stagnation or Progress? (311â&#x20AC;&#x201C;326)

5. Final remarks LCA is still not adequately applied in forestry because of broad variations of wood products originating from the forestry sector, and because the production of raw material is usually not included or is ignored in LCA process, together with the fact that LCA for forestry usually takes substantial amount of time (productions phase correlates to rotation period) and is highly data intensive (land use). The main study objectives are usually based on forestry by-products (chips, pellets, etc.), and environmental impacts of the previous forestry processes are derived from literature or calculated, but starting from the latest stage of the forest product chain.

It can be concluded that the nature of a raw material (being the starting point of a process) or a product (being the ending point of a process) and allocation of environmental burdens are strongly influenced by the applied harvesting system and harvesting method.

For example, in cut-to length harvesting, pulp wood designated for energy use should be burdened with environmental load from the beginning of the production (silvicultural processes), whereas logging residues used for energy generation should bear environmental load from forwarding onwards. Opposed to that, full-tree systems, employing skidders and processors on the landing or cable yarders with processing heads concentrate the logging residues at the landing site, setting the system boundaries from comminution phase onwards.

Production in forestry can be roughly divided into: 1) roundwood, 2) long-meter firewood, long stackwood and 3) slash. Therefore, the results should be based on 1 m3 o.b. as the most common functional unit in forestry for roundwood or 1 t biomass for longmeter firewood and slash, thus creating system boundaries that were quite vague before. Information about the moisture content and wood density should be given in order to be able to calculate additional functional units. Also, other nutrient flows, besides carbon and nitrogen, should be included in the whole life cycle assessment process. Data on road construction and maintenance should be taken from a higher level (forest administration office or region) on a yearly basis and divided into specific research areas included in the LCA study due to high differences in data of previous research and overall complexity.

Total energy invested in the whole production process of three major forestry products is usually not available or not reliable enough, and fuel consumption increased by 20% can be used as energy inputs since it is an easily measurable parameter not to mention the Croat. j. for. eng. 38(2017)2

LCA Studies in Forestry – Stagnation or Progress? (311–326)

most influential one. This way it is possible to simplify future LCA processes primary based on forestry.

Based on the number of published studies and different approaches used (raw material definition, system boundaries, allocation procedures, functional units), future trends in the LCA research of forestry production and use will need substantial harmonization (and maybe simplification) of rules and procedures to reduce the variability and enable the comparison of research results and provide solid ground for coherent conclusions.

Acknowledgments The study was carried out within the framework of »Optimization of harvesting systems and forest traffic infrastructure on strategic and tactical level of planning« financed by the Croatian Ministry of Agriculture and non-market forest values.

6. References Athanassiadis, D., 2000: Energy consumption and exhaust emissions in mechanized timber harvesting operations in Sweden. Science of the Total Environment 255(1–3): 135–143.

Athanassiadis, D., Lideslav, G., Nordfjell, T., 2002: Energy use and emissions due to the manufacture of the forwarder. Resources, conservation and recycling 34(3): 149–160.

Ayers, R.U., Kneese, A.V., 1969: Production, consumption and externalities. The American Economic Review 59(3): 282–297.

Berg, S., 1997: Some aspects of LCA in the analysis of forestry operations. Journal of Cleaner Production 5(3): 211–217.

Berg, S., 2001: Problems with terms and definitions in inventories for forest sector LCS. In: Energy, Carbon and Other Material Flows in the Life Cycle Assessment of Forestry and Forest Products; Achievements of the Working Group 1 of the COST Action E9. European Forest Institute Discussion Paper 10: 23–26.

Berg, S., Karjalainen, T., 2003: Comparison of greenhouse gas emissions from forest operations in Finland and Sweden.

Forestry 76(3): 271–284.

Berg, S., Lindholm, E.L., 2005: Energy use and environmental impacts of forest operations in Sweden. Journal of Cleaner Production 13(1): 33–42.

Beuk, D., Tomašić, Ž., Horvat, D., 2007: Status and development of forest harvesting mechanisation in Croatian state forestry. Croatian Journal of Forest Engineering 28(1): 63–82.

Bhat, I.K., Prakash, R., 2009: LCA of renewable energy for electricity generation systems – A review. Renewable and Sustainable Energy Reviews 13(5): 1067–1073.

Börjesson, P.I.I., 1996: Energy analysis of biomass production and transportation. Biomass and Bioenergy 11(4): 305–318.

Bosner, A., Poršinsky, T., Stankić, I., 2012: Forestry and life cycle assessment. P. 139–160 in Global perspectives on susCroat. j. for. eng. 38(2017)2

A. Đuka et al. tainable forest management, C.A. Okia (Ed.). InTech, ISBN: 978-953-51-0569-5.

Cherubini, F., Bird, N.D., Cowie, A., Jungmeier, G., Schlamadinger, B., Woess-Gallasch, S., 2009: Energy- and greenhouse gas-based LCA of biofuel and bioenergy systems: key issues, ranges and recommendations. Resources Conservation and Recycling 53(8): 434–447.

Cherubini, F., Strømman, A.H., 2011: Life cycle assessment of bioenergy systems: State of the art and future challenges.

Bioresource Technology 102(2011): 437–451.

Cosola, G., Grigolato, S., Ackerman, P., Monterotti, S., Cavalli, R., 2016: Carbon Footprint of Forest Operations under Different Management Regimes. Croatian Journal of Forest Engineering 37(1): 201–217.

Curran, M.A., 2016: Life Cycle Assessment. John Wiley & Sons, Inc., Kirk-Othmer Encyclopedia of Chemical Technology: 1–28.

Dixon, R.K., Brown, S., Houghton, R.A., Solomon, A.M., Trexler, M.C., Wisniewski, J., 1994: Carbon pools and fluxes of global forest ecosystems. Science 263(5144): 185–190.

Dixon, R.K., Wisniewski, J., 1995: Global forest systems: an uncertain response to atmospheric pollutants and global climate change. Water Air Soil Pollution 85(1): 101–110.

Djomo, S.N., Kasmioui, O.E., Ceulemans, R., 2011: Energy and greenhouse gas balance of bioenergy production from poplar and willow: a review. Gcb Bioenergy 3(3): 181–197.

Enache, A., Stampfer, K., 2014: Environmental performance of forest roads due to construction, maintenance and use – case study analyses in Romanian mountain forests. 37th Council on Forest Engineering Annual Meeting. Moline, Illinois, 10 p.

Engel, A.M., Wegener, J., Lange, M., 2012: Greenhouse gas emissions of two mechanized wood harvesting methods in comparison with the use of draft horses for logging. European Journal of Forest Research 131(4): 1139–1149.

Finnveden, G., Hauschild, M.Z., Ekvall, T., Guinée, J., Heijungs, R., Hellweg, S., Koehler, A., Pennington, D., Suh, S., 2009: Recent developments in Life Cycle Assessment. Journal of Environmental Management 91(1): 1–21.

Fisher, R.F., 1995: Soil organic matter: clue or conundrum?

W.W., McFee, J.M., Kelly, (Eds.), Carbon Forms and Functions in Forest Soils, Soil Science Society American, Madison, WI, USA, 1–12.

Frischknecht, R., Jungbluth, N., Althaus, H.-J., Doka, G., Dones, R., Heck, T., Hellweg, S., Hischier, R., Nemecek, T., Rebitzer, G., Spielmann, M., 2005: The ecoinvent database: Overview and methodological framework. International Journal of Life Cycle Assessment 10(1): 3–9.

Frühwald, A., 1995: LCA – A Challenge for Forestry and Forest Products Industry. Proceedings of the International Workshop organised by the European Forest Institute and the Federal Research Centre for Forestry and Forest Products, 3 – 5 May. Hamburg, Germany, 9–14.

Goedkoop, M., De Schryver, A., Oele, M., 2007: SimaPro 7.1 Tutorial. Report version 3.2, PRé Consultants, Netherlands, 79 p.

323

A. Đuka et al.

LCA Studies in Forestry – Stagnation or Progress? (311–326)

González-García, S., Berg, S., Feijoo, G., Teresa Moreira, M., 2009: Environmental impacts of forest production and supply of pulpwood: Spanish and Swedish case studies. International Journal of Life Cycle Assessment 14(4): 340–353.

Hohle, A.M.E., 2011: Energy consumption by energy wood supply. FORMEC 2011 – Pushing the boundaries with research and innovation in forest engineering, October 9–13, Graz, Austria, poster presentation.

Gucinski, H., 2001: Forest roads: a synthesis of scientific information. DIANE Publishing, Portland, OR, USA, 103 p.

Holtzscher, M., Lanford, B., 1997: Tree diameter effects on cost and productivity of cut-to-length systems. Forest Products Journal 47(3): 25–30.

Guineé, J.B., Gorée, M., Heijungs, R., Huppes, G., Klejin, R., de Koning, A., van Oers, L., Wegener Sleeswijk, A., Suh, S., Udo de Haes, H.A., de Brujin, J.A., van Duin, R., Huijbregts, M.A.J., 2002: Handbook on Life Cycle Assessment: Operational Guide to the ISO Standards. Kluwer Academic Publishers, 692 p.

Guineé, J.B., Huppes, G., Lankreijer, R.M., Udo de Haes, H.A., Wegener Sleeswijk, A., 1992: Environmental life cycle assessment of products. R. Heijungs (final editor), B&G, Netherlands, 97 p.

Guineé, J.B., van Oers, L., de Koning, A., Tamis, W., 2006: Life Cycle Approaches for Conservation Agriculture, Part I: A definition study for data analysis. CML report 171, June 2006, University of Leiden, Netherlands, 1–67.

Hakkila, P., 2004: Developing technology for large-scale production of forest chips – Wood energy technology programme 1999–2003. Technology programme report 6. Final Report. TEKES, Helsinki, Finland, 58 p.

Häkkinen, T., Mäkelä, K., 1996: Environemtal adaption of concrete. Environmental impact of concrete and asphalt pavements. VTT Research Notes 1752, Espoo, Finland, 61 p.

Hall, C., Lavine, M., Sloane, J., 1979: Efficiency of the energy delivery systems: I. An economic and energy analysis. Environmental Management 3(6): 493–504.

Hamburg, S.P., Harris, N., Jaeger, J., Karl, T.R., McFarland, M., Mitchell, J.F.B., Oppenheimer, M., Santer, S., Schneider, S., Trenberth, K.E., Wigley, T.M.L., 1997: Common questions about climate change. United Nation Environment Programme, World Meteorology Organization (1997), Nairobi, Kenya, 24 p.

Harmon, M.E., Ferrell, W.K., Franklin, J.F., 1990: Effects on Carbon Storage of Conversion of Old-Growth Forests to Young Forests. Science 247(4943): 699–702.

Heinimann, H.R., 2012: Life cycle assessment (LCA) in Forestry – State and Perspectives. Croatian Journal of Forest Engineering 33(2): 357–372.

Heinimann, H.R., 1999: Okobilanzierung von forstlichen Produktionssystemen—Beziehungen zu Umweltmanagementsystemen und Ubersicht uber das methodische Konzept.

Schweiz Z Forstwes 150(3): 73–80.

Heinimann, H.R., Maeda-Inaba, S., 2003: Quantification of Environmental Performance Indicators Epis for Forest Roads.

High Tech Forest Operations for Mountainous Terrain, Austro2003, October 5–9, Schlaegl – Austria, 13 p.

Helin, T., Sokka, L., Soimakallio, S., Pingoud, K., Pajula, T., 2013: Approaches for inclusion of forest carbon cycle in life cycle assessment—a review. BCB Bioenergy 5(5): 475–486.

Heller, M.C., Keoleian, G.A., Volk, T.A., 2003: Life cycle assessment of a willow bioenergy cropping system. Biomass and Bioenergy 25(2): 147–165.

324

ISO 2006: ISO 14040 Environmental management – Life cycle assessment – Principles and framework (2nd ed.). Geneva, Switzerland, 20 p.

Johnson, L., Lippke, B., Oneil, E., 2012: Modeling Biomass Collection and Woods Processing Life-Cycle Analysis. Forest Products Journal 62(4): 258–272.

Karjalainen, T., Asikainen, A., 1996: Greenhouse gas emissions from the use of primary energy in forest operations and long-distance transportation of timber in Finland. Forestry 69(3): 215–228.

Khan, A.A., de Jong, W., Jansens, P.J., Spliethoff, H., 2009: Biomass combustion in fluidized bed boilers: Potential problems and remedies. Fuel Processing Technology 90(1): 21–50.

Klein, D., Wolf, C., Schulz, C., Weber-Blaschke, G., 2015: 20 years of life cycle assessment (LCA) in the forestry sector: state of the art and a methodical proposal for the LCA of forest production. The International Journal of Life Cycle Assessment 20(4): 556–575.

Klvač, R., Kolarik, J., Volona, M., Drapela, K., 2013: Fuel Consumption in Timber Haulage, Croatian Journal of Forest Engineering 34(2): 229–240.

Klvač, R., Ward, S., Philip, M.O., Owende, P.M.O., Lyons, J., 2003: Energy Audit of Wood Harvesting Systems. Scandinavian Journal of Forest Research 18(2): 176–183.

Knechtle, N., 1997: Materialprofile von Holzerntesystemen – Analyse ausgewählter Beispiele als Grundlage für ein forsttechnisches Ökoinventar. Graduate thesis, ETH Zürich, Department of Forest and Wood Research, Switzerland, 65 p.

Knechtle, N., 1999: Materialprofile von Holzerntesystemen als Ausgangspunkt für Ökoinventare. Schweizerische Zeitschrift für Forstwesen 150(3):81–87.

Lal, R., 2005: Forest soils and carbon sequestration. Forest ecology and management 220(1): 242–258.

Lindholm, E.L., Berg, S., Hansson, P.A., 2010: Energy efficiency and the environmental impact of harvesting stumps and logging residues. European Journal of Forest Research 129(6): 1223–1235.

Lindholm, E.L., 2006: Energy use in Swedish forestry and its environmental impact. Licentiate thesis, Uppsala, Sweden, 28 p.

Lindholm, E.L., 2010: Energy Use and Environmental Impact of Roundwood and Forest Fuel Production in Sweden. Doctoral Thesis, Swedish University of Agricultural Sciences, Uppsala, Sweden, 1–81.

Loeffler, D., Jones, G., Vonessen, N., Healey, S., Chung, W., 2009: Estimating Diesel Fuel Consumption and Carbon Dioxide Emissions from Forest Road Construction. 2008 Forest Inventory and Analysis Symposium, US Department of AgCroat. j. for. eng. 38(2017)2

LCA Studies in Forestry – Stagnation or Progress? (311–326) riculture, Forest Service, Rocky Mountain Research Station, United States of America, 11 p.

Michelsen, O., Solli, C., Hammer Stromman, A., 2008: Environmental Impact and Added Value in Forestry Operations in Norway. Journal of Industrial Ecology 12(1): 69-81.

Miner, R., Gaudreault, C., 2013: A review of biomass carbon accounting methods and implications. National Council for Air and Stream, improvement. Technical bulletin 1015, 33 p.

Mroueh, U.M., Eskola, P., Laine-Ylijoki, J., Wellman, K., Juvankoski, E.M.M., Ruotoistenmäki, A., 2000: Life Cycle Assessment of Road Construction. Finnra report, 17/2000, Helsinki, Finland, 59 p.

Näslund-Eriksson, L., Gustavsson, L., 2008: Biofuels from stumps and small roundwood – costs and CO2 benefits. Biomass and Bioenergy 32(10): 897–902.

Nowak, D.J., Crane, D.E., 2002: Carbon storage and sequestration by urban trees in the USA. Environmental pollution 116(3): 381–389.

Odum, H.T., Kemp, W., Shell, M., Boynton, W., Lehman, M., 1977: Energy analysis and the coupling of man and estuaries.

Environmental Management 1(4): 297–315.

Pandur, Z., Šušnjar, M., Zorić, M., Nevečerel, H., Horvat, D., 2015: Energy Return on Investment (EROI) of Different Wood Products. Precious Forests – Precious Earth, M., Zlatić, (Ed.), Intech, Rijeka, Croatia, 165–184.

Papa, I., Pentek, T., Lepoglavec, K., Nevečerel, H., Poršinsky, T., Tomašić, Ž., 2015: Metodologija izradbe detaljnog registra primarne šumske prometne infrastrukture kao podloge za planiranje i optimizaciju radova održavanja šumskih cesta. Šumarski list 139(7–8): 311–328.

Pentek, T, Poršinsky, T., 2012: Forest Transportation Systems as a Key Factor in Quality Management of Forest Ecosystems.

Forest Ecosystems – More than Just Trees, J.A., Blanco, L., Yueh-Hsin, (Ed.), InTech, Rijeka, Croatia, 433–464.

Perić, M., Komatina, M., Bugarski, B., Antonijević, D., 2016: Best Practices of Biomass Energy Life Cycle Assessment and Possible Applications in Serbia. Croatian journal of forest engineering 37(2): 375–390.

Petit, J.R., Jouzel, J., Raynaud, D., Barkov, N.I., Barnola, J.M., Basile, I., Bender, M., Chappellaz, J., Davisk, M., Delaygue, G., Delmotte, M., Kotlyakov, V.M., Legrand, M., Lipenkov, V.Y., Lorius, C., PÉpin, L., Ritz, C., Saltzman, E., Stievenard, M., 1999: Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399(6735): 429–436.

Potočnik, I., 1996: Mnogonamenska raba gozdnih cest kot kriterij za njihovo kategorizacijo. Disertacija, Biotechnical faculty University of Ljubljana, 241 p.

Richter, K.,1995: Life cycle assessment of wood products. In: Life cycle analysis – a challenge for Forestry and Forest Industry. Proceedings of the International Workshop, European Forest Institute and the Federal Research Centre for Forestry and Forest Products, 3–5 May, Hamburg, Germany, 69–79.

Schneider, S.H., 1989: The changing climate. Scientific American 261(3): 70–79.

Croat. j. for. eng. 38(2017)2

A. Đuka et al.

Schwaiger, H., Zimmer, B., 2001: A comparison of Fuel Consumption and Greenhouse Gas Emissions from Forest Operations in Europe. Energy, Carbon and Other Material Flows in the Life Cycle Assessment of Forestry and Forest Products.

Achievements of the working group 1 of the COST action E9, European Forest Institute, Discussion paper 10, 952-9844-921, Joensuu, Finland, 33–53.

Schweinle, J., Thoroe, C. 2001: Vergleichende Ökobilanzierung der Rundholzproduktion in verschiedenen Forstbetrieben. Mitteilungen der Bundesforschungsanstalt für Forstund Holzwirtschaft. Nr. 204, Kommissionsverlag Buchhandlung Max Wiedebusch, Hamburg, Germany.

Schweinle, J., 1999: Analyse und Bewertung der forstlichen Produktion als Grundlage für weiterfuhrende forst- und holzwirtschaftliche Produktlinien – Analysen. Mitteilungen der Bundesforschungsanstalt für Forst- und Holzwirtschaft, No. 184., Kommissionsverlag Buchandlung Max Wiedebusch, Hamburg, Germany.

Sedjo, R.A., 1989: Forests to offset the greenhouse effect. Journal of Forestry 87(7): 12–15.

Sever, S., Horvat, D., 1996: Razvoj i organizacija hrvatskoga energetskog sektora, Knjiga 6 Gospodarenje šumama u Hrvatskoj – Proizvodnja i potrošnja energenata i energije. Coauthors: Anić, I., Fabijanić, G., Figurić, S., Hodić, I., Krpan, A.P.B., Matić, S., Meštrović, Š., Oršanić, M., Polaček, M., Poršinsky, T., Puljak, S., Risović, S., Tomljanović, S., 1996: Energy Institute Hrvoje Požar. Zagreb, Croatia, 1–76.

Sharrard, A.L., 2007: Greening Construction Processes Using an Input-Output-Based Hybrid Life Cycle Assessment Model. PhD thesis, Carnegie Mellon University Pittsburgh, Pennsylvania, 343 p.

Siegenthaler, U., Stocker, T.F., Monnin, E., Luthi, D., Schwander, J., Stauffer, B., Raynaud, D., Barnola, J-M., Fische, H., Masson-Delmotte, V., Jouzel, J., 2005: Stable carbon cycleclimate relationship during the Late Pleistocene. Science 310(5752): 1313–1317.

Sokolović, Dž., Bajrić, M., 2013: Otvaranje šuma. Forestry Faculty University of Sarajevo, 248 p.

Spielmann, M., Scholz, R.W., 2005: Life Cycle Inventories of Transport Services. The International Journal of Life Cycle Assessment 10(1): 85–94.

Stampfer, K., 2010: Forest Engineering – Course Script. Institute of Forest Engineering, Department of Forest and Soil Sciences, University of Natural Resources and Life Sciences – BOKU, Vienna, Austria.

Svenson, G., 2011: The impact of road characteristics on fuel consumption for timber trucks. In P., Ackerman, H., Ham, E., Gleasure, (Ed.) Proceedings of 4th Forest Engineering Conference: Innovation in Forest Engineering – Adapting to Structural Change. Stellenbosch University, 172 p. Šikić, D., Babić, B., Topolnik, D., Knežević, I., Božičević, D., Švabe, Ž., Piria, I., Sever, S., 1989: Tehnički uvjeti za gospodarske ceste. Scientific Traffic Board of Yugoslav Academy of Science and Art, Zagreb, Croatia, 78 p.

Treloar, G.J., Love, P.E.D., Crawford, R.H., 2004: Hybrid lifecycle inventory for road construction and use. Journal of construction engineering and management 130(1): 43–49.

325

A. Đuka et al.

LCA Studies in Forestry – Stagnation or Progress? (311–326)

Vusić, D., Đuka, A., 2015: Repositioning of Forest Biomass as a Renewable Source of Energy – Implications to Harvesting Operations in Croatia. Proceedings of the Conference CROJFE 2015: Forest Engineering – Current Situation and Future Challenges. 18th–20th March 2015, Zagreb – Zalesina, Croatia.

subprojects 2–5 of Land and Geography project of Nordic Industrial Fund, 25 p.

Wegener, G., 1994: Der positive Beitrag der Holznutzung zur CO2 Bilanz. In: Scientific conference »Holz im Rahmen der Internationalen Grünen Woche«, F.W. Bröker (Ed.). Mitteilungen der Bundesforschungsanstalt für Forst- und Holzwirtschaft, Kommissionsverlag Buchhandlung Nr. 176, Max Wiedebusch, Hamburg, Germany, 187–197.

Whittaker, C.L., Mortimer, N.D., Murphy, R., Matthews, R.W., 2011: Energy and greenhouse gas balance of the use of forest residues for bioenergy production in the UK. Biomass and bioenergy 35(11): 4581–4594.

Wernet, G., Bauer, C., Steubing, B., Reinhard, J., MorenoRuiz, E., Weidema, B., 2016. The ecoinvent database version 3 (part I): overview and methodology. The International Journal of Life Cycle Assessment, [online] 21(9), 1218–1230. Available at: <http://link.springer.com/10.1007/s11367-016-1087-8> [Accessed 15.11.2016.].

Wessman, H., Alvarado, F., Backlund, B., Berg, S., Hohenthal, C., Metsäteho, S.K. Lindholm, E.L., 2003: Land use in eco balance and LCA of Forest products. Report summarize of

Whittaker, C.L., Mortimer, N.D. Matthews, R.W., 2010: Understanding the Carbon Footprint of Timber Transport in the United Kingdom. North Energy Associates Limited, Northumberland, United Kingdom, 66 p.

Yoshioka, T., Aruga, K., Nitami, T., Kobayashi, H., Sakai, H., 2005: Energy and carbon dioxide (CO2) balance of logging residues as alternative energy resources: system analysis based on the method of a life cycle inventory (LCI) analysis.

Journal of Forest Research 10(2): 125–134.

Zah, R., Boni, H., Gauch, M., Hischier, R., Lehmann, M., Wager, P. 2007: Life cycle assessment of energy products: environmental assessment of biofuels. Final Report, EMPA— technology and society Lab, Auftrag des Bundesamtes für Energie, des Bundesamtes für Umwelt und des Bundesamtes für Landwirtschaft, Bern, Switzerland, 20 p.

Author’s addresses:

Received: March 22, 2017.

Accepted: May 9, 2017.

326

Andreja Đuka, PhD. e-mail: aduka@sumfak.hr Dinko Vusić, PhD. e-mail: dvusic@sumfak.hr Prof. Dubravko Horvat, PhD. e-mail: dhorvat@sumfak.hr Assoc. prof. Marijan Šušnjar, PhD. e-mail: msusnjar@sumfak.hr Zdravko Pandur, PhD. * e-mail: zpandur@sumfak.hr Ivica Papa, PhD. e-mail: ipapa@sumfak.hr Department of Forest Engineering Faculty of Forestry University of Zagreb Svetošimunska 25 Zagreb CROATIA * Corresponding author Croat. j. for. eng. 38(2017)2

Subject review

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations Bruce Talbot, Marek Pierzchała, Rasmus Astrup Abstract This paper provides an overview of recent developments in remote and proximal sensing technologies and their basic applicability to various aspects of forest operations. It categorises these applications according to the technologies used and considers their deployment platform in terms of their being space-, airborne or terrestrial. For each combination of technology and application, a brief review of the state-of-the-art has been described from the literature, ranging from the measurement of forests and single trees, the derivation of landscape scale terrain models down to micro-topographic soil disturbance modelling, through infrastructure planning, construction and maintenance, to forest accessibility with ground and cable based harvesting systems. The review then goes on to discuss how these technologies and applications contribute to reducing impacts on forest soils, cultural heritage sites and other areas of special value or interest, after which sensors and methods necessary in autonomous navigation and the use of computer vision on forest machines are discussed. The review concludes that despite the many promising or demonstrated applications of remotely or proximately sensed data in forest operations, almost all are still experimental and have a range of issues that need to be addressed or improved upon before widespread operationalization can take place.

Keywords: sensors, automation, operational efficiency, forest operations, precision forestry

1. Introduction Technology is revolutionising our access to information about forest resources, landscapes, and individual forest machine performance (Ziesak et al. 2014).

The improved information includes both higher spatial and temporal resolution of data and information, as well as access to previously unattainable information (Holopainen et al. 2014). In an economic sense, the forest sector is obliged to support developments that make management processes and operations more efficient. Forest operations management, therefore, needs to grasp these newly available technologies and knowledge in ensuring continual improvement.

Remote sensing is defined as the acquisition of information about an object without making physical contact with it, but there is an underlying understanding of ranges or technologies implied. During forest Croat. j. for. eng. 38(2017)2

operations, nearby objects such as trees, stems, rocks, streams, and gullies also need to be measured from machine or human borne sensors, the so called proximal sensing (Mulla 2013). Proximal sensing is in the early stages of a potentially revolutionary change as cheap and robust sensors and technologies are increasingly applied in the collection, storage, and interpretation of data. Such data can be analysed and applied instantaneously or fed into Big Data systems that evaluate status and trends at local, regional or national levels (Lokers et al. 2016). For example, technologies inherent in smart phones and tablets today include distance ranging, orientation through inertial measurement units (IMUs) including magnetometers, gyroscopes and accelerometers, as well as Global Navigation Satellite Systems (GNSS’s) and cameras (Tomaštík et al. 2016). In forestry, smartphone based sensors and apps have been demonstrated in a variety

327

B. Talbot et al.

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

of uses ranging from the measurement of forest stands or log piles (Vastaranta et al. 2015) to distinguishing between work elements in cable yarding operations (Pierzchała 2017). Over the past two decades, airborne Light Detection and Ranging (LiDAR), commonly called airborne laser scanning (ALS), has become the standard practice for forest inventory in the Nordic countries (Næsset 2004). Spaceborne Radio Detection and Ranging (RADAR), ALS and airborne photogrammetry are now widely applied for estimating forest biomass, and a number of models exist for operational forest inventory (Rahlf et al. 2014, Gobakken et al.

2015). Developments in technology and the resulting improvements in forest inventories, in combination with better terrain information, have the potential to enable precision forestry (Holopainen et al. 2014), as well as improve the control and automation of forest harvesting systems (Ziesak et al. 2014).

Mechanised systems account for a large and increasing share of timber harvesting, where they simultaneously provide stable platforms for the deployment of sensors with regard to power supply, protective housing, temperature regulation, lighting, as well as data storage, viewing and transmission (Talbot and Astrup 2014). In this way, forest machines can potentially serve as data collection platforms to help reduce field survey costs (Olivera and Visser 2014). Adding additional sensors to forestry machinery offers a multitude of potentially beneficial future applications (Ziesak et al. 2014). When it comes to applications in forest operations, the field places special demands on system ruggedness, compatibility, simplicity and robustness in terms of measurement accuracy and reliability. However, many sensors and technologies in the early stages of development are already being effectively applied in more rudimentary settings (Gallo et al. 2013, Visser et al. 2014).

This paper provides a brief overview of how different remote and proximal sensing technologies are being employed with respect to forest operations and how these are relevant for improving operational or environmental efficiency. The overview includes the most applicable remote sensing technologies for forest operations and their basic functionality, while a more categorical specification of these technologies might be found in e.g. Fardusi et al. (2017) or Holopainen et al.

(2014). The existing literature is reviewed and discussed in terms of relevant research for terrain assessment applications, infrastructure planning and monitoring, and finally, ground and cable-based harvesting, including the avoidance or measurement of biological and environmental impacts. The paper concludes with a brief summary and outlook for the future.

328

1.1 Sensor deployment and its relevance for forest operations The sensor platform refers to the sensor carrier, which could be a satellite, aeroplane, unmanned aerial vehicle UAV or a ground based vehicle or human.

For example, Liang et al. (2015) distinguish between platforms for laser scanners as being Airborne Laser Scanning (ALS), Terrestrial Laser Scanning (TLS), Mobile Laser Scanners (MLS) and Personal Laser Scanning (PLS), while Bauwens et al. (2016) add the concept of Hand-Held Mobile Laser Scanning (HMLS).

One of the main considerations in sensing the forest environment is the influence of the sensor deployment on the information gained. Each platform used offers a range of benefits and disadvantages, including the area of coverage per deployment, and the spatial and temporal resolution (Table 1).

Table 1 General characterisation of sensor deployment platform to spatial coverage and temporal resolution (adapted from Pierzchała (2017)) Sensor deployment platform

Coverage

Spatial resolution

Temporal resolution

Global/National

Low

Medium to high

Regional

Medium

Low

Local

High

High

Site

Ultra high

High

Furthermore, there are two main areas within which remote sensing technologies can be discussed:  those relating to the operating environment  those relating to forest operations themselves.

The operating environment determines the selection and use of machine systems, while the second area deals with issues influencing e.g. productivity or data capture during the actual operations.

Remote sensing technologies such as ALS, satellite and aerial photography, and satellite based radar enCroat. j. for. eng. 38(2017)2

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

able large contiguous forest areas to be mapped in a uniform way (Rahlf et al. 2014), forming a basis for the development of efficient planning systems. The utility of using ALS forest and terrain data in harvest planning is discussed by Akay et al. (2009) and Heinimann and Breschan (2012), both of whom emphasize the benefits of the high resolution digital elevation models (DEMs) that have become available. These high resolution elevation models, with one or more point references per m2, have revolutionised the basis for evaluating harvest system accessibility and performance analysis, enabling the use of high precision methods.

The proximal measurement of the forest operations environment (trees and terrain) commonly utilises LiDAR and/or photogrammetry, but the platforms used in deploying them differ and the data resolution is generally considerably higher due to the close proximity. Ground-based measurement provides vertical information on the stem that is not possible to obtain from the air. Examples of the use of terrestrial laser scanning (TLS) in doing pre-harvest tree and stand level assortment bucking have been demonstrated by e.g. Ducey et al. (2013) and Kankare et al. (2014). The stem proportions derived from TLS have been shown to correspond well with stem measurements obtained from the harvesting head (Astrup et al. 2014). Better information on stand-level assortments is useful in estimating the stumpage value of a stand and can be sourced in matching orders with harvest schedules in precision wood supply (Bergdahl et al. 2003).

Terrestrial platforms include the deployment of stationary sensors, sensors on manned or unmanned ground-based vehicles (UGVs), or on humans (Lauterbach et al. 2015, Bauwens et al. 2016, Rönnholm et al. 2016). Terrestrial deployment platforms often utilise the same sensors as aerial applications, but differ in terms of costs, payloads, energy sources, and resolution. The forest canopy poses a considerable challenge for terrestrial forest mapping (Blum et al. 2016).

Ground-level surveying in forests with the use of GNSS is limited due to signal occlusion caused by dense crowns (Wing and Eklund 2007). This occlusion results in multipath error and discrete »jumps« in position estimates, making high accuracy positioning challenging, even with a differential global positioning system (DGPS) (Naesset and Jonmeister 2002, Sawaguchi et al. 2003).

Imaging sensors can be deployed on ground platforms either on vehicles intended for data capturing (mobile mapping) or on forestry equipment itself. An example is the sScale system from Dralle AS (Dralle and Tarp-Johansen 2010), which is an imaging system that can be mounted on a vehicle to measure timber Croat. j. for. eng. 38(2017)2

B. Talbot et al.

piles, or the use of ATVs or UGVs for stand and treelevel measurement and inventory (Öhman et al. 2008, Miettinen et al. 2010, Liang et al. 2014). The harvester head itself is used as a sensor platform to measure tree sizes, and Kauhanen (2008) improved this functionality with image based data.

When properly calibrated, harvesting heads accurately measure diameter at 10 cm intervals along the entire stem length, and, with hundreds of millions of trees being harvested with CTL technology annually, harvesting heads represent a central data collection hub. Immediate uses of such data include local estimations of growth, and yield data at sub-stand level and the development of spatially explicit stem taper equations (Olivera and Visser 2016). Such data are automatically geo-referenced at the resolution achieved by the harvester GNSS. An area of great potential that remains to be fully solved is finding methods for matching single-tree data from ALS with that of the harvester head, as discussed by Lindroos et al. (2015) and Hauglin et al. (2017). Currently, certain harvester brands provide an estimate of the harvester head position relative to the base machine, calculated from hydraulic cylinder extension measurements and crane geometry. Improvements in absolute single-tree precision are, therefore, fully dependent on the accuracy of the GNSS data on the base machine. While this could be resolved with DGPS systems, the practical interim solution is likely to lie in the statistical segmentation of individual trees out of small groups identified in the immediate vicinity (Holmgren et al. 2012).

1.2 Infrastructure planning, construction and monitoring Airborne LiDAR provides high-resolution ground terrain models that represent a considerable improvement on which to base estimations on something so detailed and costly as road planning, construction and maintenance. To this end, Aruga et al. (2005a) developed a forest road design programme based on a LiDAR digital elevation model (DEM) that could optimize the horizontal and vertical alignment of a road segment through the minimisation of construction and maintenance costs, using a tabu-search heuristic. Expanding on that, both Akay and Sessions (2005) and Aruga et al. (2005b) show how the addition of a model for predicting surface run-off from roads, which has important connotations both for environmental impact and for road maintenance, provide additional depth to the potential areas of application of the method. Contreras et al. (2012) demonstrate a model using a high resolution LiDAR derived DEM (1 m) to calculate the required earthwork on a number of hypothet-

329

B. Talbot et al.

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

ical roads. The detailed DEM made it possible to use a higher number of cross sections in the evaluation than otherwise, resulting in differences in estimates of 2–21%, as compared with conventional planning.

Several applications of airborne LiDAR data in the monitoring and evaluation of existing forest roads have been demonstrated. Craven and Wing (2014) considered the influence of 4 different canopy conditions on the accuracy of estimation of road geometry based on LiDAR data, and showed mean vertical error of 0.28 m and horizontal error of 1.21 m, when considered against existing road centrelines. Road slopes were estimated to within 1% and error in horizontal curve radii was estimated with an absolute error of 3.17 m. The follow up work by Beck et al. (2015) used varying intensity values and return densities in classifying roads and demonstrated a high level of accuracy in doing so. In other applications, LiDAR has been used in detecting, monitoring or extracting the geometry of existing roads to evaluate whether they meet certain specifications. For instance White et al.

(2010) extracted alignment and gradient data from a mountain forest road, showing deviations of 1.5 m in position, 0.5% in slope and 0.2% in terms of length when compared with field survey data. A similar approach applied by Azizi et al. (2014) resulted in more than 95% of the road length being classified within 1.3 m of the field surveyed normal. These developments represent considerable time and effort savings in providing detailed road geometry, providing essential complementary data to conventional field surveys.

However, Krogstad and Scheiss (2004) list pitfalls of a blind adoption of these models including inconsistent data returns depending on canopy density and a resultant data smoothing that can provide a false basis for road design, as well as subsurface issues not reflected in the topography.

Beyond planning, construction and the retrieval of road geometry data, monitoring forest road conditions includes gathering information on their surface condition, the condition of the drainage system, the existence of vegetation, and seasonal damage. Existing roads represent partially open areas, which typically results in higher resolution LiDAR ground returns than under the forest canopy, which is the most common case for road planning. In their work on road quality control, Kiss et al. (2015) show the effect of resolutions ranging from 0.1 m to 2.0 m on the ability to correctly assess various parameters. Even at the lowest resolution, road surface was correctly classified in 66% of the cases, while ditches were correctly classified in 60% of the cases. Gaining an overview of the existence and condition of proper drainage is obvi-

330

ously of prime importance, although this is a dynamic factor and difficult to capture at the low temporal scale offered by airborne LiDAR. There are also examples of higher resolution proximal road surface and road geometry modelling. Svenson and Fjeld (2016) applied a profilograph, a vehicle based system with LiDAR scanners, IMU and GPS, in extracting surface roughness and road geometry from a 320 km long stretch of mixed road classes. The derived information was used to predict fuel consumption and derive preferred routes during timber hauling (Svenson and Fjeld 2016). At a slightly lower resolution, Hrůza et al. (2016) demonstrated the use of a UAV and photogrammetry in assessing the condition of the wearing course of a forest road. Experiences gained in that study led the authors to recommend the use of mobile terrestrial systems as preferential for this type of work.

1.3 Machine access planning and layout Machine access planning is largely about supporting decisions on which harvesting system to deploy and how best to go about doing that. Procedures for ground based harvesting and their potential for exploiting remotely sensed data are somewhat different than for cable harvesting, but both work toward maximising efficiency and minimising external impacts.

1.3.1 Ground based harvesting Ground based harvesting is typically carried out with a cut-to-length (CTL) system (harvester and forwarder) or tree-length system (feller-buncher/skidder). Optimal planning of how the skid trails should be laid out is determined to a large degree by topography and soil bearing capacity. Examples of the use of LiDAR derived elevation models in doing this include Søvde et al. (2013), who used heuristics in finding extraction trails for a forwarder while restricting the degrees of pitch and roll through a cost penalty, and Strandgard et al. (2014), who assess the influence of slope on the productivity of a self-levelling processor. Sterenczak and Moskalik (2015) optimise a forest skid trail network through a novel combination of tree segmentation and terrain analysis, where the trees identified in the ALS dataset were used in estimating loads, while the gaps were used as potential nodes in the trail network. The model presented by Contreras et al. (2016) extends on these concepts, and includes the evaluation of a soil recovery cost in determining trail layout. However, despite the high resolution of LiDAR based terrain models as compared with their predecessors, and the detailed micro-slope maps they can produce, ALS data is not sufficient to provide esCroat. j. for. eng. 38(2017)2

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

timates of surface unevenness. Surface unevenness remains one of the most critical factors determining accessibility and productivity of ground based harvesting systems and is currently more easily measurable in a post-harvest context.

Proximal scanning also shows strong potential in providing decision support during operations, for example in rapid detection of stand density and tree positions, assisting with thinning tree selection (Brunner and Gizachew 2014), allowing for data to be collected on individual tree selection by harvester operators (Brunner and Fredriksson 2012), or modelling which tree the operator might select beforehand (Fredriksson 2010). An overview of how remote sensing data can be used in improving the productivity of mechanised harvesting systems is provided by Alam et al. (2012).

1.3.2 Cable based harvesting Planning of cable yarding corridor layout must maximise the utilisation of each machine setup while considering the suitability of load paths. Before the advent of LiDAR derived terrain models, desktop planning risked missing critical terrain points as it was not possible to discern the actual terrain form between contour lines, making it necessary to perform manual profile surveys in order to confirm the degree of deflection attainable in each span. Also, the surveyor needed to make an »a priori« listing of profiles to measure, as only a smaller sample of the site could be covered practically. Detailed LiDAR derived terrain models (1 pt.m-2) now allow for complete analysis of harvesting sites to be made. Examples of such use have been demonstrated by Søvde et al. (2015), who search for the optimal location of landings, and Dupire et al.

(2015), who use LiDAR DTMs to predict the load path in a given corridor.

However, terrain alone does not determine the optimal layout of the cable corridor, as the location of suitable end trees (tail spars) and intermediate support trees also need to be verified. The pre-selection of these from LiDAR data has been shown to be both possible and effective (Scheiss 2005). Furthermore, Heinimann and Breschan (2012) describe how LiDAR can be used in gaining volume estimates for each planned cable corridor, a process which could ultimately feed back into the cable layout algorithms presented by (Dupire et al. 2015 and Søvde et al. 2015).

For both ground-based and cable harvesting, the identification of suitable landings is an important part of harvest planning. Complex spatial patterns can be determined from LiDAR data (Risbøl et al. 2014), and one related task is the detection and assessing of potential landings in terms of area, shape, and surface Croat. j. for. eng. 38(2017)2

B. Talbot et al.

evenness (Søvde 2015). This makes it possible to estimate their proximity to forest roads, and their suitability in terms of wetness and other potentially detrimental factors.

1.4 Avoiding or measuring soil disturbance Arguably one of the most significant applications of LiDAR derived terrain models has been in facilitating the mapping of areas of anticipated high moisture and, therewith, potentially high susceptibility to soil damage by vehicles. The topographic wetness index (TWI) essentially quantifies the influence of topography on hydrological processes on the basis of slope and upstream contributing area, and can be best visualised as representing flow accumulation. Cartographic Depth-To-Water (DTW) algorithms on the other hand basically indicate the anticipated vertical distance between ground water or open water surfaces at any given point in the surrounding terrain. Both have shown to be robust in delineating soil, vegetation and drainage type (Murphy et al. 2011) and are increasingly used in applications of high relevance to forest operations, such as assessing accessibility and the risk of causing rutting and compaction (Murphy et al.

2008). Ågren et al. (2014) found that both provided useful soil wetness predictors but that TWI delineations are sensitive to scale and landscape variations, while DTW produces a resolution-consistent wet-area delineation. Campbell et al. (2013) evaluated the use of DTW in predicting rut depth on a high resolution DEM and found good consistency although this has not yet been effectively demonstrated in forestry. Challenges remain in determining the scale of analysis, satisfactorily including effects of soil texture and geology and handling seasonal conditions (Ågren et al.

2015) or even daily variations in machine-specific forest soil trafficability (Vega-Nieva et al. 2009). For example, Niemi et al. (2017) achieved soil damage prediction accuracies of over 85% when including an existing soil map in their wetness index calculations.

These indices constitute a considerable improvement to forest management data, especially when combined with mathematical programming based decision support systems such as BeST Way in showing the optimal layout of main access trails, as shown by Westlund et al. (2015). In a further step, Pohjankukka et al. (2016) demonstrate the use of machine learning in avoiding soft areas, as bearing capacity known at given control points is used in training a model in estimating bearing capacity in other parts of the stand. This study represents the early phases of what is likely to become a rapidly growing application of the autonomous utilisation of remote and proximally sourced data in for-

331

B. Talbot et al.

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

est machines. The notion of machines sharing such information for a given site has already been put forward by Ala-Ilomäki et al. (2012).

The measurement of wheel rut depth after forwarding has been shown to be feasible with photogrammetry (Haas et al. 2016, Pierzchała et al. 2016), however, there remain a number of challenges to using the method effectively. If not measured iteratively, the original soil surface needs to be estimated and interpolated from the adjacent margin, which may not always be accurate. Also, photogrammetry generates a surface model and not a terrain model, which can result in problems in distinguishing between e.g. a brash mat, surface water, and the real soil surface.

1.5 Improving information on key cultural and biological features in avoiding damage A central part of planning and executing forest operations lies in avoiding change or damage to cultural remnants, special habitats, or the transgression of property borders. Remote sensing and especially airborne LiDAR has the potential for providing better geographic information on the key features of importance in forest operations planning and execution. LiDAR has been used in the detection of cultural heritage sites (Risbøl et al. 2014). The use of LiDAR has also shown to have some success in habitat characterization (Vierling et al. 2008, Sverdrup-Thygeson et al.

2016), where, with improvements in predictability, the segmentation and the delineation of boundaries indicating areas to avoid or treat differently, may yet become a mainstream part of harvest planning. By providing such polygons on high resolution DEMs, methods can be developed to calculate the operations cost taking regard of special biotopes (Søvde et al.

2014), in providing forest managers and society at large with a quantitative tool on which decisions can be based.

1.6 Autonomous machines, machine navigation and vision The use of autonomous or remotely operated machines has gained a solid foothold in applications from agriculture to open-cast mining (Mousazadeh 2013).

Forestry brings a special set of challenges, most notably a complex operating environment with poor GNSS coverage, and operation in an environment that is open to the public, and therefore subject to demanding safety requirements. Nevertheless, there are good reasons for pursuing the development of autonomous machines, not least the social (isolated work environment) and economic (one operator can control multiple machines) benefits offered (Hellström et al. 2009).

332

Given the limited GNSS coverage available under tree canopies, other localization approaches such as Simultaneous Localization and Mapping (SLAM), which attempts to locate the machine with reference to its surroundings, while simultaneously mapping its surroundings, offer some potential for the future. These concepts have been demonstrated on forest machines (Miettinen et al. 2007, Öhman et al. 2008, Tang et al.

2015). Onboard sensors, such as 2D LiDAR scanners, radars and stereo-cameras, are essential in providing navigational support for autonomous machines. In a step toward fully autonomous forwarding, Ringdahl et al. (2011) were able to demonstrate accurate path tracking in repeating a route already traversed, although this did include a significant GNSS component. With regard to application of machine vision and sensor fusion in forest operations, Pierzchała (2017) demonstrated the use of cameras, an accelerometer, IMU and GNSS unit in identifying work phases in a cable logging operation, Lideskog and Karlberg (2016) used machine vision techniques to develop strategies for efficient mound positioning in connection with soil scarification, while Matej (2014) used computer vision in determining the tilt angle of a forest machine, based on the assumption that tree stems it was imaging were vertically orientated.

2. Conclusions This review presented a range of current applications of remote and proximal sensing techniques and their relevance to forest operations. Forest inventory is now routinely carried out with LiDAR in an operational setting, and in this way directly impacts the planning and implementation of forest operations. A fundamental issue identified throughout the review was that, while many papers demonstrate new methods or applications for utilising remotely and proximally sensed data, these methods were not necessarily mature or used in an optimal combination, and there remains a series of challenges to realising almost all the applications discussed. In the same light, the review shows that the potential for making improvements and operationalizing some of the developed approaches and techniques is considerable and should be a focus of forest operations research in the years to come.

The development of remote and proximal sensing technology and techniques will provide a previously inconceivable amount of data. Especially the machinemounted sensors that unceasingly collect vast amounts of data will provide the forest operations researcher with a large and continually increasing basis from Croat. j. for. eng. 38(2017)2

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

which to extract useful information. These data can, with the application of sensible analytical approaches, provide significant opportunities for decision support as well as operations monitoring and evaluation.

At the same time, these possibilities will challenge the forest operations researcher with demands on exceptional skills related to data analysis. The approach to answering new research questions will change from one of gathering data to one of how to use the vast amounts of freely generated data effectively. The forest operations researcher of the future will, in addition, be required to have a certain degree of expertise related to sensors and connectivity of such sensors, described as the internet-of-things. Also, together with remotely and proximally measured big-data come special demands with regards to ethics and data security. Detailed forest and personal information related to land owners, managers, forest contractors, machine operators, forest workers and researchers will be instantaneously accessible via the internet, and protocols for the generation and handling of this data will require continuous modernisation.

Finally, it is anticipated that procedures for incorporating remotely sensed cultural heritage, environmental, and biological data will be continually developed as they become a central part of harvesting planning in the future.

B. Talbot et al.

Aruga, K., Sessions, J., Akay, A.E., 2005a: Application of an airborne laser scanner to forest road design with accurate earthwork volumes. Journal of Forest Research 10(2): 113– 123.

Aruga, K., Sessions, J., Miyata, E.S., 2005b: Forest road design with soil sediment evaluation using a high-resolution DEM. Journal of Forest Research 10(6): 471–479.

Astrup, R., Ducey, M.J., Granhus, A., Ritter, T., von Lüpke, N., 2014: Approaches for estimating stand-level volume using terrestrial laser scanning in a single-scan mode. Canadian Journal of Forest Research 44(6): 666–676.

Azizi, Z., Najafi, A., Sadeghian, S., 2014: Forest road detection using LiDAR data. Journal of Forestry Research 25(4): 975– 980.

Bauwens, S., Bartholomeus, H., Calders, K., Lejeune, P., 2016: Forest inventory with terrestrial LiDAR: A comparison of static and hand-held mobile laser scanning. Forests 7(6), 127.

Beck, S.J., Olsen, M.J., Sessions, J., Wing, M.G., 2015: Automated Extraction of Forest Road Network Geometry from Aerial LiDAR. European Journal of Forest Engineering 1(1): 21–33.

Bergdahl, A., Örtendahl, A., Fjeld, D., 2003: The economic potential for optimal destination of roundwood in North Sweden–Effects of planning Horizon and delivery precision.

International Journal of Forest Engineering 14(1): 81–88.

Acknowledgements

Blum, R., Bischof, R., Sauter, U.H., Foeller, J., 2016: Tests of reception of the combination of GPS and GLONASS signals under and above forest canopy in the Black Forest, Germany, using choke ring antennas. International Journal of Forest Engineering 27(1): 2–14.

This paper was written with financial support from the Norwegian Research Council through the project Sustainable Utilization of Forest Resources in Norway (225329/E40) and the H2020 Bio-Based Industries supported project, TECH4EFFECT (Grant Agreement No. 720757).

Brunner, A., Gizachew, B., 2014: Rapid detection of stand density, tree positions, and tree diameter with a 2D terrestrial laser scanner. European Journal of Forest Research 133(5): 819–831.

3. References Akay, A.E., Oğuz, H., Karas, I.R., Aruga, K., 2009: Using LiDAR technology in forestry activities. Environmental Monitoring and Assessment 151(1): 117–125.

Akay, A.E., Sessions, J., 2005: Applying the Decision Support System, TRACER, to Forest Road Design. Western Journal of Applied Forestry 20(3): 184–191.

Ala-Ilomäki, J., Lamminen, S., Sirén, M., Väätäinen, K., Asikainen, A., 2012: Using harvester CAN-bus data for mobility mapping. Proceedings of the Nordic Baltic conference on forest operations – OSCAR. Lazdina, D., Jansons A., Konstantinova, I. Riga, Latvia, LSFRI Silava 25(58): 85–87.

Alam, M.M., Strandgard, M.N., Brown, M.W., Fox, J.C., 2012: Improving the productivity of mechanised harvesting systems using remote sensing. Australian Forestry 75(4): 238– 245.

Croat. j. for. eng. 38(2017)2

Brunner, A., Fredriksson, C., 2012: A thinning model describing individual tree selection by harvester operators. Manuscript.

Campbell, D.M.H., White, B., Arp, P.A., 2013: Modeling and mapping soil resistance to penetration and rutting using LiDAR-derived digital elevation data. Journal of Soil and Water Conservation 68(6): 460–473.

Contreras, M., Aracena, P., Chung, W., 2012: Improving Accuracy in Earthwork Volume Estimation for Proposed Forest Roads Using a High-Resolution Digital Elevation Model.

Croatian Journal of Forest Engineering 33(1): 125–142.

Contreras, M.A., Parrott, D.L., Chung, W., 2016: Designing Skid-Trail Networks to Reduce Skidding Cost and Soil Disturbance for Ground-Based Timber Harvesting Operations.

Forest Science 62(1): 48–58.

Craven, M., Wing, M.G., 2014: Applying airborne LiDAR for forested road geomatics. Scandinavian Journal of Forest Research 29(2): 174–182.

Dralle, K., Tarp-Johansen, M.J., 2010: Method and a system for automatic measurement and tracking of logs, industrial wood and boards, Google Patents.

333

B. Talbot et al.

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

Ducey, M.J., Astrup, R., Seifert, S., Pretzsch, H., Larson, B.C., Coates, K.D., 2013: Comparison of forest attributes derived from two terrestrial LiDAR systems. Photogrammetric Engineering & Remote Sensing 79(3): 245–257.

Dupire, S., Bourrier, F., Berger, F., 2015: Predicting load path and tensile forces during cable yarding operations on steep terrain. Journal of Forest Research 21(1): 1–14.

Kauhanen, H., 2008: Close Range Photogrammetry-Structured Light Approach for Machine Vision Aided Harvesting.

Proc. ISPRS XXI Congress, Beijing, China.

Kiss, K., Malinen, J., Tokola, T., 2015: Forest road quality control using ALS data. Canadian Journal of Forest Research 45(11): 1636–1642.

Fardusi, M.J., Chianucci, F., Barbati, A., 2017: Concept to Practice of Geospatial-Information Tools to Assist Forest Management and Planning under Precision Forestry Framework: a review. Annals of Silvicultural Research 41(1): 3–14.

Krogstad, F., Scheiss, P., 2004: The allure and pitfalls of using LiDAR topography in harvest and road design. Joint Conference of IUFRO 3.06 Forest Operations under Mountainous Conditions and the 12th International Mountain Logging Conference Vancouver, B.C.: 1–10.

Fredriksson, C., 2010: Modellering av skördarförarens trädval vid gallring. Master, Universitet for miljø- og biovitenskap.

Lauterbach, H.A., Borrmann, D., Heß, R., Eck, D., Schilling, K., Nüchter, A., 2015: Evaluation of a backpack-mounted 3D mobile scanning system. Remote Sensing 7(10): 13753–13781.

Gallo, R., Grigolato, S., Cavalli, R., Mazzetto, F., 2013: GNSSbased operational monitoring devices for forest logging operation chains. Journal of Agricultural Engineering 44(2s): 140–144.

Liang, X., Jaakkola, A., Wang, Y., Hyyppä, Y., Honkavaara, E., Liu, J., Kaartinen, H., 2014: The Use of a Hand-Held Camera for Individual Tree 3D Mapping in Forest Sample Plots.

Remote Sensing 6(7): 6587–6603.

Gobakken, T., Bollandsås, O.M., Næsset, E., 2015: Comparing biophysical forest characteristics estimated from photogrammetric matching of aerial images and airborne laser scanning data. Scandinavian Journal of Forest Research 30(1): 73–86.

Liang, X., Wang, Y., Jaakkola, A., Kukko, A., Kaartinen, H., Hyyppä, J., Honkavaara, E., Liu, J., 2015: Forest Data Collection Using Terrestrial Image-Based Point Clouds From a Handheld Camera Compared to Terrestrial and Personal Laser Scanning. IEEE Transactions on Geoscience and Remote Sensing 53(9): 5117–5132.

Haas, J., Ellhöft, K.H., Schack-Kirchner, H., Lang, F., 2016: Using photogrammetry to assess rutting caused by a forwarder–A comparison of different tires and bogie tracks. Soil and Tillage Research 163: 14–20.

Hauglin, M., Hansen, E.H., Næsset, E., Busterud, B.E., Gjevestad, J.G.O., Gobakken, T., 2017: Accurate single-tree positions from a harvester: a test of two global satellite-based positioning systems. Scandinavian Journal of Forest Research: 1–8.

Heinimann, H.R., Breschan, J., 2012: Pre-harvest assessment based on LiDAR data. Croatian Journal of Forest Engineering 33(2): 169–180.

Hellström, T., Lärkeryd, P., Nordfjell, T., Ringdahl, O., 2009: Autonomous Forest Vehicles: Historic, envisioned, and stateof-the-art. International Journal of Forest Engineering 20(1): 31–38.

Holmgren, J., Barth, A., Larsson, H., Olsson, H., 2012: Prediction of stem attributes by combining airborne laser scanning and measurements from harvesters. Silva Fenn 46(2): 227– 239.

Holopainen, M., Vastaranta, M., Hyyppä, J., 2014: Outlook for the Next Generation’s Precision Forestry in Finland. Forests 5(7): 1682–1694.

Hrůza, P., Mikita, T., Janata, P., 2016: Monitoring of Forest Hauling Roads Wearing Course Damage using Unmanned Aerial Systems. Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis 64(5): 1537–1546.

Kankare, V., Vauhkonen, J., Tanhuanpää, T., Holopainen, M., Vastaranta, M., Joensuu, M., Krooks, A., Hyyppä, J., Hyyppä, H., Alho, P.,Viitala, R., 2014: Accuracy in estimation of timber assortments and stem distribution – A comparison of airborne and terrestrial laser scanning techniques. ISPRS Journal of Photogrammetry and Remote Sensing 97(1): 89–97.

334

Lideskog, H., Karlberg, M., 2016: Simulated continuous mounding improvements through ideal machine vision and control. Silva Fennica 50(2): article id 1386.

Lindroos, O., Ringdahl, O., La Hera, P., Hohnloser, P., Hellström, T.H., 2015: Estimating the Position of the Harvester Head–a Key Step towards the Precision Forestry of the Future? Croatian Journal of Forest Engineering 36(2): 147–164.

Lokers, R., Knapen, R., Janssen, S., van Randen, Y., Jansen, J., 2016: Analysis of Big Data technologies for use in agro-environmental science. Environmental Modelling & Software 84: 494–504.

Matej, J., 2014: Determination of forestry machine’s tilt angle using camera and image processing. Computers and Electronics in Agriculture 109: 134–140.

Miettinen, M., Kulovesi, J., Kalmari, J., Visala, A., 2010: New measurement concept for forest harvester head. Field and service robotics. Springer Berlin Heidelberg: 35–44.

Miettinen, M., Ohman, M., Visala, A., Forsman, P., 2007: Simultaneous Localization and Mapping for Forest Harvesters.

Robotics and Automation, 2007 IEEE International Conference on. IEEE: 517–522.

Mousazadeh, H., 2013: A technical review on navigation systems of agricultural autonomous off-road vehicles. Journal of Terramechanics 50(3): 211–232.

Mulla, D.J., 2013: Twenty-five years of remote sensing in precision agriculture: Key advances and remaining knowledge gaps. Biosystems engineering 114(4): 358–371.

Murphy, P.N., Ogilvie, J., Meng, F.-R., White, B., Bhatti, J.S., Arp, P.A., 2011: Modelling and mapping topographic variations in forest soils at high resolution: A case study. Ecological Modelling 222(14): 2314–2332.

Croat. j. for. eng. 38(2017)2

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

B. Talbot et al.

Murphy, P.N.C., Ogilvie, J., Castonguay, M., Zhang, C.-F., Meng, F.-R., Arp, P.A., 2008: Improving forest operations planning through high-resolution flow-channel and wet-areas mapping. The Forestry Chronicle 84(4): 568–574.

to detailed ground and canopy surface models in the design of forest transportation systems. Scientific cooperation for forest technology improvement – FORMEC. Ed. B., Košir, Ljubljana, Slovenia: 52–65.

Naesset, E., Jonmeister, T., 2002: Assessing point accuracy of DGPS under forest canopy before data acquisition, in the field and after postprocessing. Scandinavian Journal of Forest Research 17(4): 351–358.

Sterenczak, K., Moskalik, T., 2015: Use of LIDAR-based digital terrain model and single tree segmentation data for optimal forest skid trail network. iForest – Biogeosciences and Forestry 8(5): 661–667.

Niemi, M.T., Vastaranta, M., Vauhkonen, J., Melkas, T., Holopainen, M., 2017: Airborne LiDAR-derived elevation data in terrain trafficability mapping. Scandinavian Journal of Forest Research: 1–12.

Strandgard, M., Alam, M., Mitchell, R., 2014: Impact of Slope on Productivity of a Self-levelling Processor. Croatian Journal of Forest Engineering 35(2): 193–200.

Næsset, E., 2004: Accuracy of forest inventory using airborne laser scanning: evaluating the first nordic full-scale operational project. Scandinavian Journal of Forest Research 19(6): 554–557.

Olivera, A., Visser, R., 2014: Integration of harvester data and geospatial information. Harvesting technical note HTN07-03.

Future Forest Research Ltd. Rotorua, New Zealand.

Olivera, A., Visser, R., 2016: Development of forest-yield maps generated from Global Navigation Satellite System (GNSS)-enabled harvester StanForD files: preliminary concepts. New Zealand Journal of Forestry Science 46(1): 3.

Pierzchała, M., 2017: Developing the use of imaging technology for forest operations. PhD, Norwegian University of Life Sciences.

Pierzchała, M., Talbot, B., Astrup, R., 2016: Measuring wheel ruts with close-range photogrammetry. Forestry 89(4): 383– 391.

Pohjankukka, J., Riihimäki, H., Nevalainen, P., Pahikkala, T., Ala-Ilomäki, J., Hyvönen, E., Varjo, J., Heikkonen, J., 2016: Predictability of boreal forest soil bearing capacity by machine learning. Journal of Terramechanics 68: 1–8.

Rahlf, J., Breidenbach, J., Solberg, S., Næsset, E., Astrup, R., 2014: Comparison of four types of 3D data for timber volume estimation. Remote Sensing of Environment 155: 325–333.

Svenson, G., Fjeld, D., 2016: The impact of road geometry and surface roughness on fuel consumption of logging trucks.

Scandinavian Journal of Forest Research 31(5): 526–536.

Sverdrup-Thygeson, A., Ørka, H.O., Gobakken, T., Næsset, E., 2016: Can airborne laser scanning assist in mapping and monitoring natural forests? Forest Ecology and Management 369: 116–125.

Søvde, N., Løkketangen, A., Church, R., Oppen, J., 2015: A semi-greedy metaheuristic for the European cableway location problem. Journal of Heuristics 21(5): 641–622.

Søvde, N., Sætersdal, M., Løkketangen, A., 2014: A ScenarioBased Method for Assessing the Impact of Suggested Woodland Key Habitats on Forest Harvesting Costs. Forests 5(9): 2327–2344.

Søvde, N.E., 2015: Algorithm for estimating the suitability of potential landing sites. Mathematical and Computational Forestry & Natural Resource Sciences 7(1): 1–8.

Søvde, N.E., Løkketangen, A., Talbot, B., 2013: Applicability of the GRASP metaheuristic method in designing machine trail layout. Forest Science and Technology 9(4): 187–194.

Talbot, B. Astrup, R., 2014: Forest Operations and Ecosystems Services in Norway – a review of the issues at hand and the opportunities offered through new technologies. Journal of Green Engineering 4(4): 271–290.

Ringdahl, O., Lindroos, O., Hellström, T., Bergström, D., Athanassiadis, D., Nordfjell, T., 2011: Path tracking in forest terrain by an autonomous forwarder. Scandinavian Journal of Forest Research 26(4): 350–359.

Tang, J., Chen, Y., Kukko, A., Kaartinen, H., Jaakkola, A., Khoramshahi, E., Hakala, T., Hyyppä, J., Holopainen, M., Hyyppä, H., 2015: SLAM-Aided Stem Mapping for Forest Inventory with Small-Footprint Mobile LiDAR. Forests 6(12): 4588–4606.

Risbøl, O., Briese, C., Doneus, M., Nesbakken, A., 2014: Monitoring cultural heritage by comparing DEMs derived from historical aerial photographs and airborne laser scanning.

Journal of Cultural Heritage 16(2): 202–209.

Tomaštík, J.J., Tomaštík, S.J., Saloň, Š., Piroh, R., 2016: Horizontal accuracy and applicability of smartphone GNSS positioning in forests. Forestry: An International Journal of Forest Research 90(2): 187–198.

Rönnholm, P., Liang, X., Kukko, A., Jaakkola, A., Hyyppä, J., 2016: Quality analysis and correction of mobile backpack laser scanning data. ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences: 41–47.

Vastaranta, M., Latorre, E., Luoma, V., Saarinen, N., Holopainen, M., Hyyppä, J., 2015: Evaluation of a Smartphone App for Forest Sample Plot Measurements. Forests 6(4): 1179.

Sawaguchi, I., Nishida, K., Shishiuchi, M., Tatsukawa, S., 2003: Positioning precision and sampling number of DGPS under forest canopies. Journal of Forest Research 8(2): 133– 137.

Scheiss, P., 2005: The application of spatially explicit digital elevation and canopy surface data for harvest planning: moving from coarse digital elevation models and stand averages Croat. j. for. eng. 38(2017)2

Vega-Nieva, D.J., Murphy, P.N.C., Castonguay, M., Ogilvie, J., Arp, P.A. 2009: A modular terrain model for daily variations in machine-specific forest soil trafficability. Canadian Journal of Soil Science 89(1): 93–109.

Vierling, K.T., Vierling, L.A., Gould, W.A., Martinuzzi, S., Clawges, R.M., 2008: Lidar: shedding new light on habitat characterization and modeling. Frontiers in Ecology and the Environment 6(2): 90–98.

335

B. Talbot et al.

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327–336)

Visser, R., Raymond, K., Harrill, H., 2014: Mechanising steep terrain harvesting options. NZ Journal of Forestry 59(3): 3–8.

Westlund, K., Rönnqvist, M., Flisberg, P., Friberg, G., 2015: BeST Way – The next generation of forwarding planning 16th Symposium for systems analysis in forest resources (SSAFR).

L.O., Eriksson. Uppsala, Sweden, SLU, Skogforsk.

White, R.A., Dietterick, B.C., Mastin, T., Strohman, R., 2010: Forest Roads Mapped Using LiDAR in Steep Forested Terrain. Remote Sensing 2(4): 1120.

Wing, M.G., Eklund, A., 2007: Performance comparison of a low-cost mapping grade global positioning systems (GPS) receiver and consumer grade GPS receiver under dense forest canopy. Journal of Forestry 105(1): 9–14.

Ziesak, M., Marques, A.F., Rasinmaki, J., Rosset, C., Nummila, K., Scholz, J., Mittlboeck, M., Pinho de Sousa, J., Häkli, J., Rommel, D., 2014: Advances in forestry control and automation systems i Europe – FOCUS: the concept idea in a

multinational EU research project. Proceedings of the 6th Precision Forestry Symposium: The anchor of your value chain. P., Ackerman, E., Gleasure and H., Ham. Stellenbosch, South Africa, Faculty of AgriSciences, Stellenbosch University: 114 p. Öhman, M., Miettinen, M., Kannas, I., Jutila, J., Visala, A., Forsman, P., 2008: Tree measurement and simultaneous localization and mapping system for forest harvesters. Field and Service Robotics: Results of the 6th International Conference. C., Laugier, R., Siegwart 42: 369–378. Ågren, A., Lidberg, W., Ring, E., 2015: Mapping Temporal Dynamics in a Forest Stream Network—Implications for Riparian Forest Management. Forests 6(9): 2982. Ågren, A., Lidberg, W., Strömgren, M., Ogilvie, J., Arp, P., 2014: Evaluating digital terrain indices for soil wetness mapping–a Swedish case study. Hydrology and Earth System Sciences 18(9): 3623–3634.

Authors’ addresses:

Received: March 19, 2017 Accepted: June 08, 2017

336

Bruce Talbot, PhD. * e-mail: bta@nibio.no Marek Pierzchała, PhD. e-mail: map@nibio.no Rasmus Astrup, PhD. e-mail: raa@nibio.no Norwegian Institute for Bioeconomy Research Department of Forestry and Forest Resources Hoegskoleveien 7 1430 Ås NORWAY * Corresponding author Croat. j. for. eng. 38(2017)2

B. Talbot et al.

Applications of Remote and Proximal Sensing for Improved Precision in Forest Operations (327â&#x20AC;&#x201C;336)

ISSN 1845-5719

338

Croat. j. for. eng. 38(2017)2