CROJFE - Volume 34, Issue 2 by Studio Bernardic

In memoriam – U spomen

Yuri Gerasimov (1946–2013)

Yuri Gerasimov was born on the 8th August 1964 and passed away on a business trip in Germany on the 30th September 2013. Yuri graduated in 1986 from the Forest Engineering Faculty of the Petrozavodsk State University. Then he worked as a researcher and did his post graduate and doctorate studies in the Leningrad Forest Engineering Academy. In 1995 he was awarded the doctor of science degree by the St. Petersburg Forest Engineering Academy. He served then the Petrozavodsk State University 1995 – 2000, first as a teacher, then as an associate professor and finally as a professor of forest technology. During that time, he also participated in cooperation and staff exchange between the Petrozavodsk State University and the University of Joensuu in Finland. Yuri moved to the Stora Enso company in 2000, first as the Head of the Representative Office in Petrozavodsk, and later as the Business Planning Manager in the St. Petersburg office.

Croat. j. for. eng. 34(2013)2

In April 2004, Yuri moved to Joensuu in Finland and back to research. He started at the Finnish Forest Research Institute (Metla) first as a project researcher in Russian forestry related R&D projects. His task was to develop methods and tools for the analysis of different forest management practices, harvesting technologies, and wood procurement methods for the conditions in Northwest Russia. In late 2009, Yuri was elected for a permanent position in Metla as a senior researcher. As a senior researcher, Yuri continued dealing the above themes, but also including forestry in Central and Eastern European countries. Based on his excellent and diverse contacts research cooperation strengthened not only with the Petrozavodsk State University, but also with many other partners. Metla’s register on performance reflects his dedication to research: on top of about 150 publications, tens of presentations in conferences and workshops. Last years were productive. Few articles are still in the peer review process, and several articles were in the planning phase. Yuri was supposed to present results of one of the most recent article in Stralsund, Germany, at the 46th International Symposium on Forestry Mechanisation, but he passed away just a few hours earlier. Yuri was a member of the Russian Academy of Natural Sciences, member of the Scientific Council evaluating PhD and DSc thesis in the Petrozavodsk State University, editor in chief of the refereed journal Resources and Technology. He was also docent at the University of Eastern Finland, the School of Forest Sciences. With respect, Timo Karjalainen Professor, Finnish Forest Research Institute

173

Original scientific paper – Izvorni znanstveni rad

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing Juha Laitila, Marica Kilponen, Yrjö Nuutinen Abstract – Nacrtak The aim of this case study was to clarify the productivity and cost of a system based on bundling logging residues at the roadside landing with the forwarder-mounted logging residue bundler. In order to find the bundling productivity, a set of time studies was carried out, in which several working techniques were tested and evaluated. The cost-efficiency of the roadside bundling system was compared with the conventional bundling system, wherein the logging of residue logs is made directly in the terrain and, after bundling, the logging residue logs are forwarded to the roadside landing with a forwarder. The harvesting cost (bundling and forwarding) of the extracted wood biomass to the roadside landing was calculated for bundling systems using time study data obtained from this study and productivity models and cost parameters acquired from the literature. The productivity of roadside bundling ranged from 48 to 53 logging residue logs per effective working hour (E0h), depending on the working technique used, and the mean time required to produce one logging residue log ranged from 83.6 to 92.3 seconds (E0h). The harvesting costs of the logging residue logs (€/m3) at the roadside landing were 11.5–13.7 €/m3 for the system based on bundling in terrain and 10.8–17.7 €/m3 for the system based on bundling at the roadside landing, when the forwarding distance was in the range 100–600 m and the removal of logging residues was in the range 30–90 m³/ha (m3 = solid cubic metre). According to our results, bundling at the roadside landing allowed a reduction in harvesting costs, when the forwarding distance of the logging residues was 100 m or less and removal was beyond 50 m³/ha. The cost savings were quite small, however, at 0.1–0.7 €/m³. Keywords: Bundling, logging residue logs, productivity, compaction, harvesting, forest biomass, logging residue bundler

1. Introduction – Uvod The system, based on logging residue logs and comminution at a plant was launched into commercial use in Finland in the beginning of 2000, when the supply of forest biomass to the world’s largest biofuelfired CHP plant – Oy Alholmes Kraft Ab – was developed (Laitila 2000, Poikola et al. 2002). Due to the long transport distances, large procurement area and enormous annual harvesting volumes, the circumstances for introducing the novel large-scale production technology were favourable on the west coast of Finland. In addition, integration of bundle production into the procurement of industrial roundwood was straightforward, and the synergies were significant because the CHP plant Alholmens Kraft is located within the Croat. j. for. eng. 34(2013)2

large pulp-, paper- and sawmill integrate of the forest industry company UPM (Laitila 2000, Poikola et al. 2002). Another benefit was that all the machines in the supply system were able to operate independently of each other, making the system more efficient and reliable (Laitila 2000, Poikola et al. 2002). In the bundling method (Fig. 1), logging residues are bundled into cylindrical bales using the compacting device mounted on top of the forwarder deck (e.g. Laitila 2000, Ranta 2002, Cuchet et al. 2004, Johansson et al. 2006, Kärhä and Vartiamäki 2006, Stampfer and Kanzian 2006, Spinelli and Magagnotti 2009, Lindroos et al. 2010, Spinelli et al. 2012a, Spinelli et al. 2012b). Feeding and compacting is usually a continuous process, and for these bundling machines (e.g. Timber-

175

J. Laitila et al.

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

Fig. 1 Forwarder-mounted Timberjack/John Deere 1490D logging residue bundler operating in a clear-cut area Slika 1. Rad forvardera Timberjack/John Deere 1490D s ugrađenim bandlerom za šumski ostatak u čistoj sječi jack/John Deere 1490D, Pika/Pinox RS 2000), compacting can be divided into three phases (Ranta 2002). In the first phase, the collected logging residues are pressed by feed rollers. The compacting then continues in a rectangular presser. The last compacting phase ends with the binding of the pulse-fed logging residue bundle and, finally, the bundle is cut into the desired length with a chainsaw. The length of the bundle (logging residue log) can be selected but it is typically about 3 m with a diameter of 65–75 cm. The average weight is 418 kg (sd. 111), the solid volume 0.5 m³ (sd. 0.09) and the energy content approx. 1 MWh (sd. 0.17) (Kärhä and Vartiamäki 2006). To accrue the benefits of compaction as early as possible along the supply chain, log-like bundles are made directly at the stump (Fig. 1) and the bundler must therefore be installed on a vehicle capable of accessing the cut area (Laitila 2000, Asikainen et al. 2001, Ranta 2002, Cuchet et al. 2004, Kärhä and Vartiamäki 2006, Spinelli and Magagnotti 2009, Spinelli et al. 2012a, Spinelli et al. 2012b). At the stand, the bundler drives to the logging residue heap or windrow and stops to load logging residues into the bundler in-feed. The bundling unit follows an automatic cycle with actions activated by internal load sensors. Loading and in-feeding work continue until no more logging residue is within crane reach. Then the forwarder-mounted bundler moves to the next windrow and resumes the work cycle (Asikainen et al. 2001, Cuchet et al. 2004, Kärhä and Vartiamäki 2006). In the studies conducted in Finland and France, the bundling productivities have ranged from 11 to 26 log-

176

ging residue logs per operating working hour (E15h) (Asikainen et al. 2001, Cuchet et al. 2004, Kärhä and Vartiamäki 2006). In the most recent Swedish followup study, the average bundling productivity was 28 logging residue logs per effective working hour (E0h) for the John Deere 1490D logging residue bundler (Eliasson 2011). After bundling, the logging residue logs are forwarded to the roadside landing with standard forwarders. At the landing, the logging residue logs are stacked alongside conventional timber assortments and transported with the standard timber trucks to the terminal or end-use facility (Laitila 2000, Asikainen et al. 2001, Ranta 2002, Johansson et al. 2006, Kärhä and Vartiamäki 2006, Stampfer and Kanzian 2006, Jylhä and Laitila 2007, Spinelli and Magagnotti 2009, Lindroos et al. 2010, Spinelli et al. 2012a, Spinelli et al. 2012b). Logging residue logs dry well and have good storage properties if handled correctly (Petterson and Nordfjell 2009, Eliasson 2011). The unloading of the logging residue logs takes place at the end-use facility with similar equipment to that for unloading saw logs or pulpwood. In the most efficient cases, the logging residue logs are unloaded directly from the timber truck to the feeding table of the stationary crusher (Laitila 2000, Asikainen et al. 2001, Ranta 2002). In the large-scale procurement of logging residue chips, bundling has proved to be cost-efficient when operating with long forwarding and road transportation distances (e.g. Andersson 2000, Laitila 2000, Asikainen et al. 2001, Kärhä and Vartiamäki 2006, Ranta and Rinne 2006). However, in current harvesting operations, e.g. in Finland, the average forwarding distances for logging residues are usually < 300 m, road transporting distances to CHP plants < 100 km and the annual consumption of forest chips per CHP plant almost invariably < 100,000 m³ (solid) (Asikainen et al. 2001, Laitila et al. 2010, Karttunen et al. 2012), and the expected breakthrough of logging residue bundling technology was therefore not achieved. Kärhä and Vartiamäki (2006) underlined that the prerequisite for increased bundling volumes is a reduction in the cost of the most expensive sub-stage of the bundling supply chain, i.e. the bundling itself. This requires, e.g., improved recovering conditions at bundling sites, increased bundling productivity and the execution of bundling operations in two work shifts using an efficient bundler and efficient operator working methods (Kärhä and Vartiamäki 2006). A less well-developed alternative in Nordic is to forward loose logging residues and bundle them at the landing. Potential benefits of such a bundling process include a higher concentration of logging residues beCroat. j. for. eng. 34(2013)2

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

J. Laitila et al.

Table 1 Number of bundled logging residue logs per working technique and temperature during the time study Tablica 1. Broj izrađenih svežnjeva tijekom studija vremena po radnoj tehnici i temperaturi Working technique I

Working technique II

Working technique III

Working technique IV

Working technique V

Radna tehnika I

Radna tehnika II

Radna tehnika III

Radna tehnika IV

Radna tehnika V

201

193

113

–10

–3

–22

–3 & –22

–22

No. of logging residue logs Broj izrađenih svežnjeva Temperature, C° Temperatura, C°

cause residue concentration and presentation have already been recognized as major variables affecting bundler productivity (Cuchet et al. 2004, Kärhä and Vartiamäki 2006) and avoidance of requirements for the expensive bundling vehicles to have off-road capabilities (Wittkopf 2004, Kanzian 2005, Stampfer and Kanzian 2006, Spinelli and Magagnotti 2009, Lindroos et al. 2010, Gallagher et al. 2010, Spinelli et al. 2012b). A truck-mounted bundler (Spinelli and Magagnotti 2009, Lindroos et al. 2010, Spinelli et al. 2012b) would also be a solution for more cost-efficient recovery of logging residues from small, scattered cutting areas due to the smaller relocation costs. One option tested in a Southern U.S. tree length logging operation, in order to reduce costs and maximize bundling efficiency, was to adapt the simplified bundler unit for a motorized trailer and feed it by the separate loader at the landing (Gallagher et al. 2010). According to Spinelli and Magagnotti (2009), working at the roadside allows for a reduction in machine moving time from 1–2 min/ton (Cuchet et al. 2004) to 0.3–0.5 min/ton, but this fact alone does not seem to entail a marked productivity gain; in fact, the forwarder-mounted bundler seems to compensate for this with a faster bundling pace, which is the result of its capacity to bundle while moving. In this case, the time is recorded as »moving«, but the machine is also bundling during part of this time, thus maintaining sustained productivity.

2. Aim of the study – Cilj istraživanja The aim of this case study was to clarify the productivity and cost of a system based on bundling logging residues at the roadside landing with the forwarder-mounted logging residue bundler (John Deere 1490D). In order to find the bundling productivity, a set of time studies was carried out, in which several working techniques were tested and evaluated. The cost-efficiency of the roadside bundling system was Croat. j. for. eng. 34(2013)2

compared with the conventional bundling system, wherein bundles are made directly in the terrain, and, after bundling, the logging residue logs are forwarded to the roadside landing with a forwarder. The harvesting cost (€/m³) of wood biomass extracted to the road side landing was calculated for both bundling systems using time study data obtained from this study, and productivity models and cost parameters acquired from the literature (Ranta 2002, Kärhä et al. 2004, Laitila et al. 2010). The bundling system cost comparison was made at the stand level, and, in the cost comparison, the forwarding distance was in the range 100–600 m and the removal of logging residues was in the range 30–90 m³/ha.

3. Material and Methods – Materijal i metode 3.1 Time study of roadside bundling – Studij vremena izradbe svežnjeva uz cestu The time study of roadside bundling was conducted in December 2009 at a roadside landing (62°19.398’N, 30°38.691’E) located in the province of North Karelia in eastern Finland. During the time study, 683 logging residue logs were bundled and five different working techniques were tested (Fig. 2, Table 1). The time study was carried out mainly in natural light during the daytime (7:00–6:00). The sky was cloudless and the temperature range was –3 to –22 C° (Table 1). The ground had snow cover of 0–1 cm during the experiments (Fig. 3). The length of the bundles was 3 m and the diameter 70 cm. In the time studies, the productivity unit logged residue logs per effective working hour (E0h). The bundled logging residues originated from a clear-cut stand dominated by Norway spruce (Picea abies), with an average age of the harvested trees of 90 years, height 24 m and diameter (d1.3) 28 cm. The minimum length of the harvested industrial roundwood was 3 m and the minimum top diameter was 7 cm (over the bark). The clear cut had been carried out me-

177

J. Laitila et al.

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

Fig. 2 Layout of the bundling study arrangements at the roadside landing using working technique I, II, III, IV or V Slika 2. Prikaz raspodjele stadija izradbe svežnjeva na pomoćnom stovarištu pomoću radne tehnike I, II, III, IV ili V chanically in April 2009 using a cut-to-length method adapted for the recovery of logging residues (Brunberg 1991, Wigren 1991, Wigren 1992, Nurmi 1994). In July 2009, after drying, »brown« logging residues were forwarded to the roadside landing and piled into stacks with a width of 7 m and a height of 5 m (Fig. 3). The total area of the clear cut was extraordinarily large (50 hectares), which made it possible to carry out the bundling study at one stand with homogeneous raw material and similar bundling conditions for each working technique at the roadside landing. The bundle properties (moisture, solid content, etc.) were not studied because they were expected, as estimated by the author, to be similar for all the studied working techniques due to the homogenous bundling material and the same logging residue bundler. It was also deemed that the properties of the logging residue logs produced at the roadside landing do not differ from those produced in the terrain, as the raw material and compacting unit/bundler are the same. The layout of the studied working techniques is described in Fig. 2. In working techniques I and II, two machines were operating at the roadside landing because the feeding of the bundler was carried out with a separate loader (forwarder) in order to steer the full hydraulic capacity of the logging residue bundler into the bundling process. The bundler was located across from the logging residue stack, and the loader (forwarder) was on the forest road parallel to the logging residue stack (Fig. 2 and 3). The piling of the bundles in the roadside stack was carried out as a separate operation with the loader (forwarder) at the end of the roadside bundling operation (I) or during the bundling process with the crane of the logging residue

178

bundler (II). In working techniques III, IV & V, the feeding of the bundler was carried out with the crane of the logging residue bundler, and one machine was operating at the landing. The piling of logging residues was carried out either as a separate operation after bundling (III and IV) or during bundling (V). In working technique III, the logging residue bundler was located on the forest road parallel to the logging residue stack, whereas in working techniques IV and V, the bundler was located across from the logging residue stack (Fig. 2).

Fig. 3 Separate loading of logging residues with the Valmet 840.3 forwarder to the feeding table of the John Deere 1490D logging residue bundler (working methods I and II) Slika 3. Odvojen utovar šumskoga ostatka forvarderom Valmet 840.3 na opskrbnu traku bandlera za izradbu svežnjeva John Deere 1490D (radna metoda I i II) Croat. j. for. eng. 34(2013)2

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

The machines used in the study were a John Deere 1490D Eco III logging residue bundler and a Valmet 840.3 eight-wheel forwarder (Fig. 3). The crane models of the bundler and forwarder were John Deere CF7 and Cranab CRF 8.1 C, and both were equipped with a special logging residue grapple (e.g. Ranta 2002, Kärhä and Vartiamäki 2006). Skilful and motivated machine operators were pre-trained for the studied working techniques and they had more than five years working experience in bundling or forwarding logging residues and logging residue logs. The time study was carried out manually using the Rufco-900 field computer (Nuutinen et al. 2008). The output was estimated by counting all the logging residue logs produced during the observation time. The working time was recorded by applying a continuous timing method in which a clock runs continuously and the times for different elements are separated from each other by numeric codes (e.g. Harstela, 1991). The logging residue bundler working time was divided into effective working time (E0h) and delay time (Haarlaa et al. 1984, Mäkelä 1986), which is a common method employed in Nordic work studies. Effective working time was divided into the following work phases in order of priority: Loading and bundling: The work cycle began when the grapple started to move towards the logging residue stack and ended when a residue bunch was lifted and placed on the feeding table or into the chamber of the bundler and the feed rollers started to pull residues into the bundler or the compressing cylinders of the bundler started to pull residues into the chamber of the bundler. Bundling (loading is idled): This began when the feeding rollers or belts of the bundler started to pull residues into the bundler or the compressing cylinders of the bundler started to pull residues into the chamber of the bundler and ended when the individual logging residue log was wrapped. The number of binding points was chosen to be six with double twines, because frozen and dry logging residue is breaking easily and requires more binding. Cross-cutting (bundling and loading are idled): This began when a chainsaw emerged from a defence case and ended when the bundle dropped off. Moving: This began when the bundler or the separate loader (forwarder) started to move and ended when the bundler and/or loader stopped moving to perform other activity. The moving time consisted of the short move from one work location to another at the roadside landing. Piling: The piling of logging residue logs onto the roadside stack while bundling or as a separate operaCroat. j. for. eng. 34(2013)2

J. Laitila et al.

tion after roadside bundling from the bundle heaps with the crane of the bundler or the separate loader (forwarder). Arrangements: Repositioning of logging residues on the roadside stack in order to improve the loading work or shake off snow, ice or other impurities. Delays: Time not related to productive bundling work but with the reason for the interruption recorded. The main reasons for the delayed times being less than 15 minutes were bundler maintenance (e.g. tightening or replacing the sawchain and adding a bundling cord to the wrapping unit of the bundler), organizational delays (e.g. telephone calls) or personal breaks.

3.2 Cost comparison of bundling methods Usporedba troškova među metodama izradbe svežnjeva The cost comparison of bundling systems was made at stand level and in the cost comparison, the forwarding distance was in the range 100–600 m and the removal of logging residues was in the range 30–90 m³/ha. At the stand, logging residues were stacked in good heaps and the heaps were located on both sides of the strip road. The nature and slope of the ground surface were normal = flat (Tavoiteansioon perustuvat puutavaran … 1990). Bundling productivity in terrain was calculated using the time-consumption model made for the Timberjack/John Deere 1490D logging residue bundler (Kärhä et al. 2004). Bundling productivity at the roadside landing was based on working technique V reported herein. The solid volume of the logging residue logs was 0.55 m³ (Kärhä et al. 2004) for both bundling methods. The length of the logging residue logs was 3 m and it was bound at six points. The effective working hour productivity (E0h) of the bundler in terrain or at the landing was converted into operating hour productivity (E15h) by the coefficient 1.274 (Kärhä et al. 2004). The bundling productivity at the landing was 33.8/E15h logging residue logs and in terrain it was calculated as 17–18/E15h logging residue logs as a function of logging residue removal (30–90 m³/ha). The figures for the forwarding productivity of the logging residues and logging residue logs from the clear cut with a heavy forwarder (Fig. 4) were calculated using the time consumption models presented in studies by Ranta (2002) and Kärhä et al. (2004), and the effective working hour productivity (E0h) of forwarding was converted into operating hour productivity (E15h) by the coefficient of 1.224 (Kuitto et al. 1994, Kärhä et al. 2004). The payload of the forwarder

179

J. Laitila et al.

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

productivity of the logging residue logs was in the range 15.2–30.8 m3/ E15h. The operating hourly costs of the forwarder and forwarder-mounted logging residue bundler were based on the study by Laitila et al. 2010 and updated to the current cost level (November 2012) with the cost index of forest machinery »MEKKI« produced by Statistics Finland (http://www.stat.fi/til/mekki/yht_en. html) in order to guarantee the validity of the cost comparison results. The operating hourly costs of the forwarder and logging residue bundler in this study were 71.8 €/E15h and 85.3 €/E15h, respectively.

Fig. 4 Forwarding of pre-piled logging residues with a heavy forwarder at the stand Slika 4. Izvoženje neuhrpanoga šumskoga ostatka s teškim forvarderom u sastojini was set at 7.2 m³ for logging residues and 25 pieces (13.8 m³) for logging residue logs. The forwarding productivity of the logging residues was in the range 5.5–11.8 m3/E15h as a function of forwarding distance and removal of the logging residues. The forwarding

4. Results – Rezultati 4.1 Results of the time study – Rezultati studija vremena In relative terms, combined loading and bundling required on average 55–68 % and cross-cutting 12–21 % of the effective working time (E15h), when bundling logging residues at the roadside landing (Fig. 5). The piling of logging residue logs took time, 11–13 %, except for working technique II (0 %), in which piling was carried out with the crane of the logging residue bundler during the other work phases (Fig. 5). With working tech-

Fig. 5 Relative time consumption of work phases (%), when bundling logging residues at the roadside landing Slika 5. Relativan utrošak vremena po radnim fazama (%) pri izradbi svežnjeva na pomoćnom stovarištu

180

Croat. j. for. eng. 34(2013)2

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

J. Laitila et al.

Moving time of the loader, s Premještanje utovarivača, s Piling of the logging residue logs to the stack, s Uhrpavanje izrađenih svežnjeva na složaj, s Moving time of the bundler, s Premještanje bandlera, s Arrangements of the logging residue stack, s Razmještavanje složaja šumskoga ostatka, s Cross-cutting of the logging residue logs, s Prerezivanje izrađenih svežnjeva, s Separate bundling of logging residues, s Zasebna izradba svežnjeva od šumskoga ostatka, s Loading and bundling of logging residues, s Utovar i izradba svežnjeva od šumskoga ostatka, s Total time, s Ukupno vrijeme, s

Radna tehnika V

Working technique V

Radna tehnika IV

Working technique IV

Radna tehnika III

Working technique III

Radna tehnika II

Working technique II

Radna tehnika I

Working technique I

Table 2 Average time consumption of the work phases per logging residue log and working technique Tablica 2. Prosječan utrošak vremena radnih sastavnica po izrađenom svežnju i radnim tehnikama

2.0

1.2

–

10.8

–

11.8

9.8

9.4

3.4

2.5

6.2

7.9

5.1

1.2

1.3

2.5

1.2

1.8

11.5

14.9

11.3

13.6

13.2

0.8

3.1

4.6

6.0

8.6

55.5

48.5

55.9

46.8

45.6

85.3

71.6

92.3

85.3

83.6

Grapple load time, s Vrijeme utovara kliještima, s Average number of grapple loads per bundle Prosječan broj utovara kliještima po svežnju

Radna tehnika V

Working technique V

Radna tehnika IV

Working technique IV

Radna tehnika III

Working technique III

Radna tehnika II

16.4

15.5

15.7

17.1

3.4

3.1

3.6

3.0

2.7

niques I and II, the shares of mere bundling were 1 % and 4 %, respectively and, with working techniques III, IV and V, the shares were 5–10 % (Fig. 5). The share of arrangements was 1–3 % for all the working techniques. The moving time for the logging residue bundler was Croat. j. for. eng. 34(2013)2

Working technique II

Radna tehnika I

Working technique I

Table 3 Average time consumption of the loading cycle (grapple load time) and the average number of grapple loads per logging residue log and working technique used Tablica 3. Prosječan utrošak vremena ciklusa utovara (vrijeme utovara kliještima) i prosječan broj zahvata hvatalom po izrađenom svežnju i korištenoj radnoj tehnici

6–9 % of the effective working time when using working techniques III, IV and V (Fig. 5). With working techniques I and II, the total moving time was 6 % out of which the loader accounted for 2 and the logging residue bundler for 4 % (Fig. 5).

181

J. Laitila et al.

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

Fig. 6 Average time consumption of work phases per logging residue log in seconds Slika 6. Prosječan utrošak vremena radnih sastavnica po izrađenom svežnju u sekundama Combined bundling and loading of logging residues at the roadside landing took on average 45.6–55.9 seconds per logging residue log (Fig. 6 and Table 2). The mean number of crane grapple loads required to produce a logging residue log ranged from 2.7 to 3.6 and the average grapple load time was in the range 15.5–17.1 seconds per crane cycle (Table 3). The total mean time required to produce one logging residue log ranged from 71.6 to 92.3 seconds depending on the working technique used (Table 2, Fig. 6). There were no big differences between the working techniques in terms of bundling productivity per effective working hour (pieces/E0h) during the time studies and the feeding of the bundler with a separate loader did not improve the bundling productivity compared with the self-loading logging residue bundler. The productivity of mere bundling was in the range 48–53 logging residue logs per effective working hour (E0h) depending on the working technique used (Fig. 7). When calculating the bundling productivity, the mere bundling included the time consumption of the work phases loading and bundling, bundling, cross-cutting and arrangements. The bundling productivity was 45–49 pieces/E0h at the roadside landing when the moving time of the bundler and loader were included in the effective work-

182

ing time and 39–43 pieces/E0h when the piling time was included (Fig. 7). The bundling productivity of working technique III was somewhat lower than that of the other techniques, but this can be explained by the fact that the bundler operator fell ill with the flu on the day of the time study.

4.2 Results of the bundling method cost comparison – Rezultati usporedbe troškova među metodama izradbe svežnjeva The harvesting costs of the logging residue logs (€/m3) at the roadside landing were 11.5–13.7 €/m3 for the system based on bundling in terrain and 10.8–17.7 €/m3 for the system based on bundling at the roadside landing, when the forwarding distance was in the range 100–600 m and the removal of logging residues was in the range 30–90 m³/ha (Fig. 8, Table 4, cf. section 3.2 in the article). According to our results, bundling at the roadside landing enabled a reduction in harvesting costs when the forwarding distance of the logging residues was 100 m or less and the removal was beyond 50 m³/ha (Fig. 8, Table 4). The cost savings, however, were quite small, 0.1–0.7 €/m³. Traditional terrain bundling was clearly more cost-competitive in all stand circumstances when the forwarding distance was more than 200 m (Fig. 8, Table 4). Croat. j. for. eng. 34(2013)2

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

J. Laitila et al.

Fig. 7 Average time consumption of work phases per logging residue log in seconds Slika 7. Prosječan utrošak vremena radnih sastavnica po izrađenom svežnju u sekundama

5. Discussion – Rasprava According to our time study, feeding the bundler with a separate loader did not improve the bundling productivity compared with the self-loading logging residue bundler, when the length of the logging residue logs was 3 m. The main reason for the result is that the combined loading and bundling work phase was interrupted after every 3–4 grapple loads to cross-cut the produced logging residue log, which means that there was no time for the loading to become a bottleneck in the bundling system even though the efficiency and hydraulic capacity of the compacting unit itself would enable higher productivity. In order to improve the efficiency of a continuous bundling process, either a more efficient cross-cutting of bundles should be developed, or the current length of the logging residue logs should increase within the constraints imposed by the off- and on-road transportation and durability of the logging residue logs. In the studies by Spinelli and Magagnotti (2009), and Gallagher et al. (2010) the highest bundling productivity was achieved with the longest target lengths of logging residue logs. In the study by Gallagher et al. (2010), the bundling productivity was 15.9 tons/E0h when the length of the bundles was 2.5 m and 17.2 tons/E0h for a bundle length of 3.5 m. Croat. j. for. eng. 34(2013)2

The productivity (39–43 pieces/E0h) achieved in this study is higher than that reported in the others studies conducted on the truck-mounted logging residue bundler under central European conditions in Germany, Austria and Italy (Wittkopf 2004, Kanzian 2005, Spinelli and Magagnotti 2009, Spinelli et al. 2012b). In Germany, Wittkopf (2004) reports productivity of 12 pieces/E15h and in Austria, Kanzian (2005) mentions productivity of 11.5–15.2 pieces/E0h. In Italy, the productivity varies between 14 and 22 pieces/E0h (Spinelli and Magagnotti 2009). In studies conducted in Germany and Austria, the length of the logging residue logs was 3 m (Wittkopf 2004, Kanzian 2005), whereas in Italy, the target lengths were 4 and 3 m (Spinelli and Magagnotti 2009). All these studies were conducted on the very same machine, the Timberjack/John Deere 1490 bundler on a 6 x 6 MAN truck (Spinelli et al. 2012b). The unit is powered by the 353 kW engine of the truck and fed by a Timberjack CF 710 crane. The truck is equipped with a modified cab incorporating the crane control seat: this constitutes a second rotating chair mounted to the right of the driving seat with an extended rear window. The overall weight of the truck-base bundler is 24 tons (Spinelli et al. 2012b).

183

J. Laitila et al.

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

Fig. 8 Harvesting cost of the logging residue logs (€/m3) at the roadside landing as a function of forwarding distance (100–600 m) and biomass removal (30–90 m3/ha) Slika 8. Troškovi pridobivanja svežnjeva (€/m3) na pomoćnom stovarištu u ovisnosti o udaljenosti izvoženja (100–600 m) i sječnoj gustoći biomase (30–90 m3/ha) Table 4 Harvesting cost (€/m3) of logging residue logs at the roadside landing when the bundling is done either in terrain or at the roadside landing. The removal of logging residues is 30, 50, 70 or 90 m3/ha and the forwarding distance is in the range 100–600 m Tablica 4. Troškovi pridobivanja (€/m3) svežnjeva na pomoćnom stovarištu kada se svežnjevi izrađuju na sječini ili na pomoćnom stovarištu. Uklanjanje je šumskoga ostatka 30, 50, 70 ili 90 m3/ha, a udaljenost je izvoženja u rasponu 100–600 m Terrain 30 m3/ha

Landing 30 m3/ha

Terrain 50 m3/ha

Landing 50 m3/ha

Terrain 70 m3/ha

Landing 70 m3/ha

Terrain 90 m3/ha

Landing 90 m3/ha

Udaljenost izvoženja

Sječina 30 m3/ha

Stovarište 30 m3/ha

Sječina 50 m3/ha

Stovarište 50 m3/ha

Sječina 70 m3/ha

Stovarište 70 m3/ha

Sječina 90 m3/ha

Stovarište 90 m3/ha

100 m

12.4

13.4

11.9

11.7

11.2

11.5

10.8

200 m

12.7

14.3

12.3

12.7

12.0

12.1

11.9

11.7

300 m

13.0

15.1

12.6

13.6

12.3

12.9

12.2

12.5

400 m

13.3

16.0

12.9

14.5

12.6

13.8

12.4

13.4

500 m

13.5

16.9

13.1

15.3

12.9

14.7

12.7

14.3

600 m

13.7

17.7

13.3

16.2

13.1

15.3

12.9

15.1

Forwarding distance

The original invention of the logging residue bundler was developed by Swedish company Fiberpak AB in 1995, and the first bundling units were mounted on standard forwarders as an attachment. In addition, the bundling unit was operated with a separate control system and steering of the bundling unit was managed with own independent hydraulic pump. Whereas e.g. in the John Deere 1490D logging residue bun-

184

dlers, steering of the bundling unit is shared with the hydraulic line and pump of the forwarder crane. Sharing of the steering system with the forwarder crane limits the maximum power execution for the bundling unit and in principle the maximum productivity of the bundling unit is not possible to achieve without installing an independent hydraulic system or stopping the movements of the forwarder crane completely. Croat. j. for. eng. 34(2013)2

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

6. Conclusions – Zaključci According to the results, bundling at the roadside landing with a forwarder-mounted bundler made it possible to reduce the harvesting costs to 0.1–0.7 €/m³ when the forwarding distance of the logging residues was 100 m or less and the removal was beyond 50 m³/ ha. In practical operations, roadside bundling should be carried out outside the road area because of the large amount of material (needles, bark, small branches) that will be dropped on the ground while bundling ‘brown’ residues. In addition, road traffic may disrupt the bundling work, especially on the public road area, which also limits the usability of truck- and trailermounted logging residue bundlers. In wintertime, the cover of snow and frozen logging residues are obviously a problem too for a roadside bundling system in Nordic conditions. In Finland, the average forwarding distances are close to 300 m (Asikainen et al. 2001, Jylhä et al. 2010), which also limits the wide implementation of a roadside bundling system. The results reported in this paper were based on theoretical time consumption models and cost parameters from earlier bundling and forwarding studies and rather limited time study data on bundling productivity at the roadside landing, which limits the generalization of the results. The study also focused on the effective working time (E0h), which is only part of the total working time. Nevertheless, the results give new estimates for the performance and cost competitiveness of the roadside bundling system in Nordic conditions and the operators involved in the study were skilled, using machinery representatives for the current machines in use. In order to guarantee the reliability of the reported case study observations (Hellström and Hyttinen 1996), the results must be compared with the results of similar case studies, and efforts should be made to verify the observed phenomenon.

7. References – Literatura Andersson, G., 2000: Technology of fuel chip production in Sweden. In publication: Alakangas, E. (edit.). Nordic treasure hunt. Extracting energy from forest residues. VTT symposium 208: 113–125. Asikainen, A., Ranta, T., Laitila, J., Hämäläinen, J., 2001: Hakkuutähdehakkeen kustannustekijät ja suurimittakaavaisen hankinnan logistiikka (Cost factors and large-scale procurement of logging residue chips). University of Joensuu, Faculty of Forestry. Research Notes 131. 107 p. (In Finnish). Brunberg, B., 1991: Tillvaratagande av skogsbränsle – träddelar och trädrester (Harvesting of forest fuels – tree sections and logging residues). Skogsarbeten redogörelse 5. 54 p. (In Swedish). Croat. j. for. eng. 34(2013)2

J. Laitila et al.

Cuchet, E., Roux, P., Spinelli, R., 2004: Performance of a logging residue bundler in the temperate forests of France. Biomass and Bioenergy 27(4): 205–210. Eliasson, L., 2011: Follow-up of the John Deere logging residue bundler. In publication: Thorsén, Å., Björheden, R. & Eliasson, L. (edit.). Efficient forest fuel supply systems, composite report from a four year R&D programme 2007-2010. Skogforsk. 113 p. Gallagher, T., Meadows, S., Mitchell, D., 2010: Optimizing the use of a John Deere bundling unit in a Southern US logging system. In publication: Proceedings of FORMEC 2010 Conference, University of Padua, Italy. 9 p. Haarlaa, R., Harstela, P., Mikkonen, E., Mäkelä, J., 1984: Metsätyöntutkimus (Forest work study). Department of logging and utilization of forest products. Research Notes No. 46. 50 p. (In Finnish). Harstela, P., 1991: Work studies in forestry. University of Joensuu. Silva Carelica 18. 41 p. Hellström, E., Hyttinen, P., 1996: Tapaustutkimusstrategia ja metsätieteet. (Case study strategy and forest science) Folia Forestalia – Metsätieteen aikakauskirja 1996(4): 389–407. (In Finnish). Johansson, J., Liss, J.-E., Gullberg, T., Björheden, R., 2006: Transport and handling of forest energy bundles – advantages and problems. Biomass and Bioenergy 30(4): 334–341. 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. Jylhä, P., Dahl, O., Laitila, J., Kärhä, K., 2010: The effect of supply system on the wood paying capability of a kraft pulp mill using Scots pine harvested from first thinnings. Silva Fennica 44(4): 695–714. Kanzian, C., 2005: Bereitstellung von Waldhackgut (Report from the trials with an energy slash bundler in the mountain).Verfahren Energieholzbundeln im gebirge. Universität fur Bodenkultur Wien, Department fur Wald- und Bodenwissenschaften. 32 p. Karttunen, K., Lättilä, J., Korpinen, O-J., Föhr, J., Enström, J., Ranta, T., 2012: Large scale container supply chain of forest chips. In publication: Bioenergy from Forest 2012. Book of proceedings – Bioenergy from Forest 2012 conference. 263 p. Kuitto, P.-J., Keskinen, S., Lindroos, J., Oijala, T., Rajamäki, J., Räsänen, T., Terävä, J., 1994: Puutavaran koneellinen hakkuu ja metsäkuljetus (Mechanized cutting and forest haulage). Metsäteho report 410. 38 p. Kärhä, K., Vartiamäki, T., Liikkanen, R., Keskinen, S. & Lindroos, J., 2004: Hakkuutähteen paalauksen ja paalien metsäkuljetuksen tuottavuus ja kustannukset (Productivity and cost of slash bundling and bundle forwarding). Metsätehon raportti 179. 88 p. Kärhä, K., Vartiamäki, T., 2006: Productivity and costs of slash bundling in Nordic conditions. Biomass and Bioenergy 30(12): 1043–1052.

185

J. Laitila et al.

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

Laitila, J., 2000: Puupolttoaineiden hankinta Oy Alholmens Kraft Ab:n voimalaitokselle (Wood fuel procurement for Oy Alholmens Kraft Ab power plant). Faculty of Forestry: University of Joensuu. M.Sc. thesis. 51 p. (In Finnish). Laitila, J., Leinonen, A., Flyktman, M., Virkkunen, M., Asikainen, A., 2010: Metsähakkeen hankinta- ja toimituslogistiikan haasteet ja kehittämistarpeet (Challenges and development needs of forest chips procurement and delivery logistics). VTT Tiedotteita Research – Notes 2564. 143 p. (In Finnish). Lindroos, O., Matisons, M., Johansson, P., Nordfjell, T., 2010: Productivity of a prototype truckmounted logging residue bundler and a road-side bundling system. Silva Fennica 44(3): 547-559. Mäkelä, M., 1986: Metsäkoneiden kustannuslaskenta (Guide for forest machine cost calculation). Metsätehon monisteita. 21 p. (In Finnish). Nurmi, J., 1994: Työtavan vaikutus hakkuukoneen tuotokseen ja hakkuutähteen kasautumiseen (The effect of working method on a harvester’s cutting productivity and bunching of logging residues). Metsätieteen aikakauskirja 2: 113–122. (In Finnish). Nuutinen, Y., Väätäinen, K., Heinonen, J., Asikainen, A., Röser, D., 2008: The accuracy of manually recorded time study data for harvester operation shown via simulator screen. Silva Fennica 42(1): 63–72. 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. Poikola, J., Backlund, C., Korpilahti, A., Hillebrand, K., Rinne, S., 2002: The prerequisites of the bundling method in large scale wood fuel procurement. In publication: Alakangas, E. (edit.). Puuenergian teknologiaohjelman vuosikirja 2002. Puuenergian teknologiaohjelman vuosiseminaari, Joensuu, 18–19. syyskuuta 2002. VTT symposium 221: 141–56. Ranta, T., 2002: Logging residues from regeneration fellings for biofuel production – a GIS-based availability and cost

supply analysis. Lappeenranta University of Technology. Finland. Acta Universitatis Lappeenrantaensis 128. 180 p. Ranta, T., Rinne, S., 2006: The profitability of transporting uncomminuted raw materials in Finland. Biomass and Bioenergy 30(3): 231–237. Spinelli, R., Magagnotti, N., 2009: Logging residue bundling at the roadside in mountain operations. Scandinavian Journal of Forest Research 24: 173–181. Spinelli, R., Magagnotti, N., Picchi, G., 2012a: A supply chain evaluation of slash bundling under the conditions of mountain forestry. Biomass and Bioenergy 36: 339–345. Spinelli, R., Lombardini, C., Magagnotti, N., 2012b: Annual usage and long-term productivity of a truck-mounted slash bundler under mountain conditions. European Journal of Forest Research 131: 821–827. 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. Tavoiteansioon perustuvat puutavaran metsäkuljetusmaksut Etelä-Suomessa (Rate for timber forwarding in Southern Finland), 1990: Metsäalan Kuljetuksenantajat ja Koneyrittäjien liitto ry. Handout. (In Finnish). Wigren, C., 1991: Tillvaratagande av trädrester efter slutavverkning med engreppsskördare – studie av en bränsleanpassad method hos Mälärskog (The harvesting of logging residues in final felling with single-grip harvester – study of a fuel adapted method of Mälärskog). Skogarbeten stencil 1991-11-08. 11 p. (In Swedish). Wigren, C., 1992: Uttag av trädrester efter slutavverkining med engreppskördare (Bunching of logging residues integrated with single-grip harvester cutting in final fellings). Skogsarbeten resultat 8. 4 p. (In Swedish). Wittkopf, S., 2004: Einsatz der Bundelmaschine Fiberpac (Working with the Fiberpac bundler). LWF Aktuell, 48: 24– 26. (In German).

Sažetak

Djelotvornost izradbe svežnjeva od šumskoga ostatka na pomoćnom stovarištu – proizvodnost i trošak U radu se prikazuje istraživanje proizvodnosti i troškova sustava temeljenoga na izradbi svežnjeva od šumskoga ostatka na pomoćnom stovarištu pomoću forvardera s ugrađenim bandlerom (John Deere 1490D). Radi utvrđivanja proizvodnosti izradbe svežnjeva proveden je studij vremena u kojem je ispitano i ocijenjeno nekoliko radnih tehnika (slika 2). Šumski ostatak za izradbu svežnjeva potječe iz sastojine gdje je obavljena čista sječa, uz prevladavanje obične smreke (Picea abies) s prosječnom dobi stabala od 90 godina, visine 24 m i prsnoga promjera (d1,3) 28 cm. Tijekom studija vremena izrađena su 683 svežnja od ostatka sječe. Duljina je svežnjeva iznosila 3 m, a promjer 70 cm. Studij je vremena proveden uglavnom u prirodnom svjetlu tijekom dana (7:00–16:00). Nebo je bilo bez oblaka, a raspon je temperature bio od –3 do –22 °C (tablica 1). Velika je površina radilišta omogućila provedbu studija izradbe svežnjeva u istoj sastojini s homogenom sirovinom i sličnim uvjetima rada za svaku radnu tehniku na po-

186

Croat. j. for. eng. 34(2013)2

Productivity and Cost-Efficiency of Bundling Logging Residues at Roadside Landing (175–187)

J. Laitila et al.

moćnom stovarištu. Iskusni i motivirani rukovatelji mehanizacije, prethodno osposobljeni za ispitivanu radnu tehniku, imaju više od pet godina radnoga iskustva na izradbi svežnjeva, izvoženju šumskoga ostatka i/ili izvoženju izrađenih svežnjeva. Ekonomičnost sustava izradbe svežnjeva na pomoćnom stovarištu uspoređena je s konvencionalnim sustavom izradbe pri čemu se svežnjevi izrađuju u sastojini (na radilištu), a naknadno se pomoću forvardera izvoze na pomoćno stovarište. Troškovi pridobivanja (€/m³) dobivene šumske biomase na pomoćnom stovarištu izračunati su za oba sustava izradbe svežnjeva na temelju podataka dobivenih studijem vremena, modelima proizvodnosti i cijenama parametara dobivenih iz literature. Trošak sustava izradbe svežnjeva napravljen je na razini sastojine (radilišta), a u usporedbi troškova korištena je srednja udaljenost izvoženja u rasponu od 100 do 600 metara i uklanjanje ostatka sječe u rasponu od 30 do 90 m³/ha. Proizvodnost izradbe svežnjeva žnjeva njeva na pomoćnom ćnom nom stovarištu štu tu (slika slika 7) kretala se od 48 do 53 svežnja žnja nja šumskoga umskoga ostatka po efektivnom satu rada (E0h), ovisno o radnoj tehnici i prosječnom vremenu potrebnom za proizvodnju jednoga svežnja (slika 6) koje se kretalo od 83,6 za 92,3 sekunde (E0h). Prosječan broj zahvata hvatalom dizalice potreban za čno no vrij vrijeme eme utovara kretalo se u rasponu od 15,5 do 17,1 seproizvodnju svežnjeva žnjeva njeva kretao se od 2,7 do 3,6, a prosječno kunde po ciklusu dizalice (tablica 3). Troškovi izradbe svežnjeva (€/m3) na pomoćnom stovarištu (slika 8, tablica 4) bili su od 11,5 do 13,7 €/m3 za sustav koji se temelji na izradbi svežnjeva u sastojni i od 10,8 do 17,7 €/m3 za sustav koji se temelji na izradbi svežnjeva na pomoćnom stovarištu kada se udaljenost izvoženja kretala u rasponu od 100 do 600 m, a uklanjanje šumskoga ostatka u rasponu 30–90 m³/ha. Temeljem dobivenih rezultata izradba svežnjeva žnjeva njeva na pomoćnom ćnom nom stovarištu štu tu omogućuje ćuje uje smanjenje troškova škova kova pridobivanja kada je srednja udaljenost izvoženja šumskoga ostatka 100 m ili manja i kada je uklanjanje šumskoga ostatka iznad 50 m³/ha. Uštede su vrlo male, u rasponu od 0,1 do 0,7 €/m³. Tradicionalna izradba svežnjeva u sastojini je više troškovno kompetitivna u svim sastojinskim okolnostima kada je srednja udaljenost izvoženja veća od 200 m (slika 8, tablica 4). U praksi izradbu svežnjeva na pomoćnom stovarištu treba provoditi izvan cesta zbog velike količine materijala (iglice, kora, grančice) koji padne na tlo. Osim toga, cestovni promet može poremetiti izradbu svežnjeva, pogotovo na javnim cestama, koje ograničavaju iskoristivost kamiona i prikolica s ugrađenim bandlerom. U zimskim mjesecima pokrov snijega i smrznuti šumski ostaci također su ograničavajući čimbenik za sustav izradbe svežnjeva uz prometnice, npr. u nordijskim uvjetima. Ključne riječi: izradba svežnjeva šumskoga ostatka, proizvodnost, šumska biomasa, bandler

Authors’ address – Adresa autorâ: Juha Laitila, PhD.* e-mail: juha.laitila@metla.fi Nuutinen Yrjö, PhD. e-mail: yrjo.nuutinen@metla.fi Finnish Forest Research Institute, (Agr. & For.) Yliopistokatu 6, FI-80101 Joensuu FINLAND

Received (Primljeno): December 12, 2012 Accepted (Prihvaćeno): March 20, 2013 Croat. j. for. eng. 34(2013)2

Marica Kilponen, MSc. e-mail:kilponenmaricaa@johndeere.com Lappeenranta University of Technology / John Deere Forestry Lokomonkatu 21, FI-33900 Tampere FINLAND * Corresponding author – Glavni autor

187

Original scientific paper – Izvorni znanstveni rad

Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation with Flail and Chipper in Western Australia Mohammad Reza Ghaffariyan, Mark Brown, Raffaele Spinelli Abstract – Nacrtak Roadside chipping is a common harvesting system in Australian plantations, which utilizes a mobile chipper stationed at the field edge to produce high-quality pulp chips for export. The studied harvesting system included a feller-buncher, two grapple skidders, a flail-debarker and a disc chipper. The study goals were to determine machine productivity, operation costs, fuel consumption, chip quality and measure the amount of slash left in the field after harvesting. The average productivity for feller buncher and skidder were about 97.26 GMt/PMH0 and 60.22 GMt/PMH0, respectively. The productivity of flail and chipper averaged at 57.80 GMt/ PMH0 and 58.18 GMt/PMH0 in this case study. The transportation productivity averaged about 57.34 GMt/PMH0. Time studies and regression analysis were used to model machine productivity. Tree size had significant impact on the feller-buncher productivity, while skidding distance was a significant variable affecting skidding productivity. Operation costs were evaluated using the ALPACA (Australian logging productivity and cost appraisal) model. This paper offers valuable information about the impact of different factors on feller-buncher and skidder productivity. Application of two skidders resulted in high total operating cost. Extracting whole trees to roadside yielded a very small amount of remaining slash distributed on the site. Keywords: whole tree harvesting, feller-buncher, skidder, flail-debarker, cost, slash

1. Introduction – Uvod The most common option in the production of woody biomass is chipping in the forest at roadside followed by transportation of the chips (Stampfer and Kanzian 2006). In Denmark in-field chipping is often used in thinning and small diameter tree harvesting (Talbot and Suadicani 2005). About 75–80 % of the annual woody biomass production in Sweden is produced in this way (Ranta and Rinne 2006, Junginger et al. 2005). Roadside chipping is a common harvesting method in Australian eucalypt plantations. It utilizes a mobile chipper to produce export grade pulp chips at the plantation. If the fundamental objective of logistical Croat. j. for. eng. 34(2013)2

efficiency is to handle the largest piece size the least number of times, roadside chipping must be considered as preferential to any other method. Chips production at the roadside in Australia can be performed either by debarking the stems at the stump using a single-grip harvester, or alternatively, by debarking the stems with a chain flail delimber and debarker at the forest road prior to chipping (Lambert 2006). The system of roadside chipping with debarking at the stump was developed by Eumeralla Pty Ltd and AFM Pacific in Australia in 1998, for Timbercorp limited. This system uses single grip harvesters to fell, delimb and debark full tree lengths at the stump and position them for subsequent extraction. From this point, a purpose built tree-length forwarder extracts

189

M. R. Ghaffariyan et al. Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199)

Table 1 Harvesting equipment for roadside chipping with Husky Precision Tablica 1. Oprema za pridobivanje drvne sječke strojevima Husky Precision Operator experience, years

Hourly machine cost, $

Pogonskih sati

Iskustvo rukovatelja, god.

Trošak strojnoga rada po satu, $

191

4 738

240.59

630C (S9)

184

3 811

0.3

278.84

Tigercat

630D (S10)

191

748

0.7

203.07

Husky Precision

FD 2300-4

309

3 993

345.68

Husky Precision

HTC 2366

441

8 624

2.5

383.15

Machine type

Make

Model

Power, kW

Hours used

Tip stroja

Proizvođač

Model

Snaga, kW

Tigercat

845C (shear head: Tigercat 2001)

Tigercat

Tracked swing-to-tree feller-buncher Gusjenični feler bančer Rubber tired grapple skidder Kotačni skider s kliještima Rubber tired grapple skidder Kotačni skider s kliještima Flail Procesor za kresanje grana i koranje Chipper Iverač

the stems to the forest road for stockpiling. Finally, the full-length, debarked trees are chipped using a chipper at the roadside. The method of roadside chipping with debarking at the forest road is currently being used in the Green Triangle Region, Albany and Bunbury in Australia. In this system, trees are felled and bunched using a driveto-tree feller-buncher. The felling can also be commonly carried out by a boom-mounted swing-to-tree feller-buncher, which has the ability to process multiple rows at a time and can place the bunches in the out-row with less machine movement. The fellerbuncher head can be installed on a rubber-tired or a tracked based machine. At the roadside, trees are delimbed and debarked using a chain flail delimber/ debarker and then chipped in the trailer. The delimber/debarker can be integrated with the chipper, such as the Peterson Pacific DDC5000 (DDC), or separate from the chipper, such as the combination of the Husky Precision Flail and Chipper (F/C). A number of different variations of these machines have been tested over the years (Lambert 2006). Two recent studies on roadside chipping operations in Western Australia reported a productivity of 33.90 GMt/PMH0 for the Peterson Pacific chipper (Wiedemann and Ghaffariyan 2010) and 51.70 GMt/PMH0 for Husky precision chipper (Ghaffariyan et al. 2011). Both studies indicated that the major operational delay was the waiting time for trucks. This delay may be reduced through improved truck scheduling. The Husky Precision chipper study (Ghaffariyan et al.

190

2011) was about chipping small trees for biomass usage and no flail was used to debark the trees. The current study investigated the chipper and flail to produce pulp export chip, which is a common system in Western Australia. To add to the body of knowledge about the productivity of this harvesting method in Australia, this study aimed to investigate the efficiency of a road-side chipping system using a Husky Precision chipper. The objectives of this study were to: Þ Measure productivity of each machine of the system, Þ Estimate the cost of each machine and of the whole system, Þ Study impact of different parameters on productivity, Þ Measure fuel consumption of each machine and of the whole system, Þ Measure harvesting residues retained on the site after logging operation, Þ Assess the quality of chips produced.

2. Materials and Methods – Materijal i metode 2.1 Study area – Mjesto istraživanja The study area was located in a Eucalyptus globulus (Blue gum) plantation in southwest Western Australia, 58 km from the delivery point for all the products, the Croat. j. for. eng. 34(2013)2

Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199) M. R. Ghaffariyan et al.

Albany Plantation Export Company (APEC) chip mill. The study area was about 1.45 ha of flat terrain. The diameter at breast height over bark (DBHOB) and total tree volume averaged at 17.8 cm and 0.21 m3. The stocking was 711 stems per ha. Table 1 describes the machine used for the harvesting system. The trees were felled, bunched and skidded to the roadside as whole trees, then processed into

pulp chips, and loaded directly into trucks for transport. The whole trees were processed by a Husky flail. The trees were delimbed and debarked using the flail. Then the debarked wood was fed into the chipper. The trucks used in this study were pocket road train type with the loading capacity of 60 tonnes. The chipping residues were returned to the site as »beehives« using the grapple skidders.

Table 2 Work elements for the feller-buncher, skidder and truck (Acuna and Heidersdorf 2008) Tablica 2. Definicije radnih elemenata feler bančera, skidera i kamiona (Acuna i Heidersdorf 2008) Machine

Work elements

Definition

Stroj

Radni elementi

Definicija

Positioning Zauzimanje položaja

Any time spent for the movement of machine to place to start felling – Svako vrijeme utrošeno za pomicanje stroja na mjesto početka sječe

Felling-bunching Sječa i uhrpavanje

Starts when felling head is attached to tree to start cutting. It finishes when operator lays the felled tree on the ground – Počinje kada sječna glava obuhvati stablo i počinje sjeći. Završava kada rukovatelj položi posječeno stablo na tlo

Traveling Premještanje

Begins when the machine starts to travel to next tree and ends when the machine stops moving to perform some other activity – Počinje kada se stroj krene premještati do sljedećega stabla, a završava kada se stroj prestane kretati i započinje obavljati neku drugu aktivnost

Clearing Raščišćavanje

Starts when the machine stops moving or felling/bunching to dispose of non-merchantable material and stops when feller/bunching or moving recommences – Počinje kada se stroj prestane kretati ili sjeći i uhrpavati radi raščišćavanja nekomercijalnoga drvnoga amaterijala, a prestaje kada se sječa i uhrpavanje ili kretanje stroja nastavi

Clear debris Uklanjanje ostatka

Any time spent for clearing debris and removal to stockpile or return to the block – Svako vrijeme utrošeno za uklanjanje ostatka nakon iveranja i njegovo uhrpavanje ili vraćanje u sječinu

Travel empty Vožnja praznoga

Starts when machine commences travel into block and ends when loading of bunch commences – Počinje kada stroj započinje vožnju u sječinu, a završava kada počinje utovarivati

Loading Utovar

Starts with grappling the bunch and picking up and ends when travel loaded commences – Počinje sa zahvaćanjem i podizanjem tovara, a završava s početkom vožnje opterećenoga skidera

Travel loaded Vožnja punoga

Starts when wheels commence turning after loading, and ends when skid distance to the landing is reached Počinje kada se kotači skidera počinju okretati nakon utovara, a završava kada se prevali udaljenost privlačenja do pomoćnoga stovarišta

Unloading Istovar

Time to drop load and turn around to commence travel empty. Starts when skid distance to deck is reached and ends when turn around is completed – Vrijeme potrebno za istovar tovara i okretanje prije početka vožnje praznoga skidera. Počinje kada se prevali udaljenost privlačenja do mjesta istovara, a završava s okretanjem

Loading Utovar

Begins when chipper starts blowing the chips into truck and ends when truck starts travelling loaded – Počinje kada iverač započne upuhivati drvnu sječku u kamion, a završava kada puni kamion započinje vožnju

Travel loaded Vožnja punoga

Starts when loading finishes and truck starts travelling loaded to the mill and ends when unloading starts Započinje sa završetkom utovara i početkom vožnje punoga kamiona u tvornicu, a završava s početkom istovara

Unloading Istovar

Starts when travel loaded ends at the mills and ends after being fully unloaded at the time of starting travelling empty – Započinje sa završetkom vožnje punoga kamiona u tvornici, a završava nakon potpunoga istovara, u trenutku početka vožnje neopterećenoga kamiona

Travel empty Vožnja praznoga

Starts when truck driver commences to travel at the end of unloading element. It ends when loading starts Počinje kada vozač kamiona započinje vožnju na kraju istovara. Završava s početkom utovara

Feller-buncher Feler bančer

Grapple skidder Skider s kliještima

Truck Kamion

Croat. j. for. eng. 34(2013)2

191

M. R. Ghaffariyan et al. Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199)

Table 3 Productivity, cost and fuel consumption of roadside chipping with Husky Precision Tablica 3. Proizvodnost, trošak i utrošak goriva pri iveranju na pomoćnom stovarištu strojevima Husky Precision Machine

Productivity, GMt/PMH0

Cost, $/GMt

Fuel consumption, l/hr

Fuel consumption, l/GMt

Stroj

Proizvodnost, GMt/PMH0

Trošak, $/GMt

Utrošak goriva, l/h

Utrošak goriva, l/GMt

97.26

2.55

32.09

0.33

60.22

12.02

91.91

1.58

57.80

5.98

44.51

0.77

58.18

6.59

72.14

1.24

14.96

4.19

–

31.33

–

3.92

Feller-buncher Feler bančer Grapple skidder (two skidders) Skider s kliještima (dva skidera) Husky Precision flail Procesor Husky Precision Husky Precision chipper Iverač Husky Precision Truck Kamion Total Ukupno

2.2 Method – Metoda 2.2.1 Time study and modeling – Studij vremena i modeliranje The elemental time study method was used to evaluate machine productivity for the feller buncher, two grapple skidders and trucks. The felling-bunching and skidding working cycles were divided into the specific elements described in Table 2. Personal, mechanical and operational delays were also recorded during the time study. Productivity was calculated by the delivered tonnes of chips (GMt) and productive machine hours, excluding all delays (PMH0). Backward stepwise regression was applied to develop the productivity predicting equations in SPSS 18. If any variable had significant impact on the residual mean square of the models, it was included in the models. The analysis of variance of each model was used to verify the significance of the model. The models were validated using witness samples, and the confidence intervals for each coefficient were calculated. By recording the total working time and delivered volume, the productivity of the flail, chipper and trucks were estimated. 2.2.2 Harvesting costs – Troškovi pridobivanja The hourly machine cost included fixed, variable and labor costs. The hourly machine cost for each harvesting machine was modeled using the ALPACA (Australian Logging Productivity And Cost Apprais-

192

al) calculator, developed by Murphy and Acuna (2009). Unit cost was determined by dividing hourly machine cost by the net machine productivity. 2.2.3 Yield and chip quality – Količina i kakvoća drvne sječke The yield was based on weighbridge data of the chips delivered to the mill. Using 8 samples of about 2 kg each, the moisture content of the chips was estimated to calculate the yield in bone dry metric tonnes (BDMt). The samples were tested for their size classification and bark content according to the APEC export chip specifications. 2.2.4 Assessment of harvest residues – Procjena količine drvnoga ostatka There were two types of harvesting residues in this study; scattered residues left at the stump site and flail residues piled at roadside. The amount of stump site residues was estimated using two lines transects 20 m apart, along which 4 square plots of 1x1 m were established every 20 m. All the slash on each sample plot was collected manually and weighed with a portable scale. Roadside residues were taken back to the field with the skidder and stacked into piles, also called »beehives«. The »beehives« were evenly distributed over the site. The bulk volume of 6 samples of »beehives« was determined by measuring the length, width, height and cross-sectional shape of each pile. The total number of the »beehives« was about 66. By Croat. j. for. eng. 34(2013)2

Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199) M. R. Ghaffariyan et al.

multiplying the average volume to the number of »beehives«, the total volume was estimated. No information on bulk density was available to convert the volume of »beehives« to weight.

3. Results – Rezultati 3.1 Productivity, cost and fuel consumption Proizvodnost, trošak i utrošak goriva Table 3 shows the measured productivity, cost and fuel consumption for each machine engaged in the test operation. Skidding had the highest cost, and incurred the highest fuel consumption per GMt. The main reason for using two skidders was to avoid waiting time

for the chipper while extracting the trees and clearing debris, which might take long time when using one skidder in the operation.

3.2 Feller-buncher productivity model – Model za izračun proizvodnosti feler bančera Tree size significantly impacted the productivity of the feller-buncher. Increasing tree size resulted in higher productivity (Fig. 1). The model is significant at α = 0.05 (Table 4). The model is: Productivity (GMt/PHM0) = 182.078 + 57.585 × ln (Tree size (m-3)) R2 = 40.2%, n = 80 Table 5 summarizes the percent incidence of each work step on total time consumption, for the Tigercat feller-buncher. Felling and bunching accounted for over 95% of work time. No delay occurred for the duration of our time study.

3.3 Skidder productivity model – Model za izračun proizvodnosti skidera

Fig. 1 Impact of tree size on feller-buncher productivity Slika 1. Utjecaj obujma stabla na proizvodnost feler bančera

Tree size did not have any significant impact on skidder productivity and therefore it was excluded from the model. Skidding distance significantly affected the productivity of both skidders (Fig. 2 and 3). From the ANOVA tables, both models were significant at α = 0.05 (Tables 6 and 7). The model for the skidder TC 630C had a higher coefficient of determination compared to the model for the skidder TC 630D, and it could explain about 49% of the total variability observed for skidder productivity. The average productivity for the TC 630C skidder was about 28.53 GMt/ PMH0 which was lower than for the TC 630D skidder, with 31.69 GMt/PMH0 although the skidder 630D covered a longer mean skidding distance (256 m vs. 190 m) (Table 8).

Table 4 Analysis of variance of productivity model for feller-buncher Tablica 4. Analiza varijance modela za izračun proizvodnosti feler bančera Sum of Squares

Mean Square

Sig.

Suma kvadrata

Stupnjevi slobode

Varijanca

F-vrijednost

Statistička značajnost

8 502.33

52.35

0.00

12 668.22

162.41

–

21 170.55

–

Regression Regresijski model Residual Rezidual Total Ukupno

Croat. j. for. eng. 34(2013)2

193

M. R. Ghaffariyan et al. Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199)

3.3.1 Productivity model for Skidder TC 630C Model za izračun proizvodnosti skidera TC 630C

3.3.2 Productivity model for Skidder TC 630D Model za izračun proizvodnosti skidera TC 630D

Productivity (GMt/PHM0) = 34.559 – 0.032 × Skidding distance (m)

Productivity (GMt/PHM0) = 37.214 – 0.020 × Skidding distance (m)

R2 = 49.0%, n = 10

R2 = 38.9%, n = 11

Fig. 2 Impact of skidding distance on the productivity of skidder TC 630C Slika 2. Utjecaj udaljenosti privlačenja na proizvodnost skidera TC 630C

Fig. 3 Impact of skidding distance on the productivity of skidder TC 630D Slika 3. Utjecaj udaljenosti privlačenja na proizvodnost skidera TC 630D

Table 5 Work element breakdown for the feller-buncher Tablica 5. Raščlamba radnih elemenata feler bančera

Share, % – Udio, %

Positioning

Felling & bunching

Travel

Clearing

Delay

Zauzimanje položaja

Sječa i uhrpavanje

Premještanje stroja

Raščišćavanje

Prekid rada

0.3

95.5

4.0

0.2

0.0

Table 6 Analysis of variance of productivity model for skidder TC 630C Tablica 6. Analiza varijance modela za izračun proizvodnosti skidera TC 630C Sum of Squares

Mean Square

Sig.

Suma kvadrata

Stupnjevi slobode

Varijanca

F-vrijednost

Statistička značajnost

163.17

7.68

0.024

Residual – Rezidual

170.01

21.25

–

Total – Ukupno

333.18

–

Regression Regresijski model

194

Croat. j. for. eng. 34(2013)2

Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199) M. R. Ghaffariyan et al.

From Fig. 3, the longer the skidding distance the lower the productivity, due to the increased travel time. The percent incidence of each work element on the skidding cycle for the two skidders is presented in Table 9. Nearly half of the work time was spent for clearing debris. The lowest percentage was for unloading, which accounted for less than 2% of the total skidding time. The delays were mainly due to waiting for the chipper to unload the bunches in front of the chipper to be accessible for the chipper grapple due to lack of free space (operational delays). The incidence of delays was 10 percentage points higher for the skidder 630D than for the skidder 630C. 3.3.3 Husky Precision flail and chipper – Procesor i iverač Husky Precision The flail worked for 243 minutes, reaching the average productivity of 57.80 GMt/PMH0. Debarking accounted for about 92% of total work time. Delays included waiting for wood (4.5% of total work time), warm up (1.6% of total work time) and waiting for chipper as the chipper was waiting for truck (2.0% of total work time). The chipper discharged directly into the trucks. Four trucks were used to transport the chips to the APEC mill. The average delay-free chipping time per truck was about 56 minutes. Net productivity averaged 58.18 GMt/PMH0. Effective chipping time accounted for 93 % of total work time. Delays were represented by waiting for wood (4.7% of total work time), waiting for trucks (0.2%) and warm up (2.0%). 3.3.4 Transportation – Daljinski transport The transport distance from study area to the APEC mill gate was 58 km. Mean net productivity and the payload was 14.96 GMt/PMH0 and 54 GMt, respectively. The average delay-free cycle time for transportation was about 4.58 hours. Elemental time breakdown for transportation is shown in Table 10. Traveling loaded had the highest incidence on total cycle time (28 %). Delays consisted almost exclusively of waiting. 3.3.5 Yield and chip quality – Količina i kakvoća drvne sječke The study area (1.45 ha) yielded 232 GMt of pulp chips, corresponding to 160 GMt/ha. Based on moisture content sampling of 43%, the actual yield in dry mass was equal to 90 BDMt/ha (Mitchell and Wiedemann 2012). The chip sample analysis showed that bark content was 0.18%, well within the limits set by APEC specifications (<0.5 %). Table 11 shows that 68% of the chip mass consisted of particles measuring between 9.5 mm and 22.2 mm (Mitchell and Wiedemann 2012). Croat. j. for. eng. 34(2013)2

3.3.6 Harvest residues assessment – Procjena količine drvnoga ostatka Scattered stump site residues accounted for 6.4 GMt/ha. In contrast, flail residues returned to the field and stacked as »beehives« represented 262 m3.

4. Discussion – Rasprava The productivity of the feller-buncher in this case study is lower than the average productivity (138.0 GMt/PMH0) reported for a similar Valmet 445 EXL tracked self-leveling feller-buncher working in the pine plantations of the South Gippsland coast of Victoria (Acuna et al. 2011). It is also lower than the 122.2 GMt/PMH0 reported for the clear fell of pine plantation in Southern Tasmania (Ghaffariyan et al. 2012). The main reason for that is likely to consist in the smaller tree size handled in this study. The fuel consumption per cubic meter is also lower than the consumption reported for a large feller-buncher (0.36 l/GMt) by Johnson et al. 2006. It is also slightly lower than the consumption of 0.34 l/GMt reported by Ghaffariyan et al. 2012 for Southern Tasmania, which is consistent with the lower productivity. The close relationship between feller-buncher productivity and tree size in eucalypt clearfell operations is supported by the results obtained in Brazil by Moreira et al. (2004), who reported a productivity of 33.5 and 36.1 GMt/PMH0 for an average DBH of 9.0 and 10.4 cm, respectively. Similar results are also reported by Spinelli et al. (2009) who studied a range of feller-bunchers used for eucalypt clearfell and obtained figures between 14 and 20 GMt/PMH0 for smaller DBH and steeper slopes than covered by this study. The average productivity of both skidders in this study is lower than the productivity (44.6 GMt/PMH0) of a similar TC 730C grapple skidder used for extracting small whole eucalypt trees in Western Australia (Ghaffariyan et al. 2011). Productivity rates in this study are also lower than the 47.5 GMt/PMH0 reported for whole eucalypt tree skidding in Brazil (Valverde et al.1996). This could be the result of the longer skidding distance, smaller payload and residue clearing in our case study. The productivity models estimated by Dodson et al. (2006) for two Caterpillar rubber-tired grapple skidders working in western juniper stands included three independent variables, namely: skidding distance, number of stems per turn and a dummy variable for stand type (mixed or not-mixed). Our skidding productivity models contain the skidding distance as a significant variable affecting the skidder productivity.

195

M. R. Ghaffariyan et al. Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199)

Table 7 Analysis of variance of productivity model for skidder TC 630D Tablica 7. Analiza varijance modela za izračun proizvodnosti skidera TC 630D Sum of Squares

Mean Square

Sig.

Suma kvadrata

Stupnjevi slobode

Varijanca

F-vrijednost

Statistička značajnost

125.02

5.72

0.04

196.59

21.84

–

321.61

–

Regression Regresijski model Residual Rezidual Total Ukupno

Table 8 Descriptive statistics of productivity model – Skidder TC 630C and TC 630D Tablica 8. Opisna statistika modela za izračun proizvodnosti skidera TC 630C i skidera TC 630D Skidder type

Minimum

Maximum

Mean

Tip skidera

Najmanja vrijednost

Najveća vrijednost

Aritmetička sredina

Skidding distance, m

TC 630C

60.00

510.00

190.00

Udaljenost privlačenja, m

TC 630D

20.00

555.00

256.04

Tree size, m3

TC 630C

0.14

0.21

0.17

Obujam stabla,m

TC 630D

0.15

0.21

0.17

Productivity, GMt/PMH0

TC 630C

18.50

39.00

28.53

Proizvodnost, GMt/PMH0

TC 630D

22.10

40.30

31.69

The productivity rates of both skidders in our case study are lower than reported productivity of 53.8 GMt/PMH0 for a Caterpillar grapple skidder 525C in clear felling operations in Eucalypt stands with the average tree size of 0.178 m3 and average skidding distance of 160 m (Wiedemann and Ghaffariyan 2010). The skidding distance was longer in our case study, which resulted in lower productivity. The average fuel consumption of the two skidders in this study (0.79 l/GMt) is higher than the fuel consumption reported by Makkonen (2004) for a grapple skidder used in Canada. However, it is also lower than reported for large clam bunk skidders (1.17 l/GMt) used in USA (Johnson et al. 2006).

Flail and chipper were two separate machines operated by two operators at the road side in this study. The chipper net productivity (58.18 GMt/PMH0) is slightly lower than recorded for the Morbark chipper working at roadside (59.4 GMt/PMH0) to chip logs from first thinning in Pine plantation of South Australia (Ghaffariyan 2012). Tree size and machine power in this study were higher than for the Morbark chipper trial, which should have resulted in higher productivity, based on the findings of Spinelli and Hartsough (2001). They found a direct relationship between chipper productivity, piece size and engine power. The lower chipping productivity in this study is likely due to the smaller tree bunches delivered to the chipper as

Table 9 Percent incidence of each work element on the total duration of the skidding cycle Tablica 9. Postotni udio pojedinoga radnoga elementa u ukupnom trajanju turnusa privlačenja Skidder type Tip skidera

Clear debris

Travel empty

Čišćenje ostatka Vožnja praznoga

Load

Travel loaded

Unload

Delay

Utovar

Vožnja punoga

Istovar

Prekid rada

Share, %

TC 630C

Udio, %

TC 630D

196

Croat. j. for. eng. 34(2013)2

Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199) M. R. Ghaffariyan et al.

Table 10 Percent incidence of work steps on total transportation cycle Tablica 10. Postotni udio pojedinoga radnoga elementa u ukupnom trajanju turnusa daljinskoga transporta

Share, % – Udio, %

Travel loaded

Unloading

Travel empty

Delay

Utovar

Vožnja punoga

Istovar

Vožnja praznoga

Prekid rada

Table 11 Particle size distribution of chip samples Tablica 11. Granulometrijska struktura uzoraka drvne sječke Size class Razred

>28.6 mm

>22.2 mm

>9.5 mm

>4.8 mm

<4.8 mm

3.36

16.31

68.24

9.88

2.03

Share, % – Udio, %

Bark Kora 0.18

a result of the hot-decking operation, where chipping/ loading occurred at the time of wood extraction to the road side. In contrast, the Morbark chipper worked trees decked in large piles (average height and length of the piles were 4 m and 66 m, respectively), allowing for relatively large bunches of wood to be fed into the chipper. Another factor may be the impact of whole tree chipping (in our case study delimbed stems from whole trees by flail) versus log chipping (Spinelli and Magagnotti 2010). The productivity recorded in this study is also higher than reported for a Peterson Pacific chipper tested in whole tree chipping for biomass (33.90 GMt/PMH0) in Western Australia, due to the smaller tree size of 0.10 m3 in the latter study (Ghaffariyan et al. 2011). In our case study area only four trucks were loaded, and chipping and trucking were characterized by a very small sample size. The amount of scattered stump-site residues (6.45 GMt/ha) was much lower than reported for sites harvested by the cut-to-length system. According to Smethurst and Nambiar (1990) stump-site residues amounted to 52 GMt/h in a clearfelled Pinus radiata plantation in Mount Gambier, South Australia. Similarly, Ghaffariyan and Andorovski (2011) report as much as 70.4 GMt/ha for the stump-site residues left after the cut-to-length clearfell harvesting of a Eucalyptus nitens plantation in Northern Tasmania. In our case study, it is important to determine whether the »beehives« are better spread over the whole site or if the flail residues could rather be refined and used as boiler fuel.

increased total operating costs. Future studies could compare the use of two skidders with the use of one skidder only. Long skidding distance, small payload and spending time for clearing debris resulted in low productivity of the skidders in this case study. According to the results, the skidding distance had significant impact upon the productivity of two skidders. Based on the productivity predicting models, the larger the tree volume the higher the feller-buncher productivity. As two separate machines were used for debarking (Husky flail) and chipping (Husky chipper), the future studies could also explore the efficiency of integrated delimber-debarker-chipper units, where the flail and chipper are combined into one machine, as an initial trial has indicated that using separate flail and chipper can result in higher total harvesting cost than using an integrated delimber-debarker-chipper (Ghaffariyan and Sessions 2012). Roadside chipping operation left a small amount of residues in the stand, being based on whole tree-extraction. The possible impacts of intense slash removal on site fertility could also be studied in the future.

5. Conclusions – Zaključci

6. References – Literatura

Based on these results, the inclusion of more machines will result in higher cost of operation and higher fuel consumption. In this case study, using two skidders

Acuna, M., Heidersdorf, E., 2008: Harvesting machine evaluation framework for Australia. CRC for Forestry, Draft Technical Report, 33 p.

Croat. j. for. eng. 34(2013)2

Acknowledgement – Zahvala This is to acknowledge the following researchers for their assistance in data collection: Rick Mitchell and John Wiedemann. The authors would like to thank the journal reviewers who provided valuable comments that helped improve this article.

197

M. R. Ghaffariyan et al. Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199) Acuna, M., Skinnell, J., Mitchell, R., Evanson, T., 2011: Bunching stems in steep slopes for efficient yarder extraction. CRC for Forestry Bulletin 17, 3 p.

Mitchell, R., Wiedemann, J., 2012: Volume recovery comparison for four different harvesting systems in short-rotation bluegum plantations. CRC for Forestry. Bulletin 27, 5 p.

Dodson, E. M., Deboodt, T., Hudspeth, G., 2006: Production, cost, and soil compaction estimates for two Western Juniper extraction systems. Western Journal of Applied Forestry 21(4): 185–194.

Moreira, F.M.T., de Souza, A. P., Machado, C. C., Minetti, L. J., Silva, K. R., 2004: Technical and economic analysis of a feller-buncher in two harvest subsystem of Eucalyptus forests. Revista Árvore 28: 199–205.

Ghaffariyan, M. R., Andorovski, V., 2011: Bundling harvest residues in shining plantations. CRC for Forestry, Bulletin 15, 3 p. (and Forest Energy observer 2011 online at: http://journal. forestenergy.org)

Ranta, T., Rinne, S., 2006: The profitability of transporting uncomminuted raw materials in Finland. Biomass and Bioenergy 30(3): 231–237.

Ghaffariyan, M. R., Brown, M., Acuna, M., Sessions, J., Kuehmaier, M., Wiedemann, J., 2011: Biomass harvesting in Eucalyptus plantations in Western Australia. Southern Forests 73(3–4): 149–154. Ghaffariyan, M. R., Sessions, J., 2012: Comparing the efficiency of four harvesting methods in a blue gum plantation in southwest Western Australia. CRC for Forestry, Bulletin 29, 4 p. Ghaffariyan, M. R., Sessions, J., Brown, M., 2012: Machine productivity, volume recovery and harvesting residues of a cut-to-length harvest system in Southern Tasmania. Southern Forests: a Journal of Forest Science 74(4): 229–235. Johnson, L. R., Lippke, B., Marshall, J. D., Comnick, J., 2006: Life-cycle impacts of forest resource activities in the Pacific Northwest and Southeast United States. Wood and Fiber Science, 37 Corrim Special Issue, 2005: 30–46. Junginger, M., Faaij, A., Bjorheden, R., Turkenburg, W. C., 2005: Technological learning and cost reductions in wood fuel supply chains in Sweden. Biomass and Bioenergy 29(6): 399–418. Murphy, G., Acuna, M., 2009: Australian logging productivity and cost appraisal model (ALPACA). Internal toolbox, CRC for Forestry, Hobart, Australia. Lambert, J., 2006: Growth in blue gum forest harvesting and haulage requirements in the Green Triangle 2007–2020. CRC for Forestry Consultant report, 119 p. Makkonen, I., 2004: Saving fuel in mechanized forestry operations. Forest Engineering Institute of Canada, PointeClaire, QC. Internal Report IR-2004-08, 10 p.

Spinelli, R., Hartsough, B., 2001: A survey of Italian chipping operations. Biomass and Bioenergy 21: 433–444. Spinelli, R., Magagnotti, N., 2010: A tool for productivity and cost forecasting of decentralised wood chipping. Forest Policy and Economics 12: 194–198. Spinelli, R., Ward, S., Owende, P., 2009: A harvest and transport cost model for Eucalyptus spp. fast-growing short rotation plantations. Biomass and Bioenergy 33: 1265–1270. Stampfer, K., Kanzian, Ch., 2006: Current state and development possibilities of wood chip supply chains in Austria. Croatian Journal of Forest Engineering 27(2): 135–145. Smethurst, P. J., Nambiar, E. K. S., 1990: Distribution of carbon and nutrients and fluxes of mineral nitrogen after clearfelling a Pinus radiata plantation. Canadian journal of forest research 20: 1490–1497. Talbot, B., Suadicani, K., 2005: Analysis of two simulated infield chipping and extraction systems in spruce thinnings. Biosystems Engineering 91(3): 283–292. Valverde, S. R., Machado, C., Pereira de Rezende, J., Paulo de Souza, A., Antiqueira, A., 1996: A technical and economical analysis of timber skidding using a skidder in a full tree harvesting system. Viçosa, Brasil. Revista Arvore 20(1): 101–109. Wiedemann, J., Ghaffariyan, M. R., 2010: Preliminary results: volume recovery comparison of different harvesting systems in short-rotation hardwood plantations. CRC for Forestry, Bulletin 9, 4 p.

Sažetak

Ocjena učinkovitosti, kakvoće drvne sječke i količine drvnoga ostatka pri proizvodnji drvne sječke procesorom i iveračem u Zapadnoj Australiji Iveranje pokretnim iveračem na pomoćnom stovarištu uobičajen je sustav proizvodnje visokokvalitetne drvne sječke za celulozu u australskim šumskim plantažama. Istraživani sustav pridobivanja drvne sječke činili su feler bančer, dva skidera s kliještima za privlačenje uhrpane stablovine, procesor za kresanje grana i koranje, diskni iverač za usitnjavanje okorane deblovine i kamion za prijevoz proizvedene drvne sječke. Skideri su osim za privlačenje stablovine na pomoćno stovarište korišteni i za vraćanje drvnoga ostatka nastaloga pri proizvodnji drvne sječke u sječinu i njegovo uhrpavanje.

198

Croat. j. for. eng. 34(2013)2

Evaluating Efficiency, Chip Quality and Harvesting Residues of a Chipping Operation ... (189–199) M. R. Ghaffariyan et al.

Istraživali su se proizvodnost, troškovi i utrošak goriva pojedinih strojeva u sustavu te kakvoća drvne sječke i količina drvnoga ostatka nakon pridobivanja drvne sječke. Prosječna proizvodnost stroja za sječu i uhrpavanje iznosila je 97,26 GMt/PMH0, a prosječna proizvodnost skidera iznosila je 60,22 GMt/PMH0. Proizvodnost procesora i iverača iznosila je prosječno 57,80 GMt/PMH0, odnosno 58,18 GMt/PMH0. Prosječna proizvodnost daljinskoga transporta iznosila je 57,34 GMt/PMH0. Za konstrukciju modela za izračun proizvodnosti pojedinoga stroja u sustavu pridobivanja korišten je studij vremena i regresijske analize. Utvrđen je značajan utjecaj obujma stabla na proizvodnost stroja za sječu i uhrpavanje te udaljenosti privlačenja na proizvodnost skidera. Troškovi su procijenjeni primjenom modela ALPACA (Australian logging productivity and cost appraisal). Ovaj rad donosi važne spoznaje o utjecaju različitih čimbenika na proizvodnost feler bančera te na proizvodnost skidera. Primjenom dvaju skidera u sustavu pridobivanja, nužnih za održavanje proizvodnosti ostalih strojeva u sustavu, nastao je visoki ukupni trošak. Pridobivanje sirovine za proizvodnju drvne sječke stablovnom metodom dalo je vrlo malu količinu drvnoga ostatka preostaloga u sječini. Ključne riječi: pridobivanje drva stablovnom metodom, stroj za sječu i uhrpavanje, skider, procesor, trošak, drvni ostatak

Authors’ address – Adresa autorâ: Mohammad Reza Ghaffariyan, PhD.* e-mail: ghafari901@yahoo.com University of the Sunshine Coast Private Bag 12 7001 Hobart AUSTRALIA Prof. Mark Brown, PhD. e-mail: mbrown2@usc.edu.au University of the Sunshine Coast Locked Bag 4 4558 Maroochydore, Queensland AUSTRALIA

Received (Primljeno): January 22, 2013 Accepted (Prihvaćeno): February 11, 2013 Croat. j. for. eng. 34(2013)2

Raffaele Spinelli, PhD. e-mail: spinelli@ivalsa.cnr.it CNR IVALSA Via Madonna del Piano 10 50019 Sesto Fiorentino ITALY * Corresponding author – Glavni autor

199

Original scientific paper – Izvorni znanstveni rad

Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization and Breakeven Analysis Storm Beck, John Sessions Abstract – Nacrtak Non-conventional products provide opportunities for the forest industry to increase economic value from forests; however, these products may require transport by specialized vehicles. The existing forest transportation network was not necessarily designed to the road standards required for these specialized vehicles. Several road modifications can be made to give specialized vehicles access to the forest transportation network including filling the ditch, removing the superelevation, reversing the superelevation, or reconstructing the roadway. For each investment, there is an associated vehicle that can traverse the road segment if the investment is made. For scheduling multiple biomass operations over a road network, we use the Ant Colony heuristic to identify the combination of optimal vehicle choices and road modifications to effectively transport non-conventional products. These combinations related to a 27% reduction in total transportation costs. For isolated biomass operations, we use breakeven analysis to make the vehicle selection and road modification option. Decisions for isolated biomass operations depend on road modification cost, transport volume, and transport costs on forest and highway roads. Keywords: ant colony optimization, biomass transport, vehicle accessibility

1. Introduction – Uvod The production of high valued non-conventional products, such as utility poles or the production of low valued products such as chips or hogfuel, provide opportunities for the forest industry to increase economic value from forests. However, most of the forest transportation system has been designed and built for long-log, stinger-steered trailers (Sessions et al. 2010) and there is little engineering record of road design or location throughout the forest industry (Craven et al. 2011). This lack of engineering records provides a challenging environment in the assessment of transporting non-conventional products. The primary challenge to hauling non-conventional products, on specialized vehicles, is determining if the vehicle can navigate the horizontal and vertical geometry unloaded and loaded, as well as turning around near the landing. These specialized vehicles include truck tractors pulling pole Croat. j. for. eng. 34(2013)2

trailers with rear self-steering axles, pole trailers with stinger-steered axles, fifth-wheel chip trailers (with and without self-steering rear axles), and stingersteered chip trailers. We define a pole trailer as a stinger-steered trailer with a bunk-to-bunk distance longer than 8.5 m, hauling logs that are longer than 13.7 m. We focus on the economical assessment of varying sized chip trailers (chip vans) throughout the forest transportation network for the remainder of the paper, although the principles are the same for other specialized trailers.

2. Problem Description – Problematika Several choices can affect the accessibility of these specialized vehicles. These choices include temporarily filling the ditch, removing or reversing the superelevation to reduce lateral tire slip, and widening the roadway. During the dry months, temporarily filling

201

S. Beck and J. Sessions Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215)

the ditches or changing the superelevation of the roadway are options that permit specialized vehicles access. Temporarily filling the ditch provides a greater road width for the specialized vehicle to pass, usually 0.5 to 1.5 m of extra road width. Single lane forest roads surfaces are insloped, outsloped, or crowned. Positive superelevation of the road surface is often constructed into forest roads to counteract centrifugal force created by vehicles in curves (Oglesby and Hicks 1982). Negative superelevation of the road surface is sometimes constructed into curves to adjust the normal forces on the driving axles to permit climbing steeper grades (Anderson and Sessions 1991). Outsloping a forest road is sometimes used to drain water from the road surface without diverting water to ditches and insloping of forest roads is done for safety when roads are icy (Bowers 2006). During the dry months, superelevation may not be needed either because side friction is greater and/or cross slope drainage is not an issue; providing an opportunity to alter the road surface to reduce lateral tire slip toward the inside of a curve. Two options exist when altering the superelevation (1) remove the superelevation and (2) reverse the superelevation. Removing the superelevation reduces the amount of off-tracking that a vehicle produces by reducing the amount of lateral tire slip due to gravity (Glauz and Harwood 1991). Reversing the superelevation could be used to counteract offtracking; allowing the weight of the vehicle and the effects of gravity on an inclined plane to counter the effects of off-tracking. Lastly, forest engineers and

Fig. 1 A 13.7 m drop center 5th wheel chip van being loaded on a forest road in Lane County, Oregon Slika 1. Poluprikolica za šumsku sječku dugačka 13,7 m (utovar na sredini prikolice) tijekom utovara na šumskoj cesti u Lane County, Oregon

202

Fig. 2 A stinger-steered chip van. Photo courtesy of Western Trailer Company Slika 2. Samokretna prikolica za šumsku sječku. Slika dobivena od Western Trailer Company managers can affect the outcome by redesigning the roadway to allow these vehicles access along the entire roadway length. This is achieved by widening the roadway and removing obstacles close to the roadway such as standing trees. Each modification option has an associated cost and benefit. For example, if a 13.7 m drop center 5th wheel chip van (Fig. 1) needs an extra half meter of road width to access a harvest unit, the ditches might be temporarily filled to allow the 5th wheel chip van access. If the ditches were not filled, the only vehicle that might have access to the unit would be a stingersteered chip trailer (Fig. 2). Not only does the amount of off-tracking vary between vehicles, so does the volume of chips or hogfuel consistent with weight restrictions that these vehicles can haul. The operating cost and traveling speed vary for each vehicle configuration, creating a multi-dimensional problem. We look at two cases. The first case involves scheduling multiple biomass operations over a road network, where trucks from several biomass operations can take advantage of the same road investment. The second case looks at isolated biomass operations, where the road investment is used by only one operation. For both cases, mixed integer linear programming can be used to exactly solve the underlying mathematical problem. However, for the second case it is more convenient to use a breakeven analysis. For larger problems of the first case, due to the solution time for mixed integer programming, heuristics such as Ant Colony Optimization (ACO) can be used to determine a high quality solution for vehicle type, path, and road modifications for transporting biomass. Other useful heuristics are described by Glover and Kochenberger (2002), Hoos and Stutzle (2005) and Geem (2009). Croat. j. for. eng. 34(2013)2

Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215) S. Beck and J. Sessions

3. Mathematical Formulation – Case One Matematički prikaz – slučaj prvi The mathematical problem is to minimize the sum of road modifications and biomass transportation costs. Let G = (N, A) be a directed network with nodes N and arcs (i,j) within A. We associate with each node i within N a number S(i), which indicates the supply or demand depending on whether S(i) > 0 or S(i) < 0. The minimal cost problem is then: Minimize

∑ FC

(i,j) ∈ A +

∑ ∑VC

t ij

t t ij ×Yij

∑ ∑VC

(i,j) ∈ A

t ij

× Volumeijt

t∈ T

∀ ( i , j ) ∈ A , t ∈T

× Volumeijt

(1)

T (i,j) ∈ A t ∈ Conservation of Flow

∑

Volumeijt −

{ ji(i,j) ∈ A}

∑

Volume tji = V t ( i )

{ ji( j,i)∈ A}

∀ i ∈N

(2)

Sale Volumes V t (i ) = S (i )

∑

∀ i ∈N

(3)

t∈ T

Road Triggers ≥ Volumeij1

∀ (i , j ) ∈ A

(4)

∑

M × Yijt ≥ Volumeij2

∀ (i , j ) ∈ A

(5)

∑

M × Yijt ≥ Volumeij3

∀ (i , j ) ∈ A

(6)

∑M × Y

t ij

t∈ T

t ∈ T( t ≥ 2 )

t ∈ T ( t = 3)

Decision Variables Yijt = {0,1} Volumeijt ≥ 0

∀ ( i , j ) ∈ A , t ∈T ∀ ( i , j ) ∈ A , t ∈T

4. Review of Ant Colony Optimization Opis optimizacije metodom mravlje kolonije The ACO (Dorigo and Stutzle 2004) is based on the analogy of ants searching for food. Ants randomly walk in search of food leaving a pheromone behind as they travel. The pheromone is a scent that influences other ants to take that path. As more ants travel over the same path the pheromone increases, increasing the possibility of an ant choosing that path. This process continues until all ants are following the same path to the food source. The ACO heuristic has been used to solve fixed cost and variable cost forest transportation problems with side constraints (Contreras et al. 2008). Outside of the forest industry, this heuristic has been used to solve vehicle route scheduling problems, capacitated vehicle routing problems, and other scheduling problems (Donati et al. 2008, Rizzoli et al. 2007).

(7) (8)

Equation (1) is the objective function. FCijt is the fixed cost to modify link ij to allow truck type t access. Yijt is a binary variable, zero if the link is not used, and one if the link is used. VCijt is the round trip variable

cost over link ij in truck type t, ($/tonne). Volumeijt is the amount of volume crossing link ij in truck type t, (tonnes). Equation (2) provides conservation of flow at each node for each truck type. V t ( i ) is the volume entering each node i for each truck type t, (tonnes). Equation (3) requires that the total supply or demand Croat. j. for. eng. 34(2013)2

at each node S ( i ) (tonnes), equal the sum of the volume transported over all truck types. Equation (4) requires that the road modification for truck type 1 (the lowest standard truck type) be made to at least pass truck type 1, if there is volume passing over link ij in truck type 1. Equation (5) requires that the road modification for truck type 2 (the moderate standard truck type) be made to at least pass truck type 2, if there is volume passing over link ij in truck type 2. Equation (6) requires that the road modification for truck type 3 (the highest standard truck type) be made to pass truck type 3, if there is volume passing over link ij in truck type 3. Equation (7) requires that the road trigger for link ij for truck type t be a binary variable, zero or one. Equation (8) requires that the volume passing over link ij for truck type t be equal to or greater than zero.

5. Ant Colony Optimization – Optimizacija metodom mravlje kolonije The ACO developed in this paper is designed to minimize the total transportation cost. The total transportation cost is the sum of the modifications costs plus the round trip variable costs multiplied by the volume of each harvest unit. If a truck is loaded at sale x, it must make it to destination z using the same truck. If different types of trucks use the same link, the one with the maximum fixed cost will be applied. Therefore, if road modifications are applied so that a 16.2 m drop center 5th wheel chip van (Fig. 3) can navigate the road, no other modifications need to take

203

S. Beck and J. Sessions Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215)

Fig. 3 A 16.2 m drop center 5th wheel chip van near Port Angeles, Washington Slika 3. Poluprikolica za šumsku sječku dugačaka 16,2 m s utovarom na sredini u blizini Port Angeles, Washington place for other truck types. The ACO regards each road modification option as a separate link. In other words, between each node, three links exist; one that has no fixed cost, one that has a moderate fixed cost, and one that has a large fixed cost; all of which end up at the same node (Fig. 4). As the algorithm progresses through each set of ants, each ant in each set has a designated modification option that it will choose from as it progresses through the network. It was chosen to have three kinds of ants; a truck type 1 ant, a truck type 2 ant, and a truck type 3 ant to diversify the search. With this formulation, each modification option has its own set of pheromones. The starting pheromones provided an equal probability choosing each link leaving a node for each truck type. As the algorithm identifies a lower total cost route from each sale, the links that are not part of that path

have their pheromones decay. We used a constant decay factor of 25 %. The ACO was compared to a mixed integer linear programming model, using a small network (Fig. 4). The large black circles are the nodes in the network. The small black circles are the road modification option for the 16.2 m drop center 5th wheel chip van, the small horizontally hatched circles are the road modification option for the 13.7 m drop center 5th wheel chip van, and the small white circles are the no road modification option for the stinger-steered chip van. In this formulation, three different degrees of road modification could be applied, no modification, moderate modification, or major modification. The no modification option will only allow a stinger-steered chip van access. The moderate modification option will allow a stinger-steered chip van and a 13.7 m drop center 5th wheel chip van access. The major modification will allow all three trucks access to the road segment. Each truck has a different hourly operating cost. The stinger-steered chip van has an estimated hourly cost of $ 95.37, the 13.7 m drop center 5th wheel chip van hourly cost is $ 90.95, and the 16.2 m drop center 5th wheel chip van hourly cost is $ 99.79 (Table 1). We assumed cost per hour is the weighted average hourly cost and did not vary with speed or road type. The modification costs vary by the magnitude of the required modifications. The moderate modification option was assumed to require removing the superelevation within the roadway and filling the ditches to allow the 13.7 m drop center 5th wheel chip van access. We assumed that these modifications would cost $ 3,281 per km. The major modification option

Table 1 Chip Van Operating Characteristics for the three truck types Tablica 1. Tehničke značajke triju promatranih tipova prikolica

Trailers Prikolice

Volume Capacity, m3

Speed on Forest Roads (empty or loaded), km/h

Speed on Highways (empty or loaded), km/h

Operating Cost, $/h Modification Cost, $/km Maksimalna brzina na Jedinični trošak, $/h Troškovi rekonstrukcije, $/km autocesti (puna ili prazna), km/h

Obujam, m3

Maksimalna brzina na šumskoj cesti (puna ili prazna), km/h

73.6

16.1

72.4

$ 95.37

93.4

16.1

72.4

$ 90.95

$ 3,281

113.3

16.1

72.4

$ 99.79

$ 9,843

12.8 m Stinger Steered Samokretna prikolica dugačka 12,8 m 13.7 m Drop Center 5th wheel Poluprikolica dugačka 13,7 m s utovarom na sredini 16.2 m Drop Center 5th wheel Poluprikolica dugačka 16,2 m s utovarom na sredini

204

Croat. j. for. eng. 34(2013)2

Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215) S. Beck and J. Sessions

half of the link length needed to be modified because forest road curves are approximately half of the transportation network. Table 2 Sale Nodes Tablica 2. Mjesta prodaje Volume of Biomass – Volumen biomase Harvest Node

Destination Node

Biomass, m3

Mjesto iveranja

Mjesta isporuke

Biomasa, m3

135,921

28,883

175,564

The sale nodes for the small network (Fig. 4) are nodes 1, 2, and 3. The associated amount of biomass for each sale (chips or hogfuel) is identified in Table 2. All of the biomass is to be delivered to only one mill (Node 10). The haul and modification costs per link are provided in the appendix (Table 5). The ACO had a stopping criterion of 1,000 iterations. The heuristic converged on its solution rather quickly (iteration 282). The optimal solution to this problem using the ACO is $ 72,140. This amounted to $ 6,225 in modification costs and $ 65,915 in hauling costs. The optimal path is shown for each sale in Table 3 and Fig. 4. There were 1,454 trips from Unit 1 to the

Fig. 4 Small example road modification network, adapted from (Sessions 1985) Slika 4. Primjer modificirane mreže šumskih prometnica (prilagođeno iz Sessions 1985) was assumed to require filling the ditches, reversing the superelevation, and widening the roadway on a few select curves. These modifications were estimated to cost $ 9,843 per km (Table 1). We assumed that only

Table 3 The Optimal Path for the Small Network Using Ant Colony Heuristic Tablica 3. Optimalni pravac izvoženja prema metodi mravlje kolonije Total Cost – Ukupni troškovi prijevoza

$ 72,139.50

–

Sale 1 – Prodaja 1

Sale 2 – Prodaja 2

Sale 3 – Prodaja 3

Truck Type – Vrsta prikolice

13.7 m Drop Center 5th wheel

16.2 m Drop Center 5th wheel

Poluprikolica dugačka 13,7 m s utovarom na sredini

Poluprikolica dugačka 16,2 m s utovarom na sredini

Best Node Path – Optimalni pravac izvoženja

–

Croat. j. for. eng. 34(2013)2

205

S. Beck and J. Sessions Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215)

Table 4 The Optimal Path for the Small Network Using Mixed Integer Programming Tablica 4. Optimalni pravac izvoženja prema metodi mješovitoga cjelobrojnoga linearnoga programiranja Total Cost – Ukupni troškovi prijevoza

$ 72,154.26

Sale 1 – Prodaja 1

Sale 2 – Prodaja 2

Sale 3 – Prodaja 3

Truck Type – Vrsta prikolice

13.7 m Drop Center 5th wheel

16.2 m Drop Center 5th wheel

Poluprikolica dugačka 13,7 m s utovarom na sredini

Poluprikolica dugačka 16,2 m s utovarom na sredini

Best Node Path – Optimalni pravac izvoženja

–

Mill, 309 trips from Unit 2 to the Mill, and 1,550 trips from Unit 3 to the Mill. The ACO solution was compared to a mixed integer solution (Table 4, Fig. 5). The mixed integer and ACO produced similar results; a $ 13 difference between the two approaches. This was the result of rounding when formulating the mixed integer problem. Both methods used the same truck types and paths to transport the biomass to the mill. This small example illustrates that the heuristic appears reasonable for determining near optimal solutions for similar road modification problems.

6. Application to a realistic forest transportation network – Primjena na stvarnoj šumskoj transportnoj mreži Following the favorable results of the small network, the ACO heuristic was used on the McDonald Forest, to determine the least cost path for future harvesting activities. McDonald Forest is located 11.3 km north of Corvallis and is managed by Research Forest staff, College of Forestry, OSU. McDonald Forest is a teaching, research and demonstration forest revolving around four themes. These themes are: Þ Short Rotation Wood Production with High Return on Investment, Þ High Quality, Growth Maximizing Timber Production,

206

Þ Visually Sensitive, Even-aged Forest, Þ Structurally Diverse Forest. Biomass utilization is gaining interest in western Oregon and several biomass-powered cogeneration plants exist within 95 km of McDonald Forest. A major cost of biomass operations is the transportation cost. With small profit margins, it is important to determine the least cost method for transporting biomass from the woods to the mill. Being able to determine the optimal trucks and haul routes that would reduce total transportation costs would be important to the decision to utilize biomass. We applied the ACO heuristic to develop a least cost path from a sample of harvest units distributed through McDonald Forest. McDonald Forest is approximately 2,914 ha with 113 km of road or about 37.3 m of forest roads per hectare (Lysne D. and Klumph, B. OSU College Forests, Corvallis, Oregon, Personal Communication, December 14, 2011). The McDonald Forest road network and possible truck routes through Corvallis are shown in Fig. 6. Thirty hypothetical timber harvests (sales) were spread through McDonald Forest (Fig. 6) for the purpose of reducing fuel loading around the urban interface. These timber harvests were assumed to produce and recover 89.7 green tonnes of biomass per hectare or 113.3 m3 of biomass with 50 % moisture content. It was estimated that each sale would harvest between 45 and 95 ha (black triangles in Fig. 6). The destination node for all of the transported biomass is a biomass plant in Eugene (48 km south of Corvallis). The estiCroat. j. for. eng. 34(2013)2

Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215) S. Beck and J. Sessions

Fig. 5 Ant Colony Optimal Haul Routes. The bold arrows indicate optimal haul routes. The large black circles indicate nodes within the transportation network. The small black circles indicate the road modification option for the 16.2 m drop center 5th wheel chip van, the small horizontally hatched circles indicate the road modification option for the 13.7 m drop center 5th wheel chip van, and the small white circles indicate the road modification option for the stingersteered chip van Slika 5. Optimalni pravac izvoženja prema metodi mravlje kolonije. Podebljane strelice označuju optimalni pravac izvoženja, dok veliki krugovi označuju raskrižja transportne mreže. Mali tamni kružići označuju šumsku cestu prilagođenu poluprikolici za šumsku sječku dugačkoj 16,2 m s utovarom na sredini, mali vodoravno iscrtkani kružići označuju šumsku cestu prilagođenu poluprikolici za šumsku sječku dugačkoj 13,7 s utovarom na sredini, dok mali bijeli kružići označuju šumsku cestu prilagođenu samokretnoj prikolici za šumsku sječku

Fig. 6 Mixed Integer Optimal Haul Routes. The bold arrows indicate optimal haul routes. The large black circles indicate nodes within the transportation network. The small black circles indicate the road modification option for the 16.2 m drop center 5th wheel chip van, small horizontally hatched circles indicate the road modification option for the 13.7 m drop center 5th wheel chip van, and the small white circles indicate the road modification option for the stingersteered chip van Slika 6. Optimalni pravac izvoženja prema metodi mješovitoga cjelobrojnoga linearnoga programiranja. Podebljane strelice označuju optimalni pravac izvoženja, dok veliki krugovi označuju raskrižja transportne mreže. Mali tamni kružići označuju šumsku cestu prilagođenu poluprikolici za šumsku sječku dugačkoj 16,2 m s utovarom na sredini, mali horizontalno iscrtkani kružići označujuju šumsku cestu prilagođenu poluprikolici za šumsku sječku dugačkoj 13,7 s utovarom na sredini, dok mali bijeli kružići označuju šumsku cestu prilagođenu samokretnoj prikolici za šumsku sječku

mated travel speed on forest roads was 16.1 km/h and 72.4 km/h on major highways (loaded or unloaded). On public highways, it was assumed that any truck combination could be used without incurring any road modification costs.

are the same as Table 1. Once the chip vans were outside of the McDonald Forest, it was assumed that any chip van could be used without incurring a road modification cost. It was also assumed that adequate turnarounds exist to permit use of each truck type.

The transportation network included 405 nodes and 2,433 links, including the existing transportation network and two modification options for each link. The existing transportation network was assumed to only permit stinger-steered trailer access. The other two trailer types required temporary road modification for access similar to the small network problem. The chip van operating characteristics in this problem

The routes for the 30 sales produced by the ACO in 10,000 iterations are shown in Fig. 8. For every sale, the ACO determined that the least cost path used a 16.2 m drop center 5th wheel chip van. The total transportation cost was $ 2,697,920 with $ 254,647 in road modification costs and $ 2,443,273 in haul costs. The road modification costs amount to 9 % of the total cost. If no road modifications had been made, only the

Croat. j. for. eng. 34(2013)2

207

S. Beck and J. Sessions Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215)

Fig. 7 McDonald Forest Road Network, Corvallis, Oregon, USA Slika 7. Mreža šumskih cesta u šumi »McDonald« Corvallis, Oregon, SAD

Fig. 8 Optimal route path for all 30 sales, McDonald Forest, Corvallis, Oregon, USA Slika 8. Optimalni pravac izvoženja za svih 30 turnusa u šumi »McDonald« Corvallis, Oregon, SAD

stinger steered chip van could have been used with a total transportation cost of $ 3,703,310 (100 % haul costs). In this example, the ability to modify the roadway to allow larger trucks access to these sales reduced the total transportation cost by 27 %. The ability to reduce transportation costs by 27 % is a large benefit when margins are as slim as they are in the biomass market. This implies that being able to reduce the haul cost with the application of road modifications could have a significant positive impact.

framework to assist in deciding the optimal truck type. When comparing cost per tonne versus highway haul kilometers, the 16.2 m drop center 5th wheel chip van is the most economical (Fig. 9). However, if the forest transportation network requires modification, the most economical chip van changes. For illustration, we assume that loaded and empty vehicles of a given type travel at the same speed and have the same hourly cost (Table 1). The cost per tonne including transport and road investment is:

7. Single Harvest Unit Analysis – Case Two – Analiza pojedinačne sječine – slučaj prvi The network example provides an example of how several nearby chip or hog fuel sales and the use of road modifications can reduce overall transportation costs when considering road investments that benefit more than one sale. However, the ability to have nearby chip or hog fuel sales may not be practical. For the case of isolated sales, we provide a decision-making

208

Cost Per Tont = +

2 × HK × OCt 2 × FK × OCt + + KPHHt × VCt KPH Ft × VCt

FK × PFKt × MCt H× V

∀ t ∈T (9)

Where: HK FK OCt

distance traveled on highway roads (one-way), km, distance traveled on forest roads (one-way), km, operating cost of chip van, t ($/hr), Croat. j. for. eng. 34(2013)2

Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215) S. Beck and J. Sessions

two trucking options) can be calculated for any two trucking options:

HM =

 2 ×OC b PFK b × MC b PKMa × MCa  2 ×OCa FK ×  + − −  KPH + VC H V KPH VC H ×V × ×   b Fb Fa a  2 ×OCa 2 ×OC b   KPH × VC − KPH × VC   Ha a Hb b

(10)

Fig. 9 Comparison of cost per tonne versus highway kilometers when traveling on highway roads. When traveling over highway roads or when traveling on the forest transportation network, where no modifications are required for all vehicles, the most economical chip van is the 16.2 m drop center 5th wheel chip van. On an 80 km highway haul, the cost savings is $ 2.71 per tonne comparing a 12.8 m stinger-steered chip van to a 16.2 m drop center 5th wheel chip van. We assumed each trailer is weight limited Slika 9. Usporedba troškova prijevoza po toni prema udaljenosti prijevoza autocestom. Prilikom prijevoza autocestom ili šumskim cestama gdje nije bilo potrebe za rekonstrukcijom najekonomičnijom se pokazala poluprikolica za šumsku sječku dugačka 16,2 m s utovarom na sredini. Na 80-om km autoceste ta je poluprikolica 2,71 $ po toni isplativija od samokretne prikolice za šumsku sječku dugačke 12,8 m uz pretpostavku poštivanja ograničenja nosivosti VCt volume capacity of chip van, t, KPHHt average operating speed on highway roads for chip van, t (km/h), KPHFt average operating speed on forest roads for chip van, t (km/h), PFKt %age of the forest road kilometers that need to be modified for chip van, t, MCt forest road modification cost for chip van, t ($/km), V harvest volume per hectare, tonnes/ha, H total harvest area, ha. Equation 9 can be manipulated to compare alternative truck options for the single sale. For example, the breakeven highway haul distance (the highway distance that provides the same cost per tonne between Croat. j. for. eng. 34(2013)2

The subscripts »a« and »b« indicate the two trucking options being compared. Equation 10 assumes that both truck options can be operated on the highway. Some counties may have restrictions over some roads that do not permit trucks or trailer combinations over a maximum length or weight. The breakeven equation between the 12.8 m stinger-steered chip van and the 16.2 m drop center 5th wheel chip van, if no road investment is required, is trivial (Fig. 9). The cost per tonne in the 16.2 m drop center 5th wheel chip van is always lower than the cost per tonne in the 12.8 m stinger-steered chip van. The breakeven highway distance between the 12.8 m stinger-steered chip van and the 13.7 m drop center 5th wheel chip van for the 90 green tonnes of biomass per hectare case as a function of in forest kilometers (FK) is (operating characteristics from Table 1 were rounded for ease of illustration):

HM =

 2 × $ 91 0.5 × 3 281 2 × $ 95  + − FK ×   15 × 29.9 90 × H 15 × 23.6   2 × $ 95 2 × $ 91   75 × 23.6 − 75 × 29.9 

(11)

Equation (11) is the highway distance (km) needed to be traveled before the 13.7 m drop center 5th wheel chip van becomes economical for a given in forest hauling distance. The breakeven distance for a harvest area of 50 ha between these two vehicles for 2 km on forest roads is 17.8 highway km. For distances less than 17.8 km, it is more economical to use the 12.8 m stinger-steered chip van. For distances greater than 17.8 km and less than 179.4 km, it is more economical to use the 13.7 m drop center 5th wheel chip van (Fig. 10). A breakeven analysis of an in forest hauling distance of 15 km is shown in Fig. 11. For the case of removing 45 green tonnes per hectare (such as a thinning operation) on a harvest unit of 50 ha and the in forest, hauling distance was either 2 km (Fig. 12) or 15 km (Fig. 13). The optimal trucking option would be the 12.8 m stinger-steered chip van for highway hauling distances less than 45.7 km, when hauling on 2 km of forest road and 342.6 km when hauling on 15 km of forest road. As volume removed is reduced, the use of road modifications to allow

209

S. Beck and J. Sessions Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215)

close vicinity, the larger transport volume justifies a greater investment and makes a larger chip van economical.

8. Concluding Remarks – Zaključna razmatranja Mixed integer programming and breakeven analysis have been applied for a long time to address forest transportation problems. The focus of this application has been in response to the worldwide interest in the utilization of forest residues for alternative energy. Unlike the primary log market, roads were not built to extract forest residues and the limited value of these

Fig. 10 Comparison of cost per tonne versus highway kilometers, when traveling over 2 km on forest road. This comparison uses 90 green tonnes per hectare for a 50 ha harvest unit. Modification costs are only applied to half of the distance traveled on a forest road. As the highway haul distance increases, a larger chip van becomes more economical. In this case, 17.8 km of highway hauling is the breakeven case between a 13.7 m drop center 5th wheel chip van and a 12.8 m stinger-steered chip van. The 16.2 m drop center 5th wheel chip van becomes economical over the 13.7 m drop center 5th wheel chip van at 179.4 km highway hauling Slika 10. Usporedba troškova prijevoza po toni prema udaljenosti prijevoza autocestom kada je vožnja šumskom cestom dulja od 2 km. Dobiveni podaci temelje se na sječnoj površini od 50 ha i 90 t svježe sječke po hektaru. Troškovi potrebni za rekonstrukciju šumskih cesta izračunavaju se za pola njihove duljine. S povećanjem udjela vožnje autocestom veća prikolica postaje ekonomičnija. U ovom slučaju na 17,8 km autoceste poluprikolica za šumsku sječku dugačka 13,7 m s utovarom na sredini postaje ekonomski isplativija od samokretne prikolice za šumsku sječku dugačke 12,8 m, dok poluprikolica za šumsku sječku dugačka 16,2 m s utovarom na sredini postaje ekonomski isplativija od poluprikolice za šumsku sječku dugačke 13,7 m s utovarom na sredini na 179,4 km autoceste

larger vehicle access tends to increase transportation costs per tonne. From the single harvest unit case, it is apparent that modifying the transportation network is not always the economical option. However, in the McDonald Forest transportation network example, it was cost efficient to modify the network to allow larger vehicles access. By grouping several biomass harvest units in

210

Fig. 11 Comparison of cost per tonne versus highway kilometers when each truck must travel 15 km on forest road. This comparison uses 90 green tonnes per hectare for a 50 ha harvest unit. Modification costs are only applied to half of the distance traveled on a forest road. As the highway haul distance increases, a larger chip van becomes more economical. In this case, 133.8 km of highway hauling is the breakeven case between a 13.7 m drop center 5th wheel chip van and a 12.8 m stinger-steered chip van Slika 11. Usporedba troškova prijevoza po toni prema udaljenosti prijevoza autocestom kada je vožnja šumskom cestom dulja od 15 km. Dobiveni podaci temelje se na sječnoj površini od 50 ha i 90 t svježe sječke po hektaru. Troškovi potrebni za rekonstrukciju šumskih cesta izračunaju se za pola njihove duljine. S povećanjem udjela vožnje autocestom veća prikolica postaje ekonomičnija. U ovom slučaju na 133,8 km autoceste poluprikolica za šumsku sječku dugačaka 13,7 m s utovarom na sredini postaje ekonomski isplativija od samokretne prikolice za šumsku sječku dugačke 12,8 m Croat. j. for. eng. 34(2013)2

Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215) S. Beck and J. Sessions

products will usually not support widespread reconstruction of the forest network. However, strategic investments in the existing road network - some temporary, some permanent, may be justified. Decision support for temporary activities, such as filling ditches and changing road cross slopes to enable large vehicle access, has not been available in the literature. When these ideas were applied to schedule multiple biomass operations over a common road net-

Fig. 12 Comparison of cost per tonne versus highway kilometers when traveling over 2 km on forest road. This comparison uses 45 green tonnes per hectare for a 50 ha harvest unit. Modification costs are only applied to half of the distance traveled on a forest road. As the highway haul distance increases, a larger chip van becomes more economical. In this case, 45.7 km of highway hauling is the breakeven case between a 13.7 m drop center 5th wheel chip van and a 12.8 m stinger-steered chip van. Not shown, the 16.2 m drop center 5th wheel chip van becomes economical over the 13.7 m drop center 5th wheel chip van at 368.9 km highway hauling Slika 12. Usporedba troškova prijevoza po toni prema udaljenosti prijevoza autocestom kada je vožnja šumskom cestom dulja od 2 km. Dobiveni podaci temelje se na sječnoj površini od 50 ha i 45 t svježe sječke po hektaru. Troškovi potrebni za rekonstrukciju šumskih cesta izračunaju se za pola njihove duljine. S povećanjem udjela vožnje autocestom veća prikolica postaje ekonomičnija. U ovom slučaju na 45,7 km autoceste poluprikolica za šumsku sječku dugačka 13,7 m s utovarom na sredini postaje ekonomski isplativija od samokretne prikolice za šumsku sječku dugačke 12,8 m, dok poluprikolica za šumsku sječku dugačka 16,2 m s utovarom na sredini postaje ekonomski isplativija od poluprikolice za šumsku sječku dugačke 13,7 m s utovarom na sredini na 368,9 km autoceste Croat. j. for. eng. 34(2013)2

Fig. 13 Comparison of cost per tonne versus highway kilometers when traveling over 15 km on forest road. This comparison uses 45 green tonnes per hectare for a 50 ha harvest unit. Modification costs are only applied to half of the distance traveled on a forest road. As the highway haul distance increases, a larger chip van becomes more economical. In this case, the 12.8 m stinger-steered chip van is the most economical for highway hauling distances less than 342.6 km Slika 13. Usporedba troškova prijevoza po toni prema udaljenosti prijevoza autocestom kada je vožnja šumskom cestom dulja od 15 km. Dobiveni podaci temelje se na sječnoj površini od 50 ha i 45 t svježe sječke po hektaru. Troškovi potrebni za rekonstrukciju šumskih cesta izračunaju se za pola njihove duljine. S povećanjem udjela vožnje autocestom veća prikolica postaje ekonomičnija. U ovom slučaju samokretna prikolica za šumsku sječku dugačka 12,8 m najekonomičnija je do 342,6 km autoceste

work, the ACO heuristic obtained an optimal solution to a small problem; and when applied to a more realistic problem, quickly provided a solution. As transport volume increases, more could be spent on road modifications to allow larger truck capacity access. Being able to modify the forest transportation network to accommodate larger trucks access could greatly reduce hauling costs. Decisions for isolated biomass operations depend on road modification cost, transport volume, and transport costs on forest and highway roads. Breakeven analysis can be used to determine the optimal vehicle type. Further research is required to determine if the associated costs used in this paper accurately represent the road modification costs required to allow these non-standard trucks access.

211

S. Beck and J. Sessions Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215)

Appendix – Dodatak

9. References – Literatura

Donati, A. V., Montemanni, R., Casagrande, N., Rizzoli, A. E., Gambardella, L. M., 2008: Time Dependent Vehicle Routing Problem with a Multi Ant Colony System. European Journal of Operational Research 185(3): 1174–1191.

Glauz, W. D., Harwood, D. W., 1991: Superelevation and Body Roll Effects on Offtracking of Large Trucks. Transportation Research Record 1303, Transportation Research Board of the National Academies. Gover, F., Kochenberger, G., 2003: Handbook of Metaheuristics. Kluwer Academic Publishers.

Poluprikolica dugačka 13,7 m s utovarom na sredini

17.91

2,600

Poluprikolica dugačka 16,2 m s utovarom na sredini

19.66

7,800

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

6.14

5.86

850

6.43

2,550

12.28

11.71

1,700

12.85

5,100

6.14

5.86

850

13.7 m Drop Center 5th wheel 1

Poluprikolica dugačka 13,7 m s utovarom na sredini 16.2 m Drop Center 5th wheel

Rizzoli, A. E., Montemanni, R., Lucibello, E., Gambardella, L. M., 2007: Ant colony Optimization for Real-World Vehicle Routing Problems: From Theory to Applications. Swarm Intelligence 1: 135–151.

Sessions, J., Wimer, J., Costales, F., Wing, M. G., 2010: Engineering Considerations in Road Assessment for Biomass Operations in Steep Terrain. Western Journal of Applied Forestry 25(3): 144–153.

16.2 m Drop Center 5th wheel

Hoos, H., Stutzle, T., 2005: Stochastic Local Search, Foundations and Applications. Morgan Kaufmann Publishers.

Sessions, J., 1985: A Heuristic Algorithm for the Solution of the Variable and Fixed Cost Transportation Problem. Society of American Foresters Symposium, Athens, Georgia 324– 336.

18.79

13.7 m Drop Center 5th wheel

Dorigo, M., Stutzle, T., 2004: Ant Colony Optimization. The MIT Press. Geem, Z., 2009: Music-Inspired Harmony Search Algorithm. Springer–Verlag.

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

Troškovi rekonstrukcije, $/pravac

Modification Cost, $/Link

Truck Type Vrsta prikolice

Troškovi prijevoza po turnusu, $/prikolica/pravac

Craven, M., Wing, M., Sessions, J., Wimer, J., 2011: Assessment of Airborne Light Detection and Ranging (LiDAR) for use in Common Forest Engineering Geomatic Applications. M.S. Thesis, Oregon State Universtiy, College of Forestry, Corvallis. Retrieved from Oregon State University: <http:// hdl.handle.net/1957/21803

Šifre pravaca izvoženja Round Trip Haul Cost, $/Truck/Link

Contreras, M. A., Chung, W., Jones, G., 2008: Applying Ant Colony Optimization Metaheuristic to Solve Forest Transportation Planning Problems with Side Constraints. Canadian Journal of Forest Research 38(11): 2896–2910.

Link Identifier

To – Do

Bowers, S., 2006: Managing Woodland Roads, A Field Handbook. Oregon State University Extension Service, Oregon State University, Corvallis, Oregon.

Table 5 Haul and Modification Cost for the Small Network Tablica 5. Troškovi prijevoza i rekonstrukcije

From – Od

Anderson, P., Sessions, J., 1991: Factors affecting the maximum grade a truck can climb around a curve. In: Proceedings, Fifth International Conference on Low Volume Roads. Transportation Research Board, National Research Council, Washington, D.C. TRR 1291, Volume 2: 15–19.

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m 13.7 m Drop Center 5th wheel

Poluprikolica dugačka 13,7 m s utovarom na sredini 16.2 m Drop Center 5th wheel

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m 13.7 m Drop Center 5th wheel

212

Poluprikolica dugačka 13,7 m s utovarom na sredini

Croat. j. for. eng. 34(2013)2

Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215) S. Beck and J. Sessions

16.2 m Drop Center 5th wheel 2

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m

13.7 m Drop Center 5th wheel Poluprikolica dugačka 13,7 m s utovarom na sredini

4.13

600

4.54

1,800

6.43

2,550

9.39

16.2 m Drop Center 5th wheel Poluprikolica dugačka 16,2 m s utovarom na sredini

8.96

1,300

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

7.95

13.7 m Drop Center 5th wheel Poluprikolica dugačka 13,7 m s utovarom na sredini

7.58

1,100

8.32

3,300

13.7 m Drop Center 5th wheel 3

Poluprikolica dugačka 13,7 m s utovarom na sredini 16.2 m Drop Center 5th wheel

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m

9.83

3,900

6.50

16.2 m Drop Center 5th wheel Poluprikolica dugačka 16,2 m s utovarom na sredini

6.20

900

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

3.61

13.7 m Drop Center 5th wheel Poluprikolica dugačka 13,7 m s utovarom na sredini

3.45

500

3.78

1,500

13.7 m Drop Center 5th wheel 3

Poluprikolica dugačka 13,7 m s utovarom na sredini 16.2 m Drop Center 5th wheel

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m

6.80

2,700

6.32

16.2 m Drop Center 5th wheel Poluprikolica dugačka 16,2 m s utovarom na sredini

6.03

875

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

6.14

13.7 m Drop Center 5th wheel Poluprikolica dugačka 13,7 m s utovarom na sredini

5.86

850

6.43

2,550

13.7 m Drop Center 5th wheel 3

Poluprikolica dugačka 13,7 m s utovarom na sredini 16.2 m Drop Center 5th wheel

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m

6.61

2,625

9.03

16.2 m Drop Center 5th wheel Poluprikolica dugačka 16,2 m s utovarom na sredini

8.61

1,250

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

5.42

13.7 m Drop Center 5th wheel Poluprikolica dugačka 13,7 m s utovarom na sredini

5.17

750

16.2 m Drop Center 5th wheel Poluprikolica dugačka 16,2 m s utovarom na sredini

5.67

2,250

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

6.50

13.7 m Drop Center 5th wheel Poluprikolica dugačka 13,7 m s utovarom na sredini

6.20

900

16.2 m Drop Center 5th wheel Poluprikolica dugačka 16,2 m s utovarom na sredini

6.80

2,700

13.7 m Drop Center 5th wheel 4

Poluprikolica dugačka 13,7 m s utovarom na sredini th

16.2 m Drop Center 5 wheel 4

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m

9.45

3,750

6.14

5.86

850

13.7 m Drop Center 5th wheel 4

Poluprikolica dugačka 13,7 m s utovarom na sredini th

16.2 m Drop Center 5 wheel 4

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m

Croat. j. for. eng. 34(2013)2

6.43

4.34

2,550

213

S. Beck and J. Sessions Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215)

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

1.81

1.72

250

16.2 m Drop Center 5th wheel 8

13.7 m Drop Center 5 wheel 7

Poluprikolica dugačka 13,7 m s utovarom na sredini 16.2 m Drop Center 5th wheel

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m

Poluprikolica dugačka 13,7 m s utovarom na sredini

1.89

750

6.50

6.20

900

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m

6.80

2,700

9.03

8.61

Poluprikolica dugačka 13,7 m s utovarom na sredini

Poluprikolica dugačka 16,2 m s utovarom na sredini 12.8 m Stinger Samokretna prikolica dugačka 12,8 m

9.45

5.06

4.82

700

Poluprikolica dugačka 13,7 m s utovarom na sredini

18.60

Poluprikolica dugačka 16,2 m s utovarom na sredini

20.41

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

9.03

Poluprikolica dugačka 13,7 m s utovarom na sredini

8.61

Poluprikolica dugačka 16,2 m s utovarom na sredini

9.45

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

0.36

Poluprikolica dugačka 13,7 m s utovarom na sredini

0.34

0.38

150

16.2 m Drop Center 5th wheel 11

13.7 m Drop Center 5th wheel 8

Poluprikolica dugačka 13,7 m s utovarom na sredini

13.7 m Drop Center 5th wheel

16.2 m Drop Center 5th wheel 7

16.2 m Drop Center 5th wheel

13.7 m Drop Center 5 wheel 10

19.51

13.7 m Drop Center 5th wheel

2,100

16.2 m Drop Center 5th wheel

16.2 m Drop Center 5th wheel 7

12.8 m Stinger Samokretna prikolica dugačka 12,8 m

5.29

13.7 m Drop Center 5th wheel

13.7 m Drop Center 5th wheel 7

Poluprikolica dugačka 16,2 m s utovarom na sredini

Sažetak

Odlučivanje o izvoznim pravcima prikolica za šumsku sječku uz optimizaciju metodom mravlje kolonije i analizu prekretnice troškova Nekonvencionalni (sekundarni) šumski proizvodi pružaju mogućnost povećanja ekonomske vrijednosti šume, dok transport takvih proizvoda u nekim slučajevima zahtijeva prijevoz specijaliziranim vozilima. Zbog činjenice da postojeća šumska prometna infrastruktura uglavnom nije dizajnirana prema standardima koje specijalizirana vozila zahtijevaju potrebno je raditi izmjene (rekonstrukcije) na šumskim cestama (zatrpavanje odvodnih jaraka, proširivanje i rekonstruiranje kolničke konstrukcije) kako bi se takvim vozilima omogućio pristup šumi. Svaku izmjenu kolničke konstrukcije potrebno je prilagoditi tehničkim karakteristikama specijaliziranih vozila koja će naposljetku tom šumskom cestom i prometovati. U ovom su radu promatrani različiti transportni sustavi za (1) višestruko izvoženje biomase, (2) pojedinačno izvoženje biomase.

214

Croat. j. for. eng. 34(2013)2

Forest Road Access Decisions for Woods Chip Trailers Using Ant Colony Optimization... (201–215) S. Beck and J. Sessions

U slučaju višestrukoga izvoženja biomase primijenjena je metoda mješovitoga cjelobrojnoga linearnoga programiranja i metoda temeljena na principu mravlje kolonije »Ant Colony Optimization« (ACO). Obje su metode korištene na malom uzorku prikazanom na slici 4. Jedina razlika između tih dviju metoda iznosi $ 13, a nastala je tijekom izračuna (zaokruživanja) kod metode mješovitoga cjelobrojnoga linearnoga programiranja (tablice 3 i 4). Metoda ACO primijenjena je pri određivanju optimalnoga specijaliziranoga vozila (prikolice) te pri izboru optimalnoga pravca izvoženja za 30 hipotetski odabranih mjesta iverenja u šumama »McDonald« u Corvallisu, Oregon, Sjedinjene Američke Države (slika 7). Metodom ACO utvrđeno je da su ukupni troškovi prijevoza najmanji prilikom korištenja najveće prikolice za šumsku biomasu, što razumijeva najveću rekonstrukciju šumske ceste na pravcima izvoženja. Na dionicama gdje nije bila potrebna izmjena kolničke konstrukcije ukupni troškovi prijevoza manji su za 27 %. Pri pojedinačnom izvoženju biomase primijenili smo prikladniju analizu prekretnice troškova. Analiza prekretnice troškova pretpostavlja da su pravci izvoženja poznati, dok su varijable operativne karakteristike prikolice za šumsku biomasu te troškovi izgradnje pripadajuće šumske ceste s uračunatim troškovima rekonstrukcije šumske ceste (jednadžba 9). Pomoću jednadžbe 10 može se utvrditi na kojoj su udaljenosti (autoceste) troškovi transporta kod promatranih dviju opcija jednaki. U ovom smo radu procijenili četiri različita slučaja, čistu sječu i prorede s kratkim i dugim duljinama privlačenja. Utvrđeno je da se zbog povećanja privlačenja drvnih sortimenata te zbog potreba za izmjenom kolničke konstrukcije pri korištenju većega kamiona nadmašuje korisnost povećanoga obujma biomase po tovaru većega kamiona. Osim toga, zbog smanjenja obujma biomase po hektaru (proreda) te troškova rekonstrukcije pojedine šumske ceste potreban je dulji transport autocestom (veća udaljenost) da bi veći kamion bio ekonomičan. Ovom je analizom utvrđeno da rekonstrukcija šumskih cesta nije uvijek ekonomičan izbor. Ključne riječi: optimizacija metodom mravlje kolonije, transport šumske biomase, pristupačnost vozilima

Authors’ address – Adresa autorâ:

Received (Primljeno): October 24, 2012 Accepted (Prihvaćeno): December 30, 2012 Croat. j. for. eng. 34(2013)2

Storm Beck, MSc. * e-mail: storm.beck@oregonstate.edu Prof. John Sessions, PhD. e-mail: john.sessions@oregonstate.edu Oregon State University Department of Forest Engineering, Resources and Management Oregon State University Corvallis, Oregon 97331 USA * Corresponding author – Glavni autor

215

Original scientific paper – Izvorni znanstveni rad

Planning Forest Opening with Forest Roads Janez Krč, Jurij Beguš Abstract – Nacrtak The article presents the model for determining inaccessible forest areas by density of forest roads. The model is based on the GIS analysis of the distances between the existing network of public and forest roads and inaccessible forest areas, sizes of excluded forest areas, and forest site potentials. In order to increase forest road density, the following must be done: (1) construct connecting roads to the inaccessible forest areas and (2) construct new forest roads with different density in the excluded inaccessible forest areas. The model provides the minimum size of the inaccessible area located at least 300 m away from the existing forest and public road. The selected inaccessible forest areas are first analyzed according to their size – plot size of at least 30 ha is used as a model default size suitable for economically justified construction of the access road that connects the existing road network to the inaccessible forest area. The analysis showed that there are still 210,385 ha of inaccessible forests in Slovenia according to the model criteria. According to the research of regional units conducted by forest experts and based on the determination of priorities for the next ten-year forest management plan, the construction of 758 km of new forest roads is planned at the national level. Keywords: forest road, density, forest operation, model, forest management plan

1. Introduction – Uvod The construction of forest road network is considered as the key element for successful forest management. It has the biggest impact on the forest production function, since it enables and also technologically defines the forest operations in the majority of cases. The importance of forest roads or forest road network can also be considered in terms of comparison, i.e. asset values that occur in forest operations. The simplified model can be of assistance where the forest, as a piece of land, is considered as fixed asset, whereas the forest stand, road network and working means represent the current assets – all used in forest operations. In view of the above mentioned factors, the forest management should provide the sustainability of these factors. The current asset values (forest stand, road network and working means) are in the approximate value interrelation of 100:10:1. The relation is calculated for the model example covering 2667 ha taking into account the following assumptions: Þ Stand: value 45,142,000 € (timber value 60 EUR/m3, volume 282 m3/ha) (SURS: Gozd in gozdarstvo 2013), (Poročilo o gozdovih za leto 2011, 2012); Croat. j. for. eng. 34(2013)2

Þ Forest roads: value 4,300,000 € (construction costs 65 EUR/m, road density 24,8 m/ha) (Robek et al. 2007), (Gozdnogospodarski in lovsko upravljalski načrti območij za obdobje 2011–2020, 2012); Þ Working means: value 500,000 € (efficiency 80 m3/day, utilization 200 days/year). The calculation was made for the projected model of forest operations with the capacity of modern working means serving as a referential value (harvester and forwarder). The average annual volumes of allowable cut of 6m3/ha/year were applied for the calculation of the necessary utilization of the available capacity of modern working means. On the basis of the above model assumptions, the necessary forest cutting area for forest operations is well congruent with the size of the average forest district in Slovenia and volumes of allowable cuts defined in the forest management plans (Poročilo o gozdovih za leto 2011, 2012). The relations show the great importance of systematic approach to the issue of forest opening conducted on a strategic and detailed level. The strategic or general level is implemented at various levels of forest management planning. This should replace the current approach of integrated planning of opening the forests

217

J. Krč and J. Beguš

with forest roads that called for the preparation of »Perspective programs for integrated opening of forests« (Program odpiranja gozdov z gozdnimi prometnicami, 1990), studies that have not been successfully implemented anywhere in Slovenia. Today, these studies can be of a great assistance and are widely used as reference points to conduct the tests of methodology presented in this article. In the forest management plans prepared by Slovenia Forest Service (SFS), the priority areas of opening are determined within the strategic planning of forest road construction. These areas represent the surfaces where the forest roads should be constructed. Due to legal procedures of adopting forest management plans, they do not include detailed routes of forest roads. At a higher level of forest management planning, i.e. at the level of regional forest management plans, the priority opening areas are not determined. However, the strategic evaluation should include the volume of new forest road construction that would provide the necessary forest road network from the point of view of timber harvesting.

1.1 Previous research – Dosadašnja istraživanja Expert and scientific literature have been dealing with the issue of estimation of forest opening together with the evaluation of the density of the existing forest road network for a relatively long time (Matthews 1942, quoting Chung et al. 2008). The research of forest opening frequently includes the relations between skidding costs and the costs of forest road construction and maintenance. Frequently, these estimations apply the assumption about the equal distribution of logging operations in a certain area. As a result, the optimum density and position of a forest road network are determined. These are calculated according to the differential ratios of timber skidding costs. Skidding costs are primarily dependant on the applied skidding method and skidding distances (Krč 1999, Košir 2000). Some studies deal with the optimum forest road network layout on the basis of the shortest path between the appropriately distributed timber sources (timber stacks at forest road) in a certain area – i.e. from the point of view of further timber transport (Anderson et al. 2004, Dean 1997). An attempt of Preliminary Planning of Forest Road Systems Using Digital Terrain Models was conducted by Liu and Sessions (Liu et al. 1993). They set the optimal road alignment using a three step method (The first step includes the identification of possible road segments, the second step deals with the minimization of the sum of construction, maintenance and transport costs, while the third displays the results to provide the verification of result by operation planner).

218

Planning Forest Opening with Forest Roads (217–228)

The ecosystem approach to road issues was comprehensively treated by Lugo and Gucinski (2000). They proposed the unified ecosystem approach to road management using an environmental gradient analysis based on three main parameters (ecological, socioeconomic and physical). Similar to our problem, Demir (2007) tried to find a systematic solution for the remaining portion of forest road network in Turkey. He used specific functional planning criteria for different forest road network systems (production forest, reforestation forest and national parks). The determination of road density in the production forest was differentiated by growing stock (20 m/ha for stands over 250 m3/ha and 10 m3/ha under 250 m3/ha, respectively). Previous studies also state different definitions of primary forest openness. In Croatia there are different levels of primary openness determined according to the planning level (global : local) and definition purpose (planning, research work). The openness has also been determined according to the relief regions (lowlands, hilly, highlands, Karst). The planned openness (from minimal to the target) according to relief region and projected plan is determined in the interval between 7 and 30 m/ha (Pentek et al. 2007). The issue of including the road sections (public and forest roads) into the selection of productive values in terms of timber harvesting is dealt with by Pentek et al. (2011). Also the position of forest road plays an important role when estimating the level of forest openness. Thus, the length of productive and connecting roads is differentiated. According to Dobre (1995), the productive forest road length is considered when: Þ Road runs through the forest; Þ Road runs along the forest; Þ Forest is located less than 200 m from the road – there are no obstacles for skidding operations; Þ Non-forest zone is longer than 200 m, but shorter than 200 m along the road. The data from recent studies were applied and partly adjusted for the preparation and verification of the following model for planning the necessary forest opening with forest roads.

1.2 Purpose and aims of planning the necessary forest opening with forest roads – Svrha i ciljevi planiranja potrebne otvorenosti šumskim prometnicama At a strategic level, in forestry management planning of opening the forest with forest roads, the main question is related to the target density and consequently the total length of forest roads. This means Croat. j. for. eng. 34(2013)2

Planning Forest Opening with Forest Roads (217–228)

that the answer should include the quantity of new forest roads that will reach the goals set at different levels of planning. Basically, two questions are relevant: where the new forest roads need to be constructed, and how many kilometers of forest roads are still necessary. It is estimated that these two answers suffice at the strategic level, i.e. regional level. In Slovenia, the contents related to the planning of forest opening with forest roads are legally defined. Thus, the Article 3 of the Regulation on Forest Infrastructure (Pravilnik o gozdnih prometnicah, 2009) differentiates the strategic and detailed level of opening. The strategic level, i.e. the level of forest management planning defines the priority areas for forest opening with forest roads, whereas the detailed level includes making the elaborate of nought line definition. Furthermore, the Article 10 of the Forest Act (Zakon o gozdovih, 1993 and following) states that the concept of forest infrastructure is shown in the spatial part of forest management plan of regional unit. At a lower level of forestry management planning – forest management plan for forest management unit provides the concept and overview for forest infrastructure. The detailed scope for planning, construction, and maintenance of forest roads is regulated by the Regulation on Forest Infrastructure. Despite some previous attempts, the strategic and integrated planning of providing forest opening with forest roads on the basis of special programs/studies has never been truly successful and never really implemented on the national level. Current regulations try to solve the actual situation by implementing the regulations on the strategic/general planning of forest opening with forest roads to the forest management plans of forest management units with the determination of preferential areas for forest road construction. Due to formal reasons1, the routes of future forest roads are not defined in the forest management plans, and it is, therefore, hard to make an objective evaluation of how many forest roads should still be constructed to provide the optimum operation of forest production simply by applying the currently known approaches. The plans include certain actual and target forest road densities that show the level of forest opening with forest roads, but still this data insufficiently indicates further actions. Preferential opening areas show the locations where new forest roads should be necessary, but this still does not mean that

The routes of individually planned forest roads should not be drawn in the forest management plans, since then each plan would require the environmental impact assessment.

Croat. j. for. eng. 34(2013)2

J. Krč and J. Beguš

the new forest road construction would not be possible elsewhere. However, the strategic decision-making should not overlook the data of how many forest roads are still needed, i.e. not just the mere location but also the quantity that would provide the realization of the set goals. This requires a unified approach to the planning that would provide the decision-making based on the objective evaluation of needs for the increasing forest road density. According to the legal provisions and relevant issue, the accessory tool was developed that provides a quick, unified, and consistent evaluation of all forests from the point of view of forest accessibility necessary for performing forest operations. This tool consequently enables the exclusion of inaccessible areas and hence the sequence of opening, i.e. increasing forest road density, is determined in gradual steps. It is estimated that the present model tool is a step forward, since its consistent procedure helps to define the location of inaccessible forests with the quantification of needs for opening.

2. Materials and methodology – Materijal i metodologija Slovenia is located between 45°25’–46°52’ N latitudes and 13°35′–16°35′ E longitudes. Slovenia is surrounded by neighboring countries (Italy, Austria, Hungary and Croatia) and the Adriatic Sea. It covers an area of 2,027,300 hectares and has a 43 km long coastline. It belongs to the group of the smallest EU countries (the longest width of 257 km is in the East– West direction, while the smallest range of 78 km is between the North-South direction). Slovenia has four geographical regions (Pannonian on the East, Mediterranean on the South-West, Alpine on the North, and Dinaric on the South part of the country). The total forest area in Slovenia is 1,184,369 hectares. This figure represents 58.4% of the total area. High quality forests prevail, while the coppice forests spread only over 39,432 hectares and account for 3.3% of the total forest area. According to 2011 figures, the share of coniferous forest in the total growing stock is 46.2%, whereas the deciduous forest covers 53.8 %. The total annual increment amounts to 8,265,936 m3, whereas the annual allowable cut equals 5,498,733 m3 (Poročilo o gozdovih za leto 2011, 2012). According to the data of the current forest management plans for regional units (Gozdnogospodarski in lovsko upravljalski načrti območij za obdobje 2011–2020, 2012), forests in Slovenia are accessible by forest roads (12,023 km) and primary prevention fire roads (489 km), which means 12,512 km or 10.6 m/ha of roads density in total. The forest area is

219

J. Krč and J. Beguš

additionally also accessible by public roads, which could be suitable for forestry operations and wood transport. In total, the Slovenian forests are accessible by 29,244 km of roads, with a road density of 24.8 m/ha.

2.1 Data selection and analysis – Odabir podataka i analiza The model was developed according to the results of previous studies and new technological trends that have influenced the needs for additional forest roads. The model performed the geographical analysis of three main influential factors: (1) distance from the forest to the existing productive forest road or public road that provides forest accessibility, (2) forest site potentials (Košir 1975), and (3) surfaces of potentially inaccessible forests. The areas for further forest construction of forest roads were determined by applying the skidding distance and site potential factors. Then the priority of opening was determined for individually selected inaccessible areas that met the criterion of minimum size. The criteria for area selection, suitable for road construction according to the values of influential factors, were somewhat adjusted in terms of previous research results, regulations, and practice (Dobre 1995): (1) distance between the forest and the road has been increased, i.e. to 300 m, (2) minimum inaccessible area size (30 ha) has been determined, (3) also the public roads that can be used as productive roads for forest production have been included. The following general data of forestry information system were applied: Þ Forest and public roads suitable for forest production and primary forest fire prevention roads with the status of a forest road (hereinafter referred to as »forest road«). Public roads suitable for forest production have been determined with the intersection of public road linear layer and border of the forest area. Herein, the condition has been included that the section includes 200 meter of influential buffer zone around the public roads. The final classification of public roads has been determined with the in-situ examination and confirmation of Slovenia Forest Service specialists. Þ Forest-stand map with the map of forest border. Þ Map of protective forests and forests with special purpose. Þ Geo-encoded data on forest sections. Þ Map of forest plant associations, acquired from the digital data on forest sections (spatial and attributive part).

220

Planning Forest Opening with Forest Roads (217–228)

2.2 Procedure of model preparation and application – Postupak pripreme i primjene modela The preparation and basic processing of the selected data from the forestry information system is conducted first. The data preparation provides the necessary information for the determination of the position and scope of inaccessible forests in a specific area. The tools for raster processing of spatial data and DBMS modules (Database Management System) are applied for the data analysis. The attributive data are processed with DBMS modules – in terms of input data preparation as well as in terms of determination of the selected areas that are smaller than minimum projected surface for additional increasing the forest road density. The combination of basic as well as derived spatial vector and raster data of forestry informational system are applied for map preparation. Maps of model-wise inaccessible forests acquired from data processing of forest information system then serve as the basis for in-situ evaluation of the suitability of computerized model results, executed by foresters at the SFS local unit level. The process can be divided into the following steps (Fig. 1): 1. Determination of all areas with inadequate forest road network – low access areas. 2. Exclusion of inaccessible areas that are smaller than 30 ha (model-default as the minimum size of inaccessible forest area that needs further increasing of the forest road density). 3. Determination of the necessary road density per hectare. The road density is defined by site potentials (Table 1). Each surface was given the additional 300 meters of forest road for the connection to the existing road network. 4. Preparation of the numbered list of the low access areas with inadequate forest road network in the attributive form, simultaneously shown on the map (identification key is the serial number of the low access area). 5. The examination of suitability of the low access areas and the determination/selection of reasonable low access areas (i.e. the exclusion of low access areas that cannot be considered suitable for further increasing of forest road density on the basis of computerized model and input data quality). 6. Determination of low access areas where roads should be constructed with priority in the tenyear period while the forest management plan for the regional unit is in force, in order to achieve the set goals. The determination is defined by forest management goals and by the total length Croat. j. for. eng. 34(2013)2

Planning Forest Opening with Forest Roads (217–228)

J. Krč and J. Beguš

Fig. 1 Main steps of the model for determining the necessary forest road density Slika 1. Glavni koraci modela za određivanje potrebne šumske otvorenosti šumskim prometnicama of forest roads that should be constructed for this purpose (determination of priority low access areas during the validity of the forest management plan).

2.3 Definition of conditions for determining low access areas – Definiranje uvjeta za utvrđivanje nedovoljno otvorenih površina Further in the text, the main characteristics are given of the model preparation and of the process for acquiring influential data used for the evaluation of the areas requiring higher forest road density. All forest and public roads (suitable for forest production) and primary forest fire prevention roads were assigned an influential buffer zone of 300 meters according to the relief conditions, thus determining an area suitable for providing good forest accessibility. It was estimated that skidding method (tractor or cable skidder) does not play a crucial role and does not influence the width of the influential buffer zone. Thus, the cable skidding operation over the distance of 400 m and tractor skidding over the distance of 600 m was encompassed (distance from the standing tree to the road). The areas outside this zone are poorly accessible by roads. The map of low access areas was made on the basis of road intersection and 300 meter buffer zone. It determines the necessary locations of forest road construction (areas outside 300 meter buffer zone along the forest roads). Since the information layer intersections offered areas of different sizes, also with very small areas included, the size limit area had to be deCroat. j. for. eng. 34(2013)2

termined to get a reasonable idea of forest road construction. The estimation showed that the smallest low access area where the construction of forest roads would be reasonable was 30 ha. It would be unreasonable to construct forest roads on smaller areas or in other words it would be more reasonable to provide their opening with skid trails. During the development of this method, the result of suitability had to be checked with a referential study. The Study of Integrated Forest Opening with Forest Roads, i.e. »Program of Forest Opening with Forest Roads« (hereinafter referred to as »Program«), developed by the Regional Unit Kočevje in 1990 (Pro-

Table 1 Target forest road densities (TFR) according to the level of the forest site potentials Tablica 1. Ciljana gustoća šumskih prometnica (CGŠP) prema razinama potencijala šumskoga područja Site potentials

Target forest road density, m/ha

Potencijal područja

Ciljana gustoća šumskih cesta, m/ha

< 5*

5–8

9–11

>12

*The exception is the Karst region where the opening was planned also for the areas with the lowest level of site potentials (fire protection function). * Iznimka je krško područje gdje je otvaranje planirano kao za područja s najnižom razinom potencijala područja (protupožarna funkcija).

221

J. Krč and J. Beguš

gram odpiranja gozdov z gozdnimi prometnicami, 1990), was used as the referential study. The study map was compared with the model map of areas that needed opening. The method was also presented to the forest road specialists and field foresters at the Regional Unit Kočevje and thus the independent opinion on the applied methodology had been acquired already in the process of model development. The comparison of the study and the map of 300 m zone showed a high degree of accordance of inaccessible areas in terms of both location and shape. Also the

Planning Forest Opening with Forest Roads (217–228)

length of necessary forest roads established by the Program was examined and a relatively high degree of agreement was confirmed – the Program foresaw the construction of 331 km of forest roads, whereas the model offered the result of 376 km. To establish the final result, i.e. the required length of forest roads to be constructed, it was necessary to determine the required forest road length for each selected area of 30 ha or larger. The assumption was accepted that the management of better forest sites was more intensive, thus justifying the higher forest road

Fig. 2 Example of the map with the numbered selected areas in the Regional Unit Kočevje with road network at the regional level Slika 2. Primjer karte s obrojčanim odabranim područjima u Upravi šuma Kočevje s cestovnom mrežom na razini Uprave

222

Croat. j. for. eng. 34(2013)2

Planning Forest Opening with Forest Roads (217–228)

J. Krč and J. Beguš

Table 2 Example of the selected areas list in the Regional Unit Kočevje according to the level of site potentials with the planned length of new roads Tablica 2. Primjer popisa odabranih područja u Upravi šuma Kočevje prema razinama potencijala područja s planiranim duljinama novih cesta

No. Br.

Level of site potentials, selected areas, ha

New Forest Roads, km

Razina potencijala područja, odabrana područja, ha

Nove šumske prometnice, km

New

Access

Total

Nove

Pristup

Ukupno

143

448.75

11.84

0.3

12.14

116.5

355.25

9.44

0.3

9.74

88.5

164

207.25

8.75

0.3

9.05

255.5

190.75

8.93

0.3

9.23

34.25

26.5

169

43.75

169.5

9.40

0.3

9.70

...

174

15.5

14.75

0.75

0.62

0.3

0.92

175

0.62

0.3

0.92

176

30.5

0.25

0.62

0.3

0.92

177

25.75

0.61

0.3

0.91

178

5.5

24.5

0.57

0.3

0.87

densities (Table 1). The level of the forest site potentials was acquired from the forestry information system defined on the basis of forest vegetation associations (Košir 1975). The value of forest site potential is a rank that represents the relative ratio of forest vegetation associations based on forest site production potential. The ranks are scaled from 1 to 17. The most productive rank is 17, while the least productive one is 1. The value is assumed as a long-term management goal, determined by the composition of tree species which is close to the potential natural state of forest stand (Košir et al. 2006). When each area was adjusted with the necessary forest road length, the numbered list of these areas needed to be prepared. As the attributive basis, this list was connected to the graphic information layer (Table 2, and Fig. 2) with all areas appropriately numbered. The attributive list of the selected areas includes the data on the site potentials and data on forest road lengths. The list is necessary to provide clear and precise positioning of the areas in the region. In the next phase of the method, it is necessary to assess the suitability of the selected areas. Not all forest areas that are more than 300 m away from forest roads are appropriate for opening. Apart from the minimum area size factor (30 ha), the shape of the selected areas must also be taken into account. The typical example are the areas between two roads that can be very narrow and long; this is especially obvious for forest roads that run along both sides of long ridges – Number 34 in Fig. 3. Croat. j. for. eng. 34(2013)2

This is the stage where the estimation with the application of computer tools does not suffice, and the estimation of the in-situ experts is required. They decide which areas are actually suitable for opening. All the selected areas the experts find unsuitable for opening, mainly due to their shape, are excluded from the list and map. The final, adjusted model product is represented in the form of thematic map and table, which offer spatial and quantitative recapitulation of the data on the structure of the excluded inaccessible forests according to the site conditions and calculated new forest road lengths for every selected area. Additionally, the access roads are included connecting the inaccessible sections with the existing forest road network.

3. Results – Rezultati Table 3 shows the results of the described model at the national level (Slovenia). It presents the quantities of model-wise excluded inaccessible areas by regional units of Slovenia Forest Service and site potentials (Košir 1975). Then the necessary scope of new forest road construction is calculated for regional units to achieve the target road density according to the site potentials of the excluded surfaces and in the extent shown in Table 1. The scope of construction in the model is divided into the access roads and the road network that provides the forest opening. The length is determined according to the forest operations on the basis of target forest road density.

223

J. Krč and J. Beguš

Planning Forest Opening with Forest Roads (217–228)

Fig. 3 Example of the area section numbered 170 unsuitable for additional opening Slika 3. Primjer izdvojenoga područja s brojem 170 koji je neprikladan za daljnje otvaranje The analysis in the Karst area additionally included the forests with lower site potentials, where the construction of primary fire-prevention roads is planned. The majority of inaccessible areas were excluded on the sites with lower site potentials, while only a good third of the total length was excluded on the sites with the site potential level higher than 10. This only confirms the established fact that the forests located at better sites had the priority for the construction of forest roads in the past. Nowadays, more attention has to be paid to the medium potential sites. All the necessary forest roads cannot be constructed in the following management period, since the evaluation is a long-term one and in terms of the present timber harvesting technology it represents the final solution of the forest roads density. Thus, the final product determines the areas that must be made accessible to reach the management goals, set in the current plan. That is why the priority areas for increasing forest road densities are defined together with the nec-

224

essary length of new forest roads. After the final examination of the specialists of SFS in the regional units and after the determination of priorities, the final result is shown in Table 4. In general, less than a fifth (17 %) of forest roads has to be constructed in Slovenia to ensure optimal forest accessibility. Here the area of Regional Unit Sežana stands out, i.e. the area of Slovenian Istra, Brkini and Kras, as it was neglected in terms of forest road construction. However, the forest roads and primary fire prevention roads are very important in this area, representing the key element of fire protection safety. That is why the sites with lower site conditions were also included in the calculation of forest road construction in this area.

4. Discussion – Rasprava The model for planning the necessary forest opening with forest roads provides a consistent and comCroat. j. for. eng. 34(2013)2

Planning Forest Opening with Forest Roads (217–228)

J. Krč and J. Beguš

prehensive evaluation of the locations and necessary lengths of new-built forest roads for strategic and tactical forest management planning at the regional as well as the national level. Further in the text, certain characteristics are presented together with recommendations for further model development and research work in the field of forest opening with forest roads. The basis for differentiation of target forest road densities is the potential situation and forest site potentials, i.e. the projected potential of forest sites based on plant associations – and not the current situation of forest stands wood volumes. That is why some deviations can occur from the established guidelines, which presume higher forest road network density in better stands (evaluated according to the growing stock). Thus, the model serves to adjust the forest road density to the site conditions – regardless of the current level of potential forest site exploitation. The problem of access, connecting roads leading into the selected inaccessible areas is dealt with in a unified and simplified way. It is an approximation that will be definitively longer (minimum distance for the access road is included) in the majority of cases, since this method variant failed to include the coefficient of road winding factor, whereas in larger excluded areas the possibility was also foreseen of a greater number

of connecting (access) roads to the selected inaccessible forest area. The adjusted criteria for the selection of inaccessible forest areas were used with the purpose to determine the priorities for locations with relatively long distance from forest to roads in terms of skidding operations. Thus, the smallest area, entitled to further investments in road network from forest funds, was provided. The additional motive for criteria adjustment is due to the fact that the consequences of technological progress have also been taken into consideration. The development of logging and skidding technology also generates larger quantities of timber to be transported (on wheels) in the skidding operations. The share of ecologically and economically disadvantageous ground skidding is thus decreasing compared to the transport on wheels, causing the increase of the acceptable distance of timber skidding. In comparison with ground skidding, timber transport is a more ecological, faster, and more efficient method of timber skidding. It is expected that the longer distance between forest and road is also justified because the model does not include public cart tracks and all types of skid trails with road elements. The gap between the current construction scope and the need for new forest road construction on the

Table 3 Results of model calculation for increasing forest road density at the national level and its structure according to the regional units Tablica 3. Rezultati izračuna modela za povećanje gustoće šumskih cesta na nacionalnoj razini i njihova struktura po upravama šuma Site potentials, excluded surfaces, ha Area, Regional Unit

Potencijal područja, isključene površine, ha

Područje, uprava šuma Total (Slovenia) Ukupno (Slovenija) Tolmin

New Forest Roads, km Nove šumske prometnice, km New

Access

Total

Nove

Pristup

Ukupno

24197

36784

68126

76570

830

832

3046

3926.35

540.90

4467.25

4106

13997

9328

1794

494.95

57.90

552.85

Bled

1056

4216

1067

1575

131.96

26.70

158.66

Kranj

2639

3750

4690

2122

226

542

251.29

37.50

288.79

Ljubljana

5779

2296

16204

6428

241

470

591.60

84.00

675.60

Postojna

1576

2392

7057

6463

813

179

354.70

36.00

390.70

Kočevje

595

810

5563

11193

166

220

365.84

53.40

419.24

Novo Mesto

518

339

6374

24453

474

641.21

63.30

704.51

Brežice

385

336

3720

6540

216.38

34.20

250.58

Celje Nazarje

1575 632

1129 876

2153 1218

223 1300

0 0

0 129

84 149

90.18 79.94

23.40 14.70

113.58 94.64

Slovenj Gradec

137

222

994

181

31.40

8.70

40.10

Maribor

825

2062

1607

2502

819

146.91

30.30

177.21

Murska Sobota

4376

786

1838

11797

350.13

18.30

368.43

Kras

3573

6314

179.89

52.50

232.39

Croat. j. for. eng. 34(2013)2

225

J. Krč and J. Beguš

Planning Forest Opening with Forest Roads (217–228)

Table 4 Presentation of the necessary and priority construction of forest roads and primary fire-prevention roads – regional forest management plans 2011–2020 Tablica 4. Prikaz otvaranja potrebnim i prioritetnim šumskim cestama i primarnim protupožarnim cestama – planovi uprava za gospodarenje šumama 2011–2020. Primary fire prevention roads

Forest roads – Šumske ceste Regional Unit Uprava šuma

Necessary for optimum opening

Primarne protupožarne ceste

Priority lengths of opening in the next decade – priority areas

Necessary for optimum opening

Priority lengths of opening in the next decade – priority areas

Potrebno za optimalnu otvorenost

Prioritetne duljine otvaranja u sljedećem desetljeću – prioritetna područja

205

–

Potrebno za optimalnu otvorenost

Prioritetne duljine otvaranja u sljedećem desetljeću – prioritetna područja

370

km Tolmin Bled

–

Kranj

240

–

Ljubljana

125

–

Postojna

149

120

Kočevje

376

–

Novo Mesto

365

–

Brežice

197

–

Celje

–

Nazarje

–

Slovenj Gradec

–

Maribor

101

–

Murska Sobota

117

–

Sežana

450

140

440

325

Total – Ukupno Total forest and fire roads

2,753

758

560

375

Ukupno šumskih i protupožarnih cesta

3,313

1,133

–

basis of the present model is wide and obvious (Robek et al. 2007). Slovenia Forestry Service keeps the database on forest roads, i.e. Records of Forest Roads (RFR) that is also legally defined in the Forest Act and the Regulation on Forest Infrastructure. The user interface has also been developed to enable the maintenance of the database, designed as relation base (Beguš 2002). The data on the new constructions or increasing of forest road density are systematically acquired through RFR. The scope of new constructions is relatively small, since 2011 witnessed only 2.5 km of new forest roads (Poročilo o gozdovih za leto 2011, 2012). Therefore, the recapitulation of the model calculation has been prepared on the level of larger spatial units (forest management units), which provides the evaluation of the necessary road construction scope to meet the target density, defined in the model. According to the existing forest road construction dynamics in Slo-

226

venia, it is reasonable to include in the model a procedure, which will enable experts to define priorities for opening up inaccessible forests that can be realistically achieved during the validity of forest management plans. Apart from the included influential factors (forest sites and skidding distances), further model development should also include additional factors related to the decision-making, planning and maintainance of road network. Some of these factors are: forest road construction and maintenance costs, forest operation technology, timber skidding methods, current stand structures, relief characteristics, skidding trail construction and maintenance costs as well as restrictions and needs depending ecologic and social forest functions. Also the forest ownership or socio-economic and size category of forest ownership definitely plays an important part. Croat. j. for. eng. 34(2013)2

Planning Forest Opening with Forest Roads (217–228)

5. Conclusions – Zaključci The present model is a tool giving satisfactory results at the strategic (national) level and is also applicable at lower forest management planning levels. The model provides the possibility to determine the need for construction of new forest roads on specific areas – not only the target density. In our opinion, the most prominent added value of the model application (in view of the existing planning process) lies in its spatial determination of the needs for the additional forest opening (by forest roads). We also assume that the model can be used in different circumstances and also at the international level. Its practical applicability was proven on the national level elaborating Regional Forest Management Plans in Slovenia (Gozdnogospodarski in lovsko upravljalski načrti območij za obdobje 2011–2020, 2012). The contemporary perspective offers the following development possibilities of the present model: Þ Inclusion of relief characteristics (e.g. recognition of ridge points), Þ Exclusion of protected sections (differentiating the length of forest roads according to different levels of protection), Þ Differential model evaluation of road construction on different mostly rock surfaces, Þ Inclusion of timber skidding direction (shorter under; longer above the road with exception of cable skidding), Þ Differentiation of the smallest size of low access areas according to site potentials, Þ Estimation of forest functions according to forest road density, Þ Determination of priorities according to the actual accessibility (public cart tracks), Þ Preparation of the module for forest road density optimization (e.g. logging and skidding costs according to construction and maintenance costs of forest roads). From the point of view of the model development, there are still many possibilities for improvement of the presented tool that would also facilitate the preparation of regional forest management plans. The development result will be a more complex model that would show a more reliable and detailed first version of low access areas and calculation of the necessary forest road densities and length of new road construction.

J. Krč and J. Beguš Beguš J., 2002: Razvoj informacijskega sistema in podatkovnih zbirk za spremljavo stanja gozdnih cest. Master thesis. Univerza v Ljubljani, Biotehniška fakulteta, Oddelek za gozdarstvo in obnovljive vire, 130 p. Liu K., Sessions J., 1993: Preliminary Planning of Road Systems Using Digital Terrain Models. Journal of Forest Engieenering 4(2): 27–32. Demir, M., 2007: Impacts, management and functional planning criterion of forest road network system in Turkey. Transportation Research Part A: Policy and Practice 42(1): 56–68. Chung, W., Stückelberger, J., Aruga, K., Cundy, T., 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. Dean, D. J., 1997: Finding optimal routes for networks of harvest site access roads using GIS-based techniques. Canadian Journal of Forest Research 27(1): 11–22. Dobre, A., 1995: Gozdne prometnice. Študijsko gradivo. Oddelek za gozdarstvo Biotehniške fakultete. 71 p. Gozdnogospodarski in lovsko upravljalski načrti območij za obdobje 2011 – 2020, Zavod za gozdove Slovenija, Ljubljana 2012. Košir, Ž., 1975: Vrednotenje gozdnega prostora po varovalnem in lesnoproizvodnem pomenu na osnovi naravnih razmer. Zasnova uporabe prostora – gozdarstvo. Zavod SRS za družbeno planiranje in Inštituta za gozdno in lesno gospodarstvo pri Biotehniški fakulteti, Ljubljana, 145 p. Košir, B., Krč, J., 2000: Where to Place and Built Forest Roads – Experience From the Model. Journal of Forest Engineering 11(1): 7–19. Košir, B., Košir, Ž., Krč, J., 2006: Natural composition of tree species as a basis for model development of stumpage price. Croatian Journal for Forest Engineering 27(2): 71–80. Krč, J., 1999: Modelni izračun vpliva ceste na povečanje vrednosti donosa gozda, Zbornik gozdarstva in lesarstva 59, Ljubljana, 121–139. Lugo, A. E., Gucinski, H., 2000: Function, effects, and management of forest roads. Forest Ecology and Management 133(3): 249–262. Matthews, D., 1942: Cost control in the logging industry. McGraw-Hill Book Company, New York, 374 p. Pentek, T., Nevečerel, H., Pičman, D., Poršinsky, T., 2007: Forest road network in the Republic of Croatia – Status and perspectives. Croatian Journal for Forest Engineering 28(1): 93– 106. Pentek, T., Pičman, D., Nevečerel, H., Lepoglavec, K., Papa, I., Potočnik, I., 2011: Primary forest opening of different relief areas in the Republic of Croatia. Croatian Journal for Forest Engineering 32(1): 401–416.

6. References – Literatura

Poročilo o gozdovih za leto 2011, Zavod za gozdove Slovenije (Slovenia Forest Service) (http://www.zgs.gov.si/slo/zavod/ informacije-javnega-znacaja/letna-porocila/index.html) (Accessed: 1 August 2012).

Anderson, A. E., Nelson, J., 2004: Projecting vector-based road networks with a shortest path algorithm. Canadian Journal of Forest Research 34(7): 1444–1457.

Robek, R., Klun, J., 2007: Recent developments in forest traffic way construction in Slovenia. Croatian Journal for Forest Engineering 28(1): 83–89.

Croat. j. for. eng. 34(2013)2

227

J. Krč and J. Beguš

Planning Forest Opening with Forest Roads (217–228)

Program odpiranja gozdov z gozdnimi prometnicami, 1990, Gozdno gospodarstvo Kočevje, Kočevje 1990 Pravilnik o gozdnih prometnicah, 2009, Uradni list RS št. 4, Ljubljana 2009.

SURS: Gozd in gozdarstvo (http://www.gozd-les.com/vsebina/odkupne-cene-hlodovine) (Accessed 1. January 2013). Zakon o gozdovih, 1993, Uradni list RS no. 30.

Sažetak

Planiranje potrebne šumske otvorenosti šumskim prometnicama U radu je predstavljen model koji određuje neotvorene šumske površine temeljeći se na analizama postojeće mreže javnih i šumskih cesta provedenima pomoću GIS-a te analizama potencijalnih bonitetnih staništa u neotvorenim šumama. Unaprjeđenje gustoće cesta ostvaruje se preko dva oblika nadgradnje cestovne mreže: (1) izrada spojnih cesta do neotvorenih šumskih područja i (2) izgradnja novih šumskih cesta različite gustoće na izdvojenim područjima neotvorenih šuma. Dakle, model daje minimalnu veličinu neotvorene površine, koja se nalazi najmanje 300 metara od postojeće ceste. Izdvojena se područja neotvorenih šuma najprije analiziraju iz aspekta njihove veličine – površina od minimalno 30 ha uzeta je kao primjerena za izgradnju pristupnih cesta koje povezuju postojeću cestovnu mrežu s neotvorenom šumskom površinom. Povećanje gustoće cesta, gradnjom novih šumskih cesta, ovisi o šumskom ekosustavu pri čemu uzimamo u obzir njegov RK kao pokazatelj bonitetnog potencijala neotvorene šumske površine. S obzirom na bonitet staništa definirane su različite gustoće cestovnih mreža na izdvojenim neotvorenim šumskim područjima. Model je testiran na primjeru Uprave šuma Kočevje gdje je već bio izrađen »Program otvaranja šuma šumskim prometnicama«, koji je pri testiranju rezultata analize poslužio kao referencija. Podrobnije je model analiziran na području gospodarske jedinice Kočevska reka gdje smo rezultate modela predstavili stručnom osoblju Zavoda za šume Slovenija, Uprava šuma Kočevje, i napravili procjenu usklađenosti između rezultata modela i ocjene stručnjaka, koja je prethodno dobivena terenskom procjenom položaja šumskih cesta u analiziranoj gospodarskoj jedinici Kočevska reka. Sukladno važećim podacima šumskogospodarske osnove (šumskogospodarske i lovnogospodarske osnove za razdoblje od 2011. do 2020. godine) Slovenija ima 12 023 km šumskih cesta i 489 km protupožarnih cesta. Gustoća šumskih i protupožarnih cesta iznosi 10,6 m/ha. Šume su dodatno otvorene javnim cestama, koje u pojedinim odsjecima služe za šumarske radove, pa je tako šumska površina u Sloveniji otvorena s ukupno 29 244 km cesta, što znači da je gustoća cesta 24,8 m/ha. Predstavljena analiza pokazala je da je u Sloveniji, prema kriterijima modela, još 210 385 ha neotvorenih šumskih površina. Stručnjaci u upravama šuma Zavoda za šume u daljnjem su postupku odabrali prioritetne površine na kojima je u idućih 10 godina gospodarenja potrebno izgraditi 758 km novih šumskih cesta. Ključne riječi: šumske prometnice, gustoća cesta, šumske operacije, model, plan gospodarenja šumama

Authors’ address – Adresa autorâ:

Received (Primljeno): August 16, 2012 Accepted (Prihvaćeno): December 30, 2012

228

Assoc. Prof. Janez Krč, PhD.* e-mail: janez.krc@bf.uni-lj.si Biotechnical Faculty Department of Forestry and Renewable Forest Resources Večna pot 83 1000 Ljubljana Jurij Beguš, MSc. e-mail: jurij.begus@zgs.gov.si Slovenia Forest Service Večna pot 2 1000 Ljubljana SLOVENIA * Corresponding author – Glavni autor Croat. j. for. eng. 34(2013)2

Original scientific paper – Izvorni znanstveni rad

Fuel Consumption in Timber Haulage Radomír Klvač, Josef Kolařík, Marcela Volná, Karel Drápela Abstract – Nacrtak The paper presents an assessment of road timber transport by trucks, which included 132 truck-and-trailer units – three types of trucks (Tatra, Mercedes Benz and Iveco) with a selection of trailers in the Czech Republic. The main aim of this work was to establish the effect of hauling distance in the individual types of timber-transport units on the fuel consumption per 100 km and on the specific fuel consumption per one transported cubic metre of timber. Any decrease of fuel consumption per unit of production can enhance environmental profile of secondary transport. Freight transport recorded conspicuous changes in the last ten years, and the analysis presented in this work provides important information useful in the planning and organization of road timber transport. During the study period, obsolete and inadequate truck-and-trailer units were continuously replaced with new units, which resulted in a considerable reduction in fuel consumption per unit of production (0.5 L/m3 ub). Keywords: haulage road, timber transport, truck, truck-and-trailer unit, fuel consumption

1. Introduction – Uvod Timber transport from the roadside landing to the customer represents a very demanding phase in the chain of timber supply in terms of energy and cost. It is characterized by several specific factors that influence its implementation and differentiate it from the goods transport by trucks. In general, we can say that it is a one-way haulage, where it is very difficult or even impossible to utilize the timber-transport unit in its return run. The machines are specifically designed and can be used only to a limited extent for the haulage of other goods. Also, they have to drive a larger part of the hauling distance on forest roads. Holzleitner (2009) and Holzleitner et al. (2011) studied the operation of timber-transport units by using the GPS/GIS system and concluded that the share of their travel on forest roads was 14%. The machines often have to drive deep into the forests and have to be adapted accordingly. They have to work in difficult field conditions and therefore they are very frequently affected by them as well as by extreme seasonal weather. This is why the trucks are often equipped with the multiple-wheel drive and heavy-duty engines. These specific technological requirements considerably increase fuel consumption of timber-transport units. Svenson (2011) mentioned a range of technical factors directly affecting the fuel consumption of timberCroat. j. for. eng. 34(2013)2

transport units and classified them into the following groups: vehicle characteristics, trailer characteristics, road geometry, road surface, goal speed, gear change, driving behavior, weather and road surface conditions. The above factors of technical and technological character have a considerable influence on the average fuel consumption of timber truck-and-trailer units, which may be double as compared with the common road goods transport by trucks (Devlin 2010). The number of information systems specialized in goods or bus transportation is high in the Czech Republic but the number of information systems specialized in timber transport is low. Hauling timber from the roadside landing features problems such as heterogeneity of the transported material, difficult utilization of vehicles at their return run, seasonal character of operations, climatic effects – all these resulting in a high rate of »empty« drives. Data processing, transport optimization and necessity of flexible response to unexpected situations put high requirements both on the information system and on timber haulage managers. This is why an information system was designed, which tries to respond to the absence of information systems in the field of timber haulage (Klvač 2006). From the economic point of view, the share of timber haulage in total timber supply chain costs may reach more than 30% (Favreau 2006). He mentions that

229

R. Klvač et al.

transport is the biggest cost item in round wood costs in Canada. In Sweden, Svenson (2011) says that 35% of total transportation costs are related to the fuel consumption of timber trucks. Economic data provided by the contractor of timber-transport units, which were the subject of our study, demonstrated that diesel fuels accounted for the highest share in total costs (30%), followed by depreciation and leasing (20%) and repairs and maintenance (16%). Wages (15%), overhead costs (13%) and other costs (5%) followed. The objective of implementation of the information system was to conduct a basic analysis of individual types of timber-transport units and based on the acquired data to find primary relations affecting transport efficiency and thus to find ways how to reduce the cost of timber haulage. Any decrease of fuel consumption per unit of production can enhance environmental and economy profile of secondary transport. As the fuel cost makes the largest part of total timber haulage costs, the aim of this work is to analyze the fuel consumption in the individual types of truck-and-trailer units used in timber transport. Any replacement of obsolete and inadequate truck-and-trailer units by new more efficient units can result in a considerable reduction of fuel consumption per unit of production.

2. Material and methods – Materijal i metode A »tailor made« information system was designed in 2003, which can receive orders placed by customers, support the decision-making process of dispatchers by using suitable truck-and-trailer units (TTU), make records of hauling performance, monitor production in progress and summarize data in the form of databases. In 2004, the system was characterized in the form of diagrams so that designers would be capable of meeting customer requirements (Klvač 2006). This information system was designed for larger companies with a greater number of vehicles dislocated on remote workplaces. All workplaces had an access to the system via client and worked with data on multiple levels related to the position in company or business interrelationship. Each position/client type had centrally set rights and responsibilities in the system. A timber transport company implemented the system at the beginning of 2005 and data on each individual transportation case started to be recorded from the end of the same year. The data was summarized for each TTU in monthly intervals for purposes of analytical assessment by the company management. The monthly indicators of TTUs were used in this study.

230

Fuel Consumption in Timber Haulage (229–240)

The structure of the assessed data related to this study was as follows: Þ Truck-and-trailer unit, inventory number provided for non-commutability of data, Þ TTU operational centre, Þ Trailer, inventory number, Þ Total travel distance, km, Þ Travel unloaded, km, Þ Travel loaded, km, Þ Backhauling, % of kilometers driven loaded, Þ Volume of transported timber, m3 ub; softwood and hardwood, Þ Number of loads per month and per day, Þ Average size of load, m3 ub, Þ Average hauling distance – one way distance, km, Þ Fuel consumption in liters per month. Parameters that were calculated based on the above data were as follows: Þ Average fuel consumption per unit of production, llm3 ub Þ Average fuel consumption per 100 km, ll100 km All data were checked at first and records containing gross errors caused by human factor at recording were eliminated. Then the data were imported and organized within the spreadsheet software (Microsoft Excel) and subsequently summarized for individual types of TTUs. In the period 2005 – 2009, considerable changes occurred in the fleet of timber transport units with obsolete TTUs being put out of operation and replaced by new TTUs where necessary. Old and technically unfit Liaz TTUs were taken out of service first. As the amount of data on these TTUs was not representative, the Liaz type of TTU was not statistically evaluated in this study. Types of truck-and-trailer units assessed in this study were Iveco (represented by models ASTRA, MP260 and STRALIS), Tatra (represented by Tatra 815 only), Mercedes Benz (models 3344, 3341, 2644 and 3348). The data were aggregated and analyzed according to truck manufacturers. The initial analysis was made with the use of pivoting (contingency) tables and graphs. GraphPad Prism 5 (Motulsky 2007) was used for non-linear regressions. The software enables a very flexible choice of the regression model, it has very good graphical capabilities and provides the possibility to compute and draw confidence intervals of the model. Prism 5 can eliminate outliers with the ROUT method (Motulsky and Brown 2006). This method is based on a new robust non-linear regression combined with outlier rejection. It is an adaptive method that gradually becomes more robust as the method proceeds. Press et al. (1988) based their Croat. j. for. eng. 34(2013)2

Fuel Consumption in Timber Haulage (229–240)

R. Klvač et al.

2 548. The total number of TTUs assessed in the period 2005 – 2009 was 134 and the units were operated at different places in the Czech Republic. In the period 2003 – 2004, we monitored only 21 trucks; this number increased in 2005 to 90. In the following years, the fleet was gradually renewed and some old vehicles were put out of operation. This is why the number of trucks monitored in 2006, 2007 and 2008 was 87, 80 and 71, respectively. In 2009, the process of renewal was completed and the final number of trucks was 51 of which 45 were Mercedes Benz. In the period under study (Table 1), more than 3.4 million cubic meters of timber were hauled from the roadside landing to the conversion depot, directly to customers or to the siding railway. They were recorded and assessed - softwood accounted for 92% and hardwood for 8% of the total volume. Total diesel consumption of monitored TTUs was 6.8 million liters. The fuel consumption is not broken down to the amount used directly in timber haulage and the amount used indirectly, i.e. driving to the working place or driving to the workshop for repair. The share of »empty kilometers« in the total number of driven kilometers was 47%. Average backhauling of TTUs (loaded vehicles) was 53%. The presented values represent and summarize a total of 136 292 cases of timber transport. The average hauling distance was changing in the course of years depending on activities of the company operating the trucks. From 2005, the number of timber yards was decreasing and the amount of timber handled at the roadside landing was increasing as well as the timber haulage from the landing directly to the customer. The average hauling distance was

robust fitting method on the assumption that variation around the curve follows a Lorenzian distribution rather than a Gaussian distribution. The Marquardt non-linear regression algorithm was adapted to accommodate the assumption of a Lorenzian (rather than Gaussian) distribution of residuals. After fitting a curve using robust non-linear regression, a threshold is needed for deciding when a point is far enough from the curve to be declared an outlier. All methodology is described in detail in Motulsky and Brown (2006). The authors state that their method identifies outliers from non-linear curve fits with reasonable power and few false positives (less than 1%). In all cases, the logarithmic function used for the regression model was in the following form: y = a  ln(x) + b (1) Where: x explaining (independent) variable, y explained (dependent) variable, a, b coefficients. The respective statistical assessments include a, b coefficients established by the regression analysis, 95% confidence interval (shaded in the graphs), R2 – determination coefficient, number of analyzed points and number of outliers. The respective dependencies are presented in summary diagrams in Microsoft Excel, in which only regression curves were plotted.

3. Results – Rezultati The total number of assessed records (i.e. monthly performances of various truck-and-trailer units) was Table 1 Mean values for all monitored TTU types Tablica 1. Značajke promatranih kamionskih skupova

Softwood – Crnogorica

3 161 533

Hardwood – Bjelogorica

256 638

Empty kilometers – Vožnja praznim kamionom, km

5 172 109

Volume of transported timber, m3 – Obujam transportiranoga drva, m3

3 418 171

Total distance, km – Ukupno prijeđena udaljenost, km

11 032 534

Fuel consumption, l – Potrošnja goriva, l

6 811 604

–

136 292

–

Average fuel consumption, l/m3 – Prosječna potrošnja goriva, l/m3

2.19

–

Average consumption, l/100 km – Prosječna potrošnja goriva, l/100 km

67.4

–

Average hauling distance*, km – Prosječna udaljenost turnusa*, km

45.05

–

Number of cycles – Broj turnusa

Kilometers driven loaded – Vožnja punim kamionom, km 5 860 425

* One way distance – * U jednom smjeru

Croat. j. for. eng. 34(2013)2

231

R. Klvač et al.

Fuel Consumption in Timber Haulage (229–240)

Table 2 Trends of important indicators in all TTU types in the studied period Tablica 2. Trendovi i važne karakteristike promatranih kamionskih skupova u vremenu istraživanja Year – Godina

2005

2006

Average fuel consumption, l/m – Prosječna potrošnja goriva, l/m

2.32

2.06

1.87

2.67

3.08

Average fuel consumption, l/100 km – Prosječna potrošnja goriva, l/100 km

69.51

68.4

70.94

61.22

61.36

Average hauling distance*, km – Prosječna udaljenost turnusa*, km

39.31

37.6

36.87

65.76

74.21

Average size of load, m3 – Prosječni obujam tovara, m3

20.59

23.45

25.96

26.13

27.57

Average backhauling**, % – Prosječna transportna udaljenost punoga kamiona**, %

2007

2008

2009

* One way distance – * U jednom smjeru ** % of kilometers driven loaded – ** Udio s obzirom na udaljenost turnusa

Table 3 Outputs and indicators of individual TTU types Tablica 3. Tehničke karakteristike promatranih kamionskih skupova TTU type – Model kamionskoga skupa

IVECO

TATRA

MB*

Average fuel consumption, l/m – Prosječna potrošnja goriva, l/m

2.26

1.93

2.71

Average fuel consumption, l/100 km – Prosječna potrošnja goriva, l/100 km

66.74

72.25

58.31

Average hauling distance**, km – Prosječna udaljenost turnusa **, km

48.97

28.98

76.11

Average loads per day – Prosječan broj turnusa po danu

2.96

3.25

2.98

25.21

22.84

28.38

Average size of load, m – Prosječan obujam tovara, m

Average backhauling*** – Prosječna transportna udaljenost punoga kamiona *** Total, km – Ukupno, km

903 845

4 014 736

6 055 543

Volume of hauled timber, m3 – Obujam transportiranoga drva, m3

285 683

1 701 892

1 408 446

* MB: Mercedes-Benz ** One way distance – **U jednom smjeru *** % of kilometers driven loaded – *** Udio s obzirom na udaljenost turnusa

increasing towards the end of the study period – see Table 2. The lowest distance was achieved in 2007 due to the Kyrill gale disaster when a substantial part of all TTUs were concentrated to work in affected areas, where the trucks mostly transported timber over short hauling distances, which considerably affected the annual average hauling distance. The average size of load was markedly increasing during the years thanks to changes in the fleet because the newly used TTUs of Mercedes Benz type featured a considerably higher capacity than the other assessed TTU types (Table 3). Table 2 shows that the increasing average hauling distance resulted in the increasing average fuel consumption per unit of production and that the fleet renewal brought a gradual decrease in the fuel consumption per 100 km. In 2008 and 2009, when the Mercedes Benz type of TTU started to dominate the fleet, the average fuel consumption per 100 km dropped dramatically by 9%. A detailed survey of indicators and outputs by individual types of timber

232

transport units is presented in Table 3, where the prominent indicator is the load size. Backhauling considerably affects transport efficiency; average backhauling increased depending on average hauling distance, which was favorably affected by the easier coordination of loads by dispatchers. Over short hauling distances, timber transport from the forest is operated more or less in one-way direction; backhauling is often unrealistic and the trucks are additionally burdened by driving to their workplace and to repair or maintenance workshops. This is why its efficiency is below 50%. With the increasing of the hauling distance, the possibility of finding suitable backhauling increases and the effect of driving to the workplace or repair is minimized (Fig. 1). Extremely low values mostly resulted from loading into wagons (when the vehicle was used for loading wagons) and its number of empty kilometers increased due to frequent drives within the terminal (timber yard). On the other hand, extremely high values resulted from a nearly ideal relation when empty kiloCroat. j. for. eng. 34(2013)2

Fuel Consumption in Timber Haulage (229–240)

R. Klvač et al.

Fig. 1 Dependence of backhauling on hauling distance (all TTU types) Slika 1. Udio vožnje punim kamionom po turnusu (svi promatrani modeli kamionskih skupova)

Fig. 3 Relation between fuel consumption per 100 km and hauling distance for the Tatra type of TTU Slika 3. Odnos između potrošnje goriva na 100 km i duljine turnusa za kamionski skup Tatra

Fig. 2 Relation between fuel consumption per 100 km and hauling distance for the Iveco type of TTU Slika 2. Odnos između potrošnje goriva na 100 km i duljine turnusa za kamionski skup Iveco

Fig. 4 Relation between fuel consumption per 100 km and hauling distance for the Mercedes-Benz type of TTU Slika 4. Odnos između potrošnje goriva na 100 km i duljine turnusa za kamionski skup Mercedes-Benz

meters represented only driving on forest roads and very short travels for another load. Details of regression analyzes were as follows: Best-fit values a = 8.352,

b = 19.19; Std. Error a = 0.1920, b = 0.6994; 95% Confidence Intervals a = 7.976 to 8.728, b = 17.82 to 20.56; R square 0.4500; Outliers (excluded, Q = 1.0 %) 3.

Croat. j. for. eng. 34(2013)2

233

R. Klvač et al.

Fuel Consumption in Timber Haulage (229–240)

Table 4 Results of regression analyses of the relation of fuel consumption per 100 km and hauling distance Tablica 4. Rezultati regresijske analize potrošnje goriva na 100 km i duljine turnusa TTU type Model kamionskoga skupa

* Regression coefficients of equation

Border coefficients, 95%

* Regresijski koeficijenti jednadžbe

Granični koeficijenti, 95 %

y = a × ln(x) + b

Confidence Intervals – Faktor pouzdanosti

Range of × value Raspon × vrijednosti

Iveco

–15.47

123.4

–17.14 ; –13.81

117.2; 129.6

0.6102

10 – 132

Tatra

–13.96

116.7

–15.30 ; –12.61

112.3 ; 121.0

0.2390

10 – 131

–10.42

101.7

–11.25 ; –9.576

98.12 ; 105.3

0.4397

12 – 178

* x – hauling distance – Duljina turnusa y – fuel consumption per 100 km – Potrošnja goriva na 100 km

Table 5 Results of regression analyses of the relation of fuel consumption per unit of production (m3) and hauling distance Tablica 5. Rezultati regresijske analize potrošnje goriva po jedinici proizvodnje (m3) i duljine turnusa TTU type Model kamionskoga skupa

* Regression coefficients of equation

Border coefficients, 95%

* Regresijski koeficijenti jednadžbe

Granični koeficijenti, 95 %

y = a × ln(x) + b

Confidence Intervals – Faktor pouzdanosti

Range of × value Raspon × vrijednosti

Iveco

0.9842

–1.399

0.8887; 1.080

–1.756 ; –1.043

0.6601

10 – 132

Tatra

1.335

–2.444

1.280; 1.391

–2.624 ; –2.265

0.6311

10 – 131

1.531

–3.749

1.460; 1.601

–4.048 ; –3.450

0.7025

12 – 178

* x – hauling distance – Duljina turnusa y – fuel consumption per unit of production, m3 – Potrošnja goriva po jedinici proizvodnje, m3

Fig. 5 Relation between fuel consumption per 100 km and hauling distance for all types of TTUs Slika 5. Odnos između potrošnje goriva na 100 km i duljine turnusa za sve promatrane kamionske skupove

234

Croat. j. for. eng. 34(2013)2

Fuel Consumption in Timber Haulage (229–240)

R. Klvač et al.

Fig. 6 Relation between fuel consumption per unit of production (m3) and hauling distance for the Iveco type of TTU Slika 6. Odnos između potrošnje goriva po jedinici proizvodnje (m3) i duljine turnusa za kamionski skup Iveco

Fig. 7 Relation between fuel consumption per unit of production (m3) and hauling distance for the Tatra type of TTU Slika 7. Odnos između potrošnje goriva po jedinici proizvodnje (m3) i duljine turnusa za kamionski skup Tatra

3.1 Average fuel consumption in relation to driven distance including the effect of uploading and unloading and proportion of time spent on forest roads – Prosječna potrošnja goriva po prijeđenom kilometru uključujući utovar, istovar te udio vožnje šumskom cestom

roads due to harsh terrain conditions, limited speed (lower gear) and worse road quality that decreases with the increasing hauling distance. None of these

Average fuel consumption per 100 km is markedly higher in the older TTU types such as Iveco and Tatra in particular (see Table 3). It is also synergy affected by uploading and unloading times as well as by the hauling distance. If the hauling distance is shorter, the average consumption per 100 km is markedly higher than over longer distances due to the effect of uploading and unloading. During the uploading and unloading, the engine of the truck (energy source) drives the hydraulic crane and the consumption of fuel thus increases without a change in driven kilometers. According to company workers (personal communication), the loading time was different when loading stems or timber shortened to transportation length (max. 35 min.) and when loading stacked assortments up to 8 m (max. 50 min.). The second effect is the proportion of time spent on forest roads. The shorter journey meant a higher proportion of travel time spent on forest roads. The trucks have a higher fuel consumption on forest Croat. j. for. eng. 34(2013)2

Fig. 8 Relation between fuel consumption per unit of production (m3) and hauling distance for the Mercedes-Benz type of TTU Slika 8. Odnos između potrošnje goriva po jedinici proizvodnje (m3) i duljine turnusa za kamionski skup Mercedes-Benz

235

R. Klvač et al.

Fuel Consumption in Timber Haulage (229–240)

Fig. 9 Relation between fuel consumption per unit of production (m3) and hauling distance for all types of TTUs Slika 9. Odnos između potrošnje goriva po jedinici proizvodnje (m3) i duljine turnusa za sve promatrane kamionske skupove

two aspects can be eliminated to determine the influence of each separately. In other words, as the two aspects are inseparable part of timber haulage, the assessment was made including the impact of them both. Both effects also correspond to the average number of daily delivered loads with respect to hauling distance i.e.: 4 deliveries at 10.7 km average hauling distance, 3 at 38 km and 2 at 200 km, respectively. Regression equations of fuel consumption for the respective TTUs are presented in Figs. 2 – 4 including discerned outliers and including confidence interval of 95% reliability. The regression equations are plotted in a comprehensive graph (Fig. 5) for the comparison of individual TTU types. The regression curves are drawn in the interval of hauling distances in which TTU types were operating. The results of regression analyses for individual types of timber transport units are presented in Table 4.

3.2 Average fuel consumption per unit of production (hauled cubic meter) – Prosječna potrošnja goriva po jedinici proizvodnje (prevezeni kubni metar) In this case, too, the respective types of truck-andtrailer units were assessed separately (Figs. 6 – 8). The average fuel consumption per unit of production (m3) was conspicuously different in the individual TTU

236

types, the reason for the difference being mainly the effect of hauling distance and the size of TTU load. The greater the hauling distance, the higher was the fuel consumption per unit of production; at the same time, the greater the vehicle capacity, the lower was the average fuel consumption. The two factors act in synergy and there are other impacts to be expected, too, such as seasonal character of the work, effect of the operator, etc. Results of regression analyses for individual types of timber transport units are presented in Table 5. The comprehensive diagram in Fig. 9 shows regression equations for the respective types of timber transport units. The regression curves are plotted only within the hauling distance interval in which the values used in the regression analysis occurred. Tatra type trucks showed unambiguously the highest fuel consumption per unit of production.

4. Discussion and conclusion – Rasprava sa zaključcima The above graphs (Figs. 2 – 5) show the dependence of fuel consumption per 100 km on average hauling distance of the individual TTU types. The average hauling distance ranged from 10 – 180 km. Older Iveco and Tatra trucks in particular had a considerably higher fuel consumption per 100 km, which Croat. j. for. eng. 34(2013)2

Fuel Consumption in Timber Haulage (229–240)

supposedly resulted from the fact that their hauling distances were relatively short (38 km) and loading and unloading was more frequent. Thus, due to more frequent loading and unloading, the fuel consumption increased although it did not show in the travel distance. The average fuel consumption per 100 km of Mercedes-Benz TTUs was markedly lower, because the hauling distances were apparently higher. Another aspect affecting the fuel consumption together with this factor was the proportion of driving on forest roads, which decreases with the increasing hauling distance i.e. loading is limited within one working day. Fig. 5 shows that the fuel consumption per 100 km decreases with the increasing hauling distance. The third very important factor is the engine category. Mercedes Benz Trucks were Euro 3 and Euro 5 class, which should guarantee lower fuel consumption. However, this is not as visible as expected and further detailed analysis of Mercedes Benz truck is necessary. Other impacts, such as the seasonal character of work, locality (road quality, relations), human factor in loading/unloading, equipment operators, drivers, etc. could not be identified but their influence can be anticipated at least to some extent. The authors consider that the volume of data is representative for estimating the mean values of fuel consumption. Svenson (2011) informs that in Sweden the average fuel consumption per 100 km is 58 liters but does not mention the hauling distance, which could be corresponding to 65 km according to the results of our study. Although the value is highly speculative, it might be realistic for such a vast country as Sweden even if it is by 40% higher than the average hauling distance of 45 km established in this study. It is however fully comparable with values recorded in 2008 and 2009, when the timber transport company that provided the data focused on longer hauling distances. Similar conditions as in the Czech Republic can be expected in Austria, where Holzleitner (2009) claims the average hauling distance of 51 km, which is in line with the values detected in this study. Fuel consumption can be reduced in different ways. Considerate driving may considerably reduce the fuel consumption. By a program that can monitor the driving regime, the Tom Tom Corporation can identify inappropriate driving manners and demonstrate a more economical regime (personal communication Tom Tom). Lofroth and Lindholm (2005) mention further possibilities of fuel economy, e.g. that haulage trucks can reduce their fuel consumption by 5 – 10% simply by fitting a wind deflector and by removing all unnecessary items such as signCroat. j. for. eng. 34(2013)2

R. Klvač et al.

boards, extra air horns, extra lamps and other unnecessary accessories. Average fuel consumption per unit of production (m3) is first of all affected by the hauling distance and by the load size – the two factors acting in synergy. The higher is the vehicle capacity, the lower is the average consumption per unit of production, and the greater is the hauling distance, the higher is the fuel consumption per unit of production. Further to the above, the Tatra TTUs would have the lowest fuel consumption per unit of production if the average values of TTU types from global assessment were compared without a more detailed analysis (see Table 3). Nevertheless, this view of the problem would be rather naïve, because these are the average values for the entire 5-year monitoring period and they are related to anaverage hauling distance calculated for the whole period of the study. Therefore, it is necessary to compare the average fuel consumption based on data presented in Fig. 9. The Tatra truck-and-trailer unit has the highest average fuel consumption per unit of production in relation to the hauling distance, the likely reason being the average size of load but also the construction of the machine, which is designed for difficult, inaccessible terrains and is fitted with older engine types. By contrast, the Mercedes-Benz TTUs exhibited the highest average fuel consumption per unit of production (approx. 3 liters per cubic meter) in the global assessment (Table 3), which resulted from the long hauling distance in the monitored period. However, it can be concluded from Fig. 9 that the MercedesBenz TTUs are more economical in terms of fuel consumption per unit of production with the load size playing once again the most important role. The load size in the Mercedes-Benz TTUs was approximately 5 m3 greater than in the Tatra TTUs. Further to the above, it can be concluded that the timber transport units cannot be evaluated only according to summarized data (Table 3) but that more detailed analyses, such as in Figs. 6 – 8, are absolutely necessary. The issue of relations between the individual indicators is very complex and it would be certainly useful to conduct a detailed survey within the respective types of trucks e.g. in relation to hauling distance, loading capacity, trailer type or region in which the TTU operated. All activities connected with the detailed characterization of these relations are focused on fuel economy. This direction is also obvious from the activities of FP Innovation, where the socalled StarTrack was designed aimed at reducing machine weight and providing maximum loading capacity. The specifications placed on the research

237

R. Klvač et al.

truck considered the local operating conditions and included the following requirements: heavy-duty aluminum rims, smaller fuel tank (but the right size for one shift), aluminum cab protector, central tire inflation (CTI), on-board weighing, in-cab auxiliary heater, on-board computer, single tractor frame rail, lightweight multi product semi-trailer and road maintenance management system. All these innovations resulted in the following improvements: Þ The Star Truck had a higher payload by 9.8% and consumed only 1% more fuel, Þ The Star Truck transported 8.6% more products per liter of fuel, Þ The Star Truck fuel cost per ton was by 8% lower than in the control truck, Þ Tire wear was by 40% lower in the Star Truck due to CTI. (Anon. 2012) The reduced fuel consumption per unit of production aims at mitigating the environmental pollution caused by emissions of greenhouse gases (GHGs). Fuel consumption by trucks is one of the largest contributors of these emissions. Komor (1995) informs that in the U.S.A., trucks account for over 80% of the freight energy use and 19% of US oil consumption. Plans to improve the technical efficiency through new technologies, careful driving and optimal driving conditions can increase the efficiency by 50 to 70%. Bandivadekar et al. (2008) believe that the increase in the consumption of oil for transport in the U.S.A. is a challenging environmental problem that needs to be addressed in terms of reducing fuel consumption based on drivers’ behavior rather than concentrating on the improvement of vehicle performance through new propulsion technologies and new fuels in the shorter term. Other methods leading to reduced fuel consumption are decision support systems and use of telemetry in combination with GPS/GIS. An example may be the study published by Devlin et al. (2007). The amount of timber extracted in the Czech Republic per year is about 15 million m3. Adequate fleet changes, improved optimization and technical modifications may be used to reduce fuel consumption per unit of production by 0.5 – 1.0 liter. This would bring a reduction of fuel consumption in timber haulage by 0.75 – 1.5 million liters of oil in the Czech Republic. Devlin (2010) claims that each liter of oil burnt in the truck-and-trailer unit is responsible for 2.67 kg of carbon dioxide emitted into the atmosphere. Based on the emission factors established by Lewis (1997), we can state that each liter of oil is responsible for additional 0.25 kg CO2 emitted during the production and distribution. Thus, saving 1.5 million liters of oil equivalent would result in a reduction of CO2 emis-

238

Fuel Consumption in Timber Haulage (229–240)

sions into the atmosphere of 4.4 million tons. The unambiguous conclusion is that optimization and use of adequate TTU types in timber transport from the roadside landing can significantly contribute to the mitigation of the negative impact of forest machinery on the environment.

Acknowledgement The paper was prepared within the framework of research projects of the Ministry of Education of the Czech Republic nos. MSM 6215648902 and OC10041, Mendel University internal project IGA and of the COST Action FP0902. The authors also wish to express their thanks to contractors for providing the possibility of data collection.

5. References – Literatura Anon., 2012: FP Inovation. Timber Transport Research – FERIC’s Star Truck Project. Logging-on newsletter. Available on http://www.loggingon.net/timber-transport-researchferics-star-truck-project_news_op_view_id_43 Bandivadekar, A., Cheah, L., Evans, C., Groode, T., Heywood, J., Kasseris, E., Kromer, M., Weiss, M., 2008: Reducing the fuel use and greenhouse gas emissions of the US vehicle fleet. Energy Policy 36(7): 2754–2760. Devlin, G., 2010: Fuel consumption of timber haulage versus general haulage. Harvesting/Transportation No. 22. COFORD, 6 p. Devlin, J. G., McDonnell, K., Ward, S., 2007: Timber haulage routing in Ireland: an analysis using GIS and GPS. Journal of Transport Geography 16(1): 63–72. Favreau, J., 2006: Six key elements to reduce forest transportation cost. FERIC. Available on http://www.forac.ulaval.ca/ fileadmin/docs/EcoleEte/2006/Favreau.pdf Holzleitner, F., 2009: Analyzing road transport of roundwood with a commercial fleet manager. In: Prknová H (ed) Formec 2009. Kostelec nad Černými lesy: Czech University of Life Sciences Prague, p. 173–181. ISBN 978-80-213-1939-4. Holzleitner, F., Kanzian, Ch., Stampfer, K., 2011: Analyzing time and fuel consumption in road transport of round wood with an onboard fleet manager. Eur J Forest Res 130(2): 293– 301. Klvac, R., 2006: Draft of Information system for timber haulage. In Charvát K (ed) Information Systems in Agriculture and Forestry. Praha: ČZU Praha, p. 1-8. ISBN 80-213-1494-X. Komor, P., 1995: Reducing energy use in US freight transport. Transport Policy 2 (2): 119–128. Lofroth, C., Lindholm, E. L., 2005: Reduced fuel consumption on roundwood haulage rigs. Skogforsk. Resultat no. 23. Croat. j. for. eng. 34(2013)2

Fuel Consumption in Timber Haulage (229–240)

R. Klvač et al.

Motulsky, H. J., 2007: GraphPad Prism Version 5.0. Regression Guide. GraphPad Software. Inc.. San Diego. 294 p. Motulsky, H. J ., Brown, R. E., 2006: Detecting outliers when fitting data with nonlinear regression – a new method based on robust nonlinear regression and the false discovery rate. BMC bioinformatics. Available on http://www.ncbi.nlm.nih. gov/pubmed/16526949.

Press, W. H., Teukolsky, S. A., Vettering, W. T., Flannery, B. P., 1988: Numerical Recipes in C. The Art of Scientific Computing. New York. Cambridge University Press. Svenson, G., 2011: The impact of road characteristics on fuel consumption for timber trucks. In Ackerman P, Ham H, Gleasure E (eds) Proceedings of 4th Forest Engineering Conference: Innovation in Forest Engineering – Adapting to Structural Change. Stellenbosch University, p. 172. ISBN 9780-7972-1284-8.

Sažetak

Potrošnja goriva pri prijevozu drvnih sortimenata U ovom je radu istraživana pristupačnost drvnih sortimenata prijevozu kamionskim skupovima, a istraživala su se 132 kamionska skupa i tri modela kamiona (Tatra, Mercedese Benz i Iveco) s različitim vrstama kamionskih prikolica. Svako smanjenje potrošnje goriva po jedinici proizvodnje može povećati okolišni i ekonomski profil sekundarnoga prijevoza. S obzirom na to da na gorivo otpada najveći dio troškova koji nastaju pri prijevozu drvnih sortimenata, cilj je ovoga rada bio analizirati potrošnju goriva promatranih kamionskih skupova korištenih za prijevoz. Svaka zamjena zastarjeloga i neučinkovitoga kamionskoga skupa novim učinkovitijim kamionskim skupom može rezultirati značajnim smanjenjem potrošnje goriva po jedinici proizvodnje. Glavni je cilj ovoga rada bio ustanoviti na koji način prijevozna udaljenost (duljina jednoga turnusa) kod promatranih kamionskih skupova utječe na potrošnju goriva na 100 km te na specifičnu potrošnju goriva po prevezenom kubnom metru drva. Dizajniran je informacijski sustav koji može primati narudžbe od naručitelja i koji pruža potporu dispečerima pri donošenju odluka da bi se odabrao najpogodniji kamionski skup. Sustav također bilježi podatke o pojedinom turnusu, zbraja ih te ih pohranjuje u baze podataka. Početna je obrada podataka napravljena usporedbom velikoga broja tablica i grafikona. Za nelinearnu regresiju koristili smo se programom GradhPad Prism 5. Taj program omogućuje vrlo fleksibilan izbor regresijskoga modela, ima vrlo dobre grafičke mogućnosti i moguće je ubaciti i ucrtati intervale pouzdanosti pojedinih modela. Navedeni program eliminira ekstreme metodom »ROUT«. U vrijeme istraživanja više od 3,4 milijuna kubnih metara drva prevezeno je od pomoćnoga stovarišta do glavnoga stovarišta, krajnjega korisnika ili do željezničke pruge. U ukupnom obujmu prevezenoga drva udio je crnogorice bio 92, a bjelogorice 8 %. Ukupan utrošak goriva za promatrane kamionske skupove iznosio je 6,8 milijuna litara. Na potrošnju goriva po jedinici proizvodnje (m3) najviše utječu duljina turnusa i obujam tovara. Ta dva čimbenika djeluju u sinergiji. Što je veći obujam tovarnoga prostora kamionskoga skupa, manja je prosječna potrošnja goriva po jedinici proizvodnje, dok s druge strane, što je veća udaljenost pojedinoga turnusa, veća je i prosječna potrošnja goriva po jedinici proizvodnje. Zastarjeli i neadekvatni kamionski skupovi tijekom istraživanoga razdoblja stalno su zamjenjivani novim i učinkovitijim, zbog čega je primijećeno značajno smanjenje prosječne potrošnje goriva (0,5 l/m3) po jedinici proizvodnje. Smanjenje potrošnje goriva po jedinici proizvodnje u konačnici znači smanjenje emisije stakleničkih plinova te ublažavanje štetnoga utjecaja na okoliš. Sagorijavanjem jedne litre goriva u motoru kamionskoga skupa u atmosferu se ispušta 2,67 kg ugljičnoga dioksida te bi se smanjenjem potrošnje goriva za 1,5 milijuna litara smanjila i emisija ugljičnoga dioksida u atmosferi za 4,4 milijuna tona. Nedvosmisleni je zaključak ovoga rada da se pri odabiru kamionskih skupova za prijevoz drvnih sortimenata, tj. njihovom optimizacijom, može značajno pridoniCroat. j. for. eng. 34(2013)2

239

R. Klvač et al.

Fuel Consumption in Timber Haulage (229–240)

jeti ublažavanju negativnih utjecaja šumskih strojeva na okoliš. Cestovni je promet u posljednjih deset godina zabilježio velike promjene, a analiza predstavljena u ovom radu daje važne informacije korisne u planiranju i organizaciji cestovnoga prijevoza drvnih sortimenata. Ključne riječi: šumska cesta, prijevoz drvnih sortimenata, kamionski skup, potrošnja goriva

Authors’ address – Adresa autorâ:

Received (Primljeno): February 18, 2013 Accepted (Prihvaćeno): August 06, 2013

240

Assoc. prof. Radomír Klvač, PhD.* e-mail: klvac@mendelu.cz Mr. Josef Kolařík e-mail: j.kolarik@email.cz Mrs. Marcela Volná e-mail: 12796@node.mendelu.cz Assoc. prof. Karel Drápela, PhD. e-mail: karel.drapela@mendelu.cz Mendel University in Brno Faculty of Forestry and Wood Technology Department of Forest and Forest Products Technology Zemedelska 3 613 00 Brno CZECH REPUBLIC * Corresponding author – Glavni autor Croat. j. for. eng. 34(2013)2

Original scientific paper – Izvorni znanstveni rad

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects of their Work Matija Landekić, Ivan Martinić, Matija Bakarić, Mario Šporčić Abstract – Nacrtak This paper provides the results of an applied research of forestry machine operators related to their work ability index (WAI) and some ergonomic aspects of their everyday work. A questionnaire on work environment and working ability was conducted in the year 2012 and included machinery operators employed in the state forestry company Croatian Forest Ltd. and in private forestry companies. Descriptive statistics and comparisons have been carried out regarding work ability index and frequency response of the respondents. The first part of the results presents a) profile of respondents b) organization of operators’ work activities and their education, c) impact of tiredness and d) impact of psychological and social factors on operators’ work ability. The second part of the results presents a) work ability results in relation to demographic categories of respondents and b) examination of differences between work ability indexes by groups of descriptive variables. Regarding the educational aspect of the sampled machine operators, the results showed insufficient level of adequate specialized education. Higher level of mental demands required to perform the job was rated with the operators employed in private companies. Demographic parameters of the respondents negatively affect the working and functional ability of forestry machine operators, and the value of operators’ WAI decreased considerably within the groups depending on work experience in the private forestry sector. Keywords: forestry, machine operators, work ability index, working environment

1. Introduction – Uvod Skidding and/or forwarding of timber are considered to be technically demanding, the most expensive, and together with felling and cutting, the riskiest everyday operations in forestry. Within the process of timber harvesting, a group of work operations during skidding/forwarding, which are also called the primary transport (Poršinsky 2005), are defined as removal of whole trees or their parts from the felling site (stump) to the roadside landing using a forest machine. Requirements for volume and quality of forest works, especially works concerning skidding and forwarding, are ever increasing. At the same time, forest work and supporting logistics processes become more complex, and certification of forest management goes with a wide number of requirements on the quality of performance, especially in terms of stand protection Croat. j. for. eng. 34(2013)2

and biodiversity conservation (Kostenholtz et al. 2008), as well as social and safety standards of workers in direct production. Development of forest technology at the end of the 20th century and structural changes in forestry sector (market growth, denationalization, etc.) have led to a significant reduction in the number of employees in the state forestry sector and also to the development of new business models in some segments, such as exploitation of forests, where entrepreneurship plays a dominant role. Forest contractors become an important link between forest owners and wood industry (Šporčić 2005) in Europe and in Croatia. In the period 2000–2010, there has been a visible increase in contractors’ services in the state forests of the Republic of Croatia (RC). Contractors have a constant share of 41.83% in skidding/forwarding operations, which means that

241

M. Landekić et al.

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

the company Croatian Forests Ltd. Zagreb (in charge for the management of state forests) extracts with its own mechanization approximately 58.17% of timber (Landekić et al. 2011b). The volume of services that private contractors provide to private forest owners in Croatia is not registered and no reliable data concerning these activities are available. Engagement of forest contractors and transition to contract work brings some benefits (flexibility, better financial result, better work performance due to specialization, etc.), but also defects such as insufficient investment in equipment and training, questionable professional level of work, low safety level, questionable effectiveness of workers’ health care, ineffective labor inspection (Šporčić et al. 2009) and potentially less skillful machine operators in the private sector. The key role for ensuring competitiveness of timber extraction (skidding/forwarding), but also for achieving a satisfying level of safety and efficiency in forest production, lies in work ability and working techniques used directly by forestry machine operators (Martinić et al. 2011). Vaguely specified legal responsibility of forest contractors and inefficiency of labor inspection leads to negligence of basic safety standards, which may result in inadequate working environment and reduced working ability of operators employed in private forestry sector. This is why in this paper the ergonomic aspects were investigated and compared of the working environment and working ability of forestry machine operators who are employed by private contractors in forestry on the one side, and those employed in the state company Croatian Forests Ltd. Zagreb (CF Ltd.) on the other side. The working ability of operators is researched and presented through the work ability index (WAI), and the ergonomic characteristics of the working environment are measured through the respondents’ answers on the training process, exposure to physical hazards (noise and vibration), impact of fatigue on work ability and also on their assessment of psychological and social factors of the working environment.

2. Issues and objectives of research Problematika i ciljevi rada Forest work is highly dangerous and risky, and the work environment that implies working outdoors hides many hazards. The group of work operations related to timber skidding and/or forwarding is technically demanding, the most expensive, and after felling and cutting to length the most risky among everyday forest operations (Ranogejec et al. 2010). Insight into the real conditions of work safety and humaniza-

242

tion of work activities in timber extraction, in the form of ergonomic, technical and organizational solutions, is a starting point for the process of improving ergonomic features of the working environment and work ability of machine operators in forestry. Satisfactory ergonomic level in terms of working environment and high working ability of forestry machine operators, in compliance with safety rules for mechanized skidding, seems crucial for creating safe working conditions in order to prevent physical fatigue, monotony of work and work-related injuries. Analysis of ergonomic characteristics of the working environment and numerical evaluation of machine operators’ work ability in forestry is the first step in the evaluation of the current situation, which should later result in proposing the necessary measures to improve the safety system and security levels of timber extraction. The main task of this research is to determine the work ability index (WAI) of machine operators employed in the state and private forestry sector and to establish the ergonomic features of the working environment, as well as opportunities to improve the current situation in forestry operations.

3. Material and methods – Materijal i metode Forest machine operators in Croatian forestry, employed by the state-owned company Croatian Forests Ltd. as well as those employed by private forest contractors, are the primary object of research. Investigation of ergonomic requirements of the working environment and research on work ability of machine operators engaged in timber extraction included the current theoretical understanding and views of operators using the survey method. The source and model used to create a survey questionnaire were the documents under the title »The Machine Operator Current Opinions and the Future Demands on Technical Ergonomics in Forest Machines« (Walker et al. 2001). Another relevant source refers to a document under the title »Work Ability Index« (Tuomi et al. 1998), which was developed by the Finnish Institute of Public Health.

3.1 Survey method and structure of the measuring instrument – Metoda anketiranja i struktura mjernoga instrumenta The survey method is a process based on a survey questionnaire, which investigates and collects data, information, views and opinions about the subject of research (Čekić 1999). There are numerous ways to conduct a survey questionnaire, and some of the most used are as follows: on-line or web surveys, e-mail Croat. j. for. eng. 34(2013)2

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

surveys, computer-aided telephone interviewing, etc. The answer to a particular question from the survey is valuable to the extent to which it is associated with the opinion or attitude on a topic. Questionnaires, in relation to interviews, are usually viewed as a more objective research tool that can produce generalizable results because of large sample sizes (Oppenheim 1992). For the investigation of the working environment ergonomics and work ability of forestry machine operators, the questionnaire developed by the Department of Forest Engineering at the Faculty of Forestry in Zagreb was used. The survey was conducted during the year 2012 as a part of activities on the project entitled »Licensing and Certification for Acquiring European Standards of Safety and Quality at Forestry Work« initiated and financed by Croatian Forests Ltd. Before the questionnaire was distributed, all required steps and preparations had been performed with the goal to ensure reliability and credibility of the survey. The questionnaire consists of five structural parts, and contains 48 questions. The first structural part of the questionnaire is related to the collection of personal data about respondents (gender, age, qualifications, etc). The second part of the questionnaire covers the current employment and organization of machine operators’ daily work activities. The third section of the questionnaire examines the presence and development of disease and fatigue, while the fourth section is related to the impact of psychological and social factors on forestry machine operators. Within the first four structural parts, questions were designed to measure separate variables. The fifth part of the questionnaire involves the assessment of work ability index (WAI) for forestry machine operators through 7 standardized questions. In the questionnaire, the following answers to the questions were used: a) yes/no answer b) multiple answers c) Likert scale evaluation and d) questions with open answers. Stratified sampling method was used to divide up the population (N = 150) into two smaller non-overlapping sub-groups: (a) operators in CF Ltd. (N = 75) and (b) operators in private forest companies (N = 75). In each sub-group a simple random sample was done by Sample Size Calculator. During the analysis and processing of collected data, the following methods were used: statistical method, method of generalization, description and comparison. Graded, collected and summarized data for the assessment of working and functional abilities of forestry machine operators (the fifth part of the questionnaire) are expressed through the work ability index (WAI). Database for entry of collected data, systematization, verification of input accuracy and primary processing of data was made in Croat. j. for. eng. 34(2013)2

M. Landekić et al.

Microsoft Office Excel. Statistical analysis was performed using statistical software: Statistica 8 and SPSS 17.0 – Statistical Package for Sociological Research. Rank correlation was used to measure the relationship between variables and alternative nonparametric analysis of variance (Kruskal–Wallis test) was used to test the differences between groups of variables. Based on individual observations, generalized conclusions were drawn up using the analysis of a limited number of subjects in the base sample.

3.2 Work ability index in general – Indeks radne spremnosti općenito Working and functional ability cannot be objectively measured with a single instrument. It always requires an assessment based on data obtained from several different sources (medical examination, survey, testing, etc.). Work Ability Index (WAI) is an instrument designed for practical application, widely used by healthcare workers or health and safety employees, like a help tool for determining employees working competence, as a basis for further measurements (Tuomi et al. 1998). WAI is a result of workers own assessment of his or her work readiness. The instrument, developed by the Finnish Institute of Public Health, is easy and quick to use, cyclically repeatable, results are obtained quickly and can be used for monitoring on the level of individuals or groups (e.g. department, age or professional groups, etc.). It is applicable within health and safety system, where it shows how well an employee is able to perform his daily job duties. It can be used for the assessment of working and functional ability in the framework of medical examination or as a survey at the workplace. The instrument based on a questionnaire is intended for workers support: a) where it can be used at an early stage to ensure proper measures to maintain working readiness or b) it can help in determining workers who need healthcare support at work. Responding to a series of seven questions (Table 1), which take into account physical and mental demands of the job, it gives the result ranging between 7 and 49 points (Table 1), which illustrates numerically the working and functional ability of each participant. This approach establishes optimal conditions to prevent premature reduction of work ability. Steps and measures directed toward restoring work ability or additional evaluations of work ability are needed by those whose work ability is graded poor (maximum score 27). For those whose work ability is moderate (score 28–36), measures to help improve work ability are recommended. Workers with a good work ability index (score 37–43) should receive instructions on how to maintain their work ability. Those

243

M. Landekić et al.

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

Table 1 Items of the Work Ability Index (Ilmarinen 2007) Tablica 1. Stavke indeksa radne spremnosti (Ilmarinen 2007) Items – Stavka

Range – Raspon

Current work ability compared with the lifetime best – Sadašnja radna spremnost u odnosu na najbolju životnu

0–10

Work ability in relation to the demands of the job – Radna spremnost u odnosu na zahtjeve posla

2–10

Number of current diseases diagnosed by a physician – Broj bolesti koje je dijagnosticirao liječnik

1–7

Estimated work impairment due to diseases – Procijenjena radno umanjenje zbog bolesti

1–6

Sick leave during the past year (12 months) – Broj dana bolovanja u protekloj godini (12 mjeseci)

1–5

Own prognosis of work ability 2 years from now – Vlastita prognoza radne sposobnosti za buduće 2 godine

1–7

Mental resources – Mentalni resursi

1–4

whose work ability is excellent (44–49) should also be informed about which work and life-style factors maintain work ability and which factors weaken it (Tuomi et al. 1998). The index can also be used to predict the threat of disability in the near future.

4. Results and findings of research Rezultati i nalazi ispitivanja In the present study, analysis of opinions and attitudes of the forestry machine operators included: a) profile analysis of the respondents in the study; b) main findings of the second, third and fourth structural part of the questionnaire related to the working environment of machine operators, c) values of the machine operators work ability index and comparison of WAI with demographic categories of respondents.

4.1 Profile of respondents – Profil ispitanika Forestry machine operators employed in the state and private forestry sector were selected as participants in the research conducted on the ergonomic features of the working environment and work ability index. First, 75 questionnaires were delivered to forestry machine operators employed in working units engaged in mechanization, transport and building, and also to the forest offices which have their own mechanization, within forest administration Požega, Našice, Zagreb and Karlovac, as a part of the company Croatian forests Ltd. Another 75 questionnaires were delivered to machine operators employed in private forestry sector, which were engaged in timber extraction on the territory of the forest administration Našice, Zagreb and Sisak at the time of survey. 44.67% of forestry machine operators answered the questionnaire (Table 2), which is satisfactory feedback in terms of the research.

244

The profile and characteristics of the operators surveyed in the state forestry sector, according to several criteria (gender, age group, qualifications), roughly correspond to the total number of employees working as machine operators in the company Croatian forests Ltd. Zagreb. A higher share of younger machine operators is visible in private forestry sector. Also, within the sample, difference in the level of machine operators’ education is notable. In the private sector, the operators’ education level is significantly higher than that of operators employed in Croatian forests Ltd. (Table 2).

4.2 Aspect of Working Environment of Forestry Machine Operators in Croatia– Aspekt radnoga okoliša kod rukovatelja mehanizacijom u Hrvatskoj

To plan for and control environmental and ergonomic aspects of forestry machine operators work, it is necessary to know what impacts them and where these impacts come from. Consequently, ergonomic characteristics of the working environment are examined and presented through the attitudes of the forest machinery operators concerning a) education and organization of operators work activities, b) impact of tiredness and c) impact of psychological and social factors on operator work ability. Identification and efficient management of environmental aspects and impacts should result in positive influence on employees in practice and also in significant environmental improvements. 4.2.1 Employment, education and organization of work activities – Zaposlenost, osposobljenost i organizacija rada For decades, training and periodically checking the qualification of forestry machine operators have been considered as the key activities for ensuring the work Croat. j. for. eng. 34(2013)2

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

M. Landekić et al.

Table 2 General information about the respondents Tablica 2. Opći podatci o ispitanicima Type of interviewees – Vrsta ispitanika

Forestry machine operators (private and state sector) Rukovatelji šumarskom mehanizacijom (privatni i državni sektor)

Number of respondents – Broj odgovora

67 (44.67%)

Time of research – Vrijeme ispitivanja

During 2012 – Tijekom 2012. godine

Profile of interviewees – Profil anketiranih zaposlenika

Croatian Forests Ltd.

Private forest company

Hrvatske šume d.o.o.

Privatna šumarska tvrtka

39 (52.00%)

28 (37.33%)

Gender

Male – Muški

100.00

Spol

Female – Ženski

0.00

<25

3.00

25.00

25–35

26.00

25.00

35–45

36.00

29.00

45–55

31.00

28.00

55<

5.00

4.00

Unqualified (worker) – Nekvalificirani (radnik)

46.00

18.00

Level of education

Qualified (worker) – Kvalificirani (radnik)

15.00

32.00

Stručna sprema

Secondary education – Srednja školska sprema

38.00

43.00

University degree – Visoka stručna sprema

0.00

7.00

Age group Dobna skupina

quality and safety of operational forest work. In most European countries, regulations oblige employers to provide adequate training to each person using the working tools and machines (Medved 1998). The research results of professional education of operators in the sample (Table 3) show that 56.41% of operators in CF Ltd., and 33.00% of operators employed by private contractors have no adequate professional experience or specialized education. The reason why the level of education is low is the lack of formal regulations and bodies that provide certification of knowledge and skills of forestry machine operators in Croatia. Knowledge, skill and experience in operating forest machinery are the basic items in the operator career development. By comparing the professional experience of the respondents (Fig. 1), it can be seen that operators employed by the company CF Ltd. have on average more work experience as forwarder operators, skidder operators and as forest cutters. The reason for this result is noticeably older population of respondents employed by the CF Ltd. (average age mean in CF Ltd = 42.33 years, and in private sector = 35.79 years) with about 1/3 higher overall professional exCroat. j. for. eng. 34(2013)2

Fig. 1 Professional experience of operators by workplace Slika 1. Godine profesionalnoga iskustva rukovatelja prema mjestu rada

245

M. Landekić et al.

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

Table 3 Type of education related to machine control Tablica 3. Vrsta obrazovanja vezana uz upravljanje mehanizacijom

Offered response – Ponuđeni odgovor

Croatian Forests Ltd.

Private forest company

Total

Hrvatske šume d.o.o.

Privatna šumarska tvrtka

Ukupno

Self-educated – Samoobrazovan

56.41

33.00

48.00

Vocational experience – Strukovno iskustvo

23.08

63.00

38.00

Specialized education – Specijalističko obrazovanje

20.51

4.00

14.00

length method (almost always with forwarders), through stemwood, halfwood to whole tree methods. Adequate work organization can proactively act to increase the level of performance of the forestry machine operators. Research results on work techniques training for the mechanized timber extraction (Table 4) show that the operators employed by private contractors have a higher level of training (29.85%). Exposure to excessive noise and vibration in the working environment was rated with 32.14% by respondents employed in the private sector and 41.03% by respondents employed in CF Ltd (Table 4). The lack of personal protective equipment (PPE) was rated with 32.14% by respondents employed by private contractors, and 50.00% said that they do not use it every day to the full extent. On the other hand, in the company CF Ltd. only 5.13% of the respondents indicated a lack of personal protective equipment, and 43.59% said that they do not use it every day to the full extent. The occurrence of pain and discomfort (Table 4) caused by the position of the body when working was noted by 16 operators employed in CF Ltd. and 12 operators employed in the private forestry sector.

Fig. 2 Number and average age of machinery operated by sampled operators Slika 2. Broj i prosječna dob mehanizacije kojom upravljaju uzorkovani rukovatelji

perience in forestry (Fig. 1). On the other hand, the sample shows that operators employed by the CF used more commonly skidders for timber extraction (Fig. 2), and that the average age of machines (forwarders and skidders) is considerably lower than the age of machinery owned by private contractors in forestry. In timber extraction, different methods are applied for removing different forms of assortments: from cut-to-

246

4.2.2 The impact of fatigue and development of disease among respondents – Utjecaj zamora i razvoj bolesti kod ispitanika Tiredness and fatigue can have a significant adverse impact on organizational efficiency and productivity as well as operators’ health and safety. Assessment of the impact of physical fatigue on the quality and productivity of machine operators (Fig. 3) is valued with five answers on Likert scale. Operators employed in CF Ltd. believe that fatigue significantly (very much – 41.03%) affects the quality and productivity of the work compared to those employed by the private contractor (17.86%). Higher percentage of the rates »much«, »medium« and »little« is recorded by the operators employed in the private forestry sector (Fig. 3). Jobs with an increased risk, especially for the development of work-related diseases, require medical Croat. j. for. eng. 34(2013)2

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

M. Landekić et al.

Table 4 Rating of organizational factors during work of machine operators (answer share, %) Tablica 4. Ocjena organizacijskih čimbenika u radu rukovatelja mehanizacijom (udio odgovora, %)

Organizational factors – Organizacijski čimbenici

Croatian Forests Ltd.

Private forest company

Hrvatske šume d.o.o.

Privatna šumarska tvrtka

Yes – Da

No – Ne

48.72

38.46

41.03

All PPE provided – Osiguranost svih OZS Ful use of PPE – Uporaba OZS u punoj mjeri

Work techniques training in mechanized timber extraction Osposobljavanje radnim tehnikama pri mehaniziranom privlačenju Exposure to excessive noise and vibration Izloženost prekomjernoj buci i vibracijama

First-aid kit and fire extinguisher in the machine Kutija prve pomoći i protupožarni aparat u stroju Appearance of pain and discomfort caused by body position at work Pojava boli i neugode uzrokovana položajem tijela pri radu

Don't know

Yes – Da

No – Ne

12.82

78.57

14.29

7.14

25.64

33.33

32.14

39.29

28.57

94.87

5.13

0.00

64.29

32.14

3.57

53.85

43.59

2.56

50.00

0.00

97.44

2.56

0.00

85.71

14.29

0.00

58.97

41.03

0.00

50.00

42.86

7.14

Ne znam

research, the symptoms of headache were noticed by 17 participants, and the appearance of symptoms of insomnia was noted by 11 participants employed by the company CF Ltd. A high proportion of respondents said that the symptoms of headache were associated with work (Table 5). A smaller percentage of respondents employed by private contractors reported the appearance of symptoms, but they put the emphasis on the connection between these symptoms and the daily work activities and demands at work (Table 5).

Fig. 3 Rating the impact of fatigue on the quality and productivity of machine operator work Slika 3. Ocjena utjecaja zamora na vrsnoću i proizvodnost rukovatelja mehanizacijom supervision of employees for early detection, monitoring and treatment of health disorders partially caused by exposure to physical and mental challenge, and hazards at work (Macan et al. 2012). As part of the Croat. j. for. eng. 34(2013)2

4.2.3 Impact of psychological and social factors on forestry machine operators – Utjecaj psihološko društvenih čimbenika na rukovatelje šumarskom mehanizacijom Psychological and social factors of the working environment may have a direct impact on the health and safety of forest machine operators, as well as on productivity. One of the most important tasks of employers is to keep the mental load and strain of employees at acceptable levels, all with the long-term goal of diminishing the probability of occurrence and development of mental stress (Landekić et al. 2011a). According to the rating of mental job demands by forestry machine operators (Fig. 4), the results show a significant proportion of large and very large (46.42%) mental demands of the job performed by operators employed in private companies in relation to the operators employed in the state–owned company CF Ltd. The reason for the increased mental demands of the job indicated by the operators employed in a pri-

247

M. Landekić et al.

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

Table 5 Headaches and insomnia symptoms experienced by forestry machine operators Tablica 5. Pojava simptoma glavobolje i nesanice kod rukovatelja mehanizacijom u šumarstvu Croatian Forests Ltd. – Hrvatske šume d.o.o.

Private forest company – Privatna šumarska tvrtka

Headache – Glavobolja

Insomnia – Poremećaj spavanja

Headache – Glavobolja

Insomnia – Poremećaj spavanja

17 respondents (43.59%)

11 respondents (28.21%)

10 respondents (35.71%)

5 respondents (17.86%)

17 ispitanika (43,59 %)

11 ispitanika (28,21 %)

10 ispitanika (35,71 %)

5 ispitanika (17,86 %)

Work–related

Povezano s poslom

Yes – Da

No – Ne

Yes – Da

No – Ne

Yes – Da

No – Ne

Yes – Da

No – Ne

70.59%

29.41%

45.46%

54.54%

70.00%

30.00%

60.00%

40.00%

resulted in almost equal frequency response of forest workers employed in the private sector and in CF Ltd. Influence of the working climate within the operators’ working environment, expressed through business solidarity, working atmosphere, good relationship with superiors and colleagues, is evaluated in Table 6. The results show a higher proportion of affirmative answers (often and always) on the subject of freedom to make own decisions while working and the existence of a sense of solidarity by the operators employed in private companies. On the other hand, a greater proportion of affirmative answers (often or always) were recorded by operators employed in CF Ltd. regarding good working atmosphere and good relationship with superiors and colleagues (Table 6).

4.3 Work Ability Index of Forestry Machine Operators – Indeks radne spremnosti rukovatelja šumarskom mehanizacijom

Fig. 4 Rating of job mental demands by forestry machine operators Slika 4. Ocjena mentalne zahtjevnosti posla kod rukovatelja šumarskom mehanizacijom vate company (Fig. 4) can be partially linked with the results in Table 6 related to time pressure, working climate, acquiring new skills or knowledge and some other factors of the working environment explored in this paper. Relative frequency of responses related to time pressure at work is presented in Table 6. The percentage share of responses shows a greater presence of time pressure at work for operators employed in a private company (often or always – 32.14%). Evaluation of acquiring new skills or knowledge (Table 6)

248

Work ability index is a tool in the form of a questionnaire used for the self-assessment of employees. Focus is put on employees and their job readiness in relation to the requirements of their current job position. It also highlights the need to adapt working conditions to the capacities and capabilities of employees. Mean score of the work ability index of machine operators, employed in Croatian forests Ltd. and in private companies, is shown in Table 7. The results of the WAI average score show a negligible difference between the operators employed in the private sector and in state forestry sector. They have good work ability (score 37–43), which should be kept at the existing level. For a more detailed insight into work ability, a thorough examination of opinions and attitudes of forestry machine operators included as follows: (a) correlation analysis of work ability index (WAI) with Croat. j. for. eng. 34(2013)2

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

M. Landekić et al.

Table 6 Evaluation of some psychological and social factors of the working environment, % Tablica 6. Procjena nekih psihološko-društvenih čimbenika radnoga okoliša, %

Questions – Pitanja

Do you feel the time pressure due to the volume of work? Osjećate li vremenski pritisak zbog opsega posla? Do you learn new things at work? Učite li nove stvari na poslu? Does your job require skills? Zahtijeva li Vaš posao vještinu? Does your job require ingenuity? Zahtijeva li Vaš posao domišljatost? Do you have any freedom to decide at work? Imate li slobodu odlučivanja u radu? Is the atmosphere at work pleasant? Je li atmosfera na poslu ugodna? Is there a sense of solidarity? Postoji li osjećaj solidarnosti? I have a good relationship with my superiors? Slažete li se dobro s nadređenim? I have a good relationship with my colleagues? Slažete li se dobro s kolegama?

Croatian Forests Ltd.

Private forest company

Hrvatske šume d.o.o.

Privatna šumarska tvrtka

Never

Rarely

Often

Always

Never

Rarely

Often

Always

Nikad

Rijetko

Često

Uvijek

Nikad

Rijetko

Često

Uvijek

20.51

66.67

12.82

0.00

17.86

50.00

21.43

10.71

10.26

33.33

35.90

20.51

0.00

32.14

42.86

25.00

2.56

23.08

74.36

0.00

17.86

21.43

53.57

7.14

2.56

38.46

33.33

0.00

28.57

25.00

46.43

5.13

25.64

41.03

30.77

10.71

17.86

35.71

0.00

10.26

43.59

46.15

0.00

10.71

50.00

39.29

2.56

12.82

33.33

51.28

0.00

7.14

67.86

25.00

0.00

2.56

25.64

71.79

0.00

3.57

42.86

53.57

0.00

17.95

82.05

0.00

3.57

46.43

50.00

Table 7 Work ability index of forestry machine operators – CF Ltd. and private contractors Tablica 7. Indeks radne spremnosti rukovatelja šumarskom mehanizacijom – HŠ d.o.o. i privatni izvoditelji Indicator

Number of operators

Minimum

Maximum

Arithmetic mean

Standard deviation

Pokazatelj

Broj rukovatelja

Minimum

Maksimum

Aritmetička sredina

Standardna devijacija

24.00

49.00

38.46

5.70

24.00

48.00

38.11

5.39

WAI (CF Ltd.) IRS (HŠ d.o.o.) WAI (private contractors) IRS (privatni izvoditelji)

demographic indicators of the respondents and (b) testing the WAI difference among a group of descriptive variables. 4.3.1 Correlation of work ability index and demographic parameters of respondents – Mjere povezanosti indeksa radne spremnosti i demografskih parametara ispitanika This section shows testing the strength and direction of respondent demographic parameters and work Croat. j. for. eng. 34(2013)2

ability index. Correlation and influence are reviewed with the aim of gaining a more comprehensive understanding of the relation between the work of forestry machine operators and their functional ability (WAI) depending on work experience, age and weight of the workers. Spearman’s rank correlation coefficient was used to assess the degree and direction of the relation between the derived indicators. The indicator of machine operator work ability index negatively correlated with three demographic

249

M. Landekić et al.

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

Table 8 Rank correlation of work ability index of machine operators and demographic parameters Tablica 8. Korelacija ranga indeksa radne spremnosti i demografskih parametara rukovatelja mehanizacijom Variable

Indicator

Age

Mass

Work experience in forestry

Work ability index

Varijabla

Pokazatelj

Godine života

Masa

Radno iskustvo u šumarstvu

Indeks radne spremnosti

Age

1.000

0.238

0.769**

–0.525**

Godine života

–

0.052

0.000

Mass

0.238

1.000

0.335

–0.247*

Masa

0.052

–

0.006

0.044

1.000

–0.443**

–

0.000

Work experience in forestry

0.769

Radno iskustvo u šumarstvu

0.000 **

0.335

0.006 *

Work ability index

–0.525

–0.247

–0.443

1.000

Indeks radne spremnosti

0.000

0.044

0.000

–

* Correlation is significant at the 0.05 – Korelacija je značajna na razini 0,05;

** Correlation is significant at the 0.01 – Korelacija je značajna na razini 0,01

the work ability index and work experience in forestry was medium negative (r = -0.443; n = 67; p<0.01), where a higher level of work ability index was recorded with machine operators with less experience in forestry, and the calculated coefficient of determination (r² = 0.196) indicated that work experience accounted for 19.6% of the variance in the respondents answers regarding the indicator of work and functional ability. The relation between the work ability index of machine operators and the mass of respondents resulted in a small negative correlation (r = -0.2474; n = 67; p<0.05).

Fig. 5 Mean value of WAI by three age groups Slika 5. Srednja vrijednost IRS prema trima dobnim grupama variables. The relationship of work ability index and age of the respondents resulted in a strong negative correlation (Table 8), r = -0.525; n = 67; p<0.01, where a high level of work ability index was recorded with younger machine operators. The calculated coefficient of determination (r² = 0.275) showed that age accounted for 25.5% of the variance in the respondents answers regarding the indicator of work and functional ability. Strength and direction of relationship between

250

4.3.2 Testing the diference of WAI between groups of descriptive variables – Ispitivanje razlika indeksa radne spremnosti prema opisnim varijablama Using the database of respondents, differences were tested between the work ability index and selected descriptive variables. The following descriptive variables were used: a) age group and b) group of work experience in forestry. The homogeneity of variance between groups of data was tested with Levene’s test (p>0.05), where on the basis of test significance level a further testing of WAI difference was conducted with parametric and/or nonparametric techniques. Due to inadequacy of variance homogeneity, alternative nonparametric analysis of variance (Kruskal– Wallisov test) was used to test the difference of work ability index among three groups of respondents according to age. Also, differences in work ability index were examined between the four groups of respondents regarding work experience in forestry. Croat. j. for. eng. 34(2013)2

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

M. Landekić et al.

Table 9 Testing the difference between WAI groups using the Kruskal–Wallis H test Tablica 9. Ispitivanje razlika IRS između grupa pomoću Kruskal–Wallisova H-testa Chi-square

Degrees of freedom

Sample size

P-value

Hi-kvadrat

Stupnjevi slobode

Veličina uzorka

P-vrijednost

Age of respondents – Godine života

16.897

0.000**

Work experience in forestry – Radno iskustvo u šumarstvu

14.776

0.002**

Descriptive variable – Opisne varijable

** The difference is significant at 0.01 – Razlika je značajna na razini 0,01

Groups by age: Þ group 1: less than 35 years of age; Þ group 2: from 36 to 45 years of age; Þ group 3: more than 46 years of age;

Testing the score values of work ability index among defined groups of respondents resulted in the following statistically significant differences (Table 9). Statistically significant difference was determined among the age groups (p<0.01), where operators with

less than 35 years of age have the highest level of work ability index (median – Md = 41.00) compared to their older colleagues (Fig. 5). Subsequent testing of the difference using the Mann-Whitney U test showed that the median value of WAI in group 1 (<35 years) (Md = 41.00, N = 25) was significantly different from group 3 (>46 years of age) (Md = 35.50, N = 20), U = 64.50; z = -4.255; p = 0.000. A statistically significant difference was not confirmed between group 2 and group 1. Forestry machine operators with more than 46 years of age employed in the state sector (CF Ltd) showed a significantly better working and functional ability in relation to operators working in the private forestry sector (Fig. 6). Forestry machine operators with less than 10 years of experience in forestry had the highest level of work

Fig. 6 WAI of machine operators in private and public sector by age groups Slika 6. IRS rukovatelja mehanizacijom u privatnom i državnom sektoru prema dobnim grupama

Fig. 7 Mean value of WAI by four groups of working experience in forestry Slika 7. Srednja vrijednost IRS prema četirima grupama radnoga staža u šumarstvu

Groups by work experience in forestry: Þ group 1: 0–10 years of work experience; Þ group 2: 11–20 years of work experience; Þ group 3: 21–30 years of work experience; Þ group 4: 31–40 years of work experience.

Croat. j. for. eng. 34(2013)2

251

M. Landekić et al.

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

Fig. 8 WAI of machine operators in private and public sector by groups of work experience Slika 8. IRS rukovatelja mehanizacijom u privatnom i državnom sektoru prema grupama radnoga staža ability index (Md = 41.00) in comparison to colleagues with more years of experience (Fig. 7). Using the Mann-Whitney U test, testing of differences showed that the median value of WAI in group 1 (<10 years of service) (Md = 41.00; N = 35) was significantly different from group 2 (Md = 37.00; n = 19) U = 196.50; z = –2.473; p = 0.013, group 3 (Md = 33.00; N = 8) U = 65.00; z = –2.704; p = 0.007 and group 4 (>31 years of service) (Md = 34.00; N = 4) U = 15.500; z = -2.238; p = 0.011. The research has also shown that the indicator of employees work ability in the state sector slightly increases with years of service, while in the private sector it declines with years of service (Fig. 8).

5. Discussion and conclusions – Rasprava i zaključci The study was designed with the purpose of determining the limits of forestry machine operators’ working and functional ability and acquiring insights in some ergonomic aspects of their working environment in forestry production. The investigated education level of machine operators showed a lack of specialized education. In an effort to overcome the weaknesses of the national training system, the Croatian forestry sector needs to look at examples of good practices implemented by the European countries, and at

252

measures advocated by Martinić et al. (2011), regarding the establishment of a national center for forestry, which would enable the implementation of the certification process of machine operator training in forestry. Also, it is necessary to develop the safety culture in the forestry sector through the development of personal responsibility for the safety, use of personal protective equipment and establishment of joint responsibility on safety including the management and employees. Psychological and social aspects of the working environment of machine operators resulted in a higher level of mental demands on the job with the operators employed in private forest companies. A higher proportion of time pressure due to the volume of work and poor working climate in the workplace of operators employed in the private forestry sector goes in favor of the obtained results on mental demand. Psychological and social factors of the working environment can influence the operators’ productivity, and therefore one of the most important tasks of employers is to keep the mental load and strain of employees at acceptable levels (Landekić et al. 2011a). Demographic parameters of the respondents (age, work experience in forestry and weight) negatively affect the working and functional ability (WAI) of forestry machine operators. Machine operators with less than 35 years of age and with less than 10 years of experience in forestry have the highest level of work ability index, which is significantly different (p<0.01) from the WAI in the oldest groups. Lower work ability of the respondends in the oldest group, according to Ilmarinen at al. (2005), can be related to the health (symptoms) and functional capacity (physical) or work factors (mental strain) at the workplace of machine operators. According to age groups, machine operators working in private companies, with more than 46 years of age, have a considerably lower WAI compared to operators employed by the company CF Ltd. Also, within the groups formed by work experience in forestry, a visible reduction in the level of WAI (moderate rating) is observed with operators working in private companies, which indicates the need for measures required to improve the current situation. The responses of all machine operators and the analysis presented in this paper represent an important contribution in the process of developing a model of security measures for the improvement of health and safety system in timber extraction through education on work ability index and responsibilities that need to be implemented in private and state forestry sector. Such approach helps to identify potential work-related health risks in order to implement appropriate Croat. j. for. eng. 34(2013)2

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

measures aimed at reducing the possibility of declining the working capacity of employees and preventing their early retirement. Maintenance and/or improvement of the working and functional ability of forestry machine operators can and needs to be carried out through periodical training during the whole working life.

M. Landekić et al.

ment for Forestry, Freiburg, Germany, September 26 – October 1, 2011. (In press) Macan, J., Kerner, I., Šetek, J., 2012: Smjernice za zdravstvene preglede zaposlenih u izdanju hrvatskog društva za medicinu rada hrvatskog liječničkog zbora. Arh Hig Rada Toksikol, 63, 555–558.

6. References – Literatura

Martinić, I., Landekić, M., Šporčić, M., Lovrić, M., 2011: Forestry at the EU’s Doorstep – How Much areWe Ready in the Area of Occupational Safety in Forestry?, Croatian journal of forest engineering, 32(1): 431–440.

Creative Research Systems, 1982: Sample Size Calculator. (Source: http://www.surveysystem.com /sscalc.htm#two)

Medved, M., 1998: Nezgode in tveganje pri poklicnem in nepoklicnem delu v gozdu. GozdV, 56(9): 379–389.

Čekić, Š., 1999: Osnovi metodologije i tehnologije izrade znanstvenog i stručnog djela. FSK, Sarajevo, 73.

Oppenheim, A. N., 1992: Questionnaire design, interviewing, andattitude measurement. New York City: St. Martin‘s Press.

Ilmarinen, J., Tuomi, K., Seitsamo, J., 2005: New dimensions of work ability. Proceedings of 2nd International Symposium on Work Ability: Assessment and Promotion of Work Ability, Health and Well-being of Ageing Workers. Volume 1280, Pages 1–436 (June 2005) Verona, Italy. Ilmarinen, J., 2007: The Work Ability Indeks (WAI). Occupational Medicine, 57 (Source: http://occmed.oxfordjournals. org) International Business Machines (IBM) Corporation, 2008: Statistical Package for the Social Sciences – SPSS 17.0® software. Kastenholz, E., Dyduch, C., Fitzgerald, R., Hudson, B., Jaakkola, S., Lidén, E., Monoyios, E., et all., 2008: Guide to good practice in contract labour in forestry. Report of the UNECE/ FAO Team of Specialists on Best Practices in Forest Contracting. Food and Agriculture Organization of the United Nations, Rome, 1–54. Landekić, M., Martinić, I., Lovrić, M., Šporčić, M., 2011a: Assessment of Stress Level of Forestry Experts with Academic Education, Coll Antropol 35(4): 1185–1191. Landekić, M., Martinić, I., Lovrić, M., Zečić, Ž., Šporčić, M., Vusić, D., 2011b: Private Entrepreneurship in the Forestry Sector of the Republic of Croatia – Status and Perspectives. International scientific conference: Competence Develop-

Poršinsky, T., 2005: Djelotvornost i ekološka pogodnost forvardera Timberjack 1710 pri izvoženju oblovine iz nizinskih šuma Hrvatske. Disertacija, Šumarski fakultet Sveučilišta u Zagrebu, Zagreb, 1–170. Ranogajec, B., Klarić, D., Zagudajev, J., Perakić, S., Plantak, S., Pavlić, V., Koščević, V., Mundweil, V., Tomašić, Z., 2010: Upute za rad na siguran način pri privlačenju i prijevozu drveta. Hrvatske šume d.o.o. Zagreb 2010, 1–62. StatSoft, Inc., 2007: Statistica 8® software. Šporčić, M., 2005: Uvid u neka gledišta poduzetništva u šumarstvu Europe. Šumarski list, 129(5–6): 287–298. Šporčić, M., Martinić, I., Landekić, M., Lovrić, M., Svakidan, M., 2009: Prikaz stanja poduzetništva u šumarstvu srednje i istočne Europe. Nova mehanizacija šumarstva, 30(1): 37–46. Tuomi, K., Ilmarinen, J., Jahkola, A., Katajarinne, L., Tulkki, A., 1998: Work Ability Indeks. Finnish Institute of Occupational Health, Helsinki 1998, 1–22. (Source: http://www. scribd.com/doc /52853348/Work-Abilty-Indeks-Book) 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 (SLU), 1–73.

Sažetak

Indeks radne spremnosti rukovateljâ šumarskom mehanizacijom i neki ergonomski aspekti njihova rada U radu se analiziraju neki ergonomski aspekti radnoga okoliša te se numerički vrednuje radna spremnost rukovateljâ šumarskom mehanizacijom kao prvi korak u ocjeni trenutačnoga stanja. Pritom je metodom anketiranja tijekom 2012. godine istražen i analiziran a) indeks radne spremnosti (eng. Work Ability Index) (tablica 1) te b) ergonomski aspekt radnoga mjesta rukovateljâ šumarskom mehanizacijom za državni (Hrvatske šume d.o.o.) i privatni šumarski sektor. Croat. j. for. eng. 34(2013)2

253

M. Landekić et al.

Work Ability Index of Forestry Machine Operators and some Ergonomic Aspects... (241–254)

Prvi dio rezultata obuhvaća a) profil ispitanika (tablica 2), b) organizaciju rada rukovateljâ mehanizacijom i njihovo obrazovanje (tablica 3 i 4), c) utjecaj umora (slika 3) i d) utjecaj psiholoških i socijalnih čimbenika na radnu sposobnost rukovateljâ (slika 4). Na temelju nalaza istraživanja prikazanih u radu potvrdilo se da se u šumarskom sektoru Republike Hrvatske, kako u državnom tako i u privatnom, vrlo malo sredstava, znanja i truda ulaže u osposobljavanje, sigurnost i zdravlje radnika. Prevladavaju nekvalificirani radnici (tablica 3) koji upravljaju vrlo skupim strojevima bez osnovnoga strukovnoga i specijalističkoga obrazovanja. Razlog je tomu nepostojanje zakonskih instrumenata u smislu obvezne certifikacije znanja i vještina za rukovatelje šumarskom mehanizacijom u RH, a posljedično i manjak organizacije pratećega sustava certifikacije. Također, rezultati istraživanja pokazuju da zaposlenici nisu obrazovani kako izvoditi radove uza što manje fizičkoga i psihičkoga opterećenja (slika 3 i 4), te kako postići optimalnu dinamiku dnevnoga rada u pogledu održavanja tjelesne kondicije, koncentracije i dr. Radnicima koji upravljaju šumarskim strojevima u većini slučajeva osigurana su propisana osobna zaštitna sredstva, ali većina ih se ne koristi njima ili se koriste njima na neispravan način (tablica 4). Drugi dio rezultata vezan uz numeričko vrednovanje radne spremnosti rukovateljâ obuhvaća a) odnos radne spremnosti rukovateljâ prema demografskim kategorijama ispitanika (tablica 7 i 8) i b) pregled razlika između indeksa radne spremnosti prema odabranim opisnim varijablama (slika 5, 6, 7 i 8). Ključni parametri ispitanika (godine života, radno iskustvo u šumarstvu, masa ispitanika) negativno utječu na radnu i funkcionalnu spremnost (IRS) rukovateljâ šumarskom mehanizacijom (tablica 8). Rukovatelji mehanizacijom zaposleni u privatnom sektoru s više od 46 godina života imaju zamjetno niži IRS u usporedbi s rukovateljima zaposlenim u HŠ d.o.o. (slika 5 i 6). Također, unutar grupa prema radnomu iskustvu vidljivo je smanjenje vrijednosti IRS (srednje bodovano) te su potrebne mjere za unapređenje postojećega stanja (slika 7 i 8). Zaključno, na temelju rezultata dani su prijedlozi za unapređenje istraživanih čimbenika radne spremnosti rukovateljâ šumarskom mehanizacijom. Kao nužna mjera poboljšanja potreban je cjelovit i kvalitetan postupak osposobljavanja kojim će se rukovatelji šumarskom mehanizacijom upoznati s opasnostima i štetnostima u procesu privlačenja drveta, s vrstama i razinom opterećenja pri radu te osposobiti za rad uporabom sigurnoga načina rada i optimalne radne tehnike. Samo korektni radni postupci povezani sa zadovoljavajućom razinom radne sposobnosti mogu preduhitriti nastanak tjelesnih ozljeda i zdravstvenih tegoba rukovateljâ. U nastojanju da se riješe slabosti državnoga sustava obuke radnika, hrvatsko se šumarstvo treba ugledati na primjere dobre prakse iz zemalja u europskom okruženju, a prema modelu Martinića i suradnika (2011) sa središnjom ulogom nacionalnoga centra za šumarstvo rada koji bi po svom osnutku trebao biti središte sustava za provedbu procesa certificiranja rukovatelja mehanizacijom u šumarstvu Ključne riječi: šumarstvo, rukovatelji mehanizacijom, indeks radne spremnosti, radni okoliš

Authors’ address – Adresa autora:

Received (Primljeno): March 13, 2013 Accepted (Prihvaćeno): September 11, 2013

254

Matija Landekić, MSc. e-mail: mlandekic@sumfak.hr Prof. Ivan Martinić, PhD. e-mail: imartinic@sumfak.hr Matija Bakarić, MSc. e-mail: mbakaric@sumfak.hr Asst. Prof. Mario Šporčić, PhD.* e-mail: sporcic@sumfak.hr Forestry Faculty of Zagreb University Department of Forest Engineering Svetošimunska 25 HR – 10 000 Zagreb CROATIA *Corresponding author – Glavni autor Croat. j. for. eng. 34(2013)2

Original scientific paper – Izvorni znanstveni rad

Time Consumption of Skidding in Mature Stands Performed by Winches Powered by Farm Tractor Janusz M. Sowa, Grzegorz Szewczyk Abstract – Nacrtak The aim of the present research was to determine the characteristics of time consumption in skidding by winch. The research was conducted in pine, fir, spruce and beech mature stands. It covered the operation of skidding from the stand to the skid trail at the distance of up to 50 m. A time study was performed for skidding operations, timber volume and thinning intensity. The average time consumption of skidding in the operational time, assessed in the examined mature stands, amounted to approximately 18 min/m3. Significant differences were observed in frequency levels between early thinnings (24 min/m3) and late ones (13 min/m3). The operational time structure for skidding by winch was characterized by a large share of auxiliary time: 71%. Out of that time, 30% was used for attaching and detaching the load and 36% for the transfer. Approximation was also done of the multiple regression equations. The equations described changes in skidding time consumption, i.e. the Empirical Efficiency Index (EST). The changes depended on environmental factors (stand, cutting category), elements of the working day structure (the share of a given time category in a shift) and task intensity (ratio of the number of harvested trees per area unit). The strongest correlations between the EST and the analyzed variables were observed for the factors related to the percentage of time required for attaching and detaching the load and factors related to operation intensity. Keywords: work consumption, timber harvesting technologies, skidding, forest utilization

1. Introduction – Uvod In the mid-1990s in Poland, the market economy led to the privatization of almost 100% of the felling work (Więsik 2000). This economy has entirely changed the idea of timber harvesting. In Poland, small, one- or two-person forest enterprises with insufficient capital prevail. Certification of forest enterprises has been conducted in recent years with the support of the State Forests’ administration (Kapral 2000). One of the crucial aspects of assessment is the method of performing forest operations in accordance with the applicable regulations and standards. For many reasons, it is very important to know the appropriate time standards for performing a specific operation. This is often underestimated although it is useful for both foresters and forest enterprise owners. It allows for proper arrangement of tender procedures and for rational planning of operations that are to be carried out. Croat. j. for. eng. 34(2013)2

The amended Catalogues of Forest Work Time Standards, which have been in force in the State Forests units since 2003, are also useful for forest entrepreneurs. The tabular data in these standards present the average working conditions in the field. The data characterize the tasks connected with specific technological systems and timber harvesting technologies. Most studies treat modeling of time consumption (productivity) in specific forest operations or in whole harvesting technologies as a relation between the volume of the harvested timber and a selected category of effective working time. This increases the precision of inference. On the other hand, it does not allow for assessing the joint effect of all factors on a given variable. The real picture of the phenomenon can be shown by the multi-criterion consideration of the analyzed relations.

255

J. M. Sowa and G. Szewczyk

Time Consumption of Skidding in Mature Stands Performed ... (255–264)

The objective assessment of a given technology is connected with measuring the time of the examined operations. Full examination of the time needed to obtain a product (effective time, operational time, shift time) allows for complete assessment of the operations. The present research assumes that the operational time will provide a sufficient generalization. In each case, it is crucial to find the categories of time which significantly affect time consumption in a shift. This is the way to indicate the operations, which should be focused on, at the stage of shift optimization. The above-mentioned interdependence of specific categories of time should be reflected in their percentages, because they show specific features of a given technology and a given stand. The introduction of work time standards has always attracted much interest (Cserjes 1989, Lukačka 1989, Döhrer 1998, Grzegorz 2003, Derek 2004, Kusiak 2006). Standardization has often been understood only as an element of control. However, it is also a tool used to plan properly the performance of economic tasks. For this reason, the use of time consumption catalogues by the State Forests encourages critical assessment of their standards.

The present research on skidding with the use of cable winches powered by farm tractors, commonly used in Polish forestry, shows the current technical capabilities of timber harvesting as performed by small forest enterprises with insufficient capital. The farm tractor is the most common equipment used for work in agriculture (in the broad sense of the term), and therefore also in forestry (Gil 2007). In Poland, about 65% of skidding operations are performed with the use of tractors.

2. Research aim and scope – Cilj istraživanja

3. Methods – Metode istraživanja

Due to the changing technical capacities of timber harvesting and considering the necessity to update the quantification of the multi-criteria influence of the selected factors on the level of time consumption, an attempt was made to provide a preliminary assessment of this phenomenon. Constructing a time consumption model for different timber harvesting technologies would allow for making realistic time standards for performing individual operations and for undertaking research on such standardization that would be useful for both the State Forests as the employer and for forest enterprises as contractors. The aim of the present study was to determine the characteristics of time consumption of skidding by Fransgård winch powered by a farm tractor (referred to as WINCH below). The modeling consisted in approximating the mathematical functions described by the following relation (1): ESTwinch = f(stand structure, task intensity, elements of working day structure) where: ESTwinch – the synthetic index of Empirical Technological Efficiency at the work stand: WINCH

256

(1)

The present research was conducted in pine, beech, fir and spruce stands. The scope of the operations, limited to the stands of early and late thinning, made it possible to optimize the time consumption model in a group of stands that had the highest share of area and volume. In such stands the performance of timber harvesting operations is particularly difficult, especially concerning the part of skidding operations from the stem to the skid trail. This is affected by the spatial structure of such stands as well as by the volume and dimensions of logs. In the stands of middle age classes, the largest problems occur with the determination of the appropriate levels of time standards, used in job tenders by the State Forests.

The research plots of the present research were situated within the Regional Directorate of the State Forests in Cracow, the Regional Directorate of the State Forests in Katowice and the Forest Experimental Station in Krynica (Tab. 1). In the areas chosen for their full density and uniformity of forest taxation features (breast-height diameter, height) and for their species composition, experimental plots of 0.5 ha and dimensions 50 × 100 m each were set up so that the longer side of each plot was adjacent to the skid trail. On each plot, at 32 circular plots of 50 m2 each, complete stock-taking was done of all trees thicker than 7 cm. The equipment used in the present research was Fransgård V6000GS winch powered by a Pronar 5112 farm tractor. The timber was harvested in the tree length system (TLS) (Laurow 2000). Cable skidding was performed in the direction towards the skid trail at the maximum distance of 50 m. No additional equipment, such as skidding tongs or skidding sledge, was used to facilitate skidding and each item was attached to the collective rope by means of standard attachment ropes with slide locks. One collective rope was used for the skidding of maximum 6 logs. The winch operator performed the skidding from the skid Croat. j. for. eng. 34(2013)2

Time Consumption of Skidding in Mature Stands Performed ... (255–264)

J. M. Sowa and G. Szewczyk

152a

6.49

245c

9.19

250d

1.74

Bor

Mountain forest

Fir

Planinska šuma

Jela

Mountain forest

Fir

Planinska šuma

Jela

Mountain forest

Fir

Planinska šuma

Jela

Mountain forest

Fir

Planinska šuma

Jela

Mixed mountain forest

Beech

Mješovita planinska šuma

Bukva

Mountain forest

Beech

Planinska šuma

Bukva

Mixed mountain forest

Spruce

Mješovita planinska šuma

Smreka

Mixed mountain forest

Spruce

Mješovita planinska šuma

Smreka

trail. The basic technical data of Fransgård V6000GS winch powered by a Pronar 5112 farm tractor are presented in Table 2. A constant time study of the operations was conducted with the working day method during skidding (Monkielewicz and Czereyski 1971, Sajkiewicz 1981). Time was measured with the use of PSION Workabout Croat. j. for. eng. 34(2013)2

0.9

0.7

1.0

1.1

Crown density Sklop

6.37

Mješovita šuma četinjača

Stocking of stand Obrast

333b

Pine

Age , yr Dob, god.

8.07

Species Glavna vrtsa drveća

300c

Forest site type Vrsta šume

5.62

Moist mixed coniferous forest

Full crown closure Gust sklop sastojine Moderate crown closure Umjereno gust sklop sastojine Full crown closure Gust sklop sastojine Moderate crown closure Umjereno gust sklop sastojine Broken crown closure

Nepotpuno sklopljena sastojina

0.6

1.1

1.0

1.2

Broken crown closure Nepotpuno sklopljena sastojina Moderate crown closure Umjereno gust sklop sastojine Moderate crown closure Umjereno gust sklop sastojine Full crown closure Gust sklop sastojine Moderate crown closure Umjereno gust sklop sastojine

Large timber, m3/ha Zrela stabla, m3/ha

Kasne

316g

Bor

Stand quality – Bonitet

Late

26.01

Mješovita šuma

Height, m Srednja visina stabla, m

Nowy Targ

Early Rane

45b

Fresh Mixed broadleaved forest Pine

DGB, cm Srednji prsni promjer, cm

Kasne

Forest area, ha Površina šume, ha

Late

Compartment Odjel/odsjek

Rane

Forest district Gospodarska jedinicat

Sucha LZD

Early

Wał Ruda

Late Kasne

Wał Ruda

Kasne

30.46

Dominikowice Dominikowice

Gorlice Late

48a

Małastów

Early Rane

4.96

Małastów

Early Rane

58d

Juszczyn

Kasne

5.19

Tylicz

Late

68d

Stańcowa

Thinning Prorede Early Rane

Stańcowa

Dąbrowa Tarno-wska

Forest Inspectorate Šumski predjel

Table 1 Characteristics of stands on sample plots Tablica 1. Sastojinske značajke na pokusnim plohama

140

200

202

III

146

III

239

182

444

I.5

503

computer with specialist »Timing« software for conducting time studies (Sowa et al 2007). The registered duration of specific operations was assigned to given categories according to BN-76/9195-01 in the National Forest Equipment System (Botwin 1993). The outline of the classification of the operational work time and the adopted symbols are presented in Table 3.

257

J. M. Sowa and G. Szewczyk

Time Consumption of Skidding in Mature Stands Performed ... (255–264)

Table 2 Technical data of Fransga rd V6000GS winch powered by a Pronar 5112 farm tractor Tablica 2. Značajke vitla Fransga rd V6000GS i ATP-a Pronar 5112 Fransga rd V6000GS 1. 2.

Height / width, mm

860/1700

Visina / širina, mm Weight, kg

550

Težina, kg Pulling force, kN

Vučna sila, kN Power consumption, kW

37–67

Potrebna snaga, kW Rope diameter, mm

Promjer užeta, mm Rope length, m

50–120

Duljina užeta, m Winding speed (at 540 min-1), m/s Brzina užeta (pri 540 okretaja/min), m/s Height (without safety shield), mm Visina (bez zaštitnoga okvira), mm Total height, mm

Tc =

0.5–1.3 1660 2400

Ukupna visina, mm Pronar 5112

10. 11.

12.

13.

Dimensions: length / width / height, mm Dimezije: duljina / širina / visina, mm Front/rear wheel track

4130/1960/2560 1570-1730/15001800

Prednji/stražnji kotači Weight, kg

4040

Težina vozila, kg

Diesel 60 kW/2300 min-1

Engine type – Vrsta motora

Where: I – number of trees before felling on circular plots, L – number of trees removed from circular plots, Wiip – index of quantitative harvesting intensity, Wmip – index of harvesting intensity in terms of volume (Wmip = timber volume removed from circular plots / timber volume before felling on circular plots *100 %). The Wsip index, expressed in this way, reflects spatial distribution of the harvested volume on a timber handling site. For specific stands, the Wszt index, which described the number of trees removed from 1 ha, was determined. In order to obtain more stable results, time consumption was calculated by relating the obtained timber volume to the operational time T02 (4) (Giefing and Gackowski 2001). T02 M

(4)

Where: Tc – time consumption, T02 – operational time, M – timber volume. In order to achieve accordance of the time consumption, calculated for specific sections, with the normal distribution as well as due to the lack of uniformity of variance, analysis of differences of the mean values of time consumption was conducted using the parametric t-Student test. Examination of the dependence of the time consumption observed at work stands on stand characteristics, felling intensity indexes, timber characteristics and factors of the working day structure was conducted using multiple regression procedures. The significance of null hypotheses H0 was determined for the level of significance α = 0.05. Statistical calculations were done using STATISTICA 6 PL program.

4. Results – Rezultati On completion of the field work, the volume of the obtained timber was calculated, stock-taking was performed of the trees remaining on the circular plots and the intensity of planned thinning was determined (2), (3). Wiip =

I × 100 L

Wsip =

258

Wiip Wmip

(2) (3)

The research was carried out in 24 plots, 3 in each selected stand for each thinning category. Felling intensity was determined at 768 measurement points (circular research plots), where stock-taking of 4,360 trees was performed. The harvesting resulted in the removal from the circular plots of 620 trees, which constituted about 14% of their number and 145 m3, i.e. over 9% of the volume of trees recorded before the operation. Table 4 presents the mean values of the index of the total harvesting intensity Wsip, calculated for the analyzed conditions. Croat. j. for. eng. 34(2013)2

Time Consumption of Skidding in Mature Stands Performed ... (255–264)

J. M. Sowa and G. Szewczyk

Table 3 Work time classification Tablica 3. Turnus rada

T02 – Operativno vrijeme rada

T02 – Operatio-nal work time

Effective worktime Efektivno radno vrijeme

T21

Auxiliary time Pomoćna vremena rada

Time of skidding

T13

Vrijeme privlačenja drva Time of waiting for help in task execution or for the end of other activities Vrijeme čekanja (za pomoć pri radnoj operaciji ili da završi neka druga radna operacija)

T22 T23 T24

Time of walking in workplace Vrijeme kretanja radnika po radilištu Time of load attachment and detachment Vrijeme vezanja i odvezivanja drva Time of unlocking skidded timber Vrijeme oslobađanja zapelih tovara

Table 4 Indexes of quantitative harvesting intensity Wiip and the total harvesting intensity Wsip in the analyzed stands Tablica 4. Udio broja (Wiip) i volumena (Wsip) posječenih stabala na primjernim plohama

Fig. 1 The structure of operational work time for skidding performed with the farm tractor Slika 1. Udjeli vremena tijekom radnoga turnusa

The Wiip index reached higher values in the early thinning stands. Its level ranged from 9.7 to 16. In all cases, the percentage of the number of removed trees was always higher than the volume removed. The analyzed index reached the highest values in early thinning in beech and pine stands. The highest value Wsip, i.e. 2.09, was observed in spruce stands in late thinning whereas the lowest one, amounting to 0.79, was observed for the late thinning in beech stands. The mean Wsip values were by almost 20% higher in early thinning stands. The highest values were observed in Croat. j. for. eng. 34(2013)2

Stand – Sastojina i vrsta prorede

Wiip

Wsip

Beech, late thinning – Bukva, kasne prorede

8.45

0.79

Fir, late thinning – Jela, kasne prorede

8.86

1.25

Pine, late thinning – Bor, kasne prorede

10.12

1.18

Spruce, late thinning – Smreka, kasne prorede

8.06

2.09

Beech, early thinning – Bukva, rane prorede

16.00

1.33

Fir, early thinning – Jela, rane prorede

10.66

1.43

Pine, early thinning – Bor, rane prorede

13.70

1.29

Spruce, early thinning – Smreka, rane prorede

9.70

1.86

Early thinning – Rane prorede

11.74

1.50

Late thinning – Kasne prorede

9.09

1.30

Total – Ukupno

10.77

1.47

spruce stands (Wsip = 2.09 in late thinnings and Wsip = 1.86 in early thinnings), where in both categories the Wsip was much higher than in the other cases. The Wsip index, when calculated individually for each research plot, was then included in the equations of regression describing the time consumption of timber harvesting (EST). During harvesting and skidding operations, a time study was conducted for the operations performed during work on the research plots. The measurement base of the duration of the distinguished operation

259

J. M. Sowa and G. Szewczyk

Time Consumption of Skidding in Mature Stands Performed ... (255â&#x20AC;&#x201C;264)

Fig. 2 Time consumption calculated for the analyzed work-stand Slika 2. Operativno vrijeme rada

categories included 7,034 cases whose total time exceeded 70 hours. Part of the measured shift time (the time category) was included in the equations describing the time consumption of timber harvesting (EST). The coefficient of the use of the operational shift time was on the level of approximately 0.75, which points to considerable reliability of the equipment used and to good work organization. Fig. 1 presents the percentages of operations observed in the operational time at the examined work-stand. The examined skidding operations were characterized by a very high (36%) share of the time of walking in the workplace (T22). When walking, the winch operator extended the collective rope and fastened a few skidded logs with attachment ropes (between 4 and 6 at a time), which imposed long walking times. The skidding operation itself was not time-consuming (T13 amounted to 20%) but the auxiliary operations of attaching and detaching logs increased the time consumption considerably (30%). Fig. 2 presents the time consumption calculated for the analyzed work-stand. The results of the differentiation analysis of the mean time consumption values in subsequent sections are presented in Table 5.

260

Skidding operations were performed with the mean time consumption of 18.45 min/m3, which was a value close to the time consumption observed in the skidding technology where winches were powered by chainsaw (Szewczyk 2009). The mean time consumption of skidding was low. Under such conditions, low timber volume was achieved in the early thinning. Statistically significant differences were observed between the levels of time consumption in the early-thinning stands (24.59 min/m3) and the latethinning stands (13.10 min/m3). The lowest time consumption was observed in late-thinning pine stands (9.85 min/m3) and the highest in early-thinning beech stands (31.31 min/m3). The level of time consumption is one of the factors that allows for determination of usefulness of a given technology to perform specific forest management tasks. For this reason, the term time consumption will be replaced below by another, proposed by the present authors, namely EST â&#x20AC;&#x201C; the synthetic index of Empirical Technological Efficiency (Szewczyk 2010). In this case, the EST coefficient is the time consumption assessed on the basis of stand parameters, skidded timber, working day structure. The parameters of the equaCroat. j. for. eng. 34(2013)2

Time Consumption of Skidding in Mature Stands Performed ... (255–264)

J. M. Sowa and G. Szewczyk

Table 5 Significance of differences between the mean time consumption values in the operational time for skidding with the use of the farm tractor in stands with late (right up) and early (left down) thinning Tablica 5. Razlike u srednjim vremenima privlačenja drva u ranim i kasnim proredama (SD – značajna razlika, ID – beznačajna razlika) Species – Vrsta drveća

Pine – Bor

Beech – Bukva

Fir – Jela

Spruce – Smreka

ID; p = 0.07

ID; p = 0.12

ID; p = 0.07

Beech – Bukva

SD*; p = 0.02

ID; p = 0.70

ID; p = 0.15

Fir – Jela

SD; p = 0.04

ID; p = 0.10

ID; p = 0.06

Spruce – Smreka

ID; p = 0.16

ID; p = 0.78

ID; p = 0.31

Pine – Bor

tions (Tab. 6) allowing for the approximation of the EST level were estimated for factors related to the features of the stand, skidded timber and elements of the working day structure. Table 6 also presents the values: R, R2, Std error, test values and the probability level p. In the applied model, most variables were stable (terrain features) while others, related to the stand (intensity indexes Wsip, Wszt) and to the working day structure (percentages of times T13, T22), were taken into consideration. Selection of a set of independent variables in equations approximating the work consumption level

was based on the assumption that it should be jointly influenced by factors related to: stand structure (stand, cutting category - early or late thinning), elements of the working day structure (the share of specific time categories in a shift) and volume of harvested timber. Therefore, developing the equations generally consisted in removing those factors that did not significantly affect the estimated time consumption from the widest possible range of independent variables. It was always done so as to make all groups of variables visible in the equations. To sum up, the method applied was multiple backward stepwise regression.

Table 6 Parameters of the regression equations of the EST index Tablica 6. Regresijska analiza indeksa EST

Nr. Br.

Equation – accuracy of adjustment

Independent variables

Parametri regresije

Nezavisne varijable

EST R

Early thinning Rane prorede

Late thinning Kasne prorede

Total Ukupno

0.66

0.78

R2pop

0.41

0.59

21.98 0.00

29.35 0.00

Error ± Pogreška ±

14.36

3.93

Variable – Varijable

Beta b

Error std.

Constant

50.34

–

T22, %

-129.1

Wszt

6.61

7.61

0.00

-0.64

20.30

-6.36

0.00

0.04

0.25

0.02

2.55

0.01

Constant

11.84

–

2.09

5.65

0.00

T13, %

27.43

0.30

7.72

3.55

0.00

T22, %

-27.27

-0.47

4.98

-5.48

0.00

Wsip

2.39

0.41

0.51

4.71

0.00

Constant

30.85

–

3.36

9.18

0.00

6.62

0.22

2.84

2.33

0.02

T22, %

-75.43

-0.49

10.90

-6.91

0.00

Wszt

0.03

0.26

0.01

2.71

0.01

0 = Late thinning – Kasne prorede 0.64

0.39

Croat. j. for. eng. 34(2013)2

26.69 0.00

11.73

1 = Early thinning – Rane prorede

Pogreška

261

J. M. Sowa and G. Szewczyk

Time Consumption of Skidding in Mature Stands Performed ... (255–264)

Predicting the level of time consumption on the basis of the volume of a single piece of timber, the number of pieces skidded in a single cycle and e.g. the skidding distance obviously yield fairly precise results under specific stand conditions. However, such an approach does not consider the spatial distribution of the volume of skidded timber on a plot or the method of work in a shift. Such factors are included in the model proposed in the present study. The strongest correlations between the EST level at the work-stand of the WINCH OPERATOR in mature stands were found for the following variables: walking time T22 (β = -0.49) and Wszt (β = 0.26). The estimated time consumption in early-thinning stands should be by about 6 min/m3 higher in comparison with latethinning stands. An increased share of walking time T22 in a shift results in a decrease in time consumption (parameter -75.43), which at first sight seems incorrect as the T22 time is the auxiliary time rather than the effective time. However, it must be noted that at the analyzed work-stand this time is connected with walking in order to attach to the rope several logs, which are then skidded together as a bunch. Thus its higher share results in higher volume of one load of skidded timber, which lowers the level of time consumption. This phenomenon was better visible in the case of assessing the EST in early-thinning stands (βT22 = -0,64), which is understandable considering lower volume of a single log. This would point to the need to carry out skidding of whole bunches of logs on a rope in stands of younger age classes. The analyzed spatial relation (multiple independent variables) is presented in this study as a polynomial relation of the first degree of multiple variables. For the purpose of assessing the time consumption for forestry operations, the linear regression model is the most frequently used by researchers (Häberle 1990, Samset 1990, Lukáč et al. 2000, Bibliuk 2004, Messingerová 2005, Sowa et al. 2009, Sowa and Szewczyk 2008, Szewczyk 2010). The total time consumption of timber harvesting technologies, taking into consideration the various operations involved, may be assessed by totaling appropriate multiple regression equations calculated for specific operations (Zečić and Marenče 2005). This may be the method of predicting the level of time consumption for different technological variants (logical from the point of view of work organization). In the present study, the EST index was expressed as several linear functions of multiple variables. There are always two groups of variables, which generally characterize a stand and the character of stand management operations (the first group), as well as the percentages of the selected elements of the working

262

day structure in the operational time, describing the basic characteristics of timber harvesting technologies (the second group). Their changes are due to differentiation of stand features and, since the examined times are generally the skidding times, they complement the variables included in the first group. Such a comprehensive approach is an innovative solution, proposed by the present authors.

5. Conclusions – Zaključci The average time consumption of skidding by means of the cable winch powered by a farm tractor in the operational time, assessed in the examined mature stands, amounted to approximately 18 min/m3. Significant differences were observed in the levels of time consumption between early thinning (about 24 min/m3) and late thinning (13 min/m3). Differences in measurements of time consumption between early and late thinnings could have resulted from different volumes of single timber pieces and from different distances between trees that remained in the stand. The structure of the operational time of skidding by means of the cable winch in mature stands was characterized by a large share of auxiliary times T2: 80%, of which the walking time T22 accounted for as much as 36% while load attaching and detaching T23 accounted for 30% of the time. An equation of multiple regression was elaborated for the purpose of describing the changes in the level of time consumption of skidding, namely the Empirical Technological Efficiency index (EST). The EST depends on environmental factors (stand, felling category), elements of the working day structure (the share of an appropriate time category in a shift), characteristics of the harvested timber (volume) and operation intensity (the Wiip indexes of quantitative harvesting intensity and the Wsip index of total harvesting intensity). The strongest correlations between the EST and the analyzed variables were established for the factors connected with the percentage of walking time T22, which is related to the binding of a larger number of timber pieces, skidded in a single cycle. High time consumption of the examined skidding technology and a large share of the time of waiting for help indicate difficult work conditions in stands of middle age classes, where the skidded logs are frequently blocked and problems occur in connection with controlling the skidding from the skid trail. The measurements of thinning intensity, used for the approximation of the EST, may be determined prior to forest management operations based on the Croat. j. for. eng. 34(2013)2

Time Consumption of Skidding in Mature Stands Performed ... (255–264)

data from the Forest Management Regulations and Standing Timber Assessment. It allows for the rational design of the most effective technological solutions. This makes it possible to apply the results directly in given field conditions of timber harvesting.

6. References – Literatura Bibliuk, N. I., 2004: Forestry transport in Ukrainian Carpathians: main stages and tendencies of development. Proceedings of conference: Forest engineering: new techniques, technologies and the environment, Lwiw, Ukraine 2004, 183–191 p. Botwin, M., 1993: Podstawy użytkowania maszyn leśnych, Wydawnictwo SGGW, Warszawa, p. 1–120. Cserjes, M., 1989: Az erdészeti munkanormák készitése és alkalmazása, Erdeszeti Kutatasok 38: 209–213. Derek, J., 2004: Normy w górę, ceny później, Trybuna Leśnika 2: 10. Döhrer, K., 1998: Prämien-Zeitlohn für die Waldarbeit, AFZ Wald, Jg. 53(22): 1350–1353. Giefing, D. F., Gackowski, M., 2001: Ekonomiczna efektywność pozyskiwania drewna krótkiego w drzewostanach III kl. wieku w zależności od zastosowanych urządzeń zrywkowych, Polska Akademia Umiejętności, Prace Komisji Nauk Rolniczych 3: 17–26. Gil, W., 2007: Badania porównawcze ciągników rolniczych jako środków zrywkowych w wybranych zakładach usług leśnych, Zeszyty Naukowe Akademii Rolniczej im. Hugona Kołłątaja w Krakowie 435: 1–128. Grzegorz, C., 2003: Nowe katalogi pracochłonności, Las Polski 24: 21. Häberle, S., 1990: Grundzüge forstlicher Zeitstudien und ihrer Auswertung, Forstarchiv, Jg. 61, H.1: 27–32. Kapral, J., 2000: Wytyczne Lasów Państwowych w zakresie kooperacji z sektorem usług leśnych oraz sposoby jego wspierania. Proceedings of conference: Prywatny sektor usługowy w leśnictwie – Stan i perspektywy rozwoju. Międzynarodowe Targi Bydgoskie »Sawo«, Akademia Rolnicza w Poznaniu, Stowarzyszenie Przedsiębiorców Leśnych w Gołuchowie. Tuchola, Polska 2000, p. 9–18. Kusiak, W., 2006: Spotkanie przedsiębiorców leśnych z Dyrektorem Generalnym Lasów Państwowych, Przegląd Leśniczy 4: 9–11.

Croat. j. for. eng. 34(2013)2

J. M. Sowa and G. Szewczyk

Laurow, Z., 2000: Systemy pozyskiwania drewna – nazwy i określenia, Głos lasu 8: 10–11. Lukáč, T., Tajboš, J., Koreň, J., 2000: Analýza prevádzkových parametrov traktora LKT 81 Turbo Eko, Journal of Forest Science 46(6): 265–274. Lukačka, M., 1989: Ako dalej v normotvornej činnosti? Les 9: 21. Messingerová, V., 2005: Technológia vzdušnej dopravy dreva v lesnictve. Technicka Univerzita vo Zvolene, Lesnicka Fakulta, p. 81–87. Monkielewicz, L., Czereyski, K., 1971: Analiza metod ustalania technicznych norm pracy przy pozyskaniu i transporcie drewna, Prace Instytutu Badawczego Leśnictwa 390: 1–77. Sajkiewicz, A., 1981: Ekonomika pracy. Państwowe Wydawnictwo Ekonomiczne. Warszawa, 446 p. Samset, I., 1990: Some observations on time and performance studies in forestry, Meddeleser fra Norsk Institut for Skogforskning 43(5): 1–80. Sowa, J. M., Kulak, D., Szewczyk, G., 2007: Costs and efficiency of timber harvesting by NIAB 5-15 processor mounted on a farm tractor, Croatian Journal of Forest Engineering 28(2): 177–184. Sowa, J. M., Szewczyk, G., Stańczykiewicz, A., Grzebieniowski, W., 2009: Pracochłonność pozyskiwania drewna w drzewostanach ze śniegołomami. Leśne Prace Badawcze vol. 70(4): 429–434. Sowa, J. M., Szewczyk, G., 2008: Czasochłonność pozyskiwania drewna z użyciem procesora NIAB 5-15 w drzewostanach trzebieżowych. Acta Agraria et Silvestria, series Silvestris (XLVI): 53–66. Szewczyk, G., 2009: Możliwości wykorzystania wskaźników intensywności trzebieży w kategoryzacji warunków pracy dla wybranych technologii pozyskiwania drewna na ręczno -maszynowym poziomie zmechanizowania. Acta Agraria et Silvestria, series Silvestris (XLVII): 27–44. Szewczyk, G., 2010: Czasochłonność zrywki konnej w drzewostanach trzebieżowych. Sylwan CLIV(1): 52–63. Więsik, J., 2000: Prywatyzacja wykonawstwa prac leśnych w Polsce, Las Polski 23: 10–12. Zečić, Ž., Marenče, J., 2005: Mathematical models for optimization of group work in harvesting operation, Croatian Journal of Forest Engineering 26(1): 29–37.

263

J. M. Sowa and G. Szewczyk

Time Consumption of Skidding in Mature Stands Performed ... (255–264)

Sažetak

Utrošci vremena prilikom privlačenja drva ATP-om u proredama zrelih sastojina Ovim se istraživanjem raščlanio utrošak vremena prilikom privlačenja drva adaptiranim poljoprivrednim traktorom Pronar 5112 s ugrađenim vitlom Fransgård V6000GS u ranim i kasnim proredama. Značajke vozila i pripadajućega vitla prikazane su u tablici 2. Istraživanje je provedeno u državnim šumama grada Krakówa, Katowica i pokusnih šumskih sastojina u Krynici u zrelim borovim, jelovim, smrekovim i bukovim sastojinama (značajke sastojina prikazane su u tablici 1). Sastojine su odabrane zbog zadovoljavajuće gustoće sklopa i dimenzija stabala. Postavljene su pokusne plohe veličine 50 × 100 m (0,5 ha površine) tako da je dulja stranica pokusne plohe bila prislonjena uz traktorsku vlaku. Na svakoj plohi postavljene su 32 plohice, površine 50 m2 svaka, na kojima su izmjerena sva stabla deblja od 7 cm prsnoga promjera. Korištena je stablovna metoda izradbe drva, a duljina skupljanja drva vitlom bila je do najviše 50 m. Proveden je studij rada i vremena pri privlačenju drva te je prosječno (operativno) vrijeme iznosilo 18,85 min/m3. Vrijeme radnih operacija mjerilo se pomoću računala PSION Workabout i programskoga paketa »Timing«. Turnus rada prikazan je kroz efektivno vrijeme rada (privlačenje drva) i pomoćna vremena (vrijeme čekanja, kretanje radnika po radilištu, vezanje tovara i odvezivanje privučenoga drva), što je i prikazano u tablici 3. Uočene su značajne (signifikantne) razlike prilikom privlačenja drva u ranim (24,59 min/m3) i kasnim (13,10 min/m3) proredama (tablica 5). Udio je pomoćnih vremena najveći, čak 71 %, od čega 30 % pripada vezanju i odvezivanju tovara, a 36 % kretanju radnika po radilištu (slika 1 prikazuje postotne udjele vremena radnih operacija). Udio ukupnoga vremena, ovisno o vrsti prorede (rane/kasne) i mjestu istraživanja, prikazan je na slici 2. Regresijska analiza izmjerenih vremena prikazuje utjecaj pojedinih čimbenika radilišta na proizvodnost, tzv. empirijski indeks učinkovitosti, tj. EST. Razlike nastaju zbog sastojinskih prilika (tablica 1), sastavnica turnusa rada i količine posječenoga drva Wiip i Wsip (intenziteti sječe prikazani su u tablici 4). Najjača združenost podataka primijećena je između vremena potrebnoga za vezanje i odvezivanje tovara i intenziteta sječe u pojedinoj sastojini. Ključne riječi: utrošak vremena, privlačenje drva, tehnologije pridobivanja drva, korištenje šuma

Authors’ address – Adresa autorâ: Prof. Janusz M., Sowa, PhD. e-mail: rlsowa@cyf-kr.edu.pl Grzegorz Szewczyk, PhD.* e-mail: rlszewcz@cyf-kr.edu.pl Agricultural University of Cracow Faculty of Forestry Department of Forest and Wood Utilization Al. 29 Listopada 46 31-425 Kraków POLAND Received (Primljeno): September 24, 2012 Accepted (Prihvaćeno): January 11, 2013

264

*Corresponding author – Glavni autor Croat. j. for. eng. 34(2013)2

Original scientific paper – Izvorni znanstveni rad

Long Term Repair and Maintenance Cost of some Professional Chainsaws Angela Calvo, Marco Manzone, Raffaele Spinelli Abstract – Nacrtak A sample of 44 professional chainsaws was used to determine service life and maintenance cost for this type of equipment. Data were sourced from a large workshop, catering for regional forest crews in north western Italy. Chainsaw service life exceeded 3000 hours, and spread over a period of 6 to 9 years. Under these conditions, maintenance cost averaged 820 €, or about 120 % of the investment cost. Annual usage was highest and maintenance cost lowest for mid-size chainsaws, in the 2 to 3.5 kW power class. Models at the two extreme ends (i.e. < 2 kW and >3.5 kW) configured as specialist tools, which resulted in a lower usage and a higher maintenance cost per hour. The largest number of interventions (45% of the total) concerned the engine and the carburettor. The average chainsaw in the sample underwent 31 maintenance interventions over its service life. The cost per intervention varied between 7 and 50 €. Intervention cost was highest for engine work, and the lowest for overhaul. Overall, parts accounted for two-thirds of the cost, and labour for the remaining third. This study offers information about chainsaw service life and maintenance cost, obtained with scientific methods and hence suitable for general use. Keywords: logging, harvesting, biomass

1. Introduction – Uvod Italian forestry is characterized by steep terrain, ownership fragmentation and the application of closeto-nature management criteria, such as continuouscover forestry (Mason et al. 1999). All these factors tend to slow down the inevitable introduction of mechanized harvesting (Febo et al. 1997), determining the current prevalence of labor-intensive operations (Magagnotti et al. 2012). Under these conditions, versatile low-investment machinery offers a suitable balance between capital and labor inputs (Picchio et al. 2009). For these reasons, chainsaws and modified farm tractors are the backbone of the Italian forest machine fleet (Spinelli et al. 2013). In fact, motor-manual felling with chainsaws is also used in the Nordic countries, where it is favored by small-scale operators, especially when dealing with biomass production (Laitila et al. 2007). Many studies have addressed the productivity and cost of low-investment operations, based on chainsaws and farm tractors (Spinelli and Magagnotti 2012). However, most of these studies are relatively weak on Croat. j. for. eng. 34(2013)2

costing. One of their main weaknesses is the adoption of conventional assumptions, which may not reflect current practice (Rozt 1987). The international scientific literature offers no updated information about the annual use, service life and maintenance cost of these machines. Year after year, authors use the same assumptions, originating from practical experience gained several decades ago, when both chainsaws and farm tractors were much different from the chainsaws and farm tractors we use today (Ward et al. 1985). When chainsaws are concerned, studies offer general cost estimates, often obtained from secondary sources (Long et al. 2002, Mousavi et al. 2006). Most of these estimates date back to the early ‘80s (Miyata 1980). In fact, chainsaws are no longer considered in the updated versions of earlier machine rate compendia (Brinker et al. 2002). The only recent study offering good detail about chainsaw cost concerns its use in sawmilling – not in forestry practice – and therefore represents a peculiar case (Smorfitt et al. 2006). While the international scientific community is working at improving cost methods for use at a glob-

265

A. Calvo et al.

Long Term Repair and Maintenance Cost of some Professional Chainsaws (265–272)

al scale, very few people are working at developing reliable input assumptions. As a result, the lack of quality inputs may thwart all attempts to enhance the accuracy of machine cost estimates. Therefore, the goal of this study was to provide reliable information on the service life and maintenance cost of professional chainsaws used in forest operations.

2. Materials and methods – Materijal i metode The study was conducted in cooperation with a regional forest administration in north western Italy. The regional administration maintains its own logging crews, tasked with performing silvicultural operations in public forests. Regional crews are specifically trained for the task, and must attend several training courses depending on the task. Before using a chainsaw, operators must attend chainsaw use and maintenance courses. Individual chainsaw operators are responsible for the good use and maintenance of the machines they are assigned, and they are equipped accordingly. Operators conduct all minor maintenance, and especially the daily and weekly routines. Major maintenance after severe failure or fatigue is performed by professional mechanics at a centralized workshop. All maintenance interventions conducted in this workshop are recorded in a logbook, together with information about chainsaw type, model, serial number, age and worked hours. Therefore, it is possible to reconstruct all maintenance interventions conducted over the whole service life of each chainsaw in

the regional fleet, as well as the duration of their service life. For this study, we have collected and organized all the information available in the workshop logbook. It contained data about 44 chainsaws, as shown in Table 1. All chainsaws in the study were professional models, produced by the two largest chainsaw manufacturers, and namely: Husqvarna (25 units) and Stihl (19 units). The data represented light, medium and heavy chainsaws. However, the data pool was not balanced in terms of machine size and make, which prevented proper comparisons between types and makes. The characteristics of the machines in the regional fleet reflected those of local forests and silviculture, which justified a strong bias in favor of medium-size chainsaws. Furthermore, the uneven distribution of age classes between makes depended on the variable success of the two chainsaw makes with public bids. For the purpose of this study, maintenance interventions have been categorized into eight main classes, corresponding to the main constructive elements of a chainsaw and/or intervention type. The following categories were separated: general overhaul; engine repairs; crankcase repairs; carburettor issues; starter; electric system; chain and chain bar; safety devices. Intervention cost was calculated by summing the costs of labor and spare parts. The former was estimated to 24 € per hour, taxes and benefits included. The latter was the actual cost indicated in the repair bills, after discounting to present value. Statistical analysis of data was conducted with the Tukey-Duncan test at the 5% level, and with linear regression (SAS 1999).

Table 1 Chainsaw characteristics in the study Tablica 1. Karakteristike motornih pila uključenih u istraživanje Displacement

Avg. use

Interventions

Power

Guide bar

Snaga

Vodilica

Prosječna upotreba

cm3

Hours – Sati

Class

Make

Model

Machines

Razred

Proizvođač

Model

Uređaji

Obujam cilindra

Zahvati

Heavy – Teške

Stihl

MS660

91.6

5.2

2 850

315

Medium – Srednje

Stihl

MS440

70.7

4.0

3 600

190

Medium – Srednje

Husqvarna

272XP

72.2

3.6

3 400

441

Medium – Srednje

Husqvarna

262XP

62.0

3.4

4 200

130

Medium – Srednje

Husqvarna

266XP

67.0

3.2

4 700

196

Light – Lagane

Stihl

MS200

35.2

1.8

1 000

103

Light – Lagane

Husqvarna

335XPT

35.2

1.6

400

266

Croat. j. for. eng. 34(2013)2

Long Term Repair and Maintenance Cost of some Professional Chainsaws (265–272)

Fig. 1 Frequency of the number of interventions by intervention type Slika 1. Učestalost broja intervencija prema vrsti zahvata Overall, this study contains information about 1 388 maintenance interventions, corresponding to a total expenditure of 36 970 €.

3. Results – Rezultati The service life of the chainsaws in the data pool ranged between 400 and over 4000 hours, with a

A. Calvo et al.

weighed average of 3 175 hours. Service life was significantly higher for medium-size chainsaws, than for the other types. Light chainsaws were characterized by a very low utilization, and especially the lightest type (Husqvarna 335 XPT). That depended on the almost exclusive use for park maintenance, and on the relatively young age of the machines in the study, which had been bought two years earlier and were still in service. Average machine age was 8 years, with wide variations. Resulting average utilization was 400 hours per year. The largest number of interventions (45% of the total) was made on the engine and the carburettor, and generally consisted of carburettor diaphragm replacements and piston-cylinder kit substitutions. In contrast, very few interventions concerned the starter (Fig. 1). The average chainsaw in our data pool underwent 31 maintenance interventions over an average service life of 3 175 hours. The cost per intervention varied between 7 and 50 € (Fig. 2). It was highest for engine work, and lowest for overhaul. As an average, the total cost per chainsaw was 840 €, over the whole service life considered in this study. The largest share of this cost was related to engine, carburettor and chain/bar issues. Overall, parts accounted for two-thirds of the cost, and labor for the remaining third. Assuming that the average investment cost of the chainsaws in the study pool was 700 €,

Fig. 2 Average cost per intervention and machine, by intervention type Slika 2. Prosječan trošak po intervenciji i uređaju prema vrsti zahvata Croat. j. for. eng. 34(2013)2

267

A. Calvo et al.

Long Term Repair and Maintenance Cost of some Professional Chainsaws (265–272)

then cumulative repair and maintenance amounted to 120% of investment cost. Conducted at the 5% level, the Tukey-Duncan test allowed grouping machines in three categories, as a function of the number of interventions per machine (Fig. 3). Machines with a power exceeding 3.5 kW were characterized by a significantly higher number of interventions per machine. The number of interventions was lowest for chainsaws with a power below 2 kW. Machines with a power between 2 and 3.5 kW were in between, requiring more maintenance than smaller machines, but less maintenance than larger ones. Overall, average maintenance time over the whole machine life was 18 hours for machines with power above 3.5 kW, 8 hours for machines with power between 2 and 3.5 kW, and 6 hours for machines with power below 2 kW. Statistical analysis also showed that the total number of interventions was independent from machine life: chainsaws in the intermediate power class (i.e. 2 kW > power > 3.5 kW) had a longer service life than heavier chainsaws, but underwent fewer interventions. Overall, chainsaw maintenance incurred a cost between 0.13 and 0.50 € per hour (Table 2). Again, statistical analysis showed that machines could be divided into three groups. Chainsaws in the intermediate power class (i.e. > 2 and < 3.5 kW) incurred a significantly lower maintenance cost than all the other chainsaws. Chainsaws with a power below 2 kW or above 3.5 kW

incurred a maintenance cost of 0.36 € per hour, i.e. more than twice as large as incurred in the intermediate power class (i.e. 14 € hour–1). In this respect, there were no significant differences between chainsaws in the two extreme power classes. Table 2 Maintenance cost Tablica 2. Troškovi održavanja Power – Snaga

Cost – Troškovi

€ hour–1

5.2

0.38

4.0

0.26

3.6

0.45

3.4

0.13

3.2

0.14

1.8

0.50

1.6

0.21

4. Discussion – Rasprava Service life is much longer than indicated in previous studies, which quote between 1000 (Miyata 1980) and 2000 hours (Piegai et al. 2010; Spinelli and Mag-

Fig. 3 Number of interventions by intervention type and chainsaw power class (P) Slika 3. Broj intervencija prema vrsti zahvata i razredu snage motorne pile (P)

268

Croat. j. for. eng. 34(2013)2

Long Term Repair and Maintenance Cost of some Professional Chainsaws (265–272)

agnotti 2011). Annual use is much below the 1 000 hours quoted by Miyata (1980), but in line with the 500 hours reported by Piegai et al. (2010) and Spinelli and Magagnotti (2011). The very long service life is probably related to the peculiar strategies of public agencies, which are often equipped with their own repair workshops and tend to internalize repair costs. In that case, maintaining the workshop is a fixed cost, and reducing the cost of repairs by a more frequent machine turnover may not lead to any savings in terms of overall management cost. Extended service life is consonant with an accumulation of repair and maintenance cost, eventually exceeding the investment cost. Even if extending service life for such a long time may prove economically viable, one should check the effect of machine aging on ergonomic performance, especially for what concerns vibration and noise (Martinić et al. 2011). Further studies should clarify if proper maintenance can prevent the decay of ergonomic performance, despite use and age. It is also worth noticing that the utilization of medium-size chainsaws is significantly higher than the use of the extreme models. That is compatible with the higher versatility of machines in the central class, which are used more frequently. Both light and heavy chainsaws configure as specialist tools, used for special jobs only. Frequent use may partly explain the lower maintenance cost of chainsaws in the intermediate power class: both the operators and the mechanics were very familiar with these machines, which facilitated preventive maintenance, diagnostics and repairs. Furthermore, spares were more readily available, due to the larger number of machines in use. Differences in maintenance cost between power classes may also be explained by their different use, and the resulting different solicitations. Heavy chainsaws were used mostly for felling large trees, whereas small chainsaws were used in arboriculture, especially for pruning and delimbing. Medium-size chainsaws were the most versatile, used for a variety of jobs, and especially felling and processing. For this same reason, machines in the different power classes would have different construction characteristics, with light chainsaws being generally less sturdy than heavy or medium-size models, in view of their less intense use. This may have an effect on machine reliability and duration. Part of the variability may also be explained by individual model design: some models may have a better design than others, leading to differences in maintenance frequency and cost. As to hourly cost, the only available comparison is the figure reported by Smorfitt et al. (2006), which Croat. j. for. eng. 34(2013)2

A. Calvo et al.

amounts to 0.27 € hour–1, after discounting to 2013 values and converting Australian dollars to European currency. That fits well into our 0.13 – 0.50 € hour–1 bracket, and offers a close match to the 0.36 € hour–1 specifically found for heavy chainsaw models, such as those used in the Australian study. Despite their specific regional source, the hourly costs obtained from this study may be suitable for general use, or at least for representing a wider reality than actually probed. By itemizing maintenance frequency and cost, our study may point chainsaw manufacturers towards specific problem areas, where technological development is especially urgent. At present, engine and carburettor maintenance are still the most frequent and expensive. In their quest for lighter and more powerful engines, chainsaw manufacturers should not forget reliability, which is still a main issue. In contrast, the maintenance of compulsory safety devices (throttle safety catch, chain brake, etc.) incurs very little cost, denying earlier complaints that the (then) new devices would represent an additional complication, and the possible cause of further malfunctions. Unfortunately, this study could not establish a clear relationship between chainsaw age and maintenance cost. Hourly maintenance cost is supposed to increase with machine use, as a result of fatigue and general decay. Estimating this relationship in numerical terms would have helped decisions about the eventual decommissioning of older machines, for substitution with new models. However, the workshop bills contained no indication of the hours worked at the time of repair, and their dates were often unreliable. Consequently, it was impossible to associate maintenance costs with the hours worked by each machine, as needed for developing a predictive model. Future studies should address this subject, which is extremely important for operation managers. Finally, readers must recall that our study concerns major maintenance performed at a workshop, and not daily maintenance conducted in the field. However, field maintenance is generally minor, and consists of sharpening, cleaning and small repairs. We can safely assume that the largest component of field maintenance cost is labor, which is accounted for by including maintenance with delay time, as normally done in field studies (Magagnotti and Spinelli 2012).

5. Conclusions – Zaključci This study offers information about chainsaw service life and maintenance cost, obtained by scientific methods and suitable for general use. Contrary to previous assumptions, the service life of professional

269

A. Calvo et al.

Long Term Repair and Maintenance Cost of some Professional Chainsaws (265–272)

chainsaws can exceed 3 000 hours and span over up to 8 years. During this period, a chainsaw will undergo about 30 maintenance interventions, for a cost that is 1.2 times its purchase price. Engine and carburettor maintenance accounts for the largest number of interventions and the highest share of total maintenance cost. Versatile chainsaws in the intermediate (> 2 kW and < 3.5 kW) power class are characterized by the highest annual use and the lowest maintenance cost.

Mason, B., Kerr, G., Simpson, J., 1999: What is continuous cover forestry? Forestry Commission Information Note 29. Forestry Commission, Edinburgh. 8p.

6. References – Literatura

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. Croat J For Eng 30: 15–26.

Brinker, R., Kinard, J., Rummer, B., Lanford, B., 2002: Machine rates for selected forest harvesting machines. Circular 296 (Revised). Alabama Agricultural Experiment Station, Auburn University, AL. 32p. Febo., P, Pipitone., F, Peri., G., 1997: The preservation of Sicilian forests with poorly mechanized logging processes. J Agr Eng Res 67: 229–233. Laitila, J., Asikainen, A., Nuutinen, Y., 2007: Forwarding of whole trees after manual and mechanized felling bunching in pre-commercial thinnings. Int J For Eng 18: 29–39. Long, C., Wang, J., McNeel, J., Baumgras, J. (2002): Production and cost analysis of a feller-buncher in central Appalachian hardwood forest. Proceedings of the 25th COFE meeting, Auburn, Alabama. 5p.

Miyata, E. 1980: Determining fixed and operating cost of logging equipment. General Technical Report NC-55. Forest Service North Central Forest Experiment Station, St. Paul, MN. 14p. Mousavi, R., Nikouy, M., Uusitalo, J., 2011: Time consumption, productivity, and cost analysis of the motor manual tree felling and processing in the Hyrcanian Forest in Iran. J Forestry Res 22: 665–669.

Piegai, F., Fratini, R., Pettenella, D., 2010: Costi macchina. Confronto tra diversi metodi di calcolo. Supplemento scientifico degli approfondimenti di Sherwood – Foreste ed Alberi Oggi, Compagnia delle Foreste, Arezzo, Italy. 30p. Rozt, C., 1987: A standard model for repair costs of agricultural machinery. Appl Eng Agr 3: 3–9. SAS Institute Inc, 1999: StatView Reference. SAS Publishing, Cary, NC. ISBN-1-58025-162-5. p. 84–93. Smorfitt, D., Harrison, S., Herbohn, J., 2006: Short-run and long-run costs for milling rainforest cabinet wood timbers. Australian For 69: 223–232.

Magagnotti, N., Spinelli, R., Güldner, O., Erler, J., 2012: Site impact after motor-manual and mechanised thinning in Mediterranean pine plantations. Biosys. Eng. 113: 140–147.

Spinelli, R., Magagnotti, N., 2011: The effects of introducing modern technology on the financial, labour and energy performance of forest operations in the Italian Alps. For Pol Econ 13: 520–524.

Magagnotti, N., Spinelli, R. (Ed.) 2012: COST Action FP0902 – Good practice guideline for biomass production studies. CNR IVALSA. Florence, Italy. ISBN 978-88-901660-4-4. Available on line at: www.forestenergy.org. 41 p.

Spinelli, R., Magagnotti, N., 2012: Wood extraction with farm tractor and sulky: estimating productivity, cost and energy consumption. Small Scale For 11: 73–85.

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? Croat J For Eng 32: 431–441.

Spinelli, R., Magagnotti, N., Facchinetti, D., 2013: A survey of logging companies in the Italian Alps. Int J For Eng. In press. Ward, S., McNulty, P., Cunney, M., 1985: Repair costs of 2 and 4 WD tractors. Trans ASAE 28: 1074–1076.

Sažetak

Dugoročni troškovi popravaka i održavanja profesionalnih motornih pila Šumarstvo u Italiji obilježavaju strmi tereni, usitnjeno vlasništvo i primjena kriterija u gospodarenju bliskih prirodi. Svi ti čimbenici pomalo usporavaju neizbježno uvođenje mehaniziranoga pridobivanja drva i pridonose trenutačnomu prevladavanju radno intenzivnih operacija. U takvim uvjetima raznovrsna relativno jeftina mehanizacija pruža odgovarajuću ravnotežu između kapitalnih ulaganja i radnih resursa. Zbog toga motorne pile i adaptirani poljoprivredni traktori čine okosnicu talijanske šumske mehanizacije. Motorno-ručno obaranje s motornim pilama primjenjuje se također i u nordijskim zemljama, gdje je popularno kod malih izvoditelja šumskih radova, osobito u slučajevima vezanim uz proizvodnju biomase.

270

Croat. j. for. eng. 34(2013)2

Long Term Repair and Maintenance Cost of some Professional Chainsaws (265–272)

A. Calvo et al.

Mnoga su istraživanja razmatrala proizvodnost i troškove nisko investicijskih operacija na temelju motornih pila i poljoprivrednih traktora. Ipak, većina je tih istraživanja relativno nepouzdana što se tiče troškova. Jedna od njihovih glavnih slabosti je prihvaćanje konvencionalnih pretpostavki koje možda ne odražavaju sadašnju praksu. Međunarodna znanstvena literatura ne nudi nove podatke o godišnjoj uporabi, uporabnom vijeku i troškovima održavanja ovih strojeva. Iz godine u godinu autori se koriste istim pretpostavkama poteklim iz praktičnoga iskustva stečenoga prije nekoliko desetljeća, kada su se i motorne pile i poljoprivredni traktori znatno razlikovali od motornih pila i traktora koji se danas upotrebljavaju. Kada se promatraju motorne pile, istraživanja daju opće procjene troškova, često dobivene iz sekundarnih izvora. Većina tih procjena potječe iz ranih 80-ih. Ustvari, motorne pile više nisu uključene u ažuriranim verzijama ranijih pregleda zastupljenosti strojeva. Jedina novija studija s detaljnim troškovima motornih pila odnosi se na primjenu u pilanama, ali ne i u šumarstvu. Dok međunarodna znanstvena zajednica radi na poboljšanim troškovnim metodama za primjenu na globalnoj razini, vrlo je malo ljudi angažirano na razvoju pouzdanih ulaznih pretpostavki. Kao rezultat, nedostatak kvalitetnih ulaza može omesti sva nastojanja da se unaprijedi točnost procjena o troškovima strojeva. Zbog toga je cilj ovoga istraživanja bio pružiti vjerodostojnu informaciju o uporabnom vijeku i troškovima održavanja profesionalnih motornih pila u šumskim radovima. Istraživanje prikazano u radu provedeno je u suradnji s regionalnom šumskom upravom u sjeverozapadnoj Italiji. Regionalna šumska uprava servisira vlastite sjekačke ekipe, zadužene za obavljanje poslova u javnim šumama. Regionalne su ekipe posebno uvježbane za radne zadaće i moraju pohađati nekoliko tečaja osposobljavanja ovisno o vrsti zadaće. Prije korištenja motorne pile rukovatelji moraju polaziti tečaj o uporabi i održavanju motorne pile. Sami su sjekači individualno odgovorni za pravilno korištenje i održavanje pila za koje su se zadužili, i za to su odgovarajuće opremljeni. Sjekači provode sve manje održavanje i osobito uobičajene dnevne i tjedne postupke. Velike popravke i održavanje zbog ozbiljnih kvarova ili dotrajalosti obavljaju profesionalni mehaničari u središnjoj radionici. Svi se popravci i održavanje u toj radionici bilježe i evidentiraju u dnevniku, zajedno s podacima o tipu motorne pile, modelu, serijskom broju, starosti i radnim satima. Stoga je moguće rekonstruirati sve zahvate na održavanju svake motorne pile u regiji te trajanje njihova uporabnoga vijeka. Za ovo su istraživanje prikupljene i organizirane sve informacije dostupne u dnevniku radionice. To je obuhvatilo podatke o 44 motorne pile. Sve su motorne pile u istraživanju bile profesionalni modeli, koje su proizvela dva najveća proizvođača motornih pila: Husquarna (25 jedinica) i Stihl (19 jedinica). U podacima su zastupljene i lake, srednje i teške motorne pile. Ipak, podaci u bazi nisu ujednačeni s obzirom na veličinu i proizvođače strojeva, što je onemogućilo odgovarajuće usporedbe između tipova pila i proizvođača. Karakteristike motornih pila u regionalnoj upravi odražavaju značajke lokalnih šuma i njihova uzgajanja, što objašnjava veliku zastupljenost srednje velikih pila. Nadalje, neravnomjerna razdioba starosnih razreda između modela ovisna je o različitom uspjehu dvaju proizvođača motornih pila na javnim natječajima. Za potrebe istraživanja postupci su održavanja kategorizirani u osam glavnih razreda ovisno o glavnim konstruktivnim elementima pile i/ili vrsti intervencije. Razdvojene su ove kategorije: opći pregled, popravci motora, popravci kućišta radilice, pitanja rasplinjača, starter, električni sustav, lanac i vodilica, sigurnosni uređaji. Troškovi održavanja i popravaka izračunati su zbrajanjem cijene rada i rezervnih dijelova. Prethodno je procijenjeno na 24 € po satu, uključujući porez i doprinose. Potonje predstavljaju stvarne cijene navedene u računima popravaka nakon diskontiranja na sadašnju vrijednost. Istraživanje je obuhvatilo podatke za 1388 zahvata na održavanju, što odgovara ukupnim troškovima od 36 970 €. Rezultati pokazuju da uporabni vijek motornih pila premašuje 3 000 sati i da traje od 6 do 9 godina. U takvim su uvjetima troškovi održavanja prosječno iznosili 820 € ili oko 120 % troška investicije. Godišnja je uporaba bila najviša, a troškovi održavanja najmanji za srednje velike motorne pile u razredu od 2 do 3,5 kW snage. Modeli na dva ekstremna kraja (tj. < 2 kW i > 3,5 kW) kao specijalistički alati rezultirali su manjom uporabom i višim troškovima održavanja po satu. Najveći dio popravaka (45 % od ukupnoga broja) odnosi se na motor i rasplinjač. Prosječna motorna pila u uzorku prošla je 31 zahvat održavanja tijekom svoga uporabnoga vijeka. Trošak po zahvatu iznosio je između 7 i 50 €. Trošak je popravaka bio najveći za rad na motoru, a najniži za opći pregled. Rezervni su dijelovi iznosili dvije trećine ukupnoga troška, a rad preostalu trećinu. Razvrstavanjem učestalosti i troškova održavanja provedeno istraživanje može uputiti proizvođače motornih pila prema specifičnim problemskim područjima, gdje je tehnološki razvoj osobito nužan. Trenutačno su popravci motora i rasplinjača još uvijek najčešći i najskuplji. U njihovoj potrazi za lakšim i snažnijim motorima proizvođači ne bi trebali zaboraviti pouzdanost koja je i dalje glavno pitanje. S druge strane održavanje obaveznih sigurnosnih uređaja (kočnica lanca i sl.) izaziva vrlo malo troška, što pobija prijašnje prigovore da novi uređaji mogu biti dodatna komplikacija i mogući izvor daljnjih kvarova. Croat. j. for. eng. 34(2013)2

271

A. Calvo et al.

Long Term Repair and Maintenance Cost of some Professional Chainsaws (265–272)

Nažalost, provedeno istraživanje nije moglo utvrditi jasnu vezu između starosti motornih pila i troškova održavanja. Pretpostavlja se da trošak održavanja po satu raste s uporabom pile kao rezultat dotrajalosti i općega trošenja. Procjena toga odnosa u brojčanim iznosima pomogla bi u donošenju odluka o eventualnom otpisivanju starijih strojeva radi zamjene novim modelima. Međutim, računi radionice nisu sadržavali bilješke o učinjenim radnim satima u trenutku popravka, i njihovi su datumi često bili nepouzdani. To je onemogućilo povezivanje troška održavanja s radnim satima svake motorne pile, kao što je potrebno za razvoj prediktivnoga modela. Buduća bi istraživanja trebala razmotriti to pitanje koje je iznimno važno za operativno rukovođenje. Također na umu treba imati da je istraživanje obuhvatilo samo velike popravke obavljene u mehaničkoj radionici, bez dnevnoga održavanja koje se obavlja na terenu. Međutim, terensko je održavanje općenito manje i sastoji se od oštrenja, čišćenja i sitnih popravaka. Sa sigurnošću se može pretpostaviti da najveći dio u trošku terenskoga održavanja čini ljudski rad, koji je unaprijed uračunat uključivanjem održavanja motorne pile u dodatno vrijeme sječe i izrade, kako je to uobičajeno kod terenskih istraživanja. Provedeno istraživanje ipak pruža podatke o uporabnom vijeku i troškovima održavanja motornih pila, koje su dobivene znanstvenim metodama i stoga su pogodne za opću primjenu. Ključne riječi: sječa, pridobivanje drva, biomasa, motorne pile, uporabni vijek, troškovi održavanja

Authors’ address – Adresa autorâ: Assoc. Prof. Angela Calvo, PhD. e-mail: angela.calvo@unito.it Marco Manzone e-mail: marco.manzone@unito.it Università degli Studi di Torino DEIAFA sez. Meccanica Via Leonardo da Vinci 44 Grugliasco (TO) ITALY

Received (Primljeno): March 17, 2013 Accepted (Prihvaćeno): April 4, 2013

272

Raffaele Spinelli, PhD.* e-mail: spinelli@ivalsa.cnr.it CNR IVALSA Via Madonna del Piano 10 Sesto Fiorentino (FI) ITALY * Corresponding author – Glavni autor Croat. j. for. eng. 34(2013)2

Original scientific paper – Izvorni znanstveni rad

Self-Levelling Feller-Buncher Productivity Based on Lidar-Derived Slope Muhammad Alam, Mauricio Acuna, Mark Brown Abstract – Nacrtak The purpose of the study was to examine the ability of LiDAR (Light Detection and Ranging) to derive terrain slope over large areas and to use the derived slope data to model the effect of slope on the productivity of a self-levelling feller-buncher in order to predict its productivity for a wide range of slopes. The study was carried out for a self-levelling tracked feller-buncher in a 24-year old radiata pine (Pinus radiata) plantation near Port Arthur, Tasmania, Australia undertaking a clear felling operation. Tree heights and diameter at breast height were measured prior to the harvesting operation. Low intensity LiDAR (>3 points m-2) flown in 2011 over the study site was used to derive slope classes. A time and motion study carried out for the harvesting operation was used to evaluate the impact of tree volume and slope on the feller-buncher productivity. The results showed the ability of LiDAR to derive terrain slope classes. The study found that for an average tree volume of 0.53 m3, productivities of 97 m3 PMH0-1 (Productive Machine Hours excluding delays) and 73 m3 PMH0-1 were predicted for the moderate slope (11–18°) and steep slope (18–27°), respectively. The difference in feller-buncher productivity between the two slope classes was found to result from operator technique differences related to felling. The productivity models were tested with trees within the study area not used in model development and were found to be able to predict the productivity of the feller-buncher. Keywords: Tasmania, productivity, self-levelling feller-buncher, LiDAR, mechanised harvesting system, slope

1. Introduction – Uvod The productivity and efficiency of a mechanised harvesting system is affected by a number of factors including forest stand characteristics (stand density, undergrowth), tree characteristics (tree size or piece size, tree form, crown size), terrain variables (slope, rocks, woody debris, ground roughness, ground strength, streams and drainage features, roads, etc.), operators’ experience, skill & work technique and machinery limitations or design (Brunberg et al. 1989, Lageson 1997, Nurminen et al. 2006, Visser et al. 2009). Knowledge of the impact of these factors on the productivity and efficiency of forest harvesting machines can assist in predicting their performance under different conditions in a cost-effective way and lead to more productive harvesting operations. Tree size (volume or weight) has been determined by many studies to be the most influential factor afCroat. j. for. eng. 34(2013)2

fecting the productivity of forest harvesting machines (e.g. Brunberg et al. 1989, Kellogg and Bettinger 1994, Acuna and Kellogg 2009). However, slope is the primary determinant of travel speed and stability of harvesting machines (Davis and Reisinger 1990). Increasing slope has been shown to be a significant factor in decreasing the productivity of a range of forest harvesting equipment (Stampfer 1999, Stampfer and Steinmüller 2001, Simões and Fenner 2010, Zimbalatti and Proto 2010). In addition, rubber-tyred harvesting machines are generally restricted to slopes <19° whereas tracked machines can operate on slopes up to 27°, with some specialised tracked machines able to operate on steeper slopes (e.g. the Valmet Snake (Stampfer and Steinmüller 2001)). Previous harvester productivity studies have usually extrapolated stand-level slope information from a limited number of points manually measured with

273

M. Alam et al.

Self-Levelling Feller-Buncher Productivity Based on Lidar-Derived Slope (273–281)

clinometers across the study site or used DTMs (Digital Terrain Models) derived from contours (e.g. Acuna and Kellogg 2009, Oliveira Júnior E. D. de. et al. 2009). LiDAR is a well-recognised technology for the production of high quality DTMs (Ackermann 1999, Wehr and Lohr 1999), which can be used to visualise and calculate slope (Giles and Franklin 1998) and as an input into harvest planning (Reutebuch et al. 2005). LiDAR slope maps have been shown to be very accurate and of high resolution compared with DTMs derived from contour maps (Vaze and Teng 2007). Use of LiDAR to generate accurate, broad area DTMs makes it possible to predict the impact of slope on forest harvesting machine productivity across an entire forest estate using models relating productivity to slope. Self-levelling feller-bunchers have recently been introduced in parts of Australia and New Zealand for harvesting operations in steep terrain (Acuna et al. 2011). Self-levelling feller-bunchers have advantages over conventional feller-bunchers such as reducing the risk of tilting, increased lifting capacity and increased operator comfort during downhill operations (MacDonald 1999, Acuna et al. 2011). The objective of the study was to use LiDAR-derived slope from readily available low cost LiDAR data to develop a relationship between slope and the productivity of a self-levelling feller-buncher, and then to use this relationship to predict the productivity of the feller-buncher for other areas, where slope had been estimated using similar LiDAR data.

2. Material and methods Materijal i metode 2.1 Study site – Mjesto istraživanja The study was located near Port Arthur, Tasmania, Australia (Latitude/Longitude: 43°10’10” S / 147°47’ 20” E). The stand was 24-year-old radiata pine plantation of 1057 trees ha-1 with no undergrowth. The study site was an area of approximately 1 hectare within a plantation being clearfelled for pulp wood production. Tree spacing was 2.5 m × 4 m and lightly branchy trees were good in forms and quality. The site had never been thinned. The forest floor consisted of moist, soft and clay loamy soils. There were some dolerite rocks that accounted for the ground roughness. Ground slope was between 7–27° with an average of 21°. One hundred and two trees of normal growth and forms covering the range of heights at the study site were selected and their heights were measured with a Vertex hypsometer and Impulse 200 laser to the nearest 0.1 m. The diameter at breast height (DBH, cm) of

274

Table 1 Means and value ranges for pre-harvest tree measurements Tablica 1. Raspon i prosječne vrijednosti izmjere stabala prije sječe Mean

Range

Arit. sredina

Raspon

Height – Visina, m

26.1

10–37

DBH – Prsni promjer, m

0.29

0.10–0.46

Basal area – Temeljnica, m2

0.07

0.01–0.16

0.61

0.06–1.84

Volume – Obujam, m

all trees on the study site was measured with a diameter tape to the nearest 1 cm. A height-diameter model derived from the measured tree heights was used to estimate heights of the remaining trees. Each tree had a unique number painted on the stem to allow it to be identified during the time and motion study. A volume function supplied by Norske Skog, Australia was used to estimate each tree merchantable volume (m3). Means and value ranges for pre-harvest tree measurements are presented in Table 1.

2.2 Airborne LiDAR system – Sustav LiDAR LiDAR data covering the study site was supplied by Forestry Tasmania, with the specifications presented in Table 2. This LiDAR data was available as it had been collected for the purpose of resource and land management used by Forestry Tasmania that manages native and plantation forests in the region. LiDAR data supplied in .LAS format were classified into ground and non-ground points. LiDAR data accuracy was verified by the data provider. A DTM was constructed with a cell size of 2 m using ground LiDAR points and slope was derived from the DTM using ArcGIS 10. The terrain slope classification used by the Forestry Commission UK (1996) (Level = 0–6º, Gentle = 6–11º, Moderate = 11–18º, Steep = 18–27º, Very steep = >27º) was adopted in the study because: (i) the classes are based around operational considerations, (ii) there was no widely accepted terrain classification system in use in Australia, (iii) this classification is very similar to that used by the Forest Practices system in Tasmania (Forest Practices Board 2000) of Hilly =12–19º, Steep = 20–26º, Very Steep =27º and above and (iv) it is an internationally recognised classification.

2.3 Time and motion study – Studij rada i vremena An operator with twelve years experience (two years with current machine) carried out the harvesting Croat. j. for. eng. 34(2013)2

Self-Levelling Feller-Buncher Productivity Based on Lidar-Derived Slope (273–281)

M. Alam et al.

Table 2 LiDAR parameters and scanning system settings Tablica 2. Parametri LiDAR-a i postavke sustava snimanja LiDAR attribute – Obilježja LiDAR-a

Values – Vrijednosti

Date of flight – Datum leta

25/05/2011

System – Sustav

ALTM (Airborne Laser Terrain Mapping) Gemini

Beam divergence – Odstupanje pulsa

0.20 milliradian

Footprint diameter – Prostorna rezolucija

20 cm

Laser mode – Mod lasera

Single pulse -2

>3 m (1 , 2nd, 3rd and last) (2.3–3.2)

Pulse return density (range) – Gustoća povratka pulsa (raspon)

Horizontal accuracy – Horizontalna točnost

0.15 m

Vertical accuracy – Vertikalna točnost

0.15 m

Pulse rate frequency – Frekvencija pulsa

70 kHz

Table 3 Description of time elements Tablica 3. Opis radnih sastavnica Time elements – Radne sastavnice Moving time: Begins when the feller-buncher or the boom starts to move to a tree and ends when machine head is clamped on the tree Premještanje: Započinje kada feler bančer ili dizalica započinje s pomicanjem i završava kada sječna glava zahvati stablo Felling time: Starts when the feller-buncher head clamps on to the tree stem and ends when the tree touches the ground Sječa: Započinje kada sječna glava zahvati stablo te završava kada posječeno stablo dodirne tlo Stacking time: Starts when the feller-buncher grabs a log and ends when it drops the log onto the pile Uhrpavanje: Započinje kada feler bančer zahvati deblo i završava u trenutku kada ga ispusti na složaj Cycle time: Starts when the feller-buncher commences moving to a tree and ends when the feller-buncher completes felling the tree Vrijeme turnusa: Započinje premještanjem feler bančera ka stablu i završava kad feler bančer posječe stablo Delay: Any interruption to the harvesting operation spending extra time. The cause of the delay (e.g. operational, personal, mechanical, or study induced) is recorded Prekidi: Svako prekidanje pridobivanja drva koje izaziva dodatni utrošak vremena. Uzroci su prekida (npr. povremeni radovi, osobni, kvarovi ili izazvani istraživanjem) zabilježeni

operation with a self-levelling tracked feller-buncher, Valmet 475EXL fitted with a Quadco hotsaw accumulating head. It was manufactured in 2004 and had worked for 7751 hours. The Valmet 475EXL is designed to operate on uneven ground and on steep slopes. The harvesting operation was recorded using a digital video camera in mainly fine and sunny conditions on the 4th April 2011. Several brief episodes of drizzly rain occurred, but did not disrupt the harvesting operation and filming. The operator was observed to fell trees in a 4 row swath directly uphill (moving at right angles to the Croat. j. for. eng. 34(2013)2

contours) or on side-hill (moving parallel to the contours) on gentle & moderate terrain and downhill on steep terrain. In fact, terrain conditions largely dictated tree harvesting pattern in steep and moderate steep slope areas. Trees were laid out in the previously harvested area at right angles to the direction of harvester movement for subsequent processing into logs. The length of each swath was approximately 100 m. One to three trees were felled at each stop. To avoid issues associated with GPS (Global Positioning System) accuracy and performance under tree cover, following the harvesting operation GPS locations of 100 stumps along the border of the study site

275

M. Alam et al.

Self-Levelling Feller-Buncher Productivity Based on Lidar-Derived Slope (273–281)

were recorded with a standard GPS measuring device. Coordinates of stump locations were used to locate the harvest area on the LiDAR-derived DTM. Timer Pro Professional software (www.acsco.com) was used to extract each time element from the video recording (Table 3). Time elements unrelated to tree size and slope including stacking, brushing, clearing and any delays were excluded from the analysis because of their random occurrences. In order to develop feller-buncher productivity models for each slope class, the following steps were carried out: Þ Stump locations on the boundary and tree spacing measurements in the plantation area were used to interpolate the locations of remaining trees using ArcGIS 10. Þ One hundred and twenty-six trees in the moderate slope area and 124 trees in the steep slope area of the study site were selected for model development using previously derived LiDAR slope. Trees with normal growth and forms

were selected from both slope areas and selection was limited to trees that could clearly be identified as being in the allocated slope class. There was insufficient area of gentle slope at the study site for productivity modelling. Þ For estimated location of each tree in the study area, both tree volume and slope class were allocated. Þ Time consumptions of each tree for slope classes were estimated from the time and motion study. Þ A correlation between tree volumes and time consumptions for each slope class was established, which in turn was used to formulate a productivity model. Þ Prior to model development, mean tree sizes for each slope class were compared using a t-test (p<0.05).

2.4 Data analysis – Obrada podataka Productivity models for the feller-buncher were developed based on the cycle times for the trees se-

Fig. 1 LiDAR-derived slope class (Field tree distribution and Field measured tree rows refer to approximate locations) Slika 1. Razredi nagiba terena izvedeni iz LiDAR-ovih snimaka (terenske izmjere redova i pojedinih stabala odnose se na približne položaje)

276

Croat. j. for. eng. 34(2013)2

Self-Levelling Feller-Buncher Productivity Based on Lidar-Derived Slope (273–281)

lected in each slope area to determine whether slope significantly affected feller-buncher productivity.

M. Alam et al.

chosen and verified to be similar to those of model development areas. The productivity of the fellerbuncher for each of these trees was calculated using cycle time and tree volume and was estimated using the productivity models developed for each slope class. For each slope class, the calculated and estimated productivity values were compared using a paired t-test. Linear regression [Y = a + b(X)] analysis was performed to predict productivity of the feller-buncher for each slope class, where X is the independent variable, field measured productivity; Y is the dependent variable, predicted productivity and a & b are the regression coefficients. Statistical software Excel 2007 was used to perform analyses.

Productivity (m3 PMH0–1) was estimated using the following formula: Productivity = (volume / cycle time) * 60 Where, Volume – tree volume (m3) estimated from field measurements Cycle time (min) – refer to Table 3. PMH0 – Productive Machine Hours excluding delay time Regression models were developed for each slope class and tested to determine the best-fit models using their mean Bias, Mean Absolute Deviation (MAD), RMSE (Root Mean Square Error), R2 and the distribution of the residuals. The best-fit models for each slope class were compared using an F-test (p < 0.05) (Motulsky and Christopoulos 2003).

3. Results – Rezultati Field measured stump locations (coordinates) were used to locate the harvest area on the LiDAR-derived DTM. The LiDAR-derived slope range for the study site was 7–27° with a mean slope of 19°, which was found to be comparable to field measurements of the study site. The slope of the study site was classified into three classes: gentle slope (7–11°), moderate slope (11–18°) and steep slope (18–27°) (Forestry Commission UK 1996) (Fig. 1). Several small areas (maximum 8 m × 6 m) of over 27° slope were added to the steep slope class as they were too small to affect the productivity of the feller-buncher.

The relationship between tree volume and moving & felling times for each slope class was tested using linear regression to determine whether tree volume was a potential covariate in a one-way Analysis of Covariance (ANCOVA). If it was found not to be, a one-way Analysis of Variance (ANOVA) would be performed. A general linear model was used to analyse the ANCOVA and/or ANOVA models (Minitab 16, Minitab Ltd.). Mean felling and moving times for each slope class were compared by a post hoc analysis of the means using Tukey’s test in order to identify the impact of each time element on feller-buncher productivity.

Mean tree volumes in the moderate and steep slope areas were not significantly different (Table 4). As there were very few instances of stacking time during the study, it was excluded from cycle time. Two trees were cut and accumulated in head in two occasions and they were also excluded from the analysis while developing the model to be consistent with overall tree selection technique. In the steep slope areas, a number of trees had fallen on other trees during harvest operation and the operator was observed to

In order to test whether the productivity models developed in the study were able to accurately predict the productivity of the feller-buncher when felling trees elsewhere on the study site, thirty-five trees not used in the model development process were randomly selected from each slope class. Topographic variability and slope ranges of model testing areas were Table 4 Summary tree volume (m3) statistics for each slope class Tablica 4. Statistički prikaz obujma stabala za svaki razred nagiba terena

Mean volume – Srednji obujam, m SD – Standardna devijacija, m3

Volume range – Raspon obujma, m Count – Veličina uzorka

Croat. j. for. eng. 34(2013)2

Moderate slope (11–18°)

Steep slope (18–27°)

Umjereni nagib (11–18°)

Strmi nagib (18–27°)

0.55

0.51

0.26

0.05–1.12

0.13–1.20

126

124

277

M. Alam et al.

Self-Levelling Feller-Buncher Productivity Based on Lidar-Derived Slope (273–281)

Table 5 Model coefficients and goodness of fit statistics for the feller-buncher productivity model for each slope class Tablica 5. Koeficijenti modela i dobrota statističke prikladnosti modela proizvodnosti feler bančera za svaki razred nagiba Model coefficients – Koef. modela

Goodness of fit statistics – Dobrota statističke prikladnosti

MBE

MAD

RMSE

Moderate slope – Umjereni nagib (11–18°)

12.3

3.8

2.5

24.0

34.7

0.60

Steep slope – Strmi nagib (18–27°)

10.8

3.5

2.0

17.7

25.8

0.61

drag them out in the processing areas. The model form that best fitted the data was a natural logarithm transformation of processed volume and the square root of productivity: (Productivity)1/2 = β0 + β1 * ln (Processed volume) As the models were developed based on the square root of the dependent variable, the goodness of fit measures for the productivity models were calculated from back-transformed model outputs (Scott and Wild 1991). Model coefficients and fit statistics are shown in Table 5. Productivity of the feller-buncher was strongly correlated with tree volume in both slope areas (Fig. 2). Productivity was greater (97 m3 PMH0-1) for the moderate slope class (11–18°) than that (73 m3 PMH0-1) for the steep slope class (18–27°) at the pooled mean tree volume of 0.53 m3. The difference between the productivity models was statistically significant (p<0.05).

Fig. 2 Productivity of the feller-buncher against tree volume for moderate slope (11–18°) and steep slope (18–27°) Slika 2. Ovisnost proizvodnosti feler bančera o obujmu stabla za umjereni (11–18°) i strmi (18–27°) nagib

Feller-buncher moving and felling times were separately found to be poorly related to tree volume and thus a one-way ANOVA was performed. The data for the felling and moving times were found to satisfy the ANOVA assumptions. Mean felling times for each slope class were significantly different, whereas there was no significant difference between mean moving times for each slope class (Table 6).

The mean feller-buncher productivity, predicted using the slope class productivity models for trees not used in the model development, was not significantly

Table 6 Feller-buncher mean felling and moving times, standard deviations (SD), min. and max. values for slope classes at the study site Tablica 6. Deskriptivna statistika vremena sječe i premještanja za razrede nagiba istraživane sječine

Najveća vrijednost

Maximum

Najmanja vrijednost

Minimum

Standardna devijacija

Standard deviation

Aritmetička sredina

Mean

Najveća vrijednost

Moving time, sec – Vrijeme premještanja, s

Maximum

Najmanja vrijednost

Minimum

Standardna devijacija

Standard deviation

Aritmetička sredina

Mean

Felling time, sec – Vrijeme sječe, s

Moderate slope – Umjereni nagib (11–18°)

9.7

5.4

4.9

56.4

11.6

7.7

3.5

44.6

Steep slope – Strmi nagib (18–27°)

14.3

8.4

5.9

56.8

12.4

6.6

4.7

45.8

278

Croat. j. for. eng. 34(2013)2

Self-Levelling Feller-Buncher Productivity Based on Lidar-Derived Slope (273–281)

Fig. 3 Predicted productivity as a function of measured productivity for moderate slope (11–18°) and steep slope (18–27°) of the model testing areas Slika 3. Predviđena proizvodnost kao funkcija izmjerene proizvodnosti za umjereni (11–18°) i strmi (18–27°) nagib istraživane sječine different from the mean feller-buncher productivity calculated from cycle times and tree volumes for the same trees (p<0.05). Measured productivity was found to be strongly correlated with predicted productivity for each slope (Fig. 3).

4. Discussion and Conclusion – Rasprava sa zaključcima The productivity of the feller-buncher in the current study was found to decrease in the steeper slope class, which was consistent with the findings of previous studies (e.g. FPInnovations 2008, Oliveira Júnior E. D. de. et al. 2009). However, there is considerable variation amongst the previous studies in the degree of decrease in productivity with increasing slope, which implies factors other than slope are influencing the results. In the current study, the decrease in productivity between steep slope (18–27°) and moderate slope (11–18°) was 24% whereas Acuna and Kellogg (2009) found no sigCroat. j. for. eng. 34(2013)2

M. Alam et al.

nificant difference in the productivity of a feller-buncher across a slope range from <10º to 20º and FPInnovations (2008) showed a 30% reduction in productivity between 6–11º slopes and 11–18º slopes based on modelled results from a number of feller-buncher studies. The greatest decrease in feller-buncher productivity was reported by Oliveira Júnior E. D. de. et al. (2009), who found an 80% decrease in productivity for a tracked feller-buncher between 0 and 27º slopes. This large productivity drop was explained by the difference in soil type (e.g. »agri-loose« soil) and difficulties in handling larger trees on steep terrain. Other potential factors accounting for the variation between the study results may include machine characteristics, operator skill and the number of stems removed per hectare, because travelling time between trees may increase disproportionately with increasing slope. These factors, however, were not investigated in this study. To isolate the cause of the productivity differences between the slope classes in the current study, the cycle time components (moving and felling times), were further analysed. The study found felling time to be the main driver for the variation in productivity, which was the result of the operator spending significantly more time per tree (over 4 seconds) felling trees in the steep slope area (Table 6). Terrain conditions largely dictated the harvesting pattern and observations indicated it had a greater impact on steeper slopes within the steep slope classification. Since operating the machine on an uphill slope is slightly more comfortable and productive (Howe 2011), the operator was observed to fell trees uphill by predominantly extending the boom and moving the feller-buncher in the moderate slope areas and lay them out for processing primarily using the boom, whereas, in the steep slope areas the operator drove downhill to fell each tree and then back uphill to deposit them in suitable areas, preferably those with moderate slope, for processing. In the steep slope areas, the operator also spent time dragging out a number of trees that had fallen on other trees. The combination of these factors contributed to higher time consumption when felling trees in the steep slope areas. The study did not investigate whether soil strength was an influential factor affecting the machine’s stability and traction in the steep slope areas, although the soil in the steep slope areas was observed to be muddy compared with that in the gentle and moderate slope areas. Stampfer and Steinmüller (2001) demonstrated that the locomotion of the harvester was dependent on the terrain slope and the soil bearing capacity. Therefore, consideration of the soil bearing capacity, while evaluating slope effects on harvester productivity, may be an area for future research.

279

M. Alam et al.

Self-Levelling Feller-Buncher Productivity Based on Lidar-Derived Slope (273–281)

The models developed in the study were able to predict the productivity of the feller-buncher felling trees from each slope class on the study site where the topographical features and slope ranges were similar to model development areas. This suggests that the models can be used to predict the productivity of the feller-buncher operating in other areas of radiata pine plantation with tree volume between 0.06–1.84 m3 and slope between 11–27°. The models may not be applicable where topographic variability is significantly different (higher or lower) from the model development areas, because the assignment of trees for each slope class based on the methodology used in the study may not represent the exact locations, which excluded the use of trees on the boundary of the slope classes and where slope was very variable and different slope variability will potentially influence the operator approach. In addition to being able to estimate terrain slope, LiDAR has been demonstrated by a number of researchers to be able to accurately predict tree volume (e.g. Hyyppa et al. 2001, Persson et al. 2002). However, the LiDAR point density in these studies was considerably greater than that for the current study, which targeted using readily available LiDAR data in the interest of exploring a methodology that would be cost-effective for practical application.

Acknowledgement – Zahvala We acknowledge the support of Mrs. Sandra Hetherington and her team of Norske Skog, Tasmania, Australia and Rick Mitchell (Western Australia Plantation Resources) for organising harvest operation required for data acquisition. We also acknowledge the support of Mr. David Mannes (Forestry Tasmania) for providing LiDAR data.

5. References – Literatura Ackermann, F., 1999: Airborne laser scanning-present status and future expectations. ISPRS Journal of Photogrammetry & Remote Sensing 54(2-3): 64–67. Acuna, M., Kellogg, L., 2009: An evaluation of alternative cutto-length harvesting technology for native forest thinning in Australia. International Journal of Forest Engineering 20(2): 17–25. Acuna, M., Skinnell, J., Evanson, T. Mitchell, R., 2011: Bunching with a self-levelling feller-buncher on steep terrain for efficient yarder extraction. Croatian Journal of Forest Engineering 32(2): 521–531. Brunberg, T., Thelin, A., Westerling, S., 1989: Basic data for productivity standards for single-grip harvesters in thinning operations. Report No 3, The Forest Operations Institute of Sweden, p. 21

280

Davis, C. J., Reisinger, T. W., 1990: Evaluating Terrain for harvesting Equipment Selection. Journal of forest Engineering 2(1): 9–16. Forest Practices Board, 2000: Forest Practices Code, Forest Practices Board, Hobart, Tasmania. Australia 7100. Forestry Commission UK, 1996: Terrain Classification. Available: http://www.biomassenergycentre.org.uk [Accessed 22 January 2013]. FPInnovations, 2008: Feller-buncher studies. Progress Report #12, Saint-Jean Pointe-Claire, QC, H9R 3J9, Canada. Giles, P. T., Franklin, S. E., 1998: An automated approach to the classification of the slope units using digital data. Geomorphology 21(3–4): 251–264. Howe, D., 2011: Cut to length on difficult terrain. Available: http://www.forestrysolutions.net/userfiles/File/CTL%20harvesting%20FOCUS%20PRESENTATION%20Dereke.pdf [Accessed 18 January 2013]. Hyyppa, J., Kelle, O., Lehikoinen, M., Inkinen, M., 2001: A segmentation-based method to retrieve stem volume estimates from 3-dimensional tree height models produced by laser scanner. IEEE Transactions on Geoscience and Remote Sensing 39(5): 969–975. Kellogg, L. D., Bettinger, P., 1994: Thinning productivity and cost for mechanized cut-to-length system in the Northwest pacific coast region of the USA. International Journal of Forest Engineering 5(2): 43–54. Lageson, H., 1997: Effects of thinning type on the harvester productivity and on the residual stand. Journal of Forest Engineering 8(2): 7–14. MacDonald, A. J., 1999: Harvesting systems and equipment in British Columbia, FERIC Handbook, ISSN 0707-8355, No. HB-12, p. 197. Motulsky, H. J., Christopoulos, A., 2003: Fitting models to biological data using linear and nonlinear regression: A practical guide to curve fitting. GraphPad Software Inc. San Diego, CA. Nurminen, T., Korpunen, H., Uusitalo, J., 2006: Time consumption analysis of the mechanized cut-to-length harvesting system. Silva Fennica 40(2): 335–363. Oliveira Júnior, E. D. de., Seixas, F., Batista, J. L. F., 2009: Fellerbuncher productivity in eucalyptus plantation on steep ground terrain. Floresta 39(4): 905–912. Persson, A., Holmgren, J., Soderman, U., 2002: Detecting and measuring individual trees using an airborne laser scanner. Photogrammetric Engineering and Remote Sensing 68(9): 925–932. Reutebuch, S., Andersen, H., McGaughey, R., 2005: Light Detection and Ranging (LIDAR): An emerging tool for multiple resource inventory. Journal of Forestry 103(6): 286–292. Scott, A., Wild, C., 1991: Transformations and R2. The American Statistician 45(2): 127–129. Simões, D., Fenner, P., 2010: Influence of relief in productivity and costs of harvester. Scientia Forestalis 38(85): 107–114. Stampfer, K., 1999: Influence of terrain conditions and thinning regimes on productivity of a track-based steep slope harvester. In: Proceedings of the International Mountain LogCroat. j. for. eng. 34(2013)2

Self-Levelling Feller-Buncher Productivity Based on Lidar-Derived Slope (273–281) ging and 10th Pacific Northwest Skyline Symposium, Corvallis, Oregon: 78–87. Stampfer, K., Steinmüller, T., 2001: A new approach to derive a productivity model for the harvester Valmet 911 Snake. In: Proceedings of the International Mountain Logging and 11th Pacific Northwest Skyline Symposium, Seattle, WA, p. 254– 262. Vaze, J., Teng, J., 2007: High Resolution LiDAR DEM – How good is it? In: MODSIM 2007 International Congress on Modelling and Simulation.

M. Alam et al.

Visser, R., Spinelli, R., Saathof, J., Fairbrother, S., 2009: Finding the »Sweet-Spot« of mechanised felling machines. In: Proceedings of USA: 32nd Annual Meeting of the Council on Forest Engineering (COFE), Kings Beach, CA, p. 10. Wehr, A., Lohr, U., 1999: Airborne laser scanning – an introduction and overview. ISPRS Journal of Photogrammetry and Remote Sensing 54(2–3): 68–82. Zimbalatti, G., Proto, A. R., 2010: Productivity of forwarders in South Italy. In: FORMEC 2010, Forest Engineering: Meeting the Needs of the Society and the Environment, Padova–Italy.

Sažetak

Proizvodnost feler bančera sa žiroskopskom kabinom temeljena na nagibu terena izvedenom iz LiDAR-ovih snimaka Cilj je istraživanja bio ocijeniti mogućnost uporabe LiDAR-ovih snimaka za određivanje nagiba terena na velikim površinama te ispitati djelovanje nagiba na proizvodnost feler bančera sa žiroskopskom kabinom. Istraživanje je provedeno u čistoj sječi plantaže smolastoga bora (Pinus radiata) u Tasmaniji, u blizini Port Arthura (Australija). Plantaža je bila u dobi od 24 godine. Korišten je feler bančer sa žiroskopskom kabinom, opremljen gusjenicama. Visina i prsni promjer stabala mjereni su prije sječe. Upotrijebljene su LiDAR-ove snimke niskoga intenziteta (>3 točke po m2) iz 2011. godine kako bi se odredio nagib terena. Pri sječi i izradbi obavljen je i studij rada i vremena radi određivanja proizvodnosti vozila, a u ovisnosti o obujmu posječenih stabala i nagibu terena. Rezultati istraživanja dokazuju primjenjivost LiDAR-ovih snimaka za raščlambu nagiba terena. Može se zaključiti da je za stablo prosječna drvnoga obujma od 0,53 m3 proizvodnost feler bančera sa žiroskopskom kabinom bila 97 m3/h (ne uključujući prekide rada) na terenu umjerena nagiba (11–18°) odnosno 73 m3/h (ne uključujući prekide rada) na strmijim terenima (18–27°). Razlika u proizvodnosti vozila zasniva se na različitim postupcima pri sječi i izradi koje je radnik morao obavljati ovisno o nagibu terena. Modeli proizvodnosti temelje se na stvarno posječenom i izrađenom drvnom obujmu. Ključne riječi: Tasmanija, proizvodnost, feler bančer sa žiroskopskom kabinom, LiDAR, strojna sječa, nagib

Authors’ address – Adresa autorâ:

Received (Primljeno): February 8, 2013 Accepted (Prihvaćeno): June 3, 2013 Croat. j. for. eng. 34(2013)2

Muhammad Alam*, PhD. Student e-mail: mmalam@student.unimelb.edu.au University of Melbourne 500 Yarra Boulevard Richmond, Australia 3121 Mauricio Acuna, PhD. e-mail: macuna@usc.edu.au Australia Forest Operations Research Alliance (AFORA) University of the Sunshine Coast Hobart, Tasmania, 7001, Australia Prof. Mark Brown, PhD. e-mail: mbrown2@usc.edu.au Australia Forest Operations Research Alliance (AFORA) University of the Sunshine Coast Maroochydore DC, Queensland, 4558 Australia *Corresponding author – Glavni autor

281

Original scientific paper â&#x20AC;&#x201C; Izvorni znanstveni rad

Performance, Capability and Costs of Motor-Manual Tree Felling in Hyrcanian Hardwood Forest Meghdad Jourgholami, Baris Majnounian, Nosratollah Zargham Abstract â&#x20AC;&#x201C; Nacrtak Motor-manual tree felling is the most labor-intensive component of all harvesting operations and frequently represents a bottleneck in wood production. The study of motor-manual tree felling was carried out in two compartments in the Namkhaneh district of Kheyrud Forest. The objects of this study were as follows: time study of tree felling operations, estimate of chainsaw productivity and costs, development of a regression model in uneven-aged stand using single-tree selection methods. The factors affecting total felling time regression model (increasing order of importance) were DBH of harvested trees, direction of felling regarding the lay and inter-tree distance. The hourly production of chainsaw felling with and without delay time was 56.4 cubic meters per hour (13 tree/hour) and 80.7 cubic meters per hour (19 tree/hour), respectively. Productivity of chainsaw felling increased in relation to tree DBH as power relation. The cost of chainsaw felling with and without delay time was 0.55 and 0.39 USD/m3, respectively. The cost of felling increased as simple exponential equation when DBH of harvested trees decreased. However, the unit felling cost for chainsaw operation decreased as the tree size increased. Total felling cycle time without delay averaged 3.14 minutes and with delay time it averaged 4.5 minutes. Productivity was more sensitive to DBH than felling direction and inter-tree distance. Keywords: tree felling, time study, regression model, production, cost

1. Introduction â&#x20AC;&#x201C; Uvod Hyrcanian forest in northern Iran is an example of biodiversity, with endemic and endangered species, and a diverse range of economic and social conditions. About 45% of the Hyrcanian forests are located in mountainous areas, where forest lands are not readily accessible with ground-based logging equipments. Felling, limbing and bucking are all done at the stump site. Motor-manual systems are used by workers equipped with chainsaws (Sobhani and Staurt 1991). Chainsaw felling is often associated with large trees and steep or rough terrain. It is used for areas where ground-based machines cannot travel or where the trees are too large for mechanical felling. Due to larger diameter and crowns of hardwoods, and the relatively steep terrain in the Hyrcanian forest, motormanual tree felling is still the only system used in the Croat. j. for. eng. 34(2013)2

region (Sarikhani 2008). The capital investment required for motor-manual felling is several hundred times less than for mechanical felling, and the felling costs per cubic meter are usually lower as well. Despite these differences, other factors such as terrain and timber conditions and total system productivity dominate the choice between the two systems for large contractors (MacDonald 1999, Sessions et al. 2007). Harvesting starts with the cutting down of trees with hand tools, chain saws, or mechanized felling machines. Felling is the most dangerous part of the harvesting operation (Conway 1976, ILO 1998, Heinimann 2004, Sessions et al. 2007). Larger trees generally must be felled manually with a chainsaw. In Hyrcanian forest regions, trees were large and heavy with huge crowns. They fall with a tremendous force, which can uproot the neighboring trees; and stems

283

M. Jourgholami et al.

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

may shatter, bounce, and roll uncontrollably. Therefore, motor-manual felling operations are the most hazardous part of harvesting operations for the labor forces in this forest. They are also a major cause of damage to the forest stand and result in the generation of a large amount of wood waste. The objective of the tree felling operation is to fell the tree with minimum damage, to avoid damaging surrounding trees, to minimize soil and water impacts, and to position the tree or logs for the next phase of harvesting. Directional felling is a specific tree-felling technique, in which the direction of fall is determined by the operator prior to cutting. Where possible, trees should be felled in the direction of existing canopy gaps in order to reduce damage to nearby standing timber. In general, trees should be felled either towards or away from skid trails, preferably at an oblique angle to the skidding direction (FAO 1976, Dykstra and Heinrich 1996, MacDonald 1999). Hartsough et al. (2001) developed the felling time prediction model based on the tabular data of felling time per tree collected on clear cutting of secondgrowth timber. Kluender and Stokes (1996) developed a nonlinear model to predict felling time for different harvesting prescriptions, using variables as distance from previous tree, proportion of basal area removed and DBH. Lortz et al. (1997) conducted an analysis of southern pine felling with chainsaw and produced several equations for estimating felling time and productivity. They found that factors affecting total felling time were DBH of harvested stems, inter-distance, and harvest intensity. Wang et al. (2004) conducted a time study on central Appalachian hardwood forest consisting of motor-manual felling and cable skidding. They reported that felling time was mainly affected by diameter at breast height and distance between harvested trees. This study showed that productivity of chainsaw felling was 362 ft3 per productive machine hour (PMH) with a unit cost of $8.0/100 cubic feet. Rummer and Klepac (2002) conducted a time study to compare two harvesting systems; mechanized and motor-manual felling operations. This study showed that the harvester was about as productive as a manual crew of five. Also, they reported that there is a strong trend of increasing cycle time as tree size increases and a regression equation was developed to predict total cycle time as a function of tree diameter. Li et al. (2006) conducted a simulation study for comparing production and cost of felling among chainsaw, harvester, and feller-buncher. They found that the unit felling cost for chainsaw operation decreased as the tree size increased. Few previous studies have addressed the production and cost of motor-manual tree felling in Hyrca-

284

nian hardwood stands. Nikooy (2007) developed a productivity model for chainsaw felling in Caspian hardwood forests, which included variables such as diameter at breast height and the distance among harvested trees. This study reported that productivity of tree felling with and without delay time was 53 and 67 cubic meters per productive machine hour (PMH), respectively. Behjou et al. (2009) conducted a time study on Hyrcanian forests. They found that felling time per tree was most affected by diameter at breast height and by the distance among harvested trees. The gross and net production rate was 20.6 m3 and 26.1 m3 per hour/one person, respectively. The objective of this study was to: conduct a continuous time study on motor-manual tree felling with a chainsaw in a Hyrcanian hardwood forest, employing regression techniques to develop models for elemental times and cycle time of chainsaw felling, and estimate the production rates and costs of chainsaw felling.

2. Study sites and methods – Mjesto i metode istraživanja The research was carried out in two compartments 219 and 223 located in Namkhaneh District within Kheyrud Educational and Research Forest. The altitude ranges from 1 000 to 1 135 m and the forest lies southwest. The slope ranges from 10 to 70% with an average of 40%. The average rainfall ranged from 1 420 to 1 530 mm/year, with the heaviest precipitation in the summer and fall. The average daily temperatures ranged from a few degrees below 0°C in December, January, and February to +25°C during the summer. This area is dominated by natural forests containing native mixed deciduous tree species such as Fagus orientalis Lipsky, Carpinus betulus L., Acer velutinum Boiss., and Alnus subcordata (Jourgholami 2013). The management method is mixed un-even aged high forest with single and group selective cutting regime. Trees to be removed are felled, limbed and topped motor-manually. Felled trees are bucked and processed with chainsaws into logs, sawn-lumber and pulpwood. The logs 5 to 15 meter long are extracted by wheeled cable skidders to the roadside landings. The fuel wood is extracted by mules. Also, in steep terrain that cannot be reached by skidders, logs are processed to sawn-lumber and then hauled by mules (Jourgholami 2012). Table 1 summarizes some characteristics of the study site. Felling was performed using a STIHL chainsaw with 4-hoursepower (hp) engine and bar length of 70 cm (Fig. 1). The field study was conducted from January to February 2011 on Kheyrud Forest during Croat. j. for. eng. 34(2013)2

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

M. Jourgholami et al.

Table 1 Study site description Tablica 1. Mjesto istraživanja Compartment

Area

Odjel

Površina ha

Trees per ha Broj stabala po ha

Volume

Total felled trees

Total volume of felled trees

DBH of felled trees

Obujam

Ukupan broj posječenih stabala

Etat (sječna gustoća)

Prsni promjer

m3/ha

num. (t/ha) – broj (stabala/ha)

m3 (m3/ha)

219

173

504

270 (10 t/ha)

872.3 (32 m /ha)

20–135

223

123

301

181 (3 t/ha)

719.5 (13 m3/ha)

20–135

Table 2 Main work phases that make up total felling time Tablica 2. Faze radova sječe Work elemental function

Definition – Opis radnih zahvata

Radni zahvati Walk to tree – Prilazak stablu Acquire Čišćenje okoliša oko stabla i određivanje smjera rušenja

Undercut – Izrada zasjeka

Backcut – Potpiljivanje

Wedging – Zabijanje klinova

Refuel and Service Punjenje goriva, maziva i popravak Delays – Zastoji

Begins when the sawyer starts toward the tree to be cut and ends when the sawyer reaches to the tree Prilazak stablu počinje kada šumski radnik sjekač krene prema doznačenomu stablu i završava kada dođe do njega Begins when the sawyer starts clearing around the tree and decides where the tree will fall and ends when the sawyer is ready to cut the tree Priprema počinje kada šumski radnik sjekač počne čistiti okoliš oko stabla, zatim odlučuje o smjeru rušenja stabla te završava kada je radnik spreman za sječu stabla Begins when the sawyer starts to make a wedge-shaped notch in the base of the tree to ensure that it accurately faces the felling direction and ends when the sawyer starts backcut Radni zahvat započinje izradom kosoga reza zasjeka, u odabranom smjeru rušenja stabla, a završava kada je sjekač spreman za potpiljivanje Begins when the sawyer starts cutting the opposite side of the direction of fall and ends when the tree hits the ground Radni zahvat počinje prilikom prerezivanja stabla sa suprotne strane od zasjeka i završava kad se stablo sruši na zemlju Begins when the co-sawyer starts to enter the wedge-shaped blade to the cutting gap and ends when the tree falls in the predetermined direction Radni zahvat započinje kada pomoćni radnik sjekač počne postavljati klinove u potpiljak te završava kad se stablo sruši na zemlju Maintenance and refueling – Održavanje motorne pile i punjenje goriva i maziva Personal delay, Technical delay, and Operational delay – Osobni prekidi rada, tehnički prekidi rada i povremeni radovi

winter; cold occasionally affected worker utilization percentages. The power-saw team normally consists of three men: a feller, an assistant, and a helper. Time and operational variables were measured using a stopwatch and recorded on paper (Bjorheden and Thompson 1995, Wang et al. 2004). A work cycle for each operation consisted of certain elemental functions and factors. The time for each function and value of each factor were measured in the field. Elemental time functions for chainsaw felling are shown in Table 2. Harvesting factors or operational variables for chainsaw felling measured in the field include disCroat. j. for. eng. 34(2013)2

tance to tree (cm), tree species, diameter at breast height (DBH) (cm), walk to tree slope (%), slope at tree stump (%), and direction of felling: code 1: felling to lean; code 2: felling sideways to the lean (0° to 90°); code 3: felling opposite the lean (90° to 180°)). A total of 233 cycles for chainsaw felling were observed in the field. Local volume equations of Namkhaneh district were used to compute the volume of felled trees. The SPSS 14.0 statistical program was applied to develop regression equation of time consumption. A regression analysis with the stepwise method between operational variables (independent variables) was per-

285

M. Jourgholami et al.

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

Fig. 2 Effect of DBH on felling time without delay of tree felling Slika 2. Utjecaj prsnoga promjera na vrijeme sječe stabla, bez radnih zastoja formed on the time study data collected for chainsaw, to determine independent variables that were significant in estimating total felling time (p = 0.01). Regression techniques were also employed to develop models for elemental times, felling cycle time and productivity of chainsaw felling. Total felling time was analyzed in stages (Lortz et al. 1997). First, each work elemental function (phase) was fit to a linear equation (Y = a + bX) using DBH as independent variable. Then, other operational variables were added to the model to show how these factors influence the felling time, and to give a more reliable model of motor-manual felling operation.

3. Results and discussion – Rezultati i rasprava

Fig. 1 Starting to undercut (A and B), cutting the backcut and wedging (C) in the study area Slika 1. Šumski radnik sjekač započinje izradu zasjeka (A i B), potpiljivanje i zabijanje klinova (C)

286

DBH of felled trees ranged from 20 to 135 centimeters and averaged 52.3 centimeters, while the volume per felled tree was between 0.2 to 29.7 cubic meters with an average of 4.27 cubic meters (Table 3). Distance between harvested trees varied from 2 to 105 meters with an average of 25.4 meters. A felling cycle consists of the following elements: walk to tree, acquire, undercut, backcut, wedging, refuel and service, and delay times. Total felling time varied from 0.6 to 29.65 minutes with an average of 4.5 minutes, while total felling time without delay ranged between 0.6 to 10.1 minutes with an average of 3.14 minutes per cycle (Table 3). Croat. j. for. eng. 34(2013)2

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

M. Jourgholami et al.

min

Tree diameter

Prsni promjer stabla

Volume

Obujam

Walk

Hod do stabla

Acquire

Čišćenje okoliša i određivanje smjera rušenja

Izrada zasjeka

Undercut

Izrada završnoga reza

Backcut

Wedging

Zabijanje klinova

Refuel & service

Točenje goriva i popravak

Delay free time

Factor

Vrijeme bez zastoja

Table 3 Statistics of operational variables of motor-manual felling in the field study Tablica 3. Statistika operativnih varijabli u provedenom istraživanju ručno-strojne sječe

0.09

0.22

Maximum

10.1

3.5

4.17

3.49

5.72

2.15

2.82

29.7

135

Std. dev.

1.95

0.56

0.49

0.62

1.01

0.21

0.43

5.23

24.59

Nagib kod panja

Factor

code*

šifra* Mean – Srednja vrijednost

Pers. Delay

0.05

Osobni prekidi rada

0.06

Tech. Delay

0.13

Tehnički prekidi rada

Oper. Delay

Operativni prekidi rada

0.6

Ukupni prekidi rada

Minimum

Total delay

52.34

Ukupno vrijeme sječe

4.22

Total felling time

0.66

Inter-tree dis.

0.23

Udaljenos između doznačenih stabala

1.25

Nagib tijekom hoda do stabla

0.74

Walk to tree slope

0.1

Slope at stump

0.15

Smjer obaranja

3.14

Direction of felling

Mean – Srednja vrijednost

min

1.31

26.07

28.5

25.43

4.5

1.36

0.17

1.18

Minimum

0.6

Maximum

105

29.65

24.36

20.21

24.36

Std. dev.

0.55

15.29

14.36

16.6

4.99

4.16

1.48

3.9

*code (felling direction as described in text) – šifra (odabrani smjer rušenja stabla kako je opisano u tekstu)

Time of the walk to the tree averaged 0.66 minutes and ranged between 0.09 to 2.82 minutes. Since the walk to the tree is directly related to stand density and harvesting method (Single-selection method), it was significantly different depending on the distance between felled trees. Acquire time averaged 0.23 minutes and ranged between 0.05 and 2.15 minutes per cycle. Time of undercut varied from 0.06 to 5.72 minutes with an average of 1.25 minute per cycle s. Backcut time ranged from 0.13 to 3.49 minutes and averaged 0.74 minutes per cycle. Some trees needed no wedging time. However, a maximum of 4.17 minutes was taken to direct large trees. Refuel and service time averaged 0.15 minutes per cycle. A total of 55 delays was observed during motor-manual felling in the field study. The delay times were ranged from 0 to 24.36 and averaged 1.36 minutes per cycle. Croat. j. for. eng. 34(2013)2

The relation between tree size and total cycle time is shown in the scatter diagram in Fig. 2. There is a strong trend of increasing cycle time as tree size increases. A regression equation was developed to predict total cycle time as a function of tree diameter. Other independent variables were tested and were not significantly related to the total cycle time (Fig. 3–5). The stepwise analysis has revealed that tree diameter (DBH) significantly affects the felling cycle time (Fig. 3). Therefore, we have developed a regression model of the total felling time without delay using tree diameter, direction of felling, and inter-distance of felled trees as an independent variable (Eq. 1). On the other hand, the total felling time was best described by DBH, direction of felling, and distance between felled trees. Statistical significance was checked by an

287

M. Jourgholami et al.

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

Fig. 3 Relation between DBH and total cycle time without delay for felling per cycle Slika 3. Odnos između prsnoga promjera i ukupnoga vremena sječe, bez radnih zastoja

Fig. 5 Effect of DBH on backcut time of tree felling per cycle Slika 5. Utjecaj prsnoga promjera na vrijeme potpiljivanja stabla

Table 4 ANOVA for regression model developed for motor-manual tree felling Tablica 4. ANOVA za regresijski model razvijen za ručno-strojnu sječu Factor

Fig. 4 Effect of DBH on undercut time of tree felling per cycle Slika 4. Utjecaj prsnoga promjera na vrijeme izrade zasjeka

F-test of the overall fit and t-tests for individual parameters (Table 4). T = –1.1997 + 0.05844DBH + 0.63097DF + 0.001778D (1)

288

Sig.

Regression

621.44

2078.15

182.2

<0.0001

Residual

260.3

229

1.14

Total

881.74

232

Where: T DBH DF D R2

= total felling time without delay (min) = diameter at breast height (cm) = direction of felling (1–3 or 0°–180°) = distance between felled trees (m) = 0.705, Adjusted R-square = 0.701; Number of observations = 233

The multiple correlation coefficient of the model shows that 70.5% of the total variability can be explained by the model. The significance level and Fvalue in the table with the analysis of the model variance confirms that the model makes sense at the probability level of 0.05. Croat. j. for. eng. 34(2013)2

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

M. Jourgholami et al.

The production of felling with chainsaw can be obtained by using the production and time data as follow: Production =

TFV TFV

Where: TFV = total felling volume, m3 TFT = total felling time, hour

The hourly production (m3/hr) with delay time was 56.34 m3/hour. The measured production for motormanual felling without delay times was 80.7 m3/hour. Hourly production of felling without delay times was higher than production (m3/hour) with delay times. Also, the hourly production of chainsaw felling with and without delay time was 13 trees per hour and 19 trees hour, respectively. The relation between tree size and felling production is shown in the scatter diagram in Fig. 6. There is a strong trend of increasing production as tree size increases. We calculated the hourly cost ($/hr) of motor-manual felling using the cost estimation model developed by the Forest, Range and Watershed Management Organization of Iran (1999). A purchase price of USD 1 400 was used in the chainsaw cost estimation model, and the annual interest rate of 18.5%. A chainsaw life of 3 years was assumed. Insurance and tax rate and utilization rate were set at 5% and 83%, respectively. The hourly machine cost was estimated at USD 31.26. Table 5 summarizes the estimates of machine costs for chainsaws. As a result, the average felling cost per cubic meter, including the delay time, was USD 0.55/m3, while the average felling cost without delay was estimated at USD 0.39/m3. The cost of chainsaw felling with and without delay time was 2.34 and 1.64 USD per tree,

Fig. 6 Effect of DBH on felling productivity Slika 6. Utjecaj prsnoga promjera na proizvodnost sječe

respectively. Approximately 25% of the total operating hours were identified as delay times during the time study, which results in an average machine utilization rate of 75%. The effect of each variable used in the model on the felling time was studied by changing one variable in its range and retaining the other variables constant at their average. Fig. 7 shows the effect of operational variables on felling costs. Increasing tree diameter and direction of felling will increase felling costs per cycle. The effect of tree size on unit cost of motor-manual felling tree is shown in Figure 8. DBH classes of 20 to 50 centimeters have a dramatic effect on felling costs,

USD/hour USD/satu

0.47

0.2

Croat. j. for. eng. 34(2013)2

0.07

0.73

0.47

1.06

14.53

Total hourly machine rate

Trošak radnoga sata

Hourly labor cost

Ukupni materijalni troškovi

Subtotal (Operating)

Lanac i oštrač lanca

Chain and file

Gorivo i mazivo

Fuel and lubricant

Održavnje i popravak

Maintenance and repair

Ukupni fiksni troškovi

Operating costs – Materijalni troškovi

Subtotal (Fixed)

Porez i osiguranje

Tax and insurance

Kamata

Interest

Amortizacija

Depreciation

Vrsta troška

Cost elements

Fixed costs – Fiksni troškovi

Ukupni trošak radnoga sata stroja

Table 5 Calculation of motor-manual felling costs Tablica 5. Izračun troškova ručno-strojne sječe

31.26

289

M. Jourgholami et al.

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

Fig. 7 Effect of tree diameter (DBH) on felling costs Slika 7. Utjecaj prsnoga promjera na troškove sječe

in Fig. 9. On average, during a cycle most time was spent on undercut, which accounted for 27.9% of the total time. Personal delay (rest and meal time) accounted for approximately 26.3% of the time. Backcut and walk to tree accounted for 16.5% and 14.6% of the total cycle time, respectively. Acquire accounted for less than 5% of the total cycle time. Technical delay accounted for only 3.9% of the total time, while refuel and service and wedging time accounted for about 3.4% and 2.3%, respectively. Undercut, backcut and delay time were the most important time-consuming elements in felling. This suggests that the productivity could be increased by diminishing the time consumption of these elements. Delay time is an inseparable part of each work phase in harvesting in Iran. Delay time accounted for approximately 30% of gross-effective hour. Technical delays, such as sharpening and dealing with the chain of a chainsaw breaking, accounted for approximately 4% of the delay time. One of the reasons for a long delay time was the use of old and obsolete equipment, unsuitable and incorrect filling of the chain saw (Mousavi 2009). Operational delay accounted for the largest share that needs to be considered. Operation delay may relate to management, supervision, and equipment

Fig. 8 Effect of tree diameter on unit cost of motor-manual tree felling Slika 8. Utjecaj prsnoga promjera na jedinične troškove sječe ranging from USD 1.2 to 0.2 per m3. With the classes above 50 cm class, the felling costs for the chainsaw changed constantly. During the study period, felling time was divided into elemental time functions (work phases) as shown

290

Fig. 9 Percentage of time distribution of tree felling elements Slika 9. Postotna distribucija vremena radnih zahvata prilikom sječe stabala Croat. j. for. eng. 34(2013)2

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

M. Jourgholami et al.

availability. A felling group might not have had all the necessary tools needed for work, which caused a prolonged delay as they had to borrow the tools from the neighboring groups. Activities such as the chain breaking and filing as well as pinching in the kerfs can be part of the working time (Sarikhani 2008), however, in this study it is considered as a technical delay. If we take into account these activities as a part of effective working hour, productivity of felling decreases approximately by 3.6%. Walking is the first element of the felling work cycle. Silvicultural treatment is one of the most important factors influencing time consumption of walking. In the single tree selection method, there were more trees in the forest than in the shelter wood and clear cutting method, and hence more time was required (Lortz et al. 1997, Mousavi 2009). In this study, only 14% of the gross-effective hour was related to walking time. In some areas, the skid trails were not marked, so the operator was free to choose the direction. It may increase skidding time and cost. It is recommended to mark skid trails before felling (Nikooy 2007, Mousavi 2009). The higher percentage of backcut is related to the use of a wedge to lead the tree in the specified direction in order to prevent damage to the residual stand and breakage to the tree being felled. The results showed that the stump diameter, direction of felling and distance were the most important variables affecting the felling time. Tree diameter and inter-tree distance influenced the time consumption of felling, productivity, and unit cost of felling. A study by Kluender and Stokes (1996) showed similar results. They found that tree diameter is the most important factor in estimating the felling time, while the distance between trees and harvesting intensity were also important. However, productivity of felling may be influenced by the operator skills, silvicultural method, tree species, stand composition, undergrowth trees and seedlings, weather condition, coldness of weather, age and brands of chainsaws, chain condition, and lean of the tree as well as slopes (Nikooy 2007, Sarikhani 2008, Mousavi 2009). However, the influences of all these factors were not documented in this study but they were mentioned by Conway (1976).

sumption and productivity of felling. Inter-tree distance also influences the time consumption and productivity of felling. The productivity of felling trees with a large diameter is higher than the productivity of felling trees with a small diameter.

4. Conclusion – Zaključak

Jourgholami, M., 2012: Small-scale timber harvesting; mule logging in Hyrcanian Forest. Small-scale Forestry 11(2): 255–262.

Motor-manual tree felling is a highly variable operation. There are many factors influencing the felling productivity. This paper identifies the most significant variables that should be recognized prior to harvesting. It has been proved that the stump diameter of the tree is the most influential factor affecting time conCroat. j. for. eng. 34(2013)2

Acknowledgements – Zahvala This paper is a one of the results of the research project No. 88001084, which was carried out in the period 2010–2012 in the Hyrcanian forest in northern Iran. The authors would like to acknowledge the financial support of the Iranian National Science Foundation (INSF).

6. References – Literatura Behjou, F. K., Majnounian, B., Dvorak, J., Namiranian, M., Saeed, A., Feghhi, J., 2009: Productivity and cost of manual felling with a chainsaw in Caspian forests. Journal of forest science 55(2): 96–100. Bjorheden, R., Thompson, A. M., 1995: An International Nomenclature for Forest Work Study. Paper presented at the XX IUFRO World Congress, Tampere, 6–12 August 1995. Manuscript. 16 p. Conway, S,. 1976: Logging practices. Miller Freeman Publication. USA. 465 p. Dykstra, D. P., Heinrich, R., 1996: FAO model code of forest harvesting practice. FAO. Rome. 97 p. FAO, 1976: Harvesting planted forests in developing countries. A manual on techniques, roads, production and costs. FOI: TF-INT 74 (SWE). FAO, Rome. 76 p. Forest, Range and Watershed management Organization, 1999: Instruction for Preparing Harvesting Plan, 39 p. (in Persian). Hartsough, B., Zhang, X., Fight, R., 2001: Harvesting cost model for small trees in natural stands in the Interior Northwest, Forest Products J 51(4): 54–61. Heinimann, H. R., 2004: Forest operation under mountainous conditions, In Encyclopedia of Forest Sciences, J. Burley, J. Evans, J. Youngquist, Editors. Elsevier Academic Press: Amsterdam, etc, P: 279–285. International Labour Office (ILO), 1998: Safety and health in forestry work. Geneva. Italy. 116 p.

Jourgholami, M., 2013: Harvesting plan of Namkhaneh district. Faculty of Natural Resources, University of Tehran, Iran, 240 p. (in Persian). Kluender, R. A, Stokes, B. J., 1996: Felling and skidding productivity and harvesting cost in southern pine forests. Pro-

291

M. Jourgholami et al.

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

ceedings: Certification–Environmental implications for forestry operations, Quebec, 1996 September 9–11, 35–39. Li, Y., Wang, J., Miller, G., McNeel, J., 2006: Production economics of harvesting small-diameter hardwood stands in central Appalachia. Forest Prod J 56(3): 81–86. Lortz, D., Kluender, R., McCoy, W., Stokes, B., Klepac, J., 1997: Manual felling time and productivity in southern forests. Forest Prod J 47(10): 59–63. MacDonald, A. J., 1999: Harvesting Systems and Equipment in British Columbia. FERIC Handbook No. HB-12. B.C. Ministry of Forests. Forestry Division Services Branch. Production Resources. 595 Pandora Avenue. Victoria, BC V8W 3E7. 211 p. Mousavi, R., 2009: Comparison of productivity, cost and environmental impacts of two harvesting methods in Northern Iran: short-log vs. long-log, Ph.D. thesis, University of Helsinki, Finland. Nikooy, M., 2007: Production optimization and reduction impact on forest by preparing harvest planning in Nav, Iran. Ph.D. thesis, Tehran University, 165 p. (in Persian)

Rummer, R., Klepac, J., 2002: Mechanized or hand operations: which is less expensive for small timber? Published in Small Diameter Timber: Resource Management, Manufacturing, and Markets proceedings from conference held February 25–27, 2002 in Spokane, Washington. Compiled and edited by D.M. Baumgartner, L.R. Johnson, and E.J. DePuit. Washington State University Cooperative Extension. 268 p. Sarikhani, N., 2008: Forest utilization. Tehran University Press, Tehran. 728 p. (in Persian). Sessions, J., Boston, K., Murphy, G., Wing, M. G., Kellogg, L., Pilkerton, S., Zweede, J. C., Heinrich, R., 2007: Harvesting operation in the Tropics. Springer-Verlag, Berlin, Heidelberg. 170 p. Sobhani, H., Staurt, W. B., 1991: Harvesting Systems Evaluation in Caspian Forest. Journal of forest engineering 2(2): 21–24. Wang, J., Long, C., McNeel, J., Baumgras, J., 2004: Productivity and cost of manual felling and cable skidding in central Appalachain hardwood forests. Forest Prod J 54(12): 45–51.

Sažetak

Izvedba, mogućnosti i troškovi ručno-strojne sječe stabala u šumi tvrdih listača Hyrcanian Zbog velikih promjera krošanja te zbog prilično strmoga terena u hirkanskim šumama stabla se sijeku isključivo ručno-strojnom metodom. Ciljevi su ovoga istraživanja: provesti studij rada i vremena ručno-strojne sječe u tvrdim listačama, primjenom regresijskih funkcija razviti modele vremena radnih zahvata ručno-strojne sječe te procijeniti proizvodnost i troškove ručno-strojne sječe. Istraživanje je provedeno u odjelima 219 i 223 koji se nalaze u okrugu Namkhaneh unutar nastavno-pokusne šume Kheyrud. Za sječu je stabala korištena motorna pila STHIL s četiri konjske snage te vodilicom od 70 cm. Istraživanje je provedeno zimi od siječnja do veljače 2011. godine. Zimsko vrijeme, osobito hladnoća, ponekad utječu na radni učinak radnika sjekača. Radni se ciklus sastojao od određenih radnih zahvata i drugih čimbenika. Vrijeme za svaki radni zahvat i vrijednost svakoga čimbenika mjereno je na istraživanom radilištu. Čimbenici koji utječu na sječu ili operativne varijable za sječu koje su mjerene u istraživanju su udaljenost od stabla (cm), vrsta drveća, prsni promjer (cm), nagib po kojem se radnik kreće prilikom dolaska do stabla (%), nagib terena kod panja (%), smjer rušenja stabla. Ukupno su snimljena 233 radna ciklusa sječe stabla. Za izradu regresijskih jednadžbi korišten je statistički program SPSS 14.0. Rezultati provedene analize ukazuju na povećanje radnoga ciklusa s povećanjem dimenzija stabla. Postupna analiza podataka pokazala je značajan utjecaj prsnoga promjera na vrijeme sječe, te je stoga razvijen regresijski model za izračun ukupnoga vremena sječe, bez zastoja, na osnovi prsnoga promjera stabla, smjera rušenja te međusobne udaljenosti doznačenih stabala. Proizvodnost (m3/h) sa zastojima rada iznosila je 56,34 m3/h. Izmjerena proizvodnost za ručno-strojnu sječu, bez zastoja rada, iznosila je 80,7 m3/h. Proizvodnost (m3/h) bez zastoja rada iznosila je više nego proizvodnost sa zastojima rada. Povećanje dimenzija stabala značajno utječe na povećanje proizvodnosti. Kao krajnji rezultat analize prosječni trošak ručno-strojne sječe, sa zastojima rada, iznosio je 0,55 USD/ m3, dok je prosječan trošak ručno-strojne sječe bez zastoja rada iznosio 0,39 USD/ m3. Povećanje prsnoga promjera stabla te smjer rušenja utjecat će na povećanje troškova ručno-strojne sječe po proizvodnom ciklusu. Stabla prsnih promjera od 20 do 50 cm imaju značajan utjecaj na troškove ručno-strojne sječe, koji se kreću u rasponu od 1,2 do 0,2 USD/ m3. Na vrijeme izrade zasjeka, završnoga reza te vremena zastoja otpada najveći dio vremena radnoga ciklusa. U ovom su radu predstavljene najvažnije varijable koje utječu na ručno-

292

Croat. j. for. eng. 34(2013)2

Performance, Capability and Costs of Motor-Manual Tree Felling... (283–293)

M. Jourgholami et al.

strojnu sječu te ih je prije provođenja radova ručno-strojne sječe potrebo vrednovati. Dokazano je da su promjer panja i udaljenost između doznačenih stabala najutjecajniji čimbenici potrošnje vremena i proizvodnosti prilikom sječe. Proizvodnost pri sječi stabala većih promjera veća je nego pri sječi stabala manjih promjera. Ključne riječi: sječa stabala, studij rada i vremena, regresijski model, proizvodnost, troškovi

Authors’ address – Adresa autorâ:

Received (Primljeno): February 12, 2012 Accepted (Prihvaćeno): July 12, 2013 Croat. j. for. eng. 34(2013)2

Asst. Prof. Meghdad Jourgholami, PhD. e-mail: mjgholami@ut.ac.ir Prof. Baris Majnounian, PhD.* e-mail: bmajnoni@ut.ac.ir Assoc. Prof. Nosratollah Zargham, PhD. Department of Forestry and Forest Economics Natural Resources Faculty University of Tehran P.O. Box: 31585 – 4314, Karaj IRAN *Corresponding author – Glavni autor

293

Original scientific paper – Izvorni znanstveni rad

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning of High-Quality Wood Plantations Enrico Marchi, Francesco Neri, Marco Fioravanti, Rodolfo Picchio, Giacomo Goli, Giuseppina Di Giulio Abstract – Nacrtak Arboriculture plantations aim to produce high-quality wood. In order to investigate the type and extent of mechanical injury that pruning causes to tree cambium as well as the effects on the healing process, different types of shear were selected and used in an eight-year-old Quercus robur L. plantation. The amount of removed, detached and crushed bark was assessed by means of image analysis immediately after pruning. After 15 months, the effect of different cutting patterns on the healing process was investigated by measuring the area of the pruned branch covered by woundwood (HI1). Five years after pruning, the same analysis was performed above and below bark (HIo5 and HIu5) and a number of parameters were assessed in order to quantify the quantity and quality (symmetry) of woundwood growth and the healing time for sealing. The action of pruning tools depends on cutting pattern and branch diameter. The greater the diameter, the longer the healing time. The double-blade tool caused less injury and showed the fastest healing process. The use of double blade pruning tools is thus recommended to improve the performance of wood quality production in arboriculture plantations. We also recommend the healing index HI1 for an early assessment of pruning damage. Keywords: pedunculate oak, pruning tools, agroforestry, wood quality, occlusion

1. Introduction – Uvod Forest trees have been pruned for centuries in order to increase wood quality and to shape the tree (O’hara 2007). Pruning meets different objectives, such as: fuelwood production; aesthetic improvement; dead branch removal, reduction or prevention of the attack of pathogens. However, the most important purpose of tree pruning is usually to increase the quality of the log to be used in sawing, peeling or slicing operations (O’hara 2007, Kupka 2007), in particular in plantation to produce high quality wood (Springmann et al. 2011). The main aim is to get at least 2.5 m of stem length without knots or related defects in logs with a diameter larger than 30–40 cm (Mohni et al. 2008). In fact, on average, the butt log represents 90% of a tree’s economic value (Kronauer 2009) and knot should be limited to the inner 8–10 cm of the log diameter. This suggests the need for early pruning. Croat. j. for. eng. 34(2013)2

After accurate pruning action, the wood growth will likely be free of defects and will consequently achieve greater wood quality than unpruned trees. Nevertheless, branch cutting is a stressful action for trees (Springmann et al. 2011) and pruning should avoid a sudden reduction of the total leaf area, so that the growth increment is kept regular. A severe pruning or pruning that removes foliage solely from the upper crown will improve stem form and reduce the size of the defected core thereby increasing clearwood production (Medhurst et al. 2006). However, it may also reduce tree growth or stimulate a tree response by developing secondary shoots (Alcorn et al. 2008), thus having a negative effect on clearwood production. The effects of pruning on wood quality involve a lot of factors: e.g. pruning season, pruning methods, tree species, as well as the diameter of the cut branches, which in high quality arboriculture plantation should

295

E. Marchi et al.

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning ... (295–304)

be lower than 3 cm (DeBell et al. 2006, Dujesiefken et al. 1998, 2005, Nicolescu and Kruch 2009). Many studies on tree pruning concern the methods of branch removal and subsequent compartmentalization (Shigo and Marx 1977) and wound occlusion. These studies show contrasting results. Several works indicate that a cut close to the stem is effective for pruning trees for a more rapid occlusion and to enhance wood quality by avoiding infections of fungi or bacteria (Brodie and Harrington 2006, DeBell et al. 2006). Other studies indicate that for a proper pruning the branch collar should not be removed from the stem, thus improving compartmentalization response (Shigo 1984, Dujesiefken et al. 1998, Smith 2006). A more rapid occlusion in trees with faster radial growth rates, both conifers and broadleaves, was observed (Roth 1948, O’Hara and Buckland 1996, Petruncio et al. 1997). Some studies revealed a more rapid occlusion after pruning live branches than dead branches and dead branches occluded more rapidly if the branch collar was intentionally injured during pruning (Brodie and Harrington 2006). The mechanical injuries caused to the bark along the perimeter of the cutting section affect the healing process, they may favour abnormal wood coloration/ discoloration and/or pathogen attacks, and, ultimately, they reduce timber quality and value (Pearce 2000, Dujesiefken et al. 2005, Brunetti et al. 2006, Nocetti et al. 2011). Infection from fungal decay organisms has always been a concern with pruning. However, many authors found little or no evidence of decay in broadleaved trees in the northern hemisphere (Chiu et al. 2002, DeBell et al. 2006) and sometimes found less decay in pruned trees than unpruned trees (Skilling 1958). Nevertheless, pruning of thicker branches may cause extensive discolouration and decay in the trunk even though the cuts are made correctly (Dujesiefken et al. 1998). In this considerable international body of literature on tree pruning, there is a dearth of studies on the effects of pruning tools. Very few studies performed until now consider the effect of pruning tool cutting pattern (Baldini et al. 1997, Marchi and Rossi 2007, Schatz et al. 2008). Other studies on different pruning tool cutting patterns were focused only on the force requirements for manual pruning (Crossland et al. 1997, Parish 1998). The use of different cutting patterns in fact could result in different occlusion time, wounding patterns, wood defects, defects extension into the bole, and decay. The aim of this paper was to investigate the relation between pruning tools cutting patterns and the me-

296

chanical injury to tree bark in Quercus robur L. plantation, and to analyze the effect of the cutting pattern of pruning tools on occlusion processes. These results will help to minimize direct and indirect unfavorable pruning effects on wood quality, and to give a contribution in improving research on the methods of branch removal and subsequent wound occlusion.

2. Materials and Methods – Materijal i metode A pruning operation was carried out in an eightyear-old plantation located in S. Barbara (Arezzo – Central Italy – 43°34’59.49” N, 11°28’23.39” E). It was a mixed stand of Q. robur, as the main species, and Alnus cordata, as a secondary species. The planting pattern was square with 3 meters between the trees and the main and the secondary species were alternate along the row and between the rows. In the pruning campaign, carried out in April 2006, three shears with different cutting patterns were used (Fig. 1): bypass (BP), draw cut (DC) and double blade (DB). The shears were selected to represent the range of tool designs currently used in Italy. Thirteen trees of Q. robur (Table 1) were selected for a total of 36 branches cut by each shear. For every single tree the branches were cut with the same shear. For each branch cut the diameter, the height above ground and the angle with reference to the North were measured (the first with a caliper, the second with a measuring tape and the last one using a Minerva compass) in order to be able to locate the cuts subsequently. In pruning, the cut left the branches bark ridge and the branch collar intact; only branches less than 3.5 cm in diameter were cut. To evaluate mechanical injury, each cutting section was photographed with a macroobjective. The camera was parallel to the cut surface. The amount of bark injury to the section perimeter was estimated by means of a CAD software (AutoCAD 2002, Autodesk Inc., USA). The bark injury (BI) was quantified by measuring the angle included in the arc of the injured section perimeter (Fig. 2). Fifteen months later, the cutting sections were rephotographed with the same method and they were compared with the pictures taken after pruning. The area of each cutting section covered by callus and woundwood (HA) was determined by means of an image analysis software (ImageJ 1.39u, National Institute of Healh, USA), and then compared with the total area of each wound after pruning (TA). A healing process index (HI) was calculated as a ratio between HA and TA. Croat. j. for. eng. 34(2013)2

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning ... (295–304)

E. Marchi et al.

Table 1 Characteristics of the Quercus robur trees at the time of pruning (2006): DBH, diameter at breast height, and diameter and height above ground level of the pruned branches (S.E. and N). The values of felled trees refer to 2006. DBH was measured overbark on standing trees in 2006 and underbark on crosscut sections in 2011. The differences among groups were not statistically significant (Kruskal-Wallis test, p-level 0.05) Tablica 1. Značajke stabala hrasta lužnjaka u vrijeme orezivanja grana (2006): DBH, prsni promjer; promjer orezanih grana i visina grana od tla (standardna pogreška i N). Vrijednosti se odnose na 2006. godinu. Prsni je promjer mjeren s korom na dubećim stablima 2006. godine i bez kore na isječcima 2011. godine. Razlike među grupama nisu bile statistički značajne (Kruskal-Wallisov test, p-level 0.05)

DB All trees in 2006 Sva stabla 2006.

BP DC DB

Trees felled in 2011 Stabla posječena 2011.

BP DC

Tree DBH, mm

Branch diameter, mm

Height a.g.l., cm

Prsni promjer stabla, mm

Promjer grane, mm

Visina od tla, cm

50.0

15.7

152.9

(4.6 N = 5)

(0.1 N = 36)

(7.8 N = 36)

60.2

15.7

142.9

(13.3 N = 4)

(0.1 N = 36)

(10.2 N = 36)

61.4

17.0

147.9

(2.7 N = 4)

(0.1 N = 36)

(10.6 N = 36)

48.2

15.7

164.7

(8.1 N = 3)

(0.1 N = 21)

(12.4 N = 21)

65.8

15.8

143.6

(16.7 N = 3)

(0.1 N = 32)

(11.3 N = 32)

61.0

17.0

147.9

(2.7 N = 4)

(0.1 N = 36)

(10.6 N = 36)

Fig. 1 Outline of cutting pattern Slika 1. Izgled načina rezanja Five years after the pruning campaign (March 2011), ten pruned trees were cut for a total of 21, 32 and 36 cutting sections for DB, BP and DC, respectively. The cutting sections were re-photographed and measured in order to compare them with the previous results and pictures (taken immediately after pruning Croat. j. for. eng. 34(2013)2

Fig. 2 Example of the method used for measuring the damage to cambium in fresh cuttings. The injury was quantified by measuring the angle included in the arc of the injured section perimeter Slika 2. Primjer metode korištene za mjerenje oštećenja kambija na svježim prerezima. Ozljeda je kvantificirana mjerenjem kuta kružnoga isječka na čijem se luku nalazi oštećenje

297

E. Marchi et al.

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning ... (295–304)

Fig. 3 Asymmetrically (a) and symmetrically (b) healed cuts, and cross section (c) with detail of: distance between cutting section and underbark stem perimeter (SC) and thickness of entrapped cork layer (DW) Slika 3. Asimetrično (a) i simetrično (b) zarasli rezovi te presjek (c) s detaljima: udaljenost mjesta reza od plašta debla bez kore (SC) i debljina urasle kore (DW)

and fifteen months after pruning) by measuring overbark HI (HIo5). Moreover the cutting sections that had not completely healed were counted and measured. The symmetry of the healing process for each cutting tool was measured and a symmetry index (HSI) was calculated as a ratio of the lower and higher value of the distance between the perimeter of the cutting section and the contact line of callus (an asymmetrically healed cut is shown in Fig. 3a and a symmetrical healed cut is shown in Fig. 3b). Some pruned brunches (4 for DB, 4 for BP and 9 for DC) were no longer visible overbark; for these branches the HSI was not calculated. The distances were measured by a ruler to the nearest 0.5 mm. Then, the bark over the cutting sections was carefully removed in order to measure the area of each cutting section covered by woundwood (HAu).The healing process index underbark (HIu5) was then calculated as a ratio between (HAu) and the total area of each woody section after pruning. Finally, each stem was crosscut at the knot level and the knots were analyzed. Two, two and four knots obtained by DB, BP and DC, respectively, were no longer recognized. On each cross section the following variables were measured (Fig. 3c): distance between the cutting section of the branch when pruned and underbark stem perimeter (SC); thickness of entrapped cork layer (DW); number of years to complete the healing process (HT). In relation to the unhealed knots, in order to not exclude these cases from the analysis, we considered that the unsealed branch would seal in 1 more year, i.e. a value of 6 years was adopted. A damaged woundwood index (DWI) was calculated as a ratio between DW and SC.

298

The value of underbark DBH in 2006 and 2011 was measured for each felled tree on the cross section at 1.3 m above ground. The values of SC, DW and DBH were measured by a ruler to the nearest 0.5 mm. Data were checked for normality (KolmogorovSmirnov test) and homogeneity of variance (Levene test). The Kruskal-Wallis non-parametric multiplecomparison test was used to test differences between non normally distributed variables (Sprent and Smeeton 2001, Lo Monaco et al. 2011, Picchio et al. 2009), e.g. tree diameter, branch height above ground, HI1, HIu5, HSI, SC, DWI, and HT. In order to test the differences between the characteristics of trees and branches at the time of pruning (2006), the same test, i.e. Kruskal-Wallis, was also applied to cut branch diameter. A one-way ANOVA was applied to determine the effects of cutting pattern on the extent of injury immediately after pruning (2006), and differences were tested by the Tukey HSD test. All differences were considered as significant when p ≤ 0.05. A multiple linear regression was applied to test the relation between type of pruning tools, years for the healing process to be complete and branch diameter. All statistics used the Statistica 7.0 software (StatSoft, Tulsa OK, USA).

3. Results – Rezultati Non-significant statistical differences were found in the diameter at breast height (DBH), in the diameter of the cut branches and in the height above ground level of the branches of the thirteen trees in 2006 and of the ten felled trees in 2011 (Table 1). Croat. j. for. eng. 34(2013)2

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning ... (295–304)

E. Marchi et al.

3.1 Mechanical injury immediately after pruning Mehanička ozljeda neposredno nakon orezivanja Three kinds of bark injury were found over the cutting section perimeter, namely: crushing, i.e. the bark appeared to be more compact; detaching, i.e. the bark was separate from the wood; and removal, where some bark parts were missing. The cutting pattern did not significantly affect the kind of injury (data not shown), while the extent of injury varied in the following order: DB < DC < BP (p < 0.001) (Fig. 4).

3.2 Healing process after 15 months – Proces zarašćivanja nakon 15 mjeseci The cutting pattern significantly affected HI1 (p < 0.001). Cuts made using double blade tools showed a quicker healing process, while draw cut and bypass showed a slower and similar healing process (Fig. 5). 72% of the branches cut by DB completely healed their wounds (HI1), and 19% closed 90% of the wound. In contrast, just 11% and 8% of branches cut by BP and DC respectively attained complete healing (Fig. 6a). Callus and woundwood developed in a quite circular pattern in the healing process of cuts made by double blade tools, whereas when using bypass and

Fig. 5 Healing process index (HI1) fifteenth month after pruning (2007) for different cutting patterns (+/– S.E.). Different letters show significant differences (Kruskal-Wallis multi-comparison test, p < 0.05, N = 36) Slika 5. Indeks procesa zarašćivanja (HI1) petnaest mjeseci nakon orezivanja (2007) za različite načine rezanja (+/– standardna pogreška). Različita slova pokazuju signifikantne razlike (Kruskal-Wallisov multiusporedni test, p < 0.05, N = 36) draw cut tools they developed an irregular shape (Fig. 7). In particular, 50% of the branches cut by draw cut shears developed callus and woundwood mainly on one side, i.e. the side cut by the blade (data not shown).

3.3 Healing process 5 years later – Proces zarašćivanja nakon pet godina

Fig. 4 Damage assessed immediately after pruning (2006) for different cutting patterns (+/– S.E.). Different letters show significant differences (Tukey HSD test, p<0.05, N = 36) Slika 4. Oštećenje ustanovljeno neposredno nakon orezivanja (2006) za različite načine rezanja (+/- standardna pogreška). Različita slova pokazuju signifikantne razlike (Tukey HSD test, p < 0.05, N = 36) Croat. j. for. eng. 34(2013)2

In 2011, cuts by DB confirmed their better ability to seal in shorter time periods. 100% of DB branches were completely healed, contrasting with only 90.6% of BP branches and 71.9% of DC branches (Fig. 6c). However, no statistically significant differences arise for HIu5 (Table 2). A large part of the wounds were sealed, and HIo5 was almost 1 (Fig. 6b). DB showed a more symmetrical healing process, followed by BP, but not significant differences (p = 0.09) were observed (Table 2). A symmetrical healing process could mean that the cambium worked well on both sides and this could lead to shorter sealing times. The distance between the cutting section of the branch and the underbarked stem perimeter (SC) was higher in the knots obtained by DB and BP than in DC knots (Table 2), i.e. BP and DB resulted in shorter sealing times.

299

E. Marchi et al.

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning ... (295–304)

Fig. 6 Distribution of healing process index (HI) classes (%) – 3 shears: double blade (DB), bypass (BP) and draw cut (DC) – Quercus robur L. – a) HI1: healing index assessed fifteen months after pruning; b) HIo5: healing index assessed overbark five years after pruning; c) HIu5: healing index assessed underbark five years after pruning Slika 6. Distribucija po razredima indeksa procesa zarašćivanja (HI) – 3 tipa škara: s dvostrukim sječivom (DB), s mimoilaznim sječivom (BP) i s jednostrukim sječivom (DC) – Quercus robur L. – a) HI1: indeks zarašćivanja ustanovljen petnaest mjeseci nakon orezivanja; b) HIo5: indeks zarašćivanja ustanovljen na kori pet godina nakon orezivanja; c) HIu5: indeks zarašćivanja ustanovljen nakon uklanjanja kore pet godina nakon orezivanja

DWI differed significantly between DB and DC (Table 2). DWI of DB and DC showed that 29% and 54% of the new wood, respectively, was damaged by entrapped cork layer. DB showed the shortest time for sealing the pruned branch (HT) while DC and BP did not differ. The healing time (HT) increased with increasing branch diameter for the three cutting patterns (Figure 8). Finally a linear regression with dummy variable was performed to test the relation between years to

complete the healing process, type of pruning tool and branch diameter (at 2006). The regression was statistically significant (p < 0.001; R2adjusted = 0.27; SE = 1.13; MAE = 0.85) and the result is shown in Equation (1) and Fig. 8: y = 3.96 – (0.76 × t) + (0.05 × D)

(1)

Where y is the number of years needed for sealing, t the pruning tool pattern (dummy variable; DC=1, BP=2, DB=3), and D the branch diameter at cutting

Table 2 HIu5, healing index assessed under bark; HSI, healing symmetry index; SC, distance between cutting section of a branch when pruned and underbarked stem perimeter; DWI, damaged woundwood index; HT, healing time five years after pruning by double blade (DB), bypass (BP) and draw cut (DC) shears (S.E. and N). Different letters show significant differences among values in a column (Kruskal-Wallis test, p < 0.05) Tablica 2. HIu5, indeks zarašćivanja ustanovljen nakon uklanjanja kore; HSI, indeks simetrije zarašćivanja; SC, udaljenost između mjesta reza grane pri orezivanju i plašta debla bez kore; DWI, indeks oštećenja drva rane; HT, vrijeme zarašćivanja pet godina nakon orezivanja škarama s dvostrukim sječivom (DB), s mimoilaznim sječivom (BP) i s jednostrukim sječivom (DC) (S.E. i N). Različita slova označuju signifikantne razlike među vrijednostima u stupcu (Kruskal-Wallisov test, p < 0.05) Tool – Oruđe DB BP DC p

HIu5

HSI

SC, cm a

1.00

0.81

18.0

(na N = 21)

(0.04 N = 17)

(1.4 N = 19) a

0.99

0.78

18.4

(0.01 N = 32)

(0.04 N = 28)

(1.1 N = 30) b

DWI 0.29

HT, years – HT, godine

2.4 a

(0.04 N = 19) 0.42

3.6 b

(0.05 N = 30) 0.54

(0.1 N = 19)

(0.2 N = 30) 4.0 b

0.95

0.65

12.1

(0.02 N = 36)

(0.05 N = 27)

(0.6 N = 32)

(0.06 N = 32)

(0.2 N = 32)

< 0.001

< 0.03

< 0.001

na: not available because all the wounds were completely healed, i.e. HIu5 = 1 – na: nije dostupno zbog toga što su sve rane potpuno zarasle, odnosno HIu5 = 1 ns: not significant – ns: nije signifikantno

300

Croat. j. for. eng. 34(2013)2

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning ... (295–304)

E. Marchi et al.

Fig. 7 Examples of a cutting section after pruning (above) and 15 months later (below). a – Double blade shear; b – Draw cut shear; c – Bypass shear. Quercus robur L. Slika 7. Primjeri mjesta reza nakon orezivanja (gore) i 15 mjeseci poslije (dolje.); a – dvostruko sječivo; b – jednostruko sječivo; c – mimoilazno sječivo. Quercus robur L.

4. Discussion and conclusion – Rasprava i zaključak

Fig. 8 Healing Time (HT in years) versus branch diameter per each cutting pattern Slika 8. Ovisnost vremena zarašćivanja (HT u godinama) o promjeru grane za pojedini način rezanja time. Only 27% of the variability of the healing time was explained by the independent variables considered. Croat. j. for. eng. 34(2013)2

The analysis of damage caused by the different tools over time showed contrasting results. The injuries measured immediately after pruning showed that BP was the worst cutting pattern. Fifteen months later, no significant differences between DC and BP were recorded, suggesting that the injury caused by DC was underestimated when recorded immediately after pruning. Therefore, damage estimation by measuring the angle included in the arc of the injured section perimeter cannot be recommended for Q. robur immediately after pruning. This may be due to the higher patch area of the DC anvil relative to the BP hook which, together with the elevated Q. robur bark thickness, may have minimized the evidence of cambium compression during macroscopic analysis. In the long term (5 years), such a compression resulted in a decline of radial growth so that the distance between the cutting section and the underbark stem perimeter (SC) was lower for DC than for the other two tools. It may be interesting to apply the same method of pruning damage analysis to trees with a thinner and softer bark. The healing index HI1 applied one year after the pruning turned out to be a better parameter for early

301

E. Marchi et al.

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning ... (295–304)

damage assessment. HI1 is in fact a non destructive method, whose results were confirmed by the assessment of the healing time HT. HT was destructively applied five years after the pruning when most of the cuttings were sealed, which resulted in non significant differences in HI5. The woundwood analysis showed a minimum DWI value of 0.29 for the DB tool and a maximum value of 0.54 for DC. This means that the portion of stem core with knots or wood with defects will increase by 0.58 cm for DB and 1.08 cm for DC relative to the tree diameter at pruning time. Double-blade shears are, therefore, to be recommended for increasing the quality of timber. The time necessary to complete the healing process (HT) was less when DB was used, suggesting that double-blade shears maximize the quantity of high-quality timber. HT for BP and DC were 1.5 and 1.7 times higher than for DB, respectively. These figures may even be underestimated because the unhealed knots were given a further one-year period to complete the healing process. The hypothesis that a symmetrical healing process implied an optimal woundwood growth on both sides and thus shorter sealing times was not confirmed. In fact, the healing symmetry index (HSI) did not significantly depend on the pruning tool. The hypothesis that the larger the pruned branch diameter, the longer the time to seal (Joyce et al. 1998, Nicolescu and Kruch 2009) was confirmed, thus suggesting that early pruning (diameter < 3 cm) should be carried out. The double-blade cutting pattern caused the least mechanical injury. We postulate that this is because both blades penetrate the wood and cut the bark tissues clean off. By contrast, the hook and the anvil of the other cutting patterns oppose the cutting force necessary for the blade to cut the tissue, but cause evident injury at the point of contact between the hook or the anvil and the bark. In DC the anvil patch area involves both the cut branch and branch collar causing bark injury close to the stem. In BP the hook patch area should involve only the cut branch but the lack of contact between the hook and the blade in the terminal phase of a cut causes injury. When the cutting parts of a by-pass tool come in touch with each other, and the branch starts bending downwards, the hook tends to slip, thus damaging the bark close to the stem. Ultimately the DB tool showed the best pruning performance and is to be recommended as short healing processes may avoid pathogen attacks which cause a reduction of timber quality and value. In conclusion, the effect of pruning tool used in branch removal on wound occlusion process is not negligible, suggesting that a higher attention to the

302

pruning tool is recommended for improving research on the pruning methods and on the subsequent physiological response of trees.

Acknowledgement – Zahvala Authors would like to thank Claudio Bidini for the execution of the pruning operations.

5. References – Literatura Alcorn, P. J., Bauhus, J., Thomas, D. S., James, R. N., Smith, R. G., Nicotra, A. B., 2008: Photosynthetic response to green crown pruning in young plantation-grown Eucalyptus pilularis and E. cloeziana. Forest Ecology and Management 255: 3827–3838. Baldini, S., Brunetti, M., Fabbri, P., 1997: Pruning tests using different tools in a wild cherry plantation. Ann. Ist. Sper. Selv. XXV and XXVI: 345–361 (in Italian, English abstract). Brodie, L. C., Harrington, C. A., 2006: Response of young red alder to pruning. in: Red alder – a state of knowledge, R. L. Deal and C. A. Harrington (eds.). USDA For. Serv., Department of Agriculture, Pacific Northwest Research Station. Gen. Tech. Rep. PNW-GTR-669: 95–102. Brunetti, M., Nocetti, M., Zanuttini, R., 2006: Effects of pruning on the wood quality of walnut. Sherwood 125: 5–9 (in Italian, English abstract) Chiu, C. M., Lo C. C. N., Suen, M. Y., 2002: Pruning method and knot wound analysis of Taiwan zelkova (Zelkova serrata Hay.) plantations. Taiwan J. For. Sci. 17:503–513. Crossland, G., Murphy, G., Martin, G., Dean, M., 1997: Energy and force requirements for six pruning shear designs. N. Z. Forestry 42(3): 22–27. Debell, D. S., Harrington, C. A., Gertener, B. L., Singleton, R., 2006: Time and distance to clear wood in pruned red alder saplings. in: Red alder – a state of knowledge, R.L. Deal and C.A. Harrington (eds.). USDA For. Serv., Department of Agriculture, Pacific Northwest Research Station. Gen. Tech. Rep. PNW-GTR-669: 103–113. Dujesiefken, D., Liese, W., Shortle, W., Minocha, R., 2005: Response of beech and oaks to wounds made at different times of the year. European Journal of Forestry Research 124: 113–117. Dujesiefken, D., Stobbe, H., Eckstein, D., 1998: Long-term effects of pruning on the trunk of lime and horsechestnut. Forstwissenschaftliches Centralblatt 117(6): 305–315. Kronauer, H., 2009: Laubholz zu Wertholz erziehen. Allgemeine Forstzeitschrift 64: 920–921. Kupka, I., 2007: Growth reaction of young wild cherry (Prunus avium L.) trees to pruning. Journal of Forest Science 53(12): 555–560. Lo Monaco, A., Todaro, L., Sarlatto, M., Spina, R., Calienno, L., Picchio, R., 2011: Effect of moisture on physical paramCroat. j. for. eng. 34(2013)2

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning ... (295–304) eters of timber from Turkey oak (Quercus cerris L.) coppice in Central Italy. For. Stud. China 13(4): 276–284 Marchi, E., Rossi, S. 2007: Pruning with different tools. Tests on English oak, walnut and cherry. Sherwood 132: 27–31 (in Italian, English abstract). Medhurst, J. L., Pinkard, E. A., Beadle, C. L., Worledge, D., 2006: Photosynthetic capacity increases in Acacia melanoxylon following form pruning in a two-species plantation. Forest Ecology and. Management 233(2-3): 250–259. Mohni, C., Pelleri, F., Hemery, G.E., 2009: The modern silviculture of Juglans regia L: a literature review. Bodenkultur – Wien and Munchen 60(3): 212–34. Nicolescu, V. N., Kruch, J., 2009; Research on the effects of various silvicultural interventions on young wild cherry (Prunus avium L.) trees. Revista Padurilor 124(3): 8–16. Nocetti, M., Brunetti, M., Giovannelli, A., 2011: Improvement of Pruning Technique to Minimize Wood Discoloration in English Walnut (Juglans regia L.). Proceedings of the 3rd International Scientific Conference on Hardwood Processing, ISCHP, 16–18 October 2011, Virginia, USA: 245–254. O’Hara, K. L., Buckland, P. A., 1996: Prediction of pruning wound occlusion and defect core size in ponderosa pine. Western Journal of Applied Forestry 11(2): 40–43. O’Hara, K. L., 2007: Pruning wounds and occlusion: A long-standing conundrum in forestry. Journal of Forestry 105(3): 131–138. Parish, R.L., 1998: Operating force requirements for manual pruning shears. Applied Engineering in Agriculture 14(4): 349–352. Pearce, R. B., 2000: Decay development and its restriction in trees. Journal of Arboriculture 26(1): 1–11.

E. Marchi et al.

Petruncio, M., Briggs, D., Barbour, R. J., 1997: Predicting pruned branch stub occlusion in young, coastal Douglasfir. Canadian Journal of Forestry Research 27(7): 1074–1082. 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. Roth, E. R., 1948: Healing and defects following oak pruning. Journal of Forestry 46(7): 500–504. Schatz, U., Kannisto, K., Rantatalo, M., 2008: Influence of saw and secateur pruning on stem discolouration, wound cicatrisation and diameter growth of Betula pendula. Silva Fennica 42: 295–305. Shigo, A. L., Marx, H. G., 1977: Compartmentalization of decay in trees. Agriculture Information Bulletin No. 495. 73 p. Shigo, A. L., 1984: Tree decay and pruning. Journal of Arboriculture 8: 1–12. Skilling, D. D., 1958: Wound healing and defects following northern hardwood pruning. Journal of Forestry 56 (1): 19–22. Smith, K., 2006: Compartmentalization today. Arboricultural Journal 29: 173–184. Sprent, P., Smeeton, N. C., 2001: Applied Nonparametric Statistical Methods. 3rd edn. Chapman & Hall/CRC, London, 461 p. Springmann, S., Rogers, R., Spiecker, H., 2011: Impact of artificial pruning on growth and secondary shoot development of wild cherry (Prunus avium L.). Forest Ecology and Management 261(1): 764–769.

Sažetak

Utjecaji načina rezanja škara na zarašćivanje pri orezivanju stabala u plantažama za proizvodnju drva visoke kakvoće Stabla se u plantažama orezuju radi proizvodnje visokokvalitetnoga drva. Da bi se odredili tip i opseg mehaničkih ozljeda koje orezivanje uzrokuje kambiju stabla te utjecaj na zarašćivanje, odabrani su različiti tipovi škara (s dvostrukim sječivom – DB, s jednostrukim sječivom – DC i s mimoilaznim sječivom BP) kojima je orezana osmogodišnja plantaža hrasta lužnjaka. Količina uklonjene, odvojene i nagnječene kore ustanovljena je analizom snimaka neposredno nakon orezivanja. Petnaest mjeseci nakon orezivanja mjerenjem zarasle površine orezane grane istraživan je utjecaj različitih načina rezanja na zarašćivanje. Na temelju zarasle i ukupne površine prereza određen je indeks procesa zarašćivanja (HI1). Pet godina nakon orezivanja jednaka je analiza provedena mjerenjem na kori (HIo5) i nakon uklanjanja kore (HIu5). Osim toga ustanovljeni su i ostali parametri radi određivanja količine i kakvoće (simetrije) zarašćivanja te vremena zarašćivanja. Utvrđeno je da utjecaj oruđa za orezivanje ovisi o načinu rezanja i promjeru grane. S povećanjem promjera grane raste i vrijeme zarašćivanja. Uporaba oruđa s dvostrukim sječivom uzrokovala Croat. j. for. eng. 34(2013)2

303

E. Marchi et al.

Effects of Cutting Patterns of Shears on Occlusion Processes in Pruning ... (295–304)

je manje ozljede i brži process zarašćivanja. Stoga se radi poboljšanja kakvoće drva pri orezivanju plantaža preporučuje uporaba oruđa s dvostukim sječivom. Također se preporučuje primjena indeksa procesa zarašćivanja (HI1) za rano određivanje štete uzrokovane orezivanjem. Ključne riječi: hrast lužnjak, oruđa za orezivanje, agrošumarstvo, kakvoća drva, zarašćivanje

Authors’ address – Adresa autorâ: Assoc. Prof. Enrico Marchi, PhD. e-mail: enrico.marchi@unifi.it Francesco Neri, PhD. e-mail: francesco.neri@unifi.it Marco Fioravanti, PhD. e-mail: marco.fioravanti@unifi.it Giacomo Goli, PhD.* e-mail: giacomo.goli@unifi.it Giuseppina Di Giulio University of Firenze, Department of Agricultural, Food and Forestry Systems (GESAAF) Via S. Bonaventura 13 50145 Firenze ITALY

Received (Primljeno): January 11, 2013 Accepted (Prihvaćeno): July 20, 2013

304

Rodolfo Picchio, PhD. e-mail: r.picchio@unitus.it University of Tuscia, Department of Agriculture, Forests, Nature and Energy (DAFNE) Via S. Camillo de Lellis 01100 Viterbo ITALY *Corresponding author – Glavni autor Croat. j. for. eng. 34(2013)2

Preliminary note – Prethodno priopćenje

Possibility of Determination of Daily Exposure to Vibration of Skidder Drivers Using Fleet Manager System Zdravko Pandur, Dubravko Horvat, Marijan Šušnjar, Marko Zorić Abstract – Nacrtak This paper describes an indirect method of determining exposure to vibration of skidder drivers using the Fleet Manager System. The hand-arm exposure of workers to vibration is expressed as energy equivalent A(8), which is determined by the procedure clearly described in the international standards ISO 5349-1-2001 and ISO 2631-1-1997. A(8) is a value that depends not only on the vibration level in certain operating procedures, but also on the duration of exposure (duration of each skidder working procedure). Research was done on the skidder Ecotrac 120V equipped with the Fleet Manager System (FMS). The role of the FMS is to measure engine speed and duration of a certain engine speed during the working day. The analysis of the working days was performed in the aim to connect skidder working elements (driving, winching, pulling out of winch rope, etc.) with engine speeds. Vibration on the steering wheel and seat of the skidder Ecotrac 120V was measured by vibrometer with triaxial accelerometer (Brüel and Kjaer 4447) at different engine rotational speeds. The exposure to vibration of a skidder driver on a daily basis A(8) was calculated using data of the summarized durations at certain skidder engine rotational speeds measured by the FMS and the measured level of vibrations at specific engine rotational speeds. Keywords: vibrations, A(8), engine rotational speed, Fleet Manager System, skidder

1. Introduction – Uvod Field scientific research, such as the research of some exploitation characteristics of forest vehicles during extraction of different forest products, the impact of extraction on some soil characteristics aimed at determining the environmental viability and its impact on ergonomic conditions in the operator’s cabin, requires a lot of time spent in the field and is hence very expensive. The Fleet Manager System (hereinafter FMS) is a system of remote monitoring and control of the vehicle operation, which enables gathering of data without disturbing the vehicle operation, i.e. it provides the possibility to research in almost uncontrollable exploitation conditions. The FMS is a very useful tool for the control and organization of the complex system of production and supply of wood chips from the place of chipping to the buyer (Holzleitner et al. 2013.) as well as during the time study. Croat. j. for. eng. 34(2013)2

The use of FMS as the tool in the process of data gathering from the vehicle aimed at measuring and determining some ergonomic parameters such as vibrations transmitted through the steering wheel on the hands and through the seat to the whole body of the operator, is not known in the literature. Goglia et al. (2012) consider that, with the methodology of determining 8-hour energy equivalent of the total value of the estimated accelerations A(8), an accurate picture of the working day should first be made and whole day shooting of the operator’s work with the film camera is one of the ways to get such a picture. The same authors state that, in practice, it is practically impossible to measure the levels of vibrations for each activity and for this reason it is necessary to make initial measurements in the test polygon under controlled conditions. Measuring vibrations on chain saws, Rottensteiner et al. (2012) conclude that the level of vibrations is con-

305

Z. Pandur et al.

Possibility of Determination of Daily Exposure to Vibration of Skidder Drivers ... (305–310)

siderably affected by wood density and that it should, therefore, be one of the basic parameters in calculating the worker’s daily exposure to vibrations A(8). Goglia et al. (2003) measure vibrations on the steering wheel of the small farming tractor while idling and under full load, and then they calculate the daily exposure to vibrations A(8), which amounts to 14.28 m/s2. According to ISO 5349-1-2001, it means that in less than two years adverse effects of vibrations can be expected with 10% of tractor drivers. Dewangan and Tewari (2009) measure vibrations on steering handles of one-axle farming tractor during driving and soil processing at three tractor speeds. The highest vibrations were measured at the lowest tractor speed and during soil processing, and actually higher vibrations were measured during soil processing. Poje (2011) concluded that the worker`s exposure to whole-body vibration were the highest during the skidding operations and for operations where a cable skidder was moving with no load (1.31 m/s2) and ramping (1.22 m/s2), and the lowest with the full load (0.91 m/s2). The A(8) value does not only depend on the vibration level in certain operating procedures, but also on the duration of exposure, i.e. on the duration of each skidder operating procedure. Therefore the FMS, as the system for monitoring and control of the vehicle operation, is highly suitable for determining the A(8) values, because the analysis of data gathered from the vehicle (time, engine rotational speed, winch/crane operation, position/moving, vehicle speed, etc.) can easily show the duration of individual operations at specific engine rotational speed. By subsequent measurement of vibrations (ahv) at characteristic engine speeds, the A(8) value can be simply calculated, as specified by the standard ISO 5349-1-2001. 2 2 2 ahv = ahwx + ahwy + ahwz

1 A(8) = T0

∑a i =1

2 hvi

⋅ Ti

(1) (2)

Where: T0 = available time of 8 h or 28 800 s, ahvi = total exposure to vibrations for i operation, Ti = duration of this operation, N = the total number of operations. Sherwin et al. (2004) state that with harvesters the vibration level is affected by the characteristics of the vehicle (engine rotational speed, engine fitted with shock-absorbers), terrain characteristics (surface obstacles), methods of wood processing (processing of

306

trees), seat characteristics, physical condition and sitting position of the operator and soil characteristics (dry, frozen, wet soil). The same authors conclude that the air pressure in tires has a considerable impact on the level of vibrations that are transmitted through the seat to the whole body of the operator while the harvester is moving on uneven terrain. Kumar (2004) concludes that the vehicle speed and the type of terrain on which it moves highly affect the level of vibrations that are transmitted through the seat to the whole body of the operator. The Directive 2002/44/EC defines the minimum health and protection requirements for the workers exposed to vibrations that are transmitted to the handarm system (HAV – Hand-Arm Vibrations) and to the whole body of the operator (WBV – Whole Body Vibrations). According to this Directive, the upper limit of the worker’s daily exposure to hand-arm vibrations is 5 m/s2, while the daily warning value is 2.5 m/s2. The worker’s exposure through hand-arm system is measured in accordance with the specifications described in the standard ISO 5349-2:2002. According to the Directive 2002/44/EC, the upper limit of the worker’s daily exposure to vibrations that are transmitted to the whole body is 1.15 m/s2, while the daily warning value is 0.5 m/s2 (8-hour working time). Scarlett et al. (2002) agree that these limit values could be exceeded with a large number of modern farming tractors. Workers’ exposure to vibrations that are transmitted to the whole body is measured by the method specified by the standard ISO 2631-1:1997.

2. Material and Methods – Materijal i metode Engine rotational speed obtained by the use of the FMS was divided into classes of 100 min–1, ranging between 800 and 2 100 min–1. As in the FMS report the engine rotational speed is shown in the dependence of time, the duration of engine operation at specific engine rotational speed was determined by further analysis. Subsequent measurement of vibrations at the same engine rotational speed classes was performed with the help of vibrometer Brüel & Kjaer, type 4447 and triaxial accelerometer, type 4520-002, on the steering wheel and triaxial accelerometer, type 4524-B, fitted in the rubber protective cover on the seat and seatback of the researched skidder. The measurements were carried out on the skidder Ecotrac 120V whose mass is approximately 7.5 tons. The researched skidder is powered by a 6-cylinder airCroat. j. for. eng. 34(2013)2

Possibility of Determination of Daily Exposure to Vibration of Skidder Drivers ... (305–310)

Fig. 1 Measuring of vibrations on the steering wheel Slika 1. Mjerenje vibracija na kolu upravljača

Fig. 2 Measuring of vibrations on the seat Slika 2. Mjerenje vibracija na sjedištu

Table 1 Time consumption per classes of engine rotational speed Tablica 1. Prikaz utrošenoga vremena prema razredima brzine vrtnje pogonskoga motora Revolution per minute, rpm

Duration

Duration – Ti

Broj okretaja u minuti

Trajanje

Trajanje – Ti

min–1

hh:mm:ss

850 (800–899)