Contemporary ENERGY Vol1 No2 (2015) by REMOO

International Journal of Contemporary ENERGY Peer-reviewed open-access E-journal

ISSN 2363-6440

Vol. 1, No. 2 (2015) November 2015 www.Contemporary-ENERGY.net

Publisher

Get It Published Verlag e.k. Allee am Röthelheimpark 14 91052 Erlangen GERMANY

T 00 49 (0)9131 917 96 14 E info@get-it-published.de W www.get-it-published.de

Copyright This journal and all published articles, including all illustrations contained in authors’ papers block, are protected by copyright. Upon an article being accepted for publication, all rights of publication, for translation, further reproduction, distribution, transmission, display, broadcast, of storage in any electronic form and producing photocopies are transferred to the publisher. Without the written permission of the publisher, any usage outside the limits of the copyright act is forbidden.

© Copyright by Get It Published Verlag e.K. Notice The publisher does not assume any responsibility for any harm and/or injury to property and persons resulting from any ideas, instructions, methods or products contained in the material published in this journal, as well as a matter of inattention or creation liability, or from any use or operations.

Cover Illustration Gewerbegebiet Illustration; Source: http://de.123rf.com; Copyright: irstone

Founding Editor & Editor-in-Chief Zoran V. Stosic

–

editor@contemporary-energy.net

Director RENECON International, GERMANY; Former Vice President ICO South East Europe at AREVA, GERMANY

Editorial Board Prof. Jan Blomgren

–

Uppsala University; CEO of INBEx, SWEDEN

Ass. Prof. Leon Cizelj

–

University of Ljubljana; Head of Reactor Engineering Division at IJS, SLOVENIA

Ass. Prof. Davor Grgić

–

Faculty of Electrical Engineering and Computing, University of Zagreb, CROATIA

Prof. Nikola Čavlina

–

Faculty of Electrical Engineering and Computing, University of Zagreb, CROATIA

Dr. Ludger Mohrbach

–

Head of Competence Center Nuclear Power Plants at VGB PowerTech e.V., Essen, GERMANY

Dr. Maximilian Emanuel Elspas

–

Head of Energy Law and Lawyer Partner at Beiten Burkhardt Law Munich, GERMANY

Dr. Dietmar O. Reich

–

Co-Head of Competition Practice Group and Lawyer Partner at Beiten Burkhardt Law Brussels, BELGIUM

Dr. Miodrag Mesarović

–

Secretary General of the SerbianWEC Member Committee; Senior Advisor to Energoprojekt-ENTEL, Belgrade, SERBIA

Prof. Ana M. Lazarevska

–

Faculty of Mechanical Engineering, University of Skopje, MACEDONIA

LL.M. Ana Stanič

–

Lawyer Principal at E&A Law London, UNITED KINGDOM

Prof. Li Ran

–

School of Engineering, University of Warwick, UNITED KINGDOM; Deputy Director of China State Key Lab in Power Transmission Apparatus Security, Chongqin University, CHINA

Dr. Changxin Liu

–

Deputy Director General of China National Nuclear Corporation – CNNC, Beijing CHINA

Prof. Xu Cheng

–

Institute of Fusion and Nuclear Technology, Karlsruhe Institute of Technology – KIT, GERMANY; School of Nuclear Sciences and Engineering, Shanghai, Jiao Tong University, CHINA

Prof. Josua P. Meyer

–

Department of Mechanical and Aeronautical Engineering, University of Pretoria, SOUTH AFRICA

Prof. Zhao Yang Dong

–

Chair Professor and Head of School of Electrical and Information Engineering, University of Sidney, AUSTRALIA

M.Sci.Engng. Jukka Tapani Laaksonen

–

Vice President ROSATOM Overseas, Moscow, RUSSIA; Former Director General of the STUK, FINLAND

M.Sci.Engng. Jože Špiler

–

Head of TechnicalServices and Investments at GEN-energija, Krško, SLOVENIA

Prof. Michael Narodoslawsky

–

Institute for Process and Particle Engineering, Technical University of Graz, AUSTRIA

Dr. Raffaella Gerboni

–

Post-Doc Fellow Researcher, Energy Department, Politecnico di Torino, ITALY

Prof. Henryk Anglart

–

Deputy Head of Physics Department, KTH Royal Institute of Technology, Stockholm, SWEDEN

Dr. Suna Bolat

–

Assistant Professor, Eastern Mediterranean University – EMU, Famagusta, North Cyprus, TURKEY

Prof. Nikola Popov

–

Faculty of Engineering Physics, McMaster University, Hamilton; President DENIPO Consulting Ltd., Toronto, Ontarion, CANADA

Prof. Milovan Perić

–

Managing Director of CoMeT Continuum Mechanics Technologies GmbH, GERMANY; Senior Corporate Consultant CD-adapco, UNITED KINGDOM

Prof. Umberto Desideri

–

Department of Energy Engineering, University of Pisa, ITALY

Prof. Chul-Hwa Song

–

University of Science and Technology – UST, Seoul; Director of Thermal-Hydraulics Safety Research Div., KAERI, Daejeon, SOUTH KOREA

Prof. Shpetim Lajqi

–

Faculty of Mechanical Engineering, University of Prishtina, KOSOVO

Dr. Camila Braga Viera

–

Post-Doc Researcher, SCK-CEN Belgian Nuclear Research Center, Boeretang, BELGIUM

Ass. Prof. Manuel Ruiz de Adana Santiago

–

Department of Applied Thermodynamics, University of Cordoba, SPAIN

Ass. Prof. Roonak Daghigh

–

Department of Mechanical Engineering, University of Kurdistan, Sanandaj, IRAN

Dr. Cristina Cornaro

–

Assistant Professor, Department of Enterprise Engineering, University of Rome “Tor Vergata”, ITALY

Dr. Naseem Udin

–

Principal Lecturer, Institute Teknology Brunei, BRUNEI

Prof. Gordana Laštovička -Medin

–

Faculty of Science and Mathematics, University of Montenegro, Podgorica, MONTENEGRO

Prof. Serkan Dag

–

Department of Mechanical Engineering, Middle East Technical University – METU, Ankara, TURKEY

International Journal of Contemporary ENERGY, Vol. 2, No. 1 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

A WORD FROM THE EDITOR-IN-CHIEF Three humans born every second resulted in September that 7.3 billion live on the planet. Unfortunately, this number of new born is reducing making the population ageing fast, so that in 2050 the share of senior is expected to be more than three times than that in 1950, and the share of youth could drop down to 61%. In addition, in the last 60 years the world population lifetime has been increased by more than 20 years. Both, the lower birth rates and increasing life expectancies means reducing number of today’s workers of about 3.6 to support each social security recipient down to 2.1 in the year 2025. Therefore, more than 9 billion people could live on the planet in 2050 — and we have responsibility today for their future.

Founding Editor & Editor–In–Chief Zoran V. Stosic

In 1760s the Industrial Revolution boosted the hunger for energy somehow careless about sustainable development as well as about environment, which preservation should not be a challenge but rather must be common sense. Our energy demand has been constantly increased and will further grow. Thus, it is our imperative role to support the natural evolution of global energy mix, for which diversified fuel mix is the most important prerequisite. Not to forget that the greatest source of energy for the future is continuing to use it more efficiently. In last 20 years, global energy savings were equivalent to more than 25 years of primary energy consumption in the United States. In addition, technical possibility to meet the growing global energy demand by using only clean and sustainable energy sources and technologies that will avoid dangerous climatic change of more than 2oC above pre–industrial levels is becoming questionable. So, there is an urgent need to promote and to fund research and innovation in traditional energy and in the development of new technologies. For both, modelling and experimental achievements are the most important supporting tools, which also need appropriate promotion and funding. Take time and enjoy reading …

___________________________________________________________________________________________________________ A Word from the Editor–in–Chief

International Journal of Contemporary ENERGY, Vol. 2, No. 1 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

___________________________________________________________________________________________________________ A Word from the Editor–in–Chief

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Eugenijus Uspuras, Sigitas Rimkevicius, Algirdas Kaliatka Application of the Best–Estimate Approach for the NPP Licensing Process

Wojciech Bujalski, Krzysztof Badyda, Kacper Rosinski Analysis of Operation of Heat Accumulator in Large–Scale Combined Heat and Power Plant

Ana M. Lazarevska, Natasha Bakreska Kormushoska, Atanas Kochov Complementary and Overlapping among Energy Performance Indicators as Part of the Sustainable Development and RECP Indicators in Cement Industry

Krzysztof Badyda, Piotr Krawczyk, Szczepan Mlynarz Numerical Analysis of the Impact of Parameters of Urea Solution Injection on Reagent Penetration inside the Combustion Chamber of an OP–140 Boiler

Simon Pezzutto, Agne Toleikyte, Matteo De Felice Assessment of the Space Heating and Cooling Market in the EU28: A Comparison between EU15 and EU13 Member States

Matej Fike, Gorazd Bombek, Aleš Hribernik Visualisation of Unsteady Flow Field in an Axial Flow Fan

Elena Bebi, Jorgaq Kacani, Edmond Ismaili, Noriyuki Goto, Atushi Fujiwara Experimental Modelling of a Wind Farm in Mamaj, Albania

Marjela Qemali, Raimonda Bualoti, Marialis Celo Voltage Stability Assessment through a New Proposed Methodology

About the Journal

Instructions for Authors

e11

Authors‘ Papers

Energy, Safety and Profitability – The Thinking Behind by Jan Blomgren

The Journal

Editorial

CONTENT

Advertisements

___________________________________________________________________________________________________________ Content

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

ENERGY, SAFETY AND PROFITABILITY — THE THINKING BEHIND Energy is a notoriously difficult area to reach consensus about. I believe the ultimate reason is that energy is such a multi-faceted field. It involves technology, economy, ethics, environment, national security and a whole range of other aspects. Humans have different values, resulting in that also when we agree on the facts, we can still draw different conclusions about the way forward, simply because we give different weight to these various factors. More than so, in some cases we do not even view the challenges the same way, and what is a cornerstone in some areas can be totally neglected in other. I attended a meeting a number of years ago in which representatives of a utility presented their work to improve profitability. What they had initiated was a strict probabilistic approach to operation and maintenance, although the word probabilistic was never used. In short, this is the same type of thinking we use when we take an insurance. The definition of risk is probability times consequence; If we know the probability our house will burn down is one in a thousand per year, and the cost for building a new identical one is, say, 200 000 Euro, we multiply the probability with the consequence (0.001*200000 = 200) and realize 200 Euro per year is a reasonable insurance rate (although the fee is a bit higher, because our insurance company would like some profit too…).

___________________________________________________________________________________________________________ “Energy, Safety and Profitability – The Thinking Behind”

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

This is straight forward, but here is another question to ponder: Suppose the probability for an event is one in a million per year (most people would consider this negligible) and the cost in case it happens is infinitely high. Mathematically, this still has infinite risk, and hence the insurance cost would be infinitely high! Now you can point out that there are no costs that are infinitely high, and therefore this example is uninteresting. However, for a commercial company, if it experiences an event that costs more than the total value of the company, it goes bankrupt. For that company, whether the cost is finite but so high it wipes out the company or whether it is truly infinite is a rather uninteresting distinction – the effect is the same: extinction. Back to the business meeting above with the strictly probabilistic handling of operation and maintenance. I got a strange gut feeling this was not the right thing to do. During the coffee break, I talked with a colleague who is an expert in hydro power, whereas my background is in nuclear power. He also had the same feeling something was not right. All the others at the meeting seemed thrilled about this concept, and we felt like two grumpy old men that did not share the enthusiasm. Suddenly, we realized the reason. All the other participants came from other energy technologies than hydro and nuclear; their daily work was in coal, lignite, gas, wind and whatnot. There is a clear distinction between hydro and nuclear on one hand, and all other energy technologies on the other: the former can have accidents with so costly consequences that a single event can bankrupt the entire company, which is just not the case for the latter. A torn-down wind turbine causes losses but not immediate bankruptcy. Nuclear power has in the public domain always lived with a reputation to be able to cause catastrophes of biblical proportions; in fact, almost any person on the street strongly overestimates both the probabilities and the consequences of an accident in a nuclear power plant. Hydro power, on the other hand, is hardly perceived by anyone in the public domain to be a large threat, in spite of its potential for major disaster. A recent investigation in my home country Sweden arrived at the conclusion that the single most costly accident in the energy production sector would be a rupture of the Suorva dam. This is a mud dam far upstream in the largest Swedish river. The dam has no electricity production installed; its purpose is to collect water to be able to level out the yearly variations in precipitation. This man-made lake is on the top-ten list in volume of Swedish lakes, and it contains two full years of precipitation when full. If this dam would rupture, it wipes out all hydro power plants in the largest river of the country as well as flood two cities with in total above 100 000 inhabitants. The total number of direct casualties would probably stay below 100 persons thanks to that it takes a few hours for the flood to reach populated ___________________________________________________________________________________________________________ “Energy, Safety and Profitability – The Thinking Behind”

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

areas, allowing time for evacuation. Along the way, it would destroy the connection point for the main power line from Northern Sweden where most of the hydro power production is to Southern Sweden where 90 % of the population lives. This might remove access of up to half the electricity of the country for an extended time, and it would remove the regulation capacity which is essentially concentrated to that particular river. The total cost for such an accident is estimated to at least 50 billion Euro, which is far more than for a core meltdown in a nuclear power plant. Being a nuclear guy, I suspect I will immediately get accused of bringing this up as a way to defend nuclear power: ”look guys, the others are worse!” This is not my intention, but I want to underline two points. First, nuclear and hydro power both have potential for disasters so large they can wipe out the entire business for the operator. This means strict probabilistic thinking cannot be applied. More about this later on. My second point is that the public perception of probabilities, consequences and risk is poorly connected to the real situation. Few people, if hardly any, points out hydro power as a potential source of large accidents, whereas most, or close to all, in society at large are (over-)aware of the potential dangers of nuclear power. Let me illustrate this second point, not with potential accidents, but calamities that have actually occurred. Most people born in the 40s know about the Three Mile Island nuclear power accident 1979 in the USA. This accident lead to destruction of the core of a brand new nuclear power plant, but lead to no large releases of radioactivity (the personnel at the plant got additional doses comparable to an extra X-ray examination by the dentist, and people outside the plant far less than that), no casualties, and no animals were harmed making this core meltdown. In 1963, a landslide into the lake above the Vajont dam in Northern Italy caused a tsunami 100 m high (!) that washed away the town of Langarone, killing 2000 people. Few people born in the 40s have ever heard about this disaster, in spite of its consequences for human lives being so much more grave than the Three Mile Island accident. Back to the first point: if you have potential for a single accident that can wipe out your entire business, strict probabilistic management will not be sufficient. In these technologies, deterministic analysis is a cornerstone. Often such analyses are based on the study of so called design basis accidents, i.e., accident scenarios for which there must be ways to handle the consequences, no matter what the probability is. Design basis accidents could be anything from events that occur frequently to scenarios that have never taken place anywhere in history, and that is actually the point: if a scenario can have enormous consequences, the probability is disregarded from a design point of view. There must be a mitigation, irrespective of the probability. ___________________________________________________________________________________________________________ “Energy, Safety and Profitability – The Thinking Behind”

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

One example from my home turf, nuclear power, is core cooling in accident scenarios in old boiling water reactors. Without going into any technical detail, a nuclear power reactor is essentially a giant boiling pan, about 20 m high and 6 m diameter with 15 cm thick steel walls, in which water is boiled to steam under high pressure. It is important that water is always circulating in the reactor during operation (in fact, also when not operating) and therefore large circulation pumps are part of the system. In the first generations of boiling water reactors, these pumps were located outside the reactor, and 1 m diameter tubes were coming out of the reactor to the pump, and then back in again. If such a tube would rupture (which is called guillotine break – I leave the rest to the imagination of the reader), you would have a 1 meter diameter hole and hot water with steam of 70 atmospheres pressure above it. This steam pressure would push all water out of the reactor in just a few seconds, leaving the core without cooling. A nuclear power core, however, continues to emit heat after the normal operation has terminated due to stored heat in the fuels and because of remaining radioactivity. This heat emission cannot be stopped, and if the core is not cooled, the temperature will increase until it melts after about half an hour or so. Fukushima is an example of a core meltdown accident, although caused by other reasons and therefore taking more time. This type of tube rupture has never happened anywhere in the World in the history of nuclear power. This fact is, however, unimportant. Still, you must have a mitigation if it would. Therefore, boiling water reactors were equipped with high-capacity showers inside the reactor tank. In case water was lost in the reactor, these showers should be powerful enough to keep the core below melting temperature. Over time, it was realized that another solution was possible. The propeller of the pump was placed inside the reactor tank, connected to a motor outside the reactor via a rod. With such an external pump, there was no longer a largediameter tube that could rupture, and thereby one of the single worst scenarios was no longer relevant. As a consequence, the showers were no longer required, and instead a simpler feed-water system was deemed sufficient. In the present volume of Contemporary Energy, an example of a deterministic safety analysis for nuclear power is presented by Uspuras, Rimkevicius and Kaliatka. Although the title is focused on safety for nuclear power, it is actually of interest also for the discussion about how to arrive at acceptable economics with high safety level, irrespective of which technology we have in focus. The authors point out that in safety analysis, we traditionally use conservative estimates, i.e., we presume the worst not to underestimate the hazards. However, what is a conservative estimate for one challenge to safety could underestimate other risks. A way out of this dilemma is to also perform best-estimates studies, in which we presume the system will work in a way we believe reasonable. Comparisons of best-estimate versus conservative analyses should be able to tell you something about your safety margins and the robustness of your plant. Let us repeat: In reality, no costs are infinite, but costs large enough to wipe out your company has the same effect as an infinite cost. If realistic events can wipe out your entire business, probabilistic thinking is not the full answer. One

___________________________________________________________________________________________________________ “Energy, Safety and Profitability – The Thinking Behind”

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

solution is deterministic limits that are introduced to avoid ”infinite” consequences. Although I have pointed out hydro and nuclear power as being unique in the sense they are both capable of inducing accidents that can wipe out the balance sheet of a multi-national utility, I believe the reasoning behind deterministic analyses can be useful also in sectors that traditionally have not considered catastrophic events in their technical operation – if such analyses are applied to the business. In his famous book The Black Swan, Nassim Taleb points out that many of the major business events have come as a complete surprise to the actors in the field. Examples are taken from bank and stock market crashes, as well as from other areas of modern society. The book makes a point not of predicting the unpredictable, but of building robust systems that can handle the unforeseen. There is always a probability that an event, unforeseen in probabilistic analyses, or for that matter identified but neglected or even outright rejected as unimportant, can significantly change your business. We have seen a number of examples in recent years of companies coming from essentially nowhere, or at least from a very unexpected starting position, that have quickly grown to global market leaders. In some cases, the opposite has happened; the market leader quickly loses its top position, and in some cases become extinct in just a few years because of a disruptive change in the market conditions and poor ability to adapt to the new situation. The Finnish mobile phone company Nokia can serve as an example of both processes. Who could have predicted some decades ago that a century-old company producing car tires and wellington boots should capture half the World market on mobile phones? At the peak of its success, Nokia prospered thanks to its design and user-friendliness that attracted in particular the young generation that turned out to be a very important customer segment, neglected by many of the competitors that focused on technical excellence with elderly men as perceived core customers. Then things changed rapidly, with the advent of phones that actually were hand-held computers. Nokia embraced this technology leap a little too late, and the company quickly fell from its top position to the minor league, and was finally sold for a minute fraction of what its stock market value had been just a few years earlier. So what has black swans and mobile phones to do with energy? Traditionally, energy has been a very slow-moving beast. This has made sense due to its nature of long-term investment. A hydro power plant can be expected to operate at least a century and maybe two or more – future has to tell. Coal-fired and nuclear plants have technical life spans of half a century or more, and so on. In the last few years, we save seen much more turmoil on the European energy scene than previously, to quite some extent caused by political interventions. Large-scale support programs for introduction of new production technology ___________________________________________________________________________________________________________ “Energy, Safety and Profitability – The Thinking Behind”

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

(read: wind and solar) have undermined the profitability of the base-load production in continental Europe. Germany has taken decisions to phase out nuclear power, whereas UK have initiated a support program to build new nuclear power, so the trends are by no means uniform. Belgium has for a long time lived under a decision not to build new nuclear power and to limit the number of operational years for the existing fleet. This has in essence constituted a phase-out scenario for nuclear power, presumingly motivating investments in other technologies awaiting the coming closure of the nuclear power plants providing more than half the electricity of the country. When being close to the final countdown, the government – suddenly, it seems to me – realized the risk of losing a significant fraction of the electricity production, and forced a rapid permission to extend the life of two reactor units by ten years, in a process whose legal correctness can be questioned. For obvious reasons, this was appreciated by the owners of nuclear power plants, whereas the investors in wind farms saw the value of their investment not living up to the expectations of increased prices thanks to anticipated but now delayed mothballing of competing power production. My conclusion is that the calm days when the rule book for the energy business changed slowly are over. We can expect similar rapid alterations of the conditions, significant changes in a short time. This leads us back to our first observation: if there are potential events that can wipe out your entire business, probabilistic thinking is not sufficient. Hydro and nuclear power has lived with that risk for a long time, because it is inherently built into their technologies. I suspect, however, that all energy technologies face this risk from now on, but for non-technological reasons. Changes in politics, customer preferences, climate or even factors we still have not identified can lead to disruptions capable of eradicating major companies. The contributions in this issue of the International Journal of Contemporary Energy are all presentations from the REMOO conference September 23-24, 2015 in Budva, Montenegro. This conference had the title Technological, Modelling and Experimental Achievements in Energy Generation Systems, aiming at providing a joint platform for meetings over technological boundaries. Such inter-disciplinary learning can hopefully better prepare us for handling the disruptive events we can expect, but have a hard time predicting what will happen, when they will happen and what they will lead to. A wide variety of perspectives were considered. Bujalski and Rosinski presents a study on optimization of the use of heat accumulation in a large combined heat and electric power production plant. In short, the demand for electricity is higher daytime and lower in the night, whereas the heat demand is the opposite. Using heat accumulators has the potential to make even more efficient co-generation, increasing the resilience of the production. Lazarevska, Bakrevska-Kormushoska and Kochov have developed key performance indicators for energy efficiency applied in the cement industry. The cement industry is a large net consumer of energy, and thereby improved efficiency – in real terms less energy used per unit cement produced – has a clear environmental aspect associated with it. Properly applied, such analyses can motivate improvements reducing the business risk due to environmental impact.

___________________________________________________________________________________________________________ “Energy, Safety and Profitability – The Thinking Behind”

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Badyda, Krawczyk and Mlynarz present work with a similar ultimate motivation – less environmental impact by increased efficiency – in a paper on improved combustion technology. In particular, the potential for reduction of NOX emissions is studied in depth. It takes energy to heat a building, but even more to cool it. Pezzutto, Toleikyte and de Felice have studied the European markets for both heating and cooling. They have found that statistical data on heating seem realiable, whereas statistics on cooling might be off by as much as 60 %. Moreover, they claim the heating market is more or less saturated, but find that there is large potential for growth of cooling. Fike, Bombek and Hribernik have simulated unsteady flow fields, with potential to improve efficiency of fans. Not directly linked to this, but with some connection points, we find the contribution by Bebi, Kacani, Ismaili, Goto and Fujiwara. They present simulations to optimize the construction of a planned wind farm, in particular on how to minimize interferences between the turbines. The introduction of varying electricity production into the European grid has been accompanied with concerns this might jeopardize grid stability. Qemali, Bualoti and Celo have analysed grid voltage stability problems in Albania – although the reason for the problem is over-demand compared to the grid capacity, or alternatively expressed, undercapacity of the grid compared to the demand. They present a novel method to analyse the stability of the grid, better suited for the present and coming situation with varying production, resistance to installation of new high-voltage lines and market deregulation. This sums up the contributions in this issue of the International Journal of Contemporary Energy, in which the papers were all presented at the REMOO conference September 23-24, 2015 in Budva, Montenegro. You can look forward to most interesting reading.

(All images reprinted with permission from Vattenfall, E.ON, Fortum and Freeimages) Jan Blomgren Associate Editor

Jan Blomgren is CEO and founder of INBEx (Institute of Nuclear Business Excellence), providing independent nuclear executive advice and business leadership training globally. The INBEx team comprises over 20 former CEOs, Director Generals and similar. He was the youngest professor ever in Sweden in nuclear physics, holding the chair in applied nuclear physics at Uppsala University. His research was focused on neutron-induced nuclear reactions, an area in which he has published over 200 papers in refereed international journals and conference proceedings. When plans to build new nuclear power in Sweden were initiated, he was recruited to Vattenfall, one of the largest nuclear power operators in Europe. At Vattenfall, he was responsible for planning the competence development needed for nuclear new-build, as well as coordinating training for nuclear power plant personnel. In addition, he was Director of the Swedish Nuclear Technology Centre, which is the coordination organization for nuclear research and education involving universities, industry and the regulator. He was involved in the creation of ENEN, the European Nuclear Education Network, in which essentially all European universities in nuclear engineering collaborate. Moreover, he has recently established a large collaboration with France on research and education. Finally, he is the father of several industry-sponsored university programs, as well as having started a number of nuclear business training programs in industry. Jan Blomgren is alone in Europe to have upheld high-ranked positions both at university and nuclear industry. He is frequently invited speaker at conferences, and was recently invited as expert advisor to the French Senate.

___________________________________________________________________________________________________________ “Energy, Safety and Profitability – The Thinking Behind”

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Authors’ Papers

Eugenijus Uspuras, Sigitas Rimkevicius, Algirdas Kaliatka

Application of the Best–Estimate Approach for the NPP Licensing Process 13

Wojciech Bujalski, Krzysztof Badyda, Kacper Rosinski

Analysis of Operation of Heat Accumulator in Large–Scale Combined Heat and Power Plant 20

Ana M. Lazarevska, Natasha Bakreska Kormushoska, Atanas Kochov

Complementary and Overlapping among Energy Performance Indicators as Part of the Sustainable Development and RECP Indicators in Cement Industry 27

Krzysztof Badyda, Piotr Krawczyk, Szczepan Mlynarz

Numerical Analysis of the Impact of Parameters of Urea Solution Injection on Reagent Penetration inside the Combustion Chamber of an OP–140 Boiler 35

Simon Pezzutto, Agne Toleikyte, Matteo De Felice

Assessment of the Space Heating and Cooling Market in the EU28: A Comparison between EU15 and EU13 Member States 49

Matej Fike, Gorazd Bombek, Aleš Hribernik

Visualisation of Unsteady Flow Field in an Axial Flow Fan 57

Elena Bebi, Jorgaq Kacani, Edmond Ismaili, Noriyuki Goto, Atushi Fujiwara

Experimental Modelling of a Wind Farm in Mamaj, Albania 65

Marjela Qemali, Raimonda Bualoti, Marialis Celo

Voltage Stability Assessment through a New Proposed Methodology

___________________________________________________________________________________________________________

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

DOI: 10.14621/ce.20150201

Application of the Best–Estimate Approach for the NPP Licensing Process Eugenijus Uspuras, Sigitas Rimkevicius, Algirdas Kaliatka Lithuanian Energy Institute Breslaujos 3, LT-44403 Kaunas, Lithuania, algirdas.kaliatka@lei.lt

Abstract

1. Introduction

According the international practice the best-estimate approach in safety analysis of Nuclear Power Plants (NPPs) is used mainly for the Loss of Coolant Accident (LOCA) in reactor cooling circuit type accidents. In Lithuania the best estimate approach was successfully applied not only for LOCA but also for reactor transients, reactivity initiated accidents and accident confinement system response analyses. This paper presents the four examples of best-estimate accident analysis developed for Ignalina NPP licensing, covering the LOCA, transients, reactivity initiated accidents and the accidents in the confinement system. In the paper the so called “partiallyconservative” approach, also used for the Ignalina NPP licensing, is introduced. The comparison of both approaches are performed and the recommendations for employment of approaches presented.

Keywords:

Best-estimate approach; Partiallyconservative approach; RIA; flow blockage; LOCA

Article history:

Received: 29 July 2015 Revised: 25 September 2015 Accepted: 09 November 2015

Deterministic safety analyses present the most part in the scope of works performed for nuclear power plants licensing activities. This analysis is performed through the calculation of plant parameters (responses) with complex computer codes, solving a set of mathematical equations describing a physical model of the plant. Historically, a conservative approach has been taken for licensing analysis, including making conservative assumptions on plant data, system performance and system availability. The conservative approach means that use of conservative codes is combined with conservative boundary and initial plant conditions. This approach gives the results deliberately biased in a pessimistic manner. Now almost in the all countries the best-estimate approach is widely used in order to avoid the unnecessary conservatisms and to properly assess and to address the existing safety margins. In this approach the use of best-estimate codes is combined with realistic boundary and initial plant conditions. Together with use of best-estimate code the uncertainty analysis is required. In Lithuania the best-estimate approach was successfully applied in licensing practices of Lithuanians only nuclear power plant – Ignalina NPP starting 2003. Two units of RBMK-1500 reactors were built in Lithuania, in Ignalina NPP. RBMK (Russian abbreviation for: “Large-power channel-type reactor”) belongs to a class of graphite-moderated nuclear power reactors and were designed in the Soviet Union in 1970s. At present both reactors at Ignalina NPP are shutdown (first unit was shutdown in 2004, second – in 2009). Recently there are no plans to build new RBMK type reactors, however there are 11 RBMK reactors operating in Russia (4 reactors in Saint Petersburg, 3 – in Smolensk and 4 – in Kursk). The main findings of this paper may be applied for the RBMK-1000 reactors, operating in Russia (because of the similar design of the reactors).

___________________________________________________________________________________________________________ E. Uspuras, S. Rimkevicius, A. Kaliatka: “Application of the Best-Estimate Approach for the NPP Licensing Process”, pp. 1–12

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

In this paper are presented four examples of bestestimate accident analysis performed for Ignalina NPP with RBMK-1500 reactors licensing:

The graphite structure is contained in a steel vessel, which is called reactor cavity.

• Limiting Reactivity Initiated Accident (RIA) for RBMK-1500;

2.1. Simulation of power distribution in reactor core and processes in fuel matrix

• Break of fuel channels inside reactor cavity; • Large LOCA case in main circulation circuit (break of main circulation pump pressure header); • Blocking of coolant flow rate in group of fuel channels (blocking of coolant flow rate in group distribution header). These selected bounding cases with its consequences covers all possible RIAs, transients, LOCA type accidents and the accidents in reactor cavity. The RBMK-type reactors do not have the continuous containment, which covers all reactor and cooling circuit. There are Accident Localisation System (ALS), which covers the part of reactor cooling system equipment (main circulation pumps, headers and water supply pipelines) and the Reactor Cavity (RC), which enclose the graphite stack with fuel channel. Both these leak tight, complex geometry systems plays a role of containment and are the last barrier preventing the radioactivity release into atmosphere. For the analysis the different best estimate computer codes were used: • QUABOX/CUBBOX-HYCA code [1], [2] – for the calculation of power distribution in fuel assembly; • FEMAXI-6 code [3] – for the analysis of processes in fuel matrix of RBMK-1500; • RELAP5 Mod3.2 [4], [5] – for the analysis of reactor cooling circuit response; • CONTAIN [5] – for the analysis of reactor cavity response.

For the modelling of processes in RBMK –1500 reactor core and the analyses of reactivity initiated accidents the QUABOX/CUBBOX-HYCA (Q/C-H) code was used in Lithuanian Energy Institute (LEI). The core model Q/C-H was originally developed by GRS for core calculations of Light Water Reactors (LWRs) [1], [2]. Since 1990 the code was adapted to the features of RBMK-1000 reactors and since 1995 additionally adapted to account for the special requirements of RBMK-1500 reactors. During this time the code was continuously used for RBMK-1500 core calculations and for the surveillance of the reactivity behaviour of the changing core loading using higher enriched uranium-erbium mixed fuel and Control Rods (CR) of new design [10], [11]. The code was used for audit calculations during the review of the Ignalina NPP Safety Analysis Report (SAR) for Unit 1 and for the preparation of the SAR for Unit 2. Later on, the Q/C-H code was applied for the independent assessment of the shutdown systems modification at Unit 2 [10], [11]. Design of fuel pellets and separate fuel rods for RBMK reactor differs very little from fuel rods manufactured for standard BWR-type reactors [9]. In RBMK reactor the fuel assembly is fit into a circular fuel channel with inside diameter of 80 mm and an active core height of 7 m. In order to achieve the required height, the RBMK fuel assembly consists of two fuel bundles placed one above another. Each fuel bundle includes 18 fuel rods placed in two circles around the carrying rod. For the modelling of processes in the fuel rod of RBMK-1500 the FEMAXI-6 code [3] is used in LEI. FEMAXI-6 code can analyse the integral behaviour of the whole fuel rod throughout its

2. Models for the simulation of processes in RBMK-1500 The RBMK reactor is a channel-type graphite-moderated boiling water reactor [9]. It has a huge graphite block structure as the moderator that slows down the neutrons produced by fission. The feature of RBMK type reactor is that each fuel assembly is positioned in its own vertical Fuel Channel (FC). The water is supplied to fuel channels bottom, where it is heated to saturation and partially evaporates. The steam produced passes to the steam separators, which separates water from the steam. The fuel channels are made of Zirconium and Niobium alloy similar to that used for the fuel claddings.

Figure 1: Model of RBMK-1500 fuel rod (bottom bundle), developed using FEMAXI-6 code

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

life as well as the localized behaviour of a small part of fuel rod (temperature distribution in the fuel rod, thermally induced deformation of fuel pellet and cladding, fission product gas release). This code, designed for vessel type reactors, was adapted for modelling of processes in fuel rods of RBMK-1500 reactor, introducing the material properties for the RBMK fuel and cladding, [12], [13]. The RBMK-1500 fuel rod model, developed using FEMAXI-6 code, is presented in Figure 1. The single fuel rod from bottom fuel bundle in fuel assembly was modelled, because the generation of power peak is met in the position 1.07 m from the bottom of the core [9].

2.2. Simulation of reactor cooling circuit response Reactor Cooling System (RCS) of RBMK has two loops, which are interconnected via the steam lines and do not

have a connection on the water part. The best estimate system thermal-hydraulic code RELAP5 [4] has been adapted to model the RBMK type reactors and used since 1993 at the Lithuanian Energy Institute [22], [23]. The RELAP5 nodalization scheme, which was used for the modelling of processes in RCS, is presented in Figure 2. The left loop of RCS model consists of one equivalent core pass. Two Drum Separators (DS) are modelled as one “branch” type element (1). All down comers are represented by a single equivalent pipe (2), further subdivided into a number of control volumes. The Main Circulation Pumps (MCPs) suction header (3) and the pump pressure header (8) are represented as branch objects. Three operating MCPs are represented by one equivalent element (5) with check and throttlingregulating valves. The stand-by MCP is not modelled. All 830 fuel channels of this left core pass are represented by an equivalent channel (12) operating at average power and coolant flow. In the right circulation loop, the

Figure 2: RBMK-1500 model nodalization scheme: 1 - DS, 2 – down comers, 3 - MCP suction header, 4 - MCP suction piping, 5 - MCPs, 6 - MCP discharge piping, 7 - bypass pipes, 8 - MCP pressure header, 9 - GDHs, 10 - lower water pipes, 11 - reactor core inlet piping, 12 - reactor core piping, 13 - reactor core outlet piping, 14 - steam-water pipes, 15 - steam pipes, 16 - check valve, 17 - ruptured pressure header, 18 - valve for break modelling, 19 – model of compartments which surround the RCS pipelines ___________________________________________________________________________________________________________ E. Uspuras, S. Rimkevicius, A. Kaliatka: “Application of the Best-Estimate Approach for the NPP Licensing Process”, pp. 1–12

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

#17

Water

Node #16 Vacuum breaker

#21 #19 #18

Steam from MSV and SDV-A

#23 #22

Water #5 #4 #6

#24

Water #3

Node #1

Node #2

Node #9

Node #10

#12

#27 #26

#19 #18

#21

MSD

Node #29 Node #9 MSD

Node #25

#20

Node #33 Node #34

Node #15

Node #6

Node #7 Vacuum breaker

Node #6

Node #32

#14 #13

Water

#28

Blow-out panels

Node #20

Water

#15 Node #31

Node #11

Water

Node #30

Water Node #35

Figure 3: The nodalization scheme of the ALS (a) and RCVS (b) model for the code CONTAIN

MCP system model consists of two equivalent core passes. One core pass represents one Group Distribution Header (GDH). Fuel channels from this GDH are represented by few equivalent channels of different power levels. For the core power of 4200 MWth, the channel average power is assumed to be 2.53 MWth, the maximum channel power is 3.75 MWth and minimum channel power is 0.88 MWth. The other core pass represents the other 19 GDHs. The channels of this pass are simulated by an equivalent FC of average power. The steam separated in the separators is directed to turbines via steam pipes (15). Two turbine control valves divert steam supply to the turbines. The guillotine break of MCP pressure header (17) in the right loop model of RCS is modelled by a valve (18). The flow area of this valve is double of pressure header flow area. The valve (18) is connected to the volume (19), which represents the compartments covering RCS pipelines. The detailed description of this model is presented in papers [8], [14].

2.3. Simulation of reactor cavity response The reactor and a large part of the reactor cooling circuit of RBMK-1500 are enclosed within the accident localization system, which consists of a number of interconnected compartments with 10 condensing pools to condensate the steam, discharged during the accident [9]. In this respect, the ALS may be called a pressure suppression type confinement. The reactor cavity in RBMK-type reactors is the compartment surrounding reactor core (fuel channels inside a graphite stack). The RC is protected against overpressure by the reactor cavity venting system, which directs the released coolant from the broken fuel channels to ALS. The RC and ALS are very important systems of RBMK reactors as they perform the containment function, i.e. form the last barrier preventing radioactive material release to the environment. The nodalization scheme of ALS and

Reactor Cavity Venting System (RCVS) is presented in Figure 3 and general description of ALS and RC – in [15].

The model, developed by employing CONTAIN code [5], consists of 35 nodes and 178 structures for heat exchange modelling. Nodes 32 and 33 simulate RCVS pipelines from the upper part of RC. The four lower condensing trays of each ALS tower were connected to separate nodes 5 and 14. In the case of fuel channels rupture, the coolant from the RC is directed only to a half of the 5th condensing tray of left ALS tower. This is why the 5th condensing tray of the left tower of ALS is divided into two parts and simulated as two pools (Figure 3). The steam to the pool 5th in node 21 is provided from the RC through 2 steam distribution devices (node 20) as well as from the top steam reception chamber through 8 steam distribution devices (node 19). The steam to the pool in node 27 is provided from the top steam reception chamber through 10 steam distribution devices (node 26). These two pools are interconnected at water part by the pipe of 100 mm diameter. The pools located in nodes 6 and 15 represent the water in hot condensate chambers. The initial level in these pools assumed to be 2.5 m (according to the design). The overflow of water from Hot Condensate Chamber (HCC) to nodes 3 and 12 (representing bottom steam reception chamber) is considered in the model. The air release from the towers of ALS (from nodes 8 and 17) to the environment is simulated employing special junctions that close in 5 min after the accident starts. The blow-out panels (nodes 7 and 16), which open if excess pressure in gas delay chambers (nodes 6 – 8 and 15 – 17) increases to 80 kPa, are installed in compartments in both towers of ALS. An assumption is made in the model that six Membrane Safety Devices (MSD) (nodes 9 and 33) open when the excess pressure in RCVS pipelines before MSD increases to 60 ± 10 kPa (i.e. in accordance with MSD design data). The drainage

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

3. Results of the Best-Estimate Analysis 3.1. Analysis of the limiting Reactivity Initiated Accident Modelling of reactivity-initiated accidents involves the simultaneous solution of equations for neutron transport, heat transport within the fuel rods and across the clad-to-coolant interface, mechanical behaviour of fuel and cladding, and coolant thermal-hydraulics. These equations are strongly interconnected and dependent on both space and time. Since it was no possibilities to solve it in full detail in core-wide analyses, the separate codes and simplifications were used: the power distribution in fuel assembly was calculated using Q/C-H code, while the processes in fuel matrix of RBMK-1500 were evaluated using FEMAXI-6 code. All possible RIAs were analysed during SAR preparation: (1) single CR withdrawal; (2) spurious withdrawal of CRs group; (3) drop out of CR with shorted absorber; (4) erroneous refuelling; (5) loss of water in the channels, where control rods are placed [19]. The accidents related to spurious withdrawal of single Control Rod (CR) were selected for more detailed investigation, because these events lead to more significant change of the core parameters. For the maximum permissible thermal reactor power (4200 МW), the withdrawal of the rod in the centre of core leads to high peak of linear power for fuel rod. In the base calculation the peak value of 530 W/cm was reached, that exceeds the acceptance criterion (485 W/cm) [17]. For this case uncertainty and sensitivity analysis was performed using GRS methodology. For the evaluating possible uncertainties of the calculation results a total number of 16 uncertain input parameters were taken into account defining model options (heat transfer parameters in fuel matrix, radial zones in fuel pin, etc.) and physical boundary conditions (CRs insertion depth, reactor power, etc.). The analysis of the calculation results allows identify the main contributors to the uncertainty of investigated output parameters as well as to determine their twosided (95%, 95%) statistical tolerance intervals. Based on gained results it was confirmed that during the single CR withdrawal accident the maximal temperatures of fuel pellet and cladding were 1865 oC and 375 oC respectively and did not exceed the safety limits. The peak of linear power in the maximum loaded fuel channel reaches 589 W/cm (Figure 4). The linear power exceeds the acceptance criterion (485 W/cm) in 25 fuel channels,

located around the failed control rod. The violation of safety limit continues ~20 s. The power in maximal loaded FC exceeds it’s the highest values in 14th second of the accident. The power peak during transient always was located in the bottom part of the core. The maximum bounding linear power from Figure 4 was assumed as power behaviour in selected fuel assembly for the analysis of process in fuel rod. The FEMAXI model of RBMK-1500 fuel rod, presented in Figure 1 was used for the simulation. The temperatures of fuel pellet and fuel cladding temperatures start to increase after beginning of withdrawal of the failed CR, due to increase of heat generation in the core. Later, after finishing of CR movement (8th sec) temperatures start to decrease. The peak temperature in the fuel pellet centre is 2270 oC and on the inner cladding surface 400 oC. These temperatures are below acceptance criteria (2800 oC and 700 oC accordingly) [17]. The fuel cladding outer surface temperature remains constant, because the condition for boiling crisis is not reached in the fuel channel. The intensification of releases of gases (Kr and particularly Xe) from fuel into gap between pellets and cladding is insignificant during this accident, thus, the increase of pressure of gases inside fuel rod is caused due to increase of pellets temperature mainly. After decreasing of neutronic linear power in the fuel rod, the pressure in the gap between fuel pellets and cladding decreases down to the initial value. Because the fuel burnup is low, the fuel grain growth, cracking of pellet, formation of fission gas bubble in fuel and following swelling of fuel pellets do not occur. The outer diameter of fuel pellets is increasing only due to radial displacement. The radial displacement of fuel cladding is insignificant, because temperature of fuel cladding remains approximately constant during the all investigated period of time. The results of FEMAFI calculation - the change of the gap between fuel pellets and cladding is presented in Figure 5. As can see in time interval 5 - 20 s the gap between fuel pellets and cladding disappears in segments No. 2 – 9 (at ~1-3 m from fuel rod bottom). The gap closure

in max. loaded FC

589 W/cm

600

Linear power (th.), W/cm

water removal from RC as well as from the top and bottom RCVS pipelines is taken into account. The arrows on the nodalization scheme in Figure 3 indicate the places where the structural leaks are simulated (break of fuel channels inside the reactor cavity).

limit

500

400

300 0

Time, s

Figure 4: Distributions of linear power (th) in maximum loaded FC (uncertainty analysis results)

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

exceed the acceptance criterion (485 W/cm) for short term, but the fuel rods remain intact.

1-th segm. 2-th segm. 4-th segm. 9-th segm. 10-th segm.

80 60

Gap, μm

3.2. Analysis of the multiple fuel channels rupture inside reactor cavity

40 20 0 0

Time, s

Figure 5: Variation of fuel-clad gap during limiting RIA

Equivalent stresses, MPa

120

0 0

Time, s

Figure 6: Change of equivalent stresses in cladding during limiting RIA (at 1.07 m from core bottom)

appears because the radial displacement of fuel cladding is much slower than displacement of fuel pellets and fuel pellet comes in contact with fuel cladding. When the pellet reaches the cladding, the fuel pellet-clad interaction phenomenon is met. However, after reactor shutdown the radial dimensions of pellets and cladding returns back to the values close to the initial conditions. The equivalent stresses of fuel cladding are presented in Figure 6, which shows that the highest equivalent stresses are in the range of 110 MPa. Peak of stress is related to fuel pellet-clad interface. According to [20], the yield stress for Zr + 1 % Nb alloy is 180 - 220 MPa for 300 °C temperature and 320 – 380 MPa for 20 °C temperature. After exceeding the yield stresses limit, the fuel cladding will be affected by plastic deformation that leads to cladding failure. In the analysed case the calculated maximal value of equivalent stress in the cladding is much lower than the yield stress. Thus, the fuel cladding remains intact. Summarizing, it could be concluded that in case of the investigated limiting reactivity initiated accident in RBMK-1500, evaluating possible uncertainties in the modelling of neutron dynamic processes in reactor core, the peak of thermal linear power in fuel assembly could

During the analysis of the multiple FCs rupture in the reactor cavity it was assumed that the reactor is operating on the maximum permissible thermal power of 4200 MWth. In this case the steam-water mixture from the ruptured FCs is released to the reactor cavity. The energy flow (mass flow and enthalpy) from the simultaneously ruptured 9, 11 and 16 FCs were calculated using RELAP5 model of reactor cooling circuit (see Figure 2). Further these calculation results were applied as the boundary data for the analysis of thermal hydraulic parameters behaviour in RC and ALS employing CONTAIN model (see Figure 3). According to the performed modification in reactor cavity, the RCVS must withstand 9 fuel channels rupture [15], [17], [21]. As it is shown in [15], the excess pressure 214 kPa inside reactor cavity is assumed as an acceptance criterion because it is the smallest allowed load on the reactor cavity structures. The pressure increase (depending of number of ruptured fuel channels) beyond this value can lead to the failure of RC, loss of core cooling and large radioactivity release into the atmosphere. Thus, the pressure inside RC was selected as the analysis output result for the uncertainty analysis. The uncertainty and sensitivity analysis was performed using the GRS methodology [6] together with the developed package SUSA 3.2 [7]. The parameters which may influence the uncertainty of calculation results were compiled into two main groups: initial conditions of discharged flow (pressure, temperature and mass of water) and CONTAIN code models, assumptions and correlations. The first group of parameters, which could have the impact on the simulation results are the following: (1) heat transfer coefficient in Condenser Tray Cooling System (CTCS) heat exchangers; (2) service water temperature for CTCS heat exchangers cooling; (3) water temperature in the 5th condensing tray; (4) water temperature in the 1 to 4 condensing trays; (5) water temperature in the HCC; (6) gas temperature in RCVS pipelines; (7) air temperature in ALS leak-tight compartments; (8) air temperature in ALS tower compartments; (9) heat transfer area in the graphite stack; (10) heat transfer area of FC; (11) MSD opening pressure; (12) volume of the accident zone; (13) the connection area between the volume inside a graphite stack and the bottom (top) part of RC; (14) water level in the lower (1 to 4) condensing trays in the left ALS tower; (15) water level in the lower (1 to 4) condensing trays in the right ALS tower; (16) water level in HCC; (17) water level in the 5th condensing tray in the left ALS tower. The second group of parameters

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

The performed sensitivity analysis demonstrated that the heat transfer area of metal, model of water deposition from the atmosphere and model of water interaction with hot surfaces located in RC have the largest influence on the pressure in RC. The uncertainty analysis showed that in the case of 11 FCs rupture the limiting pressure was not exceeded even in the most unfavourable combinations of the initial conditions and modelling parameters (see the top curve in Figure 7 b). The extrapolation of the bottom curves of the calculated pressures for the rupture of 9, 11 and 16 FCs (Figure 7) shows that maximal number of ruptured FCs, when limiting pressure is not exceeded, is 19 FCs. Thus, summarizing the results of the uncertainty analysis, it is possible to conclude that the capacity of RCVS comprises from 11 up to 19 FCs, i.e. 15±4 FCs.

3.3. Analysis of large LOCA in RBMK-1500 Considering the consequences of all LOCA type accidents and taking into account the rupture size and peak temperature of the fuel cladding [16], it was found that the worst consequences for RBMK-1500 are in the case of MCP pressure header (the largest diameter pipeline) break with a failure to close the check valve of one GDH at maximum permissible reactor power level of 4200 MWth. The analysis of such large LOCA was

Limit pres sure

300 Press ure, k Pa

250 200 150

9 FC rupture

100 50

350 0

10 Time, s

300 Pressure, kPa

The uncertainty analysis with one-sided tolerance limits (59 runs) was performed for each of the selected cases 9, 11 and 16 FCs ruptures. The calculated pressures in RC in the case of multiple fuel channel ruptures are presented in Figure 7. Figure 7 a) and b) shows that the maximum pressures of all calculated variants for the rupture of 9 and 11 FCs do not exceed the criterion of RC integrity. It is also shown in Figure 7 that calculated curves of pressure in RC form three groups. The maximum pressures were received assuming, that all the liquid fraction of discharged through the ruptured FCs coolant remains in RC in a dispersed condition and interact with hot surfaces of steal and graphite in RC. This case is not expected in reality. The bottom group of curves represents pressures calculated considering the models for water deposition from the atmosphere, but without consideration of direct interaction of this water with hot surfaces in RC. The results of the middle group of curves (Figure 7 b), which represents the greatest part of the calculated cases, reflect the most probable behaviour of the pressure in RC.

350

Limit pressure

250 200 150 100

11 FC rupture

50 450 0

400

10 Time, s

9 FC rupture

350 Pressure, kPa

(CONTAIN code parameters) contains the following: (1) model of water deposition in the reactor cavity (i.e. in one case it is assumed that water discharged in the reactor cavity filled by helium-nitrogen mixture is deposited to the sump, and in the other case that water is transported between the nodes by the gases flow); (2) interaction of water with hot surfaces located in RC.

Limit pressure

300 250 200 150 100

16 FC rupture

50 0 0

Time, s

Figure 7: Calculated behaviour of pressure in the reactor cavity (uncertainty analysis results obtained using SUSA generated runs from CONTAIN calculations): a) 9 FCs rupture case, b) 11 FCs rupture case, c) 16 FCs rupture case

performed according to the “best-estimate” approach. Thermal hydraulic analysis was performed by employing the best estimate thermal-hydraulic RELAP5 Mod3.2 code Ignalina NPP model (see Figure 2). In order to identify the main contributors to the uncertainty of the calculation results, the GRS methodology [6] together with the developed package SUSA 3.2 was used [7]. More details about the application of uncertainty and sensitivity analysis methodology for the best-estimate analysis of RBMK are presented in [8]. The parameters, which may impact the uncertainty of calculation results, can be divided into two main groups: •

Initial and boundary conditions with values that may be impacted by measurement errors;

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________ 800

• RELAP5 code models, assumptions and correlations.

Straight after the MCP pressure header break, the water from MCP piping and DSs is discharged through the break (18, see Figure 2). The reactor emergency scram system is activated due to the pressure increase in the compartments where the coolant is discharged. GDH check valves (16), which are located downstream to the break, close and prevent the loss of coolant from the fuel channels of the affected RCS loop. The coolant flow stops in the affected RCS loop. However, within approximately two seconds cold water is supplied from the Emergency Core Cooling System (ECCS) fast acting subsystem to these channels. This enables supplying reliable cooling of these channels. The coolant flow in the FCs, connected to the GDH with the check valve which failed to close (16) becomes stagnant for a moment and later changes its flow direction. Due to the coolant flows behaviour, the fuel cladding temperature (in channels connected to GDH with the check valve which failed to close) increases more than in the other channels of the affected RCS loop within the first seconds of the accident. When the saturated water from DS reaches the overheated fuel assemblies, fuel cladding temperature drops down. At the beginning of the accident, channels connected to the GDH with the check valve which failed to close are cooled by saturated water flow; however, later (after DS gets empty) they are cooled only by saturated steam. It should be noted that the first fuel cladding temperature increase asserts only at the very beginning of the accident and takes a very short time, i.e. no more than 15 seconds. Later, all fuel assemblies are reliably cooled by water, supplied from

700 o

Temperature, C

650 600 550 500 450 400 350 300 250 -10

-5

10 Time, s

Figure 8: Fuel cladding temperatures in 3.0 MW power FC at the location of 2.75 m from the core bottom obtained using SUSA generated runs from RELAP5 calculations 16 14

Number of channels

The six following plant parameters, which may have the greatest impact on the simulation results due to knowledge from earlier performed benchmarking calculations, are selected for the analysis: (1) pressure in the DS; (2) coolant flow rate through the MCPs; (3) feed water temperature; (4) amount of steam for in-house needs; (5) reactor thermal power; (6) delay time for reactor scram initiation. The deviation values for these parameters are known from the Ignalina NPP documentation: they vary from 1.5 to 2%. Additionally, the 8 RELAP5 code parameters and models, such as water packing scheme, vertical stratification model, counter current flow limit model, thermal front tracking model, mixture level tracking model and others were selected for the analysis. It was assumed that the selected RELAP5 code parameters vary in the area where mainly two-phase flow conditions might occur: in the vertical heated channels, above the heated channels, steam-water pipes modelling elements and break location. Other areas (especially with single-phase conditions) are excluded due to the fact that these parameters do not have impact on the results in such regions.

Acceptance criterion 700 oC

750

No failure

Failure is possible

12 10 8 6 4 2 0 1.04 1.28 1.56 1.85 2.13 2.41 2.68 2.96 3.18 3.28 3.36 3.41 3.45 3.75

Channel power, MW

Figure 9: Real distribution of FCs power in the most loaded GDH at 4200 MW power level

ECCS. The results of the analysis [16] shows, that the peak cladding temperature in fuel channels with 3.0 MW power is close to the acceptance criterion for fuel cladding (700 oC) [17]. Therefore, this code output quantity was selected for the uncertainty analysis. The aim of the analysis was to evaluate the number of channels with affected fuel rods. Due to the fact that for the selected case only the upper limit of technological parameter is important, only one-sided tolerance limit is used in the uncertainty analysis. For the uncertainty analysis and according to Wilk’s formula, one-sided tolerance limit (with 0.95 of probability and 0.95 confidence) requires at least 59 runs to be performed [6], [18]. The behaviour of the calculated fuel cladding temperature in 3.0 MW power FC for all 59 calculation runs is presented in Figure 8. As it is shown in this figure, the fuel cladding temperatures band does not exceed the acceptance criterion of 700 oC. Therefore, while evaluating possible uncertainties of calculation, acceptance criterion is exceeded in the fuel channels with power higher than 3.0 MW and, thus, fuel cladding integrity in these FCs can be violated. For the evaluation

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

of the number of affected fuel channels, the real distribution of FCs power in the GDH taken from Ignalina NPP data was employed in this paper. Figure 8 shows a histogram of the reference channel power distribution in one GDH at the maximum permissible thermal operating power (i.e., 4200 MW). As it may be seen in Figure 8, there is a group of 12 fuel channels with power exceeding 3.0 MW; therefore, the integrity of fuel claddings in remaining 31 FCs will not be violated with 95% of probability and 95% of confidence. This information about the number of FCs (with possibly affected fuel claddings) was further used in the analysis of radiological consequences. Since the number of possibly affected fuel rods is small, this does not have any considerable impact to the radiological consequences.

energy generation in the affected fuel channels decreases, fuel channel cooling mode from the post-CHF returns back to the bubbly regime and fuel cladding and channel wall temperatures decrease. For the present transient analysis the same parameters (initial and boundary plant conditions, RELAP5 code models, assumptions and correlations) as for the analysis of large LOCA were selected. The uncertainty analysis with one-sided tolerance limits (according to Wilk’s formula 59 calculations are required) was performed. The analysis results - behaviour of the most important RELAP5 output result - fuel cladding

3.4. Analysis of coolant flow rate blockage in group of fuel channels

2 7 3

5 4

Figure 10: GDH blockage event. Structure of coolant flows: 1 - DS, 2 - MCP pressure header, 3 – manual valve, 4 – ECCS bypass line, 5 – ECCS header, 6 – affected GDH, 7 – fuel channels, connected to affected GDH

750 o

Acceptance criterion 700 C

700 650 600

Temperature, ºC

The blocking of coolant flow rate in group distribution header leads to the considerable coolant flow rate decrease in a group of 38-42 fuel channels connected to the affected GDH. Previous analyses showed that the consequences of this event are the worst (the highest fuel cladding and fuel channel wall temperatures are reached) in comparison to other transients. During the analysis of coolant flow rate blockage it was assumed that up to the beginning of the initiating process (before GDH blockage), the reactor operates at the power of 2900 MWth. The coolant is supplied through the core by two MCPs in each RCS loop. This reactor state was selected because in such conditions reactor cooling of the core is the most complicated. 2900 MWth is the maximum allowable power level when four MCPs in both circulation loops are in operation, i.e. the worst power and coolant flow rate ratio is selected. During such type of events this fact has high effect on the results. In calculations it is assumed that the manual valve (see (3) in Figure 10) in pipeline connecting MCP pressure header (2) and one GDH (6) is closed by mistake. It results in the drop of the pressure in the failed GDH (6). Under the influence of the pressure difference between MCP pressure header and failed GDH, the coolant from a MCP pressure header (2) starts to flow through bypass line (4) into the ECCS header (5) and further is directed into the failed GDH (6). Thus, 38 – 43 fuel channels, connected to this GDH are cooled only by water supplied through this ECCS bypass line (4). Due to this fact the coolant flow rate through affected FCs decreases. The decreased coolant flow rate removes less amount of heat from the fuel assemblies, thus, this leads to the critical heat flux in these fuel channels. Fuel cladding and channel wall temperatures rapidly increase before the reactor is shutdown due to the coolant flow rate decrease in one GDH. After reactor shutdown

550 500 450 400 350 300 250 -5

Time, s

Figure 11: Fuel cladding temperatures in maximum loaded FC, obtained using SUSA generated runs from RELAP5 calculations

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

temperature in the location 3.75 m from the core bottom – for all 59 calculation runs is presented in Figure 11. As it is seen from the presented figure the acceptance criterion (temperature of 700 oC for fuel cladding) is not exceeded in any FC with 95% of probability and 95% of confidence.

Temperature, C

___________________________________________________________________________________________________________

4. Comparison of “Best-Estimate” and “Partially-Conservative” Calculation Results

• The pressure in DS is equal to 6.95 MPa. It is the maximum possible pressure in the DS. This pressure is bounded by activation of equipment, protected the RCS from the overpressure (the lowest set point of activation of this equipment is equal 6.96 MPa. • The coolant flow rate through each MCP is assumed equal 6500 m3/h. This coolant flow rate is minimum possible and it is limited by the position of throttling regulating valves. • The feed water temperature is assumed pessimistically high and equal 195 oC. This value is equal to the maximum possible temperature of feed water 190 oC, taking into account 1 % of measurement error.

• The reactor thermal power is assumed equal to the maximum allowable reactor thermal power level 4200 MW increased by 1.06 times (3 % of measurement error and plus 3 % due to the first active control system interaction). For the “partially-conservative” calculations, the RELAP5 code modelling parameters, which had been recommended by user manuals and were established during the RELAP5 model validation procedure [14], are used. The comparison of “partially-conservative” calculation and upper boundary of “best-estimate” results (with realistic boundary & initial conditions plus uncertainty

Acceptance criterion of 700 С

"Partially-conservative" approach Upper boundary of "Best-estimate" approach

-10

-5

Time, s

Figure 12: Analysis of large LOCA. Comparison of the peak fuel cladding temperature in the fuel channel of 2.62 MW, connected to GDH with failure to close check valve in case of “partially-conservative” calculation and “best-estimate” calculation with uncertainty and sensitivity analysis

750

Acceptance criterion of 700 С

700 Upper boundary of best-estimate calculation 650

Partially-conservative approach

600

Temperature, 0C

The uncertainty and sensitivity analysis by employing the GRS methodology [6], [7] requires a certain number of calculations to perform. The use of “partiallyconservative” approach leads to minimization the number of calculations. In this case the best-estimate code RELAP5 with conservative boundary and initial plant conditions were employed. The conservative initial conditions are assumed as the worst possible initial conditions or the conditions increased (or decreased – depending on what value results in more conservatism) by possible measurement errors. For the MCP pressure header rupture with failure to close check valve of one GDH analysis at maximum permissible thermal reactor power level of 4200 MW the conservative initial conditions were assumed:

800 750 700 650 600 550 500 450 400 350 300 250

550 500 450 400 350 300 250 -5

Time, s

Figure 13: Analysis of GDH blockage event. Comparison of the peak fuel cladding temperature in the maximum loaded channel in case of “partially-conservative” calculation and “best-estimate” calculation with uncertainty and sensitivity analysis

and sensitivity analysis) for MCP pressure header rupture case is presented in the Figure 12. In the “partially-conservative” calculation the maximum temperatures is 10 – 15 oC higher as the upper boundary of results using “best-estimate” approach with uncertainty and sensitivity evaluation. The comparison of “partially-conservative” calculation and upper boundary of “best-estimate” results (with realistic boundary & initial conditions plus uncertainty and sensitivity analysis) for GDH blockage event is presented in Figure 13. The results of “best-estimate” approach shows that none of acceptance criterion are exceeded. As it is seen from the figure, the “partiallyconservative” values of peak cladding temperatures are approximately 6 – 8 centigrade higher and the acceptance criterion for fuel cladding temperature is reached.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Design Safety Limit

Acceptance Criterion (Regulatory Requirement) Margin to Acceptance Criterion, calculated using “partially -conservative ” approach

Margin to Acceptance Criterion, calculated using “best-estimate ” approach Upper Limit of Calculated Uncertainty Range, calculated using “bestestimate” approach

Actual Safety Margin

Value, calculated using “partially-conservative” approach

Calculated Uncertainty Range using “best-estimate” approach

Real Value

Value, calculated using “best-estimate” approach Figure 14: Illustration of the margin to the acceptance criterion

The comparison of results of “partially-conservative” calculation and “best-estimate” calculation with uncertainty and sensitivity analysis enables to draw a conclusion that in most cases both approaches either “best-estimate” or “partially-conservative” can be applied for the accident analysis. The latter approach looks tempting, since in this case only one calculation is sufficient; while in the case of “best-estimate” approach at least 59 calculations are required. Thus, “partiallyconservative” approach requires considerably less computational time. These two methods (“best-estimate” and “partiallyconservative”) can be compared by comparing the margin to the acceptance criterion (see Figure 14). The results, calculated using “partially-conservative” method should be more conservative in comparison with the results of “best-estimate” approach (using realistic boundary & initial conditions plus uncertainty and sensitivity analysis). Analysed accident situations consequences can be acceptable if calculated parameters’ values are below the acceptance criteria. Thus, if the results obtained using “partiallyconservative” method do not meet acceptance criteria (as it was in the GDH blockage case), the complete analysis by employing “best-estimate” approach is necessary.

5. Conclusions The analyses presented in this paper were performed for the licensing needs of RBMK-1500 reactors. However, the main findings may be applied for the

RBMK-1000 reactors, operating in Russia. The performed analyses demonstrated a wide range of application of the best-estimate methodology with uncertainty and sensitivity analyses. The deterministic calculations were performed using QUABOX/CUBBOXHYCA code (for the calculation of power distribution in fuel assembly), FEMAXI-6 code (for the modelling of processes in fuel matrix of RBMK-1500) and best estimate thermal-hydraulic code RELAP5 Mod3.2 (for analysis of reactor cooling circuit response). The integrated analyses code CONTAIN 1.2 was used for the analysis of reactor cavity response. The GRS statistical method based on propagation of input errors, together with the package SUSA was used for the uncertainty and sensitivity analyses. The presented applications demonstrated that the best-estimate methodology, used during nuclear power plants licensing activities, allows to avoid the unnecessary conservatisms and to assess and address the existing safety margins. The use of uncertainty analysis allows to assess the capacity of safety systems more accurately. This method also allows to analyse the influence of separate parameters on the calculation results, that creates the necessary conditions for further improvement of used computational models. The performed comparison between two approaches shows that in general “partially-conservative” approach provides higher peak temperatures. This approach could be recommended for safety calculations if obtained results do not violate safety criteria. Otherwise, “bestestimate” approach would be appropriate. It is necessary to point that for the “partially-conservative” approach proper boundary and initial conditions should be selected. This selection requires experience from

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

user, since for different phenomena, which are observed during the accidents, initial and boundary conditions can differ.

rods model testing employing the best estimate approach, Kerntechnik, Vol. 75, Iss. 3, (2010), pp. 72-80. [13]

Juseviciute A., Kaliatka T., Kaliatka A., Uspuras E., Usage of FEMAXI–6 program code for RBMK-1500 nuclear fuel rods simulation, Energetika (in Lithuaniain), Vol. 55, No. 2, (2009), pp 65-76.

[14]

Kaliatka A., Uspuras E., Thermal-hydraulic analysis of accidents leading to local coolant flow decrease in the main circulation circuit of RBMK1500, Nuclear Engineering and Design, Vol. 217, N 1–2, (2002), pp 91–101.

[15]

Cesna B., Rimkevicius S., Urbonavicius E., Babilas E., Reactor cavity and ALS thermal-hydraulic evaluation in case of fuel channels ruptures at Ignalina NPP, Nuclear Engineering and Design, Vol. 232, (2004), pp 57-73.

[16]

A. Kaliatka, R. Urbonas, M. Vaisnoras, Best estimate thermal-hydraulic analysis of loss of coolant accidents, Proceedings of 4th Baltic Heat Transfer Conference, Kaunas, Lithuania, 2003, pp. 8.

[17]

International Atomic Energy Agency, Accident analysis for nuclear power plants with graphite moderated boiling water RBMK reactors, Safety Reports Series No. 43, IAEA, Vienna, Austria, 2005.

[18]

Wilks S. S., Statistical prediction with special reference to the problem of tolerance limits, Annals of Mathematical Statistics, No. 13, (1942), pp. 400-409.

[19]

Pabarcius R., Kaliatka A., Marao A., Analysis of fuel rod behaviour during limiting RIA in RBMK plants, Kerntechnik, Vol. 75, Iss. 6, (2010), pp. 329-336.

[20]

Antikain P. A., Metals of equipment and pipe lines for nuclear power plants, Moscow, Russia, 1984.

[21]

International Atomic Energy Agency, Safety margins of operating reactors. Analysis of uncertainties and implications for decision making, IAEA-TECDOC-1332, IAEA, Vienna, Austria, 2003.

[22]

Uspuras, E., Kaliatka, A., Almenas, K. Verification of Design Characteristic of the RBMK-1500 Accident Confinement System, Nuclear Engineering and Design, 1995, Vol. 153, No.3.

[23]

Uspuras Е., Кaliatkа А. Accidental transient processes at NPP with RBMK-1500. Моnograph, Каunas, 1998, 194 p. (In Russian).

References [1]

[2]

Langenbuch S., Maurer W., Werner W., Coarse mesh flux expansion method for the analysis of space-time effects in large LWR cores, Nuclear Science and Engineering, No. 63, Vol. 437, (1977). Langenbuch S., Maurer W., Werner W., HighOrder schemes for neutron kinetics calculation based on a local polynomial approximation, Nuclear Science and Engineering, No 64, Vol. 508, (1977).

[3]

Suzuki M., Light Water Reactor Fuel Analysis Code FEMAXI-6 (Ver. 1), Japan Atomic Energy Research Institute, Tokyo, Japan, 2005.

[4]

Fletcher C. D., et. al., RELAP5/MOD3 code manual user’s guidelines, NUREG/CR-5535, Idaho National Engineering Lab., Idaho Falls, USA, 1992.

[5]

Murata K. K., Change document for update C11af: CONTAIN 1.2 pre-release bug fixes, 1995.

[6]

Glaeser H. G., Uncertainty Evaluation of ThermalHydraulic Code Results, Proceeding of int. Meeting on Best-Estimate Methods in Nuclear Installation Safety Analysis, BE-2000, Washington DC, USA, 2000.

[7]

Kloos M., Hofer E., SUSA Version 3.2. User’s Guide and Tutorial, GRS, Garching, Germany, 1999.

[8]

Vileiniskis V., Kaliatka A., Uncertainty and sensitivity analysis of MCPs’ trip events at Ignalina NPP, Nuclear Engineering and Design, Vol. 224, (2003), pp. 213–225.

[9]

Almenas K., Kaliatka A., Uspuras E., Ignalina RBMK-1500. A Source Book. Extended and Updated Version, Lithuanian Energy Institute, Kaunas, Lithuania, 1998.

[10]

Bubelis E., Pabarcius R., Tonkunas A., Modelling of continuous withdrawal of CPS control rods without reactor scram using QUABOX/CUBBOXHYCA code, Energetika, No. 1, (2004), pp. 49-56.

[11]

Tonkunas A., Pabarcius R., Clemente M., Listopadskis N., Uncertainty and Sensitivity Analysis of the CPS-CC Voiding in the RBMK Reactor Ignalina-2, Kerntechnik, Vol.71, Iss. 3, (2006), pp. 104-112.

[12]

Marao A., Kaliatka T., Kaliatka A., Uspuras E., Adaptation of the FEMAXI-6 code and RBMK fuel

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

DOI: 10.14621/ce.20150202

Analysis of Operation of Heat Accumulator in Large–Scale Combined Heat and Power Plant Wojciech Bujalski, Krzysztof Badyda, Kacper Rosinski Institute of Heat Engineering, Warsaw University of Technology 21/25 Nowowiejska Street, 00–665 Warsaw, Poland, bujalski@itc.pw.edu.pl

Abstract

1. Introduction

This work presents analysis of operation of heat accumulator in one of the biggest European CHP plant. It includes basic information about optimization process for combined heat and power plants and cooperation heat accumulators with such a system. The main focus in this work was given to propose general rules of heat accumulator behaviors, which have been developed by optimization software created by Department of Power Engineering Machines at the Institute of Heat Engineering of Warsaw University of Technology. In the experimental part the unique method of accounting cycles was presented. It was assumed that two qualitatively different cycles can be observed during proper operation. The most intensive usage of heat accumulator was observed in spring and autumn seasons. This time of the year was also characterized by the highest speed of charging/discharging process. During summer there were only a few single cycles and operation of whole accumulator was regulated in weekly mode. Finally, validity of optimizer’s decision was proven because charging process was always covered with occurrence of the highest spot market prices.

Currently, the very common is an application of a heat accumulators integrated with district heating systems [1]. The profitability of this solutions is based on energy production’s optimization [2, 3, 4, 5, 6]. It is well known, the CHP plants, which are based on back pressure turbines, generate electricity with close-coupled dependence on the amount of generated heat. In the case of energy sale on power exchange, the electricity price is low during the night, while expensive during the day. The electricity production by CHP plant is an opposite – during the day, when the energy is expensive, the amount of generated electricity is low, while during the night i.e. period when the electricity prices are low, the electricity production by CHP plant is very high. Therefore, the optimization process bases on the shifting of energy production from night period to peak hours. This shift is mainly possible to due to an application of heat accumulators.

Keywords:

Heat accumulator; Optimisation; CHP

Article history:

Received: 28 July 2015 Revised: 10 September 2015 Accepted: 09 November

In the available literature, there are reported many analysis of a heat accumulator cooperation with a district heating network. Željko et al. [7] reports the influence of heat accumulator on CHP plant’s income, on the basis of optimization code – ACOM. The target of the objective function was to select appropriate load coefficient for each devices for minimization the value of costs’ function. The electricity price was assumed to be known, however it was split into two separate periods with different values. The application was created especially for one of the Croatian CHP plants. Paper [8] reports an analysis of operation of CHP plant equipped with the heat accumulator. Heat demand was assumed to be known and electricity prices were time depended. The aim of the analysis was to adjust generation to appropriate water inlet temperature and strategy of heat accumulator’s management. The main target of objective function was costs minimization, which are covered by district heating network operator.

___________________________________________________________________________________________________________ W. Bujalski, K. Badyda, K. Rosinski: “Analysis of Operation of Heat Accumulator in Large-Scale Combined Heat and …”, pp. 13–19

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

The problem was solved by using a projected Lagrangian algorithm. Paper [6] presents an analysis for selection an optimal capacity of CHP plant combined with heat accumulator in the German spot market, competitive energy market. The paper shows an impact of store volume of heat accumulator on generation and their impact on CHP’s income. All of the presented analysis, which are based on net present value and simple payback time, were obtained by using a commercial software energyPRO. In paper [5], it is reported a short–term optimization model for CHP plant equipped with a heat accumulator. The proposed model is based on MILP method. Author considered three different scenarios. Paper [9] report long-term strategy for optimization of power systems equipped with a heat accumulators. The proposed model takes into account nonlinear characteristics curve of CHP plant’s operation. The optimization process was solved by MINLP method. The presented papers relates to planning period and assumes the complete use of storage capacity of heat accumulator. However, in reality, the system operation has a lot of constrains and real operation is different than predicted one. Figure 1 presents a real changes for heat consumption and electricity prices. Figure reveals that the changes of both electricity prices and heat consumption are not clear as well as there are

very often far away from the results of simulation forecasts. In addition, CHP plants has some operational limits e.g. failures. Thus, the complete implementation of scheduled operation is not possible. Authors implemented the optimization system of heat accumulator’s operation in the CHP plant, which has a highest capacity in the Poland [10]. However, the most interesting point is an analysis of real operation of heat tank and if (how) it is possible to use the total potential of storage tank in operating conditions.

2. CHP PLANT The all research were done at existing object i.e. Siekierki CHP plant. This plant is one of four main sources, which are supplying the Warsaw district heating network. The Siekierki CHP plant (see Figure 2), the biggest one in Poland and the second biggest in Europe, is the largest heat source supplying Warsaw Heating System – the most extensive one in European Union. The Siekierki CHP plant is composed of 9 steam turbosets, two of them extraction-condensing ones (ST1 of power 52, ST8 of power 125 MW). Other turbines are three large counter-pressure units of 100 MW (large power unit with turbines ST7, ST9 and ST10) and three

Figure 1: The exemplary heat demand covered by one of the polish CHP plants and electricity prices on stock market ___________________________________________________________________________________________________________ W. Bujalski, K. Badyda, K. Rosinski: “Analysis of Operation of Heat Accumulator in Large-Scale Combined Heat and …”, pp. 13–19

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 2: Simplified diagram of heat sources technological system in the Siekierki CHP plant with indicated connections to the heat accumulator. (B – boiler, ST – steam turbine, PRS – pressure reduction station, X – heat exchanger)

extraction counter-pressure of power 30 MW (turbines ST2, ST3, ST5). The fourth one of power 30 MW (ST6) is not equipped with a controlled steam bleeder, therefore it operates as a counter-pressure turbine. The five turbosets ST1, ST2, ST3, ST5 and ST6 compose the older part of the facility that is supplied with steam by four steam boilers through a header system (header part of CHP). The remaining four units (large power unit) work in the newer part of the boiler house. Six water boilers extend the technological system of the CHP plant. Conspicuously, the configuration of heat sources is complex. The connection of the accumulator and the system enables cooperation with any cogenerating unit (Figure 2). Basic technical data of the accumulator is presented in the Table 1. The task of the accumulator is to improve flexibility of applied turbosets mainly in order to increase electricity generation in high (peak) demand time of the day. In this case it is more complicated than in case of large accumulators in other countries due to several reasons:

Table 1: Basic parameters of the heat accumulator installed in the Siekierki CHP plant No. Parameter

Unit

Value

Volume

30 400

Container height

Container diameter

Heat storing capacity

MWh

1600

Heating power

300

Insulation thickness

500

Charging/discharging speed

t/h

4500

• Two counter-pressure turbines are used in the analyzed facility; this type of turbines is not commonly used in plants with accumulators;

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

• The ratio of heating capacity of the plant to the accumulation capabilities of the accumulator is significantly different from used in other cases (Figure 1); • The discussed facility is an example of an installation having complex technological structure and high level of heating power; • The typical load variability is different from observed for most Danish, and even German and Austrian cities due to climate conditions of Warsaw. Climate in Warsaw is relatively harsh what results from location of the city (e.g. long distance from the sea comparing with Danish cities).

3. The analysis of a heat accumulator’s operation The analysis of heat accumulator’s operation is based on the data gathered by an electronic acquisition data system (PI Process Book). The examined time period was one year, i.e. since 1st January 2013 to 31th December 2013; with an one hour time interval. The available data consisted of the following parameters: the power of each device, the scheduled production and ambient conditions. Those data ensured possibility to analyze an operation of heat accumulator at different periods. The first stage of analysis was to determine time periods, when the heat accumulator’s behavior is either constant or unique. Poland is situated in the temperature climate with a clear distinction between seasons. Thus, the proposed division of time periods distinguishes the following periods: winter, summer and transition period. The division into periods was not based on the calendar, but on the basis of detailed analysis of temperatures at each month, therefore: a) the summer season — June – August b) the transition season — April – May, September – December c) the winter season — January – March. Figure 3 shows an average absolute values of temperatures (temperature to maximum temperature at given time period). It is clearly presented that the highest changes for heat demand occurs during transition period. The next examined elements was both degree of utilization and types of thermal cycles of heat storage at each month. The detailed analysis of charging status of heat accumulator (expressed in % of maximum capacity) revealed existence of two different cycles. There were

denoted as “low cycles” and “high cycles”. The main difference between those cycles lays in a three consecutive function extrema’s of status of heat accumulator load. In the case, when difference was a two times in a range from 20% to 50% of maximum capacity, the cycles was recognized as “low cycle”, while the difference two times was above 50%, the cycles was classified as “high cycle”. This way of analysis allowed to determine the frequency of occurrence of each cycle at every month. As a part of research, the average duration time of each cycle was calculated. The results are depicted in Figure 4. The examination revealed that the storage tank is the most intensively used during the transient period. In the other words, it is characterized by the largest number of cycles, while the duration of the cycles is very short, barely exceed 30 hours. In addition the ratio of “huge cycles” to “small cycles” is the smallest. During the winter period, the dynamics of accumulator’s operation decreases, while the ratio of “huge cycles” to “small cycles” increases. The average time of duration of “huge cycle” is ca. 40 hours. The decreased dynamics of storage tank utilization is a result of constant heat demand, which enables the operation of most of base load power plants with full capacity. In this situation, the optimization of operation of back pressure turbines, which operation is not flexible, does not provide considerable economic benefits. During summer period there are noticeable significant reduction of operation’s dynamic of storage tank. In July and August, there are only a few cycles, which duration time is very long. The observed cycles are much different than those in other periods, because the charging of storage tank does not always coincide with the highest prices of electricity at stock market. The observations suggest that during summer period, the operation of heat storage tank is regulated in weekly mode. Hence, the heat tank operation in July and August, i.e. the period when the average duration of each cycle exceed 150 hours, was scrutinized. This number corresponds to duration of a week, which is consistent with the assumption. It should be noted, that time of use of heat accumulator never reaches 100% during the month, because there are some moments, when the heat accumulator is not operating within tens of hours. The next examined issue was a dynamic changes of a status of heat accumulator at every day in each period. The first stage was a determination the average % of heat accumulators’ charging state at every day. Unfortunately, the specific level of charging is not a meaningful value. Thus, the derivate of the charging status over a time was determined. In order to verify the validity of optimization’s decisions, the average prices of electricity in each period was added and presented in the graphs. The results are shown in Figures 5–7.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 3: The average days changes of heat demand for three periods in year (Q – average hourly heat demand at given, Qmax – the maximum heat demand at given period)

Figure 4: The amount, type and average duration of cycles in 2013

The results confirm the validity of optimizer’s selection and conclusions obtained by the method of cycles’ calculation. The most dynamic changes are observed in transient period. During the night, the change of tank’s

charge status reaches – 6.5 %/h. The maximum rate of discharging occurs at 3.00-4.00 a.m., while at this moment, the electricity price at stock market is ca. 27 €/MWh. On the other hand, during the 10:00 a.m. to

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

charging/discharging dynamics is not so intensive, i.e. the maximum charging/discharging rate is 5.1 and – 4%/h respectively. The electricity prices, in the period mentioned above, was not lower than 40 €/MWh during the charging period, while during the period with highest rate of discharging, the electricity price was 29 €/MWh. The summer period is characterized by the smallest dynamic of change of heat accumulator. The extreme speed variations of heat accumulator filling are in range -2% to 1.8%/h. Also, in this case, the maximum speed of charging occurs when the electricity prices is the highest on the stock market. Figure 5: Dynamic changes of an accumulator charging status and electricity prices during winter 2013

Figure 6: Dynamic changes of an accumulator charging status and electricity prices during transient period 2013

4. Summary During the heat accumulator’s examination, one of the most difficulties tasks is a correct estimation of storage tank potential after an installation. The proposed analysis allows to qualitatively and quantitatively describe the degree of heat accumulator use. The paper reports an analysis of heat accumulator use and their operation cycles were defined, i.e. “high” and “small” cycles. The “high” cycle denotes the cycle, where the degree of use of heat accumulator is high, i.e. full charging the discharging of heat accumulator. The “low” cycle stands for a cycle, where the potential of heat accumulator was not fully used. The detailed analysis revealed that the highest utilization of heat accumulator occurs during transient period. Thus, the number of “high” cycles increases, while there are only few so called “small cycles”. During 30 days in a month, there were ca. 20 “large cycles”. The analysis confirmed that during the summer period the heat accumulator was regulated in weekly mode. The presented results allow to develop a methodology for assessing the actual degree of heat accumulator’s utilization in a new constructed CHP plants. It shows that, during the transient period, the accumulator is able to do ca. 20 cycles per one month. In other periods, the number of cycles is significantly lower. Authors will examine other facilities to develop more universal rules.

References [1] Figure 7: Dynamic changes of an accumulator charging status and electricity prices during transient period 2013

Gustafsson S.I., Karlsson B.G.: Heat Accumulators in CHP Networks, Energy Conversion and Management, vol. 33(12), 1992, pp. 1051–1061.

[2]

9 p.m., when the electricity price is not lower than 38 €/MWh, the heat accumulator is intensively charge with rate exceeding 6%/h. During the winter period, the

Gatzen C.: The Economics of Power Storage, Theory and Empirical Analysis for Central Europe von Christoph Gatzen, Oldenbourg Industrieverlag, Munchen, 2008.

[3]

Henning D.: Cost minimization for a local utility through CHP, heat storage and load

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

accumulator, Energy, vol. 31(13), 2006, pp. 2285 – 2292.

management, International Journal of Energy Research, vol. 22(8), 1998, pp. 691–713. [4]

Medrano M., Gil A., Martorell I., Potau X., Cabeza L.F.: State of the art on high-temperature thermal energy storage for power generation. Part 2— Case studies, Renewable and Sustainable Energy Reviews, vol. 14(1), 2010, pp. 56–72.

[5]

Rolfsman B.: Combined heat-and-power plants and district heating in a deregulated electricity market, Applied Energy, vol. 78(1), 2004, pp. 37– 52.

[6]

Streckiene G., Martinaitis V., Andersen A.N., Katz J.: Feasibility of CHP plants with thermal stores in the German spot market, Applied Energy, vol. 86(11), 2009, pp. 2308–2316.

[7]

Željko Bogdan, Kopjar D.: Improvement of the cogeneration plant economy by using heat

[8]

Zhao H., Holst J., Arvastson L.: Optimal operation of coproduction with storage, Energy, vol. 23(10), 1998, pp. 859 – 866.

[9]

Tveit T.M., Savola T., Gebremedhin A., Fogelholm C.J.: Multi-period MINLP model for optimising operation and structural changes to CHP plants in district heating networks with long-term thermal storage, Energy Conversion and Management, vol. 50(3), 2009, pp. 639–647.

[10]

Badyda K., Bujalski W., Milewski J., Warchoł M.: Heat Accumulator in Large District Heating Systems: Simulation and Optimisation, ASME Conference Proceedings, vol. 2010, 2010, pp. 39– 44.

International Journal of Contemporary ENERGY, Vol. 1, No. 1 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

DOI: 10.14621/ce.20150203

Complementary and Overlapping among Energy Performance Indicators as Part of the Sustainable Development and RECP Indicators in Cement Industry Ana M. Lazarevska1), Natasha Bakreska Kormushoska2), Atanas Kochov1) 1)

Faculty of Mechanical Engineering, Ss Cyril and Methodius University Karposh II b.b. POBox 464, 1000 Skopje, Macedonia, ana.lazarevska@gmail.com 2) Cementarnica USJE, AD Skopje, Macedonia

Abstract The cement industry is one of the most energy intensive of all manufacturing industries. In that respect, saving energy in the cement production process is one of the key elements of sustainability and has equal importance from an environmental and an economic point of view. The Key Performance Indicators (KPIs) for energy provide a measure on how companies are moving towards a more eco-efficient use of fuels. This paper focuses on investigating energy KPIs developed within different models. In this case, comparison is made between energy KPIs presented as part of the Sustainable Development Indicators and Resource Efficiency and Cleaner Production (RECP) model. The aim of the herein presented work is to determine the complementarities and overlapping among energy performance indicators and their inter-relation with other sets of indicators utilized for measuring contribution towards sustainable development i.e. environmental, economy and social indicators. The attained set of indicators can further be utilized as a pool of indicators relevant for assessing how implementing the RECP model contributes towards increasing energy efficiency of the production process and thus, affects the overall sustainability of the company. The paper argues that the nuclear energy may improve the public trust significantly and at the same time improve the safety record by a much stronger commitment towards the science based decision making in the industry and the regulatory organizations.

Keywords:

Article history:

Energy KPIs; RECP; Sustainable development; Cement industry

Received: 13 July 2015 Revised: 14 September 2015 IntroductionAccepted: 09 November 2015

Energy and material intensive industrial sectors are facing the challenge to engage in the holistic concept of industrial ecology [1], thus to integrate sustainable development (SD) and the RECP principles through carefully selected and controlled practices where byproducts and waste materials from one industry become inputs for another. The cement industry is one of the most energy intensive of all manufacturing industries. Cement production relates to large amounts of raw materials and fuels on the input side and substantial carbon dioxide emissions on the output side. In that respect, saving energy in the cement production process is one of the key elements of sustainability because it is equally important from environmental, economic and social point of view. Therefore, the cement industry is one good example on how industries become inventive and smarter in the use, reuse, and recycling of raw materials, energy, and wastes. By utilizing both the useful mineral content and recoverable calorific value a synergy between alternative fuels and raw materials is created. Namely, industrial by-products and waste materials are incorporated in the manufacturing process, some as part of the final product, others as fuel. The latter reduce the necessary amount of fossil fuels, implying lesser environmental impact. A further positive impact on the environment is through involving cement kilns in waste management and hazardous waste disposal, thus lowering demands on landfills and/or incinerators, implying lesser groundwater pollution, greenhouse gasses generation and hazardous ash residues. Having the above in perspective, on one side, stakeholders are concerned about the safe usage of byproducts as alternative fuels and their proper management, and on the other, about compliance of the related discharged emissions with corresponding protocols (especially when related to hazardous waste), towards securing controlled and minimized impact of those by-products on the wider community and

___________________________________________________________________________________________________________ A. M. Lazarevska, N. Bakreska Kormushoska, A. Kochov: “Complementary and Overlapping among Energy Performance …”, pp. 20–26

International Journal of Contemporary ENERGY, Vol. 1, No. 1 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

environment. Built upon the principles of SD, ecoefficiency [2] and industrial ecology, the Cement Sustainability Initiative (CSI) [3, 4] has emerged as a result of mutual commitment between the cement industry and the concerned stakeholders worldwide towards ensuring proper management and safe use of fuels and materials in the cement industry, thus minimizing environmental impact. The following elements of the SD concept have been selected as a basis of the CSI, i.e. managing resources (energy, materials, waste), implementing technology advancements (process optimization, waste coprocessing and energy/materials recovery, ecoinnovation), protecting ecosystems and reducing cement industry “footprint” on the environment [5], and promoting quality of life (higher quality of the construction projects, maintaining health and safety, and high-quality employment opportunities, while carrying for the social and economic needs of the local communities). Delivering reliable decisions that address the challenge to resolve and prevent potential environmental pollution originating from industrial facilities, while maintaining high product quality and optimal production costs, is closely related to identifying pollution sources in all media: water, air and soil and comparing them versus economic and social parameters. This challenge extends to determining suitable and explicit indicators (often integrated) that facilitate proper monitoring [6]. Finally, optimizing this set of indicators is a prerequisite when efficient and sound decision making is the goal. The research presented herein focuses on investigating energy related indicators developed and utilized within different models. In this case, comparison is made between indicators used in the SD and the Resource Efficiency and Cleaner Production (RECP) model, with a final goal to determine complementarities, overlapping and interrelations among them.

2. Examples of deteriorated infrastructures Sustainable Development (SD). As per the Brundtland Report [7] Sustainable development (SD) encompasses two key issues: (1) the concept of needs and (2) the environment's ability to meet present and future needs dictated by the state of technology and social organization. The three SD pillars are economic growth, environmental protection and social progress. Authors

researching sustainable energy resource management [8, 9] recommend a fourth pillar which addresses technology aspects. This category often serves as a pool of alternatives towards realizing the analysed SD concept. Information relating SD is available and continuously reported with IUCN, UNEP, WWF [10, 11, 12], the Open Working Group on SD Goals1, etc. When it comes to delivering decisions on how to use various Earth’s resources, one has to take in account not only how much of these resources are used, but as well what processes were used to obtain those resources, who has access to those resources, are there going to be enough resources left for the next generations, and will the environment be left as it is today? Thus, as every other system, it is important to understand that when assessing a particular system’s contribution to SD this process involves and is limited by temporal and space boundaries, while its implementation is closely related to the quality of the analysed system. Resource Efficient and Cleaner Production (RECP). As per UNEP (1990) the valid definition of the term Cleaner Production is: “… continuous application of an integrated environmental strategy to processes, products and services to increase efficiency and reduce risks to humans and the environment”.2 The RECP concept continuously applies integrated and preventive strategies to processes, products and services, towards increasing efficiency and reducing risks to humans and the environment, i.e. works towards advances in: − production efficiency – through optimization of productive use of natural resources (materials, energy, water) at all stages of the production cycle; − environmental management – through minimization of the adverse impacts of industrial production systems on nature and the environment; − human development – through minimization of risks to people and communities, and support to their development. The seven absolute indicators defining the RECP concept, i.e. materials, energy and water, for resource use, waste, air emissions and waste water for pollution and production indicator for product output and their interrelation are presented on Figure 1. They are the basis for defining the corresponding six relative indicators normalized vs. the reference production indicator, i.e. for resource productivity (materials, energy and water productivity) and for pollution intensity (air emissions, waste and waste water intensity).

https://sustainabledevelopment.un.org/index.php? industrial-production/cp/cleaner-production.html menu=1549 http://www.unido.org/en/what-we2 UNEP, CP Programme, http://www.unep.fr/scp/cp/, UNIDO do/environment/resource-efficient-and-low-carbonCP Programme, http://www.unido.org/en/what-weindustrial-production/cp/cleaner-production.html do/environment/resource-efficient-and-low-carbon___________________________________________________________________________________________________________ A. M. Lazarevska, N. Bakreska Kormushoska, A. Kochov: “Complementary and Overlapping among Energy Performance …”, pp. 20–26

International Journal of Contemporary ENERGY, Vol. 1, No. 1 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 1: The seven absolute indicators defining the RECP concept and their interrelation (Source: UNIDO/UNEP, 2010 [13])

Table 1: Main characteristics and selection criteria for Key Performance Indicators

KPIs should

(1) conform to a company’s strategy (2) be easy to understand (3) allow for action (4) be contextual (5) be non-redundant

Interconnection through

KPIs should enable (1) personnel management (empower employees for sound work performance)

Management Decision making Interface

(2) operational management (towards improving performance) (3) emergency procedures (prevent and/or address corresponding risks) (4) communication with/from the stakeholders (local community; governmental and other relevant authorities NGOs, etc.)

3. Conceptual approach towards SD and RECP in cement industry The commitment that cement industry has taken in order to practice and promote eco-efficient and sustainable cement production in line with the principles of industrial ecology and corporate social responsibility (CSR), lead to defining a variety of indicators serving as a pool of indicators from where companies decide and choose which of them are the best reflection of their company policy. As noted in Section 2, since SD addresses all aspects of RECP, in this paper, we utilize the SD matrix to compare and present this overlapping. The authors performed a thorough survey among criteria and indicators used in the cement industry and globally, thus the selected and herein presented set is the one that gives the sound description of all aspects of the SD and RECP concepts applied to the cement

industry, with as least as possible redundancy regarding those aspects. Following sources have been consulted and reviewed to compose the above mentioned set: IUCN/UNEP/WWF [10, 11, 12], UNIDO [13], the WBCSD/ CSI [2, 3, 4, 14, 16], World Bank Group/International Finance Corporation (IFC) [i], the Overall Equipment Effectiveness Foundation, the European Cement Research Academy (ECRA), the Eco-Cement Project [15], general and sector-specific reporting guidelines promoted by the Global Reporting Initiative (GRI) and eco-efficiency management tools, internal guidelines and practices in the cement industry etc., withstanding with relevant international standards and country-specific legislation and regulative. It is clear that with reference to the latter, i.e. compliance with the existing global, regional and local legislation, the analysis should be performed on a case-by-case basis. Namely, every analysed facility subject to assessment on whether and to what extent it complies with the SD and/or RECP concept, including cement industry, should be in

International Journal of Contemporary ENERGY, Vol. 1, No. 1 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

compliance not only with the country-specific legislation (relating financial, environmental, social etc. issues) where the facility is located, which may as well have to be in compliance with the regional and global agreements, protocols, directives, but as well with the international standards, regulative, protocols, procedures defined and required by the industry group

it is affiliated to (e.g. as in the case of the global cement groups, members of the CSI). Since neither the three SD pillars nor the RECP concept define a direct indicator in the first level of the hierarchy referring to legislation compliance, therefore in our analysis we have foreseen sub-indicators in each of the four groups presented in Table 2 through Table 5, that

Table 2: Relevant environmental criteria and indicators in the cement industry

Table 3: Relevant economic criteria and indicators in the cement industry

International Journal of Contemporary ENERGY, Vol. 1, No. 1 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Table 4: Relevant social criteria and indicators in the cement industry

Table 5: Relevant technology criteria and indicators in the cement industry

indirectly take the influence of the compliance with the legislation into account. E.g. in all indicators in Table 2, the compliance with the environmentally related legislation is interwoven as per the country-specific and/or facility-specific situation, while indicator 2.12 is a measure on whether a financial frame that has to be

allocated in order to enable compliance with the existing relevant legislation exists and to what extent. Similarly, sub-indictors referring the compliance with legislation from a social (e.g. sub-group Health & Safety (H&S) at work in Table 4) and from a technology (e.g. sub-group Product Quality in Table 5) point of view are foreseen.

International Journal of Contemporary ENERGY, Vol. 1, No. 1 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

4. Results and discussion Table 2 through Table 5 are a systematic representation of the interconnection/overlapping between SD and RECP concept, whereby in the columns “SD and/or RECP concept” with the corresponding seven RECP indicators it is indicated whether and what type of compliance exists. The last column notes whether the chosen indicators have relation to energy demand/consumption, thus weather they hold potential to be used as energy performance indicators. Further research on practical case-studies shall prove whether the proposed selection and categorisation is justified. Table 2 provides a sound presentation of relevant environmental aspects in the cement industry. The authors would like to stress that when choosing a categorization approach for the sub-category “wastes”, herein we focused on the ways wastes are used in the process (i.e. whether their environmental impact shall be reduced/increased by means of their inclusion/exclusion in the cement production process), although this category can be sub-categorized from various points of view (e.g. ways of disposal, types of contaminants, originating point etc.) [18]. As presented in Table 3, in this study we consider economic challenges in the cement industry through three main sub-categories addressing production, environment and externalities. Among those, only the production and environment related ones have direct implications on energy demand (consumption), thus can serve as energy related indicators. Table 4 represents relevant social aspects in the cement industry, categorized in three subcategories: community relations, human resources, health and safety. It is notable that direct relation to energy demand/consumption does not exist, while impact on product output in indirect. Table 5 covers relevant technology aspects categorized in the following sub-categories: energy sources, process management [19], materials and product quality. As expected, all of them are related to energy performance of the facility, while RECP compliance is noted through the energy usage and product output indices.

requirements of the multi-criteria analysis, thus enables addressing goals towards more sustainable and ecoefficient cement production by applying the theory of multi-criteria decision making. Moreover, as a future research challenge, such hierarchy, with suitable modifications, holds potential to address issues relating risk assessment and management in the cement industry.

References [1]

[2]

World Business Council for Sustainable Development (WBCSD), Eco-efficiency – creating more value with less impact, ©WBCSD, 2002.

[3]

WBCSD and CSI, The Cement Sustainability Initiative: Our Agenda for Action, ©WBCSD and CSI, Switzerland, July 2002.

[4]

WBCSD and Cement Sustainability Initiative (CSI), Guidelines for the Selection and Use of Fuels and Raw Materials in the Cement Manufacturing Process, ©WBCSD and CSI, Dec 2005.

[5]

The Cement CO2 and Energy Protocol, Ver. 3, CO2 and Energy Accounting and Reporting Standard for the Cement Industry, ©by WBCSD CSI / ECRA GmbH.

[6]

Nospal, Aleksandar, Lazarevska, M. Ana, 2008: Environmental protection and industry: Parameters necessary for environmentally related decision making, GeoSpatial Visual Analytics: Geographical Information Processing and Visual Analytics for Environmental Security”, (Eds. De Amicis, R., Stojanovic, R., Conti, G.), Springer + NATO Public Diplomacy Division, Science for Peace and Security Series C: Environmental Security, Science + Business Media B.V., Dordrecht, The Netherlands, 2009, pp. 97–112.

[7]

Brundtland, G. et al., Our Common Future, The World Commission on Environment and Development. Oxford: Oxford University Press, 1987, p. 24.

[8]

Afgan, N. H., Carvahlo, M. G. and Hovanov, N. V., Energy System assessment with sustainability indicators, Energy Policy, 28, pp 603-612, 2000.

[9]

Zhou, P., Ang, B. W. and Poh, K. L., Decision analysis in energy and environmental modeling: an update, Energy, 31, pp 2604-2622, 2006.

[10]

International Union for Conservation of Nature (IUCN), United Nation Environment Programme (UNEP), World Wide Fund for Nature (WWF), World Conservation Strategy: Living Resource

5. Conclusion In this paper, a pool of relevant indicators utilized in the cement industry is investigated and a structured overview seen through the prisms of sustainable development (SD) and resource efficient and cleaner production (RECP) is provided, whereby, relation to energy performance is noted. The complementarity and overlapping among the two concepts is clearly identified and presented by utilizing the SD matrix, whereby a hierarchical approach is applied to address each SD aspect. This approach is in concordance with the

International Journal of Contemporary ENERGY, Vol. 1, No. 1 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Conservation for Sustainable Development, ©IUCN-UNEP-WWF, 1980. [11]

[12]

IUCN, UNEP, WWF, Carrying for the Earth: A Strategy for Sustainable Living, ©IUCN/UNEP/WWF, Gland Switzerland, 1991. IUCN, 2006. The Future of Sustainability: Rethinking Environment and Development in the Twenty-first Century. Report of the IUCN Renowned Thinkers Meeting, 29-31 January, 2006.

[15]

ECO-CEMENT Project, Deliverable 2.23 WP2 KPIs Eco-Cement Project (FP7 funded project), 2013 (http://eco-cement.eu/download.html).

[16]

WBCSD and CSI, Safety in the Cement Industry: Guidelines for measuring and reporting, Health and safety, Ver. 4.0, ©WBCSD and CSI, May 2013.

[17]

World Bank Group/IFC, Environmental, Health, and Safety Guidelines for Cement and Lime Manufacturing, World Bank Group/IFC, April, 2007.

[13]

United Nations Industrial Development Organization (UNIDO), UNEP, Enterprise-Level Indicators for Resource Productivity and Pollution Intensity: A Primer for Small and Medium-Sized Enterprises, November 2010, ©UNIDO, 2010.

[18]

Bakreska Kormushoska Natasha, Atanas Kochov, Defining indicators and boundaries of the recp system for cement industry, Proceedings, Proceedings of the Conference for International Energy and Environmental Protection in South Eastern Europe, Zlatibor, Serbia, 2015.

[14]

WBCSD and CSI, The Cement Sustainability Initiative: Progress Report, ©WBCSD and CSI, Switzerland, June 2005.

[19]

Raffaele Iannone and Maria Elena Nenni, Managing OEE to Optimize Factory Performance, Operations Management, INTECH, pp. 31-50.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

DOI: 10.14621/ce.20150104

Numerical Analysis of the Impact of Parameters of Urea Solution Injection on Reagent Penetration inside the Combustion Chamber of an OP–140 Boiler Krzysztof Badyda, Piotr Krawczyk, Szczepan Mlynarz Institute of Heat Engineering, Warsaw University of Technology 21/25 Nowowiejska Street, 00–665 Warsaw, Poland, piotr.krawczyk@itc.pw.edu.pl

Abstract

1. Introduction

The implementation of new emission standards under the Industrial Emissions Directive is helping to drive the introduction of new technologies by power producers. Producers in Poland using the common OP-140 type pulverized fuel fired boilers are looking for solutions to reduce NOx emission in flue gases.

OP-140 is a pulverised fuel fired boiler design. These boilers are mainly operated in municipal and industrial heat and power generating plants and they have not been equipped with any flue gas cleaning systems, except for particulate matter removal. Current emission standards permit concentrations of 600 mg/Nm³ at 6% O2 for nitrogen oxides in the exhaust gas from those boilers [10].

Selective Non-Catalytic Reduction (SNCR) technology is one of the deNOx technologies tested for possible application in OP-140 type boilers. Optimization of the NOx emission reduction process in this technology consists of a selection of corresponding parameters of sprayed reagent in a boiler internal space. As part of project POIG.01.03.01 – 14 – 035/12 a simulation of thermal-flow conditions in a combustion chamber was carried out using Computational Fluid Dynamics (CFD) tools, among others simulation of thermal-fluid conditions in an OP-140 type boiler as well as the effect of operating parameters of the system that sprays out the reagent water solution and the influence of nozzle geometry on the spraying quality of this solution. This paper provides the results of aforementioned CFD simulations and experiments of SNCR technology operation for the considered optimal operating parameters (for selected levels of the OP-140 type boiler output) of this system.

New legal regulations, including the Industrial Emissions Directive (IED) [5], will impose a requirement to considerably reduce emissions of pollutants, including nitrogen oxides. This has increased interest in the use of SNCR technology with OP-140 boilers. IED requires combustion plants with capacity rating exceeding 50 MWth to reduce their nitrogen oxides emission. The reduction obligation affects boilers with ratings from 15 MW upwards if they operate with a single emission stack in a group of at least 50 MWth. Depending on the type of source, the new standard will come into force in either 2016 or 2022. In Poland to date de-NOx systems are to be found exclusively at large commercial power plants. Boilers with outputs around 120 MWth have not been equipped with de-NOx systems.

Keywords:

CFD simulation; OP-140; Pulverized fuel fired boiler; SNCR; Spray system

Article history:

Received: 31 July 2015 Revised: 20 October 2015 Accepted: 09 November 2015

Selective Non-Catalytic Reduction technology for NOx emission abatement from industrial boilers involves injecting reducing an agent directly into a hightemperature zone of the combustion chamber. Reactions between the NH2 radical and nitrogen oxides contained in the flue gas yield nitrogen, carbon dioxide and water [9], [3], [13]. The radicals are generated through the thermal decomposition of reducing agents. The following

___________________________________________________________________________________________________________ K. Badyda, P. Krawczyk, S. Mlynarz: “Numerical Analysis of the Impact of Parameters of Urea Solution Injection on …”, pp. 27–34

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

substances are used in this capacity in the SNCR process[4]: • Urea (NH2)2CO in the form of an aqueous solution or solid, • Ammonia NH3 as a gas or aqueous solution NH4OH, • Cyanuric acid C3N3(OH)3 (cyclic trimer of the cyanic acid HNCO), • Ammonium salts (sulphate (NH4)2SO4, carbonate (NH4)2CO3), • Mono-, di-, trimethylamine CH3NH2, (CH3)NH, (CH3)N, • Hydrazine N2H2. To date, industrial applications have used urea and ammonia. The other compounds have been tested in laboratory reactors or at pilot plants, but so far have not been widely used on a larger scale and their application would require more research and development. Literature data [3], [9], [13] reveals that the optimal temperature ranges are 900-1150°C for urea and 8701100°C for ammonia. At lower temperature levels the reaction rates are low and ammonia is carried along unreacted with the exhaust gases causing ammonia slip. At higher temperatures the reagent oxidises, leading to additional NOx generation. Therefore efficiency of a NOx reduction process depends on parameters such as reaction temperature and type of reagent used, but also on the characteristic of the spraying nozzles, duration of contact between the reagent and the treated gas and the chemical composition of the treated gas. The method of introducing the reagent into the combustion chamber should be individually adapted for every boiler type, according to its design, size and combustion technology used (grate, pulverised bed). Correct installation of a Selective Non-Catalytic Reduction system reducing nitrogen oxides emission from an industrial boiler requires proper selection or design of injection equipment, taking into account properties defined by flow characteristics, spraying angle, droplet diameter, spray range, mutual interference between sprays from different nozzles and the influence of flue gas lift. Before injectors can be selected or designed, it is necessary to specify their desirable operating parameters, taking into account the local conditions of individual boilers (furnace geometry, flue gas velocity and temperature). The aim of this study was to determine optimal spray nozzle operating parameters, such as average reagent droplet diameter, initial droplet velocity and spray angle, to ensure best if not full coverage of the furnace's

cross-section for specified boiler geometry. Computational Fluid Dynamics methods and ANSYS Fluent Software [1] were used to achieve this aim. The geometrical and operational parameters used in the modelling process were sourced from actual boilers and they were as follows: • Furnace dimensions (width/depth/height): 6 m / 6 m / 27 m, • Air-fuel mixture injection velocity: 25 m/s, • Secondary air injection velocity: 35 m/s, • Total fuel mass rate: 3.27 kg/s, • Water walls temperature: 627 K. The simulation results provided information concerning the influence of nozzle operating parameters on reagent evaporation rates, and therefore also the spray range in realistic temperature conditions. They also made it possible to observe the influence of exhaust gas velocity on reagent droplet trajectory, depending on droplet size. Those relations are very important in the context of achieving the ultimate objective: ensuring reagent supply at optimal concentration to every point of a cross-section of the furnace.

2. Calculation methodology The following algorithm was used to achieve the defined aim of the study: • Performing thermal flow calculations for the processes occurring inside the furnace of a OP 140 boiler in order to pre-determine the desired temperature window and to find exhaust gas velocity and temperature distribution in that area. • Separating part of the furnace volume where the temperature conditions are favourable for reagent injection, together with assigning proper boundary conditions at relevant boundaries; • Running a number of reagent injection simulations in a domain which is a separated part of the furnace of the analysed boiler for different: • Average reagent droplet sizes, • Injection velocities, • Injection angles. The simulations were performed for a single injector feeding the reagent into the boiler. This approach was chosen, as the aim of the research was determining reagent behaviour (in the form of droplets and upon

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

evaporation) in the boiler rather than analysing NOx content reduction. Result analysis informed conclusions about the influence of injection parameters on the spray range and reagent concentration distribution in the boiler.

3. Simulation parameters The geometry used for analysing influence of injection parameters on reagent behaviour in the boiler was separated from the whole furnace volume; the part with temperature conditions for reagent feeding deemed favourable was selected. The separated geometry is marked in green in Figure 1. Relevant boundary conditions were specified for the side, top and bottom surfaces of the modelled volume (i.e. walls temperature, flue gas inlet and outlet respectively). The flue gas conditions involved: • Velocity distribution, • Pressure distribution, • Temperature distribution, • Turbulence distribution. These parameters were assigned following results of the earlier furnace modelling.

Figure 1: Computational geometry used to analyse the impact of injection parameters on reagent behaviour in the boiler volume (marked green)

A base grid consisting of ca. 140,000 box elements, with density additionally increased near the domain boundaries, was defined for the modelled volume. The density of the base grid was also increased in the area adjacent to the selected reagent injection location. Grid independence tests were performed, and following a series of preliminary calculation runs one variant – optimal in terms of calculation time and accuracy – was selected. The element count of the final grid was ca. 265,000 (Fig. 2). The settings used for solving the problems included activation of energy equation and discrete ordinates (DO) radiation model [11]. The Discrete Phase Model was used to model reagent injection [2]. The reagent is introduced into the furnace at high velocity, which may lead to secondary droplet disintegration due to aerodynamic forces. SSD Breakup Model was used to take this possibility into account [7]. The introduced droplets were considered as Multicomponent elements (mixture of urea and water). In order to determine the amount of reagent introduced by a single injection nozzle into the boiler it was necessary to make a number of assumptions. It was assumed that:

Figure 2: Computational grid used to analyse the impact of injection parameters on reagent behaviour in the boiler

• The average content of nitrogen oxides in the exhaust gas of an OP-140 boiler is 600 mg/Nm³, which upon conversion into dry gas and 6% oxygen content yields 447.8 ppm;

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

• Reagent would be injected through six injectors; • Amount of generated flue gas is ca. 170,000 kg/h. The general reaction between the urea and nitrogen oxide may be defined as follows [9]:

• Equal to the required stoichiometric concentration ensuring reduction to the assumed nitrogen oxides content of 600 mg/Nm³, • Two times higher than stoichiometric,

CO(NH2)2 + 2NO +0,5O2 = 2N2 + CO2 + 2H2O

• Four times higher than stoichiometric,

According to this equation, in stoichiometric conditions 1 mole of urea reacts with 2 moles of NO. As the molar mass of the urea is 60 g/mol and molar mass of NO is 30 g/mol, in stoichiometric conditions their masses are equal.

• Eight times higher than stoichiometric;

Therefore urea demand for an OP-140 boiler running at nominal load (assuming stoichiometric proportions) is ca. 0.0214 kg/s, i.e. ca. 77 kg/h. If six injectors are used to deliver a 15% urea solution, the reagent flow through single injector is ca. 0.0253 kg/s.

4. Calculation results Selected modelling results shown in this paper apply to a simulation of reagent injection into a volume which is a part of an OP-140 boiler furnace for various: • Reagent average droplet sizes – 100 μm, 200 μm and 500 μm; • Injection velocities – 50 m/s, 100 m/s and 200 m/s; • Injection angles – 20° and 40°.

• Exhaust gas volumes contained in the isosurfaces defined above were calculated. In reality, the urea decomposes into ammonia and cyanic acid [8], [12]. Observed isosurfaces of urea concentration should be treated as a visualization of gaseous reagent products. Analysis of shapes, placement and volumes of the flue gas contained in individual isosurfaces provides much useful information concerning urea behaviour (in the gaseous phase) in the boiler. It was assumed that the flue gas volume between the isosurfaces would be one of the tangible parameters describing the quality of injection for a certain set of parameters. It was assumed that the growth of volume contained in a specific isosurface should be interpreted as a higher potential of uniform urea supply to the flue gas contained in the furnace.

4.1. Influence of injection velocity on reagent behaviour in the boiler

• Changing droplet diameter as the penetration progresses.

Three values of injection velocity were considered in this study: 50 m/s, 100 m/s and 200 m/s. Modelling results indicate that the injection velocity has only limited impact on the reagent droplet range. Of course increasing injection velocity increases the range, but only until a certain point. At a certain velocity, the Weber number exceeds the critical value and aerodynamic forces acting on droplets cause secondary disintegration [6]. This leads to a rapid reduction in average droplet size in the stream and faster evaporation. This in turn reduces range. This phenomenon is presented in the diagram below (Figure 3).

Nevertheless, what is essential for the course of NOx reduction into N2 using urea is the urea behaviour after evaporation, i.e. in gaseous phase. Therefore it was decided to perform a detailed analysis of urea concentration distribution in the modelled volume for individual analysed injection scenarios. In order to do that:

It can be seen that for the 200 μm droplets the range at the injection velocity of 200 m/s is shorter than for velocity of 100 m/s. This effect is caused by secondary droplet disintegration. This phenomenon is not observed for 100 μm droplets, as for such a small value the Weber number does not exceed critical value in the analysed velocity range.

In order to evaluate the quality and usefulness of reagent injection for a certain set of injection parameters, it is necessary to observe reagent behavior both in liquid phase (urea solution droplets) and in gaseous phase (urea upon evaporation). The results shown below, which refer to the movement of urea solution droplets in the flue gas, specify: • Horizontal and vertical droplet penetration; • Droplet velocity;

• A number of isosurfaces was created, representing points of constant urea concentration for values,

As for droplet stream movement, the influence of injection velocity on droplet range is not great below the critical value of the Weber number.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Increasing velocity, just like increasing droplet size, moves the zone in which the reagent turns into gas further away. Accordingly, the gaseous reagent cloud appears further from the injection point (Figure 4).

4.2. Influence of droplet size on reagent behaviour Modelling results lead to the conclusion that the biggest factor affecting range is reagent droplet size. As is shown in Figure 5, reagent droplet size change roughly doubles if the diameter is increased from 100 to 200 microns. The figure below presents a comparison of reagent droplet diameter distribution in a stream for injection velocity of 50 m/s, spray angle of 20° and two different initial droplet diameters of 100 μm (Fig. 5a) and 200 μm (Figure 5b).

Considerable differences may be seen not only in droplet range, but also their lift. Larger diameter droplets, prior to full evaporation, slow down sufficiently to be lifted with the flue gas. This may be a harmful phenomenon, pushing the reaction zone outside the optimal temperature range. At smaller diameters, the droplet evaporation zone is placed lower. Additionally, this phenomenon intensifies along with increasing injection velocity, but only up to the value where secondary droplet disintegration starts. Figure 6 presents the behavior of reagent droplets in secondary disintegration conditions. Due to disintegration, droplet diameter drops rapidly, thus leading to quick reagent evaporation and considerable range reduction, regardless of initial injection parameters. Modelling also revealed (Figure 7) that 500 μm droplets are definitely too large for use in the analyzed boiler type (at injection velocities in the investigated range). Even at injection velocity of 100 m/s those droplets approach the opposite boiler walls. In such a case the evaporation zone is lifted to a level considerably above the injection level, thus placing the gaseous phase of the reagent beyond the desired flue gas temperature range. Figures 8 and 9 present exemplary droplet velocity and temperature distributions for the following injection conditions: initial velocity 50 m/s, initial droplet diameter 200 μm, and spray angle 20°.

Figure 3: Range of reagent droplets as a function of injection velocity and average droplet size – modelling results

The presented results show that the velocity of sprayed droplets drops quite quickly. When the velocity drops below a certain value, the droplets in their final life period – when their diameters are reduced in reference to the initial value – are lifted by the flue gas flow.

Figure 4: Isosurfaces of urea mass concentration in the flue gas for the droplet diameter of 100 μm and injection angle of 20° (a – 50 m/s; b – 100 m/s; c –200 m/s). View from the top of the boiler ___________________________________________________________________________________________________________ K. Badyda, P. Krawczyk, S. Mlynarz: “Numerical Analysis of the Impact of Parameters of Urea Solution Injection on …”, pp. 27–34

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 5. Distribution of reagent droplet diameters in the stream for injection velocity of 50 m/s and spray angle of 20° (a – injection of 100 μm droplets, b – 200 μm droplets)

Figure 6: Distribution of reagent droplet diameters in the stream for initial diameter of 200 μm, injection velocity 200 m/s and spray angle of 20°

Figure 8. Reagent droplet velocity distribution for initial velocity 50 m/s, initial droplet diameter 200 μm, and spray angle 20°

Figure 7: Reagent droplet diameters for injection velocity of 50 m/s, reagent concentration of 15%, spray angle of 40° and initial droplet diameter of 500 μm

Figure 9. Reagent droplet temperature distribution for initial velocity 50 m/s, initial droplet diameter 200 μm, and spray angle 20°

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 10. Distribution of reagent droplet diameters for injection velocity of 100 m/s, initial droplet diameter of 100 μm and injection angle of: a) 20°, b) 40°

The presented distribution of droplet temperature values shows three phases of reagent transition from the aqueous solution into gaseous form. The first short section of spray involves heating the droplets up to the water evaporation temperature. Then a constant temperature is maintained due to water evaporation. Upon complete water evaporation, urea temperature rises again to its evaporation temperature.

4.3. Influence of spray angle on reagent behaviour Another analysed parameter was the reagent spray angle. Two values were analysed: 20° and 40°. It was shown that increasing the spray angle causes a decrease in range for all analysed injection velocities and droplet diameters. Table 1 shows that, in addition to range, the width of the spray and droplet lift are also affected by the spray angle. If the spray angle is doubled, spray width and lift increase by several percent.

Table 1: Influence of spray angle on geometric spray parameters

Injection velocity [m/s]

Range 20°/40° [m]

Lift 20°/40° [m]

Width 20°/40° [m]

2.40/2.23

1.74/1.97

1.91/1.92

100

2.51/2.56

1.93/2.26

1.65/2.14

200

2.13./1.86

1.07/1.00

1.17/1.16

In contrast, spray angle was found to have no significant influence on the flue gas volume contained in individual isosurfaces of reagent concentration in the gaseous phase. Modelling results suggest that spray angle is not an essential parameter, but it does have a considerable impact on the behaviour of sprayed droplets or gaseous reagent (Figure 10).

5. Conclusion Having analysed a number of parameters describing reagent injection into the analysed boiler design, it may be concluded that the parameters which have an essential influence on the behaviour of reagent droplets in the boiler are primarily droplet diameter and velocity. The appropriate combinations of mentioned parameters result in wide spectrum of injection range. Spray angle is of secondary importance. According to the modelling results, to achieve a reagent droplet range sufficient to reach the middle of the analysed furnace does not require the use of droplets much larger than 200 μm. Injection velocity should be ultimately decided after verifying the correlation between the droplet size and velocity just after the injector, using an experimental facility. Analysis of the modelling results concerning reagent spread in the gaseous phase leads to the following conclusions: • Reagent behaviour is dependent less on injection parameters and more on the flue gas velocity profile in the furnace; • The volume of flue gas contained between consecutive isosurfaces of reagent

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Nuclear Engineering and Design. Volume 225, Issue 1, October 2003, Pages 37–48.

concentration is relatively insensitive to the analysed injection parameters; • The injection parameters which impact the location where the gaseous reagent appears are: droplet size and injection velocity. Analysing such a wide spectrum of injection scenarios provided knowledge and insight on potential reagent behaviour in the boiler, both in the form of droplets and in the gaseous phase.

[7]

Hai-Feng Liu, Xin Gong, Wei-Feng Li, Fu-Chen Wang, Zun-Hong Yu: Prediction of droplet size distribution in sprays of prefilming air-blast atomizers. Chemical Engineering Science. Volume 61, Issue 6, March 2006, Pages 1741– 1747.

[8]

Kawauchi S., Takagi M.: Study of Spray Distribution and Evaporation and Thermolysis Processes of Reductant in Urea SCR. 2011 AP4: Study of Spray Distribution and Evaporation and Thermolysis Processes of Reductant in Urea SCR.

[9]

Korupka J., Gostomczyk M.: Badanie selektywnej redukcji niekatalitycznej tlenków azotu. Ochrona środowiska (4) 63, 1995 str. 17 – 21.

[10]

Rozporządzenie Ministra Środowiska z dnia 22 kwietnia 2011 r. w sprawie standardów emisyjnych z instalacji. Dziennik Ustaw Nr 95 Poz. 558.

[11]

Stamnes K., S-Chee Tsay, Wiscombe W., Jayaweera K., Numerically stable algorithm for discrete-ordinate-method radiative transfer in multiple scattering and emitting layered media. Applied Optics, Vol. 27, Issue 12, pp. 2502-2509 (1988).

[12]

Tae Joong Wang, Seung Wook Baek, and Seung Yeol Lee: Experimental Investigation on Evaporation of Urea-Water-Solution Droplet for SCR Applications. American Institute of Chemical Engineers December 2009 Vol. 55, No. 12 (3267– 3275).

[13]

Thanh D. B. Nguyen, Young-Il Lim, Seong-Joon Kyung-Seun Yoo: Kim, Won-Hyeon Eom, Experiment and Computational Fluid Dynamics (CFD) Simulation of Urea-Based Selective Noncatalytic Reduction (SNCR) in a Pilot-Scale Flow Reactor. Energy & Fuels 2008, 22, 3864– 3876.

Acknowledgement This research is supported by the POIG.01.03.01-14035/12 project which is co-financed by the European Union under the European Regional Development Fund.

References [1]

ANSYS Fluent software - http://www.ansys.com/

[2]

ANSYS Incorporated, ANSYS FLUENT User’s Guide.

[3]

Birkhold F., Meingast U., Wassermann P., Deutschmann O.: Modeling and simulation of the injection of urea-water-solution for automotive SCR DeNOx-systems. Applied Catalysis B: Environmental 70 (2007) 119–127.

[4]

Brouwer, J., Heap, M. P., Pershing, D. W., & Smith, P. J., (1996), “A Model for Prediction of Selective Noncatalytic Reduction of Nitrogen Oxides by Ammonia, Urea, and Cyanuric Acid with Mixing Limitations in the Presence of Co”, Twenty-Sixth Symposium (International) on Combustion/The Combustion Institute, 21172124.

[5]

DIRECTIVE 2010/75/EU OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 24 November 2010 on industrial emissions.

[6]

Duan Ri-Qiang, Koshizuka S., Oka Y.: Twodimensional simulation of drop deformation and breakup at around the critical Weber number.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

DOI: 10.14621/ce.20150105

Assessment of the Space Heating and Cooling Market in the EU28: A Comparison between EU15 and EU13 Member States Simon Pezzutto1), Agne Toleikyte2), Matteo De Felice3) 1)

Institute for Renewable Energy, EURAC research Viale Druso 1, 39100 Bolzano, Italy, simon.pezzutto@eurac.edu 2) Energy Economics Group, Vienna University of Technology, Austria 3) Energy and Environment Modeling Unit, ENEA, Italy

Abstract

1. Introduction

The aim of this investigation is to fill in the knowledge gaps of the space heating and cooling market in Europe. While the space heating market is already well researched, there is a lack of information concerning air-conditioning. Following a bottom-up approach, data of the actual space heating and cooling market are collected and analysed. Next, air-conditioning market information was retrieved through a topdown approach and compared with bottom-up acquired data. Results indicate the energy consumption for space cooling in the EU28 to be almost 70 TWh/y, while the European Commission underestimates this value by approximately 60-70%. According to the findings from the bottom-up approach, the ratio between total potential and existing energy consumption is around 9:1 for air-conditioning and almost 1:1 for space heating. The EU15 is responsible for practically the entire space cooling consumption (~93%) of the EU28, with about 63 of the total 68 TWh/y. In contrast, the whole potential of recent member states comes out to be enormous, around four times its present energy consumption for air-conditioning. The space heating market is almost fully saturated, whilst the space cooling market in Europe is characterized by a high potential for growth, especially in households.

By 2020, the EU is aiming to decrease greenhouse gas emissions to 20% below 1990 levels. The energy produced by renewable energy sources (RES) is expected to be 20%. Additionally, a 20% upgrade of energy efficiency is foreseen [1]. The primary energy consumption in Europe amounts to around 1800 Mtoe/y in 2010. This is mainly caused by different types of heating and cooling applications (almost 900 Mtoe/y, not only related to building´ space heating and cooling but also industrial heat), followed by transportation and electricity with about 540 and 360 Mtoe/y respectively. Buildings account for around 720 Mtoe/y. The majority of energy consumption within the European building stock is attributed to space heating (SH) and space cooling (SC) purposes [2-7]. In recent years, the 28 EU member states invested a lot to quantify the energy consumption of the different sectors [8-11]. However, almost no data is available for the air-conditioning (AC) part. A proper investigation to determine this type of information has not been carried out yet.

Keywords:

Space heating; Air-conditioning, Potential, Market, Europe

Article history:

Received: 22 July 2015 Revised: 18 September 2015 Accepted: 09 November 2015

The majority of the data collection focused on the experience of preceding investigations. Especially, the “ARMINES - Mines de Paris/Mines Paristech Graduate School” was involved in a number of publications and projects to investigate the present topic, such as the EECCAC (Energy Efficiency and Certification of Central Air Conditioners) report [12]. A number of other relevant sources were used: the Intelligent Energy Europe project SOLAIR (Solar Air-Conditioning), the Sixth Framework Programme project POLYSMART (Polygeneration with advanced small and medium scale thermally driven air-conditioning and refrigeration technology) and the International Energy Agency project ANNEX 34 (Thermally Driven Heat Pumps for Heating and Cooling) [13-15].

___________________________________________________________________________________________________________ S. Pezzutto, A. Toleikyte, M. De Felice: “Assessment of the Space Heating and Cooling Market in the EU28 …”, pp. 35–48

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

The indicated exemplary data sources provide information for SH, SC and domestic hot water (DHW) preparation, related energy demand and consumption values and market figures within Europe. However, a clear picture of the AC market in the EU is missing. Domestic hot water preparation issues have been treated as well, because their needs are covered by the same kind of equipment as SH. Through a bottom-up approach, SH, SC and DHW preparation demand and consumption data has been analysed in the residential and service sectors of the EU15 and the remaining EU13 member states. Additionally, following a top-down approach, the EU28 AC consumption has been investigated and set against the respective bottom-up approach´s results. Principal complications faced during this study are: the fact that the terms energy demand and consumption are erroneously used interchangeably and only laboriously obtainable data are available for the European SC market. Chapter 2 explains in detail the information collection and analysis processes used. In chapter 3, the energy demand and consumption in the residential and service sectors has been subdivided into the SH, AC and DHW preparation categories. Due to the disparate information availability between the EU15 and EU13 nations, these two agglomerations has been analysed separately in a first step: within section 3.1 and 3.2 respectively. Comparison, discussion and conclusions regarding energy consumption for all three fields (SH, SC and DHW preparation) of the EU28 residential and service sectors are provided as well in chapter 4.

2. Methodology The data regarding specific energy demand and consumption for SH, AC and DHW preparation (kWh/m² y) was collected, divided by member states (EU28) and ordered within the households and service sectors. Energy for appliances was excluded by the survey (e.g. cooking). Concerning the collected information, it is important to distinguish between energy demand and consumption. The energy demand is the net energy requirement to cover SH and SC needs. In contrast, the energy consumption represents the energy input at the devices required to satisfy the demand. As such, the two quantities differ by disparate conversion factors [16]. Therefore, concerning SH and DHW preparation, because the efficiency of boilers is < 1 (0.8-0.9 for currently installed technologies in Europe), the energy consumption data is higher than the energy demand.

The energy efficiency ratio (EER) for electrically driven SC equipment is > 1 (around 2-3 for currently installed technologies within the EU). Because of that, the energy consumption for AC is lower than the SC demand [17]. It has to be mentioned that only the comparison of SH and SC demand is correct, while electricity consumption (in heat pumps and air-conditioners) can only be fairly compared with fuel consumption (e.g. gas in a gas boiler) if an adequate conversion into primary energy terms is performed. This is because the two energy carriers have a different content of grey energy when employed by the final consumption. The primary energy (usually expressed in terms kWh or toe) accounts for the consumption of fossil resources, providing a basis for clear comparison of different energy carriers. The heated and cooled floor area, as well as the whole floor area in the residential and service sectors, was identified for the different EU member states. In the case of graphs shown with a unit of kWh/m² y (figures 1, 3, 6 and 8) the average line is obtained by weighting the mean of the single nations´ energy demand and consumption on the heated or cooled floor area of the respective country. In the case of DHW preparation the entire floor area is taken into consideration (figures 1 and 6). In the charts shown with a unit of kWh/inhabitant y (figures 2 and 7), the average line is obtained simply by calculating the mean of the energy demand and consumption values for the different EU15 and EU13 countries. The following values with a unit of kWh/inhabitant y (residential sector) or kWh/employee y (service sector) mainly indicate the specific energy use habits of the inhabitants or employees in the various EU15 and EU13 member states. The columns given in kWh/inhabitant y have been calculated by dividing the energy demand or consumption per application type (SH, SC or DHW preparation) in TWh/y by their respective amount of occupants within the households sector (figures 2 and 7). In figures 4, 5, 9 and 10, the columns of total SH and SC per country in TWh/y have been obtained by multiplying the average energy demand and consumption per country in kWh/m² y with the respective heated or cooled floor area in million square meter (Mm²). These show the related distribution of the energy demand and consumption among the EU15 and EU13 nations. The potential SH and SC demand as well as consumption were identified using the same calculations, with total floor area instead of conditioned floor area. Therefore, the technical potential has been calculated. Following values regarding DHW preparation demand and consumption per member state in TWh/y have been calculated by multiplying the average DHW preparation

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

demand and consumption per country in kWh/m² y with the respective entire households or service sector floor area of each country in Mm². For the SH and DHW preparation part, sufficient information is available, whilst there is a lack of data concerning SC. On the other hand, information with regard to the energy consumption for AC in the EU is difficult to collect. At the moment, a huge amount of data concerning the AC market in Europe is based on estimations [18-20]. Not all collected information has been used to form the statistics. Data which lie outside a range of plus or minus one standard deviation around the average of the respective data pool have been discarded. Due to the impossibility of creating complete energy statistics by collecting climate corrected information, this type of data has been excluded by the investigation. Values characterized by a reference year more than a decade ago have not been taken into consideration. Specifically, the data used to create the following figures and tables covers the period of 2005-2015. In the following statistical figures, with a unit of kWh/m² y, the numbers straight over the top of the columns indicate the amount of information utilized to calculate the values for each column, the error bars show their standard deviation and the percentages above their coefficient of variation (CV). In the case of charts with a unit of TWh/y, the percentages at the top of the columns indicate the CV of the data used to form the respective columns and the error bars represent their standard deviation. Within chapter 4 of the present investigation, solely energy consumption and not energy demand was considered. Main reason for this is the shortage of energy demand information for the EU13, making it impossible to compare the EU15 and EU13 member states. For the EU13 countries solely energy consumption and not energy demand was utilized. All available energy consumption data is considered and in the case, only energy demand values are available, these have been transformed into energy consumption. Space heating and DHW preparation demand information was transformed into energy consumption by dividing with a value of 0.85 (average efficiency of currently installed boilers within the EU = 0.85) [17]. Space cooling demand data was transformed into AC consumption by dividing with a value of 2.7 (actual average seasonal energy efficiency ratio in Europe = 2.7) [12], [20-24]. The seasonal energy efficiency ratio (SEER), not EER, was used for this calculation, because it reflects the real consumption conditions of the AC equipment for the SC season over a year. In contrast, the EER is measured under rated load conditions [12], [25-31]. The SEER indicates the whole heat amount removed (Q) from the

conditioned space during the entire annual AC season, divided by the entire work input (W - electricity) of the SC machinery during the same season [25]. See Eq. (1). SEER = Qcold, season/Welectricity, season

(1)

Due to the aforementioned shortage of information, the whole energy consumption indication, given in TWh/y, for the EU28 service sector includes solely offices for the EU13. However, EU office buildings have the highest entire SC energy consumption ratio of the European building stock [32-34]. With regard to the EU13 nations, no calculations with a unit of kWh/employee y were performed, because it is not meaningful to determine the habits of the work active population per EU13 member state, while considering only the offices instead of the whole service sector. If the whole energy indication of the various EU15 and EU13 sector could not be calculated, due to a lack of data (e.g. entire SH consumption of the EU13 office sector), related values were excluded from final calculations.

3. Results 3.1. Space heating, space cooling and domestic hot water preparation demand and consumption of the residential and service sectors (EU15) Figure 1 shows the relation between SH and DHW preparation demand of the residential sector in kWh/m² y between the different EU15 nations.

The average demand for SH and DHW preparation in the residential sector, comes out to be about 147 and 21 kWh/m² y respectively. This results in a proportion of about 7:1. The respective SH consumption appears to be just slightly higher: 156 kWh/m² y. Figure 1 shows a certain homogeneity regarding the SH demand per country. Member states with colder climates demonstrate more SH demand than nations with warmer climates. Of all states, Luxembourg has the highest value. It also is characterized by low energy prices and the highest purchasing power per person [38-39]. The CV percentages shown in Figure 1 demonstrate that the selected data to form the bars are rather similar. The average value is around 18%. Regarding the DHW preparation, Finland returns the highest number (53 kWh/m² y). The Netherlands also shows a huge value (37 kWh/m² y). One possible

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 1: Space heating and domestic hot water preparation demand per country, residential sector, kWh/m² y [4], [12-15], [26-29], [35-37]

Figure 2: Space heating and domestic hot water preparation demand per country, residential sector, kWh/inhabitant y [4], [12-15], [26-29], [35-37]

Figure 3: Space heating and cooling demand per country, residential sector, kWh/m² y [4], [12-15], [26-29], [35-46]

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

explanation for the high energy demand by the Dutch, is a high utilization of electricity for DHW preparation [40].

countries. This also applies to the values for the tertiary sector within Western Europe.

Analysing the data published within the past decade shows a reduction in specific residential SH and DHW preparation for all EU15 countries. The same applies to respective data within the service sector of the old EU member states.

If the SH and SC demand of the residential sector are compared between the different EU15 member states, with a unit of TWh/y, an even larger proportion between these emerges ~78:1 (about 1863 TWh/y of SH demand and around 24 TWh/y for SC).

The energy demand for SH in the residential sector of the EU15 is around 1863 TWh/y. In comparison, the energy demand for DHW preparation of the same sector is about 331 TWh/y. Hence, a relation of approximately 6:1 emerges.

If the energy use habits of the different EU15 citizens are compared regarding the SH and SC demand again a wide gap results.

If the energy use habits of the different EU15 citizens are compared concerning the SH and DHW preparation demand, an even wider gap results. See Figure 2. The mean value concerning the energy demand for SH in the residential sector is about 4863 kWh/inhabitant y. If this value is compared to the average number of the energy demand for DHW preparation of the residential sector (~932 kWh/inhabitant y), the number for SH is more than five times higher than that for DHW preparation. The highest SH demand value in Figure 2 is again given by Luxembourg. Moreover, the bar for France appears relatively smaller in Figure 2 compared to Figure 1 for the same kind of application. French dwellings are typically smaller than the EU15 mean; France has about 89 m² average floor area per apartment and the EU15 mean per living unit is around 93 m² [41]. The respective value within the service sector amounts to approximately 4120 kWh/employee y. If this number is compared with the corresponding average value of the residential sector a reduction of ~15% in the service sector is recognizable. The focus goes to the comparison between SH and SC demand in kWh/m² y. See Figure 3. The SH and SC demand of the residential sector, results to be 147 and 41 kWh/m² y respectively. The resulting proportion is around 3.5:1. In comparison to SH demand, the scenario for SC changes completely. The countries located in warmer regions of the EU15, show a higher demand for SC application and the other way around. Surprisingly Luxembourg, United Kingdom and Belgium show relatively high values for AC application. These reach almost the number of France. For SC demand, the average CV percentage is quite small, with an average of about 12%. In contrast to SH, when analysing the different data published within the past ten years, an increase in the residential specific AC demand is evident in all EU15

Compared to the average value of SH demand purposes in the residential sector (about 4844 kWh/inhabitant y) the respective energy demand for AC is around 93 times lower. The focus goes to a comparison between the SH and potential SH demand and SC and potential SC demand per member state in the residential sector of the EU15, with a unit of TWh/y. See Figure 4 and 5. If the differences between the energy demand and potential energy demand are compared in the two graphs above, the mismatch results to be higher at the AC sector: 422 TWh/y for the SH and 563 TWh/y for the SC sector. Especially the proportion between the total SC and entire potential SC demand is much larger than the related SH indications: around 1:1.23 for the SH and 1:24 for the SC sector. If the same comparison is carried out for the service sector, the difference results to be higher at the SC sector again: for around 299 TWh/y. Especially the proportion between the whole AC and entire potential SC demand in the service sector is much larger than at the SH sector, about three times more. Hence, in total (residential and service sector) the potential AC demand results to be higher than the potential SH demand in the same sector for a value around 440 TWh/y. The whole energy demand (residential and service sector) of the total SH, SC and DHW preparation has to be taken into account as well. The highest position is held by SH with approximately 2514 TWh/y, followed by DHW preparation with around 331 TWh/y and AC with about 149 TWh/y. Thus, the SH is almost eight times larger than the DHW preparation and nearly 17 times higher than the SC portion. As already mentioned above, the SC sector shows the largest whole energy demand potential of these. Tables 1 and 2 summarize and complete the actual and potential values of energy demand and consumption for SH, AC and DHW preparation in the residential and service sectors of the EU15.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 4: Space heating and potential space heating demand per country, residential sector, TWh/y (as not visible in the graph above: space heating demand and potential space heating demand of Luxembourg = 4 and 6 TWh/y respectively) [4], [12-15], [26-29], [35-37]

Figure 5: Space cooling and potential space cooling demand per country, residential sector, TWh/y (as not visible in the graph above: space cooling demand of Austria, Belgium, Denmark, Finland, Ireland, Luxembourg, Netherlands, Portugal and Sweden = 0.3, 0.3, 0.1, 0.2, 0.05, 0.006, 0.4, 0.9 and 0.2 TWh/y respectively; potential space cooling demand of Germany and Italy: 120 TWh/y both) [4], [12-15], [35-37]

Table 1: Space heating, space cooling and domestic hot water demand and consumption, bottom-up approach, actual values, residential and service sectors (EU15), TWh/y [4], [12-15], [35-37] Bottom-up approach results, actual values (EU15), TWh/y Residential sector Service sector Demand Consumption Demand Consumption Space heating 1863 1982 651 NA Space cooling 24 17 125 NA Domestic hot water preparation 331 409 NA NA

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Table 2: Space heating, space cooling and domestic hot water demand and consumption, bottom-up approach, potential values, residential and service sectors (EU15), TWh/y [4], [12-15], [35-37] Bottom-up approach results, potential values (EU15), TWh/y Residential sector Service sector Demand Consumption Demand NA Space heating 2285 2421 651 NA Space cooling 587 425 424 NA Domestic hot water preparation 331 409 NA NA

Figure 6: Space heating and domestic hot water preparation consumption per country, residential sector, kWh/m² y [4], [11-15], [37]

Like shown in Table 1 and 2, the potential of the EU15 member states is enormous, exceeding its present energy demand for AC applications by seven times. The ratio between the potential and present energy demand in the residential sector is more than 20 times that ratio for the service sector.

As was already the case for the EU15, the specific SH and DHW preparation consumption data in the residential sector show a reduction for all EU13 counties during the past decade. The same applies to respective data within the service sector within the new European member states.

The mean consumption for SH and DHW preparation purposes in the residential sector, results in about 160 and 33 kWh/m² y respectively. This is a proportion of about 5:1.

As already mentioned, the energy consumption for SH in the households sector of the EU13 is around 438 TWh/y. In comparison to that, the energy consumption for DHW preparation of the same sector is almost 92 TWh/y. Hence, the relation between SH and DHW preparation is approximately 5:1 again.

With regard to SH consumption in Figure 6, again a certain homogeneity per country is given. Furthermore, the countries in warmer climates clearly show lower SH consumption values. The highest value for SH consumption is given by Latvia, and the lowest one by Malta with approximately 215 and 19 kWh/m² y respectively. Possible reasons for Latvia´s high value include cold climatic conditions, and the deterioration of its building stock [47-48]. Moreover, Malta´s low value is a result of its warm climate [21], [4751]. From the average CV percentage shown for SH consumption in Figure 6, it is visible that the selected data to form these bars is very similar. The average CV is around 9%.

If the energy use habits of the different EU13 citizens are compared regarding the SH and DHW preparation consumption, once more a wide gap emerges. See Figure 7. The average value concerning the energy consumption for SH in the residential sector of the EU13 member states is about 3787 kWh/inhabitant y. If this value is compared to the mean value of the energy consumption for DHW preparation consumption of the same sector (~942 kWh/inhabitant y), the number for SH is more than four times larger than that one for DHW preparation.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 7: Space heating and domestic hot water preparation consumption per country, residential sector, kWh/inhabitant y [4], [11-15], [37]

Figure 8: Space heating and cooling consumption per country, residential sector, kWh/m² y (as not visible in the graph above: space cooling consumption of Lithuania = 0.4 kWh/m² y) [4], [11-15], [21], [27], [37]

Attention now turns to the comparison between SH and SC consumption in kWh/m² y. See Figure 8. The average SH and SC consumption of the residential sector, results to be 160 and 10 kWh/m² y respectively. The resulting proportion is enormous: ~16 to 1. The situation for SC is completely different than the scenario for the SH consumption. The countries in warmer climates show higher SC consumption, and the countries in cooler climates demonstrate lower SC consumption. There are three exceptions: Estonia, Latvia and Hungary. Those three EU13 member states are located in rather cold climatic zones of Europe, and show SC values exceeding the respective weighted average given in Figure 8 [49-50]. In the case of SC consumption, the average CV percentage is quite small, with an average of 15%.

Like seen in the EU15, the specific SC consumption data in the residential sector of the EU13 member states registers an increase over the past ten years. This statement is also valid for the respective information concerning the service sector of the new EU member states. If the SH and SC consumption of the residential sector are compared between the different EU13 member states with a unit of TWh/y an even larger proportion between these emerges: 438:1 (about 438 TWh/y of SH consumption and around 1 TWh/y for SC). If the energy use habits of the different EU13 citizens are compared regarding the entire SH and SC consumption again a wide gap comes out. Compared to the mean value of SH consumption purposes in the residential sector (about 3787 kWh/inhabitant y) the energy consumption for AC in the

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

residential sector in kWh/inhabitant y results to be around 51 times lower. The focus goes to a comparison between the SH and potential SH consumption and SC and potential SC consumption per country in the households sector of the EU13, with a unit of TWh/y (Figure 9 and Figure 10). If the difference between the energy consumption and potential energy consumption are compared in the two graphs above, the resulting mismatch is higher for the SC sector: 5 TWh/y for the SH and 14 TWh/y for the SC part. Especially the proportion between the whole SC and total potential SC consumption is much larger than that one of the SH sector: around 1:1.01 for the SH and 1:15 for the SC sector. Hence, the potential AC consumption of the residential sector within the EU13 results to be higher than the potential SH consumption in the same sector for a value of around 9 TWh/y. The entire SH, SC and DHW preparation consumption in the residential and office sectors of the EU13 has to be

taken into account too. The highest position is again kept by SH with approximately 438 TWh/y, followed by DHW preparation with around 92 TWh/y and AC with about 5 TWh/y. Thus, a relation of almost five times is given between SH and DHW preparation and almost 88 times between SH and SC. As mentioned above, like it was already the case within the EU15, also here the SC sector shows the highest total energy consumption potential of these. Tables 3 and 4 summarize and complete the actual and potential values of energy consumption for SH, AC and DHW preparation in the households and office sectors of the EU13. As shown in Table 3 and 4, the potential of recent member states is enormous, exceeding its present energy consumption for AC applications by approximately four times. The ratio between the potential and actual energy consumption in the residential sector is more than 13 times that ratio for the office sector.

Figure 9: Space cooling and potential space cooling consumption per country, residential sector, TWh/y (as not visible in the graph above: space cooling consumption of Croatia, Czech Republic, Estonia, Latvia, Lithuania, Poland, Romania and Slovakia = 0.01, 0.02, 0.002, 0.01, 0.001, 0.002, 0.01 and 0.1, TWh/y respectively) [4], [11-15], [21], [27], [37]

Table 3: Space heating, space cooling and domestic hot water consumption, bottom-up approach, actual values, residential and office sectors (EU13), TWh/y [4], [11-15], [21], [27], [37-53] Bottom-up approach results, actual values (EU13), TWh/y Residential sector Office sector Consumption Space heating 438 NA Space cooling 1 4 Domestic hot water preparation 92 NA ___________________________________________________________________________________________________________ S. Pezzutto, A. Toleikyte, M. De Felice: “Assessment of the Space Heating and Cooling Market in the EU28 …”, pp. 35–48

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Table 4: Space heating, space cooling and domestic hot water consumption, bottom-up approach, potential values, residential and office sectors (EU13), TWh/y [4], [11-15], [21], [27], [37-53] Bottom-up approach results, potential values (EU13), TWh/y Residential sector Office sector Consumption Space heating 443 NA Space cooling 15 6 Domestic hot water preparation 92 NA

4. Discussion and conclusions Space cooling accounts for an important part of the European Union households’ energy consumption and is particularly meaningful for its service sector. According to the bottom-up approach´s results, the ratio between total potential and existing entire energy consumption for air-conditioning is ~9:1. The potential energy consumption for space cooling exceeds the respective actual value by an enormous number of approximately 535 TWh/y, which corresponds to almost eight times the actual air-conditioning energy consumption. European space cooling consumption quantification, following a bottom-up approach, shows a value of around 68 TWh/y. The mentioned air-conditioning indication significantly exceeds the European space cooling market information acquired by top-down values from the European Commission (2009 and 2012 for both last named sources). It estimates between 19 and 39 TWh/y [34], [43], [53]. Various documents confirm the range provided by the European Commission. For example, Kranzl et al. 2014 name the European Union air-conditioning consumption to reach a value of almost 30 TWh/y [54]. In contrast, a number of further top-down gathered data indicate the space cooling branch in Europe to be around 55 and 88 TWh/y: e.g. Marinhas 2010, Constantinescu et al. 2006 as well as Weiss and Biermayr 2009 respectively [55-57]. Other indications, collected by the top-down approach, show the European Union air-conditioning market to be significantly larger, exceeding the last mentioned space cooling consumption information for about two to three times. For example, Boermans T. et al. 2015 indicates the air-conditioning consumption in Europe to be approximately 130 TWh/y [58]. Kalz and Pfafferott 2014 quantify the space cooling consumption in the European Union by almost the same size and Sanner et al. 2011 characterize the air-conditioning consumption in Europe by more than 170 TWh/y [59-60]. Dalin 2006 estimates a number of almost 250 TWh/y [18].

If the entire space cooling consumption is compared between the EU15 and EU13 member states, interesting facts emerge. The EU15 is responsible for practically the whole air-conditioning consumption (~93%) of the EU28, with about 63 and 68 TWh/y respectively. In contrast, a number of sources indicate the EU15 airconditioning market to be around 80% of the total European one [1], [2], [12]. For the EU13 service sector only the offices portion was analysed. However, offices are characterized by the highest entire space cooling consumption of the European service sector [32-34]. Europe´s specific air-conditioning consumption (residential and service sectors) has constantly increased during the past decade. A huge discrepancy emerges if the space cooling consumption is compared between the EU15 and EU13 member states in the households sector. The EU15 air-conditioning consumption accounts for about 30 kWh/m² y, while the respective value in the EU13 is approximately 10 kWh/m² y. If the residential space cooling consumption is compared between the EU15 and EU13 an enormous mismatch comes out, with 17 and 1 TWh/y respectively. Principal reason for this is the minor diffusion of airconditioning applications within the EU13. Surprisingly, Cyprus shows the highest residential space cooling consumption value of the EU13 with about 0.3 TWh/y and is highly ranked within the EU28 as well - in fourth position together with Portugal, which has about ten times more inhabitants [18]. Primary reason for that is the high percentage of cooled floor area in Cyprus´ households: more than 74% [26]. The elevated specific value for air-conditioning purposes in Cyprus contributes to that point as well. Specific space heating consumption (residential and service sectors) is characterized by a constant decrease during the past ten years in Europe. Interesting facts emerge if the specific space heating consumption are compared between the EU15 and EU13 member states in the residential sector. There are very similar weighted averages, with 156 and 160 kWh/m² y for the EU15 and EU13 respectively.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

In Europe, the entire potential energy consumption for space heating purposes is almost equal to its actual one: around 1.2:1. The European Union´s energy consumption is mainly marked by space heating with a value of at least 3169 TWh/y. The EU15 residential space heating consumption exceeds that one of the EU13 by about 1544 TWh/y, with around 1982 and 438 TWh/y respectively. A wide gap appears, when the specific space heating use habits of the different EU15 and EU13 citizen are compared in households. Approximately 5457 kWh/inhabitant y for the EU15 and 3787 kWh/inhabitant y for the EU13 space heating consumption are revealed. Main reason for that is the lower economic availability in the EU13 countries [39]. The average space heating consumption of the EU28 member states is approximately 4681 kWh/inhabitant y. Indication for domestic hot water preparation consumption differs with 26 kWh/m² y for the EU15 and 33 kWh/m² y for the EU13. Prime reasons for this are relatively colder climatic conditions in the EU13 and more energy efficient boilers allocated in the EU15 [12], [24], [31], [61-66]. The same comparison regarding domestic hot water preparation consumption values shows a mismatch of around 317 TWh/y, with approximately 409 TWh/y and 92 TWh/y for the EU15 and EU13 respectively. The ratio of potential and actual domestic hot water preparation consumption is 1:1 in Europe with 501 TWh/y each, as it has been assumed that the consumption of domestic hot water preparation is given for the whole residential and service floor area. With regard to the habits of occupants and their impacts on domestic hot water preparation consumption, a difference of around 13% emerges with around 1079 and 942 kWh/inhabitant y for the EU15 and EU13 respectively. This higher domestic hot water consumption for the EU15 seems to be in contradiction with its more efficient boilers mentioned above. A possible reason for the greater energy consumption of domestic hot water preparation within the old European Union member states could be a higher amount of litres domestic hot water utilized within the households of Western Europe. One member state has to be pointed out concerning the analysed market of the EU13: Poland. Poland´s position within the EU13 has a number of similarities to that of Germany within the EU15. Concerning energy consumption, Germany and Poland result in having the highest values in the absolute majority of cases, with usually approximately 1/3 of the entire energy consumption per type (space heating, air-conditioning and domestic hot water preparation). The economic situation of a country has an influence on its whole

amount of energy consumption as well. In terms of gross domestic product, Poland and Germany are the largest economies within their respective European countries´ agglomerations [18], [39]. Regarding the total energy consumption (residential and service sectors) of space heating, space cooling and domestic hot water preparation within the entire EU28, the highest position is held by space heating with approximately 3169 TWh/y, followed by domestic hot water preparation with around 501 TWh/y and airconditioning (68 TWh/y). Thus, a relation of approximately six times is given between space heating and domestic hot water preparation and around 46 times between space heating and space cooling. Not all collected information appear to be trustworthy even if the mentioned data has been retrieved from reliable sources solely. For the space heating and domestic hot water preparation part, sufficient information is available; for the air-conditioning one, a major lack of information exists. The space heating market is practically fully saturated, while the space cooling market is characterized by a huge potential, especially in the households sector. For the EU15, much more data has been found than for the EU13. This especially concerns the energy consumption for air-conditioning purposes. In particular, the information collected for the EU13 cooled floor area was only laboriously obtainable and its reliability has to be further investigated. Thus, it is recommended to analyse the European space cooling consumption and related cooled floor area in more detail, with major focus on tertiary buildings and particular attention to the EU13. The energy efficiency of buildings as well as the different building typologies have a major influence especially on space heating and cooling consumption in the residential and service sectors. Because of this fact, it has been decided to extend the carried out investigation, taking into account the differences of the European building stock. Another improvement suggestion regards the comparison of European Union countries´ building energy demand for space heating, space cooling and domestic hot water preparation without the influence of the climate condition - calculating climate corrected energy demand values. This transformation could be based on the specific energy demand of the single nations in the European Union. Mean heating degree days of the single European counties (or regions) and average heating degree days from all 28 European Union member states might be gathered from Eurostat statistics for the time range of 2000 to 2009 [67]. By doing so, the performance of the building stock allocated within the different European Union nations

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

can be differentiated, considering the specific climate conditions.

References

It would be interesting to find out to what extent the airconditioning market in Europe is of importance concerning employment, its contribution to the European Union´s gross domestic product and how this market evolves, once the economic crisis is over.

[1]

EC, The 2020 climate and energy package, http://ec.europa.eu/clima/policies/package/ind ex_en.htm

[2]

Benejam, G.M. et al., Bottom-up characterization of the Spanish building stock for energy assessment and model validation, 2012.

[3]

Usher, W., Service Sector Module, https://wiki.ucl.ac.uk/display/EnergInst/Service+ Sector+Module#ServiceSectorModuleEnergyservicedemand

[4]

BPIE, DATA SEARCH, http://www.buildingdata.eu/data-search/results

[5]

Economidou, M., Energy performance of the existing building stock, World Sustainable Energy Days 2012, Wels, Austria, 2012.

[6]

Bossedouf, D., Energy Efficiency Trends in Building in the EU, ADEME, 2012.

[7]

Aebischer, B., Impact of climate change on thermal comfort, heating and cooling energy demand in Europe, 2007.

[8]

EC, FP7, http://cordis.europa.eu/fetch?CALLER=FP7_ENE RGY_PROJ_EN&DOC=51&QUERY=0140cf455228 :5934:27badc5

[9]

EC, RESEARCH & INNOVATION, http://ec.europa.eu/research/index.cfm

[10]

EC, Projects, https://ec.europa.eu/research/energy/eu/index _en.cfm?pg=projects&fp7page=10

[11]

EEA, Energy consumption, http://www.eea.europa.eu/data-andmaps/indicators/energy-efficiency-and-energyconsumption 5/assessment

[12]

Adnot, J. et al., Energy Efficiency and Certification of central Air Conditioners (EECCAC), ARMINES, Paris, France, 2003.

[13]

Melograno, P.N., Pezzutto, S., IEA ANNEX 34 Italian country report, EURAC research, Bolzano, Italy, 2010.

[14]

Villa, E., Steege, A., Market Report for Small and Medium - Sized Solar Air-conditioning Appliances, target GmbH, Hannover, Germany, 2008.

[15]

Aprile, M., The market potential of micro-CHP, University of Salento, Lecce, Italy, 2004.

Acknowledgement Our gratitude goes to Prof. Herbert Braun and Prof. Herbert Weingartmann (University of Natural Resources and Life Sciences, Vienna) for the scientific supervision of the present investigation. We are thankful to Prof. Tobias Pröll (University of Natural Resources and Life Sciences, Vienna) and Dr. Raphael Bointner (Vienna University of Technology) for the corrections of the present study. Finally, we would like to bring our deepest appreciation to Dr. Ulrich Oberegger Filippi (EURAC research) for the calculations carried out together.

Nomenclature AC CV DHW EER E.g. kWh M m² NA SEER SC SH RES Toe TWh W Q y

Air-conditioning Coefficient of variation Domestic hot water Energy efficiency ratio Example given Kilowatt hours Million Square meter Not available Seasonal energy efficiency ratio Space cooling Space heating Renewable energy sources Tonnes of oil equivalent Terawatt hours Work Heat flux Year

Synonyms Actual Air-conditioning EU15 EU13 Residential sector Service sector Total

Real, present Space cooling Old European Union member states, Western Europe New European Union member states Households sector Tertiary sector Whole, entire

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

[16]

Capehart, B.L., Encyclopedia of Energy Engineering and Technology, CRC Press, London, England, 2007.

[17]

[18]

[31]

Recknagel, R. et al., Handbook for Heating and Air-conditioning Technology, Oldenbourg, Industry publishing, Munich, Germany, 2010.

Riviere, P. et al., Sustainable Industrial Policy Building on the Ecodesign Directive - Energy Using Products Group Analysis/2, ARMINES, 2012.

[32]

Dalin, P. et al., The European Cold Market, Final Report, WP 2, Euroheat & Power, Brussels, Belgium, 2006.

Parker, S.P., Mc Graw-Hill dictionary of scientific and technical terms, Mc Graw-Hill, Saint Paul, USA, 2003

[33]

Staller H., Stiller A., New technical solutions for energy efficiency building, SCI-Network, 2011.

[34]

JRC, EU energy efficiency measures contribute to stabilize electricity consumption - drop in domestic consumption, JRC, Brussels, Belgium, 2009.

[35]

Pérez-Lombard, L. et al., A review on building energy consumption information, Energy and Building, Volume 40, (2007), Issue 3, pp. 394-398, doi:10.1016/j.enbuild.2007.03.007.

[36]

Dalin, P. et al., The European Heat Market, Final Report, WP1, Euroheat & Power, Brussels, Belgium, 2006.

[37]

Swedish National Board of Housing Building Planning, Housing Statistics of the European Union, Swedish National Board of Housing Building Planning, Karlskrona, Sweden, 2004.

[38]

Ministry of Economy Luxembourg, Luxembourg action plan for renewable energy, 2009.

[39]

EC, GDP per capita in PPS, http://ec.europa.eu/eurostat/tgm/table.do?tab =table&init=1&language=en&pcode=tec00114& plugin=1.

[40]

Oxford University, Country picture: Netherlands, 2002.

[41]

MINISTRY OF REGIONAL DEVELOPMENT CZ, Average useful floor area per dwelling and per person, 2009.

[42]

Balaras, C.A., European residential building and empirical assessment of the Hellenic building stock, energy consumption, emissions and potential energy savings, Building and Environment, Volume 42, (2007), Issue 3, pp. 1298-1314, doi:10.1016/j.buildenv.2005.11.001.

[19]

EREC, Renewable Heating & Cooling, http://www.erec.org/policy/sectoralpolicy/heating-cooling.html

[20]

EEA, Energy efficiency and energy consumption in the household sector (ENER 022) - Assessment published Apr 2012, http://www.eea.europa.eu/data-andmaps/indicators/energy-efficiency-and-energyconsumption-5/assessment

[21]

Riviere, P. et al., Preparatory study on the environmental performance of residential room conditioning appliances (airco and ventilation), ARMINES, Paris, France, 2008.

[22]

Carreno, A.M., COOLING BENCHMARKING STUDY REPORT. ECONOLER, Montreal, Canada 2007.

[23]

Bettenhauser, K. et al., Climate Protection trough Reducing Building´ Space Cooling Energy Demand, German Federal Environmental Agency, Berlin, Germany, 2011.

[24]

Riviere, P. et al., A METHOD TO REDUCE EUROPEAN CHILLER HOURLY LOAD CURVES TO A FEW POINTS, 2005.

[25]

ANSI/AHRI, Performance Rating of Unitary Airconditioning & Air-Source Heat pump Equipment, ANSI/AHRI, Arlington, USA, 2012.

[26]

Skytte, K., Ropenus, S., Regulatory review and international comparison of EU-15 member states, 2006.

[27]

World Bank, Total Population (in number of people), http://data.worldbank.org/indicator/SP.POP.TO TL

[28]

Veit, B., Policy development for improving RESH/C penetration in European member States, 2009.

[43]

UK Government, Population by Country of Birth and Nationality, http://data.gov.uk/dataset/population_by_coun try_of_birth_and_nationality

EC, Employment statistics, http://epp.eurostat.ec.europa.eu/statistics_expl ained/index.php/Employment_statistics

[44]

McNiel MA, Letschert VE, Future air conditioning energy consumption in developing countries and what can be done about it: the potential of efficiency in the residential sector, ECEEE 2007 SUMMER STUDY, 2007, pp. 1311-1322.

[29]

[30]

Pout C., Hitchin E.R., The future environmental impact of room air conditioners in Europe, 2008.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

[45]

IEC, Leaner and cleaner heating and cooling, http://www.iec.ch/etech/2012/etech_0712/tec h-1.htm

[46]

Dupont, M., Adnot, J., Inspection and auditing of air-conditioning facilities in Europe - A new efficiency target, 2005.

[47]

OECD/IEA, Energy Efficiency Refurbishment, 2005.

[48]

Building and energy conservation bureau, About us, http://www.ekubirojs.lv/?cat=1&lang=en.

[49]

Haller, M. et al., The Reference Framework for System Simulations of the IEA SHC Task 44 HPP Annex 38, 2011.

[50]

European Climate Foundation, European Climate Foundation, 2012.

[51]

World Book, Living in the EU, http://www.asparis.net/lowerschool/2dgrade/ maps/climate/europe.htm

[52]

Bertoldi, P. et al, Energy efficiency Status Report, JRC, 2012.

[53]

EC, COMMISSION REGULATION (EU) No 206/2012 of 6 March 2012 implementing Directive 2009/125/EC of the European Parliament and of the Council with regard to ecodesign requirements for air conditioners and comfort fans, 2012.

High-Rise

[54]

Kranzl, L. et al., Policy pathways for reducing energy demand and carbon emissions of the EU building stock until 2030, Energy Economics Group, Vienna University of Technology, 2014.

[55]

Marinhas, S. et al., Air-conditioning, air-handling and refrigeration equipment: European-wide certification, standards and European directives on energy efficiency, Rehva Journal, 2010.

[56]

Constantinescu, N. et al, Reducing Europe´s consumption of fossil fuels for heating and cooling, Euroheat & Power, Brussels, Belgium, 2006.

[57]

Weiss, W., Biermayr P., Potential of Solar Thermal in Europe, 2009.

[58]

Boermans, T. et al., A heating & cooling strategy for the European building sector until 2050, Ecofys, Cologne, Germany, 2015.

[59]

Kalz D., Pfafferott J., Thermal Comfort and Energy-efficient Cooling of Non-residential Buildings, Springer, Cham, Germany, 2014.

[60]

Sanner, B. et al., 2020-2030-2050 Common Vision for the renewable Heating & Cooling sector in Europe, RHC-Platform, Brussels, Belgium, 2011.

[61]

Pezzutto, S., Analysis of the space heating and cooling market in Europe, PhD thesis, University of Natural Resources and Life Sciences, Vienna, Austria, 2014.

[62]

Mahlknecht, H. et al., Refurbishment and monitoring of an historical building, AICARR International Conference, 2009.

[63]

Peel, M.C. et al., Updated Köppen-Geiger climate map of the world, http://people.eng.unimelb.edu.au/mpeel/koppe n.html

[64]

McKnight, T.L., Hess, D., "Climate Zones and Types". Physical Geography: A Landscape Appreciation, PEARSON, New York, USA, 2000.

[65]

Diczfalusy, B. et al., Technology Roadmap, Energy-efficient Buildings: Heating and Cooling Equipment, International Energy Agency, Paris, France, 2011.

[66]

Zangheri, P. et al., Heating and cooling energy demand and loads for building types in different countries of the EU, Polytechnic University of Turin, end-use Efficiency Research Group, 2014.

[67]

Eurostat, Heating degree-days by NUTS 2 regions - annual data, http://appsso.eurostat.ec.europa.eu/nui/show. do?dataset=nrg_esdgr_a

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

DOI: 10.14621/ce.20150206

Visualisation of Unsteady Flow Field in an Axial Flow Fan Matej Fike1), Gorazd Bombek2), Aleš Hribernik2) 1)

Faculty of Energy Technology, University of Maribor Hočevarjev trg 1, 8270 Krško, Slovenia, matej.fike@um.si 2) Faculty of Mechanical Engineering, University of Maribor, Slovenia

Abstract

1. Introduction

An unsteady flow field with rotating stall cells in an axial flow fan has been investigated experimentally. In order to capture the behaviour of the rotating stall cell, measurements of the flow field at the rotor inlet and velocity field within the rotor blade passage at an 80 % span were carried out with the PIV system. Those data were processed by socalled “phase-locked averaging” technique, which enabled to capture the flow field of the rotating stall cell in the reference co-ordinate system fixed to the rotor. As a result, a sequence of 18 images was composed at the three different flow rates and the behaviour of the rotating stall cell has been analysed.

In everyday operation the turbo-machinery is not only used in its design point, however significantly lower flow rates may take place especially when the pressure losses within the system increase. Although small to moderate flow rate reductions do not change the operational characteristic much they may influence the stability of turbomachinery which is demonstrated as a rotating stall, a characteristic flow instability of axial and centrifugal compressors and fans at reduced flow rates [1]. This instability plays a prominent role as a precursory phenomenon of a surge which deteriorates operational characteristics severely and may even damage compressor blades [2, 3]. Day [4] proved by hot wire measurements that rotating stall originates in the instability of a local structure within the compressor’s cascade-flow. This forms one or more rotating stall cells rotating in a circumferential direction at a fraction of the rotor speed [5].

Keywords:

Axial flow fan, PIV; Unsteady flow; Rotating stall cell, Phase locked averaging

Article history:

Received: 31 July 2015 Revised: 20 October 2015 Accepted: 09 November 2015

Rotating stall has been investigated over several decades via both analytical and experimental methods. Most of the experimental investigations based on velocity and pressure measurements have used hotwire [4, 6] or multi-hole pressure probes [2]. Their results verified the periodicity of the rotating stall cell. Hara et al. [2] and Shiomi et al. [6] processed their experimental results with the double phase-locked averaging technique, which allowed for predicting the velocity and pressure fields at the compressor’s inflow and outflow cross-sections, respectively, [6] and made the capture of a rotating stall cell and its propagation within the compressor inflow section possible. However, the spatial distribution was low and the averaged results were limited to the non-rotating section of the compressor and the flow structures within the rotor blade rows (passages) remained hidden. The purpose of this investigation was a visualisation of a flow field within the rotor blade passage of an axial flow

___________________________________________________________________________________________________________ M. Fike, G. Bombek, A. Hribernik: “Visualisation of Unsteady Flow Field in an Axial Flow Fan”, pp. 49–56

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

fan operating under rotating stall conditions. Three different operating (OP) points under rotating stall conditions were analysed. A PIV system was used to capture the velocity field at an 80% span of the rotor blade. The Particle Image Velocimetry (PIV) triggering was synchronized with the observed blades’ passing, and obtained PIV images were then phase-locked averaged, and a sequence of 18 images was composed.

2. Experimental set–up 2.1. Fan geometry and instrumentation Experimental investigations were performed on a test rig. The test rig was built according to ISO 5801:2007 [7] and comprised an inflow section, an axial fan and an outflow section. A transparent duct (Din = 290 mm) formed the fan’s outer casing and extended 1000 mm into the inflow section. Coaxially, an inner cylinder, the diameter of which corresponded to the fan’s hub diameter, was inserted so as to straighten the flow-field ahead of the fan in the axial direction. An axial flow fan was used without a stator and inlet guide vane. The rotor has comprised 10 blades with a NACA 65 profile. The chord length and stagger angle were constant and did not change with the span. The rotor drive was hubmounted, and the three-phase electromotor rotated the rotor at 1470 rpm. Detailed fan geometry specifications are summarized in Table 1. The outflow section was made of a 2000 mm long duct (Din = 290 mm). It ended with a movable lattice that was used as a flow throttle.

2.1. PIV system The Dantec low-speed PIV (Particle Image Velocimetry) system was used to capture the flow-field within the blade passage and at the rotor entry (Figure 1). The maximum system operating frequency was 4.5 Hz. A camera and laser were placed on a lightweight traverse

Table 1: Design parameters of the test fan No. of blade Casing radius Hub radius Chord length Tip clearance Blade profile Stager angle Hub/Tip ratio Rotation speed

10 145 mm 92.5 mm 80 mm 2.5 mm (constant) NACA 6508 45 o 0.65 1470 rpm

system originally used for LDA measurements. A twocavity Nd: YAG laser was used, operating at high power with 50 mJ pulse energy. A fog-generator was used for seeding in order to produce seeding particles with average diameters of 1 μm. The laser was placed upstream at ca. 1000 mm from the rotor. A CCD camera with 1280×1024 pixels resolution was used, and the area covered was ca. 81×65 mm. The time-interval between the laser pulses was 20 μs, and 32×32 pixel-size interrogation areas were used for velocity calculation. Cross-correlation and adaptive correlation were used, both with 25% overlaps. The blade passage velocity fields were measured within the axial-circumferential plane at 80% of the rotor blade span (Figure 2). The captured area covered the whole blade passage and extended approximately 20 mm into the fan entry. A signal relating to the stall cell propagation was needed in order to utilise the phase-locked averaging technique. A signal from the pressure transducer measuring the dynamic pressure 10 mm upstream of the rotor’s leading edge (Figure 2) was used for this purpose. The output of the photo-encoder mounted on the rotor shaft was used to trigger the PIV system, and the 2-D

Laser sheet

Laser Camera

Figure 1: Schematic view of the PIV ___________________________________________________________________________________________________________ M. Fike, G. Bombek, A. Hribernik: “Visualisation of Unsteady Flow Field in an Axial Flow Fan”, pp. 49–56

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Dynamic pressure sensor Averaging line Flow direction

Measurement area

Figure 2: PIV measurement area

velocity-field was, therefore, always captured within the same blade passage. The triggering accuracy was approximately ± 0.05 degrees. The generated trigger signal and the rotating stall reference signal (dynamic pressure ahead of the rotor) were both acquired by the DAQ card based data acquisition system with a 100 kHz sampling frequency.

3. Results and discussion 3.1. Overall characteristic of the test fan The overall fan performance was measured and is presented in Figure 3. It was obtained by moving the exit lattice from fully opened to almost closed, and measuring the pressure rise and fan flow rate. The fan design point DP was located in the middle of the stable region at flow coefficient φ = 0.32 and pressure rise coefficient ψ = 0.325. When the flow was reduced below the design point, the pressure rise coefficient ψ increased until reaching the stall point E. Below this point, a discontinuity occurs, and the operating point is suddenly moved to point G. Further flow reduction is accompanied by continuous pressure rise decrease towards points G, H and I, where the fan surge starts. With the increase of the exit lattice opening, the flow and pressure rise increased, and the fan’s performance characteristic started moving back, crossed points H and G and reached point F, where the second discontinuity occurred, and the operating point moved suddenly into a stable region (point C). Characteristic hysteresis was formed in this way, as reported by [8]. The region between the operating points F and H formed the rotating stall region, which was experimentally investigated.

3.2. Simple averaging PIV measurements were performed at operating points A, B, D, F, G, H and I. Around one hundred and sixty PIV images of the same blade passage were made for each

Figure 3: Fan performance curve

operational point. The results were averaged without phase-locking during the first step. The results obtained at six characteristic operating points are presented in Figure 4. No flow separation from the suction side of the rotor blade took place at operating point A and D. At operating points F, G and H, the conditions seemed similar, although the rotating stall was presented; however, any unsteadiness was smoothed by the averaging process and the rotating stall remained hidden. Due to the surge, the blade passage was severely blocked at operating point I most of the time, and averaging intensified these conditions even more. At this point, we can conclude that simple averaging does not ensure any realistic results when flow unsteadiness, such as rotating stall, is presented. The smoothing effect erases any low amplitude flow fluctuation. Figure 5 presents relative velocity’s standard deviation fields within the blade passage at 80 % span. The standard deviation of velocity is within the range between 0 and 1 m/s under normal-flow conditions (operating points A and D). At the operating points, where the rotating stall was detected, the standard deviation increases significantly. It reaches its maximum of 9.5 m/s at the operating point G, where the oscillations in relative velocity between normal flow and rotating stall cell were the greatest. At the operating point I, where surge occurs, the relative velocity’s standard deviation in comparison to the rotating stall is reduced to around 4 m/s.

3.3. Phase–locked averaging When rotating stall occurs, the flow pattern within a blade passage alternates between a through flow and a backflow. If only one rotating stall cell exists, two domains can be observed: normal-flow and rotating stall cell. Whilst in the normal-flow domain the through flow

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

= 0,28 – OP A

= 0,18 – OP G

= 0,24 – OP D

= 0,14 – OP H

= 0,19 – OP F

= 0,09 – OP I

Figure 4: Averaged relative velocity fields within the observed blade passage at 80% span (every tenth vector is shown) – operating points (OP) A, D, F, G, H and I

= 0,28 – OP A

= 0,18 – OP G

= 0,24 – OP D

= 0,14 – OP H

= 0,19 – OP F

= 0,09 – OP I

Figure 5: Relative velocity’s standard deviation fields within the blade passage at 80 % span– operating points A, D, F, G, H and I ___________________________________________________________________________________________________________ M. Fike, G. Bombek, A. Hribernik: “Visualisation of Unsteady Flow Field in an Axial Flow Fan”, pp. 49–56

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Table 2: The frequency of the rotating stall reference signal Operating point F

Flow coefficient 0,19

Frequency (Hz) 15,8

G H

0,15 0,14

14,8 14,8

Figure 6: Raw rotating stall reference signal, generated sine-wave signal, and the PIV system trigger signal – operating point F

a) Before sorting

a dynamic pressure variation ahead of the rotor which can be used as a rotating stall reference signal when performing phase-locked averaging. It was obvious from our previous research that the rotating stall cell formed at the tip of the rotor; therefore, a small twin-tube probe with a total pressure tap in front and a static pressure tap at the back was inserted 10 mm from the fan tip and 20 mm deep from the outer casing (Figure 2). Figure 6 shows the dynamic pressure signal from the twin-tube probe. Its periodic shape allowed for the assumption that the rotating cell propagates at a constant speed. For different operating points the FFT procedure was applied to obtain the characteristic reference signal frequency. The results are presented in Table 2. The stall cell frequency depended on the flow coefficient. At the operating point H and G, the frequency of the rotating stall cell was 14.8, which is 60% of the rotor rotational frequency. With increasing flow-rate, the rotating stall cell speed increased by up to 65% of the rotor speed (operating point F) and with further-flow increase the rotating stall cell decayed (operating point C), whilst the reference signal frequency equaled the rotor frequency. Below operating point H no characteristic frequency was detected. It indicated the rotating stall transformed into surge (operating point I). The phase-locked averaging was applied for the fan’s operating points, where the rotating stall cell frequency was obtained. A special focus was on the operating point F, where the shape of the raw rotating reference signal was the most periodic among all operating points with the rotating stall. At that operating point F additional PIV images were obtained, and the final number of processed images was 1020. Acquired raw rotating stall signal was not appropriated for the averaging process, therefore a simple sine-wave signal was used instead. A sin-wave signal with amplitude of 0.5 V was generated of the raw using the characteristic frequency rotating stall reference signal. It can be represented by a simple time–dependent function as: ( ) = 0.5 ∙

b) After sorting Figure 7: The sequence of averaged velocities along the averaging line before and after sorting – operating point F

prevails, the back flow is locally present in a domain of the rotating stall cell. This alternation between the through flow and back flow may simply be observed as

(1)

The quality of describing raw rotating stall reference signal with generated sine-wave signal depended on the operating point. The generated sine-wave signal for operating point F is presented in Figure 6. It fits the general periodic form of the reference rotating signal well. When the flow was reduced below the operating point F, towards points G and H, the periodicity of raw reference signal with sin-wave signal dropped. Consequently, the quality of phase-locked averaging procedure was reduced. The PIV trigger signal is also presented in Figure 6. The PIV system operated at a 2.7 Hz frequency, thus approximately every sixth

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

passing of the rotating stall cell was captured; however, the captures were not synchronized with the rotating stall position. In order to obtain phase-locked averaging, every single PIV image had to be correctly sorted according to the rotating stall cell position. This sorting was performed by a simple procedure. The time-delay (∆ ) between the trigger point and the nearest point, in a positive time direction, where the increasing sinewave crosses value of 0.5 was first found for each PIV image. Then the angular position of the PIV image in respect to the rotating stall cell position was assigned, and finally the images were sorted in respect of their angular positions from the minimum to maximum values. To test the effectiveness of the sorting procedure, an averaging line in the measurement plane perpendicular to the blade chord (Figure 2) was drawn. For each PIV image the average velocity along this line was calculated. Figure 7 presents the sequence of averaged velocities before and after sorting at the operating point F, proving that the images were successfully sorted. The next step of phase-locked averaging was the averaging of previously averaged velocities along line. It was performed within constant 20° angular steps and resulted in 18 consecutively averaged velocities. The averaging minimized the velocity fluctuations, which can be seen in Figures 7b and 8, respectively. The domains of normal-flow and rotating stall cell (RSC) are clearly distinctive. The rotating stall cell at the operating point F occupied approximately 40% of the rotor circumference. Figure 9 presents that at operating point G and H approximately 70% and 90% have been occupied by rotating stall cell, respectively. The averaged velocities dropped significantly within this region. Their fluctuations were extremely high at the operating point F. With the flow reduction towards operating point G and H, the fluctuation became lower. The last step of phase-locked averaging was the averaging of sorted PIV images. It was performed within

= 0,18 – OP G

constant 20° angular steps, and resulted in 18 consecutively averaged PIV images. Only two characteristic relative velocity fields (images at 0° and 200°) within the blade passage at 80% span for all tree operating points with the rotating stall cell are presented in Figure 9. Each velocity vector represents relative velocity, i.e. absolute measured velocity reduces by a circumferential one. In this way, the representation of the flow around the rotating fan’s blade is much more adequate. Images with angular position 0° represent the blade passage flow under normal-flow conditions. The flow entirely follows the contour of the suction side of the rotor blade and no swirls are shown in the flow fields. Images with angular position 200° represent the rotating stall cell passing the observed blade passage. At the operating point F the swirl is due to the averaging process small and it exists only at the trailing edge. When the flow is reduced towards the operating point G and H, the averaged swirl increases, the flow separation point is moved to 40% and 50% of a chord length, respectively. The whole intake domain is blocked (axial velocity component is 0).

4. Conclusions This paper has presented the visualization of flow in an axial flow fan under the unstable aerodynamic condition. The PIV measurement system was used to capture velocity flow fields within rotor blade passage. A FPGA chip was used to trigger the PIV measurement system corresponding to the blade passing reference signal. This enabled the same blade passage to be captured at a desired instant. The measured velocity fields were simply averaged first, which gave a totally unrealistic impression of the flow structure within the blade passage. The influence of the presence of rotating stall was only noticed with a slight velocity reduction and only a highly unstable surge flow operation regime was detected. Analysis of relative velocity’s standard deviation fields within blade passage showed the possibility of detecting significant fluid flow changes

= 0,14 – OP H

Figure 8: The sequence of averaged velocities along averaging line after sorting – operating points G and H ___________________________________________________________________________________________________________ M. Fike, G. Bombek, A. Hribernik: “Visualisation of Unsteady Flow Field in an Axial Flow Fan”, pp. 49–56

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Normal flow

0°

Rotating stall cell

200°

= 0,19 – OP F

0°

200°

= 0,18 – OP G

= 0,14 – OP H

Figure 9: Two characteristic averaged relative velocity fields within the blade passage at 80% span (every tenth vector is shown) – operating points F, G, and H

when influenced by rotating stall cell. The intensity graphs showed that the standard deviation increases significantly under rotating stall conditions. Although the relative velocity’s standard deviation improved the understanding of flow patterns within the blade passage, it could not be used for flow animation of the rotating stall cell. With the dynamic pressure measurements at the rotor tip the rotating stall cell frequencies were detected. This made possible the phase-locked averaging of PIV measured velocity fields. Sequences of averaged velocities along chosen line were composed, presenting a flow field changing over 20° steps. The domains of normal-flow and rotating stall cell were clearly distinctive. With the flow reduction, the domain of rotating stall cell became larger. Sequences of images were also composed. Two characteristic images (0° and 200°) presents the blade passage flow under normal-flow and rotating stall cell conditions, respectively.

References [1]

Bianchi Stefano, Corsini Alessandro, Sheard G. Anthony, Torora Cecilia, A Critical Review of Stall Control Techniques in Industrial Fans, ISRN

Mechanical Engineering, Volume 2013, (2013), DOI 10.1155/2013/526192. [2]

Hara, T., Morita, D., Ohta, Y., Outa, E., Unsteady flow field under surge and rotating stall in a threestage axial flow compressor, Journal of Thermal Science,Vol. 20, No. 1, 2011, pp. 6-12.

[3]

Guardain, N., Montagnac, M., Burguburu, S., Numerical simulation of an axial compressor with non axisymmetric casing treatment, Progress in propulsion physics, Vol. 1, 2009, pp. 593-608.

[4]

Day, I.J., Cumpsty, N.A., The measurement and interpretation of flow within rotating stall cells in axial compressors, Journal of Mech. Engineering Sci., IMechE, Vol. 20, No. 2, 1978, pp. 101-114.

[5]

Cumpsty, N.A., Greitzer, E.M., A simple model for compressor stall cell propagation, Transaction of ASME, Journal of Engineering Power, Vol. 104, 1982, pp. 170-176.

[6]

Shiomi, N., Cai, W.X., Muraoka, A., Kaneko, K., Setoguchi, T., Internal flow of a high specificspeed diagonal-flow fan (Rotor outlet flow fields with rotating stall), Int. Journal of Rotating Machinery, Vol. 9, 2003, pp. 337-344.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

[7]

[8]

ISO. European Standard, Industrial fans – Performance testing using standardized airways, ISO 5801:2007 including Cor 1:2008, CEN, 2008. Shuey, M.G.E., Numerical near-stall performance prediction for a low speed single stage compressor, M.Sc. Thesis, University of Cincinnati, Cincinnati, OH, 2005.

[9]

Fike, M., Bombek, G., Hriberšek, M., Hribernik, A., Visualisation of rotating stall in an axial flow fan, Experimental Thermal and Fluid Science, Vol. 53, 2014,

pp.

269-276,

DOI

10.1016/j.

expthermflusci.2013.12.020

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

DOI: 10.14621/ce.20150207

Experimental Modelling of a Wind Farm in Mamaj, Albania Elena Bebi1), Jorgaq Kacani2), Edmond Ismaili3), Noriyuki Goto3), Atushi Fujiwara3) 1)

Polytechnic University of Tirana, Production and Managing Department, Tirana, Albania, ebebi@fim.edu.al 2) Polytechnic University of Tirana, Rectorate, Tirana, Albania 3) Faculty of Engineering, Mie University, Japan

Abstract

1. Introduction

Developing a wind farm is a challenging task, especially in developing countries or countries in transition where there is not a good database for wind energy implementation. In this study, a wind farm development is presented with a unique approach. Wind characteristics and wind energy potential of a site in the southern part of Albania were analysed utilizing the 10min wind speed data collected from 2 towers during the period March 2013- June 2014. Wind energy assessment was then processed through WASP software (Wind Atlas Analysis and Application Program). Further, experimental verification and optimal arrangement of wind turbines were performed. A scale model of the site was set in the test section of a wind tunnel, facilitated from Mie University in Japan. Wind over the site was measured through the PIV (Particle Image Velocimetry) method. Small scale wind turbine models were placed over the site and wind speed was assessed for different arrangements. The wake characteristics developed behind the wind turbines were studied. Optimum number of wind turbines and arrangement were determined.

The rapid growth of the wind energy industry has led to cost reduction challenges. Reducing the cost of energy produced by wind power depends on many factors such as, site selection, site layout design, predictive maintenance, and optimal control system design [1]. The wind farm layout design is an important component of ensuring the profitability of a wind farm project. An inadequate wind farm layout design would lead to lower than expected wind power capture, increased maintenance costs, and so on. Normally, if a turbine is within the area of turbulence caused by another turbine, or the area behind another turbine, the wind speed suffers a reduction, and therefore there is a decrease in the production of electricity. In other words if there is a lot of interference or wake generated by the wind turbines, the possibility of mechanical failure would increase as well as the need for more maintenance actions, resulting in an inevitable reduction in power output. A rule that has gained support among the scientific community is not to place turbines at a distance less than 5 rotor diameters in an attempt to minimize the impact of the wake in the turbines, and to control the reduction on its power output. In addition to considering the impact that turbines can have on each-other, it is important to take into account the terrain, weather conditions and wind conditions in the region, such as speed and wind direction.

Keywords:

Wind power; Wind farm, Wind turbines; PIV; Experimental modelling

Article history:

Received: 26 July 2015 Revised: 28 October 2015 Accepted: 09 November 2015

Several studies have been conducted in recent years in order to maximize energy production and the efficiency of the turbines [2], [3], [4], [5]. These efforts have focused primarily on finding the best location of turbines within wind farms, based on their efficiency. Grady et al. [2] and Mosetti et al. [3] used a genetic algorithm to minimize a weighted sum of wind energy and turbine costs. The wind farm is divided into a square grid to facilitate the encoding of a 0–1 type solution. Lackner and Elkinton [4] presented a general framework to

___________________________________________________________________________________________________________ E. Bebi, et al: “Experimental Modelling of a Wind Farm in Mamaj, Albania”, pp. 57–64

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

optimize the offshore wind turbine layout. Castro Mora et al. [5] also used a genetic algorithm to maximize an economic function, which is related to turbine parameters and locations. Similarly in [5], the wind farm is represented with a square grid. One of the shortcomings of the approach presented in [5] was that wake loss was not considered. In [2], [4], the wind energy calculation is not based on the power curve function, and wind direction was not fully discussed in their optimization models. These investigations focus their interest in finding a configuration in which the effect of a turbine on the other is reduced. A commonly used model is that proposed by Jensen [6]. In this model the wake was treated as turbulence, which occurs in the area behind the turbine after a gust of wind passes through the rotor, reducing this model to consider only the momentum that takes place in the turbine downstream. Beyer et al. [7] used three wind farm cases with different wind speeds and directions and evaluated them with

Figure 1: Top view of Mamaj, village

expert guess configurations that exist for those wind farms. At the design stage, the Annual Energy Production (AEP) is the expected (planned) wind energy production. The annual energy production (AEP) is affected by the turbine availability, i.e., the number of operational hours in a year. Maximizing the AEP is an effective approach for reducing the cost of energy production. In this study, a small wind farm development is presented with a unique approach. Wind characteristics and wind energy potential of a site in the southern part of Albania were analysed utilizing the 10 min wind speed data collected from 2 towers during the period March 2013 - June 2014. Wind energy assessment was then processed through WASP software. Wind farm layout design was optimized by improving the AEP minimizing the wake losses. Further, experimental verification and optimal arrangement of wind turbines was performed.

Figure 2: 3-D model of interested area

Figure 3: Site with highest wind potential

Figure 4: The wind frequency rose

___________________________________________________________________________________________________________ E. Bebi, et al: “Experimental Modelling of a Wind Farm in Mamaj, Albania”, pp. 57–64

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Total gross AEP [GWh]

19.005

Total gross AEP [GWh]

25.312

Total net AEP [GWh]

18.790

Total net AEP [GWh]

24.359

Proportional wake losses [%] 1.13

Proportional wake losses [%] 3.76

Figure 5: Wind farm of 3 wind turbines

Figure 6: Wind farm of 4 wind turbines

A scale model of the site was set in the test section of a wind tunnel, facilitated from Mie University in Japan. Wind over the site was measured through PIV (Particle Image Velocimetry) method. Small scale wind turbine models were placed over the site and wind speed was assessed for different arrangements.

2.2. Experimental setup Figure 7 shows the experimental configuration. The experiments have been carried out in a Gottingen type single return wind tunnel with main parameters as follows: - Cross test section 600mm x 600mm (Figure 8);

2. Materials and methods 2.1. Site description and modelling The site for developing the wind farm is located in Mamaj, north of Tepelenë, Albania (Figure 1). It consists of a natural valley near the river Vjosa and relatively low height hills of not more than 150m (Figure 2). Most of the area under study is covered with herbaceous vegetation, which based on the Davenport classification corresponds to the roughness height 0.03m. From previous research [8] it was found that the site with the highest wind potential is the area over the hills as shown in Figure 3. Prevailed wind speed was blown from the south-east direction (1200-157 0) and northwest direction (3000-3300) (Figure 4). Wind farm prediction with WASP software demonstrated that maximum Annual Energy Production with minimum wake losses was achieved in case of VESTAS V100/1800 wind turbines with arrangement of 3 and 4 wind turbines as in Figure 5 and Figure 6 respectively.

- Length of test section: 5100mm; - Maximum reachable wind speed: 35 m/s. To achieve the wind similiar to natural wind condition, a turbulence grid was mounted to the outlet of the wind tunnel. Measurements were performed for a wind speed 7m/s. For experimental purposes a model on scale 1:1250 was prepared as shown in Figure 9. Restricted from the test section dimensions, and based on the fact that the site with the highest wind potential was on top of the hills, the model was considering only the hill area of the site. Experimental model was prepared in a circular shape with a diameter of 580mm and was placed at a distance 1750mm from the tunnel outlet in the downstream direction. Model was mounted on turnable table, as in Figure 10, making possible positioning of the model against the flow and obtaining variable wind speed direction. Outside the simulated area there are some terrain configurations that may affect the inflow wind. To take into account this effect, an obstacle was placed in the

___________________________________________________________________________________________________________ E. Bebi, et al: “Experimental Modelling of a Wind Farm in Mamaj, Albania”, pp. 57–64

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

600 mm

Lase sheet

Figure 7: The experimental configuration

600 mm

CCD Camera

Figure 8: The cross test section

Figure 9: The experimental model

Figure 10:Turning table

Upstream

Real site

Downstream

Test section

Figure 11: Effect of terrain and modeling

Figure 12: The model with 3 wind turbines

Figure 13: The model with 4 wind turbines

___________________________________________________________________________________________________________ E. Bebi, et al: “Experimental Modelling of a Wind Farm in Mamaj, Albania”, pp. 57–64

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 14 shows an aspect of a real time measurement.

upstream of the model in order to obtain similiar inflow condition as the actual terrain. In addition, the hill site was smoothly joined with the flat site (Figure 11).

Measurements were performed for a wind speed of 7 m/s. Laser sheet was positioned at a height 60 mm over the top of the model, assuming that all the hub heights were over this heights in order to exclude the effect of wake of the nacelle of the model wind turbines. Different wind directions, according to the wind rose frequency distribution on section 2.1, were obtained by turning the model against the wind and measurements were performed for every 15 degrees in the interval 120o-180o and 300o-360o respectively (Figure 15).

Small turbines on a scale of 1:1250 of real turbines VESTAS V100/1800 were mounted on the models in the arrangement of 3 and 4 turbines as discussed on section 2.1, and depicted on Figure 12 and Figure 13, respectively.

2.3. Measurement apparatus and procedure Flow field over the model was measured through Particle Image Velocimetry method. Seeded particles were feeded from tracer inlet. Images of the particles on the laser sheet were captured through a CCD camera. Main data of the PIV system are shown in Table 1.

3. Results and discussion Wake loss is an important factor in considering wind farm layout design [9]. When a uniform incoming wind

30 0

300

180o

150o

20 0

-297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5

-297.5 -292.5

-297. 5 -297. 5

-287.5 -282.5 -277.5 -272.5 -267.5 -262.5 -257.5 -252.5 -247.5 -242.5 -237.5 -232.5

-297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5

-227.5 -222.5 -217.5 -212.5 -207.5 -202.5 -197.5 -192.5 -187.5 -182.5 -177.5

-297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5 -297. 5

-297. 5 -292. 5

0 0

-287. 5 -282. 5 -277. 5 -272. 5 -267. 5 -262. 5 -257. 5 -252. 5 -247. 5 -242. 5 -237. 5 -232. 5

0 0 0 0 0 0 0 0 0 0 0 0

-227. 5 -222. 5 -217. 5 -212. 5 -207. 5 -202. 5 -197. 5 -192. 5 -187. 5 -182. 5 -177. 5

0 0 0 0 0 0 0 0 0 0 0

-297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5

-172.5 -297. 5 -167.5 -297. 5 -162.5 -297. 5 -157.5 -297. 5 -152.5 -297. 5 -147.5 -297. 5 -142.5 -297. 5 -137.5 -297. 5 -132.5 -297. 5 -127.5 -297. 5 -122.5 -289.8 44 -117.5 -289.8 69

-172. 5 0 -167. 5 0 -162. 5 0 -157. 5 0 -152. 5 0 -147. 5 0 -142. 5 0 -137. 5 0 0 -132. 5 -127. 5 0 -122. 5 7.6 64899 -117. 5 7.6 3927

-297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5

-112.5 -289.8 76 -107.5 -289.8 76 -102.5 -289.8 89 -97. 5 -289.8 94 -92. 5 -289.8 86 -87. 5 -289.9 02 -82. 5 -289.9 14 -77. 5 -289.9 48 -72. 5 -289.9 58 -67. 5 -289.9 51 -62. 5 -289.9 6

-112. 5 7.6 32059 -107. 5 7.6 32788 -102. 5 7.6 2048 -97. 5 7.6 13766 -92. 5 7.6 21155 -87. 5 7.6 05199 -82. 5 7.5 92839 -77. 5 7.5 58586 -72. 5 7.5 48665 -67. 5 7.5 55604 -62. 5 7.5 46752

-297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5

-57. 5 -289.9 75 -52. 5 -289.9 78 -47. 5 -289.9 68 -42. 5 -289.9 61 -37. 5 -289.9 69 -32. 5 -289.9 68 -27. 5 -289.9 62 -22. 5 -289.9 63 -17. 5 -289.9 66 -12. 5 -289.9 86 -7. 5 -289.9 92 -2. 5 -289.9 98

-57. 5 7.5 30346 -52. 5 7.5 26125 -47. 5 7.5 35158 -42. 5 7.5 4207 -37. 5 7.5 33865 -32. 5 7.5 34829 -27. 5 7.5 40455 -22. 5 7.5 39076 -17. 5 7.5 36332 -12. 5 7.5 1581 -7. 5 7.5 09837 -2. 5 7.5 03339

-297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5 -297.5

2.5 -290.0 2 7.5 -290.0 6 12. 5 -290.0 77 17. 5 -290.0 86 22. 5 -290.0 66 27. 5 -290.0 77 32. 5 -290.0 91 37. 5 -290.0 99

2. 5 7.4 80844 7. 5 7.4 41486 12. 5 7.4 23429 17. 5 7.4 14178 22. 5 7.4 34046 27. 5 7.4 23689 32. 5 7.4 08742 37. 5 7.4 00922

200

10 0

Flow

100

2 B

Flow 0

3.0

- 10 0

-1 0 0

1 - 20 0

-2 0 0

9.0 - 30 0 - 30 0

- 20 0

- 1 00

10 0

2 00

3 00

-3 0 0 -3 0 0

-2 00

- 1 00

100

2 00

3 00

Figure 16 (a, b, c, d): Velocity distribution and wake behind the wind turbine in case of 4 wind turbines for the southeast wind directions ___________________________________________________________________________________________________________ E. Bebi, et al: “Experimental Modelling of a Wind Farm in Mamaj, Albania”, pp. 57–64

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 17 (a, b): Velocity distribution and wake behind the wind turbine in case of 4 wind turbines for the northwest wind directions.

300

165

150

3 200

100

Flow

Flow 2

3.0

1 -100

-200

9.0

9.0 -300 -300

-100

100

200

300

135

-200

300

120

200

100

Flow

3 100

Flow 0

3.0

-100

-200

9.0

9.0 -300 -300

-200

-100

100

200

300

c) A

-300 -300

-200

-100

100

200

300

Figure 18 (a, b, c, d): Velocity distribution and wake behind the wind turbine in case of 3 wind turbines for the southeast wind directions

___________________________________________________________________________________________________________ E. Bebi, et al: “Experimental Modelling of a Wind Farm in Mamaj, Albania”, pp. 57–64

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Table 1: The main data of the PIV system Laser Pulse energy CCD Camera Quality Picture size Sampling frequency Sampling points

GEMINI PIV 15 120 [mJ] Hisense 11M 4000*2672 pixel 258[mm] x 172[mm] 1 [Hz] 100

encounters a wind turbine, a linearly expanding wake behind the turbine occurs [3]. A portion of the free stream wind speed will be reduced from its original speed.

It is important to emphasize that choosing 3 and 4 wind turbines for the wind farm layout design came as a result of using large wind turbines as well as the small area of the region of high wind potential.

4. Conclusions The main topic of this research work was the experimental verification of the wake developed from the wind turbines in a small wind farm and to find the optimal layout of the wind turbines in view of the minimal wake losses. Two main configurations were experimentally simulated in a wind tunnel:

In this section we present the velocity field and the wake developed behind the wind turbines obtained from PIV measurements. Figure 16 (a, b, c, d), shows the velocity distribution and wake in case of arrangement of 4 wind turbines, for the wind direction of 180o, 150o , 135o and 120o , respectively. Colour bar depicts the change of velocity values. It is noted that for chosen layout of the wind turbines, turbine number 4 is slightly affected from the wake developed from the turbine number 1 for the wind direction 135o. For the wind direction 120o it seems that turbines 2 and 3 are affected more from the wake developed from turbine 1. For the prevailing wind directions 135o–160o it seems that the wake does not affect much the operation of wind turbines in the proposed layout. Figure 17 (a, b), shows the velocity distribution and wake in case of arrangement of 4 wind turbines, for the northwest wind direction of 330o and 315o, respectively. It seems that the inflow wind of the turbine 3 is affected from the wake developed from the turbine 1 for the wind direction 330o, while there is no significant wake affect for the most prevailing northwest wind direction at 315o. Figure 18 (a, b, c, d) demonstrates the flow and wake distribution in case of layout of three wind turbines for the southeast wind directions 165o, 150o, 135o and 120o , respectively. It is noted that inflow wind for every wind turbine is undisturbed from the wake developed from the other wind turbines for the most prevailing wind southeast wind directions. For the wind direction 120o, turbine 2 is completely under the wake developed from the turbine 1. Practically, such wind direction has a 0% occurrence. Proposed wind turbine layout arrangement confirms the prediction from the WASP analysis for the maximum AEP achieved with the minimum wake losses.

Layout with 4 wind turbines;

Layout with 3 wind turbines.

From WASP software analysis it was found that layout with 3 wind turbines had lower wake losses compared to the layout with 4 wind turbines. It was experimentally verified that in case of 3 wind turbines layout, inflow wind conditions for every wind turbine were almost undisturbed from the wake developed from other wind turbines. However, the wind farm with 4 wind turbines has much larger AEP compared to the wind farm with 3 wind turbines. To determine the best layout a detailed economical analysis is necessary.

Acknowledgements We thank the Fluid Engineering Laboratory for Energy and Environment of the Mie University in Tsu City, Mie Prefecture, Japan for their support and availability for the realization of the experiments and data collection.

References [1]

Wiser R, Bolinger M. Annual report on U.S. wind power installation, cost, and performance trends: 2006. Available from. Golden, CO: NREL, US Department of Energy, 2017, http://www.nrel.gov/wind/pdfs/41435.pdf,

[2]

Grady SA, Hussaini MY, Abdullah MM. Placement of wind turbines using genetic algorithms. Renewable Energy 2005;30:259–70.

[3]

Lackner MA, Elkinton CN. An analytical framework for offshore wind farm layout optimization. Wind Engineering 2007, 31:17–31.

___________________________________________________________________________________________________________ E. Bebi, et al: “Experimental Modelling of a Wind Farm in Mamaj, Albania”, pp. 57–64

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

[4]

Mosetti G, Poloni C, Diviacco B. Optimization of wind turbine positioning in large wind farms by means of a genetic algorithm. Journal of Wind Engineering and Industrial Aerodynamics 1994;51:105–16.

[7]

Beyer HG, Ruger T, Schafer G, Waldl HP, 1996. Optimization of Wind Farm Configurations with Variable Number of Turbines. Proceedings of the European Union Wind Energy Conference (EUWEC), Sweden, 1069-1073.

[5]

Castro Mora J, Calero Baron JM, Riquelme Santos JM, Burgos Payan M. An evolutive algorithm for wind farm optimal design. Neurocomputing 2007;70:2651–8.

[8]

E.Bebi, J. Kaçani, E. Ismaili (2015) The assessment of wind potential in Mamaj, Tepelenë, Albania. IJEES Volume 5/2 (2015) pg 277-286.

[9]

[6]

Jensen NO, 1983. A note on Wind Generator Interaction. Riso National Laboratory, Roskilde, Denmark.

Neustadter HE. Method for evaluating wind turbine wake effects on wind farm performance. Journal of Solar Energy Engineering 1985; 107:240–3.

___________________________________________________________________________________________________________ E. Bebi, et al: “Experimental Modelling of a Wind Farm in Mamaj, Albania”, pp. 57–64

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

DOI: 10.14621/ce.20150208

Voltage Stability Assessment through a New Proposed Methodology Marjela Qemali, Raimonda Bualoti, Marialis Celo Polytechnic University-Tirana, Electrical Engineering Faculty, Power System Department Sheshi”Nene Tereza”, Nr.4, Tirana, Albania, m_qemali@yahoo.com

Abstract

1. Introduction

In the last ten years, it is worth to mention that there is an ongoing discussion in literatures, workshops, conferences, seminars that emphasise the fact that soon the power systems will operate at the limits of capacities. This has become a cliché to countries of the world but a reality for Albanian power system. On 23 February 2010, as a result of an atmospheric discharge at Vau Dejes hydropower plant, the 220kV line Vau Dejes-Koplik was disconnected due to the distance protection action. A long scenario of corrective actions resulted to a system blackout because of the exhaustion of the power system installed capacities. Many others similar scenarios which have occurred in the last years have highlighted the fact that it is no longer worth to rely on the N-1 security criterion, as the classical security standard. In this paper we propose a new methodology with the scope: to define the time period when the voltage stability criterion is fulfilled. Time period through 24 hours a day, for minimal and maximal loading is determined. The methodology has been successfully tested on IEEE 14 bus system and then applied in Albanian power system. NEPLAN software is used to evaluate the voltage stability criteria for the proposed methodology. The obtained results give important information on Albanian power system situation and many recommendations to be taken in consideration from power system operators.

The complexity of transmission systems management is increasing. In this way it gets more difficult the determination of the security limits and measuring the distance between the operation point and the nearest critical limit operation. In recent years we have often heard through workshops, conferences, seminars that soon the power systems will operate at their capacity limits. It is turned into a cliché for countries of the world but a reality for the Albanian power system.

Keywords:

Voltage stability; N-1 criteria, Albanian power system

Article history:

Received: 26 July 2015 Revised: 29 October 2015 Accepted: 09 November 2015

The supported actions and penalizing measures undertaken by the Albanian government in terms of non-technical losses reduction have given their impact on performance of Energy Distribution Company, as well as on the value of the consumed energy. In [1] is shown that for the year 2014, the energy demand has decreased by 164 GWh compared to the previous year. The registered peak load has decreased from 1475 MW on 2013 to 1448 MW on 2014. The Peak load continues to exceed the installed capacity of Albanian power generation (about 2%). So, it is not possible to use the N-1 criterion, as the classic standard criteria for security system operation during all working hours of the year. The system is no longer able to "survive" without an action after a disturbance. In order to implement corrective actions, there have been builted and developed different protection and control philosophies. Equipment such as tap changer transformers, and static VAR compensator have been added to the system in order to increase its control ability. Power system is becoming more complex. The reasons for this situation are some phenomena which can not disappear in the near future [2]: 1. It is more difficult than ever to build new overhead lines. People are more and more afraid of hypothetical electromagnetic effects, or they just don’t like to see big towers in the landscape.

___________________________________________________________________________________________________________ M. Qemali, R. Bualoti, M. Celo: “Voltage Stability Assessment through a new Proposed Methodology”, pp. 65–73

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

It is very difficult to explain the need for new interconnection lines to people who already have access to electricity at a reasonable price and with high availability. Alternative solutions are technically complex, costly, and need even more time to be deployed. 2. The second main reason is the massive integration of renewable, generally intermittent generation in the system. Power flows in the grid are created by difference in location between power sinks and sources. With a significant amount of intermittent power generation, the predictability of the sources (location and levels of power injections) decreases and strongly affects the predictability of power flows. Furthermore, these new power plants are generally small units connected to the distribution grid. Transmission system operators (TSOs) therefore have difficulty observing these power injections, and they have no direct control over them. Another factor is the inconsistency between the relatively short time needed to build new wind farms (two or three years) or install photovoltaic panels (months) and the time it takes to go through all the administrative procedures required to build new lines (more than five years). Some TSOs have proposed to the regulators to implement a mechanism that encourage the installation of these new generators in areas where the whole system has enough physical/technical limit of reserves availability to accommodate the new injections. These locations do not generally match those of the large load centres, so a transmission network is still required, and this network will still have to deal with the variability of the power flows. 3. The third reason is related to the liberalization of the energy markets. Generators, retailers, and consumers view the transmission system as a public resource to which they should have unlimited access. This leads TSO's towards profit maximization, optimizing the assets utilization. This optimization is limited by security considerations, however, because large blackouts are unacceptable in our modern societies due to their huge economic and social costs. Since TSOs are responsible for maintaining the security of the power system, they must define the security limits that should be respected. As in any constrained optimization problem, the optimal solution toward which the market evolves tends to be limited by these security constraints. A transparent definition and assessment of the distance to these security limits thus becomes of paramount importance. To maintain the security of the supply in this

context, TSOs must adapt the transmission systems by considering the following technologies: long-distance HVAC underground cables with large reactive compensator; HVDC underground cables in parallel with the ac grid with smart controls of the ac/dc converters; HVDC grids, first to connect offshore wind farms efficiently and then to provide cheaper interconnections between distant areas. In this paper we propose a new methodology to define the day period that the limit of voltage stability criterion is fulfilled. The methodology has been successfully tested on IEEE 14 bus system and then applied in Albanian power system for minimal and maximal day loading.

2. Relationship between voltage stability security coefficient and N-1 criteria. Case study: 14 node IEEE network In [3] is suggested a limit of 130% for the voltage stability security coefficient. System security assessment can be performed by different security criteria’s. The N-1 criterion is used in the most study. We have realised that the 130% limit of stability is closely related to N-1 security criterion [4]. In order to illustrate this relationship we have applied N-1 security criterion to the standard IEEE 14 buses (Figure 1). The critical elements are identified calculating the branch participants’ factors. Branch participation factors indicate which branches consume the most reactive power in response to an incremental change in reactive load. Branches with high participations are either weak links or are heavily loaded. Branch participations are useful for identifying remedial measures to alleviate voltage stability problems and for contingency selection [5]. Figure 2 shows the calculated branches participation factor for the standard IEEE 14 buses. We have identified Branch10 (connected between bus 5 and bus 6) as the critical one. To evaluate the voltage stability limit we have applied the N-1 security criterion for the branch10. The P-V curves method [6] is used in order to evaluate voltage stability limit with application of N-1 criteria. The P-V curves are produced by running a series of load flow cases. P-V curves relate bus voltages to load within a specified region. The benefit of this methodology is that it provides an indication of proximity to voltage collapse throughout a range of load levels. The nature of voltage collapse is that, as power transfers into well-bounded region are increased, the voltage profile of that region will become lower and lower until a point of collapse is reached. The voltages at specific buses in the region can vary significantly and some specific bus voltage could appear acceptable. The point-of-collapse at all buses in

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 1: IEEE 14 buses network

Figure 2: Branches participation factors for standard network IEEE 14 buses ___________________________________________________________________________________________________________ M. Qemali, R. Bualoti, M. Celo: “Voltage Stability Assessment through a new Proposed Methodology”, pp. 65–73

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 3: P-V curves when Branch10 is connected

Figure 4: P-V curves when Branch10 is disconnected ___________________________________________________________________________________________________________ M. Qemali, R. Bualoti, M. Celo: “Voltage Stability Assessment through a new Proposed Methodology”, pp. 65–73

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

the study region, however, will occur at the same power import level, regardless of the specific bus voltages (Figure 3, 4). We can see that the voltage stability limit decreases from 171 % to 123.5 % when Branch10 is disconnected. We can see that the voltage stability limit decreases about 47.5%, which emphasises the fact that this branch is operating in critical conditions. While if the same operation is done for the other network branches their voltage stability limit decrease corresponds within the conditions imposed by [3] that the limit for a stable system operation is 130% [7]. For example for branch 8 the voltage stability limit decreases about 14%. Finally we can conclude that the conditions imposed by [3] that limit for a stable system operation is 130%, is closely related to N-1 system security criterion. This relationship between voltage stability security coefficient and N-1 criteria gives us the possibility to use the verified value of voltage stability criteria in the further paper objective: to define the time period when the voltage stability criterion is fulfilled.

3. Proposed Methodology In this section is illustrated the new proposed method. The objective of the method is the evaluation of the day period (within 24 hours) during which power system operates respecting the voltage stability criterion of 130% [1].

3.1. Required data a- Network hourly load profile. The hourly load profile for the day with the maximal load will be used to evaluate the worst case. b- Voltage stability limit evaluated through P-V curves maximum loading point. The analysis should be performed for different hours of network load curve and the power system loading limit values should be recorded.

3.2. Data processing In this step a relationship between the hourly load profile (section 3.1-a) and P-V curves maximum loading point (section 3.1-b) is determined. a- Initially has to build the hourly load duration curve. b- The ratios between the P-V curves maximum loading point with the corresponding power value is determined. c- To determine the polynomial data approximation, the smallest squares method is used. The following expression is determined:

where b and

+ ⋯

+⋯+

(1)

are constants.

The degree of the polynomial curve compatibility with data in the case of the smallest squares method is provided by R coefficient, whose value 1 determines full compliance of the selected curve with the data. d- We obtain the variation of the power system loading limit change during 24 hours. Knowing that the voltage stability security coefficient is 130%, can be determined the time period in which the voltage stability security criteria is fulfilled. e- The final step consists on substituting the value of voltage stability coefficient 130% in equation (1) and determines the loading coefficient for which voltage stability security criteria is fulfilled. From the equation solution is obtained the limit power values for which the Albanian power system maintains the value of voltage stability security coefficient of 130%. The proposed methodology is applied on the Albanian power system. Figure 5 illustrates the Albanian Power scheme which includes 65 buses and 20 generators.

4. Voltage stability assessment through a new proposed methodology The Albanian transmission system has the longitude profile with lack of generation in the South area. The analysis of load flow results shows a shading profile of the voltage level and voltage stability problems. In paper [8], we have analysed voltage stability through different methods. The proposed methodology is applied on the Albanian power system. The NEPLAN software package has been used to study the voltage stability of the Albanian power system. To illustrate the proposed method, the maximal, average and minimal Albanian power system load of the maximum loading day as worst case is used. Figure 6 represent the lowest P-V curve obtained from voltage stability analysis for three above loads. The critical curve corresponds to Babica 110 node, located in the southern part of Albania. This is an expected result cause Albania power system has a longitude profile from north to south. The main generation units are located in the north Albania while the main consumption is located in the centre and southern part [9]. During the maximal load, the loading limit is 24.75%, lower than 30% which should be the minimal allowable loading limit during the static voltage stability evaluation. This result shows that during

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

Figure 5: Albanian power system

140 120 100 80 60 40 20

Babice110_Max

Babice110_Avg

240

230

220

210

200

190

180

170

160

150

140

130

120

110

100

Babice110_Min

Figure 6: P-V critical curves for maximal, average and minimal load ___________________________________________________________________________________________________________ M. Qemali, R. Bualoti, M. Celo: “Voltage Stability Assessment through a new Proposed Methodology”, pp. 65–73

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

1400

1200

1000

800

600

400

200

0 1

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Generation

Consumption

Figure 7: Hourly generation and consumption curve

maximal load, Albanian power system operates almost in the voltage stability critical limits. While during average and minimal load t the Albanian power system has emerged from the critical zone in terms of the minimal allowable loading limit. Figure 7, shows the hourly loading generation and consumption duration curve for the maximum loading day of year 2013 (15 December 2013). As the next step is determined the values ratio between the P-V curves maximal loading limit obtained from the voltage stability simulation results with the corresponding power value of the operating load as it is represented in Table 1.

Table 1: Maximal loading limit and rated power ratio Operating Regime

Consumption (MWh)

Maximal loading limit P (%)

Ratio P (%)/MWh

Maximal

1359

124.75

0.1

Average

976.5

158.75

0.2

Minimal

594

240

0.4

The second order polynomial approximation of the above selected data (maximal loading limit with the corresponding power value of the operating regime) is as follow: =

Where b and

+ ,

(2) are constants.

Figure 8 represents the second order polynomial approximation for data referred to Table 1, with its corresponding expression (3): = 23.62

36.87 + 138

(3)

Substituting the value of voltage stability security limit coefficient = 130 in Eq. (3) it is obtained the limit power values by which the Albanian power system maintains the value of voltage stability security coefficient of 130%. The value for is defined as: = 1.475

(4)

In the following is evaluated the rated power value which corresponds to the value of x as expressed in Eq. (4) according to the hourly load profile as below: ∗ 0.475 +

(5)

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

From Eq. (5) is evaluated that the power limit value is 1158 MW and this one represents the maximal limit value to which the Albanian power system maintains voltage stability security coefficient value 130%.

5. Conclusions In this paper was presented a new methodology that can reliably assess the security operation of transmission system in conditions when power systems complexity is increasing and they require new methods for the evaluation of the operation security.

So, for Albanian power system loading lower than 1158 MW, Albanian power system satisfies the criteria of static voltage stability. Figure 9 represents the voltage stability security coefficient value and the day period evaluation with secure operation.

The power loading limit and the corresponding voltage stability secure operation day period that satisfies the static voltage stability criterion from P-V curve methods is defined in order to maintain the voltage stability security coefficient to 130%.

In Figure 9, the green area represents the day period with secure operation, which means the hours with the power value lower than 1158 MW. The time limits during which the voltage stability security coefficient maintains the value of 130%, is around 13 hours. So, only 13 hours for maximum loading day Albanian power system operates within the permitted level of security toward voltage stability. During the other 11 hours (mainly corresponds to the maximal operating load), the power system operates in a critical area in terms of fulfilling the criteria of voltage stability.

The advantage of this methodology consists on the fact that it can be applied easily in the dispatching centre. So, the system operator, day ahead can provide the power loading limit and the voltage stability secure operation. In this way, during the risk period (when voltage stability security coefficient is less than 130%) can be undertaken adequate measures in order to maintain the security of power system.

y = 23,625x2 - 36,875x + 138

250

200

150

100

0 1

Loading coefficient

Ratio

Poly. (Loading coefficient)

3 Poly. (Ratio)

Figure 8: The second order polynomial approximation ___________________________________________________________________________________________________________ M. Qemali, R. Bualoti, M. Celo: â&#x20AC;&#x153;Voltage Stability Assessment through a new Proposed Methodologyâ&#x20AC;?, pp. 65â&#x20AC;&#x201C;73

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

240 220 1158

200 180 160 140 120 100 80 60 40 20

0 20 18 19 17 21 16 15 22 14 10 13 11 9 12 8 23 7 24 1 Hourly consumption

P(%)

Poly. (P(%))

Figure 9: The day period evaluation during with secure operation

References

[6]

M.H. Haque, A fast method of determining the voltage stability limit of a power system, Electric Power Systems Research, Vol. 32, pp. 35-43, 1995.

[1]

ERE, ERE Annual Report 2014, Regulatory Body(ERE), Tirane, Albania, 2014

[2]

Panciatici, Patrick, Bareux, Gabriel, and Wehenkel, Louis, Operating in the fog, IEEE power & energy magazine, (2012), pp. 40-49, DOI 10.1109/MPE.2012.2205318.

[7]

K. Vu, M.M. Begovic, D. Novosel and M.M. Saha, Use of local measurements to estimate voltage stability margin, IEEE Trans. on PS, Vol. 14, No. 3, pp. 1029-1-34, 1999.

[3]

Kundur, Prabha, Power system stability and control, McGraw-Hill, Inc., New York, USA, 1994.

[8]

[4]

M.H. Haque, On-line monitoring of maximum permissible loading of a power system within voltage stability limit, IEE Proc. – GTD, Vol. 150, No. 1, pp. 107-112, 2003.

Qemali, Marjela, Bualoti, Raimonda and Çelo, Marialis, Voltage Stability Analysis in the Albanian Power System, Asian Journal of Engineering and Technology, Volume 02, (2014), Issue 02,pp. 119128, ISSN: 2321 – 2462.

[9]

Bualoti, Raimonda, Qemali, Marjela, Gjini, Leonard, Hobdari, Nako, Çelo, Marialis, A Method For Defining The Location Of Facts Devices – Type SVC, MAKOCIGRE 2009, 6. Conference - Ohrid, Maqedoni. (ID0361) October 4-6, 2009.

[5]

C. D.Vournas, Voltage Instability: Phenomena, Countermeasures, and Analysis methods, Proceedings of the IEEE, Vol.88, Issue: 2, Feb. 2000, pp.208 – 227.

International Journal of Contemporary ENERGY, Vol. 1, No. 2 (2015)

ISSN 2363-6440

___________________________________________________________________________________________________________

The Journal

.e8

About the Journal

.e9

Instructions for Authors

.e11