J and t ESMAP fNIR(A~ SI( TOR MANAG[M(NT ASSIS TAN( I PR)(RAM Proceedings of the Regional Seminar on Electric Power System Loss Reduction in the Caribbean Kingston, Jamaica July 3-7, 1989 * ENERGY SECTOR MANAGEMENT ASSISTANCE PROGRAM PURPOSE The World Bank/UNDP/Bileteral Aid Energy Sector Management Assistance Program (ESMAP) was launched in 1983 to complement the Energy Assessment Program which had been established three years earlier. Tho Assessnent Program was designed to identify the most serious energy problems facing some 70 developing countries and to propose remedial action. ESMAP was conceived, in part, as a preinveatment facility to help implement recommendations made during the course of assessnent. Today ESMAP is carrving out proinvestment and prefeesibility activities in about 60 countries end is providing a vide range of institutional and policy advice. The program plays a significant role in the overall intemational effort to provide toechnical assistance to the energy sector of developing countries. It attempts to strengthen the impact of bilateral and multilateral resources and private sector investment The findings and recommendations emerging from ESMAP country activities provide governments, donors, and potential investors with the informotion nr ded to identify economicoally and environmentally sound energy projects and to accelerate their preparation and implementation. ESMAP's policy and research work analyzing cross-country trends and isues in specific energy subsectors make an important contribution in highlighting critical problems and suggesting solutions. ESMAP's operational activities are managed by three units within the EnergV Strategy Management and Assessment Division of the Industry and Energy Department at the World Bank. - The Emy Effaiencv and Strateyv -Un engages in energy assessments addressing institutional, financial, and policy issues, design of sector strategies, the strengthening of energy sector enterprises and sector management, the defining of investrnent programs, efficiency improvements in energy supply, and onergy use, training and research. - The Houshold and Renewable Enemy Unitaddresses technical, economic, financial, institutional and policy issues in the areas of energy use by urban and rural households end small industries, and includes traditional and modem fuel supplies, prefeasibility studies, pilot activities, technology assessments, seminars and workshops, and policy and research work. - The Natuid Gs Dedv*ooment Un.j addresses gas issues and promotes the development and use of natural gas in developing countries through preinvestment work, formulating natural gas development and related environmental strategies, and research. FUNDIN6 The ESMAP Program is a major intemational effort supported by the World Bank, the United Nations Development Programme, and Bilateral Aid from a number of countries including Australia, Belgium, Canada, Denmark, Finland, France, Iceland, Ireland, Italy. Japan, the Netherlands, New Zealand, Norway, Portugal, Sweden, Switzerland, the United Kingdom, and the United States. FURTHER INFORMATION For further information or copies of the completed ESMAP repons listed at the end of this document, contact: Energy Strategy Management OR Division for Global and Interregional and Assessment Division Programmes Industry and Energy Department United Nations Development Programme The World Bank One United Nations Plaza 1818 N Street N.W. New York, NY 10017 Washington, D.C. USA 20433 USA ES1MP/OL&DZ PIOCEKDIIGS OF THE RXGIONAL SE8IR ON ELECTRIC POVER SSTEN LOSS REDUCTION IN THE CAIBBEAN Kingston. jamaji July 3-7, 1989 Energy Efficiency and Strategy Unit Industry and Energy Department World 1hank Washington, D.C. PROCEEDINGS OF THE REGIONAL SEMINAR ON ELECTRIC POWER SYSTEM LOSS REDUCTION IN THE CARIBBEAN, KINGSTON, JAMAICA, JULY 3-7, 1989 Table of Contents Group photograph of participants £ag Preface Introduction i-iii Seminar Agenda iv-vi List of Participants vii-xii List of Observers xiii Opening Address: (1) Gabriel Sanchez-Sierra - OLADE 1 (2) Alastair J. McKechnie - World Bank 3 (3) Orville W. Cox - JPS 7 PAPERS SPEAKERS Power Losses and Development Levels Trevor Byer 11 Detecting Power Losses Using the OLADE Energy Balance June Budhooram 25 Sources of Losses Winston Hay 42 Important Parameters in Loss Calculations Winston Hay 51 Electric Power System Losses - Huntley Higgins & 69 Jamaica Public Service Co. Ltd. - Case Study Raymond Silvera Demand Management Alfred Gulstone 83 Efficient Generation - Diesel Klaas Kimstra 96 Ways to Reduce Transmission and Distribution Losses and Ways to Reduce Transformer Losses Barry Kennedy 114 Engineering Economics of Loss Reduction Projects Barry Kennedy 158 Economic Analysis of Power Loss Reduction Projects Luis Gutierrez 200 Losses Control in Electric Systems - General Concepts and Deflnitions Renato Cespedes 225 A Statistical Approach for Evaluating Some Non-Technical Losses In Power Systems Angel Zannier 250 Corrective Measures for Non-Technical Losses Willy Pacheco 266 The Loss Reduction Experience of the Barbados Light and Power Company Ltd. over the Period 1964-1988 Claude Franklin 293 Micro-Computer Calculation of Technical Losses Okorie Uchendu 302 St. Vincent Electricity Servlces Ltd. - Case Study Lennox Morris 332 ESMAP/OLADE REGIONAL SEMINAR ON ELECTRIC POWER SYSTEM LOSS REDUCTION IN THE CARIBBEAN KINGSTON, JAMAICA JULY 3-7, 1989 Jol j l l}llWslf4 i\iW _ § - -' l v-. . - " r l t 5~ ;gii U Yt.~~~~ , ,,.v1 PREFACE On July 3-7, 1989, the World Bank/UNDP Energy Sector Management Assistance Program (ESMAP) and the Latin American Energy Organization (OLADE), jointly sponsored a seminar for the Caribbean region on the reduction of losses on electric power systems. The purpose of the Seminar was to share with participants the experience of the two organizations on economic methods of reducing both technical and non-technical electric power system losses. Participants, drawn from 20 Caribbean countries, included high-level management personnel in electric utilities and senior civil servants who administer their governments' policies in the electric power sector. ESMAP, the Energy Sector Management Assistance Program, is an international effort in the energy sectors of developing countries, sponsored by the World Bank, the United Nations Development Programme (UNDP) and several bilateral and other multilateral aid groups. As its name implies, ESHAP's major role is to provide technical assistance to developing countries in the efficient management of their energy sectors. OLADE, the Latin American Energy Organization, provides energy assistance to its member countries from Latin America and the Caribbean. The expenditures for the Caribbean seminar were borne by the sponsoring agencies. ESNAP's costs were met from contributions to the program made by the Governments of Switzerland and the United Kingdom. Funds made available to OLADE by the European Community defrayed the expenses which OLADE incurred. The Government of Jamaica graciously hosted the week-long seminar in Kingston. This volume contains papers prepared for the seminar. It has been compiled as a record of the proceedings of the seminar and in the hope that it may be rZ use to persons who were not participants in the seminar but are concerned with efficient power system operations, particularly in maintaining losses at economically acceptable levels. INTRODUCTIOh ESNAP findings in a number of developing countries indicate that electric utilities often issue invoices for less than 80 percent of the energy sent out from the generating stations. The 20 percent difference is classified as losses. In some extreme cases losses exceed 40 percent. In industrialized countries losses do not normally exceed 10 percent. Power system losses may be divided into two categories, technical a.- nontechnical. Technical losses are those which result from current flowing through the transmission and distribution lines, transformers, customer serviced lines, etc. There are also nontechnical losses which are due primarily to human factors including se-,h problems as consumer invoicing, accounting, metering or meter-reading errors, and power theft. Electricity losses represent an economic cost to the country since the resources used to generate and distribute electricity are not being utilized to the greatest productive advantage. They also represent a financial cost to the utility which is thereby deprived of revenues due to it for energy consumed. SMAP__a-d OLADE Partnership Much of E.tMAP's previous work has been focused on making recommendations as to specific measures by which power system losses may be reduced and on preparing evaluations which indicate the approximate costs and benefits of the measures. ESMAP has extended its work in this area by presenting a series of seminars in power loss reduction to qualified personnel in developing countries. The first two of these seminars took place concurrently in Abidjan, C8te d'Ivoire, in November 1987. Participants were drawn from 26 sub-Saharan African countries. One of the seminars was presented in the English language, the other in French. OLADE has also been active in promoting loss reduction in Latin America and hosted a seminar on this topic in October 1988 in Bogota, Colombia. Participants were drawn from the Spanish-speaking countries of the region. As a number of the English-speaking Caribbean countries are also members of OLADE, the organization was beginning to plan a similar seminar in the Caribbean when it was discovered that ESMAP was also planning in the same direction. It was decided that a joint seminar would not only make more efficient use of the resources available but would also demonstrate the commonality of purpose of these two organizations active in the regional energy sector. ESMAP and OLADE agreed that the seminar would be targeted to participants drawn from senior management positions in electric utilities, particularly those persons with responsibility for power distribution and commercial operations. However, civil servants who monitor the performance of the utilities and/or determine electricity tariffs were also considered. In addltion to the Anglophone countries of the region, Haiti and the former Dutch colonies were invited to nominate persons for participation in the seminar. - li - All participants expenses which resulted directly from the ssAinar were borne by the organizers. These expenses comprised the cost of travel and accommodations, including overnight accommodations en route to or from the seminar, where necessary, and the provision of a daily allowance to meet subsisterce costs. Seminar Objectives and proceedinga The seminar was developed to mseet four objectives. The foremost was to improve awareness of the economic and financial costs of power system losses. The seminar was also to be useful in indicating methods by which power system losses may be reduced to economic levels. Because the participants had some experience in confronting these typea of problems and the sessions were intermingled with experts having worked on resolving these problems, the seminar was to be r forum for interchange of ideas on the subject. Finally, the meetings were to be opportunities to encourage innovative approaches to system efficiency improvement. The seminar was pre.ented in Kingston, Jamaica, July 3 to 7, 1989. There were 48 attendees from 20 countries, the countries represented being Antigua, Anguilla, Aruba, Bahamas, Barbados, Belize, Bonnaire, British Virglr. Islands, Cayman Islands, Dominica, Grenada, Guyana, Haiti, Jamaica, Monteserrat, St. Lucia, St. Vincent, Suriname, Trinidad and the Turks and Caicos Islands. Thirty- three of the participants are employed by electric utilities and 15 are in the civil service. Five women were included among the participants. The topics for the formal sessions of the proceedings included the relationship between national development and power losses, the energy balance of the Caribbean, sources of technical and nontechnical losses, identification and reduction of technical and nontechnical losses, economic evaluation of loss reduction projects, computer calculation of technical losses, and efficient operation of diesel generating stations. A major focus of the presentations was the economic appraisal of losses, both technical and nontechnical. They further recognized that projects designed to reduce losses always require investment of some funds. Discussions centered on the fact that the investment required to reduce losses by a given increment increases as the losses get lower. A point, thus, will be reached at which the investment required to produce further reduction in losses Bill exceed the benefits which will derive from the reduction in losses. this point will be the economic level of losses and is dependent on a number of factors peculiar to each system. Partlcipants agreed there is no level of loses which can be selected as the standard to be used as the goal of all utilities. Each loss reduction project must, therefore, before Implementation, be subjected to careful cost benefit analysis to ensure that the benefits are commensurate with the costs. The methodologies by which the costs and benefits of loss reduction projects are evaluated absorbed much of the time devoted to the formal sessions of the seminar. - iii - Three countries in the region, Jamaica, Barbados, and St. Vincent presented case studies of thc'r own experience in loss control. The case studies indicated the successes and frustrations which each utility had experienced and the programs being developed to deal with the areas in which improvement was still required. These case studies were included as one of the means by which the objective of providing a forum fcr the useful interchange of experiences in and ideas for loss reduction programs could be achieved. Field trips also formed part of the seminar agenda. Jamaica Public Service arranged for visits to be made to a number of its installations which proved to be of great interest to many of the participants. These installations included the Rockfort Diesel Barge, the System Control Center, and the Roaring River Hydroelectric Station. Evaluation and Euture Programs Participants were asked to evaluate the relevance and effectiveness of each session and to assess the extent to which the seminar achieved its overall objectives. The average ratings on all counts was encouragingly high, above 70 percent. Among the recommendations made by participants was a similar seminar on the topic of efficient thermal generation, both diesel and steam, which they believed would be of benefit to the regional utilities. ESMAP and OLADE were satisfied with the success of their initial joint effort in seminar presentation. The various sugge:tions made by the country delegates could very well form the foundation on which ESMAP and OLADE will build their next joint venture in the Caribbean. REGIONAL SEMINAP ON ELECTRIC POWER SYSTEM LOSS REDUCTION IN THE CARIBBEAN Wyndham Hotel Kingston, Jamaica: July 2-7, 1989 DAY/D= T1 OU S ACTZvZ?Y SPZAK SUNDAY JULY 2 1500-1600 Registration 1900-2100 ReCepti*D YM Y JUl' 3 030-0U35 RegistratLon 0900-0930 Opeing Address - LAD Gabriel Sanohes-Sterra, Rexcutive Secretary. OLADE 0930-1000 Openin Address - ISWAP Alestair J. XoKohali, Division Chlef, World Bank 1000-1030 Jpening Cerem Orville W. Cox, eiautive ChaLxua, Jamie Public Sezlce CO. 1030-1045 reak 1045-1120 Power Losses and Trevor y"r, Priecipal D.v.lopment Levls Evaluation OffLeor, World Bank l110-12S0 nerw alane of the Csrlbban June 3adhooraw. Chief, Energy Dlane Prora, MADE 1250-1400 1_ 1600-100 Soureso of Losses WLnston Bay, Senior Power asineer, World Bank 150-1531 avow 151-1iSIS _qe rtant Los Pazmts Vinston laY. Senlor Powr Enginer, World Baek 1515-1L50 General. KostrutLem DAY/DATS TDMKIHOURS ACTrVITT 5P1Am TUESDAY JULY 4 0900-0945 Case Study Hu"tley HiBg4ss, Direetor, Engineering * Projeotat Raymond Silver&, Dlrector ef Districts, Jamaica Public Serviceo Co. Ltd. 4S5-1030 Doand Management Alfred Culaton., Senior P.;ir Enineer, World J na: 1030-1045 Break 1045-1145 Efficient Diesel Oeneratian fleas gtmatra, EnaLnecrig Kauager, Sto:k-W.rkspoor DLesel, The Netherlands 1145-1245 Ways to Reduce DistrLbution Losses Barry Kennedy, Consultant. Electrloity Loss Reduction 1245-1400 Luneh 1400-1500 Engineering keonlemic of Loss Barry Keanedy, Consultant, ReductLon Project ElectricLty Loss Reductlon 1500-1515 Break 1535-1615 Problem Solving Barry eanedy, Consultant, ElectricLty Loss Reduction 1615-1630 Anauncemats, genaral Discuse lon 1930-2200 Dime Guest Speaker, Rt Ron Hugh Small, iisdater of Mining and Energy WEDNESDAY J.LY 5 All Day 111 lrip to JusAm Public Servie Co. Inatallatla.ss • Roaring River lyds. Pe"r Station * L_h e Rockfort Dlesel Power Statiln * JPS System Control Centre - vI - DAY/DATE TZIlOURS ACTXVZT! 8s n AKU TSUVDAY JULY 6 0900-0945 Cost/Den fit Analysis of gon- Lul Outierres, Senior Technical Loa Reduction Prosro Energy Econoeist, World bank 0945-1045 Identlileation *nd Clasification Renato Ceepedeo, ProZessor of Non-TSchnic4 Losses Universidad Mac ional do Colombia 2045-1115 Sank 1115-1215 Evaluation of Won-technical Losses Angel Zannier, Manager, Electicity Program, OLADE 1215-1400 Iunh 1400-1500 Corrective Measures for Non- Willy Pacheco, )perational technical Losses Manager, Bolivian Power Company 150U-15S0 reAk 1530-1630 Case study Claude Franklin, Chief Dlstrlbution EngLneer, Barbados Light and Power 1630-1730 Announcements end General Discuseionas FRIDAY JULY 7 0900-1030 MLcro-cqput*r Calculation of Okori. Udhendu# Technical Losses Power Enginer, World Bank 1030-1045 Break 1043-1230 Pael Discuasmon oa Loss Reduction Moderators Rafael moscote, Di7isLon Chief, World Bank 1230-1345 AMb 1345-1445 Case Study L.mio Morris, Plannin Enineer, St. Vincent ElectrliLty Servlces Ltd. 1445-1500 Brek 1500-1600 Coues Evaluation and Certificate Presentations -vii - REGIONAL SEMINAR ON ELECTRIC POWER LOSS REDUCTION IN THE CARIBBEAN Wyndham Hotel Kingston, Jamaica - July 2-7, 1989 LIST OF PARTICIPANTS Title Phone Organisation FAz <:OINTRtY PARTCPAM Addr ss. Tolex ANISGUA & BARUDA Earl Gardner Trainee Engineer Pt(809) 462-1391 Antilu. Public Utilities Authority Ft(809) 462-2516 P.O. box 416, St. Jobnx Antise W. I. Seymour Ilackman Crabbs Power & Desalination P:(809) 462-4990 Plant maaet r Pt(809) 462-2516 Antique Public UtilitLs Authority Tt2090-TELCO P.O. Box 416 St. Johns LAtisue. W.I. AIGUILLA Crfton wiles Electrlial Enginoer P:(809) 497-2651 ElectrLeLty Dpartment Vz'809) 497-3651 Govenment of Anguilla T:9313 ANGO= LA The Valley Anguills, W.I. ARUBA Raclro Stemn do Cuba Senair Power Znglneer Pt2978-24600 Water-en Energlebedrijf Fs P.O. Box 575, Oranjestad TS100-WEBAR AW Aruba THE AABMIAS Edward oleasle &ast. Cbenerl Mnaser, Ps(809) 328-7700 systam Development Jt(809) 325-6852 Bahaas ElectricLty Corporation T:297-2-20396 ElecttLe Nassau Bohan"s Cyril Thompson Mnager. Pt(809) 328-7700 Distributinm Maintenance Vs (q9) 323-6852 Babma IlectrLaity Corporation TS297-2-20396 Electric P.O. Sri *-7509 Nassau Bahaas" - viii Carl Roll. Consutmr Accountant PS:809) 328-7700 Dabemao Electrilcity Corporation Ft(809) 323-6"52 P.O. Box 3-7509 T:297-20396 Nasau Behesa ZARBADOS Yvette Davis Planning Egineor P:(809) 429-3000 Barbados Ltght & Power Co. Ltd. 7:(809) 436-9933 P.O. Zon 142 TKilcvattI2241 W1 Bridgetown Barbados Michael Labhley Trainee Distribution Enginser Pt(809) 429-3000 Barbados Light & Power Co. Ltd. Ft:809) 436-9933 P.O. Box 142 T:KilovattI2241 VB BrLdgetovn Barbados Claude Franklin Barbados Light & Power Co. Ltd. P:(809) 429-3000 P.O. Box 142 Ft(809) 436-9933 bridgetoun T: Barbados BELIZE P xnando Co7e General Manager Engineering P:501-2-45234 Solis1 Eletriceity Board F:501 2-30891 115 Barrck Road T:333 272 32 B*lise CLty 3.1Ls Central Aerica BORAIRS Joban CGikus Managing Director P:599-7-8756 WEB onaire Ft599-7-8756 Magrietstraat TS Bonaire The Netherland Antlles 331S1SS VIRGIN. Ronnie Skolton Cneral Manager PC(809) 494-3911 ISlANDS B ElePctrlicty Corporation 7:(809) 494-4291 P.O. Box 260 1:7974 TRICITY DV Road Town TortoXL brltlsh Virgln lsl nde No. Mani" Allen Assistant Secretay Pt(609) 494-2213 Ministry of Conmicatlons a works Fs(809) 494-4435 Road Town S17959 CRAD Tortola Britleh Vlrgin Islands - ix - N.. Marvs Thompson Chlef Accounteft P:(809) 494-3911 3V1 ElectriLity Corporation tS(809) 494 4291 P.O. Box 268 Ss7974 TRICIT SV Road Town TortolalBritish Virgin Island& CAYHAN ISLANDS John Sav Trainee Engineer P.(809) 949-5300 Caribbean Utlilties Co. Ltd. F,(809) 949-5203 P.O. Box 38 Tt4235 CUCDRAN CP Grand Cayman Cayman Islands BAI.I. DOPINICA Wentm Dorsett TransmissLon & Distribution EngiLner P:(809) 448-2681 Dosiale- ElectricLty Sorvices Ltd. F:(809) 448-5397 18 Castle Street, P.O. oan 5 3 T:8655 DOHLEC DO Roseau Dominlia Coomnvwealth of DomnLLca GREMADA George Rrdiz Electrical EngLneer P:(809) 440-3166 Grenada ElectrLcity ServLcoes Ltd. F,(809) 440-4106 P.O. Box 381 1T3472 GRENLEC GA St. George's Grenada GUYANA Ms. Varlyn KlaUss Strategic Planning Manager P:592-2-57778157091 Guyana ElectricLty Corporation F:592-2-71978 40 Main Street T:GY 2250 eorgetown Guyana Lalta Durbal Enginser P:592-2-66993/58569 Guysna National Energy Authority F:02-71211 295 Qusmina Street T.2253 UMDA GY Cuomingeburg Georgetown Guyana Oscar Spencer Dep. Chief Electrieal Engineer Pt592-2-66171166181 Guyana Sugar Corporation Fs592-02-57274 22 Church Street TsGY 2265 Georgetown Jacob Ebmllton Act. Sup. Power * UtilLties Dept. P.592-4-2779 Guyana Mlnng Eaterprlse Fs592-4-2795 NbXenAL S:2243 Linden GuYa- - x - MAITI Plerre HtonuS Read Energy Plann1S Tern Ps509-6-1517, 6-1163 Bureau des ins.. et de Il'nrgie Ft P.O. box 2174 1SIIARLE 2030246 Port-au-Prince, Haiti Erle Nicola" Coordintor Poawr Loss Activities P :509-3-4600 (zut 208) Electricit d'HaitL p13-S750 PO IBox 1753 S 2030113 Port-au-Prince, Haiti Jean Pauyo System Planning Supervisor Pas09-3-4600 (Ext 247) Electrieite d'Haiti V 3-8750 P0 Box 1753 1S2030113 Port-au-Prince, Haiti JAMAICA Huntley Higgins Director Engineering & ProJects P:(809) 926-3190 Jamaiea PubiLe Service Co. Ltd. 1s(809) 926-6710 6 Knutsford Bouleard T12180 JAKSBRV P.O. Box 54 Kingston Jamaca Raymond Silver& Director of Distrtcts P:(809)-92-68049 Jamaiza PubiLe Service Co. Ltd. PtC809)-92-66710 6 Knutsford Boulevard Ta2180 JMSERV P.O. Box 54 tingston Jamaica W. I. Mrs. Melody Daley Dlrector, EconomLe Planing Unit : (609)-962-6917019 SErg DlvLsan Ft:(809)-926-2835 MlnLitry of ining & Energy t:2356 PETCORP P'J Resource Center 36 Trafalgor Rpad Kingston Jamie& V.I. Ainsworth Lawson ChLf Eginee Ps 809)-926-9170 Enerw DLvtosin 1s809)-926-2855 Ministry of Mining & Enrw 1s2356 PETCORP PCJ Resource Center 36 Trafalgar Road Kingston J_lice V.X. - xi - NONTSERRAT Ronald Bowman Managing Dlrector P:(809)-491-244112 Montserrat Electricity Services Ltd. F: C/o (809)-491-3599 P.O. Box 16 TS5708 MONLEC MX Church Road Plymouth Montserrat W.I . Lenox Browna Chief Engineer Pt(809) 491-244112 Montserrat Electricity Servlces Ltd. PCIo (809)-691-3599 P.O.Doz 16 TS5708-MONLEC MX Church Road Plymouth Montserrat W.I. ST. LUCIA Errol Bartley Chlef Engineer/Project Mangeor P:(809) 452-2324 St. Lucia Electricity Service. Ltd. F:Clo (809)-45-23313 P.O. Box 230 TS6421 LUCELEC LC Castries St. Lucia W.I. Ronnie WLilLam Trans. & Distribn. Engineer P C809) 452-2324 St. Lucia Electricity Services Ltd. F:CIo (809)-45-25313 P.O. Box 230 T:6421 LUCELEC LC Castries St. Lucia W.I. ST. VINCENT Paul Soleyn Assistant Planning Engineer (T&D) P:(809)-456-4123 THE GRENADIhES St. Vincent Electricity Servlces Ltd. F:(809)-456-2436 P.O. ox 856 TS7581 VINLEC tingatown St. Vincent L Light ........ > G T UE FC4 F Nat. Gas E > & > > Mech. ...N D I L>-N DHeat NTL ETC. >Etc. GL TDL FUL Fig. 4: Simplified Power System Flow In physical terms we have: (a) GL - Generation Losses - 100 - nG (nG is the efficiency of the Generating System. In terms of the transformation from a primary source of energy to electricity). nG - 100 EP/I (b) LTD - Transmission & Distribution Losses - 100 - nTD (nTD is the efficiency of the T&D System) ntd - 100 FC/EP (c) FUL - Losses in converting Electricity to energy forms, heat, light, etc. - 100- nFU (nFU is the efficiency of the converting device: bulb, stove, etc.) nFU - 100, UC/FC It is intuitively clear that within the context of power loss reduction one can improve economic efficiency by simply increasing the productive efficiency of the generating system by reducing the technical losses LEG without compromising the quantity of energy (kWh) generated. In fact, to maintain equal levels of electricity generation less fuel will be needed thereby reducing fuel costs in the case of thermal installation. 34 a Parallely, there is an optimization of capital investment as well as improved operation of the generating system by reducing their power lodses (MW) which have a direct economic benefit which at a minimum improves the quality of service and possibly permit more load to be served and/or delay the expansion of generation and transmission facilities. It is worth noting here that in high loss systems the outlays required to achieving energy and power savings are generally very much less than the cost of increasing supply capacity. Within the Transmission and Distribution System, the losses are mainly due to heating in the system: On the average, these losses should technically be below 10% of gross generation while economically optimal loss levels may be as low as 5%. These T&D losses are usually very high and may approach some 20% of gross generation of which three-fourths occur at the distribution level. It is not surprising then, that these are economic losses and do not only adversely affect the financial state of the utility iteself, but also the national economy. The energy that is lost, or more accurately wasted, due to technical inefficiency, could satisfy additional incremental demand or load. This may generate even more savings of national resources that are assigned to produce electricity. It is true that the principal reasons for the existence of unacceptably high levels of technical losses in many of our countries are the decline in the financial positions of the power utility and scarcity of foreign exchange resources which has led to reduced investment in system maintenance and rehabilitation in spite of the fact that the prices of copper and aluminium the main components of the distribution hardware system have declined considerably. Today the utilities can put in a lot more of the relatively cheap hardware to reduce the more expensive losses. On the demand side, at the point of the consumer and the consuming energy devices (electrical machines and appliances), technical losses FUL also occur. In fact, the energy conversion process, from electricity to other forms of useful energy has inherent energy and power losses. It becomes evidently clear that electricity and power loss occur not only in the power delivery system, but also at the end-use stages. It is convenient here to briefly mention measures to improve end-use losses before discussing the convenience of their optimizaion in the transmission and distribution system. Energy conservation at the end-use stage may be achieved by two principal methods: (1) improving the technical efficiency of energy using devices and appliances, and/or (2) modifying the shape or characteristics of the load through demand management techniques. Nore often than not, the public utilities are more concerned with the latter method, which is based on the fact that it is more costly to supply electricity during peak periods (seasonal and daily) rather than off-peak periods. Therefore, changing the shape of the power utility's curve by shifting electricity consumption from peak to off-peak periods will effectively reduce the cost of supply and at the same time, conserve energy. Until now, discussions have been oriented to technical losses in all the components of a power system. However, besides technical losses there are - 35 also non-technical losses associated with the system. These non-technical losses refer to the energy that is consumed and not billed by the utility. The main source of these losses are billing errors, meterlng errors, unregistered customers and outright theft. Non-technical losses are primarily financial losses to the utility. Their main impact is evidently on the financial position of the utility itself. The revenues that are lost when electricity is consumed but not paid for by the consumers impose a heavy burden on the financial viability of the utilities both directly and indirectly. Referring to Figure 4, non-technical losses are represented by the energy flow NTL. It is noteworthy, that they are not losses in the technical sense. NTL represents that portion of electricity which is consumed and not billed. Non-technical losses distort the optimal pattern of electricity consumption, which represents an additional cost to the economy. In utilities where tarlffs reflect costs (operating and investment), non fraudulent consumers are charged for the service relatively more than they should and as a consequence, they tend to consume less because the higher cost of those who do not pay for electricity is passed to them through higher tariffs. Conversely, those who do not pay for electricity, tend to consume more than if they had to pay for it. This feature creates a distortion in the economically optimal electricity consumption pattern. - 36 - 4. THE ECONOMIC AND FINANCIAL BENEFITS FROM LOSS REDUCTION PROGRAMS It has already been mentioned that a reduction ln power and energy losses, but technical and non-technical, may have lmpacts to the utilities and to the economy at large. Nevertheless, it may be convenient to analyse in more detail, possible direct and indirect benefits and to identify the beneficiarles, as a result of comprehensive loss reduction programs. From the national economLc perspective, it is more convenient to analyse loss reductlon programs, taking into account all power system components as an integrated unit, comprising the generation, transmission, dlstribution and final user sub-systems. This approach, although unusual to most power system analyses, have the advantage of considering the electricity sector as another component of the macroeconomy. In thls regard, this component produces an intermediate pr4duct which is used for the productlon of other goods and services in other economic sectors. Diagramatically, this unit can be represented as follows: I Electricity* > UE ** TL - Total Power System Losses * Includes generatlon, transmission, dlstribution and final user sub-systems. ** TL includes CL, TDL and FUL. FiLgure 5 In the above Figure 5, TL may be interpreted as total physical losses, eLther power (NW) or energy (kWh). In monetary terms, these losses may be "financial or economic", dependlng on the conversion factor utillzed (market values or shadow prlces respectively). Therefore, lt becomes evldent that a reduction in total power system losses results in economic and financial beneflts for the whole economy. In a similar manner, loan reductLon programs '.n each one of the components of a power system imply ecoromic and financial benefits. accruing to various benefLciarLes comprising: (a) At the macroeconomic level - Ministeries of Finance; Planning; and Economy who can target savings in the development of other sectors l.e. health, education, transport, etc. (b) At the intermedLate level - The Mlnistry of Energy which can re- allocate these savlngs for intra-sectoral development. AddLtionally -37 less fuel requirements will imply less import pressures for OIDC's, while more exports revenues for OEDC's. (c) At the micro level - Power utilities and Electricity consumers, both benefit because on the one hand, the operational aud investment costs, for power delivery may be considerably reduced, while on the other, the consumer by increasing his electricity consumption efficiency receives the same service at a lower cost. Programs aimed at reducing losses for specific components of a power system, may have different impacts on the various beneficiaries mentioned above. 4.1 Loss Reduction at the Final User Level Although the loss reduction programs at the final user level may appear to be beyond the scope of power utilities, it is important to highlight the potential benefits derived from such actions. BE . L Final User FC (FU) > UE NTL I 4 FUL Figure 6 Figure 6 graphically represents the final consumer system, with its corresponding input and output energy flows. A reduction in the final use losses (FUL) is directly proportional to a reduction in the final demand of electricity (FC) entering this sub-system. The resulting effect of this phenomenon are: (a) The consumer consumes less energy and subsequently lowers his energy bills (b) The utility, receives less revenues, but at the same time, reduces its operating costs and has the potential to delay investments. 4.2 Loss Reduction at the Power Production and Deliverv System A reduction of technical losses at the power delivery system (generation, transmission, and distribution), shall produce economic and financial savings to the various beneficiaries as follows: - 38 - Power Production and Delivery ' FC System (C. T6D)) * GL + T&DL - ESL Figure 7 A loss reduction in the electricity production and supply system is guaranteed to: a. Reduce operating expenses in the system. b. Delay investments in expanding generation, transmission and distribution facilities. C. Improve quality of service. d. Permit more load to be served with existing capacity. 4.3 Non-Technical Loss Reduction So far the present analysis has been dealing with the benefits derived from reducing technical losses. However, a more careful examination of Figure 4 reveals that there exists other potential areas to further improve efficiency by reducing non-technical losses, which would: (a) Improve the financial performance of utilities by means of increasing revenues, derived from previously non-billed consumers. (b) Normalize the electricity consumption pattern, as result of billing previously fraudulent consumers. 5. POTENTIAL ECONOMIC BENEFITS AS A CONSEQUENCE OF AN AGGRESSIVE LOSS REDUCTION PROGRAM IN THE CARIBBEAN On the basis of information extracted from the OLADE's energy balances for the Caribbean member countries, estimates of the potential benefits derLved from an active technical loss reduction program, at the generation level, have been calculated. The purpose of that calculation is to highlight the often neglected potential to improve efficiency at this level and to have an order of magnitude of resulting benefits. To do so, data corresponding to the spectrum of primary energy input into the power plant, auto-consumption and gross generation for the period 1977 to 1986, has been utilized. Table 1 presents a summary of the basic data. - 39 - Table 1 : PRINCIPAL ENERGY FLOWS IN THE ELECTRICITY SUBSECTOR IN THE CARIBBEAN (BOE x 10A3) Year PRIMARY ENERGY LOSS AND GROSS INPUT AUTO-CONS. GENERATION 1977 19731 14660 5071 1978 19883 14609 5274 1979 21341 15738 5603 1980 21020 15220 5800 1981 21267 14916 6351 1982 21214 14781 6433 1983 23285 16523 6762 1984 23789 16842 6947 1985 23324 16075 7249 1986 25836 18267 7569 Source: OLADE Energy Balances for Latin American and the Caribbean 1970-1986. A diagramatic scheme of energy flows presented in Table 1 may be represented as follows in Figure 8 below: Imports Schematic Representation of the Gross J -z=_ Electricity Generation System >-- Energy in the Caribbean Gen. Net Gen. I nput Exports Auto-Consumption L and Losses Figure 8 Using the above data, loss figures for the generating system have been calculated (including auto-consumption). No particular trends have been observed and with an erratic behavior they range from 69% in 1985 to 74% in 1977, with an average of 71% for the period in question. Net generation has baen adjusted mathematically, using the least square linear regression method of their natural logarithm. A 6.6% annual growth rate has been obtained with a correlation coefficient of 0.98. A net generation projection has been performed on this basis and two different estimates of primary input energy have been calculated. The first assumes that losses during the next decade will remain in the average value of 71% observed during the period 1977-1986. The second estimate assumes a constant loss value of 66% - 40 - implying on average a 5 loss reduction program in the generation systems of the Caribbean. Results are presented in Table 2. Table 2: PROJECTIONS OF NET GENERATION AND INPUT ENERGY FLOWS FOR DIFFERENT LOSS VALUES IN THE CARIBBEAN Net Gen. Input Energy Input Energy Year (GWh) Loss - 71% Loss - 66% BOE x 10^3 BOE x 10A3 1989 14116 30702 26118 1990 14762 32106 27313 1991 15437 33575 28562 1992 16143 35111 29868 1993 16881 36716 31234 1994 17654 38396 32663 1995 18461 40152 34157 1996 19306 41989 35719 1997 20189 43909 37353 1998 21120 45918 39062 The above results are the basis to evaluate the potential benefits of an eventual loss reduction program. Energy savings have been quantified and evaluated considering US$20 per barrel of fuel. Assuming an average 0.53 load factor, potential power expansion delays have also been estimated, somewhat in the order of 715 NW during the next decade, with their corresponding economic savings evaluated at US$500/kW. Loss reduction costs at the generation level, have been estimated at US$200/kWh delayed. Results are summarized in Table 3. Table l: EVALUATION OF POTENTIAL SAVINGS Year Energy Savings Power Net Savings NK of US$ NH of US$ 1989 92 287 1990 96 300 1991 100 314 1992 105 328 1993 110 343 1994 115 359 1995 120 376 1996 125 392 1997 131 410 1998 137 435 Total 1,131 3,544 - 41 - 6. CONCLUSIONL AND RECOMMENDATIONS From what has been discussed, the following conclusions may be draun: (a) That against the background of the recent critical socio-econoaic developments in the region there is an urgent need to formulate susb- sectoral electricity policies not only at the sectoral energy leve. but also within the macroeconomy. (b) That given the problems that have affected the electricity subsector, there is an urgent need to improve its efficiency in economic, financial, technical and administrative terms. In this regard, loss reduction policies can play a major role by means of reducing power and energy requirements, resulting from improved efficiency at the power system production and delivery stage, as well as at the final consumer level. (c) The loss reduction programs would appear to be the most beneficial policy option within the short term provided that implementation costs do not exceed the alternative expansion plans. In other words the cost of saving a kW/kWh should be less than the cost to the economy of providing an additional kW/kWh. (d) That the estimated potential benefits of a '088 reduction program for the generating systems of the OLADE's Caribbean member countries merit further consideration. Although it has been reported that, up to three-fourth of losses do occur at the distribution level, it may be worthwhile from a national economic perspective, to further analyse the loss-saving potential at both the generation and final user levels. - 42 - SOURCES OF LOSSES By Mr. Winston Hay, Industry and Enegy Department, World Bank Introduction As electric power is transmitted from the generating station and distributed to the ultimate consumer, the loss of a certain amount of the energy sent out from point of generation cannot be avoided. Previous papers have shown that these losses result in economic costs to the country and financial costs to the utility. In this paper, we will take an overview of the sources and niature of losses and discuss briefly some of the approaches by which they may be reduced. Figure 1 is a very basic diagrammatic representation of an electric power system. The power is generated in the station (section A) after which the voltage is increased in the step-up transformer (B) to the level at which the transmission lines operate. At the end of, or at points along the transmission line (C) transformers in the substation (D) reduce the voltage to a level appropriate for the primary distribution feeders (E). Distribution transformers (F) along the feeders further reduce the voltage to energize the secondary distribution lines (G) from which the service drops provide the supply to the consumers' meters (H). This diagram incorporates the major components of all systems. Small systems may not have transmission lines, the primary feeders being fed directly from the generating station. Most systems will also have a number of large consumers who obtain their supply directly from the primary feeders or even, perhaps, from the transmission line itself. Traditionally, such systems have been designed with the major objectives being to: (a) supply the maximum consumer demand with a constant and acceptable voltage level at all points of supply, even at the end of the distribution feeder; (b) ensure a reliable supply to the consumer; (c) minimize investment and operating costs; (d) ensure the safety of person and property of consumers and utility employees alike; and (e) easily accommodate expansion of the system as demand increases. Losses were seldom a consideration in the planning and implementation of such systems. With time, however, it began to be realized that the least investment cost would seldom represent the least total cost, and that losses represent real costs to the utility and to the country. This realization began to gain wide attention with the increase in the price of energy in the 1970s. .43 - 8 . tt S i~~~~~~. 11 -'-- - - - -, SI Ji|= _ _I _ _ ,; I -44- All of the components which form the building blocks of our model system will result in some degree of energy loss. In this pape- we will not deal with losses in the generating stations, although these are usually significant and must be included in an= thorouLh evaluation of system losses. System Energy Losses We will define system energy losses over any given period of time as the difference between the energy sent out from the power stations and the energy invoice otherwise accounted for during that System Energy - Energy Generated at - Energy Delivered Losses Power Stations to consumers Losses are sometimes spoken of as being only the difference between the energy sent out and that for which the utility invoices its consumers. However, not all consumption is invoiced. Very few utilities invoice themselves for their own energy consumption. It is, however, important from the standpoint of good energy accounting that all consumption be metered. In order to know the true extent of losses, the authorized consumption, invoiced or otherwise, must also be known. In discussing power system losses, it is customary to speak of two different classifications--namelv Itechnical" and Onon-technic-1.n Technical Losses The technical losses are those which inevitably result from the flow of electricity through the components of the system--through transformers, the transmission system, primary and secondary distribution lines, the service drops and even the consumers' meters. In this process, a certain amount of the electricity will be converted into heat. Technical losses can be further subdivided into 'load losses" which result from the resistance of the circuits in which the current flows, and 'no-load' losses which occur primarily in inductive equipment such as meters, motors, and transformers, and which are independent of consumer demand. They continue even when there is no such demand. From the standpoint of electric power systems no-load losses normally need to be considered only for transformers, since metering losses are very low and there are not many motors as components of the utility's power system outside of the customer load. The no-load losses of a transformer comprise the eddy current, hysterisis and dielectric losses, as well as the resistance losses resulting from the exciting current in the prLmary winding. The technical losses on electric power systems are predominantly resistance or the so-called "12R" losses. These losses may. therefore, be lowered by reducing the current, the circuit resistance or both. However, since the losses vary with the square of the current but directly with the resistance, loss reduction measures which reduce the current are proportionately more effective than those which reduce the resistance. For any given system demand the current may be reduced by increasing the operating voltage or by improving the power factor. The circuit resistance may be reduced primarily by increasing the size of conductors, changing the conductor material and/or replacing existing - 45 - transformers with others of lower load losses. The ways in which these loss reduction measures may be accomplished are dealt with in greater detail in subsequent presentations. Discussions of ways and means by which losses may be reduced must necessarily consider details of the sources of losses, and it may be useful to take a somewhat superficial look at these sources at this point. SteR-Up Transformer Returning to Figure 1, and tracing the flow of current from the generating station to the consumer, the first major system component will be the step-up transformer. The losses in this transformer will consist of the no-load portion which is constant as long as the transformer remains energized, and the load portion which will vary with the square of the current, and therefore with the square of the load if a constant power factor is assumed. Once the transformer has been purchased and installed, there is little that can be done to reduce its losses, short of replacing the transformer entirely. This lack of flexibility in controlling transformer losses makes it important to properly evaluate the losses before the transformer is purchased. The no-load losses are sometimes overlooked in these evaluations, but recognition of the fact that they continue effectively for 8,760 hours/year will indicate that they could possibly exceed the load losses. In any case, they will be significant. Transmission System The current next flows through the transmission system. Transmission losses will vary with the square of the current and directly as the resistance of the line. The magnitude of the current will be determined by the total load (consumer demand plus losses), the operating voltage, and the power factor. Line resistance will be a function of conductor material, conductor cross section, and the length of the line. Losses can therefore be reduced by increasing the operating voltage, improving the power factor, changing the conductor to one made of material of lower resistivity, increasing the conductor size, and/or reducing the length of the line. Of these alternatives, increasing the operating voltage and increasing the size of the conductors are the approaches most often used on transmission lines. Running an additional parallel line is effectlvely the same as increasing the size of the conductor. It is also possible to reduce the resistance by changing the material used for the conductor, but there are other factors which influence the choice of conductor material and these are often more important than variations in resistivity. It is interesting to note in this respect, however, that all-steel conductors were once fairly widely used but are now seldom, if ever, encountered on new power lines. This is because of the relatively high resistance of steel in comparison to its weight. The options for shortening the length of a transmission line are generally restricted but ought not to be overlooked. Power factor is best corrected at the source of the demand for reactive power. Any power factor correction on a transmission line should therefore be 46 - to supply the reactive power demands of the line itself. Transmission lines have both capacitive and inductive effects but these tend to be small except on long, high-voltage lines. They also tend to offset each other, the dominant effect depending on the length of the line, the voltage, and the load. Power factor correction om transmission lines is therefore normally undertaken only on long, high-voltage lines. It needs to be carefully implemented to maintain slightly lagging power factor at the source since leading power factor will result in negative generator voltage regulation. The transmission voltage is reduced to that of the primary distribution system in the substation transformer. Loss considerations for this transformer are the same as for the generator step-up transformer. Primary Distribution System Similarly, the sources of losses, and consequently the possibilities of loss reduction, on the primary distribution system are essentially identical to those of the transmission system. The emphasis may, however, be different. Line reactance is of even lower significance for losses. Opportunities for shortening the lengths of primary distribution lines are generally encountered more often than is the case for transmission lines. This is because of the possibilities of switching loads from one line to another in close proximity, and extending additional feeders from the same or another substation. Power factor correction is also most often installed on the primary lines. Although the reactive demand may be predominantly on the secondary systems, it is generally more economic to install the capacitors on the primary system. The capacitance of a given capacitor varies with the square of the applied voltage. A capacitor of a fixed size is therefore more effective on the higher voltage of the primary system than on the secondary system. Primary lines are sometimes run as two phase or single phase extensions. Increasing the number of phases in such instances will lower the losses by reducing the current per phase. Distribution Transformers The distribution transformer links the primary and secondary distribution systems. Again, the no-load and load loss considerations apply. Often the ratio of average load to peak load is lower for distribution transformers than it is for substation or generator transformers, so that the no-load losses assume even greater significance for overall losses. It Is dangerous to generalize, however, and the economics of transformer loss reduction should be applied individually, or in groups which are known to comprise units of very similar load characteristics. - 47 - Secondarv Distribution Loss reduction on secondary distribution lihes is subject to the same considerations as primary lines. The possibilities of reducing losses by shortening the lines are even greater, as it now often becomes simply a question of installing distribution transformers at more frequent intervals. The other options of increasing conductor size or the number of feeders remain applicable. If power factor correction is undertaken on secondary lines it will be especially effective at the points of supply to large conL.Amers with high reactive demands. Lines of less than three phases are more often encountered, so that there will probably be greater scope for reducing losses by increasing the number of phases. On the other hand, the supply voltage is normally standardized, effectively ruling out voltage increase as an approach to loss reduction. The final link in the supply chain is the service drop from the secondary line to the consumer's meter. From the standpoint of losses, the important consideration in this instance is to ensure that the line is adequately sized. However, it is false economy to attempt to save investment costs by carefully sizing the service line to each consumer's existing demand. That demand will almost always increase with time, but checks on adequate service line sizing to existing consumers are very seldom made before problems develop. Inductive Reactance The energy losses in a transmission or distribution line are determined by the resistance of the conductors. However, the reactance of these lines is also important because of its effect on voltage drop. The reactance of a line often results in a greater voltage drop than that produced by its resistance. Although the reactive voltage drop does not result in energy loss it can have a critical effect on the quality of supply to power consumers and also on the quantity of energy sold. The latter will, of course, have an effect on the financial returns on the utility's investments. The reactance of lines can be easily calculated. In general reactance varies inversely with the radius of the conductor and directly with the conductor spacing. In both instances the variation is a logarithmic function. OTHER SOURCES OF LOSSES Two additional sources of technical losses are frequently mentioned. They are high-resistance joints and vegetation in contact with the lines. In all probability, these do not contribute significantly to overall losses. Both are far more likely to affect reliability of supply than losses. High resistance joints are especially a problem with aluminum conductors because the oxide on the surface of the metal is not a good conductor of electricity. The energy lost in the joints is transformed into heat, and the joint consequently gets hot. If a lot of energy is being dissipated, the metal will melt and the joint will - 48 - become open-circuited. Unless burning of conductor joints is a frequent occurrence, the losses at these joints are probably negligible. Similarly, vegetation in contact with uninsulated power lines will burn away if they conduct significant amounts of energy. This is especially noticeable on lines which operate at or above 12,000 volts. Nevertheless, good housekeeping demands that conductor joints provide good contact and that power lines be kept free of vegetation. In concluding these discussions of technical losses there is a point not yet focussed on but which ought not to be overlooked. It is that losses increase the power demand on the system and that this increase In demand is itself the cause of further losses. The losses at one end of a secondary distribution line will be the cause of an increase in the power generated and in the total current flow through, and therefore in the losses in the various transformers, the transmission and primary distribution lines, the secondary distribution system upstream of the end point. Technical losses, therefore, are themselves the cause of further losses. Non-Technical Losses Non-technical losses represent energy consumed for which the utility does not receive revenues. The primary sources of non-technical losses are: (a) Unmetered supplies. These may be due to direct connections made by consumers who wish to enjoy the benefits of electricity without having to pay the cost. It can also result from direct connections made by the utility because of meter shortages, and which were never regularized. Customers can be removed from the accounts records while their premises are still connected to electricity. In some countries supplies to government offlces, public light, and certain other public facilities are not metered or even reliably estimated. (b) Defective metering. This may be the result of the meter having been tampered with in an attempt at fraud, or the meter may be incorrectly connected, have been installed with defects or have developed defects after installation. (c) Meter reading errors. The meters are either incorrectly read or not read at all. (d) Billing deficiencies. Errors may be made in the calculation of the bills, bills may not be prepared for certain consumers because meter readings were not received or for a variety of other reasons, bills may be prepared but not delivered, etc. Non technical losses have a direct impact on the finances of the utility affected. With the increase in energy prlces in the 1970s, a large number of utilities experienced an increase in losses as more and more consumers attempted to avoid having to pay the increased costs of electricity supply. Many -49- utilities, particularly in developing countries, have not been able to implement the controls which would reduce the theft of power to a minimum. If total system losses exceed about 158 of net generation, it is almost certain that non-technical losses constitute a significant contributor to the total. This is not to imply that these losses are insignificant if the total percentage is lower than 15. It is best to make a thorough calculation of the technical losses and so develop a realistic estimate of the extent of the non- technical. Tar2et Loss Levels Reduction of losses, especially technical losses, will require investment of capital. For any given system, there is therefore an economic level of losses at which the benefits to be obtained from additional investments would not be recovered in the benefits of further reduction in losses. This optimum loss level is, however, dependent on a number of factors which, in combination, are peculiar to each utility. No specific target level for technical losses can be given as being applicable to all utilities. In general, however, it can be confidently said that if the technical losses exceed 10% of net generation, there is room for economic reduction of losses. The case of non-technical losses is simpler. There the target level ought to be zero. Unlike technical losses, non-technical losses are not inevitable, and very often dramatic improvements can be achieved in this area without significant investment of capital. Reduction of non-technical losses is primarily a matter of good management. There are many utilities, including several in developing countries, which have levels of non-technical losses which are effectively zero. Conclusio.ns Per capita electricity usage is one of the measures of economic development and quality of life. Losses, however, are a useless waste of resources. If losses were totally eliminated, the consumer would still be able to have his electricity demand satisfied but the resources required to satisfy that demand would have been appreciably reduced. Technical losses cannot be completely eliminated but they can be controlled. Non-technical losses can be effectively reduced to zero. The extent to which a utility can economically reduce losses is therefore one of the standards which may be used to gauge the efficiency of its operations. In particular, of its management. mw- - 50 - BILIOGRAPY 1. Energy Efficiency: Optimization of Electric Power Distribution System Losses. By: M. Munashinghe and W. Scott World Bank Energy Department Energy Paper No. 6 July 1982 2. Electric Distribution Systems EngineoerLng Manual, Volume 1. By: EBASCO/Elctrical World McGraw Hill Publications 1982 3. Power System Efficiency Through Loss Reduction and Load Management By: Asian Development Bank 1985 - 51 - INPORTANT PAURETSERS IN LOSS CALCULATIONS By Mr. Winston Hay, Industry and Energy Department, World Bank Xj.groductio Any program to reduce technical losses will involve the investment of money. Typically the incremental effect of each dollar invested decreases as the losses decline. For each system or subsystem, there is therefore an optimum level of losses at which the costs of further reduction in the level of losses will exceed the benefits which result. There is, however, no specific loss level which is optimum for all systems. That optimum is influenced by a number of characteristics which vary widely from system to system. Some of these characteristics, such as the incremental costs of demand and energy losses, have been dealt with in a previous paper. Energy losses may cost less, for instance, on a system supplied from indigenous hydro plants than on one in which the power is generated by steam plants fired with imported oil. The economics of loss reduction will also be influenced by factors such as the size of the system, consumer density, average consumption, etc. Especially important is the pattern of load demand with time, as characterized by a number of system parameters. We will cancentrate on some of these in this paper. Table 1 below shows the hourly demands on a hypothetical feeder for a given day. The demand profile is graphically presented in Figure 1. !abl.1S E¢ur Lad Hour Load - - ' (kV) '(kMl) 1 320 13 400 2 310 14 400 3 300 15 420 4 300 16 420 5 s30 17 360 6 350 16 550 7 350 19 660 * 30. 20 620 * 380 21 520 10 400 22 400 11 400 23 360 12 400 24 330 The shape of the curve in Figure 1 is characteristic of feeders with a preponderance of residential consumers. The demand is at its minimum in the early mornLng hours, Increases slightly durlng the business hours and experiences a sharp increase with the onset of darkness. The absolute peak occurs between 7:00 p.m. and 8:00 p.m., and the demand then declines relatively gradually to the early morning. Figure 1 FEEDER NO. I HOURLY DEMANDS TOTAL DEMAND AND CONSUMER DEMAND (RESIDENTIAL) xn~~~~~~~~~~~~~~~~~~~~~uDni 2W 101rA sxJ o i, -. - I -. I . I £ I I I I I I I I I ! I a p I p U;~ 2 3 4 b 6 7 6 9 10 11 IJ 13 14 l!h It. ZIfl 1Y iv Sti 2n sJ2 24 IknecIb)o - Ibis ta~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - 53 - The nature of the consumer demand determines the shape of the time/demand curve. Figure 2 represents the weekday demand profile of another feeder on the same hypothetical system. This feeder is one on which the demand of commercial consumers is primarily responsible for the shape of the profile. The demand is also at its minimum during the early morning hours; but the steep increase occurs at the beginning of business hours. The load remains relatively constant until the businesses begin to close. There is a slight increase in demand as darkness falls and lights are turned on, but the nighttime load is always well below the average during the business hours. The peak demands continue for longer periods than was the case with the former period so that the average demand will be higher. Peak Demand The data points used to plot Figures 1 and 2 are such as would be taken from a logsheet. Each value will be the instantaneous reading at the moment the operator recorded them. These are hourly readings, and the peak, or maximum demand experienced during the day being reviewed, is shown as occurring at some specific hour, 8:00 p.m. in the case of Figure 1 and 1:00 p.m. for Figure 2. In practice, the demand could have varied quite appreciably at periods in between readings and the actual feeder peaks could have been experienced at some time other than that at which the readings were made. The true peak demands may have been experienced for 15 minutes, 5 minutes or even less. Similarly, it is quite possible that during the period of lowest demand the true minimum load may have been lower than that shown on the graphs. The two effects may tend to cancel each ot1 r, but the peak demand of a feeder or of an entire system is more important because it determines the capacity which will be required of the feeder or system. It must therefore always be kept in mind that the true instantaneous peak demand may be higher than that recorded at hourly intervals. The actual figure may be more frequently at times at or close to that or peak demand, or by installing conttinuous recording devices. Load Factor Load factor is the ratio of average load to maximum demand. Load factors are usually expressed as percentages and may be calculated for any given period of time but are normally for a day, a week, a month or a year. For loss reduction purposes the annual load factor is the most useful as a year represents a full cycle of seasons and is also the interval most often used in planning studies. The average load may be defined as the continuous and unchanging load which would draw the same amount of energy over the given period of time as the actual load. It is calculated by dividing the actual energy transferred during the given period of time by the period of time. Care ust be taken that the units employed are consistent. For Instance if the energy Is stated in kilowatt hours, then the unit of time should be hours and the average load will be obtained in kilowatts. FIgure 2 FEEDER NO. 2 HOURLY DEMANDS (COMMERCIAL) 1tUX_ 4I glw u I 2 3 4 5 6 / a 9 10 I1 12 13 1.1 It, lo I/ 18 19 Al 21 22 23 24 &WtNaUk Oaa iV"I \ - 55 - If we return to the load readings of Table 1 and assume that each of the hourly demands shown in Table 1 did in fact continue for a full hour, then the energy transferred by the feeder on that day would have been the sum of the individual hourly loads, or 9,680 kWh. The average load would be 403 kW. The peak demand of the day was 660 kW-at 7:00 p.m. The load factor, or ratio of average of average load to peak demand for this feeder on the day shown would be: 403/660. x 100 - 61% or 9,680/24/660 x 100 - 61%. It is to be noted that the only data required for the calculation of the load factor for any system or piece of squipznrnt Are: (a) the period of time over which che load factor is calculated; (b) the maximum demand during that period; and (c) the energy transferred during that period. It is therefore not necessary to record hourly or any other regular demands. A kilowatt hour integrator and a maximum demand meter will together provide the energy data required. In the example above the period of time over which the load factor was calculated is one day. As was mentioned previously, any time period may be used as the basis of load factor calculations, and care must be taken to ensure that units are consistently employed. Losses The load on any system will be determined primarily by the consumer demand. The current flow to meet this demand will result in losses. Load losses are proportional to the square of the current flowing in the system and can therefore be taken as being directly related to the squares of the demand in kilowatts, assuming constant voltage and power factor. On this basis, if feeder or system load losses are known for any given demand, they can be calculated for other demand levels. System losses are normally calculated initially for the annual peak demand. For the feeder whose demands are shown in Table 1, the losses at peak demand (660 kW) were calculated to be 48 kW. The consumer demand at peak is then 610 kW. Table 2 below shows the losses and consumer demands for each of the hourly readings recorded in Table 1. - 56 - Thable' WEEDER DEMANDS, LOSSES, AND CONSUMER DEKANDS oatder Conumar Feeder Consumer rout Load Loss Dmand tour Load Loss Dmad (kW) (kW) (kW) )kW) (kW) (kW) 1 320 11 309 13 400 18 382 2 310 11 299 14 400 18 382 3 300 10 290 15 420 19 401 4 300 10 290 16 420 19 401 5 330 12 318 17 380 16 364 6 350 13 337 18 550 33 517 7 350 13 337 19 660 48 612 8 380 16 364 20 620 42 578 9 380 16 364 21 520 30 490 10 400 18 382 22 400 18 382 11 400 18 382 23 350 14 346 12 400 18 382 24 330 12 318 Figure 1 graphically represents not only the total feeder demand, but also the consumer demand over the 24 hour cycle, the difference between the two demands being losses. The increasing effect of losses on total demand as the consumer demand also increases will be evident from the curves in Figure 1. Duration Curves A duration curve relates the variation of a given parameter to time over a specified period. For any given value of the parameter, it is either equalled or exceeded. Duration curves for demands and losses are useful to illustrate the chronological patterns of these variables. Table 3 assembles the data required to plot the duration curves for total demand and losses for the hypothetical feeder of Table 1. As losses are proportional to the square of the demand, the latter may be used to establish ratios for losses. In this way, knowing what the demand profiles are, we can develop a duration curve for losses without knowing what the actual loss values are. - 57 - Table 3s DEMAND AND LOSS DURATION Demand Sq. of Demand cI Frequency b/ EINix £1 Dur dl Eq/Ex e/ Pk. Dn fl Squared i Pi. Dm 1 (kW) tX) (S) (kW2 ,x) X 660 1 1 4 100 435,550 100 610 1 2 4 a 94 386,400 86 550 1 S 4 1S 83 302,500 69 520 1 4 4 17 79 270,400 62 420 2 5 6 25 64 176,400 40 400 6 12 25 50 61 160,000 37 380 3 1S 12 63 58 144,400 33 360 1 16 4 67 5 129,600 30 350 2 16 8 83 50 108,900 25 330 2 20 8 83 50 108,900 25 320 1 21 4 88 48 102,400 24 310 1 22 4 92 47 96,100 22 300 2 24 8 100 AS 90,000 21 aI Column 1 lists the varlous hourly demand readlngs recorded, arranged ln descending order of magnitude. hi Column 2 indicates the number of times that the specifle kiloavtt demand vas recorded durLng the 24 hour period. gi Column 3 shows the number of times that the speclfic demand was equalled or exceeded. 41 Column 4 Lndicates the period over which the demand was experienced, expressed as a percentage of the total period (24 hours). *1 Column 5 lndicates the period during whlch the demand wan equalled or exceeded, again shown as a percentage of the total period. The values in this colum represent saumations of the percentage durations calculated In Colum * for the speciflc ae0 higher demand levels. f/ Colutn 6 shows the recorded demand an a percentage of peak demand. &I Colu 7 lists the squares of the instantaneous demands. The losses wvil be proportional to these values. hi Colum 8 show the ratio of the square of the instantaneous demand, expressed as a percentage. Thls percentage will also be the ratio of the losses at that demnd level to those at peak demand. Table 3 is based on the assumption that each demand reading recorded was maintained for a full hour. All values have been rounded off to the nearest whole number. The data in Table 3 has been used to plot the Demand and Load Loss Duration curves shown in Figure 3. Duration curves can be plotted in units other than the percentages used in Figure 3. For instance, the abscissa (x-axis coordinates) could have been in hours, and the ordinates (y-axis coordinates) in kilowatts. The area under each curve would then be kilowatt hours. Percentages were used in this Instance because the average demand or loss can then be easily determined as a percentage of the peak value of the corresponding parameters. Igure 3 DURAUON CURVES - DEMAND AND LOSS 1w 60 20 _ .__ O 20 du1 idI iCJ * b~~~~~~~~MdtfakJa*V4IWe I 59 - The 500 point on the time axis of the load duration curve will indicate the average load and therefore the load factor. On Figure 3, this value will correspond to the 61% previously calculated. Loss Factor In evaluating system losses, it is important to know not only the extent of power losses llowatts) which occur at times of peak system or feeder demand, but also the energy losses (kilowatt hours) which occur over a given period of time, usually a year. Modern microcomputer technology makes the calculation of peak demand losses relatively easy. Calculation of energy losses is more complicated however, since it is not feasible to make loss calculations for each of the hourly loads experienced during the period being evaluated. The energy losses are therefore usually calculated by application of what is known as the 0loss factor." The loss factor is defined as the ratio of the average loss to the peak loss. The ratio is often expressed as a percentage. In other words, the loss factor is the percentage of the time which the peak load would require to produce the same energy losses as are produced by the actual load cycle during the same length of time. It may be calculated as follows: Loss factor - kWh of Loss During a SRecified Time Period x 100 (Hours in Time Period) (Peak kW Loss) As losses vary with the square of the demand, the latter parameter may be used to represent the losses, effectively being equal to the loss multiplied by a constant. Then: Loss factor - Summation of (Hourly Demand)2 x 100 (Hours in Time Period) (Peak Demand)2 In the example of Table 3: Loss factor - 4.109.800 x 100 - 39% 435,600 x 24 Note that in the summation of the squares of demand the frequency of occurrence (Column 3) must be considered. Loss factors can therefore be calculated without knowing what the actual losses are, provided the variation of demand with time is known. If the loss factor and the losses at times of peak demand are known for a speclfic period of time, then the energy losses in that period may be calculated by multiplying the peak losses by the number of hours in the period and by the loss factor. The loss factor must, however, be approprLate to the period being considered. Calculations of loss factors for a given day and for a year are likely to yield very different results. The loss factor applies of course only to load losses, that is those losses which vary with the square of the load. No-load losses are independent of the load and are therefore not affected by the loss factor. For this reason, the terms Oload loss factor" and "copper loss factorm are sometimes utilized. - 60 - Load losses being proportional to the square of the demand, it follows that the average losses will bear some relationship to the average load and hence to the load factor. Note however that the loss factor is seldom equivalent to the square of the load factor. This is so because the average of the squares of a group of numbers is not necessarily the same as the square of the average of those numbers. The loss factor is never greater numerically than the load factor, and never less than the square of the load factor. The 50% point on the time axis of a loss duration curve will give the average of the squares of the hourly loads, or the square of the load factor. In the instance of Figure 3, this will correspond to 37%. This is not the loss factor in this example. The loss factor is the ratio of the area under the loss duration curve to the area enclosed within the rectangle formed by the two 100% boundaries. The loss factor for the load profile of Table 3 (or Figure 3) is 39%. Unlike the load factor, the loss factor is not determined by peak demand and energy transferred alone. The variation of load with time is also important. Accurate calculation of the annual loss factor for any system or feeder would require the summation of the squares of the demands for each hour of the year. This would be a very tedious process and is seldom, if ever, done. More commonly, a typical sample period is taken and allowances made for seasonal variations, weekends, holidays, etc. Even then, the information required for certain feeders may not be available and estimates of the loss factor have to be made. In contrast, the data required to calculate the load factor is often readily available. An empirical relationship between load and loss factors has been established which generally yields results within tolerable limits of accuracy. This relationship states that: Loss factor - c (load factor)2 + (1-C) (load factor) where I"C has a value between 0.15 and 0.3 depending on the load characteristics of the feeder or other system component being considered. In the absence of any data on which to base the choice of "C,w a value of 0.2 may be used. Loss Eauivalent Hours The loss factor represents the percentage of total hours which the maximum loss would have to last to equal the actual loss. Thus, if the loss factor over a year's time equals 0.20 or 20%, the equivalent peak-load loss lasts for 20% of the 8,760 hours in a year, or 1,752 hours. This value is called the annual loss equivalent hours. It is obviously possible to calculate daily, weekly or monthly loss equivalent hours, but the annual basis is normally the one of greatest practical significance. For the example in Table 3, the loss equivalent hours for the day would be 0.39 x 24, or 9.4 hours. Peak Responsibility Factor On each component or subsystem of a total system, the peak losses occur at the time of maximum demand of that component or subsystem, but this may not be at the same time as the overall system peak. For instance, the time of maximum demand on a specific feeder may not be coincident with the overall system peak. In that case, the contribution of the feeder to the system peak will be only a fraction of the peak on the feeder itself. The contribution of any - 61 - component to overall system peak is determined by the "peak responsibility factorw which is defined as: Load_on Component at Time of System Peak (kW) Component Peak Load (kW) This factor is a measure of the probability of any given component experiencing its peak load at the same time as the system peak. A step-up transformer for a peaking generator will normally have a peak responsibility factor of 1. Conversely, the transformer serving an exclusively daytime load on a system with predominantly residential demand will achieve its peak loading coincident with the system peak, and its peak responsibility factor will be less than 1. Figure 4 is a plot of the demands of individual feeders and of total system demand for a hypothetical system consisting of four feeders. The system peak occurs at 7:00 p.m., which is also the time at which Feeders 1 and 3 experience their peak demands. These two feeders therefore have peak responsibility factors of 1. Feeder 4 however reaches its peak of 930 kW at 2:00 p.m. At the time of system peak its demand is 670 kW. The peak responsibility factor of this feeder is therefore: 670/930 - 0.72 The peak responsibility factor assumes importance in determining how loss reduction measures carried out on a sub-system will affect the peak demand of the system as a whole. As an example, improvements carried out on a feeder which reduce the peak losses on that feeder by "xM kW, will reduce the overall system peak by less than "x" kW if the peaks are not coincident. This consideration will affect the economics of system upgrading to reduce demand losses as well as of evaluation of such losses for equipment purchase. Energy losses on a per unit basis are of the same importance as peak losses, except in those instances where off-peak fuel costs are lower. This could be the case, for instance, in a utility which uses gas turbines to meet the peak system demand but relies on more efficient diesel units to supply the off-peak or base demand. The peak responsibility factor is a ratio of loads. The contribution of component losses to overall system peak losses will therefore be a function of the square of the peak responsibility factor. Peak Loss ResDonsibllitv Factor The peak loss responsibllity factor of any system component is the ratio of the losses experienced by that component at the time of system peak to the losses at the time of the component peak load. From the previous paragraph, it will be seen that the peak loss responsibility factor is the square of the peak responsibility factor. The internal losses of any individual system component will depend on its demand profile and technical characteristics. They (the internal losses) are therefore independent of the peak loss responsibility factor. The demand of that component, however, will increase the load on other system components between it and the power source. Therefore, other things being equal, a higher peak responsibility factor will have the effect of increasing the overall system loss. - 62 - Diversity Factor A summation of the peak demands of individual consumers on any given circuit will produce a total which is greater than the peak demand actually experienced by that circuit. This is because the peak demands are not coincident. For instance, each householder does not use his appliances at exactly the same time as his counterparts. Appliances with cycling demands, such as refrigerators or water heaters, will not cycle in unison, The importance of this diversity in demands can be appreciated by considering the tremendous increase in system capacity-generation, transmission, and distribution--which would be required if all connected loads were imposed simultaneously. The diversity between maximum demands is measured by the 'diversity factor." This may be defined as the ratio of the summa.ion of the maximum demands of a set of consumers to the actual maximum demand of the whole group. If diversity factor is abbreviated "DF," then: DI + D2 + Ds +..+D4 DF - D 1+2+34%.. .+n) where D1, D2, D3 to D, are the maximum demands of consumers Numbers 1, 2,3 to n respectively, without regard to time, and D(1+2+3 .+n) is the maximum demand of the group. Referring again to Figure 4, the sum of the maximum demands of the four feeders is 660 + 840 + 1,250 + 930 - 3,680 kW. The system peak is 3,320 kW, so that the diversity factor of the feeders is 3,680/3,320 - 1.11. Diversity factors can, of course, be much larger than unity. The diversity factor is used to determine the maximum demand resulting from a group of individual loads, or from combining two or more of such groups. These combinations may represent a group of consumers supplied by a single transformer, a group of transformers supplied by the same primary feeder, a group of primary feeders supplied from the same substation, etc., back to the ultimate point of supply. The diversity factor is useful in determining the effect of different classes of consumers on the overall load demand profile. Numerically it is always greater than unity. Coincidence Factox The "coincidence factor" is the ratio of the maximum demand of a group of consumers to the summation of the groups' individual maximum demands. It is therefore the reciprocal of the diversity factor, and is numerically always less than unity. Figure 4 HOURLY DEMANDS - FEEDERS AND TOTAL SYSTEM 25W }~~~~ * ;m o 2 3 4 6 78 11 1'.; 1.1 14 16 R. 11 18 19 20 11 22 23 24 $,.2AM.lL II V i.41W4.i 4 - 64 - Coincidence and diversity factors are usually applied to a group of consumers with similar power demand time profiles. These factors will have accuracy limitations if calculated for a small number of consumers. However, as the number of consumers increases, the contribution of each load to the group's maximum demand decreases and therefore increases the accuracy of the calculated factor. Figure ' indicates the stabilization of the coincidence factor with increasing numbei of consumers. Demand and Energy Losses It is probably worthwhile here to review briefly the costs of system demand and energy losses. There is general agreement on the evaluation of energy losses as these are stated in kilowatt hours or other units of energy, and can in turn be translated into the cost of fuel required to generate these losses, or cost of energy purchased or value of energy not available for sale. The treatment of demand costs is sometimes not so easily agreed on. At any given point in time, the system losses will impose a kilowatt demand on the system over aRnd above that resulting from the consumers. This additional demand must be accommodated by the distribution, transmission, and generation systems. There are those who feel that demand losses form such a small proportion of total system demand that they have no influence on the timing and sizing of new generation, transmission or distribution installations. However, the generation, transmission, and distribution systems have no means of determining the source of the demand or distinguishing between a demand which will result in income to the utility and one which will be dissipated as heat to the atmosphere. No matter how small the demand losses they must be included in overall demand in system planning. In many instances, the level of losses are by no means small in comparison to the overall demand and form a significant compc ent of the capacity planning to meet future needs. Technical losses in excess of 15% of peak system demand are not unusual in developing countries Reduction in the level of losses will therefore postpone the timing of new plant, and that postponement can be converted into a value in dollars and cents. The actual value of the reduction of losses by 1 kW will vary with the circumstances peculiar to each utility. It will also be dependent on where on the overall system the loss reduction occurs. Lower distribution losses will reduce inve3tment in the distribution, transmission, and generation systems. If the transmission losses are reduced, however, only transmission and generation investaents will be affected. It follows, therefore, that the greatest incremental benefits are obtained by reducing losses on the distribution system. T.itl does not mean, however, that losses elsewhere ought to be neglected. Loss Evaluation Subsequent papers will describe in detail how losses on power systems are to be calculated, particularly using microcomputers It is, however, desirable that we review here some of the basic considerations in such evalu&tions. Loss calculations require a certain minimum set of data. Important among those are: (a) peak demand, seasons, and annual; FIgure 5 COINCIDENCE FACTOR TYPICAL VARIATION WINH NUMBER OF CONSUMERS 0 I 09 .0 - 06 .01 o' S ... 04 03 02 _ 01, oil a A aa U 1 2 3 4 S 6 7 8 9 10 11 12 13 14 1b 16 11 19 A) 21 21 23 24 ;1b 26 21A 3U umb.Md ct'AuclhusAs kWVjbm Sw* V4m0 5 - 66 - (b) daily demand profiles, weekdays, and holidays--season; (c) energy sent out, at least annually, preferably monthly; (d) physical configuration of system, distances, spurs, phases, conductor materials, transformer locations and specifications, etc. For economic analyses of loss reduction measures, information will also be required concerning the incremental cost: (i) kilowatt of investment in generation, transmission, and distribution systems; and (ii) per kilowatt hour of energy supplied at the transmission and distribution levels. The cost of energy supplied at the distribution level is higher than at the transmissitnt level because of the losses in the distribution system. Power is rarely supplied directly from the generating stations to the consumers, but in those instances where it is, the costs are lower than would have been the case for supply from the transmission system, again because of the losses in the transmission system. The losses for a total system must be developed from analyses of the individual system components. Thus, each secondary distribution system, each primary feeder, each sub-transmission lines, and each transmission line must be calculated independently. The total system losses will then be the sum of the individual components, due allowance being made for the peak responsibility and factors in determining the system total. On large systems, the calculation of all of these system components individually may present practical problems because of the length of time that would be involved. In such instances, a realistic number of typical components for each category ought to be selected and the calculated loss values taken to be representative of the sub-systems It is to be noted that the data required as mentioned above (load profiles, peak demand, energy sent out) will be required for each feeder or transmission line being calculated, and not merely for the overall system. It is therefore desirable that transmission lines and primary feeders be equipped with the appropriate instrumentation If a serious loss evaluation is to be undertaken. If the instrumentation does not presently exist, careful thought augt be given as to how reliable estimates may be developed from the information currently available and from spot checks which can be made In an acceptably short period. In order to determine whether a glven loss reduction measure is economic or not, the benefit of the lower losses must be compared wlth the cost of the proposed measure. The benefits will be the sum of lower demand and energy costs. For simpler schemes, the calculation may be made on a straightforward payback period basis. This nmthod io best suited to schemes which have very high returns--payback periods of about a year. For longer payback periods, the present value of costs and benefits ought to be determ&ned over a period of 7- 10 years. Due allowance must be made ln such calculations for load growth on the feeder or other systeu components during the period oz evaluation. The benefits of any loss reduction scheme will increase as the load grows, and it would be short-sighted to design efficiency improvement measures without - 67 - evaluating their effect on future increases in peak demand and energy supplied. Whenever the decision is made to implement a system improvement project, some time will elapse between the decision and its implementation. The demand will normally have increased during that period and other system changes may have occurred. System improvements must therefore be designe.d for the conditions which will prevail at, and subsequent to, the time of implementation and not for the conditions which exist at the time of the decision. Loss reduction schemes need frequent reevaluation to assess the effect of changing conditions, such as energy prices, load growth, etc. A scheme which was evaluated as being uneconomic a year ago may well prove attractive today. Conversely, schemes ought to be reassessed just before the commitment to expenditures is made to assess the effects of changed conditlons on the economics of the project. - 68 - BIBLIOGRAPHY 1. Energy Efficiency: Optimization of Electric Power Distribution System Losses. By: M. Munasinghe and V. Scott Energy Department Paper No. 6 World Bank Energy Department 2. Improved Methods for Distrlbutlon Loss Evaluation. By: Westlnghouse ElectrLc Corporation for Electric Power Research Institute 3. Electric Dlstrlbution Systems. By: EBASCO/ElectrLcal World McGraw Hlll Publicatlons Co. 4. Distribution Systems, Volume III. By: Westinghouse Electrlc Corporation 5. Distribution System Efficiency Improvement Guldebook. By: Bonnevllle Power Administration - 69 - ELECTRIC POWER SYSTEM TAFlSES JAMAICA PUBLIC SERVICE COMPANY LIMITED BY Messrs. R.A. Silvera and H. G. Higgins Paper Presented at Caribbean Regional Electric Power System Loss Reduction Seminar GENERAL Jamaica Public Service Company Limited (JPS) is a publicly owned electric utility and is the sole supplier of electrical energy in Jamaica, operating under licence. A. Technical Characteristics of the System Generation The company has a generating capacity of 442.5 MW supplied from two oil fired steam stations of 302 MW, one slow speed diesel station 40 MW, five gas turbines 80 MW and five run of the river hydro electric stations 20.5 MW (see Appendix 1). Substations There are approximately fifty-four (54) substations which step down from 69 kV transmission to distribution voltage levels with a total capacity of 655 MVA. Seven (7) bulk power transmission substations (138/69 kV) with a total capacity of 427 MVA also exists. Transmission Lines There are 171 circuit miles of 138 kV lines comprising nine circuits, interlinked with 445 circuit miles of 69 kV transmtssion lines. The system is predominantly of wood pole construction, but include2 90 miles of 138 kV and 42 miles of 69 kV on steel towers (see the attached map in Appendix 2). Distribution SXstem The primary distribution system consists of approximately 7000 miles of lines operating at 24 kV, 13.8 kV, 12 kV and 4 kV. The system is presently being upgraded islandwide to operate at a standard voltage of 24 kV. Conductors used are all Aluminium Alloy in sizer 394.5 KCM, 155.4 KCM and 77.5 KCM. There are 118 primary distribution feeders existing. Secondary voltage levels are 415 V and 220 V 3 phase and 220V/llOV single phase. - 70 . B. Customer Management System JPS maintains nine (9) district regional offices for the management of customers accounts. Applications for electricity service, establishment of customers contracts, extension of service, the conduct of meter reading and the collection of revenue are all functions conducted at District level. The bulk of customers meters are read once every two (2) months. The meters of large customers (rate 50) are read monthly. Revenue billing is computerized and customers are billed monthly or bi-monthly depending on the level of kWh consumption. Actual revenues billed to customers are monitored, reviewed, adjusted where necessary and collected at the district level. Receipting machines are used at District Offices in the revenue collection process to create and batch the collection data for the crediting of customers' accounts. Remote computer terminals at District Offices permit on- line access to recent customer accounts history on the central computer. These terminals also facilitate the conduct of revenue adjustment billing functions at the district level. C. Electricity Demand and Load Forecast Since 1980 extremes in the demand for electricity have been experienced. The system recorded a growtn rate of -4% in 1984 to 9.6% in 1986 in terms of peak demand and -1.15% in 1984 to 11% in 1987 in terms of energy sales. Table 1 supplies details on electricity demand, forecast, sales and loss data. D. £nnegrgLosses In recent years JPS electric system power loss, comprising technical and non-technical ('unaccounted for') losses has been in the region of 19% to 21% of net generation. Various efforts have been made to quantify these losses and apportion percentages in the known categories. The most recent of these is a World Bank funded study on 'Electric LOss Reduction" undertaken by EBASCO, a USA consulting firm. This was submitted to JPS in July 1988. The approach used in the study was selection of a number of primary distribution feeders for detailed analysis and extrapolation to project losses for the entire system. - 71 - The EBASCO report allocates losses in the respective areas which total 19% of net generated energy, as follows: % of Total kWh Energy Generated 1. Transmission system, including substation transformers 1.14 2. Primary distribution lines 7.03 3. Distribution Transformers 1.52 4. Secondaries and services 4.18 5. Unaccounted (non-technical) 5.13 19.00 The annual cost of power losses on the system for 1989 is expected to be in the regLon of J$160 m. E. Non-Technical Losses The company haw determined that its non-technical losses arise from both internal and external causes. (a) Internal Causes: (1) Defective Meters : mechanical or electrical failure in the component parts of the meter (10%) (2) Incorrect meter multiplier constants - caused from incorrect field information or clerical effort (30%) (3) Incorrect wiring - improperly installed wiring connections to meters (5%) (4) Incorrect metering - wrong meter installed in socket (5%) (5) Other - revenue billing errors such as incorrect meter readings, closed accounts with advanced readings and meters with advanced readings not appearing on the billing system (unbilled supplies). (15%) (b) External Causes: (1) Line Taps and bridged meters - unauthorLzed connections made to the line side before the metering point, includ- ing also bridged meters (10%) - 72 - (2) Direct connections - unauthorized direct connections to the company's lines to serve customers installations where no meter is existing. (5%) (3) Tampered Meters - unauthorized interference with the meter mechanisms (backing pointers and adjusting gears etc.). (20%) The percentages above reflect the estimated portion of total non-technical revenue loss, due to the causes indicated. There is high incidence of electricity theft in Jamaica and is found among all income groups and consumer types. Electricity theft is a criminal offence in Jamaica and offenders if found guilty in a court of law, is liable to a fine and/or imprisonment. F. Loss Reduction Programme - Non-Technical The company established a Customer Service Investigation Unit (CSI) in 1982 to address the problem of electricity theft. Technicians from the unit accompanied by armed police conducted widescale investigations primarily in low- income communities. Those found in possession of an illegal supply of electricity were arrested and charged. The programme however, was not very effective economically. Technicians also spent long hours in courts. As well, police action in the communities targeted had only short-term effects and only a limited amount of revenue was collected during the four years of the Unit's operation. In consequence, in November 1986, the comnany established a 'Revenue ProtactionO pilot project with an expanded terms of reference. This project was subsequently enlarged into a permanent department with regional representatives in districts and was charged with responsibility for protecting -,he company from revenue losses of all kinds. The Revenue Protection Department subsequently instituted a number of programes. These include the following: (a) CT Meter Audit Programme - a programme to audit all metering facilities which have current transformers installed. (b) Investigation of whole current meters with consumption levels exceeding 2,000 kWh/month. (c) Irregularity Investigation - where investigations are conducted randomly, also on the basis of customers reports received aTid at the discretion of the Department's management. A major meter resealing programme has subsequently been identified as one of the devices to contort irregularity. As at May 31, 1989 the Department had conducted 9,760 investigations and discovered among them 1,403 irregularities. About 817 of these accounts with irregularities hav& been adjusted in the amount of J$4,495,000.00 of which J$3,310,000.00 has been collected. The .evenue realizable from the adjusted accounts is estimated at J$1,740,000 annually. - 73 - G. Loss Reduction Programe - Technical The company has recognized the benefits to be derived from power loss reduction in the technical areas and has outlined a series of projec..s to achieve the reduction of such losses on the system. These are discussed in order of priority. 1. Voltage Standardization Programme This programme was started some years ago with the objective of converting all primary distribution circuits to operate at 24 kV. This project will have the greatest impact on energy losses when completed, due to the resulting substantially lower line currents and will also impact on the standardization of material and inventory. This exercise as well, has the potential to significantly increase distribution line MVA capacities thus deferring capital investments in upgrading of distribution lines and installation of new substations. The progrAmme is approximately 30% completed. 2. CaDacitor Installation An active capacitor installation programme is currently being pursued on the primary distribution system. The objective of this programme is to achieve and maintain a system power factor of 0.95 against the current system power factor of 0.88. The reduced MVA demand will also assist in deferring investments in new generating plants. 3. Primary Distribution - Reconductoring The company has standardized on conductor sizes for the distribution system and has recognized the need to reconductor all circuits operating with conductors below the economic loading limit. The studies and engineering for this project is expected to be included as part of a distribution master plan to be prepared for the company. Significant net savings is expected to be derived from this exercise. 4. Distribution Phase Balancing A source of lozs on the system is in the imbalance of phase currents on the primarj distribution system. It is planned to balance phase loads on feeders so as to reduce energy losses arising from the high currents. In addition, the company has pursued the following strategies or adopted as a matter of policy - (a) Overlay of the old 69 kV transmission grid system with a 138 kV network. Energy losses on the transmission system is thus reduced resulting from the transmission of bulk power at a higher voltage on larger conductors. (b) Procurement of energy efficient transformers to serve substations, customer pad mounted and distribution pole mounted installations. - 74 . (c) Use of larger conductors and service wires on secondaries and optimizing the locations of pole mounted transformers. CONCLUSION The combination of technical and non-technical loss reduction programme are expected to reduce total energy losses to a targeted level of 15% of net generation in the medium term. The company has recognized the financial benefits to be derived and other advantages to be gained from the pursuits of these programmes and has herefore committed itself fully to their implementation. - 75 - QgRendix1 EXISTING JPS GENERATING UNITS Type Station Unit Year Nameplate Maximum No. Installed Rating MW Continuous Rating (MW) 1 1968 33.0 30.0 Old 2 1969 60.0 60.0 Harbour 3 1970 68.5 55.0 Steam- 4 1972 68.5 60.0 1 1953 12.5 Electric 2 1955 12.5 Hunts Bay A 3 1959 15.0 4 1961 15.0 5 1962 20.0 20.0 Hunts Bay B 6 1976 68.5 68.5 Total Steam Electric 373.5 293.5 Maggotty 1959 6.4 6.3 Upper White River 1945 3.2 2.6 Hydro- Lower White River 1952 4.8 4.7 Electric Roaring River 1949 4.1 4.1 Rio Bueno 1956 2.5 2.5 Total Hydro Electric 21.10 20.2 Hunts Bay 1 1969 15.3 13.5 Hunts Bay 2 1970 15.3 13.0 Gas Bogue 3 1973 21.1 20.0 Turbines Hunts Bay 4 1974 21.1 15.0 Hunts Bay 5 1974 21.1 18.5 Total Gas Turbines 93.8 80.0 Slow 1 1985 20.0 20.0 Speed Rockfort 2 1985 20.0 20.0 Diesel _ _ Total Slow Speed Diesel 40.0 40.0 Total System 17 528.3 433.7 Jamaica Public Service Co. Ltl TRANSMISSION 138/69 KV SYSTEM * ¢ A" I 1 @ A N S It A 14IL4WXV L~~~~~~~~~~TAmASI~ 3W9V Ul ' s( FM~~~~~~ MM"Gol n_ t TRAftVMU#Va - 77 - Appendix_3 PRIMAKY DISTRIBUTION SYSTEM 1. 7,000 Circuit Miles 2. Operating Voltages 24 kV - Urban/Rural 13.8 kV - Kingston Area 12 kV - Rural Mainly 4 kV - Kingston Area 3. Conductor Sizes - 394.5 KCM ) 155.4 KCM ) All Alum. Alloy 77.5 KCM ) 4. 11E "Irimary Distribution Feeders 5. Secondary Voltage - 415 V WYE ) ) 3 Phase - 220 V Velta ) - 220/110 V - Single Phase JAMACA PUBLIC SERVIC CUIPAN le0tc:lat7 Demnd, Forecat, Losses S Salet Data Gross not Raiggatial Cami Inus Coilndus Scti year Ceaner- C190 *f 190 > 190 > 190 >190 > 190 Watercontent. % v/v 0,05 0.05 0.05 c 0.05 < O.OS c 0.05 TON. mg KOH/9 25.0 24.3 23.6 22.4 22.2 21.5 Combustion soot 0.27 0.21 0.20 0.26 0.21 0.23 Dispersancy 88.4 89.3 84.1 85.0 75.5 81.8 Calcium. % m/m 0.99 1.00 0.94 0.84 0.82 0.94 Barium, % m/m 0.00 0.00 0.00 0.00 0.00 0.00 Phosphorus. % m/m 0.04 0.04 0.04 0.04 0.04 0.04 Zinc. % WmJ O.5 0.05 0.05 0.04 0.04 0.04 METAL-ANALYSIS: Silicon. mg/kg 0 0 0 0 0 0 Iron. Mg/kg 0 0 0 0 0 0 Alumtnium. mg/kg 4 4 4 5 5 5 Chromium. mg/kg 2 2 1 1 1 1 MolyOdenum.mg/kg 0 1 0 1 2 0 Copper. mg/kg 1 1 1 1 1 1 Tin, mg/kg 0 1 0 0 0 1 Lead, mg/kg 0 2 0 7 0 0 NiCkel. mg/kg 6 5 4 4 4 5 Manganese. mg/kg 2 1 2 1 1 1 Silver. mg/kg 0 0 0 0 0 0 Vanadium. Mg/kg l 17 15 12 13 14 Commmnts--: 95 9S 95 95 95 95 Shell Nederlana verkoodmAts3Cnaoojj 9.V. Dealt by MLI Phone COMMENTS - 111 - CURRENT SAMPLS: 95 ALL ANALYSIS FAIURES ARE WITHIN NORMAL LIMITS AND 00 NOT NEEO TO BE COMMENTED. THE OIL IS FIT FOR FURTHER US! PREVIOUS SAMPLES 95 ALL ANALYSIS FIGURES ARE WIlHIN NORMAL LIMITS ANO OO NOT NEEO TO BE COMMENTED. THE OIL IS FIT FOR FURTHER US! When Investigating anw orodUcing tne analysis figures we presume that the saeoie fully represents the indlcatea o11 cnarge. exceDt In the case of gross negligence on our side we 00 not accept resDoniDbility tor tne remmmnaat ions given. Shell Neaeriand VS. OA"&tzCrapp¶j O.V. Establishea In Rotterdam. m.reg. notteraam ll 29 07 - 112 - Annex -2 VIBRATION-LEVELS ( + ) 9TM620 DIESEL ENGINE. W~~~~~~~~~~~~~~~~~ 1@XiX i 14.1 (~~~~~~% . S )1 i splacement 600 400 -. 0 300 - \\x > 100: %DIamage probable so - 40 -20 ' e -DbGo/ no go criteria U- 7o t49C I ^ Aceptable i ; IDOTE pubi. nr. 434 Oct. 1986 SM I 5 10 20 30405080100 200300 Vibration frequenclvlHz) Effective vibrations ot cylindef had leveI of OMese' enqtnes - 113 Annex3 VIBRATION-LEVELS COOLING-WATER CIRCULATION PUMP A /7 /////////, /17 iTV - - l-tt AlL hcblae5 En_\l*,DLA_JsL I I2N TfiT _ _ _ . I. H I Nj I IN I I I I Go/ no go criteria 1. ' H'L ___ ±_ _ _ . - . * r. na du , 0 . f . tLlr.ss - ; I §- I 39!5 -7N _ . _. _ -_,_. ., ~ ~ ~ ~ D ., . 64 ", .w -n ft . m _ _ ." _ --~~~~~-1D _ .- - 114 - WAYS TO REDUCE TRANSMISSION AND DISTRIBUTION LOSSES AND WAYS TO REDUtE TRANSFORMER LOSSES By Mr. Barry Kennedy . ~~~~~~~~~~~rn C'D C',, CD a- goo 0 (0 - St.T ' NON-GENERATING PUBLIC UTILITIES LOSSES DIST. TRANS 36.5% UNACCTD 7.8% BULK SUBST. 2.2% TRANS/ST 10.5% SECY SYS 8.1%, FEEDERS 34.9% INVESTOR OWNED UTILITIES LOSSES BULK SUBST. 4.0% DIST. TRANS. 16.2% TRANS/ST 32.3% I.- I s~~~~~~~~~~~~~~~~~~~~~~~~~ FEEDERS 38.4% SECY SYS 9.1 % - 118 - Gon.uallon(i | lh~~~~rnsfortnor Step-up 1 JI) Copper Losses Propodional to Load lransformeor, (2)lron Losses are Constant High Voltage j frnumission < Power Flow Loses (ISR) 1tansmlsslon Substalion Power Flow Losses (AR) Sub- Transmission lIstribution 8bSUafon nsnlormer snd Regulator Dstributlon (1) Coppor Losses Proportional to Load Subs1allon f-ve- ( 2)lron Losses ero Conslant Dlstribullon Power Flow Losses (I'R) ransflormer Distilbution I) Copper Losses Proportional to Load Tlhnsdormr 1 2)kon Losses are Consabnt Powlr Flow Losses f'fl) Secondary System REGION'S T&D COMPONENTS o 1.3 million distribution transformers (under 7.5 kVA to over 500 kVA) o 1,200 substation transformers (1.0 MVA to over 60 MVA) '0 o 60,000 circuit miles of primary feeder (normal voltages from 11 kV to 33 kV) o 24,000 circuit miles of nonfederal transmission lines (operating voltages of 34.5 kV to 230 kV.) .. . .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ FEEDER LOSS FORMULA Feeder Lass Formula (Peak kW)2(Resistance per phase per mile)(Loss factor)(8760) kWh loss = (kV)2(Power factor)2(Number of phase)(1000) - 121 - c 0 0 _~~ 3~ 0 a) a)8 U - 122 - Sample System Scle I! 1 Inch 2 Miles; N \lsta b_ it~~~~~~u Sub Highland Urban Area N E Substatkon - 7.212.5 DItr-butln 115 W6TVurard_on RUraINec POWER FACTOR CORRECTION Net Load 1A14 Units of Currnn 1000 kVW ___________ 1000O kVAR D - 1414 kVA Generation T lanamussion Sub Distribution °Load Trnumlssion (A) No Power Factor Correction 1000 kVAR Capacitor 1.00 Unlis of Curnenl Load Net Load 1000 kW o kVar 1000 kVAR (0) With 100% Power Factor PENINSULA LIGHT CO. LOCKER ROAD FEEDER MOST COST EFFECTIVE METHODS METHOD (MILL/KWH) 1. Reconfigure - 2. Add Capacitors 18 5 3. Reconductor 50 4. Replace Transformers 66 Supply Curve Methodology Estimelt Unit Costs Estimate Numbers end Establish Operating Esiablish Operating and Rlepl""menl Cosls Types ol Equlpmenl Cherecletistice of Characteeislics of (Section 2.0) (Section 3.01 Exisling Stock Elficient Stock (Section 4.0) (Section 4.0) EstimateLovelied Estimele Age Cost of Replscmeni Distribullon of Equipment Eitiating Slock ISection 5.01 (Section 6.01 Estia total Loveised Estimate Leve"lied coilto InsRelce EuIsting Coat of Reltiing Calculate Losses Stsndsrd-Eliciency Existing Stock (Section 4.0) Equipment with High- ISeclion 5.0c Efficihny Equipment (Secilon 6.05 EstJiml LoveNsed Cost pev kWh of Loss Rcove,y .AMW_ (Seclion 6.0) Cteele Stiply Curves |e iSecdions 6.0 and 6-0) - 126- u 8~~~ ., . - s~~~~~~ e ' iA .g,~~~~~~~~ 0 8s, 1 \ _ tos~~~~~ ; , 1, t I , l.-,.. .. ., ,1 . . ,, ., , J X, N R X _ v " ° X @ t b o (WUt/I ~~~ Distribution Transformer Supply Curve 12 11 10 Loss Recovery 78 AMW 9 Cosl 5 5.6 cens/kWh a 4 3_. 2 3I 0 10 20 30 40 60 60 70 80 90 100 110 120 Loss flecovery (AMWJ Reconductor Supply Curve 12 Loss Recovevy 99 AMW 7 Cosi = 5.5 cents/MWh I S 1 . 4 0 to 20 30 40 50 60 70 80 90 100 110 120 Loss Recovery (AMW Distribution Voltage Upgrade 7 '4 3 4 Cosl Recovery = 276 AMW u 3 Cost = 2.0 cents/kWh 453 2 - - -_ - - - -- - - - - c 20 40 60 90 - oXo 120 140 16 1i 200 220 240 20 -.o Loss Recovery (AMWI FIGURE 6.6. Supply Curve for Voltage Upgrade of the Existing 12.5 kV System to 34 kY Composite Supply Curve Loss Recovery 380 AMW 0~ Cost = 5.5 centskWh Loss Recovery 335 AMW | 4 - - -Cosi = 4.0 cents/kWh Loss Recovery = 293 AMW I Cost = 2.0 cents/kWh I I 2~~ ~~~~ [ . l1§ 2 0 t00 200 300 400 Loss Recovery (AMW - 131 v T & D, R & D Evaluation Method INTERVIEWS LITERATURE SURVEYS TECHNOLOGY IDENIFICATION TECHNOLOGY DEFINITION - ;RIZ~CAEGRIATION TECHNOLOGY OF GROUPING TECHNOLOGIES AND 3TATE-AF FUTURE SELECTION SELECTION OF r PRnJE MOST PROMISING * PHASE I'61 15 TECHNOLOGIES DETAILED ANALYSIS SE'. ECTION OF OF TECHNOLOGY BPA CUSTOMER TECHNOLOGY APPLICATION BENEFITS MCDEL .-SSESSMENT . - CATEGORIZATION OF STATEEOF-CHLHIES_-ART _ J t . ~TECHNOLOGIES BENEF-ITS STATE OF THE ART RANKING TECHNOLOGY 9. ALUMINUM CARBIDE CONDUCTOR 1. IMPROVE VAR SITING 10. OPTIMAL COMPONENT COSTS FOR 2. OPTlMUM FEEDER RECONFIGURING TRANSMISSION 3. DYNAMIC PHASE LOAD BALANCING 11. STANDARD TO INCLUDE LOSSES 4. DISTRIBUTION AUJTOMATION 12. CONSERVATION VOLTAGE 5. ANALYZE LOAD REDUCTION CONTROLMA*NAGEMENT 13. IMPROVE DISTRIBUTION 6. DISTRIBUTION AUTOMATION/LOAD ANALYSIS/PLANNING MANAGEMENT 14. IMPROVED INSULATORS 7. AMORPHOUS STEEL TRANSFORMERS 15. IMPROVED FUSES & IROVED SILCON STEEL DISTRIBUTION TRANSFORMERS TOP STATE-OF-ART TECHNOLOGIES 1. DEMAND SIDE MANAGEMENT 2. AMORPHOUS STEEL DISTIBUTON TRANSFORMERS 3. CONSERVATION VOLTAGE REDUCTION 4. DISTRIBUTION AUTOMATION 5. SUPERCONDUCTrVY - 134 - WHAT ARE DEMAND-SIDE ACTIVITIES O Defined as any means by which a utility modifies its customers' load shapes either directly or indirectoly o Includes: - Load management - Conservation - Electrifcation - Load growth - 135 - GenertIon (SŽ) Sle.tp Ihandormer Stop-up V V (I) Copper Los"s Proportional to Lord lTansformer.r. 2)iron Losses are Conslanl High Voltage lhnsmission I Power Flow Louse (PlR) bansmlssion Substallon y Power Flow Losses (I8R) Sub. Trnsmisslon Station lhanslomner and Regulator Distribution ) C opp er Losses Proportional to Load Substation ( 2i)kon Losss are Consaint Dlstrlbullon - Power Flow Losses ORR). Pdmary DlstrIbution 11 Copper Losses Proportional to Load lknslormor 12)[ron Losses ae Constant 4 Power Flow Losses (13R) Secondary System A Method for Evaluating the General Load and Loss Effects of Demand-Side Management on the T&D Delivery System CL BROOKS. B.W. KENNEDY WESTINGHOUSE ELECMRIC BONNEVILLE POWER CORPORATION ADMIISTRATION SENIOR MEMBER MEMBER J.W. SANDERS BENTON COUNTY (WA) PUD NO. 1 Application of Method o BENTON COUNTY P.U.D. NO. 1 IN STATE OF WASHINGTON o SYSTEM MODEL - 3 URBAN AND 1 RURAL SUB - ABOUT 1S0 MW - LOAD DATA FOR ALL DISTR. TRANS. - LOAD FLOW : 500 NODES : FPCED MENDANCE LOAD : 3-PHASE UNBALANCED - FOUR SUBPROFILES : DIVIDED BY RURAL AND URBAN AREAS : DIMDED BY 6-MONTH WINTER AND SUMMER SEASONS Load Charactevtses 150- M t I( Peak SummerPeakDay r - - 11 ^ AveWe Lood on P ~ l6A veage td 6 E Winter Peak Day A a-Approidmoted K 11tF . Summer Peak / Load A I Aveae Load-n 0 Aea Avnera Summer Peak Dav Avet adc LAdem - -ntr oa ' i* Averae SumrmerLoad J F M A M J J A S 0 N D MONTHS OF THE YEAR 139 - Sampl Sysem Scale I Inch 2 Miles N ~~~~~~~~~~~~~_~~~~~~~~~~I c -r n o i Subi H4hSub-NN\ Uiban Area. o'XN< .1 Subda, -7.V12.6>,W*x_ 1 ^w115WTWrwnb.dm _i Rual~~~~~~ ^ 140 - 3 C,^ 138 Results o PEAK DROPPED 128,429 KW TO 106,596 KW (17.0%) o WNER URBAN LOAD FACrOR INCREASED 0.414 TO .499 (20.5%) O PEAK LOSSES DROPPED 6,887 KW TO 5,263 KW (23.5%) o ENERGY LOSSES DROPPED 18,219 MWH TO 17,376 MWH (4.6%) Benton Cnmty - highs d eaks AMORPHOUS STEEL TRANSFORMERS o RP 1290-1 AMORPHOUS STEEL FOR TRANSFORMERS o RP 1529-1 AMORPHOUS STEEL CORE DISTRIBUTION TRANSFORMER - 14~ z ' 0SX Qa I  I-. 11 3 I I I I 1 I- I I U 1 1 t I .1 'as 4 51! u-I4 Kg ii \IiJ Iii. 4.- 1 0 I 5 .7 1 ii I I I II.  I ii I ii. .3' .1 I N - 145 - CVR Regional Supply Curve Zone E. All feeders are more than 12 miles long. Representative of a feeder sytom that Supplies a rural area. 10 NWPPC Conservation-Cost-Effectiveness Threshold 5.04 Zone D. Same feeder-length and service-area characteristics as Zone C. Depicted as distinct zone to show supply curve trend from 1 to 5.04 C/kWh. 1.0 Zone C. Most feeders are 3 to 12 miles long. Representative of a feeder - R system that supplies a mixture of rural and suburban areas. ; 0.1 c | Zone B. A mixture of feedes that are either E less than 3 miles or in a range from a 1 3 to 12 miles l. Rresentative E of a fetder sytm that supplies a moderately populated area. 0.01 Zone A. Most fedw are l than 3 miles long. Representiwe of a feeder system tha suplies densely populad area. 0.001 ' I ' 0 40 80 120 160 200 240 280 Conservation Resource (AMW) DISTRIBUTION AUTOMATION o RP2021 ECONOMIC EVALUATION OF DISTRIBUTION AUTOMATION SYSTEMS o RP5689 DEMONSTRATION OF DISTRIBUTION AUTOMATION SYSTEM o RP1420 DEVELOPMENT AND TESTING OF MULTIPLE FUNCTION ELECRONIC WATITHOUR METER o P850 DEMONSTRATION OF ALTERNATE COMMUNICATION SYSTEM FOR DISIRIBUTION AUTOMATION o RP1535 BROADCAST RADIO SYSTEM FOR DISTRIBUTION AUTOMATION o RP1472 INTEGRATED CONTROL AND PROTECTION OF DiSTRIBUTION SUBSTATIONS AND SYSTEM SUPERCONDUCTIVrrY OVERVIEW 1. WHAT IS SUPERCONDUCTIVITY? IX*HY DO SOME MATERIALS BECOME SUPERCONDUCING BELW A CRICAL TEMPERATURE? 3. APPLICATIONS AND LIMITATIONS OF CONVENTIONAL (EAUIC) SUPJiRCONDUCrORS. 4. THE NEW METAL-OXIDE SUPERCONDUCI!)RS: a. SATE OF DEVELOPMENT/WHO IS INVOLVED b. CHALLENGES FOR R&D 1 WHAT IS SUPERCONDUCDIVITY? CONVENTIONAL (METALLIC) SUPERCONbUCTORS ARE POOR CONDUCrORS AT ROOM TEMPERATURR WHEN COOLED BELOW A CRITICAL TEMPERATURE, TWO REMARKABLE PROPERTIES APPEAR: 1. Zero Electrical Resistivity 2. Expulsion of all magnetic flux (Meissner Effect) 1987 OnsHe Review Paciflc Northwest LaboratorV Heooing Up Tento"N Indkxa ns I at 24OK (-28 F) Feb. 4987 98K.(-283 F) Nitroen Lquefled at Mt/ ) _?~~~~~7K (320- , > New Oxkes 5 I {/ ~Compounds-. 1911E . \\Vs\\\\ t\\\\\\\\\\\\\ \\\\\\\\\\\\\ C mpounds, " Absolute Zero 1920 1940 1960 1980 Transformer Loss Formula Em &-aL +L i X 100 A WHERE E = EFFICIENCY OF LNIT (%) A a INSTALLED NAMEPLATE CAPACITY OF UNIT (KVA) NLL = NAMEPLATE NO-LOAD LOSS OF UNIT (KW) IL NAMEPLATE LOAD LOSS OF UNIT (KW) -151- 700 -100 600 Efficiency i ^ ~~~Total losses 500 90 4001 300 80~ 200 Copper loss 100 Eddy current loss Iron loss70 Hysteresis loss 0~~~~~~~~~~~~ 0 5 10 :15 20 25 Lod In Wa - 152 - BASIC TRANSFORMER MODEL Core Losses |Demand (1) (kW) (kW kVAr) jIC ID| ; | ~~~Total I (Resistance I I I Model Reactance I * Single Phase I - Three Phase I _ Bank of L Tmransformers L. m ._____m_, Demand Losses (kW) Energy Losses (kWH) Probable Loss of Life (%) (1) Demand may be: Single phase Three phase Mixed single and three phase ANNUAL TRANSFORMER LOSSES FOR A 5000 MW UTILITY MILLIONS OF KWHR TRANSFOtRMER TYPE IRON COPPER GENERATOR STEP-UP 18 8 BULK POWER SUBSTATION 67 138 DISTRIBUTION SUBSTATION 97 114 DISTRIBUTION 328 127 TOTAL 510* 468 * 1.4% of Electricity Generated. -154- Figure 3a Transormer Losses Standard Transformers 4SW0 (97.8% Efficiency) 4COD Rating Watt LOss S 43 10 67 IS ~~90 35I 1 130 SO ~~2'15 100 370 3000 3u X 1~2000 I 1tS~~~~~ KVA S,2000 8D 100 10 14 0 Loading In % of Namev&re Rarng Page 44 - 155 - Figure 3b Transformer Losses High-EfFiciency Transformers (98.4% Efficiency) 3000 N0Load Losses Rating Watts LOSS 5 30 2500 10 47 15 64 25 88 S0 150/ ,2W 100 225 aooo0 A/3 SW 01 80 100 120 140 160 10 200 Loading In % of Namepue Rang. Pag 4s 156 - Figure 3c Transformer Losses Very High4icuiency Transformers 1800 Very (98.7% Efficiency) 1600 No-Load Losses 1500 Rating WatM Loss 1500 5 26 1400 10 40 is 53 25 69 50 113 1z0 | 1X ~~150// 1200 100 X' 810010 *0 10 6 8 0 Iwo ii~ ~~~~~o 840 500 400 200 Loading In % of Nameplat Rating Page 46 Ak Core Conventional Amorphorous COSTr OF LOSES NO-LOAD LOSSES 48 WAlTS 18 WAlTS &L5 WATr XS5 WATT $240 $ 90. LOAD LOSSES 284 WAiTS 249 WAITS $284 $249 TOTAL LOSSES $524 $339 PURCHASE PRICE S695 TOTAL OWNING COST $1034 $1034 - 158 - ENGINEERING EC-ONOMICS OPELOSS REDUCION pROIECTS By Mr. Barry Kennedy 0~~~~~~~~~~~~~~~~~~~~~~~000 CGnductor Size Fiure 1 Pawe7 To calculate the conductor losses: 1. Conductor size and material 2. Peak load on the conductor In kilowatts (kW 3. Power factor 4. Voltage . 5. Load factor .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4 The peak load must be modified by two factors: * Growth Factor * Distribution Factor page - 162 - figure 2 Growth Pactor and Distribution Factor Gr'2owt P.cW DistIbutIn Puceor 4 1. 1.0 SCAMPLE* -< IXWp-X * KW5-200 KWf -600 .9 KWLm 100 KCWF .9 .8 KWL . ~ a.3- - tw 3 20020 a-S ~~~~~.6 b'm.5 From From 4 Nomooram 5 Nomogram I91.91 d.764 .4 3 . .3 2 .1 _577 0 Sour>:C RMA Manual 609, May, 1960 . _ _ . Growth Factor: * Ratio future load to present load * Enter the nomograph of Figure 2 t.~~~~~~~~~~~~~~~~~~~~~~ passe - 164 Figure 2 Growth Pactor and Distribution Factor avowal Pasto DltlbusUan Pactw, ar fO9f o ncr 41 4 EAMPL 1. 1.0 EXAMPLE KWOwW200 C9 tW,a2 7 Kwf -60 tL'100 3~6 a- KWf .9 .8 bKW, SW .7 100 a-3 .6 bm.5 From From 4 Nomogm * Nomogram g-i- On.764 .4 2 3 . _.3 .2 1.5 2 e_ 1 ..577 Source: kEA Manuaf 60.9. May. 1960 _I N.~ _ Distribution Factor: * Ratio load at the end to load at source * Enter the nomograph of Figure 2 L Pages E a K~~~~~ 1 g ofiffa!z I aIn * I.~~~~~~~~~~~~~I 3 UV U 20 Load Factor Formula: total annual energy a (peak month's demand) 8760 l Pag 0 Loss Factor Formula: loss factor=.9(load factor 2+.o1(lad factor) co Pagee I Loss Formula: kWh loss (Peak kW)'(Resistance per phase per mile1xLoss factorn(8760) IkVl'lPower factorl2Nurnber of phasesl1000)l Page 9 Loss Formula Reduced: kWh loss = (Peak kW)2(Conductor constant) 0 Pages~~~~~ T*blA IA CONDUCTOR CONSTANTS-ACSR SINGLI-PHASE UINS sysem Load 266.8 356.4 477 566.5 Factor 14 12 110 210 310 410 KCMIL KCMIL KCMU. KCMIL .30 .0519 .0327 .0206 .0163 .0129 .0103 .00610 .0644 .0D454 .00390 .40 .0660 .0541 .0341 .0271 .021S .0171 .0134 .0107 .00752 .00645 .50 .1265 .0609 .0509 .0405 .0321 .0255 .0201 .0159 .0112 .00964 .60 .1794 .1130 .0711 .0565 .0449 .0356 .0280 .0223 .0157 .0135 .70 .2388 .1503 .0941 .0753 .0597 .0474 .0373 .0296 .0209 .0179 .60 3065 .19"0 .1215 .0966 .0766 .0609 .0479 .0380 .0268 .0230 .90 .3627 .2409 .1517 .1206 .0957 .0760 .0598 .047W .0335 .0287 1.00 .4673 .2942 .1852 .1473 .1168 .0928 .0730 .0580 .0409 .0350 V-PHASE UM IS .30 .0259 .0163 .0103 .00617 .00648 .00515 .00405 .00322 .00227 .00194 .40 .0430 .0271 .0170 .0135 .0107 .00654 .00672 .00533 .00376 .00322 .50 .0642 .0404 .0255 .0202 .0161 .0126 .0100 .00797 .00562 .00482 .60 .0897 .0565 .0356 .0283 .0224 .0178 .0140 .0111 .0078S .00673 .70 .1194 .0752 .0473 .0376 .0298 .0237 .0107 .0148 .0104 .00695 .80 .1533 .0965 .0608 .0483 .0383 .0505 .0239 .0190 .0134 .0115 .90 .1914 .1204 .0759 .0603 .0478 .0380 .0299 .0237 .0167 .0143 1.0l .2536 .1471 .0926 .0736 .0584 .0464 .0365 .0290 .0204 .0175 .-P.AS URI .3 .0173 .0109 .00685 .00545 .00432 .00343 00270 .00215 .00151 .00130 .40 .0286 .0180 .0114 .0W093 .00717 .00569 .00448 .00356 .00251 .00215 .50 .0428 .0270 .0170 .0135 .0107 .00651 .00669 .00532 .0037S .00321 .60 .0598 .0377 .0237 .0189 .0150 .0119 .00934 .00742 .00523 .00449 .70 .0796 .0501 .0315 .0251 .0199 .0158 .0124 .0098O .00696 .00597 .80 .1022 .0645 .0405 .0322 .0255 .0203 .0160 .0127 .0094 .0766 .90 .1276 .0803 .0506 .0402 .0319 .0253 .0199 .0158 .0116 .00956 1.00 .1SS8 .0980 .0618 .0491 .0389 .0309 .0243 .0193 .0136 .0117 Application: a. Description of the line section b. Wire size c. Length of the section d. System load factor e. Annual peak power (kW) (effective) f. Voltage_ g. Power factor .SO 9 Assume: * 7.2/12.5 kV system ** 90 percent power factor. - I Pno 9 Power Factor Multiplier: (.90)2 t.J (PF)2 Page 10 Voltage Multipivlng Factor: (7.2 kV)2 (actual voltage)2 . .~~~~~~~~~~~~~~~~~~~~~~~~~~~ pae 0 Method: 1. Transformer nameplate rating, I.e., 10 kVA, 25 kVA, etc. 2. Number of customers served by the transformer ..~~~~~~ Determine if customers high-use or low-use: * Low-use customers load of approximately 1,500 kWh *14 * High-use customers approxmiately 3,000 kWh. Pag 12 Table 2 TRANSFORMER LOADING AS A PERCENT OF RATING AND LOSS FACTOR LOW-USE CUSTOMERS Nof Nameplate Ratlng-KVA customers 10 . 15 2s so 100 Loss Factor 1 120 80 48 24 12 .026 2 - 144 87 43 22 .030 -.5. - 197 119 59 30 .035 4 - - 148 74 37 .039 s - - 169 85 42 .045 6 - - 190 95 4 .051 HIGH-USE CUSTOMERS X I - 149 99 45 22 .02 2 - - 162 e1 40 .034 3 - - - 110 55 .039 4 - - - 135 68 .044 5 - - - 157. 79 .051 6 - - - 177 89 .056 Note: Dashe Indlcate transformers loaded above 200% of ratino. This Is not reconnmenoWed. Page 43 - 179 Pigure 3a 45M ~~~~~Efficiency)~ 4aw~~ MaLd LOU." w / 10 is 90 25 130 to0~5 215 30= ism Inn4. - 180 - (98.4% EffiClfncv) mtwu"m -1 10 47 25 U8 10 I I 3sm .w zt 60 1I o t 120 140 160 180 200 I.Wn . .i.~~ ~ ~ . s - 181 - ?rmmforNIw Losse , 6.7% Gfftcencv) iiI Muis c AX !10. mLm 9 26 110S 100 401 1 LO w * 0f a1 Is~ so~~15 Iwo ~ aSg9M O inrnRui Economics of Loss Reduction Method: 1. Calculation of kWh savings 2. Calculation of annual benefits 3. Calculation of annual costs Paoe 21 The time period: * The useful life remaining in existing equ i pment, which will be replaced before It Is necessary to do so by higher .efficiency equipment. Paoe 27 Table 3C COST PER ADDITIONAL DOI !AR OF INVESTMENT FACTORS FOR DETERMINING AVERAGE INVESTMENT COST cost of capital StUdy Perlod 5% 10% 15% tYearsl New Old New Old New Old 5 0.0651 0.2310 0.1061 0.2638 0.1523 0.2983 10 0.0651 0.1295 0.1061 0.1627 0.1523 0.1993 20 0.0651 0.0802 0.1061 0.1175 0.1523 0.1598 . Page 43 Table SC COST PER ADDITIONAL DOLLAR OF INVESTMENT FACTORS FOR DETERMINING AVERAGE INVESTMENT COST Cost of Capital StUdy Period 5% 10% 15% EYearsi New Old New Old New Old u' 5 0.0651 0.2310 0.1061 0.2638 0.1523 0.2983 10 0.0651 0.1295 0.1061 0.1627 0.1523 0.1993 20 0.0651 0.0802 0.1061 0.1175 0.1523 0.1598 Page 4n - 186 - Typical Costs * Reconductor S35,000 to S45,000 per mile. * Convert Single-Phase to Three- Phase (Add Two Phases) S30.000 tO $35,000 per mile. * Convert V Phase Line to Three- Phase Line (Add one Phase) $25,000 to $30,000 per mile: * Reinsulate 1 0 7.2/12.5 kV Line to I 0 14.4/24.9 kV S5,000 per mile. - Reinsulate 3 0 7.2/12.5 kV Line to 3 0 14.4/24.9 kV $15,000 per mile. * Change Out Transformers 10 kVA S550 each 15 KVA 585 each 25 kVA 680 each 50 kVA 975 each * Add Capacitors $3.50 to s7.50 per WVAR. . .S TabIe SC COST PER ADDITIONAL DOLLAR OF INVESTMENT FACTORS FOR DETERMINING AVERAGE INVESTMENT COST Cost of Capital Period 5% 10% 15, (Years$ New Old New Old ilew Old 5 0.0651 0.231b 0.1061 0.2638 0.1523 C.2983 10 0.0651 0.1295 0.1061 0.1627 0.1523 0.1993 20 0.0651 0.0802 0.1061 0.1175 0.1523 C.1599 Table 3b SAVINGS FROM LOSS REDUCTIONS ON CONDUCTORS AND LOAD-LOSS REDUCTIONS ON TRANSFORMERS FACTORS FOR DETERMINING AVERAGE VALUE PER kWh BASED ON BPA'S PF-1 RATE Dollars per kWh SVstem Load Factor/ At Given Cost of Capital Study Period (Yearsi 5% 10% 15% 40% System Load Factor 5 S0.0353 $0.0346 SO.0340 10 0.0429 0.0413 0.0400 20 0.0551 0.0504 0.0465 50% System Load Factor 5 S0.0277 SO.0272 50.0266 10 0.0335 0.0324 0.0313 20 0.0432 0.039S 0.0364 60% System Load Factor 5 S0.0231 ' S0.0230 $0.0224 10 0.0281 0.0273 0.0263 20 0.0363 0.0332 0.0307 PS4e8 - 189 - RECONDUCTOR COSTUPPUCTV ass CALCULATIONS a" us"o _ ~~wo_ml. .~~~~S .A(D DWfllfl AnnUal oeIwnWn Annua WowlUft w. WWIUUwK lo. .-_ U. .I SC I .4, CM" Annual CM SQAffl Benefit (SB Ann"m 8WMW, OgWr Afnua SIVWfl lKWV u m"" wannw io one weng jaue~li UShIS WOO ar14t lo 14 arne_0c se ,<~~LC3 SeM wtnw - 190 - TRANSFORMER REPLACEMN COST-EFFECTIVENESS CALCULATIONS DeWftin~e NLad and Ls L . wor z Taoie 2 0. 43), Pgums a. D.-C (0. 44,. 45. 0). two - (t.D)/(Dx + D.) -Xt (1/D,) . The demand equation in this case is as follows: t -a - b D where b -(1/e) (t/D). Once the demand curve is estimated, the consumption of the regularized users La calculated by replacing t with the vith-project tariff t. and solving for the unknown quantity D.. The savings can then be calculated as the product of the marginal cost of replacing the connections mc and the reduction in consumption (D,, - D,). - 220- The reduction in WTP (WTP, - WTP.) also varies according to the type of unpaid consumption, i.e. whether entirely free or partly paid for. If the tariff is null (t,, - 0), the loss of consumer surplus CS is: CS - tw [D. - Dw] When the tariff is greater than zero (O < tw, < tw), the reduction in Ulit CS is calculated from: Price CS ft - t tw] [D. - D,]. C. UnRaid ConsumRtion with Gross benefit Improved Reliability In cases where the consumer ts receives a low-quality service, the w result of regularizing his situation and incorporating him into an D D kWh upgraded network may represent an ____ improvement and a net increase in his I consumption, rather than a Figure 7: Benefit to Previous reduction. In balance, the increase Clandestine Users From Improved Quality in demand in response to an of Service improvement in the quality of supply may offset the reduction in consumption resulting from the price elasticity of demand. Residents in low-income neighborhoods often connect themselves to a secondary circuit with their own wires, frequently of different sizes, the joints are done manually. The wires are taken to their homes, in some cases several blocks away, by hanging them on walls, trees, roofs, tall poles and whatever is at hand. Other residents connect themselves on the original wires, becoming for all practical purposes informal secondary circuits. Their condition making them not only a constant danger but also contributing significantly to technical losses. In these circumstances, the voltage level is very low. In many cases, turning on an appliance in one of the user's residence often reduces significantly the voltage of the other consumers in the circuit. In other cases users cannot turn on any electrical appliance during the hours of maximum demand. The benefit to the clandestine user is the shaded area between the two demand curves shown in Figure 7. The low-quality fraudulent consumption corresponds to D.. at a null tariff. Consumption increases to D. as a result of the quality izprovemnt, despite the fact that a tariff t, is now being charged. The benefit can be determined by calculating the area between these curves, using the same procedures described previously. - 221 - The net benefit is determined unit by subtracting from the benefit the Pio cost of the incremental energy, mc(D, - D,o) Considering the three i previous cases (i.e.: mc - t,, mc - tv, and mc < t.), the utility is in the most favorable situati'wn when marginal cost pricing is the rule no:\ (Figure 8), where the difference' between WTP and the cost of resources is at its maximum. In other words, ~ lik the equation 0DmD w 0aBD, - OmcBD, - t,aB Figure St cost of Additional Energy From Improved Quality of Service When Tariff reaches a maximum when t, - mc. Zquals Marginal Cost When the tariff is subsidized, the improvement in quality of service Unit produces a relatively larger increase Price in the cost of supply than the increase in consumer's surplus. The * dark triangle to the right of and above the demand curve in figure 9 shows what the company loses for being in a suboptimal situation, Mo i.e., from charging less than its \ production costs. Finally. when the tariff is D kWh higher than the cost of supply, the I___ increase in demand as a result of the Figure 9s Cost of Additional Energy From improvement in service quality will Improved Quallty of Service When Tariff be smaller than in the previous Below Marginal Cost cases, and, depending on the magnitude of the over-pricing, the with-project demand Dw may be smaller than D,. (see the graph below). Therefore the change in consumption of illegal users when thelr situation is normalized, under an improvement of quality, varies inversely to the differential b. tween tariff and marginal cost. It should be pointed out that with the project, not all clandestine users are going to connect legally to the grid during the first year after completion of the project. Some will during the firat year, others durlng the second and so on. Similarly their consumption when they have to pay for it will be lower than when they were not paying for it. The increase in consumption due to a quality liprovement affects only some consumers, I.e. the better off. - 222 - V. CONCLUSIONS The purpose of this paper was to present a methodological framework for the economic analysis of PLRP.. We therefore describe the methodology for calculating the benefits of such projects, distinguishing between technical and non-technical losses. The main benefits of PLRP are first, the savings in costs of system operation and expansion, and second, the improvement in service quality. The project may also enable a higher level of demand to be met, but generally this is a secondary benefit. The principal economic benefits of NTLRP (non-technical loss reduction projects), also known as revenue enhancement projects because of their positive effects on the company's finances, are the resource savings and the improvement in the quality of supply. An additional benefit of both kinds of project is the increase in demand as a result of tariff reductions in relation to the without project situation. These benefits are the result of the following effects: (a) a Unit reduction in illicit consumption and Prioe an increase --larger than the reduction-- in paid consumption; (b) a more efficient meter reading, billing and collection; (c) a reduction in technical losses as a result of the tw net reduction in consumption and the MO M reduction of NTLs; (d) an increase in | sales revenue and consumer's surplus as a result of quality improvements and the reduction in marginal costs. 0ow I Uw kWh No electric utility in the |__ world can sell all of the energy Figure 10: Cost of Additional Energy produced. The costs of attempting to From Improved Quality of Service When reduce losses to the physical minimum Tariff Above Marginal Cost would exceed the benefits. There is always an acceptable economic level of losses. This level varies from country to country, depending on supply conditions and the structure of relative prices. The level is reached when the cost of reducing losses by 1 kWh is larger than the long-term marginal cost of supplying it. In most cases this level is around 10% of total requirements. Higher levels generally represent an unnecessary use of resources and indicate ineffLcient past policies, especially as regards management and tariffs. In all the cases examined, the optimum operating level occurs when tariffs are equal to the marginal cost, the latter including future expenditure on maintaining the system ln good condition. If tariffs had been established on this basi., they would probably be no need to invest in projects to reduce losses. In the final analysis, PLRP are nothing more than corrective measures required by the lack of preventive maintenance and adequate policies in the past. Although it is difficult to speculate about what would have happened if other conditions had prevailed, there is no doubt that when losses are reduced to an acceptable level and appropriate tariff, planning, and maintenance policies are in place, there is no need for such projects. - 223 - I. BiblkUGorp 1 R. Anderoson y L. Taylor (1986), 'The social cost of unsupplied electricity: A critical review`, Enere EsnngigaW, Vol , go , July pp 139-46. 2 R.W. Bandaranaike (1981), 'Rural electrification and the optiml quality of electricity supply" (Univ. of Maryland: Ph.D. Thesis, abrll). S. Bh ntal y S. Ravid (1982), "A aple method for evaluating the marginal cost of umsupplied energy", The Btll Journal of econo;mcs, Vol 13, pp 249-53. 4 D. Blackorby, *tal. (1984), "Consmw .' surplus end welfare change in a aLaple dynamic model", R3j-jv of lconomic Studies, pp 171-6. 5 G. Brown y M. Johnson (1969), "Public utility pricing and output under risk", American EconoMic Review, Vol 59, pp 119-28. 6 S. Castagnino (1980), "Gufs. par& *I anilisis del efecto distributivo de proyectos de transmisi6n y distribuci6a de electricid'd" (Wash., D.C.: DID, Unidad do Metodologla, dicimbre). 7 M. Crew y P. Xleindorfer (1976), "Poak-load priacng with a diverse technologye, sell Journal of Econsmics, Vol 7, pp 207-31. 8 M. Crvw y P. Xleindorfer (1978), "RliabLlity end public utility pricLng", American Economie Review, Vol 68, pp 31-40. 9 J.M. Currie, *t.al. (1971), "The concopt of conomic surplus end its use in economic analysis", EconomLc JournsL, dicLembre, pp 741-99. 10 A. Darlin (1986), "Economic contrlbution to the projoet reportt Global loan program for the rehabilLtation of electrie energy subtranmisuion and distribution systems" (Washington D.C.: BID). 11 A. Deaton y J. Muellbauer (1980), Economies and Coniumer Behaviour (Cambridge: Cmbridge Univ. Press). 12 Federal Power Commission (1970), The Methodolos, of Load Forecastins, 1970 National Power Survey (Washington, D.C.: DOE). 13 L.E. Gutidrreu-Santos (1984), CoantrLbuci6a econ6mica al informu de proyecto de Mansbao-bejucal-Tavera y Centro de Deepacho do Cargat Anilisie eoon6iLo" (Wash. D.C.: BID). 14 L.E. Cuti6rres-Santoe y M. Webb (1977), "Comentarios sobre la evaluaci6m de proyectos con referencia Al sector energ6tioo mxicane", El TrLmestre Econmico, Vol.44(2), 3It.174 (Mexico, D.F.s FCE, abril- Junio). 15 L.E. GutLrres-Santos y 0. Westley (1979), *Economic Analys of Electricitv SuoDl Projects (Wash. D.C.: DID). 16 L.E. Gutidrres-Santos (1983). "Electricity Demand Forecasting: A Review of Currat NethodologLes" (Wash., D.C., BID, D.p. AnAlisis de Proyaetos, dictembre). 17, L.E. Cutlrres-Santos (1984), "'I cost* do la conflabilidadt (Wash., D.C., BID, Dop. Aknlisla de Proyectos, aprll). 18, L.E. Outlrrez-Santos (1986), "E1 anAlLsis ecounmlco do los proy ctos de transmIsdni" (Wash., D.C.g DID, Dsp. AnAlisie de Proy ctos, dic ebre). 19 AC. Barberger (1970), "Marginal cost pricing an social investment Criteria for electricLty undertakings', publicado en Haberger (1972). 20 A.C. Nrberger (1971), 'Three basic postulates for applied welfare econDmics an interpretative essay", Journal of Economic Literature, Vol.9, pp 785-97. 21 A.C. Darberger (1972), Prolct Evluation (Landan: Maomlll). 22 R. Kelvoreon (1976), Econometric Models of S. Eera, Demand (Mass.: Lexington Books). 23 B.S. S autbaker y L.D. Taylor (1970), ,Co r D.mn . thl Unitd Stts: nlysis and Projections (Cembridge, Mass.: Iara Univ. Press). 24 0.3. Ruck (1979) "Load Forecast Bibliography: Phase I", Proce-dins of the 1 EEB PE6 Sums Meti, Papr So. F7950-3 (Vancouver, July). 25 P. Jar_mllo y 3. Skokaic (1973), "Costo soclal do l-e restricctones do mergfa el4etrice, ME, Planificaci6n el1etricat Anhlisis de la sesuridad del abaste ie2tSo, (Santigo do Chle: ME, Oficina do Plenlfle Lcsi, agosto) - 224 - 26 D. Julius y Mets System In. (1981), Enerr! Pricina An ewlooine Countries: A Rovlow of the Lmummw, Enrgy Dep. Paper No.1 (Wash., D.C.: Banco Hundial). 27 R. Just , D. Duoth y A. Sebits (1982), Avplied Welf-r Zeongoios and Public Poliow (Now Jersey: Pretice Hall). 28 1.4.D. Little y J.A. Mirrleos (1974),.Pro1ect A Lraisrl gnd PoD loina Countries (Loandons HBin_em). 29 D. Mattson, etal. (1966), 'Econowy versus raliability Ln Svden, SIUEE 8Ss3tru, may 30 G.W. Kabenusi (1976), 'Measuring gains and losses', Journal of Politict l Economy, 84, pp 641-6. 31 C.W. MW nnie (1983), Measurin EconoaicL Welfare: Now Methods (Cambridge: Camb. Univ. Press). 32 N.J. MHihan (1973), Cost Benefit Analysis, 2nd.Zd. (Londom Allen & Unwin). 33 E.R. Morey (1984), 'Confuser surplus', American Ebonodia R viy, Vol 74, 3o0l, pp 163-73. 34 M. Munainghe yM. Collerson (1979), 'Econmic criteria for optimiting power system reliability levels", Tbe Dell Jounal of Econolcs, Vol 10, SprIng, pp 353-65. 35 M. Munasinhb (1979), The Economles of Powr System Reliability and Planning (Baltimore: J. Hopkins Univ. Press). 36 M. Munasinghs (1987), R_ural ElectrifiLation for Delonment: Pollev Analisis and Anulications (Soulder: Wastviev Press). 37 M. Munasinghe, J. Gillin y M. Mason (1988), A Rlevw of World bank dendina for Electric Power, Industry and Energy Department Working Paper No. 2 (Washington, D.C.t The World Bank, March). 38 U. MunaI nshe y W.G. Scott (1982), Eneraz Efficiency: Optimization of electric power distribution system losses, Enrgy Department Paper No 6 (Washlngton, D.C.: The World Bank, July). 39 QWUDI (1972), Guldelines for Prol-ct Zvaluagio (Now York: OIU). 40 ONIUD (1978), GuLde to-Practical Proieoct Azmraisal: Social-Cost Benefit AnalXiasin Develooin Countrles (Now York, ONU). 41 A.P. Sanghvi (1981), 'Customer outage costs La investmnt planning models for optiising generation systm epasion and reliability', Tractions of the C Eadin Electrical Association, Vol.21. 42 A.P. Sanghvl (1982), Economic costs of electricity supply interruption s US and foreign experience', fEnra Egonomics, Vol 4, No 3, July, pp 180-98. 43 A.P. Senghvi (1983), 'Optiml electricity supply reliability using customr shortage costs', *g£X3Z geonoeL1g, Vol , go , April, pp 129-36. 44 L. Squire y B.C. van der Tek (1971), Economic Anal_sis of Pro eot_ (Baltimore: World Bank/Johns Hopkins). 45 R. Sugdpn (1979), 'The masuremt of consmers' surplus in practical cost-benefit um .ysis', Anplied Konmice. Vol.5 No.4, pp 139-46. 46 M.L. Teleon (1975), 'The economics of alternative levels of relabilLty for electric power geraeration system, Mh Bell Jounal of Eonomis, Vol 6 , No 2, Autm, pp 679-94. 47 R. Turvy (1968), Qotlaa Pricine and Inves Electrict SUnv (London: Allen & Unwin). 48 R. Turvey y P. Anderson (1977), ElectrSiit cm a eae Studies (Wash. D.C., Banco Mundil,ls J. Hopkns Univ. Prss). 49 Ja. VieVal, t-al. (1986), 'Metodologfa aplicada a La 1atimali6 de p6rdida y energa no facturada en at eetma *l6ctrLeo de subtransmii& y d1stribuoi4m de Is costa atldntcea , CsMltorj& Vol (togota, Colombia). 50 1. Webb (1977), The detemxination of reserve generation capacity crlteria In electrLcity supply ycte', E Vol 9, March, pp 19-31. 51 3. Webb y D. Pearce (1985), Econc ne t f Per S l, y Dep. Per No.2 (Wash. D.C. enco DImdil, eptiombre). - 225 - LOSSES CONTROL IN ELECTRIC SYSTEMS GENERAL CONCEPTS AND DEFINITIONS By Dr. Renato Cespedes G. 1. Iroduction The growing costs of investments in the expansion of electric energy power systems and of its operation has motivated the need of increasing the efficiency of the electric power systems in the region. One of the means to reach a higher efficiency is to reduce the electric losses. The electric losses of the Latin American and Caribbean countries are, in general, higher than those observed in other developed countries and in consequence it has been concluded that it is imperative to reduce losses. Some studies have pointed out that reducing a Kilowatt of losses has a higher benefit to cost ratio than installing a new source to produce the same power. Several studies in various countries have been carried out in order to solve the 'losses problem". This problem has several aspects such as: - From a general economy point of view of a country it is considered that the losses is an inefficiency of the system but that only a part of the losses is really dissipated energy being another part energy that is effectively used but not paid to the utilities. In this respect more emphasis is given to the so called technical losses that will effectively reduce peak load demand and will reduce operation costs. - From the utilities point of view the losses problem is more related to a problem of loss of revenue and therefore it is focused with more emphasis to the 'financial losses" (non-technical losses) and to the short term and fast recovery of investments that reduce the losses. * From the users poLnt of view the problem is considered more a problem of the utillty alone and since Internal losses of the end users are in general not measured, no attention is put to a 'rational use of the energy". The losses problem is in general a complex one that has several aspects to be solved, including: - Determination of the losses sources, causes and assessment of the losses level for each one of the causes for the actual system. - 226 - - Determination of optimal long run economical losses levels in order to identify areas where potentially loss reduction actions could be more effective. Identification of the losses control actions that have a higher benefit to cost ratio. Classification of 'hese actions and establish a "Losses Control Program". In this document the losses problem is presented in a general form with emphasis in the identification of the different causes of losses and the analysis of the statistics that are related to energy losses with the objective to define common points of interest to all countries in the region regarding this problem. This document presents some aspects already discussed in the "Simposio Latinoamericano de Perdidas Electricas" of Bogota Colombia, October 1988, which is part of the losses reduction program of the Organization Lationoamericana de Energia, OLADE. 2.0 Demand and Losses 2.1 Losses Amounts In order to gain insight into the losses problem this section presents some statistics of losses taken from different sources. Figure 1 presents the evolution of the electric losses of a company. This company includes generation, transmission (220 and 115 kV), subtransmission (115, 57.5, 34.5 kV) and distribution (most at 13.2 kV for primary circuits and 440 and 208 volts from secondary distribution). It generates internally approximately 60% of its demand of 1300 NW. The figure presents the evolution of the demand and the total energy losses during the period of 1976 to 1986. They grew from 13% to 24.5% of the demand with an annual average growth rate of 12.8% while the demand increased at an average rate of 5.6% during the same period. For comparison purposes the 'square of the demand" curve is also included in Figure 1, consldering the demand square for each year and normalized so that it matches with the losses value at the initlal year. It can be observed that the losses grew almost at the same rate as this curve. Since most of the losses change with the square of the current it can be considered that this would reflect a system with almost no reinforcement. Unfortunately, the losses breakdown in technical and nontechnical losses is not known for the same period. The company report establishes that for 1976 the technical losses were of 11.2% and 6% of non-technical while for 1988 the same were of 11.2% (constant losses for all this period!) and 13.2% respectively. The previous data illustrate the need to know with at least a fair precision the different losses amounts in order to direct adequately the available losses reduction resources. - 227 - 2.2 Comparison of Different Losses Levels Table 1 presents, as an example, the losses reported by three different sources named Colombia, 1978, Cadafe and "Ideal".. This last one corresponds to the World Bank report source from which the column "Maximum tolerable) has been also extracted. Since not all the sources present the data in the same way some adjustments were made in order to have comparable data. For instance, for the Cadafe case the average values of the presented range were selected; also for Cadafe and the "Ideal" case the percentages were corrected to present them with respect to the demand instead of with respect to the generation. The comparison of the different data allow to establish the following conclusions: The "ideal" amount of losses (assuming that this vailue is applicable to the other cases) is much lower than the 12.6% of the Colombian case and the 10% of the Cadafe case. = It can be noted that the losses distribution among the different voltage levels between the "ideal" and the other cases. In the Colombia and Cadafe case the higher proportionally losses are due to distribution feeders and transformers. Figures 2, 3 and 4 illustrate these results. The previous data do not try to establish that the "ideal" case shall be considered as a target for the losses level of a country but it stresses the importance of comparing the losses levels of different countries together with the information of experiences, standards, design considerations that are applied in some countries and that contributes to have low losses levels. This can be considered by others in order to apply corrective actions correspondingly. Finally, the losses reduction programs cannot be generalized. The technical literature present cases in which the benefit to cost ratio arrives to as high as 15/1 for losses reduction actions. However, no generalization can be made based on other systems results and each case shall be analyzed independently in order to arrive to the correct solutions for each case. 3.0 Classification of Losses in Electric Systems 3.1 Introduction An electric power system is integrated by a complex set of generators, high voltage transmission lines, power transformers, distribution feeders, etc. Each element through which an electric current is circulating or is energized, that is connected to the power source like the case of an energized transformer, contributes to increase the system losses. The power (or energy) losses can be defined for each component of the system as the difference between the input power (or energy) minus the output power (or energy) of the component. The losses are in consequence the result of a reduced efficiency of the transport or transformation of a system element. - 228 - The instantaneous efficiency of each system component can be defined in terms of the input and output power. In the same form the efficiency over a time period can be defined in terms of the input and output energy. The following relation can be established: Losses (8) - 100 - Efficiency (%) (1) The previous relationship is applicable to power or energy according to the utilized efficiency being the losses less when the efficiency approaches its limit of 100%. It is important to recognize that in general all the system elements have different efficiencies which may change with the operating conditions. In addition, only for design the losses of each element are not considered independently being it necessary to group the losses according to causes and according to the sources that produce the losses. The previous conducts to the need of classifying the losses with several purposes including: - iThe evaluation of the losses by appropriate methods applicable to each case. The geographical location of the losses to establish the contribution of each part of the system to the total system losses. The contribution of the components at different voltage levels to the total losses. The contribution of elements with different functions (transport, transformation, etc.) to the losses. 3.2 general Clas1 ification 3.2.1 rovr,i&Sseu As mentioned before the power losses are those that are produced instantaneously in the power system. Figure 5 illustrates the demand curve and the corresponding demand losses. 229 - Demand * * * t 2** * ** * *2 22 t* . 2s P-ak 22 t* *2* 2 Load 22 :2~ t2*2 *2 Time Losses 22*22222* { .~~~~*2t S /22* 2 9 / 22 *** P.Losses 2*2 I S2// *2222221/ III// - Time Figure 5 The generation of the pover system shall be programed to supply the demand and the losses: G(t) - D (t) + LP(t) where G(t) is the generation at time t D(t) is the demand at time t LP(t) are the power losses at tim t It is evident from the previous equation that the losses increase the system load and require additional generation. In other term a reduction of losses reduce the total load and if thia loxs reduction is performed at the system peak, investments to feod the total load can be postponed. The power system losses can be classified in the following form: -Technical -Corona -Peak -Load depen- -By company Losses -Joule effect losses dent -By zone or -Eddy currents -Off peak -Fixed region and histeresis losses - 230 - 3.2.2 Ena.rg Losses Energy losses can be expressed in terms of power losses in the following form: Ie - Int (LP(t)) dt - Ave(LP(t)) T (2) where Le are the energy losses in the period T Lp(t) are the power losses as function of time t Ave ( ) indicates average Int ( ) indicates integral Being the eneregy equivalent to useful work, the energy losses is that part of the energy that is not uaed or is dissipated in the system (Joule effect for example). The following equation can be deducted from equation (1): Generated energy - Demand energy + Energy losses (3) In a utility the generated (or received) energy and the demand (or supplied) energy are measured by means of energy counters located at the different points where the energy is the generation minus the demand and should be the same as those that could be calculated from equation (2). In case that the energy losses are greater than the calculated ones this additional loss is also an "energy loss" for the utility although it is energy that was effectively delivered. This type of losses called non-technical losses cause financial losses to the utilities siLce the service provided by the utility is not paid adequately by the user. Considering the previous concepts the following classification can be established for the energy losses: ay type AM cause AXyU AXa vi]ariation AX site -Technical -Corona -Losses of -Load depen- -By company Losses -Joule effect perlod: daily dent -By zone or -Non tech- -Eddy currents monthly, ect. Flxed region nical and histeresis 3.3 Technical losses 3.3.1 As funtion ef Ileen - Transportation losses * In transmission lines * In subtransmission lines * In primary distribution circuius * In secondary distribution circuits - 231 - Transformation losses * In transmission/subtransmission * Xn subtransmission/distribution * In distribution tragnsformers 3.3.2 By losses causes * Corona effect losses * Joule effect losses * Eddy currents and histeresis 3.3.3 Sumay TsbLe By Types By cause 100 Transmission Corona Subtransmission Transport- Technical Primary Feeders Losses Joule Secondary Feeders Trans./Subtra. Transforma- Eddy Curr. tion Subtr./Distrib. & Histeresis Distribution 0% 3.4 Non Technical Losses For the non technical losses the following classification is proposed: Consumption of users that are not registered by the company or theft: Comprises all the direct connection of the users of energy to the power system without having subscribed a contract or agreement with the utility for the energy use. In this group are also included those users that had a contract with the utility but that having been suspended, reconnect themselves to the network without permission. This type of users have of course no energy measurement. Energy measurement error of subscribers with energy meters. Comprises all those measurement errors of meters, reading of counters and billing of subscribers excluding the cases of equipment adulteration. In these losses those due to the non-simultaneous measurement of counters are included. - 232 - - Error of estimated consumption of subscribers without energy meter. Corresponds to all those subscribers that for any reason are billed by an estimation of the consumed amount. Includes those cases of temporary users which the utility decides to bill without a meter. - Energy theft by utility subscribers. Comprises all those cases of which the user, being a utility subscriber, adulterates tha measurement equipment or takes the energy directly. - Error in the own utility consumption. Corresponds %.o the energy used by the utility but not meterd. Includes substation auxiliaries, street lights, etc. 4. Enrsy Statistics One important aspect related to energy losses is the statistics that reflect the energy produced and sold over a certain period. The statistics are normally presented as energy balances and included in the annual report of the utilities. This section presents the various aspects regarding energy balances that are related to energy losses. Figure 6 presents a typical energy balance in a graphic form. This balance does not include the generation losses since it takes as input the gross generation at generators terminals. Also it does not consider the end use of energy at the user level. 4.1 Partlcular Ccnsiderations (a) The generation of the electric energy is a process that has losses at the production level. In effect, the thermal generation and in a leaser extent hydrogeneration requires energy for the auxiliaries. This ioplies that in order to reflect the efficiency of the transportation, transformation and distribution of the electric energy, the energy used by auxiliaries be subtracted from the generatiou which is the net generation of the system. The previous will alow to compare on the same bases systems that have hydro and thermal generation avoiding those systems that have less effLelent generation and could present less losses in pe,centage. (b) Som electric systems generate all the energy they require to supply their own demand while others purchase additional energy for its supply. These purchases shall in consequence be added to the net generation in order to have a statistics of an available energy input to the system. (c) The sales of energy to other utilities correspond to commercial agreements that in general do not reflect the needs to the own demand of a partlcular system. In consequence lt is convenient to substract - 233 - them from the available energy in order to calculate the demand energy of the system. The demand energy varies slowly and in general increases constantly while the energy sold may present substantial changes from one period to another. (d) The period considered for the energy balance may be variable. In general the most important is a year. In some systems, in order to analyze the seasonal variations, it is important to have balances at shorter periods, for instance a month. In these cases rather than considering a month as period it is important to consider a moving year period ending at the month of interest avoiding in consequence the problem that are found in process like billlng that cover longer periods of time. 4.2 Percentage of Losses One figure that is commonly used in energy statistics is the percentage of losses. It has been found that in order to find thls number different basis is considered by different utilities including: - Case A: The available energy that is the net energy input to the network including energy purchases. - Case B: The demand energy but considering as part of the demand the auxiliaries consumption in generation. - Case C: The demand energy without the auxiliarles consumption in generation. From the generation point of view the losses in percentage shall be calculated according to case A that is considering the total energy input to the systems. However, -;his figure may reflect variations due to the sells to other companies. Therefore it is important to also compute the percentage of losses from the demand point of view, that is, according to case C. 5.0 Other Inmgrtant Aspects Related to Non-Technical Losses This document presents as appendix the paper 'Assessment of electric energy losses in the Colombia Power networkw which presents the work performed in order to determine the losses-causes and amounts-in Colombia and the best approach to correct them. This study reflected that of a total of 19% of losses 2/3 were technical losses while 1/3 corresond to non technical ones. The non technical of 6.43% are distributed as follows: - Non calibrated meters 1.03% - User modIfied meters 0.930 - Damaged meters 0.63* - Estimated consumption 0.92% - Theft and others 2.92% Total 6.43% - 234 - As noted one important source of non-technical losses for this case is found associated to the energy meters. In fact the calibration of meters and the replacement of damaged meters would reduce the losses in approximately 2%. In addition the installation of meters in order to avoid the estimation of consumptions would reduce the losses to approximately 1%. Theft in general, including the modification of meters is responsible for approximately 3.5% of losses. This amount can be reduced only if several aspects are considered including: - The necessary legal backup to the utilities in order to permit the punishment of theft. - The periodic visits to users with energy meters in order to detect the cases that imply that the meter is modified. - The periodic visits to users with energy meters in order to detect the cases that imply that the meter is modified. - The measurement of blocks of energy in areas of generalized theft in order to determine amounts of the energy not paid and determine the best way to legalize these areas. - The supply of energy at tariffs that consider the reliability and quality of the service. 6.0 Conclusion& This document gives an overview of some important aspects of the 'Losses problem" including the classification of losses and the statistics associated to them. These are considered important since they permit to define the causes and amounts of losses which are crucial questions to be answered before starting efforts in order to reduce the losses detected in any system. Finally it is considered of utmost importance that general definitions are adopted in the Latin American region in order to start jointly an effort towards the solution of this problem which without doubt can be solved by us with our own resources. - 235 - BILIgas 1. M. Nunasinghe and W. Scott, Energy Efficiency: Optimization of electric power distribution system losses. World Bank, Energy Department Paper No. 6, July 1982. 2. R. Cespedes et all. "Estudio de Perdidas en el Sector E16ctrico Colombiano", resumen presentado en congreso de la Asociaci6n de Ingenieros Electricos y Mecanicos, ACIEM, Bucaramanga, 1981. 3. R. Cespedes et all, "Assessment of Electric Energy Losses in the Colombian Power Network", IEEE Power Apparatus and Systems, November 1983. 4. Empresa de Energia Electrica de Bogot&, EEEB, "Programa de reduccion de P6rdidas" periodo 1987-1992. 5. Sistecom Ltda., 'Estudio de Perdidas del Sector El6trico Colombiano", 1980. 6. R. Cespedes, "Now Method for the Analysis of Distribution Network" to be presented at the IEEE summer power meeting Long Beach California, July 1989. 7. L. Mazzacan, "Metodologias de Evaluaci6n y Reducci6n de Perdidas Tecnicas en un sistema Electrico", Simposio lationoamericano de control de perdidas, OLADE, Octubre de 1988. FIG. I ELECTRIC LOSSES EVOLUTION OVER 10 YEARS 2.1- 2 1.9 1.8 1.7 1.8 1.5 1.4 'S 1.3 1.2 1. 0.9 0.8 0.7 0.8 0.5 1978 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 Mw orGWh o Demand in Mw + Total losses O Cuadrado demando - 237 - FIG. 2 COLOMBIA 1978 P.Tocnkas en X 3.10 (24.6Z) 4.02 (31.9Z)/ TRANSmISMON \ SECONDARY subTRANSMISSION FEEDERS IUNES 0.93 (7.4%) DISTRIBUTION PRIMARY TRANSFORMERS FEEDERS 1.87 (14.8;9 \ * 2.70 ~~~~~(21AXS) - 238 - F7GA 3 CADAFE P.Tocnico. an x 1.98 (20.OX) SECONDARY TRANISSION \ 4.23 (42.68) FEEDERS LINES 0.44 (4.4x) \ / i ~~~~PRIMARY / \ t~-EEDERtS DSRIBU. \ x / tRAN8FoR"* \ / ~~~~2.14 (21.6XJ 1 RANSFORtM 1.13 (11.4X)~~~~~~~~~~~~~ - 239 - FIG. 4 SISTEM "IDEAL" Perdidas T.cnicas (X) 0.43 (BA.X) 0.43 (O. IX*),1 iEONDAR DIT"R. ,/\ FEED: TRAN TRANSMISSION & SUBTRASMISSION 3.21 (45.3X) PRIMARY FEEDERS 2.59 (38.5X) 0.43 (6.1X) - 240 - Fig 6 Gross Generation Smas Purchased Energy coumo _sy Net Generation power Plant Ava ea' ¶rAgy Aux-iliries ConsumptionA Demand Energy Faunas ~~~~~~~~~~~~Sell to waU%c maa|aa n er UtiLtLes Technical DelLvered energy Losses ntternal coasumption at substatLoa and thau\r . ~~~~~~~~~~ot,ters Non-technLcal losses MIG. 6 BtlUed energy REPIESENTA CION GRAF ?CA DEL BALANCE ENERGE-0_ GRAP!C REPRESENTA7:C= CF THE ENERGY BALANCE T*l* it : au Mmtt ROLE W EWBY LOSSES COlOHSIS-1979(l) OwW (2) *IOERL8 (3) U3i1mr (3) TECNICL LOSSES X OEMH. X TECH MIC(N) C(UREC. XTECN. fUSC OIWC ZTECN. ASIC CORRC ZIEM. Transmission lines 1.99 35.? 3.00 3.99 20.0 1.40 1.50 21.2 2.60 3.20 22.6 Subtrarw.issian lines 1.12 9.9 1.60 1.71 24.2 3.20 3.65 2.8 StAstation tranfor_m 0.93 ?.4 0.40 0.44 4.4 0.40 0.43 6.0 0.90 0.91 6.5 Priwi Fod"s 2.70 21.4 .1.4 -2.5 2.14 21.6 2.42 2.59 36.6 4.01 4.5? 92.3 Distribstion TransFormrs 3.87 14.8 0.7 -1. 5 1.1) 11.4 0.80 0.96 12.1 1.60 1.83 12.9 Secondary Feeders 4.02 31.9 3.1 -4.6 4.23 42.6 LOSSES StmTOTuUL 12.62 390.0 7.4-10.6 9.92 100.0 6.62 7.09 100.0 12.40 14.16 100.0 WdI-TE(ENICRL LOSSES 6.43 tmmRL 19.05 (1) mEstudio de Purdidas Sistoma Electrico Col1obiano (2) L. e,aacan. letodblogias de Evaluwcian j R.duccion do Perdidas (3) "."WNasinghl, En.rW EfFicincy: Optimization of electric distribution... :er (3) p dk losses converted to .nmrgj with 1.24 as factor (m) Basic: data according to somaG normalized uith generation Corrected: data normalized with d&%nd aondix A - 242 Page I of a tMOS T.sam on e.Wr AMAIInVWt sad syue. Vol. PA54IQ. 4w. I I.l sebet 1 ^SESX5: Or SIMMICAL E tE IC SSES 1I THE COLOMBIAN io SYS=s U. C$spedes U. DOrIa 3. 11enin.bo A. Rodrigues Hauer CItE Senior tb"er IEU Senior Ilember tCEt E mber ICS Statecom Leta. Ssteco Lcda. VCA UA seges, Colosbia So, l, .polebia lieS delletSn Coleobia le.doltSn. Colombia &isyact The electrical energy losses for the cifte objective wvre defined for the studys ,neire ColomiaZn powe network coi rising distribuLton .ottage IovcIs up to Lbh high voleUge tranSiRloian 1e- gvaluate the nfoamation available aC Lbh dLffe- ale are calculated. The enerty loses are classified rt utilities. memor of the Colotbitn Intercova ,s "physical" lose correspondintg to Joule effect - neeted power network in order so determine i CI2R).Corona and core transftor losses and "black" - applications ti the stud. osases which are defined as the difterence betvesn the mnergy available at the consunr loel and the energy- Estime L b t enerGy losseo by utility. bt la- effectively. billed by the different electric utilities go leve1 end by cause et o wish a methodologty in Colombia. A new meshcdology for calculating the - defined according to the avable inforatim. 'physical" loae.e is presented which is extensively ba &ed in the use of cesputarized methods including state s Evaluate the eceoic inpact of the "argy ie- sceioation for the bigh voltage network and radial lo- 4es. id flow for the distributtion levels. The "blacke los- ses including mewtring errors. theft end billing or- Determine and alyze ia term of economy for - nrs are calculated vtch statistical etlhods also using the utilities. possible corrective actior.6 thut- computerized tools. The results obtained highlight - wold control and reduce heb loes. the importance of the losses in terms of lee of reve- nue for the electrical enrgy utilities. Te scope of the study cered the entire Colon- bias powe system antd the methodology used ase desig sitoDuC?loQt ed La order te ote the study objectivs subiuect to - te ti_m schedule and the huean resources allocated - The electrical energy lossee it a power system - to the study. This comprcnae resulted in a methodo- are produced by diffrent causes including: Joule of - logy tbat incorporated co-outer baed tools which oia fact. core losses in transformers deficienies in the tlified the losse estimation proces and mat the Xn- Catering system, theft. etc. These causes can be grle quired accuracy level. ed into wo subgroupst the first including those that create loasses that are inherent to the efficiency of This paper presents the methodlogy, used in the the system to produce energy ad to transport ths a - losses study of IS for the Colombian power sstem. nergy to the end consumer. i.e.. up to the point where The methods employed to determine *U the identitied- this entrgy is sold by the electrical utility to the - types of losses are discussed at the lizht of the ob- e'aor er: tle second group comprising the causes that- tained results. Mor emphasis is given to this as - .reate the differenca between the energy available ac pct cas to numerical rtealts whieh ae parcticular to the consumer level and the energy actually billed by th power system studied. the electrical utility. The first group of causes re- ults in the "physica l" osaes of the power system; the camsifncmTIo or LOs= -s904nd group is repoesible for the "black los es as *aead in this paper. noe en;rgy losses in a FowST system were caosi- fied according to t followiag criterias Inrercoeuin xIleUterica g.A. * SA. the comp ay - that coordintes the national intercoenected power sy_ a. By utility of the ColombLan power naet"ork. tee of Colombia decided to undertake *a coplete *uudy- an order to aSess se e aCnunc of e rney losses in the b. By voltage level where the losses Occur. .utire Colombian power system. The tc major gals of ESA for this study weret a. Dy type of caus producing the losses. * TMe classification of the losses according go - feurteen utUlitties of the Colombian power system their importance it terms of economical impact - were cenidered in the study; thes ucilities n be and the feasibility of corrective actions tat elasltfid asJ: would reduce the loSses. * Two companies with only high voltage trasais - he detisitioe of policies that would centrol - aim lime and generatifn power plauts. *ad reduc the losses in the future. * live companies cemprising genertion. transaia - le order to nrech thes goals she following ape- aim and distribution of electricl energy. * SeCve utilities with power networks ac Che sub - tramsrn eion and at the distgibutieo, lels. Ite emerry lose were evaluated for each utli- ty eaiderinr .al v.lsa..- level, comprised in each - aem. to addit g.-.. hel .- e.u for tlb (utLrrased Co - Iombian pwer tnel.nrk wo,.e obtatnerd by totaliaing the Partial trsults. *i-Ja'sr w.AS rIll-I.lsel1.1 i.l it". I - Dy veltawr Iyvil tIhr em'nrrj IO:SM'1S were cla"i - : .'Z S.:.*t""6; ".v.r lia',;.*' g .-!f'ed as: *_ ! | ~l't * | *A - 1.^ , ~~~~~~~~~~~~F lIt S AWlselZ* w 1.1 Aoondix A_ - 243- Page 2.of 8 3510 * Loeses In the transeuao'ic #rates. 1) Tranisseion line losses (220 and 113 KV). Li) Transforsur lossOS. ill) Subtrans.ission lne losses (66. 37.3 sud 34.5 Xv). * Loeses in the distribution sysem". O Primary feeder lossea tl.S. 13.2. 11.4 sad 4.16 XV). .0% it) Dietribution transforacr losses. Lit) Secondary feeder loses (all low volt*ges - Included). ay type of cause the losses werac cussfied "o I * t'hyaical" losses which correspond to: 1) Corona Wfaect losses at high voltage. .Al ii) Joule effect louses in lines and feders - I and in transfor.ors. ill) Transformer core loves. got * USlack" losses due to the folloving reason s f) Decalibration of energy asters duc to nor - sal veor or due to vrong calibration by the utility. Ui) Intencional decalibracion performed by the customers. iSi) lypsas of the energy mteur_. iv) Daeaged ntaers (blocked totor). v) Ersors of billing cuscocers that permat- ly or tomporarily are billed withot e_ergy meters. vi) Thnft of enorgy of persos not registered - as cUStomers iA the UtilitieS. . _~~~~~~~~~~fg 4. 6MOsnf _p gom 6 It should be noted chat the aeerty lost in tee - generacion plant and substation auxilaries ts not in cluded in che clasification of losses by voltsge 1a v:l or by cause. Tha rtesoe for not including theo - energp los*eb in this study wa because the electrical utilities cozsider thac the amavt of them ar well, - t?e easminacis of the infonatiLn uavailable for known and little or nothing cas be done it order to re the setudy or collected during the study gave the fo- deco then. For this reason the energy co"iderd a lowing ults the input to the power system Is the energy generated- tgrss generation) minus the energy consumed in the - * the inforsatios La more complete as the voltage- awialries. The net enrgy input to the system to ma levl increas. vhich alews for a good occurs- sed "Available" energy and to in consequence coulder- cy in the calculation of lesses at the treanis- ed " lOOZ for th*s study. sion and subtrausmissio levels. figure 1 presente a diagram representiag the o - * At the distribution lcvel the infcrmtio - nergy losses as classified for this study. rarely Complete osd psrticularlv at the secoeda- cy voltage levels where the expecced loaes were The "black loses calculated in this study ae greter, the information available at iost of - Identified in figure 1. All the types of loss" pro- the utlities in Colosbia corresponds to globsl- mened in thin figure were evaluated eparately with- figue (total legthb for exmple) or is lnexis- the eceptin of theft and bypass louses wbIh were - te. considered as the "slack" varibl, and were n cowd - quence evaluated as the differcoce between the enegCy- So informatin en theft ad decalibrotim was - available for blling. the total billed enrg and the almot noexistent at the tim of starting the - clcaleuted *bl.k" losses. study. EMEIODOC 70R LOSS!S EVALMAT!O?V According to thbs reslt, th stuy ws divided- in e osecutive phasess Genera Consideration. * . Ce one was dedicated to evaluate the "phtyi - The classifieation of loses retflecto the need to ecal losses ancluding all transussioe and die - have ceveral methodologie in order to evaluate the i- tributioe losese. dentified losses by cause. In adJiaun it was nececoe ry to adapt the methodoLogy accord4;n4 to the informs - During Phase Two the "black louses vere evalua- tine available in the poWer *ysCes wt;iL.eC *n4 to - ted per cause as dentified iL the mectioo doe - the characteristics of the power oystce of cach utili- cribnag the claasificati"o of tho lowses. tyS Table I presents a suomary of the methodologies- - 244 Page 3 of 8 3311 lad for the evaltucsion of ach type of louses accord- using the available neaaurcento and In ordor to sint .43 to the available in(@orutS9ioa. Tha14 shothdolotgie mlio che M3as9nrei!nc effort in the field, the use o. are explained In or. dtail Sn the anet Sectioua. State actiatian techniques van adogted. The losses were cestimaed bcth to power (7'.)- and in energy (Klrh), A state escimatton projrau wh(CIh uc ea,winYMg -wnmIaIInma.Anseggs the loast equare veighted matlud CIL.C23 was deve lipod. This program can accept all typesat poef - systes measuremencs includLogs . * ,*t*. *. Voltage (KY) at all bue-bars 4ll, soms" * wso_ all as - $gm beamm* Active sad reactive flows in lines, tranufoguer $" sa b , * ed injectioan (loads and goneration). ".N an.ei.sa snle -st.* Currents (Amps) to lines, trnsformers and ia3t tions (loads and genoration). Angles of all bun-bars. TamIl *, :".:::"' wainsaEu Sv *t '_n e ' m- SThe deloped program allo. the use of all sa- surements logged manually by tho substation and plant am*ff t59010#7 - S*U~Od * PidSeI ~ operators wich a perLodicity of half or one hou-. * of Ut.s. t.w *g ; 4s e * t O a ,a Ir this data, it appeared that enough redundancy in gam 44eaSh at foffo n available In order to toke fuU advantage of stato as =er *I Sft"O timatian. The major advantaps of state 41tim1t.ion - lMM144610,11 4*Wo,%Io 4". under the con4itLon described were: ''o'o"M "*' S provides a reliable power fle solution which ,,,pC,. toininizes the influence of the normal errors la Su . bedded in the measurements. Crrorn were escizated AIWOU * bev "Itsme. _ * gs.tuawta Oweg*o 1sev to be 2: for voltage seauucaeute. A to LZ for- ";Say gm.4of.s t4141as i. active paver ad 27 for reactive poer. *U * ea.. U * "at 6 * 2t provides a useful maner to point out gross - 5t4au aselLs btiiSae t. eel4 errorS like configuration changes not reported go in the log sheets by the operators. The program - amI o g cm b s1s* identified 'wrone masuuretnt like flow _, ~~"" "'u'"" With wtong flow direction. * let! order to asses the amount of losses a saple- .,3ste ffeCt ase in the Trainsmission and !sbtrns - week was selected. This tize period was considered - o.'ssicn Systcm the oint s to include different oporiting coditioss. Of the " hours of the week, the 91 hours vic tho- Us infotuiaion available for the estimats of mst varlable lod coditions were analyzed usLin the thesa 'losne couisXcod oft field asuremnrs. The analysia we perfotad a f * Energy mceor readp. * lead ings of ether paer systm data Uk. volts - gee, active and reactive power in lines and - trnsformes. * low system topology and parameters. r The direct use of the ncasusoe-tcs for lose - - _ / _ - valuation vwa cosidered not fet ible because of &be I. - I fol,lowingg reasons i V * lf couplets measurenta are takeu Is all the re - _ quired nie. _ _ - _ _ - - * Thea pertodicity of the mesresrato eto ng frnom c ze compan to arather; thin lpacts e inter - _ _ _ - - eeeti l dato._. I u _i_ _m The mabursests are not ken simultaneousy. _ _ _ * The aeeuracy of the readigs in further impacted ey human errors. L_ Figure 2 presents am an ex4ople, the active pow- "St" pus.. " "ions to e Ism maugas. er seaurementn tohkn at both enJ. ef a 20 IV line. : W This figure illustrates tse isrv..sisblity to ase dire tt, and ro labLy the field ^ sa r t.-c ;. to caleulate- the CranosiuLas"M losses. an tk. dirtrsoi.ce bet*'cen - both J.sew%UQ*otg would W IJ ; .V.v.r r% *....a wh iih dO nOt correspond to reality. tn order to overecv_ he .h*..daffieulty. still *2S12 - 245 - sPgo 4 of 8 * The power lmsaem wYfrv determined for eaeh I,usr f Vor the feeders for whieh thr l4e%.poly 18.1b k,nvm with clip astnt. esiust ion prorran. A.: ,nalvois- *nd convsiduring the siF5plifica-iuns ,iarertIjc a 3 - subroutiae of the ,aroUr.w cataabli.isdirectly - idai p.ver (icw program wva devoltoiurg- hsezd or. WCZa - the lOatloA. Ivy cow4'Uny. by volt.esC l.vel and cal- renie C43. A SaPple e' uor# tlian 201 for fo..4; o: the culatcd separately the louses of lines ond tran. Analyzedl utillties we ototuied. * The enorRy losses were determined by urical - the dare used by the progrom is the foll.,vinal Integration of the pover losenR for the studied- * Dtac logsed at the eubscet±ons: bosntar volL&a.- tbi period. The trapezoidal rule of inteCra - phaso currents, ative end 'tartive pevur for - *~ iasd. the feederor (where available). energy canbtua.d - The results obtained for the energy losses were- by the feeder (whore available). as follows: * leader topology and transformer lo3dm as eumsur- ed in the field or calculated accordithg t. diu - * STransinsion linac: 2.0: of the available energy tribution factors vhich assign to each tf.'nefwr- for the entire Cotoub"Wn system; per utility, te mw ar portion of the total feeder load ba.vd on animurm vas 0.2S and the uaxaim. 3.5X. the transformer capacity. * Subtransmission lines: 1.12 for the onLire Colon The pwer louse wera calculated only for the biWa system; per util&ty, the minimum was O.1S - peak load detted for each feeder during toe to t and the taxi2ma 1.6S. week. The enerqy losses wert calculateJ rton cha - * Transformers: 0.91 for tFo entire Colorbian sy- peak losses, tie tize period consi6ered and the load tem; post of the utilities preeenzed losses ia - and losses factors defined asi the range of 0.5 to 0.6. - 168 it FIs* ' l t8 '~~~~~~~~~~~~~~~~~~~~~ JIM fc 0 e}* 1 8 E Corona tffect tooses L a" k-l V I63 kl 06p@ak The Corona losses were calculated fot all the Ii wheres new vith volt.oge of 220 W., the ocuinu voltaae levl . In operation in Colombia. The method of Co-ber gad Za t(k) to the current easured for hour k of thc saample fauseUa was selected after a eomparison Oith the m - week. thods proposed by Sugiaoeo and Cari and Clade tl. *as it gave intCreodiate results becween the calculatioa - I peak is the peak current of the saple weat. for new sand 'old" conductors proposed by Carl and - Clado whilA the Suginoto metnod Save lever values - F is the losses factor than the other two mtchods. Ths C4rena effect lsaas- L calculation took into account the configuration of the F is the load factor lines, tho charactcristics or tb,e conductors and the - netcerological conditions. Thc latter omen vera Xs - The energy losses can be calculated in perceata- thered fronweathercontrol statieca located cests to - te as: It. areas of the transmissiun lines; data of the same- sar-ple woek as for the transmission line losses esti- Energy C2) * look (2) s mtion was used. Tha enargy losses in porceutagc were Losses lssese F e*rtiated to be 0.8% of the available energy. Prin.rv 'Ditrituticn Feeders TS. reslt obtained. for the prrcary distributico feeders were: 2.7S of the available enercy with a - All the feedor.s in Colombia peprate *a radl 1S It aniams of 7.4X a*d a mini ua of 0.9Z depending oa - Mse thus. the anClysil for evaluating the torsos in th utility analysed. this part of the system was consequently simpler than the one used for the tranuaission and eubtransuisr,ion- Df,orib tion Transformers levels. The folloving *io,lificatione were made: tor distribution transferness lossee evaluation- * The power factor of all the loads was asoan.d to 8the same sLuplif icacits as for the primary fetd4r o be equal to the power factor of the d ende t veto nade. Joule eet lossee and corer .bured at tol substation. lo sses were caculated frem equivalent trscvf'orsr- asucad at this subscation. circuits ihith were detercined for each tranwfoemer - * The unbalence betwren pauecs was noglected snd a tyac.. The losses in th diacrbutf on traasfcrmer ue- one linc equIvalent was. oed for losses calcula- it estinred to be 1.92 for the Colombian ovaysta with clon. h Values per utility in the range of 1.52 to 2.2. * The amouat of load In each distribution transfo! Seondary Distribution Feeders _ r was dettcruined according to the available i S forcations maseored peak loads or load discribu- Ibmea .4tdoloz wpleed for the prirsry fee tiou eccordloa to Cronsform.r capacities. dos va applied for the evaluation loases ln the er- IThe vlLaEe was considured constant for the fee- condary distribution feeders. Field ceasutemen:t and ders. customer distribution factors were used instead of - substation measurements and transforver distribution- The previous simplifications do not introduce - factors. The evaluation losses for the Colombian *Je significant errort as o azterained for the power systems tea ii this area where estirat&d to be 4.02 with a - analyzed. In particular, nor tousaiering the voltare- range between 2.42 ead 6.3S on a per utility basis. dtop introduces an trrOr of lue: than J.Z whicl, is no gligible given thc accuracy of the reitainC data used. O I ck Lossee These losses were eveiutted considering all poust ble ources of enerSy loses; as presented an figure 3. -246- Page 5 of 8 23S3 ml.TT q=1 "I-0 nQI - rSeeWW'IA .in www~~~~~~~= tz1 tbl figure the eaUse thA produc* po .AUCt po;aets vere determ"ed and wftb thi does tdu difte - .adccVrary losses 4r Idetified. Due to the Se - role models w re capared. Due to the fact that zot ctsicial nature of the prroblema *selple of custmonr - of the av7ag c osupcion currnts of the custowrs- wAs selected. tho *sie of tho saple w" decernined - are in the range of nrmoal 1 urrec (boceeu 1 10 1d - to e in1l the order of 0.51 egtehe tocal waber of cus- 30 .) and that in thi area the Ln ar =441 ,erovidid- cotri of eac ceaepay. As the rnge of cusco ers for actunce reoulte. ttLis model was slcted for teS let; hem utilltles can iderW was bem*n 50.000 cnd5a.OO-00 " eva tiou. iors tbt &WI eLs varied fea 230 to 2500. A tots a^tatuLag qusons related to the coadtiors of te h_ T Shr touts ase ted in thi study for d eletrclInsalation d4 the enargy asters as de - calibrated seters were 1.1 of teh er $zble enrgy- ttNccad 4arinS floldtn cuts potforsd Lathe copapo- tor teh Coloubiao system but tho anaLysLa per utULity ties" test workshops es.Lboraecd La order to detect prvded rsto rangig from--O.1 (doclbra n doa the teug of 1 and to evaluate the oss" for the- vo bl0 to utLUty) to 3.4S. -lected *=PI. The crus o- c' installations were Insp ctte thai * rv ud oate verzt raowd and es librated by the corersponding utility and e irzt llg- or changd itf ound to be daget. oThefllowiag s c- gits piret th nothodoloV used toe veLae the lo& ls gas am the results obttaine. 2b Ue loor, lto nte,r- 4o calLbratSmot f -ais cestil or ;&a ts; iecerrc elstntr dn "tfrte d by theY -titUty are considered uodcer isa came". IrLp, 4 pr / wte a tytScl ewet of Lbratio s fusceLon of ammlo the "weretagp of nomial current. niTsm m ro1 st ' mmt mWhigh cur t are ideari:isd to ebSu f lgure. ~{_ rl For ver Ifv ckroents the decilibrco en ecaBches tOOZ - 4" to the _eer * -Orly consmpion. Three 1oL to APProeaXte the _coer curve were cznsiderti in the tu_ 4!; these were a quadratic. a lao.,se hoic &ad a IL - Nw r moe. The three &,40 tesr -n :'!s"ated in figu- _ re 5. In this figure u" - - ts t f a -eter th|. Lo - 1002 ealibrated at lhOZ ani ;,4" ofi -witaA ecurrent - rU n u Oto hew. for teM evahaacLon at lesses three points- , . Wffr Jecfr42VLd for teeCIGS each o21 0( tef h 8Pted me r the selected points were I :0: . : anJd200S zoof at ses w zo t j eimpl currynt. Ts decaliectiin di.ues toe these - -247 - Page 6 of 8 ttue _rn wars calculated. 11w eustomruSin the eain arr . hilleuI without m_ter wvr tentidered to havde the :.M consuption as the "typical" customer. The lo- eetf were VVStiSaLVd an the difference betwen the esti .a-ed consumption and the billed energy. Ie lose0s- - - evalugated due to this cause were 0.91 of the availabl- * enerry for the Coltombian system vith a minirmi of O0 and a maxim of 1.6 oan A per ?onpany basis. The - 1euputer was extensively utsed tor this calculation as - a . F f f l the numbet of analyzed custtoers vwa over 50.0'l. Theft and lyoses - I X / These losses were evaluated as the difference - M if 7 - between thteavailable enrgy for billing (see figure w s.wsm ... -k 1). the billed energy sad the evaluated "black" ler - 0) . // se_{ ga. The losses estimated were 2.92 for the Coloe - biad system with a max*m of L3.5Z and a minimum of a _ ~ - -f - *~--&- l.82 according to the results per utility. Additional Methods Emloge a.,... For primary and secondary distribution feede - _ .X _ . - where only global data was available statistical - ru. me VW= e Wmmmrn Mm correlatin mthoes vere usd. Different modela were tested throvgh correlatime with feeders which losses- were tvluac*t by radial load flov. rom this analy- Il1eral Becalibracion sis the patrmeters of the selected mdelc were detcr- amind uad used for the systaw vhere the global data- the energy eterst illegally deemalibrated were de w" available. ctarinad by two methods: by inspection and by statis4l cal analysis. Sy Is"pection the meters that presented deficiencies in mtering one: or two phaes were noted- as such. lesides. other moters which presented stat The aney losses tm a utilty are were and mere tically unexpected high deealibration ware classified- importat a th eosts of producing enrg and as th- as to produce losses due to illegal decalibration. price of nstUlln$ new geerating capacity are acra The brek point between natural and il1ea gldecalibra- sing. Dle to this fact. it Is of mjor portance for tio. was established to be in the range comprised bet- tho electrical utilities to masse h amount of losses ven -10S and -23Z of deeaibration depending on th- and to kow, witb enougb accuracy. where an in what utility studietd The losses evaluation for illcgjl de aountthehf losses fre produced in order to tae corrfe calibratd terd ist rs do in the same manner as for - e tsssoi- naturally decolibrat4d meters. The results obtai.ed - loss of rvenuew. This study presetts a complete m are 0.9Z of the available energy for the CaLambian eye thodology for the assessent of loses according to cm. The range on a per utility bais vs Aster -iae ete described cassifi" tion. The state ctimation - to he betwee 0.1 ead 2.1S techniques with mually loggd data appers to be the mt appropriate tbhod Sm order to obtain good resuiw Damaed Meter witb fairly accurate informaion and with a mininm of masuremat effort La the field. The mthods e*ployed A damaged m_ter was considered as oe that don for distribution losses evaluation basd on a radial - met measure any energy because of blocekd rotor fstr lo flow gave accurate results but as the a*mut of - ample. The approach to determine the losses for tb - data to be processed at thin voltage level is very - electrical utilicy was to consider that In averape be high, it is Imperative to work an a ample of feedere- meter in damaged in the middle of the period comptised and extrapolate th results to the rest of the power - betwen two consecueive readings of tbe m_ter. As nor 5y5tem. aesly a blocked meter Is only det*eted by two identiea readinp it vas establised that the losses for the ean be *mjor contribution of thLi paper in the at - pany correspond to the energy consumd *m ta _me pa tewet to classify sad to evaluate tho %lack" leases - rind comprised between one and a half periods of meter aS presented. From the rs ults obtained for the Colon readings. Bamed an thi censidarstion the aunz of - bim slatrs it is concluded that 2J3 of tbe losses ar lusses were estimted to be 0.6Z of the evailable ener du to "physical" losses ad 1/3 to "blackW losess. Vy for the Colombimn system with a maximam of i. Sa The analysis of possible corrective actions als per - a minima of 0.22 on a per utillcy b d. formd as part of the study. reveled that the evalua ed loses could be diminished considerably with mdedr lncorrectlv tstimtaed Cnsuamtion to econoic lnvestments and high potential be fits - for the utilities theo rttse are in accordnce with During thi study it was determined that s of recently published papers CS3 C63* C73. C83 ad thrE. t utility regitered cstonera sre billed for a Usi fore highlight the imortance of this type of study - ad amount of enrgy due to their low consuWption (us - for other wulties. an pemaently with this billing approach) or due to special conditions like temporar service, change of- A ED0DIEM -erg meter, etc. For these custoera a statiatical- mthod was designed considering the geographical area* fr the cotious uapot ldturconxing al thpases So. - the installed load of the customer and cite consumption for the cinuous cupprt during all the ph"ea of of customers with similar charactersc ice but which we this study and all utilities in Colombia for their ae- re billed with energy meters. The energy consumption tive participation. of a "typical" customer per area compriszng around 500 -248- Page 7 of 8 .~~~~~ Sf .IC. Steww et .al. *?w.r yeti astatsta ,iAstian", * Fares t. t ad Ittm ESrans. to Ver Ap. Sys. Vol. AS49, Jlan. 1310. 2 1. DurS. %mgae. u as_u- ants mae ass tet ata- te eutiatien opeimL ad pearar. tt. IEs &W* Nor Ptwern "Stig. July 157T. 3 LJ.Cln. C.x. Cagy, cftes mLutnI cf Coron 3anses Wdar Main$ birc Vcal imsevyp rings da" eheftkng of a ustbed to 9"O"ag Ceoona Lsete", InE Tra"sactiona, Vel. t 49. go. 516 pp. 853- u0o HRayJURS 1970. n.zzS i t &.. Scalclaten of fte*v Losses a* Dstrisebus, System". pwsnmad to the PA 81e Matins, July 3379. U r, lSa _ yuaam eeoLea laneiLt fum Loea indudwin Iaue" Tr.asmiason and Distribu- * _, Vol. 34 M. 7. Jul 182. X-IJ.143a014h, 'Lead beaLsciag radia se ystem dstdWbCufm lsa.a", Talos and Dts:ribu- !!M., mL. 35. No. 7. JuLy 1881. 1 LJ. as C v, OyF.e tdv S L yes L . Tra,l,", sLoI an" Distribution, Tol 34 Ms. 7, J4u7y -.32. 8 3j. CLngte. oapandy C__n dL of timy 70"Gre get Le vidctlena, Transmission and UistierAbue VoL. 34 No. 7, JuLy .aJ92. .~~~~~~~~~~~~~~~~ AndLx A - 249- Page 8 of 8 COWlMBIE rD A N < . ~~~~~~~~~, ma*bkok..407C.d . mm-.? ; .&m _ _ _ _ _ _ _*_ _ * - m m m . CI e ve, u /. - - 250 - A STATISTICAL APPROACH FOR EVALUATING SOME NON-TECHNICAL LOSSES IN POWER SYSTEMS By Mr. Angel Zannier Head Electricity Program Latin American Energy Organization (OLADE) ABSTRACT A statistical method to measure those non-technical losses which are caused by alteratio s in watthour meters is presented. The methodology is suitable to determine: (i) the number of fraudulent customers as well as the number of non-fraudulent customers per consumer category; (ii) the amount of non- billed energy as a consequence of alterations on meters, per consumer category; and (iii) a fair billing system, so as to recover losses due to detected infringements. This methodology is based on sampling techniques and statistical theory, has been originally developed by Dr. Jose Luis Calabrese and tested at the Empresa de Energia Electrica de Bogota in Colombia 1/, with promising potential for further use in other power utilities in Latin America and the Caribbean. The paper is oriented to power engineers and therefore does not concentrate in mathematical demonstrations or pure statistical analyses. 1. INTRODUCTION It is well known that methodologies to measure, evaluate, reduce, and control losses in power systems, mainly concentrate on technical losses. In practice however and taking into account the Latin American and Caribbean context, it is very common to find power utilities in which non-technical losses are higher than those of technical origin. The present paper is an attempt to establish a suitable method for measure those non-technical losses arising from alterations in watthour meters, and where an exhaustive inspection of meters is not feasible due to costs and time constraints. In those cases, one is normally guided by rough estimates of loss figures and the usual approach is in accordance with rules of thumb. One practice that is quite common in most utilities in the region is that of deriving non-technical loss figures as a subtraction of calculated technical losses from total measured losses. Although that methodology is correct in principle, it may hide several procedural errors, the most obvious of them being that of using different time periods among the measured produced energy and the corresponding billed energy, to derive total losses. Another METHODOLOGIES DE EVALUACION DE PERDIDAS NO TECNICAS. Jose Luis Calabrese, OLADE 1988. - 251 - source of uncertainty refers to the methods used to calculate technical losses. They may range from state estimation techniques to simple Ohm's law application in distribution feeders, including the well known load flow algorithms. Although those techniques may be quite sophisticated from the computational standpoint, the data to be used is normally error prone and engineering judgement is commonly used to assume "representative* active and reactive power demands in the nodes of the corresponding networks. The calculation of non-technical losaes as the difference of two not very reliable figures, gives therefore a doubtful result. It is well known that in economic and engineering disciplines, sometimes rough estimates are sufficient to make decisions. In the reduction of losses in power systems, however, a good knowledge of the absolute loss figures and a disaggregate of them may help management and decision makers in their efforts to reduce losses in the most cost effective way. This fact, highlights the urgent need to iaprove measurement technique for technical and non technical losses. One approach that may result promising is that of using representative sampling techniques and statistical inference to the uniwrerse. The use of this approach is gaining impetus worldwide, as a result of recent improvements in computer technology and their corresponding costs reductions. In Latin America and the Caribbean a significant portion of non- technical losses has been attributed to altered watthour meters at all consumer category levels. These losses not only adversely affect the financial performance of the utilities, but also distort the composition of future demand. In fact, in utilities where prices (tariffs) do reflect costs, those non fraudulent consumers may be paying more than they should, affecting in consequence their future levels of consumption, given that in the long run, price elasticities are not zero. 2. BASIC DEFINITIONS Before embarking on a description of the proposed methodology, it may be worthwhile to discuss some basic defLnitLons and assumptions which will be used throughout the paper. 2.1 Billing Factor A billing factor may be defined as: BF -Eb/(T * PL) (1) where WEbW represents the billed energy durlng the perlod WTO, to a consumer whose installed active power is *1i. The billing factor may be interpreted as the relation of the actual bllled energy to the theoretical maximu possible billed energy for a given consumer In a given period of time TO. Using the above definition, and considering two customers with - 252 - similar electricity consumption behavior, that is, with the same installed power "Pil and consuming the same amount of energy WEel during period NT", but assuming that one has a fraudulent watthour meter, their corresponding billing factors may be calculated as: BEf - Ebf/(C * Pi) (2) BFn - Ebn/1,T * Pi) (3) where Of" represents for the fraudulent customer and NnO the non-fraudulent one. Since it is assumed that both customers consume the same quantity of energy "Ec" during period *T", it may be inferred that: Ebn > Ebf (4) and therefore: BFn > BFf (5) Expression (5) means that given two customers with the same electricity consumption behavior, the billing factor of the non fraudulent one is always greater than that corresponding to the fraudulent customer. This conclusion should be borne in mind as it will be utilized in statistical terms for customers of the same consumer category, with the same electricity consumption behavior. 2.2 Relationship Between NBFN and PiL It is worth noting, at this point, the differences between the billing factor, defined in (1), and the commonly used load factor of expression (6). LF - Ec / (T* Pm) (6) In expression (6), wEcw represents the energy consumed during period OT", by a customer (load) with a maximum demand of "Pa". The billed energy 'Eb" used in definitions (1), may be equal or less than "Ecw, depending on whether the customer is non fraudulent or fraudulent, respectively. Installed power *Pi" is normally greater than maximum demand 'Pal and in some cases may be equal. The commonly used load factor may be interpreted as the portion of energy that is actually consumed in relation to the energy that may be consumed with a constant maximum demand lasting for the whole period 'TN. The billing factor, may be interpreted as the portion of energy that is actually billed in relation to the energy that could have been billed if the customer would hatre used its installed power during period 'TO. - 253 - Normally, the load factor is used as an inherent indicator of the load ln the sense that it measures how the maximum demand is utilized. Similarly, the billing factor may be used as an inherent indicator of the customer, because it measures how the installed power is used. From expressions (2) and (3) it may also be concluded that there is no correlation between the billing factor WBFN and the installed power "Pi". In fact, two customers with the same installed power WPi", consuming the same amount of energy TEc" during period T" may have lifferent billing factors MBDf' and "BFn", depending on the alteration or not in their watthour meters respectively. The last remarks, in statistical terms mean that the covarience between the billlng factor and the installed capacity is nil. Mathematically: Cov (BF,Pi) - 0 (7) Expression (7) will be used in making statistical inferences on the basis of a sample and extended to the universe in a given consumer class. 2.3 Correcting Detected Fraudulent Energy Bills Considering the hypothetical case of the two consumers with the same electrical behavior, and considering that one of them is fraudulent and the other is not, expresslon (5) has been derived. From expressions (2), (3) and (4), one obtains: Ebn - BFn * T * Pi (8) Ebg - BFf * T * Pi < Ebn (9) Having assumed that both customers consume the sase quantity of electricity Ecw, and considering that the non-fraudulent customer has been correctly billed, the non-billed energy for the fraudulent customer may be calculated as: NBE - Ebn -Ebf -(BFn - DFn -BFf) * T * Pi (10) or: NBE - d *T * Pi (11) where d - BFn - BFf Is the difference between billing factors for the non fraudulent and fraudulent consumers respectively. Conceptually, it is therefore possible to calculate the unbilled energy of a fraudulent consumer, knowing its - 254 - installed power 'Pi', its billing factor wBFf and the billing factor of an equivalent customer. This principle shall be expanded in statistical terms to the universe. 3. METHODOLOGICAL FORMULATION 3.1 Methodological Approach The first step for the formulation of the present approach for determining non technical losses due to altered watthour meters, is the division of customers ln consumer classes. Disaggregated information of customers must be pursued as much as possible. An initial approach may be that of utillzing consumer categories which are established by the utilities. It is highly advisable to disaggregate customers within each consumer category, so as to classify them in consumer classes. Having classified the customers into consumer classes, a sampling process must be carried out within each class. The sample components must be chosen at random, and its size can be initially determined using well known statistical formula to do so. One commonly used formula 2/ is that of expression (12): n -pq N (12) (a/2)-2 (N-1) + pq (2 where n - sample size p - probability of occurrence of the phenomenon in question q - 1-P N - population size (size of the consumer class) a - sampling error given by the losses evaluator (for instance 0.05 for a confidence interval of 95%) In the present formulation, since the sampling process may have two posslble outcomes, that is altered or unaltered meters probabilities p and q are both O.S. Error waO may be lnterpreted as the probability of those cases for whlch the sample does not represent the population (consumer class). For instance an error a - 0.05 would mean that In 5S of cases, the calculated sample does not represent the unlverse (populatlon class). Table 1 lists dlfferent sample sLzes for dlfferent confidence lntervals, for three different sizes of consumer classes. 2/ Household Energy Consumptlon in Rlo de Janeiro Shanty Towns. Alfredo Behrens, International Development Research Center, June 1988. - 255 - ZTblje1: SAMPLE SIZES FOR DIFFERENT CONFIDENCE INTERVALS WITH THREE DIFFRENT SIZES OF CONSUMER CLASSES P _ 0.5 AND Q - 0.5 Confidence Size of the Universe (N) for each a Interval Consumer Class n n n 0.01 99% 5.000 909 99 0.02 98% 2,000 714 96 0.03 97% 1,000 527 92 0.04 96% 588 385 86 0.05 95% 385 286 80 0.06 94% 270 218 74 0.07 93% 200 170 67 0.08 92% 154 135 61 0.09 91% 122 110 55 0.10 90% 99 91 50 It is noteworthy that the greater the consumer class, the smaller the relative size of the sample for a given confidence interval. Thus for instance, for a given confidence interval of 97% the sample would be of 1,000 customers for a universe of 10,000 (10% of the class), 527 customers for a universe of 1,000 (more than 50% of the class), and 92 customers for a universe of 100 (more than 90% of the class). It Is also important to bear in mind that the greater the required confidence interval, the greater the size of the sample. Once the sample size has been defined and customers belonging to that sample are selected at random, a site survey should be conducted in order to analyze their electrical Installations. First a careful inspection of watthour meters should be carried out. From it the number of fraudulent and non- fraudulent customers may be determined, and their proportion may be extended to the consumer class. The survey must be concluded by an installed active power calculation for each surveyed customer. For that purpose, nameplates of electrical machines and appliances may be utilized if possible. If that Information is lacking, measurements may be done machine by machine, using preferably wattmeters or else clamp ampmters and voltmeters and estimated power factors. The later may however introduce errors, and should thus be avoided. Having the results of the site survey, and records of electricity consumption, by customers billing factors may be calculated for fraudulent and non fraudulent consumers. The sample may be subdivided into two sub-classes, namely fraudulent and non-fraudulent. The mathematical expectation of real energy consumption, for the non-fraudulent sub-class, may be calculated as: - 256 - ErRCnI - T * E[BFnl * E[Pin] (13) and for the fraudulent sub-class It can be assured that: E[RCfE 2 T * E(BFfJ * E[Pif] (14) It is important to note at this point, that expressions (13) and (14) are implicitly assuming that there is no correlation between the billing factors and the installed power. This fact has been discussed previously when analyzing two consumers wlth the same electricity consumption behavior with its corresponding mathematical conclusion as represented in expression (7). Should there exist correlation between the statistical variables OBFW and 'PLO expression (13) and (14) should take into account a further term with the covariance between those variables. Observing expressions (13) and (14) it can be said that they are statistical extensions of expressions (8) and (9). The expected real consumption for the fraudulent consumers, may be calculated using the mathematical expectation of the billing factor corresponding to the non-fraudulent consumers, as: E[RCfJ - T * E(BFnj * E[Pif] (15) Expression (15) is a statistical extension of the concept inherent in expression (10). In fact, in expression (10) the real consumption of a fraudulent consumer has been calculated using the billing factor of a non- fraudulent consumer, with the same electrlcity consumption behavior. In expression (15), the same procedure has been adopted, but this time in statistical terms. Subtracting expression (14) from expression (15), the mathematical expectation of the unbilled energy due to watthour meter alterations, within the sample, may be calculated as: E[NBE - T * E[PifI * ( E[BFnl - E[BFf] ) (16) - 257 - On replacing mathematical expectations with calculated mean values, the mean unbilled energy may be calculated as: M[NBE] - T * MNPif) * ( M[BFnj - M[BFf)) (17) The amount of electricity not billed within the consumer class under consideration, thus may be calculated as: TNBE - N * Nf * T * M[PifJ * (MLBFn) - M[BFfD) (18) where: N - Number of customers in the consumer class under consideration (universe). Nf - Proportion of fraudulent customers, within the consumer class under consideration. It is equal to the proportion of fraudulent consumers within the sample. That is FN - nf/n, being wnfl the number of fraudulent c-nsumers in the sample of Onw consumers. Results of expression (18) may be used to guide management and decision makers on evaluating the convenience or inconvenience in pursuing further searches of altered meters in each of the defined consumer classes. In fact, by valuating TNBE in economic terms and considering the costs of an exhaustive search for each consumer class, a cost benefit analyses can be conducted, so as to determine a merit order of future actions in accordance with descending benefit/cost ratios. In fact, those consumer classes with high benefit/cost ratio should be surveyed with high priority, leaving those with smaller values as lower priorities, within the actions to reduce non-technical losses. Summarizing what has been proposed so far, the steps to pursue are as follows: (1) Definition of consumer classes, to the highest disaggregation level possible. (2) Definition of sample sizes per consumer class. (3) Conduct of site surveys to detect fraudulent consumers and to calculate the installed active power per consumer. (4) Calculation of consumers with fraudulent and non- fraudulent watthour meters, per consumer class extending the proportions of the sample to the universe. (5) Calculation of billing factors for fraudulent and non-fraudulent consumers. (6) Calculation of mean values for billing factors and installed power. - 258 - (7) Calculation of the total non billed energy (TNBE), resulting from alterations in watthour meters. (8) Cost benefit analyses to prioritize actions for the reduction and control of fraudulent, non-technical losses. 3.2 Fair Billing of Fraudulent Consumption In order to formulate a methodology suitable for the fair billing of fraudulent consumption it is convenient to mathematically adjust the statistical distribution of billing factors. In the study conOcdted by Dr. Jose Luis Calabrese at Empresa de Energia Electrica de Bogota, he found *Gamma functions* were most appropriately adjustable to the statisticel distributions of billing factors, Gamma functions have the following mathematical expression: f(BF) . AAr (BF)^(r-1) * e^(-A*BF) (19) J (r) where A > 0 is a constant that may be calculated knowing the mean value "M[BF]" and the variance V[BF] of the billing factor distribution. r > 0 is a constant that can also be calculated knowing the mean value and the variance of the distribution. e Base of the Neperian logarithms. J(r) A function of constant TMr", which may be calculated as: J(r) - | t^(r^l) dt (20) Constants 'Al and "rM may be calculated solving the following system of equations: H[BFJ - r/A (21) V[BFI - r/A^2 259 - The solution of the above system of equations is: A - M.[BFJ / V[BF] (22) r - (N[BF]^2) / V[BFI Figure 1 shows a typical shape of a Gama distribution function partially skewed to the left. It is important to note at this point, that depending on the relative values of constants "Al and TMr, the Gamma function can change its skewness from the left to the right, or even to become a normal distribution. In practice however, it would be suspected that skewness would be placed to the left in those consumer classes with high installed power but with low consumption. The procedure for flnding a system of fair billing to infringing customers, consists in adjusting the statistical distribution functions for the billing factors of fraudulent and non-fraudulent sub-classes, as illustrated in Figure 2. There it can be seen that the fraudulent customers billing factor, distribution function is shifted to the left with respect to those of non fraudulent customers. This fict is in accordance with expression (5), in which it has been demonstrated that the billing factor for a non fraudulent customer is always greater than that corresponding to a fraudulent one, with similar electricity consumption behavior. Having the billing factors distribution function, and assuming that a fraudulent customer has been detected by a site survey, and using his records of electricity consumption and his calculated installed active power, his particular billlng factor BF1 can be calculated. Using BF1, from the distribution function of fraudulent consumers a distribution value of fl may be calculated, as illustrated in figure 2. fl represents the portion of fraudulent consumers from the consumer class under consideration that statistically have the same electricity consumption behavior and therefore have the same billire factor BF1. fl may also be interpreted as the portion of non fraudulent customers that statistically have the same electricity consumption behavior with a billing factor of BF2. The non billed energy for the detected fraudulent consumer, may therefore be calculated as: NBE - T * PI * (BF2 - BFI) (23) or: NBE - T * Pi * dl (24) where dl - BF2 - BF1 may also be obtained from a function d(BF) as illustrated in Figure 3. Function d (BF) may be calculated mathematically as: d(BF) - f(BFn) - f(BFf) (25) - 260 - In prectice, however, it would be highly convenient to perform the calculations using microcomputers and st-ndard spreadsheet or data base programs. Although the described methodology may be interpreted as a statistical approximatLon, it is conceptually more appropriate than those commonly applied. In the present approach the fraudulent consumer is compared to a statistically equivalent non fraudulent consumer. On the traditional approach it would have been compared to a hypothetical one with arbitrarily selected values of load factors, coincidence factors, etc. The Zair billing methodology, recently proposed, may be summarized as follows: (1) The statistical distribution functions of the difference in billing factors may be calculated using data from the site surveys. In mathematical terms, see expressions (19), (22) end (25). (2) After conducting the cost benefit analyses described in sectlon 3.1, proceed with exhaustive site surveys in those consumer classes that justify it. (3) Past fraudulent consumption of particular customers may be calculated using expression (24). (4) Special tariffs and fines to discourage further fraudulent consumption and watthour meter alterations may be applied. 3.3 Sample Size AdJustment To verify the best fit of billing factor experimental values to a Gamma function, it may be convenLent to proceed with the "Kolomogorov - Smirnov" test. For that purpose, tgae cumulative distribution function of billing factors may be calculated as: BF F(BF) - f f(BF) dSF (26) 0 Absolute deviations between experimental and fLtted values may then be calculated as: D(BF) - ADS[F'(BF) - F(BF)J (27) The maximum acceptable value for D[BFJ should be 0.27 allowlng an error of 1%. Statistical tables may be consulted so as to flnd the maximum D(BF) value compatible with the desired error. If the resultlng error is greater than required, theoretLcallv a new sampling process shotId be. carried out. For that purpose the sample *ize should be enlarged. - 261 - 4. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 4.1 Summary A methodology has been proposed to statistically determine: (i) The number of fraudulent customers as well as non fraudulent customers, by consumer categories. (ii) The amount of non-billed energy as a consequence of alterations on watthour meters, per consumer categories. (iii) A fair billing system, so as to recover losses due to detected infringements. The proposed methodology may be summarized as follows: (1) Definition of consumer classes. (2) Definition of sample sizes per consumer class. (3) Conduct of site surveys. (4) Calculation of number of consumers with fraudulent and non fraudulent watthour meters, per consumer class. (5) Calculation of billing factors for fraudulent and non fraudulent consumers. (6) Calculation of mean values for billing factors and installed power. (7) Calculation of the total non-billed energy (TNBE), resulting from alterations in watthour meters. (8) Cost benefit analyses to prioritize actions for the reduction and control of fraudulent, non-technical losses. (9) Calculation of the statistical distribution functions for the difference in billing factors. (10) Proceed with exhaustive site surveys of those consumer classes where justified. (11) Calculation of historical electricity consumption for each detected fraudulent customers. (12) Application of special tariffs and f 'es to discourage further fraudulent consumption. - 262 - 4.2 Conclusions and Recommendations From what has been discussed, the following conclusions may be drawn: (1) There ls a need to improve measurement techniques for technical and non technical losses in a reliable and disaggregated manner so as to guide management and decision makers in their efforts to reduce losses in the most cost effective way. (2) Representative sampling techniques and statistical inference theory, may be effective in detecting, measuring and disaggregating fraudulent non technical losses. (3) The proposed methodology has been tested at the Empresa de Energia 6lectlica de Bogota. Further analyses and tests, in other utilities of the region should be pursued. IL id a~~~~~~ .5 r0 4 In tD Id 1 "4 ow ~~~~~~~~~~~~~~~~~~~~~~~~~~~Id a~~~ II.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0 .4~~~~~~~~~~~~~~~~~~~ tur 4 \ - 2u4 - Fig 2 ; --F I I I I I I I I I I I I I I I I I I BUF, 3I?z SF po.reeult. 7z - 265 - IL Po . _~~I 4~~ /' on a -II . * - * . I . * . 5 _ - 266 - CORRECTIVE NEASURES FOR NON-TECHNICAL LOSSES By Mr. Willy G. Pacheco Bolivian Power Company Ltd. ABSTRACT Results of work started a year ago by Compania Boliviana de Energia Electrica S.A. - Bolivian Power Compamy Ltd. on reduction of non-technical losses in the city of La Paz, Bolivia and its surrounding areas are presentod. The purpose of this exercise being to contribute to other electric utilities in Latin America and the Caribbean with details of experiences gained in this important and often neglected discipline. Case studies of the various activities carried out thus far are given and results obtained are summarized. Cost/benefit analyses are included and recommendations as to courses of action to be taken by utilities about to engage in similar work are provided. I. INTRODUCTION Compania Boliviana de Energia Electrica S.A. - Bolivian Power Company Ltd. (The company) is one of the few remaining private utilities in Latin America. It is owned by shareholders of which approximately 70% reside in the United States and 30% in Canada. It has two divisions, the La Paz Division and the Oruro Division. The La Paz Division generates and distributes electric power to the City of La Paz and surrounding towns and the Oruro Division supplies electric power to mines operated by an agency of the Bolivian Government (Corporation Minera de Bolivia) and sells wholesale power to the Company's subsidiary, Empresa de Luz Electrica de Oruro, S.A. - ELF "ihich distributes electricity to the city of Oruro. The La Paz Division operates under a non-exclusive franchise dated October 20, 1950 granted by the Municipality of La Paz, which extends to September 1990. The Oruro Division operates under Specific Regulatory Agency Resolution of August 30, 1968. ELF operates under a contract with the Oruro Municipality which extends indefinitely and can be terminated with two years' notice by either party. At present the Company is negotiating a renewal or extension of its franchise with the Bolivian Government and the City of La Paz. Electricity rates are determined by the Direccion Nacional de Electricidad - DINE, a government entity, whose main functions are to regulate, coordinate and promote the development of the electric industry. - 267 - The latest average electricity rates established by DINE for the Company are in the order of US$0.036 per kWh distributed as follows: Domestic 2.94 US cents per kWh Comercial 5.47 U N Industrial 2.97 N Street Lighting 3.18 Rural Electrification 1.94 ' Present and past tariff policies have in general pretended to establish subsidized tariffs to a great majority of the lower income consumers without foreseeing the source of the subsidy. Low tariffs have undoubtedly affected the distribution system expansion of the city of La Paz, particularly in high growth areas in the outskirts of the city where electricity theft has risen. As a result total losses in the city of La Paz system have increased from 12.63% in 1981 to a high of 19.33% recorded in 1987. This paper will describe the loss reduction efforts started by the Company in March 1988 relating to the reduction of only non-technical losses. Technical losses in the complete generation, transmission and distribution system are also being the subject of a detailed study but results will not be available for another year. St: is estimated thbt technical losses in the La Paz system are approximately 9 to 10 percent of net generation. _I. NON-TECHNICAL LOSSES - REVIEW In order to establish a guideline of what is to follow a brief reveiw of the definition of mnon-technical losses' and a description of all of its components will first be given. Non-technical losses represent energy consumed for which a power utility does not receive revenues. It consists of all losses absorbed by the utility during its commercial operation, from the time energy is consumed, followed by accurate billing, until full payment is received. Basically tLen, there are three major components that form part of non-technical losses. These are: Consumption losses - Related to the accuracy with which the utility records the electrLcity consumed. Billing losses - Related to how accurate the billing process is. Collection losses - Related to how much of the billed consumption is actually collected. Consu=mtion Losses These can be split in two categories: - 268 - (a) Non-measured consumption: Refers to consumption which is neither measured nor recorded in the customer-file by the utility and could come from various sources such as illegal connections, inaccurate estimation of consumption, and delays in the installation of metering equipment. Illegal connections in developing countries, no matter how well administered the utilities may be, are likely to represent the largest component of non-technical losses. Customers connected, without the company' s knowledge, consume more energy than necessary. Illegal connections are commonly referred to as electricity theft. Inaccurate estimation of consumption of those customers connected directly to the distribution system, because of lack of meters, also represent high losses. This is particularly the case in countries that have to import energy meters because these are not manufactured locally and where their purchase depends on the availability of foreign exchange. Finally delays in the installation of metering equipment on those customers temporarily connected will add to losses as energy consumed until the meter is installed may be incorrectly estimated or not estimated at all. (b) Measured consumption. Which is measured consumption but not recorded with complete accuracy. In this category the following cases are included: Fraudulent tampering of meters, defective meters, improper hook-up to the distribution system, improper installation of measuring equipment, meters not registered in the customer's file and, management system discrepancies. Due to management system discrepancies large sources of error occur when the meter reading process is inadequate and does not allow for accurate recording of consumption measured by the meter. Errors in meter constants are common and often reading errors are made because of non-standardization of meters used. Eilltng Loss Billing losses are usually linked to two basic phenomena: Inaccurate information in the customer file and discrepancies in the billing process. Inaccurate inforamtion ln the customer file creates losses if: the inf-Ymation contained in the customers' contract is missing or inaccurate in the bil±ing record, rates applied are Incompatible with the service characteristics, equipment installed is different to the one registered and inaccurate notation of changes in the customer dwelling are made which will cause difficulties in the reading of meters and delivery of bills. Discrepancies In the billing provess us ally exist if the system does not follow up on non-billed customers or billing does not occur for several - 269 - periods. Also if customers who benefit from special rates or grants are not controlled. Likewise if billing irregularities or anomalies detected in the billing process are not investigated. Discrepancies in the billing process als: arise if bills are corrected without proof or control, if the process cannot ensure that all recorded consumption is billed and the system does not allow for periodic processing of all customer accounts. Collection Losses Two elements form part of these losses, unpaid bills and inefficient management of payments. The main reasons for unpaid bills are bills not delivered to the customer, customer inability to pay the bill, and inadequate collection procedures used by the utility. Inefficient managemenet of payments usually result in theft of money by utility employees, loss of revenue due to timing between billing and collection (particularly true in countries with high index of inflation) and inappropriate credit of accounts due to incorrectly identified customers. Specific details for each case of the three major components of non- technical losses described above are well known and will not be described further. It is important, however, to emphasize that non-technical losses are important to deal with as they have a direct impact on the finances of a utility. Unlike technical losses, non-technical losses are not inevitable, and very often great improvements can be achieved in this area without significant investment of capital. Reduction of non-technical losses is basically a matter of good management. Their target level should be zero. 1II. SYSTEM CHARACTERISTICS General The generation, transmission and distribution system of the La Paz Division forms part of what is known as the Northern Electrical System of Bolivia. This system is connected to the National Interconnected Grid operated by Empresa Nacional de Electricidad - ENDE. The interconnected grid now serves six of the nine Departments of the country. Electricity for the La Paz division is generated in eight hydro- electric plants located in Zongo Valley and one in the city of La Paz. Total installed firm generation capacity at present is 113.9 KW>. Three transmission lines operated at 115 kV and one at 69 kV come out of the Valley feeding a 69 kV ring-circuit that surrounds the city to which 13 distribution substationssn are connected. The transmission network consists of approximately 430 Iks of line and also extends to other neighboring towns. Primary distribution voltage in La Paz is 6.9 kV, 50 Hz. About 455 kDs of primary lines and 1,020 kbs of secondary lines form part of the distribution network. The secondary voltage is 230/115 volts, although as of 1984 the secondary voltage has been standardized to 220 volts. - 270 - Power transformer installed capacity in distribution substations is 152 NVA and distribution transformers installed capacity in primary lines is 240 MVA. With a peak load of 141.1-1 MVA (127.5 MW) the distribution transformer ratio is approximately 1.70. Generation Load Factor during 1988 was 54.6%. Consumer and Demand Growth The La Paz Division has at present approximately 150,000 consumers which according to existing rate structures can be classified in five categories. Table No. 1 lists the five categories and includes data on consumers per category and the distribution of total sales per category for the period April 1988 - March 1989. Figure No. 1 shows the correspoding percentage of total sales for each group. Table No. 1 Type No. of Customers Consumption (MWh)* Residential 130,080 269,632 Commercial 19,215 92,632 Industrial 628 85,437 Rural Electrification ** 2 9,721 Others *** 187 16,265 Total 150,112 473,687 * 12 month cumulative (March 1988-March 1989) ** Rural Electrification sales are made to two cooperatives whA hi serve approximately 15,000 consumers. *** Others include small towns, street lighting and exports to the country of Peru. - 271 - DISTRIBUTION OF SALES Commercial 20% Domestic 57% Others 3% Rural Elect. 2% Industrial 15% Figure No. 1 Cconsumer growth had an average Increase of 4.14% over the last ten years. Details are plotted in Figure No.2. Consumer growth has been Influenced to a large degree by the country's economy which has had ups and downs characterized by a big recession that started in 1981 and recovered during 1985-86. During 1988 consumer growth has increased sharply to over 6%. Annual kWh consumption per consumer in the domestic category of the city of La Paz is plotted in Figure No. 3. Over the lat decade it has been on average only 1,872 kWh per year, which is equivalent to 156 kWh per month - among the lowest, if not the lowest in Latin America. - 272 - Energy Sales for the period 1979-1988 are shown in Figure No. 4. The reduction In sales recorded in 1982 and 1983 were likewise caused by the economic recession. By the year 1985 cumulative inflation had exceeded 14,000%! Energ Losges - La Paz Diviso The devaluation of the Bolivian Peso has had direct effect on losses In the La Paz DivisLon system. Restrictions imposed at the time by the government on availability of foreign exchange limited the Company to provide the necessary resources to attend new requests for service and to carry on with the normal system expansion. As a result consumers in areas near the outskirts of the city were forced to make their own extensions with inadequate and low cost materials which undoubtedly increased technical losses. Furthermore the social unrest that was created by this condition has had a marked effect on the morality of the low income population making them more aggressive. As a result illegal connections and meter tampering have increased considerably. Figure No. 5 thows the increase in total losses in the La Paz system during the period 1979-1M88. From 1981 when recorded losses were 12.6% these have risen gradually to over 19% by the end of 1987. The rather chaotic economic situation during this period had also obliged the Company to provide direct (unmetered) services to a large number of %onsumers in need of service. Lack of meters due to dificulties in obtaining the required foreign exchange to import them was partly the cause. By the end of March 1988 the Company had approximately 8,500 consumers connected directly. In Figure No. 6 month by month variation of 12 month cumulative losses are plotted from 1986 onward to show the sharp increase in losses. Transmission, distribution and total losses are shown separately for the analysis that is to follow. -273- 150 -F- - - W* - -_-= FiQaurw No. 2 ia.a1nun (JISnMumPlrnt - zDhmestiri 220O 2000 - - - ~~~.00- - - ~~~1 ~ 1 I~~~C IL.7 1951a 1985lq 1987 FiQure No. 3 - 274 - ?Erwergj i,is - 2 0Dnz 3iEusn rn 480 11 1 1 460 440 420- - ______ tI ~ I I- _____ 1979 1981 1983 1985 1987 1980 1982 19854 1986 1988 Figure No. 4 Anrgy-~ ~iIosst - - Dta 3ia, Biuisiatn 22- D-ljI- - i i i-, 12 I I r. VHriT I jI 10 1 a 1979 1981 * ° 983 19tg4 198 , ¶s 987 18 Figure Po. 5 - 275 - iEurrg i!tssrPS - i!ta paz ttso (12 &dtl Ounmkliu) 18 1 TITt 614 10 4 Ti I I. I- I I I I. I M I I I I1T S N J M M J S N J M M J S N J M 86 1987 1988 89 Figure No. 6 - 276 - IV. LOSS REJUCTION PROGRAM General The large increase in losses registered starting towards the end of 1986 induced the company to form a special department with exclusive dedication to the control of losses and their reduction. The specific functions of this sectlon were to: - Identify major sources of losses. - Establish a loss reduction program. - Coordinate with technical divisions of the Company on all loss reduction measures. - Start a follow-up program on loss control to keep track of improvements made and maintain activities proven effective and economically sound. The loss reduction unit or Departmento de Control de Perdidas" started work during March 1988. After a year of operation 12 month cumulative losses were reduced from 19.33% to 17.62% with savings estimated in excess of US$350,000. Present estimates show that two more percentage points can be reduced in non-technical losses by the end of the next twelve month period. Identification of Souces of Losses MTe first stop needed before preparing the loss reduction program was to identify the major sources of losses in the La Paz Division system. This required a thorough study of the two main components of losses -- technical and non-technical losses. (a) Technical Losses As mentioned earlier technical losses are being the subject of a separate study which is still being carried out. Basically, what is being done is the following: - The city has been divided into a grid of mosaics 500 x 500 meters each. - For each mosaic a complete physical inventory of all electrical equipment installed within its boundries is being prepared. - Each mosaic in turn is being subdivided into a grid 20 x 20 meters. - All poles structures, primary and/or secondary lines that interconnect them, and distribution transformers are being included. - 277 - - Thio information will form part of a computer data base for which special software is being prepared. - - The ultimate aim being to identify each consumer to a given pole structure and thus be able to carry out transformer management studies, technical loss studies, voltage drop calculations, etc. This work has been under way for several months and completion is estimated to be by the end of 1989. b. Nion-Technical Losses Because of its various components the study and loss reduction efforts related to non-technical losses are being carried out jointly by several departments of the Company under the control of the loss reduction unit. Case studies to determine the magnitude of each component were first made to establish priorities and prepare a loss reduction program on non-technical losses under which the loss reduction unit is now operating. The importance of having clear, concise and periodic information on losses and other pertinent information on system production, consumer growth and sales should be emphasized, as this is basic information needed to carry out any work on loss reduction. A sample of the Company's monthly statistical report is given on Figure No. 7. All main parameters are given. 12 month cumulative data on g:3neration and sales is included. The latter are useful in determining trends and making comparative analysis because they reduce the effect of seasonal variations, as well as, the time factor involved in the readings of energy meters which because they are not coincidental can introduce appreciable error if only short periods are considered. For the particular case of the La Paz Division system all distribution substation loads are metered on the low voltage side of the power transformers. Transmission losses are calculated by subtracting the sum of all these metered energy from net generation. Transmission losses, therefore, incldue losses in power transformers. It is important to point out also that any defective meters or errors in the readings of these particular meters reflect directly on transmission losses because of the way these are calculated. Distribution losses in the statistlcal report are calculated by subtracting total energy sales from net generation minus transmission losses. - 278 - PANCUh NIA UULIVIAIJA JE EIILIttUA ELEC1BIlCA S.A. *OLIVIAte roPEcn comrAz v LtuAu i 0IV111011- LA Pitt O1ONT11LY STATiSTICAL REPORT FOR HAflCII * 19 99 MEGAWAT, T 1OU1 SlAtiOUl ;YS1EEM COIIICIDEIIIAL FEAR: _ _oss SERVICE VAIE- 31 min: 19. 3 MOUTYILY YEAn 1o0nAE jZM lIS-CUM h1Wh mW MVA __"___c__ _ 1_ J460.20 4.043.20 _17,032.25 9.35 __2L 50q_ ..3.-. - tt"so , ~~~7.46.40) -1*034.60 . .9#331.9( 15l.70. -- _400- 52Cr.46 . sutac4 2 (1 5 7.20 ~§~2Q6400 S*617 7 3_.64_ ............_3.006_...... 9 _7 W,j -nt 1793.90 5''3-.1 2 ...70- -2 cMuiaN*@u3 .,12261.60 .34.668.00 .91.710.23 -.....................................5 98. --23.0 c___ _ ^ 13.494.60 36o583.20 121,566.40 1 7.53_ A ______2___ 9-_I 11ROSS piton a.LQP 169 .a4Z,40 jJ4§561 40**4' 10.0 1,, ........6 .20 ...2S7 ..666 MMUMisgifiC7E iiiiiiiEE-Q2 - 025O JULM ___2_77___S/_ SlNo i@SimpoAI 1. 0480 _.oy_ 57 .00 4g7 . I - I . 17 00 19.72 I tanosn tUOpgt 11,004.00 28,164 .00 45.107.40 ,.4; _ NETV -IMSUO 112574W906 1-77 7W§7 ~ ______ ScYStEm t .1_589.7 _ J61.45t .u .DL1sL*_ .. .......... .. ...OL.. 'I w!0401_6sLOSSES 2,330.44 _9Q7._?7 ..0.qO27...94 `. to!aa.t 327-.50 OQSttIARCtI/89 P31313 3 LOSSES .....I-.-M 9 . .48*6-82 aL8,7flL30 1 _ _ _.______'_ I____________ ____________ .vSOU LqE . GONICUSI f1it Ad 1S CUWS TOTAL SALNS 39,317.10 119.027?.t 47..Q2L-j . 112 . . -- ¶IS. lertstoc- 42 64Q-gQ 2, 702.17 .-e,624_63 -10 .400 e heoftC3S ,...A 4j194..65 _152U1.S .61.932.70 _ .3,060 0cutsfe-_ , _ 15, 303.21 47-."1.02 194.299.57 66.344 '_-___'i__ Ss -~3 71.53 i .I68.64 lee05618 25.375 TOMt L0UsC_ 21.200.18 _67 ,083.21 269,632.40 .130.0600 * ' 2.073 GE(**LSUS, L%-C 3,563.95 11,022.30 42,370.4i ...?.o12 2 .2379 toft ,1.U439U 1 2.u 0.9S 50.26174 1,403 --_ 35.02S , u 587:8,sr6 ~ 1-.;69.31 .b;762.26 =-408 ____23 "oust"s^tt^Z*et.t 45061. 37 o i .p 0ss 5 S6 ,473.09 o.. 217 - . 260. 24Q spcciat cwte!*Cts9 * _2,450.0I .-*095.09 .22.101.00 ... . 3 - 7.. .393 ,_613 sftitaE II IN 1208. 34 _3Y~3 1.76.91 .1z. 26.49 ~~60.96 - 214.67 _180 19, *WSL i it"1f;6§ 4 .2i9i.i5 9.721.05 _ 2 4.860.525 szon,z^,^"tzzzog tSz.B2 435.83 ls1Ot1.70 #, _300 _ ,,4 709 *^s76 s-4.. 83 njSa -;Wh3S...I.. -.63.27 - ......3.77:. _. .. *. E S , g o s_____. __, _ ._..__ _._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ £ojIISzzSUEIJO/SOI 47!If If5.aE3.. ,- ..2025. ______.____, !MICI?A aCOvt _ 0l-__ C lt Uff I//I//I//b, o8stO 11' * SOla VIA I NU * ws UUUICIP*LBF;_ __161.46 ,, . 62 7//lU//U 1 R 0 . -- 2.67 I S5. 51 j __________*_, 1,192.22 700 567LLLLLLL . '^_j .561 ..s 728.O I _____ ,,(,OCg ° ,,_, . , _t ? 3 .r24 i,.89 LA PAZ EA'kg DIU ..in0. at 3.244 i 2.P:eL... ;;tt_ _ _ _ _ _- . .. ....-"- 2 00... .... _. _ ,; ;I@ .....7.. in.: 127.50 *'&141.11 tear - .. 14.189 623.04 .. _ ____ . "7**-Xt^t _ tt. osQ _. . Figure N~o. 7 - 279 - CASE STUDIES To provide guidelines and prepare an effective loss reduction program the Loss Reduction Unit began work by carrying out individual case studies to establish the incidence on losses of: (a) Direct connections (consumers connected to the network by the Company but without meters). (b) Illegal connections (consumers connected to the network without the Company's knowledge). (c) Metering errors (either because of defective meters or improperly made installations). (d) Consumer billing (to check for accuracy of information in consumer files). a. Direct Connections As mentioned in Chapter III the number of direct connections to consumers due to lack of meters had increased to over 8,500 by the end of March 1988. This number represented close to 6% of the total number of consumers in the La Paz Division. Even though energy consumed by these customers was calculated and billed based on installed capacity on each consumer premise, it is a known fact that unmeterd customers tend to consume on the average more energy than they are billed for. To check this fact a case study was made on Palca Town, a small rural village near the city of La Paz that, out of a total of 191 consumers, had 154 direct connections. Sales to this town are being controlled by metering installed on the main feeder. Table No. 2 shows details on sales, energy distributed and losses for the last 12 mouths. Note the effect on losses of two actions taken by the Company: (a) increase in kWhrs billed to each consumer connected directly, and (b) the installation of meters. Before July 1988, when losses exceeded 40%, direct connections were being billed 50 kWhrs per month per consumer. Average use of energy per consumer in other rural towns is only 32 kWhrs per month. Starting in July the decision* wa taken to raise to 80 kWhrs the monthly consumption billed to each consumer that was connected without a meter. Apparent losses which were extremely high due to incorrect estimatLon of consumptlon were reduced after this action to about 11%. - 280 - Table No. 2: PALCA TOWN SALES AND LOSSES Consumers Sales Energy Month With meter Without kWh Supplied kWh Losses April '88 37 154 11627 19600 40.7 May 37 154 11229 18520 40.0 June 36 154 11743 21840 46.2 July 34 159 16069 18000 10.7 Aug 34 161 17732 20000 11.3 Sept. 188 10 10859 10680 -1.7 Oct 191 5 9955 10968 9.2 Nov 193 3 11432 12360 7.5 Dec 190 1 7933 8840 10.3 Jan. '89 190 1 13095 14000 6.5 Feb 179 1 7994 8600 7.1 March 176 1 8863 9600 7.9 Starting in August 1988 meters were installed on all direct connections. Results of these two actions are shown in Figure No. 8 where sales and losses are plotted (3 month cumulative readings). The drop of energy supplied is worthy of notice. In June 1988 average monthly consumption in this town (with 80% of the consumers with direct connection) was about 115 kWhrs per consumer. After meter were installed average consumption dropped to only 54 kWhrs and losses have dropped to below 8%. Based on this experience all direct connections are now having meters installed. By April 1989 direct connections have been reduced from 8,500 to less than 200 and work continues to eliminate them completely. b. Illegal Connections The detection of illegal connections is more complex as additional metering usually needs to be installed to control losses and this work requires close supervision of the specific area belng chocked. The city of La Paz being situated in a narrow valley of the Andes Mountaiks has limited room for urban expansion. This fact has obliged urban planrers to start construction of high-rise apartment buildings and a large portion of the population now dwells in them. This evidence was seen as an opportunity to start a program to detect illegal connections at low cost by installing metering equipment at the service entrence of such buildings and keeping close control on individual consumer meteis. - 281 - "also cown ipan1ss (3 AUrntli Mazmuiniiue) 50 45- 4 ~~40 " A m J i A S0 0 N D F 1988 - 1989 tzLL- =OL211 ~1OSZ (3 3anrU QauzmzLut) 50 - 40 - i J - - I M . J S °N D 1988N --1989 Fi gur No. N - 282 - Basically this program is being carried out as follows: - Supply to a building is cut-off while control metering is installed to it. - All consumer meters are read at this time and special seals are installed to prevent tampering while tests are made. - Building supply is normalized and after a period of 15 days all meters are re-read after cutting-off once more the supply. - Based on results obtained individual services are checked, illegal connections (ii found) are suspended and action is taken to prevent recurrence. - A second similar control is made after about two months and results compared. This program which was started in September 1988 has proven very effective in 63tecting illegal connections. About 27 buildings have thus far been checzed and losses in them have been reduced from an average of 5.5% measured daring the first control to 3.15% measured during the second control. It ws estimated that there are over 150 high-rise buildings now in the city. The program will continue until all of them are completely checked. c. Metering Errors Measured consumption bu_ not recorded with complete accuracy represent a large component of non-technical losses and have always been of great concern to the Company. Errors are usually due to: - Defective meters. - Lack of limited calibration of meters. - Improper installation of measuring equipment. - Incorrect application of meter constants on major co sumers metered on the high voltage side. - The existence of a vast variety of meters (some with four, others with five digits). etc. One of the major tasks assigned to the Loss Reduction Unit has been the revision and subsequent improvement of all these actions in direct coordination with the Metering and Commercial Departments. Essentially the program started with a rigorous control on meter insta.lations of all major consumers. Control that included: meter testing, - 283 - check-up on multiplying constants, and verifying if the load connected agreed with tariffs being applied to each consumer. Meters installed In generation plants and distribution substations were tested during hay - July 1988. In some cases errors in excess of 2% were measured (as against + 0.05% which ls the hlghest pormsosible error for these meters according to company standards). Meters wlth high error were found to correspond to those with omany years of service which, due to aging, are difficult to calibrate and have to either be replaced or sent to servicing for a complete overhaul. Approximately 23% of the meters of all major consumers of the La Paz Division have thus far been checked. Savings due to this action have been estimated to surpass 9.7 GWhra in sales since the beginning of this program. d. Congumer Billing The introduction of all meter records in the main frame computer with which all blling is carried out has helpisd sort out inconsistencies in the consumer records. This work had started at the end of 1987 and was completed in July 1988. All meter reading is done on a monthly basis by company employees. Reading routes have now been clearly identified and the loss reduction program has included in its activitie3 the proper training of personnel to carry out this job. Billing frequency is also monthly. Intercalary bills are not emitted. Estimates of consumption are only made in those cases where the meter could not be read becau se the hovae was closed in which case a bill is sent based on this stimate and later adjiwted when a reading can be made. Delivery o' bills is carried out by a Contractor who also is in charge of disconnecting services that are ln arrears. The second consecutive unpaid bill automatically places the consumer in the arrears list. Inconslstencies between noter and consumer records were found while introducing information in the computer. This has helped detect many cases of Incorrect billing, as well as, losses. Also meter changes Improperly registered have caused incorrect application of multiplying constants. The implementation of a now fully computerized system for all commercial operations of the Company has contributed an well to the efficeincy of collection, which at present is only about 45 days from the time the bill ls emitted. Futur Work Based on experlence gained during the first year of operatlon of the Loss Reduction Unit, work will continue on the following fronts: Direct connection will be totally eliLmnated. - 284 - Once all hif -rise buildings are checked for illegal connections this work will continue on sectors of the city where low income people live. It is estimated that a very large percentage of non-technical losses correspond to these areas where uncontrolled low voltage line extensions have taken place. Work has already been started to replace these privately owned extensions. Meter testing will continue with particular emphasis on the replacement of outdated metering equipment and the implementation of a revised program on meter calibration and statistical inspection of recently acquired units. Control of non-billed accounts will be started. This is a subject that has received little attention thus far but, nevertheless, needs action. Coordination of activities between the Loss Reduction Unit and the various departments inside the organizational structure of the Company will be improved. This is a must to Improve efficiency in any loss reduction work. Personnel working in the Loss Reduction Unit will be increased to improve its operations. Efforts will be continued to receive legal support from the Government as at present there are no national laws that deal specifically with the theft of electricity in Bolivia. V. RECOMMENDATIONS The reduction of non-technical losses should be a major objective for all electrical utilities, particularly those that have detected high losses, as these have a direct impact on their finances. To attain a significant reduction of non-technical losses first priority must be given to the optimization of the commercial infrastructure of the utility, even before any loss reduction program is begun. Organizational aspects of the utility's commercial process are essential in the identification and then the control of these losses. It is indispensable that covmunications and interrelations inside the organizational structure be excellent and well maintained. There must be frequent contact and discussions between the different units responsible for meter installation, the readings of meters, the billing, rate revisions, collection and all other units that have direct contact with the consumer. Most utilities today have found that it is justiflable and advisable to create inside their organization a unit, with complete autonomy and freedom of action, whose only job is specifically to detect and control both technical and non-technical losses. - 285 - The Loss Reduction Unit formed by Ccmpania Boliviana de Energia Electrica S.A. - Bolivian Power Company Ltd. only a year ago has started a loss reduction program based o., specific case studies. Results thus far are encouraging as n'n-technical losses have been reduced by approximately two percentage points. The unit formed cons dts of four well trained technicians and one engineer who in direct coordination with other company departments (which have used no more than twenty other employers) have made this program successful. For those utilities about to engage in similar work regarding the detection and reduction of non-technical losses the following hints and recommendations might prove valuable. In order of importance the steps to be followed are: - Make sure that statistical information on generation (if any), sales, and losses is reliable and presented in an easy to read schedule. - Testing and verification of all important meters used on the calculation of energy delivered to the distribution system should come next. This will ensure that recorded losses are correct and any follow-up action on loss reduction is properly made. - Start meter testing on major consumers where the likelyhood of large improvements in losses is greatest. - Revise estimated unmetered consumption. If direct connections exist efforts should be made to have meters installed in them. = Verify the correctness of 11 consumer files making sure that all basic information dealing with the customers themselves, their location, installation and metering characteristics, conditions of sales, and the service agreed to by them for billing purposes is correct. - Verify that all consumer meters are being read and bills are emitted in timely fashion. - Only then start with a program related to the detection of illegal consumers, meter tampering, and electricity theft. For those interested in loss reduction efforts being carried out at present in other Latin American countries a list of articles that were presented in a Symposium on Control of Losses held in Bogota, Colombia in October 1988 is included in the Bibllography section (Chapter VII items 4 through 12). Reference to other recent articles (items 13 through 16) to be presented during August 1989 by the Subcommittee of Distribution of Electrical Energy of CIER (Comision de Integracion Electrica Regional) in Cochabamba, Bolivia are also included. - 286 - V;. COST/BENEFIT ANALYSIS FOR NON-TECHNICAL LOSS REDUCTION PROGRAHS1/ One of the main reasons for the existence of unacceptably high losses in distribution systems of developing countries has been the decline in financial position of power utilities because governments have prevailed in establishing low electricity tariffs which have led to reduced investment and system maintenance and indirectly stimulated consumption. Frequently electric power utilities are used ab instruments of economic policy to fight inflation. Scarcity of I reign exchange resources have, in the majority of castis, also limited system expansion. Loss reduction projects should not exist if distribution networks would have been adequately planned and received the necessary financial resources. These projects are proof that past policy errors have forced utilities to postpone investments for network rehabilitation and maintenance. High energy costs and problems in the balance of payments of developing nations have increased interest in energy conservation and reduction of waste in electricity consumption. Loss reduction projects represent an important way of energy conservation and result attractive in developing countries that posses mature electric systems because of their low cost relative to benefits that are gained. An important rule to remember: Losses become excessive when it is cheaper to reduce th:m rather than increasing the capacity of the supply. Each kWh lost represents to the country the long marginal cost of supply. Thereby it is worth investing in its reduction only if the cost of its reduction is less than the marginal cost. The purpose of a cost/benefit analysis is to prove that the benefits of loss reduction are more, or at least equal to, other sacrificed benefits in the economy when assigning resources to the project. Cost/benefit analysis provides a better way of assigning resources than traditional methods of cost optimization by allowing, first, to compare loss reduction projects with other projects on generation, transmission and distribution; and second, by providing means of ranking projects on electricity distribution with other projects In the country in need of the sometimes limited resources available. 1/ Abstract of an early paper presented at the Symposium of Control of Losses held in Bogota, Colombia on October 1988 by its author Mr. Luis E. Gutierrez (an English version of which is being presented elsewhere in this semi.usr). - 287 The two types of louses- technical and non-technical, require a somewhat di!ferent treatment as cost/benefit analysLs is applied to projects, not programs. Technical losses, on one hand, are primarily economic losses. They are considered to be losses to the national economy and do not just adversely affect the financial health of the utility Itself. The energy that is lost as heat in tho transmission and distribution system could have satisfied adeitional (incremental) demand or load. If loises were absent there would be no need to use additional scarce national resources to supply the incremental demand. By reducing technical losses there is less need for power generation and the requirements for transmission and distributlon are reduced. Tochnical losses can be reduced by using better designs, selection and localization of new substations, installation of capacitors, addition or reinforcement of primary and secondary lines, better combination of distribution transformers, etc. Non-technical losses, on the other hand, are primarily financial losses to the utility. These can be reduced by controlling theft, meter reading errors, and billing and collection improvements. Revenues lost impose a heavy burden on the financial viability. Reducing non-technical losses is one of the best and most cost-effective options to improve the financial positinn of utilities by providing more revenues. Non-technical losses also affect the economy by distorting the optimal pattern of electricity consumptt.on. Those who do not pay usually consume more. Those who do pay consume less if the higher cost of the non-payment by others is passed un to them through higher tariffs. When projects for reduction of both techutical and non-technical losses are considered for the same distribution network it is co-ivenient to first analyze the later and then, assuming that non-technical losses have been eliminated, technical losses should be examined. This is necessary so as to not overestiamte banefits of technical loss reduction projects as by reducing non- technical losses. In what follows a description of cost/bonefit analysis for onily non- technical loss reduction projects will be given. For cost/benefit analysis of technical losses please refer to references listed in the bibliography. trActical ConsLderations As described earlier non-techrical loss reduction projects aim at reducing: (a) electricity theft in its various forms, and (b) non-registered use of electricity. One of the main economic benefits of these projects correspond to savings in resources of generation of energy consumed free", that once detected will stop being produced. However, since not all unbilled onergy will stop being generated (as part of it will continue to be supplied - the difference being that with the project it will now oe billed). one must subtract the reduction of consumer surplus resulting from the increase in price. The utility of course will obtain important financial gains by selling energy otherwise lost without the projects. - 289 - A second economic benefit results from the increase in demand of existing consumers because of tariff reductlon, which will (or should) follow, due to improvements in the financial situation of the utility. The net benefit of non-technical inss reduction projects can be summarized in the following equality: [Net Benefit] - (Savings in Resources] - (Reduction of Consumer Surplus] - [Project Costs] (Eq. 1) To fully understand the components of this equation some explanations follow: Consumers that receive more energy than they pay (either because of electricity theft or meter malfunction) have a benefit equivalent to what they would be willing to pay for the energy they now consume free. We shall term this energy as WTP. (where the subscript wow stands for without a project). When the project is started this benefit becomes a cost to the project; the loss of consumer surplus. Similarly the cost associated to this free energy becomes a savings to the project as it will be shown later. The net benefit (NB0) to that economy of continuing with the unbilled free energy can be expressed as: NB, - WTP. - mc R, where Inc* is the marginal cost of supply and R, is the unbilled energy (non- technical lo0ses). The purpose of the project being to reduce the unpaid energy R. The beneflts of the porject Bp are given by: NB, - Bp - NBp NB, - em [Rp - R3J - [WTPp - WTPpJ - Cp which is the same expression as equatlon (1). The first term [Savings in Resources] corresponds then to savings due to the reduction of non-technical losses. It is estimated from the reduction in electrielty use whlch ln turn is calculated from known values of marginal cost (mc) and unbllled consumptLon (R). The level of consumption of consumers that become OnormalLzed" R,, if not readlly available, has to be calculated from demand curves. 289 - Unit Price to M~~ R - UR R. kWh Figure 9 Figure 9 shows a simplified demand curve where for a tariff tp there is a demand Rp. Rp in the graph would correspond to the increased demand of illegal consumers that don't pay (t0-O). In this particular case the demand curve can be represented as: t - a a/R, J D where a I t (*-l)1/* t - marginal tariff for these type of consumers (O < t < a) R. - illegal consumption D - demand at price t * - elasticity price of the demand (dD/dt) (t/D) For those cases where tampered meters are found that only register a fraction of the consumption, or where a defective meter was found, where part of the metered energy (D.) is pald at the existing tariff (tp) while the rest (D.) is not paid. The relevant tariff tp corresponds to an average price (paid over the total. consumption). In both cases, the relevant tariff is weighted for the fraction of the consumption paid and, therefore, hi3her than zero (t* > 0). tp - (tpD^) / (Dx + Dj) te tv (I/D.) The equation for the demand in this case can be expressed as: twr - aDb D where b -(1/*) (t/D) - 290 - Once the demand curve, is estimated, the sumption of "normalized" consumers is calculated, substituting "t for the project tariff tp., and then solving for the unknown quantity Rp. The [Savings in Resources] component can then be calculated as the product of the marginal cost of replacing the connections mc and the reduction of consumption (R. - Rp). The second component of equation 1, [Reduction of consumer surplus], (WTP0 - WTPp) will also vary depending on the type of unpaid consumption. If it is totally free, or paid 'n part. If the tariff is nill (tS - 0) the loss of consumer surplus is given by: E - 1/2 tp [R. - Rp When the tariff is more than zero (O < to < tp), the loss is calculated from: E - 1/2 (tp - to [R. - Rp] The last component of equation 1, [Project costs], should include all costs relevant to the project. These can include equipment, systems, and procedures such as: calibration equipment, control meters, service drops, computers for billing and collection, measuring instruments, training and supervision of personnel, design and implementation of procedures to penalize illegal consumers and improve collection, as well as, billing techniques. Conclusions The implementation of non-technical loss reduction projects is best done by analyzing separately each type of consumer. It is more rewarding to start with industrial consumers who normally consume more energy and thus provide greater improvement In any loss reduction effort. Likewise consumers with higher income levels should be dealt with first. A set of recommendations has been provided in Chapter V for those utilities about to engage in similar work. These were based on experiences gained by Compania Boliviana de Energia Electrica S.A. - Bolivian Power Company Limited during the first year of operation of a Loss Reduction Unit formed to carry out this work. Results thus far have been highly rewarding. VII. BIBLIOGRAPHY Papers presented at the Regional Seminar on Reducing Electric Power System Losses in Africa, held in Abidjan, Ivory Coast in November 1987. - 291 - 1. Aubin Raynald, Daoust Denis. Hydro-Quebec International, Montreal, Canada. "Non-technical Losses". 2. Hay Winston. , Industry and Energy Department, The World Bank. "Sources of Losses". 3. Munasinghe Mohan, Boroumand Jahangir. The World Bank. "Policy Framework and Economics of Electricity Loss Reduction". Papers presented at the Symposium on Control of Losses, held in Bogota, Colombia in October 1988. 4. Yuraszec T. Jose. CHILECTRA. 5. Perich Edmundo. ENELEVEN. "Experiencia de Chilectra Netropolitana S.A. en el Control del Hyrto de Energia". 6. Cespedes R., Duran H., Hernandez H., Rodriquez A., "Assessment of Electrical Energy Losses in the Colombia Power System". IENEE Transactions on Power Apparatus and Systems, Vol. PAS-192, No. 11, November 1983. 7. Cespedes Renato. Universidad Nacional de Colombia. "Perdidas en Sistemas Electricos - Clasificacion Y Definiciones". 8. Empresas Publicas de Medellin. "Medidas Remediales para el Control y Recuperacion de Perdidas Negras en el Sistema EE.PP.M. en el Periodo 1987- 1990" . 9. Posada Anibal. Interconexion Electrica S.A. - ISA y Lega Armando, Financiera Electrica Nacional, S.A. - FEN. "Perdidas en Sistemas Electricos - Desarrollo de de la Problematica en Colombia". 10. Empresa de Energia Electrica de Bogota. "ProgramA de Reduccion de Perdidas - Periodo 1987-1992 (Documento Informativo)0. 11. Gutierrez Luis. Banco Interamericano de Desarrollo. "Criterios y Procedimentos para el Analisis Economico de los Proyectos de Reduccion de Pordidas". 12. Nunasinghe Mohan. The World Bank. -Eeonomic Principles and Policy Electricity - Loss Reduction". Papers to be presented by the Subcommittee of Distribution of Electrical Energy of CIER (Comision de Integracion Electrica Nacional) in Cochabamba, Bolivia during August 1989. 13. Campusano Gabriel, Donem Guillermo. Empresa Electrica de Antofagasta. 'Estrategia utilizada en la ciudad de Calama en la Reduccion de Perdidas de Energia". - 292 - 14. Rocabado Orlando. Luz y Fuurza Electgrica Cochabama S.A. M., vAspectos Comercialos. 15. Santos Denisar. Comite Coordenador de Operacoes Norte-Nordeste - CCON. vFraude e/or Desvio de Energia Eletrica, Experiencia da Companhia Energetica de Pernambuco - CELPE". 16. Di Jora Custodio, Moreira Luis. Light - Servicos de Electricidad S.A.. *Prevencao e Represao a Furtos de Energia". - 293 - THE LOSS REDUCTION EXPERIENCE OF THE BARBADOS LIGHT AND POWER COMPANY LIMITED OVER THE PERIOD 1964 - 1988 By Mr. C.L. Franklin - C. Eng., M.I.E.R.E. Chief Distribution Engineer The Barbados Light and Power Company Limited In order to be assured that we are all thinking in like terms, it may be best to start with . short description of the utility concerned. The Barbados Light and Power Company is the utility which is responsible for the generation, transmission and distribution of all electric power in the Island of Barbados. The island has an area of 431 sq.km. - 166 sq.mls. for those who prefer it that way - and a population of approx. two hundred and fifty three thousand (253,000) people. The whole island is now electrified, that is to say, electricity is available to anyone no matter where he may live on the island. There are two (2) generating sites approx. 5 miles apart connected to a common grid system. All generation is at 11,000 volts. The major portion of the power is stepped-up and transmitted at 24,000 volts to 10 substations at the various load centres in the island. At these substations power is stepped-down to 11,000 volts again, for distribution via some 28 primary feeder circuits. There are 3 distribution primary feeders at 24,000 volts. Distribution is achieved via transformers to provide services at 230/115 volts single phase in the rural and sub-urban areas, and 200/115 volts network in the urban areas. One of the reasons for stating the above is that there are some utilities - though not in the Caribbean I dare say - who buy their power in bulk, and are happily unconcerned with the effects of generation on their system, provided it is adequate. Losses in terms of our utility is defined as the difference in the metered quantity of units generated less the quantity of units used for generation (station services) and the metered quantity of units sold. This may be more aptly put as net generation - i.e. that energy which leaves the 11KV generator bus/bars for transmission and distribution - and the cumulative quantity of metered energy sold. System losses are generally expressed as a percentage of net energy generated. Having defined losses as net energy generated minus quantity of sold energy, we may now look at the reasons for losses in an electric utility system; determine how these losses are generated; and explain how we at B.L. & P. endeavour to keep these losses low. - 294 - Losses may conveniently be divided into two categories - Technical losses and Non-technical losses. TECHNICAL LOSSES Technical losses are due to the laws of physics, and can be thus accounted for. Engineering-wise these are therefore measurable and controllable. I am afraid that technical losses will never be completely reduced to zero - Superconductivity or no Superconductivity. The criteria for technical losses may be catigorized under the following sub-headings. TRANSMISSION Voltage Conductor sizing and spacing SUBSTATIONS Siting Layout PRIMARY FEEDERS Lengths Voltage Conductor sizing P.F. Correction Regulation DISTRIBUTION TRANSFORMERS Small transformers and short secondaries Versus Large transformers and long secondaries Consideration of loss analysis in bid evaluation SECONDARY LINES Conductor sizing Lengths SERVICE DROPS Conductor sizing Lengths METERS Accuracy In the theoretical sense we are all very familiar with the underlying precepts governing the above. Most have to do wlth Ohms Law, I2R, RHO-L/A, reactive power etc., all covered In the basic principles of electricity which have been crammed into us at one time or another. In the practical design of a viable electric utility system however it becomes slightly more complicated as ponderance has to be given to such things as optimization, standardisation, least cost, quality of service to consumers, affordability, and a whole host of other economic and sociological factors. - 295 - This brings us every opportunity to talk briefly about the other category of losses posited - NON-TECHNICAL LOSSES These are in the main those losses which are related to the logistics of management; the efficacy of the management policies deployed; and the efficiency with which these are implemented. In other words, these losses are usually subjected to the vagaries of human failings and the systems which are put in place to control the same. Some sub-headings here may be as follows:- Meter reading Service Inspections Exception Reports Contraband or theft This last - theft - is possibly the most direct form of loss a utility suffers. It is extremely difficult to detect at times, depending on the ingenuity of the thief. Many a saga can be written based on the romance - 'Meter Inspector versus Contraband". You will note that I have only addressed those non-technical losses which will affect the metered quantity of energy sold, i.e. through inefficiency of inspectors, meter readers - or through deliberate energy diversion. Losses due to administrative problems and poor business principles I consider outside the ambit of my terms of reference. In point of fact I doubt that I will be able to concentrate very much in this paper on non-technical losses, interesting though that subject is, except maybe during question time. I have a feeling that the deliberations of this seminar are predominantly on the control of technical losses. It is in the control of technical losses that the engineer - whether he be planning, transmission or distribution makes his greatest contribution to an efficient electric utility system. To keep within the confines oi my terms of reference, I will put the question for you. "What do we at B.L. & P. do to be in the happy situation where the 'losses" on our system are the lowest in the Caribbean, and compare verv favourably with utilities in some of the more developed countries? Let us look at the chart of 'losses' plotted against net generation, starting around 1964 (Fig.l). This was a time of very high load growth - greater than 15% per annum compounding. You will note the high level of losses endured at the time. I joined the Company in early 1965, and I remember well. The installed capacity was 20MW, with a peak load of 14MW. There was 8MW of generation at 11KV and the remainder was at 3.3KV. There was still a certain amount of 3.3KV distribution. The programme at the time was to move away from 3.3KV distribution and to convert completely to 11KV distribution. With this higher voltape level, ampacities were reduced for the same power requirements, and we achieved a constant reduction in losses as this programme progressed. All generation additions were now made at a busbar voltage of 11,000 volts. - 296 - Fig 1 THE BARE'DOS LIGHT . POWER CO LID LI-NE LOSS-ES 1964 - 198 SP.GDN UNITS 18.2 - 4.4 MW *Dd SP.(. DN UNITS 8&9 -9.2 MW 0i) SP.GI)N UNITS 3,4&S -6.6 MW ( dd CENTRAL 24KV SUBSTATION. 0 ST.THOMAS 24KV SUBSTATION ®D SP.GDN UNITS S2 -20 MW D OLD WORKS 24KV SUBSTATION ( dd SP.GDN UNITS S1 -20 mw SP.GDN UNITS 6L7 -9.2 MW Qd SP.5DN UNITS 10&11 -25 MW 0 INACCURATE METER ON AUXLIARIES SP.GDN UNITS 12 -12.5 MW ->~~~~~~~~~ 11¢,N I - r JIIII§t6 !.iJ~~~ - l I 1i - . - a uH j: !_ _ : *!1 1 IIIZ'Ir|............... - !1i ib I I l~~~~~~~~~~~~~A r 4 rvI|||i%i A§\|/ l3^1_|^ z - =-a I I- - f-- 7iN:F1 : I , , . T= F 1 ' T _ l _ _ _ _ i~~~~~~~IC 114 _ iNo G _N_t_L N 5 bd 14 =4 0V 46 0% IF 7= 7= _1- 7= 70 an _- fb M d 7 ._ __ _ l[- _FF!YR I Lub * a ,,- I - - - - I2 IL.. ..AL. - - P _~~ 1 -...im,.- 1 I =a a= *.__ -r 1 j a 1 1 .1 a -_ .s . >* s7wu7s7a waat7 e7 BX6 l^ a *- - -f-Fl; a1R - 297 - I must admit that during those years little effort was put into Loss Reduction per se. The price of oil was low compared with post 1974 prices. The only attempt at Loss Reduction was indirect, as it pertained primarily to the maintenance of voltage levels along the line, and at customer loads. I speak now specifically to the installation of capacitors. As the system expanded to the northern and eastern parishes, a higher voltage was required, and in the mid 60's the first 11KV pole-mounted capacitors were installed in St. Peter to obtain better voltage regulation. Progressively more capacitance was installed on all 11KV feeders as voltage levels dictated. The primary reasons for capacitor installations are:- 1. Reduction of 12R losses 2. Improved Voltage Regulation 3. Recovery of system capacity resulting in deferred capital expenditure for expansion. 4. Releasing circuit capacity for application of additional load. Over the years B.L. & P. have maintained a planned programme of installing capacitor banks on our system. At the present time, there are some 20MVar of capacitance installed. In mid 1979, our first substation was built, located somewhat to the West of the centre of the island. It was fed by 9KM of 336MCM aluminum line at 24KV, stepped down to 11KV to feed surrounding areas. One year later, a similar substation was built at Old Works, to cater for the load growth in the southern areas. Looking at our chart (Fig. 1) we note the dramatic drop in our losses - some 2-1/2 percentage points. In system design, siting of substations cannot be done willy-nilly. Substations must be located as near as possible to the load centres to be most effective, for obvious reasons. The Substation is also a very practical position for installing regulation facilities on the system. All of our substations are equipped with automatic voltage regulators. Between 1970 and 1971 we see a departure from the usual trend, and our losses appear to increase. This turns out to be a false picture and is due to a defective meter on our auxiliaries. This made the net generation to appear higher, and a subsequent high loss relationship. Meter accuracy you see is very important. This will be addressed more fully in a later section. Generation is added as load growth demands, and in 1972 a third substation at Central was built fed by two lines - approx. 7.5KH of 336MCM aluminum O/H wire. Our losses went down to the impressive low of less than 9%. Previous to this in 1969, a system expansion study was done and was based on forecast load growth. It was recommended that 69KV transmission lines be introduced, and some of the 1LKV distribution feeders be converted to 24KV operation. The higher voltage levels would increase line capacities and reduce line losses. A few of our distribution feeders have been converted to 24KV, but because the load growth has been lower than that which was forecast, the 69KV transmission project was deferred. - 298 - So far we have noted the certain influences which impact on losses resulting from load growth: - Increase in generated voltage; transmitting into load centres at higher voltage levels; siting of substations; and regulation on primary feeders. We have also alluded to the effect of installing capacitors on distribution feeders. It would serve little purpose to give a dated historical account of the growth of the Barbados System. I will just acquaint you of our current position (Fig. 2). We presently have an installed capacity of 132 MW, a peak demand of some 79MW. We have a 24KV grid system linking 10 substations from each of which power is distributed at 11KV or 24KV. Our present losses are in the vicinity of 8% of our net generation. Let us now look at some of the other aspects of distribution engineering which affect losses. Conductor sizing is an obvious one. Over the years approx. 20% of my routine capital expenditure was on an i :em we call "strengthening" which is in the main reconductoring of existing H.T. & L.T. mains, always to a larger conductor size. This year we budget to spend some $3.8 million on this item. To date we have standardized on 336 MCM for all otur 3 phase mains and 1/0 Al. for H.T. radials and L.T. mains. This is an uprate from the No. 3 copper and the No. 2 Al. which was previously prevalent. The greater part of our transmission lines is built at 795 MCM Al. or 500 MCM copper. Our cable for service drops has also been uprated. Whereas as recerX as 3 years ago the average service drop to a house was in No. 10 concentric copper. We have now standardized on a No. 8 duplex copper cable. Let us now look at what I tend to consider the most important item in the distribution transformation set-up - the distribution transformer. Now each distribution transformer can be considered in the same light as a substation in our transmission system with respect to losses. I know that there are some systems in the Caribbean which use large 3-phase transformers and long secondaries - with all the attendant losses both in the transformer and along the lines. We opt for several small transformers with short secondaries. I remember visiting one area in the United States and it appeared to me that a transformer was utilized to serve at the most two dwellings. We haven't got to that situation in Barbados yet, because the load in the average home is too small - but we are getting there. It may be of interest too to note the type of distribution systems we use in Barbados - Rural and Sub-urban domestic loads are fed by single phase H.T. distribution. A radial of one H.T. phase is run into a village - for instance - with a continuous neutral wire which is earthed at multiple points along the line. Single phase distribution transformers are then installed rated at 6.35KV primary 230/115 volts secondary. We find this a very convenient and more economical method of supplying rural domestic loads. Before I leave distribution transformers, I wlll touch briefly on losses inside the transformer. Thls depends to a great extent on the design of the transformer, materials used etc. Some transformers have higher per unit losses than others. To try to control this, what we do is to use the loss analysis quoted by the manufacturer in his tender as one of the criteria in the bid evaluations. - 299 - SUBSTATINS@ FUTURE EXISflNG ST. LUCY IsN NORTH .' ST. PEER !\ JA.ES sr. JOSE * ST. THOMAS S r. JOHN 7 I a-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I !~~~~~ I;-- - ,z. S f r MC) IELXST. GEORGE SPRINC ST. FHILIP PILE *'ARO Doi- . ' X : ,-U R C H "; H 24OOO VOLTS 'tANSKtSiN SYSDt -~~~~~T.f .- . .i. c3SSl t er s ) > (Loss Data, characteristics) | programs Line Flows, Voltage drops, etc.) Figure 1: ELECTRIC TECHNICAL LOSS MODELING Figure 1 shows an illustration of electric loss modeling. The network characteristics are collected and processed by the analytical programs. Loss data, line flows, voltage drop and levels are some of the outputs of the analysis. Database creation can be accomplished manually by typing the data directly into the computer terminal or automatically through digitizers and customers' billing records in integrated software packages. Real time data acquisition systems like Power-Donut (developed by NITECH) can collect, digitize and transmit transmission and distribution line operating parameters, including load, to a remote station. This station can either process and store the information, or retransmit it through telephone lines to an analysis center. The data files so created are used in desired loss analysis using standard electrical science equations and numerical methods. The equations are iterated until a pre-selected convergence factor is achieved. The analysis performed includes load flows, switching, load balancing among phases, capacitor size and location optimization, economic loading of conductors and transformers, short circuit, reconductoring, and transformer evaluation. 1. Load Flows an-d Losses: A network supplied by two or more substations has different ways of supplying the loads connected to it. Line sections provide equalizing means through which excess power is transported to different loads. For example: - 305 - Consider the equivalent circuit of a hypothetical network (Figure 2) with node curent IA, Voltage VA and admittance Y. 'A IB A B 0-> AB V V A B .AB _AB 2 2 D C Figure 2 : Line Flow Network 1. ;MgCu Dn (a) At node A ,A i (VA - VB) YAB + VA ( 2 (b) On section AB ,AB - (VA - VB) YAB (2) (c) On section AD IAD - VA YA1B (3) 2 2. yglta&t (a) At node A A VAB VB (4) - 306 - (b) Drop on Section AS - VA - VB (5) 3. Load (Power) Flow (a) From node A to B: PAB ' JQAB(6 AD - ~~~~~~~~~~(6) where AB m real power QAB - reactive power (b) From node B to A :PB ' JQBA (7) where PBA - real power QBA - reactive power (c) Power loss between node A and B: Algebraic Sum of PAl-, JQAB and PBA M JQBA (8) Because of the non-linearity and complexity of the equations an iterative method such as the Newton-Raphson, developed into computer algorithm, is used in their solution for given convergence criterion. What most of the computer software programs do is to solve similar equation systems as given above. 2. Switchint Aralysis is done to simulate system behavior when distribution line sections are transfered from one node to another in the same or other feeder. The nodes and sections to be switched are provided to the computer program which automatically performs the operation. 3. L&ad balancinff amon uhases is achieved by putting loads equally on all phases depending on the nature of the problen. The loss reduction benefit of load balancing can be explained by the following example. Assume a current reading of 40 amperes In phase A, 130 amperes in phase B and 100 amperes in phase C, each phase having 0.706 ohm resistance and 80 amperes neutral current with 0.234 ohm resistance. The total loss for the unbalanced 3-phase line is equal to 0.706 [402 + 1302 + 10021 + 802 x .234 - 21.6 kW - 307 - Balancing the line current to 90 amperes/phase gives a total loss of 3 x 902 x .706 which is equal to 17.16 kU. The loss reduction is 4.46 kW. 4. Canacitor Size and Location Ogtimization: Capacitors generate reactive power to improved power factor and reduce line currents. In so doing they reduce losses. In loss reduction programs, capacitors of different sizes are placed in different sections in the network. Next, computer programs calcula..e the losses, and then, the capacitor is located when the desired loss level is achieved. Equally, capacitor size can be given to the program which then locates it where loss is at a minimum. 5. Economic Loading of Conductors and Transformers: Winding losses on a transformer are at a minimum if it is loaded and equally on all phases. To illustrate this point one can suppose an area is supplied by two substations of 10 NVA rating each with 50 kW winding loss at this load. Loading the first transformer at 12 NVA and the second at 6 MVA results in a total winding loss of: [2 VA]2 X 50 k + [ 6 2 X 50 kW - 90 kW 'O MVA 10 MVA If both transformers are loaded equally at 9 MVA the total loss will be 2 x [lO9A2 x 50 kW - 81 k. The loss savings is 9 kW. 6. Fault Current AnalXsis: In fault current analysis line-to-ground (single, double and three phase) and line-to-line currents are calculated depending on the type of the fault by the method of symmetrical components. Similarly the positive, negative and zero sequence fault impedance is calculated. 7. gBgonductoring: The current carrying capacity of a transmission line is directly related to the line's area and inversely to its resistance. Thus 'ncreasing the conductor's area to an economical level will reduce 12R losses. Reconductoring involves replacing the conductors of sections with others of larger cross section to liprove current carrying capacity. - 308 - III. SYSTEM REQUIREMENTS Most loss calculation programs require a microcomputer that has the following features: (1) MS DOS 2.0 or higher (2) Hard disk drive of 10 MB or more capacity (3) 640 kilobytes of memory (4) Color monitor and board (5) a printer Software packages for the electric utility industry are common in the market. A list of programs for generation, transmission, and distribution (including technical loss analysis) as well as their vendors is extensively covered in Elgctrigal Worl , November, 1988 issue. - 309 - IV. CASE STUDY OF A NETWORK WITH DPA/G 2/ The network (Figure 3) consists of two substations.- The procedure for loss analysL ia then to: a. divldb the network into separate feeders (circuits), eight ln this case; b. establlsh database for Control, Feeder, Conductor, Section, Node, Map, Equipment, and Voltage. It is suggested that nodes be established at the following locations: 1. At conductor size changes 2. At changes from overhead to underground 3. Where number of phases change 4. At large power and/or concentrated loads greater than 100 kVA. 5. At the beginning of branch 6. At the end of branch 7. Where future construction will need a node 8. At end of all lines 9. Near all voltage regulating devices and capacitors 10. At sectionalizing devices (fuses on short taps may be skipped). 2/ See Footnote 1. The author selected this software package for demonstration purpose only. --0~~4 1 -| 4a 4 e-, e,_, , C,, iz,f,. ..°JL tr.,.. L...L...... * , (A | i. 4,0 "u" -- FIUR 3:-, DITIBTO PRMR ANLII 3 PIAIPUAVt | I #~~~~4j ;eX FP 400 47SJ; *-*1v' * e , 00-+ # w# E.oo , " ,_,,,,,,, Fl6URE3:~~EAMPL DItIUINETWORKAA3I _ PIAS PRiMAY 136( :a_ .*". ES4HPLE NETWORK ._ .............. ~~'g-. da£ aR ....~~~~~~~~~~~~~~I - 311 - The followLng nine programs help in creating the database: CRECON -- creates and initializes the control file EDTCON -- allows editing of the control file CREATE -- creates and initializes the database random access files. DATAIN -- inputs the raw data into the database files. Programs utilizing the digitizer to enter data are also available CHECK -- locates cross referencing errors, incomplete data and unassigned node and section numbers BALMAP -- generates a digital map (for internal storage) of each circuit PHSCHK -- checks the phaslng information of each section in preparation for by-phase by the analysis program PHSVOL KWHIN - has the option of reading a formatted sequential file for the kWh, number of customers, spot loads, and connected kVA or prompting for terminal entry of kWh and number of customers to be placed in the section file ALLOC -- allocates feeder demands to the line sections in proportion to either connected kVA or peak month kWh. Allocation by-pha#e is an option. - 312 - Figure 4 shows the flow chart for establlshing the database. Typical data entry forms are given in Annex 1. The control file links the DPA analytical programs with the database files. c. run the following analysis and switching programs: BALVOL computes voltage, line loading and losses assuming loads are balanced among the phases. User may change load levels for the entire circuit and/or for selected line sections temporarily or permanently in the section file. Regulators and capacitors may be added or deleted, conductors replaced and phasing changed in the same manner. PHSVOL -- computes voltage, line loading and losses by phase assuming unbalanced load. User may change load levels for the entire circuit and/or selected line sections temporarily or permanently in the section file. Regulators and capacitors may be added or deleted, conductors replaced and phasing change in the same manner. CAPLOC -- optimizes the location of capacitor banks based on minimizing losses. Voltage, loading and losses for off-peak conditions are also computed if requested. - 313 - FIGURE 4: FLOWCHART FOR ESTABLISHING THE DATABASE NUMBER KAPS (1) Maps must be numbered and data EXTRACT RAN for the feeder, conductor, section, node DATA and equipment files. (2) Fill out the DATA INPUT FORMS if data are to be entered manually. If you are digitizing, have the data ready to enter . you digitize. RUN (3) Run CRECON to initialize the control CRECON file. Run EDTCON to set up database extension, .i , n location on disk of database node and - section liait, feeder and map limit, type RUN of terminal being used and any other [ TCON option that will be different from the preset values. RiUN (4) Run CREATE to initialize the database CREATE files. INITIALIZE TH4 DATABASE ONLY ONCE. (5) Run DATAIN to enter data into the data- base files. Conductor file and spacing table must be formed before the section data are entered. Capacitors, regulators and primary transformers my be entered at this time. If you are digitizing, CHECK refer to the Digitizer Users Manual. (6) Run CHECK to check for cross-referencing ,COSNO RUN errors and val&d records in the section REfERENN GE nd node files. (7) un CHANGE to correct errors in the ES rsection and node files. CHANGE may be executed at any time to edit the latabase. continued on next page 314 - continued from provious page (8) When cross-referencing is correct, run EALM^P 5 BALMAP to form the digital map of each RUN circuit. Run CHECK igain to be sure all CHECK branches are included in aps. Run PHSCHK to check phasing information of itUN each section if PHSVOL is to be executed. s , ~PHSCHK. RUN @! (9) Run KWHIN to enter kWh and customer data KWHIN via the terminal or formatted file. If the file is read, spot load data and connected kVA may also be entered. SUBO)EM (10) SUBDEN may be run to calculate Individ- ual feeder domandsif the-total substa- tion demnd but not the feeder demand is known. (11) Run ALLOC to allocate feeder demand to ALLOC \ .............osections in proportion to either con- nected kVA or peak month kWh. Run TLIST or PRINTA at any time to verify the RUN RUN database. TLIST PRINtR P IDATABASE I S NWM COWPLETE - 315 - FAULT -- uses symmetrical components to compute fault currents for each line section. Output includes minimum and maximum line-to-ground, phase-to-phase and three phase fault levels. SWITCH -- switches the loads between circuits or between branches of the same circuit. PLTGEN -- generates a graphical display of selected DPA/G calculated values for plotting with AutoCAD. d. and analyse the results. An analysis of the outputs of the five analytical programs is given below. In most cases, only a summary computor printout is provided. (i) Balanced and Unbalanced Load AnalXsis A summary of balanced and unbalanced load analyses calculated by BALVOL and PHSVOL, respectlvely, is presented in Table 1. The detailed output can be found In Annex II.1 and 2. -~~~~~~ - I. - 316 - Table 1: BALANCED AND M4BALANCED LOAD ANALYSIS tStI ULYUL (12.543 22JU-69 1054L SUIMANC OLS UJU L OSS 30CU1 JUL 3-71U 99 -'US Of a3.", iw ZNS VMTWeS MM NDJWS V11 LOA NAIWM LOStJg S=.o. winu M oU 9:.W. us.c. IVA m Ana in.1.,111U.VinU 243 16." 109.17 316 76." 530.0? 357.30 391.55 (a) Balaed PmI 1 u2.b-C) 2-JU-69 IOSM. seAn al EamC LS UUUCUO Jus 2-_ .1t9 3mrA AL FMU PAma VGLUA - -uMSo L LOSSE .us. To ae Lg.gv sm.m. MAN. A aW DwA UZ.iM-Te. n a i6 2.62 3.s5 169 61.26 1O.13 0.6? 5.2 b.1.W,l-tR-U S. 133 . 32.5 160 26.s5 l .11 2.17 1.63 in.1.NS,-nUD C I" 6.92 12.0 160 GO." 27.5' 23.63 16.50 (b) Ubalaced - 317 - Balanced analysis of Feeder 1, Substation 1 gives a maximum voltage drop of 16.83 to a level of 109.17 volts, maximum wlre load of 78.680 of its capacity in sections 243 and 216, and total loss of 357.30 kW for the feeder (Table la). Similarly, data on Table l.b show that Feeder 4 (West Feeder), Substation 1 is slightly unbalanced. The load variation among the three phases is less than 1%. (ii) Capacitor Placement A base run was obtained without a new capacitor in Feeder 2 (East-feeder), Substation 1. The total power loss obtained is 481.9 kW (see Table 2). Two banks of capacitors (300 and 100 KVAR) were placed by the program in Section 209 and 210 respectively. The total loss dropped to 417.4 kU. When the line was loaded at 30% of its peak load with the new capacitors still in place, no excessive voltage was recorded. A savings of 53.4 kW resulted from the addition of the 300 KVAR capacitor bank. The 100 KVAR capacitor bank further reduced the losses from 428.5 to 417.4 kW. - 318 - Table 2: CAPACITOR PLACEMET VUWK C@AP M.S-P) 2n-JU-6 Zou*L 5311343 aI r3e LO wc2= JULy 2-..190 VU 2 XU.1AAXt-1, U 5611 WISU SAYrNS CAPA?so -: cm (3n t12 r AR) am *51.9 0.0 0. 209 *42.$ S5.4 200. 30 417.4 3l.1 100. * uamm cAnc (U2.S-) f n-JU-so U20IAL SWI* - .3ZC Z EB LOU .190=W JULY 2-7.1989 me,aai cW AiL 79nus sIN U VMTALZ OM NXAI5 W0N LO MAIWGS LOSSU .wi. Nm nma LE0D SW.XD. 1.CAI. EVA g WAR SM.L.X -vS4 = 210 *540.57 6.43 Su 1.20 723.40 431.93 542.2? .1.EA$?-VtU 210 51.56 94.08 206 60.32 633.14 417.1? &19.72 X=.i.un-11 an 2.1s 1W.25 210 zo.o 103.33 o1.00 o 4.9g - 319 - (iii) Fault Analvss. The fault statistics Ls given in Annex II. 3. Feeder 5, Substation 2 has a line-to-line voltage of 12.47 kV and source impedences (ohm) R, - 0.61 , Xi - 0.4 1 - 0.1 . ,e .- 0.1 The computation algorith assumed a minimum and maximum phase-to-ground resistance of 35 and 0.0 ohms respectively, in calculating the phase-to- ground fault currents. The largest fault current in all phases occurred in source node 32, since it is the nearest to Substatlor 2. (iv) Switching Studies. Program SWITCH joined Section 219 of Feeder 1 to node 60 of Feeder 4 and separated Section 217 from Node 116. Sections 217 to 221 inclusive wlll now be supplied from Feeder 4 instead of Feeder 1. Table 3 shows the map computed by the program after the switching exercise. Figure 5 is the graphical presentation of the two feeders before and after switching. The program SWITCH therefore allows loss and voltage values to be compared if line sections are transfered from one feeder to another. _ 320 - Table 3: MAPS OF FEEDER 4 AND 1 AFTER SWITCHING OPERATION Im S3%t (Y2.SQC) 22,JU3-u9 R=ZOM SEM MS C Lo z=X m 3-7.,^"§ 50.1.V3S?-linU me VON= 3 I V s 8 1 49a -L I " I 50 -2 a -1 I 35 ' 51 3 -1 3 132 3 2 3 -2 1323 * 5 3 *-1 3 120 3 50 _ -2 a -1 S 154 U 54 3 -2 3 s55 3 55 5 -2 3 157 U 5? a -1 100 6 " a -1 a 2al 11 a -1 3 213 * 3?7 * 21? X S00 3 -2 1 -1 S 220 3 120 3 -a S 221 5 1n g -2 * -1 3 101 U CS 3 -2 * 102 * 02 3 -2 8 150 9 50 S 139 a 59 13003* SMZ?C3 (V2.S-C) 22-JW-S9 uwCLz suAs 2 U= C LS R203 JOLT 3-7-1919 5o.1,Nom-12cm No UWUU 1 S 210 110 3 222 * 22 * -1 * 225 * 123 224 3 124 * -1 3 225 3 12 3 -2 S 226 U 126 S 227 U 12? 5 -1 5 22 US 1 3 -2 * 229 p 12 5 -1 3 230 3 120 5 -2 3 -3 3 251 3 12 3 -2 S 232 U 12 3 -1 3 2323 5 % 3S -2 S 234 124 -2 O 225 3 125 -1 S 250 Y 130 3 -2 5 -1 823? U 123 5 -1 3229 0 I1 2 -2 S22 I 1393 20 0 140 S241 3 1413 *-1 3 242 3021 -2 S 242 5 24 3 -2 S 24 * 14 3 247 U 147 * -s - 321 - FIGURE 5 MORE SWITCHING *zFEED I <~~~ .5 J 4' ,,, 1 I~~~~~~~~~I 4 3 _ C~~~~~~~~r=Z. R 4 AMER SWITCHING FEEDER 1 I. I.. ~ ~ ~ ~ ~ ~ ~ ~~4 I IP~~~~~~~~~~~' .~~~~~~~~~~~~ ./ *zF F , lS S *fo *~~~~ 4, 4,w -- i 1 7DER 4 _- pI __ _ not st srcStt ta c SWDt src s Scott dScott Cantats - 322 - V. CONCLUSION The availability of software packages for the electric utility industry that can run on microcomputers has made the calculation of technical losses and implementation of loss reduction programs very feasible. Microcomputers have several advantages over manual, slide rules and calculators in loss calculation. They have high calculation speed; large memory that can store data and perform complex and iterative tasks; and present out in graphical or tabular form. Life-cycle savings of $15.0 or more for every $1 spent on loss recletion projects are not unusual. Use of microcomputer programs expedites the development of economic loss reduction projects and reduces the overall costs. 4 - .TI I I f III III 11 r III II I rr1r 3OWN 1:303 F TrM O _ _ ligillifif 1 3NwN U3033i 2 - _-- I I 16 I I I I ri m I I I I I i. I I I rlo I I I I I I'l II II q m I I . e I I I I 1 _31 803 . . 1l~~~~~~~~~~~flll# .1 fil 311VW l30ll3J 1! 1-1 1 1 1 1 1=1 ~~~~~~~~~~~~1 III I II If I- I it 11 11 I| I e I I I Ie 1. I 3nV 1 103 | .TA I II 10 I I I I I I 11 1 1 11I 1 1 I I Io I 1 III le I I I I0 I1 I I I I I I I-r I I-11N 33 , I I T I i I ~~~~~~~~I I I - I I I I0 I I I I ^rT* I I I I 111 111 I I lW 33 0 0 0 a~~~~~~~~~~~~~~ I1 nV 3 C4 ~~~~~~~~~~1 1|1l I11 If|1 Iz| il ll31WN 83033A _ . . , l _ I g t 11 1~~~~~~~~~~~~~~~~ 1311N V3033: .1 3VN V13033i 1-1 lei I I 111~~ I,ll 101111111111 11 113" N 03 TM I 3nV~~~~~~~~AN 030331 I 11 I I 11 -llllllitilli l l -lt lifl I 13llN M3033 flilf lielit111*11 1-174111--1- 11I i ill lllUlW lit lill 1illif 311""B033 MM v lvm ox to ~~~~~I u(vANSI '13A31 RUIAM 9-1 L in~ ~~~ ~~~~~~~ : i ON13013 5915O Yvuoi infdNI ViVa &130331 - 324 - Page 2 of 2 _~~ __=____ OIANOWS o S... ii ,,. airir 4imiThN 0EMll|lXtXl U w t-X <;i j.. f-ag-j. i .. jX ?~ ~ ____ G ___ ____________X -325- 120AN ULw"m (Y2.5) 22-JUN69 as@iwsz Smay or .3cuIC LOWS 3330012 JUL.Y 2s.1.i,a ANNEX 11. . vmuM I 621u33i VOLISCX a 12.47 XV 1.11 10 1.23 S530!S LOI M3U1 CME - LMOA IN 63ZC=O-- - LOAD TM 8=00O - VOLTA603 120 am1 -- WLlS -- cMU z Ft Cainr 2am OMi SW IVA AMW CMs COW UW gVYM AMPS CMI $=C ACCU LEMU KW KYAR EVA I 3mm mmO SVISUTAIZO TOTALS 4229. 2140 236. 572. 12.0 2537.3 291.6 216 116 2.2 ASC 310 AC 27. 49. 34. 2 .2.0 76.? 4205. 3151. 225. 572. 4.2 4.2 121.0 119.S 134.1 217 117 10.6 ABC 4 AC 402. 240. 166. 12. 40.0 29.2 295. 606. 22. 101. 2.6 7.0 119.0 16.6 5.5 216 1n6 2.2 ABC 4 AC 406. 241. 167. 14. 41.0 16.6 225. 233. I9. 61. .1 7.6 116.2 2.6 .9 219 300 523 SIC 4 AG 41. 51. 35. 2. 2.0 2.1 25. 16. 1. 3. .1 7.9 116.1 .0 .0 220 120 5.1 A 4 AC 106. 90. 62. 16. 10.0 11.1 45. 31. S. 10. .4 6.2 U17.8 .2 .1 221 12 2.3 A 4 AC 76. 73. 51. 12. 7.0 9.0 27. 26. 6. 7. .2 6.1 117.9 .1 .0 222 122 5.3 ASC 310 AC 102. 95. 6 3 . 6.0 64.6 2299. 2456. 191. 470. 2.4 1.5 116.5 79.6 S9. 4 222 123 1.0 ABC 2 AC 235. 157. 109. 9. 22.0 50.6 1S26. 1036. 8?. 226. .5 6.1 117.9 7.3 2.7 224 124 4.? ABC I AC 269. 173. 120. 10. 26.0 45.6 1364. 920. 7Ur 205. 2.2 10.2 113.8 27.0 13.7 225 125 2.6 A 4 AC 21.4. 145. 100. 25. 22.0 16.2 72. 50. 12. 22. .5 10.7 115.3 .3 .1 226 126 2.0 SIC 2 AC 266. 162. 126. I1. 23.0 25.5 1014. 66. 59. 157. .7 10.9 115.1 6.6 3.3 3EUMA2Ot 1 (EZ2IThU 120.00) =3 I3If ON 227 AT 3003 126 907. 617. -10.9 .0 126.0 9.0 227 127 6.6 ABC A AC 207. 192. 122. 10. 21.0 24.5 611. 520. 42. 129. 2.1 2.1 1,22. 9 23.9 7.9 225 12 7.9 C 4 AC 245. 1GO. Ill. 26. 26.0 16.9 6t. 56. 12. 26. 1.1 4.2 121.6 .6 .2 229 129 5.2 AC 4 AC 121. 102. 72. 9. 12.0 21.2 479. 229. 39. 72. 2.0 5.1 120.9 6.1 2.6 220 1us 6.5 C 4AG i22. 99. 09. 17. 12.0 11.9 50. 24. S. 32. .6 2.7 120.3 . 2 .1 231 121 6.6 C A AC 142. 119. 62. 20. 16.0 14.4 6. 42. 10. 16. .7 5.6120.2 .4 .1 222 122 5.0 A 4 AC 102. 67. 60. 15. 11.0 24.0 156. 109. 20. 21. 1.4 6.5 119.5 1.9 .6 222 1523 6.0 A 4 AC 122. 102. 71. 17. 14.0 12.5 52. 26. 9. 14. .6 7.1 116.9 .2 .1 224 124 6.6 A A AC 26. 9. 6. 2. 6.0 1.1 5. 3. 1. 6. .1 6.6 119.4 .0 .0 225 125 4.2 SIC 310 AC 236. 1M. 110. 9. 22.0 22.4, 1576. 1188. 93. 224. 1.3 6.9 117.1 14.6 16. 6 226 126 6.1 3 45 At124. 100. 69. 17. 12.0 12.4 50. 35. 9. 12. .6 9.5 116.5 .3 .1 TRASmOROM 11 13 U3091 ZS? At30M 122 138.7 1161. 789. 6..0 .0 110.6 22.2 105.3 21Z VOLTAGE IS 24.90 "DILl 237 12? 2.4 AMC I AC 222. 150. 104. 3. 21.0 19.2 1062. 727. 22. 169. .5 IS.? 110.2 5.5 2.6 225 126 1.6 3 4 AG 24. 23. 22. 2. 2.0 2.2 17. 12. 2. 2. .0 15.7 110.2 .0 .0 229 129 6.2 SIC 2 AC 26. 171. 119. S. 24.0 16.2 666. 599. 27. 146. .5 16.2 109.6 6.2 2.1 240 140 5.3 ABC 2 AC 231. 214. 149. 7. 24.0 13.3 68. 462. 21. 122. .2 16.5 109.5 2.2 1.1 241 141 5.3 ABC 2 AC 221. 150. 104. S. 21.0 9.6 464. 226. 1S. 86. .2 16.6 109.2 1.1 .6 242 201 3.0 ABC 4 AC 61. 76. 24. 2. .7.0 1.7 29. 27. 1. 7. .0 16.6 109.2 .0 .0 242 24 5.2 SIC 2 AC 365. 226. 229. 10. 60.0 5.7 165. 115. 5.6. .1 16.6 109.2 .1 .1 VOL!AGZ3ZS U1.470 KIL. 246 16 7.0 SIC 6 AC 272. 114. 121. 10. 27.0 9.1 122. 92. S. 20. .4 9.2 116.7 .6 .2 247 147 4.0 A 4 AC 2a1. 46. 22. 6. 2.0 3.7 22. 16. 4. S. .2 9.5 114.5 .0 .0 3M OVC ZR a flmRATUc3(8 Q13M .303 ASI 1OWX3M0 FACTOR -326- noAKow v 9501. 2.) 2JU89ANNEX 11.2a uzcozouz suaum on uzczm c LOSS auUCizop .iu 3-7 1989 FMIR 4 50.1 .VT-7I=u VOLTAGE 12.47 KV LlnK TO LInK SEC END LOl PEASE C016 ---- LOAD Is 811111-- *- LOaw 160 5=010o- VOLTAGZ 120 3BASS - LOSSES - NOO K VT O0M6 512 COM KM UVAR AeS cbS? CONo IN zVA ~AeS COST sxcT A=6 LWLV KWd KVAR ZVA 2 VWO amO SUNSTAZTIO TOTALS ON MUAS A 594. 52. 79. 90 .7 2.2 PEASEz a 373. -Si. i0. St. 2.2 1.0 lEAS 6 948. no9. 132. 163. 126.0 25.4 14.5 1 223 2.8 A 310 AC 50. 32. 13. S. 2.0 25.3 576. 45. 77. 90. .3 .3 125.7 2.4 2.7 1 5~~~0. 15. 7. 2. 1.0 16.6 235. -53. 49. 59. .2 .2 125.8 1.0 1.1 6 10~~l. 49. 22. 7. 3.0 43.9 944. 21. 1no. 163. 1.3 1.3 124.? 5.7 7.5 149 49 1.5 A 2 AC 50. 29. 13. 4. 4.0 41.3 545. 83. 73. U8. .5 .7 125.3 2.S 1.4 a 155. Lie. 49. 16. 15.0 26.6 201. -20. 40. 4. .2 .4 125.6 .9 .4 c 25. 15. 7. 2. 2.0 G9. 4 905. 249. 126. 180. 1.5 2.8 122.2 8.4 4.2 CAPA r1OR IN SIC? 149, P3*1 A 100 VAR 109. ADJUSTN) CA14AC101 1 S=C 149. PEASE S 100 &VAR( 110. A0306150) CAPACITOR IX SEC? 149. PLUME C 100 KVAR 105. A0J0S150) 150 so 5.2 C 4 AC 413. 12. 5. 2. 45.0 1.2 5. S. 1. 45. .1 2.9 125.1 .0 .0 151 51 .8 A 4 AC 75. 41. 1a. 6. 5.0 11.2 67. 39. 12 10. -.1 .5 135.4 .1 .0 a I15. 62. 37. 12. 10.0 12.5 U1. 36. 12. 13. I1 .4 125.5 .1 .0 a 25. a . 4. I1. 1.0 57.6 542. 242. 80. ISO. .7 3.5 122.3 2.7 .9 152 52 2.4 C 4 AC 100. 72. 32. U . 10.0 7.5 35. 1s. 5 . 10. .1 3.5 122.4 . 0 .0 153 53 1.0 A 4 AC 70. 04. 20. 10. 5.0 5.9 22. 15. 5. 5. -.1 .3 125.5 .0 .0 a 45. 40. 16. S. 3.0 4.1 20. 9. 3. 3. .0 .5 122.) .0 .0 O 45. 40. 16. 5. 3.0 49.3 443. 199. 48. 77. .7 4.3 121.7 2.4 .8 156 25 3.5 6 4 AC 310. 229. 103. 34. 42.0 24.6 115. 52. 17. 42. .7 4.9 121.1 .5 .2 154 54 2.3 C 4 AC' 120. - 95. 43. 14. 16.0 10.3 46. 21. 7. 16. .2 4.4 121.5 1 .0 155 55 2.1 a 4 AC 125. 94. 43. 14. 16.0 10.3 4. 21. 7. 15. .2 4.4 121.6 1 .0 157 57 1.7 A 4 AC 195. 98. 44. 14. 14.0 40.5 371. 59. so. 74. .7 1.5 124.5 2.2 .7 a 100. 63. 28. 9. 9.0 12.6 92. -49. 15. 40. -.1 . 3 125.7 .2 .I C 28. 7. 3 . 1. 1.0 32.4 329. 43. 45. 45. .6 3.5 122.4 1.6 .6 160 80 1.9 A 4 AC 175. 148. 74. 24. 25.0 30.7 235. 53. 32. 60. .7 2.1 123.9 1.1 .' a 15. 5. 3. 1. 1.0 9.7 57. -29. S. 31. -.I .2 128.68 1 .0 c 90. 54. 29. 10. 10.0 9.5 32. -28. 7 . 10. .0 3.7 122. .0 .0 CAPACITOR IN SIC 150. MMAS A 100 VARt C 107. ADJUSTED) CAPACITOR IN SMC 150. PUEAS 100 &VAR 110. ADJUSTE) CAPACITOZR S INSCT 1O, PEASE C 100 DVARt 104. AD.ZUS?5) 151 61 2.2 1 A AC 220. 53. 24. B. 30.0 5.5 27. U2. 4. 30. .1 .3 125.7 .0 .0 152 62 2.) A 4 AC 325. 152. 48. 22. 34.0 15.1 is. 34. II. 34. .3 2.4 123. .2 .1 158 so 1.4 C 4 AC 165. 145. 55. 22. 20.0 27.7 167. 64. 26. 34. .4 4.1 121.9 .6 .2 159 59 1.9 C 4 AC 115. 114. St. 17. 14.0 1.2.2 57. 25. 9. 14. .2 4.3 121.7 1 .0 unor FIRER 2 ITECATIIC(S) 553 .50X AS CONVDOZNCZ 1ACT01 - 327 - ANNEX L1.2b PrOG RM PSTOL (v2.3-C) 22-JUl-69 UXWNAL S3I1MMA ON SLCTRIC LOSS MDUCTION JTULT 3-7,190 6U3.1,VET-71=ZI IMIR VONM 4 ULOVATT.KLOYAR AND CeuTr nLO ST flAS£S AM KIUTSA S= M~~~~~ILOVATTS 213.0VA25 AUflZS W0. A a C A i C A a C 3 XS.V=AXO TOTAULS $59.9 372.9 966.6 52.2 -S5.6 229.2 76.9 69.6 121.6 76.2 1 577.7 326.8 966.1 45.0 -53.0 226.2 76.7 46.8 123.2 76.3 149 566.5 200.7 90. 9 62.0 -29.6 269.2 72.0 60.0 125.6 77.9 IS0 .0 .0 5.9 .0 .0 2.6 .0 .0 .9 .9 3.1 6.7 60.7 341.6 26.9 26.2 262.3 12.6 11.7 0.3 6.0 112 .0 .0 35.9 .0 .0 16.1 .0 .0 5.6 2.6 123 22.1 19.6 662.2 16.8 6.9 196.5 6.8 2.9 44.1 62.2 15 .0 .0 115.1 .0 .0 51.6 .0 .0 17.3 17.2 154 .0 .0 67.9 .0 .0 21.5 .0 .0 7.2 7.2 155 .0 .0 67.9 .0 .0 21.5 .0 .0 7.2 7.2 117 270.7 91.6 229.1 59.6 -66.? 62.6 69.9 15.2 66.9 28.3 16o 226.6 56.6 22.0 52.7 -29.5 -27.6 22.6 0.5 6.7 20.6 U61 .0 26.7 .0 .0 12.0 .0 .0 2.9 .0 3.9 162 76.2 .0 .0 36.2 .0 .0 U1.2 .0 .0 11.2 158 .0 .0 167.6 .0 .0 66.0 .0 .0 23.0 26.0 IS9 .0 .0 57.2 .0 .0 25.6 .0 .0 6.6 6.6 - 328 - ANNlEX 11.3a PROCRAM FAUT (V2.3-S) 22-JoN-Us RECGIOAL SVUAR ON ELC2AIC LOSS RlDUCTION JULY 3-7 1969 11DmZ 5, SUB.2,NZ.F1ZDZ SUBSTATION VOLTAGE 12.47 KV LINE TO LINE SOURCE I@ZDANCES (OHMS) R1 * .610 XI - .400 * ASSUMES 35.0 OCKS R0 * .100 X0 * .100 *A SSUKES 0.0 03KS ,,,,-,^-- CUHUAT .. ….......... NODZ fODE WNM gi/n POSITIVE SEQ. ZErO StQ. PS-TO-CU I-TO-PU 3-Mt LOCATION S1gt IROM R X a X IHO KAX** (AMS) (AMS) SUB (09)) (On) (AMPS) SOUR-S 203. S3519. 6546. 9870. 32 397 AC 1.056 .662 .526 .209 .690 203. 9309. 7374. 6515. 39 4 AC 3.696 1.947 .960 1.637 2.295 195. 3117. 42 397 AC 3.016 .836 1.000 .016 2.901 201. 3061. 4738. $470. -S 4 AC 7.636 2.141 1.414 2.046 4.306 194. 2230. 397 AC 7.656 .965 1.316 .890 6.375 200. 2854. 3794. 4361. 96 6 AG 11.616 2.912 1.936 3.032 6.783 19. 1559. 112 397 AC 12.936 1.264 1.947 1.435 7.323 197. 1616. 2699. 3116. 113 310 AC 16.216 1.967 2.756 2.664 10.452 192. 1256. 1640. 2125. .13 6 AC 20.066 2.866 3.048 3.464 11.376 167. Ioqp. 119 4 AC 22.176 3.694 3.379 4.586 12.640 161. 931. 142 4 AC 22.704 6.151 3.462 4.672 13.161 180. 699. 163 310 AC 21.646 2.437 3.266 3.100 12.465 169. 1046. 1524. 1760. 144 4 AC 25.364 6.235 3.635 5.099 14.733 179. 623. 165 A AG 29.366 6.291 6.527 7.384 17.301 169. 653. 146 310 AC 29.304 3.435 4*462 4.563 17.021 161. 761. 1101. 1272. 149 4 AC 33.528 5.560 3.142 6.848 19.555 171. 620. 625. 932. 293 110 AC 33.264 6.325 5.097 5.617 19.394 176. 6"6. 933. 1077. 286 110 AC 38.544 5.444 5.944 7.024 22.557 169. 356. 774. $93. 260 110 AC 43.624 6.564 6.790 6.430 2S.720 163. 482. 660. 762. --. SOURCE TERNINALS 6.564 6.790 6.630 25.720 163. 462. 660. 762. 0** TRANSFRME 10 IN SECTION 293 AT NODZ 260 T* HUE VOLTAGE IS 24.900 IVLL * LOAD 74IMNALS 36.093 68.274 43.530143.754 106. 120. 131. 151. 275 110 AC 69.104 37.213 89.121 46.9$7164.917 105. 116. 129. 149. 272 110 AC 54.386 38.333 69.967 46.343170.080 103. 116. 127. 147. 276 4A C 53.064 39.141 69.741 47.078169.323 103. 116. 277 4 AC 54.386 39.783 89.948 47.792170.126 103. 116. 276 6 AC 59.136 62.096 90.693 50.363173.017 101. 114. 279 4 AC 59.928 42.-41 90.817 50.791173.499 101. 113. 150 4 AC 58.872 61.967 90.651 50.220172.8S7 101. 116. -' TIM VOLTAGZ Is 12.470 9=LL 261 4AC 49.z46 9.134 7.639 11.206 21.647 153. 606. $26. 40S. 2i2 4 AC 53.366 11.647 6.386 13.856 31.776 144. 3S5. 283 4 AC 53.592 11.318 8.361 13.713 31.S76 145. 357. 442. 512. 2 4 AC 55.966 12.675 6.734 14.998 33.023 141. 235. 265 4 AG 55.660 12.216 6.651 14.713 32.702 142. SA0. - 329 - ANN= G II . 3b 0"0UL SDIXIU ON LI C SX Rzm JT 3-7.,l" Ynm 5, sa.m2.Ugju SUWSTZIOI VOLIA= 1n. 47 NV 10 TO L1 S =DWDACZS (OMS) U1 - .410 X . .400 * asSS 35.0 0UE a0 - .100 3 .100 A 6A3 0.0 0 .S---__M?M am1 MODS 13 iluT MImR =Q. gU gm. f3- 70-U PR-Wpm 3.- &cA?zW sin * a x a x Ma MAX- (MM) (MS) gm M( S) (CMS) (AM) 57 & AG 43.824 6.014 6.M 9.879 a2s.72S i. 45. s3. us. 2U 4* C "." 9.2n 7.207 11.307 V7.329 135. 42. ae 4 *A a.200 .an 7.175 .1.45 27.1U6 33. 424. 3". 418. 21 4 AC 44.312 10.199 7.304 12.307 20.7'23 10. 297. Sol * Ae 0o.424 1.2a 7.6 13.449 29.n77 144. 373. 151 * AC 44.952 10.410 1.416 12.433 24.79f 14U. 390. 4*0. S33. 132 4 AG 49.512 M0.997 7.779 3.3.194 29.296 147. 376. 463. 34. 3 4 A*C 50.932 11.483 7.940 13.73 9.996 145. 36. 447. 316. 894 4 AC S8.80 7.023 3.*4 0.614 22.763 14. s52. 295 4 AC 404.3 7.922 6.2SS 9.416 2."9 136. 48. 3 4 *AC 34.34 6.034 3.290 7.419 a0.232 1. 592. 34 A AG 39.864 6.424 6.137 10.235 23.4S 137. 479. 1SS 4 AG 41.976 9.432 6.44 11.417 14.727 133. 443. 156 4 AG 45.$60 10.423 6.496 12.374 2.90 150. 422. 137 4 AG 4S.144 1.194 4.944 13.131 25.653 143. "1. UO 4*A 47.236 12.222 7.27s 14.273 .3 1S4. 376. 1f9 4 AG *M.2 12.093 7.234 14.130 27.777 1S. 379. I16 4 AG 4.S76 12.664 7.432 14.967 #.740 143. 342. 141 4 AG -9.3 68.347 6.034 9.9U 23.122 1U. 4-9. - 330 - REFERENCES 1. P.J. Watson. "Microcomputers - Their Use in Metering", Third International Conference .on Metering, Apparatus and Tariffs for Electricity Supply, Conference Publication Number 156, 15-17 November 1977, pp 123-128. 2. N.E. Chang. 'Manage Losses with On Line Computer", Electrical World, Volume 187, Number 6, March 15, 1977. pp 93-95. 3. Kenneth W. Priest. "Cut Losses Through Unbalanced - Feeder Analysis". Electrical World, Volume 195, Number 12, December 1981. pp 93-94. 4. Scott and Scott DPA/G Example System and Users Manual Version 2.5; Seattle, Was?.ington 98121. 1989. 5. American Public Power Association. Distributing System Loss Evaluation Manual, Washington, D.C. 20037, August 1986. 6. D.G. Flinn and Westinghouse Electric Corporation. Improved methods for Distribution Loss Evaluation. Electric Power Research Institute. Palo Alto, California, 1983. 7. American Public Power Assoc-.ation. Distribution Circuit Analysis (DISTANL) Module, PowerManager Software Package. APPA. Washington, D.C. 20037. 1987. 8. C.L. Wadhwa. Electrical Power Systems. A Halsted Press Book, John Wiley and Sons, New York, 1983. 9. Olle I. Elgerd. Electric Energy Systema Theory : An Introduction. McGraw-Hill Book Company, New York, 1982. 10. Electrical World. "The 1989 T&D and Generation Software Catalog". F,lectrical World. Volume 202. Number 11, November 1988. 11. NITEC, Inc. Power-Donut System Software, NITEC Inc. Fairfield, Connecticut 06430, September 26, 1988. 12. Y. Alkaoto and T. Heryashi. "Computerised Integrated Automatic Control System for Electric Power Networks." International Conference on Large High Voltage Electric Systems Report 39-04. Paris, France, August 28-September 3, 1988. 13. G.J. Anders et al. 'Delaying Investments in Transmission Facilities Through Cable Uprating.' International Conference on Large High Voltage Electric Systems Report 37-08. Paris, France. August 28- September 3, 1988. - 331 - 14. A. Rodriguez and R.J. Yinger. 'Conductor-Temperature Monitoring Increases Transmission Line Ratings.* Transmission and Distribution, March 1989. pp 144-148. - 332 - ST. VINCENT ELECTRICITY SERVICES LTD. - CASE STUDY By Mr. Lennox Morris ST. VINCENT AND THE GRENADINES - GEOGRAPHY 1. St. Vincent is the second smallest Windward Island, with an area of only 133 square miles. It is situated approximatley 100 miles north of Grenada, 21 miles south of St. Lucia and 100 miles west of Barbados. 2. There are seven major Grenadines Islands governed by St. Vincent. These islands extend in a chain between St. Vincent and Grenada and include Bequia, Union Island and Canouan; hence St. Vincent and the Grenadines. The total area of the Grenadines is approximatley 35 square miles. 3. St. Vincent is volcanic in origin and is therefore mountainous. There is a range of mountains stretching over almost the entire length of the island from La Soufriere in the North to Mt. St. Andrews in the south. This mountainous terrain inhibits the development of inland St. Vincent. Industrial, commercial and housing development is thus confined largely to the coastal areas and, more so, to the southern coastal areas of the island. 4. The total population of St. Vincent and the Grenadines is approximately 120,000. VI NLE'S SCOPE OF OPERATION 5. The public electricity utility in St. Vincent and the Grenadines is the St. Vincent Electricity Services Limited (VINLEC). It is a Government- owned company which has the monopoly on the supply of electricity on St. Vincent and Bequia. Under contract with the Government, VINLEC also operates the electric system in Union Island. VINLEC, on behalf of the Government, is now doing a feasibility study on the electrification of Canouan. GROWTH OF VINLEC SYSTEM GeneXati2n 6. The first power station in St. Vincent was opened in 1930 in Kingstown to supply the town and its environs from two 40 kW generating sets. The station was operated and owned by the Crown Colony Government with assistance from the appropriate colonial agencies (Colonial Development and Welfare, Crown Agents). 7. The expanslon of the system was slow at first, mainly due to the low level of industrial activity. By 1953 the supply extended from Ratho Mill on the Windward side to Camden Park on the Leeward side of the island. The - 333 - installed capacity at the power station had increased to only 368 klW but apparently the plant was running at 10% overload during the evening peak period. 8. However, the need for a cheaper source of electricity had already been recognised. Around 1949 work was started on the South Rivers hydroelectric scheme by CD&W on behalf of the St. Vincent Government. The Colonial Development Corporation was invited to complete the project and was given the mandate to owo and operate the public electricity supply as a monolopy under legislation drafted for that purpose. South Rivers Hydro station was commissioned in 1953 with two 275 kW impulse turbines. Another turbine of 320 kW capacity was added the following year. 9. As demand incteased steadily a new hydro electric scheme was planned for the Richmond River. This project was started in 1959 and completed in 1961. The station provided an installed capacity of 1100 kW and came on stream in time to relieve the then strictly enforced load shedding to industrial consumers during the evening peak. 10. The two hydro stations are the run-of-river type (South Rivers has a small balancing tank). Their output is therefore greatly reduced during the dry season (to about 40&). Therefore, to meet increasing demand and provide sufficient firm capacity during dry season, a second diesel station was built at Cane Hall. Commissioned in 1971, this station provided an installed capacity of 3646 kW. Meanwhile the Kingstown station had expanded to provide 2136 kW. 11. The rate of increase in demand slowed considerably during the years 1972 to 1977. This was due in part (1974 - 1977) to the world energy crisis. 12. The St. Vincent Government began negotiations for participation in the Company about the same time that this slow-down began. This participation was obtained in 1973 tr"hen St. Vincent Electricity Services Limited (VINLEC) became a limited liability company with 49% Government participation. 13. 1974 saw steep increases in the prices of oil, spare parts, equipment and services and in spite of the fuel surcharge levied to recover the additional expenditure resulting from increases in fuel oil prices over the October 1973 level, the Company faced a substantial deficit in its operations by year end. From this point onwards to the early 1980's funds remained in short supply. 14. The coming on stream of the flour mill and the beginning of development of the Camden Park Industrial Estate in 1978 caused a substantial jump in demand. This growth in peak demand came to a standstill in 1980 when load-shedding became a must, due to prolonged dry spells and machine maintenance problems. 15. Commonwealth Development Corporation (CDC) in Its quest for funds to Increase diesel capacity was not very successful. They therefore had to be contented Lth the purchase of three diesel standby sets of 600 kW each in 1983. These sets were installed at Cane Hall and came on stream in January 1984. 16. The Government of St. Vincent and the Grenadines became the sole owner of VINLEC in June 1985. In that same year another diesel unit of 3200 kW capacity was added at Cane Hall and construction of the Cumberland Hydroelectric scheme was started. - 334 - 17. The Cumberland hydro scheme comprises three power stations in a cascade arrangement on the Cumberland River. The first station commissioned in June 1987, provided one unit of 1464kW rated output. The second station, commissioned in October 1987, provided 1280kW, and the third station, providing 950kW, came on stream in March 1988. Transmission and Distribution 18. The Company's Transmission and Distribution system had not grown at the same rate and with the same level of planning as generation. 19. In 1951 CDC was required by law to build an llkV overhead line to link the South Rivers and the Kingstown power stations and to build spur lines into Georgetown, Mesopotamia and Camden Park. This comprised some 37.5 km of lines mainly on the Windward side of the Island, only 6.4km from Kingstown to Camden Park being on the Leeward side. 20. In 1961 when Richmond hydro plant came on stream anothe-r 21km of llkV lines were added to the system to link Richmond on the North Leeward side of the island, to Kingstown. 21. These two llkV lines linking the two hydro stations to Kingstown, formed the backbone of the primary distribution system in St. Vincent. It is worthy of note that some areas of these two lines remained untouched in terms of maintenance or modification until 1986. 22. After the Cane Hall Power Station was built another section of llkV lines approximately 6.5 km long was built to link the Cane Hall and Kingstown plant. The rest of the primary distribution followed load growth. In some cases 6.3 kV single earth return lines were built to supply electricity to relatively small loads in remote areas. 23. In 1987 a 33kV line was built to link Cumberland Hydro Station to Cane Hall Station and to a distribution substation at Camden Park. 24. To date, the transmission and primary distribution system comprises 30km of 33 kV lines, 130km of llkV three phase, 30km of llkV two phase and 8km of 6.3kV single phase lines. Secondary distribution is done at 400V three phase or 230V single phase 50 Hz. It is difficult to assess the length of LV circuits since no LV mapping was done until very recent times. IlACTORS CONTRIBUTING TO HIGH LEVEL OF LOSSES Organisation and Staffing 25. In the earlier life of the Company under CDC's management, the company was run strictly as a profit-making concern. The local management comprised one chief executive below whom were only people at the craftsman's level. The chief concerns of the Company seem to have been to keep the machines running and to collect the Company's revenue. No organised Transmission and Distribution (T&D) staff was retained by the Compiany until about 1960. - 335 - Transmission and Distributton 26. In 1960 during the building of the Richmond to Kingstown llkV line, two T&D crews were organised under a T & D Superintendent. The qualification for entry into one of these crews was experience as a sailor, the reasoning being that if one could climb a mast he would also be able to climb a pole. No recognition was given to the fact that a little technical knowledge would at least be helpful. The trend was therefore set, and until the early 1980's a linesman was someone who could climb and work on a pole. T&D operated two 12-man line crews until the early 1980's. 27. The Company's distribution policy seems to have been to get power to the consumer at the cheapest possible cost. This is illustrated by the Richmond to Kingstown llkV line. The voltage at Richmond had to be maintained at 12.2kV in order to obtain llkV at the Kingstown busbars. This is a result of a decision to economise on conductor size. Kingstown and its environs are the main load centres. 28. The primary distribution system was expanded without any engineering planning and without any standards. Several single 6.3kV ground return lines were constructed to take supply to remote areas. As the load in these areas grew these 6.3kV lines were further extended, until in 1983 there were some 25.5 km of such lines. One would be amazed to know that the conductor used for most of these 6.3kV lines was 3/80 guy steel. 29. LV distribution circuits fed from fairly large transformers were extended span-by-span to keep in step with growing housing development. This led to situations where consumers at the end of the LV line had severe low voltage problems while those close to the transformer suffered from over-voltage. Needless to say no one paid any attention to conductor size or type. In 1983 some of these LV circuits were found to be over a mile long. CDC Attitude to Kaintenance 30. After 1973 when Government bought into the Company it was generally known that Government would eventually move towards total take-over. By 1974 the Company was In deficit. CDC therefore had nothing to gain by pumping money into VINLEC's operations. Their level of maintenance dropped to the bearest minimum; that being whatever it took to hold the system together. Lack of spares and proper materials led to the further deterioration of the T&D system. Twisted connections, bridged HV fuses and installation of rebuilt American transformers were some of the results. Hurricane Allen 1980 31. In 1980 St. Vincent wa struck by Hurricane Allen and extensive dAmage was suffered by VINLEC's T&D system. The T&D staff lumediatley set about the task of repairing the system. With material already in short supply, every bit of available scrap material was used in the effort to get the system operating again. Although this was necessary at the time it of course caused losses to increase even further. - 336 - Fire of 1280 32. So far, attention has been focused on factors affecting technical losses, but non-technical losses were also high. The fire of 1980 which destroyed VINLEC's commercial office contributed tremendously to the high level of non-technical losses. All consumer and meter records were destroyed. 33. Until 1983 there were still consumers connected to the system whose records were not re-established since the fire, and were therefore not billed. Some people seized the opportunity to become illegally connected to the system. Theft of service became more widespread since people were aware that no consumption pattern could be established. 34. Up to 1982 billing was done by hand and by the time the billing system was computerised that same year bills were five nonths in arrears. Although the computer system had the capability of identifying changes in consumption pattern, active monitoring of consumer accounts did not begin until 1984. Meterine Inaccuracy 35. A limited metering survey by VINLEC in 1981 indicated that losses due to metering inaccuracy was high enough to cause concern. However, no meter inspection programme was implemented until 1982. Before this time, once a meter was installed, it was not replaced or tested unless severely damaged or the customer complained about being overbilled. It is small wonder, then, that some old five and ten amp meters that were long ago damaged by constant overload are only now being removed from the system. Several meters exposed to weather or clogged with dust due to cracked or loose glass would also have contributed to los-es. POE LOSS SY Background 36. By 1979-80 VINIEC knew that a serious power loss problem existed, but with load shedding a common practice by then, the immediate concern was to secure funds to increase the diesel generating capacity. CDC in 1980 approached the Caribbean Development Bank (CDB) to finance a new diesel generating set. 37. However, CDB insisted that a study to determine the least cost proposal for generation be conducted. The study, conducted in 1981, had such terms of reference that it was able to investigate other aspects of VINLEC operations besides generation. This study identified losses as an area that required greater investigation. 38. In 1983 the Caribbean Community (CARICON) commissioned a power loss reduction study for St. Vincent and the Grenadines, using USAID funds. The objective of the study was to identify the sources of losses in the distribution system and to make recomendations for reducing chem to the optimu level. The study was addressed both to the effect of losses on the economy and to the financial implications to VINLEC of the resulting loss reduction programme. . 337 - 39. The study was conducted by Adair & Brady International Inc., Consulting Engineers operating out of Florida, USA. Findings 40. In 1982 the power system served 11,384 consumers. The peak demand of 5.5 MW occurred in December, and the total energy sold for the year was 21,778 MWh. The energy generated was 29,033 MWh, 11,027 MWh of which was generated by hydro plants and 18,606 MWh by diesel plant, with company's own use amounting to 652 MWh, distriburion losses amounting to 6,603 MWh or 23.3% of net generation. The year-to-date figure up to March 1983 put losses at 26.3%. 41. It should be noted here that the actual percentage loss may have been higher than that reported, since it was later discovered that some of the stations kWh meters were slow. 42. The approximate distribution of losses on the St. Vincent system at the end of 1982 (23.3% of net generation) as reported by Adair & Brady was as follows: a) Technical losses in: Step-up transformers 2.0% H.V. lines 1.6% Distribution transformers 2.4% L.V. services 6.5% TOTAL 12.5% (b) Non-technical losses total 10.8%, caused, in descending order of importance, by: Theft of service Inadequate billing Meter inaccuracy 43. The exact division of the losses among the non technical categories was not known but the impact of the first two items was clearly illustrated by Adair & Brady's findings during the field work associated with the study. 44. Nine consumers who were either suspected of stealing electricity or were not being billed were identified. Together they accounted for 3.1% of the system losses and represented an annual loss of revenue to the utility of over EC$400,000 per year. In comparison, the net income for 1982 was just over EC $800,000. Loss Reduction Target 45. A target of 7% losses by 1988 was set by Adair & Brady, the mix being 6.5% technical and 0.5% non-technical. Adair & Brady estimated that the planned reduction in non-technical losses would have caused a drop in peak demand of 290kW by 1986 and 450kW by 1988. The planned reduction in non-technical losses was expected to effect a net saving in 1986 of EC$976,529 and annual savings of EC$3,199,154 by 1990. - 338 - 46. Their proposed loss reduction programme required Afn investment of approximately EC$13.2m between 1985 and 1988; it was expected that of this amount approximately EC$11.9m would be recovered by 1990. !t was also estimated that savings to the consumer would be approximately EC$O.19 per kWh by 1990. Recomll2ndalions 47. The main actions recomended to achieve the proposed loss reduction targets were as follows: For reduction of technical logses 1. Install 300 kVAR banks of capacitors as follows: 7 in 1985; 3 in 1986; 1 in 1987; 1 in 1988 2. Install capacitors in future to maintain power factor established in 1988. 3. Break L.V. distribution into smaller sections by extending the H.V. distribution and using more transformers. 4. Apply new standards which incorporate larger conductor sizes and define maximum run lengths for L.V. distribution. For reduction of non-technical losses 1. Move meter installations of all Industrial and large commercial consumers to the exterior of their buildings, change meters to socket mounted kVA demand types. 2. Move some commercial and domestic meters to the outside of building each year, replace meters with socket-mounted meters. 3. Institute a policy requiring all new meter installation to be socket-mounted and conveniently located outside of buildings. 4. Institute and enforce a new meter sealing policy. 5. Survey consumers to re-establish billing data base. For Inroverment of Productivity 1. Construct new distribution centre, including offices, stores, workshops and adequate parking space for company vehicles. 2. Restructure the line crews to form a specialist pole-planting crew and separate three-man line construction crews. 3. Train line crews in the use of modern techniques; for example, hot line connection of transformers and fuses. 4. Obtain and use correct fittings and tools for alumimium to copper connections. - 339 Factorg Affecting Prgiet Schedulg Final Reprgt 48. At the very outset a number of factors contributed to delays in the implementation of a power loss reduction programme in St. Vincent. The first delay came in the delivery of Adair & Brady's final report to VINLEC; although the study was completed in April 1983, the final report was not received by VINLEC until early 1984, nearly one year later. Funding 49. A power loss reduction project was now defined for VINLEC. About this time, however, Government, on behalf of VINIEC, was in the process of negotiating funding for the Cumberland Hydroelectric Development project. A substantial amount of funds were therefore expected to be put into VINLEC by several loan and donor agencies. It was decided that these agencies should combine and make one input into VINLEC to cover the projects then pursued by the Company; namely Cumberland Hydro, Power Loss Reduction, Transmission and Distribution Extension and Transformer Improvement. 50. These extended negotiations contributed to some delays in securing funding for the loss reduction programme. 51. The Caribbean Development Bank (CDB) decided to fund the project, and the projects engineer was recruited and the power consultant selected in the latter half of 1985. It was expected that this would have been done in mid-1984 and that 1985 would have seen the project well under way. The project was therefore about a year behind schedule by that time. Procurement and Standards 52. The project engineer's first task on arrival in St. Vincent would have included obtaining and evaluating bids for supply of materials and planning a programme for the execution of the loss reduction work. At this time, however, material specifications and construction standards were not available. Standards and material and equipment specifications were discussed with the consultant and having come to agreement on what was required, procurement began. 53. It must be noted here that the final copy of the construction standards manual was not delivered to VINLEC until 1988. The requirements for the majority of these Standards were obtained from VINLEC technical staff and introduced with only minor modifications. Even after the delivery of the final copy several Standards had to be redrawn by VINLEC. 54. The first batch of meters ordered with project funds did not arrive in St. Vincent until late 1986. The line hardware ordered for the start of construction did not arrive until 1987. The line trucks, so essential to high productivity on the project, did not arrive until mid 1987, and capacitors, expected to have a significant impact on the reduction of HV losses, did not arrive until 1989. (The proposed locations and voltage setting for switched banks were delivered about the same time.) - 340 - 55. The lack of proper storage facilities in VINLEC presented another problem in the availability of materials during the first half of the project. Materials had to be ordered in smoll quantities because of inadequate storage space. As speed of construction increased the materials ordered were quickly used up and ever so often stocks of essential items were exhausted. In some instances the materials ordered were not entirely suitable. For example, the first batch of consumer meter bases ordered had non-tinned alumimum terminals. Use of copper conductor with these meter bases have resulted in serious corrosion problems. Initial Construction Activities 56. Before the arrival of any materials ordered under the loss reduction project, the project engineer decided to start reconstruction of the distribution system in the Murray's Village area just outside Kingstown proper, using VINLEC materials and the T&D Construction Crew. It was the first time that a scheme of this type was planned within VINLEC. The planning included mapping and voltage drop calculations. 57. The construction work in this area went very slowly, due to the inefficient work methods of the VINLEC crew and the number of times this crew had to be diverted from the project to do breakdown maintenance work. It became obvious during that time that there existed a need for skilled linesmen, more efficient crews and petty contractors. The Murray's Village scheme took over one year to build. A scheme of comparative size today will require four months at the most to build. TXaining 58. The staffing needs of the project required that the Company's staff be increased considerably. The staff required included linesmen, metering staff and meter inspectors. Workers skilled in these areas are not readily available in St. Vincent (probabably the case in most Caribbean countries). Workers, therefore, had to be recruited and trained. Most of the trainees were high school graduates and graduates from the two-year electricity or electronics programme of the St. Vincent Technical College. 59. The first batch of 12 trainee linesmen were recruited in April 1986; a second batch of 8 in January 1987. 60. Seven metering staff were recruited in October and November 1986, after the arrival of meters, and training of these recruits began in November 1986. 61. All training was conducted in St. Vincent. This was the result of a decision taken by VINLEC to utilize, as far as possible, its own resources; in any case, VINLEC considered that the overseas training available was unsuitable. 62. In November 1986 a foreign metering instructor was employed for two weeks to train seven trainees and existing metering staff. In November 1987 a foreign linesman trainer was employed to conduct training in modern construction techniques; his services were further engaged in October 1988. The latter found - 341 - it difficult to work with locals because he found them slow to respond at times, and they in turn often found his language offensive. 63. On-the-job training of linesmen was done during construction work, s0 that the speed of construction in the initial stages was not as great as would have been hoped. Delays in the arrival of training equipment and materials also contributed to a longer training period. For example, the line trucks were delivered late and training on their use and maintenance could only have been carried out after their arrival. Redistribution of Funds 64. The costing of the individual project components proved to be somewhat inaccurate in some areas. For example, in the area of L.V. distribution reconfiguration, the extent of the H.V. rebuilding required was not fully appreciated. Furthermore, in the rebuilt areas the H.V. system had to be extended, because of social and political pressures, to include new consumers. In this area, too, the extent to which petty contractors had to be used was not anticipated. 65. Funds therefore had to be shifted between some project components to take care of some of these dificiencies. The process of authorization of the shifting of these funds resulted in further project delays. PROGRESS AND RESULTS 66. When Adair & Brady visited St. Vincent in 1983 they estimated VINLEC's power losses at 23.3% of net generation. They suggested a target of 7% for a loss reduction project. During the development of the project the target level for reduction of system losses was set initially at 10%. This target has not yet been achieved, but the year-to-date calculation up to the end of May 1989 put system losses at 14.58 of net generation. In fact, the losses over the last six months have been just about the 12% level. 67. Peak demand loss reduction together with the resultant lowering of peak demand is difficult to evaluate. This is due to the direct relationship between load demand and improved voltage levels resulting from loss reduction activities. It is estimated that improved voltage has increased load demand in the order of 2%. In fact, VINLEC has recorded increases of up to 30% in some customers' consumption due solely to voltage improvement. 68. To date, some 8C$16.33 million has been spent on the power loss project (see Appendix 13). Of this cost, just over EC$6.Om has been indirect cost; for example, cost for T&D Centre, vehicles, training, etc. This leaves direct cost at EC$10.24m. 69. Savings due to loss reduction to date total approximately EC$1.4m. This means that of the EC$16.33m, only EC$14.93m has to be recovered. At present loss levels (14.5%) and present production costs, this gives a pay-back period of about 12 years for total expenditure and 7 years for direct costs. - 342 - TECHNICAL LOSSES High Voltaze Llnes 70. In 1982 losses in HV.lines accounted for 1.79% of net generation. By the end of 1988 HV losses ha4 been reduced to 1.66%, due mainly to the new 33kV line from Cumberland and reconducting of some llkV lines. 71. To date, some 32km of three phase and 19km of two phase llkV lines have been rebuilt. As an added benefit to loss reduction, the llkV rebuild gives VINLEC greater flexibility in system configuration. New ring circuits allow smaller areas to be isolated during planned maintenance shutdowns. 72. The feeders identified in the Adair & Brady report as having the most HV losses have been given priority in the rebuild programme. These are Sion Hill, Richmond and Belmont, in that order. CaRacitors 73. C I Power's analysis of capacitor requirements differed tremendously from Adair & Brady's recommendations. They estimated that 1050 kVAR of switched capacitor and 1200 kVAR of unswitched capacitor were required to provide the required power factor correction to the system. 74. As mentioned before, these capacitors, along with installation details, have been received by VINLEC within the past four months. Towards the end of May three banks totalling 750 kVAR of unswitched capacitors have been installed. Installation of the rest require pole changes at some locations, and the establishment of exact voltage levels on feeders once all static banks have been installed. The effect of the installed capacitors on the system is not yet known. However, it is estimated that the installation of all capacitors would result in a loss reduction of 0.3% of net generation and 2% of the system peak demand due to voltage Improvement. Power Transformers 75. The Cumberland project added five new power transformers to the system. This caused an increase in power transformer losses of approximately 0.15%. Not much can be done about reducing power transformer losses, although such losses will decline as the system peak demand loss is reduced. Distribution Transformers 76. In 1983 when Adair & Brady conducted their study there were 236 distribution transformers on the system providing a total capacity of 14907 kVA. These were mainly relatively large transformers of the high loss type, feeding extensive LV circuits. Unfortunately, 8 more of these transformers were added to the system, providing another 2120 kVA of transformer capacity before the switch to low loss transformers was made. This resulted in an increase in distribution traznsformer losses from the 1982 level of 2.4% to 2.9% in 1988. During this time 190 new type transformers (6315 kVA) have been added to the system and account for losses of only 0.4% of net generation. - 343 - 77. Economic analysis has shown that the minimum size transformer to be Installed on the system is 15 kVA. Recently, high voltage metering has been applied to three large consumers. This would cause a small reduction in distribution transformer losses. Low Voltage Lines 78. In 1982 the losses in LV lines stood at some 6.5% of net generation. This high loss was identified as caused by long lengths of LV lines and undersized LV and service drop conductors. 79. To reduce this high loss the LV distribution has been reconfigured using more and smaller transformers, and LV circuits have been restricted to maximum run lengths of 300m (1,000 ft). So far some 123 km of single phase and 6 km of three phase LV lines have been built under the loss reduction project. The rebuild has taken place in the high L.V. loss areas on the Sion Hill, Richmond and Calliaqua feeders. 80. The actual loss reduction achieved by the reconfiguration is not known. However, in the areas of Murray's Village and Glen it is estimated that the loss reduction achieved is about 0.17%. Using this as a base, it would appear that LV losses have been reducdd from 6.5% to about 4% of net generation. 81. In 1982 Adair & Brady estimated the level of non-technical losses to be 10.8% of net generation, the major components of these losses being theft of service, unbilled consumers and incorrect metering. The loss reduction programme had reduced non-technical losses to about 3.8% of net generation by the end of 1988. 82. The Adair & Brady report mentioned that VINLEC's meter inspection section was under-staffed. At that time only one meter inspector was responsible for checking the defects reported by meter readers, investigating consumers whose billing records changed significantly and attending to consumers' complaints requiring a field check. This workload was certainly too much for one person and many suspected cases of theft detected by the billing section were not investigated. This led to under-utilisation of the computerised customer management system and high levels of theft. 83. The meter inspection section was upgraded to a full crew (two persons and a vehicle) by the start of the loss reduction programme. One additional crew was employed under Power Loss. Up to the end of March 1989 some 11,861 consumer meter installations were inspected, approximately 77% of the total number of 15,500 consumers in St. Vincent. Of these, less than 900 meter installations met all the required standards. Tampered meters accounted for 91 cases. Some 25% of the meters surveyed required immediate attention. Theft of Service 84. Illegal use of electricity has been punishable by Law since 1951. In fact, the penalty then was very severe; a fine of up to $500 or two years in prison with or without hard labour. 85. In 1973, the 1951 Act was replaced by a new Law, wherein the penalty for stealing electricity was a maximum fine $500 or up to six months in prison. - 344 - The Law also gave VINLEC the right to recover the cost of electricity stolen and to disconnect supply from any consumer found using electricity illegally. 86. Prosecution for theft of service has proved to be difficult. VINLEC has therefore used the Disconnection Clause in the Law of 1973 to put pressure on consumers found stealing to pay the cost of electricity stolen. This approach has worked in 60 of the 91 certified cases of tampering identified during the meter survey. A total of EC$98,160, representing 176,617 KWh of stolen electricity, has been recovered since January 1988. 87. Also, in 1989 judgement was delivered against a large commercial consumer found stealing eleccricity in 1983. This theft of service was discovered during the power loss study. Disconnection pressure failed to have any impact on this consumer and civil action was pursued by VINLEC. In 1989 the consumer was required by the Court to pay EC$53,336 to cover stolen electricity and meter replacement. (See Appendix 20). 88. It must be noted here that meter inspectors found a number of other meters which appeared to have been tampered with. However, in the view of the Company's lawyer, they cannot be categorized as tampered cases unless tampering can be proved beyond doubt. Meter ReRlacement/Relocation 89. Meter replacement/rilocation measures were initiated early in 1987. It was recommended that priority be given to relocation of three phase meters installed on services to high consumption customers, where a defective meter would have greater impact on loss levels. The main objectives of the measures were: = to replace bottom connected meters with socket type meters, which are not only water-tight but difficult to tamper with; - to relocate meters from inside to the outside of houses and buildings in order to detect and prevent meter tampering; - to relocate meters difficult to read to a more accessible location and so avoid the non-reading of meters 90. To date, 35 three phase demand meters and 138 three phase non-demand meters have been relocated or replaced. This leaves only 12 three phase customers whose meters have not been relocated. The programme is still in progress and all three phase customers' meters should be relocated by end of July 1989. 91. Approximately, 1800 single phase meters have been replaced or relocated to date. 92. Initially, areas for single phase relocation work were chosen at random. .s re-confiZuration work progressed, single phase relocation was scheduled 4.or tie areas already re-configured. Within the past year all customez. ,tb vonsumption over 800 KWh per month were also targeted for speedy relocation. Zo date, 190 of these consumers have had their meters relocated. These include some of the larger three phase comsumers. - 345 - 93. Local viremen were hlred on contract to do the relocation work. Meters were installed by VINLEC crows. 94. The elements of the project geared towards productivity improvement have been completed and have made a tremendous impact on VINLEC's T & D operations. Office Sgace and Stores 95. Before the implementation of the power loss project suitable office space for technical staff, workshops and stores facilities were sadly lacking. T & D crews collected materials from the Stores building in the Kingstown Power Station yard. There was little parking space and, consequently, the delivery of materials was slow. In fact, if a truck parked near the gate had a flat tyre, all the other vehicles behind had to wait in the yard until it was repaired. It was estimated that each crew lost about an hour every day collecting materials. 96. The construction of the T & D Complex has provided comfortable offices for all T & D engineering and administrative staff. The complex also include adequate stores and vehicle parking space, auto repair workshop, transformer maintenance workshop and transformer storage shed, meter testing facilities, carpenter workshop, canteen, locker rooms and a conference room. Vehicles. Tools and Eauimepnt 97. Three modern line trucks and other transportation vehicles have been purchased, and though there are no spares vehicles, the crews are of such size that if one vehicle is taken out of service the crew members can be usefully distributed among the other crews. 98. Sufficient tools and equipment have been purchased to allow the T & D Crews to use modern techniques and methods ln line construction and maintenance work. 99. All of the T & D Staff have undergone at least one year of prescribed training. Training is an on-going activity in T & D and the available materials and facilities makes training so much easier for the trainees and the T & D engineering staff. granisational Structre 100. The acquisition of proper facilities, transportation, equipment, tools and training, along with improvement of engineering staff, has enabled the T & D Department to restructure Its organisation for maximum efficiency and control (see Appendices 15 to 19). - 346 - Meter Testing 101. VINLEC's meter testing facility was in 1982 located in the Kingstown Power Station compound. Meters were calibrated with the use of voltage and current measurements, instead of a rotating standard. Meters were only tested when a customer's consumption was in dispute. The test %laboratory' 'was a naturally ventilated room next to the carpenter shop. In this environment it was impossible to prevent dust from entering the meters. This situation was improved during the power loss study when the Government meter testing facility at Camden Park was loaned to VINLEC. This test lab has since been turned over to VINLEC and the equipment installed at the new T & D Complex at Cane Hall. 102. The meter testing staff comprises two lab technicians and two members of a field testing crew. Other Benefits 103. The line rebuilding conducted under the loss reduction programme would ensure that over the next 10 years the amount of line maintenance required on the VINLEC system would be greatly reduced. Also the reliability of the system would be greatly improved and the number of trouble calls would decrease. EXPENDI2ER 104. The estimated cost of most projects and the actual expenditure are often quite different. This was true in the case of the power loss reduction project. Initially, it was proposed that the meter test facility given to VINLEC by the Government be upgraded, and over EC$30,000 was allocated for this. 105. However, since the loss reduction project included the construction of a T & D Complex it was thought that meter test facility should be at the same location. The money allocated for the meter test facility was therefore shifted to the T & D Complex allocation. 106. Only 55% of the money allocated to metering has been spent. In retrospect, this may have been due to an error in programme scheduling, in that greate: emphasis should have been placed on metering. The figures will show that the gains from non-technical loss reduction have been tremendous. 107. The amount spent on purchasing tools and equipment has doubled the amount estimated, and is a reflection of how under-equipped the T & D staff was before the project began. 108. The cost of training may appear to have been over-estimated, but this is not really the case. Alongside the CDB loan, there was an IDA loan which included a training component; therefore, some of the training requirements for the power loss project was provided under training financed by IDA funds. VINLEC had also improved its technical staff to an extent which will facilitate more in-house training, thus reducing the aeed for more foreign lecturers. The amount of money spent on line reconfiguration has almost doubled the estimated amount. This is due to the fact that during the orignal study it was assumed that the existing HV system required only minimal work. As stated before, this was erroneous. In addition, the terrain in St. Vincent makes line building an expensive exercise. - 347 - 109. Interest costs were reduced, as the withdrawal of funds was delayed for several reasons, some of them mentioned before. 110. Appendix 12 shows that scarely any money has been spent on capacitors. This will soon change, however, as the capacitors have arrived in St. Vincent and installation has now begun. 111. The amount spent on engineering services was less than that allocated. This is because the need for engineering services declined as VINLEC engineering staff improved. MAINTAINING LOW LOSS LEVELS Technical Losses 112. The Planning Engineer in the T&D Department will be directly responsbile for monitering technical losses and maintaining them at the low levels established during the loss reduction project. 113. The elements of a control plan to maintain low technical loss levels would include the following: * Proper planning of new HV Lines including loss consideration - Monitor.ig of feeders to assess load and need for load balancing - Installation of capacitors to maintain established system power factor - Close monitoring of transformer loads and voltage levels on LV ciruits - Proper planning of LV systems including voltage drop and loss considerations - Continuing proper mapping of both HV and LV network - Carrying out load flow studies to determine optimu' system configuration and to avoid voltage and kVAR flow problems - Carrying out maintenance work to the same standard as new construction - Monitoring fault reports closely to Identify potential problem areas - Planning and executution of programe to correct all discovered defects and abnormalties - Strengthening the Maintenance division. - Establishing a record-keeping system for data collection Procurement 114. Attention must be focused on material procurement, as it has been seen that lack of proper material and equipment can lead to improper work - 348 - methods and faulty installations. This in time will result in loss of productivity and increase in losses. 115. The T&D Engineer will be directly responsible for procurement of T&D materials and equipment. He will ensure that: Minimum stock levels are maintained; Materials ordered meet VINLEC's specifications; Transformer bid evaluation take losses into account; Emergency materials are in stock during the hurricane season. 116. He will also keep abreast of new product information so that the best materials and equipment can be obtained. Non-Technical Losses 117. The non-technical losses on the VINLEC system have been reduced considerably. However, the project programmes have to be completed in order to reach ar close to 0% as possible. The Metering and Protection Engineer will be directly responsible for the execution of a control plan, designed to maintain low non-technical loss levels. The elements of the control plan are already in place. These can be identified as: Review of billing records Replacement/Relocation of meters Meter testing Inspections Meter reading reports Review of Billing Records 118. The computerised billing system used by VINLEC since 1982 is an adequate tool for proper consumer management. The system allows a consumer's accounts for the preceding 12 months to be reviewed. It also provides a list of possible billing errors or problems by comparing three months' billing and identifying, through print-out, any consumption pattern which changes rapidly (up or down) from previous months. The computer also processes meter records; the meter number, date of installation, location and condition of meter are all available on the computer. Of course, this information depends on input from meter installation crews, meter inspectors and meter readers and must therefore be constantly updated. 119. The computer also prints out meter defects by category on a monthly basis. The more serious faults can therefore be immediately addressed. 120. Identiflcation of anomalies In consumption pattern and print-out of meter defects must go hand In hand with inspection. All suspect services should be investigated and a programme drawn up and executed to deal with meter defects on a monthly basis. Meter ReRlacement/Relocation 121. Meter replacement and/or reloaction will continue until all meter installations on the VINLEC system meet all required standards. Once this has - 349 - been achieved close attention will be paid to metor reading reports to ensure that the system is maintained in th'.s condition. Mtger TZ2tinu 122. A comprehensive programme for field and laboratory testing of meters will be established and maintained to ensure continuous accuracy in measuring power generated and consumed. 123. At present all new meters are tested before installation and meters are field tested or laboratory tested as requested by the Generation and Billing Departments. The programme will be expanded, however, so that all power station meters, together with metering from industrial and other large consumption customers, will be tested once a year. Other consumer meter testing will be initiated on an "as necessary' basis, with p-iority given to suspect Installations where there is evidence of tampering. OCLUSIONi 124. VINLEC staff awareness of the benefits of reducing losses and maintaining low power loss levels has been heightened tremendously. This is as a direct result of the execution of a loss reduction project in St. Vincent. One can say with confidence that the loss levels on the VINLEC system will never reach 15% again. -3SO- A P P E N D I X Appendix Map of St. Vincent .................I...............e . Generating Capacity .................. 2 Transformer Capacity Total ............................ 3 Transformer Capacity New 00....... ................. ...... 4 Transformer Capacity Old ..... *....... .............. 5 Single Line Diagram 1984 ..6........................... 6 Single Line Diagram 1989 .............................. 7 Peak Demand 1980 - 2001 ...................... o... 8 Load Growth 1980 - 2001 9 Load and Demand Growth 1980-2001o..................... 10 Cumulative Twelve Months System Losses .oo..o.....o..oo 11 Allocation of Funds ...o.. .o.e oo.................... * 12 Use of Funds 1986 - present 13 Table of Extensions Required 14 T & D Supervisory Staff o.oe.o.o....o.o.o..........s.. 15 T & D Planning Section ......e.o.eeoo.o.ee... oo.o... 16 T & D Maintenance Section ..o.ooo..oo..........o....ooo 17 T & D Construction Section **.eoooo..o......ooo.oo.6*.. 18 T & e Outstatons 19 test Case ... 20 - 351 - a POW E TA?IOWl * SBSSTATIO) _,.. kV NV LN, --mm ii jtV NV MIe ml TOWNS tA N HG V - - I TRANSMISSION SYSTEM SERVUSLID b' R"s* E ' ST VINCENT ELECTRICITY SERVICES LTD. GENERATING CAPACITY STATION TYPE UNIT NO. UNIT SIZE *W8CLLED ___________ CA PACITY SOUTH RIVERS HYDRO NO.1 320 KW NO.2 275 KW ___________ NO.3 276 KW 870 KW CANE HALL DIESEL NO.1 1126 KW NO.2 1280 KW NO.3 1260 KW NO.4 3200 KW NO.6 600 KW NO.7 600 KW _________________ ___________ NO.8 600 KW 8648 KW . __~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. CUMBERLAND HYDRO NO.1 1464 KW NO.2 640 KW NO.3 840 KW NO.4 475 KW .___________ _ NO.5 475 KW 36894 KW RICHMOND HYDRO NO.1 650 KW _________________ ~NO.2 660 KW 1100 KW KINGSTOWN- DIESEL NO.1 315 KW NO.2 360 KW NO.4 460 KW 1036 KW __________________ - 1 _______ _ 16346 KW *Kilngstown power statlon Is near retirement. Only three of the four units are operatlonal and they are maintained for oemrgency vio only. SI.JLIc!31 aELTICHT -MVYIu UJNI;TED 8IOUr1eTu CihI F.t '3Av ic. 3H1IE8 WK tauSi tE MM TOMal. iva WAC gt 3u 'HE sIsr. 'i?1CfrE3 gasz,a,aaggaa=sZXR*ummgZ,sx aaZuuaaaz=amumat3 TOIL :SOO ;II.: 1.s o :so:o:o : o0 S a00 t10, 100 It 1 1 : 3 : 50so: & : lb:t. 01 0.) 2 I., : wsc-n S, A : AQ0IE .VA :01I ' 2. % IVA A :KIW KV IRIllA :IVA ;MA IVA MtA :WJk MK K; tA %ATMVA Mt': :tA MVA HRAMS I -- :~::~~ras-c&'uarug :01 2 :0:0 : 1 S10 1 S I : : : t : 1;:o: i:s: i : I: iO5E ;Ie: N12 0S:E S~32cAIwEsPahK-a1oIm 2 ; t: Sa: S: Ia a: o2: S1 01 01 91 Pll: 0: 3: FN:7 1 : I: :I ;2N :I332 02 :p ::5) c;a - FL:112 L 010 2 I 1 I I0 12 02 * I 11 : I I 0: I I 3 I: 0 0: 17: 1: 01 3 471 S2 : I: i t 3 TkiAo1 1 UIMEN76 PM 2 I I2 02 01 01 0 01 I 0 1 11 OS 02 0 40 1: 0: 71 4 2 0 0 I1 25I 7104 0: 01 It1 CDER TPR - I I 0 I I 1I I I I 0 1 I I 1 2 S I 0 21 0 1 0 1 J1 402 S; 7Is IS 32 I1 21 U2 32 01 0t :4toIIU. - CANlEC. | I1 a 2 I II I I 00 *1 01 01 11 01 0I1 3I I I1 13 1 2 0 10 1 4: 312I 1f:: 6 : lis) Lis dAlL - Sim OHL I I I I *: II II II I II I I II 1 5It 0: 01 I II 3 I 4 t. 10 I 0:I 14 I 0: 0: V 242: D:0 I _ flD1M.IIIIOmu - El I 11 so 22 11 I1 1 71 S0 21 21 II 3t 22S 12 : I 1 14 II 2 S 5: 2 0: so I I.t sotI i sI 1CAEhM I A - "am I I I I IV I t IE II 11 O I 1: 0 :) I I I It I 1 t I ff 4I I l t{fAi om - PA IIII I IOS 0t I II OS Ot S9 01 I 01 0I 01 1 1 3I IS I 12 114 : 10:0 : 5I 4: 114S OS a01 I -_ I (091X1 NEU CtII IIIII I I I1 I I I 0 01II 0 1 11 *l l 0 101 S *t 1 1 1 20t 3 : OS II II* 15 31*:l i t1TAL RISlWI I I I 1. IIS I I S It 1 1 1 t * 41 92!:I tt SI 22:1 29: I I 121 14b II II 2S25S U4S4 434 ItGUKl seht.h I ltAPAtIll 100 IUO ISM t1 l0 11 t tlM D 11 tl:200 160l S2100 Sl12 t2 lIZ UM 1450 I 25 I= St3 :1?M5I 20 11?5 S 2333 t m I so I sa sI tI I-- I.erWJM 100 IRaa AO INeR IU . MI FACT 1241E9 - 354 - A0endix it .. - - - - - ~~- - 4 - . - p- - a. .VD O. O. o |e FX- - - - - - - - - - - 2 a. v~~~v v ....a..- _a._ . _ ___ § -~~~~~~of .. _ _ _ _4 _s a._______i_ LI I a a -- F, - - - - - -1 r mE ~ ~ j AP Zg~ ~ --~ - -1- - - - - - - - X a _i ssI n I - IE - ; ii1IIu ,e- -- . - - - - - -a-- . n"ig -.E a.. _ . . . a|--_3 Ls R~~~~~~~~~~~~~~.a V.am E g g g -g- - - I- - - I - - - - - - - -wS" … - …-- - - 1- -1-1-1- il~ -3 - - -ll--- - - - - - - - - { ig~~ -i - - - - -I -1-- - -l - …~~------ - - - - i--Si![1!! - ----a - a .-. -. a - -- - o - - - - . - - - --- ~~~ - ---. - - - I- I- - - - - - - - - - - -- -.-. - - - - - - - - ----t. - -- ---- --- 'A… _ --ongoi S-I .~~~ ~ - ae _ _- - w 0~~4 _0 _P _4 _0 _0 s -u - a 5 --1-0-~~~~~~t - - - 1S -1---W3 }--|- CD lo - 1 1 i Pa SSE|--|- t 0 -3 r.. a -i-|-I.. - .tt. . . #,-nlul~' S1.; -f l 1'4 - -w I~~~~ '4E S. | M . | =-X - a-I c! -0 av~~~~~~~~a. -~ ~~ 556 ST VINCENT ELECTRICITY SERVICES LTD SINGLE LINE DIAGRAM 1984 U~~~~~~~~~~~~~~~~~~~~~~~SI S*Ue,E" N UNION RICHONO] , A . *~~~~~~~~~~~~~~PRING1 CANE RO CN 0H , rllrl'l 1ll 1ll , s"~~~~~~~laoou I . ~~LOWMAN HILL ON TAOS KIN( X - I I f s0 CALLIAOUA PROSPECT P TOW ER LOSS REDUCTION PhRLL TO FLOUR MIlLL Y HE I | O~~~~~ARIDEN POWER LOSS REDUCTION PROJECT - 357 - mndix Z ST VINCENT ELECTRICITY SERVICES TD. m . optL I04M4 ev'st0*4 oposeese ~~~~~~P#09*4an,ee I " "~~~S ,'.,. i ......~ ~~~~.e..,. Cevsspess. ..... , < ...-egg,.. ........ , "wewe ~ g,aee. 3g :P, eg. t...j^.~~~~~~~~~gpge. ..... " \ ....~~~~~~~~.. , . . Ic sesseses 410809#96-f~~~essem. g1,04t,g, lif$Ate aueIs ts~ ~ ~~~vlt 611fli eV*8w#u#0wl#t#w* gateso I tSsDSe ge ,o,*g . ........ . , w# tblll fermn.e tJ -l 11 ""@I@||t2"@r ^tt ewmvtiuts oOwj.oo I STV!NCENT ELECTRICITY SERVICES LTD. DEMAND GROWTH 1980 - 2001 DEMAND (10W) 20 6 I ..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ la~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~a 10~~~~~~~~~~~~~~~~~~~~~~ 0~~~~~~~~~~~~~~~~~~~~~~ o I I I I I I I I * I I m I t L I. 80 81 82 83 84 85 86 87 88 t90 91 92 93 94 96 98 97 98 9 00 01 YEAR beroes 1 POWER LOSS REDUCTION PROJECT I n an imMulm t73 w_mu l 51111 11 g - 11 II it It II ast sI t Ia I I. t nmuI a s m :1 UuurI a I I It m I : I It I I I It3:: _m m so) at .:it l tIm Ia:: 313 Ln :vot: mutt La I uta I iU:: _n s :t mu tt LN c ml1 O :It 1I4 I i : W. m n:: u :l - : LvIt ALM: iA. I m:2 t11 :11 : tl: 191 L It 1.11 I I O It i. I St1 I LI t I1 1 31 N I1 t9.0 1.1 t:1 12S I 1.22 t .1 : II 4 1 *1: SA X I ti:1 :1193:1 : 9.11913 1.92l.a.:i 11571 a.m ju: 3t et: .3£ i a. 211 3.10 : 1.: t 12.5t 1.3:1 3.13: 21.21:2 1.3: 9.11t 1.612 7.1 It mol : .9i: :a I Ln It 1."1 1I1 11 LO I 31I t3.1711 LOLA 0t1 L.51 H 11 1L5 I 0.3 3.01 :29.0IW A S .it 0.01i 5At23 tt10.11 :I 191:1 9.14: In"1 11 2. t 1.0I 1222 -1.312 3.311 911 -LIII 1 .01 391 0.012: 23.32 -LU1: 12.51 tN:: 29.911 t,.U 11 2.0311 :3.21 S 35: -LI I it 11:ag1 s 1.i: t lL * : Las 326a I12. a 2 1 1LS931 U Iu.nllt LS. I 1 5.0a1 : 29.2: ItS." I: £1.5 11.131 S1." I SL it 9.31: La: 3.101 1.92 111t 31 12.11: 112M I O." 41 La I 5 1 1SI U 11 9.011 u t 19.n111 0.31 4I I11 0IL2t 2A.931 IL%.9: 12:.5 10.9t35.11I s L a.it :12.31 AsLJ# I Ls : 9.73n: it19111 13.U11 U211 I .331 3.331 327n3I 33.3It 9.n 1 Olt L011 0.L57 4I1Q I L11 .212 Ld51 1.5I 9.5It: 1.1: 3I .0311 1.11U.101 1.251 IL.11It Ill tit 3.6 I Wu1 1 11 .I01 2661 W."t1 SLtI 013 LI11 0.391 101 LU ItSL3'l2 t10.01:: 11.32 UU51 01X 41.392 3 . .7 1 7.31it U.171 1.53? a3.12 IU:tl: 13: :.i5s1 LO 1I lU I 1a3£ 1 611 0.2 I 23.1211 LdI13 15. 113 3:1 I0.31 1.51 9I491 : u.nnIt L013:1 61.0: L33 I,10:1 It M. I I I to I I 11 I 11 I I 11 1 I1 t 1$ 1 11 I11 i 11 1 213t9MU 11.4t1 1 101 0.4111 s 31 11lo9 9.I: .al2 92a 11.311 0.6L1 NI 29.01 1 "9. 1 WN2 It 91 3LSIt 0L.5I 47t.11: 4.n1It 0663IL I7 0.5St . St I990it 33.59 I112I I AMOt IL 5 1913131 I L333 9.93 I DI X.01 0.91£ I II 55.* it26L152 1.211it 0.5 I 1521 L3.1 S33 5213 11.1211 13.01 9.4921 9.21It 211991 St f 1012101 I .LO31 Ike11 2 I . I La 11 13.3 I t 1.1 1 II i V 1 1 I 0 .011 3.52 113SIX I.11 it .111 Jl5.0 I £0.52 9.3 3l It 1199 I212.1011 u I LO311 31.611 3M3I L.il1 I.1 as ait.21 LitI I LO. it#LV.)211 l .a:t 0.51 3.512 S.Ut 13.91:2 1.U 4: 1..41 1.3t1*1 n012 It I93 3"S I . l6 11333I O t1111.31 351 LO 11 11.11 161 I1 It013 I .11 I I 1.01: u9.13 I .L13:2 LS5 L.5S1 t1911 I.9"I 1.13 It ALMI 11.11 .041:3 ItI1991 N 9l1 1 250 I .L 11 PA1 30N2 1 2a 1t1 St U I1.703I 1.21t I 1 9ILO tl 33 I 1.29321 L0 t 2.511 11.731 IL911 t1.2921 :t 3I 122 I 99l :1 it 1" tl IU I MIS I L* 11 3BA I t01 4X 11 2Vd 3 d I LIS 11 1.3 1 1218 IOti W1 LO It LS 2D 11U It IL IU" It LU :t W I ILN I L Itt 11392U n .u4 191 "1£ 1 OI 1t 921 9.2513 SO I H I LO 11 1.431 U I LO1 46.1£ LO1It S 2 LS 11 RA 1 l.1 U11 L.It1 191431 U3.l I.U:13t 11 99121 3."16 2S19 9.011 IL% I 2139 LO 11 22.11 193 3U33 3.63 t3 9 L 5..1111 5LItL 11. It Ll Nn I11 4.IW1 .LU:11 11191 1 .31 213t LO3 1 1 LLbIt3311 ILSt 29.29 11 L 3111 3.3 3. U II 1.1 3 3 3 1. .111 It L.Ut U .UI U : & It 611t I1" 31.5 Qb li b t11 3n I lni I LO 1l1 N U I 1.3 t..11 I.t I I1 O .33lU IL I U 3.51 2.312 93.9 I 9.011 an0113 11 U.01U 11.n1 Ia012 I11499to 3.?#I 1223 I .011 JLU1I 312£ I LOI 1.991 Xt 16.011 L Los£ 1 t 9 21.1a i &M :1 LS1I LS1St 91.N1101.16lS 5.3111it.U11I11.012I5&3211 It30111 t VA1 331 9 .L1O ItU 1 au I LO 11 9.0 I X 3 9.111 1O1 3: 9.32£t 91".I 1U.1 it 3S.1 2.I LSIt .9L I l." 3. & It n.n 311.11 L5 111 1136113 3Ut. 31)13m 9.131 t.1 We Is: Ur Pna.1 t 31 bt111 a3 I0 11 U£ 9.32 1.91£ 1912 a.51t a 2itt.922:1uan U an121 11.01 1n.32 5.n IS I ST VINCENT ELECTRICITY SERVICES LTD LOAD & DEMAND GROWTH 1980 - 2001 120 40 100 0~ ~~~~~~~~~~~~~~3 80 N L D A;AR L ~ ~ ~ ~ ~ ~ ~ ~ _ 80 8182 8384 8688 87 8889 9091 92 9394965969798 9900 01 YEAR |- DOMESTIC + COMMERCIAL INDUSTRIAL 0 ST. LIGHTING TOTAL SALES TOTAL GENERATION POWER LOSS REDUCTION PROJECT I CUMALATIVE 12 MNTHS SYSTEM LOSSES AS A S OF NET GENERATION 25 24 23 22 21 ' 20 - L j 19-.- *.18 9 17 16 14 13 e * 11 stijl l ff11 ps lilit mit,,, I _ JFMAMJ JASONDJFMAMJ JASONDJFMAMJ JASDJFMAMJ JASD I 86 I 87 1 88 I 89 MONTH/Yf - ACTUAL LOSSES + TREND ALLOCATI'JN OF FUNDS :: : liD : 293' ; 95! ?" : : ta ;IL :oi 1 10 UiE E* E T:si,; 3ili :::.ome :::::::: I1?.(T6P.; : I : ! 2.4. ! 77.3.31 :: :: 12 C O cITM : 72:..? : 4:077Z. :92 I4.02 I l425l.! : 2.2.! 1:2 5icaru usE : IS542.04 : 1e.31 21424. I 65190.4: 40'4;5.3I : tt htt9 tCIS!KtIS I :!S.r 1 : 195.;: 0!3v.. I:2:4. 347514.4 2r.s tlC:7.U ':. .:': :7!31.s: I!749P..1 :: 3%?.;o, :: 22 I : : :I UL5 CWETU 5C l : , 1.43.13: : ; 46.I3 :: tICS70.1: :2 siwnuqv I :2443A I 013.4 531:4.09 : :: : 1: PVC1a2 I 913324 : W6.1 1 1132064.95 1 413747.o I 3141941.4 :: 11 ftty CatrKts I 1100.0 I 45.31 : 179441.95 I 3j1. 1 :1ti*)..2 :1 11 11 1 * : : :: : ::eLSwU1fftES 0 1 1 t I I 2:: 1: 461unqse 2 43151.2 : I5101.10 1 14211.7 : 41402.2 I 314311.16 2: Pvtra,nn I *72342 I 244 .4 1*9213.62 5392.l3 t I1113.11 4 : : tty C..... I ae.. : 4t43.3: 2040.: 2N4757. 7.o 1144.11 :1 : I WX I :1 :, : *:IgISi I r I : I : :: ): It etity CaIrats I I 13.06 200.00 2139.0ft : Me_ Li CNRIIVS 1 10zt14m1 1 MIN I4t V1 27t1.79; It I I I I 2 : 21 IIIL ML IID3CIW I 2 2. 446770.1 1 2320 43 2 37 392 IM 22. : ::4113 It I I I I5 t1 II3 , lUlLB I 29vA. I "5714.4 I 2553 I 14412.0 2 3922.1 it 043149.3 11 I I 1 1 1 : 2 IIsI FUE t3102:.n3 2 9657.s001 59153.00 2 720.1 570.21l II I I I 2 2 :: :: 1t2 111 t 7 I721 I 37.2 I 233i43 2 1 91142.I4 Uo 2724912 22 II I I I I :i 110 1I111 1 2" 1 157ilU 1i 1155.2 1 1 :2 21 N II I I I I 2 I 1: 1 II7 MIL I T135.% I 6194911. 19360.191 I 125341.1142 5399.SWAN 11 I I a I I t 1i1 IIII LU FME toI CEI I 17740.0 I I I 177150.0 I. 202346.4I1 1I I I 1 I 1 II II 11 I 1 1cR I 141,44: 2:340 39.2 t Sl 532391.3 U 00244 .4 II I t I I It 1: 1111 1311 AVS? I 2 .011 3513.11 6124422 2211.942 7555.44: U 342.32 it I: I I I it :I II 231135? 1 2 5.n I VW4.24 I 61nn.33 I 116157.0 1 3 S131000.00 1 : 22 2 I 1 I it 21 it Firm Dow1 I 97.61 2 24320.412 2552 1 271:3.61:1 12 2 11 I I I I 21 1 I: *h1CL1 1 142.0 I 1111.7 I: 1410.11 13740.09 1 22 it f 1 7 .I W $ *0 : I2.|2: 241I.92 I 6 21A ::122n- :2 1014, 2E I5*731.tl: 024493.44 t 7592.32 2 161414 2 11312.61 :: II30000 o ST VINCENT ELECTRICITY SERVICES LTD. POWER LOSS REDUCTION PROJECT 9, 235Ts s * i 1'5 l~~~~~~~~~~~~~~os ACTUAL COSTS ESTIMATED COSTS 0 ST YINCENT ELECTRICITY SERVICES LB. POWR LOSS REDUCTION PROJECT. : | 1986 1987 1988 1989 TOTAL ': ESTINATED COSTS .: :: __ _ - --- ------------___ _____ ___-_ ::A. NON TECHNICAL LOSS : 15964.73 305390.66 847546.45 365225.77 1534017.61 :t 2782967.70 Il 219. TECHNICAL LOSSES 1 293291.95 1577796.75 4081555.34 1444127.07 7396771.11 ' 4559239.39 I! t: ' at t U'C. T & D CENTRE ' 1185295.92 1711054.70 1025454.42 29791.53 3952396.57 tt 3806851.60 '1 ":D. TRAIN. EQUIP & VEHICLES 1 187165.06 64B999.80 52913.62 0.00 889078.4B 1: 924059.59 :: OtE. ENSIN. & PROJ. NAAGEENTI 574525.9q 446594.90 264213.96 22611.94 1307946.49 12 162682.72 Ii :"F. INTEREST N ISC. : 31405.93 334066.63 890279.53 660.61 1246411.6 t' 3131000 Ii iIs I I'l 1: :: TOTAL 1 2287549.19 5024693.44 7151962.32 1 1962416.92 116326621.86 t: 16B30000 :. :' I : : t a t 'a IS 8 I t S I a. a.~~~~~~~~~~~~~~~~~~~~~ - 365 - A,Dnefdix 13 Page 2 of 4 ST.VINCENT ELECTRICITY SERVICES LTD. POWER LOSS REDUCTION PROJECT Millions 8 6 4 21 1988 1987 1988 1989 TOTAL YEAR - NON TECH TECH T&D CENTRE - EQUIP MNQMNT D INTEREST ST VINCENT ELECTRICITY SERVICES LTD POWER LOSS REDUCTION PROJECT TaD CENTRE 34% TECH 31% SDD@% -~~~a NON TECH B | INTEREST IS 1986 EQUIP 18% 1987 U. ST VINCENT ELECTRICITY SERVICES LTD POWER LOSS REDUCTION PROJECT -.IsI . s t7~~~4% to% , / t2% , 1989 TO APRIL I2t 1988 I!t TABLE OF EXTENSIONS REQUIRED I - ' PCNBDILR EII b.S II ' 1W.-1 smsB 'SS | | a ||kM~~ccm . HV|I. LVl.| HV| 1|LY | a2/39 | 124IU 10 .5 10 1 Wip VIllA 18 .2 .7 1 4 4 10 2 In fto. C14I 31 .75 L50 15 10 20 3 In Pv to-mSm 3ltE ) 28 .75 1.75 15 10 25 3 8d3 Vt 89 UM?tSWI Nm ) 17 .9 2.25 18 15 30 2 Sd3 t 89 R.UIROI FM 22 .5 1.50 10 10 20 2 I Pro tf1siUiA 36 1.00 .00 20 20 40 4 Sd4 'r 89 OARSamE 17 .3 1.05 6 6 35 1 Ctipl&eb UPl& D1Cl4o fiLB - 2 .2 4 (bple3 a111 . 28 .75 2.50 15 15 35 2 Cmpleted LAJ 17 l.9D 30 Oxiplet4d WLuptE- 14 1.50 30 iTiplete, QpAS I0AR 21 .5 2.00 10 10 30 2 S 4 rt 89 (IUW1U'tI 28 .5 1.50 10 10 23 2 I 1 rt 90 5 .5 1 10 atd 1 t 90 SEUN Vol" 12 .4 1.05 8 8 13 1 (bpeted e '1Hm (A 7 .5 . 10 1. td 4 Qrt 9 LFF J. UJES 13 .1 .6 2 2 10 1 Scd4 )t d9 IEJ G;W 17 .5 10 2 ed4 Ot 89 RImi IUL 8 .75 1.'= 1I 15 10 3 (btplebec lC(P cU= 7 .75 L00 15 15 5 3 in rg3 FPANU ROM 5 .3 6 1 i Epros [El:ITB]U 12 .15 .75 3 3 12 1 In E . [TM 7mL 13 .3 1.05 6 6 i5 2 9I1 Or ±9 MAIR 26 .6 2.G) 12 12 40 2 Umpletsi c1BALmUQA 30 .5 2.00 1 10 10 3D 2 Ctnpleted KINSJM 10 .5 5 in nrogre; ¶UiML 455 10.2 33.05 3 205 181 440 48 T&D SUPERVISORY CHART J F HUGGINS I GENERAL MANAGER R T DINNICK T&D ENG L MORRIS 0 SHILLINGFORD| W JOHNSON TECHNICAL l PLAN. ENG | CONST. ENG. T&D SUPER CLERKS F PEREIRA B WILLIAMS E WILLIAMS F BIBBY A.PLAN. ENG. A.T&D 3UPER. TRAIN.ENG. TECH.CLERK P SOLEYN K BOWMAN A.PLAN.ENG. PROG.CLERK T & D PLANNING SECTION L MORRI8 PLAN EN . F |ERrR P SOLEYN l A.PLAN.ENG A.PLAN.E G.| J BYRON - 8 ROWN ORFTSUN CHRTHAND N RSON8 INSPECTION l RELOCAKTO! CONNECTIONS METER/TEST A. DRFtSMN $RSV TRNSFRS FIEL/ITEST O I LNSMANI 1LNSMANI I ELECT 1 1 ELECT II 2 LNSMAN II 2 LNSMANII I ELECT 1I I A.ELECT 2 i.LNSMAN 1 A.LNSMAN 2 A.ELECT O t1HANDYMAN O0 _° ro Lo -0 Lo~~~~~~~~~~ T & D MAINTENANCE SECTION W JOHNSONF r&2 SUPER. IE WILLIAMS TRAIN. ENG. W LYTTLE R KNIGMTS CHRGHAND CHRGHAND _ I .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~' MAINTENANCE TREE COMPRSR TRANSFORMER STREET TRlMANNG |L tI SHTI NG .1 SN.LNSMAN 0 O O O O 1 LNSMAN I 0 tLNSMANI 0 t1LNSMtAN It _O 0 -1LNSMAN It O 2 A.LNSMAN O 0 1 A.LNSMAN 1 A.LNSMAN I HANDYMAN 1 HANDYMAN 1 HANDYMAN I HANDYMAN I HANDYMAN CLL O LABOURER 0 _0 CAME= HALLx T &D CONSTRUCTION SECTION C SHILLINrAFORD CONS T. ENG. 8 WILLIAMS A. TO SUPER CHRGHANO I SN LN8MAN I 8N LNSMAN I SN LNSMAN . O O O O 1 GANGER 0 O I LNSMAN I1 I LNSMAN II 0 0 1 AINSMAN 2 A.LNSMAN 2 A,LkSMAN I TRIMMER 0 I HANDYMAN I HANDYMAN 1 HANDYMAN 2 HANDYMAN 1 HANDYMAN ° ° ° 0 1 LABOURER I . -. -., ---_ - .- , - . . . , _ _ , , . = E~~~~~~~1 T&D- T & D OUT STATIONS W JOHNSON I~~~~~~~~~~~~~~~~~~~~~~~~' IE WILLIAMS . ~~~~~~~~~rRAIN,.EN6. CUM13ERLAND SOUTH 8EOUIA UNION CANE HALL RIVERS! ISLAND -., STATION W HARRY VkCA NT STATION R TRENT ENGINEER CHRGHAND CHRHGANO SUPER. OPS.SUPER ;° ,0 . I LMANIf 1 LNMAN I 1 LNSNAN I 03 1 A.LNSMAN 2 A.LNSMAN I A.LNSMAN I A.LNSMAN 3 A.LNSMAN 1 LABOURER 0 IHANWYAN O O O1 tLABOURER I LABOURER I LABOURER O; - 374 - &Pendix 2Q Page lof 8 IN THE HIGH COURT OF J-USTICE SAINT VANCFNT AND THE GRENADINES SUIT NO. 184 OF 1983 BETWEEN: ST. VINCENT ELECTRICITY SERVICES LIHITED PLAINTIFF AND MORRIS AND COMPANY LIMITED DEFENDANT SUIT NO. 186 OF 1983 BETWEEN: MORRIS AND COMPANY LIMITED PLAINTIFF AND ST. VINCENT ELECTRICITY SERVICES LIMITED DEFENDANT Mr Anthony Crick and Mr Keith Sutherland for St. Vincent Electricity Services Limited (VINLEC). Mr O.R. Smith Q.C., Mr Dave DaSilva with him for Morris ComRanv Limited. (December 8. 20, 1988: January 12 IZ, 1988 or 1989) JUDGEMENT Joseph Carlos On 4 November 1983 these two matters were by Order of Court consolidated to be heard as one and in them St. Vincent Electricity Services Limited "VINLEC" claims that Morris and ComDany Limited dMORRISS owes VINLEC $53,336.70 being $52,811.70 for electricity consumed for the period May 1981 to May 1983 and $525.00 being replacement cost of one meter. Morris alleges that on the 28 June 1983 VINLEC disconnected its electricity supply and thereafter has withheld from him a supply of electricity to its premises. As a result Morris suffered financial loss and damage and continues to do so and claims: 1. A declaration that VINLEC is not entitled or empowered to disconnect the supply of electricity to the MLorrig' premises at Ba Street in Kingstown. 2. An injunction to restrain VINLEC ... from disconnecting the electricity supply to MorrLs' premises and/or in any way from interfering with and/or depriving Misxx1a of a supply of electricity to its premises and/or withholding from Morris a supply of electricity to its premises. 3. An Order that VINLEC do forthwith reconnect the electricity supply to Moris, said promises and cause a supply of electricity to be restored to MHrris' said premises. 4. Damages for breach of contract and/or for trespass to goods. 5. Further and other reliefs. 6. Cost - 375 Angendixa2 Page 2of 8 VINLEC admits disconnecting Morris' supply of electricity but asserts that it did so by virtue of the powers conferred on it in Act No. 14 of 1973, Norris having refused to pay the sum of $52,811.70 in respect of electricity consumed by Morris during period Nay 1981 to May 1983 as per bill submitted by VINLEC to Morris on 14 June 1983. VINLEC also alleges that between 26 May 1983 and 27 May 1983 VINLEC discovered that the meters on Morris' premises were tampered with so as to prevent them from correctly registering the quantity of electricity consumed. Morris, in what can be described as a bare denial defence, simply denies owing VINLEC the sum of $53,336.70 or any sum at all. With leave of the Court granted to VINLEC during the hearing of this matter, VINLEC amended its statement for claim in Suit No. 184 of 1983? to specifically plead the fraud alleged by them in their defence to Morris Suit No. 1J6/1983 i.e. the allegation that Morris tampered with VINLEC's meters thereby preventing them from accurately registering the electricity consumed by Morris. To this Morris filed an amended defence denying that it fraudulently consumed any electricity generated by VINLEC. I now give my reasons for granting this amendment. The application was by VINLEC to amend its Statement of Claim in Suit No.184/1983 to include a plea of fraud against Morrig: This appliction was made after the case for VINLEC was closed and as the witness Dick was about to continue his evidence in chief after the adjourned hearing. Learned Q.C. for Morris objected to the amendment being allowed on the grounds that what we were dealing with was a consolidated action and that what was now being raised could not have been said not to have been apparent to VINLEC before this time. Hr. Smith contended that the introduction of a charge of fraud at the stage was too late. Mr Crick in reply argued that the pleadings, especially the defence filed by VINLEC in Morris' Suit No. 18j.1983, the affidavit of lJhn Hazell, and indeed, the evidence of John Hazell relate to an issue of the fraudulent tampering by Morris of VINLEC's meters thereby preventing them from correctly registering the quantity of electricity consumed by Horris. He contended that the issue of fraud was the real controversy between the parties. Having perused the pleadings in this consolidated matter and, bearing in mind the affidavit of John Hazell and his evidence led in this Court, I agreed with the contention of Mr Crick that the real controversy between the parties is whether or not Morris tampered with the meters as alleged by VINLEC for the purpose of defrauding VINLEC. To my mind, this was the basis for the respective claims of both parties in their different suits, this having been disclosed in the pleadings since VINLEC filed their defence in Suit No. 186/1983 in September of 1983 or for that matter since John HBazel swore to an affidavit in that suit in July 1983. My view is that what VINLEC now seeks, is merely a tidying up of the pleadings to include therein the specific plea of fraud, such an allegation of necessity requiring to be specifically pleaded (See 018 R 8 of the Rules of the Supreme Court (Revision) 1970. Having so found granted the application of VINLEC for the amendment as prayed for. I also ordered that VINLEC pay Harris all costs thrown away as a result of the amendment. And, upon the application of Mr Smith, I adjourned the matter to - 376 - A^Rfndix 20 Page 3of 8 give Morris an opportunity to file an amended defence if necessary within 14 days from the date of that order. In granting that application the Court had foremost in its mind the general principles for the grant of leave to amend as is set in the Supreme Court Practice of England (1985) Vol. 1 Page 340 at para. 20/5 - 8/6. This Court recognised the guiding principle of Cardinal inportance on the question of amendment that generally speaking all such amendments ought to be made for the purpose of detemining the real question of controversy between the parties to any proceedings or of correcting any defects or error in any proceedings. See Baker Ltd. V. Medway Building and Supplies Ltd. (1958) 1 WLR 1216. In Shoe Manufacturing Co. V. Vultrain (18%6) 1 Ch 108 at P112 A. L. Smith L.J. expressed Oemphatic agreement' with Bowen L.J.? in Cropper V. Smith (1883) 26 Ch. D 700 when he at Pp 710-711 said: 'It is a well established principal that the object of the Court is to decide the rights of the parties and not to punish them for mistakes they make in the conduct of their cases by deciding otherwise than in accordance with their rights ... I know of no kind of error or mistake, which if not fraudulent or intended to over reach, the Court ought not to correct if it can be done without justice to the other party. Courts do not exist for the sake of discipline but for tShe sake of deciding matters in controversy and I do not regard such amendment as a matter of favour or grace. It seems to me that as soon as it appears that the way in which a party has framed his case will not lead to a decision of the real matter in controversy, it is as much a matter of right on his part to have it corrected if it can be done without injustice, as anything else in the case is a matter of right." I can find nothing in the proceedings before me to suggest that the introduction of this plea at this late stage of the proceedings is fraudulent or intended by VINLEC to overreach. It appears to me to be a blunder on the part of the Solicitors for VINLEC, a blunder which was neither fraudulent nor intended to overreach. As I said earlier, it appears to me to be a mere tidying up of VINLEC's pleadings and I do not see any injustice to Morris in granting this amendment. indeed, from all that I've stated above I am satisfied that justice required that this amendment be granted. The entire trial up to the time of this application went along the line of this plea of fraud. In Atkinson V. Fitzwalter and others (1987) LALL BR 483 it was held in the Court of Appeal of England that the general principle to be applied in considering an amedment, however late, was that it should be allowed if justice required it, provided the other party could be monetarily compensated for any inconvenience, and tIe fact that the amendment alleged fraud was not of itself reason to refuse to allow it to be made. I am of the view that the amendment merely seeks to clarify the existing issues in dispute and does not raise any new issue for the first time. I therefore do not see that the allowing of this L:endment will impose any new issue for the first time. I therefore do not see that the allowing of this amendment will impose any strain or anxiety on i2rria. We are not dealing here with personal litigants but with business corporations and Morris is not - 377 A2onni.x 20Q Page 4 of 8 being called upon by this amendment to face any new issue: (See Ketteman VHansel Properties Ltd. (1988) 1 ALL BR 38). For the above reasons, in the exercise of this Courts Judicial discretion, a discretion guided by my assessment of where justice lies, I granted the amendment in the terms as aforesaid. I would now proceed to deal with the substantive matter. John Haell, a qualified engineer since 1971 and Manager of VINLEC since August 1981, testified on behalf of VINLEC. He was cross-examined at length in a searching cross examination by Q.C. Mr Smith and having seen and heard him I can find nothing to fault his testimony and I accept it on a preponderance of probabilities and being true. From his evidence I make these findings of facts. VINLEC is a private company incorporated by the Electricity Supply Act No. 14 of 1973. Its principal function is to supply electricity to consumers in the State of St. Vincent and the Grenadines. MsaoLs is one of VINLEC's consumers. Morris operates the business of a bakery. supermarket. grocery. hardware And a luber var. The supply of electricity to those different businesses is by VINLEC and there are three meters installed in the building at Bay Street. Kingstown, for the purpose of registering the amou.-at of electricity consumed by Morris, upon a reading of which by VINLEC. Morris will be required by VINLEC to pay for such consumption. In the year 1983 VINLEC realised that for several years they were experiencing a loss of more than 10% between the power generated by its utility and sent out to consumers and quantities of power recorded and billed as having been used by these consumers. This was abnormal so they undertook an investigation as to the source of these losses with a view to reducing them to accepted levels. As a rule of thumb such losses should occur in two areas. Technical losses ocurring in transformers, transmission lines and distribution systems and non-technical losses relating to deficiencies or inaccuracies in metering or diversion of power from being registered on meters. The purpose of the investigation was to decide what specific measures VINLEC should take to generjlly reduce non-technical losses. The investigation was carried out by a consultant together vith staff from VINLEC, a metering specialist from Barbados and certain technical and Government Staff. In the process of the investigation of non-technical losses, a number of accounts were chosen at random for investigation as to whether the metering of electricity used by these accounts were accurate or if not accurate to what extent. One of these accounts was Ngrri'. On 26 May 1983 the investigation team left for Morris to look at this particular account and late that afternoon the team made a report to John Hazel. As a result Hazell went to Morris' premises where he met the Consultant, the metering specialist and the other members of the investigating team. - 378 -£endix Page5 of 8 On an interior wall in the premises inside the building adjacent to an enclosed area used as a Supervisor's Office, Hazell saw the three meters. The meters were opened in his presence and he saw two pieces of copper wires which were foreign to the construction of the meters inserted in two of the meters, in a manner which would prevent the full electrical energy entering the premises from being registered on those meters. In his presence, the copper wires were taken out of the meters and handed to him and he produced them in evidence at this hearing. Hazell indicated his find to Pat Velox whom he was told was in charge. The witness Hazell then gave a physical demonstration to the Court with the aid cf an old meter as to how the inspection of such a piece of copper wire in a meter can prevent the meter from registering the full amount of electricity consumed by the consumer. Based on this witness' expertise, this Court accepts his evidence and finds as a fact that current can be diverted from being registered in the meter by this process. I also fi . as a fact that the pieces of copper wire produced in evidence by the witness Hazull were found in the meters by VINLEC in the circumstances as decribed by jHazll. I also find as a fact that the meters in which these wires were found were located on Morris premises with MoXrjs having effective custody or control of them. The evidence of Alfred Dick and Frederick Richards, the Assistant Manager and Supervisor respectively of Morris, adequately supports this finding of mine. These witnesses testify that these meters were located in an office which they occupy in the hardware section of Morris, that the meters are at eye level with Dick when standing and that no one can enter that office without their permission. Also that only VINLEC's employees would go to those meters and that was for the purpose of reading Having made these findings of facts, I hold as a matter of law that VINLEC has satisfied this Court prima facie, that Morris had without legal right, wilfully prevented the meters from duly registering the full quantity of electricity supplied by VINLEC. To support this finding of mine I refer to the Electricity Supply Act 1973 (the Act) at S 19(3) which states as follows: 'If upon any premises or land in the occupation of a consumer having effective custody or control of a meter or installtion there is connected or adjacent to any electric line or meter any wire or device capable of wrongfully abstracting, diverting, consuming or using electricity or of preventtng any meter from correctly registering any quantity of electricity supply by the Company, the existence of such wire or device ahall be accepted by a Court as prima facie evidence that such consumer bas without legal right abstracted or diverted electricity or (as the case may be) has without legal right wilfully prevented a meter from duly registering any quantity of electricity supplied by the Company." VINLEC having made out this prima facie case aeainst Morris, I do not agree with the submission of Q.C. Saith that it is not for Morris to prove that they did not put the wires there. My view is that the prima facie evidence having been produced by VINLEC a burden then lay on the shoulders of Morris to account for the presence of the wires in the meters. I have looked in vain throughout the entire evidence to see if this burden was discharged by Morris - 379 - Aggendix 2Q Page 6 of 8 on a balance of probabllities and I can find not a scintilla of evidence in this regard. ALI I can find are veiled lnsinuations against VINLEC without any proof, by Mr Smith during his cross-examination of VINLEC's Manager, Mr John Hazell. Indeed, I find it passing strange that Pat Vel2x, a Director of Horris would remain silent and show disinterest when confronted with the evidence of the wires at the very time when they were found in the meters. One reasonably would have expected some form of protest or some show of indignation from her. But her evidence is that she remained silent and showed disinterest. Having regard to these observations I hold as a matter of fact that Morrig has not discharged the burden placed on them to negative the prima facie evidence against them on a balance of probabilities. I therefore find as a fact that Morris, by means of copper wires inserted in VINLEC's meters on MorLij' premises, had without legal right wilfully prevented the meters form duly registering the full quantity of electricity supplied to them by VINLEC. The next issue to be decided is what was the loss suffered by VINLEC as a result of tampering with the meters by Morris. Here again I accept the evidence of Jioh azell that such a loss can, on a balance of probabilities, be quantified at $52,811.70. His evidence an this aspect which I accept is that VINLEC looked at Morris' pattern of consumption from their records and concluded that the diversion started from May 1981. These records have been produced in Court and, in my view, they support the conclusion of the witness and I so find as a fact. They were admitted in evidence by consent and the accuracy of their contents was not challenged in cross-examination or otherwise. I accept Hazell's evidence when he said VINLEC examined Morris' records prior to May 1981 and their assessment was that for six months or so preceding May 1981, M8rris' average consumption of VINLEC's electricity was approximately 11,000 units per month whereas, its average consumption following May 1981 up to the time the diversion was discovered was of the order of 6000 units per month. The records of VINLEC produced in Court support this examination of VINLEC and, from this witness' evidence and the documentary evidence, I accept and find as a fact, that the loss suffered by VINLEC as a result of this diversion of current was in the vicinity of some 5,000 units per month from May 1981 to May 1983 when the diversion was discovered which, as above stated, I would quantify as $52,811.70. In arrivlng at this conclusion I take into consideration the evidence of all the wLtnesses who testified ln this matter especially the evidence concerning the different appliances that needed to be served by electricity and the evidence of DlUk that with twice the amount of appliances now in operation, the average consumption of electricity is in the 'cinity of around 22,000 units per month. Having quantified the loss at $52,811.70, I find as a fact that Morris owed that sum of money to VINLEC. I find as a fact that VINLEC informed MorrLs of this debt by letter dated 14 June 1983 and requested payment by 27 June 1983. VINLEC, receiving no response fro Morris by way of payment or dispute of the debt or otherwise within the prescribed time, they disconnected n2riar' electricity supply. VINLEC's Manager Hazell was taken to task by Mr Smith for disconnecting Morri within two weeks of Morris receiving the demand letter - 380 - b2diLI0 Page 7 of 8 when it is the normal practice to allow at least 30 days before disconnection. I accept ftgellls explanation for this when he said that 30 days was only given where there was no lndi-cation of irregularity. The next issue to be decided in this consolidated matter is whether such a disconnection by VINLEC was lawful or not. In this regard I refer to So 19(2) and 22 of Act: Sec 19(2) states as lollows: wIf any person without legal thought unlawfully disconnects, damages or removes any electricity line, meter switch, fuse or other works or apparatus belonging to the Company or alters the index of any meter belonging to the Company, or otherwise prevents any such meter from correctly registering any quantity of electricity supplied by the Company, such pers.n shall be guilty of an offence and for every such offence he shall be liable on summary conviction to a penalty not exceeding one hundred and fifty dollars for the first offence and not exceeding two hundred and fifty dollars for any such subsequent offence, and without prejudice to the foregoing, the Company may recover from such person the amount of any damage by it sustained and may also (notwithstanding any agreement or contract previously existing) discontinue any supply of electricity to such person." And j_22 states as tollows: "If any e'onsumer shall Le in default with any payment due by him to the Company in respect of electricity the Company (without prejudice to any other remedy available to it) shall be at liberty to discontinue the supply to electricity to such consumer until such time as such payment together with the Company's reasonable charges for the reconnection of such Consumer's electricity st-vices have been paid." I interpret these two provisions of the Law in the = to mean that VINLEC can discontinue electricity supply to a consumer if the consumer has been in default of payment of his electricity bills or if such consumer has been found to have tampered with the electricity meter. VINLEC's Manager, JnhiVhHazellLs evidence seems to suggest that VINLEC disconnected Morris under S 22 of the Act I.e. for Morris' default in making the payment. To my mind, having regard to ay aforementioned findings, VINLEC would have been justified in disconnecting IolKix on either or both of the ground aforementioned. I find no merit in the submission of Mr. ZSmith that VINLEC could not have disconnected Harris under S 19(2) because when the disconnection took place new meters which were not tampered with were already in place. The evidence ehows that after the dicovery of the diversion negotiations started and were continuing between the parties and I cannot see how the placing of new meters on the premises by VINLEC would have taken away their right in law to disconnect for the breach by gxrris of S 19(2) of the Act. There is no pleading by Ugrria or aequiesence or waiver by VINIEC of MorLris illegal act, and indeed, the evidence shows outrage on the part of VIhLEC. I also do not agree with the submission of Mr SMigh that disconnection under S22 would also have been unlawful. Learned Queen's Counsel gave as his reason for so submitting that the debt of $52,811.70 was a disputed debt. I agree with his submission that a disputed sum cannot be a payment due but, I find as - 381 - &Wenfix 20 Page 8 of 8 a matter of law that as a result of the non response by Morria to VINLEC's demand for payment, the debt could not have been said to be disputed, and the disconnection took place in these circumstances. Also, having regard to the circumstance of illegality or irregularity which this Court finds in Morris, I am satisfied that the two weeks given by VINLEC for Morris to pay the debt was an extremely generous act on VINLEC's part. To my mind, the debt only became disputed after the disconnection took place and then the parties came to a 'without prejudice" agreement whereby upon Morri. paying VINLEC $19,000 the electricity was restored as at July 1983. Having made these findings I find the disconnection by VINLEC of Morris' electricity to be lawful and in accordance with the provisions of the Act. Taking this matter in its totality and on a preponderance of probabilities T find VINLEC's allegation of fraud on the part of Molria proved to the hilt. In these circumstances, and, having found legal justification for VINLEC's disconnection of Morris' electricity I can find no merit in MoXria' Suit No. Ilk of 1983 against VINLEC and I order that it stand dismissed with costs to VINLEC to be taxed if not agreed and tnat Judgement be entered for VINLEC against Morris in Suit No. IDA of 1983 for $52,811.70 being payment dur for electricity consumed by Morris and not paid for and $525.00: being the cost of one meter which I find VINLEC had to replace as a result of the unlawful act of MoWrs. VINLEC will have the costs of the action to be taxed if not agreed. I therefore make the order that Judgement be entered for VINLEC in the consolidated matter for $53,336.70 less the sum cf $19,000.00 already paid with cost to be taxed if not agreed. SUPREME COURT JUDGE ENERGY SECTOR MANAGEMENT ASSISTANCE PROGRAM Aaivities Completad Country Project Date Number ENElRX EFFICIENCY AND 8TRAEGt Africa Regional The Interofrican Electrical Engineering College: Proposals for Short- and Long-Term Development 3/90 112/90 Participants' Reports - Regional Power Seminar on Reducing Electric System Losses in Africa 8/88 0871P8 Bangladesh Power System Efficiency Study 2/86 031/85 Bolivia La Paz Private Power Technical Assistance 2/90 111/9( Botswana Pump Electrification Prefeasibility Study 1 /88 047/8P Review of Electricitv Service Connection Policy 7/87 071/87 Tuli Block Farms Electrification Prefeasibility Study 7/87 072/87 Burkina Technical Assistance Program 3/88 0S2/86 Burundi Presentation of Energy Projects for the Fourth Five-Year Plan (1983-1987) 5/85 036/85 Review of Petroleum Import and Distribution Arrangements 1/84 012/84 Burundi/Rwanda/Zeire (EGL Report) EvAluation de l'Energie des Pays des Grands Lacs 2/89 098/89 Congo Power Development Study 5/80 106/90 Costs Rica Recommended Technical Assistance Projects 11/84 027/84 Ethiopia Power System Efficiency Study 10/85 045/85 The Gambia Petroleum Supply Management Assistance 4/85 035/85 Ghana Energy Rationalization in the Industrial Sector of Ghana 6/88 084/88 Guinea- Rece';4amended Technical Assistance Bissau Projects in the Electric Power Sector 4/86 033/85 Management Options for the Electric Power and Water Supply Subsectors 2/90 100/90 Indonesia Energy Efficiency Improvement in the Brick, Tile and Lime Industries on Java 4/87 067/87 Power Generation Efficiency Study 2/86 050/86 Diesel Generation Efficiency Improvement Study 12/88 095/88 Jamaica Petroleum Procurement, Refining, and Distribution 11/86 061/86 Kenya Power System Efficiency Report 3/84 014/84 Uberia Power System Efficiency Study 12/87 081/87 Recommended Technical Assistance Projects 6/85 038/85 Medagear Power System Efficiency Study 12/87 075/87 Malaysi Sabah Power System Efficiency Study 3/87 088/87 Mauritus Power System EfficieneV Study 5/87 070/87 Mozambique Household Electricity Utilization Study 6/90 113/90 Panome Power System Lose Reduction Study 8/83 004/83 Papua Now Energy Sector Institutional Review: Proposals Guine for Strengthening the Department of Minerals and Energy 10/84 023/84 Power Tariff Study 10/84 024/84 ENERGY SECTOR MANAGEMENT ASSISTANCE PROGRAM AotvWk Completed Country Project Date Number ENERGY EFFICIENCY ANPl STRATEOY (Continued) Senegal Assistano Given for Preparation of Dooumente for Energy Sector Donors' Meeting 4/86 056/88 Seyche%ie Electric Power System Efficiency Study 8/84 021/84 Sri Lanka Power System Lose Reduction Study 7/83 007/83 Syria Electric Power Efficiency Study 9/88 089/88 Energy Efficiency in tho Cement Industry 7/89 099/89 Syria Energy Efficiency Improvement in the Fertilizer Sector 6/90 115/90 Sudan Power System Efficiency Study Management 8/84 018/84 Assistance to the Ministry of Energy and Mining 5/83 003/83 Togo Power System Efficiency Study 12/87 078/87 Tunisia Interfuel Subst;tution Study 5/90 114/90 Uganda Energy Efficiency in Tobacco Curing Industry 2/86 049/8e Institutional Strengthening in the Energy Sector 1/85 029/85 Power System Efficiency Study 12/88 092/88 Zambia Energy Sector Institutional Review 11/88 060/88 Energy Sector Strategy 12/88 094/88 Power System Efficiency Study 12/88 093/88 Zimbabwe Petroleum SuppIy Management 2/90 109/90 Power Sector Management Assistance Project: Background Objectives, and Work Plan 4/85 034/85 Power System Lose Reduction Study 6/83 005/83 OUQSEHOQL. RURAL AND RENEWABLE ENERGY Burundi Poet Utilization Project 11/85 046/85 Improved Charcoal Cookstove Strategy 9/85 042/85 Cape Verde Household Energy Strategy Study 2/90 110/90 China Country-Level Rural Energy Assessments: A Joint Study of ESMAP and Chinese Experts 5/89 101/89 Fuolwood Development Conservation Project 12/89 105189 Costa Rica Forest Residues Utilization Study, Volumes I & 11 2/90 108/90 C8te d'lvoire Improved Biomass Utilization-Pilot Projects Using Agro-Industrial Residues 4/87 089/87 Ethiopia Agricultural Residue Briquetting: Pilot Project 12/88 062/88 Bagasse Study 12/86 003/86 The Gambioa Solr Water rioeting Retrofit Project 2/85 030/85 Solar Photovoltic Applications 3/85 032185 Ghana Sawmill Residues Utilization Study. Vol. I & 11 10/88 074/87 Global Proceedings of the ESMAP Easm and Southem Africo Household Energy Planning Seminar 8/88 085/88 India Opportunities for Conmnercialization of Non-Conventional Energy Systems 11/88 091/88 Indonesia Urban Household Energy Strategy Study 2/90 107/90 Jamaica FIDCO Saowmill Residues Utilization Study 9/88 088/88 Charcoal Production Project 9/88 090/88 ENERGY SECTOR MANAGEMENT ASSISTANCE PROGRAM Aotivides Completed CountrV Projoot Date Number HOUSEHOLO RURAL. AND RENEWABLE ENERGY (Continued) Konya Solar Water Heating Study 2187 060/87 Urban Woodfuwl Devolopment 10/87 078/87 Malawi Technical Assistance to Improve the Efficiency of Fuolwood Use in the Tobacco Industry 11/83 009/83 Mauritania Elements of a Household Energy Strategy 7/90 123/90 Mauritius 8aganse Power Potential 10/87 077/87 Niger Household Energy Conservation and Substitution 12/87 082/87 Improved Stoves Project 12/87 080/87 Pakistan Assessment of Photovoltaic Programs. Applications and Markets 10/89 103/89 Peru Proposal for a Stove Dissemination Program in the Sierra 2/87 064/87 Rwand, Improved Charcoal Cooketove Strategy 8/80 059/88 Improved Charcoal Production Techniques 2/87 065/87 Senegal Industrial Energy Conservation Project 6/85 037/85 Urban Household Energy Strategy 2/89 098/89 Sri Lanka Industrial Energy Conservation: Feasibility Studies for Selected Industries 3/88 054/80 Sudan Wood Energy/Forestry Project 4/88 073/88 Tanzania Woodfuel/Forestry Project 8/88 086/88 Smnall-Holder Tobacco Curing Efficiency Project 5/89 102/89 Thailand Accelerated Dissemination of Improved Stoves and Charcoal Kilns 9/87 079/87 Rural Energy Issues and Options 9/85 044/85 Northeast Region Village Forestry and Woodfuel Pro-investment Study 2/88 083/88 Togo Wood Recovery in the Nangbeto Lake 4/88 055/86 Uganda Fuelwood/Forestry Femsibility Study 3/86 053/80 Energy Efficiency Improvement in the Brick and Tile Industry 2/89 097/89 Zaimnb Urban Household Energy Strategy Study 8/90 121/90 Zimbabwe Charcoal Utilization Prefeasibility Study 6/90 119/90