Tanzania - Power loss reduction study (Vol. 1 of 2) : Transmission and distribution system - technical loss reduction and network development

ESM 2C9 VolI. P1 WD<DnOq Urnieen P<;onEDO V(DOSC E3Q0p1<3Q 'Ornerm / W rDd Bank aWpLDESMAAP Energy Sector Management %asistance Programme Tanzania Power Loss Reduction Study Volume 1: Transmission and Distribution System Technical Loss Reduction and Network Development Report No. 204A/98 June 1998  JOINT UNDP I WORLD BANK ENERGY SECTOR MANAGEMENT ASSISTANCE PROGRAMME (ESMAP) PURPOSE The Joint UNDP/World Bank Energy Sector Management Assistance Programme (ESMAP) is a special global technical assistance program run as part of the World Bank's Energy, Mining and Telecommunications Department. ESMAP provides advice to governments on sustainable energy development. Established with the support of UNDP and 15 bilateral official donors in 1983, it focuses on the role of energy in the development process with the objective of contributing to poverty alleviation, improving living conditions and preserving the environment in developing countries and transition economies. ESMAP centers its interventions around three priority areas: sector reform and restructuring; access to modern energy for the poorest; and promotion of sustainable energy practices. GOVERNANCE AND OPERATIONS ESMAP is governed by a Consultative Group (ESMAP CG), composed of representatives of the UNDP and World Bank, other donors, and development experts from regions benefiting from ESMAP's assistance. The ESMAP CG is chaired by a World Bank's Vice President, and advised by a Technical Advisory Group (TAG) of four independent energy experts that reviews the Programme's strategic agenda, its work plan, and its achievements. ESMAP relies on a cadre of engineers, energy planners, and economists from the World Bank to conduct its activities under the guidance of the Manager of ESMAP, responsible for administering the Programme. FUNDING ESMAP is a cooperative effort supported over the years by the World Bank, the UNDP and other United Nations agencies, the European Union, the Organization of American States (OAS), the Latin American Energy Organization (OLADE), and public and private donors from countries including Australia, Belgium, Canada, Denmark, Germany, Finland, France, Iceland, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, Portugal, Sweden, Switzerland, the United Kingdom, and the United States. FURTHER INFORMATION An up-to-date listing of completed ESMAP projects is appended to this report. For further information, a copy of ESMAP Annual Report or copies of completed projects, contact: ESMAP c/o Energy, Mining and Telecommunications Department The World Bank 1818 H Street, NW Washington, DC 20433 U.S.A.  TANZANIA POWER LOSS REDUCTION STUDY VOLUME 1: TRANSMISSION AND DISTRIBUTION SYSTEM TECHNICAL LOSS REDUCTION AND NETWORK DEVELOPMENT November 1992  I  ABBREVIATIONS km Kilometer kV Kilovolt kVA Kilovolt ampere kVAr Kilovolt ampere, reactive kW Kilowatt kWh Kilowatt hour MVA Megavolt ampere MVAr Megavolt ampere, reactive MW Megawatt MWh Megawatt hour var reactive volt ampere US$ United States Dollar ACRONYMS ESMAP Joint UNDP/World Bank Energy Sector Management Assistance Programme IDA International Development Association JICA Japan International Corporation Agency SIDA Swedish International Development Agency SVS Static var compensation System TANESCO Tanzania Electric Supply Company Ltd. TIRDO Tanzania Industrial Research and Development Organization  I  TABLE OF CONTENTS EXECUTIVE SUMMARY . .................................................. i Project Objectives . .................................................. ii Achievements and Conclusions ......................................... ii System Loss Analysis . ................................................ iii Reactive Compensation Requirements for the Transmission System .... ......... iv Reactive Compensation Requirements for the Distribution System .... .......... vii Total Project Costs . .................................................. x 1. INTRODUCTION ......................................................I Background ......................................................... 1 ESMAP Assistance ................................................... 2 2. THE TANESCO POWER SYSTEM ........................................ 3 Generation ......................................................... 3 Transmission . ...................................................... 4 Distribution ........................................................ 4 Consumers and Demand . .............................................. 5 System Operations . .................................................. 5 System Expansion . .................................................. 6 3. ANALYSIS OF SYSTEM LOSSES ......................................... 9 Network Losses in the Distribution System ................................ 10 Transmission System Losses ............ ............................... 13 Loss Analysis for Total System ......................................... 14 4. REACTIVE COMPENSATION FOR THE TRANSMISSION SYSTEM .... ........ 21 Background ......................................................... 21 Methodology for Load Flow Studies ..................................... 22 Preliminary System Improvements ....................................... 23 The Study of Network Operation at Different Load Levels .................... 27 Results of Load Flow Studies ............ .............................. 29 System Performance for 1991 ..................................... 30 System Performance for 1992 ..................................... 30 System Performance for 1993 ..................................... 31 System Performance for 1994 ..................................... 32 System Performance for 1995 ..................................... 32 System Performance Beyond 1995 ................................. 34 Effect of the Singida-Arusha Connection ............................ 34 System Performance for 1998 ..................................... 35 Power Export Possibilities to Kenya ................................ 36 System Operation at Base Load ......................................... 37 System Operation During Outages ....................................... 37 Slippage of Planned Investment ................................... 39 Loss Reduction Benefits . .............................................. 40 Selection of var Compensating Systems ................................... 40 Compensation Requirements at Ubungo ............................ 41 Compensation Requirements at Arusha ............................. 42  5. ECONOMIC ANALYSIS OF PROPOSALS FOR POWER FACTOR COMPENSATION AT THE TRANSMISSION LEVEL ...................... 48 Benefits of Investment in Reactive Compensation ........................... 48 The value of unsupplied or poorly supplied energy ..................... 50 The value of loss reduction benefits ................................ 51 Cost of outage savings .......................................... 51 Cost of Compensation Equipment ........................................ 51 Calculation of benefits and benefit to cost ratios ............................ 52 Compensation to be provided at Ubungo/ilala ........................ 52 Compensation to be provided at Arusha ............................ 53 6. DISTRIBUTION SYSTEM DEVELOPMENT AND PLANNING GUIDELINES ..... 56 Background . ....................................................... 56 Economic Analysis of Development Proposals .............................. 60 Existing Distribution Networks ......................................... 62 Planning Concepts ............. ...................................... 63 Planning Guidelines ............ ...................................... 64 7. REACTIVE COMPENSATION ON THE DISTRIBUTION SYSTEM .... .......... 68 General Considerations and Methodology ................................. 68 Economics of Capacitor Installation for Power Factor Control ................. 70 Recommendations for Reactive Compensation on the Distribution System .... .... 74 Summary of Recommendations ................................... 75 8. DEVELOPMENT PROPOSALS FOR DAR ES SALAAM ...................... 77 The present distribution system ......................................... 77 Network Development proposals ........................................ 78 9. SYSTEM DEVELOPMENT PROPOSALS FOR TANGA REGION .... ........... 83 The present distribution system ......................................... 83 Network Development proposals ........................................ 84 10. DEVELOPMENT PROPOSALS FOR THE MOSHI REGION .................. 88 The present distribution system ......................................... 88 Network Development proposals ........................................ 89 1 1. DEVELOPMENT PROPOSALS FOR THE ARUSHA REGION ..... ............ 91 The present distribution system ......................................... 91 Network Development Proposals ........................................ 92 12. REHABILITATION, RATIONALIZATION, AND NETWORK EXPANSION ...... 95 Network rehabilitation and reinforcement ................................. 96 Rationalization of L.V. Systems ......................................... 98 Network Expansion to New Development Areas ............................ 99  List of Annexes The following Annexes provides details of the computations used in the analysis: Annex A - Distribution System Characteristics Annex B - Transmission System Load Flow Studies Annex C - Economic Evaluation of Reactive Compensation Requirements For Transmission and Distribution Systems Annex D - Planning Methods and Guidelines for Distribution Systems Annex E - Economic Analysis for Distribution System Development Annex F - Distribution System Network Drawings This report was prepared in 1992 consequent to a study conducted during 1990 to 1992 in association with a distribution planning unit established in the Tanzania Electric Supply Company (TANESCO). The report is in two volumes; volume 1 dealing with the technical studies conducted on transmission and distribution systems and volume 2 dealing with non-technical loss issues. The Bank team carrying out the study consisted of Messrs. Winston Hay, Chrisantha Ratnayake, and Ms. Paivi Koljonen. Volume 1 was prepared by Mr. Ratnayake and volume 2 by Mr. Hay. The assistance and participation of the distribution planning unit in TANESCO in carrying out the detailed studies on which this report is based is readily acknowledged.  I I  EXECUTIVE SUMMARY 1. The Tanzania Electric Supply Company (TANESCO), a state-owned utility, is responsible for the generation, transmission, and distribution of electricity in mainland Tanzania. Over the past several years, TANESCO's operations have suffered from poor system performance in the East and Northeast regions at times of peak demand. Despite use of all available thermal generation and curtailment of supply to certain major consumers, network voltages remain excessively low, and the system suffers from frequent outages that may affect the entire network. Concomitantly, equipment damage and burnouts occur in the installations of the supply authority and its consumers. In many areas, requests for new consumers cannot be met because of the poor system voltages. Symptomatic of the problem are energy losses of some 21 percent of net generation, a level considerably higher than the economic limits for the TANESCO system (after allowing for nontechnical losses). The shortcomings of the system have resulted in substantial economic losses to the country and financial constraints on the company. Moreover, high load growth rates are expected over the next few years, and the situation will worsen considerably without timely corrective action. 2. The major cause of the low system voltages has been identified as deficiencies in the transmission system. The network supplying the eastern and northeastern parts of the system is heavily overburdened and has clearly exceeded technical and economic loading levels, causing the distribution system to receive power at voltages far below the required level. In contrast, the western sections of the network experience high system voltages because of the long and comparatively lightly loaded 220 kV lines, thus further complicating the control of the transmission system voltages. 3. Two major development projects have been implemented for the distribution systems within the last five years, alleviating supply conditions to a large extent. But the rapid increase of load during this period and the absence of systematic network planning have caused many parts of the system to revert once again to poor operating conditions. A major contributory factor has also been the difficulty faced by TANESCO in undertaking sufficient network improvements outside of externally funded projects because of the lack of foreign exchange. Thus, considerable inputs are still required in the distribution systems to bring the networks to a satisfactory operating condition. 4. In 1989, TANESCO requested the joint UNDP/World Bank Energy Sector Management Assistance Programme (ESMAP) to assist in the development of a program of activities that would improve network voltage levels, reduce overall energy losses and generally improve the quality of supply to consumers. A study was commenced in late 1989 with funds provided by the Swedish International Development Authority (SIDA). A team of consultants comprising of a specialist each in the fields of transmission, distribution, commercial operations, load forecasting, and nontechnical losses was engaged for short termn visits to assist in the study. In addition, a local consultancy firm was engaged to undertake the study of the economic costs of outages with respect to industrial consumers. Counterpart staff provided by TANESCO also worked  . - in close coordination with the ESMAP staff and consultants and provided valuable inputs. Project Objectives 5. The project seeks to find solutions to redress the poor supply conditions in TANESCO's power network and to reduce the existing high level of system losses. In particular possibilities of any short term actions to improve network voltages and enable new consumers to be added to the system at locations where such difficulties exist are to be explored. A major goal is also the development of TANESCO's competence in the field of distribution system planning so that timely action would continue to be taken to keep the networks at acceptable technoeconomic standards. Specific objectives are as follows: * Identification of the main reasons for the poor system performance in the east and northeast regions of the network and development of solutions for rectification * Identification of the sources of technical and nontechnical losses and development of reliable estimates of the contribution of each to overall losses * Development of projects to reduce technical losses to economic levels v Introduction to TANESCO of state-of-the-art techniques of data collection and distribution system planning and training counterpart staff to ensure the continuity of the applications. 6. The study report is issued in two volumes, undertaking the technical and nontechnical problems, respectively. The present report, volume (i), Transmission and Distribution System Technical Loss Reduction and Network Development deals with the diagnoses of problems and provides proposals to improve the technical condition of the transmission and distribution systems. Achievements and Conclusions 7. The principal cause of the low system voltages experienced in many areas of the east and northeast was identified as the insufficiencies of the 220 kV transmission lines from Kidatu to Dar es Salaam as well as the 132 kV lines supplying the northeast loads. The study has identified solutions that can be applied in the short term to enable the system performance to be strengthened considerably. These consist of introducing reactive compensation equipment both in the transmission and distribution systems. The report recommends 45 MVAr of switched capacitors at Dar es Salaam and 30 MVAr of variable compensation using a Static var Compensation unit (SVS) at Arusha for the transmission system; it recommends 22 MVAr of capacitors for installation in the distribution system. Some subsidiary improvements to the transmission system, consisting of a 10 MVAr reactor at Mbeya (to contain the voltage rise in the western sections of the grid system) and shifting of an existing reactor from Ubungo to Ras Kiromany (for voltage control and loss reduction), are also identified.  8. In addition to the above recommendations some improvements of an operational nature was carried out while the studies were in progress. These consisted of the rearrangement of generator tap settings at the power stations and the introduction of a circuit breaker to control a reactor in service at Ubungo. 9. System studies were conducted on the distribution networks in the four major load centers: Dar es Salaam, Tanga, Moshi, and Arusha. Although the reactive compensation measures recommended (para 7) will provide some immediate relief, more substantial developments involving major changes to the network configurations are required in the longer term. Such requirements are presented as a package of proposals for which international funding will require to be secured. These development proposals will improve system performance considerably and reduce network losses to economic levels. 10. The study was conducted in close cooperation with TANESCO staff and emphasized training of the counterpart staff in the techniques and methodologies of distribution system planning. A study unit was established to undertake collection and analysis of network data on the distribution system. Microprocessor based instrumentation was provided to record and store network load data for subsequent downloading to a computer. The unit was provided with computers and state-of-the-art software packages to establish a mapping data base and undertake the required system analysis of the transmission and distribution networks. The hard work and dedication of the counterpart staff was recognised by TANESCO's management and the unit was soon institutionalized in the organization structure within the operations directorate responsible to oversee the functions of the zonal distribution organizations. System Loss Analysis 11. Important load characteristics on the networks were collected by conducting extensive field measurements using electronic recording loggers, and these data were used to compute the losses of medium voltage (MV) and low voltage (LV) feeders. Using the data together with the monthly billing information from the sales summaries enabled development of a power and energy flow table. The table provides information on the losses (as well as supplies) at each voltage level as a percentage of the net power generated and indicates that the overall technical losses of the system amount to 19.3 percent for peak power and 11.5 percent for energy, based on net generation. Because the overall energy losses are approximately 21 percent, it can be inferred that the nontechnical losses account for 9.5 percent of the energy produced. 12. The transmission system presently exhibits peak losses of about 8.0 percent and energy losses of about 4.4 percent. These loss levels will drop with the commissioning of a second circuit to Dar es Salaam (1995) and the Singida-Arusha connection (possibly by 1996). A further reduction will occur with the commissioning of the new Pangani power station, expected in 1995. The peak losses will fall to about 5 percent with all these developments in place. However, loss levels will again increase when the Kihansi power station is commissioned because of the increased  - iv - load transmitted to the main load center at Dar es Salaam. 13. The overall loss levels of the MV system are not excessive. This is attributable to the development work carried out by two recent distribution system improvement projects-the Dar es Salaam rehabilitation project, funded by the Japan International Corporation Agency (JICA) and the Power Project IV, funded by IDA. However, a number of feeders still have excessive loss levels, and the situation will worsen considerably with the expected load growth. Proposals to reduce the losses of these feeders to economically acceptable levels are contained in sections 8 to 11 of this report. 14. In comparison with the losses of the MV system, the losses of the LV networks are extremely high, with average values of the samples studied in the four regions ranging from 6.5 to 10.5 percent for peak losses. These high levels, combined with the fact that a substantial portion of the system load flows through the LV system, mean that LV line losses account for 45.7 and 35.1 percent of the overall peak and energy losses, respectively. Two major factors contribute to the high loss levels in the LV networks: the high imbalance between phase currents in the three phases and the long lengths of LV feeders used in the network design. The former can be corrected without any capital investment, and savings of perhaps up to 25 percent of the existing losses on the LV lines could be made if such an exercise is carried out. To reduce the remaining LV line losses, the number of transformers in the network should be increased substantially, so that the lengths of LV lines can be reduced. The shortening of these lines will also reduce the area of supply of the individual transformers and increase the overall reliability of the system. An estimate for such requirements is included under LV rationalization in section 12. Loss improvement in the LV system should also include a program to purchase lower loss transformers by evaluating the annualized cost of losses together with the purchase price (a common practice among utilities but not presently followed by TANESCO). Further, the existing high core losses in a number of underloaded transformers can be reduced by implementing a transformer load management program to reallocate transformers to better suit the loads supplied. Reactive Compensation Requirements for the Transmission System 15. Preliminary studies have indicated that the poor system voltages in many parts of the system are caused by the deficiencies in the transmission network. TANESCO has plans to rectify these deficiencies by constructing new transmission lines (a second circuit to Dar es Salaam from Kidatu and the Singida-Arusha connection). However, at the commissioning dates presently feasible for these lines, a number of years of continued poor system conditions will have to be endured, and load additions at important locations will have to be suppressed unless other developments are introduced. Load flow studies were therefore conducted on the transmission system to ascertain possible solutions of meeting the expected load additions in the immediate future. The results of these studies (described in section 4) indicate that reactive compensation equipment can be used to improve system performance substantially, particularly in the period before the new lines are commissioned. These reactive compensation measures are recommnended  for Dar es Salaam and Arusha, the two extremities of the eastern and northeastern sections of the transmission network. With these measures in place, the voltage regulation problems in the transmission system will be resolved, and supply can be provided to the distribution networks at acceptable voltage levels for all normal operating conditions. 16. Several benefits will accrue from these proposals, but only a few can be quantified accurately. The compensation recommended for installation at Dar es Salaam will ensure that the expected additional loads in the city and its environs can be met at acceptable voltage levels without resorting to additional thermal generation (which would otherwise be necessary) during the period up to the commissioning of the second circuit from Kidatu. Similarly, the compensation to be provided at Arusha will ensure that the network can bear the expected load additions in the northeast until the Singida-Arusha connection is introduced. The compensation to be provided at Dar es Salaam will also provide a substantial reduction of losses and improve system reliability. 17. After the commissioning of the new transmission lines, both installations will continue to play a useful function. The installation at Dar es Salaam will provide significant savings during times of line outages. Further, after the Kihansi power project is commissioned (by 1998), the capacitive compensation required at Dar es Salaam will increase substantially (up to about 75 MVAr) because of the increased power transfer over the Kidatu-Morogoro-Dar es Salaam lines, and the proposed installation will form a part of this requirement. After the Singida-Arusha line is commissioned, the installation at Arusha will serve a voltage control function by operation in reactive mode (particularly at times of low load). 18. The benefits of connecting additional load (as will be possible with the Arusha SVS) and the reduction of thermal generation requirements (with capacitors at Dar es Salaam) have been valued at US $0.10/kWh, equal to the average incremental cost (AIC) of supply at distribution level in the TANESCO system. Arusha, too, will experience a reduction of thermal generation that will otherwise be required (to support a limited load possible), but the value of such avoided generation has not been costed in view of the close correspondence between thermal generation costs and the value of new sales. The alternative cost of supporting the additional load is in fact considerably higher in Tanzania (see Table CL.I for the cost of supply using small diesel generators). The value of the savings on outage energy units (used for capacitor installation at Dar es Salaam) has been estimated at US$1.00/kWh, ten times the AIC of supply, a figure that compares favorably with studies done elsewhere. A research study conducted as part of the project by the Tanzania Industrial Research and Development Organization (TIRDO) indicated a value of US$2.29/kWh to industrial consumers for unannounced outages. Cost-benefit analyses conducted with the values indicated above ($0. 10/kWh for additional load supplied and $1.00/kWh of outages saved) provided benefit/cost ratios of 28.8:1 and 5:1 for the proposals at Dar es Salaam and Arusha, respectively. Even if the rates used for the benefits are reduced by 50 percent, the high economic viability of the proposals is obvious. In addition, a number of benefits of the proposals have not been quantified: the reduced incidence of damage to equipment owned by TANESCO and its consumers, improved quality of service and increased consumer satisfaction, reduced operation and maintenance expenses, and reduced adverse effects of delays to planned investment in generation and  - vi - transmission. 19. The reduction of adverse effects of delays to planned investment in generation and transmission is worthy of particular consideration in the context of TANESCO's previous experiences. The present plans are for the transmission line from Kidatu to Dar es Salaam to be commissioned in two sections by 1993 and 1995, respectively, and for the Singida-Arusha line to be commissioned in 1996. In the past difficulties faced by TANESCO in obtaining the necessary foreign exchange, and other shortcomings in project management have forced considerable slippage of planned investment. The benefits of the present proposals will increase substantially in the event of such delays; in the absence of reactive compensation at the scale proposed, TANESCO will be forced to resort to extensive load shedding and increased thermal generation over the period of such delays. In any event heavy demands will be placed on thermal generation to meet the hydro generation shortfall until 1998, when the Kihansi project is expected to be commissioned and the present plans for meeting the extent of the shortfall are far from satisfactory. Thus, any marginal relief of the need for thermal generation will be of enhanced value, as it is highly probable that TANESCO will not be able to meet the future demand. Therefore, the reactive compensation measures proposed could be justified even solely on the basis of providing sufficient security for the possible slippage of the planned investment in major generation and transmission projects. 20. The proposed installation at Arusha is a Static var System (SVS), employing Thyristor-controlled reactors. It will provide the best technical solution to the complex problems at this location. If TANESCO is unable to procure the necessary funds for this installation in a timely manner, switched capacitors up to the same rating should be purchased instead. This solution will provide a lesser technical performance but will still provide high economic benefits, at a benefit/cost ratio of 6.4, assuming only half the value of benefits attributed to the prefered solution. This alternative proposal is made because of the urgency of the requirements. In fact, compensation requirements at both locations are already overdue; early implementation will enable TANESCO to recoup higher savings, and delays in implementation will cause substantial losses. On the average for both installations, four months of delay are equivalent to the loss of half the value of the installation costs. This indicates the need for immediate action to capture the full benefits of the proposals. 21. Three other recommendations are made to improve the perfonnance of the transmission system. The first is shifting the existing 20 MVAr reactor at Ubungo to Ras Kiromany, the mainland terminal of the submarine cable supplying power to Zanzibar. This proposal will provide substantial loss reduction and voltage control benefits. Considering only the former, a benefit/cost ratio of 20:1 and a pay back period of 4.5 months have been demonstrated. The second is installing a 10 MVAr switchable reactor at Mbeya to enable better control of the transmission voltages in the southwestern section of the network. The third is providing 10 MVAr of switched capacitors in the distribution network at Tanga; the compensation will provide substantial relief to the northeastern sections of the system. This proposal is justified by the loss reductions that will be achieved along the 132 kV line to Tanga at a benefit/cost ratio of 21:1.  - vii - Reactive Compensation Requirements for the Distribution System 22. Long-term improvement of the conditions in the distribution system is dealt with in sections 8 to 11. However the developments identified will require at least three to five years for implementation because of the necessity to arrange for external financing as well as other logistical difficulties in procurement and project management. However, it is possible to provide early relief for the high losses and excessive voltage drops by installing capacitor banks in the distribution system. Providing compensation in the distribution system will also reduce the compensation requirements in the transmission system. In fact, progressive improvements of the existing power factors at the grid substations have been taken into account in the studies that determined the requirements of the transmission system. 23. The lead time for the purchase of these capacitors (particularly those to be fixed on distribution lines) is extremely short (about six months), and installation can be undertaken by TANESCO's own staff as the items are received. The selection of appropriate locations for the capacitors and the resulting loss reduction benefits at various voltage levels in the distribution system are discussed in section 7, and a summary of the recommended installations with related benefit/cost ratios is provided in Table 7.1. Total savings of 3,180 MWh per annum can be expected from the application of the capacitors recommended for the distribution system, giving an overall benefit/cost ratio of 10.6:1. The loss reduction represents a 3 percent reduction of the existing overall distribution system technical losses (estimated at 100 GWh), equivalent to 0.2 percent of the annual energy produced. Distribution System Development and Training of TANESCO staff in Distribution Planning 24. Distribution systems in the four major load centers of the country-the regions encompassing Dar es Salaam, Tanga, Moshi, and Arusha-were selected for detailed study. These four regions collectively account for 75 percent of the energy consumed in the grid-supplied system. The study was conducted by a team of TANESCO engineers established in a newly formed distribution planning department. A program was carried out to collect all relevant distribution system data, such as the conductor sizes and connected loads, and network maps were compiled for the study areas. The loading patterns at selected locations were measured using a number of instruments-in particular, electronic loggers capable of storing the required characteristics for subsequent downloading to a computer. A package of computerized mapping and distribution planning software was provided to set up the data base and undertake the analysis. TANESCO staff has mastered the operation of the software and completed the establishment of the digital mapping data base in all four of the regions included in the study. They are also proficient in operating the load flow programs that interlink with the data base and have conducted a number of studies using the software. In addition to these studies, simpler techniques using calculations performed on computer spreadsheets (see Annexes A and D1) have been used to compute the required characteristics of all the MV feeders as well as a sufficiently representative sample of LV feeders in the study regions.  - viii - 25. The network planning procedure consists of modeling the system to be studied either using the specialized software package or the spreadsheet methodology and determining operating characteristics such as technical losses, loading levels, and line voltage drops. Studies were also made to determine the ability to extend normal supply configurations (to provide alternative supply sources that can be utilized during network outages). The present situation as well as the network conditions in future years were studied, and the effects of various development possibilities examined. The studies were also used to prepare guidelines corresponding to optimum economic performance levels to facilitate the planning procedures. These guidelines include tables and graphs (see Annex D2) that can be applied conveniently to determine the operating characteristics (losses and voltage drops) at various loading levels. 26. The study has addressed two major long-term goals related to institutional development. First, the distribution networks that constitute a vast number of segments of lines and equipment have been collated in a computerized data base that can be updated continuously as network improvements proceed. Second, TANESCO engineers have been trained in conducting the required technical and economic analyses. Software and other study aids have been supplied to predict network performance, such as losses and line-end voltages. In addition, methodologies of computing economic benefits of proposed development options have been established. The unit established is expected to continue the task of monitoring the distribution systems in the future and planning the required developments in a timely manner to maintain the networks at optimal performance levels. The unit is also expected to play an important role in refining the proposals, monitoring the work, and updating the computerized data base as the work is implemented. 27. Sections 8 to 11 of this report deal with the major distribution system development requirements for the four regions. The proposals presented consist of the introduction of new grid substations (132/33 kV) and primary substations (33/11 kV) and uprating the voltage of certain feeders. These efforts are designed to enhance the "bulk supply" function of the distribution systems and to provide economically acceptable loss levels and reliability standards. The individual proposals have been subjected to economic analyses, and the related cost/benefit tables are presented in Annex E. The benefits accounted for are the reduction of losses, the ability to supply new consumers (where supply restriction existed formerly), and the reduction of outages. When all three components are evaluated, the benefit/cost ratios generally exceed 5:1. When the latter benefits mentioned above are not readily quantifiable, the proposals are justified on the basis of the loss reduction benefits alone. 28. Section 12 deals with improvements required to the distribution system supplementary to the major developments dealt with in sections 8 to 11. These requirements consist of three categories of work: * reinforcement of MV systems combined with network rehabilitation, * rationalization of LV systems, and * system expansion.  - ix - 29. The reinforcements required in the first category consist mainly in reconductoring MV line sections presently conductored with gauges too small for the economic dispatch of the load carried. In addition, some new lines are also deemed advantagous in order to reduce the feeder routes in the network. Many of the lines to be reconductored are also antiquated and need extensive rehabilitation. It is for this reason that reinforcement and rehabilitation are conjoined under the same category of work. The rehabilitation will cover requirements of both the MV and LV lines (including those which are not being reconductored) as well as requirements at substations and transformer stations. Rehabilitation-in effect, delayed maintenance efforts-constitutes the major portion of the costs. 30. Specific economic analyses to justify rehabilitation efforts is in reality quite superfluous; if the utility is to remain in business it must keep the existing networks serviceable enough to supply existing consumers. Also if major maintenance is further postponed, the networks will deteriorate rapidly, resulting in burnouts and damage to healthy equipment, with potentially high costs. Finally, no convenient procedures are available to measure the varied benefits of rectifying delayed or neglected maintenance, and engineering judgment often determines the nature and extent of rehabilitation requirements. Nonetheless, an analysis of the possible savings of outage costs that can be achieved by conducting rehabilitation can be gauged from Tables E6.1 to E6.4 of Annex E for the four regions, respectively. At an estimated outage saving of 1 percent of the power supplied and a cost of US$1.00/kWh, the benefit to cost ratios are in the range of 7.5:1 to 17.7:1 for the four regions. The net benefits remain positive even if the outage savings are reduced to 0.3 percent of power supplied and valued at US $ 0.75/kWh along with a reduction of the sales growth rate to 0.04 percent (from 0.08 percent). 31. The second category of suplementary work (see para 28) is LV rationalization, required in view of the high losses existing in the LV systems. The loss analysis detailed in section 3 indicates that 45.7 percent of the total power losses and 35.1 percent of the total energy losses occur in the LV lines. A substantial portion of these losses can be removed by balancing the load of the three phases. At Dar es Salaam in particular, more than 30 percent of the total LV line losses can be attributed to unbalanced conditions. The remaining losses can be reduced to economic levels by introducing a greater number of transformers to reduce the line lengths of the LV network. Estimated requirements for carrying out this work are included under the LV rationalization category. An economic analysis using estimated loss savings is provided in Table E7 in Annex E. 32. System expansion work (identified as the third category of suplementary work in para 28) is designed to extend the networks to cover new areas that have very promising load growth potential. These areas are in close proximity to the existing supply systems, and in many instances unorthodox extensions from the existing networks are already being erected to meet the pressing demand for power. Each of the new areas to be electrified is being subjected to an economic analysis (see Table E8 in Annex E) by computing the benefits from expected sales (less the upstream costs incurred in supplying power to the distribution network). The results of such analyses for the developments studied in detail so far are presented in the report and indicate very  x - high returns. The total cost of the areas that require electrification have been compiled based on presently available data. TANESCO's planning staff will continue the task of designing and analyzing each individual network expansion proposal and ensure that only the economically acceptable projects are included. Total Project Costs 33. The total costs of the proposed development work in the transmission and distribution systems are presented below. Costs of Development Work in Tanzaa's Trnz=m ion and Development Sysems Costs of Reactive Compensation Requirements US$'000 Switched capacitor banks at Ubungo/Ilala 825 SVS unit at Arusha 5,000 10 MVAr reactor at Mbeya 200 Shifting of existing reactor from Ubungo to Ras Kiromany 10 Reactive compensation requirements in the distnbution system (inclusive of 10 MVAr at Tanga) 175 Contingencies 790 TOTAL 7,000 Cost of Distribution System Development Dar es Activity Salaam Moshi Anmsha Tanga US$'OOO MV network development 7,002 1,543 1,332 4,344 14,221 Transmission stns., LV reinf. 3,014 1,279 996 1,544 6,833 Network expansion 5,703 1,088 1,002 2,038 9,831 MV and LV rehabilitation 5,233 1,689 1,568 1,892 10,363 Contingencies 1,753 TOTALS 20,951 5,599 4,879 9,819 43,000 TOTAL COST FOR DEVELOPMENTS IN TRANSMISSION AND DISTRIB1TION SYSTEMS: $50,000  1. INTRODUCTION Background 1.1 The United Republic of Tanzania is situated on the east coast of Africa, its eastern coast lying between 20 and 110 south of the equator. -The country comprises a large mainland area and a number of islands, the largest of which is Zanzibar. Mainland Tanzania, with an area of about 940,000 square kilometers, is bordered on the east by the Indian Ocean; on the north by Kenya, Lake Victoria, and Uganda; on the northwest by Rwanda and Burundi; on the west by Lake Tanganyika, Zambia, Malawi, and Lake Malawi; and on the south by Mozambique. The most recent census (1988) assessed the population of Tanzania as 23 million, 1.5 million of whom lived in Dar es Salaam, the nation's largest city and commercial capital. The overall annual rate of population growth in recent years has been 3A percent, but the rate of urbanization is appreciably higher. It is estimated that the current (1991) population of Dar es Salaam is about 2 million persons. 1.2 The economy of Tanzania depends heavily on agricultural products, but there is some mining activity, primarily in diamonds and other gems, gold, and coal. Significant increases in industrial output have occurred over the last few years, and tourism is also an important and growing foreign exchange earner. The per capita gross national product was estimated at approximately $110 in 1990. 1.3 The Tanzania Electric Supply Company Limited (TANESCO) is responsible for the generation, transmission, and distribution of electricity throughout mainland Tanzania. Although TANESCO sends electricity to Zanzibar through an interconnecting submarine cable, power supplies within Zanzibar are administered by a separate authority, the State Fuel and Power Corporation. 1.4 Over the past several years TANESCO's operations have suffered from low system voltages in the eastern and northeastern regions at times of peak demand. This has necessitated operation of diesel generating plant solely for voltage regulation. Despite the use of thermal generation. TANESCO has had to curtail supply to certain major consumers and reduce the rate at which new consumers are added to the system in areas where the distribution lines are overloaded. System disturbances often result in extensive outages sometimes affecting the entire network. Consumers who require a stable supply of power are therefore compelled to procure standby generating equipment. In addition energy losses on the system have increased steadily, currently exceeding 20 percent of net generation. The effects of these undesirable conditions have contributed to substantial economic losses to the country as well as to financial constraints on the company's operations.  -2 - ESMAP Assistance 1.5 In 1989, TANESCO requested the joint UNDP/World Bank Energy Sector Management Assistance Programme (ESMAP) to assist in the development of a program of activities that would improve system voltage levels, reduce overall energy losses and generally improve the quality of supply to its consumers. Funds provided by the Swedish International Development Authority enabled ESMAP to respond positively to TANESCO's request and the study was initiated in October 1989. ESMAP staff and consultants have worked closely with TANESCO counterpart staff in analyzing the problems and in identifying economic solutions. In the course of the study extensive use has been made of computerized techniques for power system mapping and analysis and as a result TANESCO's staff has developed significant skills in these areas. 1.6 This report describes the results of the studies on the transmission and distribution systems. Network losses are analyzed and solutions identified for the low voltages in the transmission system and poor supply conditions in the distribution systems at Dar es Salaam, Tanga, Moshi, and Arusha, which collectively constitute 75 percent of the total grid supplied load in the country. The recommendation for the voltage improvement in the transmission system consist of the provision of reactive compensation to enable the network to be operated at satisfactory voltage levels until planned transmission system developments can be introduced. The recommendations for the improvement of the distribution systems contain proposals for the major developments required to provide supply at economic loss levels and improve the system reliability. They also cover improvements necessary for the rehabilitation and rationalization of the existing networks in the study areas as well as system extensions required to provide supply to adjacent areas where significant urban development is already in progress. A separate report deals with the high nontechnical losses observed in the system. It describes the efforts made to reduce these losses by conducting systematic field operations and provides solutions to manage nontechnical losses in the longer term.  -3 - 2. THE TANESCO POWER SYSTEM 2.1 The majority of mainland Tanzania is supplied from an interconnected power grid. There are a number of isolated systems supplied by diesel generators but these are very smaU in comparison to the interconnected grid and account for less than five percent of the electric energy consumed annually. This report will be concerned only with the interconnected grid system and especially with the problems of low voltages experienced in the northern and northeast regions. Map IBRD 22162 shows the geographic distribution of TANESCO's generation and transmission system. Generation 2.2 More than 98 percent of annual generation is from six hydroelectric stations with aggregate rated capacity of 327 MW, as shown in Table 2.1 below. Table 2.1 Station River Rated Capacity (MW) Kidatu Great Ruaha 200 Mtera Great Ruaha 80 Hale Pangani 21 Pangani Falls Pangani 17 Nyumba ya Mungu Pangani 8 Kikuletwa Pangani 1 2.3 There are a number of diesel stations and a single gas-turbine generator (not in operation at present) connected to the grid but these are normally used only for emergency generation or voltage support. The aggregate rated capacity of thermal generating units on the grid system is about 100 MW. The capacity available at the end of 1990 was only about 30 MW. Since then a program of rehabilitation of the thermal station at Ubungo, adjacent to the major 220/132/33/11 kV substation supplying Dar es Salaam has increased the available capacity by a further 31 MW (with all six diesel units rehabilitated).  -4 - Transmission 2.4 Figure B4 of Annex B is a one-line schematic of the transmission system, which consists of lines operating at 220, 132 and 66 kV. Over three quarters of the power generated is transmitted by the single circuit 220 kV line (of 132 km) running east from Kidatu to Morogoro. Bulk of the load continues eastward thereafter on a single circuit line (of 178 km) to the principle load center, Dar es Salaam. The remaining power is transformed to 132 kV at Morogoro and a line operating at this lower voltage runs parallel to the 220 kV line between Morogoro and Dar es Salaam. This line has a switching station approximately mid way at Chalinze from which the north eastern network is supplied. Areas in Western Tanzania are supplied from a separate 220 kV system by lines running from Kidatu and Mtera to Mbeya on the southwest and Mwanza on the northwest. The north western network also has two 132 kV spurs supplying Tabora and Musoma. 2.5 The 132 kV system is the backbone of the transmission grid north of Dar es Salaam, with a line extending from Chalinze through Hale to Moshi and Arusha, with a spur from Hale to Tanga, the second largest city of Tanzania. All generating stations on the Pangani River system feed into this 132 kV network. The TANESCO grid is interconnected with that of Zanzibar at 132 kV through overhead lines on land and a submarine cable from Ras Kiromany on the mainland to Ras Fumba in Zanzibar. 2.6 The only 66 kV line currently in regular operation is that between the Nyumba Ya Mungu generating station and the 132/66 kV substation at Kiyungi near Moshi. The 66 kV interconnection between Kiyungi and Arusha is still operable but since the completion of the Kiyungi/Arusha 132 kV line the 66 kV circuit is not normally energized. Subtransmission lines operating at 33 kV emanate from the grid substations shown in Figure B4 (Annax B) and are used to energize distribution substations as well as to supply certain consumers directly. 2.7 . All transmission lines are of single-circuit construction with limited or no supply alternatives in the event of unavailability of any one line. Distribution 2.8 In city areas the majority of primary distribution lines operate at 11 kV with the secondary voltage being 400 V three-phase or 230 V single-phase. 33 kV is used as a subtransmission voltage for bulk supply to distribution substations, but some distribution transformers and major consumers are also directly energized from this voltage. In rural areas 33 kV forms the bulk of the primary distribution voltage.  -5- Consnmers and Demand 2.9 At the end of 1990 TANESCO had approximately 155,300 consumers with annual average energy consumption of about 10,000 kWh. The peak system demand was 264 MW. As is shown in Table 2.2 below, peak demand and annual energy consumption over the past five years have been increasing at an average annual rate of 11 percent. The number of consumers in Dar es Salaam and its immediate environs at the end of 1990 was about 70,000 with a peak demand of approximately 100 MW and annual consumption of 590 GWh. Table 2.2. Power and Energy Demand TANESCO Grid System, 1984 to 1990 11985 11986 1987 11988 11989 11990 Peak demand MW 176 183 200 219 253 264 Generation GWH 915 1041 1169 1303 1436 1565 Sales GWH 696 822 871 1005 1110 1254 Consumers 000s 126 133 111 129 144 155 Losses GWH 219 219 298 298 326 311 Losses (% gen) 24.0 21.0 25.4 22.9 22.7 19.9 % inc. Peak 3.9 9.4 9.5 15.5 4.4 % inc Gener/n 13.8 12.1 11.5 10.2 9.0 % inc Sales 18.2 6.0 15.3 10.5 13.0 % inc. Consu/m 16.0 12.0 7.6 Note: There are number of inaccuracies in the data compiled at TANESCO and the figues should be taken to represent appraximate values. The consumer records were adjusted in 1987 resulting in the removal of a number of 'dead' accounts. System Operations 2.10 Since the mid-1980s, TANESCO's operations have been handicapped by low system voltages in Dar es Salaam and areas in the northeast. At times of peak system demand that normally occurs between 9 a.m and 3 p.m. in the daylight hours and 6 and 9 p.m. in the evening (see Figure 3) the voltage on the 220 kV busbars at Ubungo, the 220/132 kV substation supplying Dar es Salaam, falls to 175 kV or less. At Arusha in the northeast it is necessary to run the diesel  -6- generating units and in addition shed certain loads at times of peak demand in order to maintain system voltages. Even then the voltage conditions are below acceptable levels. The long and relatively lightly loaded transmission lines in the west limit the level to which the voltage on the generator busbars at Kidatu and Mtera can be raised without resulting in unacceptably high voltages at the peripheral grid substations. 2.11 The low-voltage conditions on the transmission system in Dar es Salaam and other areas in the east are compounded by overloaded distribution circuits. Many consumers frequently experience single-phase supply voltages of less than 180 volts instead of the nominal 230 volts. The overloaded distribution circuits are, in part, the result of inadequate distribution planning and the fact that investment in distribution over the past decade has not been commensurate with the increase in consumer demand. TANESCO has had to restrict the rate at which it adds new consumers to the system in certain areas of Dar es Salaam because of the overloaded condition of existing distribution circuits. The overloaded distribution lines and poor power factor control by some consumers have contributed to a steady increase in system energy losses. As can be seen from Table 2.3 above, system losses have remained at a high level (around 20 to 25%) in spite of two major distribution development projects undertaken during the last five years. However not all of these losses are due to overloaded lines or other technical considerations. A significant proportion is due to the failure of the metering and billing system to capture all of the units consumed. 2.12 The low voltages experienced in the high load-density areas of eastern and northeastern Tanzania often result in lost production and equipment damage. The high voltages in the western areas place undue stress on consumers equipment and are believed to have resulted in failure of a number of grid substation transformers. Inadequate investment in the distribution system not only increases losses and restricts the rate at which consumers are added to the system as previously mentioned but also contributes to poor physical condition of much of the distribution plant. In consequence certain areas experience frequent supply outages many of which are of long duration. The various undesirable characteristics of current operations on TANESCO's transmission and distribution systems cause substantial economic loss to the country as well as financial penalties to TANESCO itself. System Expansion 2.13 TANESCO's system expansion plans are based on a study undertaken by the Canadian consultants Acres International and completed in 1985. The demand forecasts have since been revised regularly by TANESCO and Acres, most recently in 1990. The projected demand for the period 1991 to 2005 is as shown in Table 2.3.  -7 - Table 2.3. Forecast of Power and Energy demand upto 2005 IActal< Projected 1990 1991 1992 1993 1994 1995 1996 1997 2002 Peak 264 297 325 352 373 395 416 438 569 Demand (MW) l Annual 1254 1395 1507 1614 1722 1842 1977 2118 2853 Energy (GWh) Annual 13.1 11.2 8.0 7.1 6.6 7.0 73 7.1 - Growth (%) __I I_I __I__I_ Load Factor 67.6 66.6 66.0 65.0 65.0 65.0 65.0 65.0 65.0 Note: The 1990 values represent actual system perfornance. The values for other years are projected. The figures for 1991 have still not been finalized by TANESCO. 2.14 It is seen that the high load growth over the last few years are expected to continue though at a slightly reduced rate. There is also a good possibility that TANESCO will experience a more rapid increase in demand than is currently projected. Generation Expansion 2.15 TANESCO's plans to increase generating capacity in the period 1991 to 2000 are as follows: * To undertake further rehabilitation of the diesel plant in the western parts of the grid system to provide an additional 20 MW by 1994. * Redevelopment of the Pangani Falls site to provide an increase of 49 MW in 1995 * Construction of a new generating station on. the Kihansi River (the Lower Kihansi plant, near Iringa) to provide 150 MW in 1998. 2.16 Comparison of planned investment in generation with projected demand indicates a severe capacity shortage if any large generating unit should be out of service over the period beyond 1992. This situation would be compounded in the event of unfavorable hydrologic conditions, especially in the Great Ruaha/Kihansi watersheds. Adverse rainfall conditions in the catchment areas combined with the non availability of hydro generation units have in fact caused large scale load shedding during many periods in 1992.  -8 - Investments in Transmission 2.17 From now until 2000 TANESCO plans to invest in two major transmission projects. The first is a second 220 kV line (300 km) from Kidatu to Dar es Salaam (probably to a new substation site and not Ubungo) via Morogoro. Construction has already begun for the section Kidatu to Morogoro. Funds are still being sought for the section from Morogoro to Dar es Salaam, which is planned to be in service by the end of 1994. The second major investment in transmission will be the Singida to Arusha 220 kV line (about 280 km) expected to be completed by the end of 1995. Improvements in Distribution 2.18 TANESCO is currently developing plans to rectify the effects of past inadequacies in investment in the distribution systems and to be well positioned to meet the consumer demands of the future. The current ESMAP study will provide the required assistance in determining the investments at the major load centers of Dar es Salaam, Tanga, Moshi and Arusha. TANESCO is independently developing distribution plans for the other service areas. Investment financing is expected from a new World Bank (International Development Agency) power credit as well as from bilateral financial agencies.  -9 - 3. ANALYSIS OF SYSTEM LOSSES 3.1 A power system can be divided conveniently into a number of sections in accordance with the voltage of operation. All major generation stations feed into the transmission system, which in the case of TANESCO consists of a radial network of 220 kV and 132 kV lines. At present, only one load (the power supply to Zanzibar) is supplied directly at the transmission voltage. Grid substations on the transmission system located at major load centers transform power to 33 kV, and feeders at this voltage are used as the first stage of the distribution system. Some of the larger loads are supplied from the 33 kV lines, while others are fed directly off transformers connected to these lines. 33 kV lines are also used to provide supply to rural areas with 33 kV/LV transformers supplying the low voltage networks. The remaining load undergoes a further transformation at primary substations (33 kV/11 kV) resulting in an additional system voltage of 11 kV. Part of the load supplied is fed at the same voltage (11 kV) to larger consumers; a second part is supplied to consumers directly from step down transformers at low voltage; and a third part is carried along LV lines and supplies the rest of the consumers. Consumers supplied at 33 and 11 kV are classified as medium voltage loads. Those supplied at LV are considered as bulk LV loads if fed directly off transformer stations and as retail LV load if fed from the LV lines. The system consists of a hierarchial structure as shown in Fig. 3.1. A substantial portion of the total network load flows along the 11 kV and LV systems. 3.2 Technical losses on the system occur in all sections of the network. Almost the entirety of these losses are caused by the heating effect of an electric current when flowing along a conductor. The exception is the losses caused in magnetizing transformer cores. In addition to the network losses described above, losses also occur at power stations due to the consumption of auxiliary equipment as well as losses on generator transformers. The overall losses (inclusive of the losses in generation stations) are termed gross losses, while the portion of the losses that occur in the transmission and distribution networks are considered as net losses. In the present exercise we will concern ourself with the analysis of net losses, that is, those occurring in the transmission and distribution systems. 3.3 Power supply systems also contain nontechnical losses. These losses are caused by deficiencies in billing and meter reading as well as by consumption not captured by meters. The difference between units sent out of the power stations and units sold (as appearing in the billing statement) are the net system losses consisting of the sum of the two components, technical and nontechnical losses described above. The present report deals with the technical component; a complementary report is being issued regarding nontechnical losses.l/ 1/ This Report is entitled Tanzania: Reduction of Nontechnical Losses.  - 10 - 3.4 The load at various sections of a power network can be expressed in power demand (MW) as well as in energy (MWh) terms. Similarly, system losses are also expressed in power (MW) and energy (MWh) terms; the energy loss being the integration of the power losses over the relevant time period. The ratios of the average to peak values for the power demand and power loss are termed the load factor and the loss factor respectively; the relationships between the various parameters are provided in Annex D. During the study an extensive measurement program was conducted using electronic instruments capable of recording and storing the required power flow characteristics and subsequently downloading these measurements to a computer. One of the instruments used, the DIP 6000, uses three clip-on current transformers and three voltage transducers to record the current, voltage, power factor and power values of three phase networks. Two others, the load logger (which can be placed on overhead lines) and the load profiler records the current of individual phase conductors. All the instruments can be programmed to perform measurements at pre determined intervals and are also supplied with software to facilitate down loading the data and performing analyses of the results. 3.5 The program of power flow measurements carried out included all the 33 and 11 kV feeders in the study area as well as a representative sample of low voltage (LV) feeders and distributors. Some typical results of daily loading profiles and other characteristics are given in Annex A. These results provided values for the load factor and loss factor of individual feeders as well as for network components at various operational levels. These values have been used when necessary to convert the power and power loss figures to energy and energy loss terms in the various computations appearing hereafter. Network Losses in the Distribution System 3.6 In contrast to the transmission system, distribution networks contain an extremely large number of line sections and the building up of the necessary data base as well as its modeling for analysis is far more complex. An important initial step has been made by introducing computer programs to build up a data base and conduct the required system studies for the distribution networks (see Section 7 for a description of the software provided). During the course of the project significant progress has been made in establishing the data bases and performing studies in limited areas of the network. This work is continuing and the study team established is expected to complete modeling the medium voltage distribution systems in the entirety of the Coastal and North Eastern Zones (consisting of the cities Dar es Salaam, Tanga, Moshi, Arusha as well as the associated peripheral areas) by end 1992. However, in order to obtain an early assessment of the existing technical losses in the system an approximate methodology (as explained in Annex D1) has been developed and applied in parallel with the more exhaustive computer applications. This procedure has been used to compute the line losses and terminal voltage drops of all medium voltage networks in the study areas for three typical loading conditions consisting of peak, mid and base load. A similar study has been made on a sample of LV systems in the four regions. The sample consisted of a total of 105 transformer stations having 367 distributors and is therefore considered to be sufficiently representative of the total LV network. Sample results of these  - 11 - studies are presented in Annex A. The losses and voltage drops have been worked out for the 1991 year loads as well as the expected load in 1995 with no developments made to the existing system. The calculations have been repeated with the proposed system developments described in sections 8 to 11. 3.7 The results of the power loss studies for both MV and LV systems indicate a wide diversity of performance among the various feeders. Summaries of the total losses obtained for each region are presented in Tables 3.1 and 3.2 below: Table 3.1. Total Power and Energy Losses in Percent of Input at Each Voltage Level, 1991 Loads Dar-es- Tanga Tanga Tanga whole Salaam Town District Region Moshi Arusha 1) 33 KV lines Power losses 2.0 2.20 130 1.91 3.80 3.7 2) 33 KV lines Energy losses 1.50 1.65 2.38 1.43 2.90 2.8 3. 11 KV lines Power losses 2.80 1.80 2.61 2.04 3.00 5.7 4. 11 KV lines Energy losses 2.10 135 1.96 1.50 2.30 4.3 5. LV lines Power losses 8.90 - - 7.80 1.04 6.5 6. LV lines Energy losses 6.20 - - 5.50 0.72 4.6 Table 32. Total Power and Energy Losses in Percent of Input at Each Voltage Level: 1995 Loads with existing system configuration Dar-es- Tanga Tanga Tanga whole Salaam Town District Rezion Moshi Arusha 1) 33 KV lines Power losses 236 2.6 1.6 2.1 4.46 435 2) 33 KV lines Energy losses 1.77 1.95 1.2 1.58 3.34 3.26 3. 11 KV lines Power losses 3.83 2.20 3.08 2.44 3.55 6.37 4. 11 KV lines Energy losses 2.86 1.65 231 1.83 2.66 4.77 S. LV lines Power losses 1035 - - 9.07 121 7.56 6. LV lines Energy losses 7.21 - - 632 0.84 5.27  - 12 - 3.8 The percentage losses provided in Tables 3.1 and 3.2 are based on the input power at the respective voltages. These percentages can be converted to the base of the input power and energy of the total system by consecutively accounting for the loads and calculating the losses along each section of the network hierarchy (described in para 3.1 and shown in Figure 3.1). Such an analysis has been conducted for the distnbution system using the monthly billing data for the year 1990 for the four regions and is presented in Table A4 in Annex A In this analysis the maximum demands (kVA) recorded in the various tariff categories have been converted to peak loads by use of estimated coincidence factors and the consumption of retail units (where the kVA demand is not recorded) has been converted using an estimated overall load factor. The estimated figures have been derieved from the data collected from field measurements and the total demand figures have been reconciled with recorded values thus indicating a good degree of reliability for the results obtained. Thereafter the average values for power and energy losses at various voltage levels (presented in Table 3.1) have been used to develop a hierarchial power and energy flow table consisting of the loads and losses occurring at each level in the network. Finally the losses and loads at each voltage level are expressed as percentages of the input to the distribution system. A summary of the results of these tables is presented in Table 3.3 below. Charts indicating the comparative losses of the four regions, the total sales and loss breakdown and the breakdown of overall distribution losses are presented in Figures 3.2 to 3.4. Table 3.3. Distribution Network Losses in Percent of Input to Distribution System FOR ALL 4 DAR ES TANGA MOSFI ARUSHA R8GNS SALAAM PCWER LOSSES 33 kV line loss 2.26 2.00 1.20 3.79 3.70 SS T/f loss 1.22 1.26 1.02 1.18 1.21 t1 kV line loss 2.44 2.32 1.34 2.32 4.53 LV t/f loss 1.16 1.19 0.88 1.31 1.19 LV line loss 5.97 6.22 4.42 7.61 4.74 Sum of losses 13.05 12.98 8.86 16.22 15.36 B'ENPGYLOSSES 33 kV line loss 1.69 1.50 0.90 2.85 2.77 SS T/f loss 1.14 1.18 0.75 z1.34 1.22 11 kV line loss 1.74 1.63 0.74 1.98 3.42 LV t/f loss 1.00 1.01 0.86 1.17 0.97 LV line loss 3.02 3.26 1.96 3.67 2.41 Sum of losses 8.59 8.58 5.21 11.00 10.80 3.9 An examination of the distribution system losses developed above indicate that the overall network loss levels for the medium voltage feeders are not excessive (particularly in comparison to those in many developing countries). The low loss values in many parts of the network is the result of two major distribution development projects carried out recently. However, there still remains a number of feeders where the loss levels are considerably high. Further with the  - 13 - expected load growth between now and 1995 the situation will worsen considerably unless corrective action is undertaken. A number of system improvements to reduce losses in areas where they are exceptionally high have been formulated in sections 8 to 11. A particular difficulty that TANESCO faces is that it can not often mobilize the funds (particularly foreign exchange) necessary to undertake development requirements in a timely manner. It is therefore very necessary to ensure that all necessary steps are taken to limit the time delays associated securing the necessary financing facilities. 3.10 The sample studies conducted for the low voltage systems indicate that the network losses in this section of the network is exceptionaLy high. The average peak losses for LV networks in the four regions yielded results between 6.5 and 10.5% of the input power. The LV line losses together with the transformer losses exceeded half the overall distribution system peak losses. In energy losses too this figure came close to the half mark (46.7%, see Fig. 3.4.). Clearly this is an area where considerable loss reduction inputs are required. One major factor is the high unbalance existing in many of the schemes studied. The losses due to line unbalance is exceptionally high in Dar es Salaam and Tanga where savings as high as 30 and 20% of the existing losses can be realized by undertaking a program of load balancing, without incurring any capital investment. The other major loss reduction improvement required in the LV networks is the introduction of a substantial number of additional distribution transformers to reduce the LV line lengths. A program of transformer replacements and interchanges need also to be carried out to enable more efficient loading levels to be achieved (to optimize transformer losses). TANESCO should also change its present practice regarding transformer purchases by including the evaluation of lifetime cost of losses thereby ensuring that low loss transformers are secured in the future. These developments are discussed in section 12. Transmission System Losses 3.11 Losses in the transmission system can readily be obtained by the results of the load flow studies conducted. Section 4 discusses these studies and Annex B presents the results of the computations made. Typical peak, mid and base load flows for each year from 1991 to 1998 have been performed. In order to calculate the energy losses a simplified load duration curve consisting of 25% of average peak load, 42% of average mid load (taken at 80% of peak) and 33% of base load has been assumed. While the peak and base load values have been taken direct from the results of the studies, the mid load values have been estimated by reference to load flows conducted with similar operating conditions. These results are presented in Table 3.4 below:  - 14 - Table 3.4. Transmission System Losses in Percent Mid- Base- Year Proposed Developments Peak LoadLoad 1991 8.0 3.0 5.5 1992 9.0 3.0 4.7 1993 Kidatu-Morogoro 2nd cct. 5.8 3.5 4.4 1994 6.0 3.8 4.8 1995 Morogoro-Ubungo 2nd cct. 5.5 4.0 4.3 New Pangani P.S. 1996 Singida-Arusha Line 5.0 4.0 4.2 1997 5.4 4.2 4.4 1998 Kihansi P.S. 6.2 43 4.6 3.12 The results indicate a number of step changes to the loss levels that occur with various network developments introduced. System losses decrease with the transmission developments of the second circuit to Dar es Salaam and the Singida-Arusha connection as well as the New Pangani Falls power station. Peak losses decrease in certain years due to thermal generation inputs (supplied at Ubungo, the main load center) to meet the hydro power deficit. With the commissioning of the Kihansi project the necessity of thermal generation is removed and the power flow along the Kidatu-Morogoro-Ubungo lines is increased resulting in higher transmission system losses. The introduction of reactive compensation (proposed in this report) at IJbungo will also contribute to substantial loss reduction both before 1995 and after 1998. Loss Analysis for Total System 3.13 The present level of network losses derived for the transmission system and the load supplied at this level (to Zanzibar) has been combined with results of the distribution system power flow developed (in Table A4 of Annex A) to obtain the overall network power flow and system losses based on net generation (ie. input to the transmission system). For this purpose the percentage figures obtained for supplies and losses for the four regions studied are assumed to represent the entire grid supplied distribution system. Summary information from this table is provided at Table 3.5. It is seen that line losses of transmission and distribution systems are approximately of equal significance. The results of the above loss analysis for the total system (based on net generation) is also presented in Fig. 3.5.  - 15 - Table 3.5. Breakdown of System Losses and Supplies by Voltage Level POWER ENERGY Values Expressed as % of:- Values Expressed as % of:- Total Total Net Gen losses Net Gen losses BREAKDOWN OF LOSSES 132 kV loses 8.00 41.52 4.00 34.85 33 kV line loss 1.98 10.27 1.50 13.03 SS tUf loss 1.03 5.36 0.97 8.43 11 kV line loss 2.06 10.71 1.50 13.03 LV t/f loss 1.03 5.36 0.88 7.66 LV line loss 5.16 26.78 2.64 22.99 SUM of Losses 19.27 100.00 11.48 100.00 Total transmission 8.00 41.50 4.00 34.85 Total Distribution 11.30 58.50 7.50 65.15 BREAKDOWN OF SUPPUES 132 kV Supplies 5.91 7.32 8.11 9.16 33 kV Supplies 8.86 10.97 12.49 14.11 11 kVSupplies 5.50 6.82 12.58 14.21 LV Supplies 60.46 74.89 55.34 62.51 SUMofSupplies 80.73 100.00 88.52 100.00 Total losses + supplies 100.00 100.00  - 16 - Figure 3.1 POWER SYSTEM SUPPLY HIERARCHY N=KQzki components - Sources gf losses Slies 'e Power station generators generator axiflaries W Generator transformers transfonner losses transmission O * system Iosses E IIYHV consumers (Zanzibar) _ Grid Substation transformers transformer losses 220 or 132/33 kV 33 kV line losses 33 kV consumers E C Primary Substations transformer losses 33/11 kV 11 kV line . ~ _losses 11kV consumers Supply transformers transformer losses 33 or 11 kV/400V LV line losses LV consumers  - 17 - Figure 3.2 0.080 0.070 U 33 kV line loss 0.060 - 0.050 C~~~~~~~~~~~~~~~~Ss T/f loss 0.030 ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~*~LV line loss 0.020- 0.000 TOTAL DAR ES TANGA MOS1 ARUSHA Distribution System Peak Power Losses (expressed as % power input at 33kv). 0.040 0.035 *33 kV line loss 0.030 0.025 Ss 5 TIf loss 0.015 - ~~~~~~~~~~~~~~*LV V/f 'loss 0.0 10 0.005 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~2LV line loss 0.000 TOTAL DAR ES TANGA MOSH ARUSHIA Distribution System Energy Losses (expressed as % energy supplied at 33 kv).  -18 - 2.26% Figue 33 *33kV linlot" or 33 kV Supplies 2omn *Ss Td? loss IE avs rw loss 1I kcV Supplies 1.16% OLV Vt' loss LV line loss 5.96% 3 LV Supplies Distribution System Peak Power Supplies and Loss Composition (expressed as % power input at 33kv) 1.69% *33 kv rwe loss CD 33 kV Supplies114 *SS TI loss 1.74% 11I kV line loss II kV Supplies 0v LV f loss 14.31% LV line loss IILv supplies 62.88% Distribution System Erler2y Supplies and Loss ComDosition (expressed as % e-nergy supplied at 33 kV)  -19 - Figure 3.4 j j g t~~~7,35% \33 kV line loss 9.32% Ss Tf loss 1 1: kV line loss , LV t/f loss LV Iine loss 8.89% 't) ;'