Seventh Power Project in Nepal (Loan 1011-NEP[SF])

7/31/2019 Seventh Power Project in Nepal (Loan 1011-NEP[SF])

1/40

ASIAN DEVELOPMENT BANK PPA:NEP 18082

PROJECT PERFORMANCE AUDIT REPORT

ON THE

SEVENTH POWER PROJECT(Loan 1011-NEP[SF])

IN

NEPAL

June 2004


2/40

CURRENCY EQUIVALENTSCurrency Unit Nepalese rupee/s (Nre/NRs)

At Appraisal At Completion At Operations EvaluationNovember 1989 June 2000 March 2004

NRe1.00 = $0.035 $0.014 $0.0135$1.00 = NRs28.40 NRs71.10 NRs73.55

ABBREVIATIONS

ADB Asian Development BankEA Executing AgencyEIRR economic internal rate of returnETFC Electricity Tariff Fixation CommissionFIRR financial internal rate of returnIPP independent power producerNEA Nepal Electricity AuthorityNRM Nepal Resident MissionOEM Operations Evaluation MissionPCR project completion report

PPAR project performance audit reportPPTA project preparatory technical assistanceREDT Rural Electrification, Distribution and Transmission ProjectRRP report and recommendation of the PresidentSWER single wire earth returnSDR special drawing rightTA technical assistanceTOR terms of referenceVDC village development committee

WEIGHTS AND MEASURES

GWh gigawatt-hour (1 million kWh)km kilometerkV kilovoltkWh kilowatt-hourMVA megavolt-ampereMW megawatt

NOTES

(i) The fiscal year (FY) of the Government and the Nepal Electricity Authority ends on15 July.

(ii) In this report, $ refers to US dollars.

Director General, Operations Evaluation Department : Eisuke SuzukiDirector, Operations Evaluation Division 2 : David EdwardsEvaluation Team Leader : Hong Wang

Operations Evaluation Department, PE-643


3/40

CONTENTS

PageBASIC DATA

EXECUTIVE SUMMARY

MAP

v

vi

ix

I. BACKGROUND 1

A. Rationale 1B. Formulation 1C. Purpose and Outputs 2D. Cost, Financing, and Executing Arrangements 2E. Completion and Self-Evaluation 3F. Operations Evaluation 3

II. PLANNING AND IMPLEMENTATION PERFORMANCE 4

A. Formulation and Design 4B. Achievement of Outputs 5C. Cost and Scheduling 5D. Procurement and Construction 7E. Organization and Management 8

III. ACHIEVEMENT OF PROJECT PURPOSE 9

A. Operational Performance 9B. Performance of the Operating Entity 11C. Financial and Economic Reevaluation 13

D. Sustainability 15E. Technical Assistance 16

IV. ACHIEVEMENT OF OTHER PROJECT IMPACTS 16

A. Socioeconomic Impact 16B. Environmental Impact 16C. Impact on Institutions and Policy 17

Hong Wang, Evaluation Specialist (Team Leader) was responsible for the preparation of thisreport, and conducted document reviews, key informant reviews, and guided the fieldworkundertaken by the consultants, Geoff Brown and Dilli Dahal. Vivien Ramos, Evaluation Officer,supported the team with research assistance and Anna Silverio, Administrative Assistant,provided secretarial assistance from Manila.


4/40

iv

V. OVERALL ASSESSMENT 17

A. Relevance 17B. Efficacy 17C. Efficiency 17

D. Sustainability 18E. Institutional Development and Other Impacts 18F. Overall Project Rating 18G. Assessment of ADB and Borrower Performance 18

VI. ISSUES, LESSONS, AND FOLLOW-UP ACTIONS 18

A. Key Issues for the Future 18B. Lessons Identified 19C. Follow-Up Actions 20

APPENDIXES

1. Engineering Report 212. Operational Performance of the Nepal Electricity Authority 243. Financial and Economic Reevaluation 254. Report on Socioeconomic Survey 27


5/40

BASIC DATALoan 1011-NEP(SF): Seventh Power Project

PREPARATION/INSTITUTION BUILDINGTA No. TA Project Name Type Amount

1($) Approval Date

837 Seventh Power PPTA 75,000 24 Dec 19861267 Institutional Support for Distribution Planning

and Commercial Operations of NEA

ADTA 780,000 11 January 1990

As per ADBKEY PROJECT DATA ($ million) Loan Documents Actual

Total Project Cost 64.0 65.8Foreign Currency Cost 42.7 47.7Bank Loan Amount/Utilization 51.0

40.5 (SDR)255.1

3

Bank Loan Amount/Cancellation 2.4

KEY DATES Expected ActualAppraisal 23 Jun13 Jul 1988Loan Negotiations 1620 Oct 1989

Board Approval 11 Jan 1990Loan Agreement 28 Feb 1990Loan Effectiveness 28 May 1990 18 Sep 1990First Disbursement 18 Jul 1991Project Completion Aug 1994 Sep 2000Loan Closing 31 Aug 1995 9 Aug 1999Months (effectiveness to completion) 51 120

KEY PERFORMANCE INDICATORS (%) Appraisal PCR PPARFinancial Internal Rate of Return - Part A

- Part B18.84.5

13.40.8

11.8Negative

Economic Internal Rate of Return - Part A- Part B

22.213.5

20.011.1

24.619.2

BORROWER Kingdom of NepalEXECUTING AGENCY Nepal Electricity Authority

MISSION DATAType of Mission No. of Missions Person-daysPre-Appraisal 1 72

Appraisal 1 105Follow-Up 1 10Loan Negotiation 1 12Project AdministrationReview 8 79Loan Transfer 1 8Project Completion 2 35

Operations Evaluation 1 54

ADB = Asian Development Bank, ADTA = advisory technical assistance, NEA = Nepal Electr ici ty Author ity,PCR = project completion report, PPAR = Project Performance Audit Report, PPTA = project preparatory technicalassistance, TA = technical assistance.1

The approved amount of technical assistance.2

SDR1,791,979 of the loan was cancelled on 9 August 1999.3 $6 million worth of material was purchased under loan-financed contracts, but not used in the Project.


6/40

EXECUTIVE SUMMARY

The Seventh Power Project (the Project) was formulated to help the Nepal ElectricityAuthority (NEA) improve the efficiency and reliability of its power system through rehabilitation inurban areas, and to meet the growing energy demand in rural areas through electrification . Thiswas in line with the Government of Nepals objectives for power sector development, as

described in its Seventh Five Year Plan (fiscal year [FY] 19841985 to FY19891990). Theoperational strategy for the Asian Development Bank (ADB) for Nepal at project appraisal in1988 focused on (i) improvement in agricultural productivity, (ii) containment of factors thatcause deforestation, (iii) enhancement of industrial development, and (iv) development ofsupporting physical and social infrastructure.

NEA began commercial operations in August 1985 and, at loan appraisal, had been inoperation for fewer than 4 years. Both ADB and the Government recognized the need todevelop NEAs institutional capabilities, so an advisory technical assistance (TA) grant wasattached to the Project in order to improve NEAs capacity for the planning of power distribution,and to strengthen NEAs newly created commercial department.

The main project objectives were to rehabilitate existing distribution networks in fivelarger towns to improve the quality of electricity supply and to reduce system losses, and toextend subtransmission and distribution networks to rural areas.The Project consisted of Part

A: rehabilitation of existing distribution networks in five larger towns, Part B: extension ofsubtransmission and distribution networks to rural areas, and Part C: construction of a concretepole factory.

The Project was implemented largely as envisaged. Most significant was the use of loansavings to extend the scope of rural electrification. The main factor contributing to the large loansavings was participation of Chinese manufacturers in international competitive bidding after1992. Other significant factors included depreciation of the dollar against special drawing rights(SDRs), the ability of Nepalese manufacturers to bid for supply of project equipment such asdistribution transformers, and increased competitiveness of Nepalese installation contractors.Furthermore, the price contingency was calculated on the assumption of a constant nominalexchange rate, and was high. These factors, taken together, not only kept costs below or nearbudgeted levels, but also allowed allocation of contingencies for the extension of ruralelectrification.

The Project experienced significant delays. The loan was closed on 9 August 1999,4 years after the original closing date. Construction of the project works was completed only in2000. A major cause of implementation delay was ADBs lengthy process for approval of NEAsrequest to use loan savings to extend rural electrification. Other delays were from (i) NEAdelays in meeting conditions for loan effectiveness and in mobilizing the consultant for projectimplementation; (ii) the time required to fully evaluate the use of single wire earth return (SWER)distribution technology; (iii) delays in award of contracts; and (iv) poor contractor performance.

Upon completion, the Project achieved the following physical outputs: (i) about 185megavolt-ampere (MVA) of new 132 kilovolt (kV), 66 kV, and 33 kV substation transformercapacity; (ii) more than 320 kilometers (km) of new 33 kV subtransmission line; (iii) almost 2,500km of high voltage distribution line; (iv) more than 2,300 km of low voltage distribution line; and(v) a new concrete pole factory with a daily production capacity of as much as 58 prestressedconcrete poles. Through the Project, electricity was provided to almost 1,000 previouslyunelectrified load centers, providing access to electricity to more than 200,000 new customers.


7/40

vii

The quality and reliability of power supply in five major towns and two significant load corridorswere improved.

Sustainability of project benefits over the longer term will depend on the quality ofmaintenance, and the availability of technical and financial resources to upgrade and reinforceparts of the system that are loaded to their rated capacity. All work necessary to keep the

network operational has been done, but the quality of maintenance varies. Engineeringsupervision of operation and maintenance is inadequate in some places.

NEAs financial performance was not satisfactory from 1999 to 2003. This was mainlybecause of NEAs failure to increase tariffs to more sustainable levels. The average tariff forFY2003 was NRs7.02/kWh42% lower than envisaged. A further problem was NEAs failure toreduce system losses to the forecast levels. Actual losses for FY2003 were 24.5%, comparedwith forecast losses of 18.7%. NEA now faces a difficult operating environment. The Maoistinsurgency in Nepal continues unabated, and NEA facilities are an ongoing target. Insurgencydamage for FY2003 cost NRs72.4 million. But the insurgencys impact on NEA operations ismore important than the financial cost of physical damage. The insurgency impact involves boththe quantifiable costs of protecting NEA assets, and the less-quantifiable costs of having toconsider practical and political implications when making management decisions.

Electrification in urban areas and areas near sealed roads has affected local economiespositively because of its benefits to consumers lives, and stimuli for industrial and commercialdevelopment. In remote areas, consumers benefits from improved lighting, access to radio andtelevision, and the use of fans would not be possible without electricity. Electricity has benefitedwomen greatly. But many poorer households in remote areas have not connected to electricalpower because the initial connection is too costly.

The Project, as originally designed, had no specific environmental objectives.Construction has been environmentally benign, with most overhead lines located on roadreserves. Project beneficiaries have seldom complained about environmental issues. TheProject was expected to have an indirect environmental impact by utilizing Nepals indigenoushydropower facilities to replace fuelwood and imported fossil fuels. But the Project seems tohave done little to reduce the use of fuelwood, although the use of kerosene, and small dieselmotors, has decreased.

The performance of both ADB and NEA is assessed as partly satisfactory. For ADB, thisassessment reflects a failure to take proactive and timely action to respond to implementationproblems that arose early in the Project, late decisions regarding the first loan extension, and anaborted decision to transfer project administration to the Nepal Resident Mission (NRM). ForNEA, the partly satisfactory assessment is based on unsatisfactory progress over the first 5years (through 1995), its failure to reduce losses to covenanted levels, and the poor financialperformance that continues to threaten the Projects sustainability.

Overall, the Project is rated successful. But TA achievements fell short of expectationsbecause of NEAs lack of commitment to immediately implement TA recommendations, and toprovide sufficient counterpart funds to make the TA more effective. Overall, the TA is assessedas partly successful.

The Project has yielded three main lessons. First, delays associated with abandoningthe use of SWER might have been avoided if a more substantive project preparatoryconsultancy had been undertaken, and the technical feasibility of using SWER been researchedmore thoroughly during appraisal. On one hand, innovation by consultants should beencouraged. But a project of this size and complexity should include a full-scale project


8/40

viii

preparatory TA that includes investigation into the use of new and alternative technologies.Second, annual review missions were insufficient during the first 5 years of projectimplementation. Third, many problems associated with poor performance of the contractorswere not unique to the Project, and should be avoided in future. An ADB procedure to evaluateperformance of both international and local contractors, similar to evaluation procedures usedfor individual consultants, would be useful. The performance of contractors would be recorded in

files for eligibility check before qualification for future contracts.

An issue relevant to ADBs ongoing Rural Electrification, Distribution, and TransmissionProject (REDT) deserves particular attention. The rural electrification component of REDT willprobably further constrain NEAs financial position, considering NEAs poor financialperformance and the unlikelihood of significant tariff increases in the near future. It may beuseful to explore the possibilities of subsidizing initial connection charges for rural customers,and reducing their costs of payment of electricity bills, to raise the tariff for rural electricity supplyto at least a break-even level.

Eisuke SuzukiDirector GeneralOperations Evaluation Department


9/40


10/40

I. BACKGROUND

A. Rationale

1. Nepals Seventh Power Project1 (the Project) was consistent with the Governments powersector objectives as described in the Seventh Five Year Plan (FY198485 to FY198990). The

Seventh Plan included development of indigenous energy resources to substitute for importedfuels, providing a reliable power supply, and reducing fuelwood consumption. The Project was tohelp the Nepal Electricity Authority2 (NEA) improve the efficiency and reliability of its powersystem through rehabilitation in urban areas, and meet the growing energy demand in rural areasthrough electrification. The Project aimedspecifically at (i) rehabilitating and expanding existingdistribution networks to meet continuing load growth, and (ii) electrifying rural areas that had noaccess to electricity. The Asian Development Banks (ADB) operational strategy for Nepal at thetime focused on (i) improvements in agricultural productivity, (ii) containment of the causes ofdeforestation, (iii) enhancement of industrial development, and (iv) development of supportingphysical and social infrastructures.

2. NEA began commercial operations in August 1985. At loan appraisal, NEA had operated

for less than 4 years. At that time, NEA had no commercial department to provide a focal point forits commercial and loss reduction activities. It was also recognized that detailed planning fordistribution development over the next 10 years was a major undertaking, beyond the capability ofthe limited staff in NEAs distribution and customer services departments. Work was limited tospecific multilaterally financed projects, and was hampered by a lack of basic system data andgeographic diagrams of the existing 11-kilovolt (kV) networks. Both ADB and the Governmentrecognized the need to develop NEAs institutional capabilities. An advisory technical assistance(TA) grant3 was attached to the project loan to support NEA in the development of institutionalcapability.

B. Formulation

3. The Project was initiated through a small-scale project preparatory technical assistance(PPTA)4 under which two consultants developed the project scope and conceptual design. ThePPTA was followed by an ADB appraisal mission in mid-1988.5 The scope of the projectcomponents was described in some detail in the appraisal report.6 For the rehabilitationcomponent (Part A of the Project), the appraisal report identified the location and size of alltransmission and area substation equipment, and provided estimates of the lengths oftransmission and distribution lines to be installed. The rural electrification component (Part B) wasdefined in more detail. The conceptual design included extensive use of single wire earth return

1Loan 1011-NEP (SF): Seventh Power Project, for $51 million equivalent, approved 11 January 1990.

2NEA, created under the National Electricity Act of 1984, is a government-owned corporation responsible for planning,constructing, and operating all public power utilities in Nepal.

3 TA 1267-NEP: Institutional Support for Distribution Planning and Commercial Operations in the Nepal ElectricityAuthority, for $780,000, approved 11 January 1990.

4TA 837-NEP: Seventh Power Project, for $75,000, approved 24 December 1986.

5Due to the Governments slow pace in resolving key issues before loan negotiation, a short follow-up mission toupdate the appraisal report was fielded in May 1989. The remaining issues were (i) transfer of the financialresponsibilities of the ongoing ADB projects to NEA, (ii) establishment of a commercial directorate in NEA, (iii)commencement of a study on verification and valuation of fixed assets, (iv) submission of a schedule for liquidation ofGovernment arrears, (v) conversion of NEAs accounts payable to the Government into equity, and (vi) an action planfor timely preparation of audited financial accounts in accordance with the previous loan covenants. Loan negotiationwas eventually held in October 1989.

6 ADB. 1989.Appraisal of the Seventh Power Project in Nepal. Manila (Report LAP:NEP 18082).


11/40

2

(SWER) technology for electricity distribution, which the PPTA consultant considered aneconomically effective way to distribute electricity in rural areas where demand is low. Becausethis would be the first introduction of SWER technology into Nepal, it was considered necessarythat the project implementation consultant carried out a detailed technical investigation andconstruction of a pilot SWER scheme in the Kathmandu Valley was included in the project scope.In both cases the project cost estimates were based on comprehensive bills of quantities.

C. Purpose and Outputs

4. The main project objectives were to rehabilitate existing distribution networks in five largertowns to improve their quality of electricity supply and reduce system losses, and to extendsubtransmission and distribution networks for rural areas.

5. The Projects rehabilitation component was to contribute to NEAs loss reduction targets of20% and lower from FY1994 onward, to improve the reliability of supply to consumers, and tomeet the growing demand for electricity. The Project also included extension and rehabilitation ofelectricity supplies in two semi-industrial corridors that had poor reliability of supply and highdemand growth. Both corridors are along main roads to the Indian border. Thus, continuing

industrial development would support Nepals export trade with India. The Projects ruralelectrification component was to support the connection of about 83,500 new consumers in ruralareas. This would support the Governments objective of bringing electricity to rural areas, andsubstituting kerosene, diesel fuel and fuelwood by electricity generated from Nepals indigenoushydroelectric resources.

6. The Project, as defined at the time of appraisal, had three main parts:

(i) Part A: Rehabilitation and extension of the subtransmission and distributionnetworks in the towns of Hetauda, Birunj, Butwal, Bhairahawa, and Dharan; and in the corridorsbetween Hetauda and Birgunj, and between Butwal and Bhairahawa.

(ii) Part B: Electrification of six unelectrified rural areas: Ilam-Damak, Dharan, Siraha,Malangawa-Gaur, Bharatpur-Parasi, and Tamghas-Sandhikarka. The unelectrified areas were tobe supplied from the existing interconnected system, and would not require construction of newgeneration plants.

(iii) Part C: Provision of two concrete pole plants in central and eastern Nepal;construction and maintenance equipment, and vehicles; and project support facilities.

7. The attached TA (footnote 3) was to improve NEAs capacity for planning of electricitydistribution, and to strengthen its newly created commercial department. It included (i)preparation of a 5-year rolling distribution development plan to decentralize distribution planningby establishing planning units in regional branch offices; and (ii) provision of on-the-job training in

consumer accounting, commercial policy development, and loss reduction.

D. Cost, Financing, and Executing Arrangements

8. At appraisal, the estimated project cost was $64.0 million, comprising $42.7 million inforeign exchange and $21.3 million in local currency. ADB was to finance 80% ($51.0 million) ofthe project cost, including all foreign exchange costs, and $8.3 million of local costs, from itsSpecial Funds resources. The loan terms included a 1% service charge, a 10-year grace period,


12/40

3

and a 40-year maturity. The Kingdom of Nepal was the Borrower, and NEA was the ExecutingAgency (EA).

9. The loan proceeds were made available to NEA under a subsidiary loan agreement. Therelending terms to NEA for rehabilitation of distribution systems, supply of materials andequipment, and consulting services included a 10.25% interest rate, a 25-year repayment period

including a 5-year grace period.7 Loan proceeds for the rural electrification component were relentto NEA at 1% interest. The Borrower assumed the foreign exchange risk of the entire loan.

10. NEA set up a dedicated project implementation unit in Kathmandu, which was responsiblefor overall project management, engineering and design, and for contract administration.Subsidiary project offices were established in the eastern, central, and western regions. Theywere responsible to the project implementation unit for stores management and supervision of theimplementation contracts. The regional project offices were also responsible for the operation andmaintenance of each subproject until its formal handover to the NEA local branch office.

E. Completion and Self-Evaluation

11. ADBs project completion report (PCR), circulated in July 2001, rated the Project assuccessful. But the PCR was prepared before local contractors had fully completed projectinstallation, so the final project cost had to be estimated.8 The PCR discussed many problemsduring project implementation and concluded that the performance of the implementationconsultant, and some contractors and suppliers, was unsatisfactory. Accounts of the Projectsenvironmental and social impacts were limited. The PCR did not assess the relevance, efficacy,efficiency, sustainability, or institutional and other impacts of the Project. Lessons learned weremainly with respect to administrative difficulties during project implementation. Recommendationswere made in terms of closely monitoring completion of the unfinished project components,financial health of NEA, and attainment of project benefits. The PCR provided no information onimplementation of the attached TA.

F. Operations Evaluation

12. This project performance audit report (PPAR) reviews PCR findings and assesses theProject in terms of relevance, efficacy, efficiency, sustainability, and institutional and otherdevelopmental impacts. The assessment is based on a review of ADB documents, discussionwith ADB staff, and findings of the Operations Evaluation Mission (OEM). The OEM visited Nepalfrom 25 February to 12 March 2004, and held discussions with representatives of the Ministry ofFinance, Ministry of Water Resources, Electricity Tariff Fixation Commission (ETFC), and NEA.The OEM visited some project sites and districts electrified under the Project in the flat areas ofthe western, central, and eastern regions.9 During the field visits, the OEM had discussions withmanagers and engineers maintaining project facilities and carried out a reconnaissance-levelphysical inspection of some facilities constructed under the Project. The OEM also carried out

socioeconomic surveys in Siraha, Chitwan, Nawalparasi, and Rupandehi districts. The views of

7Both ADB and the World Bank required that the relending terms be on a commercial basis.

8The PCR mission placed significant emphasis on correctly reporting project costs, and indicated dissatisfaction aboutnot being able to accurately quantify costs. But the estimate of the total project cost is probably reasonably accuratebecause it was based on the noncompleted amount of the local construction contracts with fixed costs.Nevertheless, there could be inaccuracies in the cost of individual project components since the allocation of someequipment supply contracts to individual project components had to be estimated. Because the cost breakdown in thePCR was comprehensive and the best cost assessment available, it is used in this report.

9 OEMs planned travel to hill districts of eastern Nepal was restricted because of security problems.


13/40

4

concerned ADB departments and offices, and those of the Government and EA, were consideredwhen finalizing the PPAR.

II. PLANNING AND IMPLEMENTATION PERFORMANCE

A. Formulation and Design

13. The Project was formulated to rehabilitate distribution networks in five major towns inorder to improve the quality of electricity supply and reduce system losses, and to extendsubtransmission and distribution networks to rural areas in the terai (flat plains) and hills(mountains that are generally not covered by snow). The Project was expected to directly benefitdomestic consumers and commercial and industrial users in urban and semi-urban areas, as wellas users of the existing irrigation facilities that were using imported fuels, and about 84,000 newconsumers in rural areas. The Project conformed to the Governments strategy for the powersector in the short and medium term, and was in line with ADBs operational strategy for Nepal.The Project was a continuation of ADBs long-term involvement in the Nepal power sector.10

14. The extent of SWER to be included in the Project was described in detail. The terms of

reference (TOR) for the project implementation consultant included a detailed technicalinvestigation to determine the appropriateness of using SWER technology in Nepal. Acomprehensive study into the use of SWER technology in the project areas was undertaken atproject inception. Technical problems were found to make successful implementation of SWERdifficult. The high population densities in the terai project areas make the use of SWERtechnology problematic and potentially uneconomic.11 The hill areas in the Project have lowerpopulation densities and are better suited to the use of SWER. But ground conditions were foundto be such that designing a suitable earth connection would be difficult, making the use of SWERtechnology potentially hazardous, particularly during the dry season. In retrospect, these technicalissues should have been addressed at the project design stage.

15. Once the decision was made not to use SWER technology, it became necessary to revise

the design of the rural electrification components. The implementation consultant reviewed theconceptual designs for the six rural electrification areas, and proposed revised designs usingconventional three-phase distribution. The review used the best costs available at the time, andreformulated each subcomponent to meet ADBs financial and economic criteria.12 As a result,some villages were deleted from the Project, and others added. The number of households thatcould potentially be connected to the rural electrification scheme increased because of thetechnical reformulation. The consultants report estimated that, after 10 years, 144,600 newconsumers would benefit from the rural electrification.

10Of 15 major towns where earlier ADB studies indicated an urgent need for rehabilitation and expansion of existingdistribution networks, 4 were rehabilitated under the Fifth Power Project (Loan 670-NEP[SF] for $20.0 million,approved 14 December 1983). Two more towns were rehabilitated under the Sixth Power Project (Loan 708-NEP[SF]

for $28.1 million, approved 20 November 1984). Electrification of five previously unelectrified rural areas was alsocompleted under the Sixth Power Project. ADB is now financing further rehabilitation of distribution networks, andelectrification of 22 rural areas, under the Rural Electrification, Distribution and Transmission (REDT) Project (Loan1732-NEP[SF] for $50 million, approved on 21 December 1999).

11SWER technology is applied most successfully in areas of low population density with few distribution substations perkm of line. Safety considerations limit the total load on an SWER circuit. The need for an isolating transformer at theconnection to the conventional distribution network means that circuit lengths of less than about 10 km are seldomeconomical. The project consultant found that, while short-term use of SWER on the teraiwas technically possible,the forecast load growth indicated that long-term SWER use might not have been sustainable.

12A positive financial internal rate of return (FIRR) and a minimum economic internal rate of return (EIRR) of 10% wererequired for the rural electrification subprojects.


14/40

5

16. Soon after redesign of the rural component of the Project, and tendering of the mainequipment supply contracts, it became apparent that equipment costs would be lower thanestimated at appraisal. This was mainly due to participation of Chinese contractors in theinternational competitive bidding process after 1992. At the same time, depreciation of the USdollar against the special drawing rights (SDR)13 meant that the loan in dollar terms had

increased. Hence, an additional $26 million was available. To utilize the loan savings, ADBagreed with NEA proposals for additional rural electrification in 22 areas.14 The rehabilitationcomponent of the Project was largely implemented as designed at appraisal. But only oneconcrete pole manufacturing plant, instead of the two plants envisaged, was considerednecessary and built.

B. Achievement of Outputs

17. The Project included installation of the following electricity reticulation: (i) about 185megavolt-ampere (MVA) of new 132 kilovolt (kV), 66 kV, and 33 kV substation transformercapacity; (ii) more than 320 kilometers (km) of new 33 kV subtransmission line; (iii) almost 2,500km of high-voltage distribution line; and (iv) more than 2,300 km of low-voltage distribution line. As

a result of the Project, electricity was reticulated to almost 1,000 previously unelectrified loadcenters, providing electricity access to more than 200,000 new customers. In addition, new 11 kVand 400 V distribution lines were constructed or rehabilitated to improve the quality and reliabilityof supply in five major towns and two significant load corridors.15 Furthermore, a new concretepole factory was built, with a daily production capacity of as many as 58 pre-stressed concretepoles. A pilot SWER scheme, designed to supply up to 600 domestic customers, was constructedin the Kathmandu Valley using project funds, but was abandoned soon after the decision not touse SWER technology in the Project. The standard of construction observed by the OEM wasgenerally good. But a number of engineering issues and problems were identified (discussed in

Appendix 1).

C. Costs and Scheduling

18. The actual project cost was $65.8 million.Loan disbursements of $55.1 million included$53.7 million in foreign exchange cost16 and $1.4 million equivalent in local currency. TheGovernment funded the remaining local currency cost of $10.7 million equivalent. An undisbursed$2.4 million equivalent was canceled on loan closure. The disbursement of $4.1 million higherthan the appraised loan of $51.0 million was caused by depreciation of the dollar against theSDR.

19. Table 1 compares project cost estimates at appraisal with actual costs at completion. Mostsignificant is the $19.7 million for extension of the rural electrification component. Thisextensionmore than 25% of the total project costis explained in the PCR as having beenmade possible by Chinese contractors participation in the bidding, which reduced bid prices. All

13In 1994, the exchange rate was $1.45: SDR vs $1.25: SDR at appraisala 16% depreciation of the dollar againstSDR.

14ADBs decision was based on a technical and economic study that NEA prepared, using the same methodology thatthe consultant used in the inception study.

15The OEM was unable to quantify the extent of the distribution line rehabilitation undertaken under Part A of theProject, but visited three of the rehabilitated areas during the field visit.

16 Material valued at $6 million was purchased under loan-financed contracts, but was not used in the Project.


15/40

6

project components had cost overruns, except for the original rural electrification component.17 Asubstantial component of the project cost was material procured under international competitivebidding, so participation of Chinese manufacturers in these material supply contracts wasundoubtedly a contributing factor. Other significant factors were Nepalese manufacturers abilityto bid to supply project equipment such as distribution transformers, and increasedcompetitiveness of Nepalese installation contractors. Furthermore, the price contingency was

calculated assuming a constant nominal exchange rate. These factors, combined, not only keptcosts below or near budgeted levels, but also allowed allocation of contingencies to the scopeextension in rural electrification.

Table 1: Summary of Appraised and Actual Project Costs($ million)

Item Appraisal Estimate Actual CostComponents

Foreign Local Total Foreign Local Total

Part A Rehabilitation 6.0 1.9 7.9 7.3 1.7 9.0Part B Rural Electrification 24.0 9.2 33.2 33.7 15.0 48.7

Original Scope 24.0 9.2 33.2 22.0 6.0 28.0Extension 11.7 9.0 19.7

Part C Concrete Pole Plant/Tools &

Equipment

2.2 0.5 2.7 3.0 1.3 4.3

Part D Consulting Services 1.7 0.3 2.0 2.2 0.1 2.3Physical Contingencies 1.7 0.6 2.3Price Contingencies 6.8 4.4 10.2Taxes and Duties 1.7 1.7NEA Project Management 0.3 0.3Interest During Construction 1.3 2.4 3.7 1.5 1.5Total 42.7 21.3 64.0 47.7 18.1 65.8

NEA = Nepal Electricity Authority.Source: ADB. 2001. Project Completion Report on the Seventh Power Project in the Kingdom of Nepal. Manila.

20. The loan was approved on 11 January 1990, with an original closing date of 31 August1995. After three extensions, the loan was closed on 9 August 1999. NEA completedconstruction of project works in September 2000, after successfully resolving a few difficult

right-of way issues.18 A comparison of the appraisal schedule with the actual projectimplementation is given in Appendix 8 of the PCR. A major cause of implementation delay wasthe lengthy process within ADB to approve NEAs request to utilize loan savings for theextension of the rural electrification component.19 The extended project scope required asecond round of procurement, which did not begin until early 1996. Other delays arose from (i)delays by NEA in meeting the conditions for loan effectiveness,20 (ii) delays in mobilizing the

17Cost overruns in Parts A and C may partly reflect the fact that large portions of these components were implementedusing turnkey contracts rather than material supply and local installation where successful bidders were mostly

Chinese and Nepalese manufacturers. The cost overrun in Part C was in spite of the fact that only one of the twopole plants included in the original project scope was constructed.

18Project land acquisition was minimal. Compensation for right-of-way of distribution and transmission lines was notincluded in Nepals standard rules and regulations at project implementation. Rights-of-ways were obtained at nocost, mainly through consultation with affected people. No complaints regarding the issue of right-of-way werereported during the field visit.

19NEA originally made this request in August 1994, and approval took 14 months. A review mission fielded in

September 1994 expressed concern about delays in implementing the original scope and recommended thatapproval be deferred until NEAs implementation performance improved. A subsequent review mission in June 1995noted improved implementation performance and recommended that the request be granted.

20 The loan did not become effective until 18 September 1990, 8 months after Board approval.


16/40

7

project consultant,21 (iii) the time required to fully evaluate the SWER technology, (iv) delays inaward of contracts, and (v) poor performance of some international contractors. A further delaywas because ADB did not approve the first extension of the loan closing date until 17 November1995. That delay stopped project work for 3 months, because contractors would not continue after

ADBs original letters of credit expired.

D. Procurement and Construction

21. The Project required 89 different contracts, and project works stretched over almost 700km from east to west. Bid documents for most project turnkey and equipment contracts as scopedat appraisal were issued in December 1992. Bids were opened in February 1993, but the firstcontracts were not awarded until August 1993; some contracts were not awarded until mid 1994.Delays were due to the volume of contracts to be processed, and to an exceptionally largenumber of representations from unsuccessful bidders. In its PCR, NEA alleges that somerepresentations were made on the basis of forged documents supporting the complainants abilityto complete the contract works. Most of these representations were received before orimmediately after ADB received the bid evaluation reports, and before the bid results were madepublic. At the time, there was a strong suspicion that somebody was prematurely leaking the

results of NEAs bid evaluations to bidders, but there is no record that ADB took action inresponse.

22. Some contractors performed poorly. First, at least two contractors, after receiving most oftheir contract payments following successful acceptance inspection at the contractors works, didnot dispatch the materials to the sites. The matter was resolved in one case, but the equipmentwas never dispatched in the other case, and the contractors bank guarantee was forfeited.Second, an international turnkey contractor was paid the amount of the contract in foreigncurrency after equipment had been delivered. But the contractor was unable to repatriatesufficient funds back to Nepal to complete the erection work because of its own governmentscontrols on foreign currency. This caused delays. Third, the contractor for supply and delivery ofconductors defaulted, claiming that his costs had increased due to a steep rise in the cost of

aluminum. The contractor was bound by a fixed-price contract, so the claim could not beconsidered. The contract had to be terminated, and the supply of material was re-bid. Finally,single-phase customer meters were found on arrival at site not to meet the specified accuracyrequirements. As a result, the contractor sent a team of four technicians with two test benches toKathmandu to retest and adjust all meters over an 8-month period. This delay led to such anacute shortage of meters that many rural households had to wait more than a year beforeprovision of a requested connection. Such irregularities affected only a few contracts, but had aserious impact on project completion since all required material had to be made available beforefull completion of erection works.

23. The supervision of contractors was also unsatisfactory in some cases. Personnel who didfactory inspections sometimes did not fully understand project requirements, and accepted

equipment that did not fully meet project needs. In other cases, contractors delivered the wrongequipment, even after factory inspection. Other delays were caused by inadequate supervisiondue to a lack of qualified personnel assigned to the Project, and by landowners refusing thecontractors access during the monsoon season to prevent crop damage. The PCR also noted that

21The consultant was appointed in 1989, before loan approval. When the loan became effective, 1 year after contractnegotiations, it was necessary to renegotiate parts of the consultants contract and approve changes in personnel.NEA initially insisted the original contract should stand unchanged. As a result, a dispute developed between NEAand the consultant, which was only resolved after ADB intervention. Consequently, the consultant was not mobilizeduntil July 1991.


17/40

8

the impact of monsoons on line erection was not adequately considered when planningconstruction works.

E. Organization and Management

24. The large number of contracts and widespread project areas made the Project particularly

hard to organize and manage. NEA set up a project implementation unit in Kathmandu withoverall project responsibility, especially for contract administration. Subsidiary project offices,responsible for construction supervision, were established in the western, central, and easternregions. Separate project stores were maintained in each subsidiary office. The Project retainedresponsibility for project works until they were formally handed over to the relevant branch offices.This approach was considered appropriate for a project of this magnitude and complexity.

25. At project completion, about $6.0 million worth of equipment procured through loan fundshad not been used in the Project. That equipment was eventually transferred to NEA stores. Thiswas 9% of the total project cost, and 12% of the rural electrification component. Much of theequipment was later used to construct further rural electrification works outside the scope of theProject, which may not have been subject to ADBs financial and economic approval criteria or

safeguard policies. Furthermore, equipment ordered under each contract had to be allocatedamong different subcomponents. This required a complexity of project accounting that may havebeen overlooked. Future projects of this type should have better inventory control.

26. NEA satisfactorily complied with covenants related to project and corporate organization,although with some delay in internal audit function. Other loan covenants included (i) electricitylosses reduced from 26.5 % in FY1990 to 20% from FY1994 onward, (ii) a self-financing ratio ofat least 23% from FY1998, (iii) a 6% rate of return on the gross book value of fixed assets fromFY1991, and (iv) a debt service ratio of 1.2 from FY2001. Total system losses remained highthroughout project implementation, and are currently 24.5%. According to the PCR, at loanclosure in 1999 NEA achieved the required ratios of self-financing and debt service coverage, butits return on rate base was only 5.4%.22 NEAs financial performance has deteriorated since then,

to the extent that NEA has not made a pre-tax profit since FY2000 (Appendix 2). Auditedaccounts were consistently submitted late throughout project implementation.

27. The project implementation consultant remained on the Project from mid-1991 until mid-1995 when the time provided in its contract was expended. According to the PCR, theconsultants performance was not satisfactory. The PCR noted that revised designs for ruralelectrification were incomplete and misleading. The consultant was also blamed for delays inseveral activities, including tender invitation for procurement and erection works.

28. The PCR assessment of the consultants performance relies on NEA assessment in itsPCR, and is not supported by evidence in ADB files. An ADB review mission in June 1992, 12months after consultant mobilization, noted a communication problem between the consultant and

NEA. The back-to-office report of the next review mission, in May 1993, noted that the situationseemed to have been resolved, and that communication between the two parties was good. TheOEM reviewed the consultants engineering design report on the first phase of the Project, and

22There are discrepancies between the financial indicators for FY1999 as reported in the PCR and in the report andrecommendation of the President (RRP) for the REDT loan. Most significant is the self financing ratio, which isreported as 28% in the PCR but only 9% in the RRP. The returns on rate base in the two reports also differ but arenot comparable because in the RRP it is measured against the revalued fixed asset base, rather than the bookvalue.


18/40

9

considered its standard high. This design report was used as a template for subsequent NEAreports seeking approval for scope extensions of rural components.23

29. Throughout the Project, ADB sent review missions roughly on an annual basis. Thesemissions were particularly important before 1996, when project progress was consideredunsatisfactory. Extension of the loan closing date, and extension of the project scope to include

more rural electrification, appear to have been rolled together.24 But these were two separateissues, and should have been treated as such. Then in mid-1996 it was decided that projectadministration would be transferred to the Nepal Resident Mission (NRM). The decision wasreversed in January 1997, mainly because of constraints in resources at NRM.25 Fifteen boxes ofproject documents were transferred from Manila to Nepal, then returned to Manila 6 monthslater. For preparation of this report, only three boxes of project files could be located. Significantdocuments, which would have given better understanding of problems during the Project, weremissing. For example, no documents could be found relating to the demobilization of the projectimplementation consultant in July 1995.

30. Project implementation suffered from a lack of more proactive supervision when theProject was in difficulty during the early years. In such circumstances, relying on annual review

missions for feedback on project performance is unsatisfactory. Regular monthly or quarterlyreview meetings with the EA, and possibly the consultant, would have been better. This couldhave provided ADB and the Borrower earlier warning of difficulties, allowing corrective actionbefore the situation deteriorated.26 The improvement in performance after 1995 seems to havebeen due to a change of NEA project manager, rather than to changes in how ADB administeredthe Project.

III. ACHIEVEMENT OF PROJECT PURPOSE

A. Operational Performance

1. Rehabilitation Components and Subtransmission

31. Projects rehabilitation components were undertaken in strategic locations that wereimportant to the national economy, either through proximity to main border crossings with India or,for Dahran, to important agricultural export industries such as tea. Therefore, the rehabilitatedsystems have been subject to high load growth, and will probably reach their technical designcapacity well before expiration of the life of the assets. Indeed, some subcomponents of theProject reinforced distribution work completed under the Sixth Power Project. Some powertransformers installed under the Project have already been replaced with larger units. The Projecthas ensured that customers in the area continue to receive a power supply of acceptable quality.

23The one NEA person the OEM spoke with who had worked with the consultant considered the consultants worksatisfactory.

24ADB approval for both extending the loan closing date and the project scope was given on 14 November 1995, at atime when little progress had been made on installation of the first phase of rural electrification. The PCR suggeststhat the main reason for the delay in approving the scope extension was NEAs failure to provide all requiredinformation. But the OEM review of the project files indicates that progress on the Project was the major concern.

25Another reason was the large number of procurement packages and representations associated with the Project. Itwas decided that the Project would be managed more efficiently from ADB headquarters.

26The OEM was informed that NRM started monthly meetings with project managers/directors of ADB-assisted projectsto discuss ongoing problems in late 1993. The monthly meetings afterward were intensified by discussing all issuesidentified in project review missions. The status of each project was reported to concerned departments at ADBheadquarters. But the project files include no records of the monthly meetings, nor of correspondence between NRMand ADB headquarters.


19/40

10

This is particularly important for industrial customers, who are an important stimulus for continuingeconomic growth. Sustainability over the longer term will depend on the quality of maintenance,and on availability of technical and financial resources to upgrade and reinforce parts of thesystem that become loaded to their rated capacities. Until now, all work necessary to keep thenetwork operational has been done.

32. Maintenance quality varies across the network and, especially for area substations, isdependent on the enthusiasm and ability of the local maintenance staff. In places wherehousekeeping was good, maintenance was clearly proactive and of high standard. In othersituations, only the minimum maintenance necessary to keep the equipment operational is done.Of concern is the fact that many staff have little regard for the proactive avoidance of failure inday-to-day network operations. For example, the low-voltage, molded case circuit breakers ondistribution substations are not designed for exposure, and are mounted in weatherproof cabinets.But the cabinet doors are routinely left open, or sometimes even removed, so that staff canoperate the circuit breakers from the ground using a stick. OEM members also saw a number ofinstances where temporary repairs had become permanent, either because replacement partswere not available, or through failure to return to the sites to make more permanent repairs. Othersustainability concerns are largely outside the control of management and staff. The 11-kV indoor

circuit breakers used throughout the Project are no longer manufactured, so spare parts cannotbe obtained. One substation switchyard and control room was flooded to about 1 meter during the2003 monsoon season.

33. Engineering supervision of operation and maintenance is inadequate in some places. TheOEM requested data on the demand and energy growth for all transmission and subtransmissionsubstations constructed under the Project. This information should be readily available, becausesubstation operating staff routinely record it hourly. The necessary data for many substations wasprovided, although after difficulty, but data for other substations were not available. In anotherinstance, the OEM found load data recorded in a substation log that was clearly erroneous, as itwas inconsistent with other data recorded in the same log. Inconsistent data appeared to havebeen reported to management on a weekly basiswith no one questioning its accuracysince

the substation was commissioned.

2. Rural Electrification

34. The Projects rural electrification components generally appear to have been installedaccording to the revised design. But operationally, those components also suffer from many of thesame problems as the rehabilitation components, particularly in maintenance quality. Servicedrops between the low-voltage reticulation and the customer metering points do not have properconnectors, and connections are made by simply twisting the service drop conductors around thedistribution line conductors. Over time, this will cause loose and high-resistance connections thatwill, in turn, cause low voltage for customers and higher system losses. In one case, a weightappeared to have been attached to the service drop conductor in order to improve the connection

quality. Illegal connections are endemic in some areas. Many of these connections were easy todetect and, in some villages, the OEM noted at least one illegal connection on each span. Otherillegal connections may also be present, but are attached more skillfully, so they appear


20/40

11

legitimate. Most poles adjacent to residences had outside lights, probably illegal, connected to thelow-voltage conductors.27

35. About 95% of the customers in less-accessible rural villages is residential. NEA claimsthat 90% of these customers pays the minimum lifeline charge of NRs80 per month, but many donot fully use their entitled 20 units per month. But, based on an inspection of billing records in only

one location, the OEM believes the average customer consumption could be higher, particularly inthe summer when many households use fans. Nevertheless, few remote residential customerspaid more than NRs200 per month. This equates to consumption of 36 units at an average tariff ofNRs5.6 per unit. NEAs average generation cost for electricity generated from its own plants inFY2003 was NRs4.03 per unit. But NEA pays large independent power producers (IPPs) anaverage of NRs6.1 per unit, which the OEM considers a reasonable proxy for the marginal cost ofgeneration.28 On this basis, revenue from electrification of remote villages is insufficient to covereven the generation cost, and does not provide for other costs resulting from the Project,including system losses, operation and maintenance, and the cost of servicing NEAs debt to theGovernment. Therefore, from a financial perspective, the electrification of remote villages isunsustainable at existing tariff levels. The current tariff structure contains some cross-subsidiesamong various consumer categories, but they are relatively restricted. Considering the

importance of rural electrification in both political and economic terms, NEA will do its best to keepthese rural systems operational.

3. Concrete Pole Factory

36. The concrete pole factory in Amlekhganj, in southern Kathmandu, continues to produce 8-,9-, and 11-m prestressed concrete poles of good quality. By January 2004, the factory hadproduced more than 77,600 units. The factory is financially independent. It sells its poles to NEAat a transfer price that provides an adequate margin over its own operating costs, but issignificantly lower than purchasing from the private sector (Table 2).

Table 2: Performance of the Pole Factory

PolesTotal Productionto January 2004

Current ProductionCost(NRs)

Transfer Priceto NEA(NRs)

Market Price forNEA(NRs)

8-meter 45,316 2,014 2,520 3,4749-meter 23,910 2,712 3,270 3,964

11-meter 8,416 4,210 4,945 8,000NEA = Nepal Electricity Authority, NRs = Nepalese rupees.Source: Nepal Electricity Authority.

B. Performance of the Operating Entity

37. Appendix 2 compares NEAs financial performance from1999 to 2003 with performance asforecast in the 1999 report and recommendation of the President (RRP) for the Rural

27When asked about this, an NEA staff explained that village development committees paid for the street lights. But theinstallation quality of many outside lights was poor, indicating they had not been installed professionally.Furthermore, the monthly tariff for an unmetered street light is NRs1,860 per kVA, equivalent to NRs74 for a 40-wattbulb. This is high compared with the NRs80 minimum charge payable by low-use consumers.

28NEAs reported average cost of generation does not include IPP payments. Regardless, rural electrification

introduces new load to the network. Therefore, financial analysis should be based on NEAs marginal generationcost. This cost will vary by season, but NEAs average IPP cost seems a reasonable proxy for the average marginalfinancial cost.


21/40

12

Electrification, Distribution and Transmission Project (REDT). NEA has consistentlyunderperformed when compared with ADBs performance expectations. 2000 was the only year inwhich NEA made a profit. This is primarily because NEA has not increased its tariff to the levelsenvisaged. NEAs average revenue for FY2003 was NRs7.02/kWh compared with the RRPforecast of NRs10.09/kWh. A further problem is NEAs failure to reduce system losses to theforecast levels. Actual losses for FY2003 were 24.5% vs the forecast 18.7%. Another issue

affecting NEAs financial position is the large amount of electricity arrears owned by theGovernment and municipalities.29

38. NEA currently faces a difficult operating environment. The Maoist insurgency in Nepalcontinues unabated and NEA facilities are an ongoing target. Insurgency damage in FY2003 wasNRs72.4 million. More important than the monetary cost of physical damage is the impact of theinsurgency on NEA operations. This impact arises from both the quantifiable cost of protecting itsassets and the less-quantifiable cost from having to take practical and political considerations intoaccount when making management decisions. An example of this difficulty is the currentrequirement for customers to travel to the nearest branch offices to pay their monthly electricitybills.30 This was made necessary by the removal of bank branches from rural villages because ofthe insurgency. NEA cannot send its staff to collect accounts directly from customers because the

staff would obviously be carrying money and thus, be high-risk insurgency targets.

39. ETFC sets NEAs tariffs, which were last revised in September 2001. NEA filed for a tariffreview in mid-2003, requesting a seasonal tariff that would reduce electricity costs in the wetseason when river flows are higher. This request was declined; ETFC rejected the concept ofseasonal tariffs. NEA is preparing a new tariff filing for likely submission in April 2004. TheGovernment assurances for REDT include semiannual tariff reviews by ETFC, but such reviewsare not occurring, at least not in a meaningful way. The perception that tariff rates are highappears widespread in Nepal. The high cost of electricity was a recurring theme in the focusgroups discussions that OEM organized. NEA staff and other government officials complained toOEM that electricity tariffs in Nepal were already high compared with those of other countries inthe region. This political climate and the extent of government representation on ETFC make

significant tariff increases unlikely in the short term. The ETFC chairperson told OEM that NEAshould improve its financial performance through efficiency gains rather than relying on tariffincreases. But while there is undoubtedly scope for NEA to improve its operating performance,the potential impact of such improvements on NEAs financial performance is limited by the factthat most NEA costs are outside of its managements direct control. A high interest rate is onesuch cost. NEA is lobbying aggressively to reduce the interest rate on its subsidiary governmentloans.31

40. NEA is making continuing efforts to improve efficiency of its operations. It hasdisaggregated internally into four business groups (generation, transmission and systemoperation, distribution and customer services, and engineering services). Each group nowoperates as a profit center. Managers of the generation, transmission, and distribution groups

29As a result of policy dialogue with ADB, the Government paid part of the arrears in December 2003. The Governmentalso made a provision in the finance ordinance in January 2004 to directly deduct one sixth of the grants given tomunicipalities for payment of electricity arrears. NEA is also maintaining dialogue with municipalities to agree on theamounts of arrears. The OEM was informed that, as of March 2004, 25 of 58 municipalities had agreed on theamount of arrears to be paid to NEA.

30OEM was told that travel to and from the nearest branch office to pay the monthly electricity account takes somecustomers a day and costs more than NRs50 for public transportwhen in many cases the bill itself may be onlyNRs80.

31 The Government is not free to adjust these interest rates, which are specified in NEAs loan covenants with ADB andother lenders (see footnote 7).


22/40

13

must sign performance contracts that will reward them according to the extent to which theyachieve key performance objectives. Within the distribution and customer services group, 19branch offices have been designated as distribution centers and operate as profit centersoutside the regional office structure. Distribution center managers are given incentives to achieveresults, measured by key performance indicators. According to NEA, a recent audit of distributioncenter performance was encouraging.

41. High losses on the NEA system are an ongoing problem. NEAs losses, includingelectricity consumed for its own use, totaled 24.5% in 2003, down from 27% at appraisal but stillsignificantly higher than the covenanted target of 20% from FY1994 onward. It is widely acceptedthat NEAs nontechnical losses are excessive, and that any initiative to reduce losses must targetthose losses as first priority. Current initiatives include the inclusion of targets for loss reduction inmanagement contracts for the distribution centers; and a scheme, supported by the DanishGovernment, to sell electricity at discounted wholesale rates to consumer cooperatives, which willassume ownership of the local distribution networks and responsibility for their operation.32 TheGovernment also recently introduced the Electricity Theft Control Act 2058 and the ElectricityTheft Control Regulations 2059, which treat electricity theft as a criminal offence and give NEAnew powers to deal with the problem. Such initiatives are to be applauded, but OEM found that

illegal connections were easy to locate in many rural areas, and that NEA staff accompanying theOEM simply accepted such connections as a fact of life.

42. NEA will no longer contribute to the capital cost of rural electrification programs becausethey are seen as uneconomic. But rural electrification is an important priority of the Government,which has a rural electrification budget of NRs590 million for FY2004. From now on, theGovernment will contribute 80% of the capital cost of rural electrification programs to NEA asequity, with beneficiaries funding the remaining 20%.33 In addition to the Danish Governmentscheme, ADBs current REDT project includes $18.4 million for rural electrification. But accessiblerural areas are now fully electrified, and rural electrification schemes in less-accessible areas arefailing to recover the cost of generation. The operation and maintenance of these programs willput increasing pressure on NEAs financial resources, and will force reliance on subsidy support

from NEAs urban domestic consumers.34

The Government is aware of the problem and is lookingto the ongoing TA35 for recommendations on how to manage it over time.

C. Financial and Economic Reevaluation

43. The financial and economic reevaluation of the Project is discussed in detail in Appendix3. Reevaluations were carried out separately for the urban rehabilitation and rural electrificationcomponents, using the same approach used at appraisal and in the PCR. The period of analysiswas FY1992FY2021. Given the recent power demand and forecast growth, the facilitiesrehabilitated or installed under the Projects urban and rural components are expected to beloaded to full capacities by 2006. After that, further reinforcements will be required to meet theload growth.

32By March 2003, 170 applications have been received for the program, and 18 contracts signed. It is hoped that thescheme will reduce theft, because consumers will be less likely to steal from their neighbors. But success of thisscheme will depend on the ability of the small local cooperatives to operate and maintain the network sustainably.

33The beneficiary contribution does not have to be in cash; it can be in labor, supply of wooden poles, and other in-kind contributions.

34NEAs current tariff is designed to avoid cross subsidy among different customer groups, but the tariff structure doesnot prevent cross subsidies within customer groups. In FY2003, domestic customers accounted for 36% of sales and37% of revenue.

35TA 3552-NEP: Management Reforms and Efficiency Improvements for NEA, for $0.8 million, approved 27 November2000.


23/40

14

44. Incremental electricity sales in each urban rehabilitation scheme from 1998 to 2003,estimated in the Feasibility Study Report of 1992, were checked against NEAs overall loadincreases during the same period and found reasonably realistic. Thus, those data were used inthe reevaluation. NEA has estimated that about 200,000 new connections in rural areas weremade under the Project from 1998 to 2003. This is about 2.4 times the target number of new

connections estimated at appraisal.

45. The financial cost of power for the urban component is assumed to be $0.055/kWh. Thatis in line with NEAs supply at 33-kV distribution levels from its own generation plants. Ruralelectrification introduces new load to the network, so the financial cost of electricity for ruralsupply is derived from NEAs marginal generation cost. A reasonable proxy for this would beNEAs average purchase cost of NRs6.1/kWh from IPPs. The economic cost of power supply isgiven by the opportunity cost of exporting power to India,36 which at the time of OEM was$0.066/kWh. The system loss, excluding transmission at the distribution level, is assumed to be15%.

46. NEAs current average tariff of NRs7/kWh is used as the financial benefit indicator for the

urban component. But a rural electrification scheme would not produce this yield because of thehigh proportion of low-use and lifeline domestic customers. Economic benefits are assumed to bederived mainly from (i) displaced use of kerosene for lighting and of diesel generators for powerand (ii) induced electricity consumption. For displaced consumption, the average economic valueof electricity consumption to displace kerosene lighting with wick lamps is estimated as $0.2/kWh,and displacement of power generated by diesel generators, as $0.15/kWh. For inducedconsumption, the willingness to pay for both residential and industrial consumers is assumed tobe the weighted average of the unit cost saving and the existing tariff.

47. Table 3 summarises results of financial and economic analyses at project appraisal,completion, and post-evaluation.

Table 3: Summary of Financial and Economic Revaluations

Item FIRR(%)

EIRR(%)

Appraisal PCR PPAR Appraisal PCR PPARUrban Rehabilitation 18.8 13.4 11.8 22.2 20.0 24.6Rural Electrification 4.5 0.8 13.5 11.1 19.2

EIRR = economic internal rate of return, FIRR = financial internal rate of return, PCR = project completion report,PPAR = project performance audit report.Source: Appendix 3.

48. The reevaluated economic internal rate of return (EIRR) for the urban component iscomparable to the appraisal estimate. The EIRR estimate for the rural component was more

favorable, mainly because of the large increase in the number of new connections compared withestimates at appraisal and PCR. The lower recalculated financial internal rate of return (FIRR) forthe urban component resulted from inadequate adjustment of tariffs to more sustainable levels.Depreciation of the Nepalese rupee against the dollar also affects the FIRR.37 The FIRR for the

36Under the Power Exchange Agreement with India, the annual power exchange between Nepal and India is nowabout 50 MW. Both countries recently agreed to gradually increase the exchange level to 150 MW.

37 The Nepalese rupee has depreciated 159% since appraisal in1988.


24/40

15

rural components is negative since the average revenue of NRs5.6/kWh from rural consumers isinsufficient to even cover the marginal generation cost of NRs6.1/kWh.

D. Sustainability

49. Electricity demand in many project areas, including remote areas reticulated for the first

time, is high compared with the forecast at appraisal. Table 4 compares the current peak load atselected area substations constructed under the Project with the available forced cooledtransformer capacity. The loads shown include distribution network losses and, in many cases,are undoubtedly elevated due to the level of electricity theft. Nevertheless, none of thesesubstations have been in service for more than 5 years.

Table 4: Substation Loads

Substation TransformerCapacity

(MVA)

CommissioningYear

Energy SuppliedFY2003(GWh)

Peak LoadFY2003(MVA)

Peak Load(% of capacity)

Itahari 8 2000 42.0 7.5 94%Inaraua 8 1999 16.5 5.7 71%

Bisnupur 8 1999 7.9 3.2 40%Parsa 8 2000 23.1 6.8 85%Ilam 3 1999 7.0 1.3 43%Chanauli 8 2000 16.6 5.6 70%Kawasoti 8 2000 14.6 4.5 56%Parasi 8 2000 13.0 2.7 34%Palpa 8 1999 9.0 2.9 36%Gajuri 3 2001 3.8 1.6 53%

Gwh = gigawatt hours, MVA = megavolt amps.Source: Nepal Electricity Authority.

50. Ample evidence shows that a reliable and good-quality supply of electricity stimulatesindustrial and economic development in areas with good road access. For example, NEA suppliesmore than 20 industries at 11 kV in the region served by the Kawasoti area substation.38 The

main sustainability problem in these areas is the availability of capital to increase capacity as theexisting distribution system becomes overloaded. Indeed, much of the electrification installedduring the Sixth Power Project has already been reinforced. For accessible areas with potentialfor commercial and industrial development, NEA should take a longer-term view of future loadrequirements in planning future additions of capacity. The situation is different in areas withoutgood road access, because significant industrial development is unlikely.

51. The rural electrification component of the Project includes some previously unelectrifiedtowns along the east-west highway that will probably benefit from ongoing industrial andcommercial development. But all of Nepals more accessible areas have now been electrified, andthe financial sustainability of project components that serve less-accessible areas is unlikely.Nevertheless, NEA will probably keep all parts of the Project operational because of the

Governments commitment to rural electrification, and the benefits of electrification to individualconsumers.

38The Kawasoti substation is in western Nepal, adjacent to the main east-west highway. The substation was builtthrough the rural electrification component, not only to supply new load but also to support the overloaded 11-kVdistribution network built under the Sixth Power Project. The transformer capacity installed under the Project wasonly 3 MVA, but a larger transformer was installed within 1 year of commissioning the substation.


25/40

16

E. Technical Assistance

52. The attached TA was implemented from July 1990 to July 1991, when a distribution and acommercial consultant spent a year in Kathmandu working full time with NEA. The distributionconsultant (i) proposed a structure for distribution planning and developed a schedule ofdistribution requirements, (ii) established a drawing office and started the development of system

diagrams, (iii) developed procedures to establish and operate a management information systemand a 5-year rolling plan, (iv) developed the structure for a technical standards committee, and (v)ran a training course on distribution planning. The commercial consultant (i) undertook adiagnostic review of NEAs consumer accounting and billing processes, and (ii) developed a planto introduce computerized billing, using commercial billing software. The consultants fulfilled theirTOR requirements, but TA achievements fell short of expectation because of the lack of NEAcommitment to immediate implementation of TA recommendations, and insufficient counterpartfunds to make the TA more effective.39 Overall, the TA is assessed as partly successful.

IV. ACHIEVEMENT OF OTHER PROJECT IMPACTS

A. Socioeconomic Impact

53. The OEM conducted a socioeconomic survey to assess the impact of electrification onindividual domestic consumers, particularly consumers that had no electricity before the Project(Appendix 4). In summary, electrification in urban areas and areas near sealed roads has positiveimpacts on the local economy because of benefits to the lives of individual consumers, and stimulifor industrial and commercial development. This economic development improves the lives ofdomestic consumers so much that many consumers in such areas pay more than the minimumlifeline electricity cost. In remote areas, consumers benefits from improved lighting and use ofradio, television, and electrical fans would be impossible without access to electricity. Womenhave been the greatest beneficiaries of electricity. But many poor households in remote areashave not connected to electrical power because the initial connection cost is so high.Furthermore, NEA will not supply electricity to houses with thatch roofs because of high fire risks

(although some such households illegally connect to neighbors electricity). The industrial use ofelectricity in remote areas is limited largely to rice and oil mills that were diesel-powered beforeelectrification. A key project objective at appraisal was to use electricity as an energy source forirrigation, particularly for shallow well pumps. This did not materialize, mainly because of the costof providing a mains connection to each pump site. With diesel engines, there is no connectioncost and one engine can service a number of pumps because it can be moved from site to site.

B. Environmental Impact

54. The Project was appraised in 1989 and had no specific environmental objectives. Theappraisal report noted, however, that the design would take international environmental andsafety considerations into account relevant to the location of transmission and distribution lines

and of substations. Construction of the Project has been environmentally benign, with mostoverhead lines located on road reserves. As noted in Appendix 4, there have been fewbeneficiary complaints about environmental issues. But there were some concerns about thesafety of 11 kV lines constructed in urban areas. In such cases, NEA took appropriate steps to

39An example was cancelation of a planned seminar on distribution planning for regional planning staff because NEA,,citing budgetary constraints, would not allow the staff to travel to Khatmandu. In order to keep up the momentum,ADB wanted to continue both components of the consultancy for 12 more months under a new TA, if NEA would usesavings from the Fifth Power Project to purchase and implement a pilot billing system. But NEA would not agree.About 4 years later, savings were used for a pilot billing system.


26/40

17

alleviate the concerns and allow the Project to proceed. The Project was also expected to havean indirect environmental impact by utilizing Nepals indigenous hydropower facilities to replacefuelwood and imported fossil fuels. The Project seems to have done little to reduce fuelwood use,although it has substituted somewhat for imported fossil fuels through reduced use of keroseneand small diesel motors. But NEA operates high-cost diesel generation plants at times of peakloads, and the additional electricity demand caused by the Project requires increased generation

from these sources. The engineering report in Appendix 1 also discusses possible designimprovements that might further reduce the environmental and safety consequences of futureprojects.40

C. Impact on Institution and Policy

55. Objectives of the institutional development TA have, by now, mostly been achieved,although more slowly than envisaged at appraisal, and with additional help from ADB and otherdonors. First, NEA now has a computerized billing system in the larger branches that generate70% of its revenue. With help of local consultants, NEA has enhanced the original billing softwareand is now rolling out the enhancements. Under the REDT project, NEA will purchase upgradedbilling software with state-of-art technology, which it plans to install in all but its smallest branches.

Second, distribution planning is still centralized in Kathmandu, even though detailed design of 11-kV distribution network extensions is undertaken at the branch level. NEA has purchased acomputerized geographic information system, and has incorporated records of its network assetsdown to the 11-kV level. NEA has also purchased, with ADB funding, software for analysis ofdistribution systems, and is interfacing the two packages. Both systems are available only inKathmandu, which may limit benefits from the geographic information system.

V. OVERALL ASSESSMENT

A. Relevance

56. The Project conformed to Nepals needs for power sector development and ADBs

strategy for assistance at appraisal to support the development of integrated power supply inurban and rural areas. The Project has remained relevant to the achievement of sectordevelopment objectives for both the Government and ADB. However, there was a designweakness related to the introduction of SWER. Overall, the Project is assessed as relevant.

B. Efficacy

57. The Project achieved its primary objectives of improving quality and reliability of powersupply in areas with rapid growth demand, and of extending the availability of a reticulateddistribution system. But project goals in terms of NEAs system loss reduction and financialsustainability, envisaged at appraisal, have not been fully achieved. Overall, the Project is ratedefficacious.

C. Efficiency

58. Full completion of the Project took more than 10 years. During the early stages, ADBconsidered it an at-risk project. While credit should be given for the fact that significantly morerural electrification was achieved than envisaged at appraisal, this was made possible by factors

40The need to design the Project to minimize environmental impact when things go wrong was not specifically noted in

the appraisal report.


27/40

18

largely outside the control of project participants. Although the rural component is not viablefinancially, the Project has achieved more favorable EIRRs than estimated at appraisal. Overall,the Project is rated as efficient.

D. Sustainability

59. The rehabilitation component and the pole factory are considered highly sustainable. Therural electrification component is sustainable only with a significant and continuing subsidy, whichwas not envisaged at appraisal. Overall project sustainability is considered less likely on the basisof NEAs weak financial position and continuous resource injections required to ensure continuedoperation of project facilities.

E. Institutional Development and Other Impacts

60. Through the Project, NEA gained enough experience in distribution planning and designso that external engineering and project implementation consultants were no longer requiredduring later project implementation and for future ADBfunded distribution projects. The Projectalso facilitated economic growth in project areas by providing reliable power. The institutional

development and other impacts of the Project are rated significant.

F. Overall Project Rating

61. Overall, the Project is rated successful. This assessment recognizes the Projects successin terms of physical achievements and socioeconomic impact at project levels, but also takes intoaccount the requirement for ongoing subsidies to sustain the rural electrification component.

G. Assessment of ADB and Borrower Performance

62. Despite the much-improved implementation performance by NEA and ADB after 1995, theperformance of both ADB and the EA is assessed as partly satisfactory. For ADB, this

assessment reflects failure to take proactive and timely action to respond to implementationproblems early in the Project, as well as late decisions regarding the first loan extension, and theaborted decision to transfer project administration to NRM. ADBs partly satisfactory performancereflects the inadequate attention that the project department gave to project administration.Forthe EA, this assessment is based on the unsatisfactory project progress over the first 5 yearsthrough the end of 1995, its failure to reduce losses to covenanted levels, and the poor financialperformance that continues to threaten the Projects sustainability.

VI. ISSUES, LESSONS, AND FOLLOW-UP ACTIONS

A. Key Issues for the Future

63. NEAs most pressing problem is its deteriorating financial performance. NEAs cashoperating surplus for FY2003 was only 42% of what ADB forecast in 1999 (Appendix 2). Morethan 77% of this surplus was required to pay its interest liability. NEA is well aware of thisproblem, and argues strongly for relief from these interest charges. But the more fundamentalcause of NEAs problems is failure to raise electricity tariffs to more sustainable levels. Theaverage internal electricity tariff in FY2003 was only 70% of the level assumed in ADBs 1999forecast, and will not change significantly in FY2004, since no application for a tariff increase has


28/40

19

been submitted to the ETFC.41 Tariff setting in Nepal remains a political issue, because all ETFCmembers are government appointees. The prevailing view, both within the Government and NEA,is that the current average tariff level of NRs7.02 (about $0.10) is high, compared with that inother developing countries in the region. In OEMs view, significant tariff increases are highlyunlikely in the near future. Insurgency problems, which seemed to escalate during the OEM,make it even more difficult for the Government to make meaningful tariff adjustments.

64. There is little doubt that system losses are excessive, and contribute significantly to thecurrent level of electricity tariffs. Total losses in FY2003 were 553 GWhalmost 25% of NEAselectricity generation. The cost of these losses is estimated at NRs2,500 million, assuming anaverage generation cost of NRs4.50/kWh, or NRs3,300 million using a marginal generation costof NRs6.0 per kWh. Using the marginal generation cost, each 5% reduction in total losses wouldsave NEA about NRs660 million in generation costs. A reduction in losses from 25% to 15%would mean an average tariff reduction of NRs0.87/kWh, or 12% if applied to all domestic tariffreductions. Thus, NEA must continue to vigorously address electricity losses, if it seriously wantsto reduce tariffs.42

65. Assuming no change in NEAs operating position, the rural electrification component of

ADBs REDT project will strain NEAs financial position and further reduce its ability to meet itsloan covenants. REDT will fund the installation of rural electrification costing $24.6 million orNRs1,720 million. Incremental electricity sales resulting from REDT are estimated to eventuallyreach 90 GWh. Based on NEAs current operating situation, REDT will, at best, generate enoughrevenue to cover the cost of the incremental generation.43 But, consistent with ADBs financialcovenants, the rural electrification component will create an incremental annual revenuerequirement of about NRs420 million.44 Considering the relatively high initial connection chargesand the transaction costs of making monthly journey to pay electricity bills (Appendix 4),possibilities of subsidizing the initial connection charges and reducing costs of payment for ruralcustomers should be explored so the tariff might be raised at least to a break-even level for ruralsupply.

B. Lessons Identified

66. The decision not to proceed with the use of SWER technology, while justified on technicalgrounds, caused delays early in the Project because it necessitated a redesign of the ruralelectrification component. These delays may have been avoided if a more substantive PPTAconsultancy had been undertaken, and if the technical feasibility of using SWER had been morethoroughly researched during appraisal. In this case, the staff consultant who did the projectpreparation suggested the use of SWER, but was not required to carry out technical investigationof proposed alternative technologies in the TORs. On one hand, innovation by consultants shouldbe encouraged. But on the other hand, for a project of this size and complexity, a full scale PPTAshould be mounted and the TORs should include investigation into the use of new and alternativetechnologies.

41The most recent tariff increase became effective on 17 September 2001.

42NEA noted that the insurgency problem makes loss reduction initiatives more difficult to implement.

43See paragraph 35. Furthermore, alternative approaches give similar results. NEAs current average tariff is NRs7.02 /kWh, but a rural electrification scheme would not yield this because of the high proportion of low-use and lifelinedomestic customers. Thus an average rural electrification tariff of NRs6/kWh, equal to the assumed marginalgeneration cost, is a reasonable assumption.

44The analysis assumes that revenue from the rural component of REDT will cover only the incremental cost of energy,and no significant reduction of electricity losses below current levels.


29/40

20

67. The main implementation problems arose during the first 5 years of the Project. TheProject was administered from ADB headquarters in Manila, and the project files show littleevidence of any useful input from NRM, even though the Project was considered problematic.Under those circumstances, annual review missions were insufficient and the me