Djibouti - Geothermal Development Project

Doment of The World Bank F0R OITICUAL USE ONLY iqu,t No 12557 PROJECT COMPLETION REPORT DIBOUTI GEOTHERMAL DEVELOPMENT PROJECT (CREDIT 2055-DJI) NOVEBER 29, 1993 MICROGRAPHICS Report No: 12557 Type: PCR Industry and Energy Operations Division Country Department III Africa Regional Office Thi document has a noriced distriblao and mny be and by nrepients only in the prfomane of tbew officid duties Its contens may not odhewe be dwjclosed d Woldd Dank auhrizato  cumacy Unit Djibouti Franc (DP) US$1.00 176.83 DF bbl barrel 42 US gallons GWh giga hour 1,000,000 kilowat hours kV kilovolt 1,000 volts kW kilowatt 1,000 watts kWh kilowatt hour 1,000 watt hours m meter 3.28 feet MW megwatt 1,000 kllowatts MWh megawatt hour 1,000 kllowatt hours -DRGM French geological and mining agency EdD Electricity company Djibouti ENEL Itlian national power company Govenment Government of Djibouti IDA Inenal Development Association OPEC Or ion of Petroleum Exporting Countries PCR Project Completion Report PPP Poject Preparation Facity UNDP United Nations Development Program USc United States cet EJn AL em January I - D)ecmbet 31  FOR OFFICIAL USE ONLY THE WORLD BANK Washigton, D.C. 20433 U.SA Office of Dizsetor-Gaeral operations Evalution November 29, 1993 MEHORABDUM To THE EXECUTIVE DIRECTORS AND TME PRESIDENT SUBJECT5 Project Completion on Djibouti Geothermal Develooment Pro1ect (Credit 2055-DJI) Attached is the "Project Completion Report on Djibouti - Geothermal Development Project (Credit 2055-DJI)" prepared by the Africa Regional Office, with Part TI provided by the Borrower. The project, which was to be co-financed by the Italian Government, OPEC and UNDP, aimed at helping meet the country's electric power demand through the use of indigenous geothermal energy. However, it was not implemented (except for a US$ 250,000 PPF advance) and most of the loan (962) was canceled, as IDA concluded that the project could not be economically justified in the face of significant delays in project start up. The latter resulted froms (1) a lack of agreement among the parties, from the outset, on the basic technical design of the project; and (ii) difficulties in mobilizing the Italian grant. In retrospect, it is clear that such basic issues should have been resolved before approval of the IDA credit, particularly since similar issues had marred the previous geothermal exploration project (Credit 1488-DJI). In Part II, the Borrower confirms its disagreement with IDA on technical issues and argues that the project should proceed, although in a more phased manner. Since the project was not implemented as it was no longer economically justified, it is not rated. The PCR is reasonably informative but no coments wvere requested from co-financiers. An audit of Djibouti's geothermal program may be carried out in the future. Attachment This docamnt has a restricted distribution snd my be used by recipients onty ft. the pirfo of their offtel at duties. Its contents my not otherwls be disclosed without WorLd tek2* 3thorut1ion.  FOR OMCIA USE ONLY PROMCr OMLEnION REPORT GEO7THERMAUL DEVELODPMENT PROJE!M:t (C,REDIT 2055-fX) TABLE OF PREFACE ......... i EVALUAION SUMMARY ................... , ., ii PART I: PROJECr REVIEW FROM BANK'S PERSPECTIvE .. 1 Project Identity .1 Background . ............................................ 1 Project Objectives and Description. 2 Project Design and Organization. 3 Project Implementation ....................... 4 Bank Perormance .5 Bornower Performance .5 Project Relationship. 6 Consulting Services. 6 Project Documentation and Data. 6 Lessons Learned .......................... 6 PART I: PROJECr REVIEW FROM THE BORROWER'S PERSPECTVE. 7 Introduction ......................................, 7 Principal Issues. 7 Different Approaches to Project Design. 7 Cause of Delay. 9 Project Implementation. 9 Conclusions .10 PART m: STATISTICAL INFORMATION .11 IDA Related Credits .11 Project Timetable .12 Financing Plan at Apprais (US$ Miflion) .13 Use of IDA Resources .14 This document has a rstricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  JEcr CMLEND PREOErT lTis is the Project Completion Report (PCR) for the Geothermal Development Project for which Credt 2055-DJI of SDR 7.1 million was approved in 1989. The Credit was closed on December 31, 1992 and the undisbursed balance of SDR 6.8 million was cancelled. The PCR covrs the period 1989-1992. The PCR was jointly prepared by the Industry and Energy Division of the South-Central and Indian Ocean Department, Africa Region, and EdD, the national power company of Djiboutl and the implementing agency of the Borrower, which contribbted Part II of this PCR. The PCR is based, intrfalia, on the Staff Appraisal Report, the Credit and Project Agreements, coepondence between IDA, the Borrower, other donors and consultants, supervision reports and interal memoranda.  Q1ZCOMPMMON RPR - IIBO-x EVALUATION SUMMARY 1. Objeie. The project's main objective was to assist Djibouti in the replacement and expansion of its electric generat capacity at least cost hough the utilization of 'Migeneous geotherma energy resources. 2. == . Overall IDA's experience with this credit was disappointing. In retrospect, the risk of failure was higb. Building upon experience gained under an earlier exploration project, the development project was to confirm the existence of the resource through the drflling of addiional wells and provide for the installation of a power plant and related eqipment. T rihelling of new weUs was to have begun in early 1990, In order for a geothermal power plant to begin operation in 1993, in time to meet the growth of demand. Otherwise an investment in additional fuel-fired capacity would have to be underta*en. Timing was therefore a key element in the justification of the project because of a window of opportmuty' created by the retirement of old diesel plants and demand growth. IDA considered the results of the exloration carried out under the Geothermal Exploration Project (Cr. 1488-DJI) sufficieady positive to proceed with a development project. The Consultants which the Italian Govrment selected t help with the development project, were however of the opinion that further exploration was needed before launching the development phase. ITis difference of opinion over the design of the project surfaced after the project was approved by the Board in 1989. After protracted discussions, no satisfy solution to this problem could be found. 3. Delays in project implementation raised doubts about the capacity of geothermal energy to satisfy the demand for electricity in a timely maner. In late 1991, IDA's concern led to an indeendent revev of the economic justification of the project. This study, which was completed In Juy 1992, concluded that the project was no longer the least cost means to satisfy the demand for electricity in Djibouti. In any event, in ealy 1992, civil unrest in the northem part of Djibout made the project area inaccessible. 4. On November 19, 1992, IDA informed the Govemment of Djibouti and the cofinancier of Its intention not to extend the closing date of the credit of December 31, 1992. The credit was effectively closed on December 31, 1992 and the undisbursed balance was cancelled. Out of the original amount of SDR 7.1 million that was approved, SDR 6.8 million were cancelled. lhe amount disbursed under the PPF advance was SDR 0.25 million. S. s Leaned. The project was risky and it failed because delays in reaching agreement among cofinanciers on a course of action for implementation undermined its justification. The lessons learned are as follows:  - liii- (a) The risks should be clearly identified both with respect to the physical availability and quality of geothrma resources and the possibility of failure of Implementadon arrangements; (b) A risky project could be justified but only if all pates concerned In its implementation have a consensus on how to make it work; (c) IDA should provide funding for consultancy services in critical areas of the pmject it finances; (c) With the Introduction of new technologies, up-front strengthening of local staff or provision of an 'in-houseT expertise is essential if the country andor the executing agency Is to have a say in project design and implementation; and (d) When time is of essence for the success of the project, conditions of effectiveness involving third parties are not appropriate and key actions should be taken before Board presentation.  CREWI 20"5mmW PART I RECT REVlhW FR AM B' PERSPECTI 1. Prje ldentit Name : Geothermal Development Project Credit Number : Credit 2055-DJI RVP Unit : Africa Region, Country Department m County : Republic of Djibouti Sector : Energy Subsector : Power 2. sgia 2.01 The presence of a geothermal resource in Djibouti has been known for a long time and various effrts financed by UNDP and other sources have been made to develop it, in the past. Becase of the high cost of imported fuel, on which Djibouti was enredy dependent, the development of the geothemal resource looked like an interestng alternative for electricity generation. In 1982, the Goverment requested IDA to help finance a new round of exploration and in November 1984, a credit agreement (Cr. 1488-DI)7 was signed for that purpose. Additional funding was provided by the Govemment of Italy, UNDP, the OPEC Fund, the African Development Bank and the Government of Djibouti. The French Government made available to the project, the services of BRGM, its geological and mining research agency. 2.02 A major delay in the startp of the project was experienced due to late signature of an agreement between the Italian Ministry of Foreign Affairs and Aquater, the firm responsible for project supervision. Because of this delay, the implementation of the project was about eighteen months behind schedule, necessitating two extensions of the closing date. 2.03 The overall results of the exploration effort were disappointing. Two deep wells, Hanle 1 and 2, were drilled in the Hanle/Gaggade area, in the western part of the country. The temperatures were found to be too low to be of interest for power development and the exploration effort in the Hande area was abandoned. Exploration moved to the Assal area where, in the 1970s, BRGM had encountered a geothermal resource in a well designated as Assal 1. This well produced a brine which was supersaturaed with dissolved solids such as sulfides, silica and carbonates, known to cause scaling problems both within wells and on steam processing equipment. Assal 3, 4, 5 and 6 were drilled, sometimes with haste. Assal 3 and 6 were found  .2- to be promisag, lthough heavily uraed with dissolved solids. With hindsight, this exploaion program lacked a clear sene of direction and was haphazad. 2.04 Ihe Project Compledon Report of the Geothermal Exploraton Project was completed in April 1991, after the ap of the Geothemal Development Projet (Credit 2055-DJI) in Ju/July 1988. The letions leaned were: (i) when a project groups many participants, the projt erntty must tak the eal role for coordination and communication among the various particpts and must be clearly accountable for decisions taken; (ii) the responsibilitiesof agents of the project execution entity must be defined and officially agreed upon very early in the project execution stage to avoid overlaps of authority and/or gaps in responsibility; (iii} consultation proces should be clearly defined and fully utilized to avoid subsequent disputes over decisions taken; and (iv) excessive delays In the unblocking of cofinanciers' funs should be cause to re- examine the project financing plan. 3. PromLee Qbcti!xj and D.caiDno 3.01 .rojT Qb3Sves lhe projects main objective was to assist Djibouti in the replacement and expansion of its electric geatng capacity at least cost through the utilization of indigenous geothermal energy res . 3.02 Pwect Do n: The Geothermal Development Project (Credit 2055-DJI) provided for. (a) Scaing Iulin to insure that generaing equipment, wel tubes and steam ahering sstem are designed in a way which takes full account of the specific naue of the corosion effects linked to the salinity of the brine; (b) Nomad Settlement Impact Study to spell out issues which may arise from a poteni settlement of nomads in the area of the drilling activities; (c) DMiIIinl in the Lake Assal area of four deep production and replacement wells (about 2000 m deep), including necesay consunables; (d) Co ena surface exploration woksa in the Lake Assal area, including grvity and electrical resistivity surveys and related studies; (e) O Installation gf 2x5 MW power plant (Imcluding wellfield equipment) and Qi) a 1MW well head unit and desalination pant () Installation-of In eJI a miss ion line fom the LakeAssal Area to Djibouti city; and (g) iehia Assistance to Eleuicite de Djibouti M. Tfe poe utility including: (i consultg services for fur&ter project preparation (PPF financed); (iH) services of project management and supervision for three years; (iii) engineering services for de supervision of the deep drilling operations and the carrying out of related speciized services (oggig, testing, etc.); (iv) training; and (v) equipment and vehicles.  -3- 4. e tmiI gmnd OQnanization 4.01 The project was appraised in June/luly 1988. It was co-financed by the Government of Italy, the OPEC Fund, TJNDP, the Government of Djiboud and EdD. 4.02 However, there was no consensus amwng cofinaciers on the project concept. The major issue was whether geothennal energy had reached the development stage in Djibouti. Of the four wells drilled under the exploration project in the Assal Area, two were satisfactory steam producing wells, which gave results indicative of a reservoir". IDA considered the results of the exploration sufficiently positive to proceed with a development project. Under this project, four ew wells were to be drilled, starting in early 1990, in order for a geothermal power plant to begin operaton in 1993. Timing was considered a key element in the justification of the project because of a "window of opportunity created by the retirement of old diesel plants and demand growth: should a geothennal plant not be ready for installation to meet demand, investment in additional fuel-fired capacity would have to be undertaken. The Government of Italy, through its consultants, on the other hand was of the opinion that the feasibility of a geothermal development in the Assal Area had not been established and the project was not ready to enter the development phase untl further exploration wells proved the existence of a reservoir, a preequisite for developing a resource of this type. 4.03 A program was agreed In October 1990 among all parties (the Italian Miistry of Foreign Affairs, the consultant, Djibouti and IDA) to proceed with the development phase if the results from drilling the two new deep wells were positive. In spite of tis, the Cooperation Agreement signed in November 1990 between the consultants and EdD made a provision for a firm decision on the development phase only after completion of four new deep wells to be fianced by IMA as well as the clean-out of two existing wells. That meant that electricity of geothermal origin would not have been available before 1997 at the earliest, nearly four years after the time set for completion of the project. On this basis, entry-on-stream of the geothermal power plant would have fallen outside the *window of opportunity" which justified IDA's involvement in the project. 4.04 The arrangements under which services were to be provided under the Italian grant for supervising the implementation of IDA's financed project were inappropriate. The Cooperation Agreeme was not so much a mutual agreement between the consultants and EdD as an enumeration of the services the consutants would provide. The consultn had the exclusive right of coordination, general supervision and technical assistance for any activity under the project. The only control EdD could have exercised over the services it would receive was tbrough its approval of progress reports and official declarations of services rendered before payment was released by the Italian Govenment under its contract with the consultants. However, given the general terms in which the consultant's responsibilities and the activities it were to carry out were described, it was nearly impossible for EdD to challenge either the quality or the appropriat of the consultant's services. The Djiboutians had little cotrol over the grant-funded technical assistance and were not in a position to influence the consutant 4.05 IDA and the consultants also differed in their approach to solving the main technical problem of developing Djibouti's geothermal resource: scaling. The Assal resource consists of a supersa brine with dissolved solids which can cause incrustation and corrosion in the wells and on steam separation and power generating equipment Various technical solutions are employed in geothermal fields worldwide for controlling these problems. To determine the most appropriate and least-cost solution for treatng the Djibutia. brine, a broad range of solutions  -4- was to have been studied, before Credit effecttveness, financed by a PPF advance. However, as procurement directives were not followed, IDA refused to finance the study. Once lDA withdrew, the study lost direction and was reoriented by the consultant toward its preferred solution: the use of chemical inhibitors. It should be noted that the Assal deep aquifer geothermal resource conts metal sulfides. These impurities have proven to be particularly difficult and epensive to remove at geothermal fields and installations elsewhere in the world. M)A considered the chemical inhibitor solution to be high cost and technically questionable. S. 5.01 The Geothermal Development Credit (Cr. 2055-DJI) was approved by the Board on June 9, 1989 and signed in October 16, 1989. Effectiveness was delayed by more than a year. The conditions of effectiveness were: (i) effectiveness of the Italian grt financing; (li) signatr of the agreeme (in form of a contract) between EdD and ENEL, satisfactory to IDA, stipulating the details of their assistance and supervision program throughout the project; (ii) availability of final results of the scaling study; and (iv) signature of the Subsidiary Loan Agreement between Government and EdD in a form acceable to IlDA. The credit became effective in April 1991. 5.02 The Project never took off the grund. The fundamental differences between the donors on project design constituted the weak point of the project. Problems in unblocking the Italian funding and the lack of clarity In the consultants' mandate added one year and a half delay to the implementation schedule foreseen at appraisal. IDA repeatedly warned that continued delays In the implementation of the project would endanger its justificaion, but to no avail. In late 1991, IDA decided to recruit an independent consultant to review the demand projections and the economic justification of the project. 5.03 The study was undertaken in April-May 1992 and the report, which IDA endorsed, was isued In July of the same year and sent to Djibouti. After reviewing the demand projections and the comparison between geothermal and diesel based power development, the study concluded at the project was no longer the least-cost means of meeftig the demand for electricity in Djibouti. The major underlying factors were: () a substantial increase in project costs since its approval by the Board in 1989, from US$38.0 million to about US$60.0 million; (ii) lower demand growth than expected. This demand could be statisfied by the existing capacity and a new capacity addition would not be necessary until 1998; and (Mii) lower fuel costs. The economic comparison of the geothermal and diesel alternatives gave the following results: - Gtbhermal: 11.38 US cents/kWh - Desel: at 15 US$/barrel 4.24 US cents/kWh at 20 USStbarrel 4.67 US centslkWh at 30 US$1barrel 5.52 US cents/kWh at 40 US$/barrel 6.37 US cents/kWh. Therefore, under the most pessimistic scenario examined (intenational petroleum price of USS 40 per barrel), the esdmated production cost of a kWh from a diesel based development would be USc 6.37 against USc 11.38 frm geothermal.  -5-Z 5.04 The closing date of December 31, 1991 was extended twice: the first extension to Jane 30, 1992 and the second to December 31, 1992. On November 19, 1992, IDA informed the Government of Djlibouti and the Italian Government of its intention not to extend the closing date of the project when it comes due. On December 28, 1992, IDA confirmed its decision not to etend the closing date and to cancel the undisbursed balance of the Credit. Out of the original amtount of SDit 7.1 mlllion that was approved, SDR 6,846650.56 were cancelled. The amount disbursed under the PPF advance was SDR 253,349.44. 6. Bank -erformane 6.01 The drllling contemplated under the project was of an exploratory nature. Exploration was combined with the development of the resource even though its quantity and quality were not well apprehended. The project was risky and It failed because delays in reaching agreements among cofinnciers on a course of action for implementation undermined its justification. 6.02 Only limited lnforination was gained from the scaling and corrosion study, additional data and analysis were needed before the development of Djibouti's geothermal resource could de iitiated. IDA recognized that an important part of the additional information could only bave been obtained by drilling two additional wells. The consultants later insisted on the drilling of four wells. 6.03 In a complex undertaking like the Geothermal Development Project, the financing of the consultats should have been provided by IDA to ensure a timely and proper implementation of the project. EdD had little Influence on the methods and the timing of activities of the consultants which considered the Italian Government as its client, not Djibouti. Ite timely availability of funds to pay for the consultants' services from the donor's grant and lae contracts signature have also been a problem, the same as those experienced with the financing from the same donor for the earlier project. 7 ,Bg r 7.01 The Djiboutian authorities had little command over the project as major actions were decided by outsiders. The Government of Djibouti was in a difficult position since, for much of the life of the Credit, it did not have any strong "in-house geothermal expertise nor did it have regular access to information on state-of-the art geothermal technology from elsewhere in the world. Tbus, for techical advice, the Djiboutians were almost totally dependent on sporadic, long-distance contact with the consultants and, to a lesser extent IDA. This made them much more subject to the consultants' direction than might have otherwise been the case. On the other hand, they could have insisted on having the consultant accountable to them and not to the donor's Government. 7.02 IDA advice not to purchase a used drilling rig was ignored and a substantial amount of money was needlessly spent by EdD in refurbishing it. The Djiboutians later intended to use for the drilling of the shallow wells, but with the project area closed for security reasons for more than a year, the rig is presumably used to drill for water.  -6- 8. 1 8.01 IDA's reladonship with the Government and the executing agency were consistenly afable. Strnous efforts were made by IDA to reach solutions acceptable to all parties and to bridge out differenes on project design with the consultants, but the results were not encouraging and fimdamental differences between the parties on the development of Djibouti's geothfermal resources mned significant. 9. ru 9.01 lhe consultants were under contract to their Government and not to EdD and therefore felt no obligation towards Djibouti. The Cooperation Agreement signed between the consultants and EdD was reviewed by IDA and found unacceptable, primarily because EddD had no remedies, should ffie consultants' perfommce proved to be unsatisfactory. The consultants persisted In using chemical inhibits and ignored IDA advice to try out other well known processes to prevent scaling. Had a contract been signed directly between the consultants and EdD, the results might have been different. 10. Proet Documntatlon and Data 10.1 Sufficient informaion and data were available in the project's file to prepare this project completion repott. The back to offlce reports of the numerous meetings held between IDA, the Italian Government, the consultants, EdD and the Government of Djibouti were particularly usefl as was the project completion report of the Geothermal Exploration Project (Credit 1488- DID). 11. 11.1. The project was risky and it failed because delays in reaching agreement among cofinanciers on a course of action for implementation undermined Its justification. The lessons learned are as follows: (a) The risks should be clearly identified both with respect to the physical availability and quality of geothermal resources and the possibility of failure of inplementation arrangements; (b) A risky project could be justified but only if all parties concerned in its implementation have a consensus on how to make it work; (c) IDA should provide funding for consultancy services in critical areas of the project it finaces; (c) With the introduction of new technologies, up-front strengthening of local staff or provision of an 'in-house expertise is essential if the country and/or the executing agency is to have a say in project design and implementation; and (d) When time is of esence for the success of the project, conditions of effectiveness involving third parties are not appropriate and key actions should be taken before Board presention.  -7 EART }IrRQIC R aFROM THE BORRO-W-09S PEW=EV 1. 1ntodcdo 1.1 The objective of the Geothermal Development Project in the Republic of Djibouti was to exploit geothermal energy on an economically and financially competitive basis, in order to: (a) reduce the cost of electric power, in particular the proportion of that cost in foign exchange; (1) establish a national resource base for present needs and for industrial development In the coming years; (c) diversify existing energy resources within the Republic of Djibouti. 1.2 It is Important to note that geothennal energy is a national resource that does not follow the flucuato of the interational market. 1.3 In the area of gethermal research and development, a wide variety of disciplines have so far been used, together yielding a wealth of scientific and technical data (surface studies, gradient holes, deep well drilling, studies of selected aspects, etc.). 1.4 However, the results of the exploratry wells and of the scaling study revealed a complex problem caused by: - scaling, resulting from the fact that the brine in this reservoir Is highly saline and contains large amounts of sulphur and silica; - uneven permeability levels in the area. On the other hand, there is a large, highly permeable aquifer at a depth of 400-600 m, with a temperate of around 130