Burundi - Bujumbura Water Supply Project

CONFIDENTIAL qeport "o. 79:) INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT INTERNATIONAL DEVELOPMENT ASSOCIATION PROJECT PERFORMANCE AUDIT REPORT on BURUNDI: BUJUMBURA WATER SUPPLY PROJECT (Cr. 85-BU) June 26, 1975 Operations Evaluation Department  PREFACE Pursuant to the Operations Evaluation Department's responsibility to undertake a performance audit of all projects supported with Bank and IDA assistance a short time after closure of the loans and credits, this report reviews the performance of the project supported by IDA Credit 85-BU of March 1966, closed in March 1973, and of the Bank Group in connection with it. The report is based on study of Bank archives and documents, on a brief visit to Burundi carried out in November 1974 and on comments by the Burundi authorities, the project consultants and Bank staff involved in the project on an earlier draft. Warm appreciation is expressed to REGIDESO, beneficiary of the Credit, for the assistance provided. Currency Equivalents (Burundi Franc) before February 1965 : 50.0 BuF = US$1.00 Feb. 1965 - Feb. 1973 : 87.5 BuF = US$1.00 since February 1973 : 78.0 BuF = US$1.00  PROJECT PERFORMANCE AUDIT REPORT BURUNDI: BUJUMBURA WATER SUPPLY PROJECT (Cr. 85-BU) TABLE OF CONTENTS Page No. Summary i - iv Background 1 Project Implementation 2 Operating Problems 5 Project Benefits 7 Economic Justification 9 Staffing and Training 12 Institutional Development 15 Financial Performance 17 Bank Administration 20 Conclusions 22 Annex Tables 1. Approximate Evolution of Main Elements and Costs of Bujumbura Water Supply Project 2. Bujumbura Population and Water Demand, 1964-84, Forecast arid Actual 3. REGIDESO Bujumbura Operations - Expected and Actual Sources and Applications of Funds 4. REGIDESO - Bujumbura Operations: Water 5. REGIDESO - Bujumbura Operations; Electric Power 6. Balance Sheet - REGIDESO Bujumbura Operations, 1965-73 Maps Map 1 - Map of Africa and Map of Burundi Map 2 - Bujumbura Existing and Proposed Water System s SUMMARY When Burundi, a very poor country of about 3.5 million population in the middle of Africa, first joined the Bank Group in its own right in 1964, it requested assistance in solving an urgent water supply problem in Bujumbura, its capital city, with a population of some 70,000 which has since doubled. In March 1966, IDA signed a Credit (85-BU) of $1.1 million, its first to Burundi and first for a water supply project in Africa. The Credit was closed, almost fully disbursed, in March 1973, having accounted for the major part of IDA dis- bursements to Burundi up to that time and about 7% of public infrastructure investment in the preceding five years, but only 1% or so of total foreign assistance to the country in this period. The Credit was designed to support a $1.7 million project, originally conceived by a WHO engineering adviser to REGIDESO, the national water and power authority with operations principally concentrated in Bujumbura, and beneficiary of the Credit. The project,was to more than double Bujumbura's existing bulk supply by tapping an entirely new source - Lake Tanganyika, on which the city stands - and to strengthen the primary distribution network by adding appropriate reservoir capacity and installing additional backbone lines. Project dimensions were changed during the course of execution, generally upwards from appraisal report projections, and the enlarged project was finally completed early in 1973, about four years behind schedule, for costs of about $1.8 million, only 6% above forecast. The central element in the project, the works for extracting and treating water from Lake Tanganyika, were put into provisional operation much earlier, in December 1969, only about a year behind target completion date. They made an immediate and dramatic difference to the continuity and regularity of water supply to the people of Bujumbura, who had been suffering from increas- ingly severe water shortages. Consumption in the poorer middle class part of the town shot up 50% in one year and nearly 150% over the four years to 1973. The project has however suffered from an.increasingly severe problem of periodic ruptures at the joints in the large-diameter polyethylene intake pipes, causing accelerated depreciation of the pumps because of the mud and sand introduced and involving cut-off of supply for several hours on successive days for repair. Three such ruptures occurred in quick succession late in 1974; they cost about $10,000 each to repair. REGIDESO has not been able to have recourse to the German contractor responsible for the intake since he declared bankruptcy in 1972, but the German Government has now offered a $500,000 credit of which about half would be spent on overhaul and reinforce- ment of the intake by putting steel sleeves on the joints. The Bank had questioned the adequacy of polyethylene pipe for this job at the time of con- tract award and it never seems to have received adequate answers from the project consultant to its emphasis on the need to verify the contractor's proposed welding technique and to provide for satisfactory testing. These points were overwhelmed by a larger debate on the appropriate length of the pipeline, on which the consultant took a rather rigid position' but which was hard to resolve rationally since no provision had been made for the necessary additional studies of variations in water quality in the lake. - 11 - Even including an allowance for these repairs and overhaul, the costs of the overall expansion project work out at USJ27 per cubic meter annual production capacity which is low compared to available figures on the costs of urban water supply expansion in developing countries. Operating costs of the bulk supply system, dominated by costs of electricity for pumping since the requirement of chemicals is small for the efficient slow-filter treatment plant installed at the instance of Bank engineers, are also quite reasonable. Nonetheless actual system operating costs per cubic meter sold are much higher than expected, partly because of maintenance and repairs, but mainly because of the very high distribution losses (about 45%) which result mainly from the dilapidated state of the distribution system. Rehabilitation and expansion here, for which only minor provision was made in the IDA project and appraisal report projections, has suffered from the shortage of skilled workmen, long delay in actually obtaining the UN technical assistance envi- saged under the project and the fact that the Municipality has not paid the bills for supply through public standpipes which are its liability, even though it has collected a related tax. The pace of expansion of the distri- bution system has increased somewhat in recent years, but it has been small and there is now a large backlog of work to be accomplished. This has also limited the expansion of consumption even though actual consumption, assuming a small part of losses to represent beneficial consumption, is reasonably close to the fairly high appraisal report projections. Moreover, doubts have been raised - which still cannot be definitely resolved due to lack of data - whether an expansion of the pre-existing Ndahangwa bulk water supply system in the hills behind the town could not have been accomplished at a fraction of the cost of the new lake system. The WHO engineer with REGIDESO from 1968 to 1972 suggested that an expansion of equivalent size could have been accomplished there for about $360,000, about one-third the cost of the corresponding portions of the project actually built. This had been ruled out at appraisal mainly because of uncertainty whether an intake structure could be built to withstand the very high flash- flood flows and the rocks they bring and for a variety of more minor reasons including the river's heavy sediment load at times and difficulties in maintain- ing the more elaborate treatment plant required for this water. Nonetheless, the old system, with its intake structure built in the 1940s, has continued to operate satisfactorily, as a supplement to the lake plant. Although it was included in his mandate, the project consultant does not seem seriously to have studied the Ndahangwa alternative and he did little, until too late, on improvement of the Ndahangwa treatment plant, which the Bank failed to have included specifically in his contract. Quite apart from the physical works, an important objective of the project and of IDA involvement in it was to build up newly independent Burundi's REGIDESO into an example of a well-run utility organization. Con- siderable thought was given to this matter, and a large number of commitments suggested by the Bank with regard to staffing, training, organization, account- ing principles and financial performance were undertaken by REGIDESO and the Burundi Government. Very few of these commitments have been fulfilled, for a wide variety of reasons, many of which were probably difficult to appreciate or foresee at the time they were formulated: the extent of the educational backlog left by the colonial regime and the absolute shortage of trained personnel, the extreme social tensions of which Burundi has been victim and the virtual impossibility of strong management under these circumstances, the inadequacies of existing expatriate staff and the UN agencies' difficulties in recruiting suitable technical assistance personnel and the Bank's own problems in sustaining the kind of detailed support that would probably have been required by the circumstances. Yet some of the Bank's provisions - for instance, the much greater emphasis given to foreign training, to which people could hardly be spared, than to local in-service training, and some of the constitutional and financial refinements required - seem to have been ill- adapted to the situation prevailing. Nonetheless, only a little due to the Bank initiatives but mainly due to the gradual increase in numbers of educated people and in strength of the Government administration, REGIDESO has strength- ened significantly over time, building up a management team - now mainly con- sisting of young Burundi nationals who first joined the organization in the late 1960s - and a staff of graduates from local schools. Important advances have been made in actual application of the company's Charter, including effective functioning of the Board and annual auditing of accounts, even though not all of the provisions are yet applied. REGIDESO has been further bolstered in the last year or two by an integrated team of German Technical Assistance advisors who are providing the first systematic in-service training. Because of the large distribution losses (especially on water but also on power), an increasingly serious collections problem, and failure to make any significant adjustment to tariffs, REGIDESO's financial situation has very seriously deteriorated over time, even though execution of its invest- ment plans does not appear to have been held up by cash flow shortage. The Credit documents specified a 6% rate of return to revalued net fixed assets to be earned, beginning in 1972 at latest, but the assets have not been offi- cially revalued and the return to non-revalued assets was less than 6% for all operations in 1973 and negative for water operations. It is estimated that water tariff revenues would need to have been about 70% higher than they actually were in 1973 to produce a 6% return to revalued net fixed assets. Water bills outstanding as of the end of 1974 were about equal to 1974 billings for private consumers, but twice 1974 billings for Government and nearly nine times 1974 billings for the Municipality. In view of the Government's large debts to it, REGIDESO has so far refused to make principal repaymehts to Gov- ernment on the IDA Credit, which were due to start in 1971. Bujumbura's water needs were great enough and the costs of project works low enough that, were it possible to quantify the benefits to water supply investments, the IDA project would likely show a satisfactory economic rate of return. But it is fairly clear that a higher-return project in the middle 1960s would have been one that was less ambitious on both physical and institution-building sides, involving a good deal of detailed technical - iv - assistance and focussing on rapid interim improvement of existing systems, on building up local capabilities and on execution of the hydrological studies, on the Ndahangwa River and the lake, that were necessary to give a better basis for decision-making for the long term. Such a project would hardly have been compatible with the Bank Group's capabilities, but it could have been more suitable for a bilateral assistance agency. Much as such an approach might have been better for the economics of Bujumbura's water supply, however, it is unclear whether it would have been practicable and preferable from a broader viewpoint. It is uncertain whether the type of simple but intensive assistance, carrying small equipment sales, that was required would have been available at the time from anywhere. The relatively quick action implied by including all stages from feasibility study through construction under an IDA Credit was almost certainly superior in results to the kind of elaborate preparatory study project that might have been chosen as an alternative. The Burundi Government had picked IDA to approach with this particular problem; there were no other projects ready for IDA assistance at the time; and the Bank Group needed to establish its credibility with a new member country by responding actively to an urgent request for assistance that was reasonably close to its capabilities to meet. A significant side-effect, at the time, of the involvement of IDA, with its emphasis on careful debt management, was to help dissuade the Burundi Government from accepting a large foreign supplier credit that was then being promoted for a hydroelectric scheme much beyond the country's needs. Compared then with the more realistic alternative patterns of development that might have occurred, the IDA-assisted water supply scheme, a bit more expensive than it might have been but basically successful - if only very slowly on the institutional side - can be considered in retrospect a sound initiative, one part of the benefits of which consists of the lessons it yielded to Burundi and to the Bank, relatively new in Africa at the time. PROJECT PERFORMANCE AUDIT REPORT BURUNDI: BUJUMBURA WATER SUPPLY PROJECT (Cr. 85-BU) Background IDA Credit 85-BU, in the original amount of US$1.1 million, was made to the Government of Burundi for on-lending to an autonomous para-statal entity REGIDESO (Regie de Distributions d'Eau et d'Electricite) for expansion of the municipal water supply system serving Bujumbura, the national capital; the Credit was signed in March 1966 and closed, with the cancellation of the $29,000 then remaining, in March 1973. Burundi, in the middle of Africa nearly 1,000 miles from the nearest convenient seaport of Dar es Salaam and with a population of about 3.5 million, is one of the poorest countries in the world, with a 1972 per capita income estimated at about $65, of which only $35 monetary. Both population and per capita income are reckoned to have been growing at only about 1% p.a. over the last decade, the former principally due to significant emigration and civil war and the latter due to a very low investment rate (less than 6% of GNP), extreme shortage of skills and social tensions so severe that they have led to major outbreaks of violence on several occasions, most seriously in 1972. REGIDESO is the national water and power authority, but its opera- tions are still mainly concentrated on Bujumbura, a city that may now have a little less than 150,000 population, where it has a monopoly in the supply of water, produced by itself, and electricity, mainly purchased from the Zairean corporation SNEL (Societe Nationale d'Electricite, former Force de 'Est). Since 1966 the IDA has made two further small credits to Burundi (in 1969 and 1974), but Credit 85-BU still represented about 40% of cumulative IDA disbursements to the country as of the end of 1974. Moreover, even though IDA has provided no more than 2 or 3% of total foreign assistance to the country, the small water supply project, with a total eventual cost of some US$1.8 million equivalent, still accounted for as much as 12% of public invest- ment in infrastructure over the 1967-72 period when it was mainly built and as much as 5% of total investment in the economy in the principal years of con- struction (1968-70). Following independence in July 1962 Burundi had become a member of the Bank and IDA in its own right in 1964. The Government speedily followed up its first contact with the Bank at the 1963 Annual Meeting with an urgent request in March 1964 for assistance in solving Bujumbura's water supply problem along the lines of a project conceived by a WHO engineer. The Bank found the project more advanced in preparation than another transportation scheme also suggested at the time and was impressed by the amplitude and rapidity of the reply, prepared by the WHO engineer for REGIDESO, to its standard questionnaire on water supply projects. After some further prompt- ing by the Government, it managed to field a small appraisal mission in -2- October, and, after the Bank had decided to put prime emphasis in negotiations on strengthening management and collections and to drop a requirement for an immediate tariff increase, the Burundi Government was invited in June 1965 to send a delegation to negotiate the credit. Due to political instability this could not be done until January 1966 but correspondence with REGIDESO about changes in its Statutes, hiring of consultants and securing technical assistance continued and a second, updating field appraisal was made in November 1965. Negotiations were quickly completed because, although there were many loan covenants and conditions - particularly about retention of expatriate staff, training, financial performance, improvement of collections, revaluation of assets, in addition to standard items - and their request for advance copies of the Bank's proposals had been refused, the Burundian delega- tion largely accepted them as already drawn up. The Bank had already approved REGIDESO's selection of, and contract with, consultants for the project - the German firm of IGIP (Ingenieur-Gemeinschaft fur Internationale Planungsaufgaben) - and REGIDESO made their contract to be effective as soon as the Credit was signed, on March 31, 1966. The Credit became effective on schedule in June 1966, the first World Bank involvement in a water supply project in Africa. Project Implementation The project was intended to more than double Bujumbura's then exist- ing bulk water supply capacity by tapping an entirely new source - Lake Tanganyika - and to strengthen the primary distribution network by adding appropriate reservoir capacity and installing additional backbone lines to carry the water produced to main distribution points, to serve areas of the city previously poorly supplied and to equalize pressures among zones. Bujumbura had been drawing virtually all its supply from the hills immediately behind the toin - the Misumba Springs, with an estimated practical capacity of about 1,300 m per day and the Ndahangwa River d version works and treatment plant with a practical capacity of about 7,700 m per day - whence the water could be distributed largely by gravity. The new project would draw water from the lake, along which the city stands, treat it and pump it up the hill for distribution. The original project may also have been intended to include some other items - for instance, some improvements to the existing bulk supply system, particularly the treatment works (as the Bank was later urgently to insist, alleging at one point that a provision of $60,000 had been made speci- ffically for this purpose, but of which no trace could be found in any of the origin4l documents) or some secondary distribution works (though explicitly not house connections or extensions of the existing grid, which were to be financed under REGIDESO's regular operations) 1/ - but the appraisal report gave no detailed description of the project or breakdown of costs, since the first assignment of the consultant was to do a fuller feasibility study of the scheme. 1/ REGIDESO management did not ask for IDA funds for this purpose; moreover, it had on hand large quantities of suitable pipe at the time of project appraisal. -3 - An approximate comparison between what was envisaged in the apprai- sal report and what was finally accomplished (necessarily lacking detail on distribution, which is unavailable, and particularly rough on the cost side due to the different breakdowns used in the appraisal report and the lack of records in REGIDESO on the breakdown of investment expenditure) is given in the following table. Table 1 Project Works and Approximate Costs, Forecast and Actual (cost figures in US$ thousands equivalent) Project Appraisal Approx. Actual Report Project Size Cost Size Cost Lake Intake Line & Structure 800m 110 2,000 m 474 Raw Water Line ) 100 ) 329 Pump Stations ) 300 Treatment Plant 14,500 m 3/d 200 21,500 m 3/d 85 Storage Reservoirs 2,500 m3 200 2,100 m3 200 Distribution 307 381 Contingencies 20% 233 - Engineering/Management Services 222 222 Interest during Construction 108 54 1o709 1,816 Source: Annex Table I. The table shows that a project significantly larger than originally planned in certain 5espects - particularly in overall production and treatment capacity of 21,500 m per day - was achieved for a cost only some 6% larger than osigi- nally budgeted. Additioal reservoir capacity actually provided (1,400 m at el. 848 meters and 700 m at el. 900 meters, to strengthen supply to the medium-high zones) was somewhat less than originally conceived, but, with a view to obtaining a better quality of water, requiring less treatment, the intake was built out 2,000 m into the lake instead of the 800 m originally envisaged by the Bank. The uncovered3 slow-filter design of the treatment plant (filter surface area of 2,600 m ), which appears to have been adopted largely at the insistence of the Bank's engineers, is in fact considered a model of economic design. Bujumbura has been receiving water from the new plant since December 26, 1969, about one year later than the final completion date envisaged for the whole project in the Bank's appraisal report. Insufficient time had been allowed in the detailed schedule contained in the appraisal report for the preliminary studies already provisionally contracted with the consultant, and the consultant concluded after some months of study that completion could probably only be in late 1969, due to likely delays in transport of equipment and materials to Bujumbura. The Bank urged faster completion, especially in view of the urgency of the city's water need. After delays in finishing the preliminary studies, and then the bidding documents, due to the decision to prepare them for two alternative versions of the intake line and the treat- ment plant, and to a relatively small extent (2-3 months) because of a dispute between the Bank and the consultant as to which version and bidder should finally be accepted for the intake line, this contract was signed in March 1968, nine months behind original schedule. Transport delays seem to have been a less serious problem than the consultant had expected but, due to poor organization and difficulties with the pipelaying equipment, the intake contractor (Harmstorf of the Federal Republic of Germany) fell seriously behind his obligations (completion of an intake to 3,000 m within one year of contract signature). Since most of the other related works in the project had proceeded quite close to their planned schedules it was decided late in 1969 to instal a temporary intake at the end of the 500 m of pipe laid in the lake up to that time, so as to enable water deliveries to start. The basic project was more fully completed during 1970. On December 1, 1970 provisional acceptance was declared of the 2,000 m intake, to which it had finally been decided to cut the Harmstorf contract after various contract renegotiations in the first months of the year, culminating in an agreement in April, following a major breakdown of the pipe-laying equipment, imposing all additional3costs and also penalties for late completion on Harmstorf. The 1,400 m reservoir at el. 848, near Kamenge, and the 5-km pipe bringing water there from the new treatment plant were completed during the year, enabling substantial poorer quarters of the city to benefit from the increased supply. Credit funds remaining after completion of all the basic elements in the IGIP project and finally amounting to some US$200,000, were committed only in 1970/71 - mainly for retention of the IGIP Resident Engineer for one extra year, at the Bank's suggestion, to help initial proiect operation and maintenance, and for construction of the additional 700 m reservoir (R 3), a booster pumping station (SP 4) to feed it and the high zone, and some 2,500 m of main distribution line. These work3 were mostly completed during 1972 and finally commissioned in March/April 1973. -5- Operating Problems Apart from repeated problems with some of the metering equipment, the responsibility for which had been disputed between the suppliers (Koppers of Germany) and REGIDESO, the systems installed under the IDA project have worked well most of the time since they wire completed; the lake supply and treatmenj system has operated at 20,000 m per day and regularly produces 16,000 m per day. However ruptures in the intake, all at the joints between the 10-meter lengths of pipe supplied, and each normally involving cut-off of several hours on successive days for repair, as well as accelerated deprecia- tion of the pumps because of the mud and sand introduced, have become increas- ingly frequent in late 1973 and 1974. Already nine such repairs were carried out before provisional acceptance of the intake in December 1970. No problems arose in 1971 and an inspection in July 1972 revealed nothing wrong, but an inspection early in 1973 identified two ruptures about 1,500 m from the shore, and it was then decided to abandon the portion of the intake beyond this point. The problem began again however late in 1973, and late in 1974 there were three breakages, all in the 500 meters nearest the edge of the lake, within a forty-day period. The German Government has now made a provisional commit- ment for about US$500,000 equivalent of aid, of which about half would be for putting steel sleeves on selected joints in the underwater pipeline. Final acceptance of the Harmstorf contract was never given, because of the problems arising, but REGIDESO has been unable to have recourse to the company since it went broke in 1972. Similar difficulties have apparently arisen with such large-diameter (600 mm or 30 inch) polyethylene pipes under water in some other places where they have been installed, but they may have been particularly acute at Bujumbura partly due to the very thickness of the pipe walls (24 mm or 1.2 inches) that the manufacturer recommended, with a view to the long and difficult voyage (Hamburg - Dar es Salaam - Kigoma, whence they were towed up the lake to Bujumbura). Test welds could be, and were, done in Germany, but Harmstorf's special so-called "hydro-jet" method of trenching and then laying a continuous line of pipe welded on board ship prevented testing each weld before it entered the water and also raised problems in maintaining proper temperature control under local conditions. Most of this could and probably should have been foreseen before the contract was ever let. In the discussion on the bid award in September- November 1967 the Bank made repeated reference, in its correspondence with the consultant, to the need to verify the contractor's proposed welding technique and to provide for satisfactory testing, but it never seems to have received a very adequate answer. Part of the difficulty was that there had been only two bidders 1/ (out of six prequalified and invited) for the intake 1/ The consultant states that the very limited number of offers was due to the Bank's insistence that the specifications provide for entrenching the intake pipe, which moreover has complicated maintenance and proved unnecessary under local conditions; however the Bank's files do not appear to support the contention that it was the Bank which required this provision. - 6 - contract, both of which had submitted bids substantially above the consultant's estimates 1/: Harmstorf, with its polyethylene pipe, bidding $567,000, and a Belgian firm Safricas bidding $342,000 for cast-iron pipe and some $365,000 for steel pipe for the 800-m alternative (strongly favored by the Bank, for which IGIP had estimated $250,000 and the Bank originally some $110,000); and only Harmstorf, bidding $842,000, for the 4,000-m alternative, then strongly favored by IGIP and costed at $566,000. The Bank was very concerned because IGIP, with the approval of REGIDESO's Director-General, had begun negotiating with Harmstorf to bring down their price for the preferred long alternative before informing the Bank of the bids received. IGIP raised technical questions about the adequacy of welding and coating of Safricas' steel pipe, the danger of corrosion and, particularly, the feasibility of its proposed laying techniques. 2/ The Bank unfortunately concentrated its arguments around the adequacy of the shorter intake (where, in the absence of adequate studies, 3/ either before or since, of water quality at different distances from shore, it seems to have been only partly right, technical opinion now favoring something beyond at least 1,000 m where the lake bottom begins to drop sharply) instead of picking up a doubt expressed by REGIDESO's Director about the suitability of polyethylene and questioning Harmstorf's new techniques. In commenting on IGIP's questions about Safricas' bid, for instance, the Bank wrote: "... To offset any shortcomings of the steel pipe, one should also consider possible shortcomings in the plastic line on which no data exists on reaction to water of the lake, its prospective life under fairly bigh water tempera- ture and the means for testing adequacy of joint bonding, These points do not particularly bother us but are mentioned to suggest that arguments can be raised regardless of the type of pipe to be used. Since steel pipe intakes have been used extensively all over the world with good success, we . feel that the most conservative approach would favor it over plastic." But the argument swung on the need for a longer pipeline, the Bank remaining unconvinced, but subjected to intense pressure from IGIP against both the Safricas pipe-laying system and the shorter line and from REGIDESO in favor 1/ All the figures cited here refer only to the intake line, the main portion of the contract, which, as then envisaged, also included the shore pump station and raw water line from there to the treatment plant. 2/ The consultant still considers that Safricas lacked adequate equipment and experienced personnel, and feels his objections to have been proven by what he considered poor performance of Safricas on small works later subcontracted by Harmstorf. 3/ The only data available were for Belgian scientific missions, for 1946/47 and 1954/55. - 7 - of a longer line, providing a safer quality of water; recognizing the ultimate responsibility of the borrower and his consultant, it cabled on November 16, 1967 abstention from further objection to award of the contract to Harmstorf for a 3,000-m intake at about $625,000, the price that IGIP had negotiated with the company. Project Benefits Despite the operating difficulties, the project has undoubtedly been of great benefit to the people of Bujumbura. Due to shortage of bulk supply capacity, afternoon cuts of water supply, especially in the more distant poorer areas but also in the wealthier central part of the city, were becoming more and more frequent through 1969, and they were sometimes longer in the rainy season, particularly at the beginning of the year, when the Ndahangwa can become a torrent and bring so much mud that the treatment plant has to be closed down for up to a few days. Although the population was probably growing at about 6% p.a., no significant increase in total water production and sales occurred between 1963 and 1969 (Annex Table IV). Investment in the distribution network, and especially in secondary distribution, has been very limited, although it has been greater in the last five years than prior to 1969. Moreover in 1964-65 a great number of actual connections appear to have been taken up, as a result of the inability or failure of new residents of some houses to pay REGIDESO's changes, and replaced with public taps; connections and subscribers (i.e. connections in official operation, with meters) both fell about 1,000 in 1965 but have since been gradually growing. The following table gives some of the key figures available, which should be taken as orders of magnitude rather than precise information. -8- Table 2 Recorded Growth of Water Subscribers, Distribution Network and Consumption Length of Length of Annual Annual No. of Connections System Sales Production Subscribers (meters) (meters) ('000 m3) ('000 m3) 1955 2,500 51,005 111,369 1,020 n.a. 1960 3,225 62,498 159,907 1,899 1964 3,384 63,759 165,367 1,761 3,171 1965 2,421 2,042 3,079 1967 2,485 68,163 167,954 2,452 3,028 1969 2,685 2,257 1970 2,956 2,705 4,440 1971 3,217 3,633 6,565 1972 3,498 3,517 6,169 1973 3,668 86,406 3,822 7,125 1974 3,976 /a 88,184 /b 209,162 /b % Growth 1955-64 35% 25% 49% 73% 1964-73 8% 36% 25% 117% 125% /a As of October 31. /b As of September 30. In 1963 it was estimated that about 26,000 people might be enjoying tap supply within their place of abode, the remaining 44,000 depending on supply from neighborhood standpipes (about 580 inhabitants per standpipe). IGIP estimated that in 1965 there might be an average of 900 people per standpipe, and they recommended bringing this down below 300, by increasing house connections and building additional standpipes. In fact very few standpipes have been built. However, it appears that the growth of population, perhaps to some 125,000 in 1973, has been accommodated as much by increased density of living in exist- ing houses (including division into apartments) as by increased numbers per standpipe. The quality of service in this respect may not have deteriorated significantly even with the rapid growth of population that has had to be accommodated. -9- But much the most important growth is in the total quantity of water sold, almost all of it since 1969. Policy is to meter all consumption in Bujumbura, but it is possible that a small proportion of current high losses, say 10%, may also Sepresent beneficial consumption. Actual total sales of water, at 10,500 m per day in 1973 (perhaps about 80 1 per inhabitant), were above the IGIP forecasts and, if some beneficial consumption went unrecorded, actual consumption may have fallen only 10-15% short of the Bank appraisal report's rather high projections (Annex Table II). The greatest benefit of the increased supplies has undoubtedly been in the lower middle class quarters; for instance between 1969 and 1970, on first commissioning of the new system, consumption there increased 50% compared with only 13% in the wealthier part of the town, including the commercial/industrial section, while the comparable four-year increases 1969-73 are 136% and 54% respectively. Had greater investment been made in the distribution network, consumption would certainly have grown faster. Funds on relatively easy terms are now expected to be available from the African Development Bank, and REGIDESO is finalizing preparations to start a $2.4 million project of long-planned distribution works, replacing public standpipes with private connections in the middle class quarters. REGIDESO hopes that under this project Government may agree to some reduction in connection charges (now about $155 equivalent for high income areas, plus $255 if a new development requiring mains to be built, and about $70 in the poorer areas, with mains already installed). A supplementary letter to the IDA Credit referred to the need to try to reduce these charges or to spread them over time, but the first was difficult in the absence of any special alternative source of funding and the latter imprudent in view of the mobility of people in Bujumbura; short-term loans have been available for the purpose from the national development bank (BNDE), but Government had refused to contribute to the costs directly. Economic Justification Beyond the urgent need for substantial additional quantities of water to remedy existing shortages and meet future population growth, the principal economic justification of the expansion project provided in the appraisal report was that it would be a useful, reasonably inexpensive comple- ment to the existing unreliable Ndahangwa River works which, it was suggested, might even be destroyed by floods. Although not explicitly discussed in the appraisal report, attention must also have been given to the correct sizi5g of the scheme since the appraisal report dealt with a project,of 14,500 m / day capacity, designed to increase system capacity (including the existing Ndahangwa works) to a level adequate to meet projected 1978 requirements, whereas 5he WHO engineer's original proposal had been for a scheme of 21,500 m /day capacity designed to a 1984 time-horizon. Considerable doubts have been raised as to whether more could not have been made, at very low cost compared with the Lake project, of the Ndahangwa River - or of other rivers in the hills above Bujumbura, such as the River Kanyesha, from which gravity delivery would also have been possible. - 10 - There are no adequate flow records 1/ for the Ndahangwa, but there is little doubt that there is sufficient flow to support a substantially larger delivery than now made; the canal which carries the water from the river off-take to the pipes to the treatment plant was originally designed to serve a small hydroelectric plant, long since disused, and it delivers several times the amount of water that the pipes can take 2/ but still much less in most seasons than the normal river flow. The WHO engineer with REGIDESO in 1969 concluded from data available for 1967 that the low flow was probably about ten times present delivery to the treatment plant. He made a very rough estimate of about $360,000 equivalent, or one-third the amount for the corresponding portions of the IDA project, for a bulk supply scheme, including river off-take, treatment plant and reservoir capacity, supplying the same quantity of water but located at el. 900 m such that it would have the additional advantage of enabling delivery by gravity for all areas below el. 870 m. He wrote that it was a very great pity that this possibility had not been seriously studied at the time of the investment decision. He also discussed various other possibilities such as a more expensive scheme involving production of elec- tricity as well as water by building a small dam further up the gorge. Some consideration had of course been given to possibilities on the Ndahangwa River by REGIDESO, the Bank appraisal mission and later, to a very small extent, by IGIP, and there were significant factors against it. REGIDESO's then Director-General and WHO adviser were both very convinced that it had limited potential. Difficulties were mainly the lack of data, erosion and landslides at the off-take, high turbidity of the water and heavy sediment load at times, reducing the effective capacity of the existing treatment plant by some 20%. A part of the canal had collapsed, and had to be rebuilt, in 1963, and it was felt that the range between low flow and flash flood was extraordinarily high, such that it was uncertain whether an intake structure could be built to withstand high flows and the rocks they brought. Moreover difficulties had arisen in the maintenance of the treatment plant with the departure of some skilled workmen after separation from the Congo and there was always the fear that chemicals (not required for the Lake scheme in significant quantity), most of which had to be imported, would become unavail- able. On presentation of the proposed IDA project, the Executive Directors were even told that "there will undoubtedly come a time in the very near future when this (the Ndahangwa) system will have to be entirely abandoned and the Lake Tanganyika source used entirely". Because the data are still largely lacking and the necessary engineer- ing studies have not been done a conclusive answer is.still impossible, but what is fairly clear is that, with some but not insuperable difficulty in keeping existing systems going while work was underway, the Ndahangwa works and treatment plant could have been expanded quite inexpensively to carry the city a further 4-5 years, probably the minimum period required for the hydro- logical studies, on the Ndahangwa and on the lake, which were needed to reach 1/ A UN adviser to the Government has started a regular program in recent years. 2/ Its capacity is actually about ten times that of the pipes. - 11 - sound decisions. The Bank went to considerable lengths, even drawing sketch- designs, to urge some such adjustments to the capacity of the Ndahangwa system in 1966-68 - such as expansion of the pipe capacity to the treatment plant, improvement of the filter system to increase its speed, and increasing the size of the outlet from the treatment plant - as means to provide more water, pending completion of the Lake project; a more significant increase in capacity would have involved also construction of additional filter capacity., The Bank engineers, WHO adviser and IGIP all made recommendations, but very little was actually done, apparently mainly because it was known that the Lake project would be coming. Production from the Ndahangwa and Misumba systems, although with variable costs less than half those for the Lake system per cubic meter' produced (mainly because of the much less use of electricity), has in fact been reduced greatly in recent years - about 2.3 million m3 of treated water in 1974 and 2.1 million m3 in 1973, compared with just over 3 million m3 in the years before 1970. If the normal capacity of the Misumba and Ndahangwa systems is assumed to be 9,000 m3/day, as still seems reasonable, then, with the comple- tion of the Lake project, Bujumbura has a treated water production capacity today of some 30,500 m3/day, about 2.8 times 1973 recorded consumption but only 1.6 times actual 1973 production because of the very high distribution losses resulting mainly from the dilapidated state of the distribution system. Assuming reduction of losses to a more reasonable 20%, this capacity would be sufficient to meet demand till about 1982 on IGIP's original projections and 1979/80 on its new 1973 projections (see Annex Table II). The Ban,k has criticized IGIP for designing the Lake project to 21,500 m3/day rather than the appraisal report's 14,500 m3, emphasing that the effective system peak supply capacity under the Bank-proposed scale of project could have been very cheaply increased by adding storage. More important seem to be the weaknesses the Bank displayed in advising and guiding the consultants, despite its prudent early stress to them in general terms to seek low-cost solutions. Although IGIP emphasized, in its original proposal, its strong capacity to analyze alternative possible solutions to a problem, the final terms of reference approved by the Bank made no reference to the need to study alternatives, only "variants', so it is not surprising it gave no serious attention to the Ndahangwa; and the variants it did consider did not include different phasings of system growth. The Bank became very concerned in 1968 that ICIP had not done more to advise on "improvement of existing works", but then it was found,that its terms of reference made only the briefest reference merely.to "minor improve- ments to the existing treatment plant", and they had to be revised. 1/ Given the importance of possible differences in water quality, in different seasons 1/ The Bank now states in this connection: "It is true that the Terms of Reference do not make mention of this, but on the occasions prior to signing of the Contract, the consultants were informed on the needs of the changes, and because of their limited nature, agreed to give REGIDESO assistance without incorporating the work in their contract. We were remiss in not insisting that the contract include this item." - 12 - and at different places in the lake, for the design of a project of the type envisaged, the terms of reference might have specifically foreseen the need for this. There is no evidence that the Bank reacted to IGIP's stress in its October 1966 Feasibility Report on the need for such studies and for distri- bution system exBansion, nor did the Bank raise a question at the time as to whether 21,500 m /day was too large or whether parts of the scheme should be done in phases. When IGIP produced later in 1966 its rather unconvincing report propounding the 4,000-m intake, there is no evidence of the Bank immediately demanding the necessary substantiation - for instance as to just what extra treatment costs IGIP now believed to be involved with the shorter intake line and as to just why the much-vaunted lower annual costs of the scheme with the longer intake would be likely (they arose mainly from a simple change in assumption from 33 to 40 years for the life of an intake line); instead, following the consultants' general emphasis that a longer intake bringing purer water would permit some economies in the treatment plant and hoping that the choice might be resolved by comparison of combined bids, the Bank chose the rather expensive immediate route of having bids called for both alternatives - still without the appropriate water quality studies being done. Staffing and Training Beyond economic justification of the particular expansion proposed, the appraisal report gave special justification for participation by IDA, to give support to the administration of REGIDESO " to augment and retain foreign staff while at the same time taking steps to begin the process of training the national personnel for the long range," thereby making it a model of a well-run organization. The appraisal mission had found an organization run essentially by an impressive Burundian Director-General, the only national with a high school education in REGIDESO, and four foreign technicians, staying on from colonial days, all overworked and with little time for training or planning. It had given considerable time and thought to a solution. After the Bank, under general policy of the time, had rejected the mission's recommendation that provision be made within the Credit (or out of technical assistance grant funds) for multi-year fellowships for foreign training, and REGIDESO had refused to pay for additional foreign staff, the solution finally developed is summarized well by the points in a series of supplementary letters agreed and signed by REGIDESO and the Government. - REGIDESO staff will be increased by one experienced engineer (envisaged to be obtained from UN sources); - His post and others now occupied by expatriates would continue to be filled by expatriates as long as IDA and REGIDESO think they should be; - 13 - - Government would secure annually at least two scholarships for foreign training of high school graduates selected by REGIDESO in business administration and engineering who would join REGIDESO on return and, after further in-service training, be considered for key positions; - REGIDESO would seek foreign help to provide in-service training for employees on both operational and administrative sides; - REGIDESO would obtain an experienced administrator to help it if collections and water losses did not achieve certain targets for improvement within a year (i.e. by March 31, 1967). None of these provisions has been fully operative, and that by no means entirely due to deficiencies on the part of REGIDESO. Closest to being fulfilled is perhaps the first one, insofar as a WHO engineer, initially requested in May 1965, finally materialized in March 1968 and stayed through March 1972, playing a crucial role in REGIDESO operations and execution of the IDA project, but never formally in the executive position implied by the supplementary letter, urged by the Bank and rejected by WHO Africa Regional Office as a matter of policy. Extensive other dealings with the UN system for provision of technical assistance were unproductive: an electrical engineer, for instance, was requested in November 1968, included into the budget by 1970, and-recruited after great difficulties late in 1971, but he took sick soon after arrival and had to leave early in 1972. The expatriates held over from pre-Independence days were relieved of executive responsibilities and made advisors to Burundian counterparts soon after the present Director- General came to REGIDESO in September 1966; they were considered expensive, not very competent, not interested in training and of limited education, and Bank supervision reports indicate a good deal of foundation for these concerns; most of them had left by the end of 1968. When the Director-General found the Burundian whom he had appointed Chief Accountant insufficiently competent he did appoint to this post in late 1968 an expatriate who had been with REGIDESO two or three years, but this man failed to return from vacation in 1973 after misappropriation of funds had been alleged. As regards the foreign scholarships, they seem to have been available in sufficient number, but the difficulty has been to find suitable high school graduates to benefit from them and then, if they were considered good enough for that, to make the sacrifice of losing such people for several years. Two men were finally sent, one a man who had been head of the Water Department and for whom a three-year scholarship was eventually obtained in October 1969. Indicative of the resultant difficulties, a subsequent Bank mission complained in its report that his absence "has aggravated the problem of insufficient supervision of the technical operations, which is especially disturbing...". The mission went on to suggest that perhaps the real shortage was in skilled labor rather than university graduates, but it appears that the supplementary- letter understanding that was least pursued by the Bank was the one referring to in-service training, and apart from a small amount of training provided to - 14 - the senior staff by the WHO engineer and more significant training by the IGIP Resident Engineer, especially during his last two years in Burundi, no programs were in fact established, principally because of the lack of time on the part of the very few suitable foreign staff available. The question of hiring an experienced foreign administrator was never pursued very seriously by the Bank mainly because, at the relevant moment, water distribution losses had apparently fallen substantially and the problem of collections was then more one of delays in payment of Government bills than of more effective billing of private customers. With the benefit of hindsight it may be said that the three principal errors in the Bank's diagnosis of the situation in 1964-65 seem to have been the assumptions with regard to availability of high school graduates, the readiness of REGIDESO to hire expatriates into executive positions and the ability of the UN agencies to provide timely technical assistance of the right caliber. All these problems were probably difficult to foresee, although some further checking of the relevant portions of the educational system might have been made (although, it should be noted, that would have been almost impossible in the scope of a two-week appraisal mission). In fact even by the late 1960s and early 1970s Burundi appears to have been producing only about 300 general secondary school graduates, all oriented to higher education, and perhaps about 100 technical and vocational school graduates - in all fields, from agriculture and administration to social and medical work. No facilities have been available for training accountants. The extent of the educational backlog left by the colonial regime, probably further exacerbated by social obstacles to training and hiring entirely on a merit basis, may have been hard to appreciate at the time. As regards REGIDESO's preparedness to hire expatriate executives the Bank might possibly have been quicker to recognize the unrealism of its expectations in this respect, for the Director-General with whom correspondence had been underway through 1965 had already firmly refused to hire the foreign engineer and administrator whom the Bank had first thought to make preconditions of the Credit, but his refusal seems to have been mainly on economy grounds and he did of course still have Belgian executives from the former colonial regime. Despite the great difficulties it has faced and the very limited contribution that these arrangements in the IDA Credit have been able to make, REGIDESO has been increasingly able to build up both a management team - now mainly consisting of young Burundi nationals first brought to REGIDESO in the late 1960s - and a staff of graduates from the Bujumbura Technical School and School of Administration. Following initiatives taken, with Bank support, in 1971 and, in implementation of promises made in 1972, the German Government is providing a technical assistance team, already including a water engineer (since March 1973) and a water technician, a power engineer and now a senior administrator. This team is providing the first systematic training of technicians, after REGIDESO had turned down the German offer of scholarships abroad for 15 technicians, expressing preference for an instructor to be sent, with a workshop, teaching facilities and funds for local scholarships, which the German authorities generously provided. Nonetheless, REGIDESO still - 15 - remains obviously very short of skilled staff, and many junior positions of authority on its organization chart remain unfilled, increasing the burden on the senior staff. Total staff has risen from about 197 in 1965 to 248 in 1974, not very large compared with the expansion of operations. Institutional Development The basic constitutional position of the REGIDESO, as a financially autonomous Government-owned enterprise under the Minister of Public Works, and its organization in two major Divisions, Administration and Technical Operations, have remained largely unchanged over the ten years the Bank Group has been associated with it, but many important changes of detail have been made and, above all, the company's Statutes, although still not fully applied, have come into much more effective operation. On examining the Charter given to REGIDESO at the time of national independence, providing for a Director (appointed by the King on the proposal of the Minister of Public Works), assisted by a Consultative Council and subject to directives from the Minister, the Bank had been concerned as to whether autonomy was sufficiently clearly established. Pursuant to discussions with the first appraisal mission, REGIDESO obtained approval from the King in February 1965 for a number of minor changes - mostly clarification of the financial principles involved - but the Director-General rejected repeated Bank urging that his independence be bolstered by giving him a fixed term of appointment, stating that his appointment was, in effect, for life. The Bank did obtain commitments from the Government, in the Development Credit Agreement, not to interfere with the performance of REGIDESO's obligations under the Project Agreement between IDA and REGIDESO, to consult with it before changing the Director-General, and not to make changes in the Charter or in the Subsidiary Loan Agreement (between the Government and REGIDESO, on- lending the IDA funds) without IDA's prior agreement. The Bank's appraisal report continued to express concern about the adequacy of the Charter but stated that it was acceptable for at least the period of the project. In September 1966, six months after the Credit documents had been signed, the Government suspended the Director-General, charging him with maladministration and establishing an investigatory committee; an electrical engineer who had studied five years in Belgium and spent a year as a Director- General in the Ministry of Communications was placed temporarily in charge, and the Government wrote to the Bank saying that it intended to appoint him Director (abolishing the post of Director-General) and to create a Board to support him. Upon receiving details of his background, the Bank approved his appointment at the end of January 1967, but it could not be finalized because, as the Government had informed the Bank, the Charter was being revised, to strengthen controls and avoid repetition of the alleged mal- administration - particularly, excessive staff benefits and growth of accounts receivable - of the recent past. The Bank was handed a new draft Charter in June 1967 but did not comment until September, then expressing grave reserva- tions about the extent of powers given both to the proposed Board and the Minister of Public Works and about lack of clarity as to which was in charge. - 16 - Further correspondence ensued in late 1967, and the new Charter, reflecting most of the Bank's principal suggestions and giving more specific responsi- bilities to the Director-General and the Board and more general supervision to the Minister of Public Works, was finally approved by a new Minister of Public Works and signed by the President (who had replaced the King on his ouster in November 1966) in October 1968 (with minor clarificatory amendments in March 1969). The man who had been temporarily in charge was formally appointed Director-General. Both the Director-General and the Charter have remained unchanged since. Much more significant in practice than changes in the concepts and wording of the Charter seem to have been the growing extent to which several provisions have been applied, probably reflecting mainly the gradual increase in availability of qualified personnel and hence in strength of Government administration. The Consultative Council of the old Charter had never met, but the Board established under the 1968 revision, and including representatives of the major ministries involved (though not, due to the Bank's advice to reduce the number of members, of the Ministry of Economy - which may be un- fortunate in view of its evident involvement in approving proposed tariff increases) and of consumers, has met several times each year, less regularly (due to frequent changes in Government composition) than would have been desirable, and given important support and protection to REGIDESO. The auditor, required under the old Charter to be chosen by the Minister of Public Works, was never appointed; the new Charter required a "commissaire aux comptes" to be appointed by the Minister of Finance from his staff, but the man so appointed in August 1970 never appeared; REGIDESO's first full audit, for the years 1970 and 1971, was completed by a Burundian "commissaire aux comptes", with the help of a European employee of the Ministry of Finance, in April 1973 and that for the years 1972 and 1973 in July 1974. The Bank's persistent stress on the need for auditing of accounts (the Project Agreement required annual audits "satisfactory to" IDA, and it was a regular topic of supervision missions' discussion) may possibly have played some role in stimulating this action. The two principal posts under the Director-General - Administrative Director and Technical Director - were never filled, for lack of staff, except that an untrained man appointed by the Minister of Public Works early in 1968 to the Administrative Directorate without the support of the Director-General has remained officially in office, but inactive and unpaid. This has created a heavy burden for the Director-General who has had in effect, with the help of the WHO adviser and the IGIP Resident Engineer, to fill these two posts, and sometimes others too, such as Head of the Water Department in 1969. Very recently a Burundian Technical Director (a civil engineer) and a Coordinator of Projects have been appointed, a graduate economist has been recruited to strengthen the accounting work, and the Director-General will now be supported on the administrative side by the German counsellor. The Bank reviewed the broad organizational structure at appraisal and, given the personnel shortages, found no particular deficiencies. The Bank did not give its support to a UNDP reorganization study completed in - 17 - mid-1967, because, along with many useful detailed suggestions which have been partly implemented (particularly in accounting), some of the proposals seemed unwise, and still do, and, in particular, the report took the view that expatriates should hold only advisory positions and be retained only if their salaries were paid under foreign technical assistance programs. REGIDESO has in fact re-shaped its organization below the top level on several occasions, most recently in mid-1974, laying out a new plan with a better distribution of functions. Financial Performance In the early years after the Credit was made REGIDESO showed a much stronger financial performance than had been projected by the Bank, mainly because it was able to meet virtually all its power requirements with Force de 'Est production, on which it has had a very good margin, purchasing the power at BuF 1.15 (USc 1.47) per kwh and selling it at an average price of about BuF 3.85 (USc 4.94). As a result of this it was able to finance from internally generated funds a relatively high proportion of fixed investment (just over 50% for the period 1966-73 compared with 42% projected in the Bank appraisal report) even though investment was substan- tially larger than projected, mainly due to inadequate allowance in the appraisal report for non-project works (see Annex Table III). Since 1969 REGIDESO's financial performance has deteriorated sharply due mainly to failure to implement any increase in the water tariff (the need for which on completion of project works was foreseen both by the Bank and by IGIP), greatly increased water losses once the larger bulk supplies became available, the higher costs of maintenance and electricity for the new water project and very rapidly increasing accounts receivable (equivalent to nearly one and a half year's billings by the end of 1973 compared with little over half a year at the end of 1969). REGIDESO had not requested a tariff increase before the end of 1974, so that, with inflation having been at about 3% per annum since 1965, its prices in real terms have been falling, but it is still considered uncertain whether the Ministry of Economy will approve the substantial price increase for which an application awaited confirmation by REGIDESO's Board at the end of 1974, because incomes appear to have grown slowly, even in Bujumbura, with the poor growth of the national economy as a whole. Collections have been a perennial problem for REGIDESO, particularly with the Municipality of Bujumbura (which is supposed to pay for water consumed through standpipes and does collect a related tax, but has never yet made a payment to REGIDESO) and with the national Government. The latter has ihowever always paid up after some delay, and in time not to cause REGIDESO serious cash-flow problems, and it has apparently now undertaken definitively to bear the charges to the Municipality. The followIng table indicates the recent growth of accounts receivable from sales. - 18 - Table 3 Growth of Accounts Receivable (Sales) - Bujumbura Operations (figures in BuF '000 as of December 31) 1969 1971 1973 Government 9,306 28,142 64,890 Municipality 25,803 17,575 26,564 Private 47,218 66,966 79,127 Gross Total 87,327 112,683 170,581 Provisions 51,037 69,772 85,505 31,290 42,911 85,076 REGIDESO has sufficient confidences in recovery that it cut its provisions for bad debts in its 1973 income statement to 5% of the year's sales from the 10% used earlier. However, the situation with regard to collections appears to have deteriorated for all main classes of consumer, except private parties in the poorer part of the town, in each of the past few years. Water bills outstanding as of the end of 1974 were equal to 1974 billings for private consumers in the central business district, twice 1974 billings for Govern- ment, nearly nine times 1974 billings for the Municipality but less than one year's billings for private consumers in the poorer part of the city. The supplementary letter on rates agreed with IDA in 1966 provided that "rates for the water, power and over-all operations shall be maintained (i) to cover all costs of the respective operations at all times; and (ii) to produce a rate of return of at least 6% on net fixed assets in operation, reasonably valued, at the earliest possible date, but in any event not later than the beginning of the financial year 1972." Rates were to be reviewed with IDA at the end of 1967, and every two years thereafter, and a complete physical inventory and revaluation of fixed assets were to be carried out for incor- poration in the books by 1970. The rate reviews were not pressed by IDA when the time came, but the revaluation was and the WHO engineer eventually carried out a rough reassessment of the water assets, but it was never incorporated in the books and, despite repeated Bank pressure, no revaluation has so far been - 19 - attempted for the power assets, although REGIDESO's Charter makes two ref- erences to the need to use updated asset values. 1/ Nor does REGIDESO calculate or use the rate of return concept, also mentioned in the Charter. Such a calculation, on the basis of its official accounts, largely unaudited through 1969 and very slightly adjusted in light of auditing since then, gives the following results for comparison with appraisal report projections. Table 4 Forecast and Actual Rate of Return on Net Fixed Assets in Operation 1966 1967 1968 1969 1970 1971 1972 1973 Water Department Appraisal Forecast /a 0.5 1.4 0.5 2.3 4.0 5.1 6.1 7.3 Actual (0.5) 2.9 2.4 2.7 (5.8) (2.3) (4.3) (6.7) REGIDESO (over-all) Appraisal Forecast /a 0.2 1.6 1.3 2.4 3.8 4.8 6.6 8.1 Actual 8.8 12.4 12.3 16.1 6.3 4.8 3.0 5.9 /a Operations in Gitega as well as Bujumbura, whereas "Actual" figures quoted refer only to Bujumbura, which accounts for about 90% of REGIDESO operations. REGIDESO, on this basis, has failed to adhere to the tariff standards agreed with the Bank for the Water Department in 1966 and each year since 1970 and for Bujumbura operations as a whole since January 1, 1972. While the Bank has placed a lot of emphasis on the need to carry out the revaluation of assets, it has also persistently argued that depreciation allowances on Water Department assets would only need to be 2.25-2.50% p.a.; REGIDESO has always stuck to its "at lest 4%" formula. It is interesting to note that recalculation of REGIDESO's rate of return performance on water operations, incorporating into the accounts (a) the rough revaluation of assets by the WHO engineer (approximately doubling their depreciated value as of the end of 1968) and (b) Bank recommendations on adequate depreciation allowance 1/ Article 22, following precisely the text suggested by the Bank in 1967, refers to the need to use "current" values of assets and their useful lives to calculate depreciation for the purpose of establishing tariff rates, while Article 26, not originating with the Bank, says that depreciation allowances should be no less than 4% of the replacement value of the assets. - 20 - would yield the following results: 0.2% actual return in 1971, -1.2% in 1972 and -3.3% in 1973. On this basis REGIDESO was still in compliance with the supplementary letter in 1971. More importantly, these calculations show that the increase in tariff revenue that would have been required in 1973 to produce a 6% return on REGIDESO's accounting basis (historic value of assets but high percentage depreciation allowance) would have actually been a little more, about 72%, than it would have been on the Bank's accounting basis, about 64%. This illustration raises a question whether, given the Bank's assumption that 2.25-2.50% depreciation allowances were adequate in water, assets revaluation was, in retrospect, really worth the considerable emphasis given to it in project supervision. It should be noted that, in 1971, REGIDESO and the Government did eventually infringe the prime legal document underlying the IDA Credit, the Development Credit Agreement, in a manner such that IDA has the legal power to premature the Credit. The Development Credit Agreement states: "Section 4.04. Except as the Association shall otherwise agree the Borrower shall not amend, assign, abrogate or waive any provision of the Subsidiary Loan Agreement." and, "Section 5.02. For the purposes of Section 5.02 (j) of the Regulations, the following additional events are specified.... (b) Before the Project Agreement shall have terminated in accordance with its terms: (ii) a default shall have occurred in the performance of any covenant or agreement on the part of the Borrower or REGIDESO under the Subsidiary Loan Agreement." This Subsidiary Loan Agreement, of February 25, 1966, provides inter alia that REGIDESO shall repay to the Government the IDA funds on-lent to it according to a specified schedule beginning in June 1971. Although the Bank has not been aware of the matter, REGIDESO, while maintaining interest pay- ments, has understandably refused to make principal repayments to the Govern- ment pending settlement of Government's large debts to it. The cumulative amount of repayments due as of December 1974, according to the original schedule, would be BuF 32,385,000. Bank Administration Given the fact that Burundi was entirely new to the Bank in 1963 and in light of the extent of problems that were identified in the first appraisal mission and became more apparent over the succeeding year and a half before the Credit was signed, it is perhaps surprising - and important to a - 21 - balanced assessment of the Bank's role - to recognize the relatively small amount of time that Bank representatives were able to spend in Burundi in connection with the project. Only in 1971, when the project was largely completed, were there two supervision missions in one year. Table 5 Bank Missions to Burundi in Connection with Bujumbura Water Supply Project No. of No. of Total Month/Year Description of Mission Persons Weeks /a Man-Weeks Nov. 1964 Appraisal 1 2 2 4 Nov. 1965 Appraisal II 3 1 3 7 Aug. 1966 Supervision I 1 1 1 June 1967 Supervision II 3 1 3 Oct. 1968 Supervision III 3 2 6 Aug. 1969 Supervision IV 3 1 3 Apr. 1970 Supervision V 2 1 2 Jan. 1971 Supervision VI 2 1 2 Aug. 1971 Supervision VII 1 1 1 Feb. 1973 Supervision VIII 2 1 2 20 /a Excluding travel time and a number of visits made in connection with the project to WHO offices and the consultant's offices at Darmstadt. The length of time elapsing between missions was not always the choice of the Bank; the important mission in October 1968, for instance, arrived only 16 months after its predecessor because REGIDESO had requested its postponement due to business trips and vacation periods, and a mission planned for 1972 had to be cancelled due to the disturbed state of the country at the time. - 22 - Nonetheless this was basically a project that had only annual super- vision visits, and the Bank seems now to consider that it should have had two visits a year. REGIDESO, on the other hand, found the Bank's supervision missions too much like unsympathetic policemen of rather irrelevant laws rather than helpful advisers with constructive suggestions for solving the consider- able problems faced. At one poirt a Bank supervision mission did work out a completely revised and updated set of conditions to be attached to the Credit, but the idea of renegotiating the supplementary letters was apparently turned down in Washington, and it is true that that, without some further assist- ance, would probably have had limited value. In the later years the super- vision missions gave increasing attention to efforts to encourage provision of technical assistance from other sources, especially bilateral, and that was probably helpful. Yet there was a remarkably large amount of misunderstanding and poor communication in connection with this project - from the time when the Director-General was refused pre-negotiation copies of the Bank's proposals and wrote six months after negotiations to enquire what the appraisal report was, through the discussions about how much "improvement of existing works" was supposed to be in the project, to the repeated allegations, not always well founded, of breach of commitments and the persistent, but never in the event correct, reporting to Washington that REGIDESO had now decided to put some expatriates into executive positions. While there were clearly one or two occasions when a more expeditious supervision mission would have been helpful, for instance quickly following the preliminary appointment of the new head of REGIDESO in September 1966 (there was in fact no mission until the "regular schedule" of June the following year) and when the consultant company was doing its preliminary work, it must be doubted whether an automatic program of two or even three missions a year, especially after 1967, would have been of substantially greater help. Conclusions Despite all the difficulties that arose in implementation and early operation, the physical project can be considered in large measure a success, not only in the sense that it was eventually built to larger capacity than planned without major additional expenditure, but also in the senses that it has workqd well most of the time and been heavily used and that it has met a clear and crucial need at low cost. The time overrun to initial completion was about a year, or more than one-third of original construction schedule, but completion brought a dramatic improvement in the regularity and continuity of water supply for the people of Bujumbura, especially the middle class. The original cost overrun was only about 6%, and even if the costs of repairs to date (about $50,000) and of the overhaul being financed by the German Government (about $250,000) are added, the overrun is less than 25%. Total capital cost of the expansion (including these additional items) averages only USJ 27 per cubic meter annual production capacity which is low compared to any figures available for urban water supply systems in developing countries, and bulk system operating costs are not unreasonable. There seems little doubt that, were it possible to quantify the benefits of water supply invest- ments, this project, as built, would show a satisfactory economic rate of return. At the same time it remains very unclear whether the project was the lowest-cost solution to Bujumbura's water supply problem, because of the consultant's mistake on the intake design, the Bank's failure to require analyses of alternative sized schemes, and nagging doubts as to whether satisfactory gravity supply from the hills could not have been developed at a fraction of the cost. Most immediately remunerative distribution works and maintenance were more than usually an orphan, for the normal reason that, as one supervision mission put it, "distribution had to wait until it was known how much would be available in the Credit after the main contracts," and additionally because of the dire lack of skilled workmen in this case; the fact that consultants, concerned with the main contracts and hired commercially, could be on the job two years earlier than technical assistance, concerned with regular operations and provided through UN sources; and the financial problem that the Municipality, which was supposed to pay for the water supplied through the part of the system most in need of expansion, the stand- pipes, was the worst-paying customer. It is interesting, with hindsight, to note that a broad doubt as to whether the project was not overdesigned or oversized in some way could have arisen from comparative study of the project appraisal report completed in March 1966 and the Bank's economic report on Burundi issued in May 1965 and prepared with a view to submission of the water project for approval by the Executive Directors. That economic report showed that the Burundian economy had been able to reach an investment rate of only about 5% of GNP in 1963 - a figure which, it was implied, might rise a little with prospects for slightly increased foreign assistance and possible small budget surpluses on current account. A 5% investment rate meant about $3.00 per head of popu- lation per year; perhaps it could have been expected to rise to $4.00. Yet this project alone, as then proposed, involved fixed investment of about $7.00 per year, in each of the three main years of execution (1967-69), per head of expected beneficiary population (about 80,000 was the projected population of Bujumbura in 1968). Actual relative figures have not been greatly different from these, except that Bujumbura's population has since been reckoned higher, reducing project investment on a per capita basis to about $5.00 per year, while national investment has taken considerably longer than then expected to rise from the $3.00 per capita level to $4.00. If the project was overdesigned, so also was the Credit: there were too many conditions and understandings, and not the right priorities, harming the effectiveness of the Bank's institution-building effort,, even though some of its emphases - for instance on auditing-and collections and, in later,years, on bilateral technical assistance and distribution improve- ment - have been useful. Much attention was given to the details of a piece of paper, the Charter, where what was needed was to get the main concepts of any such piece of paper operating. Much argument was spent on the need to give foreigners executive positions, in an organization that was deciding, with some reason from its own experience, to take the route of learning by experience even at some cost in efficiency. There was overconcentration on - 24 - training abroad, to which people could barely be spared, for high level staff, to the relative neglect of preparation of skilled workers who turned out to be in shortest supply. There was excessive emphasis on largely offsetting financial requirements which are still, nearly ten years later, considered somewhat irrelevant, perhaps understandably in a country which is not yet producing accountants and in an organization so short of trained staff that it is a significant accomplishment to keep established systems running. Perhaps the most useful, if oversimplified, way to present the conclusions of this study on the "institution-building" issues peripheral to the physical project but central to the Bank effort in this Credit is in tabular form, with the Ds representing very low, if any, emphasis and the As and A+s indicating very heavy and heaviest emphasis. The first three columns attempt to show IDA's actual emphases to the Burundi Government and REGIDESO in various periods and the last the emphases that this perform- ance audit suggests should have been given, mainly with an eye to the earlier period of appraisal, negotiation and early disbursement. - 25 - Table 6 Actual and Suggested IDA Emphases in Bujumbura Water Project Actual Mar. 1968- Oct. 1968- Suggested with Appraisal Oct. 1968 Mid-1970 Hindsight Autonomy (Charter) A+ A+ D D Foreign Executives A+ B A D Foreign Advisers D B A A Foreign Training A C B D Local Training C D C A+ Improve Existing Works D A A A+ Secondary Distribution D D A+ A Cheaper Connections B C C C Reduce Water Losses B B B B Collections A A+ A B Tariff Increases D D C C Assets Revaluation B B A D Auditing B A A A In retrospect, then, the best project that might have been under- taken was one that would have been significantly less ambitious on both physical and institution-building sides, involving a good deal of detailed technical assistance and focussing on rapid interim improvement of existing works, on building up local capabilities and on execution of the hydrological studies, on the Ndahangwa River and on the lake, that were necessary to give a better basis for decision-making for the long term. Such a project would hardly have been suitable for the Bank Group, without a large reservoir of technipal assistants to call upon and with limited staff-time provided to go into intricacies and to supervise consultants. It could have been more suitable for a bilateral assistance agency. Much as such an approach might have been better for the economics of Bujumbura's water supply, however, it is unclear whether it would have been practicable and preferable from a broader point of view. It is uncertain whether the type of simple but intensive assistance, - 26 - carrying small equipment sales, that was required would have been available at the time from anywhere. The relatively quick action implied by including all stages from feasibility study through construction under an IDA Credit was almost certainly superior in results to the kind of elaborate preparatory study project that might have been chosen as an alternative. The Burundi Government had picked IDA to approach with this particular problem; there were no other projects ready for IDA assistance at the time, as was pointed out to an Executive Director who questioned why the priorities of road re- habilitation and agriculture, stressed in the Bank's economic report, were not receiving first attention; and the Bank Group needed to establish its credibility with a new member country by responding actively to an urgent request for assistance that was reasonably close to its capabilities to meet. A significant side-effect, at the time, of the involvement of IDA, with its emphasis on careful debt management, was to help dissuade the Burundi Govern- ment from accepting a large supplier credit that was then being promoted for a hydroelectric scheme much beyond the country's needs. Compared then with the more realistic alternative patterns of development that might have occurred, the IDA-assisted water supply scheme, a bit more expensive than it might have been but basically successful - if only very slowly on the institu- tional side - can be considered in retrospect a sound initiative, one part of the benefits of which consists of the lessons it yielded to Burundi and to the Bank, relatively new in Africa at the time. BURUNDI: Approximate Evolution of Main Elements and Costs of Bujumbura Water Supply Project (cost figures in US$ '000) original WHO IBRD Project IGIP. Oct. 1966 IGIP Dec. 1966 IGIP Dec. 1967 Approx. Actual 1963-64 Proposal Appraised 1964-65 Feasibility Study Recommended Solution Final Eng'g. Report Project Size Cost Size Cost Size Cost Size Cost Size Cost Size Cost Lake Intake Line 2x 40 800 m 110 815 m 250 4,000 m 566 3,000 m 600 2,000 m 474 and Structure 300 m Raw Water Line ) ) 40 42 80 ) ) 329 ) Pump Stations *) 378 ) 497 320 248 Treatment Plant 21,500 ) 14,500 200 21,500 250 21,500 150 21,500 75 21,500 ) m3/d ) m3/d m3/d m3/d m3/d m3/d )l,066 Storage Reservoirs 2,000 120 2,500 200 1,400 57 1,400 57 1,400 114 2, 00 m3 m3 m3 m3 m3 m ) /a /b Distribution 8,700 232- 307-- 273 273 284 ) Contingencies 80 20% 233 10% 137 10% 141 9% 124 Engineering and /c /c /c Management Services - 222 114 114- 143- 222 Interest during Construction _ _ 108 108 108 108 54 /d Total 22,000 850 14,500 1,709 21,500 1,726- 21,500 1,771 21,500 1,776 21,500 1,816 m3/d m3/d m3/d m3/d m3/d m3/d /a Including extensive secondary distribution additions. lb . May have been intended to include some secondary distribution works. /c Excluding "management services" envisaged under original item. /d Raised in October 1967 to $1,957,000 to allow for extra treatment believed necessary. Annex Table II BIRUNDI: IBujumbura Population and Water Demand, 1964-84, Forecast and Actual 1964 1965 1970 1973 1975 1978 1980 1984 Original WHO Proposal Population 155,000 1/c/d Consumption (m 3/day) 7. losses Production (m3/day) 25,000 " peak 27 Capacity (m3/day) 31,700 Bank Appraisal Population 63,600 67,400 89,300 106,400 119,600 142,500 1/c/d 76 80 115 125 129 135 Consumption (m3/day) 4,825 5,425 10,274 13,137 15,315 19,236 7X losses 44 37 24 20 20 20 Production (m3/day) 8,688 8,630 13,518 16,627 19,142 24,044 % peak Capacity (m3/day) 9,000 24,044 ICIP Feasibility Study Population 87,000 116,500 140,000 155,900 185,000 208,500 1/c/d 69 72 74 76 77 Consumption (m3/day) 8,040 10,080 11,530 14,060 16,000 7, losses 20 20 20 20 20 Production (m3/day) 10,050 12,600 14,410 17,600 20,000 ". peak 40 capaci ty (m3/day) 28,000 Act ral I1opulation ('000) 120-140(?) I/c/d 75-87 Consumption (m3/day) 4,825 5,595 7,564 10,471 7/ losses 44 33 38 46 Production (m3/day) 8,688 8,436 12,164 19,521 2/ peak Capacity (m3 /day) Revised IGIP Projections (1974) Population 120,000 127,000 160,000 I/c/d 105 121 Consumption (m3/dav) 13,330 19,360 ,; losses 25 15 Production (m3/day) 17,800 22,800 "1 peak 40 Capacity (m3/day) 31,920 Annex Table 111 BURUNDI: REGIDESO Bujumbura Operations Expected and Actual Sources and Applications of Funds (in BuF '000) Expected Construction Period Actual Construction Period (1966-69) (1966-73) Expected Actual Expected Actual Sources Gross Internal 51,104 116,255 168,464 243,344 less Debt Service - - 43,220 19,602 51,104 116,255 125,244 223,742 Consumer Deposits - 1,986 - 5,085 IDA Credit 96,250 65,678 96,250 101,278 Reduction of Working Capital 6,736 - - Total 154,000 183,919 221,494 330,105 Applications Project 149,500 100,110 149,500 166,886 Other Investment 4,500 20,192 16,500 48,940 Water n.a. 816 n.a. 3,486 Power n.a. 2,012 n.a. 18,680 General n.a. 17,364 n.a. 26,774 Total Fixed Investment 154,000 120,302 166,000 215,826 Inc. Working Capital - 63,617 55,494 114,279 154,000 183,919 221,494 330,105 BURUNDI: RIIDESO - Rajumbura Operations: Water ('000 BuF, except where otherwise indicated) 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 -----Actual---- ---------------------Appraisal Forecast----- Water produced ('000 M3) 2,351 2,417 2,975 3,171 3,150 3,150 3,150 3,150 4,000 4,934 5,337 5,692 6,069 % losses 19% 29% 24% 4h% 37% 32% 29% 27% 25% 24% 23% 22% 21% Water sold ('000 r3) 1,910 1,708 2,256 1,761 1,980 2,1h0 2,236 2,300 3,000 3,750 4,110 4,440 4,795 average rate (BuF/m3) 6.14 5.70 7.00 7.45 7.50 7.50 7.50 7.50 7.50 9.00 9.00 9.00 9.00 Revenue from sales 11,519 9,459 15,695 13,090 14,850 16,050 16,770 17,250 22,500 33,750 36,990 39,960 43,155 Other revenue - - - - - - - - - - - - Total revenue i1,519 9,79 1,695 I 7- T 1770 17 f 7 W 39,960 43,155 operating costs 6,218 6,099 7,147 7,229 11,500 11,800 12,345 13,295 15,580 17,530 18,575 19,465 20,395 Depreciation 3,66 3,292 4,280 3,831 2,234 2,262 2,262 2,262 2,262 5,625 5,659 5,693 5,730 Provision for bad debts 132 - 3,052 1,994 2,221 1,992 1,661 1,706 1,980 2,947 3,220 3,472 3,742 Interest - - --- - - - 5 294 19 4 777 ,3 Total expenses 10,006 391 1 167E 263 193 Net income (loss) Water 1,513 68 1,216 6 (1,105) (4) 502 (13) 2,678 2,354 4,344 6,553 8,949 Average water cost (BuF/m3) operating cost only 3.26 3.57 3.17 4.11 5.81 5.51 5.52 5.78 5.19 4.67 h.52 4.38 4.25 op. cost + depreciation 5.17 5.50 5.07 6.28 6.94 6.57 6.53 6.76 5.95 6.17 5.90 5.67 5.45 Net income before interest 1,513 68 1,216 46 (1L1a ) 502 U) 2.678 4 4 L 8.949 Water produced ('000 .3) 3,079 3,028 3,014 4,440 6,565 6,169 7,125 % losses 33% 19% 22% 38% 39% 45% 46% Water sold ('000 m3) 2,042 2,078 2,452 2,362 2,261 2,761 4,033 3,633 3,822 average rate (BuF/m3) 7.73 7.90 7.90 7.88 7.91 7.77 7.14 7.92 7.78 Revenue from sales 15,783 16,412 19,369 18,607 17,891 P1,440 28,776 27,876 29,718 Other revenue (net) 894 684 456 717 2 367 1,648 2 468 1 152 1 888 Total revenue 71,T 17, 19,82 19,32 2 ; Operating costs 11,518 11,498 11,583 11,917 12,706 16,596 19,896 21,592 29,974 Depreciation 4,169 4,245 4,774 4,351 4,492 10,512 11,447 12,300 12,255 Provision for bad debts 1,578 1,641 1,936 1,861 1,832 2,187 2,878 2,863 1,486 Interest - - - 3,738 467 528 5 Total expenses 3 ,3 t M 19,030 33,033 3e,* 4 Net income (loss) Water (588) (289) 1,532 1,195 1,228 (9,945) (0,584) (13,010) (17,564) Average water cost (BuF/m3 sold) operating cost only 5.6i 5.53 4-7? 5.05 5.62 6.01 4.93 6.14 7.84 op. cost + depreciation 7.68 7.58 6.67 6.89 7.61 9.82 7.77 9.64 12.05 Net income before interest _(1588) (289) 1,532 1,195 1,228 (6207) 3,977) (7.727) 12,1) BURUNDI: REGIDESO - Bujumbura Operations: Electric Power 1961 1962 1963 1 1965 1966 1967 1968 1969 1970 1971 1972 1973 ----------Actual------------------------r- Electricity purchased (mln kwh) produced (mln kwh) Total (m n k h) =2 - - - 17.0 -2 " % losses 10% 14% 13% 13% 13% 13% 13% 13% 13% Electricity sold (m1n kwh) 13.7 12.4 13.2 12.3 12.8 13.4 14.1 14.8 15.6 16.3 17.2 18.0 18.9 average rate (BuF/kwh) 3.74 3.60 3.60 3.60 3.60 3.60 3.60 3.70 3.70 Revenue from sales 26.7 24.0 24.5 34.1 47.9 48.4 50.8 53.4 56.0 58.8 61.8 66.7 70.0 Other revenue - - - - - - - - - - - - - Operating costs 19.0 18.5 19.3 25.9 31.0 36.3 38.1 40.0 42.0 44.1 46.3 48.7 51.1 Depreciation 6.7 6.0 7.8 7.3 7.0 7.4 7.4 7.4 7.5 7.5 7.6 7.7 7.7 Allowance for bad debts - .1 2.0 7.1 6.0 5.0 5.3 5.0 5.3 5.6 6.0 6.3 Net income (loss) 1.0 0.) .7) 1 2 1_ 73 6) - 9 2.3 7- I.9 ------------- -------------------------------- Actual --------------------------- Electricity purchased (mln kwh) 13.3 14.8 16.6 21.2 23.0 23.4 23.4 produced (mln kwh) 1.0 0.8 0.1 - 0.2 0.3 Total (min kwh) 1T3 15.6 - 21.2 23.6 % losses 13% 10% 20% 19% 19% l5% Electricity sold (mln kwh) 13.1 12.6 13.6 15.0 15.4 16.9 18.7 19.1 20.1 average rate (BuF/kwh) 3.96 3.86 3.81 3.74 3.83 3.83 3.84 3.85 4.15 Revenue from sales 51.8 48.7 51.9 56.1 58.9 64.6 71.4 73.6 80.5 Other revenue 1.8 1.5 2.9 1.3 1.6 5.1 1.4 0.8 2.8 Operating costs 26.1 26.5 28.9 30.5 31.9 42.7 44.5 48.2 50.0 Depreciation 6.9 7.0 7.1 6.8 7.1 6.5 7.7 5.9 64 Allowance for bad dets 5.2 4.9 5. 5.5 6.0 6.5 7.1 7.4 4.0 Net income (loss) 1=. -13.16.8 l.1. 1 17. 12.9 22.9 3.76 3.6 3.6 3.0 360 360 .60 .70 3.7 BURUNDI: Balance Sheet - REGIDESO Bujumbura Operations 1965-73 (in BuF '000) 1965 1966 1967 1968 1969 i970 1971 1972 1973 ASSETS Current Assets Cash and Banks 37,043 28,982 14,806 11,981 36,780 19,992 21,569 17,710 21,959 Treasury Bonds - 15,000 52,500 71,000 41,000 41,000 60,000 40,000 30,000 Accounts Receivable (Sales) 48,143 68,589 77,802 80,688 82,327 118,988 112,683 142,809 170,581 Less: Allowance for Bad Debts 16,84 34,316 36,332 43,220 51,037 59,754 69,772 79,993 85,505 31,294 34,273 41,470 37,468 31,290 59,234 42,911 62,816 85,076 Other Accounts Receivable 13,655 13,840 10,896 12,754 26,547 24,043 27,800 31,923 43,088 Inventories 32,436 34,644 44,015 51,358 49,723 45,356 50,179 68,180 72,768 Others - 1,081 255 143 183 387 491 300 170 Total Current Assets 11,2 12,2 163,92 1 4,70 18,2 190,012 J21295 22,2 253,061 Fixed Assets Water Installations 95,850 96,331 98,212 99,012 99,761 238,025 253,553 276,748 276,727 Less: Depreciation 3820 4244 47,2 5204 5682 6733 7878 9100 O 103.5 Net 57,650 53,886 50,993 46,971 42,940 170,692 174,773 185,668 173,392 Power Installations 164,397 164,878 184,456 185,712 187,156 193,590 195,976 200,367 203,067 Less: Depreciation 7 7 85,545 92,823 1 3 107,549 115,4 1 Net 92,933 86,461 98,911 92,889 86,263 86,041 80,932 77,335 75,480 Work in Progress 8,742 9,100 7,257 53,700 100,475 - - - Total Net Fixed Assets 159,325 149,447 157,161 193,560 229,678 256,733 255,705 263 3 248,872 Total Assets 273Z753 277,267 321,103 378,264 415,201 446,745 468,655 483,932 501,933 LIABILITIES & EQUITY Accounts Payable 4,296 5,371 8,383 9,964 7,566 19,162 10,798 20,202 27,813 Deposits 5,766 6,059 6,605 7,163 7,752 8,404 9,026 9,821 10,851 Project Loan - - 6,134 41,554 65,678 76,743 88,552 101,278 101,278 Other Liabilities 2,219 3,878 5,461 7,266 210 210 238 210 - Equity 6 2 2 312,317 333 995 342,226 352,021 363 4 299 Total Liabilities & Equity 273,753 277,267 321,103 378,264 415,201 446,745 468,655 483,932 501,933 ------------------------------------------------------- BUR UNDI tra-ission mains BUJUMBURA EXISTING AND PROPOSED WATER SYSTEM j9 Kir. River Source 100 m3 Reser-o'r R- Water Li-e cCJ - e - oi LJg R esier vo is 2500.~ 4 \50mý3 JANURY 965 BRD146 u1:3 JANIJARY J965 I BRD0-1462 ru