Document of The World Bank FOR OFFICIAL USE ONLY Report No. 767 7 PROJECT COMPLETION REPORT BURUNDI LOCAL CONSTRUCTION INDUSTRY PILOT PROJECT (LCI) (CREDIT 1230-BU) MARCH 23, 1989 Industry and Energy Operations Division South-Central and Indian Ocean Department Africa Region r This document has a restricted distribution and may be used by recipients only in the performance of I t -' nff -i I dmiti ' It- onnt t- * v lt nt rwi dii; -I 1 'it --it Wnrlui n nk' mt kri-z tic Currency Equivalents Currency Unit - Burundi Franc (FBU) US$1.00 - BuF 125 (June 1987) US$1.00 ' BuF 114 (Annual Average 1986) US$'.00 - BuF 105 (June 1985) GLOSSARY OF ABBREVIATIONS ASECO - Assistance au Secteur de la Construction (local name of project and project unit) BNDE - Banque Nationale de D<veloppement Economique (Development Bank) DETN - Direction des Etudes et Travaux Neuf ENACCI - Entreprise Nationale de Chaux et de Ciment MTPEM - Ministere des Travaux Publics, de l'Energie et des Mines (Ministry of Public Works, Energy and Mines) ONIMAC - Office National d'Importation du Materiel de la Construction ONL - Office Nationale de Logement REGIDESO - (Regie des Eaux) National Water and Electricity Utility SOMIBUROM - Societe Miniere de Burundi et Romanie Fiscal Year January 1 - December 31 FOR OFFICIAL V5 ONLY THE WORLD SANK Washington DC 20433 USA OikO of Owecgtw-C.mqj Opm.atew EvaiiUM March 23, 1989 MEMORANDUM TC THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Pass-Through Audit Report on Burundi - Local Construction Industry Pilot Project (LCI) (Credit 1230-BU) Attached, for information, is a copy of a report entitled 'Pass- Through Audit Report on Burundi - Local Construction Industry Pilot Project (LCI) (Credit 1230-BU)l prepared by the Africa Regional Office. Full evaluation of this project has not been miace by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosee without World Sank authorization. FOR OMCIAL USE ONLY PROJECT COMPLETION REPORT BURUNDI LOCAL CONSTRUCTION INDUSTRY PILOT PROJECT (Credit 1230-BU) TABLE OF CONTE14TS Patae No. Preface ...*.* i Basic Data ...ii Highlights .....................................................ii PART I - INTRODUCTION ......................................... Project Objectives and World Bank Group Role .... .............. 3 PART II - PROJECT IDENTIFICATION AND PREPARATION .... .......... 3 PART III - ECONOMIC AND SECTORAL SETTING ...................... 4 Macroeconomic Setting ......................................... 4 The Construction Industry ..................................... 5 PART IV - CREDIT IMPLEMENTATION ................................ 8 Development and Improvement of Local Construction Materials .... 8 Assistance to Local Contractors ................................ 12 PART V - INSTITUTIONAL PERFORMANCE AND PROJECT MANAGEMENT ...... 14 PART VI - CONCLUSIONS AND LESSONS LEARNED ...................... 16 ANNEXES 1. Summary description of brickoven sub-projects .... ........... 18 2. S-mmary results of feasibility studies ...................... 20 3. Credit allocation of subloan components ..................... 23 4. Comments from the Borrower .................................. 24 EXHIBITS 1. ASECO Information and Promotion Brochure ................... 39 2. Photos displaying aspects of brickoven operations .... ...... 44 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. - i - PROJECT COMPLETION REPORT BURUNDI LOCAL CONSTRUCTION INDUSTRY PILOT PROJECT Credit 1230-BU PREFACE This report presents an evaluation of Credit 1230-BU (SDR 4.6 million, US$5.2 million), which was a pilot operation and combined a broadly structured construction industry project with a DFC credit line, using Banque Nationale de Developpement (BNDE) as the financial intermediary. The Credit was approved in April 1982, signed in June 1982 and declared effective in November 1982. The original terminal date for submission of subloan applications under the DFC component was extended twice. The closirg date was extended once from June 30, 1986 to March 31, 1987. A substantial portion (30%) of the Credit was cancelled in 1985 and the remaining balance was fully disbursed except for SDR 47,000. The Bank's South-Central and Indian Ocean Department has prepared this report on the basis of a draft written by the borrower, information contained in the Project Appraisal Report, a review of Bank files including supervision and progress reports and information gathered during recent m4.ssions to Burundi. A specif4c project completion mission was not carried out. This report presents a factual review of the Borrower's use of credit procoeds and of the project iiapleknentation between 1982 and 1987. In accordance with the revised project performance reporting procedures this report has been read in the Operations Evaluation Depart- ment (OED) but the project was not audited by OED staff. The draft Completion Report was sent to the Borrower for comments. A copy of their response is included as Annex 4. - li - PROJECT COMPLETION REPORT BURUNDI LOCAL CONSTRUCTION INDLUSTRY PILOT PROJECT Credit 1230-BU BASIC DATA As ot Dece b r 31 1907 original Dtabitrt Concel!tdl/ Repaid Uutsttandina (3611) -73DR) - (SW) - - Credit No. 12e-SUI 4,606,0 3,253,068 2,546,912 3J of which SDR1.3 millon were cancelled 2 years prior to the closing dote. Cumulative Diaburs-emets (SDR m IIliIon) (I) Estimated 1.0 2.4 5.7 4.6 - - (ii) Actu.al 0.4 1.0 1.6 2.5 8.1 8. (iii) (it) in S of (1) 46 41 48 64 - - Proiect Data Original Dates Actual Revised First Mention in Files 11/31/77 03/31/78 Negotiations 02/02/82 02/16/82 Board Approval 07/81 04/07/82 Credit Agreement Signed n/a 06/01/82 ;ffectiveness 07/82 11/12/82 Completion of Commitments 12/30/84 12/30/85 Credit Closing 06/30/85 01/31/87 MISSION DATA Identification 11/77 1 2 2 03/03/78 Preparation 07/79 ? ? ? n/a Pre-Appraisal 03/80 ? ? ? n/a Appraisal 09/80 2 4 8 10/22/80 Supervision 05/04/82 ; 1 1 06/10/82 0 11/11/82 2 2 4 12/15/82 09/83 2 1 2 10/26/83 05/84 1 1 1 06/11/84 10/84 1 1 1 01/15/85 05/85 1 2 2 06/10/85 10/85 1 1 1 12/24/85 06/86 2 2 2 08/29/86 11/86 1 1 1 01/14/87 - iii - PROJECT COMPLETION REPORT BURUNDI LaCAL CONSTRUCTION INDUSTRY PILOT PROJECT (Credit 1230-BU) HIGHLIGHTS The Local Construction Industry Pilot Project was one of 6 free- standing IDA credits that addressed a country's construction industry as a whole, compared to other, more subsector specific construction components of sector credits (road maintenance, school construction). Following the recognition that Burundi depended heavily on imported construction materials, had obsolete construction standards and lacked adequate financial and technical support to permit growth of domestic contractors, a project concept was designed to respond to these constraints through a number of mutually reinforcing components. The project (of SDR4.6 million) was appraised in September 1980 and had two major objectives, namely the development and improvement of lecal construction materials and the assistance to local contractors (para. 1.08). To achieve these goals, specific c.mponents were identified, namely the construction of limekilns and brick ovens, feasibility studies to assess the potential for other local construction materials production and technical assistance to the Ministry of Public Works to train staff and to review the country's construction standards. In addition, assisting contractors, a line of credit (SDR1.7 million) was made available through the Development Bank (BNDE) to provide financing for the purchase of equip- ment by contractors. This represented the major component of the project (and hence a former IDF division's responsibility for :he project) and was coupled with additional technical assistance to contractors (in the form of accounting, management, bid preparation and work site organization assistance) and review of general conditions of contract for works. During project execution t-o changes in scope occurred: (i) the limekiln component, under whiLh two ilns were to be constructed by private promoters, was modified. The consti :tion of one kiln was dropped altogether and the second one was bu.st by the project unit due to lack of promoters (paras. 4.1-4.3); (ii) following a substantial slowdown in economic and construction activities, demand for equipment financing did not materialize and only two subprojects with a total financing of SDR137,000 were approved under the credit line component. Therefore, SDR1.3 million (out of the total of SDRI.7 million) of the credit line component were cancelled (paras. 4.14-4.16). - iv - On balance, the project produced mixed results. The brick oven component and the revision of the construction standards were successful as they led to efficient production of high quality bricks for local construc- tion and to a new construction code adapted to local circumstances (paras. 4.5, 4.13). On the other hand, the credit intermadiation concept, which was the centerpiece of the operation and around which the other components evolved, did not produce tangible results (para. 5.3). Also, the construc- tion of the limekiln must be considered a failure as it remains inopera- tive, without a promoter to take it over and without sufficient demand for lime (para. 4.2). B'JRUNDI PROJEC, COMPLETION REPORT LOCAL CONSTRUCTION INDUSTRY PILOT PROJECT Credit 1230-BU I. INTRODUCTION 1.01 The Local Construction Industry Pilot Project was the first systematic effort by the Bank to assist in the development of the local construction induetry in a country of the former Eastern Africa Region. Tr. idea of providing assistance to the sector had been prompted by the Bank's recognition that the construction industry plays a significant role in the development process. Another rationale was that the implementation of Bank-financed projects had in many instances been hindered by weaknesses of the industry. Typically, this occurred on projects involving construction works such as schools or low-cost housing of siles too small to interest the large construction enterprises and which small local firms were neither technically nor financially equipped to carry out. 1.02 Following preparatory work on Rwanda which had signalled interest originally for this pilot operation but subsequently had withdrawn from further pursuing the project, Burundi was selected for an initial effort. Reasons were the strong interest the Government showed toward Bank assistance in the sector and the perceived acute needs of the local industry which were characterized by having been largely foreign- controlled, overstretched and unable to meet the demand. Furthermore, 901 (by value) of the materials used by the formal sector were imported. 1.03 Experimental in nature, the project did not purport to address all the needs of the construction industry but only those which had been identified as priority and for which effective results appeared possible to achieve within the framework of a relatively small operation. On that basis, the following components had been included in the project: (i) development and improvement of the local production of construction materials; (ii) assistance to the Ministry of Public Works, Energy and Mines (MTPEM); and (iii) financial and technical assistance to local enterprises to expand their activities in the building sector where their activity was relatively small. The project did not aim at promcting local enterprises' role in road construction or maintenance which most of the local firms were not prepared to handle at the time. The overall responsibility for the execution of the project was assumed by a Project Unit attached to MTPEM, and staffed with a number of foreign experts and their local counterparts. The financial assistance, which initially amounted to almost half of the credit amount, was channeled through the local development bank, through which a previous IDA DFC-type line of credit (IDA Cr. 731-BU) had already been relent between 1978 and 1983. - 2 - 1.04 The Bank's 'nvolvement in the project began in August 1978 when, at the Governmsent's request, a joint Bank/UNIDO mission visited Burundi to identify a possible operation. Following positive results, a Bank preparation mission visited Burundi in July 1979 and was followed by consultant visits in December 1979 and in May 1980. The final appraisal mission occurred in September 1980, and its results (Apprsi3al Report No. 3343-BU) led to the approval of Credit 1230-BU of US$5.2 million, which included financing of subloans for limekiln and brickoven constructions, and for *quipment investments of construction companies. Furthermore, the project covered consultant services, feasibility studies and project unit operating costs. The Credit was approved on April 13, 1982; it was signed on June 1, 1982 and became effective on November 12. 1982. 1.05 Commitments and disbursements were considerably slower than anticipated mostly due to lack of demand for equipment financing as a result of an economic slowdown, e anced soon after credit effectiveness which caused a virtual halt of local construction companies' activities. As a result, 30% of the credit amount was cancelled in January 1986 as no other use of the funds was envisageable within the scope of the project unless it were to become an oversized technical assistance operation. As of December 31, 1983, the original target date for submission of subloan proposals, only one subproject (for less than 5% of the amount allocated under the subloan component) had been committed (and was subsequently cancelled). The terminal date for subproject submission was extended twice to December 31, 1984 and December 31, 1985. The original closing date was June 30, 1986 which was extended once to March 31, 1987. Final disbursements under the credit were made on February 2, 1988, resulting in the cancellation of SDR 46,911.68. 1.06 The project component allocations as initially appraised and as revised during the life of the project, were as follows: Initial 1st Revision 2nd Revision Allocation (December 86) (Aug. 1986) (SDR) (SDR) (SDR) 1. Subloans for 2t0,000 280,00 140,OOO/ limkilIn construction 2. Subloans to 1,000 310,000 318,700 brickmakers 8. Subloans to 1,700,0 1,500,000 137,0942/ construction companies 4. Consultant services 1,36O,000 1,360,00 2,048,412 S. Feasibility Studies 480,000 480,000 226,006 6. Operating Cost of 200,000 200,000 366,006 Project Unit 7. PPF refund 270,666 270,0"C 62,794 S. Unallocated 2ffW.o 200 ff 17000 TOTAL 4,e,00,MO 4,800,000 3,300, 1 Originally planned an subloan, but carried out under force account due to lack of promotorn. Z Cancellation of SDR 1.3 million included. - 3 - 1.07 The proceeds under categories (2) and (3) of the Credit were onlent--under a DFC-type arrangement--by the Government to the development bank (Banque Nationale de Developpement Economique, BNDE) at 8Z and the rate to the final borrower was 12% (both rates remained unchanged during the life of the project). The subloan portion of the Credit was to be repaid based on a flexible amortization schedule conforming to the aggregate amortization of BNDE's IDA subloans. The Government assumed the foreign exchange risk without a charge to the subborrower, and, in addition assumed 751 of the credit risk on all subloans. Other features of the credit's subloan component included an individual free limit of US$400,000 per subproject to ensure financing of a larger number of companies, and subproject financing was limited to 752 of total cost. Apart from the usual signing of a Subsidiary Loan Agreement between the Government and BNDE, there were two other conditions of effectiveness related to Government's sharing of operating costs. Project Objectives and World Bank Group Role 1.08 The objectives of the project were derived from the main constraints of the local construction industry, namely: (i) the 1'.avy dependence on imported construction materials due to insufficient production of local materials; (ii) the obsolete construction standards; (iii) the lack of adequate financial and technical support to permit the growth of domestic contractors; and (iv) the absence of qualified local manpower. Resulting objectives were: at the sectcr level, to foster the production of local materials, to reduce the high dependence on imports and reduce construction costs by adapting the building code, which was based or European standards, to local conditions, and to strengthen the capacities of MTPEM to supervise the implementation of construction projects; at the enterprise lev'el, to assist in the development of small- and medium-sized contractors and of producers of local materials through financial and technical assistance. 1.09 Additional funding for completing a small portion of the technical assistance component was provided after the project's closing date undar the Project Preparation Facility (PPF) of IDA's Second Urban Project. Also, the recent Small Enterprise Apex project (approved on March 29, 1988) does not exclude financing of subloans to construction enterprises slould additional demand arise from that subsector. II. PROJECT IDENTIFICATION AND PREPARATION 2.01 Project identification took place in November of 1977 when a reconnaissance mission reported an overwhelmingly favorable response from the authorities. The main sector constraints that were initially identified and later on refined centered around (i) the lack of materials and supplies, (ii) the difficulty, unreliability and irregularity in transporting these materials and supplies when available, (iii) the lack of financing, particularly for the small contractors which seemed to have serious difficulty in obtaining cred It for equipment purchase and working capital, and (iv) the 1lack of capable, technically trained personnel both at management anc& skilled worker levels. In August 1978, a follow-up UNIDO/World Bank mission visited Burundi at Government's request, and identified a project to assist local construction companies. Supported by a detailed report (issued March 1978) on the situation of the construction - 4 - sector, the mission confirmed potential for subcontracting in public works, housing construction in rural areas and good possibility of G-overnment allocating specific parts of public work projects to small local contractors. Moreover, need for developing products and services in construction such as bricks, woodworks and equipment repairs was identified. 1/ 2.02 Following the above findings, a Bank preparation mission visited Burundi one year later and its full report, dated September 18, 1979, reviewd ir detail the UNIDO recommendations, confirmed them and outlined the possibilities of IDA assistance. In December 1979 and again in May 1980, consultants went to Burundi to carry out additional preparatory work. Their findings and recommendations provided substantial background for and led to the appraisal mission of IDA's first pilot project in the region to assist the local construction enterprise sector. 2.03 One and a half years elapsed between appraisal and board presentation and six months between board approval and effectiveness. In proportion to the size of the project, the preparation was intensive both in terms of overall staff time and preparatory missions. III, ECONOMIC AND SECTORAL SETTING Macroeconomic Setting 3.01 Background. With an average GNP per capita of about US$200 in 1980 (US$240 in 1987, when the project closed), Burundi remains one of the poorest countries in the world. In the time span of project execution, its population grew from about four million to five million, of which only 2.3% are wage earners, the remainder depending mainly on subsistence agriculture. The country is landlocked, densely populated and endowed with few natural resources. Agriculture accounts foL approximately 50 percent of GDP and generates income for about 85 percent of the population. Coffee being the main export, it generates four-fifths of total export earnings. Burundi is quite vulnerable to the transport conditions in neighbouring countries and even though progress has been made in the first half of the eighties to improve the transport infrastructure and to simplify transit formalities, transportation costs remain high resulting in high cost of construction materials. 3.02 Developments. During the years preceeding project preparation, Burundi experienced a phase of rapid economic growth (6.5 percent per year), which was in large part the result of the coffee price boom in the mid-1970s and the subsequent accumulation of public savings, which, in turn, allowed for a considerable expansion of public investment and growth in related 1/ One noteworthy observation is that none of the six construction enterprises visited and reviewed during project preparation sur,'ved during the implementation period. Also it is interesting to note that an internal IDA review of the UNIDO report already alluded to the demand issue in questioning whether the volume of construction work should solely be based on a five-year Government plan (which then proposed a US$100 million equivalent volume p.a.) when the actual volume p.a. at the time amounted to only US$30 million equivalent. sectors. During this period, the volume of investment grew by 75 percent, and the investment rate rose from 9 percent of GDP in 1976 to more than 12 percent of GDP in 1979. This effort, which was led and implemented by the public sector, had important repercussions on the pattern of demand and mobilization of domestic and external financial resources. In hindsight, the project can be seen as having partly responded to this demand pattern given the rapid growth of public sector expenditures between 1978 and 1979. Public consumption and investmenL expanded three times as much as that of the private sector during the same period. Government expenditure more than doubled and increased from about 16 percent of GDP in 1976 to 21 percent in 1979. Capital expenditures, particularly for infrastructure and for the financing of public enterprises, contributed most to this growth which can be identified as having contributed to the demand under the project as a number of these public investments had construction related components attacned to them. 3.03 At the turn of the decade, however, right at the time of appraisal, the government failed to adjust to the changing environment, characterised by a deterioration of the terms of trade and increasing pressure over the balance of payments. Consequently, an expansionary policy led to serious financial desequilibria, together with administrative controls and, more importantly, to severe economic disto-tions coupled with a substantial slow- down in economic growth. Owing to a significantly lower coffee crop, a fall in the volume of investment and a showdown in Government spending, GDP growth decelerated (1.4 percent per year between 1982, when the project became effective and 1985, when the bulk of project funds was either comnitted or cancelled). A first IDA structural adjustment credit s"iressed some of these issues but by the time it was approved in 1986, the LCU project was completed. It should be remarked, however, that the project's President's Report already cautioned somewhat the otherwise upbeat demand projections when it noted that because of the expected financial constraints, it was highly unlikely that during the 1980-1984 period the volume of investment can be more than half of that envisaged in the country's economic plan. Another caution, which however was not expressed in preparatory documents should have been that Burundi's level of urbanization at the time of appraisal was still low and thus effected the demand for housing. The Construction Industry 3.04 Basic Characteristics. At the outset of the project, towards the end of the seventies, Burundi's construction industry was small, underequipped, concentrated, and highly dependent on foreign skills. The rate of urbanization in Burundi has been kept very low, as compared to other African countries. This characteristic did not favor the development of a domestic construction capacity (either formal or informal). In 1979, it contributed to about 4% of GDP, 30Z of industrial output, and employed about 10,000 people, representing about 8Z of total wage employment. Construction investments accounted for a fairly stable proportion of between 502 to 60% of total investments in the 2-3 years prior to project appraisal. Most construction enterprises were working at full capacity. The sector was marked by a lack of local materials, transport difficulties and lack of skilled manpower and financing. 3.05 At the time of project preparation, the construction sector output was based on the following 3 segments: 13Z through Government force account - 6 - mainly for maintenance work, 70% through contracts to foreign-controlled firms, basically for major public works and the remaining 17% through a combination of some twenty privately-owned Burundian firms (few of which survived the economic downturn during the early eighties), religious missions and self-employed artisans. This distribution changed over time in favor of force account activities, as this relative share increase following the completion of some major foreign contracts (new airport etc.) 3.06 Construction Materials. Scarcity of local materials had long been a constraint to the development of Burundi's construction industry. Although bricks, tiles, stone, sand, gravel, wood, bamboo and limited quantities of quick lime were and still are produced locally, the country still imports about 90% of its construction materials. In 1979, the CIF value of these materials represented about US$13 million. Until to date the market for these products is characterized by irregular supply, sharp variations in prices, and a general shortage of such basic materials as cement, galvanized sheets and reinforcing rods, which are the three main materials used in construction. Irregularity in supply still leads to high construction costs and renders it difficult to complete contracts on schedule. 3.07 To alleviate the problem caused by the variations in price levels and import volumes and in an effort to improve the country's supply of construction materials, the Government created in 1978 the National Office for Import and Marketing of Construction Materials and Domestic Equipme.'t (ONIMAC), a public company responsible for regulating the import of construction materials and stabilizing the price of these goods. However, during project execution, the Project Unit (PU) often experienced difficulties in receiving agreed levels of quantity from ONIMAC (of cement, e.g.) in a timely manner. 3.08 In addition, the Government was trying to expand the local production of lime and of bricks. Burundi has large deposits of limestone which could have been used in manufacturing possolanic cement. Societe Miniere de Burundi et Romanie (SOMIBUROM), the public mining company created in 1978 for the development of Burundi's mineral resources was entrusted with the geological exploration and the development of the country's major lime deposits. However, the company went bankrupt a few years later, i.e. soon after the project became effective (para. 4.3). Also, a program to build three new limekilns was implemented with assistance from the United Nations Industrial Development Organization (UNIDO), but none of these limekilns ever operated. The Government had also established a new public company to rehabilitate and operate a cement plant and bring its annual production capacity to 20,000 tons of pozzolanic cement. The plant, however, is closed for a number of years. Bricks were and still are produced in most regions of Burundi by private manufacturers and by the National Housing Office (ONL). In 1980, 25 million bricks were produced. They were generally of uneven sizes and forms (see Exhibit 2) and their reas.stance to compression and erosion was limited. It is estimated that 30 to 50% of the production is lost because of poor quality. The proje~ct therefore envisaged to improve significantly the quality of bricks, in particular by using some simple e.aipment for the preparation of clay anA by building shelters for storing bricks before they are baked. This was to lead to a larger utilizatioin of bricks in construction and to substitute for abo.ut 2,000 tons of imported cement a year. Indeed, the Kamenge brick oven - 7 - constructed under the project does produce 2 million (less than 10% of the national production volume) of good quality bricks (para 4.5). 3.09 Manpower. The lack of qualified manpower was a constraint to growth and one of the reasons of foreign predominance in the construction industry. In addition to a few local civil engineers and architects trained abroad, the yearly average output of graduates from local technical schools prior to project implementation was limited to 13 surveyors and general construction teclnicians, 12 carpenters and masons, 10 electrici_ns, and about 50 semi-skilled carpenters and masons. Contractors have been hiring some operators a'.nd foremen in Zaize and other neighboring countries, but shortages of specialized workers appeared difficult to overcome. The project contributed to training additional staff 'but in view of the subsequent downturn of construction industry activities, many skills were lost as trained counterparts switched over to other sectors of the economy. According to appraisal estimates, the total annual output of specialized workers represented, prior to the project, only 45% of the projected manpower requirements of the construction sector. 3.10 The Public Sector. The Ministry of Public Works, Energy and Mines (MTPEM) is the key public agency in the construction sector and also the most active one. In its public works section, it is composed of a Roads Directorate, a Building Directorate and a Directorate for the planning of infrastructure and housing programs. The Roads Directorate executes all maintenance and small new works (below BuF 100 million) for roads with its own construction staff which includes more than 3,000 workers, and it supervises preparation and execution of larger road projects by contractors. The Building Directorate establishes and enforces policies regarding the construction industry, assists in the preparation and supervision of major Government building works and is responsible for the maintenance of all 1,500 Government buildings. The capability of the Buildings Directorate was reinforced during the project through technical assistance (para. 4.12) to the directorate's Studies and New Works Department (DETN), which is in charge of policies regulating contract conditions and construction standards. 3.11 The construction-related activities of other Ministries (mostly Agriculture, Army and Education) is much smaller. The Implementation Unit of the First Education Project has undertaken a large construction program for primary schools, which unfortunately continued when construction firms assisted under the LCI project were ready to substitute for these force account works (para. 4.19). Construction components of other IDA projects were also relying on force account (urban, agriculture) and, unfortunately, continue to do so. The remaining public sector institutions involved in construction work are REGIDESO, the parastatal responsible for water and power development, which executes all maintenance and most new works in the water sector and all maintenance and about 25% of new works in the power sector and the National Housing Agency (ONL) which was created in 1974 to complement the limited capacity of the private sector in the construction of public housing programs. It suffered from poor financial management, and is now to be closed down. 3.12 Construction Enterprises. Burundi's private construction sector was and still is dominated by foreign contractors, including subsidiaries or branches of large foreign companies and a handful of smaller enterprises owned and managed by expatriates. In general, they have adequate equipment - 8 - and qualified staff, and their management positions and about a third of professional positions, are held by expatriates. They benefit from the parent companies' technical and commercial assistance, particularly for the direct importation of raw materials in periods of shortages. Similar to neighbouring Rwanda, Burundi has, as of late, experienced Chinese construction firms that are capturing an increasing market share. 3.13 Regarding domestic firms, there were about twenty enterprises headed by a Burundian owner-manager at the beginning of the project. Three were medium-scale enterprises and the others small-scale. They formed the target group for both financial and technical assistance under the credit and worked mostly on Government buildings, public housing projects and, in the case of small-scale enterprises, on some high income private housing. They were generally underequipped and labour-intensive and had a weak liquidity situation. The appraisal report had found that the main constraints that these local enterprises faced in their development were the lack of finance and shortages of raw materials. There was virtually no subcontracting of work to these firms by the big foreign contractors. In the end, few of them survived (para. 4.16) owing to the dramatic downturn of the economy. Finally, there is still a substantial number of self-employed artisans in both urban and rural areas. While lacking proper tools and displaying uneven competence, they are engaged mainly in tha construction of simple, traditional family homes. 3.14 Building Regulations. Government procurement procedures remain to be of particular importance since public contracts represent about 80% of total value of construction work. Prior to the project, general regulations on construction activities were based on pre-independence Belgian legislation. They were complex, and neither systematically publicized nor enforced. Construction standards had become obsolete and did not take into account local construction techniques and materials. The overall procurement procedure of the Government is generally satisfactory but special contract conditions were often scanty or incomplete, due to inadequate technical specifications. The absence of classification of contractors led to inefficiencies in contractor selection and inadequate contract supervision. Most of these deficiencies, however, were successfully addressed during project execution. (para. 4.13). IV. CREDIT IMPLEMENTAT1ON Development and Improvement of Local Construction Materials 4.1 Construction of Limekilns. Within the overall attempt to develop construction materials, and in addition to improving brick production (para. 4.5) and preparing feasibility studies (para. 4.9) with a view to assess the potential fcr local construction material production, the limekiln component was to finance the construction of two limekilns. At the time of appraisal, it was expected that the kiln would be owned and operated by private contractors who were to receive financial, technical and managerial assistance under the project. The project unit was to select promotors on the basis of their technical and managerial capabilities and on their readiness and ability to contribute, in the form of equity, 25% of the investment cost, the balance to be provided under the credit's financial component managed by BNDE (para 5.3). The project unit was furthermore to - 9 - promote the use of lime in the construction sector as this was not accepted practice among contractors and builders when the project was launched. 4.2 The limekiln component can clearly be labeled as a failure. For two years after effectiveness the project unit was unable to identify a promotor that remained seriously interested beyond initial curiosity, even though a broad press and radio advertisement campaign was mounted. As a result, IDA was asked, and agreed (amendment signed on Oct. 4, 1984) to a government request to modify the subcomponent objective. Instead of expecting private promotors to comit themselves from the very beginning, the project unit itself was to construct one kiln under force account and only upon its completion and successful demonstration transfer the kiln to private promotors. Kiln construction began in May 85 and was completed in June 1986 but as of the preparation of this report, no promotor has yet expressed serious interest in purchasing it, despite the fact that its proper functioning has been demonstrated. 4.3 Reasons for promotor reluctance to build--under the initial concept--and/or to purchase (a) kiln(s) on a turn key basis--under the revised concept--can be traced to the substantial number of unresolved issues that arose in connection with the kiln construction. i) the fact that SOMIBURM (para. 3.08) went bankrupt: under a subsidiary credit agreement, this public joint venture between Burundi and Romania was supposed to develop the extraction of limestone in quarries. Following the initial absence of SOMIBURM- provided raw material supply, the project unit selected later on, under the force account concept, a private firm to blast a quarry in the Ruhanga area 65 kms north of Bujumbura and chosen for its low magnesium content and nearby available space for tree-planting to supply fuelwood for the kiln; ii) absence, supply delay and volume miscalculation of fuelwood: In view of Burundi's deforestation problems and the increasing shortage of fuel wood in the country, the project was to include the planting of eucalyptus trees in the vicinity of the selected sites to supply fuelwood to the two originally proposed kilns. The negotiated agreement was for the country's Water and Forest Administration to plant 37.5 ha of eucalyptus per annum and to begin planting in 1982 based on an understanding that the production was to be reserved for the limekiln and that a price was to be charged that would in full reflect the plantation investment and maintenance cost. Although the appraisal report stated that managerial and technical capabilities of the Water and Forest Administration were strained at the time, it was nonetheless concluded that in view of the relatively small size of the plantation "there should be no problem". The Water and Forest Administration never planted a single tree as it lacked financial resources and could not obtain a loan. Following these shortcomings, the project unit took over the planting of trees as it was displaying substantial interest in making the limekiln work. However, it had been estimated orignally that 1,900m3 of wood p.a. were needed for 'ropGrly operating one kiln, based on a mixed fuel consumption of 50% wood and 50% peat. Following unsatisfactory test results, peat had to be abandoned as a heating source. Therefore and in view of lower than expected plantation yields of - 10 - the eucalyptus, 184 ha had to be planted instead of the original 37.5 ha. However, 119 ha were destroyed during a brushfire on July 14, 1987. iii) Marketing uncertainties: Promotion of lime could not be started until October 1987, when the project unit was able to begin marketing the 196 tons of lime produced at Ruhanga between September 17 and October 15, 1986. (Production thereafter was stopped because of limited storage capacity). Following good test results of the lime for various applications in construction, an ordonnance was signed by MTPEM making use of lime compulsory but this was done only about a year later, on August 8, 1987. At the same time, tl.e project unit had difficulties in product marketing as it was not a comercial company itself but an integral part of MTPEM, albeit with some degree of autonomy. 4.4 The kiln was operating for a second time, following some expansion work, which was essentially storage oriented and another 373 tons were produced. However, difficulties in procuring fuel wood and limitations of yet more storage space forced the final halt the operation on June 30, 1987. (Total direct construction cost was FBU 22,058,642, equivalent to about SDR 170,000, which exceeded the allocated total by about 15%). This figure includes neither project unit overhead, nor a substantial amount of man/hours by the experts. 4.5 Construction of Brickovens. The positive outcome of the brickoven component counterbalances the sobering results of the lime kiln related activities. The component clearly achieved its objectives in as much as financial, technical and managerial assistance was provided to brick manufacturers for improvii.g the quality of bricks produced in Burundi. The financial assistance disbursed was almost double (SDR 191,000) the amount originally allocated (SDR 110,000), and the three benefitting private promotors produce good quality bricks and are responding to a consistent level of demand. 4.6 Despite the fact that brick utilization in construction has a long tradition in Burundi, the implementation of the component was not without complications: i) A two-step marketing effort had to be launched (project unit/promotor and promotor/customers) in order to create an awareness that high quality bricks, albeit more expensive per unit compared to traditional artisan bricks, are less expensive per qm2 construction as their larger size and regular dimensions provide for consistent reduction of cement mortar consumption (reduction by 2/3 of the quantity consumed for cement, versus hand-made bricks). Moreover, the quality of finished walls renders external plastering unnecessary (exhibit 2). ii) An agreement on quarry access near Bujumbura was negotiated with but never fulfilled by the government. This caused unforeseen transportation cost, hauling clay from another distant quarry as the first plant had been located next to the orignally proposed quarry, to which it was denied access despite the fact that the promotor had fulfilled all conditions to gain this access. - 11 - iii) The drop out ratio of potential promotors was high. Contrary to the limekiln case, a total of 20 promotors presented themselves to the project unit. However, nearly all possessed neither a recognized clay deposit nor land on which to luild their plant. Therefore, applications from counties ("commune") were also studied and finally 8 investment applications were proposed to BNDE for financiag. However, the two communes abandor.ed their projects later on (after IDA had approved them) following changes of governors. Another promotor had legal difficulties whereas two others were unable to assemble the necessary 25X equity participation. 4.7 Three brickovens were finally constructed. (Annex 1 and Exhibit 2) The bricks produced (average annual production 2,200 t per plant) are of good quality due to regulated cooking and are accurately gauged as ccmpared to traditional hand-made bricks that are irtegularly dimensioned and cooked. 4.8 Apart from the assistance to promotors from the project unit in winning supply contracts (e.g. the EEC building in Bujumbura), the good visual appearance of the ready product helped the promotion of bricks. As a result, the first p:omotor (at the Kamenge quarry near Bujumbura) decided soon after the first credit to double his production capacity by building a second oven. It is interesting to note though, that to fund this extension, the promotor did not apply for another subloan through BNDE because of --according to him--time-consuming formalities and excessive collateral requirements. Consequently, he borrowed from his commercial bank, albeit at higher rates. The initial and prompt suzcess of this operation drew, no doubt, the attention of the other promotors. A synopsis of all tnree brickoven plants is described in Annex 1. 4.9 Feasibility St -viies on local material production. To further study the potential of local production of construction materials, a SDR 480,000 component was set aside, leaving to the project unit the preparation of priority topics. A study program was set up, endorsed by the local authorities and approved by IDA. Tender invitation took place on May 20, 1984 and a contract was awarded on June 25, 1985 to a Canadian firm for a total of SDR 176,000. Five studies were carried out relating to the manufacturing potential of pozzolamic cement, sanitary articles, sawn and polished stone slabs, earthenware tiles and the use of bamboo in construction. 4.10 The studies conclude that in all five areas, manufacturing units could not be estab.ished in view of unsatisfactory profitability as the unit cost would be too high in view of the limited absorptive market capacity. (Summary results of the studies are compiled in Annex 2.) 4.11 Technical Assistance to the Ministry of Public Works (MTPEM). The technical assistance to MTPEM was part of the consultant's program under the project and consisted of two major parts: (i) assistance--basically in the form of lectures--to the Studies and New Projects Directorate (Direction des Etudes et Travaux Neufs--DETN) of MTPEM to upgrade knowledge of staff which are responsible for preparing tendering documentation, evaluating bids and supervising contract execution (except for works executed by force account); and (ii) review of construction standards. - 12 - 4.12 The lectures given to DETN staZf at the project unit's offices (located across town from MTPEM) had irregular attendence (50%) due to lack of att-nlees' interest and MTPEM managers' support. This outcome is unfortunate as the strengtheniag of DETN's staff would have been of particular importance because the Government has a determinant impact on the construction sector (it purchases 902 of all construction contract value) and because contract supervising technicians often have a decisive role tn directing the work and staff of construction enterprises. 4.13 Concerning the review of construction standards, this activity, following an initially shaky start, successfully implemented. The term "construction standards", translated into French, is too vague and subject to interpretation as either technical specifications or as genfral conditions of contract. The project unit chose the second meaning and reiiewed the general conditions of contract and the procurement regulations. A first specialist lost an entire year while attempting--without success-to substitute French regulations for the then existing ones, issued a decade -go by the Belgian Colonial administration and not very adapted to African construction needs and circumstances. A second specialist completed the updating of the Burundi regulations without attempting to modify all those articles that were still applicable and usable in their original version. This more efficient updating was easier to implement and undoubtedly facilitates, now that it is completed (and officially issued, printed and widely circulated) the construction activity, mainly by stipulating fair rules to indemnify the contractor in case of delayed payments for the works performed. Assistance to Local Contractors 4.14 The assistance to local contractors was divided into financial and technical assistance, the former having been the most important component of the entire project with an allocation of SDR1.7 million, reflecting the nature of a DFC-type oriented operation. The technical assistance was supposed to be an accompanying feature, helping primarily those contractors who would borrow to finance their equipment replacement and expansion under the line of credit channeled through BNDE. At the time of the appraisal it was estimated that 10 to 15 local enterprises would be able to meet specific criteria, which were made a precondition to qualify for subloans. These were not overly stringent and comprised (i) a limit on net fixed assets of BuF 100 million (in 1980 prices); (ii) a potential to develop into ernerprises with qualified management and staff, and suitable accounting systems; (iii) the inability to obtain term financing through normal commercial bank channels; and (iv) having been established and operated for at least 3 years in Burundi. 4.15 However, in the end, less than 8% (SDR135,000) of the originally allocated amount were disbursed for 2 contractors (Annex 3) The central reason why so few enterprises showed interest in equipping themselves was that they did not hove sufficient contracts that would justify the equipment purchase and therefore provide enough cash flow to service the debt. In order to avoid a disproportionate use of funds under the credit for categories other than the onlending one, which was the prime purpose of the project, IDA decided to cancel the non-utilized portion of financial assistance to contractors, after the submission date for subloan applications (extended once, from December 31, 1985 to December 31, 1986) had expired and after it became evident that no additional demand existed. The cancellation of SDR1.3 million, carried out in January 1986, met initial resistance from - 13 - MTPEM but was justified to avoid further reallocation of this component to the overall technical assistance compone.nt. Otherwise the entire project might have turned into an oversized To credit. 4.16 According to the project unit, there were numerous reasons that contributed to the tremendous reduction in the number of enterprises qualifying as potential subborrowers under the credit, and a short summary of that development is given below. "The first two Burundian construction enterprises were established in 1975. At that time and during the following years, labor supply was abandoned and contract amounts were quite substantial in size. Given a reasonable profit margin on those contracts, it was very attractive to become a building contractor. Moreover, the financial terms (payment of an advance on placement of order) enabled a contractor to open a worksite without needing funds of his own. As a result, many enterprises were set up by persons who did not necessarily possess a vocation as building contractor. Most of these persons who had suddenly become 'contractors", knew nothing about this profession, which is not a simple commercial activity consisting of buying and selling products but a complex processing industry which requires extensive technical knowledge. Lacking this competence, and without any idea of management or organization, these enterprises were doomed to fail, except of course those that had managed to surround themselves with competent assistance. Intoxicated by the amount of money they took in, contractors often used it to buy personal effects that they regarded as success symbols. Since they were quite unable to keep track of their financial situation, the day came--alas, rather late--when they realized that they no longer possessed the necessary resources to complete their ongoing projects. They then resorted to the system of underbidding on new contracts which were beyond their competence, in order to complete previous projects. This system, repeated regularly, aggravated the situation, to the point the contractors were unable to honor commitments and faced bankruptcy as a consequence. As a result, close to 20 enterprises disappeared between 1975 and 1982." 4.17 The technical assistance to local contractors originally was to be made available to enterprises upon request. To carry out this task, two experts, a construction and a management specialist were posted for 2 years, along with 2 local counterparts. Lacking sufficient financing requests from individual contractors, the project unit decided to offer group courses, and also invited firms that did not borrow under the project's credit line. The first phase of these courses, organized to improve the contractors' knowledge of management, procurement and work planning did not gain the confidence of the construction companies. In a secord phase, courses were moved from the classroom to the construction sites, resulting in a much higher degree of acceptance and prome'ter openness. To reinforce the on-site assistance, the project unit published three handbooks on management, cost control and site organization which the few firms that were still in existence at the time of - 14 - publication (about 7 altogether, grouped in the association of construction sector enterprises) found useful. Overall, this assistance led to works that resulted in improved customer satisfaction and timely executed contracts. 4.18 In addition to the above, more company specific oriented higher level program, the project unit launched training courses for construction workers. Responding to the contractors' criticism that the established national training center (Centres de Formation Professionnelle) courses wele overly theoretical, a small-scale, isolated 14-month program for about 20 carpenters and masons was organized in the project unit's backyard. The impact of this very low key concept, however, was dismal, as it was dismantled following the departure of the supervising expert and as the trainees did not find jobs in the ever declining construction industry. 4.19 A final effort was made, during the last phase of the project, to render assistance geared towards helping individual construction companies to form temporary Joint-ventures among themselves when responding to an invitation for contract tender. This initiative partly resulted from the reluctance of an IDA-financed education sector project bureau (BPE, Bureau Projet Education) to entrust school construction works to local enterprises because of unfortunate earlier experiences. Following intervent:cn by IDA staff during LCI supervision missions, the issue was resolved and contractors supported through the LCI project were awarded several contracts and received reference certificates for the works carried out. V. INSTITUTIONAL FERFORMANCE AND PROJECT MANAGEMENT 5.1 The overall responsibility for the execution of the project and the coordination of the various actions included was assumed by a project unit within MTPEM. As the project dealt with a number of private entrepreneurs, ministerial departments and parastatals (including BNDE), this unit was directly attached to the minister's office to ensure autonomy and efficiency. 5.2 The overall responsibility for project execution was with MTPEM, which displayed a consistent level of interest in the project, albeit with some administrative delays in approving progress reports, confirming credit amendments (limekiln, para. 4.2) and growing technical assistance contract extensions for the consultant in the project unit. 5.3 Financial intermediation did not develop into the predominant activity that it was supposed to under the project concept. As shown in Annex 3, total 'subproject' related disbursements reached SDR327,000, just slightly above 10% of the originally allocated amounts for this feature, combining limekiln, brickoven, and equipment financing for contractors (as per summary table in para. 1.06). As a result, the role of BNDE was a limited one under the project. Furthermore, BNDE was not--as agreed under the project concept--to take the lead in helping local contractors, lime- and brick manufacturers prepare project appraisals, and concentrated on general administration of the subprojects, the repayment of which presents some arrears problems. (Most subprojects are in arrears except for one contractor who has prepaid his loan in full). The former East Africa IDF division had issued a PCR on BNDE on January 30, 1987 (IDA's first line of credit to BNDE, Credit-731-BU) and has recently reviewed BNDE's performance as part of the SAR report 7025-BU of February 25, 1988 for the Small Enterprise Apex Project, which was approved on March 29, 1988. - 15 - Therefore, BNDE's performance is not commented upon in the context of this PCR, except for stating that, as a prudent development bank, it had to take a conservative view when considering project approval, but showed a sufficient degree of flexibility and cooperation. BNDE's prudent approach was right as the performance of subprojects did not match the targets set. 5.4 The most important role under the project was assigned to the prolect unit itself, called ASECO in the local context. The staffing of the unit included a directcr (foreign), a co-director (local), a lime-, a brick-, a :onstruction and a management expert plus their counterparts, as well as a training specialist, a construction engineer and a constructio. technician, the latter two also with counterparts, bringing total staffing to 18 plus auxiliary support. 5.5 Pursuant to Section 3.03 of the Credit Agreement, the Government of Burundi signed an agreement with a foreign consultant company (under ICB) on April 20, 1982, to provide specialists :equired for the staffing of the project unit and the execution of the project for a period of 22 months. The contract was extended for a period of 12 months beginning June 1, 1984. Following initial delays in project execution, a second contract, with the same company, wa.. approved on May 12, 1986 for a period of one year beginning August 1, 1985. It was again extended on February 26, 1987 to prolong the consultant's mission to December 31, 1986 (sic) to run from July 1, 1986. 5.6 An amount of US$240,000 as part of the advance for preparation of the second Urban Development Project was granted by IDA to extend the mission of a number of experts to june 30, 1987. An additional eight man-months was agreed on under the Technical Assistance Credit 1456-BU to the Planning Ministry for three project experts to complete their work on the limekiln, the brickovens and assistance to enterprises to help finalizing some remaining works. Recruitment of the Burundian counterparts posed a number of difficulties, as it was Impossible to assure them any type of employment after the end of the project (only the local co-director was a civil servant). 5.7 Following the contract extensions and the operation of the project unit for twice as long as expected, the total ftnancing share for this project activity rose from an initial 30% to 68.. (rhis increa..e, however, is also due to the reduction of the total credit amount following the substantial cancellation of SDRI.3 million). Overall, the performance of the consultant can be rated as acceptable. (See table below) The main achievements of the consultancy services were the revisiorn of the construction code and the assistance to brickoven promoters and contracters to organize their construction ectivities and prepare bids. Poor Fair Good Purchase of equipment by contractors x Construction of lime kiln x Brickovens x Training for Public Works ministry agents x Training for domestic contractors x Training for construction workers x Revision of construction code x - 16 - 5.8 Regarding proiect execution matters, and following the above mentioned extensions, the project lasted twice as long as originally planned. The reporting, provided by the project unit, and--overlapping partly with IDA Credit 733-BU--by BNDE, was adequate albeit delayed at times. Procurement and disbursement provisions were adhered to and did not undergo changes. Auditing of the project unit was carried out by a foreign auditor and resulted in clean audits consistently throughout the project. VI. CONCLUSIONS AND LESSONS LEARNED 6.1 Although the technical and financial assistance provided under the project provided some useful benefits, the project has, on an overall basis, not produced the expected results. At the same time, it should be recognized, however, that the project was of a pilot nature and that it was noted from the beginning to be complex, multi-faceted and therefore difficult to implement. Furthermore, the SAR did not fail to allude to the inherent risks of the construction sector and in particular to the uncertainties about the reaction of beneficiaries to the proposed assistance. Since the majority of construction industry output is capital investment, cemand tends to fluctuate considerably more than for most other industries. The Burundi contractors, thus, had to face the risk of sharp fluctuations in the volume of work and, hence, in the amount of equipment needed. This was due to the beginning of economic slow-down shortly after credit effectiveness but also to the fact that Burundi's level of urbanization was very low, a fact that was probably not suffic ently recognized at appraisal and that began to change only recently. 6.2 At the time of project preparation, construction sector activity was high and the main constraint hampering the development of domestic construction firms was indeed the lack of financial resources for equipment purchases. Therefore the concept of the project to build upon that need with additional features responded well to the Bank's previous recognition that financial assistance to contractors would be effective only, if it is reinforced with technical assistance, with improvements 'n governments' procurement procedures and practices, and with addressing more fully the range of institutional, financial and administrative problems constraining the construction sector development. 6.3 The lessons that could be drawn from this experience in order to improve the design of similar future projects include: {i) Allocation of sufficient works: If the country has limited private demand for construction, Government should be ready and willing to allocate works from existing development programs to the local contractors whom the project intends to help. To be effective, any development project of domestic contractors should, therefore, be tied to a program of ongoing works. (Masons are taught while building walls, and not in academic classes). In the Burundi case, the design of the project relied on a continuously flourishing construction market providing sufficient demand to keep the contractors in business. Unfortunately, the downward trend of the economy cancelled out this possibility. A safety network cculd have been built-in by allocating the domestic contractors a preferential share in the few projects which continued being implemented (such as road maintenance or school construction). As this was not done, domestic contractors became more and more idle which severely - 17 - weakened the project. Consaquently, similar projects should be based on thorough studies of the sustainable demand based on which the construction industry will develop, and be tied to an appropriate program of works. (ii) Strong commitment of the authorities: The project suffered from delays in support from the Authorities, who did not react in a timely fashion when the situation deterioriated. Therefore, it appears that it might have been wrong to place the project under the direct supervision of the minister himself. It would have been wiser to get a high level manager involved in the project, such as the 'directeur de la construction," who could have provided routine support to the project manager. (iii) Limited dispersion of activities: The projoct should have concentrated on a smaller number of activities with a clear order of priority. Too broad an array of targets made project implementation and pezformance difficult. It is easier to select a consultant with a specialized technical field than tc find a generalist who covers very different activities satisfactorily. -18- Annex 1 Page 1 of 2 PROJECT COMPLETION REPORT BURUNDI LOCAL CONSTRUCTION INDUSTRY PILOT PROJECT Credit 1230-BU Summary description of brick-oven subprolects I. Kamenge Brickworks (Simon Mahwera) BNDE accepted the Kamenge brickworks application on January 20, 1984. The esouipment, comprising a gasoline-powered mixing machine and 2 manual brick presses with matrix, was ordered in February 1984. The civil engineering works started in February and were completed in late October 1984. Firing of the first kiln load of about 37,000 bricks took place on November 30, 1984. Originally designed to operate with a single * upright kiln holding about 50,000 bricks per charge, the brickworks quickly expanded since the standstill time for charging, cooling and recharging, on the order of 20 days, was slowing down production, the * construction of a second kiln took place in April 1985 . To meet an ever growing demand, two additional Terstaram-type manual presses were purchased and four upright kilns were built, raising monthly production from 100,0G0 to 250,000 solid bricks during the summer period when the relative air humidity is very low. Supplemental equipment, comprising a type 4 U clod-breaker, a 15-hp type 4V crusher, an SZ 2 wetter/mixer, a type VL 15 molder with vacuum pump and three filieres, a cutter and two 45-hp diesel drive motors, were ordered in December 1985. The four machines and the motors arrived in August 1986 and irstallation was completed in November 1986. The clay preparation and machine-setting tests were carried out between January and April 1987. The trials for bakirg hollow bricks were conducted on the reversed-flame kiln between FebtLary and May 1987 and gave satisfactory results. Pr,4uction with the new machines has not yet started. It awaits competion of construction of the reversed-flame kiln, which is in progress. The molding machine was set to a production of 16 tons per day. Despite the difficulties encountered, the prospects appear quite good. - 19 - Annex 1 Page 2 of 2 II. Kayanza Brickworks (Barthelemy Barutwanayo) The application for the Kayanza Brickworks was approved by BNDE on November 29, 1985 and transmitted to IDA for approval on December 4, 1985. IDA apDroved it on January 29,1986. The machines were bought from the Commune of Bururi, which had abandoned its own project following a change of governor. They comprise a clod-breaker, wetter/mixer, a diesel drive motor and two manual presses. Total ennual production is about 2,200 tons, with expansion possible to 4,400 tons. Construction of the storage shed and kiln and installation of the machinery were completed in May 1987. Baking of the first kiln load of 35.000 bricks took place in July 1987 in the absence of the project unit's brickmaking technician, who had left on June 30, 1987, but who returned for a short follow-up period in early 1988. The project is facing high production cost problems and, together with BNDE, is looking for further technical assistance. III. Second Bujumbura Brickworks (EGB construction company) The application for the Bujumbura Brickworks was sent to BNDE on October 3, 1985 and approved on Novemuer 29, 1985. It was approved by IDA on January 29, 1986. The original project proposed by the project unit was of the same capacity as the Kamenge Brickworks and had two purposes: to provide good- quality bricks in adequate quantity to supply the Bujumbura market, and to ensure competition with the Kamenge Brickworks in order to maintain product quality. Following the decision o, the promotor, much more efficient equipment was ordered in April 1986. Production trials started in January 1989 with assistance from the equipment supplier. - 20 - Annex 2 Page 1 of 3 PROJECT COMPLETION REPORT BURUNDI LOCAL CONSTRUCTION INDUSTRY PILOT PROJLCT Credit 1230-BU Susmmary Results of Feasibility Studies I. Pozzolamic cement The basis of the study was a resurrection of ENACCI, the former public cement company, which, located at the Tanganyika Lake, has never operated since the floods of 1964. IDA discourages the establishment of new public sector manufacturing units, especially as part of the policy context of the recently approved Second Structural Adjustment Credit. The sunmmary analyses below is, therefore, for the record only. No private investor so far has expressed interest in providing the necessary funds. The rate ot replacement of portland cement was estimated at 25% until 1990, and demand was estimated at 26,473 tons. Two production alternatives were considered: conversion of ENACCI, and installation of a new production unit. The cost, not including operating expenditures, comes to BuF 34,024,822 for rehabilitation of ENACCI and BuF 179,879,500 for a new production unit. Three output levels were examined: 9,150, 18,300 and 27,400 tons/year. With a price of BuF 18,000/ton anid an annual output of 27,400 tons the IRR is 28.44% for a new production unit and 59.0% for conversion of ENACCI. While economic rates of return were not provided, the sensitivity analysis yielded the following results: - If the selling price per ton falls from BuF 18,000 to BuF 16,200, the IRR falls from 28.44% to 22.4% for a new production unit and from 59% to 48.50Z for conversion of ENACCI. - If the selling price falls to BuF 14,400/ton, only conversion of ENACCI shows an attractive IRR (37.02%). - If production falls from 27,400 to 9,150 tons/year and the selling price is set at BuF 18,000/ton, the IRR falls from 48.50Z to 9.55Z in the case of a new production unit and from 59% to 30.60% in the case of conversion of ENACCI. - If operating expenses rise by 10%, 20% and 30%, with a selling price of BuF 18,000/ton, the ENACCI conversion project remains profitable only in the case of a 10% increase. - 21 - Annex 2 Page 2 of 3 II. Manufacture of sanitary articles Seven items forming a standard set of sanitary articles were selected. Average consumption was estimated at between 230 and 300 a year up to 1990. Production capacity was set at 750 standard sets a year and 200 wall urinals, which assumes exports of 450-520 bathroom suites, something rather unlikely, given transport problems, and as unfavorable price/weight ratio. Installation and operating costs amount to BuF 179,593,198 and BuF 18,140,166 respectively. Project implementation time is estimated at about 23 months. The IRR remains low for the output levels in question: 6.98% and 3.50%. This poor profitability is due to the high amount of the investment required to set up the plant. III. Manufacture of sawn and polished stone slabs On the basis of the investments in the construction sector, the potential area to be built was estimated at about 140,000m2/year. Assuming 50-percent implementation of this projection, the area to be built falls to 70,000m2/yer. The utilization rate of floor slabs is between 17.1% and 21.4%, which generates a surface of between 12,500m2 and 15,00m2/year. The profitability calculation was based on three production hypotheses of 12,000, 15,000 and 30,000m2/year. With the IRR set at 25%, the selling prices for outputs of 12,000, 15,000 and 30,000m2/year are BuF 9,260, 7,620 and 4,390/m2 including taxes and BuF 6,760, 5,610 and 4,390/m2 excluding taxes. Although the project is financially viable, the substitutes for sawn and polished stone slabs are currently cheaper (FBu 1,100/m2) and this price discrepancy would discourage purchase of the new product. IV. Manufacture of earthenware tiles Two products were selected: earthenware tiles (carrelages de gres) and pantiles (tuiles "canal"). Using the same market hypotheses as for sawn and polished stone slabs, profitability calculations were performed for outputs of (a) 20,000m2 of earthenware tiles and 5,000m2 of pantiles and (b) 40,000m2 of earthenware tiles and 10,000m2 of pantiles. For production of 25,000m2, installation cost is BuF 80,802,312 and operating cost BuF 17,914,870. For production of 50,000m2, installation cost is BuF 93,827,078 and operating cost BuF 23,037,400. The rate of return for 25,000m2 production is 1.99Z including taxes and 6.51% excluding taxes. The rate for 50,000m2 production is more attractive: 16.78% including taxes and 29.53% excluding taxes. However, the Burundian market is too small to absorb such a production volume. -22 - Annex 2 Page 3 of 3 V. Use of bamboo in construction The manufacture of woven ceiling panels was adopted for this study. The installation cost of a unit with a production of 105,300m3/year was estimated at BuF 41,707,486. Operating costs (raw materials, energy, labor) of BuF 23,877,840 must be added. The project is not profitable since the selling price per m2 would have to exceed the price of the substitute products (triplex and asbestos cement board). In addition, the Burundian market is not large enough to absorb the full plant's output. -23- Annex 3 PROJECT COMPLETION REPORT BURUNDI LOCAL CONSTRUCTION INDUSTRY PILOT PROJECT Credit 1230-BU Credit Allocation of Subloan Components Subproject Name Date approved committed disbursed No.- _ (SDR) (SDR) A. Subloans for Brickoven Construction BM3 Simon M.hwera (Kamenge) 3/7/84 38,000 37,837 BM5 Briqueterie de Bujumbura 1/28/86 188,500 125,454 BM6 Briqueterie de Kayanza 1/28/86 51,400 28,280 + subsequent increase 3/4/87 4,500 Sub-total 282,400 191,571 B. Subloans to Contractors CC-1 LAGECO 6/21/84 75,093 75,093 CC-2 NKAMIKANIYE 6/21/84 62.000 59.977 Sub-total 137,093 135,070 TOTAL 419,493 326,642 * BM1, 2 and 4 wer4 cancelled because promotor failed to produce required equity participation of 25%. Annex 4 Page 1 of 17 89E0475 February 7, 1989 - 24 - French (Burundi) 175/20-OEDD2 TS:sn COMMENTS FROM THE BORROWER Banque Nationale pour le D6veloppeaent Economique Bujumbura, Burundi Ref: 16/DR Bujumbura, January 5, 1989 World Bank Washington, D,C. For the attention of Mr. Alexander Nowicki Division Chief, Operations Evaluation Department Ref: Your letter of November 14, 1988 Subject: Project Completion Report on Local Construction Industry Pilot Project (Credit 1230-BU) Gentlemen: Further the above-referenced Project Completion Report we have the honor to set forth our comments below. We consider that BNDE fulfilled all the obligations deriving from the project agreement signed on June 1, 1982 by IDA and BNDE. If "as a prudent development bank, it had to take a more conservative view when considering project approval" (para. 5.3 of PCR), with hindsight we have to admit that it was right to act prudently since the results subsequertly obtained have not unfortunately matched the targets set and the optimistic forecasts of the LCI project unit. The fact is that, contrary to what is stated in the report where it notes that BNDE "concentrated only on general administration of the subprojects, the repaymenc of which does not present problems" (para. 5.3 of PCR), all the loans granted under Credit 1230-BU, with the exception of one that has been paid in advance, are in arrears, in some cases to such an extent that losses for the bank are to be expected. We att.ach an annex detailing the present status, both financial and economic, of the subprojects financed by BNDE; we also attach recent visit reports and an expert's report on the requirements for the Kayanza brickworks to be viable. Yours sincerely, Is/B. Boisselet s/B. Kamwenubusa Director of Financial Operations President - 25 - Annex 4 Page 2 of 17 cc for information: Minister of Public Works and Urban Development Minister of Energy and Mines Minister of Finance World Bank Resident Representative in Bujumbura Encl.: 4 - 26 - Annex 4 Page 3 of 17 ANNEX Status of SubRroiects Financed by BNDE under Credit 1230-BU (December 1988) BRICKWORKS 1. BM-3 KamenRe Brickworks (Mr. Simon Mahwera) Loan Recovery Situation Amount of loan granted on January 20, 1984 BuF 4,616,000 Outstanding as of November 30, 1988 4,250,910 Arrears as of November 30, 1988 2,804,410 Prospects The promoter recently completed (after 18 months) the extruded brick production trials and he is only now beginning to market the output. Despite the difficulties encountered, the prospects appear quite good. See also attached visit report. 2. BM-5. Bujumbura Brickworks (SOFABRI) Loan recovery situation Amount of loan granted on November 29, 1985 BuF 24,000,000 Outstanding as of December 31, 1988 27,201,925 Arrears as of December 31, 1988 3,927,390 Prospects Contrary to what is stated in the Project Completion Report, this brickworks has not yet started producti4... Trials are to start in January 1989 with assistance from the equipment supplier. The promoters regret not having been able to benefit from technical assistance from the Local Construction Industry project whose last expert left before the startup of the subproject. It is still too early to be able to give an opinion on the economic viability of this subproject, the promoters of which-who also run a public works enterprise-seem to be competent. See also attached visit report. 3. BM-6. Kayanza Brickworks (Mr. Barthelemy Barutwanayo) Loan recovery situation Amount of loan granted on November 29, 1985 BuF 7,000,000 Outstanding as of November 30, 1988 7,930,839 Arrears as of November 30, 1988 2,335,294 - 27 - Annex 4 Page 4 of 17 Amount of supplementary loan granted on 12/17/86 1,000,000 (out of BND funds) Outstanding as of November 30, 1988 1,291,399 Arrears as of November 30, 1988 353,300 Prospects The brickworks has been operating intermittently since the end of 1937; the production cost of the bricks is greater than the selling price. The promoter regrets that the project technical assistance, which helped him with the construction of the kiln, was withdrawn at the time of the production rrials because the expert left for good. The prospects are presently very bleak. The promoter and BNDE are currently looking for techuical assistance, which could be provided by the EEC Industrial Development Center. See also visit report and CERATEC recommendations. CONSTRUCTION FIRMS 1. CC-1. LAGECO Loan recovery situation Amount of loan granted on April 20, 1984 BuF 9,115,302 Outstanding as of December 15, 1988 7,695,086 Arrears as of December 15, 1988 7,695,086 Prospects The firm has not obtained any con.racts since 1987 so its activity was practically nil in 1988. The State's commitment in respect of LAGECO is about BuF 6 million. The prospects appear very bleak. See also visit report. 2. CC-2. NKAMICANIYE Amount of loan granted on May 22, 1985 BuF 7,500,000 This loan was repaid in full in advance in 1987. Bujumbura, January 5, 1989 - 28 - Annex 4 page 5 of 7 Visit Reports 1. Kamenge Brickworks 2. SOFABRI Brickworks Visits made by Mr. B. Boisselet and Mrs. Beatrice Bukware 1. Kamenge Brickworks Date of visit: December 9, 1988 Present Situation The brickworks started its activit-.es at the beginning of 1985 under the supervision of the LoLal Con-truction Industry (LCI) project. It ran into a large number of problems due largely to lack of experience, to the cost of the brircs produced being higher than anticipated, and to the sizable new investment requirements (drying and storage shed, new presses, second kiln). The promoter informed us that the trials which have taken practically 18 months have now been completed: -~ using Kamenge clay, they can produce extruded bricks at a rate of 3,000 bricks a day and BuF 12 per brick. - using Muzinda clay (binder) mixed with Kamenge clay, they produce Dressed bricks at a rate of 4,600 a day and BuF 8 per brick. The future seems promising because they look like obtaining two big contracts: - (1) SIP, for 1,700,000 bricks - (2) COTEBU, for 2,000,000 bricks Since its bricks are of good quality, the Kamenge Brickworks will certainly be well placed to gain a part or all of these contracts. Financial Situation The brickworks is experiencing financial difficulties connected with the reasons mentioned above. Arrears amounted to BuF 2,804,410 as of December 9, 1988. The promoter has undertaken to pay two installnents in December 1988 and two per quarter in 1989 in order to pay off the arrears. 2. SeVABRI Brickworks Visit made on December 19, 1988. The brickworks has had a line of machines for brickmaking since October 1987. Construction of the kiln was started in May 1987 but the shed has been ready since 1986. Annex 4 - 29 - Page 6 of 17 The kiln is of the "reversed flame" type. The plans were made by the LCI project but construction was done by the promoters. They did not receive LCI technical assistance as was originally intended because the LCI project was terminated in June 1987. The equipment supplier will send an expert for the startup of production in January 1989. Actual production will be started in February 1989. A stock of wood is already available and the promoters expect to build up a stock of clay between now and January 1989. SOFABRI will benefit from the Kamenge Brickworks' experience. Financial Situation Loan granted BuF 24 million (Board meeting of Nov. 29, 1985) Interest rate 12X 11 years including one grace year Arrears as of December 19, 1988: BuF 2,858,666, comprising interest due on March 31, 1988 and two installments (first installment due June 30, 1988). The promoters intend to call a meeting of their Board befor -he end of 1988 to determine how they can remedy this situation. An additional grace year has been proposed to SOFABRI, which would make the first installment due on June 30, 1989 (cf. letter of December 23, 1988. Bujumbura, December 27, 1988 /s/Beatrice Bukware Visit Report Enterprise visited: LAGECO Visit made by: Mr. B. Boisselet and Mrs. Beatrice Bukwsre Persons met: Messrs. Fidele Nsabimana and Germain Semondo LAGECO obtained a loan of BuF 18 million, of which only BuF 9,115,302 was released, the rest being canceled. Repayment has been due for a year, but nothing has been received since April 1988. The entire balance as of December 2, 1988, of BuF 7,695,086, has been declared due and payable and LAGECO has been so notified by BNDE letter of December 5, 1988. The purpose of the visit was to discuss with the promoters whether they could make any concrete proposals regarding reiayment of this loan before BNDE initiates legal action. The promoters stated that LAGECO's activities were at at standstill during 1987-88 because it received no contracts. Final acceptance of the last -30- Annex 4 Page 7 of 17 works took place in September 1988 and payment of the ;S% withheld, amounting to approximately BuF 6 million, is now due and this sum will be paid by the State in the course of 1989. We have asked LAGECO to send us a letter providing the following information: - the amount payable by the State - copies of accounts receivable - proposal as to the proportion of each bill paid that will be transferred to BNDE - BANCOBU's consent to transferring the proposed amount to our account - LAGECO's future prospects. Once this information is received, BNDE will be able to decide whether it can allow rescheduling of the debt or whether it should proceed with legal action. Bujumbura, January 5, 1989 /s/ Beatrice Bukware - 31 - Annex 4 Page 3 of 17 BNDE REPORT ON THE SITUATION OF THE KAYANZA SEMI-INDUSTRIAL BRICKWORKS Visits made on January 29 and February 26 by Mrs. Beatrice Bukware and Marc Rotsaert. 1. Proiect background (a) On November 29, 1985, the BNDE Board granted Mr. Barthel6my Barutwanayo a loan of BuF 7 million for establishment of a brickworks at Kayanza. The terms and conditions of this loan were: Interest: 12X p.a. Term: 6 years, with one grace year Refinancing: IDA (LCI project) (b) The installation work was started in September 1986 and a supplementary loan of BuF 1 million was granted on the same terms and conditions on December 17, 1986. (c) The first production trials began in June 1987. The kiln was constructed with LCI tec'nical assistance but production was handled by the promoter once the LCI expert left for good. 2. Present situation of Pro1ect 2.1 Production After a trlal period, the promoter concluded a contract with BPE for the supply of 800,000 bricks for construction of the Kayanza Teacher Training School. The selling price was fixed at BuF 5 per brick delivered to the worksite. This price was also subject to change depending on the quality of the bricks and to the extent that transportation from the brickworks to the site was done by BPE, with the result that the output and sales data were as follows as of January 30, 1988: Bricks baked Losses Bricks sold Price Total sales 45,000 21,500 23,500 4.5 105,750 55,000 13,000 42,000 3 126,000 59,000 7,000 52,000 3 156,000 55,000 15,000 40,000 3 120,000 50,000 20,000 30,000 4.5 135,000 53,000 12,000 41,000 4.5 184.500 317,000 88,500 228,500 3.75 BuF 827,250 3nnex 4 Page 9 of 17 2.2 Operating costs According to the promoter, the operating costs break down as follows: - Materials and supplies BuF 765,500 Water + electricity 30,000 Fuel 41,000 Raw materials 65,500 Kiln fuel 135,000 Transportation 436,000 Travel 60,000 Stationery 3,500 - Other services 79,075 Building maintenance 5,000 Vehicle maintenance 65,000 Postage, telephone 2,500 Bank charges 6,575 - Personnel 740,000 Wages 715,000 Medical 25,000 - Taxes 22,000 Local taxes 22,000 BuF 1,606,575 3. Analysis of the situation We observe that the brickworks' productivity is too low and that the present selling price is less than the cost of the bricks. (a) Productivity Up to the present, the promoter has on average been baking one batch of 55,000 bricks per month. Losses are very high: 28S To remedy these problems, the promoter has now stopped production in order to take the following steps: - Provision of an access road to the clay pit, since the existing road is impassable when it rains - Construction of a shed to provide more space for drying bricks prior to baking (space created: 180 m2) - Construction of a shed for storing clay out of the weather (space created: 144 m2). Annex 4 - 33 - Page lO of 17 As to the loss rate, this is explained by the promoter's lack of experience in brickuaking, especially since he was no longer able to use technical assistance from the LCI project, which had been wound up in the meanwhile. The loss rate for the last batch was still 23%. (In comparison, the loss rate for the Kamenge brickworks is 1A.) In addition, the organization of production is complicated by the need for transportation, which is hard to hire on a day-to-day basis in Kayanza. A duuptruck is needed for: - hauling fuelwood, which has to be obtained from the vicinity of Kayanza (four truckloads per kiln) - hauling peat, which comes from Bujumbura - bringing in clay, which entails a 20-km round trip. Thirty truckloads are needed to fill one kiln. Production (b) Cost To the extent that the operating cots reported by the promoter represent the true situation since the start of the operation, the production cost breaks down as follows: - Variable costs Water + electricity 30,000 Fu-el 41,000 Raw materials 65,500 Kiln fuel 135,000 Transportation 436,000 Maintenance 5,000 Wages 715.500 BuF 1,428,000 - Fixed costs Overhead 191,150 Finance charges 508,800 (6 months) Amortization 600,000 BuF 1,299,950 (6 months) In the present situation, the production cost would be: 1,428,000 + 1,299.950 = BuF 11.9 brick 228,500 bricks sold In the future, assuming tiiat the loss rate can be brought down to 5S and that two batches are produced per month, the production cost could come down to: - 34 - Annex 4 Page 11 of 17 Bricks baked 317,000 Bricks sold 301,150 Variable costs BuF 4,742 bricks Fixed costs 1,299,950 - BuF 2,158 brick 301,150 x 2 Production cost BuF 6.9 brick Selling price BuF 7.5 brick Gross profit BuF 361,000/month Comments The cost of transportation adds to the production cost: not counting the trips made by the BPE truck, the cost of which ihas been incorporated directly into the selling price, transportation represents 30% of the variable costs. The promoter would do well to obtain a dumptruck, because it would: - facilitate organization of the work - make the project independent - increase productivity - bring production costs down - an could also conceivably be used for haulage activities outside of the project. The promoter is presently negotiating with a private individual for a second-band tiuck at a cost of BuF 3 million. Since he does not have the funds for this purchase, we have recolmmended to him that he check out a leaRing formula. Conclusions The first order provided valuable experience for the promoter in that it enabled him to gain a better idea of the difficulties of his project and to place on the market a good quality brick with a compressive strength of 200 kg/m2. It will nevertheless be impossible for this brickworks to show a recurn. Repayment of the loan will therefore be highly uncertain. We accordingly suggest that management reschedule the debt over ten years, including two grace years. As to purchase of a truck, could not the bank's management consider leasing to Mr. Barutwanayo the truck belonging to Sobritumaga that is currently the subject of a dispute with Mr. Kamatari? /I/ Marc Rotsaert Rapporteur 35 - Annex 4 Page 12 of 17 REPORT ON VISIT TO KAYANZA BRICKWORKS ON OCTOBER 19, 1988 Purpose of visit: To inform the representative of the Belgian company "CERATEC" regarding the situation of the project to enable CERATEC to propose a technical assistance plan for restarting the project and making it profitable. Participants: Mr. Barthelemy Barutwanayo, promoter Mr. Vanoverstraeten, CERATEC Export Manager Mr. Boisselet, Director of Financial Operations, BNDE Mr. Rotsaert, BNDE The participants first visited the brickworks and then went to the claypit. The CERATEC repr'sentative checked the dimensions of the kiln structure and the characteristics of the equipment. Because the claypit is so far away (about 8 km), CERATEC would like the promoter to gather clay samples from nearer the brickworks and send them to it in Belgium, so that CERATEC can determine what products can be made from it: tiles, bricks and floor slabs. The CERATEC representative pointed out that it is difficult for brickworks located too far from their claypits to turn a profit. It was suggested to the promoter that he should send samples of his products to potential customers and determine the prices they would be prepared to pay. The promoter explained that he had several customers interested in floor tiles (for which he did not have molds) and "tuiles flamandes." Availability of the truck that was the subject of the promoter's present loan application for haulage activities outside of the brickworks was examined. In the current situation the truck would be used mainly for hauling clay and fuel. Conclusion The CDRATEC representative will send BNDE proposals for possible improvement of the project's situation. Bujumbura, October 21, 1988 s/Marc Rotsaert - 36 - Annex 4 Page 13 of 17 CERATEC S.A. Rue du Touquet 228 B-7793 Ploegsteert Belgium Ref: PDB/DDC/JRM 402.01 November 25, 1988 Mr. Boisselet Director of Financial Operations BNDE P.O. Box 1620 Bujumbura, Burundi Subject: Kayanza Brickworks Dear Sir: Following Mr. Vanoverstraeten's recent visit to Burundi where he visited the Kayanza Brickworks, we attach some recoummendations which we hope will help you to resolve the problems this project is experiencing. Only a full-scale study by an expert would make it possible to consider all the details of the situation and give you precise specific advice as to the steps to be taken and the possibilities open to you. Regarding training, Mr. Vanoverstraeten will certainly have told you that we regularly organize training courses (practical and theory), a detailed program of which is also attached. We will, of course, be pleased to provide any further assistance you may require, and will make a point of visiting you ths next time we are in Burinmdi. Yours sincerely, /s/D. Deconinck /s/Ph. de Bruyn Technical Adviser Managing Director 37 Annex 4 Page 14 of 17 RECOMMENDATIONS FOR THE KAYANZA BRICKWORKS, BURUNDI The project in its present form is very probably not viable. Its main handicaps are: 1. The distance between the claypit and the brickworks is too great and the cost of transportation too high. 2. The use of a high-powered diesel generator or motor is extremely costly. To the extent possible, connection to the electricity system is much more economical in the long term. 3. Baking: with the present kiln, and with any type of intermittent kiln, about 1,200 kcal is need to bake one kilogram of dry clay, and only firewood can be used, whereas with a continuous or semicontinuous kiln only 400 or 600 kcal is needed for 1 kg of dry clay. Moreover, any fuel can be used, for instance sawdust, coffee husks, etc. 4. Personnel training; output Without training it is usually difficult to obtain very good results, and especially to achieve a high output. Project startup is then a very slow process, which ultimately proves extremely costly and can cause the project to fail. In the present case, output is extremely low, notwithstanding the fact that the project was started two years ago. 5. This project as it has been set up does not make a lot of sense, being a mixture of incompatible machines and processes. One uses either: (a) vacuum extrusion, in which case the production line will be made up of a feeder, a roller, a mixer, a vacuum extruder and a cutter. The clay must contain 20% moisture and is mixed to form a plastic and homogeneous mix. (b) dry pressing, in which case the moisture content of the clay must not exceed 10-15X, otherwise the clay will stick to the molds. In addition, the clay must be broken into fine particles (particle size from 0 to 20 mm) so that the molds can be properly filled with a constant quantity of clay and maximum production achieved. Mixing is pointless with this technique. A production line normally consists of just a clod-breaker and presses (manual or automatic). For the Kayanza Brickworks, the most appropriate technique would therefore seem to be dry pressing. However, to enhance profitability the following points should be corrected: (a) Claypit The distance between the pit and the works is too great, especially for so low an output.; It would be advisable to see whether a 38 - Annex 4 Page 15 of 17 source of clay could not be opened up closer to the works. If this proves impossible, the production cost should be analyzed to determine the incidence of transportation. On the face of it, purchase of a truck for so low an output would hardly be worthwhile. Renting two trucks and a hydraulic shovel for one day a month would perhaps be the best solution. However, this would still leave the problem of firewood supply. (b) Preparation The mixer is quite unnecessary. I. should therefore be eliminated in order to save labor and energy. (c) Diesel consumption could also be reduced. The clod-breaker used is a very heavy one that needs a high installed power. Other types of breakers operate with a Lister 5-hp motor and only use 5 liters of diesel a day. (d) With a little training, it should be possible to considerably increase production, up to the maximum capacity of the manual presses (300 bricks/hour per press). Stepping up of production will in any event be essential for the enterprise's survival. Output could also be increased by procuring an automatic press (1,400-2,000 bricks/hour), provided there is a sufficient supply of clay and production is well organized. (e) Baking, as it is done at present, is very good. Construction of a reversed-flame kiln will not lessen fuel consumption. However, in the future, once the other problems have been solved, construction of a rectangular kiln (see Annex) might be the ideal solution and would permit a quite sizable fuel saving, which could provide a decisive advantage compared with traditional production units. In general, it is very diffi ult to produce bricks at a cost lower than that of traditioaal bricks. It is the quality difference that makes it possible to justify the higher price, provided the price difference is not too great. With a clod-breaker and presses (manual or automatic), this is perfectly possible as long as the above points are corrected. Production of items such as tiles is generally very profitable and can be done with very little equipment. However, if a high output is desired the investment required is very sizable (extrusion line, with or without vacuum equipment). Annex 4 39 - Page 16 of 17 89E0523 February 27, 1989 French (Burundi) OEDD2 OTS:jl Republic of Burundi Ministry of Energy and Mines Office of the Minister Bujumbura, Burundi Our Ref.: 760/CAB/082/G.B./S.A/89 Bujumbura, February 10, 1989 World Bank Washington, D.C. For the Attention of Mr. Alexander Nowicki Division Chief, Operations Evaluation Department Subject: Prolec' Completion Report on Local .onstruction Ilidustry Pilot Pro_ect (Credit 1230-BU) Copy sent for information to: Prime Minister and Minister of Planning, Bujumbura Dear Mr. Nowicki: Thank you for the draft version of the Project Completion Report on the Local Construction Industry Pilot Project. As you emphasize in the Highlights of the report, Burundi is heavily dependent on imported construction materials. It was to eliminate or at least lessen this dependence that the Government asked IDA for the funds to set up Burundi's own local construction materials industry. This financing was granted in the context of the LCI project. However, for the reasons set forth under Conclusions and Lessons Learned in the PCR, this project did not produce the results expected of it. You accordingly decided to terminate the LCI project before the credit was fully utilized. However, imported construction materials are today an evkn heavier drain on Burundi's resources than before (because imports are more costly and the Burundi franc has depreciated). Utilization of local raw materials is therefore now a vital necessity. Annex 4 - 40 - Page 17 of 17 I think that certain components of the LCI project, in particular development and improvement of prod.ction of local construction materials, should be examined, taking into account the difficulties already encountered, of course. Yours sincerely, /8/ Gilbert Midende Minister of Energy and Mines cc for information: Min.ster of Public Works and Urban Development Minister of Finance Banque Nationale pour le Developpement Economique World Bank Resident Representative Director General of Geology and Mines 41 Exhibit I -41 - # Page 1 of 5 clay mixer 43 X ,1 manual brick press d o | ~~~~~~~~~~dryin8 of preased brickc prior to being heated 4_ Exhibit I _ ~~~~~~~~~~~~Page 2 of 5 (C lC_ loading bricks into ov':n f4. e1- nf nv n - 43 - Exhibit 1 i ME Page 3of 5 4-~ '41~ 4- ~ Nf' ,, ' p..~~~~~ construction of necond oven --_ -, a* - i,l. lt ' ',5 1' 27~~~~7 construction of second oven - 44 - Exhibit 1 I F~~~~~~~~age 4 of 5 - regular artisan brick qai ofbcfr prjtfncdbcovI I~~~~~~ quality of brick from project financed brick oven - 45 - Exhibit_I Page 5 of 4*e, Xr VI , S~~~~ -p. ~~~~~~~~~~~~~~~* A *..~~~~ -* I~~~~~~~~~~~~~ Exhibit 2 -46- Page 1 of 3 touTES LES CONSENMAB SONE e A LOCALDC LE CUW1rLE EUT WD m L FAlOrATI 0u PiOWUrr p__b"w dassubsbu ma* gs asd m a Wn nWuaue g am x WA Om x a so S 4mmn paase Cms 2S kg. ds q des W"s aua cvsmm de s uSW '1 flean Ia mduMW,sk. ~~~~~~~~~~~~MTRE : a-1 d_sW* 6 CYKf 04 ba Ille -10U _ sr_iL - t dissu*gas m mpr de 4 A 10 t/-h __ ~~~~~~~~~~~- 1 nikxmi c25d5dl0it - I m0bw iss ee amascde as mm dm' A posifi et aassais Plinse ou 2 nAma5s OMscta &Ac Nbh.s deci -lI p~sess avjes c 2 masnde 4. ncwis IGENE CMWL - 2 hIn di,*0bd. 45 mO - ri - I hF a de 0cosp - n de 370m'. TEMPS DE CONSTRUCTO : tO amis' EMPLOIS CREES :22 Ps _ ESTISSEMENT TOTAL O: s 50O.C S US * EcIlg 400 S tn. W_ TE000 T0UOS/ __ bl~~~~~~~~MTERIEL I d__vtIff & cyW t be 6tOUh I ma.W. 4.5 mc de 4 6 10 t/PL p amu 2 moin Motsis m &mc _J de et4 - Spaa NmAM amc 4 mouse de rscimgso OBMOW - 2 aL,abd ft s 2 MI4 f X - I bosBd tgS a lob1 af de 0W0 'iF; TEMS 0 CONSTRUCTION la m:ll" - REES iv 40 _ p; S tf sfES1USEBdT TOTAL Din 0_ am S 1La ' mbs. S us" UW 0300 T1MIl MATERL - 1 are*g cyb* et ba s s-l0 Lt% - M O~ ls ,wdl sdes4 610 L/k 3- Sr m wAses mc as vde * p a __ , , _ NE CNL - 3 han nma dsse s2ml6 mas OSIE ~ S- I hw ad 'cin' -I-S d*M'. _.. TEMPS EOEN nN : 0ofa s10m TOTAL O 01_ U00 S US. * ElnaSs
Группа Всемирного банка · Project Completion Report
Burundi - Local Construction Industry Pilot Project
Открыть оригинал документа
Полный текст размещён на сайте публикующей организации. lawenc.com индексирует метаданные и ведёт на официальный источник.
Полный текст
Основные сведения
Организация
Группа Всемирного банка
Тип документа
Project Completion Report
Страна
Бурунди
Источник
Всемирный банк