Dahomey - Second Highway Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 3986 PROJECT PERFORMANCE AUDIT REPORT BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) June 30, 1982 Operations Evaluation Department 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. 6 FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) TABLE OF CONTENTS Page No. Preface ............................................................ i Basic Data Sheet ................................................... ii Highlights ......................... .............................. iv PROJECT PERFORMANCE AUDIT MEMORANDUM I. BACKGROUND .....1..................................... 1 II. PROJECT RESULTS ....................................... 2 III. ISSUES ....... ................................. ...7...... 7 IV. FUTURE IDA SUPPORT FOR TRANSPORT INVESTMENTS ............. 13 V. CONCLUSIONS .......................................... 15 Annexes A. Benin and Togo - Vehicle Operating Costs Excluding Taxes - 1972............................................ 16 B. Benin and Togo - Vehicle Operating Costs Excluding Taxes - 1976..... .......... ............ ......... 17 C. Benin and Togo - Vehicle Operating Costs Excluding Taxes - 1980................................. ......... 18 D. Proposals for Use of Unspent Credit Funds ................. 19 PROJECT COMPLETION REPORT I. Introduction ......................................... . 21 II. Project Identification and Preparation ................. 22 III. Project Implementation ............................... 24 IV. Economic Reevaluation .................................... 33 V. IDA Performance .................................. .....35 VI. Conclusions .......................................... 36 Tables 1. Traffic and Transit Trade at the Port of Cotonou 1977 37 2. Highway Network, 1979-80 ..................................... 38 3. Bids Received .......................................*....... 39 4. Road Maintenance Expenditures, 1970-78 .... -............ 40 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. -2- Table of Contents (cont'd.) Page No. 5. Project Costs............................................. 41 6. Schedule of Disbursements................................. 42 7. Average Daily Traffic..................................... 43 8. Vehicle Operating Costs........................ ........ 44 9. Reevaluation of the Economic Rate of Return of Parakou-Malanville Road (318 km) and the two Sections of the Cotonou-Bohicon Road (17 km) ............ 45 ADDENDUM TO PROJECT COMPLETION REPORT ............................... 46 Map PROJECT PERFORMANCE AUDIT REPORT BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) PREFACE This report presents a performance audit of the Benin Second Highway Project, for which the Executive Directors approved Credit 415-DA in the amount of US$11.8 million on April 24, 1973 and US$9.0 million on March 10, 1976. The original credit was fully disbursed in April 1977, while only US$6.4 million of the supplementary credit was drawn down and the balance was cancelled in March 1982. The report consists of a Project Performance Audit Memorandum (PPAM) prepared by the Operations Evaluation Department (OED) and a Project Completion Report (PCR) prepared by the IDA's Western Africa Regional Office. The Borrower provided material for the PCR to a mission undertaken by the Region in June 1979 and updated traffic count data in June 1981. The bases for the PPAM comprised: the PCR, discussions with IDA staff, the transcripts of the Executive Directors' meetings at which the project was considered, and the project files and documents. OED staff did not visit the country. The audit agrees with the PCR on many aspects. However, it points out the shortcomings in the methodology used for the economic analysis as well as the appraisal underestimation of traffic development on the Parakou-Malan- ville Road, the variations in estimates of vehicle operating cost savings, and the over-estimation of supplementary credit needs. In addition, the audit suggests that the competition for traffic between new and existing transport investments will shortly call for Bank Group policy decisions regarding Benin and its neighboring country, Togo. Finally, the audit notes issues raised by the attempts to apply the large remainder left unused under the Credit after the original (revised) project components had been completed. The draft report was sent to the borrower for comments; however, none were received.  - ii - PROJECT PERFORMANCE AUDIT BASIC DATA SHEET BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) KEY PROJECT DATA Appraisal Actual or Item Estimate Reestimate Total Project Cost (US$ million) 21.1 32.1a Overrun (%) - 53 Credit Amount (US$ million) Original 11.8 11.8 Supplementary 9.0 9.0 Total 20.8 20.8 Disbursed - 18.2 Original Credit - 11.8 Supplementary Credit - 6.4 Cancelled - 2.6 Co-financinl 8.0 12.0 Project Completion Date 11/30/76 05/15/77/c Proportion of Time Overrun (%)- 17% Economic Rate of Return (%)--Weighted Average 13 16 Parakou-Malanville Road: 13 16 Benin traffic benefits only 13 14 Benin and Niger traffic benefits together 17 24 Cotonou-Bohicon Road 17 15 Cumulative Estimated and Actual Disbursements (US$ million) FY74 FY75 FY76 FY77 FY78 FY79 FY80 FY81 FY82 Estimated 2.3 6.6 10.5 11.8 11.8 11.8 11.8 11.8 11.8 Actual 0 4.3 8.3 14.2 16.1 16.5 16.8 17.7 18.2 Actual/Estimated (%) 0 65 79 120 136 140 142 150 154 /a The project was scaled down. The Credit Agreement was amended on Decem- ber 21, 1978 to reflect the deferral of technical assistance for transport coordination, and on May 16, 1980 to finance the purchase of equipment for maintenance and for consultants' services under the Feeder Roads Project. lb By US Agency for International Development. /c Original project components only.  - iii - OTHER PROJECT DATA Original Item Plan Actual First Mention in Files - 1969 Negotiations Date 02/73 02/20-27/73 Board Approval Date Original 04/73 04/24/73 Supplementary 03/76 03/02/76 Credit Agreement Date Original 07/73 07/03/73 Supplementary 03/76 03/10/76 Effectiveness Date Original 10/31/73 03/28/74 Supplementary 06/10/76 10/05/76 Closing Date Original 06/01/77 04/29/77 Supplementary Not Specified 03/31/82 Borrower Government of Dahomey /a Executing Agency Department of Public Works, Ministry of Equipment Fiscal Year of Borrower January 1 - December 31 Follow-on Project Name Third Highway Project Credit Number 746-BEN Amount (US$ million) 10.0 Credit Agreement Date November 18, 1977 MISSION DATA No. of No. of Staff Date of Item Month/Year Weeks Persons Weeks Report Appraisal 02/72 1.5 3 4.6 05/73 Supervision I 05/73 1 1 1 08/73 Supervision II 02/74 2 1 2 03/73 Supervision III 11/74 1.5 2 3 12/74 Supervision IV 07/75 1.3 1 1.3 10/75 Supervision V 10/75 1.4 1 1.4 12/75 Supervision VI 04/76 0.3 1 0.3 04/76 Supervision VII 07/76 0.7 2 1.4 08/76 Supervision VIII 12/76 1.6 1 1.6 03/77 Supervision IX 04/78 1.8 2 3.6 06/78 Supervision X 10/78 1 2 2 01/79 Supervision XI 06/79 1.5 2 3 07/79 Supervision XII 01-02/80 2 2 4 03/80 COUNTRY EXCHANGE RATES Name of Currency (Abbreviation) CFA franc (CFAF) Year: 1973 (May) Exchange Rate: US$1 = 230 1974 ) US$1 = 241 1975 ) US$1 = 214 1976 ) (Average) US$1 = 239 1977 ) US$1 = 246 1978 ) US$1 = 226 1979 ) US$1 = 213 1980 ) US$1 = 211 /a The country's name was changed from Dahomey to the People's Republic of Benin in 1975.  - iV - PROJECT PERFORMANCE AUDIT REPORT BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) HIGHLIGHTS Credit 415-DA for US$20.8 million supported rehabilitation and maintenance programs to reduce vehicle operating costs on roads in Benin. The main objective of the project, which was financed by IDA, USAID, and the Government, constituted rehabilitation of the road (318 km) that is part of the "Benin Route" between the seacoast and landlocked Niger. The improved road was to assist in development of the northern part of Benin and to facili- tate transit traffic. While this objective generally was achieved, the design width of this road has proven to be insufficient for the higher than expected traffic that has developed. As a result, wear is occurring on the curves, and a continuous routine maintenance problem exists. Rehabilitation of two associated main road sections (17 km) encountered technical problems, which were overcome initially although after five years one section is again experi- encing distress. The institution building objective for road maintenance (continued from the previous project) was largely unfulfilled and that for transport coordination was deferred to the subsequent project upon the Government's request. This had detrimental effects for the project under review. The original project components were completed with a 17% time overrun. The actual project cost was US$32.2 million versus the appraisal estimate of US$21.1 million. The 53% cost overrun was due mainly to the civil works, which were affected by higher than expected bid prices, worldwide inflation following oil price increases, and in one case, unexpected technical problems. To finance the overrun, IDA provided a supplementary credit of US$9 million to augment its original credit of US$11.8 million, USAID provided additional finance of US$4 million to enhance its original loan of US$8 million, and the Government furnished additional funds of US$1.1 million on top of its initial provision of US$1.3 million. IDA's supplementary credit was in excess of the needs of the original project components, and it was ultimately determined that the surplus funds would be allocated to equipment for the shoulder maintenance brigade and to consultants' services under the Feeder Roads Project (paras. 44-52 and Annex D). The outstanding balance in the Credit Account of US$2.6 million, or nearly 30% of the Supplementary Credit, was cancelled. The weighted average economic rate of return at reevaluation is 16% compared with the appraisal estimate of 13%. The following points are of particular interest: - the methodology of economic analysis in the appraisal of the Para- kou-Malanville Road did not include the correct "without case" against which to evaluate alternative improvements and the procedure for estimating induced traffic yielded unsatisfactory estimates (paras. 25-29); - v - - the forecast traffic volumes on the Parakou-Malanville Road were seriously underestimated (paras. 30-33 and PCR, paras. 4.02-4.05, Table 7, and PCR Addendum); - the narrow range of probability analysis, the lack of a systematic approach to traffic counting and analysis, and the lack of origin and destination information were responsible for ineffective traffic forecasting (paras. 32-37); - the future use of the Parakou-Malanville Road by transit traffic for Niger, especially to serve the requirements of its uranium mines, will depend on competing investments, some financed by IDA (paras. 38-42); - the variations in estimates of vehicle operating cost savings prevent the development of confidence in the inputs for calculating project benefits (para. 43); - the attempts to apply the large remainder left unused under the credit after the original (revised) project components were com- pleted raises issues regarding effective control and use of surplus funds in a loan and regarding the application of loan funds within the Executive Directors' authorization (paras. 47-52); and - the interrelation of road, rail, and port projects will shortly call for IDA policy decisions on transport investments in Benin and its neighboring country, Togo (paras. 53-55). - 1 - PROJECT PERFORMANCE AUDIT MEMORANDUM BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) I. BACKGROUND The Transport System 1. The main feature of Benin's 1/ transport system is the south- north axis, which now consists of 440 km railway from the capital and main port of Cotonou to Parakou and a 318 km road from Parakou to Malanville on the Niger border. Another road runs parallel to the railway, but it is not yet completely paved. The south-north axis, known as the "Benin Route," links Cotonou with Niamey, the capital of Niger. It is the main route for imports into the western and central parts of landlocked Niger, eastern Mali, and also carries considerable traffic within Benin. The south-north axis is connected to important secondary and feeder roads. A competitive and only slightly longer route, known as the "Togo Route," has been completed recently, partly with IDA assistance, between Lome, the capital and main port of Togo, and Niamey. IDA Support for Highways 2. Highway projects supported by IDA have evolved from the Dahomey Land Transport Survey of 1970, for which the UNDP provided financing and the Bank Group acted as executing agency. The Survey's recommendations led to the Highway Maintenance and Engineering Project (referred to in this report as the First Project, Credit 215-DA for US$3.5 million of September 1970), which OED has already evaluated,2/ and to the Second Highway Project (Credit 415-DA for US$20.8 million of July 1973 and June 1976), which is the subject of this audit. 3. The main component of the Second Highway Project was to constitute rehabilitation of the Parakou-Malanville Road (about 80% of the estimated pro- ject cost). Rehabilitation of two short sections of the associated Cotonou- Bohicon Road was to comprise the second largest component.3/ PCR, para. 2.05 provides details about the other components (consultants' services for 1/ The country's name was changed from Dahomey to the People's Republic of Benin in 1975 (PCR, para. 1.02). 2/ Project Performance Audit Report No. 1768 of October 25, 1977. 3/ Most of the rest of the Cotonou-Bohicon Road has been improved under the Third Highway Project (Credit 746-BEN). Of the Bohicon-Parakou extension of this Road, the Bohicon-Dassa Zoume section is being paved with financ- ing from the European Development Fund (EDF), while the Dassa Zoume- Parakou section has been the subject of a feasibility study, with posi- tive results, also by EDF. - 2 - ccnstTrction supervision, a highway maintenance program, and transport coordi- nation as well as procurement of materials). The total cost of the project as appraised in February 1972 by IDA and USAID, which was to act as cofinancier, amounted to US$21.1 million. IDA was to provide US$11.8 million, USAID US$8.0 million, and the Government US$1.3 million toward the cost. The Credit Agreement of July 1973 only became effective in March 1974 because of a delay in meeting cross-effectiveness conditions for the USAID loan as a result of a dispute over the breakpoint between the IDA-and USAID-financed sections of the Parakou-Malanville Road. This delay did not seriously hamper smooth execution of the project. However, problems related to bidding procedures did do so (para. 5 and PCR, paras. 3.07-3.08), as they held up contract award. 4. Subsequent projects included the Feeder Roads Project (Credit 717-BEN for US$5.5 million of June 1977), the Third Highway Project (Credit 746-BEN for US$10.0 million of November 1977), and the Fourth Highway Project (Credit 1142-BEN for US$11.3 million of May 1981). These projects are gen- erally supporting road construction, rehabilitation, and maintenance; techni- cal assistance; and preinvestment studies. In addition, two griculture projects included construction and rehabilitation of feeder roads- II. PROJECT RESULTS Implementation 5. The project was carried out during a period of profound political and economic change in the country (PCR, para. 1.02). Only the institution building objectives were adversely affected by these changes (PCR, para. 6.01). The civil works objectives were achieved, despite the far from satis- factory bidding procedures (PCR, paras. 3.02-3.04). The Government and the consultants who assisted in the bid analysis selected the second lowest from among the three lowest evaluated bids. This selection was partly because the unit price of the lowest bidder was considered unrealistic and partly because the crushed stone base proposed by the second lowest bidder was preferred (for technical reasons) over the soil cement base proposed by the lowest bidder. 6. Since the bids had been called for either type of base, IDA asked the Government to justify the economics of selecting the more expensive type. To allow time for this, the Government asked the bidders to extend the valid- ity of their bids by one month. Only the highest bidder, who had proposed a soil cement base, agreed. The Government eventually accepted this bid, even though it was US$150,000 higher than the lowest bid and did not provide for the preferred type of base. Clearly, it would have been better for the Government and the consultants to have decided on the technical merits of crushed stone versus soil cement base before calling for bids. Although the difference in the costs of the two types of base would have been small, soil cement is more difficult to widen if subsequently found to be necessary (PCR, paras. 3.10-3.11). The contract for the USAID-financed section of the road 1/ Hinvi Agricultural Project (Credit 144-DA for US$5.2 million of March 1969) and Zou Bourgou Cotton Project (Credit 307-DA for US$6.1 million of May 1972). - 3 - required a crushed stone base. It is noted that the bids for the USAID- financed section turned out to be 36% above the appraisal estimate because of worldwide inflation and USAID increased its loan to US$12.0 million to cover the higher cost of work. 7. Contracts specifying a 30-month execution period were signed on March 29, 1974 for both the IDA-and USAID-financed sections of the Parakou- Malanville Road. Work on each section was completed only two months late in relation to the orders to proceed, despite political changes, closure of the border with Nigeria in March 1976, and interruption of bitumen supplies. The USAID-financed section was completed satisfactorily in February 1977 and the IDA-financed section in May 1977. 8. After completion of the IDA-financed section, reservations devel- oped concerning the appropriateness of the design standards. The Credit Agreement included a schedule of design standards that called for the Parakou- Malanville Road to be constructed to a width of 9 m, including a 6 m pavement and 2 x 1.5 m shoulders. The PCR (para. 3.10) explains that pavement widening was not applied at the curves even though about 20 curves had been scheduled at appraisal to be realigned on a larger radius. Also, higher than expected traffic, particularly heavy vehicles, has caused wear on the pavement/shoulder junction, which requires continuous routine maintenance. Equipment for shoulder maintenance brigades was added to the project to help remedy this problem (paras. 13, 21, and 48). The PCR draws the lesson from this that "for roads carrying heavy vehicles at high speeds a 6 m pavement width is inade- quate." The audit discusses, in paras. 25-42, the lack of consideration of a 7 m pavement in the economic analysis and of effective traffic forecasting. 9. Rehabilitation of the two sections of the Cotonou-Bohicon Road was completed in February 1977, or 31 months from the order to proceed rather than 12 months as specified in the contract. The delay was due to unexpected technical problems on the Godomey-Abomey Calavi section as a result of two serious ground failures that destroyed both the road where it serves as a causeway and the main water supply pipeline of Cotonou. The PCR (paras. 3.12-3.19) fully reviews these problems, which Bank staff played an important role in resolving. According to the PCR, the consultants did insufficient testing and lacked technical leadership. Although the quality of the road surface is not up to design level, it is considered to be acceptable in the circumstances. The Sehoue-Zakpo section also experienced some problems. When cracking of the sand asphalt surfacing occurred near Zakpo, the contractor agreed to repair the cracks and to accept a two-year maintenance period until July 1979. Since then, deformation and edge cracking due to foundation settlement have occurred; repairs are proposed under the Third Project (PCR Addendum, para. 3.01). 10. The project under review provided for the continuation of technical assistance in road maintenance begun under the First Project. This took the form of 83 man-months of technical assistance and the purchase of materials worth US$250,000. However, as the PCR (paras. 3.21-3.22) describes, the physical performance of the maintenance program deteriorated. This was largely because of the lack of counterpart funds. Road maintenance expendi- tures remained constant in current prices, despite severe inflation after 1973, and the road fund tax was not revised. At least half the equipment was immobilized due to lack of spares and fuel as well as inexperienced operators. Less than a third of the funds for materials were used by the time the consul- tants left. In these circumstances, the consultants were not able to achieve their objective of strengthening the management of the road maintenance organization and had to limit their work to planning maintenance and quantify- ing needs. This was valuable preparation, however, for the Third Project. Further support for road maintenance is included in the Fourth Project. 11. The subject project also provided for technical assistance in transport coordination to the Directorate of Land Transport (DLT), which had been established in 1969. Similar assistance under the First Project was not effective because of insufficient qualified local personnel, funds, and sup- port. At the Government's request, the assistance planned under the Second Project was deferred, pending the transfer of responsibility for transport coordination to another body, the Directo ate of Studies and Planning (DSP) in the reorganized Ministry of Transport.' The effects of this deferral are discussed in paras. 36-37. The Third Project provides once more for this assistance. Project Cost and Timing 12. The actual project cost was US$32.2 million, or 53% more than the appraisal estimate of US$21.1 million (PCR Addendum). IDA provided US$17.7 million (50% more than the appraisal estimate of US$11.8 million), USAID provided US$12.0 million (50% above the estimate of US$8.0 million), and the Government provided US$2.4 million (85% over the estimate of US$1.3 million). The actual and estimated costs are not strictly comparable, however, because the original project was scaled down (para. 11). The cost estimated for the deferred technical assistance for transport coordination was low (US$161,000) but, when subtracted from the appraisal estimate of the project cost, does not result in a significant increase in the overrun. 13. The likelihood of a US$10.1 million overrun for the IDA-financed portion of the original project items, as a result of inflation, dollar devaluation, and quantity increases and additions in civil works, was identi- fied in mid-1975. To cover the overrun, in March 1976, IDA agreed to provide a Supplementary Credit of US$9.0 million and the Government additional finance of US$1.1 million. In fact, IDA actually provided only US$4.9 million and the Government US$1.0 million. Part of the surplus of US$4.1 million in the Supplementary Credit was used to purchase equipment for shoulder maintenance (US$360,000) under this project and to finance consultants' services under the Feeder Roads Project (in the amount of US$988,000, of which IDA financed US$889,000 and the Government provided US$99,000L'). The overestimation of supplementary credit needs is further discussed in paras. 44-52. 1/ The Credit Agreement was amended accordingly on December 21, 1978. 2/ The Credit Agreement was amended accordingly on May 16, 1980. - 5 - 14. All original project components actually executed were completed in May 1977 with only a 17% time overrun (vis-a-vis the Project Completion Date of November 1976 specified in the Credit Agreement). The additional com- ponents were finished in 1981. Disbursements 15. Because of the delay in the start of road construction due to bidding procedures, no funds were disbursed in FY74 and disbursements in FY75 and FY76 were 65% and 79% of appraisal estimates, respectively. The original credit was fully disbursed, while only US$6.4 million of the supplementary credit was drawn down. The US$2.6 million balance in the Credit Account was cancelled in March 1982. Government's Compliance With Lending Covenants 16. The Government did not comply fully with the covenants in the Credit Agreement. In particular, it did not provide adequate road maintenance (para. 10) or keep adequate statistical records (para. 35). Moreover, the Government did not impose penalties for trucks exceeding axle load limits, with the result that overloading continues even though weighbridges have been installed (PCR, para. 3.34). Also, it did not furnish the expected detailed program of measures to strengthen the DLT (paras. 11, 36 and 37) or to improve the management of the Organisation Commune Benin-Niger (OCBN) railway, for which assistance was provided under the Port of Cotonou Project (Credit 826-BEN for US$19.3 million of October 1978). 17. The requirement under the Credit Agreement that, prior to under- taking any extension of the Cotonou-Parakou railway line, the Borrower should furnish a feasibility study establishing its economic justification, has been carried out. IDA has not, however, accepted the findings of the study (para. 53). 18. The failure to comply fully with the covenants may be ascribed in large measure to the severe inflation from 1974 onwards and to political changes in the portfolios of ministries and their personnel. Economic Reevaluation 19. In the Appraisal Report, costs and benefits were quantified sepa- rately for rehabilitation of the Parakou-Malanville Road (including the USAID- financed portion) and of the Cotonou-Bohicon Road. The Parakou-Malanville Road (318 km) yielded a rate of return of 17% for Benin and Niger together and of 13% for Benin alone. Benefits for the two countries together consisted entirely of savings in vehicle operating costs (voc), while those for Benin alone comprised low voc savings to which were added estimated net receipts in foreign exchange by the port, the stevedoring companies, the forwarding agents, the railways, and the truckers as a result of the incremental transit traffic. The developmental effects of the rehabilitated road on domestic traffic were based on the assumption that generated traffic would vary from 10% to 20% of "normal" traffic depending on road section and would develop over the five-year period following completion of this project. -6- The reestimated return is 24% for Benin and Niger together, based on traffic data collected in 1981 and actual construction costs. Clearly, the returns to Benin alone are difficult to distinguish because all heavy trucks cannot be assumed to be going through to Niger. However, taking into account savings in operating costs of domestic vehicles in the same proportion to total savings as estimated in the PCR (60.5%) and excluding net foreign exchange receipts from transit traffic, the return to Benin alone would be 14%. On the basis of individual road sections, the heavily trafficked Para- kou-Ndali section, which appears to have been underdesigned (para. 31 and PCR, paras. 3.10-3.11), and the Ndali-Malanville section, which serves the lower traffic needs of northern Benin and Niger, yield rates of return to Benin and Niger together of 28% and 23%, respectively. The returns to Benin alone are 18% and 13%, respectively. The generally higher reestimated returns vis-a-vis appraisal estimates are due to faster than expected development of both Benin and Niger traffic, which outweighed the substantial increases in construction costs. 21. The reestimated returns may be reduced when account is taken of the need to repair damaged pavement, which already necessitates the use of shoulder brigades (para. 8), or to widen pavement and embankment. The PCR (para. 3.11) considers, however, that certain measures can be taken to lessen shoulder damage on the existing road, e.g., center-line and edge marking and regular shoulder maintenance. Moreover, the equipment for shoulder mainte- nance purchased under this project (paras. 8, 13, and 48) should help improve the situation. A similar situation applies in the case of the Togo Second Highway Project.1L/ 22. The reestimated returns also may be reduced as a result of the arbitration claim against the Government of Benin by the USAID-financed contractor (PCR, para. 3.08). The contractor claimed US$7.5 million and is reported to have been awarded US$4.0 million. This would reduce the returns to Benin and Niger together to 23% (a one percentage point reduction) and to Benin alone to 12% (a two percentage point reduction). 23. With regard to the two sections of the Cotonou-Bohicon Road, the Appraisal Report gave economic returns of 19% for the Godomey-Abomey Calavi section and 14% for the Sehoue-Zakpo section (17 km). Considering 1976 traffic counts, higher vehicle operating cost savings, and the 69% increase in construction costs over appraisal estimates, the PCR reestimated the economic return at about 15%. Future traffic is difficult to forecast because of fluctuations in traffic counts in the years since appraisal and the absence of counts in 1979 and 1980. On both sections, traffic rose sharply up to 1974, especially on the Godomey-Abomey Calavi section, but recent counts show marked declines. When the Dassa Zoume-Parakou section is paved (para. 3), however, and joins these sections as part of the "Benin Route," traffic almost cer- tainly will rise faster than originally expected. The viability of the roadworks will not be in doubt, provided that the higher traffic levels do not unduly shorten the life of the road. Also, the PCR suggests that, as traffic 1/ PPAR No. 3987, Togo Second Highway Project (Credit 450-TO of November 1973 for US$8.7 million), June 30, 1982. - 7 - increases, settlement of the road on the Godomey causeway will have to be monitored continuously. Evidence exists of significant recent failures on the Cotonou-Bohicon Road, which soon may require extensive rehabilitation. In the event, the economic return on the road will be considerably reduced. 24. Since the Cotonou-Bohicon Road was a relatively small improvement scheme giving returns similar to those of the main rehabilitation scheme, the weighted average returns for the whole project are the same as for the Para- kou-Malanville Road at appraisal and at reevaluation (13% and 16%, respec- tively). III. ISSUES Shortcomings in Methodology of Economic Analysis 25. The methodology used in the appraisal of the Parakou-Malanville Road, in retrospect, had several shortcomings. The "without case" against which to evaluate alternative improvements was not correctly identified and the procedure for estimating induced traffic yielded inaccurate estimates. As shown below, this had an effect on the choice of design standards for the road. 26. The Land Transport Survey of 1970 selected rehabilitation of the Parakou-Malanville Road, linking the railhead with the Niger border, as an "extremely urgent project." The Survey gave "very careful study" to the possibility of upgrading to two lanes the then badly deteriorated one-lane paved road, but could not justify the higher standard because of low traffic volumes. A benefit-cost ratio of 6.7:1 was calculated for improvement of the one-lane and of 0.2:1 for the marginal improvement to two lanes. The possibi- lity of "downgrading" the road to two-lane gravel standard was studied, giving a benefit-cost ratio of 7.6:1, and this was curiously described as the "second best alternative." 27. Under the First Highway Project, consultants carried out detailed engineering for rehabilitation of the road to its original one-lane paved standard. Technical surveys revealed that deterioration was more extensive than had been expected. The consequent higher cost of rehabilitation and common sense judgements of the consulting engineers and of the Bank engineers involved led to serious questioning of the decision to rehabilitate. Accord- ingly, the same consultants who had carried out the Land Transport Survey were asked in 1971 to carry out a complementary economic study of the road. They found that traffic would grow faster than originally expected and that a two-lane paved road would be economically viable. However, the one-lane alternative still had a higher benefit-cost ratio than the two-lane alterna- tive (5.01:1 compared with 3.3:1). The economic returns were 41% for one-lane compared with 30% for two-lanes. The consultants concluded that construction to a two-lane standard was justified. 28. The fact that the one-lane road had a higher economic return still caused some concern, but fortunately common sense prevailed over somewhat misapplied economic analysis. This is a clear case where rehabilitation of - 8 - the one-lane road or an accelerated maintenance program should have been taken as the "without case" in the benefit-cost analysis. Either was the minimum work required for Benin to continue to have a major road to the agricultural north of the country and to serve as a major transit route to Niger. The "without case" should then have been tested against alternative levels of improvement - two-lane gravel and two-lane paved, with alternative pavement and formation widths. A two-lane gravel road appears to have been the best solution in the original analysis of 1969, although it was not recognized as such, and, with higher traffic forecasts, a two-lane paved road clearly was feasible in 1971. In retrospect, the economic analysis should have covered a 7 m pavement and a wider formation at least for part of the road. There is often a reluctance among consulting engineers to consider alternative stand- ards and, in this case, serious detrimental effects resulted for the project. A similar situation occurred under the Togo Second Highway Project.1/ 29. The appraisal methodology for treating the developmental effects of the road also was not very satisfactory. Induced traffic was estimated as a percentage of normal traffic (10%-20% depending on road section) and bene- fits were established by assigning the voc savings of normal traffic to induced traffic. This procedure provided a rather unsatisfactory estimation of the volume of, and understated the benefits of, induced traffic. Domestic traffic on the road inside Benin might have been estimated more accurately by deriving induced traffic from a calculation of the transport effects of economic development in the area, although this would have required more research time. Appraisal Underestimation of Traffic Development on the Parakou-Malanville Road 30. Actual traffic for the opening year on the Parakou-Malanville Road was considerably above the forecast in the appraisal. The appraisal of 1973 forecast average daily traffic volumes for 1976, the opening year of the road, at 51-109 vehicles up to 10 tons (passenger cars, pick-ups, and trucks), depending on the section of the road, and 54 heavy trucks carrying Niger transit traffic the entire length of the road. The PCR of 1979 gave the latest available average daily traffic count in 1977 over the whole road as 183 vehicles up to 10 tons and 67 heavy trucks. Total traffic was thus 87% higher than expected. Domestic traffic of light vehicles increased by 129% over expectations, while heavy vehicles apparently increased only 24%. A similar underestimation of traffic development occurred in the case of the Togo Second Highway Project.2/ 31. These counts, although not fully reliable, strongly confirm very heavy local traffic on the first 60 km of road (Parakou-Ndali) (para. 8). They suggest that the shoulder damage that is occurring is being caused by friction between heavy trucks and lighter traffic. Not all heavy vehicles were Niger transit traffic, but they apparently more than doubled after 1972 instead of remaining more or less stable, as assumed in the appraisal. The 1/ PPAR No. 3987, Togo Second Highway Project, June 30, 1982. 2/ Ibid. - 9 - latest available traffic count, in July 1981, shows a marked increase of 446 vehicles up to 10 tons and 138 heavy trucks between Parakou and Ndali and 228 vehicles up to 10 tons and 106 heavy trucks at Guessou Sud, the first station beyond Ndali. 32. The very rapid growth in domestic traffic could not have been pre- dicted from the traffic counts of the 1960's. Although light vehicles rose slightly over the period, trucks from 2-10 tons actually declined. Heavy vehicles did, however, increase 7.7% per annum in the period 1961-71 and the error in not anticipating this trend was not due so much to underestimating the Niger uranium boom, as suggested in the PCR, as to accepting, with the encouragement of various consultants' reports, that Benin would lose much of its Niger transit traffic to other routes, especially the Niger River. An elaborate probability analysis was carried out at appraisal but within far too narrow a range. The best estimate, which assumed that Benin would lose 70% of its original share of Niger traffic by 1990, was given a probability of 40% and the lowest estimate of Niger River traffic (only 14% lower than the best estimate) was given a probability of 30%. The probability of no traffic using the Niger River because of technical problems, which is what actually happened, was not assessed. 33. In 1970, a Dutch consultant had determined that transport on the Niger River from Niamey to Yelwa in Nigeria (north of the Kainji Dam) was technically feasible, provided a rock clearance program was carried out and the Malanville bridge was modified. However, at that time, there had not been a pilot scheme to pinpoint the operational difficulties on the river. Canadian consultants subsequently identified navigability problems below the Kainji Dam, which could only be resolved by building a barrage at Jebba. Construction of this barrage is included in Nigeria's development plans, but the timing has not yet been determined. 34. A wider range of probability or sensitivity analysis, taking a sceptical view of an untried transport route such as the Niger River, would have provided more realistic transit traffic forecasts for the economic analysis. The failure to estimate the even faster growth in domestic traffic might have been avoided by a different methodological approach at appraisal (para. 28). 35. The lack of a systematic approach to traffic counting and analysis and the lack of origin and destination information also was responsible for the ineffective traffic forecasting. Traffic counts have been carried out in Benin twice a year since 1961, except in 1979 and 1980. However, data do not appear to have been collected continuously for all stations. Very little analysis of these figures appears to have been carried out to distinguish between through, local, and diverting traffic. Moreover, origin and destina- tion information was almost completely lacking. Consultants carried out limited interviews in 1969 and 1971 but, because of insufficient time and inadequately trained local personnel, did not learn much. 36. The Land Transport Survey had recommended technical assistance for transport coordination, which became a component of the project under review. The requirement to collect data to help assess the transport network - 10 - a covenant in the Credit Agreement. To facilitate transport coordination as well as monitoring of investment projects, establishment of a system of traffic counts and origin and destination surveys and training of Beninese staff to provide continuous data analysis is essential. Of particular impor- tance for the country, because of its heavy reliance on transit traffic, is information on growth trends, types and weights of vehicles using the roads, and origins and destinations of road traffic. 37. The serious effects of the failure to set up an effective transport coordination agency have already been felt. Lack of reliable traffic counts and origin and destination information contributed to the underdesign of a section of the Parakou-Malanville Road. Moreover, in the future, when a paved road exists between Parakou and Cotonou, road-rail competition will take place for the same traffic and an even greater need will exist for data collection and analysis to facilitate transport coordination, planning of investments, and formulation of road/rail tariff policies. 38. As stressed in the presentation of the project to the Executive Directors, rehabilitation of the Parakou-Malanville Road was not only impor- tant to Benin's economic development but also critical to Niger's. According to Port of Cotonou statistics, the best source of information on this subject, total Niger transit traffic through Benin grew by 6.2% per annum between 1970 and 1980 to 296,000 tons. This amounted to about 80% of total Niger traffic in the early 1970's and to about 60% later in the decade. The main alterna- tive routes for western and central Niger are by road through Upper Volta to Lome in Togo (part of which was paved under the Togo Second Highway Project) and by road and rail through Upper Volta to Abidjan in Ivory Coast. The "Nigerian Route" has developed not on the Niger River as forecast but on Nigerian roads and railways. It was used as a natural export route for groundnuts from Eastern Niger in the early 1970's until groundnut exports ceased due to drought and now is used mainly for some petroleum product requirements of the uranium mines in northern Niger. It has been an irregular route because of lack of confidence in Niger about the security of Nigeria's railways and ports. The "Abidjan Route" has not developed as forecast in the appraisal because of the much greater distance and cost, pilferage on the railway, and capacity problems in the Port of Abidjan. 39. The main competition for the "Benin Route" has in fact developed from the "Togo Route" which, although slightly longer (1,240 km to Niamey as against 1,058 km) and costlier, takes 3 days against 15 days on the "Benin Route" where road/rail transfer at Parakou is required. Also, the Port of Lome has offered better services to shippers than the Port of Cotonou since nationalization in 1973 of most port services in the latter. Niger transit traffic through Lome increased from 19,000 tons in 1975 to 93,000 tons in 1980. 40. Considerable uncertainties remain about future transit traffic to and from Niger, the principal ones being how fast uranium mining will grow in Niger and whether other transport routes will be improved. Studies in 1970 and 1978 of the alternative routes helped considerably to form a regional view of alternatives, although the study forecasts may be disputed. - 11 - 41. The importance of Niger's uranium traffic was fully recognized in the appraisal of this project. Niger now has the fifth largest reserves among uranium producing countries and ranks fourth in world production. In the appraisal, each ton of uranium ore exported was estimated to require 24 tons of imports (petroleum products, sulphur, and other supplies). Exports were forecast at 1,500 tons by 1974 and 3,000 tons in 1980 with corresponding imports of 35,500 tons and 73,000 tons, respectively. As it happened, produc- tion in 1980 was reported to be 4,100 tons from the two mines, Somair and Cominak, presently in operation. A third mine, Societe Miniere de Tassa n'Taghalque (SMTT), estimated to produce a further 1,500-2,000 tons of output, is due to start up in 1982 but further development will depend on world price levels, which declined recently. Production might begin by 1985 at the large Imourarem mine (1,500-3,000 tons) and before 1990 at the Azelit mine (1,500 tons) and one other mine if prices and demand are favorable. The maximum output of uranium by 1990 is now foreseen to be 12,000 tons, with a cor- responding import requirement of 290,000 tons if the 1:24 ratio were main- tained. However, technologies change and this ratio might not remain valid. 42. These tonnages are likely to be divided mainly between the "Benin Route," the "Togo Route" and the "Nigerian Route." A considerable portion of petroleum product requirements already use the "Nigerian Route" because of that country's large oil resources and growing number of refineries. Although the Niger government has decided to pave the Tahoua-Agadez-Arlit Road, thus giving the mines a paved route to the Benin border, the petroleum companies are reported to be saying that they still will use the unpaved Agadez-Zinder route to Nigeria, even though a 650 km detour is necessary in the rainy season. Eventually, the Agadez-Zinder Road is due to be paved as part of the Transsahara Road. Thus, Benin might lose much of its petroleum traffic as Nigerian refinery capacity increases. Exports of uranium ore and inputs other than petroleum probably will continue to use the "Benin" or "Togo Routes," however, because of doubts about the security of the "Nigerian Route." In the longer term, a significant part of uranium-related traffic on the "Benin" and "Togo Routes" may divert to the Transsahara Road, either southward through Nigeria or northward through Algeria. Variation in Estimates of Vehicle Operating Cost Savings 43. Vehicle operating cost savings are major elements in the benefits identified in appraisals of road projects. It is disturbing that these are recalculated from time to time by different consultants, sometimes on differ- ent bases and for different vehicle types, and that considerable differences exist in data for the same country and between data for similar neighboring countries. To permit greater confidence in these inputs, data (empirical if possible) should be collected continuously and comparisons should be made between countries with similar conditions. This is another subject that should be studied by a transport coordination agency in Benin to ensure an accurate and consistent approach. Some examples of the variations in Benin itself and in Togo, a similar neighboring country, are shown in Annexes A-C .1/ 1/ PPAR No. 3987, Togo Second Highway Project, June 30, 1982. - 12 - Overestimation of Supplementary Credit Needs 44. An IDA Supervision Mission of July 1975 determined that the cost of the IDA-financed portion of the project would rise to about US$23.2 million, representing a cost overrun of US$10.1 million (77%). Some 76% of the overrun was attributed to existing and projected inflation, 4% to dollar devaluation, and 20% to quantity increases and additions to civil works. The possibility of an overrun had not been mentioned until that time. One reason appears to have been that the consultant's monthly progress reports omitted final cost estimates until March 1976. The Supplementary Credit of US$9.0 million provided by IDA to cover the higher project cost (para. 13) proved to be in excess of actual needs. 45. Upon completion of the original project components in mid-1977, it became apparent that a significant surplus (initially estimated at US$3.7 million) would be left in the Credit Account. IDA's supervision mission of December 1976, which immediately preceded this completion, did not report a downward revision from US$23.2 million in the estimated cost of the IDA- financed items. Only as a result of the next mission, 16 months later, was that cost scaled down to US$19.3 million. Ultimately, the final cost of the original items was US$16.7 million, representing an overrun of US$4.9 million vis-a-vis the appraisal cost estimate. 46. Overestimation of the project cost is explained by several factors, including: (a) the generous safety margin allowed for the cost of repairing the causeway on the Cotonou-Godomey Road; (b) the modification of the embank- ment design for the causeway, which reduced the cost; (c) the lower than estimated cost to repair the main water supply pipeline of Cotonou; (d) the slowdown in the rate of inflation; and (e) the improvement in the exchange rate of the US dollar. 47. The record since mid-1977 discloses considerable reluctance on the part of the Government and the IDA to cancel the surplus funds and extensive efforts by both parties to find alternative uses. The IDA felt that cancella- tion should be avoided in view of the potentially negative effects on rela- tions with Benin and of the recent improved economic performance of the country. At the same time, the IDA repeatedly considered the need to cancel the surplus funds. 48. Among the proposals for using the surplus funds were those listed in Annex D. In the end, some US$360,000 was allocated to equip the shoulder maintenance brigade and some US$900,000 to support consultants' services under the Feeder Roads Project (para. 13). 49. The history of the initial underestimation and subsequent over- estimation of the costs of this project, the Supplementary Credit, and the surplus funds in the Credit Account lead to several interesting observations and questions. The proportions of the overrun in total costs and the subse- quent underrun were significant and closer surveillance would have provided earlier notice of their occurrence. Nevertheless, the amounts were substan- tial only in relation to the very limited resources of Benin and would, in larger countries, have readily been absorbed by the Government. Because of - 13 - the limited resources, IDA felt constrained to assist the Government in financing the extra costs and to provide a generous allowance so as to fore- stall another Supplementary Credit. This, in turn, created a surplus that neither party wanted to cancel because of the stringency of Benin's resources. 50. The very size of the surplus increased the reluctance to cancel it. Had the surplus been modest, disposition would have been more manageable. The overestimation of costs in effect caused nearly as much difficulty as the underestimation. Efforts to improve the accuracy of cost estimation and to monitor costs continuously thus are important to smooth project implementa- tion, particularly in small countries. 51. The surplus also presented both the Government and IDA with diffi- cult choices of fiscal and project management. These funds emerged fortui- tously and unexpectedly and were not subject to the normal budgetary and programming processes. Nor was there an overall plan for the use of the US$3.5 million that ultimately became available. Without such a plan, it became difficult to set targets for efficient application and controlled expenditure of the surplus. Finally, use of the funds was constrained by extremely restrictive parameters with regard to timing, objectives, and so forth. These constraints presented formidable obstacles to efficiency and imposed significant additional burdens on the technical staff of both the Government and IDA. 52. Finally, the history of this case raises issues regarding control of credit funds within the Executive Directors' authorization. Some of the difficulties could have been avoided had the Description of the Project been less specific or had that Description been interpreted broadly, not literally. The audit does not question the formulations which were adopted or the inter- pretations placed on them. However, it does suggest that policy guidance may be required on this aspect of IDA operations. IV. FUTURE IDA SUPPORT FOR TRANSPORT PROJECTS INVESTMENTS 53. The traditional dependence of Niger traffic on the OCBN is being eroded by political interest in route diversification, the relative speed and efficiency of providing road transport through Togo, and the low productivity and rising transfer costs of the public sector controlled OCBN route in Benin. Past insistence by Benin on use of the railway versus the closed nature of access to Cotonou port have served to reinforce these trends. Governmental interest and effectiveness in dealing with these issues and improving the private sector market environment in Benin are likely to influence IDA's policy decisions in Benin and Togo. Two interrelated projects in Benin dealing with this Niger traffic are discussed below: A. A feasibility study was carried out in 1979 for the extension of the OCBN railway from Parakou to Niamey in Niger (about 600 ki). If built, this railway would run parallel to the road rehabilitated under the present project and for this reason the Credit Agreement contained a covenant requiring the study to establish, to the mutual satisfaction of the Government and IDA, the economic justification - 14 - of such an extension. The study found the economic return would be 12.7%-13.1%. However, an IDA review considered the forecasts highly optimistic, the costs too low, and the methodology too limited in focus. The review recalculated the economic return at 4%-5%. Following the uncertain nature of market demand for uranium stemming from advances in the recycling process, the Governments of Niger and Benin have virtually abandoned serious consideration of this pro- ject. The only possible future development that could change this assessment would be if phosphate deposits on the Benin-Niger border are found to be exploitable on a large scale. B. A feasibility study for paving the Dassa Zoume-Parakou Road (about 250 km) was recently carried out under EDF finance and the project was found to be justified (para. 3). This is the last remaining unpaved section of the road from Cotonou to the Niger border and would thus result in road competition for the railway. Previously, Government policy had been to keep the road to a gravel standard appropriate to local traffic and to restrict the use of joint road-rail bridges to protect the railway. However, the developmental needs of the southern part of the country and the desire to unite the north and the south have led to a reversal of this policy. IDA is interested in cofinancing this project with EDF, but, having learned the lessons of the project under review, it has recommended a 7 m pavement and 2 x 1.5 m shoulders. It also proposes an axle load study for the Parakou-Malanville Road to serve as the basis for determining appropriate pavement strength for the Dassa Zoume-Parakou Road. 54. The proposed project for paving the Dassa-Zoume-Parakou Road would have important repercussions: (a) on the railway, which would undoubtedly lose freight and revenues; (b) on the Port of Cotonou, which will have to change its handling procedures, infrastructure, and equipment to cope with existing or higher traffic volumes; (c) on the OCBN system, which operates a virtual public sector monopoly covering all modes and services, with the Niger Govern- ment Transport Co. (NITRA) alone having a share of Niger traffic mainly between Parakou and Niger; and (d) on the Port of Lome in Togo, which would compete with Benin's new all-road route for available traffic. The "Benin Route" would have the advantages of distance (a saving of 182 km) and of not traversing a third country (Upper Volta) that levies taxes at its borders. The feasibility study assumes somewhat unrealistically, that improvements to the "Benin Route" would divert from the "Togo Route" freight amounting to 174,000 tons by 1984 and 236,000 tons by 1990, necessitating two more berths at the Port of Cotonou. 55. Where new projects like the above result in lower transport costs for landlocked countries such as Niger, taking transport time into account, diversions of traffic may rapidly occur. Other considerations are: (a) a desire to diversify outlets even at greater cost, e.g., to Abidjan via road in Upper Volta to the RAN railway, to Dakar via road and the Bamako-Dakar railway, and to Nigeria or Algeria via the Transsahara Road; (b) political and financial attraction to other Francophone countries and; (c) company decisions on preferred routes, e.g., by uranium mining companies, petroleum companies, - 15 - and shipping lines. In the longer run, however, there may be a greater attraction of trade with the larger countries, e.g., Nigeria and Ivory Coast, which are building up their own manufacturing and service industries. 56. In the meantime, questions will arise shortly about the investment and technical assistance requirements of the Ports of Cotonou and Lome, about road and rail tariff policies, about possible disinvestment in the OCBN and about assistance to the road trucking industry. Other important policy ques- tions concern transit traffic. Should Benin pay entirely for road improve- ment that is largely justified by the voc savings of foreign trucks? Should Benin follow Upper Volta's lead in levying transit taxes at its border? These problems emphasize the importance of the technical assistance in trans- port coordination, which was intended to be provided under the present project and has since been transferred to the Third Project (paras. 11 and 36-37 above). V. CONCLUSIONS 57. The main objectives of the project under review were met in that the Parakou-Malanville Road was satisfactorily rehabilitated and has been able to serve higher than expected traffic volumes, apart from problems of shoulder damage that gives evidence of underdesign on certain sections. The institu- tion-building objectives regarding road maintenance and transport coordination were not successfully accomplished, mainly because of the Government's request to drop them. Clearly, there was more Government support for the civil works components than for the institution-building aspects. The reevaluated economic rates of return indicate that the civil works components were justi- fied, despite large construction cost overruns due to inflation and technical problems. The major part of the cost increase was due to worldwide inflation, which could not have been foreseen in 1973, but some technical problems could have been avoided if identified earlier by the consultants. Overestima- tion of further financing needs resulted in surplus funds in the supplementary credit account, about 30% of which ultimately was cancelled after expenditure on additional items.  - 16 - ANNEX A PROJECT PERFORMANCE AUDIT REPORT BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) Benin and Togo - Vehicle Operating Costs Excluding Taxes - 1972 Toga/a Benilb Gravel Good "Ruined" Good or Poor Paved Paved Paved Paved Road Road Savings Road Road Savings ---------------------(CFAF per km)--------------------- Cars and Pick-Ups n.a. n.a. n.a. 16.4 23.3- 25.5 6.9 -9.1 Cars 14.2 31.4 17.2 n.a. n.a. n.a. Pick-Ups 14.0 29.4 15.4 n.a. n.a. n.a. Trucks (2-10 tons) n.a. n.a. n.a. 32.4 51.7- 57.4 19.3-25.0 7 ton Trucks 27.4 57.6 30.2 n.a. n.a. n.a. Heavy Trucks n.a. n.a. n.a. 74.8 125.2-139.9 50.4-65.1 11 ton Trucks 48.9 99.7 50.8 n.a. n.a. n.a. Truck-Trailers 84.9 170.5 85.6 n.a. n.a. n.a. /a Source: Consultant's Report, "Route Blitta-Sokode Etude de factibilite", 1972. /b Source: PCR, Benin Second Highway Project (Credit 415-DA), Table 8, August 1979. PROJECT PERFORMANCE AUDIT REPORT BENIN SECOND HIGHWAY PROJECT CREDIT 415-DA) Benin and Togo - Vehicle Operating CoEts Excluding Taxes - 1976 To go./ Benin/b Gravel Good Poor Good or Poor Paved Paved Change in Savings Paved Paved Change in Savings Road Road Savings 1976/1972 Road Road Savings 1976/1972 ----- (CFAF per km)----- (%) -------(CFAF per km) ------ (%) Cars and Pick-Ups 27.4 43.8 16.4 - 5 n.a. n.a. n.a. n.a. Cars n.a. n.a. n.a. n.a. 22.8 33.1 10.3 + 29 Pick-Ups n.a. n.a. n.a. n.a. 23.1 33.5 10.4 n.a. Trucks (2-10 tons) 7 ton Trucks 50.7 97.8 47.1 +56 n.a. n.a. n.a. n.a. 10 ton Trucks n.a. n.a. n.a. n.a. 94.1 154.4 60.3 +172/c Heavy Trucks 11 ton Trucks 71.1 137.8 66.7 +31 n.a. n.a. n.a. n.a. Truck-Trailers 88.9 172.2 83.3 - 3 114.8 198.5 83.7 + 451/ /a Source: PCR, Togo Second Highway Project (Credit 450-TO), October 1979. Data came from Appraisal Report on Togo Third Highway (Credit 693-TO), March 1977 and were prepared by consultants (French) for Aneho-Tabligbo Road feasibility study. /b Source: PCR, Benin Second Highway Project (Credit 415-DA), Table 8, August 1979. /c Vehicle types not fully comparable. PROJECT PERFORMANCE AUDIT REPORT BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) Benin and Togo - Vehicle Operating Costs Excluding Taxes - 1980 Togo/a Benin Good "Ruined" Good Paved Paved Change in Savings Paved Gravel Change in Savings Road Road Savings 1980/1976 Road Road Savings 1980/1976 ------ (CFAF per km)------ (%) ------(CFAF per km) ------ (%) Cars and Pick-Ups Cars 37 50 12) 35 43 8 -22 -36 c Pick-Ups 34 43 9) 37 46 9 -13 Trucks (2-10 tons) ) 10 ton Trucks 108 150 42 n.a. ) 94 131 37 -39 Heavy Trucks ) 11 ton Trucks ) Truck-Trailers 205 315 110 +32 149 223 74 -12 /a Source: Consultant's Report, "Road Maintenance Analysis in Togo, 1980." Appraisal Report on Togo Fourth Highway Project (Credit 1139-TO), Annex 5, April 1981. /b Source: Appraisal Report on Benin Fourth Highway (Credit 1142-BE, April, 1981). PROJECT PERFORMANCE AUDIT MEMORANDUM BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) Proposals for Use of Unspent Credit Funds Estimated Available Proposal Date Funds Disposition (US$ million) Improve the dike section on the Cotonou- 01/06/78 3.7 Withdrawn by Government. Bohicon Road (estimated cost US$700,000) and protect culvert inlets and outlets on that road (estimated cost US$500,000) Purchase equipment for shoulder maintenance Late Over Agreed by letter of 12/21/78 whereby Part III brigade (estimated cost US$310,000) 1978 3.0 of Project Description was amended to include "and establish- ment and equipping of a highway shoulder maintenance brigade" Funds were reallocated, inter alia, to increase to US$500,000 Category IV of Schedule 1 of the Credit Agreement, (Construction materials and maintenance equipment for Part III of the Project). Improve the Cotonou bypass road in Early 3.4 Considered not feasible to be incorporated as part of Second east-west direction (estimated 1979 Highway Project. cost US$3 million) Construct a new bridge over the Oueme River 10/79 N.A. Considered "not compatible with objectives" of Second at Save on the Cotonou - Parakou Road Highway Project. (estimated cost just under US$4 million) Bridge financing for technical assistance to maintain operations under Feeder Roads 02/80 3.9 Agreed by letter of 05/16/80 to allocate US$600,000-650,0 Project (Credit 717-BEN) after IDA funds to finance consultants' services, which had been included in under that project had run out and before the Feeder Roads Project provisions for establishment, staffing and funds under a follow-on project could equipping the Feeder Roads Division of the Directorate of Roads and become available (estimated cost US$2.2 Bridges, on-the-job training of the Directorate's staff, and prepa- million for period October 1980-September ration of a long-term program for feeder roads. These were 1981) financed under the Second Highway Project, Category III of Schedule 1 of the Credit Agreement (Utilization of the Borrower of consultants' services in highway maintenance and transport coordination), - 21 - August 1979 PROJECT COMPLETION REPORT BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) I. Introduction 1.01 Benin has traditionally served as a transport corridor to the sea for landlocked Niger. The main component of the Second Highway Project aimed at improving the road in northern Benin which serves as a link in this corridor and also carries considerable domestic traffic. The "Benin route" from Cotonou to Niamey (Niger) totals 1,033 km, 114 km shorter than the alternative Togo route through Lome. From the port of Cotonou, goods are carried 438 km by rail to Parakou and then 318 km by truck to Malanville on the Niger border and into Niger by roads. The railway company, jointly owned by Benin and Niger, also has a monopoly of truck traffic to Niger and charges fees for using the joint road-railway bridges. This system results in tariffs about 25-35% lower than on the Togo route and up to 50% lower than through Abidjan. The Abidjan route is however gaining in importance because of better services. The Parakou-Malanville road is northern Benin's only outlet to the south, and over the years domestic traffic has become more important than the transit traffic. This trend is expected to continue as more of the land, well suited to cash crops, comes under cultivation; at present, with only 2% of the area cultivated, it produced about 60% of the country's cotton. 1.02 The Second Highway Project was carried out at a time of profound political and economic change. Since independence in 1960, Dahomey had suffered from internal political unrest, rooted in tribal and economic differences between the three ancient kingdoms, Abomey and Porto Novo in the south, and Bariba in the north. A prime objective of the military government which came into power in 1972 was to unite the country politically. As part of the unification policy, the name of Dahomey was changed to the People's Republic of Benin in November 1975 and a single political party was estab- lished. 1.03 The Bank Group's strategy has been to promote the Benin route road maintenance and feeder roads. The Feeder Road Project (Cr. 717-BEN) provides for improvement and/or maintenance of 1,270 km of roads, many connected with the north-south main road, and institution building; the subsequent Third Highway Project (Cr. 746-BEN) rehabilitates the remaining sections of the north-south road, Godomey-Bohicon-Abomey, (107 km) and provides for regravel- ling or resurfacing of 470 km of roads connecting to the Parakou-Malanville - 22 - road and further assistance for road maintenance and institution building; and the port of Cotonou is being expanded (Cr. 836-BEN) to enable it to handle the expected increase in domestic and transit traffic, for Niger and Nigeria. II. Project Identification and Preparation Identification 2.01 The Second Highway Project evolved from the Land Transport Survey financed by the UNDP with the Bank as executing agency. The survey was carried out jointly by two Canadian consulting firms, and was completed in 1970. With regard to highways, the survey stressed the need for road main- tenance, assigned highest priority to rehabilitation of the Parakou-Malanville road serving both northern Benin and Niger transit traffic, and recommended improvement of two short sections of the Cotonou-Bohicon road, one of the major trunk roads in the south. Preparation 2.02 Based on the survey recommendation, the Association financed a highway maintenance and engineering project in 1970 (Cr. 215-DA, $3.5 million), which included detailed engineering for the three sections mentioned above. Consultants were selected to prepare the studies. They found that the Parakou- Malanville road (318 km), a one lane 3.5 m wide road, had deteriorated and re- construction would involve more work than previously anticipated. The Govern- ment requested an update of the economic study to ascertain whether a two lane road would be justified. The reevaluation was completed in 1971 concluding that a two lane paved standard was economically justified and indeed better adapted to projected traffic. The Government agreed with the Association on this recommendation and requested the consultants to complete the engineering on the basis of a 6 m wide carriageway. Appraisal, negotiations and expected financing 2.03 A Bank mission appraised the project between February 9 and 24, 1972. As USAID had expressed interest in co-financing, one of their engineers joined the mission. During negotiations, one year later, between February 20 and 27, 1973, the Government requested an increase in the amount of technical assistance to be provided to the Directorate of Public Works and Land Trans- port and additional road maintenance assistance. Moreover, and in view of the financial difficulties facing the country, the Government requested an increase in the proposed IDA credit to cover the impact of the devaluation of the US dollar (about 10%) and to finance part of the local costs. In view of the Government's willingness to enact fiscal austerity measures, IDA agreed to increase the proposed credit from $8.2 million to $11.8 million, to cover - 23 - the entire foreign exchange cost and 43% of the local cost, net of taxes, of the IDA financed items. USAID agreed to finance $6.5 million foreign exchange cost, and $1.5 million equivalent in local cost, for the reconstruction of a northern half of the Parakou-Malanville road. The Government was to cover the remaining local cost estimated at $1.3 million equivalent. The total project cost, estimated at $21.1 million, was to be financed as follows: --------------$ million---------------- Local Foreign Total % of Total USAID 1.5 6.5 8.0 38 IDA 1.0 10.8 11.8 56 Government 1.3 - 1.3 6 3.8 17.3 21.1 100 Actual'costs and financing are discussed in paras. 3.25-30. Objectives and covenants 2.04 The main objective of the project was to ensure that the main north-south road would be adequate to serve Benin's internal traffic and sufficiently competitive to retain Niger's transit traffic, which provides benefits to Benin as well. In addition, the project would ensure the con- tinuation of the highway maintenance program. In order to avoid over- investment in the north-south corridor, the Credit Agreement requested the Government that, prior to undertaking any northern extension of the Cotonou- Parakou railway paralleling the road to be rehabilitated, a feasibility study should be done establishing to the Association's satisfaction the economic justification of such extension (Section 4.06). Also, to achieve a better coordination of the transport system, the railway's management in particular emphasizing operations of the Cotonou-Parakou line, was to be improved, and the Directorate of Land Transport was to be strengthened to enable it to play an active role in defining inter-modal policies and invest- ment planning (Section 4.05). Usual covenants on road maintenance were agreed on: installing and operating weigh-bridges along the Parakou-Malanville road, and ensuring adequate dimensions and axle loads of vehicles (Section 4.04), keeping adequate records (Sections 4.01 and 4.03) and maintaining adequately the road system (Section 4.02). Project Description 2.05 The project consisted of: (a) rehabilitation of the Parakou-Malanville road (318 km); (b) rehabilitation of two sections of the Cotonou-Bohicon road, from Godomey to Abomey Calavi and from Sehoue to Zakpo (totalling 17 km); - 24 - (c) consultant services for supervision of construction works; (d) continuation of the road maintenance program (begun under the First Highway Project), consisting of (i) consultant services and (ii) materials for culverts and bridges; and (e) consultant services to Directorate of Land Transport. Board approval and effectiveness 2.06 The Board approved the project on May 24, 1973. The Credit Agree- ment was signed on July 3, 1973, but not declared effective until March 29, 1974, when cross-effectiveness conditions regarding the USAID loan were met. The delay was mainly due to a tedious dispute over the exact breakpoint between the IDA and USAID financed sections of the Parakou-Malanville road. At appraisal it was envisaged that each agency would help finance 160 km, but the final engineering completed in July 1973, considered the Sori bridge as dividing point resulting in a northern section of 169.5 km and a southern section of 148.1 km. Only after considerable and undecisive correspondence between the parties was there an agreement, and a fixed price contract for the USAID portion of works could be signed, which was a condition of effec- tiveness for the IDA credit. The Government complied with its effectiveness conditions on October 10, 1973. Project completion was expected by Novem- ber 30, 1976. III. Project Implementation A. Construction Bidding 3.01 Project start-up was not hindered by the delay in credit effective- ness. Nineteen contractors were prequalified for construction of the three sections financed by IDA, and five presented bids covering the three lots. Bids were opened on October 15, 1973 and the Government retained the three lowest for evaluation: the low bidder (French), followed by a joint venture of a French firm and a German firm, and by another French firm. There was less than 5% difference in the proposals' cost (Table 3). Three US contractors were prequalified for construction of the northern USAID-financed section of the road. Contract Award 3.02 In a letter to the Association of November 17, 1973, the Government proposed to award the contract for lots 1, 2 and 3 to the second lowest bidder (excluding housing costs), preferring this bid to the lowest bid because it proposed a crushed stone base course while the lowest bidder proposed a soil-cement base. The consultants advised Government that while the crushed stone base was 4% (about $300,000) more expensive, it would render the road more durable and save on maintenance costs. They also advised the - 25 - Government that the unit price of the first French firm for a soil-cement base seemed unrealistically low being the same as they had bid on another job in 1969. Moreover, the Government felt uneasy about the many reservations made by the firm in the "technical note" accompanying their bid. Finally, the Government saw an opportunity to break the monopoly of French construction firms by bringing in a part-German contractor. 3.03 Because all bidders had been pre-qualified, and bids had been called for either a soil cement base or a crushed-stone base, the Association could not approve the choice of the second lowest bid, and requested the Government to justify economically the more expensive one. However, as the validity of the bids expired on February 12, 1974 the Government invited the bidders to extend the validity of their bids to March 15, 1974. The two lowest bidders refused, while the highest bidder accepted. The Government was then left with little option but to award the contract to the only remaining bidder, which the Association approved on March 8, 1974, at about $150,000 more than the lowest bid (including housing costs). The contract was signed on March 29, 1974. The total contract price was $8.5 million equivalent of which IDA would finance $7.6 million, within the credit allocation of $7.75 million excluding contingencies. The procedure had taken five months from bid opening to award. The cost of the soil-cement solution in the contract awarded was higher than the lowest soil-cement bid; the discussion on the technical and economic merits of crushed stone should have been decided before the bids Wre invited. 3.04 The contract for the northern section of the Parakou-Malanville road (169.5 km) was awarded to the lowest bidder, an American firm, in March 1974. The contract price with the crushed stone base alternative was $10.9 million or 36% above the appraisal estimate and higher than the second French contract because it was a fixed unit price contract, whereas the second French contract included a price escalation clause. USAID increased its loan from $8 million to $12 million to cover the increase and contingencies. Construction Supervision 3.05 The contract for supervision of all construction works (lots 1 - 4) was awarded to the same consultants that had prepared the documents, in May 1974 with notice to proceed on June 24, 1974. Construction execution 3.06 Both French and American contracts were signed on March 29, 1974, with an execution period of 30 months. In this period, the contractors faced problems arising from political events: a revolution, a closure of the border with Nigeria (March 1976) and the consequent rupture of bitumen sup- plies. Prices also increased considerably due to the oil crisis. Neverthe- less, works were completed with only minor delays to a satisfactory quality, although in retrospect the design standards adopted seem questionable (paras. 3.10-3.11). - 26 - Parakou-Malanville (318 km) 3.07 Parakou-Sori (148.1 km). Works on this section (lots 2 and 3)were carried out by a French firm. The order to proceed was issued on September 23, 1974, and despite the above-mentioned problems, works were completed only 2 months late, on May 15, 1977 and were provisionally received on May 24, 1977. An inspection of the road in March 1979 by an IDA mission showed no serious deficiencies following two wet seasons. 3.08 Sori-Malanville (169.4 km). Works on this section (lot 4) were carried out by the American firm with USAID financing. The order to proceed was issued on July 1, 1974, and works were satisfactorily completed only 2 months late on February 7, 1977, and provisionally received by Government the same month. Besides the problems mentioned in para 3.06, progress in the early part of construction was hampered by conflicts between the contractor and the supervising consultant: the contractor protested about the personal behavior of the consultant resident engineer and the impartiality of his technical decisions. At one point production of the crushed stone base was halted because of difficulty in meeting the requirements of the sand equivalence test as specified in the bidding documents. The American firm held that by all the normal parameters the material was good, but the resident engineer insisted on compliance with the sand equivalence test specifications which is not in common use outside the French zone of influence. Eventually (August 1975) the consultants head office agreed to reduce the sand equivalence norm from 50% to 30%. Relations improved instantly when the resident engineer was replaced, and the work was finished on an amicable basis. The contractor also suffered a strike, the impounding of his explosives during the revolution and a number other problems that culminated in a claim against the Government of $7.5 million which he is pursuing before international arbitrators in Paris. The contractor had chosen an unfortunate time of rising inflation to win a fixed price contract. Additional works 3.09 In three letters to IDA dated February 2, 1976, the Government requested the following additional works to which IDA agreed: (a) installation of a truck weighing station outside Parakou at an estimated cost of $80,000 in foreign currency and $33,000 in local currency (CFAF 7,500,000); (b) construction of a parking lot for trucks at Parakou at an estimated cost of $86,000 (CFAF 19,310,000); and (c) replacement of telephone poles on the southern, IDA- financed section at an estimated cost equivalent to $100,000 (CFAF 22,680,000). - 27 - At the same time the Government requested USAID to finance a weighing station at Malanville and truck parking lots at Malanville and Kandi. The truck weighing station was to be installed by the American firm but it was decided that the existing scale could be repaired at a significantly reduced cost. The parking lot construction and telephone pole replacement were done by a French firm. Choice of road width 3.10 At appraisal the economic analysis showed the reconstruction to two lane standard of the Parakou-Malanville road was justified, and a 6 m pavement with 2 x 1.5 m shoulders was adopted. The road was completed to this standard in 1977. In retrospect, the design width of the road was insufficient. The greater than estimated traffic with its high proportion of heavy vehicles has proved too great for the 6 m wide pavement to whose bends no curve widening was applied. The result is that wear is taking place both on the inside of curves, and on the straights. For one heavy vehicle to pass another on the straight it must put a wheel on the shoulder and to negotiate a right hand bend, an articulated vehicle cannot avoid putting a rear wheel off the pavement. The result is that wear of the pavement/shoulder junction is taking place, creating a continual routine maintenance problem. 3.11 It is interesting to review how this situation came about, and what could be its cure. The point of departure was a single lane road, and the original analysis was based on data gathered in 1967 and updated in 1971, but the road did not open to traffic until May 1977. During this period the Niger uranium boom began, the size and number of large commercial vehicles increased and total traffic was 70% more than expected; also, the idea of 7 m pavement widths for international routes became more widely accepted. But it is notor- iously difficult to widen an existing embankment, costly to widen an existing cut, and next to impossible to add a narrow widening strip to a soil-cement base. There is, however, one partial solution which could limit damage through traffic management measures, acting on the psychology of the drivers rather than the structure of the road. This would be centre-line marking and edge marking on curves, which was discussed with the Directorate of Roads and Bridges. Even more important, although in practice the more difficult to get done, is regular shoulder maintenance. To assist in this effort, the Association agreed in December 1978 to finance the purchase of equipment to establish a shoulder maintenance brigade under the same project. Two sections of the Cotonou-Bohicon road (17 km) 3.12 The contract period for lot 1 was 12 months, following the order to proceed on July 7, 1974. Works began July 26, 1974 were completed in February 1977, 31 months later, and were provisionally received on July 26, 1977. The delay was due to unexpected problems during the raising of an embankment on the Godomey Dike which twice caused sudden and deepseated ground failures which destroyed both the road and Cotonou's main water supply pipeline. This is further discussed below given its technical interest. Otherwise, there was - 28 - only one minor problem, the cracking of the sand asphalt surfacing on certain sections near Zakpo. The contractor agreed to repair the cracks, resurface the affected sections and accepted a two-year maintenance period until July 1979. The soil-cement base and sand-asphalt surfacing were inspected recently and were in good condition. The remainder of the 85 km road is being recon- structed or resurfaced with financing from Kuwait and OPEC Funds under the Third Highway Project. Godomey Dike problems 3.13 Site conditions and design. At a point 300 m from the start of the Godomey-Abomey road, the Djonou Lagoon enters Lake Nokoue. The existing road crossed the narrow neck of land on a causeway (in French, Digue) under which culverts had been placed at various times, but nevertheless the road was always flooded at the height of the wet season. For this reason, the engineering design called for a bridge and a raised approach embankment on the north side. The bridge span was fixed at 6.6 m and gave a free height of lm at peak flood. The deck and the adjacent embankment were to have a finished level of 4.8 m above sea level or approximately 2.2m above the level of the existing road. On the southern side the ground drops down to the bridge without any embankment, and on the northern side climbs gently to meet the proposed road level some 500m north of the bridge. The embankment where the problems occurred runs from the northern abutment of the bridge to a point some 300m to the north. 3.14 Site investigation. A total of five boreholes and five SPT's (Standard Penetration Tests) were made in the affected area. They showed 5m-8m of very loose alluvial sandy silts with high organic content underlain by a bed of clay which varied in thickness from 3m at the bridge site to over 15m only 150m to the north. SPT values were very low never exceeding 30 blows per foot at 15m depth. The bridge design report discusses the foundation options intelligently and cautiously and finally opts for long friction piles - a correct solution. The road design report on the other hand accepts the hydrologists' recommended flood level, the bridge designers deck level and states "... and there are no problems for the placing of the pavement structure". Unfortunately there were. 3.15 Construction. Road construction began in July 1974, and the first reports of embankment failure occurred in April 1975, when a deepseated settlement of about 1m took place over a 200m length, which was accompanied by the rupture of Cotonou's main water supply line (400 mm 0). The consultant recommended construction of a berm on either side of the road formation, which involved the placing of 25,000 m3 of fill. Filling was in progress when a second settlement failure and rupture of the water line occurred in May 1975, leaving the city without water for three days. The Government stopped all further earthworks on the Digue, and a temporary pavement was maintained for the next six months. The consultants then proposed in January 1976 to recom- mence raising the Digue but this time to a final level 1m lower than the original design level. This was to be done in four lifts with 4 to 8 weeks interval between them depending on settlement rates as monitored by a line of bench marks which were to be established. - 29 - 3.16 At this stage the Bank became concerned and the intervention of the next supervision mission placed the further actions on a cautious practical footing in order to arrive at an acceptable temporary solution, while intro- ducing the idea of a full site investigation program on which to base an eventual "final solution". The first 50cm layer was placed on July 1976; but the graph of the measured settlements continued to be linear, suggesting little gain in shear strength with consolidation. Also in July, a testing program was worked out between the Government, the Bank and the consultants. 3.17 The consultants program proposed a battery of tests to be done on samples shipped to Germany. However IDA and the Centre National d'Essais et de Recherches des Travaux Publics (CNERTP) insisted that: (a) the nature of the materials involved made it unlikely that an "undisturbed" sample at its natural moisture content would ever reach the laboratory in Germany; (b) there was a great opportunity for training within CNERTP where the knowledge but not the equipment required resided. With IDA's approval CFAF 9 million was spent on triaxial and other fairly sophisticated testing equipment to do the test "in house". . 3.18 Another 50cm was added in December 1976 but there was still no flattening out of the settlement curve and the Bank engineer advised the cessation of further filling and the placing of a temporary pavement of inter- locking concrete paving blocks known locally as "Pave Trief" and which can follow differential settlements. Construction of the pave trief pavement (450m) was completed in April 1977. The site was again visited by a Bank mission in March 1978. While the existing vertical alignment is not up to the original design standard, it is acceptable for existing traffic speeds and volumes. The Bank engineer concluded and DRP officials concurred that the risks inherent in a further attempt to reach the design level were not justified by the small improvement in surface riding quality, and it was decided to try no further raising of level. Settlements were estimated at 2mm per month. Although DRB assures that monitoring continues, they have not produced any survey data since that time. 3.19 Conclusions. The original site investigation was adequate: what was not was its interpretation. Although there was not enough evidence presented on which to base a full earthworks design, there was enough to call for further testing. There was also a lack of communication between the designers and public utilities leading to a failure to identify the possibility of damage to the water pipeline. Once the problem had presented itself there was indecision. The first failure occurred in April 1975; no decision to implement a testing program was made until July 1976. There was lack of technical leadership on the part of consultants. It is worth noting that the course of action that led to the solution finally adopted was sug- gested by the Bank and was based on common sense and field observations, not on a battery of expensive and sophisticated laboratory tests. - 30 - B. Extension of the Road Maintenance Program 3.20 The First Highway Project (Cr. 215-DA, $3.5 million, 1970) provided almost $2 million under the 1970-1974 maintenance program for new field and workshop equipment and $1 million for technical assistance, with the aim of strengthening the Directorate of Roads and Bridges (DRB) both in equipment and management. Since this was not fully achieved in the time available, the Second Highway Project provided an additional 83 man-months of technical assistance to prolong the work of consultants (Canada) and $250,000 for construction materials (steel culverts, cement, etc.) for the drainage improvement which were to be part of the extended program. 3.21 Under this project the maintenance program was to be financed by the Government using the Road Fund which was re-established in 1970 by the introduction of a tax of CFAF 4 per litre of gasoline. However road maintenance expenditures of CFAF 401 million in 1974 increased to only CFAF 404 million in 1975 and CFAF 427 million in 1976 and were actually below the 1972 budget level of CFAF 474 million (Table 4). The tax on motor fuel remained fixed at CFAF 4 per litre and the Government was unable to increase its contribution to road maintenance to counteract inflation. 3.22 Physical performance was negligible; because of lack of funds, spare parts were increasingly difficult to obtain and equipment was immobilized. Physical improvements to roads fell short of expectations because insufficient funds were made available, and far from recouping the shortfall as the second project was intended to do, performance was even worse. As of November 1975, consultants reported that 56% of the heavy maintenance equipment were out of service and about half of the trucks were down. The high incidence of equip- ment breakdown was due to the inexperience of operators. As a result of these difficulties, maintenance works were only carried out for about two months of the year. The consultants did some useful desk work in maintenance planning and quantification of needs; an extension to this work prepared in detail the maintenance resurfacing and reconstruction elements of the Third Highway project. By the time the consultants left the country less than 30% of the funds provided for materials under the project had been disbursed - a remark- able failure to take advantage of an opportunity to build stocks of essential materials like steel culverts, cement, bitumen, etc. 3.23 In December 1978, in view of the shoulder maintenance problem that had arisen on the Parakou-Malanville road, the Association agreed to the purchase of equipment for a shoulder maintenance brigade, estimated to cost $350,000. Procurement has been started. C. Consultant Services to the Directorate of Land Transport 3.24 Because of impending reorganization of the Ministry of Equipment, and probably also because of the fluid state of the country during the revolution, the Government asked for a deferral of the 23 man-months of technical assistance for the strengthening of the Directorate of Land Transport. This assistance is being provided to the new Directorate of Studies and Planning, now part of the Ministry of Transport, under the Third Highway Project. - 31 - D. Supplementary Credit 3.25 By mid-1975 it had become apparent that the project would incur a considerable cost overrun. Supervision missions estimated the cost of the IDA-financed part of the project at $23.2 million instead of the $13.1 million estimated at appraisal. The estimated 77% cost overrun was based mainly on actual OPEC price increases and world wide inflation and the anticipation that such increases would continue and would add $7.7 million to project costs. Further, the US dollar was devalued from CFAF 230 at appraisal to CFAF 224 average per dollar in 1975, and it was expected that the exchange rate would remain at about 225 for the remaining disbursements; the total effect was estimated at an additional $400,000 in project cost. Additional civil works were estimated to cost $2 million. The appraised and revised estimates of the IDA financed parts of the project are given below. Cost Estimate Revised Cost Estimated at Appraisal Estimates Cost Overrun February 1972 July 1975 July 1975 --------------- t---US$ million------------ Civil works 10.8 20.3 9.5 Consultant services for construction supervision 1.3 1.8 0.5 Consultant services for highway maintenance and transport coordination 0.7 0.8 0.1 Materials 0.3 0.3 - Total 13.1 23.2 10.1 3.26 At the time (July 1975) the works were well advanced, and there was no practical way to reduce the scope of the project. The Government was unable to finance the expected cost overrun, which was beyond their control, and requested a supplementary IDA credit of $9 million, which would raise the credit to $20.8 million. The percentage of civil works costs financed was raised from 81% to 89%. The Board approved the supplementary credit on March 2, 1976. The amended credit agreement was signed on March 10, 1976, and became effective on October 5, 1976. 3.27 As it turned out, project costs proved to be less than the revised estimate, as inflation abated, the US dollar recovered its value to an average of CFAF 248 in 1976 and 249 in 1977, and bitumen deliveries from Nigeria improved replacing more costly bitumen purchases on the world market. About $3.9 million was left undisbursed under the credit, and in 1978 the Director of Roads and Bridges informally asked a supervision mission whether part of the balance could be applied to the construction of a 5.5 km link between the - 32 - new Cotonou bridge and the main highway leading west and north, estimated to cost about $2.5 million. However, the Association deeming that the objectives of the project had been met, and that the balance was due to excessive contin- gency allocations, indicated to Government in April 1979 that, according to the Bank Group rules, the balance should be cancelled when project disburse- ments are completed. E. Final Project Costs and Disbursements 3.28 Appraisal estimated and actual costs are detailed in Table 5. The total project cost was $31 million equivalent or 47% above appraisal estimate of $21 million. The project components financed by IDA cost about $18.6 million, as compared to the original estimate of $13.1 million, with an overall increase of 44% given a 35% increase in the cost of the Parakou- Malanville road and about 68% increase in the Cotonou-Bohicon sections. However, the southern section of the Parakou-Malanville road was 10 km shorter than envisaged, technical assistance for transport coordination was deleted, and equipment for a shoulder brigade included in final project costs. The IDA contribution will be about $16.9 million, instead of $11.8 million as envisaged, in line with the overall cost increase. 3.29 The northern section of the Parakou-Malanville road, financed by USAID, cost $12.4 million or 55% more than the appraisal estimate of $8.0 million; USAID increased its loan to $12 million, covering practically the full cost. The American contractor has filed a claim against Government of Benin in the International Court of Justice (Paris) for an additional amount of $7.5 million equivalent. 3.30 Actual and estimated disbursements are shown in Table 6. Delays were due to slow placing of contracts, subsequent construction delays and by slow processing of claims by Government. The original credit closing date, November 30, 1976, was postponed first to December 31, 1978 and then to December 31, 1979. Practically all disbursements for the project had been completed by December 1977; the second postponement was agreed upon to allow disbursements for equipment for the shoulder brigade that the Association approved in December 1978. F. Consultants Performance 3.31 The supervising consultant's performance was uneven. Their failure to interpret correctly the results of their own soil survey was mainly respon- sible for the Godomey Dike failure, and their subsequent handling of the repairs was little satisfactory (paras. 3.14, 3.17). There were also person- ality problems between consultant staff and contractors (American), although rela- tions improved when some of the former were replaced. The consultant firm was also very slow in finalizing accounts. Consultants for the road maintenance program on the other hand, made the best of a difficult situation. Although they were unable to achieve very much by way of physical improvements to road maintenance nor to even begin their assistance to DLT because of Government uncertainties and lack of funds, they produced a useful study of road main- tenance needs with which the Government has expressed satisfaction. They also prepared in detail most of the elements of the current Third Highway Project and were re-hired by the Government for its implementation and supervision. - 33 - G. Government Performance 3.32 The Directorate of Roads and Bridges is weak in organization and lacks staff, and institutional performance was bedevilled for much of the project period by frequent changes of senior staff. Individually, DRB senior engineers are diligent and competent, but they lack both support structure and staff, particularly for procurement and finance. Both con- sultants and contractors complained of very slow processing of payments by the Government. At one point one of the French contractors was claiming almost $500,000 in interest on late payments, but the claim has been dropped. Because of the political situation and the austerity measures which followed, little was achieved by way of institution building under the project, but at least some ground- work was done which hopefully will bear fruit under the Third Highway Project. H. Compliance with Project Covenants 3.33 The Government complied only in part with Section 4.02 of the Credit Agreement which required that the highway system should be adequately maintained and that funds and equipment should be provided "as needed". Maintenance funds and spare parts were well below the needs of adequate maintenance. However, the covenant did not establish annual minimum levels of funding. The Third Highway Project, therefore, specifies a minimum budget amount for annual highway maintenance. 3.34 The Government complied in part with Section 4.04 requiring control of truck axle loads. Weighing stations were installed at Parakou and Malanville. The Government did not however, impose penalties for truzks exceeding the weight limit and overloading continues. 3.35 The Government did not furnish the Association with a "detailed program" to strengthen the Directorate of Land Transport and the railway management as specified in Section 4.05 and the technical assistance for this item was dropped. The Association therefore agreed to include technical assistance for transport planning in the Third Highway Project (Cr. 746-BEN) and technical assistance for improving railway management is included in the Port of Cotonou project (Cr. 826-BEN). 3.36 During the project period there was no need to invoke Section 4.06 which requires the Government to furnish the Association with a feasibility study by a qualified consultant establishing the economic justifications of any proposed extension of the railway from Parakou to Niger. However a feasibility study, financed by FAC, is now under way for a project to extend the railway to Niamey. IV. Economic Reevaluation 4.01 The appraisal report did not include the overall rate of return of the project, but individual components' return ranged from 13% to 19%. Current reevaluation shows a global return for Benin of 15% within the same range, but deriving from a 16% return for the Parakou-Malanville road, the main component, instead of 13% as expected, (or 30% instead of 19% when including benefits to Niger from its transit traffic 1/), and from a 15% return instead 1/ Excluding from the cost the pending claim by the contractor. - 34 - of about 17% from the rehabilitation of the Cotonou-Bohicon sections. The identified uncertainty was carefully evaluated in a probability analysis and proved to be appropriate. However, it would have been difficult to foresee the actual traffic development and uncertainty still remains as to future pattern of traffic on the project roads (para 4.04). Such results, however, are a good reminder of the need to be cautious in ranking proposed projects based on a ex-ante rate of return analysis. Parakou-Malanville road 4.02 At appraisal, the economic evaluation was essentially directed at comparing the cost of the proposed rehabilitation with expected savings in transport costs for domestic traffic as well as for Niger transit traffic. The growth rate for domestic traffic was estimated at 5% for the period 1976-86, and 4% thereafter; in addition, the reduction in transport costs was expected to induce increased agricultural production and thereby increase traffic by about 10 to 20%, depending on the road section. Although with the rehabilitation the Benin route was expected to remain among the most economic alternatives for Niger, the transit traffic was considered highly uncertain and was treated through an elaborate probability analysis; the best estimate was that it would grow by about 50% by 1980 and stabilize thereafter. For 1976, expected to be the opening year of the new road, the traffic forecast was 59 heavy transit trucks for Niger, and from 109 to 51 "domestic" vehicles per day, the flow decreasing northwards with an average of about 82. Actually, the road was completed in 1977 and carried an average of 250 vpd, considerably above the expected total of 141, although Niger transit traffic approximated the level expected at appraisal. Traffic evolu- tion is shown in Table 7. 4.03 Unit vehicle operating savings increased about 40% between the 1972 appraisal estimates and 1977, when the road was opened (Table 8). Actual construction costs were about 50% above appraisal estimates. However, given the 70% higher traffic, the reappraisal rate of return is about 30% for the region and 16% for Benin alone, even when excluding any revenue from transit traffic; such exclusion now seems warranted as it is difficult to apportion transit revenues to road, railroad and port improvements undertaken during the same period (Table 9). 4.04 The reevaluation assumes that future growth will be consistent with appraisal rate, i.e. 5% until 1986 and 4% thereafter. Rural development projects in northern Benin and continued economic growth in Niger with con- comitant increase in transit trade sustain the assumed traffic increase to 1987. Beyond 1987, the traffic pattern may diverge from the assumed rate if two proposed transport investments take place: (a) the paving of the remain- ing section of the Cotonou-Niamey road (between Dassa Zoume and Parakou), and (b) the extension of the railway from Parakou to Niamey. Cotonou-Bohicon road sections 4.05 Based on vehicle operating and road maintenance savings and the rehabilitation costs, the economic return was estimated at appraisal to be 19% for the 6 km Godomey section and 14% for the 14 km Sehoue section; the appraisal report however does not mention the traffic growth rate considered in the analysis. - 35 - 4.06 Traffic development has been very uneven (Table 7). Interruptions or detours during construction of the Godomey section (because of the water main rupture) reduced traffic from 3,076 vpd in 1974 to 1,896 vpd in 1975; but then a considerable increase took place in 1976 when traffic climbed to 2,026 vpd. Traffic on the Sehoue section reverted in 1976 to its 1971 volume of about 420 vpd, although it had reached more than 730 vpd the previous years. A.07 Estimates of unit vehicle operating savings increased by about 40% from appraisal to the time the new road opened; at the same time actual con- struction costs increased about 67% compared to appraisal estimates even though final length reconstructed was 17 km instead of 20. Assuming a future 5% annual traffic growth, the combined economic return of the two sections is estimated at about 15%. Despite such lower than expected return, the recon- struction was fully justified and may be even more so should traffic recover previous levels faster. V. IDA Performance 5.01 The Association approved a carriageway width of only 6m for the Parakou-Malanville road. Because of the narrow lane width and considerably higher than expected traffic, road shoulders are frequently used by overtaking vehicles and shoulder wear is already heavy. The resulting maintenance prob- lem, in view of Government's record, is unlikely to be well-handled. The Association should therefore consider recommending wider roads in future projects for international routes used predominantly by heavy trucks. 5.02 The Association might have been less formal in the determination of "the lowest evaluated bid", as compared to the "lowest bid" (an about $300,000 difference excluding housing costs) that it would approve for the construction works. Because of delays in making decisions and the consequent expiration of bid validity, the Government had to award the contract to the highest of the three bids retained, and accept a soil-cement base instead of the Government's technically preferred crushed-stone base offered by the second lowest bidder. 5.03 IDA supervision missions played an effective role in solving tech- nical problems during project implementation, particularly the problem of settlement of the Godomey Dike section. The Association was also able to assist in overcoming the difficulties in bitumen deliveries. 5.04 Supervision missions voiced concern about insufficient maintenance funding and lack of effective control of vehicle weights, but could not force the Government to take the necessary steps to remedy these issues. A follow- up on these matters, as well as technical assistance for transport planning, is pursued under the Third Highway Project. 5.05 IDA took timely action in providing a supplementary credit when construction costs increased beyond available financing. The lower than anticipated cost overrun of the IDA-financed portion of the project was due to fortuitous developments which could not have been foreseen at the time of the supplementary credit approval, and cancelling surplus funds in line with the Bank policies is appropriate. - 36 - VI. Conclusions 6.01 The objectives of the project were met, except for institution building which could hardly have been achieved even through reinforced components as political instability prevailed. This objective, however, is being pursued under the follow up Third Highway Project. Each component is well justified with a current global 15% rate of return estimate, despite considerable cost overruns. The $2 million overrun due to technical problems on the Godomey Dike could most likely have been avoided if the consultants had identified the problems early; but remaining cost increases were beyond project control. The Parakou-Malanville road is probably underdesigned with a 6m pavement width, given the increase in number and size .of vehicles above estimates and the question also remains whether a crushed-stone base would have been better, more so in view of the eventual need of widening the road. - 37 - TABLE 1 BENIN Traffic and Transit Trade at the Port of Cotonou 1977 (tons) Destination/Origin Imports Export Total General Cargo - Petroleum Benin 433,043 117,136 67,620 617,799 Niger 137,482 70,636 5,788 213,906 Nigeria 213,286 - - 213,286 Mali - 1,636 - 1,636 TOTAL 783,811 189,408 73,408 1,046,627 Source: Port of Cotonou Authority, 1978 - 38 - TABLE 2 BENIN Highway Network, 1979 - 80 (kilometers) Roads Paved Laterite Earth Total A. Maintained by DRB 883 2,473 1,532 4,893 1. Interstate 7911/ 1,342 - 2,133 2. National 97 1,131 - 1,228 3. Secondary and feeder roads - - 1,532 1,532 B. Maintained by Min. of Interior and by local authorities - - 2,317 2,317 1. Feeder roads and tracks - - 2.017 2,017 2. Urban roads - - 300 300 TOTAL 888 2.473 3,849 7,210 1/ Includes Bohicon-Dassa-Zoume-Savalou (119 Km) to be completed in 1980 and the new Nigeria road Source: Directorate of Roads and Bridges, and Mission Data April 1979 - 39 - TABLE 3 BENIN Bids Received (CFAF) Bid 1 Bid 2 Bid 3 Contractor: Fougerolle/Colas Razel/Grun & Billfinger Socigt6 des Grands (France) France/Germany Travaux de 1'Est (SGTE) Lot 1 Two sections of the Godomey- Bohicon road (17 km) 230,885,984 304,175,839 289,179.720 Lots 2 & 3 Parakou-Sori road (148.1 km) 1,609,553,763 1,611,632,883 1,637,797,296 Total Construc- tion 1,840,439,747 1,915,808,722 1,926,977,016 Housing 1/ 91,526,680 78,447,500 39,009,600 TOTAL 1,931,966,427 1,994,256,222 1,965,986,616 1/ The cost of housing was not evaluated. Sources: Letter from Government dated November 17, 1973 (Bids 1 & 2) Contract 20/74 - Bid 3. - 40 - TABLE 4 BENTN Road Maintenance Expenditures, 1970-1978 (Million CFAF) 1970 1971 1972 1973 1974 1975 1976 1977 Expenditures for Road Maintenance 228 401 479 402 401 404 468 445 Source : Directorate of Roads and Bridges BENIN Credit 415-DA, Second Highway Project Project Completion Report Project Costs Appraisal Estimated Final Cost 2/ %Us) (US$O0C) 1/ (US$'00) CFAF Millions % % Cost IDA USAID GOVT TOTAL IDA USAID GOVT TOTAL IDA USAID GOVT TOTAL IDA F.E Overrun 1. Civil Works Lot 1 Cotonou-Bohicon Road (17 km) 1,662 - 202 1,864 2,805 - 347 3,152 671 - 83 754 89 74 69 Lots 2 & 3 Parakou-Sorri Road (148 km) 7,944 - 967 8,911 10,705 - 1,323 12,028 2,560 - 316 2,876 89 77 35 Lot 4 Sorri-Malanville Road (169 km) - 8,000 - 8,000 - 12,000 430 12,430 - 2,870 103 2,973 0 75 55 Sub-Total Civil Works 9,606 8,000 1,169 18,775 13,510 12,000 2,100 27,610 3,231 2,870 502 6,603 2. Supervision Lots 1 - 4 1,237 - 83 1,320 2,030 - 130 2,160 486 - 31 517 94 76 64 3. Consultants Services for Road Maintenance 667 - 68 735 759 - 76 835 180 - 20 200 90 90 14 4. Construction Materials and Equipment 254 - 7 261 450 - 14 464 108 - 3 111 97 97 78 T,,n Pro,,,t Cmst I I 766 81000 1,327 2110111 16 769 12 1Z 101 42, . Total without US Aid Project 13,091 18,639 4,458 90 81 42 1/ Appraisal Rate of Exchange - $1.0 - CFAF 230 2/ AhNumed Avernge Ratce of Exchange - $1.0 - CFAF 239.2 - FFR4.18 DM2.35 . CAN$1.07 3/ Includes $256,000 subcontracts for pave' trief and new water main - 42 - TABLE 6 BENIN Credit 415-DA, Second Highway Project Schedule of Disbursements ACCUMULATED DISBURSEMENTS DISBURSEMENTS US$'000 EQUIVALENT AS A PERCENTAGE OF IBRD I APPRAISAL ACTUAL (UP TO FISCAL ESTIMATED OR LATEST SEMESTER) AND YEAR AND ACTUAL TOTAL APPRAISAL INEW ESTIMATE FOR FUTURE SEMESTER DISBURSEMENTS ESTIMATE SEMESTERS (1-2)x100 or (4-2)xlO 1974 1st 2nd 2,300 1975 1st 2,400 4,750 51 2nd 4,300 6,600 65 1976 1st 4,775 8,800 53 2nd 8,272 10,500 79 1977 1st 10,672 11,375 91 2nd 14,200 11,800 128 1978 1st 16,000 136 2nd 16,304 138 1979 1st 16,304 138 2nd 16,530 141 1980 1st 16,900 1 Closing Dates: 12/31/79 11/30/76 1/ includes $310,000 for shoulder maintenance brigade. - 43 - TABLE 7 BENIN Average Daily Traffic A. Parakou-Malanville Road Type of Vehicles 1973 1974 1975 1976 1977 Passenger cars and minibus 43 43 61 42 54 Pick-up trucks 21 29 40 39 61 Trucks 2-10 tons 26 36 54 49 68 Trucks exceeding 10 tons and tractor trailers 51 46 68 58 67 141 154 223 188 250 B. Two Sections of Cotonou-Bohicon Road Godomey Section 1974 1975 1976 Passenger car and pick-ups 1,789 1,024 918 Light trucks 665 489 583 Medium trucks (2-10 t) 569 346 451 Heavy trucks, tractors trailers 53 37 74 Total 3,076 1,896 2,026 Sehoue Section Passenger cars and pick-ups 365 413 301 Light trucks 151 179 45 Medium trucks (2-10 t) 200 134 58 Heavy trucks, tractors trailers 20 38 21 Total 736 764 425 Source: Directorate of Roads and Bridges (1978) Note: Counts represent average of two traffic counts. Traffic on the Parakou-Malanville road is counted at site traffic stations. The swings in traffic 1975-76 on the Godomey section is due to traffic diversion during construction. - 44 - TABLE 8 BENIN Vehile Operating Costs A. Estimated at Appraisal 1/ CFAF per vehicle/km (at Paved road in good Gravel standard condition "with" road or paved road the project in Poor condition "without" the _______________ project Passenger car and light commercial vehicle 16.4 23.3 - 25.5 Truck less than 10 tons 32.4 51.7 - 57.4 Heavy truck and tanker 74.8 125.2 -139.9 B. Estimated at Re-evaluation at 1976 Prices Passenger car (Peugeot 500) 22.80 33.06 Pick-up truck (Peugeot 404) 23.12 33.52 Ten ton truck (Berliet) 94.13 154.37 Tractor-trailer (tractor with 24 ton trailer) 114.76 198.50 1/ Appraisal Report, Table 6, p. 2. Note: The medium truck size differs in the two sources of data. -45 - TABLE 9 BENIN Cr. 415-DA, Second Highway Project Re-evaluation of the Economic Rate of Return of Parakou-Malanville (318 km) and the two Sections of the Cotonou-Bohicon Road (17 km) 1976 prices - CFAF'000 i Parakou - Malanville (318 km) Two Sections Cotonou - Bohicon (17 km) Total Costs Savings in V.0.C. Savings in V.0.C. Total Costs Construction, Domestic Regional traffic Construction, Savings in V.0.C. Supervision & Traffic Benin & Niger Supervision & Year Maintenance Maintenance 1975 2,426,831 - - 9 - - 0 - 324,395 - 0 - 76 695,895 284,139 1,003,773 90,675 70,287 77 2,147,648 809,008 1,334,805 286,698 89,197 78 771,719 852,418 1,404,215 103,104 94,477 79 152,640 899,775 1,484,303 8,256 99,044 80 152,640 947,131 1,564,391 8,256 104,325 81 152,640 994,488 1,644,480 8,256 110,604 82 152,640 1,049,737 1,735,247 8,256 116,884 83 152,640 1,104,986 1,831,352 8,256 123,020 84 699,600 1,172,075 1,938,137 8,256 129,300 85 152,640 1,235,217 2,044,921 8,256 136,436 86 152,640 1,302,305 2,151,706 37,840 144,428 87 152,640 1,353,608 2,237,133 8,256 150,136 88 152,640 1,404,911 2,322,561 8,256 155,987 89 152,640 1,456,214 2,407,988 8,256 161,696 90 152,640 1,507,517 2,493,416 8,256 168,975 91 152,640 1,574,606 2,605,539 8,256 176,253 92 699,600 1,677,2,2 2,712,324 37,840 183,389 93 152,640 1,704,836 2,819,108 8,256 190,667 94 152,640 1,767,978 2,925,893 8,256 197,803 95 152,640 1,846,906 3,054,034 8,256 205,081 96 152,640 1,925,833 3,lo2,ii.) 8,256 213,787 Economic Rate of Return: Parakou-Malanville: Domestic traffic = 15.65% Regional traffic = 29.95% 2 Sections of the Cotonou-Bohicon Rd: = 14.75% Total Project: = 15.35% NOTES: Construction costs are distributed according to rate of disbursements as follows: 41% in 1975, 10% in 1976, 36% in 1977 ana 13% in 1978. Supervision costs by Consultants is attributed 90% to the Parakou-Malanville road and 10% to the two sections of the Cotonou-Bohicon road. Routine maintenance cost is estimated at CFAF 480,000/km. No routine maintenance cost is included for the first year after completion of construction, as this cost is covered by the contractor. The cost of single surface bituminous treatment every 8th year is estimated at CFAF 2.2 million per km. - 46 - November 1981 ADDENDUM TO PROJECT COMPLETION REPORT BENIN SECOND HIGHWAY PROJECT (CREDIT 415-DA) I. Introduction 1.01 A PCR for this project was circulated on August 31, 1979, which stated that though the main components were complete, there remained a sur- plus of about $4 million and that disbursement was still continuing for equipment procurement. This addendum brings the PCR up to date. II. Parakou-Malanville Road 2.01 This road has continued to perform well, except for the shoulder maintenance problem caused by the narrow pavement (PCR,para. 3.10). Traffic has exceeded estimates. Due to data availability actual traffic figures for two counting stations representing two sections on the road 1/ have been used from 1976 to 1981 and projected for later years using the same growth rates as in the PCR. The recent traffic data do not allow differentiation between benefits accrueing to Benin and to Niger so the 60 : 40 split has been assumed as before. The rate of return has been re-estimated using the same basic data and assumptions as used in the PCR except for the mod- ified actual traffic data. Some other minor modifications were also made which may have contributed to the moderate reduction in the ERR presented in the PCR 2/. The re-estimated rates of return are 28% for Section 1, 23% for Section 2 and 24% for the road as a whole. The returns to Benin only were 18%, 13% and 14% respectively. The traffic counts do not permit an accurate determination of generated versus normal traffic growth. However, if for sensitivity purposes one assumes that 50% of the estimated VOC savings were from generated traffic, thus reducing the annual benefits by 25%, then the overall ERR would be 18%. Revised tables 7 and 9 for the Parakou-Malanville road are attached. 1/ The sections are Parakou - N'Dali (60 km) and N'Dali - Malanville (259 km). In the re-estimation data from one counting station on each section was used rather than an average of six stations for the road as a whole in the PCR. This change was necessary due to the limited number of stations for which 1981 data were available. 2/ A lower unit VOC saving for the 2 - 10 ton trucks was used (the average of a pick-up truck and a 10 ton truck instead of a 10 ton truck) and only 50% of the potential VOC savings were considered for 1976, since only part of the road was ready. - 47 - III. Two Sections of the Cotonou-Bohicon Road 3.01 The southern section (Godomey Dike) is still in acceptable condi- tion and the geotechnical problem that was so troublesome during construction appears to have been solved (PCR, paras. 3.14-3.19). However, the northern section, Sghoug-Zakpo (17 km) is now showing distress - deformation and edge cracking - as a result of foundation settlement. Ground conditions over much of this road are similar - large thicknesses of poorly consolidated alluvial deposits - and the two sections reconstructed under this project were those showing serious deterioration in 1972. Subsequently the problem showed up in other areas, which are being reconstructed under the Third Highway Project. Funds permitting, the Sehoug-Zakpo section will be repaired under this project. Traffic on these sections has been affected by construction on the rest of the road, but there is no reason to doubt that the work remains justified; the rate of return has not been re-estimated. IV. Use of Surplus Funds 4.01 The First Feeder Roads Project (Cr. 717-BEN) experienced large cost overruns and it was decided to prepare a follow-up project as soon as possible. However, it was necessary to finance the technical assistance team from when funds were to run out until the new credit became effective. In reply to a memo from RVP dated March 20, 1980, Mr. Stern agreed to the use of surplus 415-DA funds for this purpose, and by September 1, 1981, approximately $890,000 had been disbursed. The second feeder roads credit (Cr. 1090-BEN) was signed on January 14, 1981, and no further disbursements under this heading are expected. V. Financial Situation 5,01 Disbursement is now complete for procurement of the shoulder maintenance equipment and technical assistance to the feeder roads project. The USAID-financed contractor has dropped his claim for interest on late payments. 5.02 The total amount disbursed at September 1, 1981 was $17.744 million leaving a balance of $3.056 million which will be cancelled. A revised table 5 is attached which shows the latest disbursement position. BENIN TABLE 5 a Tsed 09.22.81 Credit 415-DA, Second Highway Project Project Completion Report Project Costs Appraisal Estimated Final Cost 2/ %/US$ (US$'000) 1/ (US$'000) CFAF Millions % % Cost IDA USAID GOVT TOTAL IDA USAID GOVT TOTAL IDA USAID GOVT T(7TAL IDA F.E Overrun 1. Civil Works Lot 1 Cotonou-Bohicon Road (17 km) 1,662 - 202 1,864 2,905 - 359 3,264 695 - 86 781 89 74 69 Lots 2 & 3 Parakou-Sorri Road (148 km) 7,944 - 967 8,911 10,807 - 1,336 12,143 2,585 - 320 2,905 89 77 35 Lot 4 Sorri- Malanville Road (169km) - 8,000 - 8,000 - 12,000 430 12,430 - 2,870 103 2,973 0 75 55 Sub-Total Civil Works 9,606 8,000 1,169 18,775 13;712 12,000 2,125 27,837 3,280 2,870 508 6,659 - - - 2. Supervision Lots 1 - 4 1,237 - 83 1,320 2,026 - 129 2,155 485 - 31 516 94 76 64 3. Consultants Services for Road Maintenance 667 - 68 735 758 - 76 834 181 - 20 201 90 90 14 4. Construction Materials and Equipment 254 - 7 261 359 - 11 370 86 - 3 89 97 97 78 5. Technical Assistance to Feeder Roads - - - 889 - 99 988 213 - .24 237 - - - Project Total Project Cost 11,764 8,000 1,327 21,091 17,744 12,000 2,440 32,184 4,245 2,870 587 7,702 55.9 72.7 47 Total without US AID Project - - - 13,091 19,754 4,729 90 81 42 1/ Appraisal Rate of Exchange - $1.0 CFAF 230 2/ Assumed Average Rates of Exchange - $1.0 = CFAF 239.2 FFR4.78 = DM2.35 = CAN$1.07 - 49 - TABLE 7 (partial) as revised 10.14.81 BENIN Cr. 415-DA - Second Highway Project Average Daily Traffic Parakou-Malanville Road Section 1 Type of vehicle Section 1 Parakou-N'Dali 1/ 1976 1977 1978 1979 3/ 1980 3/ 1981 Passenger cars and minibus 57 57 68 95 133 186 Pick-up trucks 49 49 75 98 127 165 Trucks 2 - 10 tons 59 59 51 67 88 115 Trucks exceeding 10 tons and tractor trailers 44 48 78 94 114 138 209 213 4/ 272 354 462 604 Section 2 N'Dali-Malanville 2/ Passenger cars and minibus 54 54 27 43 70 113 Pick-up trucks 89 89 52 59 66 75 Trucks 2 - 10 tons 68 68 43 42 41 40 Trucks exceeding 10 tons and tractor trailers 80 80 75 84 94 106 291 291 4/ 197 228 271 334 1/ Using counts at Tamarou 2/ Using counts at Guessou Sud 3/ Interpolation between 1978 and 1981 counts 4/ The average traffic for the road indicated in the PCR on the basis of six counting stations was 250vpd. - 50 - TABLE 9 (partial) as revised 10.14.81 BENIN Cr. 415-DA, Second Highway Project Re-evaluation of the Economic Rate of Return of Parakou-Malanville Road (1976 prices - CFAF'OOO) Construction and Maintenance Costs VOC Saving I Section 1 Section 2 Total Section 1 Section 2 Total Parakou- N'Dali- N'Dali Malanville 1975 456,458 1,970,373 2,426,831 -- -- 76 130,898 564,997 695,895 77,454 493,645 571,099 77 403,973 1,743,675 2,147,648 154,908 987,289 1,142,197 78 145,160 626,559 771,719 212,708 807,029 1,019,737 79 28,712 123,928 152,640 265,009 897,779 1,162,778 80 28,172 123,928 152,640 332,122 1,007,137 1,339,259 81 28,172 123,928 152,640 416,349 1,149,865 1,566,214 82 28,172 123,928 152,640 437,166 1,207,358 1,644,524 83 28,172 123,928 152,640 459,025 1,267,726 1,726,751 84 131,595 568,005 699,600 481,976 1,331,112 1,813,008 85 28,712 123,928 152,640 506,075 1,397,668 1,903,743 86 28,712 123,928 152,640 531,379 1,467,551 1,998,930 8' 28,712 123,928 152,640 552,634 1,526,253 2,078,887 E 28,712 123,928 152,640 574,739 1,587,303 2,162,042 89 28,712 1235928 152,640 597,729 1,650,795 2,248,524 90 28,712 123,928 152,640 621,638 1,716,827 2,338,465 91 28,712 123,928 152,640 646,504 1,785,499 2,432,003 92 131,595 568,005 699,600 672,360 1,856,920 2,529,280 93 28,712 123,928 152,640 699,260 1,931,197 2,630,457 94 28,712 123,928 152,640 727,230 2,008,445 2,735,675 95 28,712 123,928 152,640 756,320 2,088,782 2,845,102 96 28,712 123,928 152,640 786,570 2,172,334 2,958,904 Economic Rate of Return Section 1 Section 2 Combined Domestic traffic 1) 18% 13% 14% Regional traffic 1) 28% 23% 24% 1)Assuming 60.5 of benefits relates to domestic Benin traffic as in PCR BENIN SECOND HIGHWAY PROJECT TRANSPORTATION SYSTEM prr. N I G ER - 0 T1-- -LAV LE GS Dehon 8 r AAV ¯- ROAe - WOR .. - - AN AE OD IHA NEAC N ssN zi mO C -D I - ----- - E N R I L A ADNISRA TION RALWAS ¯ .DSRC DNE 15- DA)K0 * EART AROADS SEONC2> H LHAPOEC NCEAAJNA S.RTPAVEOAD _ISRABTTO N T ARAOCKS CPCIR * EOVAT -f-4--,-j A,L0AS 100 5. TIS O 2OK