Document of The World Bank FOR OMCIAL USE ONLY Report No. 6937 BOTSWANA PROJECT COMPLETION REPORT FIFTH ROAD PROJECT (LN 1941-BT) September 15, 1987 Transportation Division Eastern and Southern Africa Regional Office This dhocmet has a resticted dstrbutioo and may be used by recipints only in dhe performance of their offia dtie bs l abtmnts may not oerwbe be disclosed without Wodd Bank authorixatIon. Country Exchange Races 1/ Appraisal Year Average 1980 US$1 = PULA 0.7772 Intervening Year Average 198i 0.8367 1982 1.0297 1983 1.0969 Completior, Year Average 1984 1.2984 Final Disbursement Year 1985 1.9026 Average for Project Implementatiola Period 1981-85 US$1 = PULA 1.2329 ABBREVIATIONS AND ACRONYMS ERR - Economic Rate of Return NDP - National Development Plan RD - Roads Department RTC - Roads Training Center TA - Technical Assistance TRH - Technical Recommendations for Highways TRRL - Transport and Road Research Labtratory VOC - Vehicle Operating Cost VPD - Vehicles Per Day I/ Source. Internetional Financial Statistics. FOR OFFICIAL USE ONLY THE WORLD BANK Washmtngon. D.C 2O433 U.S.A. Oflte ei. iietn'.fwtal September 15, 1987 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Ccmpletion Report on Botswana Fifth Road Project (Loan 1941-BT) Attached, for information, is a copy of a report ent.tled "Project Completion Report on Botswana Fifth Road Project (Loan 1941-BT)" prepared by the Eastern and Southern Africa Regional Office. Further evaluation of this project by the Operations Evaluation Department has not been made. (+ ,,-o7 Attachment This document hu a restrited dbistibution and may be use by recipicals onb in te perfornmac of their official duties. Its contents may not otherwise be dissbkod witlwut World Bank autlfnion. Fo0 OMCIAL US ONLY BOTSWANA PROJECT COMPLETION REPORT FIFTR ROAD PROJECT LOAN 1941-BT TABLE OF CONTENTS PAGE NO. KEY PROJECT DATA............................................. BVALUATION SUMMARY ....... ................................. iv I* INTRODUCTION............................................ 1 II. PROJECT IDENTIFICATION. PREPARATION AND APPRAISAL ....... 3 Identification and Preparation......................... 3 Appraisal ............................................... . .... 3 III. PROJECT IMPLEMENTATION AND COST .................................. * 7 Implementation ................................................ 7 The Project Cost ......................... **** *9*9* 9 Table I - Fifth Road Project Appraisal Estimate and Final Cost ...................... 10 IV. INSTITUTIONAL PERFORMANCE .............. 11 V. ECONOMIC REEVALUATION .......... ................ 12 VI. ROLE OF THE BANK AND LESSONS LEARNED ................... 16 ANNEES 1. Vehicle Operating Costs............................ 17 2. Traffic Data ......*...*..*............**.....*........**...*..... 18 3. Economic Reevaluation .............................. 19 3A. Government Comments on the Project Completion Report of the Third and Fourth Road Project ...... 20 3B. Third Road Project (Loan 1174) - Revised Economic Reevaluation for the Gaborone- Molepolole Road ............................... 25 MAP IBRD 15120 This document has a restricted distribution and may be used by recipients only in the performc of their offlcial duties. Its contents may not othewise be discblosed without World Bak authoulation. (i) BOTSWANA PROJECT COMPLETION REPORT FIFTH ROAD PROJECT LOAN 1941-BT PREFACE Mi) Loan 1941-BT for Botswana's Fifth Road Project of US$17.0 million was signed on February 6, 1981 and was made effective on May 4, 1981. The Loan was closed October 1, 1984 and the last disbursement was made in April 1985, six months after the closing date specifitd in the Loan Agreement. A total of US$12.4 million was disbursed under the Loan and the undisbursed balance of US$4.6 million was cancelled. (ii) The Bank participated in Botswana's 1979-1985 National Development Plan for Roads by financing: (i) Construction to bituminous paved standard of the Dumela-Zimbabwe border road (76 km); the Selebi Phikwe-Sefophe road (25 km); and the access road to the Selebi Phikwe airport (7 km); (ii) Spot improvement to the Sefophe - Tsetsebjwe road (56 km) and Sefophe - Dikabeya road (93 km); (iii) Construction of the soils and materials laboratory building in Gaborone and nine centralldistrict road maintenance depots; and (iv) Consulting services for: construction supervision of roads in (i) and (ii) above; preparation of a pavement strengthening programi study of road design standards; technical assistance to the Road Department (RD) and Road Training Center (RTC). (iii) The Bank, since 1964 (Credit 63-BEC) has provided US$51.9 million for projects in the road sector for construction to bitumen paved standard of mair. roads; selective betterment/spot improvement of minor roads; construction of road maintenance facilities; purchase of road maintenance and laboratory equipment; preparation of the district road maintenance program; preparation of a feeder roads program; improvement and expansion of road maintenance organization; and strengthening of RD and RTC. (iv) This PCR was prepared by the Bank staff based on a mission to Botswana in December 1986, information contained in the project appraisal report, a review of the Bank files including supervision and progress reports, information obtained from the Government and interviews and discussions with the RD. (v) In accordance with the revised procedures for project performance audit repcrting, this Project Completion Report was read by the Operations Evaluation Department but the project was not audited by OED. The draft report was sent to the borrower for c,'ments; huwever, none were received. (it) BOTSWANA PROJECT COMPLETION REPORT FIFTH ROAD PROJECT LOAN 1941-BT KEY PROJECT DATA Appraisal Item Expectation Actual Total Project Cost (US$ million) 25.0 17.9 Underrun (2) 28 Credit Amount (US$ million) 17.0 17.0 Disbursed (US$ million) 12.4 Cancelled (US$ million) 4.6 Date Physical Components Completed 4/84 2/84 Proportion Completed by: Appraisal Completion date (2) 100 100 Actual Completion date (X) - 100 Economic Rate of Return (2) 13.0 26.0 a) Dumela-Zimbabwe 11.0 26.0 b) Selebi Phikwe-Sefophe 17.0 20.0 c) Access Road to the Selebi Phikwe Airport 22.0 5.0 d) Spot Improvement 35.0 82.0 OTHER PROJECT DATA Item Original Plan Actual First Mention in Files or Timetable 09/29/78 Government's Application n.a. Negotiations 11110-14/80 11/10-14180 Board Approval Date 0116/81 01/6181 Loan Agreement Date 02/6/81 02/6/81 Effectiveness Date 05/04181 05/04/81 Closing Date 10/1/84 10/1/84 B r.ower Govt. of Botswana Executing Agency Min. of Works, Transport & Comm. Fiscal Year of Borrower 4/01 - 3/31 CUMULATIVE ESTIMATED AND ACTUAL DISFURSEMENTS US$ MILLIONS FY82 FY83 FY84 FY85 Appraisal 5.9 13.1 17.0 Actual 3.85 9.01 11.09 12.36 Actual as percentage of Estimate 65 69 65 73 Date of Final Disbursement April 1985 (iii) MISSION DATA Month/ No. of Man Date of Item Year Weeks Persons Weeks Report Idontification 1/ 07/79 2.0 1 2.0 7/18/79 Preparation 2/ 08-09/79 2.2 2 4.4 10/9/79 Preparation 12/79 0.8 1 0.8 1/23/80 Preparation 2/ 03/80 0.4 1 0.4 3/26/80 Appraisal 6/10-24/80 3.0 2 6.0 11/20/80 Subtotal 8.4 13.6 Supervision I 2/ 10/81 1.4 2 2.8 12/15/81 Supervision II 2/ 02/82 0.8 1 0.8 4/22/82 Supervision III 10/82 1.4 2 2.8 11/18/82 Supervision IV 2/ 03/83 0.6 1 0.6 3/31/83 Supervision V 2/ 11/83 1.6 1 1.6 1/25/84 Supervision VI 10/84 1.4 1 1.4 11/16/84 Subtotal 7.2 10.0 Total 15.6 23.6 1/ In conjunction with a review of Bank assistance in the Highway sector. 2/ In conjunction with supervision of Third and Fourth Road Projects. (iv) BOTSWANA PROJECT COMPLETION REPORT FIFTH ROAD PROJECT (LOAN 1941-BT) EVALUATION SUMMARY Introduction The Bank has participated in Botswana's road subsector since 1964 when the First Road Project, Credit 63-BEC (US$3.6 million) was provided. Since then, a total of rive loans/credits have been provided totalling US$51.9 million, including the Shashe Infrastructure Project, Loan 776-BT (US$3.5 million for roads). All five road projects were closed at the time of this PCR preparation. The on-going Selebi Phikwe Technical Assistance Project includes funds for detailed engineering design for a road linking Sefophe to the South African border and feeder roads for agricultural development in the Tuli Block area (para. 1.04). Consistent with the long-standing Government objectives in the road sector, the Fifth Five-Year (1979-1985) National Development Plan (NDP) had as its goalst t.he completion to an all-weather paved standard of the North-South road which links Botswana with its two trading border countries and to increase the paved road network by 520 kw. by 1985; the establishment of more effective maintenance organization, facilities and management systems; and the development of a local construction industry (para. 1.05). Objectives The Fifth Road Project's objectives were to support specific goals of the NDP by financing: (a) the completion to paved standard of an extension of the North-South road to link with Zimbabwe; (b) improved road access in arable portions of the Tuli Block and adjacent areas; and (c) strengthening the Roads Department (RD) by financing construction of facilities for the soils and materials testing laboratory and road maintenance depots and by providing a variety of consulting services for RD and instructors for the Road Training Center (RTC) (para. 1.06). Tna project thus included funds for construction of civil works comprising roads and buildings and for consultant services for construction supervision of roads and buildings; preparation of a pavement strengthening (v) program and training of Roads Department (RD) staff in pavement evaluation; study of appropriate design standards for Botswana; and technical assistance for RD and Roads Training Center (RTC) (paras. 2.06-2.12). Implementation Exoerience Road works were started and completed within the implementation schedule estimated at appraisal, while some delays occurred in the completion of building works. The quality of work was generally good. The performance of the consultants was good (paras. 3.02-3.12). The actual costs of the project in US dollars were significantly less than appraisal estimates while actual costs in Pula were close to the appraisal estimates. This resulted in a savings of US$4.6 million (27% of the loan amount) which was cancelled (paras. 3.13-3.14). Results The roads which had marginal benefits at appraisal now show higher benefits probably due to low estimates from the feasibility studies (e.g., for Dumela-Zimbabwe border roae 11% ERR at appraisal compared to a 26% ERR of the economic reevaluation), except for the Selebi Phikwe Airport access road (7 km) where the reevaluated ERR was only 5% because of low actual traffic (-43% compared with SAR) and a high construction cost (+70% compared with SAR) (paras. 5.01-5.14). 8ustainability The roads improved under the project are likely to be maintained at adequate levels even though additional efforts will be required to provide skilled maintenance personnel to fully benefit from the investment made (para. 4.02). In addition, axle load limits are fully enforced (para. 6.02). Under the circumstances the road investments under the project are likely to achieve more than the average level of sustainability. Findings and Lessons An important lesson to be learned from the project is the need for an adequate advance study of the priorities in the sector prior to appraising a project. This is evident from the major improvements made in the project composition at appraisal as compared to project preparation (para. 6.01). The project included construction of axle-load control stations and the Loan Agreement provided a covenant requiring enforcement of axle-load control. Here it is important to note that the weigh stations were completed on time and axle-load control was fully enforced, which is a rare event in developing countries. An assessment of Botswana's success in enforcement of axle-loads would no doubt provide valuable experience for the Bank in assist- ing other countries to succeed in their effort to enforce axle-load control (para. 6.02). BOTSWANA PROJECT COMPLETION REPORT FIFTH ROAD PROJECT LOAN 1941-BT I. INTRODUCTION 1.01 The transport sector of the Republic of Botswana is served by a 6,900 km gazetted road network; 718 km of railway line; one international airport, 3 small domestic airports and several small airstrips; and a ferry operating across the Zambesi River between Botswana and Zambia. The eastern half of the country, which is more populated and developed than the rest of the country, is well served by a developed road system and the 718 km railway line linking Botswana with the Republic of South Africa and Zimbabwe. The western half of the country, which is sparsely populated and has less potential for development, depends for the most part on earth and sand tracks and several scattered small air strips. The o,900 km gazetted road network (430 km paved at independence in 1966) is composed of 1,042 km of bitumen paved; 1,563 km gravel and 4,295 km earth/sand surfaced tracks. 1.02 The Bank has prticipated in Botswana's road subsector since 1964 when the First Road Project, Credit 63-BEC (US$3.6 million) was provided. Since then, a total of five Loan/Credits have been provided totalling US$51.9 million, including the Shashe Infrastructure Project, Loan 776-BT (US$3.5 million for roads). All five road projects were closed at the time of this PCR preparation. The on-going Selebi Phikwe Technical Assistance Project includes funds for detailed engineering design for a road linking Sefophe to the South African border and feeder roads for agricultural development in the Tuli Block area. 1.03 Since the Fifth Road Project became effective, the Ministry of Works and Communication, to which the Roads Department directly reports, was reorganized (in December 1986) to become the Ministry of Works, Transport and Communications. The reorganization which now pools together ander one ministry all the transport modes could have the advantage of providing a better coerdination for developing an integrated inter-modal planning and prograwming for the transport sector. 1.04 Since independence in 1966 the emphasis of the National Development Plans (NDP) in the road sector has been consistent: (i) to reduce the high costs of transport in Botswana resulting from low density of population and the country's landlocked position in part by developing viable international transport systems; -2- (ii) to maximize the benefits from providing transport infrastructure, taking social as well as economic goals into account, and (iii) to reduce, as far as feasible, the vulnerability of the transport system and hence the economy as a whole, to events beyond the control of the country. 1.05 Consistent with the long-standing Government objectives in the road sector, the Fifth Five-Year (1979-1985) National Development Plan (NDP) had as its goals: the completion to an all-weather paved standard of the North-South road which links Botswana with its two trading border countries and to increase the paved road network by 520 km by 1985; the establishment of more effective maintenance organization, facilities and management systems; and the development of a local construction industry. 1.06 The Fifth Road Project's objectives were to support specific goals of the NDP by financing: (a) the completion to paved standard of an extension of the North-South road to link with Zimbabwe; (b) improved road access in arable portions of the Tuli Block and adjacent areas; and (c) strengthening the Roads Department (RD) by financing construction of facilities for the soils and materials testing laboratory and road maintenance depots and by providing a variety of consulting services for RD and instructors for the Road Training Center (RTC). 1.07 The NDP received in addition to support from the World Bank assistance of about US$30 million from: the European Economic Community (EEC), the US Agency for International Development (USAID), Germany's Kreditanstalt fur Wiederaufbau (KFW), the African Development Bank (AfDB), the Swedish International Development Agency (SIDA), the Norwegian Agency for Development (NORAD) and the Danish International Development Agency (DANIDA). - 3 - II. PROJECT IDENTIFICATION. PREPARATION AND APPRAISAL A. Identification and Preparation 2.01 The Fifth Road Project of Botswana was first discussed in September 1978 during the Bank's annual meeting. In this meeting the Botswana delegation explained that Government's priorities for the road subsector would be specified in the Fifth Five Year (1979-1985) National Development Plan (NDP) expected to be published in mid-1979. Possible project components raised by Government officials included the Tuli Block area roads whose feasibility studies and detailed engineering were financed under the Third and Fourth Road Projects and the Francistown-Orapa road which had suffered from rapidly growing traffic levels. 2.02 A Bank mission in September 1979 identified the project components based on the NDP which at that time was in final review stage. Except for the Nata-Maung road (306 km), feasibility and/or detailed engineering for the roads proposed for inclusion in the project were financed under the Third and Fourth Road Projects. The preparation mission recommended the following components for inclusion in the Fifth Road Project. (a) Civil Works: (i) the construction to paved standard of the Dumels-Ramokgwebana (65 km) scJction of the Dumela-Zimbabwe Border Road; the Selebi Phikwe-Sefophe (23 km) and Palapye jct. - Moeng (42 km) roads in the Tuli Block area; and (ii) Spot improvements of gra;vel/sand surfaced roads of about. 450 km in the Tuli Block area; and (b) Consultant services for: (i) construction supervision of (al above; (ii) detailed engineering of the Nata-Maung road (306 km); and (iii) assistance to implement i Roads Department training program. 2.03 The Dumela-Ramokgwebana road (65 km), the Selebi Phikwe-Sefophe road (23 km) and the Palapye jct. - Moeng road (42 km) had ERRs ranging from 8.5% to 12.3%. The 450 km roads proposed for spot improvement had high ERR's ranging from 152 to 65%. While Government's proposal was for inclusion of the entire 76 km of Dumela-Zimbabwe border road, the mission recommended that the last 11 km to the border be left out because of its low ERR (about 22). The low ERR for this section was mainly due to restriction of traffic to eross the border because of the political situation and security problems which existed at that time. B. Appraisal 2.04 The Project was appraised in June 1980 as agreed upon during the project preparation mission and the appraisal report was completed in November 1980. The composition of the Project included subproject components which vary significantly from those proposed during project preparation (Para. 2.02). Because of additional information and data obtair.ed, it was possible at appraisal to define better the composition of the project. The project consisted of the following components: -4- (a) Construction to bituminous paved standards; (i) Dumela - Zimbabwe border road (76 km); (ii) Selebi Phikwe-Sefophe road (25 km) and a 7 km access road to the Selebi Phikwe Airport serving the Tuli Block area; (b) Spot improvement of the Sefophe - Tsetsebjwe (56 km) and Dikabeya-Sefophe (93 km) roads in the Tuli Block and adjacent areas; (c} Construction of RD's soils and materials laboratory building and nine road maintenance depots; (d) Consultant services for: (i) Construction supervision of (a), (b) and (c) above; (ii) Preparation of a pavement strengthening program for about 80 km of main roads and training of RD staff in pavement evaluation; (iii) Determination of appropriate road design standards for Botswana and preparation of an RD design manual; (iv) Technical assistance to RD; and (e) The recruitment of two instructors for strengthening the Road Training Center (RTC) training staff. 2.05 The underlining reasons for the changes made at appraisal in project composition compared to that at project preparation are briefly discussed in the following paragraphs (Paras. 2.06 to 2.12) (a) Construction of Roads to Bituminous Paved Standards Dumela-7imbabwe Border Road 2.06 At the time of the project preparation missions (September and December 1979) the last 11 km section of the Dumela-Zimbabwe border road (76 km) was not found to be justified (Para. 2.03) because of the low traffic counts (16 vpd) and the resultant low ERR of 22. However, at appraisal time, ten months later, the traffic restriction at the border being lifted, following Zimbabwe's independence, traffic on this section had increased from 16 vpd to 75 vpd. Moreover, the traffic was forecasted to increase to 115 vpd by 1983, the time when the construction of the road t.z paved standard would be completed. Based on the new traffic forecast, the ERR of the last 11 km increased from 2% prior to September 1979 to 7% at appraisal time in June 1980 giving an average ERR of 11% for the total 76 km to the Zimbabwe border. Thus the full 76 km length of the road to Zimbabwe border was included for financing under the project. Selebi Phikwe-Sefophe Road & 7 km Access Road to Selebi Phikwe Airport 2.07 The Selebi Phikwe-Sefophe road was identified to be 23 km in length at project preparation time. At appraisal the length was found to be 25 km and a new component of a 7 km access road to the newly completed Selebi Phikwe Airport was added. The Selebi Phikwe-Sefophe road (25 km) had an ERR of 17% and the 7 km access road to the Airport had an ERR of 22%. The above ERRs were to construct these roads to a 6.Om wide pavement with a single bituminous surface dressing and a structural strength for a maximum of 8.2 tons single axle-load usi.ng South Africa TRH 4 design method. Palapye jct. - Moeng Road 2.08 The Palapye tct. - Moeng (42 km) road identified for financing during project preparation in September 1979 was left out from the project at appraisal due to higher priority subprojects and because of its marginal ERR. (b) Spot Improvement 2.09 The spot improvement subproject proposed for inclusion under the project during project preparation was curtailed at appraisal due to higher priority subprojects and because of a low ERR. The roads included were Sefophe-Tsetsebjwe (56 km) and the Dikabeya-Sefophe (93 km). These two roads had estimated ERRs of 23% and 56% respectively. The high benefits for these roads were mainly due to the low level of spot improvements recommended for these roads which were estimated to cost about US$7,000 per km. The works to be carried out were limited to improveient of drainage, shaping and grading and the provision of minor structures at the most seriously affected sections to enable trucks to pass during all but the heaviest rains. (c) Construction of Buildings Road Department's Soils and Materials Testing Laboratory 2.10 The construction of a soils and materials testing laboratory building not discussed during project preparation was justified because the existing laboratory space, housed at the Meteorology Department, was inadequate. Furthermore, the RD had to expand its soils and materials testing activities to accommodate the need to test bituminous material and to identify additional sources of road materials required for the spot improvement program planned for the road network. Moreover, construction of a new soils and materials laboratory building adjacent to the RD's training center was considered appropri>e since it would have the advantage of using the facilities for training of RD's staff in road material testing. Therefore, the financing of the construction of an independent soils and materials testing laboratory was considered to have a major impact on institution building in the roads subsector. Through this provision RD would be able to carry out required material tests departmentally thereby reducing the high cost of having such tests carried out by consulting firms as was the case in the past. -6- Road Maintenance Depots 2.11 Even though there was no provision made during project preparation it was found essential at appraisal to provide funds for construction and enlargement of nine central/district road maintenance depots to serve RD's needs for equipment maintenance and repair and for storage of spare parts and road maintenance materials and tools. Since independence in 1966, Botswana's road network had expanded from about 430 km to about 6,900 km thereby necessitating central/district road maintenance depots. A USAID financed highway maintenance study had also recommended the need of providing adequate road maintenance depots. (d) Consultant Services 2.12 The other two components added during appraisal were the pavement strengthening program and the preparation of a manual for road design standards applicable for Botswana. The pavement strengthening program was to assist the RD in carrying out a road condition survey on existing roads (about 80 km) constructed to paved standard soon after independence which had started showing deterioration. Carrying out of pavement condition investigations before these roads showed serious deterioration was technically justified to avoid high reconstruction costs if these roads were left unattended. The preparation of the manual for road design standards was necessitated by the Goverment's concern about the variety of design standards recommended by consultants employed for detailed engineering of roads. In particular, South African and UK pavement design methods were indifferently selected, without due regard to their comparative merits in Botswana. III. PROJECT IMPLEMENTATION AND COST 3.01 The project was approved by the Board on January 6, 1981, was signed on February 6, 1981 and became effective on May 4, 1981. There were no special conditions for Loan effectiveness. A. Implementation Dumela-Zimbabwe Border Road (76 km) 3.02 The contract for construction of the Dumela-Zimbabwe border road was awarded in October 1981, five months after Loan effectiveness and the work was completed at end of November 1983 as compared to the contractual completion date of October 22, 1983. The contract was awarded on the basis of International Competitive Bidding (ICB' in accordance with Bank guidelines; bids were received from nine prequalified firms and the contract awarded to the lowest evaluated bidder. The time taken from the daze of Loan effectiveness to contract award and from contract award to completion of works is considered efficient by any standard and the quality of finished work was good. 3.03 Variation orders for civil works amounting to Pula 1.9 million were issued during the construction period. This amount was to cover costs of additional works carried out on junctions leading to villages, cattle grids, access roads and gates to telecommunications repeater stations, a short realignment of the route, construction of an approach road to the Tati River Bridge and construction of a short section of dual carriageway and additional surface area for a weigh-bridge station at Ramokgwebana (Zimbabwe border). The additional works were fully justified. 3.04 The actual construction cost for this road was Pula 12.6 mi:lion (Table 1) including Pula 1.9 million for variation orders and Pula 1.3 million for payment of claims. Selebi Phikwe-Sefophe Road (25 km) and 7 km Access Road to Selebi Phikwe Airport 3.05 The contract for the construction of the two roads was awarded in September 1981, four months after the date of effectiveness and construction work was completed in March 1983, three months after th6 contractual completion date of December 1983. The contract was awarded on the basis of International Competitive Bidding (ICB) in accordance with Bank guidelines; bids were received from five prequalified firms and the contract was awarded to the lowest evaluated bidder. During the construction period variation orders were issued to carry out: surfacing of an access road to the Independence road; short realignment of the route and removal of black cotton soil over and above what was originally envisaged. The contractor was given a three-month time extension to compensate time required for the additional works. Both the time elapsed from the date of Loan effectiveness to contract award and time taken to complete the construction is considered efficient. -8- 3.06 The actual construction cost for the two roads work was Pula 4.01 million (Table 1) including Pula 0.2 million for variation orders for civil works. There was no claim by the contractor on this contract. Spot Improvement (149 km) 3.07 The contracts for spot improvement of the two roads were tendered together with the Selebi Phikwe-Sefophe road and were awarded to the same contractor for completion within the same contract period. The actual spot improvement cost was Pula 0.68 million as compared to Pula 0.45 million estimated at appraisal. The quantities for the spot improvement works were estimated on experience because quantity estimates using detailed engineering were not warranted for this type of spot improvement. Thus unit rates quoted for work activities were the contractual commitments for this work rather than unit rates for spec0fied quantities. Soils and Materials TestinR Laboratory Building and Road Maintenance Depots 3.08 The project provided for the construction of a new soils and materials laboratory for RD as well as for the construction of one new road maintenance central depot, one new road maintenance district depot and the extension of seven existing road maintenance depots. Consultants for the detailed engineering, preparation of bidding documents and construction supervision for these components were employed in September 1982, fifteen months after the date of Loan effectiveness. The prolonged delay in the start of these components was mainly because there was no advance preparatory work undertaken prior to appraisal as had been the case for the road components. Secondly, RD was understaffed and the road components were given priority. Tenders for the works were received in June 1983. With the exception of the road maintenance depots in Francistown and Nata, all contracts were awarded in July 1983. The Francistown road maintenance depot had to be retendered in December 1983, since only one bid was received which was much higher than the Government estimate. The work for the Nata road maintenance depot involved only a loading ramp, and the RD made arrangements to have the work carried out by force account and the work was financed by the Government. Consultant Services 3.09 Consultants were employed for construction supervision of the Dumela-Zimbabwe Border road, and the Selebi Phikwe-Sefophe road and the 7 km access road to Selebi Phikwe airport. For the Dumela-Zimbabwe border road three firms submitted proposals and contract was awarded to the lowest evaluated bidder on October 9, 1981. For the Selebi Paikwe-Sefophe and the 7 km access to the airport roads, five proposals were received and the contract was awarded to the lowest evaluated firm (joint venture of two firms) on August 12, 1981. 3.10 A consultant for the preparation of a pavement strengthening program was selected from proposals received from five firms based on the evaluation of technical proposals. The contract was awarded in December 1981 and work was completed in October 1983. About 470 km of paved roads were tested and analyzed and in the final result only 5 km of roads were recommended for strengthening. - 9 - 3.11 For the study of Design Standards for Botswana, six proposals were received and a consortium of four firms was selected And a contract signed on September 3, 1981. Government subsequently requested, and the Bank agreed that the consultants also study and revise as necessary the Ministry of Works' specifications for earthworks with special emphasis on the unstable calerete soils frequently encountered in Botswana. 3.12 While the above consultant services were financed under the project, the TA for the RD and RTC was financed by Government. Overall, the performance of consultants was of good professional standard. B. The Prolect Cost 3.13 The comparison between appraisal estimate and the final cost for the project is shown in Table 1. The total expenditure under the project amounted to Pula 19.42 million. In US dollars total expenditure was US$17.85 million which is 28% less than the appraisal estimate. This is due to changes in the exchange rate between the Pula and US dollar. The exchange rate used at appraisal was one US$ equivalent to Pula 0.78 while the average exchange rate over the implementation period was one US$ equivalent to Pula 1.233. Moreover, for road construction works, monthly certificate payments were composed of 40% in Pula, 45% in South African RAND and 15% in pounds Sterling at a fixed exchange rate of one pula equivalent to 1.002 RAND and one Pula equivalent to 0.566 Sterling Pound. 3.14 Disbursements under the project were US$12.4 million which represents 73% of funds available under the project. The last disbursement was made in April 1985, six months after the project closing cate of October 1. 1984. The undisbursed balance of US$4.6 million was cancelled. BOTSWANA PROJECT CONLETION REPCRT Fiftn Road Project (Ln. 1941 - BT) Fifth Road Project Appraisal Eltstate and Final Cost App a st;tit.e Final Cost 1 Ap,ramnsl Est.ate r#.na; Coet 1/ Local Foreign Total Local Fore5gn ToStal Local Fore;gn Total Local Foroe;q Total S. CtVIL VORS (AR0405) A. Censtruction to Oitumen paved standard (I) Oumela-Z,mbabse border (76 li.) 2.08 6.72 4.60 3.78 8 82 12.60 3.64 6.61 12.30 3.39 7.90 11.20 (11) Selebt Phiku-eSefopho (25 ha) 0.66 1.53 2.19 0.95 2.2t 3.16 0.84 1.97 2.81 0.44 2.30 3.21 (;) Access road to Selebi Plih;e AIrport (7 he) 0.15 0.35 0.51 0.2S 0. 0.85 0.20 0.45 0.6S 0.i6 0.61 0.87 B. Spot lprovementS (1) 5 fop%e-T&tsebj,r road (56 km) 0.07 0.16 0.23 0.00 0.18 0.26 0.09 0.20 0.2" 0.09 0.21 0.30 (Ii) Dtsbays-Sefoph* road ((93 km) 0.04 0.09 0.13 0.13 0.29 0.42 *f 0.05 0.11 0.16 0.15 0.36 OS1 Sub-total 3.80 8.86 12.66 S.19 12.10 17.20J' 4.87 11.34 16.21 4.88 11.38 16.26 II. ciVL HORKS (8UILDDNS) A. Construction of an RD soils and materials laboratory 0.18 0.27 0.45 0.10 0.27 0.I5 0.23 0.3S 0.58 0 16 0.24 0.40 S. Construction and enlargeent of 9 road maintenance depots 0.19 0.28 0.47 0.24 0.35 0.59 0.24 0.36 0.60 0.13 0.20 0.33 Sub-total 0.37 0.SS 0.92 0.42 0.62 1.04 0.47 0.71 1.18 0.29 0.44 0.73 O Subtotal I and lI 4.17 9.41 13.58 5.61 12.72 18.33 5.34 12.0S 17.39 S.17 11.82 16.99 ?tU. CONSULTANT SEVtCES A. Construction Supervison of I 0.20 0.78 0.90 0.16 0.63 0.79 0.2S 1.00 1.25 0.12 0.46 0.58 8. ConstructIon Supervision of It 0.01 0.01 0.01 0.01 C. Preparation of a pavement, strengthening program and tra;ntn* tD etaff in pavement evaluation 0.02 0.07 0.09- 0.04 0.14- 0.18 0.02 0.09 0.11 0.03 0.13 0.16 0. Study of appropriate design Standards for Baotaans 0.02 0.07 0.090 0.02 0.10 0.12 0.02 0.09 0.11 0.02 0.10 0.12 E. Technical Assistance to RD 0.03 0.1i 0.)6 0.04 0.16 0.20 Sub-total 0.27 1.06 1.33 0.22 0.67 1.09 0.33 1.3S 1.68 0.17 0.69 0.86 Iv. RTC NSTRUCTORS 0.06 0.2S 0.31 0.08 0.32 0.40 B re Cost January 1981 4.S0 10.72 15.22 5.83 13.59 19.42 5.75 13.72 19.47 5.34 12.Sl 17.8S V. CoNttNCCIES A. Physical (lox) 0.45 1.07 1.52 0.59 1.37 1.95 8. Pric. (18o) 0.8o 1.93 2.74 1.04 2.47 3.50 Sub-total 1.26 3.00 4.26 1.61 3.04 5.45 TOTAL S.76 13.72 19.40 5.03 13.59 19.42 7.38 17.56 24.92 5.34 12.61 17.85 Total excluding taes and duties 4.20 13.72 17.92 5.40 17.SS 22.96 11 Source. Botswana Roads Department - 11 - IV. INSTITUTIONAL PERFORMANCE 4.01 The project included a component for the construction of a soils and materials testing laboratory to strengthen RD's institutional capability to meet its increasing needs to test bituminous materials and additional sources for road-making materials required for the spot improvement program planned for the entire gravel/sand surface road network and for paved road shoulder maintenance. The RD now has adequate and well laid out laboratory space and sufficient laboratory equipment to carry out departmentally most of the testing needed. The Transport and Road Research Laboratory (TRRL) of UK has been closely associated with RD's laboratory and RD staff have benefited significantly. However, because of shortages of local staff RD does not seem to have sufficiently benefited from the transfer of knowledge. For the benefits to accrue from the investments made for the construction of the laboratory building and purchase of laboratory equipment, it would be very important for RD to train more laboratory technicians using its well established training center. Provided efforts are made to adequately staff the new laboratory the RD should be able to carry out most of the soils and materials testing required for road work as well as to assist other sectors such as those in the building area. 4.02 The construction and/or enlargement of the eight road maintenance depots financed under the project helped strengthen RD's institutional capability. These facilities provided adequate working space for servicing and repair of road maintenance equipment and storage spaces for spare parts and road maintenance materials and tools. Road maintenance units are now close to work sites thus avoiding long distance crew mobilization costs as well as reducing daily travel time to work sites. However, the RD has to make additional effort to provide the required number of skilled maintenance personnel to fully benefit from the investment made. 4.03 The consultants engaged for the study of an appropriate design standards manual completed the work in September 1983. The availability of design standards applicable to Botswana's needs aided design works both by consultants and departmental staff. It also helped to reduce the decision making process on technical issues as well as enabling the standardization of construction methods. The manual also served as a source document for training road construction and maintenance supervisory staff. 4.04 The Loan covenants were met by the Government. Axle-load weigh stations were constructed and axle-load control was fully enforced. A high level of performance was achieved in road maintenance and budgetary allocations have been sufficient. Due to shortage of staff, the RD was not able to prepare the project completion report as specified in the Loan Agreement. - 12 - V. ECONOMIC REEVALUATION Objectives 5.01 The appraisal report (SAR) states that "the project's objectives are to support specific goals of Botswana's National Development Plan for 1979- 1985 which calls for the upgrading of the basic road network, encouraging rural development and strengthening road maintenance. Specifically, the project willt (a) complete the paving of an extension of the North-South road to link with newly independent Zimbabwe; (b) improve road access in arable portions of the Tuli Block and adjacent areas; and (c) strengthen the Road Department by constructing facilities for soils and materials testing and road maintenance depots and by providing a variety of consulting services and instructors for the Roads Training Center. 5.02 The subject of the economic reevaluation was the same project components listed in the SAR, namely: (a) construction to bituminous paved standards of the following roads: (i) Dumela-Zimbabwe border road (76 km); and (ii) Selebi Phikwe-Sefophe road (25 km) and a 7 km access road to the Selebi Phikwe airport serving the Tuli Block area; and (b) spot improvements of the Sefophe-Tsetsebjwe (56 km) and Dikabeya-Sefophe (93 km) roads in the Tuli Block and adjacent areas. Overall Assessment 5.03 Most of the reevaluated components (except the access road to the Selebi Phikwe airport) exceed by far the appraisal expectations. The combined ERR of the project at appraisal was 13% compared to 262 in the present reevaluation. Nethodology 5.04 The economic reevaluation was done comparing the "with" and the "without" project scenarios. The economic reevaluation was carried out in 1985 prices with necessary conversions of the construction and supervision costs using appropriate inflator factors to foreign and local costs. No shadow pricing of foreign exchange was applied since there is no marked divergence between the official and the market exchange rates between the Pula and the US dollar. - 13 - 5.05 The quantified benefits used in the reevaluation are vehicle operating cost (VOC) savings and maintenance cost savings, except for spot improvement works where the latter were not considered (as in the SAR). 5.06 Actual construction costs net of taxes were used for the recalculation of the ERRs. For spot improvement works on the Dikabeya- Sefophe and Sefophe-Tsetsebjwe roads, however, actual costs provided were lumped in one stream; thus, the reevaluation combines the analysis of the two roads. Supervision costs for the Selebi-Phikwe road, the airp.'rt road, as well as for spot improvements were provided by the Government under one umbrella payment since they were performed by the same consultant. Therefore, supervision costs were assigned to each road at a share of about 62 of total costs. 5.07 Vehicle operating cost' unit savings were broken down according to the three categories listed in the actual traffic counts: a) cars/vans; b) commercial vehicles; and c) buses (Annex 1). The economic life of the components was assumed to be, as in the SAR, 20 years for paved roads and 7 years for the spot improved earth roads. Based on the inputs mentioned above the ERRs for the individual components are presented below: Dumela-Zimbabwe Border Road (76 km) 5.08 The actual traffic data (Annex 2) was provided by Government for the road as a whole, Instead of a breakdown into four sections as in the appraisal report. Therefore, the present analysis is based on the road as a whole. A uniform 7% annual growth rate was applied in the SAR. Actual traffic,2 however, was 492 higher in 1984, and 85% higher for 1985 than estimated in the SAR. Therefore, in the reevaluation the actual growth rate was adjusted to 122 annual growth up to 1992, and 7% thereafter. 5.09 The reevaluated ERR for this component is about 26%, as compared to about 11% at appraisal (Annex 3). The difference between these two ERRs can be explained by the marked increase in traffic (para. 5.08), which was underestimated at appraisal time. An unlikely reduction In benefits by 202 would reduce the ERR to 22%. Selebi Phikwe-Sefophe (25 km) 5.10 The actual traffic count for 1984 was about 23% higher and in 1985 15% higher than the appraisal estimate, therefore a uniform growth rate of 1/The VOCs were presented in the consultancy study of the Viability of Roads in the Tuli Block (Supplementary Study), Final Report, June 1986, prepared by ZMCK Consulting Engineers, page 13-20. 2/Twenty percent of the traffic on this road is generated traffic to which half of the benefits of VOC savings of normal traffic were assigned, as per SAR. - 14 - about 8% was used in the reevaluation as opposed to the SAR, which applied a 7% rate from the opening year of the road up to 1992 and 4% thereaftet.3 As in the SAR the present analysis considers a savings of 5 km in trave' distance. The reevaluated ERR for this component is about 20%, compared with the 17% estimated at appraisal (Annex 3). A reduction in benefits by about 20% would yield an ERR of about 16%, making the construction of this road economically justified. Selebi Phikwe Airport Road (7 km) 5.11 The average daily traffic for late 1986 was 128 vehicles, as compared with 223 vehicles estimated at appraisal, i.e., a 43% shortfall, due to a decrease in economic activity in the mining area. The reevaluated ERR for this road, assuming the same traffic growth as in the SAR: 52 until 1990 and 2.5% thereafter, was about 6%. Assuming a lower growth rate of 2% throughout, the ERR is about 5% compared to the appraisal estimate of 22% (Annex 3). This discrepancy can be explained (in addition to the traffic decrease) by the 70% increase in construction cost per km. This is the only road under the project with a reevaluated ERR of less than 10%. Dikabeva-Sefophe (93 km) and Sefophe-Tsetseblwe (56 km) 5.12 The actual traffic for 1984 was 60 vehicles per day, or about double the estimate at appraisal. Therefore, in the reevaluation, the assumed annual growth rate was increased to 4% as opposed to 1.5% - 2% at appraisal. The reevaluated ERR for this component is about 82% as compared with a 35% weighted average ERR for the two roads ±n the SAR (Annex 3). This difference is explained by the fact that actual traffic almost doubled by 1985 from 32 to 61 vehicles per day. An unlikely reduction in benefits of 20% would yield an ERR of 70%. Timing 5.13 Regarding the timing of implementation, the first year returns for three of the components (a. Dumaela-Zimbabwe; b. Selebi Phikwe-Sefophe; and c. Dikabeya-Sefophe and Sefophe-Tsetsebjwe) exceed the current estimated opportunity cost of capital for Botswana of 6%. They range from 14 to 45%, which indicates that these components were overdue. As for the access road of the Selebi Phikwe airport, the first year's return (5%) is below the assumed opportunity cost of capital indicating that the implementation of this road was premature. To be economically justified, this road should have been constructed to a lower standard. 3IFor 1987, the average daily traffic reported was 279 vehicles. This figure seems to be out of line with the trcffic figures for 1984, with an average daily traffic of 164 vehicles, and for 1985, with an average daily traffic of 165 vehicles. The present evaluation is based on 198415 figures. If the analysis had been based on the 1987 traffic level, the ERR would have been higher. - 15 - Third Road Project LN 1174-BT Revised Economic Reevaluation for the Gaborone-Molepolole Road 5.14 Government comments on the Project Completion Report of the Third and Fourth Road Projects4 reached the Bank after that PCR was issued, therefore, the comments could not be incorporated at that time. The Government comments (Annex 3A) indicate that the traffic volume used for the reevaluation of the Gaborone-Molepolole road in the framework of the Third Road Project was low and did not conform with the records contained in the Transport Statistics of the Ministry of Works and CommuU.Lcations and with counts undertaken by the Road Department. Based on the new data submitted, a new reevaluation was prepared (Annex 3B). A/Document * SECM85-865, dated July 19, 1985 - 16 - VI. ROLE OF THE BANK AND LESSONS LEARNED 6.01 In comparison to the project preparation mission report the appraisal composition of the project was better balanced. Provisions were mAde to finance high priority road sections and the institutional building 8pect and the RD*s need for road maintenance facilities were adequately covered in the SAR. An important lesson to be learned from the project is the need for an adequate advance study of the priorities in the sector prior to appraising a project. This is evident from the major improvements made in the project compocition at appraisal as compared to project preparation. The development of a well-balanced project at appraisal was possible because of additional data and information made available after the project preparation mission. 6.02 The project included construction of axle-load control stations and the Loan Agreement provided a covenant requiring enforcement of axle- load control. Here it is important to note that the weigh stations were completed on time und axle-load control was fully enforced which is a rare event in developing countries. An assessment of Botswana's success in enforcement of axle-loads would no doubt provide valuable experience for the Bank in assisting other countries to succeed in their effort to enforce axle-load control. 6.03 The RD has a critical shortage of local engineering staff. As a result, the RD at present engages about 22 expatriate staff mostly to fill permanent line positions. Hence, there is an urgent need for training high school graduates in engineering courses abroad, while developing a local training capacity. - 17 - Annex 1 BOTSWANA PROJECT COMPLETION REPORT FIFTH ROAD PROJECT (LN 1941-BT) VEHICLE OPERATING COSTS 1/ (Thebe 2/ per km) TyRe of Surface Bitumen Gravel Earth/Sand Heavy Sand Type of Vehicle Passenger car, pickup, light vehicle 28.2 39.2 50.2 69.6 Trucks 98.6 134.0 173.7 247.9 Minibus/Bus 58.0 80.0 105.0 150.8 11 ln Botswana currency, 1985 prices. Based on ZMCK Consulting Engineers, Viability of Roads in the Tuli Block (Supplementary Study), Final Report, June 1986, pp. 13-20. The study assumes that VOC apply to flat terrain throughout the country. 2/ 100 thebe - 1 pula Exchange rate in 1985s US$1 = pula 1.9026 - 18 - Annex 2 BOTSWANA PROJECT COMPLETION REPORT FIFTH ROAD PROJECT (LN 1941-BT) TRAFFIC DATA Averate Daily Traffic Road Year SAR PCR 2 Change Dumela-Zimbabwe Border 1984 202 301 + 49 1985 216 399 + 85 Selebi Phikwe-Sefophe 1984 134 165 + 23 1985 143 164 + 15 Selebi Phikwe Airport Road 1986 223 128 - 43 Dikabeya-SefopheI Sefophe-Teeteebjwe (weighted average) 1984 31 60 + 94 1985 32 61 + 91 Source: Republic of Botswana, Ministry of Works and Communications, Statistics Unit, Transport Statistics, Gaborone, 1984, 1985 and 1986. - 1U - Annex 3 BOTSWANA PROJECT COMPLETION REPORT FIFTH ROAD PROJECT (LN. 1941-BT) ECONOMIC REEVALUATION Coaarison of ERRs (2) Reevaluated ERR with First Benefits Year Roads SAR LCR -20C Raturn 1/ Dumela-Zimbabwe Border 11 26 22 15 Selebi Phikwe-Sefophe 17 20 16 14 Selebi Phikwe Airport Road 22 5 3 5 Dikobeya-Sefophe Sefophe-Tsetsebejwe 35 82 70 45 Combined ERR for all roads 13 26 21 18 1/ A 6% discount rate was applied. X ;<\\yt/ - 20A- nnex 3A' Page I of 5 Fgurth Road PrgQ1ects Loalns 1U7A Ld S 119PIM I-NTRQIŽUgT IQN Para 1.01, Line 11: It is not clear which north-south road is being referred to. The distance from the South African border (Ramatlabama) to the Zimbabwean border (past Ramokgwebana) is 648 km. From Ramatlabama to the Zambian border at Kazungula is a distance of 1044 km. If the 780 km. north-south road referred to is the Gaborone - Nata road (referred to in para. 2.02), the distance is again incorrect and it is doubtful whether termination of the road inside the country lends itself to the term north-south. IFSOOMM i : Shashe Infrastructure Project. 2 PROJECT PREPARATIONJ APREAISAL AND APPERQVAL. Para. 2.08, Lines 11 - 15: It should be noted here that the feeder roads study, funded under the Fourth Road Project, did not lead to any of the project components included in a Fifth Road Project', although 3 roads were designed under the Fourth Project and are still awaiting funding for- construction Mochudi - Sikwane Maun - Shorobe; Pitsane GoodHope. the Tuli Block study referrred to in para. 2.10 was more consistent with Government's priorities and recommendations from the study formed the basis for components included in the Fifth Road Project, namely Selebi-Phikwe - Sefophe and spot improvements to the Sefophe - Tsetsebjwe and Sefophe -- Dikabeya roads. Heading for Para. 2.13 : Broad Project Objectives Tables on p.6 : Descriptions of the Third and Fourth Road Projects show the situation as per the loan agreements. The situation following the amendments to the loan agreements could be usefully shown here - i.e. after the addition of new project components. 7': PROJECT IMPLEMENTATIQN aND CQST. Para 3.11 Does "spot improvements" or "a higher level of betterment" adequately describe the works performed, which often ,amount to virtual full - scale construction ? A better technical description of the works is required. S z ECONOMIC RE-EVALUATION_ Para 5.01 : In most feasibility studies carried out in Botswana,economic costs of construction and maintenance are calculated from financial costs by allowing for taxes and duties and also by the shadow-pricing of unskilled labour, given the considerable underemployment of labour in the country. This does not appear to be Bank practice or is this an omission ? Also, no residual value is attributed to the road structure at the end of its design life, whilst this is again common practice in fiasibility studies in the country. Use of the Bank's methodology is therefore bound to give lower economic returns to projects - 21 - Annex 3A page 2 of 5 than those predicted by the ruling methodology for Botswana. Para. 5.02 : The traffic figures quoted do not conform with records contained in the Transport Statistics of the Ministry of Works and Communications or with special counts undertaken by the Roads Department. There appears to have been a serious mistake if the figur-es were provided by the Government side. Tr-affic at the permanent station S-14 (Molepolole East), indicative of traffic on the Gaborone --Molepolole road east of Molepolole, is 390 v.p.d. in 1984, whilst at the permanent station S-18 (Metsemotlhaba), representative of the Metsemotlhaba - Molepolole section west of Metsemotlhaba, it is 462 v.p.d. An average for the road tection might be estimated at 426 v.p.d. This is considerably higher than the 324 v.p.d. quoted for the 44 km. of road constituting the bulk of the Gaborone - Molepolole road in the PCR. The practice of comparing traffic on the 8 km. Gaborone - Metsemotlhaba section in the re-appraisal with that expected at appraisal for the 5 km. Gaborone - Mogoditshane road is a poor one. How was adjustment made for the different size of catchment area served ? Average daily traffic on the 5 km. section is much higher than on the 8 km. section as shown by the figures below, reproduced from the Roads Department's special traffic count in 1984, which shows traffic levels varying considerably over the road : Stsatieu tsetion Of Wtich ReprCsentative 1984 vRd Gaborone West Gaborone - Mogoditshane 2804 Mogoditshane Mogoditshane - Metsemotlhaba 1749 2km. west of BDF Camp Mogoditshane - Metsemotlhaba 1018 Metsemotlhaba ( S-18) Metsemotlhaba - Gaborone Fork 528 Metsemotlhaba Metsemotlhaba Molepolole Fork 462 Molepolole East (S-14) Molepolole - Gaborone Fork 390 For the entire length of the Gaborone - Metsemotlhaba section, the traffic appears to be considerably in excess of the 522 v.p.d. quoted in the PCR's economic re-evaluation. The traffic data used in the re-evaluation appear to have been derived from the permanent station S-18 Metsemotlhaba - Gaborone Fork, which grossly underestimates total traffic on this road. Figures contained in the M.W.C. Transport Statistics do not support the contention that traffic fell by 20 - 33% from 1977- 79. At the permanent traffic stations, growth between 1977 and 1979 was as follows : Molepolole East - Gaborone Fork (S 14) : 24.6% Metsemotlhaba - Gaborone Fork (S 18) : 16.4% It can only be assumed that like has not been compared with like - i.e. a special traffic count nearer to Gaborone taken in 1977 may have been compared to a standard count taken further away from Gaborone at Metsemotlhaba in 1979. The figure of 11% p.a. growth between 1979 and 1984 is correct if the permanent stations S-14 and S-18 were those used in the calculations. The figures for normal traffic growth of 10% up to 1990 and 7% thereafter quoted in the PCR are extremely optimistic. The National Transport Plan (Hoff and Overgaard 1983) estimated normal traffic growth for the country as a whole.at 10% p.a. for the period 1976 -1981, roughly equa'. to average G.N.P. growth over the period. There was no reasor. to expect this relationship to change over the near future. The Ministry of Finance and Develooment Plannina has estimated expected G.N.P. growth during - 22- - Annex 3A Page 3 of 5 N.D.P. VI (1985 - 1991) at 4.9% p.a. on average. Estimates of traffic growth beyond 1990 are pure speculation, but in the absence of any major new mining projects, an assumption can be made that G.N.P. and traffic growth will be roughly the same as during N.D.P. VI - i.e. roughly 5% p.a. An assumption of traffic growth of 5% p.a. for the period 1985 - 1999 therefore seems more realistic, assuming that the effects of upgrading the Gaborone - Molepolole road on traffic generation have worked themselves through. Para. 5.03 : The upgrading of the Gaborone - Molepolole road was not completed until June, 1979. It is assumed therefore that benefits from V.O.C. and Maintenance savings were calculated as beginning from this date and that correct adjustments were made ffor cost escalation. I assume that the statement that all costs and benefits are in 1976 prices is a mistake and that it should say 1979 prices as in para. 5.06. Given the observation of traffic growth between 1977 and 1979, the last sentence of this paragraph is unjustified. With a revaluation of V.O.C. savings a lower than expected I.R.R. would not be so much the result of poor project performance as of poor data sources at appraisal. Para 5.04 : The estimates presented are very dubious. Should not the social and wider economic benefits of this project be stressed more than vehicle operating cost savings ? There is no mention of employment creation and its impact on rural incomes for instance. Para. 5.05 s Again, projected growth rates to the project's end are very optimistic and should be revised. Table 5.1 s The gaps in the table relative to estimates for 1975 and 1976 need to be explained. How were savings for minibus and pick-up traffic calculat2d ? - use of the same savings as for passenger cars would induce a bias towards lower v.o.c. savings compared to previous appraisals but this wouid reflect the absence of important data rather than the use of the most up-to- date data. Is lack of data a factor in the lower V.O.C. savings used in the re-evaluation ? Table 5.2 : The table gives very little data and makes it difficult to check the results. A table with as much data as Table 10 in the Third Project Appraisal Report would allow a much easier comparison. -23 - - / Annex 3A s / Page 4 of 5 * v7 /~~~AAUO The data used to calculate expected rates of return for the Gaborone - TMolepolole road below differ from the data used by the World Dank in the Project Completion Report for the ThiId and Fourth Road ProJects. The main difference is in the traffic data used, where data taken from the Ministry of Works Transport Statistics and froam special traffic counts taken on various sections of the road by Roads Department in 1984 give a very different picture of the normal traffic flows on the road. In all cases, the 1984 figures are considerably higher than those used in the Bank's re-evaluation. The vehicle operating cost data are also different and are based on the latest available estimates made by the Transport Statistics Unit of the Ministry of Works and Ccmmunications, converted into 1979/80 prices using the G.D.P. Deflator. Data are given for more vehicle categories than in the Bank's re- evaluation and the data used relate more directly to the vehicle categories identified in the Roads Department's traffic counts. There is no distinction made betweer. passenger cars and L.D.V.'s in the traffic counts and so it was necesscry to create a composite car/L.D.V. category for which a vehicle operating cost figure was calculated using the proportions 50% L.D.V. and 50% car. These proportions are roughly the same as for the combined registered vehicle population of Molopolole and Gaborone. Maintenance cost savings are also calculated differently and follow the recommendations of a recent feasibility report (Viability Of Roads In The Tuli Block i ZMCK Consulting Engineers, June 1984 ) which utilised data from the Roads Department Maintenance Branch's cost accounting system to calculate maintenance cost savings. The maintenance scenario assumes heavy traffic on the road with a regravelling period of 3 years and a resealing period of 7 years. A fogspray is applied to the bitumen surface in years Z and It following construction. Maintenance _ost savings constitute a small proportion of the benefits derived from road upgrading. The estimate of construction costs used is the same as that adopted in the Project Completion Report. The same assumptions concerning the surface composition of the road before upgrading have also been used. No allowance has been made for a residual life of an upgraded road although this is common practice in Botswana. The full benefit of V.O.C. savings for generated traffic has been utilised in the calculations although economic theory more correctly prescribes incorporating half of the realised savings for generated traffic. Adherence to this rule would reduce the rates of return considerably, given the dominance of vehicle operating costs in the benefits total and the high levels of generated traffic. However, the aim is to produce an estimate rather than an exact figure. It has been necessary to make a number of large assumptions in order to arrive at the estimates shown in Tables 3 and 4. These were necessitated because of the lack of detailed traffic data for the year 1979. It was necessary to estimate past traffic levels for much of the Gaborone - Metsemotlhaba section given the massive under-estimation which would be derived from use of data for the Metsemotlhaba - Gaborone Fork section (permanent station S-18). Data from this ,tation appear to have formed the.basis for the Batik ' re-appraisal-. The assumptions made about traffic on the Metsemotlhaba - Molepolole section are less heroic and thia result can be presumed to be fdirly robust. For the shorter 6 km. section it seems li:ely that there will be a large measurement - 24 - Annex 3A Page 5 of 5 error. If traffic has been estimated too high for the immediate post-1979 period, a considerable upward bias will have been imparted to the results given the very high discount rate needed to give zero net present value. If vehicle operating cost savings for generated traffic were takeff as half of face value because of the consumer suirplus argument, there would be a further over- estimation because less of the traffic increase up to 1984 could be considered as generated traffic. However, given the correct high traffic flows on the firBt 8 km. of road in 1984 and 15 years of design life remaining ( although this could be shortened because of the very high growth in the number of equivalent standard axles crossing the road, despite the dominance of light traffic ), the rate of return for the first 8 km. will clearly be very high. The rates of return given in Tables 3 and 4 are much higher than those given in the P.r.R. despite the fact that much lower traffic growth rates are assumed for the period 1985 - 1999. - 25 - Annex 3B THIRD ROAD PROJECT LN. 1174-BT Revised Economic Reevaluation for the Gaborone-Molepolole Road 1. A new reevaluation was undertaken following the revised figures for the road sub-section Gaborone-Matsemotlhaba (8 km); instead of the traffic of 324 v.p.d. taken In the PCR of 1985, 426 v.p.d. was taken as per government comments (taken in the PCR of 1985) and for the Matsemotlhaba- Molepolole subsection (44 km) 2,177 v.p.d. was taken as per new counts weighted by the road sub-section. Government comments also suggest that the annual growth rate to be used is 10% from 1984 to 1985 and 5% thereafter. However, due to the higher traffic volume than anticipated, the present reevaluation took a somewhat higher growth rate i.e. 10% from 1984 and 1985 and 7% thereafter. 2. The economic reevaluation of the road construction, based on 1979/1980 prices, indicates the following rates of return. For the Gaborone-Metsemontlahaba sub-section an ERR of 42% while for the Metsemotlhaba-Molepolole sub-section the revised ERR is 20%. In the 1985 PCR the corresponding ERR's were 13 and 12% respectively. The combined ERR for the entire road is now about 24%. r- - f.... ......... ,l, 1 20° 0, > ~~A N G O L A $t R ANGOLA KAVNG Is.: -,* ! ._-9 I 1'-'sNAxx ! ~~-~ - 1j2 v | 00~K A VA N GOC \i- ! -- c, E R TA S ~~~~I K.-. / Z0 I ~~~~~~~~~~~~~~~~\ .'tS ! -->. / )Rs~~~~~~~~Kong -24° | HuAun>\2/ \ ~~~~Hukunts Rus Q,I A '\ X > AS. L, AHA Iy /, ' .X 2* . N IBRO 15120 I~ I/ y s | ~ * ' SEtTFMtteR 18 t !280 REPULBC OV BOTSWANA iXoziJtgs, Z A M B I A 1' FIFTH ROAD PROJECT TRANSORT NETWORK FiFt5 RAQA P1OJECT C r,6riiniotlio to bltwotmo pv.d shndord ,> gtPREvioU'ANK PROJECTS ni boStd tm4ftt oCtdt 6':S.§O \ \ I ~~~~~~~~~~~~~~~~~~~~~~Cnt axt n Ato AlI r o4ot o mI l stondat4 ZIMBABWE~ %i:tron 33- \\ El ~~~~~~~~ + ~~~~ * ~~~~~-- c6ftstuallo n to bttaint tsved stmondrd 3 1htt4 Ad ttoltct (laot. I 174-Ot) \ . ...... Xstno¢tte¢bht,lnou>w d stot.dord 4 F4WIk %a.* PaIqct (lao 1408k-T) \~~~~~~~~~~~~~~~~~~ - - Cootuflon ho hItumttioo pov$ oboodo 5oWs w 9 wc6t Project Moan 776 8T) \\kA \ 1~~~~~~~~~~ -C1Owo.slovt * lWbiqnototno* nee oi .- \1 EXiSTIN0 ROADS 'o to - B- thanmoii .wf.4oInrood .5 weto Bu1awqyo ~ ti2~IOS -hJpM t*r roO -~~~~~~Ttm ------ 01st WWI_. 5 *- A ,A t ,- okgwo ba n .. .. Ptstreo ANS ' (C ,oppe, r l ,, . T h e o } ~~~~~. (So4. d,ta _ _" (0iomaeds) P--.^' w 6 ^ o-P. m beloak Lretihakhane .rotv kri*s ' I ow i "Mmadnoare $ ) E ! {Seleb h c \ 22e- SeroY@X.a;, | ¢ T setset7t1~d Ee Serow * %setoe2Moeng. Shoshonqg \5. Ro dh Letlhakeng / PLokaIj!j Cd :NJ d \I/ g( ~~JKeN--G@ Molepolol*<_ M h i \I b G¢s°M1° Mle 2 kae4;- frt 8tavsf w 'TI5ht5 0 . r4d7o,r _ A ';2 3~~~~~~~~~~~~~~~~~~~~~~~~~~~~0 10 wrt01b, i*p boJmeJe !. r'^Y # # ,$ , . ~~~~~~~~~~~~~ANGOL.A Z AMtBIA - r. , .h .1MBABWE I. 0 -t OLA t M AOF -l bjbfekir4g. - ab g <, - Il^L,jl* # !gBOTSWANA/f SWAZILANDI AJIaN~ ~ REPUBLIC ICtiomtoro 0 40 80 J~2O l60 - 4~K Ws00 0 20 40 . 60( )' 1 rkt r-T 9 0 He e Q * ., J < v . -an . s,2p,. ___________.________________