Peru - Second Road Construction Project

CIRCULAIING CUP : TO BE RETURNED TO REPORTS DES Document of The World Bank FOR OFFICIAL USE ONLY Report No.1148 PROJECT PERFORMANCE AUDIT REPORT ON PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) April 21, 1976 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.  FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) TABLE OF CONTENTS Page No. PROJECT PERFORMANCE AUDIT BASIC DATA SHEET PROJECT PERFORMANCE AUDIT SUMMARY 1-4 ATTACHMENT: PROJECT PERFORMANCE AUDIT REPORT I. INTRODUCTION A.1 II. THE BANK AND THE LOAN A.1 III. PROJECT IMPLEMENTATION A.5 IV. PROJECT COSTS A.12 V. ECONOMIC JUSTIFICATION A.13 VI. THE ROLE OF THE BANK A.15 VII. CONCLUSIONS A.17 ANNEXES 1. Highway Expenditures and Road User Contributions, 1966-72 2. Forecast and Actual Project Costs 3. Sources of Construction Cost Overruns for the La Oroya-Aguaytia Road 4. Estimated Vehicle Operating Cost Savings 5. Forecast and Actual Traffic on the La Oroya-Aguaytia Road Map: Peru Second Highway Construction Project Exchange Rates: Peruvian Soles (S/.) 1966 .............................. US$1 = S/. 26.82 September 1967-September 1975 ..... US$1 = S/. 38.70 Since September 1975 ..............US$1 = S/. 45.00 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  PROJECT PERFORMANCE AUDIT BASIC DATA SHEET PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) Amounts (in US$ mln) As of Zik%I/ Original Disbursed Cancelled Repaid Outstanding Loan 425-PE 33.0 33.0 -5.5 27.5 Project Data Original Plan Revisions At-- a /a Conception in Bank-- 10/63 Board Approval 9/01/65 Loan/Credit Agreement - - 9/17/65 Effectiveness 9/30/65 11/30/65 11/30/65 Physical Completion Early/70 6/30/71 Early/73 7, of original project actually complete 38% - 717 Loan/Credit Closing-- 6/30/71 12/31/72 (12/31/73) 3/74 /c Total Costs (mIn) US$55.6 - US$59.8- Econ. Rate of Return 137 11% 5%-10% Mission Data Month, No. of No. of Date of Year Weeks Persons Manweeks d Report Appraisal 8/64 3 1 3 8/65 Technical Appraisal 2/65 1 1 1 8/65 St'-total 4 Supervision 1 2/66 2 1 2 2/24/66 Supervision I1 5-6/66 2 1 2 6/23/66 Supervision 111 1-2/67 2 1 2 2/23/67 Supervision IV 8-9/67 1 1 1 9/08/67 Supervision V 4-5/68 2 1 2 5/31/68 Supervision VI 4-5/69 2.5 1 2.5 5/16/69 Supervision VII 7/69 .5 1 .5 9/03/69 Supervision VIII 5/70 1 1 1 6/16/70 Supervision IX 8/70 .5 1 .5 9/16/70 Supervision X 2/71 1.5 2 3 3/03/71 Supervision IX 5/71 1.5 1 1.5 6/15/71 Supervision XII 6/71 .3 1 .3 6/15/71 Supervision XIII 2/72 .7 2 1.4 2/29/72 Supervision XIV 6/72 1.5 1 1.5 6/12/72 Supervison XV 2-3/73 2.7 2 5.4 4/23/73 Supervision XVI 5/73 .5 2 1 6/08/73 Supervision XVII 10/73 .5 1 .5 10/20/73 Supervision XVIII 5/74 _1__ 2 _21- 5/28/74 Subtotal 23.7 30.1 Follow-on Project Loan 706-PE of US$30 m1n, signed 9/14/70 for Fifth Road Project. /a Date of Conception in Bank is date Bank first recorded project was being considered for financing and began to follow up that decision in a serious continuous way (Project Negotiations or Country General Files). /b Actual Loan/Credit Closing Date is date of last disbursement out of the loan and/or credit, as given by Controller's Department data. /c Includes construction and supervision costs only. Final costs of other components not available. /d Based on seven-day field week.  PROJECT PERFORMANCE AUDIT SUMMARY PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) 1. This report presents an audit of performance under the Peru Second Road Construction Project, for which Loan 425-PE in the amount of US$33 million was fully disbursed in March 1974. It is based on correspondence and reports in Bank files, and on discussions with Bank staff who partici- pated in the appraisal and supervision of the project. In February 1975, a three-day visit was made to Peru and the valuable assistance of the Ministry of Transport and Communications and the consultants is gratefully acknowledged. A Project Completion Report, prepared by the Latin America and Caribbean Regional Office in October 1975, was also useful in the preparation of this report as were the comments provided by Bank staff and the Government. THE BANK AND THE LOAN 2. Loan 425-PE, the Bank's fourth investment in Peru's highways, and the second for road construction, evolved from a Government request in October 1963 for assistance in expanding the highway network. In response, the Bank agreed to finance construction of the La Oroya-Aguaytia section (486 km) of the Central Highway, which was and still is the only complete transversal road linking the coast with the eastern jungle region of Peru. At appraisal it was heavily travelled and it was the only road for which final engineering was almost completed. 3. A Bank appraisal mission visited Peru in August 1964 and the loan agreement for a US$33 million equivalent loan was signed in September 1965. The loan financed the foreign exchange cost and part of the local currency cost of a US$55.6 million project comprising: (a) construction/ reconstruction, paving and supervision of the 486 km between La Oroya and Aguaytia (93% of project costs); (b) final engineering studies of high priority roads (5% of project costs); and (c) a study of road user charges and continuation of maintenance advisory services provided for under a pre- vious loan (2% of project costs). 4. The main loan covenants concerned: (a) employment of consultants to: supervise construction, advise on maintenance operations, undertake final engineering studies, and study road user charges and adopt a timetable to implement recommendations; (b) division of the road into seven contract sections to be awarded according to Bank guidelines; (c) provision by the Government of resources for the project as needed; (d) enforcement of limits on vehicle dimensions and axle loads and addition of weighing stations; and (e) continuation of strengthening of maintenance. PROJECT IMPLEMENTATION 5. Project implementation was not smooth and numerous difficulties were encountered (Attachment A.5). Construction of the La Oroya-Aguaytia - 2 - Road was expected to begin in 1966 and be completed in 1970. In 1969, sub- stantial cost increases, prompted a reduction in the project's scope. One of seven road sections was deleted from the project and the remaining six sections were constructed to gravel standards rather than being paved as envisioned at appraisal. Construction of the revised project was completed in early 1973, three years behind the original schedule and a year and a half behind the revised schedule. 6. The main problem which delayed construction was the extraordinarily difficult topography of the region the road crossed, and inclement weather characterized by unusually heavy rainfall. Frequent and severe landslides slowed work and made additional drainage structures necessary as well as rede- sign of slopes. These conditions persisted throughout the construction period and when the gravel road was completed, it still required extensive mainte- nance to keep it open to traffic until it was paved under a subsequent Bank loan. A secondary source of delay was created by the Government's restricted financial situation which was aggravated by a 45% devaluation of the sol in 1967. These financial difficulties prevented the Government from paying the contractors and while the contractors at first continued to work, some of them eventually stopped. In early 1968, after repeated warnings, the Bank discontinued disbursements until all sums due to contractors had been paid. After arrears were met by the Government, work resumed but was reduced for the remainder of the year to alleviate the Government's financial commitments. 7. Performance with respect to the other components financed under Loan 425-PE was generally good although they too were affected by the financial constraints. The road user charge study was completed after minor difficulties, other studies financed by the loan were completed and the performance of the maintenance advisors was satisfactory. Adherence to the loan covenants was also generally satisfactory. The Government had under- taken and was successful in increasing road user charges and ensuring that vehicle axle load limits were enforced. Other undertakings however were restricted by the fiscal situation. For instance, while it had agreed to further strengthen traffic regulations by increasing the number of weighing stations, strengthen road maintenance and provide funds for the project as needed, these endeavors were limited. PROJECT COSTS 8. The actual construction cost of the road, S/. 2,080 million (US$53.8 million), represents an overrun of about 99% above the revised 1969 cost of S/. 1,047 million (US$27 million) (Attachment,A.12). Forty per- cent of this overrun is due to price increases, 30% to quantity increases, 15% to complementary works to prevent deterioration before paving, and 15% to the interaction of higher prices and larger quantities. The supervision cost of the project of S/. 236.2 million represents an overrun of 165% above the appraisal cost of S/. 89.0 million. (Supervision costs were not reesti- mated in 1969 when the project was revised) This overrun results from the - 3 - devaluation of the sol and the greater amount of supervision and additional engineering required. Finally, the cost of maintenance advisory services and final engineering studies of future roads were actually less than expected because their scope was reduced so that funds could be applied toward construction. ECONOMIC JUSTIFICATION 9. The expected overall rate of return at appraisal was 13% and that estimated in 1969 on the additional investment to complete the reduced project was 11%. Reassessment of the economic justification is difficult because information is lacking. Although the road was expected to have important developmental effects, these can be quantified only in terms of traffic and a complete assessment of the economic justification will only be possible several years from now. Based on available information, at audit the overall rate of return is conservatively estimated in the range of 5%-10, depending on the economic life of the road, while the audit rates of return on indi- vidual sections range from 18% to negative values. 10. The lower than expected overall rate of return results from cost overruns and low vehicle operating cost savings because of reduced standards. These negative factors more than offset the development of traffic which was generally higher than forecast except on the two sections farthest from the coastal region. On these two sections the rate of return is lower than that on sections close to the beginning of the road because of the inter- action of higher construction costs per km and lower traffic levels. Much higher traffic on sections farthest from the beginning of the road than that expected at appraisal would be required for the economic return of the project to improve significantly. CONCLUSIONS AND THE ROLE OF THE BANK 11. The primary objective of this project, reconstruction and paving of the existing road between La Oroya and Aguaytia was not achieved. In retrospect, this objective was unattainable within the construction schedule and costs formulated at appraisal. The extraordinarily difficult topography and climate of the region traversed by the road, which was not revealed in the final engineering and Peru's financial situation prevented this objec- tive from materializing. 12. The role of the Bank in selecting this project was satisfactory given the information available at appraisal. In its review of the final engineering, the Bank could not have foreseen the geological instability and climatic conditions overlooked by the consultants which made completion of a paved road impossible within thE scheduled construction period. Nor could the Bank have foreseen the extensive maintenance and emergency works needed to keep the road open. 13. Because the road crosses a comparatively new geological area, excavations or widening of the road usually produced landslides and time was required for the road to stabilize. These geological conditions raise the issue of the optimum design standard for the road. Bank staff have suggested that design standards in these conditions should be even more modest than those proposed at appraisal (6 m carriageway and 1.2 m shoulders). A maximum of 4-5 m road with some shoulders is suggested, as any additional width entailing excavation, triggers landslides in the unstable mountains. 14. The road user charge study was included in the project descrip- tion shortly before negotiations. While the Bank attached importance to increasing road user charges, it did not take this up with the Government, until immediately before negotiations. Consequently, the Government would not commit itself to increase the charges as the Bank wished but did agree to study the matter. While this did not adversely affect the outcome of the project and road user charges were increased anyway, the Bank should try to bring up important issues early on so that they might be fully discussed with the Borrower. 15. Disbursement procedures used by the Bank under the loan were flex- ible and adapted to Peru's financial constraints. Peruvian law prohibited direct payment to contractors, a form of disbursement normally used, so the Bank disbursed through a revolving fund. But the Government's financial difficulties impeded disbursement of its share of funds to the contractors some of whom consequently stopped working. After repeated warnings, the Bank suspended disbursements until arrears were met. The suspension was the only alternative available. But the Bank helped the Government out of these difficulties by agreeing that arrears to contractors could be partially covered by promissory notes by reducing work on the road for the remainder of the year to reduce the Government's financial burden and by postponing award of a contract for one section. These efforts on the Bank's part con- tributed to the successful outcome of the project. 16. Revision of the project in 1969 was a good decision and the Bank played an important role by rejecting the consultants'draft report on the revisions because it was poorly prepared, and by insisting that the document be redrafted, which it subsequently was. 17. A broader approach with regard to the project benefits, i.e., integrating the road into a regional development plan, could have made the project more successful. This is particularly true in the sections farther away from the coast, where even if the traffic projections at appraisal had materialized their economic return at audit would have increased only marginally. However, in retrospect, it is doubtful that such a program could have been implemented because of the financial difficulties of the Government. ATTACHMENT PROJECT PERFORMANCE AUDIT REPORT PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) I. INTRODUCEION 1.01 In October 1963, Peru's newly elected Government requested the Bank to take the lead in coordinating and financing a highway expansion program, to which it attached high priority. The Bank had already assisted Peru in highway development by making three loans totalling US$20.5 million. The first, Loan 127-PE for US$5 million made in 1955, supported the reorganization of the highway maintenance organization. The second, Loan 271-PE for US$5.5 million made in 1960, supported a joint venture with the Agency for Inter- national Development (AID) to reconstruct the Aguaytia-Pucallpa section (170 km) and carry out final engineering of the Tingo Maria-Aguaytia section (105 km) of the Central Highway (840 kin) (see map). The third, Loan 300-PE for US$10 million made in 1961, supported the continuation of the maintenance program, including the first inventory of the condition of some 15,000 km of roads, final engineering of an additional section of the Central Highway be- tween La Oroya and Tingo Maria (381 km), feasibility studies, and an initial study of the national road system by the consultants Coverdale and Colpitts (US). 1.02 In January 1964, the Bank met with AID and the Inter-American Development Bank (IDB), which the Government also had asked to participate in the highway program. This interagency meeting and subsequent meetings recommended implementation of the study prepared by Coverdale and Colpitts. But the study, which was of a general nature, had to be revised and expanded. In the interim, the agencies agreed to finance studies of 1,800 km of roads to provide the basis for the contemplated investments and construction of those roads for which final engineering already existed. II. THE BANK AND THE LOAN 2.01 In mid-1964, in compliance with the interagency decision to finance road construction, the Bank informed the Government that it would assist in financing the only road for which final engineering was well-advanced, the La Oroya-Tingo Maria-Aguaytia section (about 486 km) of the Central Highway. This road, the only complete transversal transport corridor in Peru and one of its most heavily travelled facilities was given high priority in the Coverdale and Colpitts study. 2.02 The Central Highway traverses 840 km between Lima and Callao on the Pacific Coast to Pucallpa on the jungle plains (see map). From Lima, the highway climbs the Western slopes of the Andes Mountains and reaches its highest point of 4,800 m (about 15,700 ft) above sea level shortly before La Oroya, an important mining center. Between La Oroya and Cerro de Pasco, the road follows the Altiplano at an altitude of 4,000 m (13,000 ft). From - A.2 - Cerro de Pasco, the road descends to luanuco at an elevation of 2,000 m (6,500 ft). After Huanuco, the road climbs again to 3,000 m (10,000 ft) and descends to the jungle plains about 50 km east of Aguaytia, from where it follows almost horizontal terrain to Pucallpa. 2.03 Final engineering of the Tingo Maria-Aguaytia section (105 km) along the existing alignment had been prepared by Brown and Root (US) under Loan 271-PE, whereas final engineering of the La Oroya-Tingo Maria section (379 km) was being prepared by Renardet-Sauti (France/Italy) under Loan 300-PE. The full report on the former and preliminary costs for the latter were received by the Bank in June 1964. After reviewing the data submitted for these two sections, the Bank sent an appraisal mission to Peru in August 1964. The mission reviewed the technical and economical feasibility of the project, collected additional data, and discussed with AID possible joint-financing of construction. 2.04 In January 1965, the Bank received the final engineering of the La Oroya-Tingo Maria section, excluding a section between Huanuco and Tingo Maria where two possible alignments were studied. One alignment followed the difficult existing route over the Carpish Pass with some relocations and the other followed the Huallaga River Valley, which had the benefit of giving better road characteristics. The Government selected the former alignment and the Bank agreed because the latter alignment entailed construction of more than 2,000 m of tunnels where perpendicular rock walls prevented other solu- tions and cost US$5 million more. 2.05 During the following month, the Bank sent a mission to Peru. The mission concluded the technical appraisal although final costs still were not available. The mission also discussed measures to enforce traffic regulations, particularly maximum vehicle axle loads. Some progress had been made under Loan 300-PE, including creation of a transport commission to regulate trans- port, establishment of five permanent weighing stations, and purchase cf port- able scales to be operated by the Highway Patrol. But these efforts were inadequate. 2.06 In March 1965, the Bank offered to finance 60% of the total project cost, including the estimated foreign exchange cost of 45%-50%, and some local currency costs. The Government delayed in accepting this offer because it was awaiting AID's participation, which would have reduced its own contribu- tion. This participation never materialized, however, and the Government ultimately accepted the Bank's terms. 2.07 Negotiations were held in June 1965. At this time, the Bank tried to obtain assurances from the Government that it would enforce traffic regula- tions, contribute 40% of the project cost, and increase road user charges. The latter undertaking was particularly important because road user charges were inadequate, with annual revenue comprising less than half the costs of - A.3 - maintaining the highway system. Fuel prices for the consumer were the lowest in Latin America and tax on fuel comprised only about 20%. In view of the large road investment program to be initiated and financed by the Bank, IDB and AID, and of the shortage of funds in the general budget, an increase in road user fiscal contributions was important. The Government agreed, but was not prepared to undertake such measures immediately. Gasoline taxation was a sen- sitive political issue because it was associated, unrealistically, with a dispute with the International Petroleum Corporation and the Government felt it was impos- sible to immediately introduce further gasoline taxes. Other road user taxes, such as a 40% tariff increase in vehicle license plates and driver permits in proportion to vehicle weight, had recently been levied so these charges could not be increased either. The Bank considered tying of loan disbursements to the Government's performance in implementing road user charges to be impractical because contractors would not work under the constant threat of interruption for reasons beyond their control. Moreover such tying would be unfair because the Bank had first mentioned increasing road user charges only in May 1965. The Bank settled for a diluted version of the commitment it would have liked to secure: the Government would first undertake a study and subsequently propose effective measures. 2.08 Negotiations were substantially completed in June 1965. However, no agreement on project costs was reached because the Government's team presented estimates which were about 20% over the appraisal figure. The reason for this escalation, as explained subsequently by Renardet-Sauti, was the resubmission by Brown and Root of costs for the Tingo Maria-Aguaytia section, (engineered almost two years earlier), because salaries and prices of materials had increased, as well as the submission of final costs for the La Oroya-Tingo Maria section reflecting a very minor increase. As salaries were expected to increase further and foreign contractors were likely to bid on the road due to the shortage of qualified local contractors, the Ministry of Transport and Communications (MTC) recommended increasing construction costs by 10% and adjusting the 15% contingency accordingly. The Government asked the Bank to increase its contribution from US$29 million to US$33 million, which was still 60% of total project costs. The Bank agreed, and negotiations were concluded. 2.09 The loan agreement for US$33 million was signed in September 1965. The total cost of the project was US$55.6 million. The Bank's contribution of 60% of the total cost covered the whole foreign exchange cost and some of the local currency cost. The main components of the project were: (a) construction, reconstruction and paving of about 486 km of highway between La Oroya and Aguaytia and consulting services for supervision of the work (93% of expected project cost); (b) final engineering studies primarily of roads for which feasibility studies and preliminary engineering were being undertaken with funds from Loan 300-PE and the Bank's tech- nical assistance grant, and of other high priority roads (5% of the expected project cost); - A.4 - (c) a study of road user taxes; and (d) a continuation of maintenance advisory services by consultants pro- vided for in Loan 300-PE (2% of the expected project cost). 2.10 The Government undertook loan commitments to: (a) employ Renardet-Sauti as consultants for construction supervision, extend for three more years their contract to provide advice on maintenance operations (Loan 300-PE), and broaden their contract to include advice on programr.ing, planning, and road design; (b) engage consultants to complete final engineering on the roads included in the project for which feasibility studies and preliminary engineering were already completed; (c) retain consultants to study road user charges and adopt a timetable to implement the reconmendations of the study; (d) divide the length of the road into seven sections with a separate contract awarded for each section after interna- tional competitive bidding in accordance with the "Guide- lines Relating to Procurement Under World Bank Loans and IDA Credits" stipulating that contractors would not bear foreign exchange risk; (e) make available promptly, as needed, the resources required to carry out the project; find additional revenue from internal sources to help finance construction and mainte- nance programs; increase road user charges to bring them in line with costs; and assess estate taxes on property owners adjacent to the road. The Highway Patrol, which enforced traffic regulations, was to receive the proceeds of a progressive license tax instituted in April 1965; (f) ensure that vehicle dimensions and axle loads were kept within the limits agreed with the Bank by organizing police patrols to enforce regulations. A study to increase the number of weighing stations was to be submitted by mid-March 1966; and (g) continue to strengthen the Maintenance Division of the Road Department and to see that roads were adequately maintained. Future budget allocations for maintenance were to be calculated on the same basis as 1965. A portion of the yearly allocation was to be earmarked for depreciation of existing equipment and for additional equipment needed to maintain new roads. These funds could be accumulated from year to year if the total amount was not spent within the fiscal year. - A.5 - III. PROJECT IMPLEMENTATION A. Construction 3.01 At appraisal, construction was expected to start at the beginning of 1966 and be completed in 1970. In 1969, because of substantial cost increases, the project's scope was reduced: one of the seven sections (about 10% of the expected project cost) was omitted and the remaining six sections were constructed to gravel standards rather than being paved as envisioned at appraisal. Construction of the revised project was completed in early 1973. Four of the six sections were completed two years behind the original appraisal schedule and eight months behind the revised schedule, while the remaining two sections were completed three years behind the original schedule and a year and a half behind the revised schedule. 3.02 The project started slowly, and construction did not oegin until the end of 1966, about seven months later than expected. The reason for the delay was the Government's slowness in awarding and signing the contracts for supervision and construction. Signature of the supervision contract was delayed because the Ministry of Finance objected to expatriate salaries being stipulated as tax-free. As Bank staff have pointed out, the contract provided for reimbursement of any taxes by the Government so that if the tax element were added to salaries it would only complicate the issue. After some discussions, the Government waived this condition and concluded a contract with Sauti Overseas Anstalt (Sauti), previously a partner of Renardet-Sauti, in April 1966. Award of the construction contracts was delayed because the Government imposed several contract conditions, including scheduling work so that the contractor was bound to complete only 25% of the total in any single year and asserting its rignt to reject bids 10% above its own cost estimate or require increased performance guarantees for bids 10% below its estimate. But the most serious condition was made in a corrigendum to bidders specifying that exchange into a foreign currency would not bear a fixed exchange rate. This contradicted one of the supplemental letters of September 17, 1965 and discouraged at least one contractor from bidding. Because bids for four sections (1, 3, 6 and 7) had been opened and contained adjustments for exchange risk, the Bank accepted these but strongly urged the exchange rate be fixed for the remaining three sections (2, 4 and 5). By June 1966, all contracts were approved, but a further delay of several months occurred when the Contraloria General del Peru refused to approve signing because of some minor and apparently insignif- icant administrative and legal issues. By the end of 1966, almost a year after contractors had prequalified, contracts were signed as outlined below and construction was started. - A.6 Actual Contract No. Section Length Firm Nationality Amount (km) (S/.) 1. La Oroya-Cerro de Pasco 122.0 Constructores Unidos Peruvian 129,160,372.70 2. Cerro de Pasco-Salcachupan 62.2 Impresit y Bertolero Italian/ 197,614,075.14 Peruvian 3. Salcachupan-Puente Rancho 78.0 Rodisa Peruvian 154,039,904.90 4. Puente Rancho-Caracol 43.7 Impresit y Bertolero Italian/ 192,942,158.38 (Carpish) Peruvian 5. Caracol (Carpish)-Tingo 54.2 Impresit y Bertolero Italian/ 184,817,476.75 Maria Peruvian 6. Tingo Maria-Divisoria 43.6 Laos y Bolzmann/ Peruvian/ 127,079,746.80 Water Resources Dev. Israeli 7. Divisoria-Aguaytia 62.4 Laos y Bolzmann/ Peruvian/ 140,898,470.60 Water Resources Dev. Israeli Total 466.1 1,126,552,205.27 3.03 Shortly after the start of construction, another delay occurred because a heavy and prolonged rainy season during the first half of 1967 at times brought works to a standstill and slowed progress. The rains revealed poorer soil conditions and more severe drainage problems than anticipated in the final engineering. Landslides developed, and removal of the rubble was difficult and time consuming, costing the life of a Sauti employee. Extensive additional earthwork was necessary although the consultants relocated portions of the road to safeguard future stability. 3.04 In addition, the Government was slow to meet its share of payments to the contractors because of the precarious financial situation of the country. The Bank's share, on the other hand, was paid promptly from a revolving fund the Bank was instrumental in establishing. This situation worsened when the sol was devalued 45% in September 1967 and domestic prices subsequently increased. By the end of the year, the situation was such that contractors who had continued work without pay began to stop work. In early 1968, after repeated warnings, the BanK discontinued disbursements until all sums due to the contractors for work performed had been paid and new disburse- ment procedures agreed upon. Also in early 1968, the contract for Section 1 (about 10% of the expected construction cost) was cancelled because the contractor had not begun to work satisfactorily and had failed to mobilize sufficient equipment. - A.7 - 3.05 To meet its debt to the contractors, the Government in May 1968 issued promissory notes repayable over five-years with 18 months of grace. The Bank agreed to the issue of these notes as the only practical solution to clear arrears. Furthermore, because the Government's budget was insuffi- cient to cover the planned work for the year, the Bank agreed to a 20%-30% reduction in works for the rest of the year with a view that, when budgetary allocations were increased, the pace of construction would also be accelerated. In addition, it was agreed that award of a new contract for Section 1 would be postponed indefinitely and work on Section 2 would be temporarily discon- tinued. 3.06 By August 1968, the Government settled outstanding debts on certifi- cates issued to March/April 1968 and the Bank resumed disbursements. The new disbursement procedure required the Government to supply evidence that con- tractors had been paid in full, i.e., both the Government's 40% and the Bank's 60% share. However, shortly thereafter, construction again slowed because of insufficient budget funds and arrears amounted to nearly US$2 million equiva- lent. Payments in dollars lagged even more than those in soles because of the Government's difficulty in prepaying contractors in foreign exchange as re- quired by the new disbursement procedures. By the end of 1968, only 32% of the loan had been disbursed as against 53% expected at appraisal. 3.07 In early 1969, it was clear that the project could not be completed as originally planned because of the delays and substantial cost increase since appraisal. Therefore, in May 1969, Sauti submitted new estimates of project costs and recommendations for a revision of the project. The Bank found these estimates unacceptable because the quantities had been derived theoretically and the reductions proposed did not take into account technical and economic implications. At the Bank's insistence, the report was revised. The cost data of the revised report showed that construction (excluding supervision and contingencies) had increased from the contract value of S/. 1,126.4 million to S/. 2,280.0 million, or 102%. About 56% of the total increase was a consequence of price increases stimulated by the effects of devaluation and about 44% was the result of increased quantities, including additional drainage structures and easier slopes. 3.08 The substantial cost increase led the Government to request that the project be divided into two phases. The first phase, which would involve construction of the road to gravel standards and would match available funds, would be financed under Loan 425-PE, while the second phase, which would include paving of the road, as envisaged at appraisal in 1964, would be financed under a future loan. The necessary engineering and revision of documents led to another delay in the project. In November 1969, the Bank approved Phase I, com- prising cohstruction to gravel standards of Sections 2-7, Cerro de Pasco-Aguaytia. 1/ Traffic levels on these sections remained low so that a good gravel road would provide an adequate driving surface for several years. Moveover the road width 1/ Section 1, La Oroya-Cerro de Pasco (112 km), where no contractor was working was postponed since the road was in all-weather condition. - A.8 - would be reduced in difficult mountain sections to avoid unnecessary work and reduce costs,and single-lane bridges, (as opposed to dual-lane ones),would be used wherever possible. The reduced project was expected to be completed within the original closing date of June 30, 1971. 3.09 Technical difficulties due to extremely difficult topography of the region through which the road passes, which is characterized by new geological formations and tropical climate with unusually heavy rainfall, continued throughout implementation of the revised project and slowed work. Landslides were frequent, 1/ additional drainage structures were required, and slopes had to be redesigned so they could be maintained. The final engineer- ing completed by Renardet-Sauti and by Brown and Root had overlooked these difficulties. 3.10 Problems with contractors also contributed to the slow work progress. The contract for Section 3 was cancelled because the contractors' financial capacity was so weak that the bank guarantee could not be renewed. The Govern- ment undertook the remaining work on this section by force account and rented machinery to supplement its own equipment. The contracts for Sections 6 and 7 also were cancelled because of the collapse of the consortium formed by Laos y Bolzmann and Water Resources Development (WRD). WRD withdrew from the partnership, claiming losses of US$4 million and maintaining that completion of the remaining work would entail further losses. Failure of the consortium was a consequence of underestimation of soil and climatic conditions, and of earthmoving requirements, poor management, changes in construction plans, and uneasy cooperation between partners. The remaining work of about 44 km on Sections 6 and 7 was too small for international competitive bidding, which would entail a further delay. In November 1971, the works were let to four Peruvian firms by negotiated contracts on the basis of the original unit prices amended by an escalation formula. The contracts incorporated a new alignment (the "Tea Garden Alignment"), recommended by the consultants to reduce slides. 3.11 Work on Sections 2, 3, 4 and 5 was completed in January/February 1972, eight months behind the revised schedule and about two years behind the schedule envisioned at appraisal. The quality of work on Sections 2, 4 and 5 was good and that on Section 3, one of the most difficult sections, was satisfactory. Work on Sections 6 and 7 was not completed until January 1973, a year and a half behind the revised schedule and three years behind the schedule envisaged at appraisal. 3.12 Emergency works needed for normal operation of the road between the first and second phase included retaining walls, drainage facilities, and channelization works. The remaining loan funds were insufficient to cover these works, and the Government financed them itself. -Vera Gutierrez, a Peruvian firm, was hired with Bank approval to undertake this work, and Sauti's 1/ For example, in November 1972, a 500 m section of mountainside entailing 1.0-1.5 million m3, fell over the road in Section 7, blocking it. - A.9 - supervision contract was extended. In addition, intensive maintenance was required to preserve and keep open the new 344 km gravel road comprising Sections 2-7 1/ until paving under the second phase. At the Bank's urging, the Government undertook maintenance for all sections except 6 and 7, which Vera Gutierrez carried out. 3.13 Even with maintenance, the risk was great that erosion of material, especially due to rainfall run-off, would result in landslides and embankment failures forcing closure of the road for extended periods. These occurrences would mean that some sections could not remain in their improved condition for more than two or three years. Consequently, the Bank pushed ahead with the loan to finance the Phase II paving works, for which a study by Sauti was included under Loan 425-PE in March 1972. The study concentrated on optimizing solutions for the road. A subsequent loan, 1025-PE of July 1974, undertook further work on some sections of road and another loan, the Lima-Amazon Transport Corridor Project, which was recently approved, will include still more works on the road. Despite these efforts the entire road still may not be completely finished. 3.14 The performance of the contractors was not as good as hoped for. The contract for Section 1 was cancelled because the contractor failed to mobilize; the contract for Section 3 also was cancelled because the contractor could not renew his Bank guarantee; and the contracts for Sections 6 and 7 were terminated because the contractors could not complete the road due to mis- management and difficulties posed by the road itself. 3.15 Sauti's performance as supervisory consultants in the early years of construction was generally good. When soil difficulties were first experi- enced, it was able to change portions of the alignment thus reducing costs and promoting a better road. At the time of the revision of the project in 1969, however, Sauti's relations with the contractors were unsatisfactory and the Government was disenchanted with its work. Moreover, Sauti's report proposing reductions of the project was poor. Part of the problem resulted from the inexperience of Sauti's staff in dealing with the difficulties posed by unstable soil. The team leader was replaced and the field staff was streng- thened, irproving both situations somewhat. Sauti's work was better after this, but the Government still tended to blame the firm for the unusual problems encountered. This should be weighed against the unusually difficult conditions of the project (a total of four Sauti employees lost their lives during construction in addition to several Peruvians employed by the contractors) and the Government's financial difficulties which caused the consultants to work without pay for long time periods. 1/ The total length of the road (Sections 1-7) is 466 km which represents a reduction of 20 km over the appraisal approximation. - A.10 - B. Maintenance Advisory Services 3.16 The services of Renardet S.p.A., retained by the 11TC under a previous loan to provide advisory services on maintenance, equipment, planning, design and soil mechanics, were extended under Loan 425-PE. The consultant's assistance was satisfactory, and in May 1968 its contract was extended. Under the extension, Renardet focused on maintenance cost accounting, maintenance equipment, and staff training. C. Final Engineering, Feasibility Studies and Road User Charges Studies 3.17 The Bank had initially considered including eight final engineering studies under Loan 425-PE to provide a solid basis for future loans. There was considerable confusion, however, over financing of the studies, and the Government proceeded with two (Sullana-Paiti and Nazca-Cuzco) on its own without advising the Bank. It presumably did so because more favorable financing was found. The shortage of funds also curtailed the original study program, reducing the number to three. The studies finally undertaken were final engineering of the Via Central-Satipo Road (110 km) and the Pasamayo By-Pass (15-20 km) of the Pan American Highway and a review of engineering done by the Government of the Nazca-Cuzco Road (650 km). Final engineering of the Via Central-Satipo Road was undertaken by Tippetts-Abbett-McCarthy- Stratton (TAMS). Refinements and checking of the final engineering of the Nazca-Cuzco Road, completed by the MTC, was undertaken by Sauti as was final engineering of the Pasamayo By-Pass. None of these projects resulted in road construction projects financed by the Bank, and it is unclear if any were financed by other entities. 3.18 The road user charge study was contracted to Wilbur Smith and Company in July 1966. The study was to be completed within six months, but initiation of the study was delayed because the MTC did not provide the agreed office space and transportation facilities. Subsequent progress was good, and the report was only slightly delayed to February 1967. Both the Govern- ment and Bank found the document unsatisfactory because the terms of reference had not been followed in important respects and the quality of analysis was inadequate. After considerable discussions between the consultants, the Bank, and the Government, the report was recast to focus on the economic aspects of road user charges rather than on a system of charges which would finance future highway expenditures, as had been done. The revised report, completed in May 1967, failed to cover some points adequately but after further addi- tional work, was accepted in November 1967. D. Loan Covenants 3.19 In carrying out its commitment regarding road user charges, the Government was able to increase them adequately (Annex 1). At appraisal the Bank recommended that gasoline taxes should be increased and between 1965 and 1974, the Government did so. Moreover, the Government introduced an element of progressive taxation into the gasoline tax structure whereby the region with the - A.11 - highest traffic volume (the Costa) is the most heavily taxed. The Government has not taxed diesel oil and lubricants, however, and truckers, who are the major contributors to the rapid deterioration of the road network, are not paying their share of road use. This situation could be improved by the Government increasing diesel taxes, a point made in the appraisal of the Sixth Road Project, and raising charges in the Lima-Callao area. An additional road user charges study is being financed by the Bank in the Sixth Road Project. 3.20 With regard to the commitment to make available the necessary funds for the project, the Government did less well. In the early years this was primarily a consequence of its difficult financial position. The Government showed a serious commitment to the project by undertaking the complementary works on Sections 6 and 7 and also by providing intensive maintenance for the entire road. It was slow, however, in settling outstanding payments to consultants and contractors and in Se,tember 1975 some payments outstanding since 1968 were still due, including a S/. 100,000 payment to Renardet for work executed in 1969. 3.21 The MTC's commitment under the loan to enforce traffic regulations appears to have been met. Vehicle.axle load limits were established and appear to be enforced. Much remains to be done in this area, however, as fines remain too low to be threatening and trucks continue to be overloaded, wearing down road surfaces. Weighing stations were not increased under the loan and the activities of the Highway Patrol established under an earlier loan (Loan 300-PE) were limited, apparently as a consequence of the shortage of funds. 3.22 The MTC was not successful in carrying out its commitment to improve road maintenance between 1966 and 1969, when road maintenance worsened and the progress made under the previous loan deteriorated. This deterioration reflected Peru's difficult economic situation when funds were insufficient to meet needs. The commitments to provide adequate budget funds for mainte- nance and to set aside an amount which would be applied toward maintenance equipment depreciation were also not achieved. 3.23 After 1970, maintenance improved somewhat because budget alloca- tions were increased (Annex 1). The major drawback however still continues to - A.12 - be inadequate maintenance allocations. The Bank's continuing interest in main- tenance has resulted in maintenance components in two subsequent projects (Loans 70-PE and 1025-PE). Moreover the Lima-Amazon Transport Corridor Project approved in December 1975 requires the Government to formulate a Five-Year Maintenance Program and submit it for Bank review. IV. PROJECT COSTS 4.01 The actual construction cost of the project is S/. 2,080 million (US$53.8 million), representing an overrun of about 99% above the revised 1969 cost of S/. 1,047 million (US$27 million) (Annex 2). Forty percent of the overrun is due to increases in prices, 30% to increases in quantities, 15% to complementary works to prevent deterioration of the road until paving under Phase II, and 15% to price escalation payments on emergency works (Annex 3). 4.02 The cost overrun varies for each section (Annex 2). The largest overruns, 169% and 118%, occurred in Sections 7 and 5, respectively. They are mainly explained by the cost of landslide removals (Annex 3). About 51% of the total cost of removals is concentrated in these two sections which account for only 25% of the total road length. The western portion of Section 5, possibly the most difficult road section, required more extensive maintenance by contractors to prevent landslides tnan anticipated in 1969. Section 7, also a particularly difficult section, was blocked by unusually large landslides which required extensive realignment near Divisoria (Tea Garden Variant) as well as complementary works, including drainage and retaining walls, to main- tain the road until it was paved in Phase II. The third largest cost overrun, about 112%, occurred in Section 6 and is also explained by complementary works required to keep the road open. The cost overrun for each of the remaining three sections (2, 3 and 4) is much lower, about 68%, and is mainly explained by price increases. 4.03 Supervision costs, accounting for 6% of the estimated project cost at appraisal, increased by about 165%, from about S/. 89.0 million to S/. 236.2 million. This overrun results from the devaluation of the sol and the greater amount of supervision and additional engineering required as a consequence of the difficulties encountered. The work over and above that originally estimated included relocation of a substantial portion of Sections 6 and 7, smaller relocations on other sections, and a review of project alter- natives in 1969. Moreover, in 1972, the additional engineering for Phase II works was included in the loan and undertaken by the consultants. 4.04 Finally, maintenance advisory services and final engineering studies of future roads, accounting for 7% of the expected project costs, actually cost less than expected because of reductions in the scope of the works so that funds could be saved and applied toward construction. - A.13 - V. ECONOMIC JUSTIFICATION 5.01 The rate of return on the project expected at appraisal was 13% and that estimated in 1969 on the additional investments to complete Phase I was 11%. 1/ Reassessment of the economic justification is difficult because information is lacking. At appraisal, the road was expected to have an important development effect, but at audit this effect can be quantified only in terms of traffic. Even using traffic, counts are available for only two years after project completion on five sections and for only one year on the other sections. Because of these factors, a clear assessment of the rate of return will only be possible several years from now and the outcome will depend heavily on future complementary investments, the extent of regional development and the growth of traffic. 5.02 The rate of return at audit has to be interpreted as a conservative estimate of the actual rate of return on the investment. 2/ The actual rate of return probably will not be lower; most likely it will be higher. Based on the limited information available, the rate of return at audit for Sections 2-7 combined ranges between 5% and 10% depending on the assumed economic life of the road. The 10% rate of return corresponds to the 20-year useful life assumed at appraisal, while the 5% rate of return corresponds to a 12-year useful life. The useful life varies on the road. Some sections could not remain open to traffic after 2-3 years unless extensive maintenance and additional investment (which are not included in the calculation of the audit rate of return) are made, while other sections could have a useful life of 20 years as anticipated at appraisal. 5.03 The lower than expected overall rate of return is explained by the substantial cost overruns, delays, and low vehicle operating cost savings because of reduced standards. These negative factors offset the favorable development of traffic which in general was higher than forecast, except on Sections 6 and 7 (Annex 5). Traffic on these two sections may be lower than expected because the land traversed has limited potential due to poor soil and 1/ The rate of return calculation undertaken when the project was revised assumed that 44% of the cost increase was due to quantities and the remainder to higher construction costs resulting from inflation. The latter increase was expected to be offset by higher benefits of the same magnitude. Past expenditures which accounted for the road being one-third completed were considered as sunk costs. 2/ The audit rate of return is based on actual construction costs, new estimates of vehicle operating costs developed by Sauti (Annex 4) and used in the appraisal of the Seventh Highway Project, and actual traffic data from 1972-73. Construction costs and vehicle operating costs were adjusted to 1972 prices to eliminate the effect of inflation. In addition, time savings have been included;however.it was not possible to include other probable benefits such as, rising vehicle operating costs in the "without situation because information is lacking. - A.14 - climate. However, even if the traffic on these two sections projected at appraisal had materialized, it seems that the economic justification of the road at audit would have improved only marginally. 5.04 The audit rate of return on sections far from the beginning of the road near the coast is lower than that on sections close to the begin- nirg of the road as seen below: Percentage of the Actual Total Rate of Rfpturn No. Section Investment Appraisal Audit- Audit - 2-7 Cerro de Pasco-Aguaytia 100 13 5 10 2 Cerro de Pasco-Salcachupan 15 n.a. 12 16 /c 3 Salcachupan-Puente Rancho 10 n.a. 15 to 16L 17 to 18- /c /c 4 Puente Rancho-Caracol (Carpish) 22 n.a. - 1 to 1-- 5 to 7- 5 Caracol (Carpish)-Tingo Maria 18 n.a. 8 12 6 Tingo Maria-Divisoria 14 n.a. negative 1 7 Divisoria-Aguaytia 21 n.a. negative 1 /a Assumes a 12-year useful life. /b Assumes a 20-year useful life. /c There is a range in the rates of return for Sections 3 and 4 because traffic information from differtnrt sources was used. This is a consequence of the interaction of higher construction costs per kilometer and lower traffic volume. The construction costs per kilometer of Sections 5-7 were almost double while traffic on them was much lower than that of Sections 2-3. Much higher craffic toward the Aguaytia end of the road than that expected at appraisal would be required for the economic return of the project to improve significantly. It is possible that traffic might materialize in the future; however, this would depend on development of the region, which in turn depends heavily on much higher complementary investments. 5.05 Although no specific agreement was reached with regard to comple- mentary investments in the region, the Government indicated at appraisal that it planned to develop the area. The Central Highway according to the appraisal report was expected to provide a vital link with the "new" lands being opened in the Selva and Ceja de Montana and " ... generate further growth of the development of the area." Investments were being made in health services and schools. Technical assistance was being provided to agricul- ture and experimental farms were operating in the Tingo Maria-Pucallpa area. Farm-to-market roads and investments in other areas were to follow. But the Government did not develop a comprehensive program of such investment and as Bank staff have pointed out, in view of Peru's financial difficulties from 1967 on, it would not have been possible to support such a program anyway. - A.15 - VI. THE ROLE OF THE BANK 6.01 The Bank's role in selecting this project was satisfactory, given the information available. The project had been assigned high priority by the Government, Bank,and in the Coverdale Colpitts study. Furthermore it was the only project for which final engineering was well advanced. 6.02 In its review of the final engineering, the Bank could not have foreseen the geological instability of the areas through which the road passed and the severe rainy season which made completion of a fully paved road impossible in the four-year construction period estimated at. appraisal. Nor could the Bank have foreseen that the road supported under this loan would need time to stabilize as well as extensive maintenance and emergency works to keep it open. The preparation of the final engineering was rushed and the zonsultants overlooked particular local conditions (soil and rains), which later created formidable technical problems. Because the region the road crosses is comparatively new geologically, any excavation or widening of the road produces landslides and time is needed for the road to stabilize. It has been pointed out by Bank staff that the consultants could not have determined the extensiveness of the instability without constructing segments of the road to see if it would stabilize. Therefore, it is difficult to think that the road could have been constructed differently from the trial and error approach used. 6.03 The Bank should have been cautious with respect to project costs, insisting that the costs and economic assessment of the project be finalized before negotiations to ensure that the project would have an acceptable rate of return. When final costs which proved higher than expected were received during negotiations, the Bank agreed to increase its contribution from US$29 million to US$33 million. However, in retrospect, it seems that the Bank also could have considered increasing the 15% contingency allowance. In subsequent loans to Peru, the Bank has wisely allowed a separate 15% contingency for quantity increases and has calculated price increases for a 36-month period using annual escalation rates for foreign costs ranging between 12% and 7.5%, depending on the item procured, and a constant rate for local currency costs of 10%. 6.04 For the most part, delays in bidding procedures were unavoidable because of the conditions the Government placed on the contractors and the Contraloria's subsequent delays in signing construction contracts after they were awarded. In regard to the issue of exchange risk, the Bank showed flexibility. The Government had released a supplementary instruction to bidders which contradicted one of the loan covenants requiring that contractors not bear the exchange risk. Because bids for four sections had already been received, the Bank agreed to let the Government accept these bids rather than require new bidding on the basis that contractors would have made an allowance for exchange risks in their prices. But the Bank insisted that a fixed ex- change rate be used for the remaining three sections for which bids had still not been received. - A. 16 - 6.05. Disbursement procedures used by the Bank under the loan were flexible and adapted to the economic difficulties of the times. An effort was made in the first disbursement method (via a revolving fund) to adapt to Peruvian law, which evidently prohibited direct payment to contractors, a form of disbursement normally used. The Government's subsequent failure to disburse its share of funds to the contractors was a consequence of Peru's difficult financial situation. The Bank's suspension of disbursements for the construction component of the loin after some contractors stopped working, was perhaps the only alternative. Bilit the Bank helped the Government out of this situation, by agreeing to a 20%-30% reduction in the works for the year, to a postponement of the contract for Section 1 and to payments of the contractors by promissory notes. The new disbursement procedures adopted when disbursements resumed, whereby the Government would pay contractors in full and then apply for reimbursemenc, also was effective and showed flexi- bility on the Bank's part. Under recent projects, Loans 706-PE and 1025-PE, the Bank disburses according to its traditional procedure because the situa- tion has changed and alternative methods are no longer necessary. 6.06 Revision of the project in 1)0. because of coSL L1,crases was a good decision. The Bank played an important role in this revision by rejecting Sauti's draft report, which was poorly prepared, and by insisting that this document be redrafted, which it subsequently was. 6.07 The Bank's oversight in not discussing road user charges until immediately before negotiations did not have a negative impact on the out- come of this issue. However,important issues such as this one should be raised well before negotiations to allow for thorough discussion and exchange of view. The Bank was correct in deciding not to force an immediate increase in road user charges by making disbursements contingent on them but rather to suggest a road user charge study. 6.08 The audit and Bank staff believe that a broader approach with regard to the project benefits, i.e., integrating developments into a regional development plan, could have made the project more successful. The develop- ment assumptions made at appraisal do not appear to have been supported by a regional study. While some sections including the Tingo Maria-Campanilla area (Section 5 and part of Section 6) and the Neshuya-Pucallpa area have experienced growth during the last ten years through agricultural settlement schemes, other sections such as the Aguaytia area (which starts at the latter end of Section 6, includes all of Section 7 and extends beyond Aguaytia) have limited agricultural possibilities because of the heavy annual rainfall (5,000 mm) which encourages plant diseases and reduces yields. The Bank might consider ways to optimize road benefits by including other complementary investments. It could suggest regional studies be undertaken and a program of investments developed, and possibly assist in such efforts. This type of effort would be particularly important where several follow-on loans have been made, as is the case with this project. 6.09 The Bank's supervision of this project, with an average of two visits per year during the seven-year implementation period, seems adequate. - A.17 - In addition the Bank had a resident representative in Peru for part of the period covering project implementation. VII. CONCLUSIONS 7.01 The primary objective of Loan 425-PE, reconstruction (basically widening) and paving of the old road between La Oroya and Aguaytia on the Central Highway, was not achieved. In retrospect, the objective was un- attainable within the construction schedule and costs formulated at appraisal. The extraordinarily difficult topography and climate of the region traversed by the road, which was not revealed in the final engineering, and Peru's financial situation prevented this objective from materializing. In retro- spect the trial and error method by which the road was constructed is pos- sibly the only way it could have been undertaken. Even with two subsequent loans for this road, 1025-PE and the recently approved Lima-Amazon Transport Corridor Project, it is still doubtful if the road will be paved to the standards set forth in the appraisal of 425-PE. When the road is completely paved, its economic justification will depend largely on complementary in- vestments occurring in the region, the extent to which the region develops, and the growth of traffic. 7.02 The project provided valuable experience to the Bank, contractors, Government and consultants. Bank staff have suggested that although the design standards at appraisal (6 m carriageway and 1.2 m shoulders) seemed modest, in cases like this where roads cross new geological formations where soil is unstable and traffic remains comparatively low, the standards should be even more modest, for example 4 m carriage with some small shoulders. 7.03 The contractors who were either Peruvian or joint ventures with Peruvian partners received excellent in-the-job training which may be applied in other parts of Peru or neighboring Andean countries. Bank staff have also pointed out that Vera Gutierrez, the small firm brought in, to complete Sections 6 and 7, is now the largest road construction company in Peru, although it is not possible to determine what impact its works under the project con- tributed to its growth. 7.04 Performance with respect to the other components financed under Loan 425-PE was generally good considering the financial constraints. The road user charge study was completed after minor difficulties and some prog- ress was made in increasing the charges. The Bank is supporting further studies in the area of road user charges under Loan 1025-PE. The other studies financed by the loan were completed and the performance of the main- tenance advisors was satisfactory. 7.05 Improvement of maintenance, which was a covenant of the loan, was affected by the shortage of funds. Weighing stations were not established for this reason, also, although the MTC's other commitments to enforce traffic regulations appear to have been met. Vehicle axle loads were agreed upon - A.18 - early in implementation of the project and were enforced; and the autonomous traffic commission was more effective than expected, considering its low budget. Much still needs to be done, however, in the area of vehicle axle load limits, as fines are too low and consequently trucks continue to overload. 7.06 The development impact of this road would have been greater if complementary investments had accompanied the road building program, although given the Government's financial constraints it would not have been able to implement a program. Nonetheless the Bank might explore ways of broadening its approach to project benefits by integrating road investment with regional development plans. 7.07 The high cost overruns and lower than expected traffic on sections farther away from the coast have made the return on the investment much lower than expected. However, even if traffic on these sections were com- parable to that of the appraisal forecast, the rate of return would have improved only slightly. Omitting these sections from the project at appraisal or in 1969, moreover, would not have been feasible because they are vital in linking Aguaytia and Pucallpa with the coastal regions. In conclusion, reassessment of the overall development impact of the road will only be possible several years from now, and the outcome will depend on future complementary investments as well as the growth of the region and of traffic. ANNEX 1 PROJECT PERFORMANCE AUDIT REPORT PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) Highway Expenditures and Road User Contributions, 1966-72 1966 1967 1968 1969 1970 1971 1972 ---------------------S-----------------------/. million, 1973 /1 A. Highway Expenditures-- Administration 99.5 147.3 156.7 151.2 200.0 230.6 268.0 Studies 41.2 37.1 23.1 13.6 30.0 28.7 28.0 Construction 1,585.9 2,116.6 837.7 668.2 873.8 1,027.0 980.0 Maintenance 156.7 159.6 168.5 127.5 310.6 428.3 556.9 Equipment 466.8 370.8 6.3 - 74.8 - 20.0 Studies of urban trans- port and traffic 10.1 18.0 20.4 ORDEZA /2 40.5 100.0 Total 2,350.1 2,831.4 1,192.3 960.5 1,499.3 1,777.3 1,973.3 B. Road User Contributions Excise taxes - 19.0 37.0 42.4 67.2 57.7 55.0 Import duties - 940.3 466.3 672.9 368.9 702.2 561.1 Gasoline taxes - 447.2 1,122.5 1,152.9 1,168.4 2,293.6 2,370.8 Vehicle registration, - licensing, road use, etc. - 100.2 124.4 152.7 181.0 139.5 183.2 Tolls - 16.8 19.0 22.8 39.9 57.0 73.0 Total - 1,523.5 1,769.2 2,043.7 1,825.4 3,250.0 3,243.1 C. Road User Contributions as a Proportion of Highway Expenditures 54 148 212 122 183 164 /1 Includes the Ministry of Transport and Communications, Comite de Reconstruccion y Rehabilitacion de la Zona Afectada (ORDEZA), and the Maintenance Equipment Agency (SEMi). Excludes -eeder roads financed by the National Fund of Economic DEvelopment (FNDE) and SINAMOS, and expenditures of the road transport regulation agency (ORETT) /2 This agency invested mainly in feasibility studies and road construction, but the detailed break- down is unknown. Sources: Bank Economic Mission Report, Transport Sector Review (Draft); June 8, 1970; Volume II, Table 2.10. Bank Economic Report WH-206a, March 31, 1971; Table 5.13 (1970 figures). Ministries of Economy and Finance, and Transport and Communications, Peru. 1972 Transport Sector Mission (1972 figures).  ANNEX 2 PROJECT PERFORMANCE AUDIT REPORT PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) Forecast and Actual Project Costs Actual Costs as a Proportion Nov. 1969 Dec. 1974 of Reduced Original Actual 1966- (Reduced Project) Actual Project Costs Completion Date Completion Date USS s/. US s/. US$ S.. No. Section -- million -- -- million --- -- million --- 1. La Oroya-Cerro de Pasco 4.8 129.1 - - 0.3 10.8 - - - 2. Cerro de Pasco-Salcachupan 7.4 197.6 4.7 182.0 8.0 308.7 170 Early 1970 June 1972 3. Salcachupan-Puente Rancho 5.7 154.0 3.2 125.0 5.4 208.6 167 Early 1970 December 1972 4. Puente Rancho-Caracol (Carpish) 7.2 192.9 7.0 271.0 11.7 454.4 168 Early 1970 June 1971 5. Caracol (Carpish)-Tingo Maria 6.9 184.8 4.3 168.0 9.5 366.5 218 Early 1970 March 1972 6. Tingo Maria-Divisoria 4.7 127.1 3.6 138.0 7.6 293.0 212 Early 1970 January 1973 7. Divisoria-Aguaytia 5.2 140.9 4.2 163.0 11.3 438.0 269 Early 1970 January 1973 Subtotal 41.9 1,126.4 27.0 1,047.0 53.8 2,080.0 199 Supervision of construction 3.3 89.0 /2 /2 6.1 236.2 Advisory services to ministry including road user charges study 1.1 30.0 L2 /2 /3 L Feasibility/Engineering future Roads 3.0 81.0 /2 /2 L3 L Construction contingency of 15% for quantity and price increases 6.3L 166.61 - - - - Complementary works on 6 and 7 - - - - 3.9-5 151.1L5 Total 55.6 1.493.0 /2 /2 /3 L Exchange Rate US$1 - S/. 26.82 US$1 - S/. 38.70 US$1 S/. 38.70 /1 Appraisal Report estimated contract price to be S/. 1,125 million. The amount in which contracts were awarded is represented here: S/. 1,126.4 therefore contingency reduced from S/. 168 million to S/. 166.6 million. /2 The final total was known to be about US$55.6 million, but column does not total because costs of items other than construction were not re- calculated and because construction items are in terms of original (1966) bid unit prices. /3 Final estimates not available. /4 Carpish Pass Section (85 km) allows 20% contingencies as it is based only on preliminary engineering. /5 Included in Sections 6 and 7 above. Supervision of construction also includes supervision of complementary work in the amount of S/. 25.5 million. Sources: Appraisal Report, TO-481a (1966-figures). Ministry of Transport and Communications.  PROJECT PERFORMANCE AUDIT REPORT PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) Sources of Construction Cost verruns for the La Oroya-Aguaytia Road (1) (2) (3) (4) (5) (6) Price Escalation Slide Removal Price Escalation Comple- Revised 1969 p Payments on Con- and Emergency Payments on Slides mentary Total Payments No. Section Contract Amount- tract Work Items Work and Emergency Works Work (1)+(2)+(3)+(4)+(5) -------------- --------------- 5. milion - ------------------------------------------------ 1. La Oroya-Cerro de Pasco 9 2 - - - 11 2. Cerro de Pasco-Salcachupan 182 77 30 20 - 309 3. Salcachupan-Puente Rancho 125 39 30 15 - 209 4. Puente Rancho-Caracol (CarDish) 271 77 73 34 - 455 5. Caracol(Carpish)-Tingo Maria 168 85 75 38 - 366 6. Tingo Maria-Divisoria 138 59 16 8 72 293 7. Divisoria-Aguaytia 163 68 82 45 80 438 Total 1056 407 306 60 151 2080 /1 In terms of original (1966) bid unit prices. Source- Ministry of Transport a-d Co=mnications.  ANNEX 4 PROJECT PERFORMANCE AUDIT REPORT PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) Estimated Vehicle Operating Cost Savings (1972 sol/km) Unimproved Road Raised to Road Gravel Standards Savings Cars 2.75 2.21 .54 Trucks 15.38 9.43 5.95 Buses 19.27 11.28 7.99 Source: Sauti. Feasibility Study of the La Oroya-Aguaytia Highway. 1973. Vehicle operating costs are for a mountainous, winding road.  PROJECT PERFORMANCE AUDIT REPORT PERU SECOND ROAD CONSTRUCTION PROJECT (LOAN 425-PE) Forecast and Actual Traffic on the La Oroya-Aguaytia Road 1964 1970 1972 1973 No. Section Actual Forecast Forecast Actual Forecast Actual ------------------------ vehicles/day ---------------------------------- 2. Cerro de Pasco-Salcachupan 250 376 422 370 446 530 3. Salcachupan-Puente Rancho 200 310 348 389 (560) 367 500 4. Puente Rancho-Caracol (Carpish) 150 244 274 303 (382) 288 n.a. 5. Caracol (Carpish)-Tingo Maria 150 244 274 277 (331) 288 548 6. Tingo Maria-Divisoria 125 202 228 n.a. 240 171 7. Divisoria-Aguaytia 125 202 228 n.a. 240 171 Sources: Ministry of Transport and Communications and Appraisal of the Second Road Construction Project, Peru. Table 4. Appraisal Report, Seventh Highway Project, 1973 actual figures for Sections 2 and 3. Sauti. Feasibility Study of the La Oroya-Aguaytia Road, 1972, 1972 actual figures in parentheses v' for Sections 3-5.  믹