Peru - Road Reconstruction Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No.3080 PERU PROJECT PERFORMANCE AUDIT REPORT PERU ROAD RECONSTRUCTION PROJECT (LOAN 706-PE) July 25, 1980 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 ROAD RECONSTRUCTION PROJECT (LOAN 706-PE) Table of Contents Page _No. PREFACE 11 PROJECT PERFORMANCE AUDIT BASIC DATA SHEET iii-iv HIGHLIGHTS v PROJECT PERFORMANCE AUDIT MEMORANDUM I. Background 1 II. The Project 2 III. Project Implementation 3 IV. Points of Special Interest 4 V. Conclusions 9 ATTACHMENT: PROJECT COMPLETION REPORT I. Introduction 11 II. Project Preparation, Appraisal and Negotiations 12 III. Project Implementation 16 IV. Costs, Disbursements and Financing 28 V. Economic Reevaluation 32 VI. Institutional Performance of the Borrower 34 VII. The Role of the Bank 36 VIII. Conclusions 38 Tables 1. Project Data 40 2. Civil Works - Construction Costs 41 3. Design Characteristics of the Roads 42 4. Feasibility Study Estimates of Traffic Compared with Actual Counts for Opening Year 43 5. Economic Reevaluation Results of Road Construction 44 Annex 1. Design Changes Recommended for Catac-Pativilca Road 45 Charts 1. Organization of ORDEZA 47 2. Partial Organizational Chart - Highway Maintenance 48 Map 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.  - 11 - PROJECT PERFORMANCE AUDIT REPORI PERU ROAD RECONSTRUCTION PROJECT (LOAN 706-F., Preface This report presents a performance audit of the Road Reconstruc- tion Project in Peru for which Loan 706-PE for the amount of US$30 mil- lion was approved by the Board on September 8, 1970. The loan was closed fully disbursed on December 31, 1978. The report consists of a Project Performance Audit Memorandum prepared by the Operations Evaluation Depart- ment (OED) and a Project Completion Report (PCR) prepared by the Latin America and the Caribbean Region. The memorandum is based on the PCR and discussions with Bank staff; the transcripts of the Executive Directors' meetings of July 14 and September 8, 1970, which considered the project, have been read and project files and documents have been reviewed. OED staff visited Peru in August 1979 to review the project experience with the Government and others familiar with the project. Among the many comments made by Govern- ment officials, the more important ones concerned the fact that, while the project was very useful, it was seen more as a contribution to economic development rather than to emergency reconstruction. They felt that im- plementation had taken too long. These views have been fully taken into account in preparing the audit memorandum. The assistance to the mission by the Peruvian authorities is gratefully acknowledged. The draft audit report was sent to the Borrower for comments in the normal course; however, no further comments were received. In general, the audit memorandum agrees with the analytical com- ments and conclusions of the PCR. The memorandum, however, provides addi- tional comments on certain implementation problems such as difficulties with route selection and design, and the economic impact of the project. It focusses especially on the Bank's role in providing assistance under emergency conditions, including possible ways of making project identifi- cation and implementation more effective and faster.  - iii - PROJECT PERFORMANCE AUDIT BASIC DATA SHEET PERU ROAD RECONSTRUCTION PROJECT (LOAN 706-PE) KEY PROJECT DATA Actual or Item Original Plan Reestimate Total Project Cost (US$ Million) 41.7 47.6 Overrun (%) - 14 Loan Amount (US$ Million) 30.0 30.0- Disbursed ) 30.0 30.0 Cancelled As of 4/30/80 Repaid - Borrower's Obligation ) 33.78 1/ Date Physical Components Completed 9/75 9/77 Proportion Actually Completed by above Date (%) 50 100 Time Overrun (%) 65 Economic Rate of Return C%) 17.8 14.6 Institutional Performance - Disappointing Cumulative Estimated and Actual Disbursements (US$ Thousand) FY71 FY72 FY73 FY74 FY75 FY76 FY77 FY78 (i) Estimated 1,650 8,250 14,850 21,150 28,100 30,000 30,000 30,000 (ii) Actual - 620 2,100 6,900 15,500 23,310 29,800 30,000 % of (ii) to (i) - 7.5 14.1 32.6 55.1 77.7 99.3 100 OTHER PROJECT DATA Item Original Plan Revisions Actual First Mention in Files 6/70 Negotiations 8/--/70 Board Approval 9/08/70 Loan Agreement 9/14/70 Effectiveness 10/26/70 Closing Date 9/30/75 12/31/77 12/31/78 2/ Borrower Republic of Peru Ministry of Transport and Communications (MTC), Organismo Regional para el Desarrollo de la Zona Afectada (ORDEZA, renamed ORDENOR-CENTRO) Fiscal Year of Borrower January 1 - December 31 Follow-on Projects Project Name Sixth Road Project Lima-Amazon Transport Corridor Project Loan No. 1025-PE 1196-PE Loan Amount US$26 Million US$76.5 Million Loan Agreement Date July 1, 1974 May 27, 1976 1/ Includes US$3.78 million for exchange adjustment. 2/ The closing date was extended originally to December 1977 to permit completion of civil works and technical assistance. The further extension by one year was made to allow the remaining balance to be used for maintenance activities on the high- ways financed under this loan.  - iv - MISSION DATA Month/ No. of No. of Item Year Days Persons Man-Days Date of Report Appraisal 1/ 6/70 10 2 20 8/24/70 Subtotal 10 20 Supervision I 8/70 3 1 3 9/16/70 Supervision II 11/70 2 1 2 12/03/70 Supervision III 3/71 9 2 18 4/15/71 Supervision IV 5/71 9 1 9 6/10/71 Supervision V 10/71 1 1 1 12/06/71 Supervision VI 2/72 1 2 2 3/14/72 Supervision VII 6/72 5 1 5 7/12/72 Supervision VIII 10/72 6 1 6 11/08/72 Supervision IX 3/73 6 2 12 3/30/73 Supervision X 5/73 1 2 2 6/08/73 Supervision XI 10/73 1 1 1 11/20/73 Supervision XII 12/73 5/8 2 13 3/13/74 Supervision XIII 5/74 5/3 2 8 5/28/74 Supervision XIV 6/74 7 1 7 7/15/74 Supervision XV 9/74 2 1 2 10/11/74 Supervision XVI 7/75 4 1 4 10/23/75 Supervision XVII 10/75 6 1 6 1/29/76 Supervision XVIII 4/76 6 1 6 7/22/76 Supervision XIX 10/76 1 1 1 1/05/77 Supervision XX 3/77 3 1 3 5/16/77 Supervision XXI 6/77 3 2 6 7/13/77 Completion 6/77 - - - 3/29/79 Subtotal 97 117 TOTAL 107 137 COUNTRY EXCHANGE RATES Name of Currency (Abbreviation) Sol de Oro (S/.) Appraisal Year Average Exchange Rate: US$1 = S/. 38.7 Intervening Years Average 1971-1974 US$1 = S/. 38.7 1975 US$1 = S/. 40.8 1976 US$1 = S/. 57.43 1977 US$1 = S/. 83.81 1978 (Completion Year) US$1 = S/. 156.34 1/ This was a combined identification, preappraisal and appraisal mission because of the need for quick loan processing.  - v - PROJECT PERFORMANCE AUDIT REPORT PERU ROAD RECONSTRUCTION PROJECT (LOAN 706-PE) Highlights After a severe earthquake had devastated part of northwestern Peru on May 30, 1970, the country's Government immediately requested the Bank for assistance. In spite of the fact that economic and expropriation problems (para. 2) had prevented the Bank from lending to Peru for several years, a loan for US$30 million was approved about three months after the disaster. The purpose of the project supported by the loan was to improve access to and within the earthquake disaster area, the Callejon de Huaylas. The project covered preparation studies, road construction and technical assistance for highway maintenance. The project was implemented largely as envisaged but the orig- inal cost estimate was exceeded by 14%. In total, the project had a time overrun of 65%, mostly due to some technical difficulties and the longer than expected time required for decision-making. The economic return estimated at the time of audit was satisfactory at 15% but below the 18% calculated in the feasibility studies which was mostly caused by lower than forecast traffic (PCR, paras 5.03-5.06). Furthermore, it appears that even though the roads as constructed optimize the return to the national economy, a different road choice might have provided higher benefits to the poorest part of the population in the valley (paras 26-34, 40). Points of special interest are: - the Bank's approach to reconstruction was concen- trated rather narrowly on transport but, seen in retrospect, appropriate to the circumstances (paras 15-17); - because of the time needed for studies and engineer- ing, the project effectively served long-range devel- opment requirements but had little relation to earth- quake reconstruction needs; - if the Bank Group intends to contribute to recon- struction after emergencies, it should not only re- duce the time required for loan processing, but also accept greater risks by reducing planning studies to a bare minimum to ensure early imple- mentation (paras 18-21, 37-38); and - more attention should be given to releasing the most experienced and knowledgeable staff for proj- ect identification after emergencies (paras 15, 38).  PROJECT PERFORMANCE AUDIT MEMORANDUM PERU ROAD RECONSTRUCTION PROJECT (LOAN 706-PE) I. Background 1. Peru has a population of about 17 million in an area of 1.3 mil- lion sq km. The per capita income in 1978 was about US$850 equivalent. The Andes, which is one of the world's highest mountain ranges, divides the country into a narrow coastal region, the mountains and the eastern plains. Because of the extreme ruggedness of the Andes, the development and maintenance of a transport system are difficult and costly. Lima, the capital and the country's commercial center, is on the coast. 2. Between 1952 and 1967, the Bank made 23 loans to Peru. These included ten for projects in the transport sector (four port, two railway and four highway projects). Implementation of these transport projects sometimes met with difficulties. The two highway construction projects (1960 and 1965) suffered from delays and a shortage of local funds. The two road maintenance projects (1955 and 1961) were initially successful but subsequently maintenance deteriorated mainly due to budgetary short- ages. Following a loan granted in mid-1967, the Bank did not lend to Peru because, after the economic mission early in that year, the Bank became concerned over the rapidly deteriorating fiscal situation in the country and about a balance of payments crisis which ultimately led to the devalua- tion of the sol in September 1967. Another factor of importance was the expropriation of a foreign-owned petroleum company in October 1968 without prompt satisfactory compensation. Regular lending was resumed only in 1973. 3. On May 30, 1970 a severe earthquake struck Peru. It affected an area of 77,000 km2 extending 700 km along the coast from south of Lima to the port of Trujillo and well over 100 km inland. The area of the greatest damage was in the Department of Ancash between the towns of Pativilca and Chimbote on the coast and the Cordillera Blanca, a branch of the western Andes. The effects of the earthquake were threefold: (a) avalanches of ice, stone and mud which covered whole towns and villages and disrupted power supply and communications; (b) tremors which caused homes and build- ings to crumble and slides to block roads; and (c) compaction of subsoils causing ground water levels to rise and flood areas near the coast. An estimated 75,000 persons perished as a result of the earthquake. 4. Initial estimates indicated damage amounting to at least US$500 million, and the Government of Peru immediately approached the Bank for a loan of US$150 million to cover various aspects of the rehabilitation pro- gram. This request was supported by various member countries which urged the Bank to provide assistance to Peru. Two-and-a-half weeks after the earthquake, the Bank sent two specialists who were, within the framework of a larger United Nations mission, to identify a priority project which would contribute to restoring and improving living conditions in the area most affected by the earthquake. IDB was separately sending missions to - 2 - identify and prepare projects. Even though the Government request was for general assistance, the Bank's mission was staffed by ransportation ex- perts and identified a road reconstruction project requiring a loan of about US$30 million. Information on other needs was also collected but there is no evidence that serious consideration was given by the Bank to financing any reconstruction outside the highway field. 5. After reviewing the various problems which had prevented its lending to Peru, the Bank decided that a loan of the magnitude proposed by the mission would be justified under the special circumstances. The loan was processed quickly and, even though studies were to be done, the loan covered the entire project, instead of following the more commonly used procedure of granting an engineering loan which may be followed later by a construction loan. This provided a clear commitment by the Bank which under the circumstances was considered important. By the time the loan was made, substantial assistance from other international or bi- lateral donors had been committed or was under consideration, covering part of the other reconstruction requirements. II. The Project 6. The purpose of the project was defined as: "To improve access to and within the earthquake disaster area, the Callejon de Huaylas". It consisted of three parts: (a) feasibility study and detailed engineering of the main roads leading to and within the Callejon de Huaylas; (b) reconstruction and improvement of the selected road(s), including supervision by consultants and, possibly, such other high priority roads as may be agreed between the Borrower and the Bank; and (c) technical assistance for the Highway Department to improve maintenance and betterment operations. 7. The total cost of the project was estimated at US$41.75 million, of which US$33 million was the estimated construction cost (including the cost of supervision). The project was supported by a Bank loan for US$30 million which was approved by the Board of Executive Directors on Septem- ber 8, 1970. The loan covered some US$5.5 million of the local costs of the project as well as US$5 million for capitalized interest and commit- ment charges over a period of five years. Because of the special circum- stances following the earthquake, a 30-year term for the loan, including a ten-year grace period, was granted. Since only very global cost esti- mates were available, it was agreed during negotiations that any surplus construction funds could be used for another high priority road project to be agreed between the Government and the Bank. - 3 - III. Project Implementation 8. In physical terms, the project was generally completed as planned. With the assistance of the feasibility studies, the Government decided, in consultation with the Bank, that the road in the Huaraz valley from Catac- Caraz and the access road to the valley from Pativilca-Catac should be reconstructed under the loan (see map). Overall implementation of the project required 65% more time than anticipated but road construction was completed sooner. Because of the urgency of the project, no economic evaluation was included in the appraisal. Subsequent feasibility studies, however, anticipated an economic return of some 18%. The recalculated economic return is about 15%; although lower, this still seems to indicate that the project is making a useful contribution to the Peruvian economy. 9. The project was implemented as an ordinary highway project and the only deviations from normal Bank procedures were loan approval before project preparation was completed and the start of final engineering for the 23 km section Huaraz-Anta when the results of feasibility studies were not yet available. The project had a cost overrun of 14%. However, when comparing the amounts of construction contracts as awarded with the final costs, the latter were in some cases substantially higher because of increased quantities (partially due to design changes) and price escalation. 10. Implementation delays were the result of many factors as described in Chapter III of the attached PCR. Generally, it seems that both the Government and the Bank overestimated the speed with which various stages of the project could be carried out and underestimated the time required for the decision-making process in between stages. It should be especially noted that the Government created a separate agency to handle the recon- struction program. This facilitated coordination of the various recon- struction efforts but at the same time reduced the chance that officials would be available who had prior experience in dealing with international lending agencies. Disbursement procedures have also been rather slow and at one time, according to Government sources, officials involved with the project suggested that the Bank make disbursements on the basis of less detailed information on various cost items and only undertake an exhaustive review towards the end of the project, but this idea was rejected by Bank staff. 11. Some reasons for implementation delays were the rather late appointment of consultants for the feasibility studies (PCR, para. 3.03); problems in deciding on the most suitable location of the access road from the coast into the valley (PCR, para. 3.06); differences of approach to road design standards (PCR, paras 3.02(c), 3.05, 3.12); revisions of cost and traffic forecasts needed in the consultants- reports (PCR, para. 3.05); inadequacies in the final design leading to extensive redesign work (PCR, paras 3.11, 3.15, 3.16); inadequate capacity of contractors, and technical problems (PCR, paras 3.21, 3.23, 3.30); and the limited implemen- tation capacity of the executing agency (PCR, paras 6.02-6.07). Some of these points are commented on in the next chapter. 12. In spite of the above problems, a highway has been provided which so far has proved to be mostly of good quality. Some difficul- ties are, however, being experienced on a number of road stretches in the Conococha-Catac section where pavement failures occurred soon after completion. 13. Initially, the Government appeared rather reluctant to imple- ment the technical assistance portion of the project which was to assist in improving maintenance and betterment operations. Once started, the technical assistance was reasonably successful and has, according to Gov- ernment officials, led to definite improvements in the maintenance system (PCR, paras 3.37-3.44). The consultants, however, overestimated the avail- ability of local funds. Thus, where the system is now in place, there is not always sufficient money to do the actual work. A serious problem ac- cording to highway maintenance officials is the shortage of workshop build- ings which have to be paid for from the local budget. 14. The Bank's efforts to improve highway maintenance in Peru have been only moderately successful considering that in 1955 and 1961, it had already financed projects which were almost entirely devoted to this ac- tivity. The 1961 project included an agreement that vehicle weight would be limited and weigh scales financed under the project be installed to en- force the new rules. Presently, there is only very limited control on ve- hicle weights in Peru, in spite of the covenant dealing with this in the Loan Agreement for the project which is the subject of this audit. IV. Points of Special Interest Appropriateness of the Project 15. Project identification under the emergency conditions prevailing in Peru was complicated by the fact that part of the project area was in- accessible; little statistical information was available; and some Govern- ment officials were occupied with emergency repair activities which pre- vented them from spending as much time with the Bank mission as would have been possible otherwise. This meant that the Bank's identification mis- sion had to rely more than usual on judgment and familiarity with the coun- try. The Projects staff who visited Peru were highly qualified, but their extensive experience was heavily weighted towards transport, leaving little room for the possibility that the Bank would become involved in more gen- eral reconstruction work at a time when the extent of the damage was still being assessed and no commitments had yet been made for any reconstruction financing. The transportation staff who normally dealt with Peru and there- fore had the greatest knowledge of the country's institutions and project implementation capability in the transportation field were not available for the mission. 16. The project which resulted was named a road reconstruction proj- ect and was to cover reconstruction but also substantial improvements as included in the five-year plan. When the first Bank mission visited, the Government was already making emergency repairs to restore access to the - 5 - area and most roads were restored to virtually the same condition as before the earthquake well before the reconstruction under the Bank-financed project started. Subsequently, the roads covered by the Bank loan were rebuilt from mostly one-lane gravel before the earthquake to two lanes with an asphalted surface. Since the Loan Agreement required feasibility and engineering studies, construction work could be started only long after the earthquake. 17. The audit attempted to review whether a different type of proj- ect, such as reconstruction of houses, schools, water supply, or irriga- tion works, would have been preferable. It concluded that, be it somewhat by accident, the selection of a road project which required substantial preparatory work has been, seen in retrospect, not a bad choice. Most Government officials were of the view that because of the limited implemen- tation capacity and the need for urban planning before starting part of the reconstruction work, it is unlikely that, during the first 3-4 years, more money could have been spent on reconstruction than was actually avail- able. They were generally pleased with the project but saw it more as a contribution to the development of the area than to reconstruction. Period Required for Studies Might Have Been Shortened 18. Processing of loan approval was rapid, and the fact that the loan covered both the study and the construction phase may have helped to avoid some delays which normally occur when appraising and processing a second stage. After loan signature, however, the rather time-consuming project preparation process started. Some five months were needed for the Government to hire consultants for feasibility studies and subsequent detailed engineering. These studies altogether required over 2 1/2 years to complete. Upon the urging of the Government, engineering of the 23 km section, linking the new airport at Anta with the regional capital Huaraz, was started before all feasibility studies were completed and was followed quickly by bidding for road construction work; however, it still required five months before construction started. For the other road sections, however, construction started only after all studies were completed and in some cases extra delays occurred because of the need for some redesign work. 19. Governmenf officials appeared rather disturbed by the long time required for the roads to be completed, but some of them were resigned to the fact that this always happens with Bank-financed projects. When a project is urgent, fast loan processing usually contributes to an earlier start of the first steps required to implement the project than would have otherwise been the case. More essential, however, is that ways be found to reduce the planning period, recognizing that the resulting proj- ect may thus not be optimal, but balancing extra speed, where required, against the risk of lower quality. 20. In this particular case, it would appear that part of the re- construction and betterment could have been started without feasibility studies and using only general engineering information. The terrain in the Huaraz valley between Conococha and Caraz is not exceptionally dif- ficult and there was no problem of having to make the important choice - 6 - between different alignments. The road work could therefore have been awarded to a contractor on a cost plus or unit price contract, while the engineering would have been refined slightly ahead of construction by con- sultants who would also have supervised construction. 21. Whether the feasibility studies for the access road between the coast and the valley could have been avoided is open to question because of the uncertainty on the most appropriate alignment (see below). The ac- cess road was to traverse difficult mountainous terrain over most of its distance which made adequate engineering before the start of construction essential. Road Design Complications and Delays 22. The road has been built to fairly high design standards consider- ing the high construction cost in mountainous terrain and the relatively low traffic volumes, ranging on different sections from about 275-650 ve- hicles per day. The problem of what standards to use has been controver- sial, with the Bank at least at first insisting on modest standards, while the consultants in agreement with the Government wanted to design a higher standard facility. The fact that the loan had already been made reduced the Bank's influence and may have encouraged the Government and consul- tants to opt for higher standards since before the start of heavy inflation caused by oil price increases, ample funds appeared to be available under the project. 23. As pointed out in the attached PCR (para. 3.02(c)), part of the design met with difficulties because of the survey methods used and dis- agreements regarding the responsibility for the alignment stake-out. Super- vision missions by an engineer visited Peru about once every four months but problems with the survey method used were not recognized. Unfortunately as a result much time and money have been used for unsatisfactory work. 24. Finally, the executing agency for the project decided to change to different consultants who were requested to complete the design for and supervise construction of the access road to the valley. The original con- sultants supervised construction of most of the road in the valley. The new consultants not only completed the design, but also upgraded the road standards, relocated the road in several sections and changed part of the pavement design. It appears that these changes were caused by a combina- tion of Government pressure and initiative by the new consultants. Vari- ous modifications were helped by a change in the Bank's supervision staff, with the new engineer less inclined to economize through the use of lower road standards. One of the partners in the consortium of consultants who originally worked on the design mentioned to the audit mission that the considerable changes made by the new consultants in the alignment and especially the road standards cast an unjustified doubt on their work, since they were under pressure to come up with a different standard facil- ity. They agree, however, that the survey work was not fully satisfactory. The audit mission was unable to discuss this with the partner in the con- sortium who was most involved in the survey work, since the firm is no longer in business. - 7 - 25. The design proposed by the first consulting firm included a double surface treatment as pavement type for the entire road. The new consultants changed this for the 43 km section Conococha-Catac into an asphaltic concrete pavement. The changes were made because the latter pavement was judged to be more suitable for the high altitude section where there are large variations in temperature. On this section, how- ever, localized pavement failures resulted soon after completion of con- struction. There is considerable controversy over the reasons for the failures but Government officials are understandably unhappy with these poor results, especially since the pavement design changes resulted in substantial extra costs; and several officials still question the need for the different pavement. Alternative Access Routes and Their Impact on the Economy of the Project Area 26. Economic activity in the Huaraz valley consists of agriculture in the valley itself and on part of the mountain slopes. A limited amount of mining of copper, zinc, lead and silver is taking place in the valley and the mountains to the east. Tourism in the area is developing, based on the mountain climate, mountain climbing expeditions and visits to lakes and ruins in the vicinity. Before construction of the new road, the vari- ous economic activities were handicapped by high transport costs. 27. When the earthquake struck, the valley of Huaraz was accessible from the coast by three routes. The shortest route from Lima left the coast near Pativilca and entered the valley over a pass at Conococha. The second road connected the center of the valley at Huaraz with Casma on the coast. Both roads were winding and steep in parts with a coarse gravel surface. Vehicle operating speeds on most sections of the roads were quite low. The third route consisted of a road from the lower part of the valley at Caraz through a canyon to a narrow gauge railway which connected with the town of Chimbote on the coast. The road had a coarse gravel surface and was narrow and winding with many short tunnels. All three roads were blocked after the earthquake. The first two were apparently reopened fairly soon, but for the latter road repairs took about two years while the railway was abandoned and replaced by a road which follows the railway alignment to Chimbote. Economic activity along the access roads is very limited. 28. Under the feasibility studies, numerous alignments to the coast were considered and ultimately two routes, one from the valley to Casma, the other to Pativilca, were retained for further analysis. However, the recommendation in the final report of the feasibility study in favor of the road to Pativilca was in conflict with official plans to promote Chimbote as a regional center. This contributed to some delays, but ulti- mately the Government decided to build the road to Pativilca. 29. The audit found that Government officials are still uncertain about which route would have been of the highest priority and several appeared to favor the road to Casma. The former route appears to be most -8- suitable for transport of agricultural products which, according to Govern- ment officials, move between the valley and the coastal region to the north (Chimbote and Trujillo), even though some produce from the valley is also sold to the Lima market. The Pativilca road is more economical for transport of most of the mining production which has to go to the ports south of Pativilca (Huacho and Callao) since Chimbote has no special facili- ties for mineral exports. This route is also the shortest for tourists from Lima to the valley of Huaraz. 30. Before the earthquake, the Huaraz region was poor and underdevel- oped with a population which in Huaraz was increasing marginally, while in other parts of the valley it decreased. Current population trends are not clear because many people left the valley after the earthquake and have only gradually returned. Some Government officials, however, feel that the new road facilities aid the outflow of the population to Lima. 31. Government officials viewed the new road as an important contrib- ution to the development of the Huaraz valley. They also thought that, for employment purposes, construction had started at about the right time because, when demand for labor weakened due to the partial completion of reconstruction work, laborers were needed for the construction of the road, even though the more skilled workers were brought in from other parts of Peru. When discussing details of the economic impact of the project, it appeared to the audit that the country as a whole seems to benefit from the new road, but the population in the region served by the project shares in these benefits only to a limited extent. 32. Mining in the road's area of influence has reportedly increased in the last few years. The reduction of about 40% in road transport costs is expected to have made a significant contribution to this, even though, for silver and lead, price increases must also have played a role. Offi- cials in Peru felt that the employment impact of the increased production was limited because of increasing mechanization and the small percentage of the population employed in the sector. The savings from decreased transport costs are mostly benefitting the mining companies. 33. Tourism in the area is growing rapidly but, again according to local officials, the employment effect is limited with total employment estimated at less than 200. Part of the tourists are foreigners whose visits to the valley are of the longest duration. Local tourists come mostly for week-ends from Lima and thus especially benefit from the im- proved road. 34. Government officials felt that the new road was of limited value to the agricultural sector in which most of the valley's population is employed. This is especially the case because irrigation possibilities are very limited, while no new land is available to be brought under culti- vation. The project as formulated in the appraisal did not define as a specific purpose the need to provide assistance to the population of the valley, only to improve access. -9- Financing of Cost Overruns 35. As mentioned earlier, the project had a cost overrun. In itself, however, this is not significant, since the original cost estimate was not backed by any engineering studies. The Government experienced difficulties in finding the necessary funds to cover the extra costs of the project and finally one of the contractors provided the financing. Compared with the interest charged on World Bank loans during that period, the financing was only 0.3% p.a. more expensive. However, the Bank normally lends for a pe- riod of 15-20 years while this loan is for a term of only 6 1/2 years, in- cluding 1 1/2 years grace. 36. When the project was discussed by the Board, the Bank's manage- ment took the position that excess funds in the loan could be used for road construction elsewhere in Peru, while also warning the Board that if there were a shortage of funds, a supplementary loan might have to be considered. There is no record of the Government having requested extra financing, nor of the Bank confirming the above commitment when it was obvious that extra funds were needed. V. Conclusions 37. This project provides another example of the Bank's difficulties in providing rapid assistance after emergencies.!/ As pointed out, the project was a successful development investment for the country, but it made virtually no contribution to alleviating the emergency problems, nor did it do much to improve the lot of the local population in the project area. Even though the processing of the loan was done very promptly, the Bank apparently was not prepared to take any unusual risks and still in- sisted on feasibility studies and detailed design work before construction was started. Faster action could only have been achieved by starting re- construction work without feasibility studies while limiting detailed en- gineering to road sections traversing difficult mountainous terrain; this procedure might be considered under future emergency conditions. In the view of the audit, this approach would have been possible for at least part of the reconstruction project and would have considerably speeded up the completion of an improved road. 38. On the basis of discussions with Bank staff, the audit concluded that, in order to follow the above described procedure, it would be desir- able that emergency projects be executed, wherever possible, without, or with a minimum of the usual project preparation studies, relying on quali- fied Bank staff's judgment. To achieve this, there would have to be a clear understanding between management and the Board that in order to start work immediately, some of the normal rules on design and procurement may have to be waived while the risk would also have to be explicitly accepted that the project may ultimately not meet the normal standards of economic 1/ Some other audits dealing with emergency projects are Nigeria Trans- port Rehabilitation Project (SecM78-520) and Iceland Fishing Harbours Rehabilitation Project (SecM78-211). - 10 - justification.1/ While the relative lack of first-hand knowledge of the country on the part of the appraisal mission in this case was not the cause of any significant shortcomings in project preparation, some aspects might have been given better consideration had the mission included persons fa- miliar with Peru; among these aspects are the estimate of the speed with which the project could be implemented and a greater consideration of al- ternative project elements designed to help the poor in the region. Given the special nature of the problems to be dealt with in these circumstances, particularly the need for speedy action and a broad view of a country's im- mediate needs, it is highly desirable that emergency projects are assigned to the staff members who are most intimately involved with the country con- cerned, even if this might entail interruption of the activities on which they are currently committed, or borrowing them from other departments. 39. The creation of a new Government agency to achieve better coor- dination of reconstruction work and later of regional development activi- ties was useful in the long run. However, new agencies usually need time to gain experience which leads initially to delays in the decision-making process. The project provides an example of this where many.officials con- cerned with the reconstruction work felt that the limited implementation capacity of the new agency has contributed to delays. 40. Judging from the limited information available, the project ap- pears to be an example of investments which result in better accessibility of a poor region but still contributed little to raising the standard of living of the poorer segment of the population, with the greater part of the benefits leaving the project area. It appears that construction of a different access road into the valley (Casma-Huaraz) would have provided more benefits to the poor, but might have resulted in lower total benefits. A more definite conclusion, however, would require further study which is beyond the scope of this audit. 41. Bank supervision missions have visited Peru frequently but there was only one visit by an economist during the first 2 1/2 years of the proj- ect (including the appraisal); it seems that a greater involvement in this field would probably have been desirable, in view of the complexity of the economic problems of the project. More guidance by the Bank's engineers might have helped to avoid some of the location problems which have plagued the project. The project also shows the importance of continuity in the supervision missions' approach to such elementary aspects as design stand- ards even when Bank staff members succeed each other. 1/ Recent projects financing reconstruction of damage resulting from natural disasters have introduced a more flexible and imaginative approach, allowing for a faster start of reconstruction work. - 11 - ATTACHMENT PROJECT COMPLETION REPORT PERU ROAD RECONSTRUCTION PROJECT (LOAN 706-PE) I. INTRODUCTION A. Background 1.01 On May 31, 1970, a devastating earthquake shook Peru. The zone of greatest damage was in the Department of Ancash (Map IBRD 14016) between the coastal towns of Pativilca and Chimbote (respectively about 200 km and 400 km north of Lima) and the Andean range known as the Cordillera Blanca, running roughly parallel to the coast and 150 km inland. Particularly hard hit was a high Andean valley between the Cordillera Blanca and the Cordillera Negra, called the Callejon de Huaylas, through which the Santa river runs. Avalanches of ice, stone and mud covered whole towns and villages and disrupted power supplies and communications. 1.02 The Department of Ancash had been an important contributor of agricultural produce to the markets along the coast; the Callejon de Huaylas had been a relatively intensely cultivated valley, with possibilities for further agricultural expansion but with development limited by chronic diffi- culties in road access. The scenic valley had also been a local tourist attraction, with many tourists choosing to use the Chimbote-Huallanca railroad to avoid the hazardous access by road. This short-line railway had a branch entering the north end of the valley and reaching Huallanca, some 50 km north of Caraz. The need for an efficient transport outlet to remove the constraint to the valley's potential development had, in fact, been recognized before the massive earthquake, and improvement of two access roads (Pativilca-Huaraz and Casma-Huaraz) had been included in the Government's investment plan for 1971 to 1975. 1.03 Considerable damage to the transportation infrastructure serving the valley occurred as a result of the earthquake. The roads from Pativilca and Casma, on the coast, winding up to the valley through the Cordillera Negra, were blocked by numerous slides, and the branch of the Rio Santa railway serving Huallanca was destroyed beyond economic repair. The 106-km road section between Huallanca and Carhuaz, running along the valley itself, suffered the most extensive damage of all; sections of this road disappeared in the avalanches or were completely blocked by landslides. 1.04 Prior to the earthquake, the annual average daily traffic (AADT) on the Hualanca to Huaraz part of the valley road (109 km) had been about 300 vehicles; that on the Pativilca to Huaraz road (207 km), about 200 vehicles; and that on the Casma to Huaraz road (149 km), about 160 vehicles. 1.05 Emergency reopening of land access, crucial to the restoration of communications within the disaster area and to relief work, was immediately undertaken by the Government. Roads were reopened, but, in many locations, for only one lane of traffic. Permanent reconstruction of the roads was of vital importance to the economy of the area and to its social and economic integration within the rest of the country. On June 4, 1970, the Government of Peru asked the Bank for assistance in financing rehabilitation and recon- struction works. The total reconstruction aid was provisionally estimated by the Government at US$516 million, and the Bank was asked to provide US$150 million. - 12 - 1.06 The project was appraised in June 1970, and, on September 8, 1970, the Board of Directors approved a US$30 million road reconstruction loan to the Government of Peru. The project was aimed at: (a) restoring access to the earthquake area through improved roads leading to and within the Callejon de Huaylas, which would help to revitalize the economy of the valley; and (b) providing technical assistance to the Government in road maintenance, as a continuation of previous Bank-financed assistance in this field. B. Previous Assistance to the Transport Sector 1.07 Four previous Bank loans had been made for highway projects in Peru. Loans 127-PE (US$5 million, 1955) and 300-PE (US$10 million, 1961) had provided technical assistance in organization and road maintenance operations. Loans 271-PE (US$5.50 million, 1960) and 425-PE (US$33 million, 1965) had as their objective the reconstruction and upgrading of the La Oroya-Pucalpa road. 1/ II. PROJECT PREPARATION, APPRAISAL AND NEGOTIATIONS A. Preparation and Appraisal 2.01 The emergency nature of the situation in the earthquake-affected area made rapid loan processing important. In mid-June of 1970, a Bank mission joined a UNDP mission already in Peru to survey the area and review the reconstruction needs. The.mission performed a combined identification, preappraisal and appraisal role. After discussions with the Government concerning how best the Bank could assist, the mission concluded that Bank financing should cover a road reconstruction project in the earthquake- disaster area, together with further technical assistance in road maintenance. The following three possible road projects were identified: Preliminary Road Estimated Total Cost (US$ million) Huaraz-Casma 11.0 Huaraz-Pativilca 22.0 Huaraz-Chimbote 30.0 2.02 The recommended technical assistance for highway maintenance and betterment of national and departmental roads was not to be limited to the earthquake zone. Much of the effect of previous Bank assistance in improving maintenance organization and operations had been dissipated by the lack of adequate maintenance budgets. The Government had neither the funds nor the 1/ Two further Bank-assisted projects including additional works for improvement and paving of this highway (Loans 1025-PE, July 1974 and 1196-PE, May 1976) are still in execution. - 13 - administrative capacity to undertake new road construction programs without neglecting the maintenance of existing roads. Inadequate detailed engineering designs of several new roads had led to large cost overr,ns and diversion of maintenance funds to construction. In consequence, past loan covenants for maintenance were not implemented, and equipment became idle for lack of spare parts while roads deteriorated further. A serious backlog of road maintenance had accumulated. B. Issues 2.03 The size of the loan, the structure of the lending program and the loan repayment terms were the principle issues during project preparation. Peru's initial request had been for US$150 million in soft terms assistance, and Peruvian officials were reluctant to borrow on conventional terms under the circumstances. The size of the loan for the project was established at two-thirds of the estimated cost of the highest cost alternative that had been identified for possible reconstruction, plus financial provision for studies and technical assistance. This resulted in a US$25 to US$30 million figure, with any surplus eligible for other roadworks. Two types of lending were considered: (a) a small, immediate loan or credit for feasibility and design studies, to be followed by a road construction loan in 1971; or (b) a single loan encompassing both operations. The second alternative, essentially a "line of credit" which provided for an early commitment of funds, was seen as more appropriate. The possibility of IDA assistance was reviewed and discarded. One important reason was that IDA was not a disaster-relief organization and should not be given that image. However, in view of the essentially humanitarian nature of this loan and of Peru's limited credit- worthiness, exceptionally favorable repayment terms were deemed appropriate, i.e., 30 years, including a 10-year grace period, at 7-1/4% interest per annum, with interest and commitment charges paid out of the loan account during the first five years. 2.04 The Government was informed of the Bank's intention to consider a loan and, on August 7, 1970, was formally invited for negotiations. C. Negotiations 2.05 Negotiations were held in Washington, D.C., August 10-13, 1970, and, on September 8, 1970, the Executive Directors approved a loan amounting to US$30 million. The loan became effective on October 26, 1970. 2.06 The principal points discussed during negotiations were: (a) that the Government's request for the amount of the loan to be left open until feasibility studies had been completed was not possible and therefore the amount of US$30 million was agreed; (b) that the size of the Bank's financial contribution should not limit the scope of the feasibility studies' recommendations and that road alternatives should not necessarily be considered mutually exclusive; - 14 - (c) that road selection would be a matter of agreement between the Bank and the Borrower; (d) that the road maintenance technical assistance consultants would work within the Ministry of Transport and Communications (MTC) and not as a separate office; (e) that the Government would furnish the Bank with a letter of intent stating that (i) the road maintenance and construction budget would be prepared jointly with the technical assistance consultants and would be submitted to the Bank for review and (ii) that the Government would set aside adequate funds for project-related costs; and (f) that the commitment charge on funds for road construction would start 365 days after loan signature rather than the customary 60 days that applied to other loan categories. D. Description of the Project 2.07 The project, as described in the Loan Agreement, dated September 14, 1970, comprised the following: Part A (i) A feasibility study of the main roads leading to and within the Callejon de Huaylas, including such location studies as might be required; (ii) Detailed engineering of the roads or road sections selected for reconstruction and improvement by the Borrower and the Bank after the completion of the feasibility study referred to in (i) above and of such other roads or road sections as would be agreed upon by the Borrower and the Bank. Part B The reconstruction and improvement of the roads or road sections so selected, including supervision by consultants. Part C Technical assistance to enable MTC to improve maintenance and betterment of national and departmental roads, with field assistance in two pilot regions, the location of which would be agreed upon by the Borrower and the Bank. The project was expected to be completed by March.31, 1975. - 15 - 2.08 The estimated costs, as shown in the appraisal report, were: Cost Estimates (in US$ equivalent) Government- Bank- Description Total Financed Financed FEC (%) I. Studies and Technical Assistance (i) Feasibility Study (including location studies) 700,000 140,000 560,000 80 (ii) Detailed Engineering: Chimbote-Huaraz 1,000,000 Casma-Huaraz 500,000 Pativilca-Huaraz 800,000 Maximum total /1 would be: 2,300,000 Assumed to be actually spent /2: 1,550,000 310,000 1,240,000 80 (iii) Technical Assistance for Maintenance 1,500,000 300,000 1,200,000 80 Subtotal: 3,750,000 750,000 3,000,000 II. Road Reconstruction /3 (including supervision by consultants) 31,500,000 10,500,000 21,000,000 67 III. Unallocated 1,500,000 500,000 1,000,000 67 IV. Capitalized Interest and Commitment Charges over Five-Year Period 5,000,000 - 5,000,000 - Total 41,750,000 11,750,000 30,000,000 /1 If all three alternatives were to be designed. /2 Estimate based on fewer than three alternatives actually designed. /3 Estimate based on only the highest-cost road alternative, being financed under the loan. - 16 - III. PROJECT IMPLEMENTATION A. General 3.01 It took two and a half years longer to complete the project than was anticipated at the time of appraisal. Preparation for the construction of civil works (feasibility studies, final design and bidding procedures) took between six and 15 months longer than scheduled. Then construction contracts ran eight to 18 months over initial estimates. Technical assistance for maintenance did not start until 1974 (three years late) and the original two years of services had to be extended an additional year to permit completion. 3.02 Coupled with the underestimates by the Government and the Bank of the time needed to move from phase to phase of work (a function of Government and Bank reviews and procedures), new problems arose in each phase that carried over into successive phases, aggravating the total delay. In general terms, the main problems were concerned with: (a) Project Management. At the time that the loan was negotiated, MTC was the Government agency responsible for supervising the execution of the project. In March 1973, a new authority was created for the earthquake-affected region called ORDEZA. 1/ It took over a year to implement a full shift in authority effectively, during which time there was dual executive control with attendant bureaucratic and decision-making problems. (b) Feasibility Studies. These were concluded on time. However, although the route selection was acceptable to the Bank, the consultants (para 3.03) failed to analyze and incorporate the potentially substantial cost savings that could (and eventually did) result from applying localized variations to design criteria and road standards on a case-by-case basis. Incorporation of this concept was passed on to the final design phase where it became a source of continual contention and delay. (c) Final Design. Resistance was encountered on the part of the consultants, and some MTC technical staff, to selectively applying exceptions to "desirable" standards, which eventually led to the need for unnecessary redesign, and, additionally, a design process based on aerial photography was adopted, in which there was insufficient expertise. Inadequate mapping controls, combined with terms of reference that failed to specify the consultants' responsibility to stake out the designed align- ment and grades on the ground, led to long delays in the start of construction when the survey stakeouts revealed substantial differences between field conditions and the mapped topography. Internal dissension within the original consulting joint venture, and resulting poor management, contributed to the problem. 1/ Organismo Regional para el Desarrollo de la Zona Afectada. (d) Construction. The start of virtually all cui.tracts was delayed by lack of staked-out work and incomplete designs. In addition, contractor mobilization was slower than expected, reflecting overoptimism on the part of the Government, the consultants and the contractors. The capacity of Peruvian contractors to execute the work was also overestimated, and the problem was compounded by work quantity increases due to slides (typical in Andean construction) and necessary additions to the original design. Insufficient equipment and difficulties in obtaining spare parts most frequently constrained contractor performance, while lengthy Government procedures in increasing contract ceilings and insuffi- cient budget allocations contributed to the total delays. B. Feasibility Studies 3.03 A contract for a feasibility study was signed on February 19, 1971 with Consultant Group A. The consultants were required to study the improve- ment of an access road from the coastal plain to the Callejon de Huaylas and the reconstruction and improvement of the road along the valley itself. 3.04 Eight different passages to the valley through the Cordillera Negra were studied and analyzed in the interim report of May 11, which concluded that only two existing routes, one from Casma to Huaraz and the other from Pativilca through Conococha, warranted further study. In both their interim and final reports (October 1971), the consultants recommended that the southern access road, Pativilca-Conococha-Catac, be selected for improvement. Although the total road investment was economically justified, the proposed construction in the most difficult mountain sections, with lower traffic and higher costs, was economically marginal. 3.05 In the consultants' interim report, the preliminary cost estimates were too low, whereas both the preliminary traffic forecasts and estimates of benefits were too high. These factors led the consultants to adopt a rather optimistic view of the road standard that could be justified and constructed. The proposed standard conformed to MTC's requirements for a 60 kph design speed and called for a road platform of 8.50 m, 6.00 m paved with a double surface treatment and shoulders of 1.25 m on each side. In the Callejon valley, the new road was to follow substantially the existing alignment, while considerable realignment was involved on the access road. In the final report, cost estimates were increased by 15% to reflect higher prices and to compensate for additional earthwork quantities, which, from Peruvian experience of landslides in mountain cuts, were to be expected. Traffic estimates were reduced. However, the concepts of a uniform standard and strict adherence to design criteria continued to pervade, resulting in some costly choices of road and bridge locations. The consultants concluded that the road improve- ment as a whole (access road plus valley road) had an economic return (ER) of 17.8%. 3.06 The Government was initially undecided on whether to accept the consultants' recommendations on the access route. As part of a decentraliza- tion policy to limit the growth of Lima, the capital, the port of Chimbote and its surroundings had been proclaimed a development area. The consul- tants' recommendation that the access route run from Pativilca to Conococha (instead of Chimbote-Caraz) reduced the possibility that, if the copper mines - 18 - at Altamina were developed, the output from the mines could be exported through Chimbote. However, by March 1972, the consultants' recommendations were accepted. 3.07 The Bank expressed general satisfaction with the economic methodology and the conclusion of the report regarding route selection, but disagreed with the strict adherence to uniform design standards which were considered high in relation to the anticipated traffic volumes. Initial AADT in 1970 was estimated at about 200 vehicles on the access road and 360-480 vehicles on the Callejon valley road, growing to 990 vehicles and 2,350 vehicles on each road respec- tively by 1990 (8% annual growth). The Bank requested, in a letter dated March 24, 1972, an annex to the consultants' report in which the reduc- tion in design standards and related costs in difficult terrain would be further explored. The consultants produced a short addendum in which they recommended that the paved width be maintained at 6.00 m, but that the width of the shoulders be reduced where the reduction would substantially lower costs. C. Detailed Engineering Design and Construction Supervision 3.08 The Bank then requested, and the Government agreed, that the detailed engineering contract include a provision requiring the consultants to seek, in each subsection, economically justifiable cost reductions by selectively lowering the design standards where this was warranted. 3.09 Detailed engineering was carried out in essentially three sections, with the Catac bridge and a short bypass added later. Design of the road section between Huaraz, the regional capital, and Anta, which had the region's only airport and was consequently important to relief operations, was con- tracted, in August 1971, to Consultant Group A, while it was still working on the feasibility study. AADT on the section was around 400 vehicles, and no alternative routes required consideration. 3.10 At this time, the decision was taken to design the roads using photogrammetric mapping, a process in which only Consultant M, of Consultant Group A, had any previous experience. The actual surveys were conducted by the Servicio Aerofotogrametrico Nacional (SAN) and the Instituto Geografico Militar (IGM) on behalf of the consultants, which spread the responsibility for subsequent problems. At the same time, internal discord was prevalent within Consultant Group A. Management was weakened when the senior Consul- tant L representative returned to France, leaving his company unrepresented in the Lima headquarters. The trouble culminated with Consultant M withdrawing from the joint venture for construction supervision, at the time the first road section was contracted, while Consultant N took no further part in design. 3.11 Design of the remainder of the road was concentrated in July 1972 to Consultant Group B,1/ with completion of the valley road section (Huaraz-Anta excepted) scheduled for January 1973 and that of the access road section for August 1973. The problems with the photogrammetric mapping were not immediately 1/ This Group included the leading firm from Consultant Group A. - 19 - apparent. They stemmed from the use of control points that were normal for geodetic survey but too far from the road to give the accuracy required for a road alignment, particularly in the difficult mountainous terrain. In addi- tion, most of the control points, such as mountain peaks, were inaccessible to the survey crews staking out the alignment. Also, disagreement arose concerning the consultants' reaponsibility for conducting these stakeout surveys, carrying over into the construction phase, at which time construction work was delayed when sizable differences emerged between actual conditions and those shown on the plans. Completion of stakeout surveys for the contracts on the valley roads eventually took until March 1974, seven months after the start of construction. 3.12 The valley road designs, completed in January 1973 and approved by MTC in February 1973, were not acceptable to the Bank. The plans failed to incorporate the desired reduction in road standards at some locations and actually increased the road width over what had been agreed previously. As a result of Bank objections, Consultant Group B requested MTC permission, by letter of March 5, 1973, to incorporate the design changes requested by the Bank. The need for seeking lower cost solutions by selectively reducing standards was accepted at policy-making levels of the Government; however, the technical levels tended to insist on, or to support, the consultants' rigorous application of Peruvian road standards. The design revisions eventually reduced estimated costs of the combined Caraz-Anta and Huaraz-Catac road sections by 15%, from S/. 300 million (US$7.7 million) to S/. 260 million (US$6.7 million), but redesign delayed the signing of construction contracts by about four months. 3.13 Supervision of construction was contracted to Consultant Group C- in July 1973. The supervisors inherited the stakeout problems that had contributed greatly to the slow start of construction. Three months after the start of construction work on these valley road sections (two contracts), both contracts had only 5 km staked out and were two months behind schedule. 3.14 Dissatisfaction with the performance of the consultants led to ORDEZA eventually requesting new proposals in November 1973 for construction supervision of the four contract sections of the access road (including the remaining redesign and stakeout). Consultant Group D was selected by ORDEZA in December 1973, but the Bank considered its staffing proposals excessive, which delayed contract signing. 3.15 Although a final contract acceptable to the Bank was not signed until June 21, 1974, ORDEZA awarded the supervision contract on February 6, 1974 and survey and redesign work started almost immediately. At the end of February 1974, Consultant Group D was given the first 10 km of each road section, which had already been staked out by Consultant Group B, to check. Many errors were found and corrected. Work then continued on the remaining 120 km. Apart from relatively minor corrections, Consultant Group D also had to make substantial realignments: 10.3 km on the first section, 14 km on the second, 10 km on the third and 5 km on the section-closest to Pativilca. 1/ This Group included the leading firm from Consultant Group A. - 20 - 3.16 In addition to changes in highway alignment and geometric design, Consultant Group D recommended a revision to the pavement design for the Catac- Conococha and Conococha-Colca sections, which was accepted by MTC and the Bank (Annex 1). The originally recommended double surface treatment and base were replaced with 10 cm of asphaltic concrete plant mix, which was considered more resistant to the harsh climatic effects of the Andes. 3.17 A summary of design contract data and responsibility is shown below (costs in millions): Date of Actual Contract Section Km Signature Completion Value Consultant US$ S/. Equivalent Valley Road 1. Anta-Huaraz 23.0 8.71 12/71 79,668 2.0 Consultant Group A 2. Caraz-Anta 46.0 7/71 1/73 129,902 3.2 Consultant Group B 3. Huaraz-Catac 34.4 7/72 1/73 177,511 4.6 Consultant Group B 3.A Catac Bridge (35 m) 5,650 0.1 Consultant Group C/MTC Access Road 4. Catac-Conococha 45.5 7/72 8/73 174,347 4.5 ) 5. Conococha-Colca 40.8 7/72 8/73 177,145 4.6 ) 6. Colca-Chasquitambo 33.3 7/72 8/73 129,826 3.4 ) Consultant Group B 7. Chasquitambo- Pativilca 42.2 7/72 8/73 130,855 3.4 ) /I On June 21, 1974, Consultant Group D was contracted for the remaining redesign and construction supervision. D. Construction Contracts (i) General 3.18 Construction of the road was contracted in seven sections. Respon- sibility for the construction and supervision was as folows: Supervising Section Contractor Consultant Anta-Huaraz Contractor Z Consultant Group A Caraz-Anta Contractor Y Consultant Group C Huaraz-Catac Contractor X Consultant Group C Catac-Conococha /l Contractor W Consultant Group D Conococha-Colca ) Colca-Chasquitambo ) Contractor V Consultant Group D Chasquitambo-Pativilca) /l Including the Catac Bridge. - 21 - 3.19 Physical progress on all road sections was significantly affected by the following factors: (a) delays in staking out the road, which also delayed property acquisition surveys and resulted in design adjustments; (b) slow contractor mobilization, insufficient equipment, and difficulties in obtaining spare parts and in importing equipment (foreign exchange problems and customs clearance); (c) increases in work quantities, primarily on the access road, because of slides of unstable slopes; (d) lack of a price escalation clause in the construction contracts (claims by contractors were allowed under the contracts, but specifically excluded were readjustments for increased prices of equipment) during a time of rapidly rising equipment costs, which resulted in hardships for contractors in making up equipment deficiencies; (e) frequent changes and, at times, vacancies in MTC and ORDEZA management positions having direct responsibility and decision-making authority over the contracts; (f) frequent delays in payments to contractors; and (g) in 1976, when Bank participation was reduced from 67% to 40%, delay by the Ministry of Finance in allocating additional funds for civil works and construction supervision, which had not been envisaged in the ORDEZA annual budget. (ii) Anta-Huaraz (21 km) 3.20 The Anta-Huaraz section was the first construction work to be contracted. On February 3, 1972, seven bids from prequalified contractors were received by MTC and, after an exchange of correspondence concerning the methods and criteria of bid evaluation, the Bank agreed to award the 18-month contract to Contractor Z, the third lowest bidder, for S/. 80 million (US$2.1 million). 3.21 In July 1972, reconstruction began slowly and remained behind schedule throughout the contract. The main problems stemmed from insufficient construction equipment, an inadequate aggregate crushing plant, delays in receiving construction materials (steel, bitumen and cement) and spare parts for the equipment and delays in payments to the contractor by ORDEZA. Construction work was also delayed because of poor coordination with other agencies responsible for the water system of the town of Huaraz and the removal of a power transmission tower. By December 1973, the contractor's aggregate production problems had been largely solved, but the rainy season was under way and the contract was extended from December 1973 to June 1974. Cost overruns at this point were about SI. 13.8 million (US$0.4 million), or - 22 - about 17% of the original contract costs. By June 1974, the road section was nearly complete, except for some unstabilized slides. The contract was finally completed in August of 1974 at a cost of approximately S/. 109.5 million (US$2.8 million), 40% above the original contract amount. The final cost of construction supervision by Consultant Group A was SI. 14.1 million (US$0.3 million), 10% of the construction cost. (iii) Caraz-Anta (46.1 km) and Huaraz-Catac (81.3 km) 3.22 Nine bids were submitted for these two contracts, although 27 contractors had been prequalified. No foreign contractors presented bids. The contracts were signed in late August 1973. Contractor Y was awarded the contract for the Caraz-Anta section for an amount of S/. 125 million (US$3.2 million); Contractor X was awarded the contract for the Huaraz-Catac section for an amount of S/. 177.5 million (US$4.6 million). Each contract was for a period of 24 months, and their combined total cost was 16% over the engineers' estimate. 3.23 Work on both contracts began very slowly in November 1973, impeded by the need for a staked-out alignment and for property surveys along the right-of-way. Insufficient construction equipment remained a problem for a long time, particularly on the Caraz-Anta road section where the contractor was eventually given a deadline, by ORDEZA, of July 1975 for completing his fleet of equipment. This was virtually the scheduled completion date for both contracts, which were, at that time, respectively 50% and 57% complete. In late 1975 and 1976, delays in allocation of funds by the Government led to delays in payments to contractors, with further ill effects on physical progress. 3.24 On the Caraz-Anta section, poor work programing by the contractor also contributed to the delay; at one stage, there were some 45 km of sub-base and earthworks exposed at the start of the rainy season. On the Huaraz-Catac contract, difficulties in importing new equipment and frequent breakdowns of the equipment working on the job were the principal causes of delay until the period of financial uncertainty in 1975-1976. Construction on both sections was finally completed in January 1977 at a cost of S/. 155.2 million (US$3.4 million) for the Caraz-Anta section and SI. 224.2 million (US$4.9 million) for the Huaraz-Catac section, 6.2% and 6.5%, respectively, above the original contract amounts. Consultant Group B supervision costs amounted to S/. 65.1 million (US$1.5 million). (iv) Catac-Pativilca (163 km) 3.25 . The 163-km access road was divided into four construction contracts: Catac-Conococha, Conococha-Colca, Colca-Chasquitambo and Chasquitambo-Pativilca. Although design revisions were still required, bids for the four contracts were invited on September 1, 1973. A decision by ORDEZA to construct a short bypass around Catac with a new two-lane bridge delayed the opening of bids, while plans were being prepared by Consultant Group C and MTC, to October 30, 1973. The contracts were awarded as follows: - 23 - Cost (millions) Section Contractor S/. US$ 1. Catac-Conococha Contractor W 174.3 4.5 2. Conococha-Colca ) 177.1 4.6 3. Colca-Chasquitambo ) Contractor V 129.8 3.4 4. Chasquitambo-Pativilca) 130.9 3.4 3.26 The signing of the contracts and start of work were delayed for redesign and alignment stakeout until mid-May 1974, when a limited program of work was started. At the Bank's request, the contractors deferred full equipment mobilization to avoid the possibility of later claims based on not having enough work to utilize their equipment fully until sufficient staked-out road was ready. 3.27 On the Catac-Conococha section, Contractor W was soon about two months behind schedule. In late 1974, progress improved and, by early 1975, the volume of work executed was about two months behind the revised schedule. This pace continued through most of the year until, in 1976, financial and contractual problems delayed the work and further slippage occurred. 3.28 Work on Contractor V's three contracts also fell behind schedule during the first two months for lack of completed designs and the Bank's request for deferred mobilization, but Contractor V rapidly caught up during the second half of 1974 and worked on schedule until the end of 1975. 3.29 Execution of works under the contracts throughout the first half of 1976 was adversely affected by the lengthy Government proceedings required to obtain an increase in contract ceiling to cover additional work quantities. In the case of Contractor V's contracts, the quantity increases were primarily for removal of slides, but also for a change in pavement design in the Conococha-Colca section. On Contractor W's contract, it was primarily the new asphaltic concrete pavement construction that necessitated contract modifications. This was also the year in which ORDEZA was having problems in receiving an increased budget allocation from the Ministry of Finance when the Bank reduced its percentage participation in financing because of the increased total project cost. 3.30 The principal technical problems involving the construction of the road occurred primarily along the Catac-Conococha and Conococha-Colca road sections where highly plastic clays had to be removed and replaced with granular material. Even toward the end of the construction period in the Catac-Conococha section, one of the earth cuts was displaying stability problems, apparently due to underground water seepage. In addition, a major embankment showed signs of substantial movement with cracks appearing on the surface of the roadway. These problems were eventually corrected; however, the Catac-Conococha section is apt to require substantial maintenance. A - 24 - more general problem that occurred on several road stretches was caused by prolonged exposure of primed base to traffic, prior to placing the asphalt pavement. This occurred on both of Contractor V's sections, and on Contractor W's Catac-Conococha section. Several road stretches had to be sacrificed, recompacted and reprimed. The Bank informed ORDEZA that it would not finance any base repair work resulting from improper construction practices. 3.31 The Catac-Conococha section was the last to be completed. Part of the section received provisional acceptance on January 24, 1977, but all paving was not completed until August 1977. Final acceptance was scheduled for January 1978. The final cost of the contract at initial unit prices was S/. 224.1 million (US$3.0 million) and, including price escalation adjustments, was S/. 323.0 million (US$6.9 million). 3.32 The Conococha-Colca and Colca-Chasquitambo road sections both received provisional acceptance on February 24, 1977 with final acceptance six months later. The final contract cost, at initial contract prices, for the Conococha-Colca section was S/. 285.7 million (US$3.9 million), an increase in cost of 61% due to large increases in earthworks, which, with price escalation added, rose to SI. 413.1 million (US$8.8 million). Colca- Chasquitambo cost S/. 135.3 million (US$1.8 million) at initial unit prices and, with price contingencies included, S/. 166.9 million (US$3.6 million). 3.33 The Chasquitambo-Pativilca road section received provisional acceptance in September 1976 and final acceptance one year later. This section cost S/. 108.6 million (US$2.3 million) at initial contract prices, a decrease of 17% from contract cost, but, with price escalation added, ended up costing S/. 153.6 million (US$3.3 million). 3.34 The Consultant Group D supervision contract cost S/. 114.4 million (US$1.5 million) at initial contract prices, 1% over original contract amount, and, with price escalation added, S/. 146.5 million (US$3.4 million). The redesign and construction supervision amounted to 19.8% of the final construction cost, which is high, but not unrealistic, considering the amount of redesign work undertaken and the contract extensions. E. Technical Assistance for Maintenance 3.35 The loan included an estimated US$1.2 million to finance 80% of technical assistance services to help MTC in improving road maintenance and betterment operations. It took three years to start this component. MTC was reluctant from the outset to initiate the technical assistance, citing its high cost. MTC also took the position that the increased budget for road maintenance, from S/. 200 million (about US$6.5 million) in 1969 to S/. 530 million (US$13.7 million) in 1972, was more important and effective than technical assistance. Eventually, in December 1971, terms of reference for the technical assistance services were finalized, and proposals from five invited consultants were received in June 1972. All proposals were rejected by MTC on the grounds that four had failed to include the taxes on salaries of experts, which had been requested - 25 - by MTC, and that the one firm which had done so was too costly. It was the Bank mission's opinion at the time that the Ministry was making use of any possible delaying tactic to prevent consultants from starting work, reflecting their continuing doubts regarding the benefits that might result. 3.36 Invitations to submit proposals were again forwarded in April of 1973 to Consultant 0, Consultant P, Consultant Q and Consultant R. Consultant 0 was eventually chosen to carry out the technical assistance component, and a contract for SI. 6.3 million (US$0.7 million) for a 27-month duration was signed on December 7, 1973. Three months had been added to the duration of the services to allow the consultants time to briefly review the maintenance situation and to adjust their proposed work programs. Procedural difficulties developed, however, concerning the tax issue, which delayed finalization of the contract. This interruption once again required additional Bank effort in stressing to the Government the importance of having assistance in solving one of the most crucial deficiencies in highway management in Peru. 3.37 The maintenance study actually began in March 1974. A Bank mission's assessment of the overall situation at the end of three months cited the following major problems: (a) MTC was not obtaining the maximum benefit from the maintenance study nor was it utilizing the full potential of the consultants; (b) the Government had not as yet assigned the MTC counterpart personnel to the consultants; (c) MTC had not coordinated its activities on maintenance planning with that of the consultants; only a month after the maintenance study started, MTC finalized its own report recommending the purchase of maintenance equipment, in the amount of US$100 million, over a period of four years; (d) the consultant had confirmed that the general condition of roads in the country did not reflect the maintenance budget (approxi- mately US$400/km/year) spent through the Direccion de Transibilidad and the Servicio de Equipo Mecanico (SEM); and (e) the economic justification of the maintenance equipment purchase proposed by MTC was not acceptable. 3.38 An inception report, issued in August 1974 by Consultant 0, dis- cussed the scope and contents of the assistance in light of the needs of the highway subsector and Government plans. By this time, some progress had been made. Objectives and policies for highway maintenance were being formulated; a full list of routine and periodic maintenance activities including the efforts required, expressed in labor and equipment needs, had been prepared. The Government had furnished counterpart personnel and was now taking a more active interest in the study. The inventory of 300 km of - 26 - roads in the two pilot areas (Lambayeque and Junin districts) was nearing completion by late September 1974. Two long-term issues became focal points: (a) the division of responsibility and authority, within MTC, between SEM, which owned all the equipment, and the Direccion de Transibilidad, which managed the maintenance organization and personnel; and (b) the desir- ability of establishing an MTC fund to finance periodic equipment renewal. 3.39 Collaboration between MTC and Consultant 0 improved steadily. By late 1975, the newly created Directorate General for Land Transport, within MTC, was supporting the consultants' efforts and displaying interest in the program and its effects on maintenance work. The programs in the pilot areas remained behind schedule, however, for lack of adequate maintenance budgets. This was aggravated by the austerity measures taken by the Government in 1976, which reduced MTC's funds for road maintenance. The restrictions necessitated rescheduling much of the consultants' work and required a contract extension from September 26, 1975 through June 1977. In addition, provision for an extension of the technical assistance, aimed particularly at improving SEM management, had been included in the Lima-Amazon Transport Corridor Project (Loan 1196-PE). 3.40 Based on the work techniques and maintenance management system which had been tested in the two pilot zones, Lambayeque and Junin, the consultants submitted, in July 1977, their final report concerning the improved system of planning, programing and executing road maintenance activities, together with recommendations for a four-year nationwide program (1978-1982). An MTC reviewing committee recommended acceptance of all the consultants' recommenda- tions. 3.41 The major obstacles to efficient implementation of an adequate maintenance program identified in the report were: (a) the lack of necessary equipment and funds for maintenance operations; (b) the lack of a reporting system which would provide MTC in Lima with adequate and timely information; (c) inadequate mechanisms for work planning and control; (d) insufficient time devoted to highway maintenance in the regions, where MTC's supervisory staff carried joint respon- sibility for vehicle registration, traffic accident reports and other traffic operational duties; (e) SEM's inability to supply and maintain highway maintenance equipment adequately while discharging obligations to construc- tion work forces and some other Government agencies; and (f) a poorly conceived SE/MTC interrelation which lacked incentives for SEM to maximize equipment availability. - 27 - 3.42 The recommended four-year maintenance program, consisting of only essential routine maintenance, was estimated to cost US$20.0 million per year. Jointly, an extensive program of contracted periodic maintenance (resealing and resurfacing) and rehabilitation, costing US$8.5 million per year, was recommended. To execute this program, assuming an improvement in SEM's capacity, equipment and spare parts costing US$50 million would be needed, with US$30 million required in the first year of the program. Other program components included workshop improvements (US$6.6 million), a training program (US$0.6 million), and technical assistance to implement the programs (US$1.6 million). 3.43 The main shortcomings of the report were: (a) instead of proposing a gradual building up of maintenance effort, the report recommended that required improvements be achieved within one year - this is unrealistically rapid; (b) the physical resource needs in terms of equipment, spare parts, workshops and tools seem exaggerated since no consideration was taken of overhauling some existing units; and (c) specific Government measures outside of MTC, which would be required to implement the program in Peru, were not identified. 3.44 Subsequent to the report, the Bank mission discussed, and MTC agreed to, the following: (a) further analysis by MTC of the current budgets for regional maintenance organization and SEM to establish the direct and indirect costs attributable to road maintenance and the level of additional funds required for the 1978 and subsequent budgets; (b) MTC to formulate a plan of action for highway maintenance along the lines recommended, which would be based on a realistic, but somewhat reduced, budget that MTC alone, without outside assistance, would be able to implement; and (c) MTC to incorporate the required administrative and organizational measures needed to implement the four-year program, with appro- priate target dates, in its plan of action. F. Performance of Consultants 3.45 Difficulties with consulting services contributed substantially to the delay in project completion. The following is an overview of the performance of the consultants. (i) Consultant Group A 3.46 The work of Consultant Group A on the feasibility study was satis- factory except for the lack of response to the request for an analysis of the economic and financial effects of selective reduction of highway standards. However, the performance of Consultant Group A, and later of Consultant Group B, - 28 - in engineering design was rather unsatisfactory. In fairness to the consul- tants, the pressures from MTC engineering staff to conform to Peruvian road standards should also be noted. A part of the problem must be laid at the door of the Government policymakers who concurred with the Bank's position regarding reduced costs but failed to have the responsible technical divisions properly instructed. 3.47 The Consultant Group A contracts, and later those of Consultant Groups B and C, all suffered in part from the management problem that arose with the transfer of authority from MTC to ORDEZA. In addition, the lack of clarity in the Consultant Group B design terms of reference concerning the consultants' responsibility for staking out their design was in part a manage- ment failure, but it also indicated a general unfamiliarity regarding the limitations of photogrammetric mapping for road design in mountainous terrain. In particular, the eventual difference in opinion on this matter should have been evident in the consultants' staffing proposal, which should have indicated any staffing requirements for right-of-way surveys and stakeouts prior to the start of construction. 3.48 For works supervision, Consultant Group C performed more adequately, and control of the works and the contractors' programing was good. The consultants reduced the number of supervision personnel to help minimize costs when the construction contracts were extended. (ii) Consultant Group D 3.49 The consultants performed well in both the redesign and the super- vision of the four contracts on the Pativilca-Catac access road. Technical field work and quality control were generally good, and the contract was helped by the consultants exercising absolute supervisory responsibility since there were no MTC or ORDEZA field officers close to the road. However, allowing Contractor V to prime 32 km of base and allowing traffic to use it for several months without completing the double surface treatment can be rated as a technical deficiency. The consultants should have taken the initiative to intervene with the contractor without prompting from the Bank or ORDEZA. (iii) Consultant 0 3.50 Given the reluctance of MTC to start the maintenance technical assistance services, the consultants performed well in gaining the client's confidence and promoting improvements in road maintenance management. IV. COSTS, DISBURSEMENTS AND FINANCING A. Costs 4.01 The final costs for the project components, compared to those of the appraisal estimates, are as follows: - 29 - Appraisal Final Difference Description Estimate Cost US$ (US$ (US$ millions) millions) I. Studies and Technical Assistance (a) Feasibility study and detailed engineering 2.25 2.10 -0.15 -6.70 (b) Technical assistance for maintenance 1.50 1.10 -0.40 -26.70 Subtotal: 3.75 3.20 -0.55 -14.70 II. Road Reconstruction (including supervision) 31.50 39.40 7.90 25.00 III. Unallocated 1.50 - - - IV. Interest Charges 5.00 5.00 0 0 Total: 41.75 47.60 6.85 14.00 4.02 Road construction and supervision covered 83% of the final cost. At initial contract prices, the cost of this component increased by 44%. The cost of each road section at initial contract prices is as follows: Construction Contract Value Final Value Difference (S/. millions) (S/. millions) (millions) Anta-Huaraz 79.7 68.4 11.3 - 14.2 Caraz-Anta 124.9 122.6 - 2.3 - 1.8 Huaraz-Catac 177.5 177.3 0.2 - Catac-Conococha 174.3 224.1 49.8 28.6 Conococha-Colca 177.1 285.7 106.8 61.3 Colca-Chasquitambo 129.8 135.3 5.5 4.0 Chasquitambo-Pativilca 130.8 108.6 - 22.2 - 17.0 Catac Bridge 5.6 4.1 - 1.5 - 27.0 Subtotal I 999.7 1,126.1 147.6 12.6 Supervision Anta-Huaraz 12.9 9.4 - 3.5 - 27.0 Caraz-Catac 32.3 37.1 4.8 1.5 Catac-Pativilca 112.7 114.4 3.2 1.5 Subtotal II 157.9 160.9 4.5 2.0 Total 1,157.6 1,287.0 130.6 11.2 - 30 - 4.03 The final cost of each road section with price escalation added, as compared with contract value, is as follows: Construction Contract Value Final Cost Increase S/. US$ equiv. S/. US$ equiv. S/. % --------------(millions)---------------- Anta-Huaraz 79.7 2.0 109.4 2.8 29.8 40.0 Caraz-Anta 124.9 3.2 155.2 3.4 30.3 6.2 Huaraz-Catac 177.5 4.6 224.2 4.9 46.7 6.5 Catac-Conococha 174.3 4.5 322.9 6.9 148.6 53.0 Conococha-Colca 177.1 4.6 413.1 8.8 235.9 91.0 Colca-Chasquitambo 129.8 3.4 166.5 3.6 37.0 5.8 Chasquitambo-Pativilca 130.8 3.4 153.6 3.3 22.7 0.3 Catac Bridge 5.6 - 5.4 0.1 0.2 - Subtotal 999.7 25.8 1,632.4 33.8 551.2 31.0 Supervision Anta-Huaraz 12.9 0.3 14.1 0.4 1.2 33.3 Caraz-Catac 32.3 1.1 65.1 1.5 24.1 36.3 Catac-Pativilca 112.7 2.9 146.5 3.7 33.7 27.5 Subtotal 157.9 4.3 225.8 5.6 59.0 30.0 Total 1,157.6 30.1 1,858.2 39.4 610.2 31.0 Comparing the original and final costs, there was an overall increase of 31% in US$ terms. The reasons for the increases were the same that caused the delays, as described in paragraph 3.11 and further elaborated throughout the report, i.e., cost overruns with respect to the base contract totals have been caused, in varying degrees, by one or more of the following three factors, to which should be added the impact of inflation during the long construction period: (a) job quantities in excess of final design estimates; (b) design modifications requiring additional quantities or more expensive solutions; and (c) price readjustment payments from delayed starting and extended construction periods. B. Disbursements 4.04 Because of delays in contracting and execution, disbursements lagged behind appraisal estimates as follows: - 31 - % of Appraisal Actual Appraisal Fiscal Year Estimate Disbursement Estimate ------------- (US$ thousands)------------- 1971 1,650 - 0 1972 8,250 620 7 1973 14,850 2,100 14 1974 21,150 6,900 32 1975 28,100 15,500 55 1976 30,000 23,310 77 1977 - 29,800 99 1978 - 30,000 100 Closing Date 9/30/75 12/31/78 /1 /1 The original closing date was extended to December 1975 to permit completion of civil works and technical assistance. Outstanding disbursement requests extended the date to June 30, 1978. At the request of the head of the ORDENOR- CENTRO (ex ORDEZA), the date was further extended to allow the remaining balance to be used for maintenance-related activities on the highways financed under this loan. 4.05 The allocation of the proceeds of the loan was revised as follows: Loan Final Category Agreement Revised /1 Disbursement ----------- (US$ thousands) ---------- I. Feasibility and Detailed Engineering 1,800 1,800 2,002 II. Construction and Supervision 21,000 22,942 22,729 III. Technical Assistance 1,200 1,200 893 IV. Interest Charges 5,000 4,058 4,375 V. Unallocated 1,000 - - Total 30,000 30,000 /1 Revised as of September 19, 1975 upon cable request of September 13, 1975. C. Financing 4.06 The Government financed approximately 20% (S/. 19,387 million) of the feasibility study, detailed engineering and technical assistance for maintenance components. Government participation in civil works construction and supervision came to slightly over 40% (S/. 776,674 million, Table 1). - 32 - V. ECONOMIC REEVALUATION 5.01 An economic justification of the project was not included in the appraisal report. Because the earthquake emergency required rapid processing of the loan (para 2.01), the decision regarding which access road to the earthquake area would be reconstructed was to be based on a subsequent study conducted as part of the project. This study, carried out by Consultant Group A and completed in October 1971 (paras 3.03-3.07), based the selection of the road to be reconstructed and its economic feasibility on the assessment of quantifiable benefits to road users. These benefits were basically derived from lower vehicle operating costs on the reconstructed and improved roads and from reductions in travel times. The study concluded that the selected access road Pativilca-Catac and the road along the Callejon Valley Catac-Caraz yielded an estimated ER ranging between 15.2% and 24.1% (for the various road sections), with the lower returns found on the access road and the higher on the inter-valley sections. The estimate for the aggregate ER for the whole project was 17.8% (Table 5). 5.02 On completion of the project, a comparison of the actual final costs of each of the road sections and the traffic counts for the first year of full operation (1977) of the reconstructed roads with those estimated by the feasi- bility study (Tables 1, 4 and 5) showed that the study estimate of the ER was slightly surpassed for the valley sections (Catac-Caraz) of the road, but was overestimated for the valley access road (Pativilca-Catac). For the project as a whole, the study's ER was some 20% higher than at completion, although the ER at completion, estimated at 14.6%, still amply justifies the investment. 5.03 Several factors account for the variations in the ER and, in parti- cular, the lower than expected return for the project as a whole. Available traffic counts for the opening year (1977) show that, for most roads, actual traffic was significantly lower than that forecast by the feasibility study (Table 4). Traffic was particularly lower than estimated on the access road to the valley (between 30% and 60% lower than estimated) and even lower in the valley sections. On the other hand, no marked changes in the composition of traffic occurred, heavy commercial vehicles remaining at the study levels of about 60% on the access road and 40% on the valley sections. The only exception was the Huaraz-Catac section, where light vehicle traffic developed rapidly because of the section's central location in the most populated area of the valley. This traffic reached the study forecast levels (500 AADT), reducing the comparative participation of heavy traffic to 25%. 5.04 The lower than estimated traffic volumes observed in the first year of operation can be explained partly by the overwhelming problem that confronted the area of influence of the road, i.e., the earthquake, that brought the project itself into existence. The good policy decision made by the Bank to approach the emergency from its long-term needs, namely the reconstruction of the area's transport backbone, implied that a long economic recovery period was expected. The slow recovery of the area, where a sizable portion of the urban population and urban infrastructure was annihilated, was later to be compounded by the critical economic and financial situation of - 3, - the country as a whole. Seen in this light, the stuuy's traffic projections appear optimistic since they were based on the expectation of rapid recovery of pre-earthquake traffic levels--a presumption that, at .he time, appeared confirmed by the relatively high traffic observed in the months subsequent to the earthquake. This traffic, not sustained in the following years, is explained by the intensive emergency activities in the area, mostly generated from the outside. 5.05 The economic reevaluation is based on the later available traffic counts, corresponding to the first year in which the road was in full oper- ation (1977). It is expected (para 5.04) that the reconstructed road would significantly help the steady economic recovery of the region, in turn gener- ating and developing additional traffic. For this reason, the traffic growth rates used by the study (8% p.a.) have been deemed appropriate and, although applied over a much lower traffic base, were used in the economic reevaluation. Recent data on vehicle operating costs were compared with the study estimates, which were found appropriate. They were, after necessary updating, retained for the economic reevaluation. 5.06 The costs of civil works used in the consultants' study were updated to 1976 prices in order to express the reevaluation costs and benefits at a constant base (Table 5). A comparison of the actual completion costs and the study estimates shows that, with the exception of one road section (Conococha- Catac), the original estimates of the consultants were higher than actual results at constant prices. This explains why, in spite of important over- estimation of traffic in the study, actual economic returns do not sensibly differ from original estimates.. The high costs of civil works used in the study's economic feasibility calculations prompted Bank requests for substan- tial design and standards revisions (paras 3.07 and 3.08) that resulted in lower civil works actual costs. However, no confirmation of economic results based on the lowered costs was carried out. 5.07 The only section in which important differences in the costs and economic returns are observed is the 41.5-km Conococha-Catac section. This segment of the road was plagued with implementation problems and was the last to be completed. Also, technical problems arose, mainly due to the nature of the terrain and climatic conditions that required substantial design changes and a more expensive asphaltic concrete pavement (paras 3.16, 3.30). Actual costs resulted in some 100% higher than original estimates used in the economic feasibility study (at constant prices), bringing the ER for this section to a low 6.5% (in comparison with the study's 17.2%). This is the only section of the project that seems to have a less than marginal justifica- tion. However, this section, traversing the most difficult terrain at 4,100 m altitude, cannot be taken in isolation since it constitutes an essential link in the access to the Callejon valley, carrying only long-distance traffic. 5.08 In spite of lower than expected initial traffic, the revisions in the road standards and a careful redesign at the Bank's request resulted in a 14.6% ER for all civil works, which is not much lower than the original study estimates of 17.8%, thus confirming that the project was a worthwhile economic investment. Because of the gradual nature of traffic generation expected in the earthquake area and considering the project's recent comple- tion, this ER could be considered a conservative estimate, and the full beneficial impact may well be understated. - 34 - VI. INSTITUTIONAL PERFORMANCE OF THE BORROWER A. Project Management 6.01 Initially, overall responsibility for the entire project had rested with MTC. ORDEZA, in late 1973, became the executing Government agency for the project, except for the technical assistance component for highway mainten- ance, which remained directly under MTC. The transition, however, took time and was marked by a number of problems. Not until 1976 was the full transfer of authority really effective. 6.02 ORDEZA was created by the Government to coordinate and manage all relief operations in the earthquake-affected area, and it became the admin- istrative model for the new decentralization policy of the Government, enjoying administrative and budgetary autonomy. Supervising and monitoring of construc- tion were carried out from ORDEZA's local office in Huaraz. MTC's Regional Director assigned to Huaraz was seconded to ORDEZA as Director General for transportation and communications, and, as such, was responsible for project management in Huaraz. Therefore, field project management responsibility was vested in one man; however, he was responsible to his superiors in both MTC and ORDEZA. 6.03 Decentralization of project management brought about contention in coordination between MTC and ORDEZA. Up until 1974, the lack of effective liaison between MTC and ORDEZA had produced a confusing and sometimes con- flicting arrangement of responsibilities and accounted for slow Government decisions. During a Bank mission in December 1973, it was agreed by the Bank and officials of MTC and ORDEZA that the Bank would address ORDEZA directly on all matters concerning the Callejon de Huaylas roads, the main component of the project, whereas, for the technical assistance in road maintenance, the Bank would continue communicating exclusively with MTC. To avoid further misunderstandings on the Government side, it was also recommended that, with regard to the correspondence and the visiting supervision missions on the roads component, contact should be established with other ORDEZA officials in addition to the Regional Director, preferably with the Director Superior or the Chief of ORDEZA. These measures helped to reduce the communications problem among those involved in the project. 6.04 The decentralization resulted initially in an agency with very limited administrative and technical capacity, reliant on MTC for manage- rial assistance while building up its own capacity. Administrative support to ORDEZA was late in coming and contributed to a lack of coordination and to little progress. 6.05 Organizational problems within ORDEZA also contributed to the slow pace. Lack of continuity and changes in key personnel disrupted project management. ORDEZA was under the management of three different chiefs during - 35 - 1974-1975, each with a new deputy. Appointment of Regional Directors by MTC to ORDEZA offices in Huaraz was also exceedingly slow. After the first Regional Director left in 1974, it took until 1975 to appoint an acting Regional Director, but the problem was not fully solved until a permanent Regional Director was assigned in April 1976. The appointment of a new Chief to ORDEZA in September 1975 and the assignment of the new Regional Director finally resulted in improved management of the project. 6.06 Problems in management and its organization adversely affected progress from the start of the project, when the preparation, negotiation and signing of contracts were slower than expected. In the early years, this reflected an extra-cautious attitude of Government officials toward the selection of consultants and contractors and toward the Government's relation- ship with the Bank; this seemed natural after the poor experiences of the new Government with practically all foreign consultants and contractors. Once this state of apprehension disappeared, however, delays in contracting, as well as in paying consultants and contractors, continued. 6.07 MTC's management of the technical assistance component for road maintenance similarly delayed its start and initial phases (paras 3.35 and 3.36) but gradually became more cooperative and efficient (para 3.39). However, in spite of the constraints imposed by shortages of funds for actual maintenance in the region, the end result in the pilot regions was disappointingly limited to very few departments, leaving the implementation of the consultants' recommendations to only a group of three MTC engineers. Subsequent to the consultants' departure, this group, in one year, has managed to implement the recommended procedural and management systems in all the departments of one out of the seven regions in the country, and is planning to extend this work to the other regions over the next four years. 6.08 The following external factors relating to financial management also had a fundamental bearing on project implementation: (a) a continual delay by the Ministry of Finance to allocate funds for the cost overrun of the project for civil works and super- vision contracts, with the effect of delays in ORDEZA's authorization to proceed with the work; (b) price variation formulas were not included in the civil works contracts which were bid in 1973 and were, in fact, specifically excluded with respect to readjustment for the increase in cost of equipment; (c) financing the readjustment payments for price increases created budgetary problems for the Government, and payments were delayed by up to five months. B. Compliance with Loan Covenants 6.09 In general, the Borrower complied with the fiscal and procurement covenants of the loan agreement. Non-compliance with procurement covenants (Section 3.04(b)) occurred in the following instances: - 36 - (a) ORDEZA did not receive the Bank's prior approval to the terms of the new contract for the detailed design for the Catac-Pativilca road, although agreement had been reached on recontracting and on the selection of Consultant Group D as the consultant; (b) ORDEZA did not consult with the Bank on including for construction a short bypass to the town of Catac, involving a new bridge. 6.10 Agreement on design standards (Section 3.06) was reached between the Bank and the Government, but the technical levels within the Ministry, dealing with the consultants, resisted variations from national standards, contrary to the agreements reached at higher levels. Eventually, the Bank's viewpoint prevailed, but only after redesign of a number of "high cost" stretches of road. 6.11 In general, the Borrower failed to comply with covenants concerning road maintenance and operations. In particular: (a) the National and Departmental roads in the pilot zones (Section 4.03(a)) were not maintained and repaired to desired standards because of lack of funds and the equipment needed to effect timely repair works; (b) the gradual increases in maintenance budget (Section 4.03(b)) that would, in five years, be sufficient to meet the maintenance needs of the Peruvian highway system adequately, did not take place. The maintenance budget was almost doubled in 1972, but little further real increase took place until 1978 when the effectiveness was undercut by abrupt devaluation of the Sole and rapid inflation; (c) although the Bank financed (Loan 300-PE) the installation of five permanent weighing stations and the procurement of portable ones, all of which have been operating, the coverage is insufficient, and no effective vehicle and axle-load weight control (Section 4.04(a)) is being implemented on much of the roadnet. 6.12 The Borrower was never able to furnish the Bank accurate financial information (Section 5.02(a)) concerning how the Regions and Department actually expended their funds (on maintenance, construction, operations, etc.) in spite of a number of requests by Bank missions. VII. THE ROLE OF THE BANK 7.01 The structure of Bank lending limits the nature of its participation in an emergency situation such as the Peru earthquake of 1970. Within the constraints imposed by this situation, the Bank's early commitment of funds provided the Peruvian Government with a firm planning parameter despite the lack of precise location, engineering and cost information. The project that was eventually selected in the disaster area served a mid- to long-term need and proved to be economically sound, while the road maintenance study provided the basis for a future project which could emphasize the elimination of critical deficiencies in road maintenance. The overall concept of the loan was therefore sound. - 37 - 7.02 There is no evidence that the structure of the loan, with feasibility studies, route selection and construction all included, hastened completion of the project. In at least one respect, it had a negative effect. The passing along, to the final design phase, of the issue concerning selectively reducing design standards to reduce costs was eventually counterproductive since it resulted in redesign and attendant delays. These delays might have been avoided had the issue been settled after a "first loan," during negotia- tions for a second "construction loan." 7.03 The insistence by the Bank missions that these economies be studied and incorporated into the designs is fully justified in light of the relatively low traffic volumes using the road today. Had the pressure by the Bank not been kept up and had the supervision by the Bank missions not detected cases of noncompliance, project costs would have been substantially higher, with no significant added benefits. 7.04 Closer attention to the consultants' proposal for design services would have signaled the difference in concept regarding the consultants' responsibility for physical stakeout of road alignment. In the future, this responsibility should always be clearly and explicitly stated in design terms of reference. 7.05 The difficulties experienced by the contractors in importation of equipment and spare parts should be avoidable since the foreign exchange is financed by the Bank. Insofar as these difficulties are procedural, constant Bank pressure on the competent and responsible Government authorities is needed to have procedures streamlined. 7.06 The real capacity of the local contractors was generally overesti- mated throughout the project by the contractors themselves, the Government and the Bank. In the future, the Government should be encouraged to work with smaller contracts, or allow longer contract periods, until the local contrac- tors can demonstrate convincingly that they have the financial capacity and equipment to execute this work on schedule. The Bank mission could do little about delays in contract execution; however, more emphasis in reporting measures to be taken by the Government to overcome problems (in monthly reports) might be a means of applying pressure on higher management levels to act on these matters. 7.07 The Bank was very much involved in generating genuine interest on the part of the Government in road maintenance, an interest that has been sustained at the level of MTC, but which may still require considerable support at the regional level where the work actually has to be executed. The need for considerable improvement within MTC's mechanical service (SEM) was recognized and included as a covenant in the Lima-Amazon Transport Corridor Project (Loan 1196-PE) where provision was made for related technical assist- ance. Available equipment in sound operating condition is a prerequisite resource for improved maintenance. 7.08 The importance of timely anticipation of changes in the percentage of Bank financial participation is underlined by the problems caused by inade- quate budgets for local counterpart requirements in 1975/76. In Peru, this - 38 - would require more up-to-date reporting of project trends (financial and physical progress) since Government budgets, which are prepared far in advance, are affected. The type of monthly report required internally by MTC presently contains much detail of little immediate interest to the Bank, which delays report preparation. This detail could well be confined to the quarterly reports only, while simplified and speedier monthly reports should focus only on crucial matters wherein Bank involvement might be fruitful. This simpler reporting should also be of benefit to MTC's top management. 7.09 The shift of responsibility for the project from MTC's central (sectoral) authority to ORDEZA (integrated) regional authority created project management problems but eventually resulted in a regional organization that was considered most successful in Peru and is being perpetuated as ORDENOR- CENTRO. If this pattern becomes standard, it would open the possibility to Bank lending on an integrated regional basis rather than by sectors. Certainly, in an emergency situation, it would facilitate a wider umbrella of assistance even though the goals would remain mid-term rather than immediate. VIII. CONCLUSIONS 8.01 The Bank loan helped to finance a project that provides satisfactory road service to and through the earthquake disaster area. It was completed seven years after the calamity. The new road was built to a standard markedly superior to that existing before the earthquake, thus removing a previous development constraint. The loan also helped to finance technical assistance to road maintenance, which resulted in the proposed plan for overcoming the more critical deficiencies in the most critical area of Peru's transport sector. 8.02 Implementation took two years longer than was estimated. In hind- sight, at least half of the delays might have been avoided by more alert and decisive project management by the Government, but the shift in project responsibility to a new organization in the process of being established decreased the possibility of this taking place. Substantial delay can also be traced to technical problems, both in design and construction. 8.03 Bank lending is poorly structured to cope with calamity work. In emergency situations, medium to long term objectives should be identified for Bank assistance. An integrated regional approach might constitute a more fruitful, but more complex, lending program in these situations, and Bank technical assistance in developing the institution to manage such a program might profitably be included. 8.04 Including detailed engineering and construction under a single loan did not accelerate completion of the project. It could have raised financial constraints and did, in fact, contribute to design modifications being deferred until the construction phase, leading, eventually, to project delays. Sequential lending would probably have avoided this problem but would have failed to define the total Bank participation to the Borrower during the immediate aftermath of the crisis. - 39 - 8.05 In Peru, in particular, three subjects should be given special attention in subsequent Bank lending operations to the transport sector: (a) the design specifications and terms of reference should be care- fully reviewed when the use of aerial photography is proposed; (b) contract lot sizes should be reduced to conform better to local contractor capacity; and (c) the Bank should take the necessary measures to ensure easy access to spare parts and equipment by Peruvian contractors engaged in Bank-financed projects. PROJECT C(IMPLETION REPORT PERU - LAN 706-PE ROAD RECONSTRUCTION PROJECT Project Data Appraisal Cost gatiisates Contract Data Contract Cost Dat. (St- nilloe Fi.-.Lna Duratimn ProvEs.- U4s os Final11 (US$n Contnactor/ Dana tFial Contract, Ejui- Final US$ Owerr- h-k Compnent ke Loral Foreign Total Consultant Signed Cotat!Rcpin Reception Value w4lent Cost E aisint*..li 1 % Overrrun overment a/. 00$ Equav 1. Feasibility Study and Detailed Engineering 450 1,800 2,250 Cons. Croup A 2/19/71 10/71 - - - - 10173 - - - - 81,451 2.1 61,451 2.1 17,266 64,184 2.0 (see pars. 3.17) II. Civil Works Huaraz-Anta 21.0 Contractor Z 5/ 2/72 18 Months 8122/74 2/27/75 79,668 2.0 109,478 2.8 29,610 40.0 39,266 70,481 1.8 Ants -trse 46.1 Contracto- Y 8127/73 24 1/05/77 7/05/77 124,902 3.2 155,296 1.4 30,394 6.2 61,303 95,066 2.1 Huaraz-Catac 34.4 No Contractor X 8/27/73 24 1/20/77 7/20/77 177,511 4.6 224,242 f 9 46,731 6.5 91,866 132,375 3.0 C.tar-C.nn-nhs 41.5 Apprainal Contractor W 5/20/74 24 1/24/77 2/08/78 174,347 6.5 322,951 6.9 148,604 03.0 164,868 158,082 3.4 Conococha-Colca 40.2 E.tieten Contractor C 5/13174 22 2/24/77 O/Z4177 177,145 4.6 413,111 8.8 235,966 93.0 175,609 235,032 5.1 Colca-Chasquitambo 33.1 Contractar 5/13/74 22 2/24/77 8/24/77 129,826 3.4 166,875 3.6 37,049 5.8 72,848 97,566 2.1 Chasquitmbo-Pativilca 48.8 Contractor Z 5/13/74 22 9/03/77 9/03/77 130,855 3.4 153,609 3.3 22,754 0.3 68,530 85,006 1.9 (Catac Bridge) (35.0 Cnntractor X 10/30/76 135 Day. 9/ /77 5,650 - 5.484 0.1 166 Sub-Total I 26.4 1,081.35 27.8 1,632,497 3 5511 5___4 - 5.9 ~ , 28.0 696,040 937.772 21.4 Construction Supervision H.ara-Anta Coes. Group A 52,949 .3 14,188 0.4 1,239 33.3 4,065 10,122 0.3 Car..-C.t.e Cons. Group B /9/7 24 -sethe 32,303-/ 1.1 65,106 1.5 32,803 36.3 25,926 39,179 5.0 Ctos-ttivl- Cosultent 6/21/7 74 ionths Con c Dt Ct 30 rc 33C79o Da (i F Sul-Total 11 157.987 -32 225....824.. L,. _6 6,3 30.2 97.900 127.922 3.3 TOTAI_ 11,000 21003,0 ,3,4 32,2 t.939.642 91.5 618,979 35.4 81,0 1.129.878 24.7 III. Technical Assistance for Maintenance 300 1,200 1,500 Cesoltrnt 0 12/107/73 27 -th - 6,300 0.7 10,607 1.1 4,300 57.0 2,121 8,486 0.9 DO. Incerest and Cce,ent 5,000 5,00050 Charges p-. 0. Deallocated _____ 1.000 1,000 Grtd r5 7 n7 17 1750 307000 412750 47.6 34.Figures prooided by Cta Ministry 8/ Tr2npor7 2 "0 C/777i/ 7 v1y 7lightly fross th66 provided by the Coet"It27t. The total difference os enly 5 1. 3Z2All contracts required .tn 1o at varying length.6 8tth the cnttract and final cost eincl.ded partial psyet in /reech Pr/o7 5on the 32llo200g -nt:(35trac 0 - 1,368 and final - 1,607. In addition, contract figure 7Pr761d6d by t h G-:rnet1vry,frCe those provided hy the cone9lt9nt as fl,188 Car01-Ca2r1, /. 44,842; Catac-PatIilcaC S. 89,163. 41 Thi. eq.uInset inst in 00$ e"' calculated using exch-nge rates provided by the Ministry of Traneport ad Comnnic,rti.... Soomse: Ministry at Tranaport and Coass,sicenions Csubltalt Group A Pinal and Quarterly DepoII Coseltant Gronp 0 Pinal and Qsuarterly Deports Appraisal neport March 1979