China - Second Shaanxi Provincial Highway Project

Document of The World Bank Report No: 25234 IMPLEMENTATION COMPLETION REPORT (SCL-39860) ONA LOAN IN THE AMOUNT OF US$ 210.0 MILLION TO THE PEOPLE'S REPUBLIC OF CHINA FOR A SECOND SHAANXI PROVINCIAL HIGHWAY PROJECT February 10, 2003 Transport Sector Unit East Asia and Pacific Region  CURRENCY EQUIVALENTS (Exchange Rate Effective December 31, 2002) Currency Unit = yuan (Y) Y1.00 = US$ 0.12082 US$ 1.00 = Y8.2768 FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS AADT - Annual average daily traffic ADB - Asian Development Bank E&M - Electrical and mechanical EAP - Environmental Action Plan EIRR - Economic intemal rate of return FIRR - Financial intemal rate of return FJTH - Famen-Jiangzhuang-Tangyu Highway GOVAI - Gross output value of agriculture and industry ICB - International competitive bidding LTWMH - Lintong-Terra Cotta Warriors Museum Higlhway LXH - Lantian-Xiaoshangyuan Highway MOC - Ministry of Commnunications MTE - Medium truck equivalent NPV - Net present value PEO - Project Execution Office RAP - Resettlement Action Plan RIPA - Roads Improvement for Poverty Alleviation SHGHAB - Shaanxi High-Grade Highway Administration Bureau SPTD - Shaanxi Provincial Transport Department THH - Tongchuan-Huangling Highway VOC - Vehicle operating costs WTE - Weinan-Tongguan Expressway XSH - Xian-Sanyuan Highway Vice President: Jemal-ud-din Kassum, EAPVP Country Manager/Director: Yukon Huang, EACCF Sector Manager/Director: Jitendra N. Bajpai, EASTR Task Team Leader/Task Manager: Yasuhiro Kawabata, EASTR  CHINA Second Shaanxi Provincial Highway Project CONTENTS Page No. 1. Project Data I 2. Principal Performance Ratings I 3. Assessment of Development Objective and Design, and of Quality at Entry 2 4. Achievement of Objective and Outputs 3 5. Major Factors Affecting Implementation and Outcome 9 6. Sustainability 11 7. Bank and Borrower Performance 12 8. Lessons Learned 14 9. Partner Comments 15 10. Additional Information 22 Annex 1. Key Performance Indicators/Log Frame Matrix 23 Annex 2. Project Costs and Financing 25 Annex 3. Economic Costs and Benefits 29 Annex 4. Bank Inputs 51 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 53 Annex 6. Ratings of Bank and Borrower Performance 54 Annex 7. List of Supporting Documents 55 Map 1: Transport System in Shaanxi Province (EBRD27047R) Map 2: High-Grade Highways in the Project (IBRD27048)   Project ID: P003652 | Project Name. Second Shaanxi Provincial Highway Teain Leader: Yasuhiro Kawabata TL Uniit: EASTR ICR Type: Core ICR Report Date: February 10, 2003 1. Project Data Name: Second Shaanxi Provincial Highway Proj. LICITF Number. SCL-39860 Cozntry/Department: CHINA Region: East Asia and Pacific Region Sector/subsector: Roads & highways (84%); Other social services (15%); Sub-national government administration (1%) KEY DATES Original Revised/Actual PCD: 12/13/1994 Effective: 07/16/1996 07/16/1996 Appraisal: 07/10/1995 MTR: Approval: 03/21/1996 Closing: 12/31/2001 12/31/2002 Borrower/lnIplementingAgencv: PEOPLE'S REPUBLI(C OF CHINA/SHAANXI PROVINCE Other Partners: STAFF Current At Appraisal Vice President: Jemal-ud-din Kassum Russell J. Cheetham Countiy Manager: Yukon Huang Nicholas C. Hope Sector Manager: Jitendra N. Bajpai Richard G. Scurfield Team Leader at ICR: Yasuhiro Kawabata Hatim M. Hajj ICR Prima?y Auzthor: Yasuhiro Kawabata 2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely. HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N-Negligible) Outcome: S Sustainability: L Instituitional Development Impact: SU Bank Pemformance: S Bor r ower Performance: S QAG (if available) ICR Quality at Entry: S Project at Risk at Any Time: No  3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: The project had the following objectives: (i) support continuing development of road infrastructure to relieve congestion, improve access to important archeological and tourist sites, facilitate mobility, and increase the efficiency of road transport; (ii) assist in the implementation of policy to alleviate poverty in the poor areas of the province; (iii) promote development of institutional capabilities of Shaanxi Provincial Transport Department (SPTD); and (iv) increase the safety of road transport. The project objectives were clear and were accorded very high priority by the Government to meet its long-term development target. The project objectives were also in line with the Bank's Country Assistance Strategy (CAS) for China presented to the Board on May 10, 1995. The Bank strategy supports continued institutional development and sectoral reforms aimed at modernizing the highway system and its management, as well as investment for high-priority links of the National Trunk Highway System and the provincial road network that feeds into it. 3.2 Revised Objective: Project objectives remained the same throughout project implementation. 3.3 Original Components The following seven components comprised the project: a. construction of the following five high-grade highways: i. Weinan-Tongguan Expressway (WTE) -- a 78.1 kn four-lane expressway between Weinan and Tongguan at the border with Henan Province; ii. Tongchuan-Huangling Highway (THH) -- a 73.6 km four-lane Class I highway between Tongchuan and Huangling and doubling of the existing 18.4 km two-lane Class I highway between Yaoxian and Tongchuan; iii. Lantian-Xiaoshangyuan Highway (LXH) - a 48.3 km two-lane Class II auto-only highway between Lantian and Xiaoshangyuan; iv. Lington-Terra Cotta Warriors Museum Highway (LTWMH) -- a 5.9 km two-lane Class II auto-only highway connecting Lintong and the Terra Cotta Warriors Museum; and v. Famen-Jiangzhuang-Tangyu Highway (FJTH) -- a 34.8 km two-lane Class II auto-only highway between Famen Buddhist Temple and Jiangzhuang, and between Jiangzhuang and Tangyu; b. upgrading of the 32 kmn, Class I Xian-Sanyuan Highway (XSH) by providing full control of access, traffic safety devices, and installation of a telecommunications system; c. construction, rehabilitation, and upgrading of three linking roads with a total length of about 8.2 km to interconnect and feed traffic to the proposed high-grade highways at interchanges; d. improvement and new construction of about 4,100 km of rural roads in a program to help alleviate poverty (RIPA) in poor counties; e. provision of equipment for: -2 -  i. control of construction quality and monitoring of the environment during the construction of the high-grade highways and their linking roads; ii. operation and maintenance of the high-grade highways; iii. maintenance of the Shaanxi provincial highway network (other than the high-grade highways); iv. institutional strengthening of SPTD and related institutions and the traffic police; f. a road safety component to help reduce accidents on Shaanxi's highways; and g. institutional strengthening and training of the staff of SPTD and its related agencies and institutes by enhancing capabilities in the high-grade highway construction supervision and operations; highway investment planning, including a study of alternative locations for the Xian bypass; maintenance of high-grade highways and other roads, including a study of highway maintenance; and implementation of RIPA and highway traffic components of the project. 3.4 Revised Components: Two subprojects under the component for upgrading the 32 km XSH were cancelled. The construction of the Jinhe interchange was delayed when SPTD had difficulty in reaching agreement with the local government authority on the interchange design concept. When it appeared that the interchange work could not be completed by the original closing date, SPTD decided to do this work using local funds. It was agreed during the October 1999 supervision mission that the telecommunications system under XSH would be included in the electrical and mechanical (E&M) systems for the THH. During the September 2002 mission, however, it was agreed that the Bank would not finance the E&M component for the THH because of a delay in procurement and a cost overrun for civil works. 3.5 Quiality at Entry: Quality at entry is rated satisfactory overall. The assessnment was based on: (i) the consistency of project objectives with China's priorities for the sector; (ii) attention given during project preparation to lessons of experience, including the past performance of the implementing agencies; (iii) the importance given to following Bank policy, particularly those related to environmental, resettlement, and road-safety aspects; and (iv) technology transfer and quality assurance. As with most Bank-financed projects of that period, the performance indicators were not explicitly developed (no baseline and benchmark figures established); they were retrofitted in the project later, however, during project implementation. 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: Despite some delay in implementation, the project objeclives were substantially met, and the achievement of the project objectives is assessed as satisfactory. A. Support the continuing development of road infrastructure The physical components -- construction of five high-grade highways and the upgrading of a highway - have been achieved. The congestion along the project-targeted corridors was relieved, and accessibility to important archeological and cultural sites improved greatly. However, the economic return of one of the highways, the THH, is much lower than anticipated at appraisal because the traffic volumes are much lower than expected and the construction costs much higher. Three highways, WTE, LXH and LTWMH, have a lower economic return than expected because of cost overruns. FJTH has achieved the economic return as expected. -3 -  B. Assist in aleviating poverty in poor areas The RIPA component -- improvement and new construction of about 3,900 km of rural roads -- has enhanced the economic development in poor rural areas. The gross output value of agriculture and industry (GOVAI) per capita in the affected counties increased from Y 1,770 in 1992 to Y 2,730 in 2002 (approximately a 35 percent increase). C. Promote development of institutional capabilities The capacity of highway sector institutions has improved. Through overseas training and study tours and participation in studies and supervision during project implementation, the SPTD's staff received satisfactory exposure to new concepts and incentives to leam new practices and technical approaches. The findings of the Xian Bypass Study were useful and were incorporated into the Asian Development Bank (ADB)-funded Xian Urban Development Project. The findings of the highway maintenance study also were found to be useful and helped the province modernize its maintenance management and operation. D. Increase the safety of road transport The road safety component had six subcomponents, including a pilot program of blackspot improvements. The road safety program raised awareness of the importance of road safety and contributed to its improvement. However, much remains to be done so that measures to reduce traffic accidents are implemented in a sustainable manner. 4.2 Outputs by components: A. Construction of five high-grade highways (US$304.0 million, Staff Appraisal Report ISARI; US$364.4 million, Implementation Completion Report [ICRI) This component is rated as satisfactory overall. Field inspection at completion showed that the construction quality of the five high-grade highways is above average. Finishing details, including landscaping and environmental improvement along the routes, received adequate attention. WTE (US$105.3 million, SAR; US$97.8 million, ICR). The WTE opened to traffic on October 1, 1999. The easternmost 6.1 km section connecting Shaanxi and Henan provinces, which was considered critical for achieving the full economic benefits of the project, was completed and opened to traffic on December 19, 2001. The delay in completion was due to the delay in project implementation on the Henan side. Buildings for toll plazas at six interchanges were completed at the same time that civil works were completed (October 1, 1999). However, rest areas and administration offices were completed later, in December 2002. The original designs for telecommunications and toll collection systems for WTE were completed in 1995. SPTD, however, taking into consideration the recent development of technology for these systems, decided to integrate these systems into the provincial high-grade highway network as one comprehensive system, and designs were revised. The installation of telecommunications and tolling systems commenced in October 2002 and will be completed by August 2003. THH (US$149.6 million, SAR; US$196.6 million, ICR). THH included construction of a 73.6 km four-lane highway between Tongchuan and Huangling and the doubling of the existing two-lane highway between Yaoxian and Tongchuan (18.4 km). Civil works for THH commenced in March 1998 and were completed in April 2001, when the highway was opened to traffic. Building works, including four toll stations and two administration offices, also were completed in April 2001. Two parking areas (Zhifong and Yijun) and a service area (Huang ling) were completed in October 2001. The installation of lights and ventilation systems in tunnels was included in the original civil works contracts and was completed when - 4 -  civil works were completed. The installation of telecommunications, tolling, and monitoring systems will commence in June 2003 and will be completed in June 2004. LXH (US$30.4 million, SAR; US$41.9 million, ICR). LXH opened to traffic in June 1999. The building works included the Lantian toll station, Helongkou maintenance office, and Muhuguan tunnel management office; all were completed by June 1999. LTWMH (US$2.3 million, SAR; US$4.1 million, ICR). LTWMH was completed in March 1998 but was not opened to traffic until October 1999. The installation of toll collection and lighting systems, which was not completed until September 1999, delayed the opening of the highway. The telecommunications and tolling systems are covered under the contract for WTE and also commenced in October 2002 to be completed by August 2003. FJTH (US$16.4 million, SAR; US$24.0 million, ICR). The highway was opened to traffic in April 1999. Additional works, including overpass bridges, were completed by end-April 2000. Building works, including two toll stations at both ends of FJTH, commenced in August 1998 and were completed in April 1999. Works for telecommunications and toll collection systems commenced in August 2000 and were completed by April 2001. B. Upgrading of XSH (US$13.7 million, SAR; US$ 12.7 million, ICR) The upgrading of XSH to an access-controlled expressway originally included 26 subprojects. Twenty-four of these were completed by December 1998, and Bank funding for two components (Jinhe interchange and telecommunications) was dropped. Construction of the Jinhe interchange commenced in November 2001 and was completed in November 2002. Works for telecommunications will be carried out together with E&M works for THH. C. Construction, rehabilitation and upgrading of three linking roads (costs included in the main high-grade highway component) Because these roads were short, they were included under the contracts of the related sections of the high-grade highways. These roads were completed at the same time that the main carriageways were completed. D. Construction supervision of high-grade highways (US$103 million, SAR; US$11.5 million, ICR) The construction supervision of the high-g,rade highways was satisfactory. It was carried out by a joint supervision team of local engineers and international consulting engineers. The joint supervision team approach, with international consulting engineering firns playing a catalytic role, proved effective in improving quality and helped the local counterparts appreciate the importance of quality control and assurance. E. RIPA (US$65.1 million, SAR; US$111.9 million, ICR) The RIPA component consisted of the improvement and new construction bf 31 roa(d systems in 32 counties with a total length of 3,919 km. The general assessment is that the improvement of rural roads has significantly improved access to remote areas and expanded growth opportunities in poor areas of the province. The impact of this component was rated as highly satisfactory by local officials, who stressed that the communities served by the roads have seen their living standards improve and access to education and health and social services enhanced. -5 -  F. Provision of equipment (US$15.8 million, SAR; US$13.7 million, ICR) Equipment required for the operation and maintenance of the high-grade highways, construction quality control, highway research, environmental monitoring, maintenance of the existing highway network, institutional strengthening, and equipment for highway safety were procured satisfactorily. All the equipment purchased under the project has been fully utilized. G. Road safety (US$1.3 million, SAR; US$6.2 million, ICR) The five-area approach that the road safety component adopted (institutional set-up, training, study, equipment, and blackspot improvement program) was implemented satisfactorily. The accident analysis study was completed by December 2001. An accident analysis computer program was completed and installed by July 2001. The highway safety audits and implementation report were completed by July 2001. The program is being implemented gradually by the Shaanxi High-Grade Highway Administration Bureau (SHGHAB) and the HAB. The blackspot remedial program, consisting of civil works, signs and markings at 15 locations, was completed by December 2001. The driver training study was completed in 1998 and has been implemented fully. The final report of an independent evaluation of the implementation of the traffic safety component was completed in November 2002, and it is confinned that the road safety component was completed satisfactorily. H. Institutional strengthening and training (US$ 2.4 million, SAR; US$ 2.1 million, ICR) Training. The overseas and domestic training programs were completed satisfactorily. The training program proved to be a very effective tool in transferring technology and management know-how and gave staff exposure to new concepts and modem construction procedures. A total of 3,162 persons received 2,782 person-months of domestic training. One hundred six persons received 106 person-months of training abroad, and 45 persons participated in the study tour abroad (for 30 person-months). The training covered a wide range of subjects related to planning, design, supervision management and operation of highways, environmental monitoring, financial management, and road safety. On-the-job training and technology transfer were provided during the construction period by the foreign supervision engineers. Xian Bypass Study. The study was completed in July 1997. It investigated the various altemative aligmnents and recommended the preferred alternative. The findings of the study served as a basis for the development of the second Xian Ring Road, which is now under construction. Highway Maintenance Study. This study, covering the organization of highway maintenance, mechanization of highway maintenance operations, improved collection and analysis of traffic volume data, and a pavement management system, was completed in February 2001, with assistance from an Australian consulting firm. The study made recommendations for improving maintenance methods and techniques in the province. Some of the recommendations have been implemented and have contributed to the introduction of market mechanisms in highway maintenance. Additional technical studies. At SPTD's request, the following four additional teclnical studies were conducted during project implementation: * Loess embankment treatment by dynamic compaction * Management system for high-grade highway construction projects * Comprehensive treatment for landslides * Stability of high slopes in loess and rock conditions and their treatment measures. All four studies were completed in June 2002. They helped SPTD enhance its technical capabilities and design of highways in mountainous terrain. - 6 -  I. Safeguard policy Environmental protection. Major environnmental impacts of this project include noise and air pollution during construction and operation. Impacts also includecd the alteration of hydrological regimes, soil erosion and its impacts on the local ecology, impacts on cultural relics, and the problems associated with the haulage of construction materials during construction. During construction, mitigating measures were implemented for air, noise, and water pollution and soil erosion. Considerable efforts were made to protect the environment and minimize social disruption and accidents during the haulage of construction materials, as agreed in the Environmental Action Plan (EAP). Because highway construction for the THH, which is in a mountainous area, involved many cutting sections, slope protection and the disposal of excavated soil were of serious concemn for SPTD and the local communities. Four kinds of grasses were chosen carefully through an on-site experiment of 15,000 square meters and were planted along the highway alignment. The huge amount of soil that was excavated was used to create 1,000 mu of land for local farmers' use. No unforeseen adverse environmental problem arose. Monitoring of air and water quality, noise, lead content of soil, and crops and herbage according to the EAP was carried out regularly. The latest monitoring report confinned that the noise levels in the residential areas were lower than the standards required both during the day and at night; air quality did not worsen and that it met standards; and the highways did not cause any significant negative impact on water in rivers along the highways. The protection of cultural heritage was handled well. For example, during the subgrade excavation along WTE, cultural relics of the Old Stone Age were found. The contractor, in accordance with the EAP and local regulations, suspended construction immediately and informed the relevant cultural relics authority of this discovery. Construction was resumed after the authority preserved the relics. Land acquisition and resettlement. The six high-grade highways (including XSH) had to meet land acquisition and resettlement requirements, for which ten external monitoring reports were submitted to the Bank. The final monitoring report states that the land acquired exceeded the estimated figures in the Resettlement Action Plan (RAP) by 9.5 percent, the floor space of demolished houses exceeded estimates by 13 percent, and the total resettlement cost exceeded the estimate by 1 percent. The resettlement assessment undertaken in 2002 found that despite minor issues with reduced compensation payments, the livelihood of all the project-affected people have been r estored or improved. The compensation rates that had been reduced were later adjusted to take account of the Bank's comments. The 152 segments of rural roads, a total of 3,919 kn, are scattered in 249 townships in 32 counties in Shaanxi Province. Although no monitoring was done during implementation, a resettlement assessment was undertaken by an independent monitor in 2002. The results indicated that minor adverse impacts on the project-affected people have been rectified and the benefits of the RIPA program were appreciated by the rural population, including those adversely affected. All resettlement objectives -- restoration or improvement of income, livelihood, and productivity of the affected people -- have been met. All activities in the RAP, including the rehabilitation program, have been completed. The Banks resettlement policies were followed during resettlement implementation and the task team was satisfied with the outcome of the resettlement assessment. The impacts of the high-grade highways and RIPA are summarized in the following table.  Major Items High-Grade Highways RIPA Planned Actual Planned Actual Land acquisition (mu) 14,818 16,229 5,806 4,767 House demolition (sq. m) 59,955 67,531 24,848 23,316 Project-affected people 17,730* 14,226 3,780 3,040 Total resettlement cost (Y million) 394.63 399.65 30.62 26.51 *The number of project-affected people was estimated on the high side on the basis of 1.08 mu per capita before land acquisition in the project area, while the land-holding capacity after land acquisition is now 1.35 mu per capita. 4.3 Net Present Valte/Economic rate of return: The economic evaluation comprised two components: the high-grade highways (six subcomponents, including XSH) and RIPA (152 subcomponents carried out in four phases). The economic evaluation covers all subcomponents as well as the entire project. The detailed economic analysis is in Annex 3. The economic internal rate of return (EIRR) of the high-grade highways is estimated at 13.1 percent, compared to the SAR estimate of 16.0 percent; the net present value (NPV), at a discount rate of 12 percent, is estimated at Y 408.2 million. The recalculated EIRR and NPV are lower than anticipated at appraisal mainly because of substantial increases in project costs and lower traffic for one of the six roads than projected in the SAR. The results are summarized in the following table. EIRR (in %) AND NPV (12%, Y million) OF HGH SAR ICR EIRR NPV EIRR NPV 1. Weinan-Tongguan Expressway 16.7 -- 15.6 465.5 2. Tongchuan - Huangling Highway 16 3 -- 8 7 (515 2) 3. Lantian -Xiaoshangyuan Highway 14.3 -- 12.6 25.9 4. Lintong - Warriors Museum Highway 14 7 -- 13 0 5.2 5. Famensi - Tangyu Highway 12 6 -- 18.4 197 5 6. Xian - Sanyuan Highway -- -- 290 168 1 Total 16.0 -- 13.1 408.2 The EIRR for RIPA is estimated at 16.7 percent compared to the SAR estimate of 15.0 percent, and the NPV, at a discount rate of 12 percent, is estimated to be Y 531.6 million. The EIRR stated in the SAR covered only phase 1 of the program, while the EIRR for the ICR covers all four phases. The results are summarized in the following table. EIRR AND NPV OF RIPA EIRR (in %) NPV (12%, Y million) SAR ICR SAR ICR Phase 1 15.0 16.2 -- 34.3 Phase2 -- 15.1 -- 218.0 Phase 3 -- 24.7 -- 502.8 Phase 4 -- 9.2 (36.4) Total - 16.7 - 531.6 The EIRR for the entire project (high-grade highways and RIPA) is estimated at 14.1 percent, compared with 15.8 percent estimated in the SAR. The NPV is Y 939.8 million. The following tables summarizes the results.  EIRR (in %) and NPV (12%, Y billion) Total Project SAR ICR EIRR NPV EIRR NPV HGH 16.0 -- 13.1 408.2 RIPA 15.0 -- 16.7 531.6 Total 15.8 - 14.1 939.8 4.4 Financial rate of return: The financial evaluation of the project comprises three sections. Section 1 assesses the revenue-earning entity (high-grade highways, tolled roads); Section 2 assesses the non-revenue-earning entity (RIPA, non-tolled roads), and Section 3 assesses the financial capacity of SPTD to sustain highway activities. The detailed financial analysis is in Annex 3. There was no financial evaluation in the SAR. For the revenue-earning entity, the results of the financial evaluation show that the FIRR for the high-grade highways, which constitute about 62 percent of the total project investment, is expected to be 3.7 percent. The NPV (at a 5.0 percent discount rate) would be Y 696.9 million. For the non-revenue-earning entities (i.e., RIPA -- about 38 percent of total project investment), the financial evaluation, given the nature of financial risks on those investments, focuses on the availability of sufficient counterpart funds and the operating expenses. All the indicators show that the financial risks are modest. 4.5 Institutional developnment impact: The overall institutional development impact was positive. A significant transfer of klowledge and technology took place during project implementation. Significant numbers of staff were trained through foreign and domestic studies and tours. Local engineers, with foreign experts, supervised the construction of high-grade highways using the FIDIC (Intemational Federation of Consulting Engineers) contractual approach. This implementation arrangement helped raise the quality of the civil works and enabled the transfer of know-how. Experience with competitive bidcling was enhanced through the use of contracting firms. The findings and recommendations of the highway maintenance study promote the introduction of market mechanisms into highway maintenance operations, primarily through maintenance by contract, in Shaanxi and enhance the mechanization of maintenance operations to improve the quality and efficiency of maintenance work. 5. Major Factors Affecting Implementation and Outcome 5. 1 Factors outside the control of governmentt or implementing agencv: Adverse climate conditions. In 1998, exceptional flooding caused serious damage to RIPA roads in southern Shaanxi. This was especially serious in the five counties of Shang Lou Prelecture. In Dang Fang County, 2,000 mm of rain fell in 6 hours -- a world record. About 42 km of roads partially disappeared, with nothing left of the alignment. Damage to the RIPA road systems was estimated at about Y 166 million. The Bank agreed to allocate additional sums of US$5.3 million in 1999 and US$7.0 million in 2000 for rehabilitation, repairs and reinforcement. Landslide-prone areas. At the design stage, it was foreseen that construction in the mountainous sections for THH would be extremely difficult because of the many long tunnels and long bridges and a viaduct. When construction commenced, the proposed alignment was discovered to traverse two landslide-prone -9-  areas. Thus, substantial changes to the design of protection works were made. The Bank provided no objection to the variations due to geological conditions that could not reasonably have been expected to be foreseen according to FIDIC Clause 12. 5.2 Factors generally suibject to governnzent countrol: Coordination between SPTD and traffic police. Outcomes for the road safety component were mixed. The expected reduction in traffic accidents subsequent to the opening of the two-lane Class 2 highways (FJTH and LXH) did not materialize. However, accidents on the divided four-lane highways (THH and WTE) were reduced by about half. Improving road safety requires coordinated action in engineering, education, and enforcement. The Provincial Communication Departnent can work only on what falls under its responsibility - the engineering aspects. With the guidance of higher levels of goverrunent, good cooperation between the Provincial Communication Departnent and the traffic police, who are in charge of enforcement and education, needs to be established. Coordination between Shaanxi and Henan Provinces. The completion of construction of the Yangjin-Baijiazhui connection in Henan Province, originally scheduled for end-1999, was delayed. Therefore, although WTE opened to traffic on October 1, 1999, the connection with the Henan section was deferred until December 2001. 5.3 Factors generally suibject to implementing agentcy control: Insufficient geotechnical investigation and inadequate design. A huge number of major design changes and variations were made, particularly in the mountainous sections with adverse geological conditions, which contributed to cost overruns. The main reasons for these design changes and variations are: (i) insufficient geotechnical investigations and boring tests at the design stage; (ii) insufficient and inaccurate detailed topographical surveys; and (iii) inadequate designs based on the inaccurate infonnation and data. Delay in implementation of THH. Commencement of the THH project was delayed by eight months because of the late approval of the bid evaluation and the late completion of resettlement activities. In spite of the remarkable volume of earthworks and number of long tunnels, SPTD tried to complete the project by the original target date (December 31, 2000). However, because of landslides during construction, SPTD succeeded in shortening the construction period by only three months. THH opened to traffic on April 29, 2001. Because withdrawals from the loan were not expected to be completed by the original closing date, the closing date was extended by one year to December 31, 2002. Most E&M works deleted from Bank financing. The Bank loan originally covered E&M works for WTE, THH, LXH, and XSH. However, during implementation the Bank and SPTD agreed that the Bank would not finance the E&M work component, except for those for FJTH, for the following reasons: (i) SPTD decided to integrate the telecommunications and tolling system for WTE into one comprehensive system for the provincial high-grade highway network and the timing of procurement did not fit SPTD's time frame; (ii) because the cost overruns for civil works for THH and XSH were substantial, SPTD preferred that the Bank finance the cost overruns rather than E&M works, which could not be completed by the loan closing date; and (iii) SPTD could not afford to spend two years to procure and install the E&M works for LXH. The monitoring system in a tunnel was installed after the civil works were completed. Lower compensation rates. As discussed in section 4.2, lower compensation rates were paid to farmers for acquired land than agreed to in the RAP. The Bank requested SPTD to correct the noncompliance with the RAP. SPTD accepted the Bank's recommendations and the Bank agreed with the revised compensation rates and SPTD's follow-up actions. - 10 -  Two implementing agencies. The Bank recommended early during implementation that SPTD assign one chief engineer to the entire project to preserve uniformity in supervision procedures and enhance the efficiency of supervision services, as well as to ease the management of foreign experts, and speed and quality of reporting to the Bank. However, SHGHAB was in charge of THH and XSH, and Project Execution Office (PEO) was in charge of four other high-grade highways. During the supervision missions, each office separately submitted a summary progress report on its own responsibility, and weak coordination between these two agencies was observed. 5.4 Costs andfintanicing: At appraisal, the total project cost was estimated at US$5:56.5 million (including contingencies of US$90.5 million), of which US$210.0 million equivalent was from the Bank loan. The total actual cost of the project is estimated at US$ 574.0 million. The civil works for the five high-grade highways, including E&M work and buildings, were estimated to cost US$304.0 million at appraisal, whereas the actual cost is approximately US$364.4 million -- an increase of about 20 percent. Because of the substantial increase in the cost of civil works, and the expansion of the RIPA component, the total actual cost of the project increased by 23 percent over the estimated cost at appraisal without contingencies. Substantial cost overruns in civil works for the five high-grade highways were incurrecl because of adverse geological and geographical conditions in mountainous areas. However, it is understood that if more detailed geotechnical investigations (including boring tests for tunnels) and detailed topographical surveys had been conducted, the design changes could have been minimized. Consequently, the expected Bank financing ratio of 37.7 percent at appraisal was reduced to 36.6 percent at completion. 6. Sustainability 6.1 Rationale for sustainability rating: The project is likely to be sustainable when the achievements made under each of the components are considered, and sustainability is enhanced by the continued efforts of SPTD. * The completion of five high-grade highways has supported the development of road infrastructure to relieve congestion, improve access to important archeological and cultural sites, and improve the efficiency of road transport in the province. The quality of the civil works of the five highways was satisfactory; * The RIPA component greatly contributed to alleviating poverty in the poor areas. The criteria, screening, and selection process for the RIPA program will be used by the local authorities as a reference for future programs; * The Highway Maintenance Study has led to a remarkable transfer of technology to local staff. The study provided important input for the introduction of market mechanisms in highway maintenance operations, especially through maintenance by contract; and * Under a pilot program of blackspot improvements, 15 locations were improved. Blackspot improvement will become routine work for the highway authority. The procedures established for identifying and analyzing blackspots, the criteria for selecting improvement measures, and implementation programming will be good for SPTD's future programs. However, efforts must be pursued in pedestrian and driver education and traffic enforcement, which recluire more coordination between PCD and the traffic police. -11 -  6.2 Transition arrangement to regular operations: The five high-grade highways were opened to traffic on the following dates: WTE October 1999 THH April 2001 LXH June 1999 LTWMH October 1999 FJTH April 1999 The operation and maintenance of these highways have been generally satisfactory except for the slower than expected increase in traffic on LXH and THH. Under the Driver Training Study, completed in December 1998, senior- and middle-level driver training courses were developed in cooperation with the Shaanxi Provincial Traffic Training School and the Shaanxi Provincial Skills Training School. These courses have become popular in China and are used in other provinces such as Anhui, Fujian, and Shanxi. More than 10,000 drivers have taken these courses since June 1999. 7. Bank and Borrower Performance Bank 7. 1 Lending: At the identification stage, the task team ensured that the project objectives and scope were consistent with both the government's priorities and the Bank's assistance strategy. In particular, the sector issues addressed under the project were selected from among the priorities identified in a Bank sector strategy report, Highway Development and Management Issues, Options, and Strategies (February 1994). Bank assistance in project preparation was highly satisfactory. The Bank assigned a highly qualified team to the project with an appropriate skill mix, including experts in areas of engineering, environment, resettlement, road safety, highway maintenance, procurement, institutional development, training, and other fields of specialization. The preparation of the project served to transfer teclnical knowledge to the Chinese counterparts. Appraisal was satisfactory. The SAR assessment of the Government's commitment to the project and of the implementing agencies' capacity during negotiations was on the mark. The economic analysis indicated that it would be economically feasible to construct five highways as proposed and to operate them as toll roads, with the exceptions of LTWMH and FJTH, with lower traffic growth. The main risks were identified correctly: (i) a delay in completing the connection between Shaanxi and Henan provinces; (ii) a delay in completing highways as a consequence of design changes; and (iii) nonrealization of traffic projections, with less diversion to the new highways from existing roads. Although a comprehensive monitoring plan for project objectives and implementation was established, indicators to measure these (baseline and benchmark figures) were not developed during project preparation. 7.2 Supervision: During the first four years after project launch in June 1996, the task manager changed four times, and the continuity of staff was not maintained. However, Bank supervision performance during implementation was consistent and satisfactory. The project had a very ambitious scope. Every time the task team supervised the project, it had to recruit consultants to supervise the RIPA project and look after the - 12 -  technical studies and the road safety component. Bank involvement was critical to the technical studies, for which staff provided useful advice in improving their quality. Resettlement and environment staff provided specialized reviews during the critical phases of the project. The task team showed responsiveness and flexibility in approving needed design changes and modifications. During the construction of the high-grade highways, many variation orders were issued. The task team regularly invited a construction expert to participate in missions. It reviewed the proposed design changes and variations and helped the client divide them into three categories -- provisional sum, day work, and variations. 7.3 Overall Bank perjbormance: Overall, the Bank's performance was satisfactory. The supervision team paid adequate attention to project compliance with Bank policies on resettlement and environment. The flexibility of Bank supervision staff helped to enrich the quality of the project by allowing the desired variations. Borrower 7.4 Preparation: The Borrower's preparation was highly satisfactory. Both the central government and the implementing agencies were highly committed to the project from the start. Their commitment was evidenced in their well-documented preparation work for this very complex project. Recognizing the importance of sector reforn, SPTD cooperated well with foreign assistance to develop the RIPA, highway maintenance, and road safety components. 7.5 Government implementation performance: The central government supported the project throughout its implementation. SDPC and the Ministry of Communications (MOC) supported the inclusion of a very challenging RIPA component of about 3,900 km to alleviate poverty in the province. The RIPA component made remarkable achievements, especially considering that the central government was reluctant to borrow from the Bank for a component that did not raise revenue. MOC helped implement capacity building measures by coordinating training activities, particularly the overseas training and study tours. 7.6 Implementing Agency: Overall project management was satisfactory. SPTD swiftly adapted the project to changing circumstances and suggested changes accordingly. THH, one of the five high-grade highways, was implemented by SHGHAB, and the other four highways by PEO. The coordination between the two implementing agencies could have been improved with stronger leadership from the management of SPTD. The most noteworthy achievement of this project was the timely completion of the 3,900-km RIPA component. 7.7 Overall Borrower performance: Overall, the borrower's performance is rated as satisfactory. Both beneficiaries and the central government showed their commitment to the project and followed through to ensure its suiccessful implementation and operation. - 13 -  8. Lessons Learned A. Complexity of project scope The major project components included five high-grade highways (totalling 240 km) in five locations; the upgrading of a 32 km class 1 highway; the improvement and new construction of about 3,900 km rural roads; a road safety program with six subcomponents; and institutional strengthening and training consisting of two technical studies (Xian Bypass Study and Highway Maintenance Study) and a training program. The project tried to address all the key transport sector issues in China. During preparation phase in 1994 and 1995, SPTD was keen to develop the proposed components, but soon it was overwhelmed with other projects funded by ADB, Japan Bank for International Cooperation, Kuwait Fund, and other financial institutions. The agency was overstretched in its capacity to implement various projects and experienced severe shortage of qualified staff. Most of the technical studies and software components were entrusted to outside consulting firms and hence did not help to strengthen the institutional capacity of SPTD. Policy reforms remained focused on specific subjects, and were scaled up for a broader impact. B. Success of RIPA program The RIPA program demonstrated positive contribution towards poverty alleviation and in providing a basic (all-weather) road access for social services and markets. A report prepared by SPTD and Shanglao Prefecture Communications Bureau described the lessons learned from the implementation of the RIPA component and documented its social and economic benefits, as sumunarized below: Social Beneflt 1995 2001 (at completion) Regular bus service to 200 townships 81 % 100 M. Road accessibility to 2,858 villages 72 % 95 '%, Average annual per capita income Y 450 Y 1,066 Rate of timely accessible health service 40 % 85 % Rate of entrance to middle school 45 % 81 "M, Other provincial govemment agencies, however, have refrained from borrowing for RIPA-like components in succeeding Bank projects because no dedicated revenue source for loan repayment can easily be identified, unlike for toll roads. The Bank continues to support road development in low-income areas under future lending and non-lending services , but is should be recognized that a RIPA-type component demands extensive supervision. The supervising of almost four thousand kilometers of rural roads spread over 32 counties was a difficult task. Procurement for these road works was through 218 contracts representing national competitive bidding procedures (33), small works (96), and force account (89). Signed contracts were post-reviewed except for two contracts for each procurement method at the initial stage of implementation. C. Number of design changes and variations The technical complexity of highway construction and operation was underestimated. This component experienced substantial cost overruns. Design changes and variations could have been avoided or minimized if designs had been based on more accurate topographic maps and results of geotechnical investigations and boring tests during the project preparation. Because of cost increases for civil works (as well as the delayed procurement for E&M works), a large share of the loan was used to cover civil works - 14-  components than was planned. As a result, Bank funds intended for the proposed E&M works were diverted to WTE and THH construction.. D. Monitoring indicators During appraisal, a number of monitoring indices for implementation progress and project objectives were established for seven project components. However, these indicators were not suppoited by baseline and/or benchmark numbers. During project implementation, the task team discussed with SPTD the indicators that could be easily measured and sustained, and established a new set of indicators. 9. Partner Comments (a) Borrower/inzplenzenting agency: A. General 1. This report summarizes the implementation experience of the Second Shaanxi Highway Project partly financed by an IBRD Loan. B. Project description 2. Project objectives are to: a. Support continuing development of road infrastructure to relieve congestion, improve access to important archeological and tourist sites, facilitate mobility and increase the efficiency of road transport in the interest of stimulating economic activity in Shaanxi. b. Assist in the implementation of policies to alleviate poverty in the pooIr areas of the province; c. Promote development of institutional capabilities of SPTD, through training and other means; and d. Increase the safety of road transport. 3. The scope of the project includes: a. Construction of five high-grade highways comprising: i. A 78.1 km four-lane expressway between Weinan and Tongguan (WTE); ii. A 73. 6 km four-lane Class I highway between Tongchuan and Huangling (THH) and doubling of the existing two-lane Class I highway between Yaoxian and Tongchuan (18.4 km); iii. A 48.3 km two-lane Class II auto-only highway connecting Lantian and Xiaoshangyuan (LXH); iv. A 5.9 km two-lane Class II auto-only highway coinecting Lintong and the Terra Cotta Warriors Museum (LTWMH); and v. A 34.8 km two-lane Class II auto-only highway between Fameii Buddhist Temple and Jiangzhuang; and between Jiangzlhuang and Tangyu (FJTH). b. Upgrading of Xian-Sanyuan Class I highway (XSH) through providing access control, and installing traffic safety devices, etc.; - 15-  c. Improvement and new construction of about 3,900 km of rural roads to help alleviate poverty (RIPA) in poor counties in Shaanxi; d. Provision of equipment for: i. Control of construction quality and for monitoring of the environment; ii. Operation and maintenance of the high-grade highways; iii. Maintenance of the Shaanxi provincial highway network; and iv. Institutional strengthening of SPTD and related institutions and the traffic police. e. A road safety component to help reduce accidents on Shaanxi's highways; and f. Institutional strengthening, training and two studies (Xian Bypass and highway maintenance studies). C. Achievement of the objectives 4. The project has attained all major objectives through successful implementation of the project components: a. Five high-grade highways with a total length of 259 km and three linking roads with a total length of 8.2 km were satisfactorily completed. The number of domestic tourists who visited the Terra Cotta Warriors Museum increased from 1,738,000 in 1992 to 2,942,000 in 2001, while the intemational tourists increased from 357,000 in 1992 to 1,795,000 in 2001. Thus, the project has basically reached expected targets. The construction according to FIDIC conditions was undertaken in 1985 when the XSH, the first World Bank-financed highway project in Shaanxi was constructed. This construction management has been expanded and adopted for other highway projects in Shaanxi. This became common practice in management of highway construction in Shaanxi. b. The upgrading of the XSH was satisfactorily completed, which resulted in a remarkable decrease in traffic accidents from 29 times in 1995 with the casualty of 23 persons and the economic loss of Y 128,000 to 16 times in 2000 with the casualty of 13 persons and the economic loss of Y 63,000. c. The RIPA program with a total length of 3,919 km has been satisfactorily completed. The infrastructure in the poor area appears to be obviously improved and the social and economic benefits were enhanced. d. The intemational construction supervision services totaled 162.11 person-months and local supervision services totaled 12,432 person-months. By jointly working with foreign experts, local engineers' awareness of construction quality was enhanced. e. Environmental monitoring was undertaken 60 times in total during the construction period (twice a month), and 30 times during the operation period (once a month). The assessment of enviromnental aspects has been rated the project satisfactory. - 16 -  f. The implementation of subprojects under the road safety study has been satisfactorily completed. The outcomes of the subprojects such as driver training, improvement of blackspots, preparation of a road safety manual, development of software dealing with traffic accidents management and analysis contributed greatly to reducing the traffic accidents on the upgraded and newly constructed roads. The software on traffic accidents management and analysis developed became standard in Henan, Xinjiang, and Jiangxi. g. A total of 17 overseas training and study tours courses totaling 131.2 person-months were implemented. Through overseas study/training, the local managerial and technical staff gained experience in expressway planning, design, construction, financing, financial auditing, and maintenance management. h. The procurement of equipment for a central laboratory and highway operation and maintenance for the related agencies was successfully completed. With this equipment, the mechanized maintenance of the expressways has been improved and the capacity of checking and testing was enhanced. i. The study on Xian Bypass was completed by the end of 1997. The outcomes of the study contributed a lot to the route selection of the Xian Bypass and promoted the implementation of the bypass project. j. The study of highway maintenance was successfully completed. The outcomes of the study played a very important role in development of modem highway maintenance methods, the acceleration of the mechanized maintenance, the improved data collection and analysis of traffic volume, development of pavement management system, and the reformn of the highway maintenance system. D. Project formulation and preparation 5. By April 5, 1995, SDPC and Shaanxi Provincial Planning Commission approved the project proposal for the Second Shaanxi Highway Project. By October 28, 1996, SDPC and MOC approved a prefeasibility study for the project. 6. Starting with the Bank's identification mission in October 1994, the World Bank was involved in the preparation of the project. In March/April 1994, the World Bank preparation mission reviewed the feasibility study reports, environmental evaluation reports, and the regional traffic anid econiomic information. 7. In July 1995, the Bank's appraisal mission reviewed and discussed additional documents provided during the preappraisal mission. Experts funded by the Italian Trust Fund were also invited to review the designs and bidding documents. 8. . In January 1996, the loan was negotiated in Washington. The Loan Agreement and the Project Agreement for the Second Shaanxi Highway Project for US$210 million were signed on April 17, 1996. E. Project implementation 9. The loan became effective on July 16, 1996, and the experience gained during the implementation of the project is summarized below by component. - 17 -  10. High-grade highways and linking roads. The construction of total 259 km high-grade highway was implemented using 17 international competitive bidding (ICB) civil works contracts and the construction of three linking roads was included in corresponding contracts. Construction of high-grade highways commenced in March 1997, except that of THH (commenced in March 1998) and was completed by October 1999 except THH (completed in April 2001). During the construction, the employer paid much attention to important nodes and quality control. The employer, together with the designer and foreign and local engineers carried out construction supervision and approved variation orders including the replacement of the embankment material from the loess to the sand gravel; lime earth treatment for the upper course of the sub grade; anchorage works to stabilize side slopes of the THH and LXH; the design change from at-grade intersection to grade separations, additional underpasses, slide treatment works and drainage works to the FJTH. In addition, efforts were made to replace the back-filling materials at abutment of all bridges, culverts and underpasses. 11. XSH. The upgrading works of a total of 32 km of highway was divided into 26 contract sections, of which 24 sections commenced in October 1997 and were completed by the end of 1998. Because of change of design concept, two contracts including an interchange and telecommunications system were not implemented together with other 24 contracts. SPTD decided to finance these contracts by domestic funds. Construction of an interchange started in November 2001 and completed in November 2002. Installation of telecommunication system commenced in August 2002 and will be completed in October 2003. 12. RIPA. The RIPA program consists of 31 road systems with 152 road sections, divided into 218 contracts located in 32 counties. Works started in October 1996 and were completed in October 2000. The total length covered under RIPA was 3,919 km with an investnent of Y 645.96 million. In 1998 and 1999, Shaanxi province was stricken by severe flooding, and some sections of completed roads were washed away. The Bank approved additional financing for rehabilitation of the damaged roads. 13. Equipment procurement The equipment procurement was implemented in two stages: the first stage started in 1996 and was completed in 1998 with total costs of US$8.31 million (of which US$1.44 million was used for imported equipment). The second stage started in December 2000 and was completed by March 2002. The cost of procurement totaled US$4.058 million. The equipment procured was for high-grade highways operation and maintenance. 14. Ventilation and illumination system for tunnels. THH has seven tunnels, and the longest one is 1,700 m. LXH has six tunnels, and the longest one is 1,670 m. Supply and installation of ventilation and illumination system were included in the contracts for civil works, and were completed at the same time that the civil works were completed. 15. E&M work. The original design for telecommunications and the toll system were completed in 1995. However, considering the technical development in this field, SPTD decided to combine the E&M systems for high-grade highways as a complete comprehensive system. Original designs were revised. Because of the delay of revision of design and the bidding procedure, it was expected that the E&M works would not be completed by the loan closing date. SPTD decided in 1999 that the E&M works of the project (excluding FJTH) were locally financed. Installation of telecomununications systems for WTE commenced in October 2002 and will be completed by August 2003. Upon approval by the Bank, the telecommunications and toll system for FJTH was procured through direct shopping. - 18 -  16. Construction supervision. Italconsult provided a total 162.1 person-months input for construction supervision, including 4 person-months of training service. 17. Staff training. SPTD and MOC conducted a total 131.2 person-months of overseas training and study tours with a total cost of US$900,000. The training subjects involved project management, construction supervision, traffic engineering, environrment. protection, highway planning, maintenance of high-grade highway, and traffic safety. SPTD also organized 2,782 person-months of domestic training during the project implementation. The training enriched staff with modem management knowledge on design, construction, operation, and maintenance for high-grade highways. 18. Environment monitoring and training. Environmental protection was a key aspect of the project implementation. The project involved 11 person-months oversees training. The employer entrusted Shaanxi Provincial and Xi'an Municipal Monitoring Stations with conducting quarterly environmental monitoring for each subproject. 19. Land acquisition and resettlement. Land acquisition and resettlement were completed satisfactorily. In accordance with the RAP, external monitoring was executed by the Shaanxi Provincial Academy of Social Sciences. The RAP was followed as planned, and living standards and income of people affected were both improved and increased. Under the six high-grade highways, a total of 16,229 mu land was acquired, affecting directly 2,947 households covering 14,226 people, in which 503 households were removed and 2,143 population was resettled. A total of 39 enterprises were affected (no removal occurred). The total compensation for land acquisition and resettlement reached Y 399.6 million, a Y 5 million increase compared with the originally estimated Y 394.6 million in RAP, or an increase of 1.0 percent. Under the RIPA program, a total of 4,767 mu land was acquired, about 18 percent less than the estimated at appraisal. The number of people affected was 3,040, about 20 percent less than estimated at appraisal. The total costs for land acquisition and resettlement was Y 26.51 million, about Y 4 million less than estimated. F. Operation of high-grade highways and RIPA program 20. Although the traffic volume was not as expected on some highways, the operation of high-grade highways is generally satisfactory. Traffic accidents on the divided highways (WTE and THH) were substantially reduced. a. WTE: WTE was opened to traffic in October 1999. The annual average daily traffic (AADT) in 2000 reached 9,999 medium truck equivalent (MTE), and 10,114 MTE in 2001, and is estimated to be 11,438 MTE in 2002. The toll revenue in 2000 was Y 58.59 million, while that of 2001 stood at Y 58.98 million. b. THH: THH was opened to traffic in April 2001. The AADT in 2001 reached 3,159 MTE, and in 2002 is estimated to reach 4,000 MTE. The toll revenue in 2001 was Y 29.96 million. c. LXH: LXH was opened to traffic in June 1999, with the AADT of 2,231 MTE in 1999. AADT increased to 3,423 MTE in 2000,3,623 MTE in 2001, and 3,393 MTE in 2002. Toll revenue was Y 11.75 million in 1999, 29.46 million in 2000, and 30.61 million in 2001. d. LTWMH: LTWMH was opened to traffic in October 1999. The AADT reached 1,896 MTE in 2000, and 2,394 MTE in 2001 and is estimated at 2,528 MTE in 2002. Toll revenue was Y 27.61 million in 2000, and Y 26.87 million in 2001. - 19-  e. FJTH: FJTH was opened to traffic in April 1999. The AADT reached 2,172 MTE in 1999, 2,620 MTE in 2000, and 2,863 MTE in 2001 and is estimated at 3,168 MTE in 2002. Toll revenue was Y 24.05 million in 2000, and Y 24.94 million in 2001. f XSH: The upgrading works were completed in October 1998. Traffic safety was enhanced and traffic conditions were improved. The AADT reached 14,158 MTE in 1999, 16,871 MTE in 2000, and 17,887 MTE in 2001. The AADT in 2002 is estimated at 18,500 MTE. Toll revenue was Y 69.53 million in 2000 and Y 65.68 million in 2001. 21. The RIPA program significantly contributed to improving access to the remote areas and expanding growth opportunities in poor areas in Shaanxi. The local officials and people confirmed that living standards in communities served by rural roads have been substantially improved and they appreciate the improved access to education, health, and social services. Since these newly improved roads are all-weather, much less maintenance works are required. G. Works by contractors and consultants 22. Civil works. The civil works of five high-grade highway were divided into 17 contracts and procured through ICB procedures. Contractors implemented the construction according to the contracts condition, specifications and the design/drawings. Their works were in general satisfactory. Through efforts of the employer, the engineer and the contractors, all five high-grade highways were completed and opened to traffic on due date, with good quality, meeting requirements in the specifications and the contract agreement. Construction quality of the highways passed the government standards. 23. Consultant supervision by consultants. The international consultant firn, ItalConsult. S.P.A. fielded 7 consultants with 162.11 person-months for construction supervision of the high-grade highways. The consultants together with local engineers paid frequent site visits to ensure the works quality and progress of the project. They also reviewed payment orders and variation orders and prepared supervision reports. The finn also carried out domestic training for 75 local engineers, and overseas training for 30 local engineers. Local supervision teams were recruited through bidding among firms with sound qualifications and good reputations. About 296 local supervision engineers were involved in construction supervision services for five high-grade highways. All these supervision engineers carried out all-round management and control for quality, progress, measurement, and payment based on FIDIC conditions. H. Works by implementation agency 24. Project preparation and implementation arrangement. Taking into account the complexity of the project components and the number of subcomponents involved, SPTD formally formed the Project Execution Office financed by the World Bank under SPTD in January 1996. PEO consisted of several departments such as the engineering management, finance and planning, equipment procurement, land acquisition and resettlement, and comprehensive administration and was responsible for bidding, examination of major variations and claims, land acquisition and resettlement, payment according to the chief engineer's measurement, reporting to the World Bank, and monitoring the works of the contractor. Designs of all highways under the project were made by Shaanxi Highway Survey and Design Institute. During the construction phase, representatives of the design institute were fielded to be responsible for clarifications of drawing and variations. - 20 -  25. Environmental Impact Assessment. Shaanxi Provincial Scientific Institute for Environmental Protection and Xi'an Highway Transportation University conducted an environunental impact assessme,nt for each highway according to the state environmental assessment regulations and envirornental protection method. These regulations and methods required that environmental protection be implemented through design, construction, and operation phases. hI 1994/1995 the PEO submitted three times the proposed environmental impact assessment reports and the environmental action plan (EAP) for all six high-grade highways to the World Bank. In December 1995, the Bank confirmed the EAP. The environmental impact assessment and EAP of RIPA were submitted to the Bank in October 1995 and cleared by the Bank. The Employer paid much attention to environmental protection at the designing phase, in the aspects of alignment selection, road crossing, bridge and culvert, drainage, slope protection, and greening work. At the construction phase, contractors were requested to properly manage environmental protection, particularly on waste yard, drainage works, planting, and greening work. The total greening area reached 1,999,700 square meters. I. Main factors affecting the project 26. Inaccurate geological investigation and higher cost. Because of the poor geological investigation at the design stage, numerous unexpected slope collapses and landslides occurred during the construction of THH, especially at 19 spots, which cost an extra Y 110 million for treatment. 27. Unfavorable weather conditions. Heavy flooding happened in 1998 and 1999 and exerted serious damage to the already built RIPA roads in southern Shaanxi. The disaster brought forth serious damage in five counties, where 42 km of roads washed out completely. Damage resulting from the flood reached Y 166 million. The Bank agreed to finance additional US$5.3 million in 1999 and US$7 million in 2000 for the rehabilitation and reconstruction of works washed out by floods. 28. Inadequate designs and higher cost. LXH was the first high-grade highway through heavy mountainous areas in the province. Because of lack of design experience for mountainous highways, the side slope treatment was not properly designed, so that geological faults often occurred during construction. To ensure the highway safety during the operation period, additional protection works were undertaken, which caused costs. J. Works by the Bank 29. Preparation of the project. From March 1993 to July 1995, the Bank dispatched four field missions to Shaanxi to assist in preparing the proposed project. In February 1996, the Bank approved the project. The appraisal report analyzed the background information and supported the proposed scope of the project. The Bank reviewed the RAP, EAP, designs and bidding documents, procurement plans, and a training program. 30. Project implementation. After the loan agreement became effective in July 1996, the Bank dispatched two missions a year to supervise the project and monitor the progress on each project component. The missions were generally satisfied with the perfonnance of SPTD. 31. Cooperation between the implementing agency and the Bank. Cooperation between the implementing agency and the Bank was well done. The Bank helped the Employer deal withl the variation orders and claims raised. The Bank agreed with the increase of side (cut) slope protection works and reinforcement works, the addition of rehabilitation works under RIPA, and the loan relocation. Under the -21 -  guidance of Bank staff from the Bank headquarters and its Beijing office, the project was completed satisfactory to the Bank. K Planning 32. At present, the Chinese government is implementing the Great Western Development Strategy to expedite development of the mid and western parts of the country. One of the most significant tasks is to expedite especially the construction of infrastructure and highway transportation. SPTD has developed a plan to expand the highway network in the province covering trunk highways, branch highways, and rural roads, and connecting with those in surrounding provinces and regions. By 2010, the highway mileage of the province shall reach 52,000 kIn, in which expressways shall be 3,010 km. The class 1 highway shall be 178 km, the class 2 highway shall be 7,300 km. With the completion of the provincial highway network, the "one-day-driving transportation circle" shall be formed to link Xi'an with Taiyuan, Beijing, Zhengzhou, Wuhan, Chongqing, Chengdu, Lanzhou, Yinchuan, Boatou, and other central cities. Over 70 percent of counties in the province shall be connected with expressways, and each county will be within five hours driving to reach major cities in the province. 33. In 2002, SPTD started to implement the tolling system with one toll card applicable to the whole expressway network in the province. Four highways, including WTE, THH, FJTH, and LTWMH, shall first be equipped with such tolling system in 2003. (b) Cofinanciers: (c) Other partners (NGOs/prihate sector): 10. Additional Information - 22 -  Annex 1. Key Performance Indicators/Log Frame Matrix Outcome i Impact Indicators: Indicator/Matrix Projected in last PSR Actual/Latest Estimate High-Grade Highways: Estimate at appraisal As of December 2002 1. Congeston on existing roads reduoed - average daily traffic (ADT) on high-grade highways: FJTH (34.8 km) 3,000 veh/day 3,000 veh/day LXH (48.3 km) 3,600 3,300 THH (73.6 km) 7,200 3,800 WTE (78.1 km) 8,200 11,000 ADT on existing parallel roads: FJTH corridor 1,600 veh/day 800 veh/day LXH 700 900 THH 3,000 3,000 WTE 3,500 1,500 2. Economic activities promoted: GOVAI Per Capita in affected counties increased: FJTH corridor Y 5,671 Y 5,735 LXH 3,195 3,299 THH 5,217 5,422 WTE 4,487 4,503 3. Road safety improved: Fatalities/injures caused by accidents reduced: FJTH corridor 10 per 10,000 veh. 26 per 10,000 veh LXH 12 16 THH 17 9 WTE 29 12 RIPA: 1. ADT increased: 1 st package 250 veh/day 490 veh/day 2nd package 290 260 3rd package 280 340 4th package 390 310 2. GOVAI per capita in affected counties increased: 1 st package Y 865 Y 993 2nd package 845 974 3rd package 847 923 4th package 855 949 3. Number of closed days reduced: 1st package 7 days 5 days 2nd package 8 8 3rd package 6 9 4th package. 6 | Institutional capacity strengthened: Training program successfully completed: 2,633 2,913 person-months Note: During the implementation of the project, the Bank and SPTD agreed that indicators to measure the achievement of the project objectives would focus more on direct and easily measurable and sustainable ones. This table covers only major indicators. - 23 -  Annex 1. Key Performance Indicators Output Indicators: Indicator/Matrix Projected In last PSR Actual/Latest Estimate Estimate at Appraisal 1. Five high-grade highways completed: FJTH (34.8 km) November 1998 (100%) April 1999 (100%) LTWMH (5.9 km) October 1997 (100%) March 1999 (100%) LXH (48.3 km) December 1998 (100%) June 1999 (100%) 1 HH (73.6 km) December 2000 (100%) Apnl 2001 (100%) WTE (78.1 km) December 1999 (100%) October 1999 (100%) 2. 32 km of Xian-Sanyuan Highway upgraded December 1997 (100%) March 2000 (100%) 3.4,100 km of rural roads improved or newly December 2000 (100%) October 2000 (100%) constructed 4. Pilot blackspot program completed December 2000 (100%) December 2001 (100%) 5. Planned training program completed June 2001 (100%) December 2001 (100%) End of project - 24 -  Annex 2. Project Costs and Financing Project Cost by Component (in US$ million equivalent) Project Cost by Component Appraisal Actual/Latest Percentage of Estimate Estimate Apria US$ million US$ million Appraisal High-grade highways Civil works 282.8 344.6 121.9 E&M supply and installation 8.5 9.3 109.4 Buildings 12.7 10.5 82.7 Xian-Sanyuan highway 13.7 12.7 92.7 RIPA 65.1 111.9 171.9 Highway safety 1.3 6.2 476.9 Subtotal 384.1 495.3 129.0 Supervision 10.3 11.5 111.7 Equipment 15.8 13.7 86.7 Training 1.2 1.4 116.7 Studies 1.2 0.7 58.3 Contingencies 90.5 - Land acquisition 53.4 51.5 96.4 Project Total 556.5 574.0 103.1 Percent of total 100.0 - 25 -  Project Costs by Procurement Arrangements (Appraisal/Estimate) (US$ million equivalent) Procurement Method Total Project Cost by Component ICB NCB Other/a NBF/b Project ________ Cost High-grade highways and linking Civil works 345.5 15.7 361.2 (143.2) (5.7) (148.9) E&M supply and installation 11.0 11.0 (8.8) (8.8) Xian-Sanyuan Highway 4.3 12.2 16.5 (1.9) (5.3) (7.2) RIPA 39.7 39.7 79.4 (14.7) (14.8) (29.5) Traffic accidents (black spots) 1.6 1.6 (0.0) (0.0) Equipment 8.5 4.1 5.1 17.7 (5.9) (3.2) (0.0) (9. 1) Consultants and training/C Supervision of construction 12.9 12.9 (4.7) (4.7) Training 1.4 1.4 (1.1) (1.1) Studies 1.4 1.4 (0.7) (0.7) Land acquisition and resettlement 53.4 53.4 (0.0) (0.0) 365.0 59.7 73.3 58.5 556.5 Total (157.9) (22.3) (29.8) (0.0) (210.0) /a Other includes force account, international shopping, and limited international bidding, and consultants and training. /b NBF means not Bank-financed. /c Selection of consultants according to the Bank's Guidelines for Usc of Consultants. - 26 -  Project Costs by Procurement Arrangements (Actual/Latest Estimate) (US$ million equivalent) Procurement Method Project Cost by Component Total ICB NCB Other/a NBF/b Project Cost High-grade Highways and Linking Civil works 347.9 3.9 3.3 355.1 (151.7) (1.7) (0.0) (153.4) E&M supply and installation 0.9 8.4 9.3 (0.7) (0.7) Xian-Sanyuan Highway 4.7 8.0 12.7 (2.1) (3.5) (5.6) RIPA 35.8 76.1 111.9 (1:3.2) (28.2) (41.4) Traffic accidents (blackspots) 0.0 6.4 6.2 (0.0) (0.0) Equipment 3.2 2.3 8.2 13.7 (3.2) (2.3) (0.0) (5.5) Consultants and training/C Supervision of construction 2.3 9.2 11.5 (2.3) (0.0) (2.3) Training 0.9 0.5 1.4 (0.9) (0.0) (0.9) Studies 0.2 0.5 0.7 (0.2) (0.0) (0.2) Land acquisition and resettlement 51.5 51.5 (0.0) (0.0) 351.1 44.4 90.7 87.8 574.0 Total (t54.9) (170) (38.1) (0.0) (210.0) /a Other include force account, international shopping and limited international bidding, and consultants and training. /b NBF means not Bank-financed. /c Selection of consultants according to the Bank's Guidelines for Usc of Consultants. - 27-  Project Financing by Component (in US$ million equivalent) : ,. -fP 5bl~:Estin~aten. iActuallLatest Estimnate P-iercentag of App Component;.. F *:.,* - ; o A, raisal. ~~en ank,Sfhaanxi "C'e'iral Bank- Shaa'nii :C-entral k High-grade highways Civil works 40.8 118.6 123.4 86.2 106.7 151.7 211.3 90.0 122.9 E&M works 1.7 6.8 8.6 0.7 505.9 10.3 Buildings 7.1 5.6 8.8 1.7 123.9 30.4 Xian-Sanyuan 7.7 6.0 7.1 5.6 92.2 93.3 highway RIPA 15.2 26.0 23.9 39.2 31.3 41.4 257.9 120.4 173.2 Highway safety 1.3 6.2 Subtotal 73.8 144.6 165.7 156.1 138.0 201.0 211.5 139.0 121.3 Supervision 6.6 3.7 9.2 2.3 139.4 62.2 Equipment 7.5 8.3 8.2 5.5 109.3 66.3 Training 0.3 0.9 0.5 0.9 166.7 100.0 Studies 0.6 0.6 0.5 0.2 83.3 33.3 Contingencies 59.7 30.8 Land acquisition 53.4 51.5 96.4 Project Total 201.9 144.6 210.0 226.0 138.1 210.0 111.9 95.5 100 Percent of total 36.3 26.0 37.7 39.4 24.1 36.6 - 28 -  Annex 3. Economic Costs and Benefits 1. The economic evaluation of the project covers the following two components: a. construction of five high-grade highways and upgrading a highway, totaling 291 km in Shaanxi province, and b. construction and/or improvement of 152 rural roads and bridges in the poor counties of Shaanxi, totaling about 3,919 km, under the Road Improvement in Poor Areas (RIPA) program. 2. The economic analysis is based on the re-evaluation of data on traffic, operational performance, costs and benefits of project components. The methodology used was similar to that employed by the SAR and is summarized as follows: a. capital investment and maintenance costs were revised to reflect December 2002 prices and are included in the cost stream; b. the benefits, also reflected in December 2002 prices, consist of savings in vehicle operating costs (VOC), travel time savings, reduced traffic congestion on existing roads, and enhanced road safety; c. a project life of 20 years has been assumed, and the capital investment periocl for all components was from 1995 to 2001, depending on the construction phases; and d. for the costs for rural roads under RIPA, the additional costs required for reconstruction of damaged sections due to severe floods incurred in 1998 and 1999 are included. 3. During project implementation, some project components were modified. The economic analysis reflects the final implementation outcomes of the project. 4. This annex comprises four parts: the economic evaluation of high-grade highways; the economic evaluation of RIPA; the overall economic evaluation of the project, including the probabilistic risk analysis; and the financial evaluation of the project. ECONOMIC EVALUATION OF HIGH-GRADE HIGHWAYS 5. The high-grade highways consist of six roads for economic evaluation. Length (kim) Length (km) Opening Existing Roads HGH Date 1. Weinan - Tongguan Expressway 87.00 78.10 Oct. 1999 2. Tongchuan- Huangling Hwy 97.00 91.97 Apr 2001 3. Lantian -Xiaoshangyuan Hwy 51.20 48.30 Jun. 1999 4. Lintong -Warriors Museum Hwy 6.30 6.02 Oct. 1999 5. Famensi - Tangyu Hwy 40.90 34.80 Apr. 1999 6. Xian - Sanyuan Hwy 32.00 32.00. Dec. 1998 Total 314.40 291.19 -29 -  Corridor Trafflc 6. The first newly constructed highway was opened to traffic in April 1999 and the last one was opened in April 2001. The traffic volume of six highways for 1999-2001 and its distribution among the existing and new highways are shown in Table 1. 7. LXH and FJTH ware opened to traffic in June and April 1999, respectively. The actual corridor traffic of these highways in 1999 was much higher than estimated at appraisal. The diversion rates from the existing roads to the new highways were 49 percent and 67 percent, respectively. The actual corridor traffic increased greatly in 2000 -- 80 percent and 38 percent higher than the estimates. The diversion rates also increased to 63 percent and 73 percent, respectively. Even in 2001, the actual corridor traffic of these highways was still much higher than the estimates, with higher diversion rates as well. 8. WTE and LTWMH were opened to traffic in October 1999. The actual corridor traffic of these highways in 2000 was 69 percent and 182 percent higher than the estimates. However, the diversion rate on WTE was lower than expected (66 percent against 70 percent). The corridor traffic of LTWMH tripled in 2001, but with a lower diversion rate. 9. THH was opened to traffic in April 2001. The actual corridor traffic during the remaining eight months in 2001 was much lower than expected (35 percent lower) with a lower diversion rate as well. The lower corridor traffic is considered to be due to the fact that the inmnediate northem section of THH was under construction. Traffic Projection 10. The increase of traffic on a new expressway, particularly when it is operated as a toll facility, is gradual in the initial operational years and then the diversion accelerates. The average traffic growth rates (during 1999-2002) on six highways were between 7 and 23 percent. Despite the increasing diversion factor ratio and fast traffic growth rates in the initial operational years, the long-term traffic growth rates on the new highways were estimated to be slightly lower than the SAR's forecast in general. The traffic forecast for the six highways is summarized in the following table. - 30 -  TRAFFIC FORECAST COMPARISON FOR NEW HIGHWAYS (in AADT) Weinan - TonaRuan Expresswav Tongchuan - Huaneline Hwv Lantian -Xiaoshanovupn Hwv Total MTE Total MTE Total MTE SAR 1999 - - -- - 2000 - 6,256 - 6,583 -- 2,606 2001 - 6,782 - 6,961 -- 2,810 2002 -- 7,353 - 7,360 -- 3,030 2010 -- 14,036 - 11,498 -- 5,533 2020 - 26,065 - 17,656 -- 9,890 Avcrage growth p.a. 2000-2020 7.4% 5.1% 6.9% ICR 1999 12,667 9,691 - 4,024 2,231 2000 13,174 9,999 3,867 - 4,334 3,423 2001 13,174 10,114 3,867 3,159 4,334 3,632 2002 14,761 11,438 4,773 4,000 4,130 3,393 2010 25,264 19,577 9,310 7,802 7,491 6,154 2020 43,291 33,546 18,815 15,767 14,429 11,854 Avcragc growth p.a. 2000-2020 6.1% 6.2% 8.7/o 8.8%/_I 6.2% 6.4% /_1: Average annual growth rate during 2001-2020. TRAFFIC FORECAST COMPARISON FOR NEW HIGHWAYS (in AADT) Lintona - Warriors Museum Hwy Famcrns - Tangvu Hwy Xian - Sanvuan Hwy Total MTE Total MTE Total MTE 1999 - - -- 2000 -- 1,969 -- 1,581 -- -- 2001 - 2,086 - 1.725 2002 - 2,210 - 1,883 2010 - 3,700 -- 3,794 -- -- 2020 -- 8,019 - 6,767 -- - Avcrage growth p.a. 2000-2020 - 7.3% -- 7.5% -- - ICR 1999 - - 3,230 2,172 18,847 14,158 2000 3,045 2,394 3,895 2,620 21,893 16,871 2001 3,330 2,526 4,228 2,863 21,741 17,887 2002 3,619 2,761 4,642 3,168 22,983 18,500 2010 6,433 4,908 8,789 5,998 48.634 39,148 2020 12,574 9,593 17,095 11,666 1(08,986 87,728 Avcrage growth p.a. 2000-2020 7.3% 7.2% 7.7% 7.8% 8.4% 8.6% Sources: SPTD and the Bank staff. 11. Actual traffic data were used for 1999-2001, and for subsequent years, the following growth rates were assumed: a. the normal traffic growth rates for new highways were estimated to be within a range of between 5 to 10 percent during 2002-2010, depending on the type of vehicle and road. Growth rates during 2010-2020 were assumed to be in the range of between 3.5 and 8.5 percent; and b. it was assumed that there would be no generated traffic. -31 -  Economic Costs 12. Financial construction costs were converted to economic costs by applying shadow prices for each of the input items. The total economic cost of six highways at the end of project implementation in constant December 2002 prices was 20.8 percent higher than the SAR's estimate. Economic Cost Comparison (Y million) ICR/ SAR SAR/i ICR (in%) 1. Weinan - Tongguan Expressway 1,159.30 1,464.14 26.3% 2. Tongchuan - Huangling Hwy 1,645.95 1,990.64 20.9% 3. Lantian -Xiaoshangyuan Hwy 334.91 431.00 28.7% 4. Lintong - Warriors Museum Hwy 25.97 44.99 73.2% 5. Famensi - Tangyu Hwy 180.56 251.27 39.2% 6. Xian - Sanyuan Hwy 150.35 43.08 -71.3% Total 3,497.04 4,225.12 20.8% /_1. The April 1995 prices (in SAR) were updated to December 2002 pnces. Economic Benefits 13. Project benefits were estimated by using the VOC equation that was developed in the Highway Design and Maintenance Standards Model (HDM-lI). The economic analysis includes the benefits derived from: (a) VOC savings on the new highways (summarized in the following tables); (b) time savings through relieved congestion on the existing roads; and (c) lower accident costs. The VOC savings includes the net VOC savings of the vehicles using the new highways. The time savings include the benefits resulting from lowered congestion on existing roads. The value of passenger time savings was estimated at Y 2.0 per passenger-hour, on the basis of updated values in a report on the feasibility study methodology for highways in China (Australian Feasibility Study Methodology Report, March 1996). The same source was used for vehicle accident rates on different classes of roads. Economic Vehicle Operating Costs (Yuan per km, December 2002 prices) Good - Fair- - Poor Flat Hill Mount. Fiat Hill Mount. Flat Hill Mount. Car 0.788 0823 0.849 0.825 0.857 0881 1.067 1.099 1 123 Medium bus 1.130 1.312 1.470 1.629 1.866 2.058 2.147 2.354 2.546 Large bus 2.004 2.444 2.835 3.045 3 667 4 178 3 689 4.289 4 800 Small truck 0919 1.149 1.321 0.974 1 190 1.352 1.190 1.379 1 568 Medium truck 1.068 1 316 1.564 1.097 1 348 I 577 1 326 1.577 1 828 Large truck 1.522 1.865 2.208 1.620 1.968 2 276 1.927 2.276 2.604 Tractor/trailer 2.206 2.909 3.517 2.332 3.035 3 627 2.776 3.442 4.053 - 32 -  Example of Economic Vehicle Operating Cost Calculation (Yuan per 1,000 veh.-kn, December 2002 prices) Medium Large Small Medium Large Trailer/ - Car bus bus Truck truck truck container New Hi0hwavs: Fuel 325 383 714 575 630 651 840 Tires 59 57 182 94 195 620 1,150 Maintenance 339 345 419 680 567 602 680 Crew 26 26 26 26 26 26 26 Depreciation 45 19 85 233 59 III 231 Subtotal 774 927 1.835 1.321 1 564 2 208 2.517 Time value of passengers 75 500 1,000 Total 849 1I470 2.835 1,321 1.564 2.208 3I.7 Existine Roads: Fuel 425 518 840 700 743 840 1,(50 Tires 63 64 208 101 221 680 1,265 Maintenance 305 383 466 756 630 669 756 Crew 60 71 71 71 71 71 71 Depreciation 41 21 94 258 66 124 257 Subtotal 252 [.213 ;L2 .156a 182 2.04 4.053 Time value of passengers 171 1,333 2,667 Total 1.123 . 4.800 1.568 1.828 2.604 4,053 Sources: SPTD and the Bank staff. Economic Evaluation 14. The economic internal rate of return (EIRR) of the high-grade highways was estimated at 13.1 percent, which is lower than the 16 percent estimated in the SAR. The lower EIRR is primarily due to much lower traffic volume on the Tongchuan-Huangling Highway and a substantial cost overrun. EIRRs and net present values (NPV) are summarized in the following table. EIRR (in %) and NPV {12%, Y million) SAR ICR EIRR NPV EIRR NPV l. Weinan - Tongguan Expressway 16.7 -- 15.6 465.5 2.Tongchuan-HuanglingHwy 16.3 8.7 (515.2) 3. Lantian -Xiaoshangyuan Hwy 14.3 -- 12.6 25.9 4. Lintong- Warriors Museum Hwy 14.7 -- 13.0 5.2 5. Famensi - Tangyu Hwy 12.6 -- 18.4 197.5 6. Xian - Sanyuan Hwy -- -- 29.0 168.1 Total 16.0 -- 13.1 408.2 Sensitivity Analysis 15. The EIRRs with different scenarios show that the changes in assumptions to be made in the evaluation could alter the yield on the investment. Because operation and maintenance costs of the high-grade highways are.relatively small in the investment, any variation in these costs would have minimum discernible impact on EIRR. The impact of changes in traffic volume was analyzed under two traffic growth scenarios. The results of the sensitivity test show that the reduction of passenger time savings by 50 percent has the greatest impact while the reduction of road congestion has the least impact on the EIRR. - 33 -  SENSITIVITY TEST EIRR (in %) NPV(12%, Y million) VOC savings reduced by5% 11.2 (268.3) Passenger time savings reduced by 50% 10.6 (485.0) Congestion savings reduced to zero 12.9 339.3 Total benefits reduced by 15% 11. (297.2) Scenario Analysis 16. Besides the normal traffic growth scenario, two traffic growth alternatives have also been considered: a lower traffic growth projection (25 percent lower than normal traffic) and a higher traffic growth projection (25 percent higher than normal traffic growth). The traffic growth and average annual growth rate under these two scenarios are as follows: EXPRESSWAY TRAFFIC FORECAST SCENARIOS High Growth Rate Low Growth Rate Average growth rate p.a. 7.9% - 10.6% 4.3% - 6.8% 17. The overall EIRRs for the different traffic growth scenarios are summarized in the following table. It shows that the overall EIRR for a higher traffic growth rate is 15.0 percent and for a lower traffic growth rate 11.3 percent. TRAFFIC GROWTH RATE SCENARIO TEST NPV (12%, Y EIRR (in%) million) High traffic growth rate scenario 15.0 1,211.4 Low traffic growth rate scenario 11.3 (236.8) ECONOMIC EVALUATION OF RIPA Background and Traffic 18. In the SAR, 4,068 km of rural roads was to be implemented in phases. The SAR specified that the first phase would cover five roads (for 321 kin). During implementation, the RIPA was carried out in four phases with 152 subcomponents covering the rural roads and bridges. The RIPA program commenced in October 1996 and was completed in December 2000. RIPA PRPGRAM SAR ICR Length (kn) Subcomponent Length (km) Subcomponent Completion Date Phase 1 321 5 319 8 March 1998 Phase 2 -- -- 2,105 89 December 1998 Phase 3 -- -- 1,081 36 December 1999 Phase 4 -- -- 414 19 December 2000 Total 4068 152 Sources: The SAR and SPTD 19. The main work under RIPA was repairing the existing roads and bridges in the poorest areas of - 34 -  Shaanxi province, particularly in the central and northeni parts of the province. Because all the RIPA subcomponents were in the same region, the economic development and traffic patterns were very similar. Based on traffic performance, traffic growth rates were assumed to be 5.0 percent per year for motor vehicles and 3.5 percent for tractors for 2002-2010, to be reduced by 1 percent for the next 10 years beyond 2010. In the SAR, the vehicle traffic growth rate was assumed to be 7.0-10 percent per year between 1995 and 2015 for all types of vehicles. Economic Costs 20. Financial costs were converted to economic costs using a methodology similar to that applied to the high-grade highways. All input items were evaluated in constant December 2002 economic prices. The overall economic cost is about 96 percent of the financial cost. During the project implementation, the project area was hit by a very severe flood disaster. The disaster destroyed not only most of the newly built roads and bridges, but also damaged the road bed, bridge piers, and the foundation of earth work. The repair work started in 1999, and the repair costs were even higher than the original budget for the project. All these repair costs were included in the economic evaluation. It is estimated that about two-thirds of total RIPA costs were for repair work. 21. Because of the major repair works, the total actual economic cost of RIPA at December 2002 constant prices was about 4.55 times higher than the SAR's estimate., RIPA COST (Y million, December 2002 prices) SAR (A) ICR (B) (B)/(A) in % Phase 1 108.53 148.62 36.9% Phase 2 -- 1,420.99 -- Phase 3 -- 730.78 -- Phase 4 -- 267.25 -- Total 462.80 Li 2,567.64 +454.8% LI. The April 1995 prices (in SAR) were updated to December 2002 prices. Economic Benefits 22. The economic benefits of RIPA were derived mainly from VOC savings due to shorter distances (if applicable), nonnal traffic, and generated traffic. 23. As a result of the project, the road classifications of all 152 subcomponents were upgraded. The better surface pavement reduced.VOC and yielded the major portion of the project benefits. Generated traffic was assumed to be 5 percent of normal traffic. In addition, the better paved roads and bridges increased vehicle speed and reduced congestion. These benefits were quantified also as a part of operating cost savings. Economic Evaluation and Sensitivity Analysis 24. The EIRRs of the four phases range from 9.2 percent to 24.7 percent. The overall EIRR for the RIPA is 16.7 percent. If there had been no flood disaster, the EIRR of the RIPA could have been as high as 47.6 percent (ranging from 29.7 percent to 55.9 percent). - 35 -  EIRR AND NPV OF RIPA EIRR (in %) NPV (12%, Y million) SAR ICR SAR ICR Phase 1 15.0 16.2 -- 34.3 Phase2 -- 15.1 -- 218.0 Phase 3 -- 24.7 -- 502.8 Phase 4 -- 9.2 -- (36.4) Total -- 16.7 -- 531.6 25. The risks considered for sensitivity analysis are: (a) slower than projected growth of traffic, (b) no generated traffic, and (c) fewer than expected benefits. The results of the sensitivity test show that the reduction of total benefits by 25 percent has the greatest impact, while no generated traffic has the least impact on the EIRR. SENSITIVITY TEST FOR RIPA -- EIRR (in %) NPV(12%, Y million) Slower than projected growth of traffic (-25%) 15.7 390.0 No generated traffic 16.4 502.3 Reduction of total benefit (-25%o 12.3 31.1 THE OVERALL ECONOMIC EVALUATION OF THE PROJECT Overall EIRR 26. The overall EIRR of the entire project (including high-grade highways and RIPA) is 14.1 percent, and the NPV (12 percent) is about Y 896.0 million. In the SAR, the EIRR for the entire project was estimated to be 15.8 percent. EIRR (in %) and NPV (12%, Y million) SAR ICR EIRR NPV EIRR NPV HGH 16.0 -- 13.1 408.2 RIPA 15.0 -- 16.7 531.6 Total 15.8 = 14.1 896.0 Project Risks 27. The overall EIRR of the project is economically justifiable despite the high repair costs for the RIPA due to the natural disaster. The lower EIRR for the high-grade highways was mainly due to cost overruns on most of the highways and less traffic demand on the high-grade highway that constitutes about 31.6 percent of the total length and 47.1 percent of the total cost of the high-grade highways. It is suggested that the project should have had better cost estimates for traffic origin-destination and construction during the feasibility study stage. - 36 -  Probabilistic Risk Analysis 28. To determine the degree of uncertainty for the project, a probabilistic risk analysis using Monte Carlo techniques was carried out. In a Monte Carlo analysis, each uncertainty factor is allowed to vary at random between set limits, and all uncertainty factors are allowed to change simultaneously. The Monte Carlo simulation provides probability distributions of the potential outcomes of decisions. By analyzing these distributions, it is possible to assess the risk associated with making various decisions (or probabilistic risk analysis). The product of the analysis is a judgment on the possible range of the decision variable and on the likelihood of each value within this range. 29. For the RIPA, the sum of capital costs constitutes only about 38 percent of the total capital investment. In view of the higher EIRR and lower capital investment of the RIPA, the probabilistic risk analysis for the project focused on the construction of the high-grade highways. The most uncertain factors associated with the economic evaluation of high-grade highways were identified as the traffic growth rate and the value of VOC. The results of the probabilistic risk analysis for the project show that the EIRR for the most likely scenario is 13.7 percent. The worst-case scenario of the analysis is 9.2 percent and for the best-case scenario, 18.5 percent. The standard error of the mean is 0.3 percent. The results of the Monte Carlo test and probabilistic analyses (Appendix 1) are summarized in the following table. Summary of Probabilistic Risk Analysis Most likely Standard error Range of EIRR EIRR of the mean 1. Weinan - Tongguan Expressway 11.4% - 21.5% 16.5% 0.3% 2. Tongchuan - Huangling Hwy 7.5% - 16.5% 11.6% 0.3% 3. Lantian -Xiaoshangyuan Hwy 10.7% - 20.0% 15.2% 0.3% 4. Lintong - Warriors Museum Hwy 11.0% - 20.9% 15.7% 0.3% 5. Famensi - Tangyu Hwy 17.7% - 19.1% 18.4% 0 0% 6. Xian - Sanyuan Hwy 27.8% - 30.4% 29.0% 0.1%'Y Total 9.2% - 18.5% 13.7% 0.3% FINANCIAL EVALUATION 30. The financial evaluation of the project comprises three sections: an assessment of the revenue-earning entity (the high-grade highways), focusing on consolidated financial statements; an assessment of the non-revenue-earning RIPA, focusing on financial risk in terms of construction, operation, and maintenance; and an assessment of the financial capa.city of SPTD to sustain highway activities, focusing on the adequacy of the funding allocations in the provincial road development plan. For the high-grade highways, the financing structure consists of the Bank loan (34 percent) and grant capital (66 percent). The financial cost of capital was assumed to be 5 percent for both sources. The SAR had no financial evaluation for the project. FINANCIAL EVALUATION OF TOLL HIGHWAYS 31. Operating entities have been established by SPTI) to operate high-grade highways. The entities are responsible for the day-to-day management, operation, maintenance, and development of the highways, including the repayment of the loan. The toll charges are strictly controlled by the provincial government. The toll charges were assumed to be increased once (15 percent) every five years (or an average of 2.8 percent per year). - 37 -  Financial Assessment 32. Tolls. The toll is charged on the basis of vehicle size (small vehicle, medium vehicle, large vehicle and extra large vehicle) and distance traveled. The current toll charges for the six highways are different and listed in the following table. Toll Charges (Yuan/ Vehicle km) Small Medium Large Extra- vehicle vehicle vehicle large 1. Weinan - Tongguan Expressway 0.431 0.689 0.819 1.121 2. Tongchuan - Huangling Hwy 0.338 0.540 0.675 0.946 3. Lantian -Xiaoshangyuan Hwy 0.293 0.440 0.733 1.026 4. Lintong - The Warriors Museum Hwy 0.625 1.250 1.875 2.500 5. Famensi - Tangyu Hwy 0.294 0.441 0.588 0.735 6. Xian - Sanyuan Hwy 0.412 0.619 0.825 1.031 33. Operations. The operating costs of the high-grade highways can be divided into working costs (wages and benefits; maintenance; operating materials and supplies; administration; others) and depreciation. All other expenses, including interest payments on the loan, are also taken into account in the calculation of the profitability of high-grade highways. 34. Profitability. The high-grade highways will generate enough revenue over the project life. However, lower profits may take place in 2002-2004. The main reason is that repayment of the Bank loan starts in 2002 and it will create financial pressure on the high-grade highways. In addition, because of the low toll growth assumptions, the rate of return on average net fixed assets for the early years of operation is low. It is anticipated that a negative profit may take place in 2009 because major maintenance work is planned in that year (see Table 2.1.) 35. Cash flow. The low toll growth assumption may cause lower profit, but it will not affect the cash flow of the high-grade highways. This is because of the large sum of depreciation reserves. In 2002 , for example, the total depreciation reserve (Y 124.13 million) contributed about 86.1 percent of the total financial obligations (Y 144.15 million), or equivalent to 6.5 times of total working costs in that year. The strong internal cash generation will be able to provide sufficient funds to support the operations, loan repayments, and even major maintenance expenditures. On the basis of the revenue forecast, it is estimated that the high-grade highways will not need any external cash injection over the life of the project (see Table 2.2). 36. Leverage and Liquidity. The low profit in the early years of operation will not affect the financial leverage of the high-grade highways because of the high proportion of equity (government contributions weighted at about 66 percent of the total project cost), which will support a healthy debt-to-equity ratio and debt-to-capital ratio (the financial leverage of the company) throughout the life of the project. Because of the strong cash flow status, except for 1999, the current ratio will be in the sound range of 2.9 and higher (to avoid short-term solvency problems, the current ratio should be at least 1.0) (see Table 2.3). The major assumptions for financial evaluations are shown in Table 2.4. - 38 -  Financial Internal Rate of Return 37. Based on the current financial status, the high-grade highways may not need any external financial assistance. However, the low profit margin produces a negative impact on the financial internal rate of return (FIRR). The result of the financial evaluation for the high-grade highways shows that the FIRR for the high-grade highways will be 3.7 percent with an NPV (5 percent) of Y 696.9 million. Scenario Analysis 38. The financial sensitivity analysis tested two scenarios: the higher traffic growth rate and the lower traffic growth rate for the high-grade highways. The results for the traffic growth scenarios are shown in the following table. FINANCIAL SCENARIO ANALYSIS FIRR (in %) NPV (Y million, 5.0%) Higher Traffic Growth Rate (+25) 5.8 492.7 Lower Traffic Growth Rate (-25%) 1.2 -1,641.4 Project Risks 39. SPTD completed the project on time within the acceptable range of cost overruns. The field visit confirmed that SPTD can provide quality services to road users while maintaining the roads in good condition. The only remaining tangible risks are the unforeseeable mixed factors (or degree of uncertainty) that will affect the results of the financial evaluation. This will be discussed in the on probabilistic risk analysis. Probabilistic Risk Analysis (Financial Simulation) 40. To determine the degree of financial uncertainty for the project, a probabilistic risk analysis using Monte Carlo techniques was carried out. The three most uncertain factors that may affect the financial evaluation have been identified as traffic growth rate, the growth rate of toll charges, and the growth rate of total working costs. The risk analysis revealed that the most likely FIRR would be 3.5 percent, while the worst and the best FIRR would be 0.8 percent and 5.9 percent, respectively. Details are in Appendix 2 and summarized in the following table. Financial Sensitivity and Probabilistic Risk Analysis Financial Simulation and Risk Analysis Range of Most Likely Std. Error of FIRR / NPV FIRR / NPV The Mean FIRR (in %) 0.8 - 5.9 3.5 0.2 NPV (5.0%, Yuan million) - 1,943 - 429 - 854 82.6 FINANCIAL EVALUATION OF RIPA 41. All the 152 subcomponents under RIPA are toll-free roads. The financial evaluation of these non-revenue-earning entities focuses on the SPTD's capacity to minirnize financial risks, i.e., the lack of counterpart funds for construction and repairs, including iuture operating expenses for the project roads and bridges. - 39-  42. SPTD's budget (actual and forecast) shows that the investment and maintenance expenditures for the RIPA constitute only a small fraction of SPTD's fund flow. Therefore, the fiscal impact is minimal. Based on the SPTD plan, capital investment in RIPA is less than 6.0 percent of the total annual SPTD revenue. In addition, the required maintenance expenditure of RIPA is less than 9.0 percent of the total maintenance expenditure of SPTD. These low ratios indicate that the project presents a low financial risk regarding construction and maintenance of the project roads and bridges. Details are shown in Table 3 and summarized in the following table. SPTD: Investments, Revenue and Maintenance Expenditures (Y million) 2000 2001 2002 2003 2004 2005 Investmcnts and Rcvcnuc: RIPA Investmcnts (a) 443 05 506.45 521.59 - - - Total SPTD Revenuc (b) 8,614.31 11,727.97 13,840.19 15,296.0 17,602.0 19,803.0 Ratios (a)/(b) 5.1% 43% 38% 0.0% 0.0% 0.0% Maintenance Expenditures RIPA (c) 34.69 40 18 41 43 40.28 41.72 41 99 Total SPTD Maintenancc Expcnditure (d) 525.00 464.17 488 90 483 0 511 () 525.0 Ratios (c)/(d) 6.6% 8.7% 8.5% 8.3% 8.2% 8.0% FINANCIAL EVALUATION OF SPTD 43. SPTD has provided its financing plan on total revenue and expenditures for the Ninth Five-Year Plan (1996-2000) and the Tenth Five-Year Plan (2001-2005). According to the Shaanxi highway development plan, SPTD has acquired sufficient funds to finance yearly increases in the length of the provincial road network and has allocated resources for the maintenance of the existing road network. In addition, a moderate self-financing ratio (34.7 percent for the tenth Five-Year Plan) will ensure the implementation of the highway development plan. The sources of highway revenue and expenditures for 1996-2005 are shown in Table 3. 44. To avoid a capital squeeze effect (i.e., limited annual capital resources forcing an emphasis on new road construction and reduced maintenance on existing roads), SPTD has given special attention to road maintenance. As for the sustainability of road maintenance with the resources allocated to it, the average yearly increase in maintenance expenditures during the Ninth Five Year Plan (6.9 percent) exceeded the growth of the road network (2.4 percent), which means an increase in average maintenanice expenditure per kilometer. SPCD: Average Increase in the Road Network and Road Maintenance Expenditures (in %) 9th FYP 10th FYP (1996 - 2000) (2001 - 2005) Average annual increase in the road network 2.4 2 0 Average annual increase in maintenance expenditures 6.9 1.I -40 -  Table 1: NUMBER OF MOTORIZED VEHICLES PER DAY (1999-2001) Existing Roads New Hioliwawi Total Traffic Diversion Total MTE Total MTE Total MTE Faictor (DF) (1) (2) (3) (4) (2) / (3) 1999 SAR Weinan - Tongguan Expressway -- -- - -- -- 5,872 - Tongchuan - Huangling Hwy -- - -- -- 6,477 Lantian -Xiaoshangyuan Hwy - _ _ .. 2,632 Lintong - Warriors Museum Hwy . _ .-- Famenst - Tangyu Hwy -- -- - -- -- 2,650 Xian - Sanyuan Hwy .- - ICR Weinan - Tongguan Expressway 12,228 10,159 11,980 9,691 24,2018 19,850 49% Tongchuan - Huangling Hwy 6,158 5,151 - - 6,158 5,151 - Lantian -Xiaoshangyuan Hwy 2,546 2,333 2,640 2,231 5,186 4,564 49% Lintong - Wamors Museum Hwy 3,425 2,693 - - 3,425 2,693 - Famensi -Tangyu Hwy 1,184 1,083 3,230 2,172 4.414 . 3,255 67% Xian - Sanyuan Hwy/lI - . 18,847 14,158 18,847 14,158 10t0% Ratos (ICR/SAR) Weinan - Tongguan Expressway -- -. - -. 338% / Tongehuan - Huangling Hwy -- -- - -- -- 80% Liintian -Xiaoshangyuan Hwy .. .. 1 73% Lintong - Warriors Museum Hwy .- .. _ . - Famensi - Tangyu Hwy .. 1. - -- -- I23% - Xian - Sanyuan Hwy .. - 2000 SAR Weinan - Tongguan Expressway -- 2,738 - 6,256 -- 8,994 70% Tongchuan - Huangling Hwy -- 2,865 - 6,583 -- 9,448 70% Lantian -Xiaoshangyuan Hwy -- 386 - 2,606 -- 2,992 R7% Lintong - Warriors Museum Hwy -- - - 1,969 -- 1,969 Famensi -Tangyu Hwy -- 1,028 - 1,581 -- 2,009 61% Xian - Sanyuan Hwy -- - - - - ICR Weinan - Tongguan Expressway 6,648 5,242 12,667 9.999 19.315 15,241 66% Tongchuan - Huangling Hwy 7,267 5,964 - -- 7,267 5,964 0% Lantian -Xiaoshangyuan Hwy 2,357 1,968 4,024 3,423 6,381 5,391 63% Lintong - Warriors Museum Hwy 4,029 3,167 3,045 2,394 7,1074 5,561 43% Famensi -Tangyu Hwy 1,171 992 3,895 2,621) 5,066 3,612 73% Xian- Sanyuan Hwy/LI - - 21,893 16,871 21,893 16,871 100% Ratios (ICR/SAR) Weinan-Tongguan Expressway -- 191% - 160P/o -- 169P/n - Tongchuan - Huangling Hwy 208% - (1% -- 63% Lantian -Xiaoshangyuan Hwy -- 510% - 131% -- 18(n - Lintong - Warriors Museum Hwy -- - -- 282% Famenst -Tangyu Hwy -- 96% - 166% -- 138% Xian - Sanyuan Hwy -- . _ . -41-  Table 1: Continue Exisfina Roads New Hiuhwaw Total Traffic Diversion Total MTE Total MTE Total MTE Factor (DF) (t) (2) (3) (4) =(2) / (3) 200!1 SAR Weinan - Tongguan Expressway - 2,883 - 6,782 -- 9,665 70% Tongchuan - Huangling Hwy -- 2,993 - 6,961 9,954 70