India - India Railway Electrification and Workshop Modernization Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 14338 IMPLEMENTATION COMPLETION REPORT INDIA INIDA RAILWAY ELECTRIFICATION AND WORKSHOP MODERNIZATION PROJECT (LOAN 2417-IN) MARCH 29, 1995 Infrastructure Operations Division Country Department 11-India South Asia Regional Office This document has a restricted distribution and may be used by recipients oniy in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY Currency Rs/US$ Official Market' Prior to June, 1966 4.76 June 6, 1966 to mid-December 1971 7.50 Mid-December 1971 to end-June 1972 7.28 1971/72 7.44 1972/73 7.71 1973/74 7.79 1974/75 7.98 1975/76 8.65 1976/77 8.94 1977/78 8.56 1978/79 8.21 1979/80 8.08 1980/81 7.89 1981/82 8.93 1982/83 9.63 1983/84 10.31 1984/85 11.89 1985/86 12.24 1986/87 12.79 1987/88 12.97 1988/89 14.48 1989/90 16.66 1990/91 17.95 1991/92 24.52 1992/93 26.41 30.65 February 1993 26.20 32.65 March 1993 31.53 Source: IMF, International Financial Statistics (IFS), line "rf.", and Reserve Bank of India. Note: The Indian fiscal year runs from April I through March 31. A dual exchange rate system was created in March 1992, with a free market for about 60 percent of foreign exchange transactions. The exchange rate was reunified at the beginning of March 1993 at the free market rate.  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT INDIA INDIAN RAILWAY ELECTRIFICATION AND WORKSHOP MODERNIZATION PROJECT (LN 2417-IN) TABLE OF CONTENTS PREFACE .......................................................... i EVALUATION SUMMARY ......................................................... i l. Introduction ........................................................ I II. TEK III Project Objectives .................................... 4 III. Achievement of Objectives ....................................5 IV. Major Factors Affecting the Project ......................................................... 10 V. Project Sustainability ............. ............................................. 11 VI. Bank Performance .....................................1..................... I VII. Borrower Performance ......................................................... 13 VIII. Assessment of Outcome ......................................................... 14 IX. Future Operation .......................................................... 15 X. Key Lessons Learned ......................................................... 16 This document has a restricted distribution and may be used by recipients only in the performance of their iofficial duties. Its contents may not otherwise be disclosed without World Bank authorization. l  ii ANNEX I Statistical Tables ................................................... 18 Table 1 Summary of Assessments ..................................................... 18 Table 2 Project Timetable ................................................... 19 Table 3 Loan/Credit Disbursements: Cumulative and Actual ........... ................ 19 Table 4 Key Indicators for Project Implementation and Operation ..................... 20 Table 5 Project Costs and Financing ..................................................... 21 Table 6 Status of Legal Covenants .................................................... 22 Table 7 Bank Resources - Staff Inputs ..................................................... 23 Table 8 Bank Resources - Missions ................................................... 24 ANNEX II Economic Analysis Workshop Component .......................... ................ 25 ANNEX III Economic Analysis Electrification Component ...................................... 32  ACRONYMS AND ABBREVIATIONS 4WU - Four Wheeler Units AC - Alternating Current ADB - Asian Development Bank BEML - Bharat Earth Movers Ltd. BG - Broad Gauge (1.676 m) BUTP II - Second Bombay Urban Transport Project CONCOR - Container Corporation DE - Diesel Electric DH - Diesel Hydraulic EL - Electric Locomotive EMU - Electric Multiple Unit FY - Fiscal Year H.S.D. - High Speed Diesel HP - Horsepower ICF - Integrated Coach Factory ICR - Implementation Completion Report IR - Indian Railways IRR - Internal Rate of Return kv - kilovolts kwh - kilowatt hour MG - Meter Gauge MGTKM - Million of Gross Ton-Kilometers Passenger and Freight MIS - Management Information System N/A - Not Applicable OED - Operations Evaluation Department oIS - Operations Information System PCR - Project Completion Report POH - Periodic Overhaul R&D - Research and Development RKM - Route Kilometers SAR - Staff Appraisal Report SEB - State Electricity Board SOE - State Owned Enterprise TEK I - Railway Modernization and Maintenance Project TEK III - Railway Electrification and Workshop Modernization Project TEK II - Railway Modernization and Maintenance Project II TEK IV - Third Railway Modernization Project TEK V - Railway Productivity Improvement Project US$ - United States Dollars   IMPLEMENTATION COMPLETION REPORT INDIA INDIAN RAILWAY ELECTRIFICATION AND WORKSHOP MODERNIZATION PROJECT (LN 2417-IN) PREFACE 1. This Implementation Completion Report (ICR) is based on a review of the files of the project, the original staff appraisal report, a partial draft ICR by the project supervisory engineer, and a detailed report and economic analysis prepared by a team of private Indian analysts retained by the Bank. 2. It was the third operation in a series of five interrelated Bank-supported railway technology transfer projects commencing in 1978 and ending in 1992. Many of the issues raised in the project under review can only be grasped within the context of the entire technology transfer series and it was decided to examine the experience with this project in the context of fourteen years of related lending. The project itself was complex and controversial as were those preceding and following it so the ICR is, of necessity, somewhat lengthy. The effort, therefore, takes on some of the characteristics of sector work and it has served the division in that capacity. 3. The information base for the detailed report by the ICR team was not readily available in IR reports. But, IR staff cooperated well with the ICR team and the team of local consultants performed very well in extracting, organizing, presenting and analyzing the information available. It is the first retrospective look at railway electrification in India. 4. In setting the context for the technology transfer series the OED audit of the first and the PCR for the second were consulted as were the appraisal reports of the fourth and fifth operations. 5. At the time this document was submitted to the Board the written version of the project experience from Indian Railways had not been received.   IMPLEMENTATION COM1PLETION REPORT INDIA INDLIN RAILWAY ELECTRIFICATION AND WORKSHOP MODERNIZATION PROJECT (LN 2417-lN) EVALUATION SUMMARY 1. The main objectives of the project were to: (i) increase capacity to meet growing traffic demand, (ii) improve utilization of motive power, rolling stock and track, (iii) improve cost effectiveness, (iv) utilize energy efficient traction, (v) improve indigenous manufacturing of electrification components, and (vi) strengthen the organization in selected operational areas. 2. The above objectives were to be pursued by: (i) completion of the electrification (3000 route kilometers) of the high density quadrilateral formed by the megacities of Delhi, Madras, Calcutta and Bombay; (ii) the modernization of six key workshops, one manufacturing unit and 37 selected maintenance depots; and (iii) the adoption of a real-time, computer-based, Operating Information System (OIS). 3. The OIS component was not implemented (and remains unimplemented) so its contribution to the project objectives is not yet being made. 4. The workshop modernization component for the six major workshops and the coach manufacturing facility was implemented under budget in US$ terms but at a 100 percent time overrun. Five of the seven main workshop and manufacturing components are functioning well and likely to be good investments from the micro analytical point of view. 5. The electrification component was physically completed much as specified at appraisal despite substantial time overruns for some sections. In US$ terms the cost overruns were not substantial for any of the sections. The traffic projections for the electrified sections were generally underestimated with actual traffic flows (passenger and freight in millions of gross ton-kilometers) on the average about 7 percent above appraisal expectations. 6. Despite completion under budget and more traffic than expected the retrospective economic analysis shows that the electrification investments were not good in economic terms. Revised analyses were done for sections I (Jhansi-Itarsi) and 5 (Balharshah-Vijaywada) following the format and assumptions utilized in the original SAR. Both sections had traffic densities of about 40 million GTM/RKM in 1990-91 and both were completed with cost overruns well below the average for all the sections. Jhansi-Itarsi had only a 19 percent cost overrun and was energized on schedule. Yet, the revised economic analysis for this section shows an IRR of about 9 percent - down from the original SAR estimate of 23 percent. For the Balharshah-Vijaywada section the cost overrun was 31 percent  ii and the time overrun was about four years. The revised IRR for this section is now about 2 percent rather than the SAR estimate of 41 percent. These two sections accounted for 27 percent of the route kilometers electrified under the project, experienced the lowest cost overruns and had the highest traffic densities. It is reasonable therefore to assume that the rest of the sections were not superior in economic performance and that the electrification component as a whole is not likely to yield a satisfactory return on the investment. 7. There are numerous reasons for this poor outcome. The implementation period that required eight years rather than four years for many of the sections is a principal cause. Another reason is that the shift from diesel to electric on the energized sections was gradual and partial rather than abrupt and complete as was assumed in the SAR economic analysis. Thus, for Jhansi-Itarsi freight traffic was 42 percent electrified in 1989-90 rising to 67 percent in 1993-94 and is expected to reach 90 percent by 1996-97. Part of this is because some freight arriving on the electrified sections from nonelectrified sections will proceed under diesel power. Part is because of the far more serious problem of a general shortage of electric locomotives. This latter problem is exacerbated by the fact that the 6000 HP state- of-the-art electric locomotives assumed in the original economic analysis were never introduced because of the failure of the technology transfer initiated under the previous lending operation. The obsolete electric locomotives that were used never achieved the efficiencies expected in terms of speed, maintenance cost, capital cost or energy saving. In many cases the old diesel locomotives did just as well or better. Finally, the collapse of oil prices in 1986 combined with rising domestic electric power costs removed many of the energy saving benefits assumed to be associated with electrification. 8. Of the four categories of assessment the Bank considers for its completed projects the one under review was neither Highly Satisfactory nor Highly Unsatisfactory. In fact, the outcome for this project was somewhere between satisfactory and unsatisfactory. Some of the workshop components that were well supervised and had working management information systems (MIS) in place showed very satisfactory rates of return (25 percent in the case of Golden Rock). At least four of the six workshop modernization components had satisfactory rates of return. But, the workshop component was only 13 percent of total completed project costs. The electrification component finally dominated the total project cost, accounting for 80 percent of expenditures. And, the rate of return on the electrification investments was likely to be in the area of 4 percent. The weighted average rate of return on the total project is therefore estimated at about 7 percent. For a Satisfactory rating the project must have: "achieved most of its major objectives and have achieved or be expected to achieve satisfactory development results with only a few shortcomings." For an Unsatisfactory rating the project must have: "..failed to achieve most of its major objectives and is not yielding and is not expected to yield substantial development results, and has significant shortcomings." 9. With two of the six major workshop modernizations and virtually all of the electrification investments unlikely to show satisfactory economic returns and very slow progress on the OIS system, the Unsatisfactory rating must be chosen.  IMPLEMENTATION COMPLETION REPORT INDIAN RAILWAY ELECTRIFICATION AND WORKSHOP MODERNIZATION PROJECT LN 2417-IN I. Introduction 1. This project was the third in a series of five interrelated technology transfer projects commencing in 1978 and ending in 1992. These five projects (TEK I through TEK V) and their components are set forth in Table I and provide part of the essential context for understanding the role and impact of TEK III, which is the subject of this report. The rest of the context is provided by setting forth the rationale for the technology transfer series and what, broadly, has been achieved with this 14 year US$ 1.2 billion effort. 2. Previous to the TEK series the Bank had made thirteen loans to Indian Railways (IR) which were essentially lines of foreign exchange credit. In 1976 Bank management decided that it had achieved all it could under a sector lending format and that a project specific format should be adopted in which the transfer of modem efficient railway technology would be the central objective. While IR had successfully introduced first generation diesel and electric technology together with four axle wagons it had not, in the absence of an effective R&D organization, been able to continuously upgrade the new technology. Neither had it been able to rationalize its management and labor practices nor its steam based system of maintenance and overhaul facilities. Too, the communications and computer technology which had revolutionized developed country railroad operations had hardly touched IR. 3. Bank staff saw the priorities as the rapid transfer of off-the-shelf locomotive and wagon technology together with a computer based Operations Information System (OIS) to provide efficient control of the costly new rolling stock and locomotives. TEK I was, therefore, originally conceived as locomotive and rolling stock technology transfer to the vertically integrated IR manufacturing operations. The first request of Bank management at this point was to corporatize the manufacturing operations of IR so that they could be treated as the industrial enterprises they were (and still are). This position was not accepted by IR or the Government. Nor did the Government accept the related notion that the second generation of locomotive designs should be imported. This conflicted with the then national policy of indigenous technology development. TEK I was then downgraded in 1978 into support of the already existing IR plans for a wheel and axle plant, the modernization of existing workshops and the purchase of spare parts. The only concessions to the original Bank objectives were to consider workshop rationalization as well as modernization and to accept assistance in the indigenous development of the second generation of locomotives through a research and development (R&D) component.  2 Table 1 Indian Rail Technology Transfer Loan Series 1978-1992 (Millions of Current US Dollars) Item TEK I TEK 11 | TEK III TEK IV TEK V 1 Totals 1978 1982 j 1984 J 1988 1992 j 1978-92 Diesel Locomotive 55 (39) 139 (-) 194 (39) Technology Electric Locomotive 45 (64) 50 (-) 95 (64) Technology Rolling Stock 38 (55) 145 (70) 94 (-) 277 (125) Technology l Workshop 95 (90) 131 (129) 226 (219) Modernization Spare Parts & Unit 45 (43) 138 (255) 183 (298) Exchange ll Electrification 143 (137) 143 (137) Track Modernization 388 (258) 388 (258) Operations Info. 15 (-) 15 (-) System (OIS) Research and 10 (1) 3 (-) 13 (1) Development l Technical Assistance 2 (2) It (8) 7 (6) 2 (2) 2 (-) 24 (18) and Training Totals f190 (191) 397 (436) 281 (272) 390 (260) 300 (-) 1558 (1159) Notes: Figures in brackets are the amounts actually disbursed. TEK I is officially entitled Railway Modernization and Maintenance Project (Credit 844-IN). TEK 11 is Railway Modernization and Maintenance Project II (Loan 2210- IN/Credit 129-IN). TEK III is Railway Electrification and Workshop Modernization Project (Loan 2417-IN). TEK IV is Third Railway Modernization Project (Loan 2935-IN). TEK V was to be Railway Productivity Improvement Project.  3 4. By the time TEK n1 began in 1982, it was clear from the failure of the R&D component of TEK I that second generation (by then third generation) locomotive technology and freight wagon technology had to be imported.1 TEK 11 looked much more like the Bank's original idea of the TEK series. Electric locomotive prototypes and bogies for high capacity wagons were to be imported for testing under Indian operational conditions and a modem diesel remanufacturing facility at Patiala was included. The big disappointment with TEK 11 was the inability to get the OIS component imported and installed. As in the earlier case of the locomotives it was thought that an indigenous OIS was entirely feasible.2 5. TEK m (the subject of this report) came along in 1984 and was really the residue of TEK 11 components that could not be accommodated two years earlier.3 It called for more workshop modernization and supported the accelerated electrification of the high traffic density double track sections of IR. The very slow progress of the workshop rationalization effort under TEK I was accommodated under the TEK HI workshop modernization component by asking for no more than what had already been agreed (but not delivered) under TEK I. The big push under TEK m was to get the OIS component going since it was by then apparent that the indigenous effort was not forthcoming. The serious problem of redundant rail labor was also being confronted by the inclusion of a comprehensive manpower study. 6. TEK IV came in 1988, four years after TEK m and was aimed at modernizing IR track and track maintenance technology by financing the import of high strength steel rails that could not be manufactured indigenously and by importing track installation and maintenance equipment that would allow near continuous use of the now saturated main lines.4 The new track would also accommodate the smaller diameter wheels and heavier axle loads of the proposed new freight rolling stock. 7. TEK V was a final attempt in 1992 to revisit the original 1978 objectives of the TEK series.; Despite the disbursement of almost US$ 1.2 billion in loans and credits over fourteen years, the locomotive and rolling stock initiatives under TEK 11 had not yet born fruit and IR continued with the technologies of the 1960s. TEK V called for the urgent and rapid transfer of diesel and electric (by now fourth and fifth generation) technologies to IR manufacturing operations as well as light weight high capacity freight rolling stock. There was a large component for the long-delayed OIS and no attempt to support the failed domestic R&D effort. By this time the problem of line saturation and locomotive and rolling stock shortages were apparent to all as was the accelerating labor bill. 8. In the general environment of national economic reform of the 1990s, Bank management returned to its 1978 position that the vertically integrated manufacturing operations of IR should be "unbundled" from IR, the transport enterprise, so that the desperately needed technology transfer could finally take place. As I For the background and eventual performance of the TEK I lending operation see: Proiect Performance Audit Report India Railway Modernization and Maintenance Proiect (Credit 844-IN), Report No. 7020, The World Bank, Operations Valuation Department, November 30, 1987. 2 For the background and eventual performance of the TEK II lending operation see: Project Completion Report India Railway Modernization and Maintenance Proiect ll (Loan 2210-IN/Credit 1299-IN), Report No. 9455-lN, The World Bank, March 22, 1991. 3 Staff Appraisal Resort India Railway Electrification and Workshoo Modernization Proiect, Report No. 4940-lN, The World Bank, April 26, 1984. 4 Staff Appraisal Report India Third Railway Modernization Proiect, Report No. 4706-lN, The World Bank, April 15, 1988. 5 Staff Appraisal Report India Railway Productivity Improvement Proiect, Report No. 10054-IN, The World Bank, March 30, 1992.  4 in 1978 (and in 1980), IR management refused to consider such a move and the scheduled loan negotiations were canceled - and never subsequently taken up. 9. As of 1994, only TEK IV is still disbursing. At the request of IR, US$ 130 millions of the TEK IV loan was canceled on the grounds that the high strength steel rail could be manufactured indigenously and need not be imported. Subsequently it was found that the domestic rail quality could not be improved, but the canceled amount could not be reinstated. 10. Since the cancellation of the scheduled TEK V loan negotiations in 1992, Bank rail lending in India has focused on non-controversial specialized operations such as that for the Container Corporation (CONCOR), a corporatized wholly-owned subsidiary of IR, and for the Bombay suburban rail operations (BUTP I). IH. TEK m Project Objectives 11. As stated in the Staff Appraisal Report (SAR), the main objectives of the project were to: (i) increase capacity to meet growing traffic demand, (ii) improve utilization of motive power, rolling stock and track, (iii) improve cost effectiveness, (iv) utilize energy efficient traction, (v) improve indigenous manufacturing of electrification components, and (vi) strengthen the organization in selected operational areas. 12. Objectives (i) and (ii) were, and remain, critical to the well being of IR and the transport sector. Objectives (iii) and (vi) are generally desirable and generic to most projects. Objectives (iv) and (v) are both desirable and explicit. 13. The above objectives were to be pursued by: (i) completion of the electrification (3000 route kilometers) of the high density quadrilateral formed by the megacities of Delhi, Madras, Calcutta and Bombay; (ii) the modernization of six key workshops, one manufacturing unit and 37 selected maintenance depots; and (iii) the adoption of a real-time, computer-based, Operating Information System (OIS). 14. The linkage between the project components and the project objectives is clear in the case of the OIS component6. The Bank had been pushing (unsuccessfully) for the OIS since before TEK I in 1978. The TrEK HI SAR, in paragraph 2.25, makes the capacity expansion and efficiency case for the OIS again and, in paragraph 6.02 specifies OIS related items as preconditions for loan effectiveness. For the workshop modernization component the linkages with objectives (i), (ii), and (vi) are clear insofar as the individual workshops are concerned but not in terms of the entire national system of workshops. Individual workshop rationalization and modernization was to take place but not system rationalization. The latter objective was pursued unsuccessfully under the TEK I workshop modernization component. The electrification component was generally supportive of all the project objectives since the investment was restricted to the increasingly congested high density lines of the megacity quadrilateral. The pursuit of objective (v) was presumably made operationally explicit by linking it with the electrification component. 6 The OIS "component" was not financed under TEK III. It was pursued as a precondition for negotiations and then loan effectiveness.  5 Im. Achievement of Objectives The OIS Component 15. The OIS component was not implemented (and remains unimplemented) so its contribution to the project objectives is not yet being made. Under TEK I, in 1978, the OIS initiative and its well-documented impact on capacity expansion and capital efficiency was promoted. Under TEK 11, adoption of an implementation action plan was a precondition for project processing. Under TEK ml, the Bank sought assurances that Indian licenses and approvals would be forthcoming for importing the OIS software and hardware. This was a precondition for negotiations that led to a one year delay in completing negotiations. The appointment of OIS consultants was then a precondition for TEK m loan effectiveness and this led to another one year delay and near cancellation of the loan. The contentious OIS initiative cost two years delay between TEK m appraisal and loan effectiveness. By the time TEK V was appraised in 1992, the Bank was still calling for the implementation of the OIS system, initially on the Northern Zonal Railroad. The Workshop Modernization Component 16. The workshop modernization component for the six major workshops and the coach manufacturing facility was physically implemented but at a 38 percent rupee cost overrun and approximately 100 percent time overrun. The facility cost and productivity results are given in Table 2.7 While the Integrated Coach Factory (ICF) at Madras experienced a 26 percent cost overrun and took nine years to complete, a retrospective economic analysis by Bank staff yields an expected IRR of 13 percent - down from the 20 percent estimated at appraisal (See Annex II). This still satisfactory outcome derives from exceeding the output target of 1000 coach POHs per year. Two other workshops were reevaluated by Bank staff - that at Golden Rock and that at Ajmer. The Golden Rock facility was completed under the cost estimate and the POH cycle time targets have been achieved but the annual number of POHs have not been achieved because of lack of demand. The revised economic analysis shows an IRR of 25 percent similar to the 27 percent estimated at appraisal. For the Ajmer workshop the experience has not been good because of cost and time overruns, reduced output figures and less than targeted cycle times. The revised IRR was estimated at 2 percent down from the 24 percent at appraisal. No other workshop investments were reanalyzed but; based on the cost, output and productivity figures in Table 2; it is likely that the facility at Jagadhri was not, in retrospect, economically justified. Thus, five of the seven main workshop and manufacturing components are functioning well and likely to be good investments from the micro analytical point of view. These are Rupee cost overruns which, because of the long implementation period, inflation rates of about 10 percent and a Rupee to US$ conversion of 11 in 1984 and 30 in 1993, are not serious in terms of real resources. If a weighted average exchange rate of Rs. 20 to the USS is used for the total cost of this component there is a substantial cost underrun in terms of real resources. See Table 5 in Annex I.  6 Table 2 Results of Workshop Component Parel Liluah Jagadhri Golden Kharagpur Ajmer ICF Total _ ~~~~~Rock _l_l Appraisal cost (Rs. million) 161.1 142.3 1880.6 280.9 311.9 235.4 540.6 1,852.8 Final cost (Rs. million) 217.8 210.6 450.0 275.4 428.2 297.1 678.7 2,558.0 Appraisal estimated Jun 88 Jun 88 Jun 88 Jun 88 Jun 88 Jun 88 Jun 88 completion date Current/actual Mar 93 Sep 93 Dec 93 Mar 93 Dec 93 Mar 93 Mar 94 completion date No. of Loco POH DE DH DE EL DE Actual at appraisal 72 24 77 0 36 Appraisal target 108 48 144 48 72 Achieved to date 108 48 120 58 60 No. of Coach POH EMU MG NEW (four wheeler units) Actual at appraisal 3,600 3,000 0 2,600 850 Appraisal target 4,800 3,600 96 108 3,600 1,000 Achieved to date 4,260 3,000 96 102 2,600 1,016 No. of Wagon POH (four wheeler units) Actual at appraisal 13,000 12,000 Appraisal target 13,600 18,000 Achieved to date 13,600 12,720 Time required, loco POH, days Actual at appraisal Appraisal target 23 45 23 47 Achieved to date 18 25 19 18 18 29.1 19 24 Time required, coach POH, days Actual at appraisal 24 21 15 17 21 Appraisal target 19 14/12 13 16 14 Achieved to date 18.8 17.1 13 16.3 18.0 Time required, wagon POH, days Actual at appraisal 10 6.9 8 Appraisal target 8 4 6 Achieved to date 6.1 6.9 6 Staff employed At appraisal 5,886 11,175 5,442 7,176 15,344 8,340 15,547 Currently 5,331 11,227 5,177 6,115 14,068 7,763 15,691 Management Information System, computer modules implemented Target 7 7 7 7 7 7 Achieved to date 5 7 4 7 6 5  7 17. While most of the project preparation for this component was focused on the seven facilities referred to above it is important to note that 57 percent of the workshop component (about 19 percent of total project costs) was originally allocated to 37 different "support" facilities scattered throughout the system. This subcomponent absorbed the Rupee cost overruns of the seven major facilities and the electrification component. but, even after reduction (to about 7 percent of total project costs) involved the purchase and installation of over 1,300 major machine tools at a cost of about US$ 74 million. It was t his subcomponent that helped make the project so complex and added greatly to the procurement burden. It is virtually impossible to supervise such an operation and equally difficult to assess the impact of putting new machine tools into these shops. 8 For purposes of weighting, the economic return on this component was assumed to be 10 percent. 18. Nevertheless, locomotive and rolling stock availability for the entire rail system increased quite dramatically from 1986 on with diesel locomotive availability now at 90 percent, electric at 93 percent, coaches at 91 percent and wagons at 96 percent (See Table 4 in Annex 1). The workshop components under TEK I and TEK m undoubtedly had much to do with this improved performance. 19. However, this same improvement could have been achieved with a greatly reduced set of national maintenance facilities. If maintenance system rationalization could have been achieved as well as modernization less than half of the facilities extant would be required for maintenance and overhaul facilities and the machine tools in the facilities retained would have been utilized much more intensively and efficiently. There would be no need to deal with 37 different facilities operating single shifts and in many cases, e.g. Golden Rock, with work loads less than the facility capacity. The problem of redundant maintenance facilities means that the modernization required to have efficient locbmotive and rolling stock availability is itself both costly and inefficient. This factor was not considered when programming machine tools for the 37 support units and could not be considered for the same reason that it was not successfully addressed in TEK I. The Electrification Comp~onen 20. The electrification component was physically completed much as specified at appraisal despite substantial time and Rupee cost overruns for some sections (Table 3). For the entire 3043 km, the Rupee cost overrun was 82 percent above appraisal estimates. The most egregious overrun was for the last two sections Bina-Katni-Bilaspur where the appraisal costs were exceeded by 229 percent. These two sections, which accounted for 19 percent of the route kilometers to be electrified, eventually accounted for 34 percent of the route electrification costs. These two sections were also the last to be energized almost four years after the original expected completion date. The other difficult sections were the third, fourth and sixth which experienced cost overruns in excess of 80 percent. In US$ terms however, the cost overruns were not substantial for any of the sections and the electrification component was completed approximately on budget in real resource terms (See Table 5 in Annex I). While originally estimated to require 67 percent of total project costs it eventually consumed 80 percent of the resources available for the project. 21. The traffic projections for the electrified sections were generally underestimated with actual traffic flows (passenger and freight in millions of gross ton-kilometers) on the average about 7 percent above appraisal expectations. On the surface this is adequate forecasting performance for capital budgeting purposes. However when the traffic flows are broken down into freight and passengers, the actual passenger traffic was 23 percent above expectations while freight was only 4 percent above projections. The This point was made in the 1987 Audit of the 1978 TEK I operation but was much too late to be reflected in the 1984 TEK HI operation. This is one of the criticisms of the Bank's system of post completion technical audits, i.e. the lessons arrive too late to be of any operational use. As an example of what is possible; the US CONRAIL system, which has a locomotive fleet about 25 percent of the IR fleet, has one workshop for overhauls and 2 sheds for running repairs and maintenance, plus 3 fueling facilities.  8 unexpectedly high growth in passenger traffic has substantial operational as well as financial implications for the railway. The operational implications derive from the fact that passenger trains are competing for track space with freight trains and that a single passenger train generally takes up twice the track capacity required for a single freight train. The financial implications derive from the fact that freight services generate the profits to cover the money losing passenger services. 22. The actual traffic densities in 1990-91 on the various sections in millions of gross ton-kilometers per route-kilometer are also given in Table 3. Here it can be seen that the average for the entire 3043 kilometers electrified was 33.3 MGTKM/RKM with a high of 41.7 million tons on section 5 and a low of 13.3 million tons on section 10. The European experience with mixed passenger and freight operations generally supported electrification when the gross tonnage exceeded 20 to 30 million per route-kilometer and, with the exception of sections 7 and 10 all meet or exceed this rule-of-thumb. There should, therefore, be a good retrospective economic argument for the electrification investment. 23. Revised economic analyses were done for sections I (Jhansi-Itarsi) and 5 (Balharshah-Vijaywada) following the formnat and assumptions utilized in the original SAR (See Annex HII). Both sections had traffic densities of about 40 million GTM/RKM in 1990-91 and both were completed with cost overruns well below the average for all the sections. Jhansi-Itarsi had only a 19 percent cost overrun and was energized on schedule. Yet, the revised economic analysis for this section shows an IRR of about 9 percent - down from the original SAR estimate of 23 percent. For the Balharshah-Vijaywada section the cost overrun was 31 percent and the time overrun was about four years. The revised IRR for this section is now about 2 percent rather than the SAR estimate of 41 percent. These two sections accounted for 27 percent of the route kilometers electrified under the project, experienced the lowest cost overruns and had the highest traffic densities. It is reasonable therefore to assume that the rest of the sections were not superior in economic performance and that the electrification component as a whole is not likely to yield a satisfactory return on the investment. 24. There are numerous reasons for this poor outcome (See Annex II). The implementation period that required eight years rather than four years for many of the sections is a principal cause. Another reason is that the shift from diesel to electric on the energized sections was gradual and partial rather than abrupt and complete as was assumed in the SAR economic analysis.10 Thus, for Jhansi-Itarsi freight traffic was 42 percent electrified in 1989-90 rising to 67 percent in 1993-94 and is expected to reach 90 percent by 1996- 97. Part of this is because some freight arriving on the electrified sections from nonelectrified sections will proceed under diesel power. Part is because of the far more serious problem of a general shortage of electric locomotives. This latter problem is exacerbated by the fact that the 6000 HP state-of-the-art electric locomotives assumed in the original economic analysis were never introduced because of the failure of the technology transfer initiated under TEK II." The obsolete electric locomotives that were used never achieved the efficiencies expected in terms of speed, maintenance cost, capital cost or energy saving. Finally, the collapse of oil prices in 1986 combined with rising domestic electric power prices removed many of the energy saving benefits originally assumed to be associated with electrification. 10 A transition period of a year or so would not have been unreasonable. Under TEK II the Bank helped finance the import of eighteen prototype AC electric locomotives for testing to provide the basis for a new locomotive design. The locomotives were purchased and tested as agreed but did not lead to a decision for large scale technology transfer and manufacture.  9 Table 3 Cost and Traffic Results of Electrification Component Section Route Appraisal Cost Final Cost 1990-91 Traffic 1990-91 1990-91 Traffic Kilometers Estimate Estimate Projected at Appraisal Traffic Actual Density Date Energized RKm Rs. Million Rs. Million MGTKm MGTKm MGTKm/RKnm 1. Jhansi-Itarsi 381 1222 1454 13640 15158 39.8 3/89 2. Itarsi-Bhusaval 301 899 1313 9498 11368 37.8 3/91 3. Bhusaval-Nagpur 393 1010 ) 9400 14090 35.9 3/89 2514 4. Wardha-Balharshah 133 376 3780 5388 40.5 3/89 5. Balharshah-Vijaywada 454 1250 1642 19970 18914 41.7 3/92 6. Bhopal-Nagda 239 576 1058 6260 6798 28.4 3/92 7. Itarsi-Nagpur 297 855 1074 6560 5086 17.6 3/91 8. Nagpur-Durg 265 807 1331 5990 6548 27.6 3/92 9. Bina-Katna-Anuppur 429 1278 ) 17230 15885 41.1 9/93 5340 10. Anuppur-Bilapur 151 345 1790 1797 13.3 3/94 TOTALS: | 3043 [ 8618 15726 [ 94118 J 101032 33.2 Notes: RKm = Route Kilometers; MGTKm = Million of Gross Ton-Kilometers Passenger and Freight  10 25. Electrification proceeded without a coordinated plan for electric locomotive acquisition and revised train operations and maintenance procedures. Despite the statement of the capacity expansion objective in the TEK m SAR, the economic analysis employed did not consider this objective explicitly and the main benefit assumptions were reduced capital cost for locomotives and wagons and reduced energy consumption because of the introduction of the new technology. The real benefits of electrification of congested mixed passenger and freight operations is now in the expansion of track capacity because of the superior acceleration and speed of electric locomotives together with their lower maintenance cost and higher availability. The question of track capacity and saturation was never addressed in the SAR or in the economic analysis nor was there an analysis of optimal train operations under state-of-the-art electric locomotion.12 Instead, the popular justification for electrification i.e. energy saving and foreign exchange saving was reflected in the economic analysis and, under the then prevailing prices of electricity and diesel fuel, provided a sufficient justification for going forward. In retrospect it was not an adequate approach to the issue. IV. Major Factors Affecting the Project 26. For the OIS component, the major factor impeding progress was the lack of enthusiasm for computer based information systems from the management of the zonal railways and from the clerical labor force. The zonal railways were fearful of losing independence and control to the central Board with such a system. The clerical workers who were engaged in manual collection and compilation of OIS type information were also fearful of losing their jobs to computers. Finally, there was a group of technologists outside the railways who believed that both the software and hardware for such a system could be developed indigenously and need not be imported. There was no real "ownership" of the OIS concept in India at least not in enough force to prevail against those who resisted it. 27. For the workshoD modernization component the major factor was the plethora of redundant workshop facilities that could not be "rationalized" because of the self sufficiency preferences of the zonal railways and the impact on the labor force. This led to the need to modernize 37 separate support facilities when a fraction of this number could have done the job. This portion of the project was very loosely planned and impossible for Bank staff to supervise. An isolated factor was the large scale conversion of meter gauge rail to broad gauge which rendered the Ajmer meter gauge workshop largely redundant in its own right. 28. For the electrification component, two major exogenous factors were the rise in electricity cost vis a vis diesel fuel cost when world petroleum prices collapsed in 1986 and the development of extremely fuel efficient diesel locomotive technology subsequent to project appraisal. The energy saving dimension of electrification has, in the 1990s, nearly disappeared. The foreign exchange savings argument also no longer holds as the Indian economy expands its foreign trade sector and now has a "problem" of excess of foreign exchange. The failure to transfer the state-of-the-art electric locomotive technology in time to make use of 12 The Chinese - not the Europeans or North Americans - have led the way in gaining the benefits of optimal train operations plus electrification on their mainline double track operations. As early as 1988 China railways was routinely moving between 140 and 180 MGTKM/RKM, almost three times the mainline Indian densities. See: Present Situation of Chinese Railways and Uoeradina Program, Permanant Way Bureau, Ministry of Railways, China Academy of Railway Sciences, Beijing, December 1989. For a more recent example of the approach to optimal train operations and investments in China see: "Railroad Capacity Planning: A Case Study of the Beijing-Shanghai Rail Corridor", by Carl Van Dyke and Llewellyn Davis, Journal of the Transportation Research Forum, Vol. XXXII, Number 1, 1991, p. 86-102.  I1 the electrification infrastructure was also a major factor. Obsolete technology together with obsolete train operations and control have made it impossible to obtain the real and urgent capacity expansion benefits of electric technology. 29. An isolated factor that affected the electrification component quite substantially was the attempt early in project implementation to introduce a new and untried 2x25 kv AT electrification system on the 580 km sections 9 and 10 (Bina Katni-Bilaspur) instead of the existing 25 kv BT system. The new system has been used successfully in France, Japan and Russia, has certain very real advantages over the 25 kv BT system and is fully compatible with it. The Bank approved this technical change and a contract for US$ 2.5 million was awarded to a consortium of consultants who submnitted their initial recommendations in early 1989. In mid 1990 when the bids for this section were opened the costs were very much higher than expected. This was the section that ultimately experienced the 229 percent Rupee cost overrun, consumed 34 percent of the electrification resources, and was finally energized only in mid 1994. However, it is logical for the Indians to eventually convert to this system so the extra cost and time for this investment may be justified as a form of necessary R&D. V. Project Sustainability 30. The workshop modernization objectives have generally been achieved and the availability figures for locomotives and rolling stock are good even by developed country standards (See Table 4 in Annex I). The same objectives could have been achieved at much less expense if workshop rationalization had been possible. In the Indian political climate this was not possible but there is no reason to believe, even under an unrationalized system, that the rolling stock and locomotive availability figures will not be sustained. 31. For the OIS and electrification component the question of sustainability of the original objectives does not arise because the objectives were not achieved and are not likely to be achieved any time soon. The experience with the entire TEK series, culminating in the aborted 1992 TEK V project, which was to again pursue these objectives, is sufficient proof of the difficulty of obtaining and sustaining the benefits of state- of-the-art rail technology in India. VI. Bank Performance 32. Bank performance was uneven as project preparation proceeded over several years. In June 1980 a Project Brief for an electrification project was prepared in which the issues of the obsolete electric locomotive technology, inadequate manufacturing capacity and the inadequate supply of domestic electric power were raised. The sole benefit of electrification mentioned at this point was energy and foreign exchange saving. In 1980 and 1981 world petroleum prices were at their all time high and the emphasis on substituting domestic coal-based energy for imported petroleum-based energy is understandable. Nevertheless, the domestic power supply was so irregular that there was even then discussion of Railway owned and operated power plants - vertical integration by IR into the troubled power sector. The proposal was not accepted but the fact that it was considered indicates the serious nature of the problem. 33. Comments on the Project Brief did raise questions concerning the exclusive energy saving approach and the crude rule-of-thumb measures being used (MGTKMIRKM). The point was that, depending on the  12 circumstances, the breakeven point for electrification was anywhere between 20 and 100 MGTKM/RKM and that Indian rail traffic was nowhere near the 100 MGTKM/RKM density.'3 There is no response to this criticism in the file. 34. In August 1980 a letter from the Bank to the Government pointed out the electric locomotive technology and manufacturing constraint and recommended that the manufacturing operations of IR be corporatized so that technology transfer and electric locomotive output could proceed smoothly as a manufacturing issue rather than a rail transport issue. The latter recommendation was not accepted and project preparation proceeded. 35. Appraisal took place in January 1981. In the March 1981 Issues Paper there was no discussion of economic justification or the shortage of electric power in the country. Only the locomotive technology and the OIS system was highlighted. In the Decision Memorandum it was agreed that the key assumptions for justification and the question of availability of electric power would be treated in the Staff Appraisal Report. Subsequently in April 1981 doubts were raised in an internal memorandum about the availability of adequate electric power and the need to extend the high voltage network to serve IR. If this were the case the costs should be included in the electrification project. There is no response to this criticism in the file. 36. By this time the Bank was committed to the project without really examining the economic justification, in particular the impact of electrification on train operations in the high density corridors. In October 1981 IR's essentially energy saving approach to analyzing electrification investments was finally given approval by the Bank. At this time the petroleum commodity price index stood at 119.3 up from 56.0 in 1978. 37. Preparation dragged on because of the failure to move on the OIS system. In April 1983 a post appraisal mission was mounted largely because of the addition of the workshop modernization component to the electrification and OIS components. In July and August of 1983 a supplementary Issues Paper was followed by a Decision Memorandum. The petroleum price index stood at 101.8 and there was no trouble in justifying the electrification investment primarily on energy savings. Workshop modernization was discussed in depth but rationalization was mentioned only in the context of a study. No mention was made of the rationalization experience of TEK I. 38. By February 1984 the Yellow Cover SAR was circulated and the main concern was for progress on the OIS system. Two of the conditions for loan approval were the guarantee by the government that there would be adequate electric locomotives and electric power for the railroads on a priority basis. Two of the outstanding concerns critical to project success were to be dealt with through loan conditions. These were subsequently dropped at negotiations. The project was finally made effective in May 1985, almost four years after initial appraisal. 39. In November 1985 the first supervision mission's Back-to-Office report was submitted. At this time Bank staff predicted a serious shortfall in electric locomotives for the electrified system. The estimate was 150-200 locomotives lacking by 1990. This was a prediction that eventually came to pass. 40. In 1986 world oil prices collapsed and the price index of petroleum dropped to 42.0 from levels in excess of 100 in the early 1980s. An internal Bank memo of March 1986 indicates that this fall in petroleum prices, if sustained, would reduce the IRR of three of the ten sections to be electrified to levels below 7 13 The Bank's own internal study on rail electrification world wide did not appear until 1984. See: Railways and Energy, by Liviu Alston, World Bank Staff Working Paper No. 634, March 1984.  13 percent. These price levels have been sustained through 1993 and are expected to rise only gradually thereafter. In retrospect it was easy to justify electrification at fairly low traffic densities with the high petroleum costs of the early 1980s and there was no need to go in search of the more difficult to obtain benefits of electrification. Also, while there was congestion there were no serious theoretical capacity constraints on the main lines in 1981 at the time of appraisal. Bank staff had no incentive at the time to simulate the train operations on the sections to be electrified. 41. During the entire period of the TEK series the domestic power supply in India has been a serious problem, restricting the growth of private industiy which has been willing to pay almost double the long run marginal cost of providing electric power based on the price they are paying for in-house power generated with diesel engines.14 IR has been given priority for the high value uninterrupted power provided by the State Electricity Boards (SEBs). The very first memorandum in the file by Bank staff in June 1980 stressed the problem of inadequate electric power but the issue was never addressed except to insist that IR be given priority in the queue. On this issue Bank staff can be faulted since the alternative use of the electric power in terms of foregone industrial production should have been considered in the economic analysis. The fact that energy efficient diesel locomotives have subsequently been developed that very nearly preclude any energy saving with electrification is something that Bank staff could not have anticipated at the time. Nevertheless, the wisdom of rail electrification in an environment of chronic power shortage was, and remains, in question. 42. The lack of progress on the workshop rationalization because of labor and political resistance was well known to Bank staff at the time. The second best approach of providing machine tools to redundant shops, while less efficient, is realistic and possibly justified given the beneficial impact on locomotive and rolling stock availability. Bank staff can only be faulted for not making this reality more explicit in the project documentation. Vil. Borrower Performance 43. Indian Railways is one of the largest vertically integrated transport-manufacturing conglomerates in the world. It operates under five major constraints that limit its ability to perform as a rational commercial enterprise might. First, it is directly under a Minister of Railways who is answerable on a day to day basis to Parliament. Second, the major procurement and investment decisions of IR are invariably picked up and discussed in the popular press. Third, the management of IR has extraordinarily short tenures in any position typically from six to eighteen months. Fourth, IR has a labor force of about 1.7 million of which at least 400,000 are not needed for current and future rail operations. Fifth, IR has become primarily a passenger railroad serving a literate and vociferous clientele who have become accustomed to underpriced (frequently free) service which has priority over freight. There is no countervailing lobby for better quality freight service. 44. While there was frequently a meeting of the minds between Bank staff and IR staff at the technical level, the agreements that were negotiated at that level were frequently reversed for high level political considerations. The refusal to import locomotive technology under TEK I is an example. The subsequent failure to transfer the OIS technology and the electric locomotive technology arose partly from the criticism 14 In 1991/92 India's overall electric power shortage was estimated at 8.5 percent and peaking power shortage about 17.7 percent  14 of the popular press which did not have more than a superficial grasp of the underlying techno-economic issues. For most uninformed observers the 1960 electric locomotive technology was adequate in that it was cheap and domestically produced. Similarly, India was producing both computers and software so, the argument went, there was no need to import the OIS system. While track capacity can easily be expanded in the main corridors by simulating train operations and revising the passenger and freight time table, both passengers and the press would object strenuously to tine table changes. Rationalization of workshops quickly ran into the problem of labor redundancies and facility closures. Finally, the frequent change in senior IR management made it necessary for Bank staff to revisit key decisions either because the new managers had different ideas or because they needed to be brought up the learning curve concerning complicated techno-economic issues that were thought, by Bank staff, to have been resolved. 45. Much of this organizational complexity and inertia was reflected in the Bank supervision reports under the general heading of "procurement problems". The floating of tenders and awarding of contracts together with subsequent implementation took much longer than is normally expected for such project components. This was then reflected in the three to five year delays in component implementation which hurt the economic returns on the investments. Part of this was difference of views between the Bank and IR regarding the form of acceptable bidding documents.15 Part was the fact that detailed specifications for the major components were not ready at the time of loan effectiveness. This factor combined with the constant changes in IR requirements reflected in requests for reallocation of loan funds. Probably most serious was the fact that any contract in excess of US$ one million (of which there were many) required the formal approval of the full Board and the Minister of Railways. This latter requirement routinely required three to seven months and in the case of more controversial items much longer. 46. IR management frequently could not or would not deliver on their more controversial commitments. In the TEK HI project under review here the failure to transfer the electric locomotive technology, manufacture sufficient new locomotives and operate the electrified sections in a manner that would yield the benefits of electrification was a major shortcoming in borrower performance. The form but not the substance of electrification is being implemented. Related to this issue is the failure to monitor and evaluate effectively the electrification program. Bank staff had a very difficult time collecting the data for reevaluating the performance of the project for this implementation review. Since electrification has been the center piece of the IR investment program for many years now it is disappointing to see that there has been no systematic attempt within IR to monitor and evaluate its actual progress and compare that with original expectations. The slow progress on the OIS initiative after 16 years of sustained effort is another disappointment as this is one of the most cost-effective investments conceivable for IR. 47. For the non controversial workshop modernization component borrower performance was much better as no vested interests were threatened and no single investment was large enough or technologically controversial enough to attract the interest of the press or public. VIII. Assessment of Outcome 48. Of the four categories of assessment the Bank considers for its completed projects the one under review was neither Highly Satisfactory nor Highly Unsatisfactory. In fact, the outcome for this project was somewhere between satisfactory and unsatisfactory. Some of the workshop components that were well supervised and had working management information systems (MIS) in place showed very satisfactory rates of return (25 percent in the case of Golden Rock). At least four of the six workshop modernization components had satisfactory rates of return. But, the workshop component was only 13 percent of total 15 This has since been resolved since a standard form acceptable to the Bank has now been adopted by IR.  15 completed project costs. The electrification component finally dominated the total project cost, accounting for 80 percent of expenditures. And, the rate of return on the electrification investments was likely to be in the area of 4 percent. The weighted average rate of return on the total project is therefore estimated at about 7 percent. For a Satisfactory rating the project must have: "....achieved most of its major objectives and have achieved or be expected to achieve satisfactory development results with only afew shortcomings." For an Unsatisfactory rating the project must have: "....failed to achieve most of its major objectives and is not yielding and is not expected to yield substantial development results, and has significant shortcomings. " With two of the six major workshop modernizations and virtually all of the dominant electrification investments unlikely to show satisfactory economic returns and very slow progress on the OIS system, the Unsatisfactory rating must, therefore, be chosen. IX. Future Operation 49. In retrospect rail electrification in India was probably premature by ten to twenty years. Given the chronic power shortages in the country, the absence of abundant low cost hydro power, the relatively low traffic densities in the congested rail corridors and the availability of energy efficient diesel technology, the best approach would have been to continue with improved diesel technology and emphasize optimal train operations through investments in communications, signaling and time table revision. However, the electrification investment is now a sunk cost and the relevant question is how to make the best use of the technology in the future. 50. Much depends on the performance of the Indian economy and the induced demand for intercity freight and passenger traffic for all modes - road, rail and air. This demand could easily double in the next ten years and there is every indication that the now dominant highway mode will not be able to continue to expand in the high density corridors as it has in the past. i6 The railway, with its unitilized capacity potential in the high density corridors, could then relieve the strain on the transport infrastructure and allow the economic growth to take place. Here the electrification investment would be extremely useful for reasons that were not considered in the original project preparation. Rather than energy and foreign exchange saving the future benefits would be the expansion of track capacity on a megacity quadrilateral that would be approaching the traffic densities experienced on the railways of China. But, to get these benefits the substance rather than the form of electrification will have to be pursued. This means state-of-the-art electric locomotives operating on track sections with near optimal train scheduling. 51. These objectives can be pursued now by importing the electric locomotives assembled and by simulating train operations in the congested corridors and revising the outdated timetables. The Asian Development Bank (ADB) has recently supported the train operations simulation studies so much of the information for new timetables is now available.17 The locomotives are the missing element. This presumes, again, that the necessary electric power will be available. 16 See: India- The Transport Sector - Lone Term Issues, Report No. 13192-IN, The World Bank, September 1994. 17 See: Study on Alleviation of Line Capacity Constraints, Indian Railways, October 1994.  16 52. The problem of domestic electric power supply will not be solved for many years though and it may be that modern, high horse power, energy efficient, high adhesion diesel locomotives will have to continue to run under wire to move the traffic. The problem with the modern diesels under high density mixed passenger and freight operations is that their acceleration characteristics are not as good as those of electric locomotives and that even under optimal train operations the same track will not have as large a capacity as it would under electric operations. This is particularly crucial as traffic approaches 100 MGTKM/RKM. Currently the saturated sections of IR are in the neighborhood of 50 MGTKM/RKM so modem diesels can still do the job and are anyway required for the rest of the nonelectrified system. 53. The major workshops are generally working as well as can be expected under a nonrationalized system. Rationalization awaits national economic reform efforts for State Owned Enterprises (SOEs) in general. Until a national exit policy together with a downsizing and disinvestment policy for SOEs is formulated, it is difficult to see Indian Railways taking the lead independently in this controversial area. 54. The OIS system is said to be ready now for implementation in the Northern Zonal Railway. If it is allowed to be introduced in substance as well as torm then the benefits should be rapid and obvious and the introduction to the other zonal railways highly desirable. The Northern zonal experience is an obvious one to monitor in the future. X. Key Lessons Learned 55. A series of techno-economic lessons have emerged from the TEK 11 experience and its larger context - the entire TEK lending series. The most obvious is that it makes no sense to extend electrification beyond the existing 10,000 km core system. The traffic densities are not there, the power supply remains questionable and the energy and foreign exchange savings are no longer an argument. 56. The need for technology transfer for diesel and electric locomotives as well as light weight high capacity wagons is greater than ever. The Bank's argument for "unbundling" the manufacturing entities of IR is increasingly relevant for the technology transfer. The resulting corporate entities would be legally able to form joint venture manufacturing operations with foreign companies who would make the designs and future upgrades available for domestic and foreign sales. The politics of technology and procurement would no longer be central to the railway and the railway management could concentrate on transportation rather than manufacturing and R&D. In light of the successful unbundling of the Container Corporation (CONCOR) from the IR conglomerate, it would seem that the IR management would be more open to the possible benefits to be gained from corporatization. 57. As computerization and computer literacy spreads throughout IR the possibility of finally adopting and making use of a real time OIS system increases. The Bank's argument for an OIS system continues to be relevant. Much of the development has already taken place over the last 16 years. It remains to be implemented to an extent which will allow the potential benefits to begin to accrue. 58. IR badly needs to begin using state-of-the-art network flow models and corridor simulation models in their planning. There is a very large payoff in congestion reduction in the megacity quadrilateral simply from developing new time tables for freight and passenger trains. Current time tables are many years old and have been subjected to incremental changes over the years that are far from optimal. The ADB has been assisting IR in this area. The Bank has been doing the same in the area of network flow models with free standing technical assistance financed by a Japan grant. This is the Long Range Decision Support system initiative. These are the same models that the Chinese have used successfully in getting the most out of their intensively used rail system. As traffic increases, gauge conversion progresses and the network effects of  17 rail investments and operations increases, these planning tools, which are now standard features of most large developed country railways, need to be adopted and used. 59. Finally, it is clear that none of these techno economic lessons are likely to be taken on board by an IR management that typically serves no more than 18 to 24 months (and frequently less) in their positions before moving on or retiring. Under the crucial TEK II loan operation the Project Completion Report points out that during project implementation there were no less than six different managers in the key technology transfer position within IR. 60. Beyond the techno-economic lessons that emerge from TEK m1 and the TEK series there is a larger set of lessons for the Bank and its approach to rail lending in India. From the point of view of project ownership and implementation there are two broad types of IR project components: the benign and the controversial. The benign components are the ones that the traveling public, the press, politicians, and organized labor see as beneficial or at least not threatening. Over the 16 years of the TEK series these are seen to be the importation of spare parts, workshop modernization, new industrial operations, electrification, and the conversion of meter and narrow gauge track to broad gauge. More currently there is support for (or lack of opposition to) corporatized import-export container operations and the upgrading of suburban rail passenger services. The common element of the benign components is that they have no impact on labor except an expansion, do not threaten the notion of national R&D self sufficiency, and require no fundamental change in current train operations. 61. The controversial components are the importation of new technologies for diesel and electric locomotives, the importation of new designs and materials for rolling stock, the importation of OIS technology, the rationalization of workshops and the corporatization of manufacturing operations. The common element of the controversial components is that they have a substantial impact on labor (and management), explicitly recognize that domestic R&D is not adequate to produce an appropriate technology, and require fundamental changes in current train operations. 62. The electrification component of TEK m was benign in form but highly controversial in substance. While the electrification infrastructure received general support and was implemented, the necessary new locomotive design and revised train operations were not forthcoming. The OIS system for which the investment is the least and the documented payoff the greatest has also failed to make rapid progress because it violates the R&D self sufficiency constraint, the impact on labor constraint and the impact on train operations constraint. Only the workshop modernization component of TEK m1 was benign. 63. In the absence of reform inducing financial crisis or exogenously imposed reforms supported by the highest level of government, it is difficult to see a convincing argument that the controversial investments will be implemented. The controversial ones are likely to be frustrated even if supported at the technical managerial level in the railways. The problem is that the controversial components, generally speaking, have the highest economic payoff for the railway and the national economy. As long as this situation obtains it is difficult to envision comprehensive Bank lending to IR any time soon. It has been given a serious try with US$ 1.2 billion in five operations over 14 years and we would be remiss if we failed to learn the lessons of the experience.  18 ANNEX I Table 1 Summary of Assessments A. Achievement of objectives Substantial Partial Negliiibie Not aoWlicable Macroeconomic policies (O) (Q) Sector policies El E El Financial objectives al ol ol Instirutional development Q Q/ a o Physical objectives E, O Poverty reduction Oa Gender concerns a7 o a Other social objectives al a ol Environmental objectives a o o Public sector management a Q El Private sector development a o o Other (specify) l al Q B. Project sustainabilitv Likeiv Unlikelv Uncertain (A3 (1)(4 Hizhlv satisfactorv Satisfactorv Deficient C. Bank ce-rormance (,/) () (/) Identiifcation C ] Preparation assistance a Appraisal O/ Supervision Ol Riehiv D. Borrower performance satisfactorv Satisfactorv Defic:ent (A (A (A Preparation Implementation a o Covenant compliance 0 E a Operation (if applicable) 0 0 Hizhly Highlv E. Asscssmerm .f nt come satisfacrorv Satisfactorv Unsatisfactorv unsatisfactory () () ( (A) 0 0 e 0  19 Annex I Table 2 Project Timetable Steps in Project Cycle Date Planned Date actual/te Identification N/A 4/80 Preparation N/A Appraisal 1/81 1/81 Negotiations 9/81 4/84 Board presentation 12/81 5/84 Signing 5/84 5/84 Effectiveness 7/84 5/85 Project Completion 9/90 3/93 Loan closing 1/91 7/93 Table 3 Loan/Credit Disbursements: Cumulative Estimated and Actual (US$ millions) FY: 1985 1986 1987 1988 1989 1990 1991 | 1992 1993 Cumulative Appraisal 12 75 175 240 270 281 281 281 281 Estimate Cumulative Actual 1 11 19 80 150 195 245 261 272 Actual as % of 8 15 11 33 56 69 87 93 97 Estimate Date of Final Disbursement: July 31, 1993  20 Annex I Table 4 Key Indicators for Project Implementation and Operation ~~~~~~~Actual|Pln|ll FY 801 811 82 83 841 85 86 87 88 891 9 909 91 92 93 Locomotive Diesel *e ** 84 82 82 82 82 88 89 90 90 82 90 90 90 Availability (%) Electric *e ** 179 76 76 77 79 88 91 92 93 84 93 93 93 Locomotive Diesel 630 610 671 664 696 746 764 739 766 759 702 725 673 633 647 Utilization Pagr _ (km per available Electric 452 453 453 457 428 438 448 458 464 482 513 480 482 488 502 locomotive-day Pagr I _ I I Diesel 307 303 347 364 358 407 416 427 457 457 454 400 445 436 426 Freight _ _ _ __ _ __ _ _ _ _ _ _ __ _ _ _ _ _ _ _ Electric 289 274 355 380 396 385 395 400 410 405 395 420 398 395 412 Freight __ __ _ _ __ _ __ _ _ _ _ _ Wagon Availability 96 95 93 93 93 94 95 95 95 96 96 96 96 96 96 Wagon Utilization (Net-ton- 972 896 1112 1123 1112 1150 1296 1420 1449 1453 1420 1450 1407 1439 1457 km per wagon-day) _ Coach Availability (%) 86 88 84 85 86 87 87 90 91 91 90 91 91 91 91 Gross Trailing Load per 1694 1721 1780 1766 1787 1843 1955 1999 2050 2054 2094 2000 2122 2191 2238 Freight Train (Tons) _ . Net Load per Freight Train 863 884 911 896 892 922 1001 1031 1053 1042 1060 1100 1079 1119 1128 (Tons) ___ _ Traffic Units per Employee 229 234 251 255 252 255 277 298 310 304 315 285 326 339 339 (1,000's) ALL GAUGES _ _ I * Also see Tables 2 and 3 in the text, and Annexes II and III ** Not available on a compatible basis  21 Annex I Table 5 Project Costs and Financing Appraisal Estimnate* Actual / Local Foreign Total Local Foreign Total Rs. million $million** Rs. million $miRlion*** Electrification of 7,174 1,444 8,618 798 15,726 786 main line segments C Miscellaneous 33 70 103 10 103 5 0 Training 24 13 37 3 37 2 S Subtotal 7,231 1,527 8,758 811 15,866 793 T Workshops and ICP 1,013 840 1,853 172 2558 128 S Depots 1,204 1,228 2,432 225 1487 74 Training 18 31 49 4 49 3 Subtotal 2,235 2,099 4,334 401 4,094 205 Total 9,466 3,626 13,092 1,212 19,960 998 Govt India 931 726 Financing World Bank 281 272 Total 1212 998 * Includes physical and price contingencies ** Using the 1984 exchange rate of $US1.0O = Rs. 10.8 *** Assuming a weighted average exchange rate for the period 1989-1993 of $US1.0O = Rs. 20.0  22 Annex I Table 6 Status of Compliance with Principal Covenants and Other Undertakings Covenant Status of Compliance Maintain passenger fares and freight rates, and In compliance for FY86 through FY92. take all other action necessary or appropriate, to enable the Railway to meet annually out of internally generated funds al operating expenses and dividend payments on capital-at-charge. (L. A. 4.03) Ensure that provisions to the Depreciation In compliance. Reserve Fund for fiscal years 1985 through 1989 shall be at least equal to the fiscal year 1984 provisions. (L.A. 4.04(a)) Other Undertakings Status of Compliance Complete a Maintenance Standards Study by Was completed in 1991. October 31, 1985. (Minutes of Negotiation, para. 6 and Annex I). Develop and install a computer based management Basic MIS has been installed in all six workshops, information (MIS) system at each of the six but some modules are not fully operational in project workshops by March 31, 1987. (Minutes some workshops. (See Table 2 in text.) of negotiations, para 29). Furnish to the Bank by July 31, 1984, an Was provided. approved detailed training program. (Minutes of Negotiation, para 22).  23 Table 7 Bank Resources - Staff Input (Staff weeks) TASK FY83 FY84 j FY85.| FY86 FY87 J FY88 I FY89| FY90| FY91 FY92 1 FY93 FY94 1 FY95 TOTAL Preappraisal .3 12.1 12.4 Appraisal 36.3 46.2 82.5 Negotiations 11.2 11.2 Supervision .9 8.0 24.6 15.3 16.5 21.8 15.2 7.0 4.7 8.5 4.6 127.1 Other 5.3 = 2.2 12.2 19.7 Total 42.0 70.3 8.0 24.6 15.3 16.5 21.8 15.2 7.0 4.7 10.7 16.8 252.9  24 Annex I Table 8 Bank Resources - Missions Month/ Number of Staff Days Specialization Performance Types of Stage of Project Cycle Year Persons Field* Represented Rating System Problems Through appraisal (1/81) 4/80 5 2 1/81 7 7 Appraisal through Board 4/81 4 4 E,T approval (5/84) 9/81 3 4 E,T 5/83 6 12 E,F,T 11/83 1 3 T Board approval through 2/85 4 3 E,F,T effectiveness (5/85) 5/85 5 4 E,F,T Supervision 1 9/85 6 5 E,F,T 2 I,T,P,F 2 4/86 4 6 E,F,T 2 I,T,P,F 3 10/86 2 3 F,T 2 T,P,F 4 11/86 2 6 F,T 2 T,P,F 5 2/87 1 2 T 2 T,P 6 5/87 1 6 T 2 T,P 7 2/88 3 6 E,T 2 I,T,P 8 10/88 4 4 E,F,T,A 2 I,T,P,F 9 1/89 4 4 E,T,A 2 I,T,P 10 7/89 3 2 E,F,A 2 I,T,P 11 10/89 3 5 F,T,A 2 T,P,F 12 5/90 5 8 E,F,T,A 2 I,T,P,F 13 12/90 3 5 E,F,T 2 I,T,P,F 14 11/91 4 7 E,F,T 1 I,T,P,F 15 2/92 4 4 E,F,T 1 I,T,P,F 16 12/92 3 11 T 1 I,T,P 17 3/93 3 19 T 1 1,T,P Specialization: E =Economist; F =Financial Specialist; T =Technical Specialist; A = Administrative Status: 1 =Problem free or minor problems; 2=Moderate problems; 3=Major problems Type of Problems: I=Institutional; T=Technical; P=Procurement; F=Financial *Estimate for this project only, since most missions concerned two or three projects  25 Annex II ANNEX II RETROSPECTIVE ECONOMIC ANALYSIS OF WORKSHOP MODERNIZATION COMPONENT INTRODUCTION 1. The project included a Workshop Modernization component covering the modernization and expansion of six workshops, the Integral Coach Factory (ICF) and 37 running maintenance sheds/ depots. 2. At appraisal, the total cost for modernization of six workshops and ICF was estimated at Rs. 1,852.8 million (including contingencies). The final cost is Rs. 2,555.0 million. 3. Table I shows the "Cost and Results of Workshops Component." It is noted that implementation of modernization took much longer than anticipated. Sub-components that were planned for completion by June 1988 have actually been completed in 1993 and 1994. 4. The project implementation of ICF, Golden Rock and Ajmer Workshops has been studied in detail and findings are contained in the following paragraphs. For each sub-component an economic evaluation has been attempted. Basic details, which formed the input for the economic evaluation for the three units are also attached. ICF 5. The project objective was to increase the capacity for manufacture of passenger coaches from 850 to 1000 coaches per year and included procurement and installation of 165 machines and the expansion of covered area for manufacturing activity. Against the appraisal cost of Rs. 540.6 million (including contingencies), the actual cost was Rs. 678.8 million. The project implementation took nine years against the original estimate of 3-4 years. 6. The project objective was realized during year 1990-91 with the manufacture of 1013 coaches. Thereafter, the yearly production has been consistently above 1000 coaches. The increase in coach production has been accompanied by increased manpower productivity which improved from 6.65 direct workers per coach in 1986-87 to 5.6 in 1993-94. Economic Evaluation 7. At appraisal, the economic evaluation projected a 20 percent rate of return for this sub- component. The detailed basis of economic evaluation at appraisal is not available. A revised IRR was estimated on the basis of costs. outputs and time frame as projected at appraisal and yielded a return of 7.8 percent when no increase in staff productivity was considered. An economic evaluation based on actual cost, implementation period and output shows an economic rate of return of 13.2 percent. However, if the increase in staff productivity is attributed to the project an IRR of 17.7 percent is obtained. The actual rate of return is higher perhaps due to the increase in coach production exceeding the target projected at appraisal and the fact that a substantial increase in output was realized from year two onwards.  26 Annex 11 Table 1 Costs and Results of Workshop Component Parel J Liluah Jagadhri Golden| Kharagpur Ajmer ICF Total l l l ~~~~~Rock |ll l Appraisal cost (Rs. million) 161.1 142.3 1880.6 280.9 311.9 235.4 540.6 1,852.8 Final cost (Rs. million) 217.8 210.6 450.0 275.4 428.2 297.1 678.7 2,558.0 Appraisal estimated Jun 88 Jun 88 Jun 88 Jun 88 Jun 88 Jun 88 Jun 88 completion date Current/actual Mar 93 Sep 93 Dec 93 Mar 93 Dec 93 Mar 93 Mar 94 completion date No. of Loco POH DE DH DE EL DE Actual at appraisal 72 24 77 0 36 Appraisal target 108 48 144 48 72 Achieved to date 108 48 120 58 60 No. of Coach POH EMU MG NEW (four wheeler units) Actual at appraisal 3,600 3,000 0 2,600 850 Appraisal target 4,800 3,600 96 108 3,600 1,000 Achieved to date 4,260 3,000 96 102 2,600 1,016 No. of Wagon POH (four wheeler units) Actual at appraisal 13,000 12,000 Appraisal target 13,600 18,000 Achieved to date 13,600 12,720 f Time required, loco POH, days Actual at appraisal Appraisal target 23 45 23 47 Achieved to date 18 25 19 18 18 29.1 19 24 Time required, coach POH, days Actual at appraisal Appraisal target 24 21 15 17 21 Achieved to date 19 14/12 13 16 14 18.8 17.1 13 16.3 18.0 Time required, wagon POH, days Actual at appraisal 10 6.9 8 Appraisal target 8 4 6 Achieved to date 6.1 6.9 6 Staff employed At appraisal 5,886 11,175 5,442 7,176 15,344 8,340 15,547 Currently 5,331 11,227 5,177 6,115 14,068 7,763 15,691 Management Information System, computer modules implemented Target 7 7 7 7 7 7 Achieved to date 5 7 4 7 6 5 l .. =.  27 Annex II GOLDEN ROCK 8. The project objectives of modernization include increasing diesel locomotive POH capacity from 72 to 144 units per year, reduction on POH cycle time for locomotives from 23 to 19 days and for coaches from 15 to 13 days and increase capacity for rewinding of traction machines. Against the appraisal cost estimate of Rs. 280.9 million (including contingencies), the actual cost was Rs. 275.4 million. The project implementation was completed over nine years against the original estimate of 3-4 years. 9. The increase in diesel loco POH and reduced POH cycle time were realized gradually and by 1990-91, 120 loco POH and cycle time of 20 days were achieved. Further increase of POH has not taken place due to staff constraints and non-availability of workload. The POH cycle time has improved further to 17.7 days in 1993-94. The electrical machine rewind has increased from 144 in 1985-86 to 378 units in 1993-94 against a project target of 374. The POH cycle time for coaches improved from 15 to 13 days by 1992-93 but has slipped back to 15 days in 1993-94. Economic Evaluation 10. At appraisal the economic evaluation estimated a 27 percent IRR for this sub-component. As the basis of the economic evaluation carried out at appraisal is not available, the same was carried out again with costs, outputs and time frame as projected at appraisal. This has yielded an IRR of 50 percent. 11. An evaluation based on actual cost, implementation period and outputs has also been carried out and shows an economic rate of return of 24.7 percent. AJMER 12. Ajmer Workshop currently handles meter gauge (MG) stock only. The project objectives of this sub-component include increase in POH of diesel loco from 36 to 72 per year and that for coaches from 2664 to 3600 (four wheel units) per year and reduction of POH cycle time for diesel from 47 to 18 days, coaches from 21 to 14 days and wagons from 8 to 6 days. Setting up of electrical traction machine rewinding capacity for 218 units per year was also planned. Against the appraisal cost of Rs. 235.4 million the actual cost was Rs. 297.1 million. The project implementation took eight years against the original estimate of four years. 13. The increase in diesel locomotive POH to 60 units per year was realized in 1990-91 and has ranged between 60 and 63 in the subsequent years. It was understood that although the workshop had developed capacity for 72 POH per year, it did not have the workload to utilize this capacity in full. There has not been any significant increase in the coach POH which was 2580 in 1986-87 and stood at 2606 in 1993-94 after declining to 2244 and 2160 in 1989-90 and 1990-91 respectively. The POH cycle time for diesel locomotives, coaches and wagons has improved to 22.5, 18 and 6 days respectively against respective targets of 18, 14 and 6 days. The facility for electrical machine re-wind has been setup but the initial production is expected in 1994-95.  28 Annex II 14. At appraisal the economic evaluation projected a 24 percent IRR for this sub-component. As the basis of the economic evaluation carried out at appraisal is not available, the same was carried out again with costs, outputs and time frame as projected at appraisal. This has yielded an IRR of 29.4 percent. 15. An evaluation based on actual cost, implementation period and outputs has also been carried out and shows an economic rate of return of 1.9 percent. A much lower actual rate of return as compared to IRR anticipated as per projections at appraisal is attributed to longer implementation period, cost over-run and delayed and only a partial realization of post-project outputs. 16. It is anticipated that with large scale gauge conversion, there shall be a significant decline in the workload for Ajmer Workshops. 17. It may be noted that the IRR obtained for the sub-components on the basis of actual data is strictly not comparable with the IRR obtained at appraisal, since the basis for the two evaluations is not identical. BASIS OF ECONOMIC EVALUATION COSTS 18. All project costs have been considered. To arrive at economic cost, where available, custom duties were reduced from the financial cost. Where data on custom duties was not available, 80 percent of financial cost was assumed as the economic cost. 19. Additional working capital required to be injected to increase production was considered as a cost at the appropriate time. 20. Maintenance cost of additional assets employed was considered at 2.5 and 5.0 percent for civil engineering and other assets respectively. BENEFITS 21. The benefits were reduced by 20 percent to arrive at economic level of benefits. Residual value of assets after 15 years was estimated on the following basis: Civil Engineering assets 50% Mechanical, Electrical and Telecom assets 10% 22. This value was considered as a benefit in the 16th year. Return of additional working capital was considered as a benefit in the 16th year. Net benefits were arrived at by reducing the annual operating costs from benefits. Benefits from increased output were estimated as follows:  29 Annex II 23. ICF: If the capacity for coach building at ICF was not enhanced, IR would have purchased the additional coaches from other Indian coach builders at a cost higher than the cost of coaches built at ICF. Bharat Earth Movers Ltd. (BEML) has been supplying BG coaches to IR for several years and BEML price is a reasonable indicator of the cost of coaches from at alternative source. It was noted that BEML prices less taxes were about 25 percent higher than cost of coaches manufactured at ICF. Product of additional coaches built at ICF as a result of capacity expansion and difference of BEML and ICF coaches provides one benefit streams. The expansion of coach building capacity and attendant process improvements have lead to increase in man-power productivity, thus reducing labor cost component of cost of coach production. Manpower cost savings for 850 coaches per year (pre-project capacity of ICF) provided the second benefit stream for the economic evaluation. For this analysis labor productivity improvement over base year 1989-90 was considered as ICF achieved 1000 coach production in 1990-91. 24. GOLDEN ROCK: Benefits for this sub-component have been estimated for additional out-turn of diesel loco POH and the reduced cycle time for POH. Benefit due to additional POH was assessed by reducing repair, maintenance and sinking fund costs as well as the operating costs from the value added during POH. Value added for POH was estimated as recovery of depreciation of locomotive over a six year period assessed as per IR General Code Volume 1 (Annex IV para 719). In case of diesel locomotives, it translated into about 25 percent of the cost of a new locomotive. Benefit for each year was assessed on the basis of additional POH carried out and current loco cost. The one time benefit due to reduced cycle time for POH was assessed by considering the total locomotive days saved converted to locomotives saved and cost of locomotives. 25. AJMER: The benefit for Ajmer Workshop modernization was assessed on lines similar to that for Golden Rock except that besides the benefits of additional POH and one time saving of locomotives due to reduced cycle time, the one time benefit of reduced cycle time for coach POH was also considered.  30 Annex II BASIC DETAILS ICF WORKSHOP Covenant Actual COST 540.6 678.7 (Rs. million) Implementation period 3 years (1984/85 to 1986/87) 9 years (1985/86 to 1993/94) BENEFITS Increase in coach production by Coach production increase 150 units per year from 1987-88 as follows: 1987-88 127 88-89 148 89-90 75 90-91 163 91-92 166 92-93 173 93-94 175 IRR At appraisal 20% Revised as per original 7.8% With staff 17.7% projections productivity Excluding 13.2% staff productivity GOLDEN ROCK WORKSHOPS Covenant Actual COST (Rs. million) 280.9 272.6 Implementation Period 3 years ( 1985/86 to 1987/88) 9 years (1985/86 to 1993/94) BENEFITS 1. Increase diesel loco POH 72 to 144 = 72 48 (since 1990/91) 2. Reduce POH cycle time (days) Diesel loco 23 to 19 17.7 Pass coaches 15 to 13 13 (1992/93) 15 (1993/94)  31 Annex II 3. Increase re-wind (units) 144 to 374 = 230 378 IRR At appraisal 27% Revised as per original 50% 24.7% projections AJMER WORKSHOPS Covenant Actual COST 235.4 297.1 (Rs. million) Implementation Period 4 years (1986/87 to 1989/90) 8 years (1986/87 to 1993/94) BENEFITS 1. Increase POH per year Diesel loco (MG) 36 to 72=36 24 (1990/91 onwards) Pass coaches (4WU) 2664 to 3600=936 Nil 2. Reduce POH cycle time (days) Diesel loco 47 to 18 22.5 Pass coaches 21 to 14 18 Wagons 8 to 6 6 3. Electrical Rewind 218 units/year production expected from 1994/95 IRR At appraisal 24% Revised as per original 29.4% 1.9% projections  32 Annex III ANNEX III RETROSPECTIVE ECONOMIC ANALYSIS OF ELECTRIFICATION COMPONENT 1. The overall economic return of the Jhansi-Itarsi route electrification project was estimated at 23.5 percent and that of Balharshah-Vijaywada route at 40.5 percent at the pre- appraisal stage. The project was envisaged to be completed by 1988-89 and thereafter freight as well as passenger traffic were assumed to be hauled by electric traction. The main benefits of the project related to: (i) higher locomotive kms per day per locomotive in use, (ii) a substitution ratio of electric versus diesel locomotive of less than one, (iii) lower maintenance cost and higher locomotive availability, (iv) higher average speed and (v) savings in energy cost etc. 2. The basic objective of this loan project as a whole was to increase capacity to meet growing demand. However at the pre-appraisal stage the benefits on this count seem to have not taken into consideration and hence not quantified for inclusion in the benefits stream. Besides, the benefits due to savings in coaches have not been considered because of assumption relating to no change in speeds with and without the project scenarios. 3. Electric traction on the Jhansi-Itarsi route partially commenced in 1989 and even in 1992-1993 only 57 percent of freight traffic were carried on electric traction and the rest by diesel engines. As regards passenger traffic the percentage share were 90:10. Discussion with the zonal railways officials indicate that there was a gradual phase out of diesel operation from 42% in 1989-90 to 57% in 1992-93. It is also felt that the maximum percentage of 90% will be achieved by 95-96. Due to strategic reasons it is not considered advisable/desirable to go beyond 90% of freight traffic. In the case of Balharshah-Vijaywada route also the build up of electric traction has been gradual and by 1993-94 85% freight and 100% passenger trains were operating with electric locomotive. 4. The revaluation of economic benefits and estimation of rate of return are tentative as the benefits of the project have just now started to appear. It may be mentioned that the project completion report for these electrification routes has not been prepared and the required performance indicators data/information have not been compiled. For the evaluation, we have therefore used the relevant data of the concerned divisions of the Central and South Central Railways. Besides, the main basic assumption of the project viz. use of new 6000 hp electric locomotives on the electrified route has not been implemented. It was also given to understand that inadequate availability of electric locos, maintenance infrastructure and staff were the main reasons for gradual build up of traffic to be hauled on electric traction. This is considered to be one of the important and crucial constraints which have significantly affected the rate of return of ex-post evaluation stage. This and the other system constraints, outside the control of the project implementing authority, have resulted into very low rate of return at the re-evaluation stage. 5. Due to the non availability of all the assumptions relating to main parameters/ operational indicators used for estimating incremental benefits under "with and" "without" project scenarios at the pre-appraisal stage, an attempt has been made to estimate the rate of return on the basis of actual performance of the identified parameters. For this a suitable program was developed and used after pretesting on the basis of pre-appraisal data for Jhansi-  33 Annex III Itarsi section available in the SAR. The resultant estimate of economic rate of return may not be strictly comparative, with that of the appraisal stage. It however indicates the impact of non realization of assumptions on the viability analysis. The analysis does clearly bring out the importance of ensuring achievement of projected benefits so as to obtain the estimated rate of return. The ex-post evaluation also questions the validity of some of the assumptions under the existing system constraints as well as need to ensure simultaneous investment in the rolling stocks viz electric locos and human resources development for achieving desired benefits. 6. For the meaningful analysis of the returns estimated at appraisal stage and ex-post evaluation stage a comparative statement of basic assumptions made for estimating costs and benefits at the appraisal stage with the actual performance of the same during 1989-90 to 1992- 93 is crucial. The following statement gives the comparative analysis. COMPARATIVE STATEMENT JHANSI-ITARSI RKm 381 COVENANT ACTUAL a. Cost of construction Rs.783 million (EC) Rs. 1110.72 million (Rs. 947.66 at 1984 prices) b. Phasing of capital cost four year - 84.85 to 7 years to 87.88 85.86 to March 92 C. Rly c. Electric traction to commence in 1989 and Project partially commenced in 1989. and 100 percent traffic to be hauled on the Gradual percentage increase in freight traction by electric and electric and electric traffic, hauled by electric traction- locos. from 42% in 89-90 to 67% in 93-94. The percentage increase will be 90% in 96.97 for freight passenger: 90% from 90-91 onwards New Electric locos are yet no introduced. d. Higher locomotive kms per day per locomotive in use Freight Freight 1989/90 1992/93 Diesel 500 kms Diesel 460 kms 406 Electric 600 kms Electric : 596 kms 529 Passenger Passen2er Diesel 700 kms Diesel 855 742 Electric: 700 kms Electric 1702 818  34 Annex III e. Traffic Growth Rate Freight (per annum) Freight 1986/87 - 1990/91 : 2.73% 1986/87 - 1992/93 :1.5% 1990/91 - 2010/11 :2.64% Passenger Passenger 1986/87 - 1990/91 : 2.69% 1986/87 - 1992/93 : 5% 1990/91 - 2010/11 : 2.42% f. Speed of Goods Engine 89/90 92/93 Diesel : 24 kms/hr. Diesel : 26 28 Electric: 36 kms/hr. Electric : 28 29 g. Lower Maintenance Cost Diesel Rs. 3.3/kms Diesel Rs. 7.69/kms Electric Rs. 2.2/kms Electric Rs. 7.54/kms New Electric Rs. 1. I/kms New Electric N.A. h. Energy Consumption Per 1000 GTKM 89/89 92/93 Diesel 3.6 liters Diesel 3.7 liters 3.9 liters Electric N.A. Electric 12.13 kwh 12.95 kwh i. Energy Prices 88/89 92/93 H.S.D. Rs. 2.8/liter H.S.D. Rs. 3.7/liter Rs. 5.9 litre Electricity Rs. 0.4/kwh Elect Rs. 1.24/kwh Rs. 2.50/kwh j. Higher Gross Trailing Load per train (Goods) 88/90 92/93 Diesel : 1800 Diesel 1635 2209 Electric : 2400 Electric 2211 2262 New Electric: 3780 New Electric Not introduced k. Loco Economic Price 88/89 Diesel : Rs. 7.3 million 9.6 million Electric : Rs. 6.9 million 12.24 million New Electric: Rs. 14.7 million  35 Annex III 1. Maintenance Cost of Loco Diesel : Rs. 3.3/km Diesel Rs. 7.69/km Electric : Rs. 2.2/km Electric : Rs. 7.54/km New Electric: Rs. 1.1/km m. Cost of Running Staff: RS. 44,500 per loco n. IRR 23.4% 9.0% BALHARSHAH-VIJAYWADA Rkm 454 COVENANT ACTUAL a. Cost of construction Rs. 250.3 million Rs. 1343.6 million b. Phasing of Capital Cost Four years 1984/85 - 1988/89 1981-82 to 1990-91 (in the original F.R. 81-82-85-86) c. Electric traction to commence in 1989 and Electric traction hauled thereafter 100% freight and passenger traffic to be hauled by electric traction Freight Pass 1990-91 57% 70 1991-92 71.5% 80 1992-93 74% 90 1993-94 85% 100 1994-95 90% 100 d. Higher locomotive kms per day per locomotive in use Freigyht 93-94 92-93 Diesel : 450 km day Diesel 423 429 Electric: 550 km day Electric 418 421 Passenger Diesel : 560 km day Diesel 720 720 Electric 792 793  36 Annex III e. Traffic growth rate % Freight (GTKM) Freight 1986/87-1990/91 3.97 1986/87-1990/91 2.6 1990/91-2010/11 4.0 1990/91-2010 2.6 Passenger (GTKM) 1986/87-1990/91 2.53 1986/87-1990/91 3.2 1990/91-2010/11 2.42 1990/91-2010/11 3.2 f. Speed of goods train (km/hr) 88-89 92-93 Diesel 24 Diesel 22.3 21.9 Electric 36 Electric 24.1 21.3 g. Energy consumption per 1000 GTKM 88-89 92-93 Diesel 3.6 litres Diesel 3.13 2.4 Electric 11.1 kwh Electric 8.52 9.61 h. Energy Prices 88-89 92-93 H.S.D. Rs. 3.17/liter H.S.D. 3.85 5.96 Electricity Rs. 0.50/kwh Electricity 1.32 1.91 Energy cost per 1000 GTKM Diesel Rs. 21.55 Electricity Rs. 19.52 i. Higher Gross Trailing Load per train (Goods) 91-92 92-93 Diesel 1800 Diesel 1677 1698 Electric 2400 Electric 1612 1589 New Electric 3780 j. Loco Economic prices Diesel Electric (Rs. million) Diesel Rs. 7.3 million Electric Rs. 6.9 million 1989-90 13.4 16.7 New Electric: Rs. 14.7 million 1990-91 14.0 17.3 1991-92 15.4 19.3 1992-93 18.2 24.4  37 Annex III k. Maintenance Cost of loco (Rs. per engine km) Diesel : 3.30 1992-93 Diesel Rs. 7.97 Electric 2.20 Electric Rs. 7.07 New Electric 1. 10 1. Cost of running staff: Rs. 44,500 per loco m. Repair and maintenance cost of fixed installations OHE Rs. 4.602 Signaling Rs. 2720 Rkm 1990-91 1991-92 1992-93 Tele Rs. 5013 Rkm 11,296 12,858 13,437 Civil Engineering 3 % cost of n. IRR 40.5% 2.0% 7. The actual data for the two routes were gathered from the officials of the two zonal railways and the published monthly operating statistics for each division. It has to be borne in mind that data specific to each route was not available and therefore the divisional operating results had to be used for the analysis. In view of this the estimated returns at the ex-post evaluation stage may be treated as indicative. 8. As stated earlier, a program was developed and pre-tested for the evaluation of economic IRR. The actual performance data for the period 1987-88 to 1993-94 was then used to estimate the IRR of nine percent for Jhansi-Itarsi and two percent for Balharshah-Vijaywada routes. 9. The reasons for the lower actual rate of return on investment are analyzed in a subsequent section. However, the comparatively lower return of 2 percent vis-a-vis 40.5 percent at appraisal, for the Balharshah-Vijaywada route is attributed to practically no difference in operational performance indicators e.g. average trailing load, speed and engine utilization between diesel and electric traction. As stated in the appraisal report, the return is most sensitive to engine utilization. For this section, the electric engine utilization in 1992-93 and 1993-94 were marginally lower than for diesel against the projection that this would be the higher by 100 km/day.  38 Annex III ELECTRIFICATION OF JHANSI-ITARSI SECTION (CENTRAL RAILWAY) 10. Jhansi-itarsi section (381 route km) is part of IR's trunk routes connecting Delhi with Madras and Bombay. It is mainly a double line section except for small stretches of single line over some major rivers. This section has a high line capacity utilization (100-120 percent) and passenger trains constitute 53-59 percent of trains operated on this section. 11. Traffic levels projected and actual are as follows: TRAFFIC LEVELS TRAFFIC DENSITY Mil. GTKM/Annum Mil. GTKM/RKM/Annum Projected Actual Projected Actual 1981-1982 11,035 29.0 1986-1987 11,390 1990-1991 13,640 14,150 35.8 37.1 1992-1993 14,220 14,516 37.8 38.1 2010-2011 20,680 54.3 It is seen that traffic growth has been in line with projections considered during appraisal. 12. The ruling gradient of this section is 1/200 with the exception of 1/80 gradient between Midghat and Budni stations of Bhopal-Itarsi sub-section. 13. The Jhansi-Itarsi section was engerised in March 1989. This electrified double line section has seen about five years for the stabilization of change over from diesel to electric traction. 14. The following salient features are worthy of note regarding the post-electrification operation on Jhansi-Itarsi section. Build up of Operations with Electric Traction 15. The use of electric locomotives on this section has increased gradually from 42 percent freight trains hauled in 1989-90 to 67 percent in 1993-94. However 90 percent of passenger trains have been hauled by electric locomotives since 1990-91. 16. It was understood that with the induction of more electric locomotives in the forthcoming years, by 1996-97, 90 percent of freight trains shall also be brought under electric traction. 17. So far, existing 4000 hp WAG5 (freight) and WAM4 (passenger) electric locomotives are in use and projected new 6000 hp high horse power locomotives have not been acquired and deployed.  39 Annex III Electric Locomotive Usage 18. Locomotive usage for the years 1991-92 and 1992-93 for diesel as well as electric locomotives is given below: LOCOMOTIVE UTILIZATION km/day/loco in use Freight* Passenger** Electric Diesel Electric Diesel 1991-1992 459 408 763 758 1992-1993 526 406 818 742 Projection as per appraisal 600 500 700 700 * Average for Jhansi and Bhopal Division of C. Rly ** Jhansi Division Speed of Freight Trains 19. There is no significant difference in the speed of freight trains and between pre and post-electrification periods as can be seen from the following table: AVERAGE SPEED OF TRAINS (km/hr) (Average for Jhansi and Bhopal Division) Freight Electric Diesel 1988-89 28.0 26.0 1992-93 29.0 28.5 20. It was gathered that potential for higher average speed with electric traction could not be realized as this saturated section has a large proportion of passenger trains which makes extended runs for freight trains difficult. Lack of dynamic brakes on about half the electric locomotive fleet has also contributed to the non-realization of higher average speeds with electric traction Trailing Loads 21. By and large the section is operated with existing 4000 hp WAM4 and WAG5 locomotives. As the authorized loads for diesel (WDM2) and electric (WAGS) locomotives for freight trains are almost identical, there is little difference in the trailing loads hauled by diesel and electric locomotives. The following tables show the authorized and actual average trailing loads on freight trains hauled by the two modes of traction:  40 Annex III AUTHORIZED TRAILING LOADS (TONNES) FOR ELECTRIC AND DIESEL LOCOMOTIVES (BHOPAL DIVISION) Electric Diesel WAG5 WDM2 Itarsi-Bhopal (4W) 2200 2200 (Roller Bg) 2600 2400 (Box/BCN) 2850 2850 Bhopal-Bina (4W) 2200 2200 (Roller Bg) 2600 2600 (Box/BCN) 2850 2850 (Box N) 4750 4750 ACTUAL AVERAGE TRAILING LOAD ON FREIGHT TRAINS (4W/Tonnes) (Jhansi and Bhopal Division) Electric Diesel 1988-1989 77/2211 73/1962 1991-1992 77/2290 79/2190 1992-1993 80/2258 80/2210 22. The electric locomotives on this section haul 21 coach passenger trains which were earlier hauled by two diesel locomotives, thereby saving locomotives. Currently six pairs of such trains are in operation on this section. ELECTRIFICATION OF BALHARSHAH-VIJAYWADA SECTION (SOUTH CENTRAL RAILWAY) 23. Balharshah-Vijaywada section (454 route km) on South Central Railway (Secundrabad Division) is part of the Madras-Nagpur-Delhi trunk route that has now been fully electrified. The electrification of Balharshah-Vijaywada section was completed in the following phases: Vijaywada - Dornakal March - December 1986 Dornakal - Ramagundam November 1987 - March 1988 Ramagundam - Balharshah September 1988 - March 1989 Thus this double line section stands fully electrified since March 1989 and has had about five years for the stabilization of change over from diesel to electric traction. 24. Although Vijaywada-Balharsah section is part of the Madras-Delhi north-south trunk route, it has substantial originating/terminating and local traffic, in addition to through traffic.  41 Annex III This section lies in a mineral rich belt and generates coal, dolomite and cement traffic. This section also handles rice and fertilizer traffic. Part of the terrain is hilly with ruling gradients of I in 100. This section is operating at near saturation with capacity utilization ranging between 90-100%. About 50% of line capacity is utilized for passenger trains and the balance for freight trains. This section has several sidings and three branch lines. However, as only the main line has been electrified, about 20 diesel locomotives are utilized for servicing sidings and branch lines. 25. The projected and actual traffic levels are given below: YEAR TRAFFIC TRAFFIC DENSITY (Mil. GTKm/Year) (Mil. GTKM/RKm/Year) Estimate Actual Estimate Actual 1981-82 13,648 30.1 1986-87 16,850 18,460 37.5 40.6 1990-91 19,970 21,080 44.0 46.3 1992-93' 21,700 22,530 47.8 49.6 Thus the traffic growth on this section was somewhat higher than the projections at appraisal. 26. The following salient features emerged during discussion with S.C. Railway officials regarding the post-electrification operation on Vijaywada-Balharshah section. Build up of Operation with Electric Traction 27. The build up of operation with electric locomotives was gradual as can be seen from the following table: YEAR PERCENTAGE OF TRAINS HAULED BY ELECTRIC TRACTION Freight Passenger 1990-91 57 70 1991-92 71.5 80 1992-93 74 90 1993-94 85 100 1994-95 90 100 (estimated) 28. It was explained that deployment of electric locomotives was constrained by the availability of: - electric locomotives - loco sheds - maintenance and operating staff 29. Furthermore, a certain proportion of freight trains (about 10 percent) are hauled by diesel locomotives which go to the section for servicing sidings and branch lines. As a result S.C. Railway has had to continue diesel locomotive maintenance facilities.  42 Annex III 30. IR have, so far, insisted that siding owners bear part of the cost of electrification of sidings. As the siding owners have been reluctant to do so, IR are now considering electrification of sidings and branch lines at their own cost as it will enable IR to withdraw all diesel locomotives from this section. Electric Locomotive Usage 31. Locomotive usage data is maintained as per locomotive base station and not for each station on which these locomotives operate. The available usage data shows that contrary to projections at project appraisal indicating substantially higher usage of electric locomotives, there is no significant difference between electric and diesel locomotive usage. For example the utilization in terms of km/engine/day in use for electric and diesel locomotives on the Secundrabad division was as follows: LOCOMOTIVE UTILIZATION km/engine/day in use Freight Passenger Electric Diesel Electric Diesel 1992-1993 421 429 793 720 1993-1994 418 423 792 720 Projections as per appraisal 550 450 560 32. It was stated that usage of electric locomotives was adversely affected due to change of traction at points were electric locomotives picked up trains brought to main line by diesel locomotives from branch lines. Invariably the electric locomotive had to wait for the incoming train. Secondly in case of vacuum braked trains, some time was lost, since the electric locomotive exhausters were of lower capacity than those fitted on diesel locomotives. As a result, an incoming diesel hauled trains sometimes required elimination of leaks before an electric locomotive exhauster could create the required level of vacuum. 33. Another reason that has adversely affected the utilization of electric locomotives is the non-availability of dynamic brakes on a large proportion of locomotives. Although, provision of dynamic brakes on electric locomotives is a standard feature, for some years, IR have manufactured electric locomotives without dynamic brakes. On such locomotives the drivers experience the loss of brake power and tend to operate trains at lower speeds than those dictated by locomotives power. It was learnt that IR are now retro-fitting electric locomotives by dynamic brakes.  43 Annex III Speed of Freight Trains 34. There is no significant difference in the speed of freight trains between electric and diesel hauled trains and between pre and post electrification periods. AVERAGE SPEED OF FREIGHT TRAINS km/hr SECUNDRABAD DIVISION Electric Diesel 1988-1989 24.1 22.3 1992-1993 21.3 21.9 35. The potential of higher average speeds with electric locomotives cannot be realized on near saturated sections having substantial passenger train traffic. It is rare for a freight train to get an uninterrupted run of say, a couple of hours on such a section. As stated earlier, the lack of dynamic brakes on a substantial proportion of electric locomotives has led to locomotive crews operating trains at speeds lower than those feasible. Trailing Loads 36. The average trailing loads hauled by diesel and electric locomotives are similar as can be seen from the following table: AVERAGE TRAILING LOAD (tonnes) 1991-1992 1992-1993 Diesel 1677 1698 Electric 1612 1589 37. IR have, so far not procured electric locomotives with higher hauling capacity and continue to operate this electrified section with WAGI, WAG4, and WAG5 locomotives. The hauling capacity of electric and diesel locomotives is similar as can be seen from the following tables: AUTHORIZED TRAILING LOADS (TONNES) FOR ELECTRIC AND DIESEL LOCOMOTIVES ON S.C. RAILWAY (Vijaywada-Balharshah section) ELECTRIC DIESEL WAG I WAG 4 WAG 5 WDM2 SINGLE LOCO 4-WH/BOX Vijaywada- 2000/ 2200/ 2200 2200/ Ramagundam 2200 2400 2500 2500 Ramagundam- 2000/ 2200/ 2200/ 2000/ Balharshah 2200 2400 2500 2500 DOUBLE LOCO BOXN 4500 4500  44 Annex III 38. It is therefore not surprising that the trailing load of freight trains hauled by electric and diesel locomotives are of the same order. They are in fact 4 to 6% lower for the electric locomotives. 39. However, in the case of passenger trains a single electric locomotives can haul a 21 coach express train while a diesel loco can haul a maximum of 18 coaches. On Vijaywada- Balharshah section four pairs of 21 coach express trains are operating where one electric locomotive has replaced two diesel locomotives thereby saving the number of locomotives deployed.  45 ANNEX III Table 1 Jhansi-Itarsi Section Retrospective Cash Flow Statement (Millions of Rupees 1984 Prices) REPAIR AND FIXED MAINTENANCE LOCOMOTIVE ENERGY MAINTENANCE SAVINGS CREW TOTAL YEAR OF FIXED CAPITAL COST COST OF IN WAGON SAVINGS SAVINGS INSTALLATION INSTALLATION SAVINGS* SAVINGS LOCOMOTIVES COST SAVINGS 1984 _ _ 1985 -369 -369 1986 -179 _ -179 1987 -144 -144 1988 -158 -158 1989 -73 -13 90 43 21 17 1 86 1990 -48 -13 0 40 16 0 0 -5 1991 -11 -13 0 45 24 0 0 45 1992 -6 -13 34 64 27 0 2 108 1993 -28 -13 15 64 28 0 2 68 1994 -13 0 68 27 5 1 88 1995 -13 0 80 30 1 2 100 1996 X -13 7 85 33 0 2 114 1997 -13 4 88 34 0 2 115 1998 -13 4 90 35 0 2 118 1999 -13 4 93 36 0 2 122 2000 -13 4 96 38 0 2 127 2001 -13 4 99 39 0 2 131 2002 -13 4 102 41 0 2 136 2003 -13 5 105 42 0 2 141 IRR = 9 percent  46 ANNEX III Table I Jhansi-Itarsi Section Retrospective Cash Flow Statement (Millions of Rupees 1984 Prices) REPAIR AND FIXED MAINTENANCE LOCOMOTIVE ENERGY MAINTENANCE SAVINGS CREW TOTAL YEAR INSTALLATION OF FIXED CAPITAL COST COST OF IN WAGON SAVINGS SAVINGS INSTALATION INSTALLATION SAVINGS* SAVINGS LOCOMOTIVES COST ___________ __________SAVINGS 2004 -13 5 109 44 0 2 147 2005 -13 5 113 45 0 2 152 2006 -13 5 116 47 0 2 157 2007 -13 5 120 48 0 2 162 2008 -13 6 125 51 0 3 172 2009 -13 6 128 53 0 3 177 2010 -13 6 133 55 0 3 184 2011 -13 6 137 57 0 3 190 2012 _ -13 7 142 59 0 3 198 2013 -13 7 147 62 0 3 206 2014 . -13 7 153 67 0 3 217 *Paradoxically the failure to electrify this section rapidly led to locomotive savings as the diesel locomotives were much less costly than the electric - contrary to the assumption in the original economic analysis. IRR = 9 percent  47 ANNEX III Table 2 Vijaywada-Ballashah Section Retrospective Cash Flow Statement (Million of Rupees 1987 prices) REPAIR AND FIXED MAINTENANCE LOCOMOTIVE ENERGY MAINTENANCE SAVINGS CREW TOTAL YEAR INSTALLATION OF FIXED CAPTIAL COST COST OF IN WAGON SAVINGS SAVINGS INSTALLATION SAVINGS* SAVINGS LOCOMOTIVES COST SAIG SVNS I__I_I I______ SAVINGS l128 1985 -128 -128 1986 -542 -542 1987 -436 -8 -62 11 39 64 2 -390 1988 -312 -8 0 11 34 -1 2 -274 1989 -8 0 11 35 0 2 40 1990 -8 -1 11 35 I 2 40 1991 -8 -8 13 39 0 2 38 _ 1992 -8 -2 - 13 40 2 i 3 48 l 1993 -8 -2 14 41 2 3 50 l 1994 -8 -2 14 42 3 3 52 1995 -8 -3 14 44 3 3 53 1996 -8 -3 15 45 3 3 55_l 1997 -8 -3 15 48 3 3 58 l 1998 -8 -3 16 49 3 3 60 l 1999 -8 -3 16 51 3 3 62 2000 -8 -3 17 53 3 3 65 2001 -8 -3 17 55 3 4 68 2002 -8 -3 18 56 3 4 70 2003 -8 -3 18 59 4 4_ 74 2004 -8 -3 19 60 3 4 75 IRR 2 percent  48 ANNEX III Table 2 Vijaywada-Ballashah Section Retrospective Cash Flow Statement (Million of Rupees 1987 prices) REPAIR AND FIXED MAINTENANCE LOCOMOTIVE ENERGY MAINTENANCE SAVINGS CREW TOTAL YEAR INSTALLATION OF FIXED CAPTIAL COST COST OF IN WAGON SAVINGS SAVINGS INSTALLATION SAVINGS* SAVINGS LOCOMOTIVES COST ____________ ~~~~~~~~SAVINGS 2005 -8 -4 19 63 4 4 78 2006 -8 -3 20 65 3 4 81 2007 -8 -4 21 67 4 4 84 2008 -8 -4 21 70 4 4 87 2009 -8 -4 22 72 4 5 91 2010 -8 4 22 74 4 5 93 2011 -8 -5 23 77 4 5 96 2012 -8 -4 24 80 5 5 102 2013 -8 -5 24 83 4 5 103 2014 -4 25 86 5 6 118 *Paradoxically, the rapid electrification of this section led to negative locomotive cost savings as the electric locomotives were much more costly than the diesel locomotives they were replacing. This is contrary to the assumption in the original economic analysis. IRR = 2 percent   IMAGING Report NCD; 14:I18 Type: ICR