India - Karnataka Irrigation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report Nio. 7544 INDIA KARNATAKA IRRIGATION PROJECT (CREDIT 788-IN) PROJECT COMPLETION REPORT FAO/WORLD BANK COOPERATIVE PROGRANME INVESTMENT CENTRE DECEMBER 22, 1988 IThis document has a restricted distribution and may be used by recipients only in the performance of their offcial duties. Its contents may not othenWse be disclosed without World Bank authorization.|  II NOS ANP ABBREVIATIONS ACD Agricultural Complex Division AD Agricultural Division ARDC *cultual Refinance and Development Corpaion (now NABARD) Branch CaWal canal below NLBC CADA Command Area Development Authority CCACutvbeCmadAa CWC Central Water Commission DDA Deputy Director of Agriculture Distributary Canal betwmen branch canal and laterals and/or outlets DOA Department of Agriculture EL Elevation ERR Economic Rate of Retun FIC Field Irrigation ChanneL below outlet supplying water to several farms GOI Government of India GOK Government of Kamataka HA Hectare IBR) International Bank for Reconstrumion and Development ICB Tnterational Competitive Bidding ID Iigation Department of GOK IDA Interational Development Association Inrigation Intensity Ratio of area irrigated to Inigable CCA I JDA JointDiorAgiculture Kharif Wet season: June to October KIP KTmataka Irigation Project Km Kilometer Latrai Irigation canal connecting distributary to oudet LCB Local Competitive Bidding LDTC Land Development Training Center NABARD National Bank for Agiculture and Rural Devclopment NARP National Agricultrl Research Project NDO New Delhi Office of World Bank NLBC Narayanpur Left Bank Canal Notified Area Area notified for supply of irrigation water in each crop season under the provision of GOK Inigation Act OFD On Farm Development O&M Operation and Maintenance Outlet Structe connecting distributary or lateral to FIC Phase I The works of UKS included in the IDA project See Table 1. Phase II The works of UKS proposed for construction in 1987-1994 PWD Public Works Department Rabi Dry Season: November to April SAR Staff Appraisal Report for Karnataka Trrigation Project SBC Shahapur Branch Canal Stage I The works of UKS required to irrigate about 425,000 ha., includes Phase I and Phase II Stage II Possible further extensions of UKS following Stage 1 UAS University of Agriculttual Sciences, Bangalore Taluk Administative sub-district UKS, UKP Upper Krishna Scheme, Project  FOR OFFICIAL ONLY THE WORLD BANK Washmngton. D.C. 20433 U.S.A. Mat of Omnts Gvajl December 22, 1988 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECTt Project Completion Report on India Karnataka irritation Prolect (Credit 788-IN) Attached, for information, is a copy of a report entitled 'Project Completion Report on India Karnataka Irrigation Project (Credit 788-IN)3 prepared by the FAO!World Bank Cooperative Programme. Full evaluation of this project has not been made by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipients only in the peformance of their oMcil duties. Its contents may not otherwise be disclosed without World Bank authoriution.  INDIA FOR OMCAL USE, ONLY KARNATAKA IRRIGATION PROJECT (Credit 788-IN) PROJECT COMPLETION REPORT TABLE OF CONTENTS Page Preface Basic Data Sheet Evaluation Summary iv I. INTROOUCTION 1 II. PROJECT FORMULATION 2 Project Identification and Preparation 2 Appraisal 2 The Appraised Project 3 Negotiations 4 11I. PROJECT IMPLEMENTATION 5 Project Start-Up 5 Revisions to the Project 5 Planning and Design 6 Procurement and Construction 7 Quality and Performance of Project Works 8 Resettlement and Rehabilitation 9 Studies 11 Project Costs 11 Project Financing and Credit Allocation 11 IV. PROJECT IMPACT 12 Intended Agricultural Impact 13 Crop Development in UKS Phase I Area 13 Development in Malaprabha & Ghataprabha Schemes 16 Use of Farm Inputs 16 Faowaudgets 16 V. ECONOMIC EVALUATION 17 Appraisal Estimate 17 Revised Analysis 17 Bases for Analysis 18 Re-Estimated Economic Rate of Return 20 VI. INSTITUTIONAL PERFORMANCE 20 Performance of GOK 20 Irrigation Department 21 This document has a restricted distribution and may be used by recipients only in the performance of their offcial duties. Its contents may not otherwise be disclosed without World Bank authofization.  CADA Engineering Wing 22 CADA UKI Agr1cultural Wing 22 GOKs Compiance with Covenants of the 23 Project Agreement Performance of IDA 24 VII. CONCLUSIONS AND LESSONS 25 Conclusions 25 Lessons for Further Development of UKS 26  TABLES 1. Development of Upper Krishna Scheme, Stage I 2. Comparison of Appraisal Estimates and Actual Project Costs 3. Land Use 4. Changes in Annual Cotton and Sunflower Crojped Area 5. Crop Yields F Cropped Area and Production Jf Stabilisation of Field Irrigation Channels in Malaprabha and Ghataprabha 8. Farm Input Use 9. Farm Budget for 1.5 ha Holding 10. Farm Budget for 4.0 ha Holding 11. Economic Analysis - Investment Costs in Financial and Economic Terms 12 Comparison of SAR and PCR Cropping Patterns Used for Analysis 13. Cropped Area, Yield and Production Used for Analy$ls 14. Financial and Economic Prices 15. Computation of Economic Prices 16. Per Hectare Crop Yields, Farm Inputs and Cost of Production 17. Economic Analysis 18. Water Charges FIGURES 1. Project Location 2. Upper Krishna Project Stige I 3. Upper Krishna Scheme Stage I Implementation Schedule Annex 1 - Comments from Government oi Karnataka Annex 2 - Comments from Government of India  - i - PROJECT COMP'.STION REPORT INDIA KARMATAKA IRRIGATION PROJECT (CREDIT 788-IN) PREFACE This Project Completion Report reviews the Karnataka Irrigation Project in India for which an IDA Credit of US$126.0 million was approved on April 4, 1978. US$8.4 million of the Credit was cancelled in June 1979 due to cancellation of tunnels from financing. Although the original official closing date was March 31, 1984, it had to be extended by two years due to revision of the project, delays in land acquisition and procurement. The final disbursement under Credit 788-IN was made on September 30, 1986. This report was prepared by an FAO/CP mission that visited India in September/October 1986, in close collaboration with Bank staff that prepared a second phase project for possible Bank financing. Additional sources of information included Karnataka Irrigation Staff Appraisal Report No. 1835a-IN of Match 15. 1978, the Project and Development Credit Agreements of Mry 12, 1978, eleven supervision reports (1978-86), correspondence in the South Asia Information Center and discussion with staff associated with the project. The report's main conclusions have been discussed with the Borrower. A draft PCR was sent to the Borrower on September 7, 1988 and the comments received from the Government of Karnataka are attached as Annex 1 and comments received from the Government of India are attached as Annex 2. This project has not been subjected to an audit by OED.  INDIA KARATAKA IRRIGATION PROJECT tCredit 788-INI PROJECT COMPLETION REPQRT BASIC DATA SHEET |EV PRQJECT DATA 113 oAraisal Actal or Actual as ^ Expectatio Current Estimates of Aporaisal ntatl Prr.ject Cost CuSS ml:!in) 284.4 478.6 168 C.-edit Amount (US* million) 126.0 117.6 93 Date Physical Components Completed April i083 July 1987 jI P oportion then Completed about 50% 95% ti.onomic Rate of Return UKS Stage I 161% not estimated UKS Phase I IOS .5wS 2/ Meleprabha and Ghataprebb, 45% not estimatea Cultivable Command Area Stage 1 425.000 ha n/a Phase 1 105.000 ha 97.330 he 93 Number of Direct Boneficiaries 472.000 Number of Project Affected Families 13.000 13.903 107 (reservoirs only) 1/ Excludes Almatti dam. Z/ Not strictly comparable with SAN. See chapter V. _UMLlLAT IE DISBURSEMENTS 21 1979/80 G0/Si 81/82 82/83 83/84 84/86 85/86 86/87 Appraisal Estimate (USS Million) 12.0 38.0 69.0 101.0 126.0 Revised (USS million) 1/ 14.0 34.5 66.0 94.0 117.6 A tual (USS million) 23.1 30.4 5r.8 64.4 74.9 94.1 113.6 117.6 A.utl as .S of Appraisal 192 sO 75 64 69 75 90 93 A tua' as S of Revised 165 88 78 69 64 80 97 100 1/ Credit was reduced In June 1979 by USSS.4 million due to cancellation of tunnels from financing. 2/ *iSburseaent continued tiLt Septomber 30, 1986 PROJECT DATES Origina ctual Preparation Report FAO/CP March 1977 March 1977 Negotiation FeOb 978 feb/Mar 1970 Board Approval Mnrch 1378 4 April 1978 Lredlt Signing 12 may 1978 12 May 197H Efect'ven.ess dat4 July 1978 August 1978 rpolt closing 31 March 1984 31 March 1986  At1ss1on Sent bw M9ntbB No. of Mandaws 502ZJAii&1zati Perfformance Tr-nd Twos& Of a Parsons itn fjejd Reoresented 1/ Ratin 11 1/ Problems Preparation FAO/CP ae 1977 Preappraisal "a Apr/MOy 1077 Appraas- HQ Sep 1977 8 AEEFIIR - - Suoerw1s10n I HQ may/June 3 1S ElI - tI HQ March '60 2 14 1 l t u III "OO Dec 80 2 I 11 2 1 UF IV "Do June st 2 8 IR 2 1 MF V NDO Feb '82 2 12 11 2fi/ 2 no VI "DO Nor '83 2 t6 II 2 3 UTO VII "Oo Sep 83 1 24 IIR 2 2 HTO VIII NDO Apr/May '84 3 27 OIR 2 1 nO IX NOO Nov 804 3 21 TAR 2 2 TF X HQ Jun s85 7 91 IIIIRAP 2 2 7 Al #OO Dec 85/ '86 6 30 11IRP 2 Jan : R FAO/CP Sep/Oct 86 3 IAE Specialisationts la Irrigation Engineert Fs Financial analyst; Ea Economist; Ru Roads Engineer; Pu Procurement Engineers Oa Projects Otficor; A= Agriculturist; * Performance rating; 1. Problom free. minor probl,ams; 2. Moderate problems; 3. Major problems; ;i Irenon 1. Ioprogvngs 2. Stationary; 3. Deteriorating; 4' Type of problems: fu finanspal; Mu Mfnagerial; T= Technical; P- Political; Os Litigatios rthere was continuing dialogue In DOlh1. Bangalore and on site between the first two forms' pervislon miss' ons. r SupoevIs1on msssion rating. Act1ng Division Chief considered 3 more appropriate Aoditional road supervsioan reports were prepared In Dec 1978. March 1979. Apr1i 1980 and madch '981. OTHER PROJECT DATA Borrower Government of InOla executing Agency Government of Karnataka Fiscal year of Borrower April 1 - March 31 Proposed Follow-up Project Upper Krishna Irrigation Project Phase 11 Name o Cur.:ency Indian Rupee (Rs) coarroncy Exchange Rate Appra-sal 1978 Rs 8.60 s US$ 1.00 1973/79 8.206 1973/80 8.076 39101)81 7.893 1981/82 U sZ9 1932/83 9.628 1993/84 12.312 1984/85 11.890 1985/86 12.20  - iv - MIA KAF,NATAKA RRIGATOP PROTEC (CRLDIT 788-ll PROJEC COMLETON REPORT EVALUATION SUMMARY | ~~Introduction The Karnataka Irrigation Project was des&gned to help complete Phase 1 of the first Stage of the development of the Upper Krishna Project, which has been concei. -ialized as a long-tem development program to irrigate the Indiai state of Karnataka's drought prone northern districts from water of the Krishna river.The implementation of the Upper Krishna Project is envisaged in two Stages. Implementation of Stage 1 has been divided into a number of Phases to take account of the large area to be developed, and the Government of Karnataka's implementation capacity. Phase 1 commenced in the mid 1960s and was expected to be completed under this project. 3bjectives The project aimed at bringing irrigation to about 100,000 ha by financing (a) civil works associated with the construction of two dams at Narayanpur and Almatti; (b) associated distributary canals, latermls and field channeLs; (c) on-farn development on about 25,000 ha; (d) construction of abcut 530 km of new roads and the improvement of about 70 km; (e) an agricultural development center, and (f) agricultural support services. In addition the project was expected to provide for the resettlement and rehabilitation of about 13,000 families disvlaced by the construction of the two reservoirs. The total cost of the project was esdmated at US$ 284.4 .million of which about 44 percent or US$126 million w is to be financed by the IDA crA-it.The credit was expected to be disbursed in 5 years. Implementation Ex2ience The main implementation problems faced by the project were as follows: (a) inadequate planning and procurement (far too many small contracts) practices; (b) management and staffing weaknesses characterized by lack of accountability and responsibility in managers and inadequately trained and insufficient staff; and (c) insufficient understanding of the magnitude and complexity of the resettlement and rehabilitation operations which resulted in this operation being poorly designed, funded and executed. In addition the construction of the Almatti dam was premature as its need for irrigation requirements was not necessary in Phase L Consequently, even after 3 one year extensions of the Credit closing date the infrastructure was not completed and there were serious costs overruns. At Credit closure, actual expenditures incurred were more than 50% higher in rupee terms and more than 35% higher in US$ terms(due to the depreciation of the rupee). About 82 percent of the civil works proposed under he project had been completed and about 60 percent of the target had been reached in providing irrigation waters to farmers. Actual disbursements under the Credit totalled US$117.6 million and US$ 8.4 million was cancelled because of misprocurement of the construction of two tunnels on the main irrigation system. EWUDa 1W At completion the project had already had a significant impact in the project area. Agricultural production has increased and with it, the incomes of direcdy benefitting  farmers. Nutyitional levels among the population is increasing as seen by an increase in quantity and varietv of food in the local markets.This trend is cxpxcted to continue and expand. On the negative side, the most serious impact has been on the project's displaced population. Available evidence suggests that most of the people affected by the project suffered a deteric;rzdon in their standard of living and in some cases in their social status. Landless laborrs have been particularly hard hit as a result of the loss of productive land base and c..nequent lower demand for labor in an increasingly competitive labor market with the entrance of fonner small land owners. The PCR estimates the economnic rate of return (ERR) of the Narayanpur dam system to be 9.5%. This is strictly not comparable to the appraisal estimate of 16% for the entire stage I of the Upper Krishna Scheme (UKS). A more comparable estimate is the one made in the PCR (8.5%) for a situation where no further work is done on the Almatti dam wiLh that made at appraisal (10%O) for a situation where no further development of the UKS takes place.T7 - . re-estimated ERR(s) at Credit closure are lower than those estimated at appraisal.HowevWr, non-quantifiable benefits would accrue to non-farmer beneficiaries from the project and to other users of the village road network created. Sustainabilq At completion there was reason to be confident that the project would be sustainable. Indeed two years after credit closing the Phase I irrigation wgets have been achieved. The introduction of suitable cultivation and water management pracdces in Phase I has laid the foundation for more efficient future development, soil conservation and for an effective and equitable use of water resources.In addidon, in the preparation and appraisal of a follow up poject, attention has been focussed on improving the recovery of costs incurred during the upenauiun and maintenance of tie irrigapion systems. Consequently, it is envisaged that financially too, the project would become increasingly self- sustaining. Encouraging progress has also been made in the follow up project in defining a more atdonal and detailed insdtutional and legal framework to implement a coherent resettlement and rehabilitation program Consequendy, there is reason to be confident that costs associated with population displacement would also be minimized in future efforts. Findings and Lessons Several findings and lessons emerge from the project experience. On the irrigation side,a common denominator of many of the deficiencies encountered during implementation was the weak institutionad base which was reflected in the lack of effective project management and lack of accountability at all levels. It is clear that design and construction must be improved and that staff turnover must be reduced by proper incentives and the provision of facilites. The harsher lessons were however, leamed in attempting to deal with the human displacement problem Involuntary resettlement of populations affected by the submergence of their habitat for irrigation, power and other puwposes is a consequence of tie development of water resources to meet the needs of a growing population. However, it is clear that resettlement and rehabilitation of project affected families should also be one of the major objectives of such a development process. The lessons learned under this project clearly indicates that a coherent, well planned and designed resettlement plan should be formulated and implemented in projects of this nature.  t INDIA KARNATAKA IRRIGATION PROJECT (Credit 788-IN) PROJECT COMPLETION REPORT 1. INTROOUCTION 1.1 Karnataka, located in the southwestern part of the Deccan peninsula, has a population of 37 million (1981 census) of which about 70% are engaged in agriculture. Its cultivable area is estimated to be 12.6 million ha out of the State's total area of 19.i million ha. In 1983/84, the net area sown to crops was 10.6 million ha and the gross food grain production reached 7 million tons, of which the main components were rice (2.3 million tons), sorghum (1.8 million tons), millet (2.5 million tons) and other grains (0.4 million tons). In dry years, however, food production is substantially less and given the expected population growth, significant increases in the production of foodgra1ns will be necessary in the future. 1.2 According to area sown in 1984/85, sorghum (jowar) is the most important crop, with some 2.1 million ha annually. Finger millet (ragi) and paddy come next with about 1.0 million ha each, pearl millet (bajra) with about 0.4 million ha, and then wheat and minor millets with about 0.3 million ha each. About 1.4 million ha were planted to pulses. Of the connercial crops, cotton with about 1.1 million ha, was the most important crop, followed by sugarcane 0.9 million ha and groundnut 0.7 million ha. Coffee and tea are grown in the mountains running parallel to the west coast (Western Ghats). 1.3 At least 50% of the state Is drought prone and supports about 50% of the population. The total area covered by irrigation is about 2.5 million hag with an net area actually irrigated in 1983/84 of 1.6 million ha. In view of the higl risk of drought east of the Ghats, high priority is given to incrusing thte are Irrigated. The ultimate potential for Irrigation, 'includtiSuse of groundwater, has been assessed at 5.5 million ha, of which the Krishns basin provides the largest potential contribution of 3.S million ha. Within the Krishna basin, the Tungabhadra scheme is nearing completion and the Ghataprabha, Malaprabha Schemes and the Upper Krishna Project are under construction. (Figure 1). l.4 The Ka.r%taka Irrigation Project (KIP), the subject of this completion repdrt, comprises a phase of the development of the Upper Krishna Scbn (UKS) as well as minor works in the Malaprabha and Ghataprabha Schemes. The Upper Krishna Scheme has been under construction since the 1960s; the first stage Involves the construction of two dams and the irrigation of 425,000 ha; it is expected that further stages will increase the area irrigated. KIP included the completion of irrigation of 105,000 ha in UKS: the estimated cost of KIP was USS284.4 million of which USS126 million was to be financed by an TDA credit.  2 1.5 The major difference between the project constructed and that considered at appraisal is the omission of the completion of one of the two dams, which is not required for irrigation of the works included in KIP. 1.6 This report Is based on a review of the Staff Appraisal Report (SAR no. 1835a-IN dated March 15, 1978), the ProJect Development, Credit and AROC Agreements, IDA Supervision Reports, and data compiled by the Government of KarnAtaka (GOK). The project was discussed with staff of GOK, Irrigation Department (ID), Command Area Development Authority (CADA) and IOA during a visit to Delhi and Karnataka in September and October 1986. II. PROJECT FORMULATION Project Identification and Preparatlon 2.1 As originally proposed the Karnataka Irrigation Project was to include Irrigation infrastructure and land development in each of the ongoing Upper Krishna, Ghataprabha, Malaprabha and Tungabhadra schemes. In 1976, the project scope was changed, partly at the suggestion of GOI and partly because of the already advanced stage of the Ghataprabha and Malaprabha projects: the nature of the work at Tunghabadra, remedial work on canals, drainage and improvement in land development in areas already under irrigation, would have required more attention in project preparation than could be provided at that time. It was suggested that the substantial works to be carried out in proposed further stages or extensions of Ghataprabha and Malaprabha and in improvements and extensions to Tunghabadra might be included in a future Karnataka II project and only field channels and on-farm development in Ghataprabha and Malaprabha be included in Karnataka I. 2.2 In determining the extent of Upper Krishna scheme works to be included ia Karnataka I project, considerations were the wish of the Government of India (C21) to limit the duration and cost of the first phase of the project, nd on the other hand the need to carry the canal system far enough to bring in a stgn1ftcanta rea of irrigated comeand. 2.3 The project was prepared Largely by GOK with coLlaboration by the FAO/WorLd Bank Cooperative Programme, who presented a draft report of the Karnataka I Irrigation and CAD project in March 1977. The preparation report covered a project which was substantiaLLy the same as that appraised. Agrsj saL 2.4 The project was appraised by a WorLd Bank mission which visited India in September 1977 and reported in March 1978 (Rep 1835a-IN), and considered, inter alia, two major issues. 2.5 The Bank's invoLvement in major irrigation projects, which typi- caLly take 15-20 years to implement, necessaritly involves "time-sLice" financing. The risk invoLved in "time-sLico financing is that either the Government wiLL not complete the project or it wiLL be compLeted at such a sLow rate that it affects the viabiLity of the project. The issue was discussed with GOK and the mission received assurances that GOK wouLd con- tinue to finance project implementation according to the agreed scheduLe at a rate not Less than th. Lavel nf * er i+1r r * * 9nfl 4 4 4 v 1  3 2.6 Repayment of project costs (infrastructure and land development costs) by the farmer at present rates of water charges and terms of credit for land development, woulo be equivalent to about 30% of project rent at full development. The GOK was implementing a Bettermert Charges Act which would assess a betterment charge of about Rs3,700 per ha to be recovered over 20 years. This would increase the total payable by the farmers to about 40% of project rent. The ApDraised ProJect 2.7 The Upper Krishna Scheme includes two impounding reser-lirs, canals and related works to irrigate an area of 425,000 ha In the fl . stage with the possibility of extending the area to at least 660,000 ha later. Construction began in 1962 and at appraisal in 1978 It was proposed that the IDA project would comprise a five years phase of the development of the Upper Krishna Scheme, referred to as Phase I of Stage I <D, and minor works in the nearby but separate Malaprabha and Ghataprabha schemes. 2.8 Under the project, it was expected that the pace of development would be accelerated significantly in comparison with the performance of the past, and improvements would be made both in the design and construction standards employed and in the organizational and administrative arrangements for command area development. By completion of the project investments in 1983, about 105,000 <2' ha of the potential UKS cultivable command area of about 425,000 ha would be irrigated. Optimization of returns on all investmentb in UKS would be achieved by the development of irrigation in the full Stage I cultivable command area of 425,000 ha, in which case an economic rate of return of 16% was indicated. The SAR stated that it was the intention of GOI and GOK to complete Stage I of UKS as promptly as possible. A period of 15 years, including the five years of the project, would be required. 2.9 Ouring the five years in which the project was to be implemented the following activities were to be undertaken: A. Within Upper Krishna Scheme 1) The completion of Almattt and Narayanpur dams. Almatti dam would have a full reservoir level of 512 m with spillway crest at level 500 above se level4 11) The-completlon of Narayanpur Left Bank Canal (NLBC) from approximately 3& to 78 kh, including the Rajankollur tunnel (3 km) and the Gundalgera tunnel (I kh); and the construction of Shahapur branch canal (SBC)(76 kn). 111) The construction of lined/stabilized irrigation distribution system and drainage system. to serve an area of about 105 ,0.M ha. ! -- - - -Phase II of Stage I is now being considered for World Bank financing. For details of areas see Table 1. <2b 105,000 ha is the area stated In SAP to be irrigated. However field channels would be completed for say 8O.UOO na (including pre-Phase I anea by 1983.  4 iv) The construction of lined/stablized field irrigation channels to individual farms serving an area of about 80,000 ha. v) The rehabilitation and upgrading of about 70 kt of existing village and link roads and the construction of about 530 km of new village and link rogds. vi) Tile construction of field drainage channels and on-farm works including land shaping, farm irrigation and farm drainage ditches for about 2i,000 ha. The establishment of a Land Development Training Centre to train staff- on the design, execution, and operation and management of land development works. vii) Implementation of a programme of monitoring the water use efficiencies and economic and social impact of the project. viii) Resettlement of familie: displaced due to construction of dams under the project. ix) Establishment of an agricultural development centre, agricultural research stations and other facilities. x) ProcL.ement of vehicles and equipment for road construction, operation and maintenance, agricultural services, the Land Development Training Centre and project monitoring. 8. Within Mal.prabha and GhataPrabha Schemes 2.10 Construction of lined/stabilized field Irrigation channels, field drainage channels, and on-farm works including land shaping, farm irrigation and farm drainage ditches, for about 26,000 ha in Malaprabha and about 5000 ha in Ghataprabha. 2.11 The project would be implemented by the Goveroment of Karnataka with the Irrigation Department responsible for major construction and Command Area Oevelopment Authorities (CADA) responsible for field channels, on-farm development and agricultural aspects of the project. .2.12 The project cost was estimated to be USS284.4 million of which 44X was to be financed by the IDA credit of US$126 million. Annual agricultural production lncreases by UKS at fu)T development of 423,000 ha (about year 2000) were estiated to be 765,000 tons of foodgrain, 109,000 tons of cotton, 110,000 tons of oilseeds and 11,500 tons of chillies. It was estimated that UKS would Increase on-fare employment from about 20 to 45 million man-days, equivalent to the creation of about 83,000 full time jobs. The incoi of 4 family with a median holding of 4 ha was estimated to increase from Rs600/year to about Rs3,30/year. The income of a landless l#arer was estimated to double or triple with the project. Neuotiations 2.13 Negotiations were held in Washington DC in February and March 1978 between delegations from G0I, 60K and the Agricultural Refinance and Development Corporation (AROC) on the one hand and IDA on the other. GOP explained its policy for resettlelment and rehabilitation of persons whose lands are submerged. The delegations discuss"d 60I's subsidy scheme for small farmers. It was agreed that the Central Water Commission (CWC) would arrange supervision of quality cont ol of Na a.vapur and Almatti dams ano  5 NLBC tunnels and that the density, classification and specification of new roads would be reviewed and discussed during review missions. Agreement was reached on. the lining of small canals, the omission of ongoing contracts from the Credit, strengthening of extension services and the inspection of completed das. The Credit Agreement was signed on 12 May, 1978. |II. PROJECT IMPLEMENTATION ProJect Start-Up 3.1 IDA found at the end of May 1978 that overall progress since the appraisal mission was generally satisfactory in particular regarding budget allocations, staffing, and construction progress. The Command Area Development Authority Administrator had set up his headquarters in Gulbarga. Following authorization and ratification of the Project Agreement by GOK and of the AROC Agreement the credit became effective in August 1978. Revisions to the ProJect 3.2 In describing the construction of Atmatti dam, the SAR gives the spiLLway crest at EL. 500 m, fuLL reservoir LeveL (FRL) 512 m but does not mention that these were intermediate eLevations, as the dam had been designed to an uLtimate FRL 524 m. Since the project began, Almatti's spiLLway crest has been changed twice. In 1982 the spilLway crest eLeva- tion was changed to 512 m instead of 500 m, but in 1985, after considering the time when ALmatti's storage contribution wouLd be required, it was decided to construct the spiLLway crest to eLevation 509 m under a subse- quent Phase II project. Almatti dam is not required for irrigation of the Phase I area and outstanding works required for its compLetion have been excLuded from Phase I. 3.3 The irrigation network has generaLLy been buiLt in accordance with SAR proposals but as a resuLt of more detaiLed survey the cuLtivabLe command area is in the order of 6X smaLLer than estimated at appraisaL. Other changes incLude the Lining of smaLL canals in bLack cotton (SC) soiLs: murrum Lining alone was considered impracticabLe and sLab Lining was subs- tituted. ALL fieLd irrigation channeLs have been Lined with concrete or stone sLabs. Areas completed earLy in the project have gated outLets to the fieLd channeLs but the design was altered in 1984 to reduce the number of gates and to simpLify operation and maintenance by providing outlet moduLes, designed to deliver a proportionaL fLow of water to the size of the served area. In order to match water suppLies with demands, the SAR proposed the instaLLation of a flow measuring device at the head of NLBC and each distri- butary but these have not been constructed. Because of generaLly favorabLe reLief for naturaL drainage, no drains were built under the project. A study to determine future drainage needs is in process and resuLts are being anaLyzed by UKP. These wiLL serve for the deveLopment of drainage pLans and their imptementation under a future phase.  6 3.4 SAR proposed that roads would be built in first quality murrum where this is -locally available. However first quality murrum was not found and pavements are waterbound Macadam. Planning and Design 3.5 The Stage I project area, was mapped by the Survey of India, who provided CADA with rectified airphoto maps at a scale of 1:2500 witn contours at a vertical interval of 0.25 m. The first sheets were avaiLable in 1980/81. In the meantine Irrigation Department (ID) were using vilLage maps at a scale of' 1:7920 for Layout of canals. Agreement between the two sets of maps was generalty satisfactory but inevitably there were discrepancies which resulted in ID constructing outlets at levels and Locations which were not best suited for construction of field channels; consequently some Land couLd not be commanded. Detailed soiL capability maps became available in 1981 after parts of the irrigation system had already been designed and as a resuLt some 'marginal' areas were included for irrigation. These maps are used predominantLy by CADA's agriculturaL teams foe Land preparation and planning of crops. 3.6 Design was carried out by ID staff in the project for all irrigation works from the dams to the outLet to the field channels and for roads. FieLd channels and on-farm development were the responsibility of CADA. CWC are responsible for the design of major structures, incLuding the dams. 3.7 In general, design was carried out to satisfactory engineering standards but supervision missions repeatedly rpeerted several areas of concern which were not aLways adequately dealt with: 4nadequate provisions were made for cross regulators on Shahapur Branch Canal (SBC); the routes chosen for canals often required deep excavation and inadequate geotechnical investigation resulted in some unnecessarily deep excavation in hard rock, which caused difficulties during construction. Several major faiLures have already occurred on Narayanpur Left Bank Canal which give rise to doubts concerning the stability cf the major earthworks on this vital waterway. At least one failure was aggravated by unvalved pressure relief (weep) holes throughout the canaL Lining in a h gh embankment which caused piping and endangered the stability of the embankment and required costly repairs. Project staff has been alerted by Bank staff to discontinue weep holes in fills. 3.8 Overall planning for the deveLopment of the project has been less than adequate. After 25 years of construction there is doubt about the ultimate scope of the project; whether the area to be irrigated should be 425,000 ha(Stage I), 660,000 ha referred to in the Phase II identification report or 1,000,000 ha or more as has been suggested by ID. Only now, at the insistence of IDA, is a detailed reservoir operation study being undertaken which will give a realistic idea of the area that can be irrigated with the storage available. The configuration of Almatti dam has radically changed at least twice since appraisal (see 3.2) and the present design has yet to be approved by CWC. Large expenditure has been incurred much earlier than necessary for the construction of Almatti dam, which is not required to store water until well into Phase II, and for the downstream end of SBC which was substantially complete long before upstream reaches.  7 Procurement and Construction 3.9 The SAR stated that about $64 million, or 48% of civil works, would be procured through ICB. In fact only about 25%, or $33 million were procured under ICB. The reason being that (i) spillway gates for Almatti dam W $7 million) were deleted from the project because of design changes; ii) Bank-financed procurement of tunnel construction was reduced by about $18 million (disbursement value $8.4 million)because of misprocurement;j/ (iii) the SAR provided for procurement of canal construction Caggregating an estimated $30.9 million, as per subparas (III) and (iv) of para 3.60(a) in the SAR3 through large or small contracts. Accordingly, works were procured through 41 odd small contracts which proved to be the cheaper alter- native. They werieconsidered too small for foreign bidders and hence were procured through LCB procedures. In retrospect, construction by small con- tractors proved to have many disadvantages. Many contractors lacked adequate skilled manpower, plant, equipment and/or finance and lacked knowledge and experience necessary to comply with given technical specifications. Project staff proved incapable to effectively supervise the very large number of con- tractors and as a result works proved sub-standard. 3.10 Construction was delayed from soon after the start of the project by countrywide shortages of cement and explosives which continued until 1983. Land acquisition, Insufficient budget allocations, procurement delays and difficulties with contractors slowed construction progress. Further delays, particularly in the tunnels, were attributed to electricity supply failures. After delays with the erection of the main spillway gates, Narayanpur dam and the initial reaches of NLBC were substantially complete by 1982 and stored water enabled the first irrigation in the 1982/83 dry season. By September 1983 NLBC was substantially complete up to Km 35.5 (the Initial reach of 35.5 km was not included in Cr 788-IN as it was to be completed before this project began) and into the Hunsagi Branch Canal which was the first area within the IDA project to receive irrigation. The development of irrigated area is shown in Table 3. 3.11 Many reaches of canal were found to have unforeseen large quantities of rock excavation and ID and contractors failed to reach the necessary agreement on rates for carrying out this additional work. Disputes were referred to arbitration but many contractors refused to continue the work until their dispute had been resolved. Where arbitrators decided in favour of the contractors, ID invariably took the dispute to court, which further delayed a settlement: in only one of about 20 such cases has the court quashed the arbitrator's award. Four cases have been referred to the Karnataka High Court and one has reached the Supreme Court in Delhi. 3.12 Progress by CADA in the construction of field channels and on-farm development caused concern to IDA supervision missions and though the pace of construction has improved somewhat the quality of construction is substandard. Land shaping and drainage have bomb-allowed to lag.seriously behind other construction. At Malaprabha and Ghataprabha there was a delay in establishing CADA and shortage of funds and cement caused slow progress. jf A sum of S8.36 million was cancelled from the credit because of misprocure- ment of the construction of two tunnels on the main Irrigation system; because after bids invited under ICB procedures were opened, GOK for reasons unacceptable to the Bank decided to ask for rebids, upon which the Bank cancelled the relevant disbursable portion from the Credit.  a 3.13 The quality of construction has been a constant problem (see 3.17), due in part to unqualified, unequipped contractors and to 10's inadequate supervislon. 3.14 When the credit closed in March 1986 construction of the civil works proposed at appraisal was approximately 82% complete: the state of completion of the ain components is sumarized below: Almatti Oam Civil Works 70% Village Centres 85X Narayanpur Dam Civil Works 92% Village Centres 84% Narayanpur Left Bank Canal 97% Shahapur Branch Canal 86% Distribution System 75% Drainage System 0% Village Roads 100% Field Irrigation Channels 64% Field Drainage 3% Malaprabha and Ghataprabha 58% 3.15 The major outstanding construction is completion of Almatti dam but, as it is not required to store water for Irrigation until well into Phase II. the delay is not critical. The erecti1n of the gantry at Narayanpur dam and the removal of stoplogs in five additional spillways bays are expected to be complete by March 1987. Remedial work is proceeding on two large failed slopes nn NLBC and on the outstanding work on SBC, the distribution system and field channels; the majority of this is expected to be complete by March 1987 with work on distributaries continuing till July and on field channets till 1988. There has been negligible work on drainage and only since drainage problems have become apparent has work begun on design of this important aspect. The length of roads described in the SAR has already been completed but additional work continues. Figure 3 shows the implementation schedule for Stage I of UKS proposed in the SAR and the current optimistic proposals of ID taking account of the actual progress to date. 3.16 By March 1986 only 18,000 ha owt of 31,000 ha of field irrigation .channels in Nalaprabba nd Ghataprabha had been lined. Channels in a further 2.50 h ar expected to be lined in I9N6/SI 4sing IDA finance but no further linin will be undertaken once the existing stocks of precast lining have been used. Land shaping and on-favw development are also behind schedule. Quality and PerforMancQ of ProJect Work' 3.17 The quality of design and construction has been variable. Major works, such as Narayanpur dam appear to be generally satisfactory but construction of line structures scattered over large distances has not been given adequate attention. IDA supervision missions repeatedly tried to raise standards of design and construction to internationally acceptable levels but with only limited success. The supervision of construction by the frequently changing ID staff gave constant cause for concern and IDA missions made many references to examples of poor construction practice; the most frequent were inadequate compaction, lack of ce_nt, poor quality of in-situ concrete lining and of precast slabs, inadequate  9 curing of concrete and inappropriate laying and jointing of slabs (para 6.5). The low quality of design construction will give rise to high main- tenance requirements in the future. 3.18 The performance of the irrigation network will depend on the maintenance accorded to it, and on the ability to control and measure adequate flows. Appropriate water management procedures based on rotational water supply have been introduced but successful implementation requires continuous attention. Resettlement and Rehabilitation 3.19 At the time of appraisal it was estimated that the two reservoirs created under Phase I at Almatti and Narayanpur would lead to the submergence of about 41,000 ha of cultivated land, displacing around 13,000 families. While no specific proposals (other than the construction of village centres) were included in the SAR for the rehabilitation of these families, it was recogni2ed that the activity would be carried out based on guidelines laid out under the existing GOK policy for the rehabilitation of such persons. This included the construction of village cer1tres (including necessary facilities such as roads, water supply, schools, etc.), provision of free house sites at these centres, and compensation for loss of land and structures at a fair market value. In addition, it was agreed that GOK would allocate, to the extent possible, land suitable for cultiva'ion to the farmers resettled from the two reservoirs. Estimates of project cost included provlsions for land compensation, and the construction of 50 village centres (25 each at Almatti and Narayanpur). 3.20 In fact, 55 village centres (called rehabilitation centres) are being construc' d at the two reservoirs under the Phase I project. At Narayanpur, internal roads and electrification have been completed for all 27 centres; the principal outstanding work is the completion of water supply facilities at 10 centres. Work Is in progress at Almatti, but none of these centres has so far been completed. 3.21 With regard to the nuaber of families displaced as a result of the Phase I construction, various estimates have been prepared during the period of project implentation. It appears that no detailed survey was carried *out during eitber the project design phase on the early stages of project implementation, to establish the numter of families which would be displaced. The ost recent estimate of families affected as a result of Phase I is provided in a "Rehabilitation Plan for Persons Displaced" prepared by CADA in July 1986. The salient figures are summarized below:  10 Narayannour Almatti 1/ (1) V1llages submerged (No.) 37 54 Farmland submerged of above villages 36% 29% (ii) Villages whose agr. land submerged (No.) 53 28 Farmland submerged of above villages 15% 3% (iii) Families Affected (No.) - Losing land only 2,P97 3,854 - Losing houses only 1,933 5,993 - Losing both land & houses 1.605 3.110 Total 6.435 12.957 V At Almatti no land has been submerged so far. Virtually all the families have been paid compensation for the assets lost at rates established within GOK norms. Almost all the land-losing families (11,466) have contested In court the amounts paid as compensation. About 5,000 cases have so far been settled in favour of the displaced families, with courts award4ng enhancements equal, on average, to the amount of the original compensation. Another 6,000 cases are presently sub-judice. 3.22 It is generally recognized within GOK that the compensation and resettlement assistance provided have not enabled the ousted families to regain or better their pre-project standard of living. This has been parti- cularly so in the case of those families who have lost farmland. While Indian law clearly stipulates that oustees receive full compensation for lost assets, in practice land assessments are based on under-reported sales values and thus has resulted in the land losers receiving low awards thereby leading them to contest the amounts in court. Consequently, only around 10% of the land- losing families have purchased replacement farmland. 3.23 The above situation has been caused by a combination of factors: I) inadequacy of initial copensation paid for land lost; Ai) gap of mnW years between ccpompnation payment and actual flooding, when the families actually have to move, durini which period much of the compensation may have already been spent (particularly during the recent drought years); iii) receipt of compensation and subsequent court enhancement, in instalments spread over time; iv) in some instances, collection of past debts from the amounts paid as compensation; v) the need for the land losers to make payments to 4a4yers from the court enhancerdents awarded.  it Finally, no detailed rehabilitation programne was drawn up early erough, either as part of the project or as a separate GOK programme which could have addressed the problem of restoring the standard of living of the ousted families. While GOK had agreed that it would allocate to the extent possible, land suitable for cultivation to the ousted farmers (which it has failed to do), no operational framework was outlined, through whici such an action could have been implemented. Studies 3.24 Many assumptions were made for project design and the SAR considered it important for further development of UKS and future project evaluation that such assumptions be compared with actual field data. The project therefore included a monitoring programme involving water use studies and assessment of project benefits. The water use studies have not yet been started. Studies for farm economy, social framework, income, investment and employment were entrusted to insti utes in Bangalore in 1983 but no results have yet been obtained. 3.25 Other issues which require study include the ultimate extent of UKS, the feasibility of pumped irrigation, Lne conjunctive use of groundwater, and conveyance losses in canals. Project Costs 3.26 Table 2 provides a comparison between actual costs incurred (including estimated cost of completion) and the estimates made at appraisal for the different project components. The completion of Phase I of the UKS, and the m,nor component for Ghataorabha & Malaprabha, will have cost Rs4901.1 million (US$478.6 million), which in rupee terms is double the SAR estimete of Rs2445.4 million. In US dollar terms the project cost represents 168% of the SAR estimate of USS284.4 million. The relatively lower cost overrun in US dollar terms reflects the depreciation of the rupee vis-a-vis the US dollar which has occurred between 1978 (Rs8.60 per USS) and 1986 (Rs12.70 per USS). 3.27 Me most important factor which has contributed to the cost overrun is the higher costs of completing the two dams, in particular Almatti dam. As against the estimated Rs594 million, the dams have actually cost Rs1786 million. While much of the cost overun is accounted for by the significantly higher costs of land acquisition and compensation, it appears that the originally anticipated costs were underestimated at the time of appraisal. The other factor which has contributed to the cost escalation Is the considerably longer period of implementation than originally estimated (now estimated to be substantially complete in nine years instead of five), comlined with higher rates of inflation in the earlier years of project implementation than those assumed at appraisal. Project financina and Credit Allocation 3.28 The IDA credit of US$126 million was expected to finance 44X of the estimated total project cost. Actual disbursements against the credit were US$117.6 million, which covered approximately 30% of the actual cost incurred up to 31 March 1986, and 25% of the now estimated total project cost (includfng costs projected for completion). Of the total credit US$8.36 Jill k & - 11 l f r r 4 .A + I f + lat* dit- -& it  12 (30 September 1986) US$35,580 remained undisbursed. Other than the first year of project implementation, disbursements were behind the appraisal schedule, principally on account of delayed implemntation (see basic data sheet). 3.29 The table below compares actual disbursements for different expenditures categories, with the proposed allocation at appraisal. Category 'Credit Allocation SAR Revised I/ Actual Balance - .... , mt-Mon) .....-. Civil Works (al On-Farm Works 7.00 7.00 0.20 6.80 b Other Civil Works 108.00 99.64 110.20 C1O.56W Equ1 ment & Vehicles 5.10 5.10 7.20 <2.10* Unallocated 5.90 5.90 - 5.90 Totai 126.0 117.64 117.60 0.04 j/ Reflects the cancellation of US$8.36 million against the tunnels. The Development Credit Agreement stated that US$7 million would be allocated to finance 55X of AROC's refinance for on-farm development (OFO). Due to the slow progress of OFD and farmers' reluctance to borrow for this work, only US$0.2 million was disbursed for this work. No fonmal reallocation of the credit was carried out to correct the situation of overdisbursements under the other categories. The credit closed on 31 March 1986 after two postponements, but disbursements continued for ancther six months for expen- ditures incurred prior to March 31, 1986. IV. PROJECT IMPACT 4.1 The IDA Project has significantly accelerated the pace of development of UKS in comparison with pre-project performance, even though the optimistic construction schedule in the SAR has not been achieved. The main impact of the proJect is the-substantial improvemt in agricultural production that is nw becmng apparent dut to the Irrigation in the dry anti depressed project area. At the same time the welfare of the approximately 6,450 families who were forced to resettle seriously declined.  13 Intended Agricultural Imoact 4.2 Phase I of Uoper Krishna Scheme (UKS). The major production impact intended by the Phase I of UKS were increases in: annual foodgrain production of about 197,000 tons from the pre-project base of 18,500 tons, cotton production of about 31,000 tons from 6,800 tons; oilseeds of 27,000 tons from 3,700 tons; and other miscellaneous crop production of about 5,000 tons from 2,500 tons. <1> These were expected to be achieved through provision of irrigation and drainage systems, command area development and strengthening of agricultural support services. The project area appraised for the Phase I development was 105,000 ha (net c-tivable command area (CCA)) out of a total 425,000 ha (net CCA) of UKS area. No projection was given in SAR as to when the full development of crop production would be attained in Phase I area. 4.3 Malaprabha and Ghataprabha Schemes. The production impact of these schemes was not quantified In SAR. It was estimated that irrigate( crop yields would increase by about 30% a, a result of the construction of lined field channels and on-farm development. Crop Development in UKS Phase I Area <21 4.4 Land use. The appraisal target was to develop 105,000 ha of irrigation system; field irrigation channels would be developed on only 80,000 ha by 1983 (the original end of Phase I implementation period). It was assumed at appraisal that irrigation water would be supplied before field channels were constructed as was the practice In other schemes. However, since this would favour farmers near the outlet who would become accustomed to receiving more than their share of water and would discourage the construction of field channels, it was decided that irrigation water would not be delivered until after the completion of field channels. Table 3 sets out annual changes in land use during the past 5 years, compared with the appraisal targets. Phase I development area has been reduced from 105,000 ha (SAR) to 97,340 ha as a result of the detailed topographical survey conducted during the course of project Implementation. Irrigable land <3). so far developed by the project covers about 60% or 59,400 ha of the reduced Phase I development area. Reasons for the delay in development are 'explained In Chapter Itt. 4.5 CropDgni Pattern. The past 5 years' actual changes in the cropping pattern in Phase I area, compared with the appraisal projection, are given in Table 5. Differences between "actual" and "appraisal" cropping patterns include introduction of sunflower (kharif and rabi) and groundnut (rabi), and reduction in cotton cultivation. Increase in both rainfed and irrigated sunflower is particularly noticeable. The current annual area plar.ted to sudlower is about 9,000 ha or nearly 10X of total Phase I net cultivated -------------------------------------- Estimated on pro rata basis using cropped area (X) and crop yields given in SAR for total UKS Stage I area (425,000 ha) since SAR does not give separate projection of crop production for Phase I area. I '2> Information on crop development is drawn from various data provided by CADA, UKP. S <3:, Land which has both irrigation potential and field channels.  14 area. Cotton area decreased gradually from about 14,200 ha (1982) to about 10.300 ha (1985). However, there has been a drastic reduction in cotton planted urea in 1986 reportedly because of severe Insect attack in 1985 due to insufficient insect control. The area planted by end of September 1986 is estimated at only about 300 ha. These changes have also occurred in neighbouring taluks (sub-district) In Stage t area as shown in Table 4 Sorghum nd sunflower are reportedly replacin cotton mainly because of the relatively higher production cost of cotton, particularly for insect control. 01fficulty in marketing of cotton is also claimed by farners in the Phase I ir-rigated area. 4.6 In additlut.# to the above changes; cultivation of irrigated paddy and sorghum (ratoon) envisaged at the time of appraisal, has not been realized in the Phase I irrigated area with the exception of a small area of paddy cultivation in poorly drained lowland. Ratooning of sorghum in rabi is considered to be unprofitable by farmers in the irrigated area. Cultivation of irrigated paddy in the Phase I area is now prohibited because of the high water requirements of paddy. Although the area of paddy does not appear in CADA recordh. the mission c4served some irrigated paddy grown within the command area. It appears that unless a proper drainage system is provided to these areas, farmers have no choice but to grow paddy. 4.7 With regard to changes in cropping pattern by two main soil types (Black cotton soils and Red soils), no data has been systematically recorded in the Phase I irrigated area. Moreover, the appraisal report does not clearly distinguish in its projection of cropping pattern between the two soil types, in spite of distinct differences in the land use of these soil types. However, Information obtained in the field indicates the following soroximate changes in cropping patterns by soil types before and after irrigaton. Slack Cotton Soils Red Soils Kiaurif R_bi Kharif Rabi Pre-!OJct: Fallow Sorghum So rhum Fallow (Ralnfed) (Cotton)jJ (Cotton)y Mllet Gran Groundnut Safflower Pulses Wheat *ith roioct: Sorghum SMrgIe so tmm Sorghun rtirigeted =Sunflower Maize M111et Groundnut Maize Safflower Groundnut Safflower Pulses Sunflower Sunflower Paddy) Gram Pulses (Cotton)!/ wheat (Cotton)1/ !/ Cotton is planted in mid Kharif and harvested until late Rabi. 4.4 It should be noted that black cotton soils in the currently irrigated Phase t area reportedly occupy only about 10% or about 2,000 ha, although about 50% of total Phase I area consists of black cotton soils. 4.9 crogping intensity. Present overall cropping intensity in Phase I area is estimated at t I05% (1985), which is nearly the same intensity projected at the time of appraisal (see Table 6). Overall irrigation intensity in the area is however low, being 36% (1985) compared to 101% of the appraisal projection, is the Phase I area Is not yet fully developed.  15 4.10 Within the irrigated area (notified area) c1 present cropping intensity is reportedly about 150%. The mission's discussion with CADA agricultural staff and interviews with some groups of farmers in the area reveal that a significant area within the area notified for irrigation Is cultivated with rainfed crops In addition to irrigated crops because the current schedule of irrigation water supply is not planned to irrigate the entire notified area. 4.11 Irrigation intensity (the ratio of area' receiving Irrigation water to notified area) within the irrigable CCA is summarized below: ADgraisal Actual (Within Notifed Area) (EntirePhaseIarea 193 1984 1985I a 98e Kharif 56 17 75 65 68 Rabi 45 58 82 52 n.a. 101 75 152 117 n.a. ma.man =on a1s *Ms= The higher intensity currently attained compared with the appraisal projection could be due to farmers spreading the water thinly over a large area or to more water than planned being supplied or to a combination of these. Available records of discharge in the canal are inadequate to correlate with the actual irrigated area. 4.12 CrogyYlds. Table 5 sets out the past 5 years' annual average crop yields in the Phase I area estimated by CADA, compared with the appraisal projection. Yields of Irrigated safflower, pulses and whiat have exceeded the-appraisal projection. Yields of the rest of the irrigated crops, except for sorghum and maize, are closer to the projected yields. It appears that the appraisal made a little too ambitious an estimate for average yields of maize and sorghum (4.0 and 3.6 tons/ha respectively), although some progressive farmers in the irrigated area have reportedly attained these high levels of yield. The yield levels presently attained in the Phase I Irrigated area are more or less identical to average irrigated crop yields In neighbouring taluks. r 4.13 Production. Oevelopuent of annual total crop production in the past 5 yeanrs n the Phase I area together with the appraisal projection of full development projection are given In Table 6. Foodgrain production, Including sorghum, millet, maize and wheat, has increased from the actual pre-project (1982) level of 24,30D tons to the present 61,600 tons (1985). Oilseeds production, including groundnut, sunflower and safflower, hat also increased from the pre-project 9,700 tons to the present 34,700 tons. Cotto'i production has decreased from the pre-project 6,000 tons to the present 4-30O tons for the reason explained in para 4.5. Production of miscellaneous crops such as pulses, gram and vegetables has increased slightly in total from the pre-project 7,000 tons to the present 8,700 tons. It is interesting <)> Area notified for supply of irrigation water in each crop season under the provision of GOK Irrigation Act. This Is the area for which irrigation systems have been completed, and for which IO is ready to supply water..  16 to note that oilseeds production in the Phase I area has exceeded the aDpralsal projected production at full development, in spite of the fact that the Phase I area Is not yet fully developed. This is due to an increase in the price of oilseeds while the cultivation of cotton entails a relatively higher risk dub to, as yet, undevelored market and fluctuating prices. oevelop nt in Mala rabha and Ghataorabha Schemes 4.14 The area developed with construction of lined field irrigation channels (FICs) under the project Is given in Table 7. The total area provided with lined FICs by the end of March 1986 was about 17,950 ha, comprising 14,250 ha in jlaprabha area and 3,700 ha in Ghataprabha area. This is about 60% of total area envisaged for development at the time of appraisal. This shortfall is due to delay in implementation. 4.15 Since no data on agricultural production were systematically recorded and little effort was made to quantify the production impact due to the stabilization of FICs, it is not possible to assess the actual effect of this component. Use of Farm Inputs 4.16 Distribution of farm inputs including chemical fertilizers, agrochemicals, and improved seeds in the UKS area is handled by Cooperative Marketing Societies and private dealers with the assl%tance of CADA agricultural wing. As shown In Table 8 use of chemical fertillzers has, in general Increased with the increase in annual irrigated area ftom 290 tons (310 ha) to 5,620 tons (42,310 ha) over the period of 1982/83-1l'85/86. cl> The present annual use of fertilizer per irrigated ha (1985/36) is about 55:50:25 (N;P;K) kg. As data on fertilizer use before irrigation is not available, comparison between "before" and "after" irrigation cannot be made. Historical data on use of agro-chemicals and improved seeds is not available, but in 1985/86, use of these inputs per ha is estimated at about 1.3 litret for agro-chemicals and about 22 kg for improved seeds. Purchase of draught animls in the UKS irrigated area through medium-term 'Oans over the past 5 year was in the range of 350 pairs to 600 pairs per year. Farm Buduets 4.17 Data made available by CAMA on the distribution of landholdings by size, for the Shorapur taluk, shows that approximately 48% of the holding are less th4n 3.0 ha in size, while 36% are between 3 and 7.5 ha. In line with this, two representative farm budgets, for 1.5 and 4.0 ha farms respectively, have been prepared to show the impact of the project on farm incomes. These are sumarized in tables 9 and 10 respectively. A "with project' cropping tntensity of 119% has been assumed (see 5.7). As can be seM, net farm income 'with project* goes up nearly five fold as compared to "future without project" in the case of the 1.5 ha farm (SAR estimated a 4.6 fold increase), and a 5.2 fold increase in the case of the 4.0 ha farm (SAR estimated a 4.3 fold increase). The higher percentage increase, as compared with the SAR is on account of the considerably lower water charges actually levied on farmers (estimated at RsSS per ha on the present schedule of ----.--------------------------- .-. Includes non IDA area covered. by Distributaries IA to 30 of NLBC, separate data for Phase 1, irrigattd area is not available.  17 charges), compared to what was proJected at the time of appraisal (Rs 495 per hectare). In addition, a higher *with project" cropping intensity has been assumed, compared to the 106-107% assumed at the time of appraisal. V. ECONOMIC RE-EVALUATION Aporaise'l stimate 5.1 At appraisal an economic rate of return (ERR) of 16% was estimated for the entire stage I of the Upper Krishna Scheme. This was based on projected costs and benefits from irrigating 425,000 ha, and an Implementation period estimated to be 15 years from the beginning of Phase I. In addition, an ERR of 10% was estimated for an assumed scerar1o where GOK would stop further development of UKS (presumably on account of non- availability of funds) once the IDA assisted Phase I project was completed. At appraisal one cropping pattern was used for estimating incremen-'l agricultural production from the entire project area, which assumed thet from a "without project" rainfed cropping intensity of 85%, farmers wou:d move to a "with project" cropping intensity of 106% (with an irrigatici intensity of 101%). Furthermore the SAR states that in the computation o. the above ERRs all sunk costs were included In the analysis (the firs. expenditures on the development of UKS were incurred in 1957/58). A separate estimate was made of the ERR for the Ghataprabha and Malaprabha component. Revised Analysis 5.2 It is not possible, at this stage, to undertake an analysis of the entire UKS (covering 425,000 ha), as done in the SAR. For the purposes of the PCR the following ERRs have been estimated: I) ERR from the Investment, including sunk costs, In only the Narayanpur Cam system, i.e. Narayanpur reservoir and the - distribution system for the area that can be served from it (assumed !to be 152,000 ha CCA). Sunk costs have been included in the analysis principally to retain a similarity with the approach adopted at appraisal. II) Above ERR, excluding sunk costs. lii) An ERR under the assumption that on account of shortage of funds and no further World Bank assistance for Phase It of the- project, GOK would stop further work on Almatti daw. and only complete the * Narayanpur dam system '1' <1> This analysis is probably no longer necessary, as there is a coumitment in GOK to complete the UKS. However, it has been done tv provide a partial comparlsor with the SAR estimate of the 10% ERR.  - ~~~~~~~~~~~~~~~~~~~~~~~~18 Adequate data are not available for re-estimating the ERR for the Ghataprabha and Malaprabha component. 5.3 The analysis of the complete Stage i area, including Almatti Dam, has not been carried out principally on account of inadequate data available at present to ,rovide a reasonable estimate of future benefits and costs. While at appraisal it was estimated that UKS Stage I would be completed in 15 years, in fact just the Phase I project, which would result in 98,000 ha being irrigated took more than nine years to complete. Without a detailed review of implementation capacity and Go 's future resource availability, it is difficult to project the likely imp. ecantation rate far into the future, as Stage I completion could take up to 20 years more. In addition, a major change which has taken place since the appraisal of the Phase I project, is the recognition that the completion of Almatti Dam for the Stage I area would require substantial relocation of Bagalkot town with an estimated population (1981 census) of around 70,000. Consultants have been appointed to prepare a plan and detailed costs of relocation. In the context of a PCR and given the time available, it is not possible to identify the economic costs (and possible benefits) of relocating a town, as it would require a detailed review of the assets of the town, the proposed plan and costs for relocatiun, and the activit pattern (present and proposed for the future) of the relocated population. Bases for Analysis 5.4 As part of UKS Phase It preparation a reservoir operations study is presently being conducted to match water availability and the optimum "with project" cropping pattern (given present farming practices and cropring patterns in the project area), given GOI's equity objective of spreading available water on the maximum area possible, without impairing the project's economic viability. Based on the results available so far from the study, it is estimated that water from Narayanpur reservoir (without additional storage at Almatti reservoir) would be adequate to irrigate 152,000 ha. This is based on an assumed with project" cropping Intensity of 119%. Consequently, Almatti da is not required for irrigating areas till the latter half ot' the proposed phase It project period. The analysis carried out in this report Is alied at estimating an ERR for the Investment -1n th Narayanpur On System. While the Phase I project would create water distributiow infrastructure for irrigating 97,000 ha and FICs fot 90,000 ha, additional costs have been projected for completing total infrastructure for servicing the entire 152,000 ha. The costs of Almatti Dam have been excluded from thoe analysis. 5.5 Investment Costs. All investment costs related to the construction of Narayanpur Dam, NLBC and SOC, and the related costs for the water distribution system (including field Irrigation channels) for 152,000 ha htvI been included In the analysis. Land shaping and field drainage costs have been included for 80X of the total area. The analysis has been carried out In 1986 prices, with past expenaitures restated in 1986 price terms by using the wholesale price index. Table 11 summarizes total investment costs used in the analysis, In both financial and economic terms. A conversion factor of 0.67 <1 has been used to convert into economic terms the In line with conversion factors being used by the World Bank it, the preparation of the Phase II project.  19 financial costs of dam construction, branch canals, distribution system, land development, and road construction, and the standard conversion factor of 0.80 1> for the costs of equipment and services, including engineering supervision and administration. 5.6 Land Acquisition and Resettlement. An estimated 13,200 ha of land would ultimately be submerged by Narayanpur dam (by 1989/90), of which about 2850 ha had been submerged prior to 1978. Of the remaining 10,350 ha an estimated 70% is assumed to have been cultivated prior to submergence (75% as rainfed and 25% as irrigated from the river using pumps). An estimated additional 9500 ha has been acquired for the construction of main and branch canals, distributarie the IO and CADA colonies etc. It is assumed that 75% was cultivated undt. rainfed conditions. For the purposes of the analysis, the economic value of the land acquired is reflected in terms of the net value of agricultural production foregone as a result of the project. In addition to the cost of construction of village centres, the cost of submerged structures has been included in the analysis. It is estimated that 4,300 families would lose houses, and their financial value estimated at Rs3O,000 each has been reflected in economic terms by applying the SCF of 0.80. 5.7 Agricultural Production. A "with project" cropping intensity of 119% has been used for the analysis (see tables 12 and 13), based on the latest cropping pattern proposed during preparation of the Phase II project, for the reservoir operations study. This broadly corresponds with the irrigation intensity achieved In the notified area in 1985 (see 4.11). Furthermore the higher irrigation intensity assumed, as compared to the SAR estimate of 101%, reflects the downward revision in per hectare irrigation water requirements presently estimated for some of the crops, and a 5% higher total water availability (See table 12). There is, however, one reservation in using this cropping pattern for the analysis, and this relates to the assumption that 18% of the "with project" irrigated area would be planted to cotton. Presently, only an insignificant part of the Phase I irrigated area is planted to cotton and rainfed cotton areas have declined, and now represent between 10% and 12% of the rainfed cropping pattern in the project area (see table 6). For the purposes of the PCR it has not been practicable to amend the cropping pattern, as it would need to be subjected to the reservoir operations study to ensure that the revised cropping pattern maintains the water supply and demand balance. Consequently, the 119% cropping/irrigation intensity has been adopted for the analysis. This compares with the SAR "with project" cropping intensity of 106%, and irrigation Intensity of 101%. 5.8 Crop budgets and the economic prices (in 1986 constant terms) used in the analysis are shown In tables 13 and 14 respectively. The economic prices of traded commodities (wheat maize, cotton, groundnut, fertilizers) are based on IBRO Commodity Price Forecasts (August 1986), and are shown in table 15; the economic prices of sunflower and safflower are derived by applying the ratio of economic to financial price of groundnut, while those o'torghum and millet by applying the economic to financial price ratio of maize. Pulses, gram and vegetables are treated as non-traded, and their economic price has been derived by applying the SCF of 0.80 to the financial price. 5.9 In estimating total incremental benefits, it is assumed that the full development yields "with projeLt" (table 16) are aLhieved 5 years after irrigation is introduced. Full development "with project" yields adopted for the anlysis are based on expeeience in Phase I irrigated area to date (1982/83-1986 kharif season).  20 Re-Estimated Economic Rate of Return 5.10 Based on the above assumptions, and that full development irrigated area of 152,000 ha is achieved in 1994/95, <1> the economic rate of return on the Narayanpur dam system, including sunk costs, is estimated at 9.5%. (See table 17). If, however, the pre-project costs (i.e. costs, incurred before 1977/78) are treated as sunk costs and excluded from the analysis, the ERR is estimated at 13%. 5.11 Finally, by including all costs incurred on Almatti dam up to 1986/87 only, as well as the past and projected c'sts for completing the Narayanpur dam system, an ERR of 8.5% has been E mated. 5.12 The above estimates of project benefits are based on the assumption that the area receiving irrigation water "with project" is the same as that on which the farmer has been practising rainfed cultivation prior to the project. However, as the "without project" cropping pattern clearly shows, there are areas left fallow in each of the kharif and rabi seasons (including the area devoted to the bi-seasonal cotton crops). If the irrigation water is used on the previously fallow areas cropping Intensity "with project" would be higher than assumed (while irrigation Intensity would still be 119%), both in the SAR as well as in the revised analysis carried out in this report, and there would be a smaller loss in the previously rainfed area production. Based on discussions with farmers in the presently irrigated area, it appears that the above reflects the actual situation being practiced. However, since CADA's record of cropped area in the Irrigate?, area does not include rainfed crops, a precise estimate of the overall cropping Intensity cannot be made. This exclusion of rainfed crops in the "with project" cropping pattern leads to a downward bias, albeit small, in the estimated ERRs. Analysis carried out reveals, that if there is only a 70% loss of without project production as a result of the project (i.e. a with project cropping intensity of around 150%), the ERR of the Narayanpur Dam system (including sunk costs) would be 10.2%, instead of the 9.5% estimated above. VI. INSTITUTIONAL PERFORMANCE Performance of GOK 6.1 GOK has made great efforts and allocated a large proportion of its irrigation budget to the Upper Krishna Scheme; consequently the pace of development of the project has accelerated substantially since 1978. However the rate of progress forecast in the SAR has not been achieved, partly due to budget allocations being less than proposed in the SAR and to GOK's Inability to provide sufficient experienced staff for the project. 6.2 Staffing the project, particularly at the junior levels, has been a constant problem on a number of scores. The Karnataka Public Works Department includes Irrigation, Comnunications and Buildings and staff are transferred between these at frequent intervals, often resulting in staff without irrigation training and experience being transferred to the project. _--------- ------------- cl, Based on ID estimated construction schtdOle (Figure 3) stretched by  21 The project, particularly the command area, is in a poor, remote and undeveloped part of Karnataka with inadequate educational, commercial and recreational facilities which discourage professional staff from living in the project area. There is a reluctance of staff to be transferred to the project and an eagerness to transfer away as soon as possible. The incentives given to staff are insufficient to overcome this. IDA missions have proposed that the Irrigation Department staff should be a separate establishment from other parts of PWO and that greater incentives be given to UKS staff to encourage them to stay. Some progress has been made towards establishing a separate staff cadre but-there is still a rapid turnover of staff. If ID becomes a separate organization consideration will have to be given to the arrangements for nonstructing and maintaining roads in the project area. This is at pre. ..t the responsibility of UKS staff but if the project Is staffed only by irrigation engineers there would be advantages to returning the responsibility for roads to PWD. 6.3 The responsibility for the project has been divided between two Chief Engineers and the CADA Administrator; since 1985 an additional Chief Engineer has been responsible for Bagalkot relocation and a further two Chief Engineers have recently been appointed for Phase II works. The Chief Engineers report to the Secretary Irrigation II in Bangalore and the CADA Administrator to the Additional Secretary CADA. The Secretaries have many other responsibilities throughout the State. There is no one person with overall responsibility for management and coordination of all aspects of the project. A full-time UKS Project Manager resident in the project areas is required to take overall responsibility for the project, improve coordination, particularly between CADA and I0, and ensure that GOK is made aware of the action necessary for the successful implementation of the project. 6.4 GOK's shortcomings regarding the resettlement and rehabilitation of people dispaced by the project are discussed in Chapter 3. Irrigation Department 6.5 The performance of the Irrigation Department has been variable. The increased pace of development of the project is mainly due to ID but the shortcomings noted In design para (3.7) and construction para (3.17). are cause for concern. There has been a serious lack of overall planning and there are now many important Issues that require early study If efficient impl_mentation of the second ond subsequent phases is to be achieved. The rapid turnover of staff and the use of staff without irrigation training and experIenc contribute to some of the shortcomings but I0 did not take adequate action to prevent the bad workmanship which was repeatedly reported by supervision missions nor to discipline construction supervisors who permitted unsatisfactory work to continue. 6.6 Ooeration and Maintenance of the completed part of the irrigation system is presently carried out by one circle of the construction division after finishing supervision of the construction of the same system. Even though the irrigated area is constantly expanding and expected to reach 100,O0 in about two years, O&M has not yet been formally separated from construction. The lack of a fonmal separation is, in effect, a disincentive to the establishment of an OSM organization in this project as long as cons- truction continues. Recently GOK agreed with a proposal by a Bank mission, to engage consultants for the preparation of a comprehensive W1KM master plan for the ultimately developed Upper Krishna Project* that would be Implemented in phases, con_ensurate to the development of irrigation. These consultants will specify appropriate 011 procedures, quantify Infrastructures muupoers and equipment requirements and their cost.  22 6.7 A procedure has been proposed by a recent Bank pre-appraisal mission of a future project for formal hand-over from the construction divisions to the OhM divisions but this has not yet been adopted by project authorities. It is proposed t* hn* over field channels to farmers after they have been successfull)voperate*and maintained for one year by CADA (or possibly by ID in futur.' seE-pars 6.10) but the formal procedure for this has not yet been agreed; outtet committees should be given a legal standing to permit them to formally take over field channels. 6.8 It is understood that it is the policy of GOK to provide water only at the rate that will ultimately be available so that farmers do not become accustomed to more than their share and consider it their right. Although NDO staff have made efforts to establish a :ter management programne this is not being strictly observed in the field. Instead of a strict warabundi system being adopted, whereby each farm receives water at a fixed rate for a given length of time proportional to the area of the farm at prescribed weekly or fortnightly intervals during the season, the system is being operatea to satisfy farmers demands. This is possible at present while there is surplus water available to meet excess demands from the small part of the system that is being irrigated but there will not be sufficient water to permit this when the system is fully developed, when water will be available to irrigate only 50% to 70% of the CCA during any one season. ID does not appear to have appreciated the importance of making all farmers aware of the water management policy and delivering water accordingly. 6.9 Many outlet structures, particularly the early gated design, have been damaged to permit Increased quantities of water to be taken. I0 nave generally adopted a lenient attitude and have not prosecuted those causing damage. It is essential if the project is to function efficiently that anyone who wilfully damages the system is brought to justice. CADA Engineering Wing 6.10 The performance of CAOA has also been variable. The initial setting up of CADAs was slow and uncertain, particularly in Malaprabha and Ghataprabha. Design and construction of field channels was carried out by staff secondqd to CADA Engineering Wing from the I0 but they were often inexperienced in irrtgat10w. In Nalaprabha and Ghataprabha CADA constructed the field channels using directly hired labour whereas in UKS small *contractswea let fow this work, frequently to landowners through whose land the channl passed or to contractors approved by these landowners. In view of the poor quality of field channels IDA now proposes that design and construction of field irrigation channels (FtC) should be transferred back to IO. The advantages of this would be consistency between main system and FIC design; designs would be subject to the ID approval procedure; It would be possible to include FIC construction in large contracts for distributaries, thereby improving the quality of construction, reducing the stAff time required for supervision and accelerating construction. CADA UKS Agricultural Wing 6.11 The appralual report did not give much account of the institutional aspects of CADA Agricultural Wing or Its linkage oith the research programme assisted by the project and OOA regular extension programme in spite if the fact that CADA for the UKS scheme wos just established at the Lime of appraisal. CADA Agricultural Wing ompr;-.e% r Agricultural Division (AD),  cj an Agricultural Complex Division (ACO), an On-Farm Development Division (OFO), and a Land Development Training Centre (LOTC). These four divisions are each headed by a Deputy Director of Agriculture (ODA). A Joint Director of Agriculture (JOA) oversees and coordinates the activities of the four DOAs. The functions of these divisions and the activities carried out during the Phase I implementation period are summarized below. 6.12 Agricultural Division is a task force, to guide farmers in irrigated farming, organize water users' groups, and. establish water scheduling among farmers, in the newly irrigated area during the first 4 or 5 crop seasons: its work is then transferred to DOA regular extension services. The performance of nO has been in general satisfactory. 6.13 OFO Division has been respons for executing land levelling ard land shaping since August 1984, when the OFO works were transferred from tte CADA Engineering Wirg. The progress of OFO works under the project was slow due to lack of field staff. The division is at present mainly responsible for setting up OFO demonstration fields, land shaping and levelling for farmers. They have assisted farmers to implement about 6,800 ha of land shaping and given technical assistance to farmers for construction of key bunds and broad bed furrows on over 25,000 ha as a temporary measure, in order to facilitate better on-farm irrigation practices before the proper land shaping is implemented. 6.14 Agricultural Complex Division (ACO) has been responsible for producing certified serds, a small-scale varietal trial and raising some horticultural nurser.es for the farmers in UKS area. An adaptive research programme was conducted by six research scientists seconded from the University of Agricultural Sciences (UAS) as a part of ACO activities. However, due to delay In construction of research facilities coupled with lack of irrigation water, active research work other than a few varietal trials has not been carried out. ACD operates four farms with a total area of about 150 ha (located at Bheemarayangudi, Kavadimatti, Almel and Almatti). These farms have not yet been fully established because of delay in land acquisition and lack of irrigation water sources. 6.15 The mission was informed by UAS, Bangalore that under the IDA- assisted National Agricultural Research II Project (NARP II) UAS is now planning t% establish a new research station to conduct adaptive research related to irrigation farming in Phase I Irrigated area (Hunsagi). Since the responsibility of agricultural research in Karnataka is in the hands of UAS, research activities in UKS should be, in principle, carried out by UAS. 6.16 Land DeveloDment Training Centre (LOTC) was established In 1979 for training project field staff In design, execution and operation and management of land development works, including construction of field irrigation channels and drains, land shaping and land levelling; farmers are trained in land development and production. To date, LDTC has trained 70 junior engineers, 165 AAs and about 800 farmers. Performance of LOTC has besA satisfactory. GOK's Compliance with Covenants of the Project Agreement 6.17 In general GOK complied with covenants of the Project Agreement except in the following respects. Inadequate budget dllocations were made, particularly between 1978 and 1982, which was contrary to Sectorn 2.01 of the Project Agreement. in the early stages of the project, audit reports were frequently late. Karnat4ksi undertoil' to complete the UKS as early as  24 technically and financially feasible (Section 3.04); there have been considerable delays to date due to financial, managerial and technical factors. Section 3.05 requires GOK to allocate to the extent possible land suitable for cultivation to farmers resettled from the Almatti and Narayanpur reservoir areas. This has not been done. 6.18 Cost Recovery. Section 3.11(a) requires GOK to assess and collect a bettenment-levy. This levy, after being contested in courts was abolished in 1980, without IDA's consent. Section 3.11(b) requires GOK to review and, if necessary, increase water and water-related charges with the objective of collectin5 full O&M costs, and, to the extent possible, the cost of infra- ttructure investment. GOK reviewed water charges but because the project was still in its five-year running-in phase, no water charges had actually been levied. l/ Assuming no change in current (1976) rates, after the termi- nation of the running-in phase farmers would be paying on the aver-ge Rs 70/ha/yr, or about 50% of GOK's current budget of Rs 140/ha/y, ur O&M of irrigated land, probably even a smaller percentage after appropriate O&M costs are determined (para 6.6). However, farmers contribute also indirectly to O&M through taxation increases. Based on records of production increases in irrigated parts of the project area market taxes wiil rise, on the average, by about Rs 190/ha/yr, thus yielding a combined revenue increase of about Rs 260/ha/yr. These may be adequate in the short run, but should be adjusted periodically to actual expenditures. 6.19 The PCR mission has not been able to establish that Almatti and Narayanpur dams were being periodically inspected in accordance with sound engineering practice as required by Section 3.12. No reports of earlier dam panel reviews were available. Performance of IDA 6.20 IDA assisted in the planning of the project and have been instrumental in causing many changes. At the time of appraisal an IDA consultant recommended that the design floods at Almatti and Narayanpur be substantially increased; five additional spillway bays were added to Narayanpur dam to cope with the increased design flood and the design of Almatti dam waS changed. The dacisions to design and supervise construction of roads by project staff and to transfer field channels to CADA were made at the insistence of IDA. They were not completely successful and consideration is now being given to transferring road construction to PWO and field channels to Ij. 6.21 The appraisal report stated that the IDA credit would assist with the completion of Almatti dam wherea- the dimension quoted refer to only pa; tial completion. It is difficult to Justify on economic grounds the inclusion of Almatti dam in the IDA project: the SAR makes no reference to the reasons for the early construction which are understood to have been related to the establishment of allocated water rights by Karnataka. The area to be submerged and the price used for land compensation were underestimated in SAR and no mention was made of the future submergence of Bagalkot town. The price assumed for compensation at Almatti was a small fraction of that assumed for Narayanpur and has been the largest contributor to the cost overrun. The SAR does not justify the inclusion of the field channels and on-farm development works in Malaprabha and Ghataprabha in the Credit except to attribute them with an unexplained ERR of 45%. The scattered nature of these works has made it impracticable to collect data to permit a reestimation of this. I/ in line witn current practice, the running-in period is five years.This, in view of Bank missions, is unjustifiably long and hence a dialogue with GOK was initiated during pre-appraisal of a possible future project, with a view of shortening it.  6.22 Eleven full supervision missions have visited the project area over a period of about 8 years. These have been predominately engineers in keeping with the nature of the project and the apparent lack of agricultural input in the fonmal supervlsion missions was compensated by informal intervention from NDO. NDO's Irrigation Manag'ment Team spent considerable time devising a workable water management operation for UKS but there has been only limited success in putting this into effect. Supervision missions assisted ID staff to improve engineering capabilities by holding workshops and informal discussions on irrigation design and construction. Although supervision missions were regularly critical of the quality of construction IDA did not insist on severe action that might have ensured that only work of acceptable quality was permitted. Neither did IDA insist on project action by the Borrower to meet its legal obligation to provide an adequate R&R progran. 6.23 Supervision missions and missions for appraisal of Phase II have drawn attention to the need for a reservoir operation study, plans for resettlement of Bagalkot and adequate provision for the rehabilitation of people who have lost their livelihoods due to the project. VII. CONCLUSIONS AND LESSONS Conclusions 7.1 The bulk of the UKS works included in the Appraisal Report, other than Almatti Dam, is likely to be completed in 1987 thereby enabling 97,000 ha in a drought prone area to be irrigated. Although the IDA credit has accelerated the development of UKS, construction will have taken four years longer than the five years expected at appraisal. The overrun is due to insufficient budget allocations, shortage of suitable staff, delays in land acquisition, shortages of materials, procurement delays and difficulties with contractors. The estimated cost at completion is double the cost estimated at appraisal, mainly due to delays, in conjunction with inflation, and underestimation of resettlement cost at appraisal. 7.2 Inclusion of Almatti dam in the IDA project is difficult to understand and has incurred substantial costs much earlier than can be economically justified. 7.3 Although the principle of resettlement of people displaced by the project was appreciated at appraisal no attempt was made to prepare a plan of specific action required to ensure that adequate compensation was promptly paid and families displaced by the project were enabled to become established and economically self-sustaining at living standards that at least matched those before the project. 7.5 GOK should institute a formal handing over of completed sections of the irrigation system to a separate O&M organization after ensuring that the system is operating as designed. All O&M costs should be recovered from beneficiaries and GOK should shorten the current grace period from five years to two years, after ensuring reliable and equitable Irrigation supplies to all water users on the system. 7.6 Although agricultural plans are prepared by CADA, they bear little relation to crops grown and irrigation operation plans developed with the assistance of NDO's Water Management Programme are not being effectively implemented. The irrigation system has been designed to distribute the 14#n4+ad wat*r available over the largest practicable area In order to  zr achieve GO's objective of equity. This will only be achieved if a suitable operation plan is devised and implemented; consequently, addressing such operatlonaV matters early In the development ef the system is of paramount importance. Water use studies and continuous monitoring of agricultural benefits are essential to enable operational improvements to be made. If operational efficiency is not achieved the full potential benefits which have been assumed will not be realised; Investments in construction near the tail of the system will have been wasted and the economic viability of further development will be undermined. Lessons for Further Development of UKS 7.7 The Irrigation Department must pay more attention to project planning and in particular the coordination of irrigation develonment, the programme for raising the water level in Almatti reservoir and the resettlement of Dagalkot, and the feasibility of pumped irrigation schemes. The ;ltimate extent of the project (stage Ir) and the optimum use of the allocated water should be studied together with conjunctive use of groundwater. Estimates of water lost in the irrigation network should be improved by measurements so that correct quantities of water are delivered to farms and losses in the system can be identified and allowed for or reduced. The agricultural soil survey results should be taken into account when designing the irrigat 3n network. 7.8 The quality of design and construction must be improved. Measures which could be adopted to achieve this include the separation of the Irrigation Department from other parts of PWO so that only staff with suitable training and experience in irrigation are employed, the reduction of staf turnover by incentives and improved facilities in the project area, and the appointment of a Project Manager with overall responsibility for the project. Engineers must be made aware of the high professional standards that are required at all times. Construction contracts should be awarded to contractors with adequate resources of skilled man-power, plant and finance, and the knowledge and experience to maintain high standards of construction: this will require much larger contracts in order to attract suitable contractors. Contract documents should be-improved to minimize litigatiovi.  INDIA tARItAtKA IRRIGATION PROJECT (CREDIT 788-1NI fl.yalanmmnt of SUn J_rl Jahn& Scheme - Steas I T9ta. kPro-Phase I , Phase I Phe.U9 lSAt.aiI Dates* -1978 1S78-11 l3 197li-197 ,1987-1992 2004 "IMatti Reservoi fOL LI 524 a 0 512 SOO c.517 524 ,alOyanpur Re- rvoir Started 100% too% werayr npur Left bank Canal NLSC 78 km 36 ka 78 km 78 ha V,strlbuterles & loteral. 47.220 ha 7.000/8.060 J/ 48,000 47.200 r -$* Irrigation Channels 47,220 ha 7.000/8.060 j/ 48.000 47.220 aO L field draineag 3/ 37.780 ha 25.000 2.700 21 37.180 h.hapur SreAch Canal SIC 70 km 0 76 km 76 km b4stributeri,s 9 laterals 122.120 he 64.000 58.170 122.120 6-eld Irrig&tion Channela 122.120 ha 39.000 40.480 122.120 OfL a field dr-inage I/ 97.700 he 97 700 M4tr,1'. InOd & Jiwarui DC rA4 km 0 III 294 '5st'4autaries A laterals 239.360 ha 8B.940 239.360 field Irrigation Channels 2319.30 he 71.210 239.360 oro A field drainage t 91.490 ha 191.490 AlMatti La*t Bank Canal 300% Di*ti lutaeres & laterals 10.300 ha 18.300 Pl.la :-rig-tion Channels 18.300 ha 16.300 OFO & field dra-nage 1/ 13.040 h a3.040 Total Ifitgabl. Area 425.000 ha 8.060 112.000 105.390 240,550 425.000 o.t Z.antitles shown *ro cumulative. IL roll Reservoir Levwl. Ll %sumes that 50% of Irrigated area requ4res land shaping and drainage. ) Are uf full land shaping with bank ass'stance by March 1986. SA&ACtual. 0D _IM  IDI 870.6 31.41 MG. 74015) 1.6 136.5 32.0 Votneas" c0.0 00 :, 1,1 60 CIvil Worhm 33*.0 W26.41 25.5 Load Coa60eetaila 39.3 3S3.32 31.010 Villaes Centiro* '2. 0.2 30.21 3.10