Sri Lanka - Major Irrigation Rehabilitation Project

Dowument of The World Bank FOR OFFICIAL USE ONLY CA /a3>- Cc Report No. 5231-CE STAFF APPRAISAL REPORT SRI LANKA MAJOR IRRIGATION REHABILITATION PROJECT November 6, 1984 South Asia Irrigation I Division This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its rontents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS US$1 = Rs 25.2 Rs 1 = US$0.0397 UNITS AhD EQUIVALENTS 1 meter (m) = 3.2808 feet (ft) 1 kilometer (km) = 0.6214 miles (mi) 1 hectare (ha) = 2.4711 acres (ac) 1 kilogram (kg) = 2.204 pounds (lb) 1 metric ton (ton) = 1.0161 long tons (1 ton) 1 m3/sec = 35.3357 cubic feet/sec (cusec) 1 bushel of paddy = 20.8655 kg PRINCIPAL ABBREVIATIONS USED AEAR - Agricultural Extension and Adaptive Research Project ARTI - Agrarian Research and Training Institute ASC - Agrarian Service Center CIDA - Canadian International Development Agency DA - Department of Agriculture DAC - District Agricultural Committee DAS - Department of Agrarian Services DD - Deputy Director IND - Irrigation Management Division ECL - Engineering Consultants Ltd. GOSL - Government of Sri Lanka ID - Irrigation Department iIMI - International Irrigation Management Institute Io - Institutional Officer LCD - Land Commissioner's Department MADR - Ministry of Agricultural Development and Research MASL - Mahaveli Authority of Sri Lanka MCM - Million cubic metres (M = million) HIRP - Major Irrigation Rehabilitation Project NILID - Ministry of Lands and Land Development NEDECO - Netherlands Development Consultants INMAS - Program for Integrated Management of Major Irrigation Systems PM - Project Manager PMB - Paddy Marketing Board SDC - Switzerland Development Cooperation TIMP - Tank Irrigation Modernization Project UOP - University of Peridenya USAID - United States Agency for International Development VIRP - Village Irrigation Rehabilitation Project WMP - Water Management Panel WMS - Water Management Secretariat FISCAL YEAR January 1 - December 31  FOR OFFICMIL USE ONLY SRI LANKA MAJOR IRRIGATION REHABILITATION PROJECT Credit and Project Sulmmary Borrover The Democratic Socialist Republic of Sri Lanka Amount : SDR 17 Million (US$17.0 Million equivalent) Terms : Standard Project To increase agricultural production in seven selected major Description: irrigation systems. The project would (a) rehabilitate existing irrigation systems serving a total area of about 46,000 ha; (b) introduce an Integrated Management Program to ensure proper operation and maintenance of the systems and distribution of irrigation supplies; (c) strengthen support services provided by various Government agencies; and (d) make provisions for monitoring and evaluation of project execution and performance. The project would also provide support for the International Irrigation Management Institute. Project The Project is expected to generate an incremental production Benefits of about 73,800 metric tons of paddy and 6,200 metric and Eiskb tons of soya, benefitting about 34,400 families. Project induced increases in family income would range from about Rs 5,900 to Rs 10,600. Sri Lanka's economy would realize 16% economic rate of return from project investments. The project faces a general risk from declining trend in long term price of paddy and a possible risk from failure to implement the proposed water management improvements. Proiect Cost: (US$ Million Equivalent) Proiect Components Local Foreitn Total Civil Works, etc 19.0 11.2 30.2 Equipment and Vehicles 1.8 6.6 8.4 Tech. Assistance & Training 1.3 0.4 1.7 Administration, etc 2.9 0.0 2.9 Tctal 25.0 18.2 43.2 (US$ Million Equivalent) Financing Plan: Local Foreign Total GOSL 11.8 0.0 11.8 IDA 7.2 9.8 17.0 CIDA 1.6 5.2 6.8 SWITZERLAND 4.4 3.2 7.6 Total 25.0 18.2 43.2 I This document has a resticted distribution and may be used by recipients only in the performance of l their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  Estimsted Disbursement: (US$ Million Equivalent) 1985 1986 1987 1988 1989 1990 1991 MDA FY IDA Annual 0.4 2.4 4.2 3.2 2.6 2.2 2.0 Cumulative 0.4 2.8 7.0 10.2 12.8 15.0 17.0 CIDA jj Annual 0.0 0.2 0.6 0.7 0.6 0.5 0.4 Cummulative 0.0 0.2 0.8 1.5 2.1 2.6 3.0 SDC Annual 0.1 1.1 1.6 1.6 1.4 1.0 0.8 Cummulative 0.1 1.2 2.8 4.4 5.8 6.8 7.6 Rate of Return: 16X Map8s IBRD No. 18021, 18022 1/ CIDA joint contribution only.  STAFF APPRAISAL REPORT SRI LANKA MAJOR IRRIGATION REHABILITAXION PROJECT TABLE OF CONTENTS _ ....................................................................... Page No. I. BACKGROUND ............................................... 1 II. SECTOR PERFORMANCE .................................... 2 Agricultural Sector .... .............. . , , 2 Irrigation Subsector ..... 3 Irrigation Management and Rebabilitation . . 3 Bank Group Support for Irrigation Development ............ 4 III. THE PROJECT AREAS .6 General. 6 Topography, Soils, and Climate 6 Population and Landownership 7 Cropping Practices. 8 Irrigation Constraint to Cropping. 9 Agricultural Institutions, Research, Inputs, and Credit 10 IV. THE PROJECT . . .12 Project Objectives . . . 12 Project Components ....................................... 13 Water Availability ................................................... 20 Status of Project Preparation ................. ........... 22 Implementation Schedule ................. ................. 22 V. PROJECT COSTS. FINANCING, AND PROCUREMENT ................ 23 Project Costs ............................................ 23 Financing ............................................... 23 Project Special Account ...... ............................ 26 Procurement ....................................................... 26 Disbursements .................................................. 27 Accounts and Audit ....................................... 28 VI. ORGANIZATION AND MANA&GEMENT ...................... .... .0. 29 Overall Project Management . ...... .................. ..O.. 29 ; Management of Rehabilitation Works . .......... ........... 30 The Integrated Management Program ................. o ...... 31 Monitoring and Evaluation .............. .. .............. 32  Paige No. VII. PRODUCTION. MARKETING. FARM INCOMES, AND COST RECOVERY ... 32 Production ....................... ........................ 32 Marketing and Prices ...... ............... ..... . . . ....................... 35 Farm Income ....................... ....................... 36 Cost Recovery and Fiscal Impact ..... ..................... 37 VIII. ECONOMIC BENEFITS, JUSTIFICATION. AND RISK .............. . 40 Project Benefits .................... ..................... 40 Economic Analysis .......................... . ............. 41 Sensitivity Analysis and Project Risks ................... 43 IX. RECOMMENDATIONS ................... LIST OF ANNEXES ANNEX 3.1 - Sub-Project Systems: Salient Features 3.2 - Climatological Data 3.3 - Population and Labor Force 3.4 - Table 1: Distribution of Holdings by Ownership Status Table 2: Distribution of Holdings by Size of Farms ANNEX 4.1 - Criteria for Irrigation Systems Rehabiltation ANNEX 4.2 - Cropping Intensities and Water Shortage Years ANNEX 4.3 - Implementation Schedule-Chart 26370 ANNEX 5.1 Table 101 - MLLD Irrigation Management Division Table 102 - Kantalai Table 103 - Moraweva Table 104 - Iranamadu Table 105 - Giant's Tank Table 106 - Rajangana Table 107 - Nachchaduwa Table 108 - Huruluvewa Table 109 - Procurement Table 110 - Support Table 111 - Land Commissioner Department Table 112 - Department of Agriculture Table 113 - Department of Agrarian Services Table 114 - University of Peradeniya Table 115 - International Irrigation Management Institute ANNEX 5.2 Table 1 - Project Components by Year (Base Cost) Table 2 - Project Components by Year (including contingencies) ANNEX 5.3 - Phasing of Equipment Purchase and Funding by Donors ANNEX 5.4 - Schedule of IDA Disbursement  ANNEX 7.1 - Croppiug Intensities ANNEX 7.2 - Summary of Cropped Area, Production, and Yield ANNEX 7.3 - Crop Input Requirements ANNEX 7.4 Table 1 - Giants Tank Illustrative 3-ha Farm Model and Estimate of Producer Rent Table 2 - Iranamadu Illustrative 2-ha Farm Model and Estimate of Producer Rent Table 3 - Rajangana Illustrative 1-ha Farm Model and Estimate of Producer Rent ANNEX 8.1 - Total Cost and Benefit Streams for Economic Analysis ANNEX 8.2 - Financial and Economic Prices List of Charts and Maps I-A - Present Values of Incremental Benefits and Costs Against Discount Rates. I-B - Incremental Benefits and Costs. I-C - Benefit Cost Switching Value at OCC. Giant's Tank Irrigation System 25845 Nachchaduva Irrigation System 25846 Buruluwewa Irrigation System 25847 Rajangana Irrigation System 25848 Morawewa Irrigation System 25849 Iranamadu Irrigation System 25850 Kantalai Irrigation System 25851 Mahaweli Supplies-Line Diagram 26292 Organization Chart 26291 Location. of Systems (Subprojects) 18021 Iranamadu Subproject 18022 ;  SRI LANKA MAJOR IRRIGATION REHABILITATION PROJECT STAFF APPRAISAL REPORT I. BACKGROUND 1.01 Sri Lanka continues to place major emphasis on irrigation as a vehicle for agricultural development, an emphasis that has a long history and is evident in both the magnitude and the diversity of investments. Modern irrigation development commenced in the last century, but rapid expansion of the irrigated area took place only after independence through settlement schemes in the sparsely populated dry zone, a process that has culminated in the massive Mahaweli program for exploiting the power and irrigation poten- tial of Sri Lanka's largest river system. In 1982, the Accelerated Mahaweli Program accounted for about 40% of national investment. 1.02 Rapid expansion of irrigation has been accompanied by a growing awareness that benefits from the existing irrigation systems are not at full potential. Though farmers and managers have responded to attempts to improve performance, this response has been limited by the physical characteristics of the schemes, their deterioration due to inadequate maintenance, and the perpetuation of out-of-date management. Questions relating to the rehabilitation, modernization, and management of existing systems have there- fore received increasing attention, and a range of pilot projects and institutional initiatives have been undertaken. These have been handled mainly through GOSL's own resources. Nevertheless, donors -- in particular USAID and IDA -- have been actively involved. The IDA-supported Tank Irriga- tion Modernization Project (TIMP), Credit 666-CE, was the first initiative to tackle the problems of major existing irrigation systems. The Village Irrigation Rehabilitation Project (VIRP), Credit 1160-CE, is similarly con- cerned with minor systems. Important lessons have been learned from these earlier projects; as a result, GOSL has revised its approach to physical rehabilitation and to the management of rehabilitated systems. The proposed Major Irrigation Rehabilitation Project (MIRP) will support this process. 1.03 In 1983 a feasibility study for MIRP was prepared by FAO/CP. Reports on the cropping potential under the project, as determined by water availability, vere prepared by Engineering Consultants Limited (ECL), and the University of Peradeniya (UOP) provided the socioeconomic data. These reports formed the basis for appraisal by a joint rDA/CIDA mission, which visited Sri Lanka in March 1984. The appraisal mission comprised J. Srivas- tava, D. J. Berkoff, R. Stevenin (IDA) and V. Carvell (CIDA). P. Aklilu assisted the mission in the preparation of this report.  -2- II. SECTOR PERFORMANCE A. Agricultural Sector 2.01 About 75% of the population lives in rural areas, and agriculture continues to account directly for about 25% of GDP, 50% of employment, 40% of Government revenues, and 60% of export earnings. The country has about 2.25 M ha under permanent cultivation, of which paddy (0.65 M ha) and tree crops (0.95 X ha) account for about 70%. In addition, shifting (chena) cultivation by clearing poor forest areas is also practiced during maha season (October to March) and could amount to about 1.0 N ha in some years. The agriculture sector's basic characteristics are determined by the two broad climatic zones of the country. The wet zone (with an average annual rainfall in excess of 1900 m) encompasses the island's southwest quadrant. It accounts for only 30% of the land area but for as much as 65% of the area under permanent cultivation, with population densities rising as high as 700 per sq km. Rain comes with both the northeast and southwest monsoons, per- mitting the double cropping of rice and the cultivation of field crops as well as perennials such as tea, rubber, and coconuts. In contrast, the dry zone (annual rainfall between 900-1,900 mm) covering the remaining land area receives most of its rainfall during maha and cropping during yala (April through September) is possible only with irrigation. As a result, population densities in the dry zone average only about 75 per sq km. 2.02 In line with overall economic trends, agricultural growth accelerated from about 2% p.a. during 1970-76 to about 5% p.a. during 1977-82. The major factor contributing to the increase was a substantial rise in paddy produc- tion, which rose from an average of about 1.4 M tons in 1970-76 to about 2.2 H tons in 1980-82. This increase can be attributed largely to farmers' response to the incentives created by the Government's liberalization reform and, to a lesser extent, to additional use of irrigation and other inputs. In contrast, the performance of other field crops has been mixed, and that of the crucial export crops has been disappointing. Combined value-added in tea, rubber, and coconuts was no higher in 1980-82 than in the early to mid-1970s. 2.03 There is considerable potential for further agricultural growth and diversification in Sri Lanka. To achieve this, emphasis needs to be given to: (i) strengthening the liberalization and incentive policies successfully implemented in paddy and extending these to other crops (field crops, coconuts) and to livestock so as to stimulate private initiative in accord- ance with Sri Lanka's long-term comparative advantage and shorter-term needs; (ii) improving Government programs in support of private agriculture so as to encourage efficient use of available land and water resources, stabilization of upland and rainfed farming, and smallholder replanting programs; (iii) resolving the management and institutional issues in the public estate sector so as to lay the basis for long-term recovery in the crucial tree crop sec- tor; (iv) allocating investment funds more closely in accordance with  -3- economic priorities and with the need for rehabilitation and proper operation and maintenance of existing assets; and (v) generating the public and private resources needed to maintain and expand productive facilities in the agricul- tural sector as well as to contribute to the economy's wider requirements. B. Irrigation Subsector 2.04 Irrigation in the wet zone, where crop intensities average about 170%, is normally confined to run-of-the-river diversions to supplement rainfall. In contrast, in the dry zone, where rainfall is less certain, most of the irrigation systems are based on storage tanks (reservoirs) designed to supplement maha rainfall. The residual supplies after the maturity of the maha crop are used to support Yala cultivation to the extent possible. The country has about 180 major tanks, each with a storage capacity in excess of about 25 MCM (20 thousand acre ft), irrigating a total of nearly 165,000 ha (408,000 acres). About the same area is commanded by some 2,500 village irrigation schemes (each supporting 80 ha or less), of which about 60% are tank scbemes in the dry zone and the rest diversion schemes in the wet zone. The past settlement programs in the dry zone have also included river basin developments, such as the Gal Oya and Uda Walave projects. The ongoing Mahaweli Ganga program envisages irrigation of about 365,000 ha of new or poorly irrigated land (effectively doubling the irrigated area) both within the basin, as well as by transfer of supplies to the neighboring dry zone. Mahaweli supplies are already reaching a number of existing tanks in the dry zone, including three to be rehabilitated under the proposed project (para 4.16). These three are receiving supplies for the last 6 to 8 years. 2.05 Actual irrigated area under paddy (both maha and yala) in Sri Lanka has increased from about 430,000 ha in 1972 to about 520,000 ha in 1982, the major systems contributing about 60% of the irrigation. During the same period the availability of irrigation facilities increased from about 351,000 ha to 433,000 ha, mainly from major systems. Actual irrigation in Sri Lanka is however, highly susceptible to rainfall, and irrigation peaked at about 536,000 ha in 1978, when irrigation facilities existed for about 402,000 ha only. C. Irrigation Management and Rehabilitation 2.06 The benefits from major irrigation systems remain below their full potential. The main cause is the traditional emphasis on the distribution of water to guarantee the main subsistence paddy crop rather than to maximize total agricultural output. A general lack of incentives for efficient water use, resulting from absence of suitable water rates and penalties, con- tributes to late cultivation of the main paddy crop (resulting in an ineffi- cient use of maha rains), to substitution of water for other inputs (mud ploughing, use of standing water to control weeds), and to excessive water application as insurance against possible future shortages. With few control structures or measuring devices on systems, water distribution can not be controlled to match the demand and is inequitable. Physical deterioration  -4- from inadequate maintenance has further adversely affected system perfor- mance. Finally, significant problems continue to be associated with coordina- tion among Government agencies and between and among farmers and officials. Rehabilitation of the major irrigation schemes and introduction of proper management of these schemes is therefore being given priority. 2.07 GOSL has decided to rehabilitate all the major irrigation systems in need of rehabilitation under a phased program. Five systems (13,000 ha) were rehabilitated under TIMP, while seven (46,000 ha) are proposed to be rehabilitated under MIRP. Feasibility studies for rehabilitation of another four systems are about to commence with Japanese assistance. The selection of the systems for rehabilitation has been based on their potential for increase in paddy production. Thus systems with relatively better paddylands and systems benefitting from Mahaweli releases have been preferred. Increase of paddy production on the selected systems having good paddy lands is also in line with GOSL policy of crop diversification, since lands marginally suited for paddy under other systems could be used for non paddy crops. Also adequate regional distribution over the dry zones to benefit all communities by the rehabilitated systems has been considered. 2.08 Lessons learnt from TIMP have guided the planning for MIRP. Dry years after rehabilitation of TIMP systems did not permit an analysis of the impact of rehabilitation on production. However, enquiries made from farmers (ARTI Study) have indicated that rehabilitation has improved water distribu- tion. Farmers lower down the system acknowledge better water availability, though general water shortages resulted in non compliance with the rotation of supplies and even structures were damaged by farmers. For implementation of rehabilitation and operation and maintenance thereafter, the main lessons have been to: (i) carry out surveys and detailed designs and plan the con- struction of works adequately in advance to avoid delay in implementation; (ii) allow sufficiently long construction periods for rehabilitation works in view of the limited time during which canals can be closed; (iii) prevent indiscriminate land and water use in the catchment areas which reduce inflow to the tanks and increase erosion; and (iv) organize an efficient system of water management. Proper management requires the creation of an awareness in the farmers to play their role in the distribution of water including properly maintaining and operating the part of the irrigation system assigned to their care instead of damaging structures in the false belief that these obstruct flow. Confidence must also be built among farmers that water will arrive on time and that distribution will be equitable, the extent to which demands are met being dependent on water availabiliy of which they would have advance information. Moreover farmers must be assured that the changes in cropping practices expected from them will be supported by the necessary assistance from agencies concerned. D. Bank Group SuDport for Irrigation Development 2.09 The Bank Group's lending in the agriculture sector is expected to account for about 33% of the 1985-88 lending program for Sri Lanka. The Bank and IDA have supported, in the context of the Maha-reli program, one project  -5- (Lu. 653-CE/Cr. 174-CE - US$14.5 M each) that has been completed, two projects that are under implementation (Cr. 701-CE, US$19 M, and Cr. 1166-CE, US$90 M), a fourth project that was recently approved by the Board (Loan US$12 M and Credit US$30 M), and a technical assistance credit (Cr. 979-CE, US$3 M) for studies relating to the long-term evolution of the Mahaweli program. IDA has also financed three water sector projects that have been completed - Lift Irrigation ICr. 121-CE, US$2 M), Land Reclamation and Drainage (Cr. 168-CE, US$2.5 M), and the TIMP (Cr. 666-CE, US$5 M) -- as well as the ongoing VIRP (Cr. 1160-CE, US$30 M). Irrigation rehabilitation has also been prominent in three ongoing rural development projects (Cr. 891-CE, US$20 M; Cr. 1079-CE, US$33.5 K; and Cr. 1363-CE, US$26.5 H). In addition, an on-going Agricultural Extension and Research Project (Cr. 931-CE, US$15.5 M) complements the activities under irrigation and other projects. 2.10 The Bank Group has provided extensive support for the irrigation sector. Economic returns of two completed irrigation projects have been satisfactory. Main points from the PPAR (No. 2801 of December 28, 1979) of the Low Lift Irrigation Project were: (i) deficiency in project design, (ii) delays in implementation due to civil works procurement delays, (iii) inade- quate extesion service, and (iv) lack of adequate support services. PPAR (No. 3730 of December 28, 1981) of Mahaweli I Project brought out: (i) Although no provision was made to improve agricultural services the project farmers responded positively to sound technological package. (ii) Prevention of encroachment is difficult unless preventive measures are provided for during project preparation and appraisal and legal action taken soon as it occurs. (iii) Water management and maintenance of irrigaton systems not appreciated during appraisal. (iv) Farmers open to fair water charges since the privatization of rice trade and price control abolished. Continued Bank group support for irrigation projects is justified to ensure that rehabilitation is done along proper lines, that the lessons from earlier projects are reflected in system design, operation, and management, and that the institutional and policy changes proposed by GOSL to improve water management are implemented. GOSL is giving priority to rehabilitation projects to maximize returns from past investments in different sectors including irrigation and Bank group support would go a long way in assisting GOSL to achieve the rehabilitation targets.  -6- III. THE PROJECT ARE&S A. General 3.01 Salient features of the seven tank irrigation systems to be rehabilitated are given in Annex 3.1, their locations shown on Map 18021, and the main channels of the system with their irrigated areas presented by line diagrams in Maps 25845 to 25851. The total net agricultural area to benefit from the rehabilitation of the schemes is about 46,000 ha (Kantalai 6,990, Noraveva 1,960, Iranamadu 9,430, Giant-s 12,460, Rajangana 5,910, Nachchaduwa 5,400, and Raruluwewa 4,090), comprising purana lands traditionally owned and settled under ancient tanks (about 26,000 ha), and settlement lands settled more than about 20 years back under Government colonization programs (about 20,000 ha). B. Topography, Soils, and Climate 3.02 ToDoxraphv. The area serviced by each system has a typical terrain, the undulating relief established by streams descending to the sea. The humps between the streass are conspicuous: land slopes are steeper in the upper parts of the systen near the tanks, but lower down the valley the area flattens to a gently sloping plain cut up by a number of shallow streams. The main canals accordingly run as one-bank contour canals in the upper part and branch out into branches and distributaries to feed the areas lying between different streams. The area under the Giaut's Tank is different from the others and the gently sloping plain right from the Tank needs many canals originating from the tank itself. Drainage of the areas is generally good, the natural streams acting as draijs right up to the end of the system, except in isolated areas where farmers have extended cultivation to beds of drainage channels. 3.03 Soils. The available soil data is based on a low intensity mapping except for Kantalai and Rajangana (medium intensity). Soils under the Giaut's system are mostly grumusols and alluvial. Both are underlain by alluvial sands at 5 to 10 meters' depth. Grumusols (clay loan to clay) are characterized by low permeability and high available water and are very suitable for rice cultivation but need puddling for land preparation. Alluvial soils (sandy clay loam to sandy clays) have higher infiltration rates and less available water. Rowever, these soils are also being used entirely for paddy cultivation. Iranamadu soils are alluvial (deep sandy, shallov to moderately deep sandy) and clayey. The alluvial soils occur at higher levels. These permit tillage under dry conditions, can be maintained under saturated conditions because of clays below, and are therefore used exrensively for rice in Maha. The clayey soils occur in valleys. These are hard when dry and need puddling for land preparation. Rantalai. Nachchaduwa. * Raiauzana. Huruluvewa. and Morawewa soils are reddish brown earths (RBE), lov humic gley (LEG), and alluvial. RBE occur on the upper slopes and are well drained because underlain by sandy clay loams.and gravel layers of variable thickness are commonly met. LEG soils occur in bottomlands. These are deep soils (grey to dark grey sandy clay loams) with poor drainage and have high  -7- water table suited for paddy; they are extremely hard vhen dry but when wet change to sticky and plastic. The area under alluvial saile is significant only in Kantalai, where these soils are poorly drained and are suitable for paddy. 3.04 Climate. The dry zone of Sri Lanka is divided into eight agro-ecological zones designated by the Land and Water Use Division of the Department of Agriculture, and the project area lies in four of these: DL 1 (Rajangana, Nachchaduwa, Korawewa, and Huruluwewa), DL 2 (Kantalai), DL 3 (Transmadu), and DL 4 (Giant's Tank). The climatic characteristics of these zones are given in Annex 3.2. Annual pan evaporation varies between 1,550 and 1,800 am (61.4 and 70.7 inches), while annual 75X availability rainfall varies from 625 to 965 mm (24.6 to 37.9 inches). Evaporation is at a maximum in the months of June, July, and August (150 to 200 mm per month) vhen rain- fall is practically nil. Rainfall is maximum in October, November, and December (120 to 225 mm per month), and in these months evaporation is close to rainfall (75 to 190 am per month). The rainfall pattern, however, is quite variable, and extended droughts occur even during the rainy season. C. Population and Landownership 3.05 Population. About 34,400 farm families, or a total population of about 193,000, depend on the seven irrigation systems. About 68% of the project area population is under 30 years of age, about half of these 15 years of age or younger. Literacy exceeds 90% (those having more than five years primary education). Annex 3.3 gives the details of the popula,ion and labor force. 3.06 Land tenure. The landownership varies from system to system (Annex 3.4, Table 1). Most farmers of purana laud have title deeds. The title deeds have also been issued to settlers on some of the older colonization schemes (notably Iranamadu which is 80 years old). In contrast, recent (about 20 years old) settlers have been issued Land Development Ordinance (LDO) permits. These permits are also normally given to encroachers, whose holdings are later regularized, and to others receiving state lands. Lands held under an LDO permit cannot be sublet or sold without Government concur- rence, although these lands can be inherited. In practice, however, unoffi- cial leasing, sale, and other arrangements are common, with the original settler often remaining effectively as a laborer on his own land. Leasing of Land accounts for the relatively high proportion of leased-in land (20-30x) in Iranamadu, Kantalai, and Nachchaduwa. Illegal encroachments are quite comnon and account for about 10-20% of the area commanded (see Annex 3.1). 3.07 Average holdings of paddyland (reported by the socioeconomic survey) range from 3.2 ha in Giant's Tank to 0.9 ha in Rajangana (Annex 3.4, Table 2). Settlers on early colonization schemes (notably Iranamadu) were given 2 ha, but in later colonizaton schemes allotment to settlers was reduced to 1.2 ha (Kantalai, Huruluveva, and parts of Moraweva) and then to 0.8 ha (the rest of Noravewa and Rajangana). The distribution of landholdings tends to be more variable and inequitable on systems with a high proportion of purana  lands (Giant's Tank, Iranamadu, Kantalai, and Nachchaduwa) than on those having higher proportion of colonization settlement lands. In addition to their paddyland, most farmers also hold an average of 0.5 ha in the highland, while in Anuradhapura district some farmers also cultivate chena land. Chena cultivatioa, however, is significant only in the Huruluweva area, where about 25% of all farmers report an average of 1.5 acres each. D. Croppiun Practices 3.08 HIaha Cropping Patterns. Practices, and Yields. The universal maha crop is paddy, and the large majority (95% or more) of farmers use improved new varieties issued by the Department of Agriculture. Decisions on the timing of cultivation and related issues are taken at the preseason cultiva- tion meeting, attended by the farmers and representatives of concerned Government agencies and chaired by the Government Agent. Given the unreliability of rainfall, the timing of cultivation always remains flexible and the normal practice is to delay cultivation until a certain water storage is available in the tanks. These cultivation meetings are very useful but abnormal delays in arrival of rain sometimes upsets the decisions completely. 3.09 Cultivation practices vary considerably between the extremes repre- sented by those of Iranamadu and Rajangana: o Iranamadu. The farmers under this system are predominantly large-scale (relatively) and commercially oriented. The soils (para 3.03) permit dry land preparation and dry sowing thus utilizing maha rainfall fully in about 80-85% of the maha crop area. This significantly reduces use of stored water, but necessitates heavy weeding inputs. This agricultural practice is possible for those farmers with the financial resources to hire tractor services for land preparation. Labor is predominantly hired (85-90X), fertilizer use and chemical weed/pest control are virtuilly universal, and a large portion (85-90%) of production is sold. O Raiangana. Farmers on this system are predominantly small-scale and subsistence oriented. Land preparation is normally delayed until water is available from the tank, since soils are less suited to dry cultivation and farmers lack the resources for tractor hire. Land preparation is predom;mantly with buffalo, and a small proportion (25%) of the area is transplanted. Fertilizer use and chemical treatments are universal, but input levels are significantly lower than in Iranamadu. Family labor contributes more than 50% of the total, and the proportion of output sold is relatively low (55-60%). These differences in crop practice result from (i) the different water supply situation (fairly good in Rajangana, relatively short in Iranamadu), (ii) the soil conditions, and (iii) the economic status of the farmers. These factors also influence practices on other systems. The uncertainty of supplies in  -9- Giant-s Tank encourages cultivation of minor tank beds (at minimal input levels) in advance of filling, as well as considerable dry land preparation (where soil conditions permit) and transplanting (over 50X of the area, often utilizing groundwater for establishing the nurseries). Water shortages at Noravewa have discouraged use of inputs (being the lowest of the seven sys- tems), but have encouraged land preparation and cultivation in advance of water issue. 3.10 Yala Crops. Yala croppiug has also traditionally been dominated by paddy, eveu under conditions of water constraint and even on permeable soils unsuited to paddy. The area planted in Yala is decided at the preseason cultivation meeting, taking into account the water remaining in the tank after the end of maha. Cultivation rights are sometimes redistributed under the bethma system 1/ to ensure that all share equitably in the permissible crop area. If adequate water is available, yields in Yala can be somewhat higher than in maha, but in general agricultural practices appear to be fairly comparable in the two seasons. Although paddy remains the main yala crop, some diversification has been attempted, most notably in Hurluwewa, where successful soya crops were grown in 1981 (over lOZ of the area) and 1983 (over 40Z of the area), and in Iranamadu, where chillies and oilseed crops are increasingly being grown. E. Irrization Constraints to Croppin2 3.11 Water Availability and Cropping Intensities. The intensity of crop- ping and crop yields on the tank-fed irrigation systems are directly related to the availability of water (catcbment yield, transfer of supplies from neighboring basins) and the storage capacity of tanks to balance runoff with crop requirements. Runoff from the catcbments varies widely from year to year, and the unreliability of supplies places a major constraint on the stabilization of cropped areas. Inter basin transfer of water has therefore been practiced, wherever possible, to augment the inflow to the tanks. The biggest of these transfers has been the diversion from the Nahaweli basin to a number of existing and new tanks in the dry zone. Hap 26292 shows schematically the diversion path of Mahaweli supplies to the three project tanks, which receive direct supplies and to another project tank receiving supplies through return flows, from Mahaweli area. The systems receiving allocations from the Nahaweli could in theory approach cropping intensities of 200% (GOSL's long-term objective). Simulation studies of the Mahaweli system done by NEDECO indicated that double cropping on all systems down- stream of Polgolla Tunnel (a total of 83,600 ha, including 22,400 ha under the four MIRP schemes) is possible with an average annual diversion through Under the bethma system, land rights are temporarily suspended to allow all farmers to share cultivation in proportion to the land owned in the area cultivated, which is usually located near the tank. The system evolved in traditional farmer-operated schemes but has been transferred to major settlement schemes.  -10- the tunnel of 875 MCM. 1/ This analysis, however, assumed much greater system efficiencies than have so far been achieved. In practice, despite diversions that have exceeded this figure, cropping intensities in recent years have fallen well short of 100% during yala. Cropping intensities on the systems outside the Mahaveli specification area remAin low due to shortage of water in yala. High evaporation from the tanks reduces the availability of water for the crops even if the tanks have leftover storage after maha. F. Agricultural Institutions, Research. Inputs, and Credit 3.12 Agricultural Institutions. Sri Lanka's agriculture is supported by a profusion of Government institutions and organizations at the national, district, and local level. The most significant of these for the project are: o The Ministry of Lands and Land Development (NLLD), which is responsible (outside areas administered by the Mahaweli Authority): (i) through the Irrigation Department (ID), for irrigation development; (ii) through the Land Commissioner's Department (LCD), for land settlement and administration; and (iii) through the Forestry Department (FD), for all forestry matters including catchnent management; o The Ministry of A2ricultural Development and Research (MADR) which is responsible: (i) through the Department of Agriculture (DA), for research and extension in relation to field crops; and (ii) through the Department of Agrarian Services (DAS), for input supply, the administration of tenancy legislation, minor irrigation, the election of village irrigation headmen, and a range of other supporting activities; and o The District Authorities, notably the Revenue Department headed by the Government Agent, which is legally responsible for land allocation within a district, revenue collection (including water charges), the convening of preseason cultivation meetings, etc.; and the District Development Councils chaired by the District Minister, together with the District Agricultural Committees (DACs), which are responsible for coordinating irrigation and agricultural matters at the district level. 3.13 Research and Extension. The DA conducts research on paddy, other annual field crops, and minor export (perennial) crops. It maintains a Central Rice Experimental Station and a network of five Regional Research Centers and twelve branch stations. Sri Lanka has a strong rice breeding 1/ NEDECO: Mahaweli Ganga Implementation Strategy Study: Final Report, July 1980.  -ll- program. In recent years, research on other field crops suited for irrigated agriculture has been stepped up. National Coordinated Research Programs have now been launched for a number of crops (including soya, pulses, maize) in association vith international and bilateral research organizations. Under the IDA-supported Agricultural Extension and Adaptive Research (ARAR) Project (Cr.931-CE), a major adaptive research program incorporating trials on farmers' fields is being undertaken. The DA's agricultural extension divi- sion is also being supported under the AFAR project through the "Training and Visit" (T&V) system. The number of extension staff is programmed to increase to provide one field extension worker for 600 farm families against 1,000 before the project began. 3.14 Agricultural Inputs and Marketing. The policy changes introduced by the present Government since 1977 have liberalized the marketing and dis- tribution of agricultural inputs and outputs, greatly expanding the alterna- tive channels open to the farmer. Fertilizer is imported mainly by the Ceylon Fertilizer Corporation, and both imports and local production are distributed through a variety of channels, including private traders, multi- purpose cooperatives, and the Agrarian Service Centers (ASCs) of the DAS. Fertilizer subsidies relative to import prices have been substantially reduced in recent years, but they still r-Ain significant (Annex 8.2 gives the financial and economic prices) and have contributed to a substantial increase in fertilizer use on paddy. Agro-chemicals, imported by the private sector and the Ceylon Petroleum Corporation, are freely available at import parity prices, and are universally used in the project area (para 3.09). About 15Z of the annual seed paddy requirement is produced by the DA on its network of seed farms and in conjunction with private certified seed growers and is distributed through the ASCs. About 50% of all seed is retained by the farmers from their own output, with the balance coming from other farmers and traders. Seed for other crops, although not so readily available as for paddy, is also available from DA sources, and improved non-rice seeds are still largely to be developed. Tractor services are widely available in the private sector, and tractor imports have largely responded to market demand. Hire charges have, however, risen substantially, leading also to significant increases in hire charges for buffalo. 3.15 GOSL operates a guaranteed price support scheme (GPS) for both paddy and a range of subsidiary field crops (including oilseeds and pulses). For paddy, prices have been adjusted to levels broadly comparable to world market levels. Together with the impact of the Government's liberalization reforms, this price adjustment has led to a major shift to private marketing and distribution, and farmers are now served by a flexible and efficient private sector, with prices guaranteed by an effective support system operated by the * Paddy Marketing Board (PHB). In contrast, marketing of other crops is much less certain: markets tend to be narrow, prices fluctuate significantly in relation to supply conditions, and overall price levels for many commodities are kept relatively low. Depressed prices are attributable to competition from coconut products (prices of which are depressed by significant export duties and trade controls, affecting in particular oilseeds and fodder crops) and to trading and exchange rate policies (affecting particularly the export  -12- items such as chillies, onions, and pulses). Some broadening of market outlets has occurred - for instance, through the commissioning of soya-processing facilities (soya has major long-term potential), and PMB is strengthening its role in price support programs. Nevertheless, further developments and policy changes will be necessary before the risks associated with these other field crops are reduced to levels comparable to those for paddy. 3.16 Axricultural Credit. Although price incentives will be the primary determinant of farmers' willingness to introduce better methods, the weak performance of agricultural credit institutions in Sri Lanka has also affected the adoption of improved cropping practices. Institutional credit is handled primarily by the two state-owned commercial banks, the People's Bank (PB) and the Bank of Ceylon (BC). PB operates through :ulti-purpose Cooperative Societies and their banking divisions called Rural Banks. BC operates directly, in particular through its branches in ASCs. National agricultural credit policy is established by a Rural Credit Advisory Board, chaired by the Governor of the Central Bank, with the Central Bank's Depart- ment of Rural Credit functioning as the secretariat. The major problem has been high default rates on agricultural loans. Reliable figures for defaults are not available; BC and PB reporting about 16% default (not considering about 54% delays in payment as default). The situation was aggravated during maha of 1977-78, when GOSL instructed the banks to relax lending criteria. To improve loan recovery, lending criteria were tightened in 1979, and the volume of lending fell sharply. The banks are trying to regularize their position, but the overall situation remains somewhat unsatisfactory. Agricultural credit is being improved with support from the Asian Development Bank. IV. THE PROJECT A. Proiect Obiectives 4.01 The project's main objectives are to increase agricultural production under the seven major irrigation systems and to introduce an effective program for their integrated management, in general through: o The rehabilitation of the systems for optimum utilization of available water; and o Support for the Program for Integrated Management of Major Irrigation Systems (INMAS) at national level, and specially for the rehabilitated systems to ensure proper operation and maintenance (O&M) and distribution of irrigation supplies and the complementary programming of agricultural activities and inputs.  -13- B. Proiect Components 4.02 The project components are broken down by implementing agency to identify clearly responsibilities and to facilitate allotment of funds. Ministry of Lands and Land Development (Base Cost US$1.26 N) 4.03 MLLD will be responsible for INNAS, which is a national program and has been designed to meet the needs of integrated management, drawing on the experience of methods of water allocation and distribution (Minipe), institu- tion building (Gal Oya), diagnostic approach to field problems (Rajangana and Kala Weva), and the resolution of problems encountered (Twenty-five Project Management). The program will establish an organization mainly compris,ng conuittees of farmers representatives and concerned government officials at various levels (Field channel, Distributary, and Project) and skeleton management staff to: (a) Schedule agricultural cropping patterns and activities includ- ing input availability, in relation to the physical character- istics of a particular system so as to maximize agricultural output; (b) Plan distribution of water among the farmers in accordance with their needs and in conformity vith crop and soil require- ments, allocating shortages in supply in the best interest of agricultural production; Cc) Activate maintenance programs and ensure that farmers fulfill their maintenance obligations defined under the new management program; (d) Ensure that farmers' discipline is maintained ia preserving the irrigation network (especially in years of short supply), emphasizing cooperation and participation but if necessary enforcing legal actions to control damage to system or unauthorized water use; (e) Arrange GOSL allotment of adequate funds for O&K and make these available to ID according to an agreed program; (f) Propose to GOSL water charges to be levied, gradually changing these to reflect irrigation service provided, through different rates for each crop and varying with crop seasons, and ensure timely payment of water charges by the farmers; (g) Establish criteria and assessing system performance, both on the water distribution system and on-farm, in each crop season; and  -14- (h) Monitor and evaluate socioeconelmlc and agricultural benefits resulting from rehabilitation and integrated magement. 4.04 All cultivation activities (crop calendar, choice of crops, input servicing, credit and marketing) will be linked to water management through the Project Manager (PK) approach, thus strengthening the farmers and manage- ment decision making for seasonal (based on anticipated inflows) and within-season (in response to rainfall) water utilization. For long term needs of integrated management, a team of specialists with close liaison with the specialists working in the line agencies would identify actions to be taken and recommend measures for implementation. These specialists would also provide technical support and guidance to PM. INMAS would thus improve coordination between ID personnel operating and maintaining the irrigation systems, other line agencies staff providing agricultural services and inputs, and the farmers. Line agencies would continue to have their separate identities; line agency staff would retain their career prospects in the parent departments, but their functional responsibility for a specific irrigated area vould be guided and monitored by PM. Farmer organizations would be promoted through Institutional Officers (10) attached to PM, and farmer decision making for cropping patterns, inputs, and rotations of irrigation supplies as well as participation in scheme management would be encouraged. District level concerns would be coordinated through the DAC. A special subcommittee of DAC, with the Regional Deputy Director (DD) of Irrigation as Secretary, will be responsible for resolution of the project problems. Interaction with the Executive Committee of the Development Coun- cil would be established by suitable representation in this subcommittee. At the national level, an Irrigation Management DivisiontIND) has been created in the MLLD, and the Division Director has been appointed. The team of specialists will be located in this division. The organization for IND is given in Chart 26291. 4.05 To assure smooth cooperation, the MLLD has defined the pattern of working of IMD and its relationships with various line agencies, ID in par- ticular. However, working relationships would take time to establish and periodic reviev of the system would be necessary to establish if management objectives were met and whether changes were required. Assurances have been obtained that GOSL shall conduct a mid-term review to determine the effec- tiveness of the administrative changes introduced in respect of INMAS, and a report on the findings of such review shall be furnished to IDA not later than December 31, 1987. 4.06 The project will finance: ti) investment costs for setting up InD such as (a) the office and audiovisual training equipment, (b) vehicles for staff movement, and (c) buildings for the project staff where available buildings fall short of requirements; (ii) recurrent costs for staff and office costs for the duration of the project; and (iii) training and research studies: (a) induction training of 27 IOs, (b) training of 7 PM; in basic management, farmer organizations and coordination, techniques of monitoring and evaluation, (c) training of 120 field staff in farmer organizations and O&M, (d) seminars for district level staff, (e) training of farmers, and (f)  -15- research studies for specific problems related to water management to be conducted by the Department of Agriculture. Local training will be given at the Agrarian Research and Training Institue (ARTI), Project for Advancement of Settlement Expertise (PASE), and the various institutes for agriculture; overseas training will be arranged at Irrigation Projects/Institutes con- nected with water management. The IND cost is estimated at Rs 50.34 million. Irrigation Degartuent (Base cost US$26.54 M) 4.07 ID will be responsible for the rehabilitation of works, the main component of the project. Rehabilitation of the tanks and the irrigation network has been planned in accordance with the criteria given in Annex 4.1. The irrigated area under each system will include areas irrigated by the main tank, by subsidiary tanks in the same hydrological system, by feeder channels to the tanks, and by return drainage flows, but will exclude the areas under lift irrigation, Government agriculture farms, and under sugarcane cultiva- tion by the Sugar Factory. The channels for irrigation in Government farms and industrial estates will be rehabilitated by the respective owners at their expense, and GOSL would make this rehabilitation requirement a condi- tion for continuation of irrigation supplies. Areas served by the lift irrigation systems are not being irrigated due to farmers' lack of interest in paying for diesel oil or electricity. GOSL has been considering various possibilities for utilizing these facilities including handing over lift irrigation facilities to the farmers. Assurances have been obtained that GOSL shall, by December 31, 1985, furnish to IDA for its review a program for the utilization of lift irrigation systems and shall take necessary action to implement said program, as agreed by IDA, by June 30, 1986. 4.08 A pilot program (Annex 4.1) will be included in the rehabilitation to ascertain the impact of technological improvements in control of irrigation supplies on about 150 to 200 ha in each of four selected systems. The pilot program will make no changes in the system layout, but works will be modified to include the technological improvements for better control on water levels in the channels and flow through offtakes and farm turnouts. An expatriate consultant will be engaged for about 4 months to assist ID in preparing type designs and implementation of improved works. Assessment of im;act would be an important aspect of the pilot program. It would also be necessary to make the farmers fully aware of the provisions in the pilot area through special contacts, so that there will be no tendency on their part to damage the structures. Special staff will be required to distribute water in the ini- tial stages until the farmer groups take over distribution. Data will be collected by the same staff to monitor the comparative advantages or disad- vantages of the pilot area over the conventional area, in accordance with a program agreed by IDA. Assurances have been obtained that in order to make the farmers familiar with the pilot area program, GOSL shall deploy, by September 30, 1986, additional staff in the pilot areas. GOSL shall collect data on the relative performance of the pilot program, in accordance with procedures acceptable to IDA, and shall, by December 31, 1987, furnish to IDA the data collected.  -16- 4.09 Rehabilitation work would involve measures for (a) increasing irriga- tion water availability, assuring safety of storage works, and minimizing losses on feeder canals, and (b) improving the channels network to assure adequate capacity to deliver the required supplies throughout the command, better control on distribution of water, and drainage. Rehabilitation of storage works and feeder canals includes: (i) Kantalai System - improvement of drainage of top of dam and downstream slope, removing trees on dam which can be removed without endangering safety and repairing gullies, and improv- ing Mineriya - Kantalai feeder channel bringing Mahaweli supplies to the system, as well as strengthening upstream face protection of Wan Ela Tank and increasing capacity of its feeder channel; (ii) Iranamadu System - raising the full supply level by one foot to utilize part of the spill (about 30% of inflow), repairs to spillway damage (1984 floods), and providing inverted filter at dam toe in areas showing seepage; (iii) Giant's System - raising the full supply level of tank by one foot and remodelling the tank feeder for 28.3 m3/sec (1,000 cfs) to divert to the tank part of the spill at Tekkam anicut (about 40% of inflow); (iv) Rajangana System - additional riprap on the upstream face of Angamuwa Dam; (v) Nachchaduwa System - improvement of the tank feeder bringing Mahaweli supplies and its connection with the upper left bank canal; and gvi) the Huruluwewa System - improvement of tank feeder bringing Mahaweli supplies. Channel network rehabilitation would include raising banks wherever needed to assure safety at maximum discharge, erosion protection, and works for controlling and measuring the flow. A sinmary of main works required and the total cost of rehabilitation of each system is given below:  -17- Remodelling Existing Total J Net and New Improved Improved Rehab. System Area Earthwork Structures Offtakes Turnouts Cost Cost/Ra (ha) (000' m ) (nos) (nos) (nos) (Rs M) (US$) Kantalai 6,990 350 154 356 4,590 100.3 560 Morawewa 1,960 35 64 163 1,345 27.6 550 Iranamadu 9,430 920 227 311 1,632 171.9 709 Giant's 12,460 1,850 294 754 8,461 192.9 602 Rajangana 5,910 680 168 29 6,545 151.9 1,000 Nachchaduva 5,400 450 112 145 1,463 147.6 1,064 Huruluwewa 4,090 250 49 84 2,591 58.6 557 Total 46,240 4,535 1,068 1,842 26,627 850.8 720 sJ Includes modifications to structures in pilot areas, and contingencies (Annex 5.1, and Annex 5.2, Table 2) 4.10 Facilites needed to rehabilitate the works and for maintenance thereafter will be created under the project. These include: (a) Equipment for (a) supplementing the available construction machinery to allow simultaneous work on all the seven schemes (Rs 182.81 million, or US$6.65 M), (b) strengthening capacity for transport of materials for the operation and maintenance (Rs 12.76 million, or US$0.46 M), (c) hydrological data collec- tion and modelling of systems (Rs 3.94 million, or US$0.14 N), (d) surveys and designs (Rs 3.22 million, or (US$0.12 M), and (e) quality control laboratory (Rs 0.67 million, or US$0.02 M) as well as vehicles (Rs 9.11 million, or US$0.33 M). Cb) Buildings (7 for Technical Assistants, 6 for Work Supervisors, and 5 for operators) to supplement the available residential accommodation for O&M (R. 6.79 million, or US$0.25 M). (c) Surveys, aerial and soil, for preparatory work for planning rehab- ilitation are currently financed under VIRP (Credit 1160-CE); funds will be provided under the project to complete the remain- ing surveys (Rs 1.93 million, or US$0.07 million). (d) Technical Assistance. (a) Consultancy to develop hydrological models for each scheme permitting use of microcomputers (Rs 4.27 million, or Us$0.16 H); (b) an expatriate consultant for 4 months for the pilot program (Rs 2.00 million, or US$0.07 M); (c) an expatriate systems management consultant for a period of about 18 months (Rs 4.33 million, or US$0.16 M) to provide technical guidance to rehabilitation (the consultant has already been engaged on terms of reference agreed by IDA and is currently  -18- financed under VIRP, Credit 1160-CE, for a period of six months, which excludes 18 months under the project); and (d) local consultants to be employed for recommending and implementing measures for improving the management information and project accounting (terms of reference have already been agreed by IDA; expected cost is Rs 9.99 million, or US$0.36 X over a period of 5 years). te) Trainint. The project will finance training of ID staff locally as well as send a few selected persons overseas for specific training in India, Pakistan, Thailand, Philippines, and Great Britain. Two engineers will pursue post-graduate degree course in hydrology and water resources development. Training is estimated to cost Rs 10.96 million, or US$0.39 M. Land Ccomissioner's Department (Base cost US$0.72 N). 4.11 LCD would be responsible for regularizing farmers encroacbments on adjacent lands, clearly defining irrigation rights, and major repairs to LCD roads in the project area. Encroachments (Para 3.06) need to be regularized to clearly establish the area irrigated under each system. For safety of structures and convenience of O&M operations, certain land encroachments have to be cleared, which is an integral part of rehabilitation work. Land owner- ships have to be ex3mined and regularized (specially in Moraweva) where original settlers, to whom land was allotted, left the area due to water shortages and their lands were taken over by others in unauthorized manner. Irrigation rights are generally clear, but uncertainty exists in some areas such as the relative rights of purana and settlement lands and the priority for the Sugar Corporation in Kantalai. Repairs to LCD roads would improve the movement of agricultural products and inputs between the farms and the market. The project would finance: (i) the incremental costs of LCD's operations (Rs 1.25 million, or US$0.5 M) to regularize landholdings under this project; (ii) purchase of two jeeps (Rs 0.72 million, or US$0.03 N); and (iii) repair of 35 miles of gravel-topped LCD roads and 255 miles of unpaved LCD roads (Rs 29.67 million, or US$1.06 N). Department of Agriculture (Base cost US$0.32 M). 4.12 The Department of Agriculture would be responsible for conducting studies on catchment management and on-farm water management. The study for catchment management (which would be conducted on Nachchaduva and Noravewa catcbments) would provide (i) the basic information needed to model the catchment water balance which will be used to develop the water balance simulation models of each system, (ii) the pattern on which catchment manage- ment organization should be developed so as to ensure erosion control as well as allocation of water and lands to different uses to maintain availability of sufficient runoff at the major tanks, and Ciii) changes to existing legis- lation, if necessary, to provide the legal means to ensure proper catchment management. Unless efficient catchment management is established, upper vater use could seriously reduce the availability of water to any system.  -19- On-farm water management studies vill be conducted in selected areas compris- ing small hydrological units to provide a clear understanding of the utiliza- tion of applied water and of measures to improve the efficiency of utiliza- tion. The project would finance (i) incremental staff and other costs (Rs 7.69 million, or US$0.27), (ii) equipment for data collection, computer for data processing and modelling, (Rs 3.32 million, or US$0.12 N), (iii) pur- chase of 3 jeeps (Rs 1.16 million, or US$0.04 K); and (iv) training of a few staff in foreign centers involved in similar studies (Rs 1.02 million, or US$0.04 M). Assurances have been Obtained that GOSL shall not permit any further development of irrigation by tanks or diverions veirs in the catch- ment areas of the seven irrigation systems included in the project, and no diversions, other than the existing, from Nalwatu Oya shall be allowed upstream of diversion weir at Tekkam. GOSL shall, by December 31, 1987 furnish to IDA its proposal for organizational set up, including legal framework, for proper catcbment control. Department of Asrarian Services (Base cost US$0.34 M). 4.13 In order to assure that farmers will have adequate and timely availability of agricultural inputs, additional facilities need to be provided. The DAS has established additional requirements (after detailed review of existing facilities) of agrarian service centers (ABCs) and stores for fertilizer and other inputs that would be needed so that their respon- sibility towards INNAS is properly fulfilled in the Project area. These are Ci) one ASC (IranamAdu), 35 fertilizer or seed stores of different capacities, and 5 garages (Rs 8.40 million, or US$0.30 M); and (ii) purchase of 3 pickups, 2 jeeps, 17 motorcycles, and 42 bicycles (Rs 4.07 million, or US$0.15 M) for improving movement of staff and emergency supplies, which will be financed under the project. Motorcycles and bicycles will be purchased by the designated DAS personnel against money advanced to them by DAS. University of Peradeniva (Base cost US$0.12 M). 4.14 UOP will be commissioned to follow up on their initial baseline socioeconomic surveys with two further studies: (a) A mid-term evaluation of project implementation, including assessment of farmers participation in and attitudes toward the rehabilitation process, with the aim of making recommendations for corrections, if any, and tb) An evaluation study following completion of the works, which would be used for the preparation of the project completion report. MLLD would assign staff to work with UOP in the data collection. The project would finance (i) the incremental staff costs of MLLD (Rs 1.15 million, or US$0.04 M); (ii) procurement of a jeep (Rs 0.4 million, or US$0.02 M) and a photocopier (Rs 0.1 million, or US$0.004 M); and (iii) the consultancy costs of UOP (Rs 4.23 million, or US$0.15 M).  -20- International Irritation Mana2ement Institute (IInI) (Base cost US$0.77 M). 4.15 III, having the goal of enhancement of independent national capacities to improve the performance of irrigation systems with special emphasis ou irrigation management, was recommended by the "Study Team on Irrigation Management", constituted by CGIAR in 1981. IIM is currently funded by a Support Group comprising 11 members iucluding ADB, Ford Founda- tion, IFAD, Rockefeller Foundation, USAID and GOSL. Its charter ba8 been approved in the fourth meeting of the Support Group (May 22, 1984). IMT is programmed to mainly (i) conduct multidisciplinary research jointly with national agencies using irrigation systems as laboratory, (ii) contribute to professional development through training and short study courses, and (ii) prepare and diseminate research and training materials. The network approach for IINI proposed by the CGIAR Study Team has been further detailed by the "Working Group on Participating and Cooperating units and other Methods of Collaboration' formed by the IIMT Board at its first meeting. Recommenda- tions of this Working Group generally involve (a) establishment of about four branches in different parts of the World with which IItH (located in Sri Lanka) will have extensive interaction, the first to be established in Pakis- tan by the end of 1984, (b) establishment of a network with participating institutions in 10 to 15 locations using a shared methodology and common focus under IIMI sponsorship and guidance, and (c) promotion of exchanges between Ina1 and institutions in several countries. IIn is expected to start functioning in Sri Lanka by the end of 1984, GOSL having agreed to renovate and expand existing housing and other facilities at Digama (near Kandy) and construct new facilities at field sites. RDI is planning to have its "Irrigation Management Course" for participants from South Asia region at II( in early 1985. The facilities at Kandy would include offices for the permanent and visiting staff, a data-processing and computer center, con- ference rooms and classrooms, staff residences and a guest house, and auxiliary facilities. Facilities at field sites would include offices for staff, office equipment, a guest house, and residential quarters. The project would finance the GOSL contribution to the creation of these facilities which will be implemented by the Irrigation Department and MASL. The establishment of IIMT in Sri Lanka would be in the project's interest and INMAS could benefit considerably from IIMI advice, especially if one of the field sites of IIMI is identified at one of the project schemes. Estimated cost is Rs 27.29 million, or US$0.99 M. C. Water Availability 4.16 Nahaweli Supplies. Nahaweli supplies diverted through the Polgolla tunnel directly augment the inflow from the tank catchments in the case of Kantalai, Huruluweva, and Nachchaduva, while return flow from area H (irrigated from Nahaweli supplies) augments the inflow to Rajangana. Allot- ment of Mahaweli supplies to all the irrigation systems receiving Nahaweli support is studied every year by the Water Management Secretariat (WMS) of the Mahaveli Authority of Sri Lanka (NASL). Allotments are determined before the maha and yala seasons, when cropping patterns are determined from the  -21- availability of water in the tanks at the beginning of the seasons, with an expectation of inflow corresponding to 80Z probability of rainfall. The allotments are subsequently reviewed every month. Actual issues are based on ten-day reviews by WKS, which are approved by the Director General of MASL. Since 1976, when Polgolla became operational, annual diversions have generally exceeded 1,000 MCM. These diversion were possible because irriga- tion and power demands on the Mahaweli had not developed. The situation would change with the completion of Kotmale (scheduled for 1984), Victoria (scheduled for 1985), and Randenigala (scheduled for 1986) reservoirs on the Mahaweli, and power generation at these dams would create a conflict of interest. Therefore, the cropping patterns on the three Mahaveli fed systems will have to be developed on an average of 875 MCM (a range of 675 to 1100 MCN depending on dry or wet year) annual diversion through Polgolla tunnel (para 3.11). 4.17 The releases from Mahaweli reach the tanks through long feeder chan- nels that are partly natural streams. The feeders often pass through small village tanks or supply unauthorized water to such tanks for irrigating local areas. Unauthorized direct tappings for irrigation are also common. Nahaveli releases for the tanks are thus reduced en route. To control such unauthorized use, the ID has recently issued orders that the management responsible for 0&M of tanks would also be responsible for the O&M of feeder channels (other than those maintained by Mahaweli). ID would in addition introduce a system of monitoring the efficiency of transmission of the feeder channels. Assurances have been obtained that GOSL shall make adequate arran- gements to determine the efficiency of transmission of Mahaweli supplies by the feeder channels to the project area in each crop season by appropriate gauging of supplies at control poiats, and shall furnish a report to IDA within 3 months after each crop season. 4.18 Water Balance Studies. For each scheme, the ECL has studied the shortages that will be faced in irrigating different cropping patterns given the water availability (based on 30 years of inflow data, improved by cor- relation techniques, and rainfall runoff relationships) including Mahaveli diversions and existing tank capacities, after the communal and industrial water demands are fully met. Annex 4.2 simmarizes the main results. The scope of ECL studies is being enlarged to cover the entire area under each interconnected system (earlier studies were for main tanks only which cover most of the area). The studies will continue till finalization of simulation models. The criteria for establishing an acceptable cropping pattern is a shortage of water in only 4 years out of 30 years, after permitting control- led deficits of 30% when reservoir storage goes below 30%. Such criteria though somewhat arbitrary, are reasonable and attempt to rationalize the planning for tank fed irrigation in Sri Lanka. Based on the hydrological studies, the assured cropping intensities with the project have been taken as: Kantalai, 150%; Huruluwewa, 150%; Nachchaduwa, 150Z; Rajangana, 188%; Iranamadu, 143%; Giant, 95%; and Morawewa, 150%. Annex 7.1 gives the present and anticipated cropping intensities with and without the project.  -22- D. Status of Project PreParation 4.19 The preparatory work for the implementation of the project is well advanced. Detailed planning for rehabilitation and detailed designs are progressing well on realistic assumptions for land areas. These assumptions will be confirmed by aerial surveys, for which flying has been done. Staff for the preparation of designs has been posted in the field divisions. Con- struction drawings are in progress and will be completed for the first-year works by end of September 1984. No land has to be acquired. Transfer of equipment with necessary operating staff Arom completed projects and comple- tion of tender formalities for contract works will be achieved in time for commencement of work after the maha rains of 1984-85. The project is there- fore well prepared to start rehabilitation works from Juae 1985, when local funds would be available. E. Implementation Schedule 4.20 Rehabilitation of systems will be carried out simultaneously on all the schemes and phased over six years (Annex 4.3). Each year, the program on an individual system would cover the complete rehabilitation of a section of the system. Areas thus developed yearly would be: Area Rehabilitated (ha) Scheme 1985 1986 1987 1988 1989 1990 Total Ksantalai 308 1,541 1,391 2,043 1,048 659 6,990 Morawewa 265 464 710 341 180 - 1,960 Iranamadu 700 900 1,500 2,000 2,200 2,130 9,430 Giant'8 Tank 475 1,858 2,958 3,603 1,500 2,066 12,460 Rajangana 684 948 1,207 1,345 936 790 5,910 Nachchaduwa 250 1,400 1,400 940 1,100 310 5,400 Huruluwewa 458 770 589 1,035 750 488 4,090 Total Project 3,140 7,881 9,755 11,307 7,714 6,443 46,240 Works for the pilot program will be implemented in 1986. Works to be per- formed by other agencies will be scheduled to match the rehabilitation program. Renovating and expansion of housing and other facilities for IIMI will be completed in 1986.  -23- V. PROJECT COSTS, FINANCING. AND PROCURENENT A. Proiect Costs 5.01 The total project is estimated to cost about Rs 1206.2 H (US$43.2 M) including Rs 333.5 N (US$10.2 M) for price escalation. Cost estimates include Rs 72.9 N (US$2.6 M) for taxes and duties. Foreign exchange costs would be about Rs 505.2 N (US$18.2 M) or 40 of the total. The project costs for each implementing agency are summarLzed in Table 5.1. Details are given in Annex 5.1, Tables 101 to 115. The cost of civil works is based on detailed quantities of work and unit costs (mid-1984 prices) estimated for sample areas covering about 10% to 15Z of each scheme. The estimated unit costs of civil works include depreciation of construction equipment. Since construction equipment is partly available by transfer from other projects (free of cost) and is partly purchased under the project, the depreciation of construction equipment corresponding to force account work is deducted from the total cost to avoid double counting for investment. Physical contingency is taken as 10% overall, and price contingencies are as shown in the foot- notes to Table 5.1. Because estimated domestic inflation rates are generally expected to remain higher than foreign inflation rates, particularly in the early years of implementation, year specific exchange rates have been assumed to maintain purchasing power parity. The project costs have also been broken down into different categories of work for financing and procurement plans. Civil works account for 70%, equipment procurement 19%, training 4%, and incremental staff and administration costs 7%. B. Financing 5.02 The project has been approved by the GOSL cabinet for financing from 1985 onwards, and will be financed by GOSL budget allocations from its own resources and funds available to GOSL from IDA (credit), CIDA (credit and grant), and SDC (grant). The GOSL budget for the project would be voted in the MLLD for IND. IND would then allocate funds to various departments participating in the implementation of the project. In view of severe con- straints on the availability of local funds, the project could not be financed earlier than 1985. Sufficient local funds to implement the project are expected to be forthcoming, and GOSL would not apply cuts after initial allotments, which had to be done in 1984. Such cuts, if imposed on MIRP works, would upset the construction program and commitments under the con- tracts. Assurances have been obtained that GOSL shall make adequate and timely budgetary provisions for the implementation of the Project in line with the financial requirements of the Project. 5.03 The financing of the project by different donors (IDA, CIDA, and SDC) and GOSL is shown in Table 5.2. GOSL contribution would cover the incremen- tal staff, engineering and administration, and taxes and duties.  Table 5.1 Project Cost'Summary (Rs '000) (US '000) Z Totel I Total % Forelsn Base I Foreign Bau Local Foreign Total Exchange Costs Local Foreign Total Exchmni Costs -:::::::: S::S::=:_ S_":g:* uaua::ana u*xagg -- --xxzS :::s=g:3 Jagansg Baaas A. HLLD IRRIDATION MOT DIVISION 25,839.0 70446.8 33,285.8 22 4 979.9 22.4 1,262.3 22 4 B. IRRIGATION DEPARTIENT to REHABILITATION SCHEMES KANTALAI 37.688.B 20.671.1 589359.9 35 7 1U29.2 783.9 2.213.1 35 7 MORAIEWA 11.059,3 6.051%2 17P110,5 35 2 41994 229.5 648. 35 2 IRANA"ADJ 67,568.1 34.555.2 102.123.3 34 13 2,562.3 19310.4 3.972.7 34 13 GIANT TANK 67,428.4 460812,7 114,241,1 41 14 2,557.0 1775.2 4,332.2 41 14 RAJAMA 59.105.2 28.21.1 87.92602 33 11 2o241o4 1tM02,9 3,334.3 33 11 NACHCHADUV 57,977.3 31,956.0 890953.4 36 11 2199.4 1W211.8 3,411.2 36 11 HURULUWEVA 22,876.9 12.041.3 34.918.2 34 4 967.5 456.& 1,324.2 34 4 Sub-Total REHABILITATION SCHEMES 323,723.9 180,908.5 504,632,5 36 64 12,276.2 6,860.4 19,136#6 36 64 2. PROJECT SUPP& RT 55,271.7 140,078,9 195,350.6 72 25 2,096.0 5i312.1 7,400.1 72 25 Sub-Total IRRIGATION DEPARTMENT 378,995,6 320,987.5 699,983.1 46 80 14,372.2 12,172o4 26,544.7 46 88 C. LAND IWMIISSIOWER DEPT 10,26207 8,934.4 19,097.0 46 2 389.2 335.0 724.2 46 2 De DEPARTMENT OF AGRICULTURE 5,896,6 2,655.3 8,551.9 31 1 223#6 100.7 324,3 31 1 E. DEPARTMENT OF AGRARIAN SERVICES 5,181,6 3,652.4 8,834.0 41 1 196.5 138.5 335.0 41 1 F. UNIVERSITY OF PERADENITA 3,041.4 226.4 3,267.8 7 0 115.3 8.6 123.9 7 0 9. INTERNATIONAL IRRIGATION NOT INSTITUTE 13,671#0 6,640,8 20#311,8 33 3 518.4 251.8 770.3 33 3 Total BASELINE COSTS 442,89880 350,443.5 793,331.5 44 100 16,795.1 13,289,5 30,084.6 44 100 Physical Contingencies 444289,B 35,044.3 79,333.1 44 10 1,679,5 1328#9 3p008,5 44 10 Price Continiencies 213,81207 119,713o4 333,526.1 36 42 6,621.0 3,551.8 10,172,8 35 34 Total PROJECT COSTS 700,989,4 505,201.3 1,206,190.7 42 152 25M0957 18,1702 43t265.9 42 144 Note: Price Contingencies (according to Bank estimates) are taken as: Years 1985 1986 1987 1988 1989 1990 Local(%) 12 11 10 9 7.5 6 Foreign(X) 8 9 9 9 7.5 6 I , .  -25- Table 5.2: Financing of Proiect 1/ (US$ million) Financini by Item Cost GOSL IDA CIDA SDC Foreign 11.2 - 7.0 1.4 2.8 1. Civil Works, etc Local 19.0 7.1 7.2 1.6 3.1 Total 30.2 7.1 14.2 3.0 5.9 (25%) (45%) (10%) (20Z) Foreign 6.6 - 2.8 3.8 - 2. Equipmeut & Vehicles Local 1.8 1.8 - - - Total 8.4 1.8 2.8 3.8 - 3. Tech. Assistance, Foreign 0.4 - - - 0.4 Training, Survey,etc. Local 1.3 - - - 1.3 Total 1.7 - - - 1.7 4. Engineering & Adm. Foreign - - - & Incremental Staff Local 2.9 2.9 - Cost Total 2.9 2.9 - - - Foreign 18.2 - 9.8 5.2 3.2 (53Z) (29Z) (18x) Total Project Local 25.0 11.8 7.2 1.6 4.4 (47%) (29%) (6z) (18%) Total 43.2 11.8 17.0 6.8 7.6 % Financing of Total Cost (27%) (39%) (16%) (18%) Z Financing of Cost less Taxes (22%) (42%) (17%) (19%) Note: Figures in parentheses are percentage financing of items by GOSL, IDA, * CIDA, and SDC. 1/ Share of financing by implementing agencies would be: bLLD ID LCD DA DAS UPO IIMI Total GOSL 1.10 9.82 0.25 0.31 0.12 0.02 0.25 11.87 IDA 0.37 15.36 0.48 0.12 0.22 0.01 0.44 17.00 CIDA 0.06 6.43 0.14 - 0.05 - 0.12 6.80 SDC 0.28 6.60 0.26 0.04 0.06 0.18 0.18 7.60 Total 1.81 38.21 1.13 0.47 0.45 0.21 0.99 43.27  -26- IDA would be financing part of the equipment and civil works; CIDA would provide parallel financing for the remaining equipment and joint financing with IDA (CIDA contribution) for civil works; SDC would provide joint financ- ing with IDA (SDC contribution) for technical assistance (para 4.10(d)), surveys (paras 4.10(c) and 4.14), training (paras 4.06(iii), 4.10(c) and 4.12(iv)) and civil works. SDC's conditions for funding are similar to those of IDA. External financing of the project amounts to nearly 78% of the total project cost less taxes (IDA 42%, CIDA 17%, and SDC 192). The relatively high share of total foreign financing is justified given the serious budgetary constraints prevailing in Sri Lanka. To cope with the investment restriction measures in line with the recommendation of the IMF and IDA, the Government's investment program is being curtailed substantially. Maximum possible foreign participation in project financing would be desirable. C. Proiect Special Account 5.04 The estimated schedule of expenditures is given in Annex 5.2, Tables 1 and 2, and reflects the implementation program discussed in paragraph 4.20. Funding of civil works by IDA, CIDA and SDC will be on a reimbursement basis (para 5.09) requiring GOSL to finance the works initially. To relieve GOSL from making available necessary funds for initially financing the works and to facilitate the project's timely implementation, a Project Special Account (PSA) in US dollars will be established for the financing of civil works. Funding by IDA, CIDA and SDC will be through the Special Account. The PSA will be established and operated under procedures acceptable to IDA. It will be operated by the Project Director (MIRP) in IMD (para 6.01). Opening of the PSA in the Central Bank will be a condition of Credit effectiveness. D. Procurement 5.05 Equipment and_vehicles (US$8.4 M). Funded both by IDA and-CIDA, equipment and vehicles will be procured under the procedures of the respec- tive donors. The equipment to be separately funded and the phasing of delivery is shown in Annex 5.3. IDA-financed equipment for an estimated c.i.f. cost of US$2.8 M will be procured through international competitive bidding (ICB) in accordance with IDA guidelines of August 1984. The equip- ment procurement would be phased under three tenders, the first covering most of the equipment will be issued in 1984, the other two being issued in 1985 and 1986. ID will be responsible for the entire procurement of equipment including the requirements of other implementing agencies. Goods manufac- tured in Sri Lanka may be granted a margin of preference in accordance with, and subject to the guidelines. Hydromechanical equipment and small off-the-shelf items costing less than US$10,000 each, for which international tendering is impracticable, would be procured after comparing prices from not less than three independent suppliers in accordance with GOSL procurement procedures acceptable to IDA. Such purchases will not exceed the equivalent of US$0.25 M. 5.06 Civil Works (US$30.2 M). Such works will be individually small, simple, labor intensive, and scattered. Furthermore, the works will be  -27- subject to the constraints of ongoing cultivation and therefore would not be suitable for ICB. Civil works vill be partly contracted under local competi- tive bidding (LCB) and partly carried out by force account (para 6.06). ID has detailed procedures for calling bids and award of contracts which are satisfactory to IDA. All canal structures and earthwork on the smaller channels will be executed under unit price contracts with local contractors. Since each year rehabilitation program would cover the channel network for a specific portion of the command, contracts will be awarded separately for each years work on channel earthwork etc, and structures. All work contract over US$100,000 would be subject to IDA review prior to award. Other con- tracts would be subject tu selective post award review. 5.07 Procurement procedures are summarized in Table 5.3 below: Table 5.3: PROCUREMENT PROCEDURES (us$ million) Procurement Method Proiect Element ICB LCB Other Total Cost Civil Works - 18.1 12.1 30.2 (8.4) (5.8) (14.2) Equipment 3.4 - 5.0 8.4 (2.5) (0.3) (2.8) Technical Assistance - - 1.7 1.7 and Training Engineering and - - 2.9 2.9 Administration Total 3.4 18.1 21.7 43.2 (2.5) (8.4) (6.1) (17.0) Note: Figures in parentheses are the amounts to be financed by IDA. E. Disbursements 5.08 Disbursement of IDA credit, CIDA contribution, SDC contribution would be for: o Civil works - by IDA 45%, by CIDA 10%, and by SDC 20%. O Equipment and vehicles - by IDA, 100X of the foreign expenditure for procurement abroad; and 100% of ex-factory costs, or 80% of the cost of procurement locally. o Technical assistance, surveys, and training - by SDC, 100%.  -28- 5.09 Rehabilitation works would be carried out by force account and a number of small contracts and payments would be made at suitable intervals as the work progress. To avoid processing of large number of applications of small denomination, disbursements for civil works would be made against statements of expenditures, which will be audited (concurrent audit) accord- ing to procedures acceptable to IDA. Supporting documents will be retained by GOSL for one year after completion and made available to supervision missions whenever required. The borrower will submit reimbursement applica- tions only when the amount expended has reached the equivalent of US$20,000 or more. All expenditures disbursed on statements of expenditure will be audited by independent auditors, acceptable to IDA. For civil works con- tracts exceeding Rs 500,000, disbursement would be made against full docu- ments in accordance with IDA's standard requirements. Disbursement against contracts for items other than civil works will be documented in accordance with IDA's standard requirements. To assist in record keeping for contracts awarded under this project, IMD would submit to the Bank on a monthly basis a register of contracts; this register would be submitted for the duration of project implementation. IDA would be holding CIDA funds in trust for CIDA's funding of civil works, and payments will be made by IDA on CIDA's behalf. The conditions of the CIDA credit and grant will be detailed in the "CIDA Contribution Agreement" negotiated between COSL and CIDA. Fulfillment of the effectiveness conditions of this agreement would be a condition of Credit effectiveness. IDA would be implementing agency for SDC funds, would provide decisions for all matters concerning technical assistance and civil works, would review withdrawal applications, and would request the SDC to make payments on withdrawal applications approved by IDA. The conditions of the SDC grant negotiated between GOSL and the SDC will be detailed in the "SDC Contribution Agreement." Fulfillment of the effectiveness conditions of the agreement would be a condition of Credit effectiveness. Disbursements for equipment and vehicles would be made against full documentation and would be subject to normal IDA procedures. It is anticipated that the final disburse- ments of the IDA credit, CIDA grant and credit, and SDC grant would be made by June 30, 1.991, about six months after the scheduled completion of project implementation. The estimated IDA disbursement schedule is given in Annex 5.4. These dibursement projections are based on the standard sectoral disbursement profiles for IDA credits, deemed appropriate for an irrigation rehabilitation project of this type. F. Accounts and Audit 5.10 IND, ID and other Departments are subject to normal GOSL expenditure and auditing procedures, which are considered satisfactory. Project accounts would be maintained by the respective agencies, separately from overall departmental accounts, to permit auditing by an independent auditor. The IMD would be preparing the project budget, on the basis of requests from all departments concerned, and would submit it to NLLD for approval by the Project Steering Committee (see paragraph 6.02). On receipt of funds from MLLD, IMD would allocate these to all departments. IND would also maintain the consolidated project accounts by collecting the disbursement information  -29- and documents from all the departments to which funds are released and would submit reimbursement claims to IDA. IMD would submit to IDA annual statement of expenditure categorywise within six months after the end of each fiscal year on a pro forma agreed by IDA. The other departments and agencies would be responsible to the Project Director (MIRP) in IMD, for preparing annual budget requests, maintaining accounts of expenditures, and preparing and furnishing audited accounts in support of their claims for reimbursement. Assurances have been obtained that GOSL shall; (a) establish and maintain separate accounts for the project; (b) have the accounts audited both on concurrent basis for SOEs and annually by an independent auditors acceptable to IDA; (c) furnish to IDA, for review, certified copies of audited accounts and the auditor's report as soon as available, but in any case not later than nine months after the end of each fiscal year, including a separate opinion by said auditors whether the proceeds withdrawn on the basis of statements of expenditure have been used for the purposes for which they were provided. VI. ORGANIZATION AND MANAGEMENT A. Overall Project Management 6.01 The newly established IMD in the MLLD (para 4.04) will be the project implementing agency. An Additional Director in this division will be eesig- nated Project Director (MIRP), with the sole responsibility to implement the project and to coordinate the activities of the various participating depart- ments and agencies. A Project Officer (MIRP) would be appointed in the ID to be responsible for coordinating all the activities of the ID for rehabilita- tion works. Appointment of the Project Director and Project Officer would be a condition for effectiveness of the Credit. 6.02 A Project Steering Committee in the MLLD, chaired by the Secretary, with representatives from ID, MADR, DA: DAS, LCD, the Ministry of Finance and Planning, the Department of External Resources, UOP, and other agencies that may be included, will be responsible for providing overall control and guidance. It will in particular on similar lines as in VIRP: (i) approve annual programs and budgets, (ii) allocate funds to the various departments and agencies involved, (iii) review and monitor project progress, (iv) resolve implementation and other problems that may emerge, and (v) approve changes, with IDA agreement, that may become necessary in project design during the course of implementation. The Project Coordinator will act as secretary to the Committee. The formation of the Project Steering Commit- tee would be a condition of Credit effectiveness.  -30- 6.03 At the district level, subcommittees of the DACs would be estab- lished, chaired by the Government Agent (GA), and with the ID's Range Deputy Director (DD) as secretary, to provide policy and program guidance to all major irrigation systems in the district, including those to be rehabilitated under the project. District departmental heads of concerned agencies, as well as the scheme PM, would be members of the subcommittees, which ultimately might also include farmer representatives (as has already been introduced for the Gal Oya project in Amparai district in the context of the USAID-supported water management project). It is GOSL policy to establish these committees in all districts where there are major irrigation schemes. B. Management of Rehabilitation Works 6.04 The ID is responsible for the development of land and water resources for the entire country except the areas which lie in the jurisdiction of the MASL. The department is headed by a Director (DI) who is assisted by an Additional Director. Major planning and design of water projects is done in the head office at Colombo, where functions are divided into seven different sections called Branches. Each Branch is headed by a Senior Deputy Director. The implementation, operation, and maintenance of projects is carried out by the field staff, who are divided into sixteen ranges. Each range is headed by a DD. The ranges are further subdivided into divisions and projects. Each of the major tanks forms one division or project which is headed by an Irrigation Engineer. ID's strength has gradually been rebuilt, after its depletion by transfers to Mahaweli and emigration, and no staffing con- straints are envisaged. Chart 26291 gives the ID organization pertinent to the project. The ID has the capacity to design and implement the rehabilita- tion works, and assistance is required only through specific consultancies (para 4.10). ID recently rehabilitated five tank schemes under IDA Cr. 666-CE, and has thus gained the experience to plan and execute such projects. 6.05 Preparation of DesiRns. Detailed DI instructions outlining proce- dures to be used for rehabilitation planning and design are available. The Design Branch of ID has provided type designs for various structures and these designs are used by the ranges for preparing drawings for specific structures as required. New type designs will be issued for the structures in the pilot area. Design units exist in each range as well as in divisions having major tanks. These units are being strengthened to assure completion of designs fairly in advance of construction. 6.06 Construction and Ouality Control. Rehabilitation will be carried out both by force account and through contracts. Force account work (about 40% of civil works) will be limited to those works for which contractors are not readily available and where construction period cannot be exactly determined in advance due to irrigation requirements. These works are mainly earthwork on canals involving long transport, control of moisture, and mechanical compaction. ID has developed expertise for these works under TIMP and VIRP, and force account execution will be more efficient. The division/project unit responsible for each irrigation system would supervise the construction work. A system to collect samples of compacted earth in embankments and  -31- samples of concrete for structures vill be introduced, and tests vill be made frequently to assure quality of works. Additional technical personnel will be posted in each project unit before the construction starts. ID expects no problems in posting an adequate number of suitable staff for construction supervision and quality control, since major construction activity in the department is rather limited at this time as a result of severe limitation of funds. Quality control of works will be entrusted to the Central Materials Branch of ID operating separately from the Range Offices, as vas achieved successfully under VIRP (Credit 1160-CE). 6.07 Operation and Maintenance. The Operation and Maintenance (O&M) of the irrigation systems would continue to be the responsibility of the ID but would be subject to the overall coordinated program of IND. Sufficient ID staff would be available for operation of each system. The ID staff would be under the Range DD for administ.ration, financial, and technical matters but would have close liaison with the IMD staff and the farmers. The farmers have been damaging irrigation structures in the wrong belief that the struc- tures were reducing their supplies. They need proper education through the Water Management Program to respect the structures and to participate actively in the maintenance of field channels. The ID practice of carrying out maintenance by permanent labor gangs needs to be changed because it is not effective, and such labor is a severe drain on available funds. In the past, sufficient funds have not been available for maintaining the irrigation systems properly. With funds available from collection of water charges (para 7.16), GOSL will be required to budget only the balance required for proper O&M, and availability of funds to ID for good system maintenance should be assured. Assurances have been obtained that GOSL shall Ci) provide sufficient funds for the operation and maintenance costs for irrigation systems rehabilitated under the Project, and (ii) furnish to IDA by January 31 each year a statement on an agreed pro forma indicating that sufficient funds have been allocated to meet the operation and maintenance costs for the Project during the fiscal year. C. The Integrated Manazement Program 6.08 The organization for IMMAS has already been discussed (para 4.04). Each system will have a hierarchy of system-level co ittees. These would be at the project, distributary, and field-channel levels, but considerable flexibility would be maintained, and the formality of the field-channel comittees would be variable. These committees would mainly comprise farmers representatives with a few government officials to assist in decision making. The committees would meet regularly to resolve problems in implementing the rehabilitation and the integrated management programs. The scheme PM would chair the project-level committee and would be responsible for the formation and smooth functioning of all the committees. The appointment of project managers (IMD cadre) for each scheme will be a condition of Credit effective- ness.  -32- D. Monitoring and Evaluation 6.09 Monitoring and evaluation will be the responsibility of the Project Director (MIRP) in IMD, who will issue necessary instructions for collection of data and assign agencies for different types of assessment, including consultancies, as required. During implementation of the project he will monitor progress and prepare quarterly and annual reports on progress of rehabilitation, including agricultural impact. After completion of rehabilitation, IMD would continue to monitor and evaluate performance. Monitoring and evaluation would then have two aspects--(i) technical perfor- mance and (ii) agro-economic impact. The assessment of technical performance would be the responsibility of the ID and the DA, while the assessment of agro-economic impact would be that of DA, assisted by the UOP. IND will prepare annual reports on monitoring and evaluation of rehabilitated areas at the end of each Yala Season and submit it to GOSL and IDA for review. 6.10 Monitoring and Evaluation of Technical Performance. Technical per- formance would be measured by (i) the efficiency of water supply at the farm and (ii) the efficiency of transmission by the channel network. The former will be monitored and evaluated by the DA, and details of the recording procedure would be worked out to have sufficient, reliable data to compare farm irrigation requirements (commencing from the date of land preparation to the maturity of crop) with the actual water delivery and rainfall. ID would monitor and evaluate the losses on the channels and bottlenecks in transmis- sion, if any, to establish the network efficiency. 6.11 Monitorinz and Evaluation of Agro-economic Imnact. UOP has done the baseline survey on the socioeconomic and agricultural conditions prevailing under the systems. IMD through the Project Managers at each system will continue to collect data which would permit comparison with the base. In addition UOP will conduct agroeconomic studies at mid-term and completion (para 4.14). VII. PRODUCTION. MARKETING. FARM INCOMES. AND COST RECOVERY A. Production 7.01 Overview. The seven project systems covering a net irrigable area of about 46,000 ha represent 28% of total lands under major irrigation tanks in Sri Lanka. The production potential of each of these schemes is high given adequate and timely supply of water, adoption of improved water management, inputs, and other supporting agricultural services. Without the project, land resources in these schemes would remain underutilized and there might even be a further deterioration of existing facilities. The expected incremental paddy and soyabean productions from the project will be due to improvements in cropping practices and to higher cropping intensity and yields.  -33- 7.02 Cropping Practices. Changes in cropping to make better use of land and water are recommended under the project, some of which have already been adopted by farmers. These include: (i) dryland cultivation and dry sowing (as practiced in Iranamadu), utilizing the maha rains for seed germination; (ii) transplanting to reduce early water demand for land preparation; (iii) the adoption of shorter-season varieties; (iv) intensive application of weedicides and adoption of row sowing to replace the use of standing water to control weeds; (v) improved on-farm water management (e.g., rotational issue at the farm level, reduced applications, etc.); and (vi) crop diver- sification, especially during yala. 7.03 The efficient land preparation at Iranamadu, the extent of transplanting at Giant's Tank, and the successful soya cultivation at Huruluvewa demonstrate the viability of the improved practices. The incomes earned from soya at Huruluvewa facilitated considerable early land prepara- tion for the 1983/84 maha and demonstrate what can be achieved from improved management. Provided that the overall context for improving irrigation management is established and that the appropriate mix of incentives, cooperation, and discipline is developed, some increases in cropping inten- sities appear possible through the systematic promotion of new agricultural practices responsive to local conditions. 7.04 Cropping Intensity. The assumptions for land use intensity are based primarily on: Ci) the hydrological simulation studies undertaken by ECL, and (ii) the farm survey undertaken by UOP and MLLD during project preparation. Annex 7.1 gives the cropping intensities under present conditions and those expected with and without the project. The extent to which projected benefits are in practice achieved will depend on the management and opera- tional measures supported under the project. 7.05 Despite the relatively high water use efficiencies arising in part from a unique distribution system and impervious paddy soils, cropping inten- sity in the Giant's scheme would reach only 95%. In contrast, croppir.g intensity for Mahaweli assisted schemes - Rajangana, Nachchaduwa, Huruluwewa, and Kantalai -- would vary from 150% to 188%. These increases in intensity would be made possible largely by crop diversification in yala and by ensuring a full or maximum possible maha paddy crop. Cultivation of soya in yala would occupy about 25Z of the land. With the completion of R-11lu Oya diversion scheme, crop intensities in Moravewa would increase significantly, even without the project, from 90Z to 130%. The ECL studies indicate ;hat, with the scheme, crop intensities could rise to 150X and considerable areas of reddish brown earth soils would allow significant diversification into non-paddy crops during yala. For Iranamadu, the ECL studies show that some crop diversification into soya is likely during yala. With the increase of the tank capacity under the project, cropping intensity is likely to reach 143%. For the total project, cropping intensity would average 138%--compared to 98% at present and 124% anticipated without the project. At full develop- ment, paddy would account for 126% of the cropping intensities, the remaining 12% being Soya. The predominance of paddy in the proposed cropping pattern would utilize the full potential of the paddy lands under the project and  -34- thus accelerate GOSL stride towards rice self sufficiency. Concentration on paddy based on comparative advantage would also lead to a gradual conversion of marginally suitable paddy lands in other schemes for cultivation of other crops promoting GOSL crop diversification policy. 7.06 Crow Yields. The present and projected crop yields and production (Annex 7.2) are based on the results of the farm survey, respective district statistics, recently completed projects and system records. Because the survey was conducted in a relatively dry year, present yields have been adjusted upvard (for Giants Tank, Huruluvewa, Iranamadu, and Rajangana). For the overall project, paddy yields are projected to increase from 2.9 tlha without the project to 3.98 t/ha at full development. Soya yield would also rise from 1.25 t/ha to 1.75 t/ha. Over 90% of the farmers in the project area have already adopted improved variety seeds, but fertilizer application varies considerably among the systems. The projected incremental yield is attributed both to better water availability and to an increase in fertilizer application. Proiect Output 7.07 Crop production under without and with project situations is estimated as: Crop Production Present Future Future Z Increase _without with (iii):(ii) (i) (ii) (iii) --Tons Paddy 155,540 158,600 232,400 46 Soya 1,280 3,570 9,800 175 Paddy Straw 234,550 239,000 346,400 45 Total paddy production at full development (reached in 1995) would account for about 7Z of the national total, assmiing 3X annual growth between 1984-1995. Giant's Tank system would contribute the highest share (23Z) of incremental paddy production under the Project, Morawewa the lowest (4%). 7.08 Under the project, soya would be introduced in line with the Govern- ment-s crop diversification policy. Although it would occupy a very small proportion of cultivated area, the production at full development is estimated to reach about 10,000 tons, an increase of 175% over the anticipated without project situation.  -35- 7.09 The principal input requirements, to attain the estimated production, are listed below: Input Requirements Present Future Future Without With Tons Paddy Seed (tons) 6,130 6,190 6,075 Fertilizer (tons) N 3,890 4,060 5,765 p 3,920 4,130 6,045 K 1,385 1,440 1,990 Hired Labor (1000 man-days) 3,310 3,475 6,140 Detailed crop input requirements per ha are presented in Annex 7.3. An average of 5OZ of paddy seed is obtained from the ASCs which would be expected to supply the requisite amounts. As part of improved cropping practices package, seed rate would also be reduced by about 8% from the present rate. According to the farm survey, delays occur in the delivery of fertilizer to farmers. With the construction of additional stores under the project, adequate inventory of fertilizer and other inputs would be main- tained to meet farmers" needs. With respect to the substantial increases in demand for hired labor, it is envisaged that both the underemployed and unemployed would be tapped. The villages in the schemes vould continue to provide about 80% of the required hired labor and the balance would be sup- plied by the respective Districts (10%) and from outside the Districts (10x). B. Marketinn and Prices 7.10 Psaddy Marketing. In the project area, private traders absorb 75Z-95% of marketable rice, and will have no difficulty in handling additional production. Government-sponsored outlets such as cooperatives and the Paddy Marketing Board (PMB) are considered less attractive to farmers. Respondents in the farm survey indicated that private traders show greater flexibility with respect to collection of the produce from the farms, give immediate cash payment on collected produce, and are not very strict on quality control. Private traders also provide production credit. 7.11 Paddy Pricing. The shift to the private sector has had major implications not only for the share of harvest passing through public sector agencies, the PMB and the Food Department, but also for the role that each plays. In general, GOSL's policy objective has been to expand the role of the private sector while concentrating public sector activities on stabi- lizing the market within acceptable limits of seasonal price variation. Critical to the success of such policies is the setting of a range between  -36- floor and ceiling prices to provide adequate scope for the private sector without adversely affecting producer or consumer interests. Following the recent Government Paddy Price announcement, the domestic price of paddy is in line with import parity price. 7.12 Rice Imports. Over the last decade, rice imports averaged 325,000 tons annually, representing about 25X of consumption. Since 1978, imports ranged between 160,000 and 210,000 tons per year in part due to the start of Mahaweli production. Provisional data for 1983 indicate that domestic paddy production reached a record of 2.5 M tons, a 15% increase from the preceding year, and that rice imports in 1984 may be less than 7.5Z of total consump- tion in 1983. Despite considerable success in boosting paddy production, population growth, rising per capita consumption, the need to maintain buffer stccks, and climatic vagaries are expected to keep Sri Lanka a net rice importer through the middle of the next decade. 7.13 Paddy Stora2e and Milling. PMB currently has about 335,500 tons storage capacity and owns 28 rice mills with a total annual capacity of 181,000 tons. To supplement its own capacity, PMB maintains a register of about 400 "quota mills", with an aggregate capacity of about 600,000 tons, that mill PMB paddy (for administered fees and reimbursement for handling and transport expenses). In addition, there are approximately 1,500 mills operated by private traders that buy and sell in the open market and also mill for consumers. In 1982 these private mills had a total capacity of approximately 1.2 N tons (about two-thirds of that year's paddy crop). PMB has recently initiated a program to close down its most obsolescent plants. USAID and other donors are financing projects designed to increase and upgrade rice-milling and storage capacity in Sri Lanka. Suffficient milling capacity will exist in the project area to absorb additional production. C. Farm Income 7.14 Average farm holdings excluding homestead plots range from 0.9 ha in Rajangana tc 3.2 ha in Giant's Tank. To assess the likely project impact on the incomes of farm families, three representative farms from Giant's (3 ha), Iranamadu (2 ha) and Rajangana (1 ha) have been analyzed. These farms reflect tenural differences and variations in cropping intensity, crop yields, and cropping practices under without-project and with-project situa- tions. A summary of the estimated farm incomes is presented in Table 7.1; detailed tables are given in Annex 7.4, Tables 1 to 3.  -37- Table 7.1: Estimated Farm Income Financial Returns (Rupees) Present Without With W-W Giants Tank (3-ha) Net Cash Farm Income /a 2,170 5,690 14,600 8,910 Per Capita I/ Per family member 400 1,050 2,700 1,650 Per agriculturally employed 1,360 3,560 9,130 5,570 Iranamadu (2-ha) Net Cash Farm income La 10,200 16,600 27,260 10,660 Per Capita /c Per family member 1,830 2,980 4,890 1,910 Per agriculturally employed 6,000 9,760 16,040 6,280 Raianaana (1-ha) Set Cash Farm Income /a 11,040 16,440 22,350 5,910 Per Capita /d Per family member 2,060 3,070 4,180 1,110 Per agriculturally employed 5,020 7,470 10,160 2.690 La- After payment of water charges. Starting with Rs 250/ha and reaching Rs 500/ha by 1991. Without project, only 50Z of these rates are assumed to be paid by farmers. Lb Average family size 5.41, of which 1.6 are actively employed on-farm. ,'. Average family size 5.57, of which 1.7 are actively employed on-farm. ,'d Average family size 5.35, of which 2.2 are actively employed on-farm. In Giant's Tank, with project income (compared to without project) would increase by 156Z, in Iranamadu 64%, and in Rajangana 36Z. Farmers in Giant's Tank, currently below the family poverty income level of Rs 300 per month, would generate adequate income with the project to meet their basic needs. The incremental incomes are significant enough to induce farmers to par- ticipate in the project. As evidence of farmers' eagerness to participate in the project, 80Z of those surveyed have expressed their wiilingness to provide labor for both rehabilitation and maintenance of the schemes. D. Cost Recovery and Fiscal Impact 7.15 Backzround. Government's low priority for maintenance of irrigation schemes has given rise to substantial losses to the economy. The annual  -38- budget allotments for the seven schemes proposed for rehabilitation, has been only about 45% of the amount needed for proper OM. Such inadequate allot- ments arose mainly from GOSL inability to collect as well as farmer's reluc- tance to pay water charges for irrigation services, although suitable legis- lation existed. 7.16 Water Charges. Extensive dialogue was held in 1981 between the Bank and GOSL on cost recovery (during negotiations for the IDA-financed Mahaveli Project III) through introduction of Mahaweli Basin-specific water charges. GOSL then agreed to start collection of water charges reaching 100I of expected O&M costs by 1991. However, collection could not be started because of the 1982 elections and also because of the drought conditions that prevailed during the 1981-82 season. Collection of water charges under the Irrigation Ordinance also proved problematical partly because of inadequacy of enforcement procedures and partly because of lack of a viable collection mechanism. To remedy these deficiencies, Government (i) established a Cabinet-level panel to recommend water charges, and (ii) through amendments to the Irrigation Ordinance (a) reintroduced magistrate court jurisdiction to try water charge payment defaulters, and (b) empowered Government Agents to recover costs of repair of damaged irrigation works. Following GOSL's deci- sion in July 1983 to introduce a new set of water charges, MLLD issued in February 1984 detailed circulars on collection procedures to all Government Agents (GAs). These circulars require GAs to identify a few currently employed permanent officers to act as water charge collectors. Project ManAgers (IND) in each scheme would extend assistance to water charges col- lectors by monitoring and supervising the field collection. Each collector would be assigned to a hydrologic area and would be given 5Z co ission on the actual collection. Water charges collected from any scheme would be used for O&E requirements of the same scheme (additional funds would be provided by Government to the extent necessary until the collected vater charges fall short of the OEM needs). Transfer of water charges funds between schemes would not be permitted. This utilization of vater charges for OEM of the respective schemes would provide the necessary incentive to the farmers for prompt payment. In the event of crop failures, over and above that can be covered by crop insurance, water charges would be waived accordingly. 7.17 GOSL approved in July 1983 nationwide water charge at Rs 2,000/ha for major lift irrigation projects, Rs 688/ha for drainage, and Rs 500/ha for gravity irrigation, to cover current average O&M costs. GOSL also approved special accounts for OEM of works into which collected water charges would be credited. However, to phase collections, GOSL decided to collect 50Z of the approved rates initially (1984), with a progressive increase per year until full rates are imposed and collected in about five years. All irrigation schemes would thus have water charges equal to full 06K costs from 1991 onwards. Thereafter, as permitted under the Irrigation Ordinance, rates would be reviewed periodically and increased to accommodate any cost escala- tion. Approval of this program represents a step forward, emanating from a recognition of past inadequacies in resource mobilization and the realization of the need for adequate budget allotments to meet 06& requirements for irrigation works. GOSL has sent to IDA in May and July 1984 interim reports  -39- on collection of water charges. In Mannar District, vhere Giant's Tank is located, collection amounts to about 45% of the assessed water charges (50X of OEM). Given the poor 1983/84 maha harvest caused by the failure of the early rains and record high raius from December 1983 to March 1984, the collection rate is considered satisfactory, considering that this is the first concerted campaign by the Government to collect water charges under the Irrigation Ordinance. Assurances have been obtained that GOSL, by July 31 of each year of the Project period, shall provide IDA with a detailed report on the progress made on the collection of water charges with regard to irriga- tion systems in Sri Lanka, and shall take all necessary action to improve its collection procedures if the reports show that the collection performance is unsatisfactory. 7.18 Collection of vater charges has already started at Rs 250 per ha in 1984. The charges would increase to the full O&M cost (alloving for increase in costs) by 1991. Toward this end, assurances (similar to that reached for Hahaweli IV) have been obtained that GOSL, by July 31, 1986, shall carry out a review designed to assess whether the water charge collection program is successful and what adjustments thereto may be necessary; to determine the current costs of operation and maintenance of the irrigation systems in Sri Lanka and to propose a program for annual rate increase in order to achieve full operation and maintenance cost recovery not later than September 30, 1991. Assurances have also been obtained that until such time as the recovery of full operation and maintenance costs is attained, GOSL shall annually provide the funds necessary for adequate operation and maintenance of the irrigation systems in Sri Lanka. Receipt of evidence satisfactory to IDA t -t O&M funds will be readily available for the fiscal year 1985 shall be a condition of credit effectiveness. 7.19 On the basis of plAned collection of water charges up to the year 2009 (life of Project), GOSL would recover 118% of incremental O&M expendi- tures or 10% of the investment for rehabilitation and incremental O&M expen- ditures expressed in present value terms at 10% discount rate. The share of incremental farm income recovered through water charges would range from 4% in Rajangana to 8Z in Giants Tank. However, 5% in Rajangana and 131 in Giants Tank of incremental project rents .J would be recovered through col- lection of water charges. The project cost recovery cash flow projection is presented in Table 7.2. AJ Project rents equal net farm income (before recovery of water charges) less imputed value of family labor and management.  -40- Table 7.2 Proiect Cost Recovers 1991- Total PV 1985 1986 1987 .1988 1989 1990 2009 @ 1oz Inflow Incr. Water - 390 1,650 3,640 6,420 8,950 11,560 67,670 Charges /a Outflow Invest. Cost /b 141,930 222,775 149,950 116,350 89,830 71,550 - 601,430 Incr. O&M Exp./c - 630 2.220 4,150 6,420 7,960 9,250 57.18 Total Outflow 141,930 223,405 152,170 120,500 96,250 79,510 9,250 658,610 Total Cost Rec. /d .2 1.0 3.0 7.0 11.0 125 10.0 Incremental O&H Cost Recovery Le 62.0 74.0 88.0 100.0 112.0 125 118.0 La Rs 250 per ha in 1986 and rising to Rs 500 per ha in 1991. However without project it is assumed that the actual collection would amount to 50Z of the with project rate. lb Costs for MIRP including physical contingency. tC Incremental O&M expenditure of Rs 200 per ha. Ia Incremental water charges as a percentage of total outflow. /e Incremental water charges as a percentage of incremental O&M expenditure. Although the cost recovery is low, given the sensitivity of the water charge issue in Sri Lanka and the Government's futile effort in the past, the deci- sion to phase in water charges gradually is realistic and pragmatic. During the first year of the application of the nev water charge, the poor 1984 mAba harvest prevented the Government from enforcing collection strictly. Never- theless, it is hoped that the stronger legislative and judicial framework nOW in place, as well as the firmer institutional support expected from Do, will permit significant improvements. VIII. ECONOMIC BENEFITS. JUSTIFICATION. AND RISK A. Project Benefits 8.01 The project would rehabilitate irrigation systems for 11% of the total irrigated area in Sri Lanka and would transform this currently under- utilized area into highly productive land. At full development the project would generate 73,800 tons of incremental production of paddy and 6,200 tons of soya. The net incremental benefit to the economy is estimated at about Rs 78.0 million annually at full development (Annex 8.1). The net benefit per  -41- hectare of agricultural land would increase by 87%, from Rs 1,950 without the project to Rs 3,640 with full operation of the project. 8.02 Based on the national poverty income level, about 40% of the project beneficiaries fall within the poverty income bracket. In particular, the farm families in Giants Tank earning 40% below the poverty income threshold, constitute the largest low income beneficiaries in the proposed project. Despite the relatively large farm sizes, the Giant's scheme suffers from severe water shortages resulting in cropping intensity of about 88% and also depressed low paddy yields. Furthermore only less than 4Z of the farm families in Giants Tank are gainfully employed in non-agricultural activities to supplement farm income. Although the Giant's scheme would still experience the lowest crop intensity following the implementation of the project, farm family incomes would rise substantially (156Z). The rehabilitation of the schemes and provision of other supporting services under the project would increase farm family income significantly and would alleviate rural poverty. 8.03 Employment Impact. The project would generate substantial employment opportunities during both implementation and production. During the six-year implementation phase, an estimated 3,000 new full-time jobs would be created yearly. In addition, total on-farm labor requirements would increase to 9.5 mi'lion man-days at full development compared to 5.4 million man-days without the project. Hired labor would constitute 65% of total labor requirements. The labor input composition is envisaged to remain similar under the project situation-hired labor would meet the core labor requirements. This is dictated partly by the presence of alternative income opportunities for farm family members, partly by the requirements for seasonal labor. Moreover, through production, marketing, and processing linkages, the proposed project would also create employment and income that have not been quantified. 8.04 Foreign Exchange Impact. The incremental outputs of rice and soya would contribute, respectively to foreign exchange saving and earning. Sri Lanka's average 1980-83 rice import was about US$40.3 million. At full development of the project, the net foreign exchange (equivalent) receipt, based on international prices of inputs and outputs, would be about US$3.1 million annually. B. Economic Analysis 8.05 The economic rate of return of the project is estimated at 16%. The project's net present value, discounted at 10% (the estimated opportunity cost of capital) over the project life, is about Rs 211.4 million (US$8.4 million). A simmary of ERRs and NPVs for the seven schemes and the whole project is presented in Table 8.1. Economic cash flow projections for the project are given in Annex 8.1. Chart I-A plots the present values of incremental benefits and costs against various discount rates and Chart I-B presents the undiscounted stream of benefits and costs.  -42- Table 8.1: ERR and NPV of Schemes Share in Scheme ERR m NPV Total NPV (Z) (Rs.N) (US$H) (Z) Kantalai 18.0 27.7 1.1 13.1 Moraweva 20.0 12.8 0.5 6.0 Iranamadu 18.0 65.8 2.6 31.0 Giants Tank 13.0 21.8 0.9 10.7 Rajangana 18.4 45.8 1.8 21.4 Nachchaduva 13.0 16.8 0.7 8.3 Huruluvewa 19.0 20.7 0.8 9.5 Total Project 16.0 211.4 8.4 100 = == seem ,m m The underlying assumptions for the economic analysis are as follows: (a) Proiect Cost. In addition to the direct investment costs of the seven schemes, expenditures for support services (MLLD, DAS, LCD, DA, and UOP) are included. These costs have been proportionally allo- cated to each scheme on the basis of irrigated area. The cost for the IIMI is excluded. The financial investment costs (local com- ponent) are converted to economic costs by a standard conversion factor (SCF), estimated at 0.85. The present value of these invest- ment costs is estimated to be Rs 461.9 million, or Rs 10,000 per ha. (b) Oueration and Maintenance Cost. O&M expenditures under existing limited budgetary appropriation conditions, are about Rs 3(0 per ha including farmers labor contributions in cleaning field channels and ID costs. These expenditures have to be raised to Rs 500 per ha for proper OM and GOSL would make budgetary provisions accordingly (para 6.07). The incremental OLN cost of Rs 200 per ha is therefore con- sidered as project cost and for economic analysis it is adjusted by SCF. (c) Proiect Life and Phasing. Project works would commence in 1985 simultaneously on all seven schemes. By 1991, the entire project area would be under improved irrigation and full development would be reached by 1995. The life of the project is assumed to be 25 years. (d) Agricultural Practices. Recommended cropping intensity and other input coefficients are described in detail in Section VII and Annexes 7.1, 7.2, and 7.3. (e) Output Prices. Sri Lanka is likely to achieve self sufficiency in rice by the middle of the next decade. Therefore, the analysis is based on the import parity price of rice up to 1994 beyond which the export parity price is used. On the basis of the Bank's commodity  -43- price forecasts, farm gate prices of traded commodities have been derived in constant 1984 values. The nontradable paddy straw is valued at the current domestic price adjusted by the SCF. Details on prices are given in Annex 8.2. (f) Input Prices. All traded inputs are valued at border prices derived from the Bank's commodity price forecasts. Import parity prices have been used for all types of fertilizers. Family and hired labor wages are based on the weighted average peak (80%) and nonpeak (201) rural unskilled labor wage rates. To derive the shadow wage rate, the peak wage rate portion is adjusted by the SCF and the noupeak by both the SCF and the specific conversion factor of 0.80. The resulting shadow wage rate is Re 21 per man-day, or 80Z of the peak market wage rate. Seed prices have a 20Z premium over the respective output prices. C. Sensitivity Analysis and Proiect Risks 8.06 Sensitivity Anallsis. The project's resilience to possible changes in the vain parsmeters, as reflected in their switching values is shown below. A locus of all possible switching values for benefits and costs is presented in Chart I-C. Sensitivity Analysis Indices Appraisal Value Switching Value Index Index Z Change I Investment Cost 100 146 + 46 Incremental O&M Cost 100 536 + 436 Incremental Production Cost 100 123 + 23 Incremental Benefits 100 87 - 13 Incremental Paddy Yields 100 84 - 16 Cropping Intensity 100 80 - 20 Paddy Price 100 86 - 14 jj Indicates the percentage change in the selected variable to generate ERR of 1OZ, equivalent to the opportunity cost capital in Sri La-ka. The project's economic viability is sensitive to shocks arising from major cost increases or reductions in benefits, which are both unlikely. For incremental benefits to drop by 13%, either the projected yields or the cropping intensity will have to drop significantly. With anticipated crop- ping intensity unchanged, the incremental paddy yield vill have to drop by 16Z (from 1.07 t/ha to 0.9 t/ha). This occurrence is unlikely given the relatively high yields attained even under existing conditions. With anticipated yields unchauged, the cropping intensity has to decrease by 20% or from 138% to 110%, which is also very unlikely. Such an occurrence would mean that with-project cropping intensity would be lower than without the project case, which is equally unlikely. Another variable endangering the viability of the proposed project is the long-tern price of paddy. Recently,  -44- paddy price in real terms has been declining and the long-term price forecasts reflect a downward trend. Moreover, Sri Lankan rice is valued 30% lower than the international reference price of rice stemming from poor quality. Given the Government's target to enter the export market by mid 1990's, when Sri Lauka is expected to become surplus rice producer, a rice improvement program would be warranted to reap the full benefits in the export market. Strengthening of rice research program which is currently underway, would promote quality improvements in the medium term. The project can also withstand a delay in generating the incremental benefits up to 1.5 years and still remain economically viable. 8.07 As described in Section III, the lessons learned from TIMP have been considered in the planning of the proposed project to minimize future project risks. However, institutional reforms and the adoption of the proposed improved water management practices are considered vital to the viability of the project and a possible source of risk. INMAS is an innovative water and land management program. Similar programs have been under implementation in recent years. The very encouraging results are partly attributed to active participation of farmers in planning and management of cropping practices and water use. In a farm survey undertaken for the preparation of MIRP over 90% of the farmers surveyed indicated that they would actively participate in the rehabilitation and maintenance of the schemes. GOSL has demonstrated its strong committment to the implementation of INMAS throughout Sri Lanka. It should be noted that INMAS is also complementary to the Government"s long standing policy of promoting political as well as economic decentralization. Moreover the provisions made under the project would ensure timely implemen- tation of the program to safeguard against any risks. The staffing require- ments for the promotion of INMAS is rather modest. Provided that the proposed training program is executed, adequate staff would be made available on time. IX. RECOMMENDATIONS 9.01 Assurances have been obtained that: (a) GOSL shall conduct a mid-term review to determine the effectiveness of the administrative changes introduced in respect of INMAS and a report on the findings of such review shall be furnished to IDA not later than December 31, 1987. (Para 4.05). (b) GOSL shall, by December 31, 1985, furnish to IDA for its review a program for the utilization of lift irrigation systems and shall take necessary action to implement said program, as agreed by IDA, by June 30, 1986. (Para 4.07). (c) In order to make the farmers familiar with the pilot area program, COSL shall, by September 30, 1986, deploy adequate additional staff in the pilot areas, and shall  -45- collect data on the relative performance of the pilot program, in accordance with procedures acceptable to IDA, and shall, by December 31, 1987, furnish to IDP'tbe data collected. (Para 4.08). (d) GOSL shall uot permit any further development of irrigation by tanks or diversion weirs in the catcbment areas of the seven irrigation syttems included in the Project, and no diversions, other than the existing from Malwatu Oya shall be allowed upstream of diversion weir at Tekkam. GOSL shall, by December 31, 1987, furnish to IDA its proposal for organi- zational set-up, including legal framework for proper catchment control. (Para 4.12.) (e) GOSL shall make adequate arrangements to determine the efficiency of transmission of supplies by the feeder channels to the Project area in each crop season by appropriate gauging of supplies at control points and shall furnish a report to IDA within three monthbs after each crop season. (Para 4.17) (f) GOSL shall make adequate and timely budgetary provisions for the implemeatation of the project in line with the financial requirements of the Project. (Para 5.02) (g) GOSL shall establish and maintain separate accounts for the project; shall have the accounts audited both on concurrent basis for SOEs and annually by an independent auditor acceptable to IDA; and shall furnislh to IDA, for review, certified copies of audited accounts and the auditors' report as soon as available but in any case not later than nine months after the end of each fiscal year; including a separate opinion by said auditors whether the proceeds withdrawn on the basis of statements of expenditure have been used for the purposes for which they were provided. (Para 5.10) (h) GOSL shall provide sufficient funds for the O&H costs for irrigation systems rehabilitsted under the Project, and furnish to IDA, by January 31 each year, a statement on au agreed pro forma indicating that sufficient funds have been allocated to meet the 0EM costs for the Project during the fiscal year. (Para 6.07) (i) GOSL, by July 31 of each year of the Project period, shall provide IDA with a detailed report on the progress made on the collection of water charges witb regard to irrigation systems in Sri Lanka, and shall take all necessary action to improve its collection procedures if the reports show that the collection performance is unsatisfactory. Also GOSL, by July 31, 1986, shall carry out a review designed to assess vhether the water charge collection  -46- program is successful and what adjustments thereto may be necessary; to determine the current costs of O&M of the irriga- tion systems in Sri Lanka; and to propose a program for annual rate increase in order to achieve full O&M cost recovery not later than September 30, 1991. Further until such time as the recovery of full O&M costs is attained, GOSL shall annually provide the funds necessary for adequate O&M of the irrigation systems in Sri Links. (Paras 7.17, 7.18.) 9.02 Conditions of effectiveness of the proposed IDA credit would be: (a) Opening of the PSA (para 5.04); (b) Fulfillment of the effectiveness conditions of the "CIDA Con- tribution Agreement" negotiated between GOSL and CIDA (para 5.09); (c) Fulfillment of the effectiveness conditions of the "SDC Con- tribution Agreement" negotiated between GOSL and Switzerlaud (para 5.09); (d) Appointment of the Project Director and Project Officer (para 6.01); (e) Formation of the Project Steering Committee (para 6.02); (f) Appointment of. project managers (ID cadre) for each scheme (pars 6.08); and (g) Receipt of evidence satisfactory to IDA that O&M funds vill be readily available for the fiscal year 1985 (para 7.18). 9.04 With the above agreements and conditions of negotiations and effec- tiveness, the proposed project would be suitable for an IDA credit of SDR 17 X (US$17 M equivalent). The Borrower vould be the Democratic Socialist Republic of Sri Lanka.  RI MAJOR IRIGATION RBEAUILITATION PROJECT staff ADnraiSal Report Sub-Proiect Schems: Salient Features Original Additional Date of Catchment Ave. Supplementary Nat Reservoir Command Area District Origin Area Inflov Supplies Capacity Area Encroached (en km) (MC1) (MCII) (ha) (ha) 1. Giant's Tauk System Mannar Ancient System 102 43 Diversion from Malvatu 41.4 8 143 273 initially restored Oya at Tekkam anicut (Including Giant's 1890-92 Cross Catchuent above (excluding storage (excluding Tank on the right Tekkam - 2,978 sq km of about 190 minor minor tank bank, and Akathi- tanks) beds) murrippu on the left bank, of Malvatu Oya) 2. 8uruluweve Anuradhapura Ancient system 199 54 From Mahaveli via feed- ILI 3.524 564 restored 1949 er canal constructed in 1973 (2.8 .3 per sec) 3. Iranasmdu Kilinochchi Colonization scheme 280 159 Nil 128.3 7.854 1.570 constructed 1902-22 (excluding x lift irri- x gation area) 1 4. Kantalai Sylst Trincomalee Ancient system ini- 210 149 From Mahaveli via MMTE 160.6 5,571 1.451 tially restored restored/constructed (excluding (Includng Kantalai 1869: further devel- 1977 (17.01 .3 per sec) 3,850 ha and Venderasan oped this century sugar estate) tank.) 5. Eoravews Trincomalee Colonization scheme Al 40 From Nelu Oya via 1 636 327 Stage I, l?59-62: diversion canal to Stage II completed 1982 be completed in 1984 6. !chchaduva Anuradhapura Ancient system 291 124 From Kal* Weva (1917-78). 44849$ restored 1906 (catchebnt From Mahaweli via feeder (interconnected with area contains canal from lala Veva RJ Turuwila Wewa and large no. of Canal since 1978 Anuradhapura city tanks) minor tanks) 7. Ranianatna System Anuradbapura, Colonization 898 237 From Mahaweli via Kala 09jJ 5.371 537 Kurunegala schem constructed (below Oya and drainage from (excluding (Including both and Puttalum 1957-63 lala Wewa) System H lift irriga- Ranjangana and tion area) Angsauwa tanks)  Sri Lanka Haior Irrlgation Rehabilitation Project Staff Appraisal Report Climatological Data _____-_ |DLI 1 DL2 2/ DL3 3/ D DL4 4 Month 1 1 I I I I - ' 2 tl |i |N t Efi N! I2 0 2il0 i s l v 0 6 1i January 85 701 3.01 4.01 711 90 1091 80 76 6.8 4.91 781 82 12701 831 721 1.61 3.41 721 8811081 831 741 1.61 4.41 76 8512241 February 1 871 711 1.51 5.01 711 90JIlU1 831 76I 2.4 | 4.61 751 82 12131 861 721 1.01 4.01 691 8911101 861 741 1.01 5.21 721 8811911 March 1 921 721 2.01 5.91 621 SCIlOII 861 771 1.6 1 5.41 741 85 11561 891 761 1.01 4.71 681 8611421 891 751 1.0 5.41 701 8911621 1 1 I I I I I II I ll II I i I I I I I April | 921 751 5.01 4.71 671 88I 821 901 781 2.6 I 5.51 721 86 11531 901 801 2.41 5.01 741 84|2221 901 78; 2.41 5.3 751 8811661 Mnay | 901 771 1.01 5.21 721 8611731 931 791 1.0 1 6.61 661 82 12491 881 821 1.61 5.31 781 8313611 891 811 1.b1 5.61 761 8412321 l l I 11 11 I 1618 I I I I 11 11 I 11ll l l l l l l June 1 901 771 1.Uj 6.21 651 8412071 931 801 - 1 8.1 611 78 13251 871 811 - I 6.11 791 8213961 881 811 - |6.11 761 8212571 July | 901 7bl - I 6.61 641 8412181 931 781 - I 7.91 611 78 12951 861 801 - I 6.61 791 8313521 871 791 - | 5.91 771 8412401 AuguLt 1 921 761 1.0I 6.81 601 8212121 921 761 1.0 1 7.81 631 80 12751 861 801 - 1 7.61 791 8413131 81! 79) - I 6.31 771 851242! I I I 1 I I I I I I t I I I I I I I I I I I I I I I I I I Septemberl 911 751 1.41 6.01 631 8411861 921 771 1.0 1 7.61 641 80 1245! 86! 801 1.01 6.31 78! 831224| 871 791 1.01 6.11 761 841240| October | 881 731 6.01 4.61 73! 8811301 88! 761 4.5 I 5.31 731 86 12001 851 781 5.U; 7.41 791 8611241 87! 771 5.01 4.51 78! 8711761 November I 861 721 7.01 3.01 751 901 661 841 751 8.0 | 3.61 791 88 12051 841 751 7.01 3.01 791 901110! 841 761 7.01 3.41 791 8811681 1 1 1 13.1 751 71 8! 71 77 9.0I 341 8! 8 1126'1 6'16' December 81| 691 6.01 3.41 751 871 82' 791 771 9.0 1 3,4| 801 85 12751 831 731 4.0 3.31 76! 89 236 83 75 4.0 3.3 791 86 2161 I1 1 I I1I1 I I1 I! I I I I I I I II II l t I TOTAL |34.9161.41 1 137.9 170.71 1 1 I 2 !24.6I6.7l 24.6 61.51 | | I .__________1 64 I1_ 1/ Data for Maha Illuppalama Note: 2/ Data for Trincomalee 3/ Data for Jaffna Temperature in F 4/ Data for Mannar Rainfall, Evaporation in inches 5/ Data for Anuradhapura Humidity in X 61 Data for Kantalai Wind In total of average m.p.h. per day in the month I/ Data for Iranamadu 8/ Data for Vanatavillu .S , a  SRI LANKA Major Irrigation Rehabilitation Project Staff Appraisal Report Population and Labor Force No. oF Z of Population X of Popula. Z of Pop, Farm Total --------- over 5 yrs in Labor Occupied Occupied House- Students & Families Population Sinhala Tamil Muslim Primary Educ. Force in Agri. in non-Agri. wives Unemployed