India - Second Andhra Pradesh Irrigation Project

Document of The World Bank FOR OFFICIAL USE ONLY C14. /166---2W Report No. 5883-IN STAFF APPRAISAL REPORT INDIA SECOND ANDHRA PRADESH IRRIGATION PROJECT February 27, 1986 This document has a restricted distribution and may be used by recipients only in the performance or their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS US$1.00 = Rupees (Rs) 13.0 1/ WEIGHTS AND MEASURES (METRIC SYSTEM) 2/ 1 meter (m) = 3.28 feet (ft) 1 kilometer (km) = 0.62 miles (mi) 1 hectare (ha) = 2.47 acres (ac) 1 million cubic meters (Mm3) = 804 acre-feet (ac-ft) 1 cubic foot per second (cfs or cusec) = 0.0283 cubic meters per second (m3/s) 1 kilogram (kg) = 2.2 pounds (lb) 1 metric ton (mt) = 2,205 pounds (lb) 1 Thousand Million Cubic Feet (TMC) = 28.3 million cubic meters (Mm3) FISCAL YEAR GOI; GOAP April 1 - March 31 1/ Conversions in this report have been made at US$1.00 to Rs 13.0 2/ The metric system has been used in most cases. However, India is still in the process of transition to the metric system; non-metric units are still widely used and have been used in this report where a conversion to the metric system may confuse the reader.  FOR OICIL USE ONLY SECOND ANDHRA PRADESH IRRIGATION PROJECT Credit and Project Su_mary Borroa: India, acting by its President (GOI) Bey,eficiar: Government ot Andhra Pradesh (GOAP) ADo2r: IDA Credit: SDR 127.5 N (US$140.0 X equivalent) IBRD Loan: US$131.0 N Terms: IDA Credit: Standard IBRD Loan: Standard Re-Lendint Terms: From the GOI to the GOAP as part of central assistance for State development projects on terms and conditions applicable at the time. GOI would beas the foreign exchange risk. Project Description: The project would assist with development ot the Sriramasagar and Srisailam Right Branch Canal irrigation schemes, located ia the poorest drought-prone areas ot Andhra Pradesh. It would modernize and complete the irrigation network in 165,000 ha, extend the irrigable area by another 163,000 ha in Sriramzasagar, and develop a greater part (65,000 out ot 77,000 ha) of Srisailam. The main objective of the project is to increase agricultural production and farmer's income in the command area ot both schemes. This would require a number of complementary activities, including (i) introduction of new plJanning and design concepts; (ii) revisions in irrigation system operation; (iii) improved management and maintenance standards; and (iv) training of operation and maintenance staff and contact farmers. The project would be implemented over a seven-year period. At full development it vould generate incresmental annual production ot 395,000 tons of foodgrains, 218,000 tons of cotton, 148,000 tons of groundnuts and other crops. Numerous unquantifiable benefits would also accrue to the two communities. The risks ot the project are primarily associated with the introduction of vater management concepts which are new to the implementing agency. Another risk could be the shortage of qualified staff, which may result in implemen- tation deLays. However, all expected risks have been taken into account in planning project implementation. I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Thi docuXm ha a * 1_d dibution and may be used by reapients only in de perfommce of thir ofiid dues IS ceneM may nuo oerwie be dibdodwinftont Wold Dank auharitu  PRINCIPAL ABBREVIATIONS AND ACRONYKS AD - Department of Agriculture AP - Andhra Pradesh APAIC - Andhra Pradesh Agro-Industries Corporation APAU - Andhra Pradesh Agricultural University API - Adjustable Proportionate Module ARDC - Agriculture Refinance and Development Corporation (see NABARD) CADD - Irrigation Utilization and Command Area Development Department (see IU and CADD) CCA - Cultivable Commanded Area CUC - Central Water Commission DC - Distributary Committee GCA - Gross Commanded Area GOAP - Government of Andhra Pradesh GOTN - Government of Tamil Nadu CaI - Government of India lCD - International Competitive Bidding ID - Irrigation Department IU and CADD - Irrigation Utilization and Command Area Development Department (see CADD) LCB - Local Competitive Bidding NC - Minor Comittee -SL - Mean Sea Level MABARD - National Bank for Agriculture and Rural Development (see ARDC) MARP - National Agricultural Research Project OC - Outlet Committee 0&N - Operation and Maintenance PAP - Project Affected Persons PPM - Project Preparation and Monitoring Wing (of ID) PID - Public Works Department R&B - Roads and Bridges RD - Revenue Department SRBC - Srisailam Right Branch Canal SRS - Sriramasagar DNC - Thousand Million Cubic Feet TV - State of Tamil Nadu T&V - Training and Visit (system of agricultural extension) WALANTARI - Water and Land Management Training and Research Institute  FREQUENTLY USED ABBREVIATIONS FOR OFFICERS AE - Assistant Engineer CE - Chief Engineer CI - Canal Inspector DDA - Deputy Director of Agriculture DEE - Deputy Executive Engineer EE - Executive Engineer JDA - Joint Director of Agriculture LDC - Lower Division Clerk SE - Superintending Engineer VEO - Village Extension Officer GLOSSARY Hot Season - March to May Irrigated Dry - Irrigated land designated for crops other than rice and sugarcane Irrigated Wet (IW) - Irrigated land designated for rice and sugarcane Kharif - Wet Season (June to September) Chak - Area served by watercourse, between 15 ha to 40 ha Rabi - Dry season (October to February) District - The lowest administrative unit where the State Government is directly involved. There are 20 districts in Andhra Pradesb State. Taluka - A sub-division of a district. Warabundi - A system of rotational irrigation  INDIA SECOND ANDHRA PRADESH IRRIGATIO PROJECT STAFF APPRAISAL REPORT Table of Contents Page No. I. BACKGROUND ... ............ 1 Agriculture and Irrigation in India ...................o.. 1 The State of Andhra Pradesh ........... 1............. I II. THE PROJECT AREAS ....... ................. ... 5 A. History and Background ............................... 5 The Srisailam Right Branch Canal Scheme (SRBC) ...... 5 The Sriramasagar Scheme (SRS) ...................... 5 B. Project Areas ...................... ................ 6 Location ....... ..........................s. 6 Climate . ................ ......... ...... ............ 7 Topography and Soils ........... we.. ........... .... 7 Population and Land Holdings.. .............. ...... 7 Communications..o. ................................. 7 Present Development ................................. 8 Agricultural Support................................. 9 Rationale for Bank Involvement ...................... 10 III. THE PROJECT .......................................... . 11 A. Project Scope and Objectives ................ o........ 11 Project Objective ......................... o................... .... 11 Summary of Project Components ...................... . 11 B. Detailed Features.... .................................... 12 The SRBC Sub-project......o.............. o.......... .. 12 The SRS Sub-project ................................ 13 Block Development and Upgrading ..................... 13 Feeder Road Network ................................ 14 Equipment ........... ......... o............................... 14 Telecommunications ........... 14 Training and Training Facilities.................... 15 This report is based on the findings of an appraisal mission which visited India in March/April 1985 and comprised of Messrs. A. Seager, E. Cazit, R. Vick, M. Herman, S. Niaz, S. Lieberman (Dank) and K. Baxter, S. Singh, and C.S. Nawathe (Bank, New Delhi). Kr. B. Albinson has contributed to sections on design criteria and water management. Word processing was done by Miss C. Batara and Mrs. K. Landerer, with the assistance of Nesdames Z. Kalinger and S. Douglas.  Land Alienation .......... ...................... ........... 16 Monitoring & Evaluation; Technical Services; Studies and Planning............. ................. 17 Design Criteria ..... .................... O........0.00-0 .... 17 Construction Methods and Current Progress...... ..... 19 Status of Project Preparation ......... ...... ......... 20 Project Execution ............... ......... O.. ...... . 20 IV. COST ESTIMATES AND FINANCING ........................... 21 Project Cost Estimates .... ........... ............ ...... 21 Financing ..o ........................................ 21 Procurement ................................................ 22 Disbursements ............................. .................. . 23 Accounts and Audits ........................ a........ . 24 V. ORGANIZATION AND MANAGEMENT .... ..... ................. 24 Introduction ... ............................ ................... 24 The Irrigation Department - Present Organization.... ... 25 The Irrigation Utilization and Commnand Area Development Department (CADD) ......................... 25 The Proposed Organization ....................... ..... 26 Staffing Requirements and Functions under the Project ................................................. 26 Dam Safety ..................... ...... ................... 27 Agricultural Supporting Services ....................... 28 Landshaping..o.se ......... .. .. .... ......... . ..... 29 Road Construction .............. ........................ 29 Planning; Monitoring and Evaluation; Technical Services .................... 29 Quality Control ........... ............................. 31 Project Coordination ....................... . 31 VI. WATER MANAGEMENT ........... ............................. 31 In troduction .alnc............................. ................. . 31 SRBC Water Balance ..................................... ............ 33 SRS Water Balance ..................... . .......V.... .... 34 System Operation and Water Management .................. 35 G aintenance .......................... . .............. ... 39 Groundwater Development ............................................. 39  -lll- VII. AGRICULTURAL PRODUCTION, FARM INCOMES AND COST RECOVERY*... .** ........... ooeee........ 40 A. Agricultural Production ............... ................... 40 Srisai1am ..*.................... ......... ....... 40 Sriramasagare .............. o.... .................. o....... 40 Market Prospects and Prices ......................... . 41 Farm Incomes ............................. ........... . 41 B. Cost Recovery o........ ......... ............... .... 41 General ............................. ......... ......... 41 Present Situation .................................* 42 Proposed Improvements ....... ......................... 44 VIII. BENEFITS AND JUSTIFICATION ............. ................. 45 Project Impact .....o..................................... 45 Poverty Impact .............. o.............. -oo ............. 46 Environmental Effects ....... .... .... ... ............ 47 Economic Analysis .......................... ........... .. 47 IX. AGREEMENTS REACHED AND RECOMMENDATIONS ................ .. 50 TABLES 1. Climatic Data 2. Population and Land Holdings 3. Cropping Patterns and Yields 4.(a-g) Cost Estimates 5.(a-g) Timetable for Contracts 6. Schedule of Disbursements 7. Proposed Credit/Loan Allocation 8.(a-b) Staffing Schedules 9.(a-b) Water Balance 10. Summary of Agricultural Production and Input Requirements 11. Economic and Financial Prices 12. Estimated Farm Budgets 13. Farm Budgets and Project Rent at Full Development 14. Cost Recovery 15.(a-b) Cost and Benefit Streams for Economic Analysis 16. Scheme for Training of Engineers in the Irrigation Department ANNEXES 1. Related Documents Available in the Project File 2. Main Assumptions for Economic Analysis 3. Detailed Cost Estimates  -iv- 4. Draft Terms of Reference for Water Charge Review Conittee 5. Training Centers and Training 6. Organization and Management CHARTS 1. Organization Chart of the Irrigation Department 2. Organization Chart of the Irrigation Utilization and Command Area Development Department 3. Proposed Staff for Monitoring and Evaluation in the Project Preparation Cell 4. Organization Chart of Srisailam Project 5. Organization Chart of Sriramasagar Project 6. Organization Chart of Conmand Area Development Authority 7. Organization Chart of the Andhra Pradesh Agricultural Extension Service 8. Canal Sizing and Surface Delivery Periods for Lands Lacking Groundwater 9. Schematic Design Cropping Pattern - SRBC Component 10. Schematic Design Cropping Pattern - SRS Component (165,000 ha) 11. Schematic Design Cropping Pattern - SRS Component (119,000 ha) 12. Schematic Design Cropping Pattern - SRS Component (88,000 ha) MAPS IBRD 18986 - Medium and Major Irrigation Projects in Andhra Pradesh IBRD 18843 - Srisailam Right Branch Canal Project IBRD 18842 - Sriramasagar Project  INDIA SECOND ANDHRA PRADESH IRRIGATION PROJECT Staff Appraisal Report I. BACKGROUND Agriculture and Irrigation in India 1.01 With about 760 M inhabitants (mid-1985), India is the world's second most populous country. The annual rate of population increase is about 2.2X. Gross Domestic Product, now estimated at Rs 1,690,000 M (US$130,000 M) (1983/84) and growing at about 4% annually, ranks eleventh in the world. Annual income per capita (1983/84) is estimated at Rs 3380 (US$260). The country has the largest single concentration of people in absolute poverty: between 200 and 250 H. Over 90% of the poor are rural. Although the industrial sector is large, sophisticated, and growing at a respectable rate, agriculture still provides 69% of India's employment and accounts for 36% of the GDP. Agricultural growth rate was 4.6% per annum during the 5th Plan (1974-79) but only 2.0Z during the 6th Plan (1979-84), due to three years of unfavorable monsoons. 1.02 At least three-fourths of India's agricultural growth since 1947 stems from the synergy of fertilizers, improved crop varieties and the expanding irrigation. The irrigated area, now over 60 M hectares, is second only to that of China. India is investing more in expanding irrigation than any other country. Investments until quite recently tended to emphasize the creation of irrigation potential: the building of dams, river diversions and major canals. Recently, however, in a change strongly encouraged and promoted by the Bank/IDA, there has been a welcome shift towards completing and improving the system down to the level where only a small number of farmers share the water (typically, to 8 ha units), rehabilitating and upgrading systems constructed to the less exacting standards of the 1950s and 1960s and introducing water management systems and operations geared to the reliability, timeliness and quality of water supply. While this shift is clearly discernible, it needs further general support, which would be achie-3d through training engineers and planners, to ensure general accept- ance. An indication of the high and rising priority attached to the sector is the share of Plan investments allocated: 8.6% in the 4th Plan (1969-74), 9.8% in the 5th Plan (1974-79), 10.6Z in the 6th Plan (1979-84), and 10Z in the 7th Plan (1985-90). The State of Andhra Pradesh 1.03 General. Andhra Pradesh is India's fifth largest State (277,000 km2) and fourth most populous (55 million). The State has a 1,000 km coastline on the Bay of Bengal. Over 76% of the population is rural. Two of India's largest rivers, the Godavari and the Krishna, cross the State, cre-ting large deltas at their outlets. They and their major tributaries rise in the Western Ghats, in Maharashtra and Karnataka.  -2- 1.04 Climate and SOil. The State lies on the fringe of the equatorial zone, between latitude 120 37' and 190 54'. It has three seasons: the monsoon (kharif, from late June to October) the cooler rabi (November Lo February) which is dry except in the coastal areas which lie on the northern limit of the northeast monsoon rainfall zone, and the hot, summer season (March to about mid-June), which is mostly rainless. Since the average coolest temperature is above 130 C, most tropical crops can be grown at all times of the year. 1.05 The State has three regions: (a) The Coastal Region, comprising largely the Godavari and Krishna deltas. This is agriculturally the richest and best developed: the soils are mainly alluviums and the richer bLack vertisols and red alfisols. Rainfall is well in excess of 1,000 -n and subject to cyclonic storms from the Bay of Bengal. (b) The Telengana Region, north and west of the coastal zone. The soils are mainly red al'isols and vertisols of lighter texture than in the coastal region; rainfall is about 700-900 mm. Cyclonic stores tend to affect this zone, but less frequently and with less severity than the coast. (c) The Rayalaseama Region, farthest from the coast, occupies the south- western segment of the State. It is the least exposed to cyclonic storms and therefore has the lowest rainfalL: between 550 and 750 mm per anm:um. The soils are mostly black vertisols. It is the poorest and least developed. 1.06 The Economy. The economy of AP is predominantly agrarian: agricul- ture contributes about 50% of the State's income and employs about 70% of the work force. Between 1970/71 and 1980/81, the rate of growth of the primary sector, including agriculture, forestry and mining was 3.2% per annum (p.a.) compared to an all-India average of 3.0%. The secondary sector: manufactur- ing, construction, electricity, gas and water supply (222 of the State's income), grew at 4.0% p.a., compared to the national average of 3.6%. The tertiary sector, including transport, trade, finance and services (25% of the State's income), grew at an annual rate of 7.0%, about the same as the national average (6.9%). However, since tertiary sector activities started from a relatively small base and are concentrated mostly in the urban areas, their growth contributed only marginally to the welfare of the majority of the population in the rural areas (77%). Overall, State income grew at an annual rate of 3.8 % p.a., slightly above the national rate of 3.62. Per capita income increased by 1.7% p.a. and was about equal to the national average of Rs. 1,267. Since agriculture is the main activity and it has close links to other sectors, economic progress is largely dependent on the removal of constraints to further agricultural growth. Expanding and improv- ing the irrigation sector will support irrigated agriculture; the undoubted and only recently emerging potential improvements in rainfed agriculture offer another avenue to increased production, especially in areas that either because of topography, or because of shortage of water supplies, are unlikely to be irrigated in the foreseeable future.  -3- 1.07 Agricultural Performance. Between 1956/57 and 1981/82, rice produc- tion grew by 6.65 x per annum in Telengana, 3.51% in Rayalaseama, and (from a relatively high base) 1.72% in the Coastal region; in the State as a whole, by 2.93%. The area under groundnuts grew in all three regions; areas under other crops (sorghum, pearl millet) either decreased or increased only mar- ginally. Increase in yields has contributed more to production than did area expansion. 1.08 Irrigation Devylopment. Water available for suIrface irrigation is estimated at 77,500 Mm . The main sources are the Codavari and the Krishna rivers (para 1.03) and their tributaries; others are the minor tanks and groundwater. About 3.0 million ha are irrigated nut of a total cultivable area of 11.3 million ha; the irrigation potential is estimated to be about 9.2 million ha. Irrigation hAs started over 100 years ago in the two river deltas and the Tungabhadra river, a tributary of the Krishna (Kurnool-Cuddapah project). During the past 35 years AP has invested heavily in irrigation. In the Fifth (1974/75-1979/80) and Sixth (1979/80-1984/85) Plan periods, irrigation investments accounted for 35Z and 27% of total development spend- ings. Returns have not been satisfactory, because resources have been spread over too many projects, resulting in long construction and gestation periods and delayed benefits. Nor has a timely and reliable irrigation service, conducive to good water management and sustained high production, been provided by the projects; supporting services (agricultural research and extension, and some of the inputs services) also needed improvement. 1.09 Bank Involvement in AP. The Bank was involved in three irrigation projects: (a) Pochampad Irrigation, Cr. 268-IN, US$39 M, closed 12/79. (b) Godavari Barrages, Cr. 532-IN, US$45 M, closed 6/81. (c) AP Irrigation and Command Area Development Composite, Ln. 1251-IN, US$145 M, closed 6/85. 1.10 The Pochampad Irrigation Project (see also para 2.05) supported works to complete the Sriramasagar dam and develop part of the command area of the Pochampad Project (now renamed Sriramasagar). It achieved most of its objec- tives [Project Performance Audit Report (PPAR) No 3499]. The Godavari Barrage Project (PPAR No. 4000) supported the replacement of a century-old weir by a modern barrage, to enable the diversions and the distribution network in the Godavari delta to go on functioning. This was the first project in which GOI has followed International Competitive Bidding (ICB) rules for major civil works. Although the eventual contractors were Indian firms, the audit concluded, and GOI agreed, that the competition engendered by introducing ICB rules has contributed to improved standards and perfor- mance specifications. 1.11 The AP I Project, when presented to the Board in 1976, was one of the largest Bank-supported projects, and the largest irrigation project in India. Its main objectives were to hasten the construction of the canal network and thereby the utilization of water impounded in the Nagarjunasagar reservoir, (live storage 5,800 Zm3, 205 TMC) and extend the water distribution network beyond the irrigation outlet down to the farm. To minimize seepage losses,  -4- for ease of management, and to achieve better water control, the project had a major canal lining component. 1.12 Largely because of insufficient data on conveyance losses, which, when systematically measured during project implementation, turned out to be up to three times hither than advised by COAP, coupled with difficulties to Limit rice growing and the cor.sequent excessive water use in parts of the project area, the Nagarjunasagar component was extensively reformulated in 1979-80. Key components of the reformulation were: (a) the need to remodel many of the previously unlined canals; (b) considerably more lining of major canals, distributaries and minors than originally envisaged; (c) reduction of project scope; (d) obtaining GOAP's commitment to a rigid water allocation system, to enable irrigation of the project area; - subsequent improvements in irrigation efficiency are expected to enabLe the entire area to be adequately irrigated even after full upstream developments in the Krishna, when water available to the project will be substantially reduced; and (e) a comprehensive system of flow measurements, to refine further and trace the utilization of water releases. 1.13 The AP I project closed in June 1985. Its objectives were mostly, - although not fully, - achieved. The lessons of the project were the need for full main canal lining and a substantial rethinking of previous water allocation and water delivery policies. The project enabled the introduction of a water management plan in parts of the area. Its principal aim was to rationalize the crop calendars by: Zi) retaining water in the dam for pre-monsoon land preparation and early planting of rice; (ii) ensuring equi- table water-sharing between farmers on all reaches of the main canals and (iii) reducing operational losses by prevailing on farmers to carry out operations on the same crops in the same block in a more concerted manner. 1.14 In addition to these three water-related projects, the State shares in several ongoing or completed national projects that support the agricul- tural sector. For the proposed project the National Agricultural Research P-oject (NARP I; - Cr. 955-IN) and the Andhra Pradesh Agricultural Extension Project (Cr. 1219-IN, US$6 H) are of special importance. The latter aims to introduce the Training and Visit (T&V) system of agricultural extension - started in the commands of the AP I Project - over the entire State, under the control of the Department of Agriculture. By now the extension service is fully unified and covers the entire State (para 2.21). Other Bank Group agricultural projects that have affected Andhra Pradesh's agricultural development were ARDC (now named NABARD) general line of credit projects which support lending for groundwater development and other on-farm invest- ments, mostly to individual farmers.  -5- II. THE PROJECT AREAS A. History and Background The Srisailam Right Branch CanaL (SRBC) Scheme 2.01 The idea of utilizing Krishna water to irrigate the drought-prone, upper Pennar Basin in the Rayalaseama region of the State is more than 100 years old. The first attempt was the Kurnool-Cuddapah project (para 1.08), which, through a system of multiple pickups from the Kundu, a tributary of the Pennar, enabled progressive utilization of return flows. By now it extends to about 100,000 ha. 2.02 The right bank of the Kundu, the venue of SRBC scheme, became poten- tially irrigable only after the construction of the Srisailam Dam (IBRD Map 18843). This large (live storage 250 TMC, 7,075 Mm3) reservoir was designed for power generation and as upstream balancing storage for the Nagarjunasagar complex, the backwaters of which reach the toe of the Srisailam Dam. Even- tually there would be 7 power turbines on the right flank and 9 reversible turbines on the left flank, each of 110 NW capacity. 2.03 Notwithstanding this great power potential, it is GOAP's policy decision that power generation will be subordinate to irrigation needs, with the caveat that the highest feasible head of the water should be retained in the reservoir to maximize power generation. 2.04 The need to skim water from the Srisailam Reservoir without passing it through the power turbines, by tapping into the flank of the reservoir, arose in the 1970s when it became clear that Madras City requires water from the Krishna to ensure its long-term municipal and industrial requirements. Accordingly, the three riparian States (Andhra Pradesh, Karnata a and Maharashtra) allocated 5 TMC each, or a tocal of 15 TffC (425 Mm ), to lamil Nadu (TN), expressly for industrial and municipal use in the city of Madras. The Goverment of Tamil Nadu (GOTH) has undertaken to finance costs of con- veying this water, over a period of about 3 months (i.e. out of peak monsoon flows), from the Srisailam Reservoir to the AP/TN border. Once the cost of transfer from the Krishna to the Pennar basin was thus covered, GOAP found it economically feasible to draw water from the Madras canal at times when the canal is not delivering water to the AP/TN border, to irrigate areas in its poorest and most drought-prone Rayalaseama Region (para 1.05). The S iramasagar Scheme (SRS) 2.05 The SRS scheme on the Godavari river was first considered in 1950; construction of the dam began in 1963. The Bank Group's involvement started in 1971, when it participated in what was then known as the Pochampad Project (para 1.10). That credit supported works to complete the dam and to develop a further 83,000 ha of the service area. Prior to Bank Group's involvement the development of about 17,000 ha near the dam was financed by GOAP. 2.06 The Bank Group's involvement in SRS continued under the AP I Project (para 1.11) which had a component to assist with the construction of field  - 6 - channels, farm roads and land levelling, primarily in areas not covered by the Pochampad Project. Recently, the COAP has completed (i) remaining works on the Sriramasagar Dam to its full capacity of 3,170 Mm3 (112 TMC); (ii) a canal on the left bank to link the Sriramasagar and Khaddam dams to irrigate 17,000 ha between the two (the Saraswathi Head component); (iii) the Laxmi canal on the right bank of the Godavari; (iv) major works, but not all minor distribution works, on the remaining 65,000 ha service area downstream of the Pochampad Project; and (v) the Mannair Dam to impound 13.4 TMC (380 Mm3) of water and act as a balancing reservoir for the system. B. Project Areas Location 2.07 The SRBC scheme is roughly rectangular in shape, (MAP IBRD 18843) about 130 km long and 5-10 km wide, with a cultivable command area (CCA) of 77,000 ha. It slopes gently from the Mallamelai hills to the Kundu river. 2.08 The SRS scheme has total potential CCA of about 679,000 ha. The Sriramasagar Dam is situated near where the Godavari enters AP from Maharashtra in the Telangana Region (Map IBRD 18842). The scheme consists of the following parts: Name Area Remarks ('000 ha) Godavari Basin Left Bank Saraswathi Head 17 under construction; links SRS and Khaddam Dams Khaddam 26 completed, uses Khaddam water (para 6.08) Saraswathi Tail 42 planned Sub-total 85 Right Bank SRS-Mannair 165 nearing completion Mannair-Km 267 178 under construction Laxmi 7 under construction Sub-total 350 Krishna Basin Km 267-284 24 planned for 1992 Km 284-347 220 planned after 1992 Sub-total 244 Total 679 When completed, the SRS and the Nagarjunasagar schemes would form a con- tiguous irrigated area of about 1.5 M ha.  -7- Climate 2.09 Rainfall averages about 700 mm in the SRBC area and about 930 m in the SRS area. Both are in the single-monsoon belt, although occasional storms occur during the northeast monsoon (November to February). Tempera- tures are fairly even throughout the yei.r except for the very hot season of late March to early June. All crops suited to the zone can be grown between June and March: rice; groundnuts; green, black and red gram; sorghum; maize; pearl millet; cotton; and chillies. Chickpeas, a photo-sensitive crop for diminishing daylight hours, can be grown only between November and March. The salient climatic data are in Table 1. Topography and Soils 2.10 Topography. The SRBC area is fairly leveL, with gentle slopes and little segmentation. The SRS area has several secqndary catchments and some quite large seasonal rivers (locally known as vagu) on which there are cas- cades of minor tanks (para 2.18). There are also several granitic and basal- tic outcrops, typical of the Deccan plateau, making for a broken topography. 2.11 Soils. Practically the whoLe of the SRBC and some 60X of the SRS area to be brought under irrigation by the present project are black ver- tisols; the balance are red alfisols. The vertisols in the SRBC area are heavier and deeper than in the SRS area. A typical vertisol cracks deeply when dry, is self-mulching, and because of slow infiltration after heax'y rain or irrigation, tends to have a perched watertable which makes it difficult to cultivate after the first few showers. Therefore, most of the fields are planted in late kharif, when the rainfall eases; the crop matures on residual moisture. Sorghum and red gram (pigeon peas) respond best to this regime. The red alfisols in both areas have sufficient depth to support all crops. Profitable farming systems that effectively utilize year-round irrigation on alfisols have been practised in the State for many years. Population and Land Holdings 2.12 The combined population in the two scheme areas is about 1,140,000; a density of about 120/kn2. The average holding size is 2.3 ha in the SRS and 3.6 ha in the SRBC project areas (Table 2). The major settlement of the SRBC area is Nandyal. Bananganapalle is a market center of importance. 2.13 Two major towns in the SRS are Warangal and Karimnagar. Other cen- ters of importance are Chalgal, Jagtial, Nizamabad (district HQ); all in the area developed in earlier stages of the scheme. Comummications 2.14 Nandyal is on a major branch of the Hyderabad-Bangalore meter-gauge railway line, which connects it with the coastal region. A State highway connects Nandyal, Kurnool, Cuddappah, Cuntur and Anantapur. Secondary roads within the SRBC project area are few, but in fair condition: no point is more  than 10 km from an all-weather road. However, there is virtually no farm road network. 1/ 2.15 The SRS area has an extensive State and National highway network. It is served by meter-gauge and broad-gauge railway lines connecting to major cities within and outside the Stare. An all-weather road network connects the maior towns and most of the major villages to the State and National highways. However, the farm road network is almost non-existent except in the top 120,000 ha of the area, where it was developed under the Pochampad project (para 1.10). 2.16 The canal service road network is poor, which makes supervision of the system difficult. The consequences are poor maintenance of the dis- tributaries and illegal diversions in their upper reaches with shortages at the tail ends. Present Development 2.17 Crops. Table 3 shows the present cropping pattern and yields in the project areas. Drought-resistant crops, such as sorghum and pulses, predominate. Well water (para 2.19) is used for cotton, groundnuts, orchards and rice. The high percentage of rabi sorghum, especially in SRBC command, reflects the dominant farming system on the vertisols (para 2.11). Present yields of unirrigated crops are above average for the State, reflecting reasonably high farming standards. 2.18 Present Irrigation. Irrigation in the SRBC area, and the SRS area below the KMannair dam, is by tanks and wells: Tanks Wells No Area ('000 ha) No Area ('000 ha) SRBC 14 2.8 /a 1,200 2.4 SRS 385 48.0 /b 4,000 8.0 /a 3% of the Gross Commanded Area (GCA). lb 17Z of GCA. The proportion of the area served by tanks is higher in SRS than in SRBC because of the larger proportion of non-commandable area in the GCA and the more broken topography. The tanks are excluded from the project but would benefit indirectly by increased recharge and return flows. The tank commands irrigate mainly rice, sorghum and groundnuts in SRBC but only rice in SRS. 2.19 The dugwells, mostly electrified, typically command 2 ha to supple- ment rainfall in the kharif. In the dry season they irrigate about 1 ha of high-value crops: cotton, groundnuts, orchards and hybrid sorghum for seed. Aquifer conditions allow 7-12 hours of pumping per day. 1/ The term "farm road" is used throughout this report to describe roads between the fields and the farmers' homesteads in the villages. "Feeder roads" connect the homesteads to all-weather roads.  2.20 Although some of the area between the Mannair Dam and Warangal has been receiving surface water, development is partiaL and incomplete. Upstream of the dam the major irrigation network is nearly completed on about 120,000 ha (except for lining of distributaries, minors and field channels) and is under construction on a further 45,000 ha. Hlowever, because of the design of the minors and field channels; deficient control structures, the absence of lining apart from the main canal, lack of effective operational control and unsuitable water regime, tail enders on the distributaries do iKOt receive an adequate irrigation service while the lower-lying areas by the head and middle reaches of the distributaries are getting waterlogged. Agricultural Support 2.21 Extension. As indicated (para 1.14), agricultural extension in the State is supported by the Andhra Pradesh Agricultural Extension Project (Cr. 1219-IN), which operates the Training and Visit (TAV) system. It is by now under the sole control of the Department of Agriculture. Statewide, staffing density averages one village extension officer (VEO) per 1,500 farmers. Coverage in the parts of the project area not yet irrigated is below this ratio; also, the quality of the service has been uneven. However, there is a nucleus of agricultural officers and subject matter specialists in Warangal and Nandyal who would form a solid base, under the State Extension Project, for introducing an effective service. The extension service as staffed in the upstream 165,000 ha of the SRS scheme is adequate. It faces a major task in rapidly extending improved farming systems and technologies to meet the demands of irrigated and rainfed farming in their zones. 2.22 Research. Agricultural research is under the control of the Andhra Pradesh Agricultural University (APAU) near Hyderabad. Assisted by the NARP I Project (Cr 855-IN), it has recently expanded its Nandyal Research Station to serve as a Zonal Research Station for Rayalaseama. A new station at Jagtial (also supported by NARP I) will serve as the zonal station for Telan- gana. Both are fully staffed and are about to initiate research programs to serve their zones. 2.23 Input Services. The seed, fertilizer and pesticide marketing system in the project areas is well developed, as everywhere in the State. Fer- tilizers are marketed mainly by the private sector; the Department of Agriculture sells only about 10. Pesticide use is limited: cotton and to a lesser extent rice are the main users. The sales points are usually the same as for fertilizers. Experience throughout India shows that the vigorous private sector does come in wherever opportunity offers. Seed suppl;es for hybrid sorghum, cotton, groundnuts and other crops are of good quality and on the whole adequate. 2.24 The institutional credit organization follows the normal pattern for India. There is an adequate network of primary cooperative credit societies; commercial banks are weil represented in the villages. The cooperative banks, which provide short and medium-term loans, have the State Cooperative Bank as the State level apex bank. There appear to be an adequate network of branch offices. The State Land Development Bank operates through branches and affiliated offices in taluka (sub-district) headquarters; they handle long- term credit (up to 15 years). Commercial banks are obligated to devote about 15X of their resources to agricultural lending. However, there has been a decline in their lending, in actual cash terms, over the last few  - 10 - years. Recovery levels of cooperative banks are presently about 75X of the demand and statewide recovery performance of the SLDB in 1983-84 was about 9oz. 2.25 Marketing, Processing and Storage. Marketing in the wholesale markets is regulated under the Agricultural Produce Market Act (1965). Facilities have matched the increasing production. Rice in excess of demand within the State is procured by GOI agencies or bought by private traders. Pulses and oilseeds (mainly groundnuts) find a ready market. Processing facilities are adequate: rice husking is a wei. developed cottage and large-scale industry and there are many high-capacity modern solvent plants to procr-s groundnuts; their numbers are growing. Maize is processed into starch; oil is extrated in plants situated in most large towns, including Warangal. Cotton ginning facilities are adequate in the ORS area but not in the SRBC area. However, as has happened in nearby Anantapur when cotton growing has taken off, it is highly probable phat the private sector will come in to provide the necessary ginning units. Storage facilities are adequate: in the public sector, they are provided by the State V1arehousing CorForation, the Central Warehousing Corporation and the State Marketing Federation. Some cooperatives also have godowns. In addition, the mnny private grain and produce merchants have their own stores. Rationale for Bank Involvement 2.26 AP is in the midst of modernizing the planning, design, implementa- tion and operation of development projects in the irrigation sector. The process, which has been assisted by the Bank, changes old norms and prac- tices. The most innovative part would be the introduction of improved system operation procedures which will ensure reliable water supplies to farmers. Past Bank involvement focussed more on improved design and construction standards and to ensure properly phased project execution, with all com- ponents being completed, as far as possible simultaneously. Strong Bank support will now be particularly important in helping the State to take the operational measures necessary to provide a reliable, predictable water supply. 2.27 The project includes an appropriate mix of ongoing and new irrigation development works which are top COAP irrigation development priorities and adequate to spearhead modernization. In addition to its involvement in this project, the Bank is also considering continued involvement in improving water management in the Nagarjunasagar scheme (paras 1.11-1.13) under a proposed Water Management Project. GOAP has decided to merge its Irrigation and Command Area Development Departments, thus rationalizing the management structure of the institutions dealing with irrigation (Chapter V). If all the above plans are successfully implemented, by the end of the project period GOAP would be well advanced in its efforts to modernize its irrigation sector. 2.28 A further rationale for Bank involvement is the poverty-alleviation aspect of the SEBC project. Rayalaseama is the State's poorest, most drought-p_one region (para 1.05). The impact of irrigation would therefore be greatest, as was the case with the Kurnool-Cuddappah project, and shown by the impressive developments in the Tungabhadra Low-Level Canal command which  - 11 - became a prosperous cotton and groundnut-producing area. The Bank's involve- ment would enable the time taken to achieve full production in the scheme areas to be considerably reduced from what would otherwise be the case. III. THE PROJECT A. Project Scope and Objectives Pro lect Objective 3.01 The main objective of the project is to raise agricultural production and farm incomes through the expansion of irrigated areas and the efficient utilization of available water supplies. To achieve these objectives, the project will assist with the further development of the SRS scheme; initiate and complete the development of a greater part of the SRBC scheme; and intro- duce improved design and operation principles as pre-conditions to a reliable and equitable irrigation service. The Bank Group's involvement in the project would extend over a seven-and-a-half-year period. Summary of Project Components 3.02 The project would have the following components: I. Conveyance System Works (a) SRBC Sub-project 1. SRB Canal from the Banakacherla Regulator to the Gorakallu Aqueduct; 2. The Corakallu Aqueduct; 3. SRB Canal beyond the Corakallu Aqueduct, for about 100 km (b) SRS Sub-project 1. Upgrading and completion of conveyance works of the Kakatiya Canal command between km 0-146; 2. Completion and new conveyance works of the Kakatiya Canal command between km 146-267; 3. Excavation of the tail section of the Kakatiya Canal to serve as an escape (km 267-284). II. Components Common to Both Sub-projects (a) block development of about 65,000 ha in the SRBC sub-project, modernization of 165,000 ha and new development of 163,000 ha in the SRS sub-project; (b) feeder road network; (c) water management (paras 6.11 to 6.27); (d) equipment for project implementation, including telecommunications;  - 12 - (e) provision of training facilities and training of professional and support staff required for project implementation; and training of farmers (paras 6.20 to 6.25); (f) rehabilitation of and amenities for oustees from the Srisailam and Mannair reservoir sites and other project works; and (g) monitoring and evaiuation, technical services; provisions for studies and planning. B. Detailed Features The SRBC Sub-project 1/ 3.03 SRB Canal up to the Aqueduct. The SRB Canal which originates at the Banakacherla Regulator will debouch into an aqueduct at km 50.22, to connect to the service area. It would be nearly 12 m deep at the offtake point, to draw a minimum of 750 cusecs (21 cumecs) when the Srisailam Reservoir is at its expected level of 854 ft (260.3 m) at the end of the cropping season (March). Excavation of about 10 km of this canal has already been awarded and would not be financed by this project. 3.04 The CorakalLu Aqueduct will cross the Corakallu valley between km 50.22 and 53.46 (3.24 km), to reach the first block of the irrigation service area. 3.05 The SRB Canal beyond the aqueduct (2,400 cusec capacity at its head reach) would consist of two parts along the service area: (a) between km 53.46 and 112.73, it traverses gently rolling country, without significant cuts or fills until the last few kilometers. This section would command about 41,000 ha in 11 blocks, each served by its own distributary; (b) at km 112.73 the canal abuts on the Mallamelai hills, where two tanks, with their independent catchments, command about 670 ha. 1/ GOAP has already began construction of works that are common to the Madras water supply and the SRBC projects. These are: (a) a 3 km approach channel within the Srisailam Reservoir, to convey water at an elevation of 854 ft (260.3 m) above MSL to the Pothired- dypadu Regulator; (b) construction of the Pothireddypadu Regulator; (c) the "Deep Cut" - a canal ;6.34 km long and at places about 30 m deep, to take water from the Krishna river to the Pennar Basin; and (d) the Banakacherla Regulator which divides the water between the Madras canal, the SRBC and an escape into a seasonal tributary of the Pennar. These works would not be financed under the proposed project.  - 13 - Provisions would be made to divert water from the SRBC in drought years to stabilize their command areas. This section of SRB Canal would command about 24,000 ha and stop at km 153 under the project. The last part of the SRB Canal and the balance of 16,000 ha would be developed later. The SRS Sub-project 3.06 The first section of the Kakatiya Canal (km 0-146) has a capacity of 8,600 cusecs, extends from the Sriramasagar Dam to the Mannair Reservoir and commands 165,000 ha, including the Pochampad Project area (para 1.10). The banks of the main canal and its lining would be raised to enable peak water requirements to be delivered for the entire command. AdditionaL struc- tures would be provided to enable implementation of a more effective water management system. 3.07 The Kakatiya Canal's second section (km 146-284), also with the capacity of 8,600 cusecs at its head reach, originates at the Kannair Dam. The Bank Group's involvement under the proposed project would focus on three segments: 1/ (a) the Mannair Dam to km 234, just beyond Warangal town, where lining as well as construction of several distributary offtake structures remain, to be completed; (b) km 234-267 where all earthwork contracts have been let and some of the distributary offtakes and gates installed. Works not yet con- tracted 2/, including lining, would be eligible for financing under the project; and (c) km 267-284 where no work has started, but some contracts have been let. This section would only be excavated in the present time-slice to serve as an escape into the Akheru river. The project would finance all remaining contracts. Block Development and Upgrading 3.08 Each irrigation block (the unit served by a distributary) would include the following works under the project: (i) the distributary and minors; (ii) field channels leading to 8 ha units; (iii) farm roads (para 2.14); and (iv) drains. A collector drain from each block, with field drains connected to it, would debouch into a main drain or river. The drains would be designed tc. evacuate a 10-year intensity storm within 72 hours. Safety would be ensured by escapes on the tail of each distributary. 3.09 In the Kakatiya command (km 0-146) upgrading the existing dis- tributaries and minors, primarily by lining to the standards set by the 1, Contracts have been awarded for the excavation of 138 km, between Mannair and the Akheru river. 2/ Between km 234-267 contracts for sections of some of the branch canals and larger distributaries have been awarded and work has started, includ- ing the last, at km 267 (D-17, CCA 61,000 ha).  - 14 - project (para 3.24) IJ constructing additional conveyance channels as required; and development of the irrigation blocks for 165,000 ha of net CCA would be completed. The surface drainage system vould connect to the minor tanks in the gross command area (para 2.18), thereby augmenting their vater supply. Feeder Road Network 3.10 Rural roads in the project area would be built or upgraded to stand- ards established by GOAP which are satisfactory to the Bank Group. The network density in both scheme areas would provide all-weather access to all villages with a population of more than 1,000. No part of the irrigated area would be more than 3 km from an all-weather road. The rural road network would average about 4-5 inha of CCA newly bought under irrigation; experience gained since the construction of the Pochampad and other projects suggests that augmentation averaging 2 m|ha of CCA is required to provide adequate communication facilities. Sample surveys show that the SRBC and SRS areas would require a feeder road network of about 260 km and 900 km. Design standards would be: (i) 3.5 a wide roads with a surfacing of waterbound macadas; (ii) no black topping; and (iii) construction of hard passages instead of bridges over seasonal streams. The subgrade design would be according to requirements dictated by the traffic (typically 15 cm deep) and the foundation according to California Bearing Ratio (CBR) values (a worldwide standard accepted in India). 3.11 To ensure that planning, design and construction of the road com- ponent can progress unhindered, an agreement has been reached vith GOAP that by April 30, 1987, and thereafter by April 30 of every year, it would submit to the Bank/IDA for review: (i) plans for the extent and location of the road reserves to be acquired in the distributary commands scheduled for development in the following construction season; and (ii) design specifica- tions and cost of the roads to be constructed. 3.12 Cost estimates of civil works assume that contractors would provide their own construction equipment. Equipment for operation and maintenance during construction and for maintenance of farm and canal service roads would be provided by the project and transfered to the O&N administration after project completion. The Roads and Buildings Branch of PWD would maintain the farm and feeder roads but not the canal service roads: these are part and parcel of the canals. Computation equipment would be required for the design office during the design phase and for subsequent operational control. Ten 16 bit microcomputers are included for this purpose. Annex 3, Tables 1 and 2 lists the items required. Teleco glcatiov8 3.13 A comprehensive telecommunication network would be provided in the 8RS area with substations at all major canal offtakes and other key points. In addition, four vehicles operating in the command area vould be equipped 1/ Distributaries and minors were not lined under the Pochampad Project (,ara 2.20).  - 15 - with two-way radios. A simpler system would link key points in the SRBC project area with the Banakacherla Regulator, the Pothireddypadu gates and the main control room at the Srisailam Dam. Two of the inspection vehicles would be equipped with radios. The radio frequencies to be used would be determined in consultation with the telecommunication authorities of GOI. An agreement has be(;n reached with GOI that it would allocate suitable radio frequencies and promptly authorize the installation of the necessary equip- ment. Training and Training Facilities 3.14 Engineering Staff. Pre-service and in-service training of engineers in the design, construction and O&M of irrigation systems, operated according to standards envisaged for this project, would take pLace in the Water and Land Management Training and Research Institute (WALAMTARI) near Hyderabad. The project would finance the cost of training and improvemetts to WALAM- TARI's facilities. It is expected that during the project implementation period over 1,000 engineers of various grades would be involved; all would require in-service training and/or refresher courses. 3.15 Support Staff and Contact Farmers. Training of sub-professional irrigation staff and farmers would take place in training centers within the project areas. One would be established in the SRBC area; three in the SRS area. An additional center in the S'S area, partially built, would be com- pleted. Each would have dormitory and resthouse accommodation, classrooms and teachers' quarters. Canal Inspectors (CI), engineering technicians and the Chairmen of Outlet and Minor Canal Committees (paras 6.21-6.26) would undergo training to enable them to function effectively as technicians and community leaders. Some of the contact farmers selected by the Agricultural Extension Service would undergo routine training in on-farm water management and distribution practices; other farmers would be brought in as time allows. 3.16 The principal orientation of the training courses for support staff would be: (a) Canal Inspectors' training would focus on canal operation and water allocation principles established under the project within outlet and minor canal commands; (b) OC Chairmen would be trained in water allocation principles within outlet commands, responsibilities for maintenance of common works, their legal obligations and instruments to implement them; and (c) Contact Farmers would be trained in improved irrigation practices, field layouts and appropriate land preparation to minimize on-farm water losses, legal rights and obligations associated with water use, and administrative and management procedures of OCs. Since the majority of the farmers are not in a position to leave their homes- teads for several days, principally because of the routine demands of live- stock, VEOs and the Subject Matter Specialists (SMS) dealing with irrigated farming would also be trained to enable them to impart, and demonstrate to farmers on their own land and under their own specific soil, topography and water supply conditions, points of field layout and land preparation. Full details of the syllabi and organization are in Annex 5. The project would finance the operating costs of the training centers during the implementation  - 16 - period; however, training activities would continue throughout tne life of the project. An agreement has been reached vith GOAP that they would present to the Bank/IDA for review training plans for engineering staff, support staff, and Chairmen of the various farmer committees, and thereafter imple- ment the agreed training plan. Land Alienation 3.17 Land Required for Project Works. Land acquisition for works financed by the present project would only be for roads and canals. These are generally seen to serve the community, rarely engender significant objection and the land reqcired is readily acquired under vell-established procedures of GOAP. No acquisition would imply the resettlement of villages and no dwelling or storage structures built oa farmland would be affected. 3.18 The Srisailai. Reservoir. The site of the reservoir was acquired gradually from the occupants from 1974 onwards but was not finally flooded until 1981. Consequently, some of the potential Project Affected Persons (PAPs) to whom GOAP paid compensation in the 1970s, disregarded the threat of flooding and have spent compensation monies while still in occupation of their original houses and farms. Others found the compensation, which may well have represented actual replacement costs in 1974, inadequate to buy replacements in 1981. While GOAP has paid the PAPs additional compensation to alleviate hardship caused by the delays and inadequate advance warning of the proposed filling date of the reservoir, it recognizes that there is room to provide additional relief and amenities. 3.19 The Nannair Dam and Reservoir. The Mannair Dam, through which all water reaching the area below khu 146 must pass, has been completed in 1985. When eventually filled to full supply level, the reservoir will submerge 16 villages, 7,365 ha of land (including the village sites) and affect the livelihood of 13,799 families or about 63,370 persons. At present, resettle- ment is still in progress, providing an opportunity to organize it in a more systematic way than has been done for Srisailam. Amenities can be made available more promptly and, what can no longer be done for the PAPs of Srisailam because of elapsed time and dispersal, vocational and rehabilita- tion training can be organized. 3.20 To ensure assistance to the affected families GOAP would (a) for Srisailm, conduct a survey in villages where most of the displaced families have moved, to determine what additional social amenities, (e.g. schools, clinics, drinking water supplies) are required to provide them with the same amenities they enjoyed previously; (b) in addition, for uannair. they would: (i) determine specific vocational training needs of affected families and (ii) prepare a Resettlement and Rehabilitation (R&R) Program, with an implementation schedule; (c) for persons directly affected by the project (para 3.17) GOAP would also prepare a R & R program and a timetable for its implementation, which would include either provisions for: (i) vocational training; or (ii) irrigable land of comparable production potential to what the PAP has held before acquisition, or (iii) stable employment, as preferred by the PAP. An agreement has been reached with GOAP that they will resettle and rehabilitate the PAPs in accordance with an action plan satisfactory to the Bank/IDA; such plan would be based on paras 3.17-3.20 and Annex 7 of the BAR. The program, along with the Action Plan covering the Srisailam and Nannair sites, would be submitted to the IBRD/IDA for approval, as a condi- tion of Loan/Credit effectiveness.  - 17 - Monitorina and Evaluation: Technical Services: Studies and Planninz 3.21 The COAP has already upgraded its Project Preparation Cell (formerly under a Superintending Engineer) to a Project Preparation and Monitori'g (PPM) Unit, in charge of a Chief Engineer. Under the project, this Unit would be expanded and put in charge of technical services and studies required for effective project design (paras 3.22-23) and implementation. PPM would, either within its own resources, or more probably by engaging consultants, conduct and oversee a long-range investment options study to identify the priorities for future irrigation development in the State. The consultants would be either independent firms, Departments of Universities - not necesarily based in AP - or a group of individuals engaged for the purpose by PPM. Additional studies may well include, but not be confined to, the folloving: (i) irrigation development options (in Telengana and Rayalaseama), taking medium as well as major projects into consideration; (ii) institutional improvements that may appear necessary after the merging of ID and the Irrigation Utilization and Command Area Development Department (para 5.08); (iii) an extensive study of the Coastal Region with a view of improving irrigation and - more importantly - the drainage infrastructure of the deltas and identifying agricultural development options and research needs, further to improve production and crop diversifica- tion in the area. Appropriate provisions have been made to finance these studies under the project. Organizational details are in paras 5.25 - 5.27. Desiuin Criteria 3.22 Desi8zn PrinciRles. Both systems would be designed to standards satisfactory to the Bank/IDA which would enable water management that ensures a reliable and equitable supply to all farmers in the project areas (paras 6.11-6.20). The components of the structured network JJ to be constructed under the project, are: (a) A main canal and branch canals, designed to be regulated to supply distributaries in rotation, in 7-day cycles, to meet the demands of the farmers'. cropping pattern within their allocation. Sufficient cross regulation structures would be constructed to allow demand in the distributaries to be met in any sequence. Ultimately, only distributaries vould be supplied from the main and branch canals. Canal capacity would be sufficient to meet the average computed peak crop requirements. J/ See proposed Guidelines for the Design and Operation of Structured Sys- tems,. World Bank, October 1985.  - 18 - (b) Distributaries, typically serving 2,000-6,000 ha, to run either full on or off, serving minors through proportional distribution flumes or modules. A measuring flume would be built near each distributary head; there would be gates only at the head of the distributaries. Capacity would be sufficient to meet peak crop requirements with continuous flow. Regulation for less than peak irrigation require- ments would be achieved by running the distributary intermittently. (c) Minors, serving groups of chaks, aggregated into 200-500 ha minor service areas. The chaks would be supplied through proportional distribution flumes. There would be no gates along the minors: they would flow either full on or off, following the regime of the parent distributary. (d) Chaks of modular size, corresponding to the estimated usable range of farm stream size, depending on soil and topography. Field channels within the chaks will serve 5-8 ha zones through simple on/off divi- sion boxes. Flow at the chak head would be continuous while the distributary is running; distribution to the 5-8 ha zones would be on a warabundi schedule (para 6.18). 3.23 Conveyance losses would be accurately determined, in part by making use of the extensive data collected in the AP I Project, and appropriate allowances made for them in final designs. Expected losses in the dis- tributary and minor systems will be aggregated and distributed so that every chak outlet receives its design duty. Losses in the chaks will be allowed for by adjusting the warabundi schedules after actual field measurements of losses so that every farmer receives an equal water duty. When eventually the system is fully lined (para 3.24), the proportionality of the division flumes will need adjustment. 3.24 Canal Lining. Lining canals and field channels down to 8 ha groups of fields not only reduces conveyance losses, but also, with proper water allocation and management, reduces drainage needs and the risk of waterlog- ging and possible salinity. More importantly, it makes for an easily managable, responsive system. Further, as experience in other projects shows, without adequate lining tailenders on distributaries and minors do not get their share of water because of conveyance losses and uncontrollable, illegal abstractions at the upper reaches, which cannot be balanced by providing excess flows once the system is fully developed. However, due to budgetary constraints, lining of the minor system will have to be deferred until the water saved is required to expand the command area. 3.25 Under the present project, only canals conveying more than 35 cusecs would be lined. The following specifications have proved successful in other projects on similar soils and similar operating regimes: Hain Canal: 10 cm M100 grade in situ concrete (with 100 cm cohesive non-swelling (CNS) backing if the canal passes through black cotton (BC) soils) Distributaries: greater than 50 cusecs: 7.5 cm M100 grade in situ concrete (with 100 cm CNS backing in BC soils) 25-50 cusecs: 7.0 c M1100 grade in situ concrete (with 60 cm CNS backing in BC soils)  - 19- In addition, all the following canals or canal sections would be lined: (a) upstream and downstream of structures, along embankments and around curves to the extenc deemed necessary; (b) in areas of seepage rates more than 5 cfs/M ft2 (for field channels, more than 2Z/100 m) (c) in fills exceeding 30 cm, in pervious soils or in fractured rock; (d) in stretches where the watertable is within 60 cm of the surface or where, because of the nature of the terrain, natural drainage is impeded. 3.26 Distributary Sizes and Commands. A major 4esign parameter would be the size of distributaries off the Kakatiya Canal. Several distributaries of the Kakatiya Canal between km 0-146 carry only 4-5 cusecs and command as little as 80 ha. For ease of control, a distributary should command not less than about 1,000 ha, i.e. take a flow of about 25 cusecs. However, of the 44 distributaries that were originally designed to carry less than 25 cusecs, 37 are already built. COAP proposes to change the design of the remaining 7 to meet the principle outlined. They propose to proceed with combining as many of the other 37 commands as possible during the implementation period of this project but may find it expedient and cost effective to postpone some of the proposed clubbing to a later stage. Accordingly, an agreement has been reached with GOAP that, except as the Association shall otherwise agree, take all such action as shall be necessary to ensure that each distributary linked to the Kakatiya Canal between kilometers 146 and 267 shall carry no less than 25 cusecs of water and command an area no less than 1,000 hectares. Construction Methods and Current Progress 3.27 Technology Used. Present construction technology uses on labor-intensive methods, which are suitable for most of the project works. Excavation in rock is aided by blasting; there, and in deep cuts (e.g. parts of the SRB Canal and the Kakatiya Canal) tipper trucks are mechanically loaded. Elsewhere, tipper trucks are filled with headloads. Canal banks would be compacted in layers by using sheeps' foot rollers. Portable mixers would prepare all concrete for the structures. 3.28 Progress on the Madras Water Supply Canal. Progress on components of the Madras Water Supply canal, which would also carry water to the SRBC project, in particular on the "Deep Cut" (para 2.04) is satisfactory. According to the water supply agreement between the two States ("Agreement entered into between the Governments of Andhra Pradesh and Tamil Nadu for the Drawal of Water from River Krishna for Madras City Drinking Water Supply") GOTN pay all costs directly attributable to conveying 15 TMC of water in three months to the AP/TN border. This includes the cost of the "Deep Cut" as it is designed and currently constructed. However, although the risk of delay because of late payment or default by GOTN is minimal, an agreement has been reached with GOAP that it would make provisions to finance works on the Madras wate.r supply canal upstream of the Banakacherla Regulator, and of the regulator itself, to the extent necessary to enable 19 TNC of water to be conveyed to the SRBC project area, starting on or before September 30, 1987.  - 20 - Status of Project Preparation 3.29 In the SRS scheme, GOAP is completing works above the Mannair Dam and has started some of the major excavations below Mannair (paras 3.03-3.05). Detailed design work keeps up with contract awards and there have been no delays. However, the time lag between the delivery of detailed designs and the requirements of the contractors and site engineers has been so narrow as to cause concern. Moreover, below the Mannair Dam, planning focuses on works above the chak outlet and not on field channels and field drains. In the SRBC scheme, the first 50 km of the canal have been designed and some contracts awarded, but final detailed design of the 66 cross struc- tures and bridges has not yet been completed. The distributiqn system has only been designed in sample blocks, to establish costs. 3.30 To ensure the smooth progress of construction, GOAP has submitted design and contracting packages, satisfactory to the Bank/IDA, that cover all proposed works in the 1986-87 construction season. Plans and design criteria to allow the temporary cultivation of rice, relying only on surface water resources, has also been submitted and found satisfactory. However, to ensure that design and contracting packages are always satisfactorily prepared in good time before the beginning of the construction season, an agreement has been reached with GOAP that it will submit detailed designs and draft contract documents for all works proposed in both sub-projects by April 30, 1987, and thereafter by April 30 of each year for all work to be let to contractors for execution during subsequent construction seasons. 3.31 To enable GOAP to meet this condition, all or part of the above design and planning work may be contracted to a competent outside agency and be part of the technical services and support component of the project (para 5.27). Project Execution 3.32 In the past, major worke like dams and large canals tended to be completed long before the distribution system; delay in full block develop- ment, down to the 8 ha le-vel, was even longer. This unconcerted development has been causing considerable delay in effectively utilizing irrigation potential. Therefore, to shorten the time between incurring expenditure on civil works and actually irrigating a block, an agreement has been reached with GOAP that it would develop the distribution system, including the field channels, concurrently with building the major canals.  - 21 - IV. COST ESTIMATES AND FINANCING Project Cost Estimates 4.01 The estimated project costs over a seven-and-a-half-year period would total US$475.8 M, including physical and price contingencies. This amount includes foreign exchange cost of about US$76.4 M (16% of the total) and about US$22.8 M (5% of the total) for taxes and duties. Estimated base cost are given in January 1986 prices. Cost estimates are given in Table 4a to 4f (details are in Annex 3) and summarized below: Project Cost Summary Z Total (Rs Million) (US$ Million) Z Foreign Base Local Foreign Total Local Foreign Total Exchange Costs A. SRIRAMASAGAR 1,628.8 452.4 2,081.1 125.2 34.8 160.0 22 48 B. SRISAILAM 1,963.2 260.7 2,274.0 151.0 20.1 171.1 12 51 C. SUPPORTING SERVICES 57.9 5.2 63.1 4.5 0.4 4.9 8 1 Total BASELINE COSTS 3,649.9 718.3 4,368.2 280.7 55.3 336.0 16 100 Physical Contingencies 365.0 71.8 436.8 28.1 5.5 33.6 16 10 Price Contingencies 1,154.0 200.6 1,354.6 88.8 15.4 104.2 15 31 Total PROJECT COSTS 5,168.9 990.7 6,159.6 397.6 76.2 475.8 16 141 4.02 Physical contingencies average 10%. Price contingencies account for over 222 of the total project cost. Local price contingencies are assumed to be 7.0Z for 1986/87 - 1987/88, 7.5% for 1988/89, 7.7Z for 1989/95, 7.5% for 1990/91 and 4.5Z from 1991/93. Foreign price contingencies are assumed to be 7.0Z for 1986 and 1987, 7.5% for 1988, 7.7Z for 1989, 7.6% for 1990 and 4.5% for 1991-93. Engineering and administration costs are estimated at 14Z of construction cost. Financing 4.03 The financing plan for the Droposed project costing US$ 451.1 M, net of taxes and duties (US$ 22.8 M), is as follows: US$ Million GOI/GOAP 182.0 IBRD Loan 131.0 IDA Credit 140.0 Total 453.0 The proposed IBRD loan/IDA credit of US$ 271 M would finance about 602 of total cost, net of taxes and duties, including the foreign exchange component (US$ 76.2 M) and 52Z of the local cost. All GOI/GOAP expenditures would be met by GOAP from its development budget, which includes a GOI contribution.  - 22 - The IDA credit would be made to GOt on standard IDA terms, and the loan from IBRD funds would be provided according to the standard IBRD terms. 4.04 An agreement has been reached with GOI that it will make the proceeds of the credit/loan available to OOAP on standard terms and arrangements for development assistance to the states. In addition, an understanding has been reached with COAP that on April 1 of each year, commencing on April 1, 1987 and thereafter by April 1 of each year, inform the Bank/IDA of the annual budgetary allocations made for meeting, during the forthcoming fiscal year, the financial needs derived from the implementation of the project, invest- ment costs, O&M costs, administrative costs and other costs related to the project. Procurement 4.05 Procurement arrangements including contingencies are summarized below: Procurement Method /a (US$ M) Force Total Cost Project Element ICB LCB Account NA Procurement Works (Irrigation, Roads 252.3 122.0 4.5 61.7 /b 440.5 and On-Farm) (169.8) (79.3) (249.1) Equipment, vehicles and 4.3 1.3 - - 5.6 supplies ( 4.0) ( 1.0) (5.0) Training & technical services - L - - 5.9 5.9 (.-9) (5.9) Rehabilitation - - - 1.0 1.0 _(1.0) (1.0) Total 256.6 123.3 4.5 68.6 453.0 /c (173.8) ( 80.3) (6.9) (261.0) /d Percentage of Total 56.6 27.3 1.0 15.1 100.0 (66.6) (30.7) t-) (2.7) (100.0) /a Figures in parentheses are the amounts to be financed by the Bank Group. /b Includes US$13.0 M for land acquisition, US$45.0 M for engineering and administration and US$3.7 M for survey and investigation. /c Costs are net of taxes and duties of US$22.8 M. /d Excluding unallocated amount of US$10 M Loan/Credit. 4.06 Civil Works. Construction works contracts under the project, includ- ing contingencies, would cost approximately US$378.8 M (US$440.5 M less US$61.7 M see footnote b/ above). Fourteen of the contracts costing approximately US$252.3 N (or 66% of the irrigation works, costing US$378.7 M) would be let through international competitive bidding (ICB). Eligible domestic bidders would be afforded a preference of 7 1/2Z. Twenty-eight other irrigation contracts, generally costing over US$1.0 M and all less than US$6.0 M, with a total value of approximately US$78 N are large enough to be  - 23 - shown in the procurement schedules (Tables 5a to 5g). However, because they would cost less than US$6.0 M per contract, they are not large enough to be of interest to foreign bidders and would be let through local competitive bidding (LCB). The procurement schedules give detailed projections on how and when the forty-two larger works are expected to be put to bid, started and completed. The remaining works to be financed under the project, total- ing approximately US$44 M, consist of: (i) on-farm development, small works and buildings with individual valtues of less than US$1.0 M, scattered throughout the project areas and let through LCB, and (ii) works that can only be done by force account, approximately US$4.5 M, primarily required for railway relocation and other works not readily contracted out. 4.07 Whenever practicable, all excavations, lining and structures in a given segment of canal would be combined in one tender; such combinations would be innovative in this project. This would produce efficiencies and economies resulting from having only one main contractor in one area, which would enable, for instance, utilizing spill from excavations to construct embankments in the same unit. 4.08 Goods. Equipment and vehicles, costing approximately US$4.3 M, would be grouped in appropriate bidding packages and procured through ICB. Qualified domestic manufacturers would receive a preference of 15% or the equivalent of the import duty, whichever is lower. Light equipment and smaller vehicles, estimated to cost US$1.3 M in total, are readily available domestically at competitive prices, and will be procured through LCB. Because of the need for service and spare parts facilities, considerable advantages would be derived from procuring locally manufactured light equip- ment and vehicles. The LCB procedures are satisfactory to the Bank. Annex 3, Tables 1-3, shows details of procurement procedures to be adopted for each type of equipment. 4.09 Procurement Review. All procurement actions for works estimated to cost more than US$0.5 M and goods estimated to cost over US$100,000 would be subject to the Bank's prior review. This procedure would cover about 80% of the value of the procurements. The balance of the contracts would be subject to Bank's post-action review. Disbursements 4.10 Disbursements under the loan/credit would be made against: (a) civil works and buildings--65% of expenditures on main canals, dis- tributary canals, minors, main drainage, roads and chak development works (link drainage, field channels, field channel protection and field drains); (b) vehicles and equipment: (i) directly imported--lOOZ of foreign expenditures; (ii) locally manufactured--lOOZ of ex-factory costs; and (iii) other locally procured--70Z. (c) training, technical assistance, monitoring and evaluation --100x; (d) rehabilitation --100Z.  - 24 - 4.12 Disbursement requests for civil works would be made against certifi- cates of completion of units and statements of expenditure submitted by GOAP, itemized by project component. Disbursements for chak development works would be submitted together with a completion certificate duly signed by the officer in charge of the contractors, stating that all works in the chak have been completed in accordance with the agreed planning and design criteria. Disbursements for payments of less than Rs 300,000 for works (including departmental works) or Rs 150,000 for goods, training, adaptive and engineer- ing research, on-farm development, monitoring and evaluation, the costs of special surveys and rehabilitation services may be against statements of expenditures. Documentation would be retained by GOAP, for inspection by the Bank Group during review missions. Full documentation would be required for all other disbursements. 4.13 The semiannual disbursement schedule and the proposed allocation of the proceeds of the loan/credit are presented in Tables 6 and 7. The disbur- sement profile is based on historical disbursement profiles for Indian irrigation projects. It is expected that disbursements would be completed by December 31, 1993. Accounts and Audits 4.14 An agreement has been reached with GOAP that it would: (i) maintain separate records and accounts adequate to reflect, in accordance with sound accounting principles and procedures consistently applied, its resources, expenditures and operations related to the project; (ii) furnish to the Bank/IDA copies of accounts and financial statements for each year, certified as to their accuracy and authenticity by an independent auditor acceptable to the Bank/IDA, not later than nine months after the end of each fiscal year; (iii) have accounts and financial statements for each fiscal year audited by independent auditors acceptable to the Bank/IDA; (iv) furnish to the Bank/IDA immediately following its finalization, the report of such audit in such detail as the Bank/IDA shall reasonably request; (v) furnish to the Bank/IDA any other information as it shall reasonably request; and (vi) enable repre- sentatives of the Bank/IDA to examine all records pertaining to the Loan and Credit. V. ORGANIZATION AND MANAGEMENT Introduction 5.01 All surface irrigation projects in Andhra Pradesh are constructed by the Irrigation Department (ID) which also operates and manages all but three schemes. The Irrigation Utilization and Command Area Development Department (IU and CADD - referred to as CADD) is at present involved in three large schemes: Nagarjunasagar, Sriramas&gar and Tungabhadra, where it is in charge of operations below the chak outlet. In other schemes operation and manage- meat beLow the outlet is done by farmers, assisted by the ID. The ID comes under the State Minister of Irrigation; IU and CADD, since March 1985, under the Minister of Agriculture.  - 25 - The Irrigation Department- Present Organization 5.02 Senior Management. The Irrigation Departmezv is headed by a Secretary. Under him, the Chief Executive Officer and GOAP's principal irrigation adviser is the Engineer-in-Chief to whom nine Chief Engineers (CE) report directly, one of whom heads the Project Preparation and Monitoring (PPK) Wing (Chart 1). The CE SRS is in charge of new constructions in his scheme area and also operates and maintains its completed parts (Chart 5). The CE Srisailam maintains the Srisailam Dam, is in charge of common works of the Madras Canal comon to the water transfer (see footnote 11 on page 12) and the SRBC project (Chart 4). He will eventually control operations. Both CEs have liaison with the administrations of project districts through Special Collectors, who function independently of the District Collectors. The CEs report directly to the Secretary of Irrigation on matters of policy. Under the CEs, Deputy CEs act as senior staff officers. Each CE has a Design and Quality Control Circle, headed by an SE. The SE Designs, under the technical control of the CE (Central Design Organiz.tion), is responsible for designs of all major works and structures. 5.03 Executive Management rests with SEs in the Circles. The Circles under CE SRS deal with dam operation, construction, construction combined with operation anwz maintenance, or operation and maintenance only. The ID's operation and maintenance (O&M) function ends at the chak outlet. The uni_ below the Circle is the Division, headed by an Executive Engineer (EE). Whereas a Circle may have more than one funct:on (typically, construction as well as O&M), a Division has only one. However, it is flexible and can convert, for example, from construction to OEM. Typically, four to seven EEs report to an SE, four Deputy EEs (in charge of Sub-Divisions) to the EE. A Deputy EE has four Assistant Engineers (AE). In construction divisions the AEs are on construction sites, supervising the accuracy and quality of the contractors' work. 5.04 In O&M Divisions, the AE supervises a group of six Canal Inspectors (CIs). Each CI is in charge of a group of six skilled laborers (laskars). The workload of the AE and CI is determined by an administrative order. The Irrigation Utilization and Command Area Development Department (CADD) 5.05 Senior Management. The Department is headed by a Secretary. A CE for Water Management, a SE for other works in the command areas, and the Director of the Water and Land Management Training and Research Institute (WALAMTARI) are his senior staff officers. Officers of middle management rank coordinate the disciplines of cooperation, credit, statistics, and on-farm and agricultural production development (Chart 2). 5.06 Operations Management. In the three schemes where CADD operate (para 5.01) there are Command Area Development Authorities (CADA) headed by an Administrator (Chart 6). The Administrator's staff consists of officzrs on deputation from line departments. CADA's function is to overcome constraints to agricultural production and to coordinate the implementation of programs formulated by the Command Area Development Board. 5.07 The Command Area Development Board deals with policy issues and is chaired by the Minister of Agriculture. It consists of the Ministers of Irrigation and Finance', the Chief Secretary, and all concerned Secretaries;  - 26 - it meets four time,, a year. The CADA Working Committee, which meets monthly, is chaired by the Secretary, CADD and consists of the heads of departments in the CADA area. Its main functions are coordination and dealing with issues of a local nature, e.g. input supplies, marketing problems and water supplies and allocations. The Proposed Organization 5.08 Combining D and CADD. The duality of the ID and CADD organization has proven less than satisfactory for effective project operation. There was no single line of control of water deliveries, schedtuling and utilization between headworks and chak outlets. Therefore, G(AP has decided to bring the two departments under a single administrative controL, as is the case in many other States. COAP has advised the Bank/IDA that it expects to complete all necessary official steps to formalize the merger by April 1, 1986. It is expected that the combined department would assign all tasks pertaining to construction of irrigation projects to the Irrigation Wing and all tasks concerning maintenance and operation of water delivery systems, water indent- ing, water management, organization and guidance of water users' committees and the coordination of farmer services not directly related to irrigation, to the CAD and O&M Wing. 5.09 Following the establishing of a combined irrigation department, GOAP intends to appoint a Merger Committee to determine exact allocation of responsibilities between the newly established wings, lines of conmand, etc. The Merger Committee would use consultants as required and visit other states to learn from their experience. An allocation has been made in the cost estimates to cover expenses likely to be incurred by this committee. Annex 6 gives full details of present and proposed organization. Staffing Requirements and Qunctions under the Project 5.10 The present design and construction organization of the CEs heading the two sub-projects would require strengthening to cope with the expected workload. The SRBC sub-project would require, at its peak, a total of 327 engineers; the SRS 1,472. The proposed staff buildup is given in Tables 8a and 8b. A training program would be required to prepare staff (in many cases already serving) for their new, more exacting roles. Special emphasis would be required to train design engineers, to create the capability that ensures plans are prepared and reviewed well in advance of the time they are required (para 3.30). Table 16 outlines the type of training to be arranged for irrigation engineers. An agreement has been reached with GOAP that it would (i) sanction and fill staff positions in accordance with an organizational plan and staffing schedule established in consultations with the Bank/IDA; (ii) train such staff to carry out effectively the project and to operate and maintain project works according to accepted good engineering standards. 5.11 Routine O&M of the SRS system would be put under separate O&M Circles. In accordance with GOAP norms (para 5.03), the upstream 165,000 ha would be controlled by an O&M circle consisting of three Divisions. As the area below the Mannair Dam is completed and passes out of the two-year com- missioning period during which the Construction Circle is responsible for maintenance, up to four more O&M Divisions would be required. Eventually, three 05& Circles are envisaged: one would maintain, operate and control the dams and have one or two O&M Divisions; the rest of the area would be divided between the other two Circles which would have only OM duties. The feeder  - 27 - roads would be maintained by the Roads and Buildings (R&B) Department of PID (para 5.23). Dam Safety 5.12 Pursuant to a resolution by the conference of State Ministers of Irrigation in 1980, proposals have been submitted to GOAP for the formation of a Dam Safety Panel, as well as a Dam Safety Cell. The Cell, in charge of an Executive Engineer and under the operational control of the Engineer-in-Chief, collects pertinent data on the behaviour of the dams (e.g. seepages, piezometric pressures) which are available to the Dam Safety Panel during their periodic inspection visits. The Panel, in accordance with GOI's general guidelines, is responsible for monitoring aUl dams in the State. An effective Dam Safety Panel is essential for public safety as well as for reliable, safe irrigation and power generation. Therefore, an agreement has been reached with GOAP that it would, by December 31, 1986 establish and thereafter maintain the Dam Safety Panel with membership and terms of reference satisfactory to the Bank/IDA and cause it to inspect periodically the Sriramasagar, Mannair and Srisailam dams in accordance with sound engineering practices, to determine whether there are any deficiencies in the condition of the structures or in the quality and adequacy of maintenance or method of operation of the same which may endanger their safety. The Dam Safety Panel would report to the Engineer-in-Chief and would have available all the engineering disciplines required and recommended by CWC. Agricultural Supporting Services 5.13 Agricultural Extension. Under the Andhra Pradesh Agricultural Exten- sion Project (Cr. 1219-IN) agricultural extension is the sole responsibility of the Department of Agriculture's (AD) Extension Wing (Chart 7). Agricul- tural extension issues affecting the timely realization of project benefits are (a) staffing densities in the project areas and (b) a senior technical control organization dealing only with problems of irrigated agriculture. Experience under the AP I Composite Project has shown that extension work is more effective if irrigated areas are served by staff who specialize in the problems of irrigated agriculture and if one Village Extension Officer (VEO) is assigned to no more than about 800 farmers. Further, experience with agricultural extension organizations in India and elsewhere shows that the best results come if the staff in contact with farmers, their supervisors, technical support and training officers are solely concerned with one type of farming system: in case of this project, only with irrigated farming. Therefore, the sub-projects would be organized into special extension areas and, with the exception only of out-of-command pockets within the CCA of each of the projects, these special areas will only have irrigated areas. Within the special extension areas agricultural extension staff would be organized on an outlet rather than a village basis, to ensure effective linkage with Irrigation Department staff and farmers groups. They would be administered under the organizational structure of the Department of Agriculture. Because the special extension areas cut across district boundaries, the Subject Matter Specialists (SMS) would be responsible to the Senior Joint Director of the Zone under which the special areas fall. The SMS tend the monthly workshops, thus assuring an effective research-extension linkage. Accord- ingly, an agreement has been reached with GOAP that they would establish and, thereafter, maintain an agricultural extension system in accordance with organizational and staffing arrangements satisfactory to the Association, which shall include: (i) two special agricultural extension units in the  - 28 - right bank command within the SRS Sub-project area, to be established by October 1, 1987; and (ii) one special agricultural extension unit in the SRBC Sub-project area, to be established by October 1, 1988. 5.14 The GOAP at present is considering the reorganization of its administrative structure, under which district administration would devolve onto elected bodies and the districts would be subdivided into Mandals, which would be smaller units than the present talukas; moreover, there may welL be some adjustment to the present district boundaries. Since it is a basic principle that the agricultural extension organization does not cut across administrative boundaries, there may well have to be some minor reorganiza- tion when the Mandal system is finally implemented. 5.15 Funding of the strengthened agricultural extension service would continue under the ongoing extension project (para 5.13). Therefore, no additional financial support would be required. 5.16 Agricultural Research. The principal function of adaptive agricul- tural research for project support would be to identify crops, cultural practices and crop sequences that enable double cropping to the maximum extent compatible with the amount and timing of water in the two project areas. 5.17 Agricultural research in the State is directed by the Andhra Pradesh Agricultural University (APAU) in Rajendranagar, near Hyderabad (para 2.22). The first National Agricultural Research Project (NARP I) has financed the establishment of zonal research stations in Nandyal and Jagtial. These stations would be the field supervising bodies of the adaptive research program. The Nardyal station has the land and soil types on which to imple- ment a relevant research program. However, the Jagtial farm does not have the space and sufficiently representative soils to fulfill the same function. Therefore it would be necessary to establish at least two small substations in the SRS area, one on vertisols (black) and one on alfisols (red), on which junior research staff, supervised and directed from Jagtial, would implement the program. 5.18 Much of the work in the proposed centers would focus on cultural practices and land preparation, i.e. would involve research and investigation of animal-drawn farm implements. Since most of the work required to serve the needs of the project areas would be applicable to most of the Rayalaseama and Telengana regions, the Agricultural Implement Research Center, adjoining the APAU at Rajendranagar, shGuld play a key role in the program. 5.19 Funding of the work envisaged would be part of the All-India Agricul- tural Research Program begun under NARP I, and continued under NARP II. Accordingly, an agreement has been reached with GOAP that it would, by December 31, 1986 cause the Andhra Pradesh Agricultural University to submit proposals to the Indian Council of Agricultural Research for the estab- lishment of suitable facilities in which to pursue an adaptive agricultural research program to serve the needs of the project and also submit it to the Bank/IDA for its comments; after taking into account these comments, finalize and implement the program.  - 29 - LandshapLng 5.20 Landshaping is an established part of irrigation development in AP. Most landforming has been for rice and therefore the compartments are shaped to zero slope. However, the need for longitudinal and cross slopes is recog- nized and has been a standard feature of work done in earlier stages of the non-rice areas in the SRS project. Farmers mostly shape their own land. The State Agro-Industries Corporation (APAIC) also play a role, although most of their work is on soil conservation in rainfed areas. Landshaping costs in the SRS scheme are about Rs 3,000/ha; to date some 75,000 ha have been shaped either by the farmers or by APAIC. Loans are available, irrespective whether farmers do their own work or engage APAIC. Refinancing is proceeding smoothly and only about 15-20Z of the farmers are not eligible for credit because of earlier defaults. 5.21 APAIC has a fleet of over 140 bulldozers and an approved program to purchase new units at the rate of 20 per annum during the Seventh 5-Year plan (1985-1990). It also fabricates and sells metal ox-drawn scraperboards that farmers require for their own landshaping. However, it does not have the capability to survey, mark out and check fcr accuracy work done by the farmers. 5.22 To ensure the availability of integrated landshaping plans, including the proper siting and alignment of field drains, the ID would prepare com- prehensive plans, in consultation with the farmers involved, to develop entire outlet commands, including landshaping required for each field. It would be essentz.& to ensure that farmers' landshaping effort is not imple- mented in a manner that would interfere with effective drainage or irrigation of adjoining fields. The staff necessary to perform this task would be provided under the project. Road Construction 5.23 Farm and feeder roads (para 3.10) would be constructed and maintained by the Roads and Buildings (R&B) Division of GOAP's Public Works Department (PWD). It has the capacity and organization to start survey and design work immediately. Work on farm roads would run concurrently with block develop- ment in both sub-project areas. 5.24 Past maintenance of canal roads has been less than satisfactory. This situation would be improved if their maintenance were to devolve on the O&H Circles of the ID (para 5.11). The task would be integrated with routine maintenance work on the cAnAls, as on the Gangetic plains. Planning; Monitoring and Evaluation; Technical Services 5.25 Strong planning and monitoring units would be required at the head- quarter level of the combined Irrigation and Command Area Department as well as on the projects. While exact details of its specific functions would be worked out by the Merger Committee (para 5.09), the present PPM Unit (para 5.02) will require considerable strengthening to fulfill its expected func- tions and responsibilities. These would include, but not be confined to: (a) all long-range planning of the irrigation sector in the State; (b) monitoring quality control activities; (c) monitoring ongoing projects; and (d) organizing and supervising special studies carried out on their behalf by consultants. The Unit, headed by a Chief Engineer, would include engineers,  - 30 - economists, sociologists, agriculturists and other professionals as necessary to perform its functions. The PPM would monitor physical and financial progress of project works, prepare reimbursement documentation to GOI and for the Bank, and ensure that tendering and contracting procedures follow agreed guidelines. Its initial task would be to conduct a benchmark survey to record present socio-economic conditions in the project area. It would also: (i) monitor system operation regarding timeliness of water deliveries, according to the prearranged timetable, and losses in transit; (ii) monitor groundwater levels, drainage conditions, the utilization of existing wells and the rate of development of new wells; (iii) collect crop data from local Revenue Offices in the CCA, irrespective whether irrigated or not; (iv) report changes in cropping patterns, crop selections, farming intensities; (v) assist in identifying factors that promote or slow down the introduc- tion of beneficial chaniges in crop and cropping pattern, thereby providing material for the adaptive agricultural research program; (vi) evaluate the impact of project works, either in-house, or by con- tracting out; (vii) commission, and supervise the execution of surveys to trace the rehabilitation of persons resettled from the Mannair Reservoir site; and (viii) prepare progress reports and the Project Completion Report. The organization of the strengthened PPM unit is shown in Chart 3. 5.26 An agreement has been reached with GOAP that it would: (i) by December 31, 1986 submit to Bank/IDA for comments a detailed list of func- tions and staff strength of the PPM Unit; and plan for M&E activities und_ the project; (ii) by June 30, 1987 complete strengthening of the PPM; and (iii) commencing January 1, 1987 and thereafter on January 1 of each year, submit project progress reports in format and frequency agreed with the Bank/IDA. 5.27 Because of the many innovative features of the proposed project, principally in the field of design (paras 3.22-3.23) and water management (paras 6.11-6.20) the ID would require the assistance of qualified consult- ants with design and operation. The function of these consultants would include, but not be confined to (i) general advice and collaboration, to translate the principles of the structured network (para 3.22) into designs and contract documents; (ii) defise operating rules, manuals and procedures to provide, as far as possible, a reliable irrigation service based on warabundi (para 6.10); (iii) to assist with studies to investigate the desirability of integrating the operation of the minor tanks in the SRS project (para 2.18) into the overall system; and (iv) to assist with inves- tigating and planning further irrigation and drainage development in the State. An agreement has been reached with GOAP tnat it would retain general consultants, starting on November 1, 1986 to assist with final design and  - 31 - documentation for constructing a structured system and to develop operating and maintenance planes and from time to time/as required.to f.vdist the PPM with its engineering preparation and planning tasks. Other consultants for X&E, rehabilita- tion, studies and other activities would be engaged as required. All consultants would be engaged in accordance with Bank guidelines. Oualitv Control 5.28 Each CE has a Quality Control Circle. While these Circles function ade- quately, none of the engineers assigned to it had special training in quality control and often does not remain long enough in the Circle to gain the requisite experience. The Quality Control Circles would be strengthened by providing special training to their engineers (para 3.14) and their monitoring tightened by the PPM Wing (para 5.25). An agreement has been reached with GOAP that it would take all action necessary to ensure quality control by its departments and agencies impler-- ing the project in respect of planning, design, construction, upgrading, oper- u-i and maintenance of the works under the project. Project Coordination 5.29 With the integration of ID and CADD, changes would be required in the constitution and function of the Command Area Development Board (para 5.07), and also the organization of the CADA Working Committee in the SRS sub-project. The Co mittee vould review progress reports by the CE in charge of the scheme area, monitor progreus of block development works and any other reports and statements pertaining to project objectives. Its conclusions would serve as a guide for future activities. 5.30 A similar Working Committee would be established for the SRBC sub-project, where the SE whose Circle is engaged in the development of the service area would be an ex-officio member. An agreement has been reached with GOAP that it would, by April 30, 1987, establish a Working Committee chaired by the Secretary CADD for the SRBC sub-project with membership and functions agreed with the Rink/IDA. VI. WATER MANAGENENT Introduction 6.01 Due to the limited surface water resources of the Telangana and Rayalaseama Regions, rice growing would severely limit the areas that can be irrigated from surface sources. Therefore, it is GOAP's policy that all nev surface irrigation projects in the two regions will be designed and operated to satisf7 the require- ments of irrigated dry crops; potential rice growers would be required to provide the additional water from their own wells. GQAP has further decided that surface supplies over and above long-term allocations (para 6.06 and 6.09) will not be made available to farmers, even as a temporary measure, during the development period of new schemes jJ. This policy is based on lessons learned in the Nagarjunasagar Project (para 1.12), where rice growing, temporarily permitted while the command area Was being developed, has proved politically difficult to stop. Farmers have taken more than their entitlement of water at the head reaches of the project to sustain their rice and therefore water reaching the tail-enders vas less than their fair share and inadequate for proper irrigated farming. JI/ During the initial development period there is a surplus of water in most schemes, as the reservoir is completed and irrigation starts before development of the service area is completed. L . _   - 32 - 6.02 Improvements in the price and yields obtainable for maize, cotton and groundnut give farmers added incentives to change to these crops. The higher degree of skill required to grow irrigatd crops other than rice is becoming more common and the trend is accelerating, particularly in the rabi season. 6.03 Localization. Irrigated areas in AP are identified as "irrigated wet" (IW) and "irrigated dry". The procedure for this identification is known as "localization". The water duty I/is 40 ha and 16 ha in areas local- ized as irrigated dry and IV respectively. In IW areas, all crops may be grown but farmers grow mostly rice. In areas designated as irrigated dry, --usually, but not always, the lighter soils,- rice and sugarcane are specifically banned because farmers near the top reaches would prevent water reaching downstream areas (para 6.01). Therefore, to ensure an equitable and reliable water supply and to introduce the changes required in SRS to serve all commandable lands to be irrigated, not only would it be necessary not to localize new IW areas, but eventually to convert all existing IW into irrigated dry lands. GOAP is aware of the problem and proposes that: (a) the new areas to be served from km 146-267, and eventually to km 284, of the Kakatiya Canal be designated as irrigated dry, with a duty of water at the chak outlet of about 1 cusec per 40 ha (to be derived exogenously from the irrigation intensity based on the average design cropping pattern and available water); (b) in presently irrigated areas, which are illegally taking more than their share of water to grow rice, to throttle the flow to the irrigated dry level as in (a) above; and (c) in the areas localized for IW, rice irrigation to continue only for a limited period, after which the outlets would be throttled to irrigated dry levels. In the interim, rice shutes would be provided, eventually to be concreted up. (d) supplies would typically be delivered intermittently (one week on-one week off) except in peak planting seasons. 6.04 To ensure that the foregoing steps are taken in a systematic manner, an agreement has been reached with GOAP that by October 1, 1986 it will introduce and thereafter apply a water management plan in the SRS scheme which would include a rotational water supply system, taking into account the need for regulating irrigated dry crops. In doing so, rice growing will not be possible using only surface water in the SRBC project and in parts of the SRS project downstream of the Mannair Dam. However, farmers would be free to grow rice if they supplement surface water from their own wells (para 6.29). 6.05 Regarding'the phasing out of rice growing on only surface supplies (para 6.03[cJ), an understanding has been reached with GOAP that they would proceed according to the following timetable: (i) June 1986/June 1987 - Preparation of Action Plan giving details for phasing out the designation of irrigated wet crops. 1/ The notional area served by a flow of I cusec.  - 33 - (ii) June 1987/June 1989 - Conversion of about 5,000 ha to irrigated dry land. (iii) June 1989/June 1991 - Conversion of about 40,000 ha to irrigated dry land. (iv) June 1991/June 1993 - Conversion of about 17,000 ha to irrigated dry land. While it may not be possible to meet the specified targets for conversion in the periods stated because of the size of the units that have to be con- verted, GOAP 's intention is to complete the process on the entire 62,000 ha by the end of June 1993. SRBC Water Balance 6.06 Supplies. The Krishna Water Disputes Tribunal Decision of 1973 has allocated to AP 800 TMC (22,640 Mm ) of the river's 75% dependable flow as measured at the AP/Karnataka border. The Decision confines planned irriga- tion development on the Krishna by the upstream riparians to be consistent with the above allocation. However, the Decision allows AP to use all excess flows until the year 2000, without creating a future right to such excess flows. This allocation is augmented within the State by regeneration from irrigation projects, estimated by the Tribunal as 11 TMC, but reduced by 5 TMC that AP (along with similar amounts from the two upstream riparians Karnataka and Maharashtra) has surrendered to augment municipal and industrial water supplies to Madras City (para 2.04). Out of its share, GOAP has allocated 19 TMC (538 Mm3) to the SRBC project. Independent assessments have confirmed the availability of this amount. 6.07 The cropping pattern expected to maximize returns to available water requires supplies to be maintained until the end of March (Table 9a). Assum- ing the Srisailam Reservoir will be operated as now planned, and given the design parameters at the head reaches of the SRBC canal (para 3.03), the assessments have confirmed that the required quantity of water can be drawn until the end of March. An agreement has been reached with GOAP that it would ensure that the SRBC project receives aL least 19 TMC (538 Y1m ) of water between June 1 and March 31 in all years when inflows at the AP/Karnataka border equal or exceed 800 TMC (22,640 Mm ); and 6.08 The Telugu Ganga Project. With the construction of the Madras canal that would deliver water to the AP/TN border, GOAP also initiated development of a distribution system which would serve lands between the Banakacherla Regulator (para 3.05) and the State border with TN. The Madras canal is designed to deliver 15 TMC in about three months to Madras and could be used to deliver water for other purposes outside this time period. Since GOAP is entitled, under the Agreement, tc water entering the State over and above 800 TMC (para 6.06), it plans to use part of this water to supplement supplies in the Pennar Basin for the Telugu Ganga Irrigation Project of about 150,000 ha. However, construction of the project has not yet been approved by GOI. Since any rabi irrigation in a proposed Telugu Ganga scheme must be available until the end of ?.arch, the system would operate on the same schedule as specified for the SRBC project (para 6.07). Therefore, it could compete with the SRBC project when all the Krishna water upstream of the AP State border had been utilized. However, based on the present rate of development irrigation works  - 34 - in the basin, full utilization is not expected for many years. In the mean- time, the efficiency of water use in old and new systems is expected to increase far above present levels. For these reasons, with the assurance requested (para 6.07) as well as the expected socio-political pressures in the SRBC area to safeguard the allocated water supply, it can be expected that the SRBC wilL in fact be provided with the amount of water required to ensure the economic viability of investments in its development. SRS Water Balance 6.09 According to the ruling of the Codavari Water Disputes Tribunal (1979) AP is entitled to all the water that reaches the Sriramasagar Dam. CWC estimate this to be 160 TMC (5648 Mm3). A further about 17 TMC is avail- able at the Mannair dam, after downstream commitments in the Mannair river are met. Thus, overall availability for the SRS scheme is about 177 TMC. Storage in the Sriramasagar Reservoir (87 TMC, 2,460 Mm3) combined with that of the Mannair Reservoir (13.4 TMC, 380 Mm") give the Kakatiya Canal much flexibility. At full development the command would require the following amounts of water: Cropping Name Area Pattern Water requirements /a ('000 ha) (Mm3) (TMC) Laxmi 6.7 sugarcane /b 116.6 4.1 Saraswati Head 17.0 no rice 130.3 4.6 km 0-146 165.0 rice /c 1,551.9 54.8 km 146-284 207.0 no rice 1,645.4 58.1 Total 395.7 - 3,444.5 121.6 Note: No allowance has been made for the Khaddam command (26,000 ha) which would continue to have its supplies met from the Khaddam reservoir, generated by the Khaddam river (para 2.08). Ia Assuming an overall water use efficiency of 50% at full development. lb Irrigated by wells in April-June. /c Requires supplementation by wells. Therefore, at flll development down to km 284 of the Kakatiya canal, about 56 TMC (1585 Mm ) would be available to serve the area beyond km 284. Alter- natively, if overall system efficiency attained would only reach about 37%, there would still be sufficient water to serve the project area, even if the 165,000 ha between km 0-146 were to receive all their supplies from surface sources. In that case, however, extending the command area beyond km 284 would not be feasible. 6.10 The 385 minor tanks within the GCA would be augmented by return flows and system recoveries induced by irrigation, thereby indirectly benefiting from the project (para 2.18). Their possible integration in the overall system would be one of the studies expected to be commissioned by the PPM Wing (para 5.27).  - 35 - System Operation and Water Management 6.11 General. Both SRS and SRBC would be managed as "utilities". The main objective would be to provide a reliable water supply to farmers. Ini- tialLy the scheme (utility) will supply the water down to the 40 ha outlet, according to predetermined schedules arranged in consultation with farmers' representatives. Below the outlet distribution would be organized by the Outlet Committees on warabundi schedules (para 6.18). It is expected that water users organizations would undertake increasing responsibility over time; they may ultimately buy water in bulk from the utility for sale to farmers, collect water charges, etc. The structured system proposed (para 3.22) would provide an on-off supply to the head of each distributary. Farmers would be free to grow crops of their choice, but the scheduling of surface water would be such as not to cover the full irrigation requirements of paddy and sugarcane: to grow these crops, they must meet the gap in their water demand from their own weLls (para 6.31). There will be no canal sup- plies in April-May to sustain sugarcane and canal supplies are likely to be insufficient for rice at all times; certainly so in the rabi. The irrigation system would however provide full supplementary kharif irrigation for a typical mix of irrigated dry crops under drought conditions, and an average irrigation intensity of about 70Z for rabi crops. Since GOAP's ID has no previous experience with structured system operation and design, provisions are made in the project costs to retain consultants, supervised by the PPM Wing, to assist ID with the preparation of the necessary designs, the for- mulation of operating rules and operation of structured systems (para 5.27). 6.12 Water Duty. The design duty for the system wouLd be derived from the peak requirements of the design cropping pattern, with suitable allowances for seasonal rainfall and system losses. The allocation of water would be uniform over the whole CCA in both SRBC and SRS except for presently authorized IW areas in SRS where temporary arrangements would be made for a higher water allocation to allow for paddy growing during the interim change-over period (para 6.11). Only peak water requirements would be met by continuous deliveries; lower than peak requirements, by intermittent deliveries. 6.13 Irrigation Schedule - Rabi. In the rabi season, when there is usually no rain, the required irrigation frequency depends on crop consump- tive use and the moisture-holding capacity of the soil. For almost all project soils a 14-day cycle would be suitable. By September 1, the quantity of water in the reservoirs is approximately known; by October 1, the estimates are final and can only increase through late storms. Therefore, the number of waterings for the rabi season is predictable. In consultation with the Agricultural Department and the farmers' representatives, the ID would promulgate, by ear'y October, an irrigation schedule for the whole rabi which would meet the average demands. Every hectare of the CCA would be allocated the same volume of water and the number of irrigation turns and their timing would be made known to farmers before the beginning of the season. An indicative flow schedule of the required surface supplies are shown in Chart 8. 6.14 The incipient and actual damage and inequities arising out of localizing part of the previously developed SRS system as IW require to be remedied. This would be achieved by introducing, as rapidly as it is expedient, a water management plan which would include a rotational water delivery system, based on supplying irrigated dry crops over the entire area  - 36 - (para 6.04). The objective would be that by kharif 1992 continuous flows would cease in all distributaries and minors, forcing farmers who want to grow rice to resort to using wells to meet both water requirements. In addition, an agreement has been reached with COAP that it would, not later than two years of the expected date in which water is to be delivered to the SRBC scheme area, prepare and furnish to the Bank/IDA, for its review and comments a detailed water management plan, which would include the rotational water supply system, and thereafter apply such plan when water is delivered to such scheme. 6.15 To ensure that farmers can make the best use of irrigation water, an agreement has been reached with COAP that it would: (a) each year, by September 1, starting in 1987, advise farmers of likely timings and number of turns when water would be available in their outlet between October and March; (b) finalize the schedule by October 1 each year, starting in 1987, and appropriately advise farmers; and (c) inform farmers of any predictable deviation from the published schedule at least one month in advance. 6.16 Irrigation Schedule - Kharif. Under the rainfall pattern of the project areas the highest return per unit of irrigation water is during monsoon drought or too long interval between rainstorms. In this case, irrigation prevents substantial yieLd reductions and crop failures. There- fore, if over 50% of the farmers request irrigation water during the kharif, it would be provided in their area if operationally feasible. 6.17 Carry -over storage, to enable an earlier start of the cropping season, or a fixed-period start irrespective of the actual commencement of the monsoon, would not be practical in SRS and SRBC schemes and it would be impossible in the SRBC project without severely interfering with power gener- ation. In SRS, pre-season planting implies harvesting all the planned crops except cotton in heavy rainfall, which would result in higher production costs and crop losses. In cases where early planting is desirable, as with some varieties of cotton, farmers will have to rely on their wells. 6.18 Water Allocation within Outlet Command will be based in the warabundi system, adapted to the conditions in the project areas. COAP has had limited success in applying the warabundi system of water allocation in the SRS scheme, because when introduced, water supplies were excessive. Now that more of the area is developed, and the expected further rapid development of the service area under the project, the lessons learned would facilitate successful introduction. The system, which has been practised in North-West India since the introduction of controlled irrigation, rotates the full flow of the chak outlet to each farmer according to a predetermined timetable each week. The quantity a farmer draws, i.e. the time period during which the full flow is available to him, is proportional to the size of his landhold- ing. As long as a farmer pays his water charges and does not use water for unauthorized purposes, he is entitled to draw water whenever the outlet  - 37 - discharges at the time of his weekly turn 1/. During any unscheduled shutdown of the system, such as unexpected rainfall or canal breaches, the "Warabundi Clock" continues to run. When supply resumes, it will be drawn by the farmer whose turn is scheduled for the day and hour of the resumption. 6.19 Warabundi is a rationing system to discribute limited irrigation supplies equitably. When water is abundant, as was the case when an attempt was made to introduce warabundi scheduling in Pochampad, strict rationing is not seem as necessary by the farmers and the plan may fail. However, even at that stage equitable and predetermined supply at the outlet is essential. As more of the potential command is developed, the full enforcement of warabundi scheduling becomes necessary for an equitable and reliable service. This stage has now been reached by the SRS scheme and is expected to be reached by the SRBC scheme before 1992. 6.20 Distribution Management. Three levels of distribution management would be required: (a) Main system comprising the dams, main and branch canals, would be reguLated 2/ to provide a supply according to a pre-scheduled demand to the distributary heads. In the rabi the demand would be according to a predetermined schedule, which would only be changed if unex- pected storms make an irrigation turn unnecessary. In the kharif the demand will fluctuate according to rainfall. Adequate decision procedures to determine kharif supplies would be developed. Deci- sions would be taken by SE Circle (Canal and Dam Operations) (para 5.11) operating from a control center. A microcomputer wil. be used to pLan scheduling and to monitor continuously the regulation of the main canal. A transient flow program will be developed to minimize wastage during either pre-planned or emergency regulation events. (b) Distribution system, including the distributaries and minors, will be running full on or off, with proportional distribution flumes or adjustable proportionate modules (APMs) at each node. Since no gate operation is involved, the duties of the staff would be limited to monitoring physical failures and unauthorized interference. Opera- tion would be supervised by the EEs' Division. (c) Below the outlet, at the chak level. 6.21 Farmers' Organizations. The system below the chak outlet would be managed by farmers. At present, the CADA in the Sas area is charged with organizing farmers in the chaks and having Outlet Comittecs (OC) elected. 6.22 The Act stipulates that OCs be elected annually. Their election is a matter of organization and therefore should be delegated to an authority 1/ Since the channels run for 24 hours, it is usual once a year change the timing (e.g. farmers who draw water between 6 a.m. to 6 p.m. change with those who dres it between 6 p.m. to 6 a.m). 2/ Sufficient control structures, as well as computation, communication and measuring equipment would be provided to ensure that the manual regula- tion facilities are able to deliver controlled volumes of water. * U. -  - 38 - normally dealing with electoral matters. Also, whether all members should be elected every year, or whether the entire committee should be elected for four or five-year term, requires consideration. 6.23 The OCs would organize the construction of field channels leading from the 8 ha outlet to each field. This section would consist of, typi- cally, about 300 m long channels; the parts of concern to all farmers within the chak are the ones serving more than one farmer 1/. Layout for this system would be provided by ID as part of the overall chak development plan; its execution would be part of the OCs' brief. 6.24 In the already operating portions of the SRS Project, the organiza- tion outlined requires to be set up as soon as practicable; the enabling Act has been on the Statute Books since March 1984. Therefore, an agreement was reached with GOAP that it would, in the already irrigated areas of the SRS command, arrange for the establishment of Outlet Committees for all outlets on or before May 31, 1987. 6.25 In areas to be brought under irrigation in the scheme areas, an agreement has been reached with GOAP that Outlet Committees would be estab- lished, and their Chairmen suitably trained, not more than six months before irrigation is due to start. 6.26 In addition to the Outlet Committees, it would be desirable to form committees dealing with larger areas that the field channel command, commit- tees to deal with minor canal and distributary commands, with whom ID offi- cials in charge of larger units could discuss, from time to time, issues affecting these larger areas. The members of these committees would come from the OCs: typically, OC Chairmen would constitute a Minor Committee (MC) and MC Chairmen a Distributary Committee (DC). The broad function of these two higher-level committees would be: MCs - to encure that outlets are not tampered and thereby either deprive OCs of water or dump excesses on them; - to represent the OC group when water allocations for the season are arranged; and - to request water in the kharif in case of draught; DCs - to advise management of the distributary command and to initiate remedial actions that would affect all farmers in the area. 6.27 While in the long run, bearing in mind the ultimate aim of managing the projects as a utility (para 6.11), these committees will come to fulfill a major function, their role for the next few years may be fulfilled by ID engineers. Therefore, an understanding has been reached with GOAP that, after enough expe-?ience has been gained with the OCs, they will review the position and initiate the necessary action to establish MCs and eventually DCs. 11 If a holding is 8 ha, this provision would not apply; the length of shared channels changes inversely with holding sizes.  - 39 - Maintenance 6.28 Three kinds of maintenance activities have to be considered: (a) between construction of a new scheme and its handing over to the operat- ing divisions; (b) routine maintenance; and (c) special maintenance, made necessary by storms, breaches, accidents in operation or major repairs that normally do not lend themselves to routine operations. COAP's budget for routine maintenance has averaged about Rs 50/ha, of which only Rs 12/ha were available for contracts, the rest being absorbed in establishment costs. This sum has proven inadequate, as shown by the r-ondition of completed works in SRS and other operating systems; it needs to be increased. 6.29 Concerning new works constructed under the project, maintenance during the first two years of their operation would be part of construction costs and in the hands of the construction Circles. This commissioning period would give adequate time for an in-depth examination and assessment of realistic O&M costs to cover routine and expected special maintenance. Accordingly, an agreement has been reached with GOAP that it would (i) by March 31, 1987, initiate, under terms of reference agreed with Bank/IDA, a detailed study to review existing rules, procedures and financial require- ments for operation and maintenance and collection of water charges from the irrigation systems; (ii) submit to the Bank/IDA, by March 31, 1988, a report summarizing the results and recommendations of the study for its review and comments; and (iii) taking into account Bank/IDA comments, implement such recommendations. Annex 4 provides the terms of reference for the study. 6.30 The maintenance of the distribution systems below the outlet would be the responsibility of the OC (para 6.24) which would share the task propor- tionately among farmers in the chak. Maintenance works such as vegetation control, desilting and minor repair to embankments can be readily provided by farmers. Repair to structures, which would require cement and sand, is expected to be infrequent. To ensure proper maintenance of field channels and their structures, an agreement has been reached with GOAP that it would: (a) cause the outlet committees to maintain the field channels and assist the process by promptly providing, for due compensation, the neces- sary building materials as required; and (b) where the outlet committees fail to maintain the field channels, ensure execution of the work by Irrigation Department and recover the cost from the farmers. Groundwater Development 6.31 Shallow dug wells are a traditional feature of Deccan agriculture. There are about 5,200 such wells in the areas to be rought under irrigation. In the project areas, they would recycle over 300 Mm of water annually from the shallow aquifers, at present recharged mainly by rainfall. There is a potential for further about 4,030 wells in SRS and for about 2,000 in SRBC, to lift a total of about 420 Mm . Under project conditions, groundwater is likely to be utilized for two purposes, which would almost certainly overlap: (a) to su?port rice growing under the projected water management regime (para 6.04) under which the canals discharge intermittently (to maintain standing water in the field, the compartment requires replenishment about every three or four days);  - 40 - (b) to grow crops that require frequent light waterings (e.g. vegetables) and perennial crops that require irrigation when the system would be normall-y closed (e.g. sugarcane and orchards). Present financial incentives to develop wells are adequate and GOAP intends to continue them. However. excessive supplies of surface water that have been available in the partially developed SRS project have not only slowed dowr. well development, but have led to the abandonment of many wells. Once the service area is more fully developed, and surface supplies better con- trolled, the momentum to construct wells is expected to resume. The agricul- tural extension and research services would further encourage the trend by developing and advocating profitable cropping systems relying on wells to supplement surface water. The yield of the wells, and the number that can be developed in the command areas, would increase substantially once irrigation is established and recharge of the aquifer increased. VII. AGRICULTURAL PRODUCTION, FARM INCOMES AND COST RECOVERY A. Agricultural Production Srisailam 7.01 Present cropping intensity in the project area is about 106%. The cropping pattern is largely subsistence oriented: foodgrains and legumes for home consumption, groundnuts and short-staple cotton as cash crops. Without irrigation, neither the present crop mix, nor cropping intensities are likely to change much; however, better extension services, armed with the appropriate messages, increasing adoption of improved farming practices and water conservation techniques may enable yields to increase somewhat with or without the project. With irrigation, production would become more stable, intensified and would shift to higher value crops; cropping intensity would reach about 135% (Chart 9). The largest output increments are expected from groundnuts and from changing to medium-long staple cotton instead of short-staple cotton (Table 10). Sriramssagar 7.02 In parts of the project area where there is irrigation, overall cropping intensity is about 120%. The total irrigated area has never exceeded 106,300 ha, implying an irrigation intensity of only 64X (1983/84 season), due to deficiencies in the water distribution system. About 22% of the area actually irrigated grow rice. In parts of the area where at present there is no irrigation, the cropping pattern, as in Srisailam, is subsistence oriented. Cropping intensity is about 108Z; double cropping is sustained by wells. With the project, cropping intensity is expected to reach 150% in the previously irrigated areas and 135Z where there has been no previous irriga- tion. The ultimate expected cropping pattern, and the way it fits into the  - 41 - cropping and farm management calendar is shown in Charts 10-12. Full development is expected to take about 20 years. The principal cashcrops grown would be groundnuts and medium-staple cotton. Ultimately, both areas would be cropped at an intensity of not less than 150X because of conjunctive groundwater use, made largely possible by the higher aquifer recharge rates due to irrigation (para 6.31). Production increments of maize, groundnuts and cotton are expected to account for 75Z of the incremental value of production. long run, be dictated by international market prices. Produc- tion will change from short-staple to medium-to-long-staple cotton, which Lesponds better to irrigation and is internationally more readily marketable. Market Prospects and Prices 7.03 Incremental production resulting from the project will be a small percentage of AP's total agricultural production and therefore would be readily handled by existing marketing channels. Prices of grundnuts and other oilseeds are likely to be kept steady by India's high income elasticity of demand for edible oils, and by GOI's continuing efforts to reduce its Large vegetabLe oil imports. Cotton prices will, in the long run, be dic- tated by international market prices. Production will change from short-staple to medium-to-long-staple cotton, which responds better to irrigation and is internationally more readily marketable. Far. Incomes 7.04 Representative farm models have been used to calculate the project's impact on farm incomes. Prevailing financial prices were used to value outputs and inputs. The results are detailed in Table 12 and summarized below: Net Farm Income /a Future Without Future With Increase due Farm Model Present Project to Proiect ---------------- (s)--------------- --- (Z) Srisailam (3.6 ha) 4,680 4,980 36,880 641 Sriramasagar (i) Presently Irrigated Area (2.3 ha) 7,340 11,570 21,910 89 (ii) New Irrigated Area (2.3 ha) 4,250 5,570 17,120 207 /a Get return to family labor and farm capital investment before tax and water charge payments. B. Cost Recovery General 7.05 Direct means of cost recovery like water charges have not been used to their full extent: their present levels are lov and procedures for their assessment inadequate. A Betterment Levy for recovery of capital costs has been sucessfully challenged in the courts and its collection suspended. The most important source of cost recovery is therefore revenue from sales taxes on incremental agricultural production. To improve the current situation, and in particular to increase the share of direct charges in cost recovery, a  - 42 - high-level standing Water Charge Committee would be established under the project to review regularly cost recovery policy and propose adjustments. Present Situation 7.06 Water Charges. Water charges are generally perceived by farmers as taxes and not as charges for irrigation service. Revenue from water charges accounts for Less than 1X of total state income and only about 602 of annual expenditures on OS". Depending on the type of irrigation provided, two basic water rates are currently charged: Rs 75 per ha cropped for IW (i.e. paddy) and Rs 60/ha for irrigated dry (i.e. upland) crops (para 6.03). Concessional rates are .xarged during the first four years, beginning with 20X of the full assessment after the initial release of water, followed by equal annual increases to the full rate in the fifth year 11. The present level of water charges in SRS irrigated areas amounts to about Rs 60/ha of net CCA which is about 12 of the gross value of production. It has been at this level since 1974. However, on August 1, 1985 GOAP published in the Government Gazette its intention to double water charges beginning with the 1986-87 crop season. The statutory time allowed to raise objections has passed and therefore an agreement has been reached with GOAP that it will issue orders not later than May 30, 1986 to revise its water charges and recover the full cost of opera- tion and maintenance and a reasonable portion of the capital costs, taking into account farmers' repayment capacity. 7.07 Assessment of Water Charges. Assessment of water charges and the exchange of information between the ID and the Revenue Department (RD) needs to be substantially improved (para 7.12). The assessment and collection of water and water related charges is the responsibility of the RD. Water charge assessment requires accurate verification of areas actually irrigated. Detailed records of land revenue and water charge assessments are maintained by the Village Revenue Officer (patwari). Provisional assessment is made for every field receiving irrigation (source-wise) during October/November, and is finalized in Kay/June. The assessments are aggregated by revenue juris- dictions for the entire State. Au important part of the records are detailed notes describing the performance of an irrigation system below the outlet. The reason why every potentially irrigable field does not receive irrigation or receives an inadequate service is recorded and used to deal with applica- tions for remission of land revenue. However, revenue remission proceedings are not communicated to ID; nor are they consulted. Exchange of information between the ID and RD is limited and irregular: ID does not provide RD with information on irrigated areas as identified on project maps, and RD does not inform ID on irrigated areas as entered in the village records. Exacerbated by a shortage of field staff in both Departments, this lack of communication permits farmers to conceal irrigated areas, evade payments, misuse water and own more irrigated land than permitted under the Land Ceiling Act. 7.08 Betterment Levy. The GOAP Irrigation Act provides for collection of a betterment levy for recovery of capital costs from farmers benefiting from public irrigation works. The betterment levy is defined as half the increase in land value due to an individual project (calculated using before-and-after land values). On the basis of present land prices in SRS, the levy should amount to about Rs 7,000/ha. The repayment period and the interest to be 1/ A 37% land cess surcharge on the relevant basic rate which used to form part of the assessment has been recently abandoned.  - 43 - charged on the outstanding amount are determined by an executive order, published in the official gazette. However, the assessment of the betterment levy has been successfully challenged in the courts. Therefore, GOAP sub- stituted it in 1976 with a special one-time land tax of Rs 500/ha, recoverable in equal installments (without interest) over a 10-year period. As pres.,stly constituted, the special land tax would recover less than 1% of the project's capital costs. 7.09 Indirect Taxation. GOAP's most important means of cost recovery from irrigation projects is currently indirect taxation. The State government relies on revenue from sales and commodity taxes, users' fees for support services, roads and bridges ("octroi" charges), fees in the regulated markets and commodities taken to the larger settlements for disposal. Indirect taxation provides 90Z of public revenues from agriculture and is steadily growing in real terms. The most important source is the sales tax on agricultural producis which averages about 4Z. Moreover, according te recent studies of the Indian tax system 1/ effective indirect taxation on inp. q indicates actual tax rates on private consumption expenditures of about 6% and 4% for above and below poverty-level rural households. 7.10 Farmers' Ability to Pay Water Charges. For the assessment of farmers' ability to pay for irrigation service, the project rent 2/ has been estimated. At full development in about 20 years, the project rent would average about Rs 7,800/ha in SRBC and about Rs 4,200/ha in SRS (Table 13). Another indicator of the net benefits resulting from irrigation is the general increase in land prices as a result of irrigation development. For example, prices range from Rs 10,000/ha to Rs 15,000/ha for dry land and Rs 20,000(ha to Rs 40,000/ha for irrigated land in SRS. 71.11 Capital and OEM Costs. The present value of the cost of irrigation infrastructure in the SRBC scheme, including on-farm development works but excluding roads, would amount to Rs 19,300/ha of CCA. Annual capital recovery (discounted at 12% interest over 50 years) would amount to Rs 2,320/ha. The cost of irrigation infrastructure in the SRS scheme is estimated at Rs 8,700/ha of CCA. This amount includes part of the Sriramasagar Dam (75% allocated to irrigation), the full cost of the Mannair Dam and the Kakatiya Canal. The sunk costs of these works have been adjusted to 1985/86 values. The annual capital recovery cost would amount to Rs 1,050/ha. A preliminary estimate of Rs 150/ha of net CCA for routine O&h requirements has been used; actual annual OK cost would be subject to a major review under the project (para 6.29). 7.12 Analysis of Present Situation. The present level of recovery of water and water related charges per ha of net CCA would be as follows: 1/ E. Ahmad and N. Stern, "Effective Taxes and Tax Reforms in India," Development Economic Research Center, University of Warwick, Discussion Paper No. 25, January 1983. 2/ Project rent is the incremental net benefits (after appropriate allowan- ces for management and risk) due to the project.  - 44 - Item Present Production Level Full Development Level SRBC SBS /a SRBC SRS