Final Paper - Executive Summary
November 2000

This report is the product of one of the case studies being undertaken by the World Commission on Dams (WCD), world-wide with a common methodology and approach that seeks to inform the Commission on development effectiveness and on a range of issues associated with the planning, design, construction, operation and decommissioning of large dams. It concerns the Ceyhan Aslantas Project (CAP), a major initiative of the Turkish Government initiated in the mid 1960s and completed in 1985. The project aimed to irrigate 97 000ha of land in the lower Ceyhan basin, generate power and reduce the occurrence of floods.

The Ceyhan River Basin is located in the eastern Mediterranean region of Turkey and drains into the Mediterranean Sea in the South. The Ceyhan river rises in mountains of 2 200 m to the north of the regional centre of Kahramanmaras, and flows southwest to the Mediterranean sea near Adana. The basin covers 20 670 km² and includes mountains, three major provinces (Kahramanmaras, Osmaniye and Adana), major towns of Ceyhan, Kadirli, Yumurtalik, Erzin and Dortyol and a potential of up to 590 000 ha of irrigable land.^[1]

The development of the basin was conceived initially in the 1966 basin-planning document prepared by IECO International Engineering Co. Inc and the CAP was one of the projects identified at this time. Four other major dams on the main Ceyhan River: Menzelet, Kilavuzlu, Sir and Berke dams have also been constructed. The Menzelet and Sir dams interact directly with Aslantas and were constructed subsequently, and were completed in 1991, while the Kilavuzlu and Berke dams are currently under construction. There are 12 more dams existing, planned or under construction on Ceyhan's tributaries.

To the knowledge of the study team, there existed no other alternate formal plans for irrigating the Lower Ceyhan Plains in the event the focal dams were not to be built. There was no alternative to the flood control system: dam, dyke and river improvement at the project time. In the design, no passage was allowed for fish. An assessment of the costs in providing alternative power sources was made during the World Bank Appraisal process. The options considered as possible competitors of the project were thermal plants with various capacities and sizes. These plants would run on fuel oil. Analysis revealed that the electric energy thus supplied would be twice as expensive as that provided from CAP.

CAP was initially conceived in 1966 during the basin planning process undertaken by IECO and formed part of a larger development project. For the purposes of this study however, the component of the project approved by the Government with financial support from the World Bank project (SAR 1973) is used as a baseline for comparison with actual outcomes. CAP therefore provides stage one of a phased process to deliver 97,000 ha of irrigation, out of a planned total of around 140,000 ha envisaged in the IECO basin plan.

Predicted versus actual costs, impacts and benefits and unexpected outcomes

The purpose of the project was to provide irrigation, flood control and electrical power and cater to pressing needs of agricultural producers, flood victims and to alleviate power shortages, which were becoming more and more frequent. Although not specifically stated in those terms, the dam project consisted of seven main categories of activities or components as listed below.

Aslantas Dam was proposed in the feasibility study prepared by IECO as a 93m high earthfill dam with 2 250 MCM maximum storage reservoir. During detailed planning, the crest elevation of the dam was reduced from 171m above sea level to 157m to save the unique historical site at the Karatepe and Haruniye thermal facilities from flooding. The final construction is an earthfill structure 78m high from the riverbed with radial gated chute type spillway and with two diversion tunnels. The earthfill dam height from the bedrock is 98m, the head for power generation is 59 meters (average) and the live storage for power and irrigation is 680 MCM. The reservoir has 1 190MCM gross storage capacity at maximum normal operation level of 146m above sea level, plus flood storage capacity of 650MCM. The upper 9 m of the water level at the dam is reserved for flood control.

Construction was planned to take 7 years but took 10. The irrigation component was 4 years later than proposed. An unusually high number of unexpected circumstances of various characters (technical, administrative, and political) delayed the working progress. Modification of the project infrastructure during the construction of the dam was one of the technical reasons for the delay. Additional work was required, increasing part of the volume of works. Due to above modifications, the quantity of works increased as follows: open excavation (+61%), underground excavation (+32%), riprap (+56%) and concrete volume (+56%).

Some of these delays could have been minimised by improved supervision and quality control from the government agencies but these agencies were hampered by a shortage of qualified local supervisory staff. The most important factor affecting the project activities was the nation-wide economic crisis in 1979 and early 1980, which induced a very severe shortage of fuel and construction materials^[2], as well as power breakdowns resulting in more than one year's delay. Construction began in 1975 and the hydropower plant began generating in 1984, three years later than scheduled. In 1973 prices, the project was appraised to cost US$327 million in 1972 (552 million in 1998 US$ prices). The final cost was $884.9 million in 1998 dollars.

Except for equipment, buildings and land acquisition costs, all project cost components were underestimated. The actual project costs were 1.7 times the estimated for dam and power plant civil works, and nearly 1.4 times for irrigation, drainage, river dykes and about 1.2 times for on-farm development, and engineering and administration costs. Some of the increase in costs can be attributed to raising of the coffer dam elevation, conversion of the smaller diversion tunnel to a bottom outlet for the dam, lengthening of diversion tunnels, breakdown of spillway gate, and construction of two more tunnels in the left abutment.

The region where the Aslantas dam is located is a second-degree earthquake zone and is considered subject to high shocks. All the necessary investigations were made and the dam was planned taking into consideration the earthquake risk of the region. The Adana earthquake that happened on 27 June 1998, was 6.3 Richter Intensity. DSI experts, after examining the site have reported that no damage was observed to the dam and its associated infrastructure.

Hydro-electric power plant Power facilities proposed by IECO in 1966 included a 85MW installed capacity consisting of two 42.5 MW units to generate 348 GWh per year. The 1973 SAR envisaged a 138 MW installed capacity, consisting of 3 units, with 500 GWh average annual generation capacity and this was built as planned.

Actual annual average hydropower generation is extremely variable and ranges from 60%-200% of predicted, according to in-flows to the dam. Based on the past 15 years generation, TEAS (Turkish State Electricity Utility) generated an average of 652.7 GWh per year, 31% more than the predicted value of 500 GWh.

Annual operation and maintenance costs were estimated at 432,000 US$ in 1985 (in 1998 US$) and these rose to $1.7 million in 1998. This was due to escalation, renewal investments and the high costs of repair for some system components.

There was no internal cost recovery planned for the power component of the project, and the capital costs and loans from the World Bank were repaid directly from the national treasury.

The flood protection component comprises Ceyhan River dykes, which were raised, reinforced and extended and flood storage capacity of the reservoir. At appraisal, dykes were estimated for a total length of some 118km considering both right and left banks, to provide flood protection for 59km for the river length These dykes were constructed on both banks of the river at a cost of $2.4 million

The proposed project infrastructure planned to provide full flood protection to 35,000 ha and was implemented as plannedhowever, the incidence of floods in the plain arising from flood peaks in the tributaries of Ceyhan River downstream of the dam still occurred. To cope with this unexpected problem, drainage pump stations and outlet gates were constructed subsequent to CAP. These six stations provided additional flood control to 5 096ha that was not predicted in the SAR.

There was no alternative considered to the flood control system of the dam, dykes and river improvement at the project time. After the construction of the dam, two important floods occurred, in addition to the one due to spillway gate failure. The main reason for these floods was the storm rain that fell directly on the Ceyhan plain sub-basin, rather than upstream in the catchment. According to a study carried out in 1997, the Aslantas dam was found to be safe for an overtopping failure in the presence of a probable maximum flood (PMF). The cost of flood control is borne by the Government and is not recovered.

Irrigation and Drainage Works include the Cevdetiye regulating weir that was constructed in 1971 by DSI and equipped with radial gates under the project. It serves the right and left banks of the Ceyhan River via concrete lined conveyance canals. On both banks of the river the project aimed to provide irrigation for a net area 97 000ha comprising 88 800ha gravity irrigation and 8 200ha pumped irrigation). The works completed under the project were:

The major revision in the project implementation was the elimination of the 8 200ha pumping of a high hilly area, due to high-energy costs at the time. 2 000ha of soil reclamation predicted in the project was not also realised.

The study shows that the CAP developed 92,880ha of the planned area in 1985 but this reduced to 84,00 by 1997 as some land was lost to a motorway construction and industrialisation. However, the net irrigation realised is 84 000ha (95% of planned). In line with the IECO Basin Plan the storage capacity of the dam and main canals diverting water from Cevdetiye weir were designed for a larger irrigation area than that equipped under the CAP project. The size of the canals and their combined capacity was designed to service a gross irrigation area close to 147 000ha, and realisation of this objective is currently being pursued in phases under separate financing. DSI has prepared and implemented other irrigation projects to exploit the full irrigation potential of the Aslantas dam. Subsequent investment projects, which were also referred to, but not financed in SAR, include CAP Phase II and Phase II Irrigation Projects comprising Yumurtalik and Dortyol Plains for a total gross area of 32 000ha.

The irrigation system suffers from drainage inefficiency. The on-farm drainage area serves 34% of the project area. Water used in unplanned irrigation is negligible at the moment but it is drawing increasing attention in recent years. The overall water use efficiency of the project is estimated to be 43% compared to the planned efficiency of 60%.

In terms of production, actual cropping intensity is 34% more than that predicted. However, due to significant change in the cropping pattern, production value has remained less than predicted. Although a 250% increase was foreseen in the value of output, project benefits increased only by 196%. While cotton, wheat and groundnut yields failed to reach the predicted values, increases in maize and watermelon yields are 200% and 50% more, respectively. There was a significant change in cropping pattern due to the Second Crop Project implemented later. As an outcome of this project production of soybean in the project area contributed 100% of the total soybean production of Adana and Osmaniye provinces in 1996. Cropping patterns are affected by declining seasonal labour forces and the development of the South East (GAP) Project that grows some crops more competitively (eg cotton).

Changes in net farm incomes were also overestimated by predicting an increase of more than 300%. The actual increase on project completion was between 3% and 78% in all four land size groups (5ha, 10ha, 20ha and 50ha), with the smallest increase in the smallest land size group. A preliminary analysis indicates that the predicted 14 000 additional man-years of labour were not created. Farm labour shows a 7% decline with the pre-project situation, in contrast to the 144% predicted increase in the SAR (1973). Rural-urban migration reduced seasonal work forces and increasing mechanisation form the backdrop to this situation. The scheme may however, have saved some farm jobs, given the general trend of-farm rationalisation as farming practices intensified nationwide.

Other important assumptions concerning structural changes (eg land reform), and institutional strengthening (eg increasing the capacity of the personnel) were not realised and affected the performance of the project. During the planning of CAP, on-farm development and farmer training services, which are two important factors for the sustainability and performance of such projects, were assumed to continue after project completion. However, TOPRAKSU was discontinued after 1985 and FTES was dissolved in 1983. Thus the sustainability of the project and the sectors assumed to benefit from the project were partially negatively affected.

Cost recovery of the project has shown rather a poor performance. However the unanticipated transfer of operation and maintenance for the irrigation schemes to Water User Associations has resulted in 70% recovery of annual O&M costs, while new central taxes on agricultural produce have made a partial improvement on the overall cost recovery.

Benefits, costs, and impacts for other major areas of predicted benefits, such as recreation, environment, regional and national development, are hard to identify due to lack of quantified predictions and monitoring data to access the actualisation.

The Land Acquisition and resettlement component was implemented by DSI and aimed at financing land acquisitions for areas flooded by the lake and the construction of irrigation infrastructure on private property. It was exclusively funded by the Government and implemented under national laws of the day. Planning in Turkey is undertaken centrally and there is only restricted local participation in the planning process.

The villages flooded by Aslantas are relatively homogeneous ethnically and originate from nomadic or transient Turkish groups resettled from Thrace, Bulgaria and the Caucasus earlier this century. In this sense the villages do not have a long historical association with the area.

Expropriation and compensation were realised for 1 000 families (app. 5 000 persons) who were displaced. While 925 families preferred to receive direct financial compensation and move on their own, 75 of the displaced families asked for rural resettlement and rejected to take their compensation. After investigation by GDRS, 47 of them were accepted as resettlers, and although all of them asked for rural resettlement, 12 received urban resettlement. A total of an estimated 320 people in these families were resettled and there is no record of conflict between host and resettled communities. For some groups, resettlement only occurred in 1990, some 7 years after flooding. Consequently, 953 people received their compensation and moved on their own.

Thirty-six settlement units in 18 villages were totally flooded (houses and land). Landowners are required to declare land value for taxation purposes every 5 years and the 1971 declaration was used as a basis for compensation. Farmers have stated that they declared lower values to avoid taxation and the compensation paid was therefore around half of the true land value. The compensation scheme paid funds directly to displaced people. The total cost for land acquisition was US$38 million for 5,781ha.

Some of the consequences of resettlement/compensation for local groups included the sudden availability of large sums of cash in communities not used to handling such funds. Homesickness, loss of cultural roots and customs have been reported for elder generation rural people resettled in urban environments yet the second generation has access to education and urban benefits.

The SAR (1973) predicted that displaced people would find jobs on the project. Some 70 people from flooded villages were employed in construction.

Concerning environmental issues, no environmental impact assessment study was carried out at the time of the project planning. Only the possibility of increase in malaria and potential for salinisation of the river near river mouth were mentioned in the SAR, although the IECO report also mentioned the potential negative impacts on migratory eels. There currently is a minor incidence of malaria in the Adana region although numbers of cases declined significantly from 69 000 in 1977 to less than a thousand cases per year in the 1990s. Eels have declined considerably since project implementation although this may also be due to pollution in the Gulf of Iskenderun. Salinisation problem in the river mouth is not considered an issue, although the delta is no longer accreting as fast due to sediment retention in the reservoirs.

Few environmental studies and information are available at present and a pre-project baseline does not exist. It is therefore hard to reconstruct the environmental changes that have occurred. From limited information resources and observations made by the experts, no serious negative impacts are reported.

The SAR (1973) predicts improved access to the Hittite stones in Karatepe National Park and also the submergence of a Crusader Castle ("one of many in the region"). The value of the castle has however, been contested. The historical thermal springs were also saved from flooding by decreasing the crest elevation of the dam.

What were the unexpected costs, benefits and impacts?

With a shift in the state policies to industrialisation, import substitution and export orientation, during the planning, implementation and later periods of CAP, the population structure has significantly changed, and a rapid urbanisation all over the country has been realised. The CAP region also affected this development and many of the people have left their lands. Although the SAR assumed that there would be land reform, which would improve the land distribution, as well as increases in incomes of the farmers, land reform was not realised, thus causing further division of land due to inheritance. Generally, the small size of farm holdings has reduced the potential efficiency of farming especially for field crops, fruit and nut crops possible from efficiencies of scale and reduced opportunities for mechanisation and other on-farm investments that are associated with larger scale operations.

Due to rapid urban population growth, there is significant demand for potable water from Aslantas dam. Osmaniye province is preparing projects to submit to DSI for water supply from Aslantas and this was not anticipated at the time. Equally, urban growth upstream has led to some industrial and domestic pollution that may lead to low quality water being fed into the irrigation system.

The significant decrease in cotton production was also unexpected. Increasing input costs of cotton in CAP area, as well as emerging GAP region with a better comparative advantage in cotton production were the main factors of this decline. Introduction of soybean was another unexpected outcome of the project. After the introduction of second cropping in the region, there was an unexpected significant shift from the predicted cropping pattern.

Fisheries production from the reservoir is significantly less than the potential projected in the Bond Plan. Commercial operations have not been successful and current fisheries are unlicensed. There are 68 households directly benefiting from commercial fishing and/or provides additional income besides their farming activities. There are 34 boats fishing in the lake without licence. Eel migration (not considered in SAR) has been disrupted and the eel population has almost disappeared in the reservoir and upstream.

Other unexpected effects of the project such as changes in habitats, river ecosystem, etc are within the category of environment. However, there is limited data to assess them in detail. Local farmers report occurrence of cutaneous leishmaniasis since 1987, a disease transmitted by sandflies.

What was the distribution of costs, benefits and impacts?

Distribution of costs and impacts are examined by distinguishing between reservoir, downstream and upstream areas. Within each area the impacts are distinguished between households, firms and institutions. The principal findings emerging from the distribution analysis are summarised below:

A total of approximately 13 000 farming families, both small and large landholders, benefited significantly from the irrigation and flood mitigation effects of the project. The agricultural and livestock production activities benefited positively from the project. There were limited benefits from fishing to landless and non-farm households in the project region. All households benefited from the electricity generated in the project. 1 000 households, who lost their lands by impoundment, were partly compensated for their economic losses but these were not considered adequate by many and they did not receive compensation for the social, cultural and psychological costs associated with relocation.

Farmers who were settled on the lands set aside for them on the command areas of the CAP are likely to have benefited from the irrigation supplied to their farms. Some households, who may or may not have been those displaced by the project, benefited by securing employment in the jobs generated by the increased agricultural production.

The recreation industry benefited from the business opportunities generated by the new recreation resources created by the scheme, but some businesses were negatively affected by the increased incidence of white flies. Municipalities and industries in the basin were reported to have benefited from the electricity supplied by the CAP.

The nation as a whole benefited from the increased agricultural production, added supply of electricity, and recreational opportunities generated by the CAP, but also bore most of the scheme's financial costs. The global community benefited from the thermal emissions avoided by the generation of hydropower by the CAP. Except for the costs of electricity and irrigation most of the financial costs of the project are borne by taxpayers who have not directly benefited from the project outputs, except for electricity.

The main social costs are borne by the displaced households, due to forced social change and relocation including these families who had to wait many years to be resettled.

The agricultural traders, agro-industry and agricultural input suppliers were another group who have directly and indirectly benefited from the project.

The project seemed to have contributed in an unquantified way to positive impacts on the education, health and nutrition, poverty alleviation and employment in the region. This was especially important for landless and non-farm households who have not benefited directly from the project outputs, except for electricity. The project has supported a trend towards increased disposable income, increased education opportunities for rural people and university education for the next generation.

Project benefits were stronger for the population in the downstream command area of the project, whereas the costs were greater for the populations in the reservoir area and at the national level (taxpayers) as a result of the subsidies on capital and recurrent costs.

The principal existence values in the project area were the historical sites, National Park and the thermal springs. The benefits were distributed across the region and the nation.

Decision-making, criteria and guidelines

At the time of planning, government gave emphasis to industrialisation for the development of the country. However, development of the agricultural sector was also important due to the self-sufficiency policy in food. One other aspect was the scarcity in energy, and urgent investments were essential in this sector. However, agricultural land is limited, and for increasing agricultural production, irrigation of agricultural land was required. Although it is not indicated in the SAR, another important factor is the policy to increase the production of raw material in order to encourage development in industrial sectors.

Starting from establishment of the republic in 1923, Turkey took European industrialisation as a model, in which development process begins with steel, textile and agri-business industries. The project area was suitable for cotton production, and the success of the Seyhan Project in the same region provided a good basis for the CAP irrigation project in the region, which would further push the industrialisation process by development of textile sector in the region and the country.

On the other hand, the World Bank was also encouraging irrigation projects for further development of the agricultural sector. Thus, the agricultural sector and the incomes of individuals involved in the sector would be further improved and the potential of deterioration in income distribution caused by the initial stages of industrialisation process would be eliminated.

Another important factor is that, in short term, agricultural development was considered to have greater impacts on the overall economic growth than the industrial impacts, which are observed over longer time scales.

The comprehensive development of water resources of the Ceyhan Basin was originally conceived in the early 1960s and is outlined in the IECO report, which was commissioned by DSI and financed by USAID.

DSI was responsible from the overall project design and implementation, and the final design was prepared in association with MARA and TOPRAKSU. TEK did not participate during the design and specifications of the dam and power plant but TEK's view was taken into consideration at the final design stage. It was also decided that DSI would be responsible for the O&M of the irrigation schemes, TOPRAKSU would implement the on-farm works and a FTES would be established and maintained for 10 years after the completion of the project.

It was agreed that after the completion of the HEPP, it would be transferred to TEK. Although the Aslanatas HEPP was in CEASs concession area, it was decided that TEK should be responsible for the operation of the HEPP.

There were no mechanisms for the participation of stakeholders in decision-making and no EIA was required nor undertaken.

Operation and Maintenance (O&M) of the power plant has been under TEK's responsibility since mid 1984. The O&M Department of DSI prepare the reservoir operation rules at the beginning of each water year. TEAS operation engineers are responsible for respecting these rules. In order to eliminate any negative impact on reservoir operations, the upper Basin projects needs to be managed and operated in an integrated way in the next years to maintain the flow of the Ceyhan River into Aslantas Dam reservoir.

The SAR predicted that DSI would remain firmly in control of O&M of all irrigation schemes. However, as an evolving policy of the Government, all schemes were turned over to the users starting from 1995. DSI had the policy of transferring O&M responsibility of smaller and more remote projects to local administrations since the mid 1950s. However, until 1993 the pace of this transfer activity was slow. Transfer rates accelerated dramatically from 1993 onward. The operation and maintenance of the irrigation and drainage works and collection of water charges were sustained by DSI until the end of 1994. After this year, the operation and maintenance of the irrigation system under the main canal level was transferred to the Water User Associations (WUAs). WUAs currently recoup around 70% of the O&M cost, but do not recover capital outgoings.

The SAR predicted that the enactment of pending legislation on agrarian reform would substantially improve income distribution. The reform has not been realised.

No guidance and monitoring activities were foreseen in the project. Although there are some credit support activities under different programs, the study showed that only the Karatepe Village Development Co-operative was established during the planning and appraisal stage of CAP, when Karatepe villagers learned that their villages would be flooded.

The creation of Osmaniye as an independent province in 1995 and its separation from Adana province did not affect CAP, since the project is under the responsibility of DSI, which is an independent institution that has jurisdiction across other than provincial administrative structures and is directly funded from the national budget.

During the planning process for CAP, almost all the criteria and guidelines relevant to various components and cycles were in existence, with the exception of the Environmental Law. The procurement of works, machinery/equipment and services were carried out in line with WB procurement guidelines for portions financed by the Bank, and in line with the Turkish Tender Law for the portions financed by the government. During the planning and design stage of the project no monitoring indicators/programmes were set. The SAR (1973), indicates that land consolidation in CAP area could be considered under the Land and Agrarian Reform Law, which was effective in 1973 and then cancelled by the Constitutional Court in 1976. During this time, government has not deemed land consolidation essential, mainly because the project area was not considered a priority area compared to some regions of the country. In 1972, there were 10 000 farm units with a total area of approximately 108 000ha of land (SAR). In 1983, according to the closing report of FTES there were 11 402 farm units cultivating 82 355ha. In 1972 while 52% of the farmers owned land less than 5ha, in 1983, 66% of the farmers have land less than 5ha. In this group most of the farmers have 2ha land in average, while the country average is 5-6ha per farmer

One of the major social issues of CAP was the expropriation and resettlement, which was carried out based on the provisions of the two main prevailing laws in the fields of expropriation and resettlement. All the work related to the expropriation was carried out by the DSI whereas the whole resettlement processes were planned and carried out by GDRS. The Expropriation Law was amended in 1983 by defining a new method of expropriation. Due to the problems experienced with the resettlement law for years, a new draft law has recently been submitted to the National Assembly and is under consideration.

At the time of project planning EIA/SIA was not practised. However, after the enforcement of the Environment Law in 1983, EIA/SIA has become a common practice after 1993 for dams with a reservoir area bigger than 15km². Policies to protect environment have evolved in the last two decades. The National Park Law (No.2873) was effective on 9 August 1983. Since then, the Karatepe Aslantas National Park is protected under this Law.

In 1970 the Ministry of Tourism has declared Karatepe-Aslantas as a tourist site to be protected. In 1974, the Yumurtalik Lagoon system, at the mouth of Ceyhan River, was taken under protection within the National Park Law and in 1987, Akyatan Lagoon was taken under protection as a wildlife reserve. Following a Government Decree issued in 1993 (Official Gazette 17 October 1993 No. 21731) Haruniye was approved as Thermal Tourism Centre, among 33 others in Turkey. In 1998 the General Directorate of National Parks of the MOF has initiated a study for preparing a long-term master plan for the overall management of the Karatepe-Aslantas National Park.

During the project planning and construction period, the energy sector was not open to private sector activities legally except CEAS concession. The BOT Law issued in December 1984 enables the private sector to finance, construct and operate such power plants.

In 1982, MARA decided to implement the Second Cropping Research and Extension Project in the CAP area. The project encouraged second cropping through various incentives, which resulted in the introduction of soybean to the project area, and an unexpected high production of maize and soybean. Starting from 1982, the government policies on seed sector played an important role on the crop selection as well as the crop pattern of CAP.

Fishing in the reservoir does not comply with the bans specified in the fisheries legislation, mainly due to the lack of control services of MARA

Lessons Learned:

The following lessons were endorsed by the stakeholders meeting in Adana in January 2000.

1. In project cost estimations, one has to pay more attention to possible future developments especially in projects with a long period of implementation. Estimations of physical as well as price contingencies should be done more accurately than employing simple constant coefficients for all components.

2. The price estimations are as important and in many cases more important than the physical design and resource estimations of the project. The magnitudes involved in large-scale projects such as CAP justify the inclusion of economists in the cost calculations.

3. The realisation of the projected cost recovery magnitudes required many institutional arrangements and the supporting legal environment, which were not realised except for the WUA's. The legal and institutional developments are as important as the physical developments in enjoying the full benefits of the project and reducing its cost to the taxpayers. Institutional and legal arrangements should not be delayed until after the project is completed and the benefits have started to be enjoyed. A project proposal should at least say what will happen if cost recovery fails and who will pay if the users don't?

4. Decisions on the reorganisation of agricultural institutions should involve participation of farmers and project staff who are the actual stakeholders. Otherwise such decision taken at central level may not respond the needs of farmers.

5. The yield increases are due a combination of improved agricultural techniques and irrigation supply. One is a capital cost (CAP) and the other is research support and extension on a continuous basis. The efficient extension services working in close relations with research and farmers result with positive impacts on all aspects of cultivation, provided that this services is sustainable.

6. Design studies should take into consideration all technical details be completed with due diligence and with concrete results. This should include geological field studies as well as technical supervision of construction works.

7. Operation and maintenance of HEPP is a vital aspect of overall dam management and needs experienced team, regular supervision and monitoring

8. When designing dams, site selection studies should also take into consideration the existing unique archaeological and historical sites.

9. As a result of uncontrolled and unlicensed fishing, the fish population in the reservoir is negatively affected. Parallel to this the yield from the reservoir is far below the potential.

10. Social and economic support should be provided and a built-in monitoring mechanism should be operated. In such a case the displaced would be better off and the differentiation between the peopleirrigating land after the project and displaced people would be eliminated.

11. The water needs can be changed during the operation period due to the changing cropping pattern and the growing population. The municipal and industrial water demand came up in the Aslantas project because of the insufficiency of groundwater resources. This extra demand will force the operation rule of reservoir. The multipurpose multireservoir basin-wide operation studies should be done periodically for decision-making process during the operation period.

12. The reservoir has been affected from the operation of upstream dam constructed. This positive effect is important especially for the reservoir sedimentation and the operation rule. Because the regulated flows come from upstream dams, the efficiency of reservoir operation will be increased. The effects of the upstream dams must be taken into account especially in planning stage. After the construction of a dam, and in the case of new upstream projects coming up, the life of the reservoir will be prolonged, and be abld to meet the changing water demands easily.

13 Irrigation O&M is more sustainable and cheaper when locally organised user groups are responsible for delivery and cost recovery of irrigation water.

14. CAP stores sufficient water to offset conveyance losses. An options assessment process would have shown whether it is more sustainable to reduce conveyance and irrigation losses, rather than building a larger storage.

15. CAP has set an excellent example in adhering to the basin planas all of the planned investments are gradually being undertaken. Such investments are also candidates for outside financing if they have not been officially contracted out.

16. It is evident that abundance of water actually hurts resource sustainability and its actual use. Farmers do not feel compelled to conserve water and wrong habits in resource use occur.

17. A specific M&E programme is essential to learn lessons, review progress and monitor whether predicted benefits are being met and negative impacts are mitigated.

The following additional lessons were suggested by the participants:

18. Land consolidation should be taken into consideration during the planning stage of the project. Actual expropriation costs would more or less be equal to the costs for land consolidation

19. Discharged water from irrigation should not be released to rivers and wetlands without treatment.

20. Irrigation schemes should be transferred to local administrations or private sector during the project approval stage, and these groups should carry out operation and maintenance

21. For smooth of implementation of expropriation, cadastral work must be completed before the project implementation

22. For an efficient utilisation of water resources, it is essential to establish a good hydrometric observation network, which will provide regular and qualified observations

23. An action plan containing time schedule, costs and activities should be prepared for the possible displaced people before the project is started

24. Identification of settlement facilities should start parallel with the resettlement studies, and preferably, possibilities of resettling the displaced near their villages should be investigated.

25. Fish passages or ladders should be considered in dam projects

The Table provides an overview of the main findings with respect to the predicted, actual, and unexpected outcomes of the CAP.

Predicted, Actual and Unexpected Outcomes of the CAP