Irrigation and Drainage Systems 3:259-266 (1987) © Martinus Nijhoff Publishers, Dordrecht - Printed in the Netherlands

259

Commissioning a paddy irrigation project water management and project viability T.R. FRANKS & T.J. HARDING Project Planning Centrefor Developing Countries, Universityof Bradford. Yorkshire BD7 1DP, UK

Received 20May 1987; accepted 10 June 1987

Key words:commissioning,water management, project viability Abstract. The results of a research investigation into water management practices during the commissioningof a paddy irrigation project are described. Investigationscentred on the water issues and areas irrigated by the project's Right Bank main canal during commissioning.The water use observed during this period was compared with that forecast for the project at full development. From the limiteddata available,feasibletargets for the phasing-inof land during commissioningare suggested, for use both by planners and managers. The effect of adopting these targets on project viability is also briefly discussed.

Background This article summarises the findings of a research programme focussed on the commissioning of the Inginimitiya Irrigation Project in central Sri Lanka. This is a 2500 ha newly-constructed paddy irrigation settlement scheme. It is situated on the banks of the fiver Mi-Oya and is supplied by two main canals, one on each bank of the fiver, off-taking from a tank (storage reservoir). The idea of conducting such a research programme at Inginimitiya was conceived during visits to the project by the researchers during the first season of project operation. It appeared to those responsible for the scheme's management that more water was being used in the first season than would be expected of a fully-developed scheme. It seemed impossible to institute a rotational system of water allocation at field level because of farmer complaints that there was not enough water, even though the designed capacity of the scheme should have made this possible. The managers were not surprised by the heavy water use in the first season but had no guidelines for determining how much additional water was likely to be required in the early seasons, nor how management practices should be modified to take account of this. The research programme was therefore designed to view the problems from the managers' viewpoint, and to provide practical answers to the following questions:

260 - What, in terms of water use, can be expected to happen during the early stages of project commissioning? - What management practices should be adopted, in the light of this? - What effect will this have on long-term scheme viability? Two significant effects of the commissioning process were anticipated. The first was that precedents would be set for inefficient or inequitable water management practices, which would subsequently be difficult to change, and which might cause difficulty in bringing water to the tail of the command area. The second was that any delays in bringing land under irrigation, which were caused by the water management practices adopted, would adversely affect the net benefit stream of the project in the first few years of operation.

Research investigations The research programme was carried out over seasons 2, 3 and 4 of project operation, some additional data from project sources being available for season 1. Primary data collection focussed on the water issues and areas irrigated by the project's Right Bank main canal, as commissioning proceeded. Analysis of the data made possible an assessment of project performance in relation to water use, and the rate at which that performance was approaching the forecast performance for the project at full development. Water use was assessed by daily measurement of discharges at the head of the Right Bank main canal and at various distributary and field channels down the system, to try to measure differences related both to location and to number of seasons of irrigation. Main canal discharges were available from the start of operation, while data collection for the smaller channels only started towards to end of season 2. Daily rainfall data, which is an important factor directly affecting farmer demands for irrigation water, was also collected. Estimates of the cultivated areas being supplied by each of the canals were available from a number of sources. These estimates differed considerably, especially in the early seasons of operation. It was concluded that the estimates based on individual farmer reports were most accurate. Secondary data collected during the research programme comprised information on the physical development of the project, its agricultural production and, in particular, its management system.

Research analysis The analysis of water use was based on two parameters, the total depth of water applied per season (often referred to as the "seasonal duty") and the maximum

261 weekly demand. In tank irrigation, the duty is a factor of particular importance, since the tank is designed to store a certain volume, to be delivered over the season. Management action may be required if, in the early seasons, the total duty, and therefore volume required, is greater than planned. Peak weekly demand is an important parameter because the canal system is designed to meet a particular peak, based amongst other considerations, on estimates of long-term efficiency of water use. If water use is higher than expected in the early seasons, the capacity of the canal system may present an absolute constraint to the area that can be irrigated at that time. An important step in the analysis was the definition of a measure of performance, known as the "performance ratio". This may be defined in two ways:

(i) (a)

Seasonal Duty at Full Development Seasonal Duty during Commissioning Peak Weekly Demand at Full Development Peak Weekly Demand during Commissioning

When commissioning of the project has been completed and it is operating normally at full development, the performance ratio approximates to 1.0 Until that time it will be less than 1.0, if water use during commissioning is greater than at full development. The amount by which it differs from 1.0 is an indicator of the extent to which water use is greater than at full development, while the rate at which it approaches 1.0 is an indicator of the speed with which full developpment can be achieved. Both of these pieces of information are of fundamental importance to managers concerned with the commissioning of an irrigation scheme. There are likely to be a number of reasons for increased water use in the commissioning stages of a project. These include physical factors such as the time required for consolidation of channel and field bunds, as well as the development of hard pans below field plots, which will reduce seepage and percolation. Land levelling which occurs as a result of repeated ploughing operations is also likely to be a major factor contributing to increasingly efficient use of water. Managerial experience, both for the farmers and the scheme managers, will be important in determining overall levels of water use. As both parties become more experienced in the operation of the scheme and the control of the canal network, avoidable losses will decrease.

Results of the analysis The analysis of the programme deals first with the data collected for the Right Bank main canal. These data show the actual water demand that the scheme

262 management was having to satisfy at the time. When compared with the forecast duty and peak weekly demand based on the calculation of project water requirements and the estimates of the contribution of rainfall, they provide a measure of the performance ration which was observed for the area supplied by that canal in the first three seasons of operation. The data obtained, presented in Fig. 1, indicate that, in the first season of operation, water use on the area cultivated within the Right Bank main canal command area was about twice its forecast value at full development, in relation both to the total amount required for the season and to the peak weekly demand. Water use decreased as the commissioning proceeded but, even in season 3, it was still about one third more than its forecast value at full development.

~/////

z 0.8t ~.

0.6 |

0-4[

e

0

" 1st

2 2nd

~

3rd

4th

5th

6th

SEASON AFTER COMMISSIONING ® DUTY AT FULL DEVELOPMENT DUTY DURING COMMISSIONING [] PEAK WEEKLY DEMAND AT FULL DEVELOPMENT PEAK WEEKLY DEMAND DURING COMMISSIONING

Fig. I. Right Bank Area. Project Performance Ratios.

In order to make the analysis more generally applicable, however, it is necessary to take it a stage further. In each successive season the cultivated area wa~ being extended. The canal was therefore supplying water to areas which were at different stages in the commissioning process. Some areas were receiving water for the first time, some for the second nd some for the third. Knowing the areas

263 in each of these categories, it is possible to make an estimate of the performance ratio at field level (for instance for a particular field channel area), and the way in which this changes as commissioning progresses. Some data on field level performance is available from the discharge measurements made on individual distributary and field channels. The values of the performance ratios obtained from this data are plotted against season in Fig. 2, which also includes the theoretical values derived from the project performance ratios. There is a considerable scatter of data from this analysis but the Figure shows that, in broad terms, performance at field level is better than performance at project level,, with an improvement in performance from the second to the third season. Water use was generally higher than would have been predicted from the theoretical derivation but the agreement is close enough to suggest that the basis of the derivation is sound. O ~- 1"0 < e,e' ~_j

CENTRE SLUICE X T2, FC53 X

~ X T2, FC10

Z

< ~: ,w 0 It.. t'~

0.8

/ /

/

f

~ X CENTRE SLUICE T1,FC16 x T1,D5 X T2'FC5&6 FC7 x T1, FC6

/[~ T2,D7 /

~. 0 . 6

/

/

x T2,FC31 x T2,D1

/ i~

x T2,FC50

O4

0"2

=

I

I

I

J

I

1st

2nd

3rd

4th

5th

6th

SEASON AFTER COMMISSIONING x Observed value [3 Theoretical values (T2, D7 refers to Distributary Channel 7 in Tract 2)

Fig. 2. Right Bank Area. Field Performance Ratios.

Development of the analysis The data presented in Figs. 1 and 2, while mutually consistent and rationally explicable, apply only to one part of one scheme, and extend over only three

264 seasons. Moreover the field level data, while showing the expected improvement from the second to the third season, indicate that the rate of improvement is lower than would be expected from the analysis of overall project performance. More general conclusions are therefore suggested, based on a rationalisation of the data. These may be used for planning and for providing management guidelines in future similar situations, but are subject to refinement as further data becomes available. The research analysis suggests that it would be reasonable to work on an idealised field performance ratio of 0.5 in the first season and 1.0 in the second season. The implication is that an individual field channel command area uses twice as much water as forecast at full development in the first season of irrigation, but that it thereafter uses only the expected amount. These assumptions make it possible to define a target phasing of land during commissioning, based on an assessment of water use over that period. If water use is twice its expected value in the first season, then, other factors being equal, only half the command area can be irrigated in the first season. In the second season a further quarter of the area can be irrigated, while still remaining within an overall water constraint. Thereafter a strict application of the theory would suggest that an additional half of the as-yet unirrigated area can be added each season (a series converging to 1.0). In practice, of course, there is sufficient flexibility in the hydraulic system to allow for some overloading and it should prove possible to bring the whole area under irrigation in a relatively short time. Table 1 summarises a target phasing, which can be used as the basis for both the planning and the management of scheme commissioning. At commissioning there may well be other reasons, such as the construction or settlement programme, which prevent all of the land being irrigated in the first season and require a phased build-up in irrigted area. It is important, however, that the target phasing of Table 1 is applied over and above any phased development introduced for other reasons. The equity and efficiency of the overall commissioning process depends on establishing the water management procedures of full development asquickly as possible. To achieve this, it is necessary to aim for a cropping intensity of 50~ on the area available for cultivation in the first season, whatever proportion of the total command area the latter may be. Table 1. Phasing for Irrigated Land during Commissioning.

Season

Area irrigated (as % of available command area)

1 2

50% 75~

3

90%

4

100%

265 A further aspect of the research programme was to assess the hypothetical implications for the overall viability of the Inginimitiya Project of adopting the phased targets of Table 1 for the commissioning period. It was estimated that this would have had an impact on project viability equivalent to a 4~o decrease in net project benefits or a 5 ~o increase in investment costs. It would not have been as significant as a possible loss of one cropping season in five and would be more than compensated by a marginal increase in the yield assumptions. It was thus concluded that the phased introduction of newly-irrigated land would have had no significant effect on the assessment at feasibility stage of the project's viability.

Guidelines for managers at commissioning Three important guidelines for managers involved with the commissioning of similar new schemes are suggested by the analysis of this programme. 1. Plan to irrigate 50% of the available land in season 1. The important aspect of an equitable and efficient method of water management at project commissioning is that the attempt should be made to meet this target through a 50~o cropping intensity for each individual farmer, rather than by supplying water to only half the system. It is acknowledged that a considerable amount of education and communication with farmers will be necessary to achieve this objective. One practical measure which would help in this respect is to make only enough seed available through official sources to each farmer to plant half of his land. Although it will be possible for them to obtain seed through unofficial sources, this will reinforce the official recommendation that only half the area should be cultivated. A second important set of management actions is to ensure that the water allocation procedures are operated on the design basis and additional water is not available on demand. 2. Institute the design water allocation procedures as soon as possible. On the assumption that water use will be about twice its expected long-term value in the first season, but that only a 50~o cropping intensity is being achieved, it should be possible to operate the whole supply system on the design basis in the first season. In particular it is important that the rotational system should be instituted as soon as possible and operated as tightly as possible, so that farmers quickly become accustomed to the total quantity of water that is likely to be available to them, both on a seasonal and weekly basis. This will also ensure that farmers who first start irrigating do not establish a precedent right to water on demand. 3. After the first season of irrigation,farmers should be left free to cultivate their full area (or, as much of their land as theyfeel able to do, with their knowledge of the likely

266

pattern of water supplies). The data available from Inginimitiya indicates that the efficiency of water use increases reasonably rapidly after the first season. It is thus possible to allow farmers to attempt to cultivate as much as they wish in subsequent seasons. By this time they will have experience of the amount of water they can expect the system to deliver and will be able to make a judgement about how much they should cultivate. Cropping intensities will gradually increase as water use efficiencyincreases. In any case it would not be possible for the scheme management to try to restrict farmers to a cropping intensity of, say, 75 ~o, because sufficient seed would be available to plant the full area, and the measurement and management intervention required to impose such a restriction would be extremely complex and time-consuming.

Conclusions Without doubt, in the first few seasons of an irrigation project, a considerable amount of negotiation and adjustment will go on between the farmers and the scheme management. Such negotiation takes place in any case, as some areas are found to be unirrigable, while in other places the command area is extended. However, if the guidelines recommended above were to be followed, such negotiation would take place within an established framework, which could be shown to be intended to provide for the efficient and equitable provision of water to the whole cammand area as quickly as possible. This framework would provide a more rational basis for decision-making than existed at Inginimitiya during its commissioning.

Acknowledgement This research was carried out in collaboration with the Irrigation Department, Government of Sri Lanka. Special thanks are due to Mr K.D.P. Perera, Director of Irrigation, and Mr P.W.C. Dayaratne, Deputy Director, Puttalam Range. Funds for the programme were provided by the U K Overseas Development Administration.

Reference Franks T.R. & HardingT.J. 1987.Commissioninga PaddyIrrigationProject- Water Management and Project Viability.Report to ESCOR Committeeof Overseas DevelopmentAdministration, UK.