The Reliability Improvement in Irrigation Services: Application of Rotational Water Distribution to Tertiary Canals in Central Asia 100 RESEARCH REPORT

Iskandar Abdullaev, Mehmood Ul Hassan, Herath Manthrithilakeand Murat Yakubov

IWMI is a Future Harvest Centersupported by the CGIAR

The Reliability Improvementin Irrigation Services:Application of Rotational WaterDistribution to Tertiary Canalsin Central Asia

100

RESEARCHR E P O R T

I n t e r n a t i o n a lWater ManagementI n s t i t u t e

Research Reports

IWMI’s mission is to improve water and land resources management for food,livelihoods and nature. In serving this mission, IWMI concentrates on the integrationof policies, technologies and management systems to achieve workable solutionsto real problems—practical, relevant results in the field of irrigation and water andland resources.

The publications in this series cover a wide range of subjects—from computermodeling to experience with water user associations—and vary in content fromdirectly applicable research to more basic studies, on which applied work ultimatelydepends. Some research reports are narrowly focused, analytical and detailedempirical studies; others are wide-ranging and synthetic overviews of genericproblems.

Although most of the reports are published by IWMI staff and their collaborators,we welcome contributions from others. Each report is reviewed internally by IWMI’sown staff and Fellows, and by external reviewers. The reports are published anddistributed both in hard copy and electronically (www.iwmi.org) and where possibleall data and analyses will be available as separate downloadable files. Reports maybe copied freely and cited with due acknowledgment.

i

International Water Management InstituteP O Box 2075, Colombo, Sri Lanka

Research Report 100

The Reliability Improvement in IrrigationServices: Application of Rotational WaterDistribution to Tertiary Canals in CentralAsia

Iskandar Abdullaev, Mehmood Ul Hassan, HerathManthrithilake and Murat Yakubov

IWMI receives its principal funding from 58 governments, private foundations, andinternational and regional organizations known as the Consultative Group on InternationalAgricultural Research (CGIAR). Support is also given by the Governments of Ghana,Pakistan, South Africa, Sri Lanka and Thailand.

The authors: Iskandar Abdullaev is Water Management Specialist at the Central Asiaoffice of the International Water Management Institute (IWMI) in Tashkent, Uzbekistan.Mehmood Ul Hassan is Head of West Africa sub-regional office and Social Scientist atthe International Water Management Institute (IWMI) in Accra, Ghana. Herath Manthrithilakeis the Head of the Central Asia office of the International Water Management Institute(IWMI) in Tashkent, Uzbekistan. Murat Yakubov is a Research Officer at the Central Asiaoffice of the International Water Management Institute (IWMI) in Tashkent, Uzbekistan.

Acknowledgements: The authors want to acknowledge the assistance of WUA ZhapalakDirector Mr. Kamilov Janibek, Dr. Bakhtiyar Matyakubov, Mr. Eshmurza Toktosunov andMr. Salijan Akhmataliev in implementation, data collection and field monitoring of rotationalwater distribution in Sokolok Canal, Kyrgyzstan.

The authors also want to thank the farmers of Sokolok Canal without whose willingnessand participation the first systematic experiment of rotational water distribution would nothave been possible. Also appreciated are the critical comments from Vadim Sokolov,Meir Pinkhasov and Ahmadjan Alimjanov (SIC ICWC), which helped to revise the originaldraft.

Abdullaev, I.; Ul Hassan, M.; Manthrithilake, H.; Yakubov, M. 2006. The reliabilityimprovement in irrigation services: Application of rotational water distribution to tertiarycanals in Central Asia. Colombo, Sri Lanka: International Water Management Institute.28p. (IWMI Research Report 100)

/ water distribution / irrigation canals / irrigation scheduling / water allocation / water users’association / monitoring / institutions / participatory management / performance evaluation/ conflict / water rates / Central Asia / Kyrgyzstan /

ISSN 1026-0862ISBN 92-9090-639-1ISBN 978-92-9090-639-1

Copyright © 2006, by IWMI. All rights reserved.

Cover photograph by Iskandar Abdullaev shows water users cleaning the Sokolok Canal,Kyrgyzstan.

Please send inquiries and comments to: [email protected]

iii

iii

Contents

Summary v

Background 1

Problem Description 2

Methods and Materials 4

Results and Discussions 8

Conclusion and Recommendations 19

Literature Cited 21

Acronyms

FSU Former Soviet UnionIFPRI International Food Policy Research InstituteIHE International Institute for Hydraulic and Environment Engineering

(UNESCO-IHE Institute for Water Education)IIMI International Irrigation Management InstituteILRI International Institute for Land Reclamation and ImprovementISF Irrigation Service FeeIWMI International Water Management InstituteIWRM Integrated Water Resources ManagementO&M Operation and MaintenanceOIP On-Farm Irrigation ProjectSIC ICWC Scientific-Information Center of the Interstate Coordination Water Commission

of the Central AsiaWMO Water Management OrganizationWUA Water Users AssociationWUG Water Users Group

v

v

Summary

Land and agricultural reforms in Central Asiancountries, following the collapse of the FormerSoviet Union (FSU), have led to a big increase inthe number of individual farm units along secondaryand tertiary canals. Given the new setting, themethods for water distribution, as applied under theformer large-scale collective farming system, havebecome irrelevant, leading to much chaos, inequityand unreliability in water supply to farmers. Thus,many farmers and water managers have had toresort, with variable success, to some alternativewater distribution methods to meet these newchallenges. Nevertheless, transparency and equityin local water use still remains an issue. With thisin mind, an action research to study an arrangedintermittent (rotational) water distribution wasundertaken in a typical distributary canal incollaboration with a Water Users Association(WUA) in the Kyrgyz Republic during 2003 and2004. The rotational water distribution methodemployed was performed in a truly participatorymanner and allowed farmers involved to always beaware of their specific time schedules, includingwhen to irrigate their fields and for how long. Thisalone has translated into huge time savings for

farmers when waiting for their irrigation turns andmore equitable water distribution between differentcanal reaches. This has also allowed those at thetail ends to increase crop yields and net incomes,resulting in better Irrigation Service Fee (ISF)collection. At the same time, there has also beena change in the nature and pattern of waterdisputes.

The work conducted on rotational waterdistribution suggests that it is the needs andconcerns of the end users that provide a goodentry point for collective action, to pragmaticallyunderstand and analyze the situation, from whereappropriate remedial strategies and methods canbe further devised and employed. This is also agood starting point to initiate farmer debates anddiscussions on public participation, which shouldultimately lead to a truly farmer-owned processand action. Legal instruments alone, though beingan important factor, per se are rarely sufficient tofully enable, sustain and institutionalize requiredchange to local communities. Unfortunately, thishas mostly been the case in Central Asianeconomies so far, and it is this that requires majorchange.

1

Central Asia has one of the oldest systems ofirrigated agriculture in the world, with the historyof irrigation dating back thousands of years. Inthe early twentieth century (pre-Soviet times),water distribution was based on the IslamicShar’ia law. According to Shar’ia, water wasregarded as a common good. The public ownedall canals and ditches collectively with the mainprinciple for water sharing being for a landownerto receive sufficient water to fill their field (Bartold1970; Mukhamedjanov 1986).

During the Soviet era, Central Asia wascovered with large irrigation schemes serving atotal of about 8.0 million ha (hectares) of irrigatedcropland. Massive irrigation and drainage systemswere designed to accommodate the needs oflarge-scale farm units owned and controlled bythe state. These large farms consisted of anumber of production units called “brigades” withwater allocated and distributed against “agro-technical operations plans.” From the mid-1960s,water distribution in Central Asia was demand-based. In the mid-1980s, this was replaced bythe “adjusted water demand principle”(“limitirovannoye vodopol’zovazniye” in Russian)or, simply, supply-based, requiring proportionateadjustments to the initially expressed waterdemands in situations of lower water availability.

After the collapse of the Soviet Union, theintegrated large-scale irrigation systems had to be

shared across the newly established independentCentral Asian states. Each nation undertook itsown agricultural, land and water reforms tosubdivide state farms into smaller, farmer-ownedor managed units. In some cases, farmers arefree to plant whatever crops they like, but cottonand wheat are still mandated, for instance inUzbekistan. In Central Asia, there are tertiarycanals that supply water to only a few farmers insome areas, and to hundreds in others.Previously, these canals served big collectivefarms with the entire system designed to suitlarge-scale farming. However, such waterallocation and distribution mechanisms hadbecome unsuitable or even redundant in the post-Soviet Central Asia, thus resulting in chaos,inequity and unreliability at all levels of theirrigation system management. This has also ledto a mismatch between water supply and actualcropping pattern needs, and an increase in thenumber of water-related disputes. So, this reportaddresses a specific context of massive inequityand unevenness in water allocation anddistribution experienced presently by the waterusers in transitional Central Asian economies, asa result of broad-scale fragmentation of thepreviously large farms. The report describesaction research aimed at making waterdistribution at the tertiary level more reliable,transparent and equitable.

The reliability improvement in irrigation services:Application of rotational water distribution to tertiarycanals in Central Asia

Iskandar Abdullaev, Mehmood Ul Hassan, Herath Manthrithilake and Murat Yakubov

Background

2

Water distribution along irrigationschemes: Institutions andinfrastructure

Main canals (“magistralniy kanal” in Russian) inCentral Asia are normally lined and very wellequipped to prevent seepage losses. Everymajor offtake along a main canal is equippedwith flow regulation and water measurementstructures. The gates at the main canal offtakesare regulated based on total water demand bywater users. Water requests are collected everythree days by the canal managers to prepare aschedule of water releases at all diversionpoints. At the same time, it is not uncommon forthe manager to make sudden changes to suchschedules, if so requested by higher-levelauthorities or in case of an emergency. If canalwater has to be lifted by pump, the reliability ofthe water supply depends on the availability ofelectricity. The main canal is normally dividedinto reaches (“gidrouchastok” in Russian), whichare supposed to be coordinated by theiroperations control units (“dispecherskiy punkt” inRussian). Since they are equipped with outdatedand inefficient radio communication sets, thosein charge of different canal sections lack real-time data and, to a large extent, now makeuncoordinated water distribution decisions. Thisleads to quite frequent flow changes being madesimultaneously in different reaches, resulting inunreliable and unequal water distributionthroughout Central Asia.

The former on-farm water distribution systemin Central Asia includes secondary and tertiarycanals. Normally, they are poorly equipped withregulation structures (such as outlets, gates ormeasuring devices) due to some built-in featurespeculiar of the old system. One on-farm irrigationsystem in the FSU used to belong to onecollective farm or a state farm (“kolkhoz” or“sovkhoz”). Since their funds were quite limited,such farms could only improve on-farm irrigationinfrastructure when all other costs were fully met.Thus, farm level infrastructure is typically scantyand poorly maintained.

Currently, daily bulk water deliveries arearranged with individual tertiary canal watermasters (mirabs), normally twice a day and basedon (i) total water requested by water users for theday, or (ii) water available from the main canal forfurther deliveries down the secondary network.

Water delivery to the farm gate (field) is thefull responsibility of a WUA’s mirab, who is tocollect all water requests from the water usersand open the gates as per the irrigation schedule,prepared by the WUA committee. It should alsobe noted that mirabs always enjoy a high degreeof freedom in supplying water to the farms. Inmost cases, mirabs are very rational and do theirbest to avert any water-related conflicts throughdiscussions with water users and appropriate on-the-spot decisions. However, due to the largenumbers of competing water requests, mirabs areunable to equally and reliably distribute wateramong numerous water users.

Problem Description

The centerpiece of water resources managementis allocation of water to different purposes andusers, complemented by distribution to thoseusers (actual implementation of allocationdecisions). Thus, the term “allocation” refers tothe assignment of rights or allowance to usewater (Uphoff 1986). A working definition for water

allocation could be a combination of actionsenabling water users and uses to obtain orreceive water for beneficial purposes according toa recognized system of rights and priorities(Taylor 2001). The three water allocationmechanisms widely used in the world are: (i)administrative, (ii) user-managed, and (iii) applying

3

of market instruments in water allocation (IFPRI1994). In a broader sense, water allocation canbe divided into two types: supply driven anddemand driven. Table 1 presents the options forirrigation scheduling and water distributionemanating from these two water allocation types.

Until the 1980s, the water requirements of theirrigation in Central Asia was demand-based(against crop water requirements) with waterdeliveries scheduled and effected centrally (i.e.,administrative mechanism) by state-run WaterManagement Organizations (WMOs). Irrigationwater was delivered in variable flows in 10-dayintervals (decade) based on crop type, sown area,soil characteristics, groundwater depth and otherenvironmental factors of irrigated areas. Tostreamline water allocation for irrigated agriculture,a “zoning system” was developed by researchinstitutes categorizing all irrigated areas. Thezoning was based on the environmentalcharacteristics of each particular zone, affectingin some way or other the consumption pattern ofirrigation water. Thus, the areas with similarenvironmental indicators were grouped under oneirrigation zone (or hydromodule), specifyingpredetermined irrigation regimes (date, amount,frequency) for all crops grown in such areas. Thisapproach was meant to facilitate the waterdistribution process in irrigated agriculture.

By the mid-1980s, given the growingconcerns over the drying Aral Sea, the waterallocation principle in Central Asian agricultureshifted from being demand-based (against cropwater requirements) to that driven rather by

supply. According to the latter, initially expressedwater demands for irrigation were later subject toproportionate adjustments considering actualwater availability in a river basin. However, theway water was scheduled and delivered to userscontinued to be the same as it was earlier whencrop water requirements were not restricted.

At present, scheduling and deliveries ofirrigation water are centrally arranged by WMOs,based on cropping patterns. Central to theprocess is preparation of draft water-use plansbased on statutory water requirements for allcrops planned for a season. These are latersubject to proportionate adjustments if overallwater availability for the season is forecast atsomewhat less than normally required. Thissystem worked quite well under the large-scalefarming system, when a much fewer number ofwater users had to be dealt with. However, keychanges in the Central Asian economies followingindependence led to multiple fragmentations ofhuge state farms among numerous newlyestablished individual farmers. Thus, WMOs findit almost impossible to (i) collect water requestsfrom so many users, and (ii) to deliver water tothem in an orderly and timely basis. Due to largenumbers of overlapping water requests from thefarmers, the WMOs are hard-pressed to prepareany workable water delivery schedules. Therefore,almost all canal outlets are left open instead, tolet water continuously flow without muchregulation. Consequently, the upper reach tertiarycanals receive more water at the cost of the tailend canals, and within tertiary canals, small fields

TABLE 1.Irrigation scheduling and water delivery by allocation types.

Allocation type Type of scheduling Type of delivery at tertiary offtake

Supply-based Proportional scheduling Traditional Irregular changing flows

(water source) Arranged Intermittent full supply

Demand-based Central scheduling On request Variable flows - short periods

(crop water requirement) (agency deciding) Arranged Variable flows - long periods

Arranged Rotation Intermittent full supply

Responsive scheduling Automatic Stepwise changing flows

(farmer deciding)

Source: Horst 1998.

4

fill up quickly and surplus water is discharged tothe drainage network, while bigger plots are neverirrigated fully throughout the season (IWMI 2004).In addition, water deliveries become unreliable inthe lower reaches due to discharge fluctuations,and simultaneous efforts of water withdrawal bycanal mirabs, as well as the users. This situation,especially, puts those at the tail ends of canals(farms and households) at a great disadvantagedue to unreliable and unequal water distribution.

Experience elsewhere suggests that effectivewater delivery in situations like these can only beachieved by fostering greater participation ofusers in the process of planning water use anddistributing water (Abernethy 1988; Horst 1987;Horst 1990). Involving water users in the planningand distribution processes requires participatoryapproaches and methods that are user-oriented,

as well as being simple enough to be understoodby farmers. Given this context, IWMI, under theauspices of the IWRM-Fergana Project1,experimented with a number of such alternativewater distribution methods that would be effectiveand build on farmers’ own initiative and capacity.One such method was successfully introducedand pilot-tested at a tertiary canal of a pilot WUAin Kyrgyzstan. The method in question is calleduser-based rotational water distribution andfeatures proportional scheduling of full canal flowfor rotational delivery to individual outlets withactive participation of the water users. This wasnot something completely new to the localpeople. Some variation of it, locally called“avron”, was practiced long before Russiaconquered Central Asia in the second half of thenineteenth century (Thurman 1998).

Methods and Materials

User-based Rotational WaterDistribution

One of the methods widely used for waterdistribution, based on timed allocation, is knownas warabandi. The method has been widelypracticed for over a century in most countries ofSouth Asia, such as India, Pakistan, Bangladeshand Nepal. In Pakistan and northern India, it isapplied over some 24 million hectares. Accordingto many reports, there are a number of variantsof warabandi, featuring a range of designs andmanagement options. Water at a tertiary level issupplied to every landowner or field at a fixedrate, for a fixed duration and at a fixed time onfixed days. The irrigation variables are fixed eitherby the O&M (operation and maintenance) agency(pucca warabandi) or by the farmers (kachaa

warabandi). The duration of the turn isproportional to the size of the farm. Oneadvantage of this method is that it matches localmanagement capacity and is intended to provideequitable access. However, some warabandisystems have increasingly been experiencingsevere problems with sustainability (salinity) (see,for instance, Bandaragoda and Ur Rehman 1995;Chaudhry and Young 1989; Latif and Sarwar1994; Lowdermilk et al. 1975; Makin 1987; Merrey1990; Qureshi et al. 1994; Singh 1981; Vehmeyer1992; Bastiaanssen and Bos 1999). Since thewarabandi practice tends to breakdown whenwater supply is not limited, some experts bothinside and outside Central Asia, have suggestedthat warabandi does not suit the specific CentralAsian context where crop diversification and landfragmentation are important factors. There is

1IWRM Fergana-Integrated Water Resources Management in Fergana Valley, project funded by Swiss Agency for Development andCooperation, implemented in the Fergana Valley of Kyrgyzstan, Tajikistan and Uzbekistan. The project is jointly implemented by IWMIand SIC ICWC since the year 2001.

5

ample evidence that no real water shortages arepresent in Central Asia (Mukhamedjanov et al.2004; SIC ICWC 2003), with most problems incurrent water distribution being rather of a socio-institutional nature, than purely technical. Wehypothesized, with proper adjustments and strongfocus on the socio-institutional aspects of waterdistribution, warabandi could be adaptedsuccessfully in the emerging context of CentralAsian irrigated agriculture.

The main hypothesis of this study holds that,when designed and applied in direct consultationwith water users, warabandi makes local waterdistribution more transparent and effective, thus,improving equity, reliability and timeliness of thewater supply to farmers sharing one distributarycanal, and reducing the number of water-relateddisputes among them.

The main feature of the proposed waterdistribution method is timing the duration of waterdelivery to each watercourse in the distributary inaccordance with crop-specific water requirements.This is a major difference between classicalwarabandi as practiced in South Asia and therotational methods pilot-tested in Central Asia.The latter, among other things, also provides theopportunity for water users to actively participatein water distribution.

The three major parameters that areconsidered when implementing rotational waterdistribution are as follows: (i) the 10-day waterduty for each watercourse based on the statutorywater requirements of each particular crop grown;(ii) the time required to release the requiredvolumes, based on the discharge rate available inthe head of the distributary canal and the size ofthe fixed outlet structure in the head of eachwatercourse; and (iii) the timing of opening andclosing each outlet.

The existing planning procedures werefollowed to determine the 10-day water duty foreach watercourse. Local water use plans arenormally prepared annually, based on crop plans(crop type and area sown), canal characteristics(delivery efficiency) and weather forecast. In thisresearch, an Excel spreadsheet-based programwas used to calculate the water duty for eachwatercourse of the study canal.

The time required (Tirr (i,j)) to release the 10-day water duty for a watercourse is calculated asfollows:

Where: Tirr (i,j) duration of water supply towatercourse “i” in the j-th decade,in hours

Virdecade (i,j) water duty for the watercourse “i”in the j-th decade, crop waterrequirement, identified from wateruse plan in m3 (cubic meters)

Qj head discharge for thedistributary canal in the “j”decade, in l/s (liters per second)

3.6 factor to convert l/s into m3/h(cubic meters per hour)

In this regime, all flow is supposed to besupplied to only one watercourse at a time. Thisis practically feasible, because the offtakes havethe same discharge rate as at the head of thedistributary canal, and so diverting the entiredistributary inflow to one offtake at a time doesnot create any problems.

The Site

To pilot-test this user–based rotational waterdistribution method, a tertiary canal was selectedin one of the WUAs located in Osh Province ofthe Kyrgyz Republic. It is called Sokolok Canalunder the WUA “Zhapalak”. The selection wasdone in consultation with the WUA. The studycanal is believed to be representative of a typicaltertiary canal in Central Asia today. The WUAwas founded in 1996 and has 2,112 ha in thetotal irrigated service area (figure 1). Thecommand area of the Sokolok Canal is 290 haand the canal is around 6 km (kilometers) long.This tertiary canal supplies water to 473 waterusers via 14 watercourses. The maximumcapacity of Sokolok Canal at the offtake is250 l/s. Its water source is the main Aravan-Akbura canal, which is one of the largest canalsin Osh Province, South Kyrgyzstan.

(Tirr (i,j)) =

(Virdecade (i,j))

(Qj.3.6)

(1)

6

Before the intervention, the studydistributary was poorly maintained. It had nowater regulation structures, so all its outletswere opened and closed manually, usingspades, stones and mud. This made waterdistribution extremely difficult to manage,leading to frequent siltation and bankdestruction. During 2003-2004, a part of thedistributary was lined under the World Bank-sponsored On-Farm Irrigation Project (OIP).

The cropping pattern in the pilot areacomprised of 43.4 percent corn, 11.5 percentwinter wheat, 3.3 percent sunflower, 3.1 percentvegetables, including onions, tomatoes, andcucumbers, while 2.1 percent of the commandarea was occupied with fruit trees. The averagelandholding within the command area is 0.48hectare per household. The sizes of individualland parcels across the watercourses were notuniform, with the following 3-land distributionpatterns observed:

1. Commands with less then 0.50 ha perlandholding on average (Watercourses 1, 2, 4,6, 8, 13, and 14)

2. Commands with the landholding sizesbetween 0.50 to 1.0 ha (Watercourses 3, 9,10, 11, and 12)

3. Commands with an average of more than 1.0ha (Watercourses 5 and 7) per landholding

Data Collection

Most baseline and follow-up data required for thisstudy were collected either from (i) primarysources such as farmer surveys, direct fieldmeasurements of maximum and minimum waterdischarges in the canal outlets (these wereobtained using mobile weirs), processdocumentation by field observers (minutes of themeetings of Water User Groups (WUG) bywatercourses), direct communication with WUAstaff (technical data on Sokolok Canal and itsofftakes, etc.); or from (ii) secondary sourcessuch as WUA records (water, land and croppinginformation from staff reports and annual reports

by the WUA Director to the WUA Council, andtechnical documentation). Data for 2002, thatwere not available from secondary sources, wereobtained by interviewing the canal master and thewater users. In addition to technical data, theexisting water distribution practices, as well asfarmers’ perceptions and attitudes of them, werecollected and analyzed because of rapid appraisalsurveys.

Assessing the Irrigation Performanceby Survey

A considerable amount of work in the world ofresearch has been dedicated in the past 10 yearsto developing an irrigation performanceassessment framework. Most irrigationperformance indicators have traditionallymeasured adequacy, equity and reliability of waterservices (Wolters 1992; Murray–Rust and Snellen1993; Bos et al. 1994). Overall, the methods forsuch evaluation have, particularly, undergonemajor changes in the last 20-25 years.Performance assessment began in the mid-70s interms of classical irrigation efficiencies (Bos andNugteren 1974; Jensen 1977). This was latertranslated into an assessment based on irrigationperformance indicators (Levine 1982; Small andSvendsen 1990; Bos et al. 1994) with the mostrecent developments leading to the principles ofwater accounting at the basin scale (Molden1997; Burt et al. 1997).

Bos et al. (1994), after a thorough analysis ofall the previous work, came up with a list ofindicators to measure the performance of (i) waterdelivery systems, and that for (ii) environmentand economic aspects of the production system.This paper builds on the former - water deliverysystem performance indicators. Those are: (i)Reliability: actual water delivery schedulesreflecting the planned or intended irrigationschedules; (ii) Timeliness: water delivery againstspecific time requested; and (iii) Equity: an extentto which each farm receives water according toirrigation requirements of the crops grown (Bos etal. 1994, 2005; Horst 1998).

7

FIG

UR

E 1

.P

ilot

cana

l lo

catio

n.

Not

e: M

ap w

as c

reat

ed b

y D

r. A

lexa

nder

Pla

tono

v, G

IS/R

S S

pec

ialis

t at

IW

MI,

Tash

kent

, U

zbek

ista

n.

8

The general consensus is that performanceevaluations can only be made at the higherhierarchical levels in an irrigation system andwhere routine and accurate flow measurementsare made. Therefore, in this study at the tertiarycanal level, we have used water users’perceptions of irrigation services as aperformance evaluation tool. The surveys wereconducted in January and November of 2003 toanalyze the situation before and after theintervention. Eighty water users were sampled (17percent of total) in each survey. Five to six waterusers were randomly selected from eachwatercourse, using the WUA’s Irrigation ServiceFee (ISF) roster for the Sokolok Distributary.They were requested to evaluate waterdistribution performance based on the above threeindicators before and after the intervention usinga 1-to-5 scale from very bad (1) to excellent (5).The survey questions reflecting those indicatorswere simply designed to make it easy for farmersto give answers.

Assessment Criteria

The efficiency of rotational water distributionbefore and after the intervention was evaluatedagainst five measurable indicators. All suchindicators for the intervention period werecompared with those for the previous cropseason. The first indicator, ranked by the water

users as the most important, was equitybetween the offtakes of the distributary canal.This was operationalized as the ratio of total orper hectare volume of water actually supplied fora watercourse, based on measured data, to thewater duty for the watercourse in the water useplan of the Sokolok Canal. The second mostimportant indicator ranked was the averagewheat yield per watercourse during theintervention period compared to the previouscrop season. Wheat yields were assessed byinterviewing household heads and farm leaders,as well as from tax records at local governmentoffices. The third indicator was the rate of ISFcollection per watercourse. As no special effortswere made by the WUA Directorate to improveISF collection during the intervention period, anyimprovement in fee collection was assumed tohave resulted from better water distribution andusers’ satisfaction with the service. The ISFcollection data for the two seasons in questionwere obtained from WUA accounts. The fourthindicator was time spent by water users to gettheir irrigation turn during the interventioncompared to the preceding crop season. Thefifth indicator was the number and nature ofwater-related disputes in the distributary canalcompared to the previous year. In WUA“Zhapalak”, conflicts and grievances had beenregularly registered since 1999, so the field staffregistered all such disputes during theintervention period as well.

Results and Discussions

Monitoring the Rotational WaterDistribution: Socio-technical Issues

The new, rotational water distribution wasimplemented in two phases. During the firstphase in the growing season (April-October) of2003, it was directly supervised by the projectstaff. Two field staff were hired in the first year of

the trial to implement and monitor the wholeprocess. They were first trained for two weeks inthe rotational water distribution to facilitatefarmers and WUA staff in the smoothimplementation. The following data were collectedby the staff: daily head discharges for thedistributary and all its offtakes, annual croppingpatterns for each offtake command, the number

9

of hours each watercourse was scheduled toreceive water, daily changes in water schedulesand the number of disputes that occurred duringthe season.

During the second phase in 2004, waterdistribution was supervised directly by the WUAstaff (water master) in cooperation with the WUGsat each offtake. The project staff only monitoredthe process.

Adopting and implementing the newmethodology was not an easy technical task. Ittook quite an effort to plan and prepare itthoroughly in close cooperation with the WUADirectorate, water masters and water users. A

survey conducted prior to the interventionindicated the need for farmers to be organizedinto Water User Groups (WUGs) along eachwatercourse. However, due to insufficient time leftin the first year, the formation of WUGs waspostponed to early 2004. The project staff metwater users from each watercourse to discussWUG issues. As a result, each watercoursenominated a volunteer to represent its users onall rotational water allocation issues. WUGs werefinally formed in the second year at eachwatercourse using a thorough social mobilizationapproach, based on the following 8-step cycle(figure 2).

FIGURE 2.Participatory cycle for rotational water distribution.

Note: Created from experimental observations and data from the Sokolok Canal.

10

In the first phase, it was the project staff whoinitiated meetings and surveys at eachwatercourse to perform diagnostic analysis of theproblems faced, build awareness and consensusamong the water users, who then elected theirown representative to supervise water distribution.

In the second phase, project staff worked inclose cooperation with the offtake representativesand canal masters to collect the requiredtechnical data on the distributary canal and all itsofftakes. The data included information on thecommand and sub-command areas, canallengths, hydraulic and flow control structures, thenumber of water users, temporal flowcharacteristics, cropping patterns for a particulargrowing season by each offtake, long-termaverages for water discharges and water level inthe head of the distributary, etc.

In the third phase, water users weremobilized to make technical improvements totheir canals. A technical survey identified themajor technical obstacles as being siltation, poormaintenance, unstable flows and unregulatedofftake diversions (table 2).

The second meeting with farmersconcentrated on technical problems that mighthinder rotational operation. Users were willing todesilt their canals and also asked for supportfrom the project in equipping their offtakes withgates. Three alternative gate structures wereidentified, with the majority favoring a pipe-basedshut-off gate structure, given its cost and utility.While technical assistance to design and produce

the required number of gates was provided by theproject, labor contributions to install them weremade by the water users with supervision fromthe project staff. Project staff calculated therequired pipe diameter for each gated outletbased on the maximum discharge of an offtakediversion and contracted a local manufacturer toproduce 14 units. Each unit cost US$35. Theoutlets were installed before the beginning of thegrowing season in March 2003. In addition, twoflow regulation gates were installed in the middleand tail reaches of the canal to help sustain itswater level so that the outlets could receive waterwithout much damage to the canal banks. Thegates are operated by the mirab (canal master),according to the agreed distribution schedules.

The fourth phase involved conversion of thewater volumes planned for each watercourse intowater turns. The duration of each water turn wascalculated using formula (1). These calculationswere facilitated by a special spreadsheet providedto the WUA. WUA staff and canal masters(mirabs) were trained to use it and thenimplement the schedule in the field.

Following this, (the fifth phase), the waterturns were transformed into 10-day draft irrigationschedules. The draft schedules were discussedwith WUGs every 10th day of each monththroughout the growing season. Preliminarydiscussions held within the WUGs along theSokolok Canal had revealed that the farmerspreferred their irrigation turns to proceed from thehead-end to the tail of canals. Following this

TABLE 2.Major impediments to rotational distribution in Sokolok Canal.

Technical Problem Description

Discharge in the canal is low due to siltation Qmax= 200 l/s, Qaverage= 120 l/s (multi-year); In the 2003 season

Q average was only 80 l/s

Canal water losses are huge 73% of head discharge is lost to infiltration

(as tested prior to the intervention)

Flows and water supply are very unstable Head discharge fluctuates throughout the day

Offtake diversions are not regulated None of the offtakes have any regulation structure and thus

have to be manually closed or opened using mud and stones

Notes: Created from experimental observations and data from the Sokolok Canal.

l/s = liters per second.

11

pattern, irrigation schedules for the first 10 dayswere drafted and approved by WUGrepresentatives from the beginning of April 2003.The format of the schedules was kept as simpleand user-friendly as possible. It clearly set outthe time for each irrigation to start and finish ineach watercourse. Starting at midnight on the firstday, an irrigation turn was scheduled to end atmidnight of the 10th day following a 10-dayrotation cycle.

In the sixth phase, after irrigation scheduleswere finally approved, they were widelycommunicated and publicized among all waterusers in the command area using metallic displayboards placed at the head, middle and tail of thedistributary canal. Such displays proved to be agood reminder to water users as to when and forhow long they would receive water.

The seventh phase was the implementationof the water rotation as scheduled and jointlyapproved. In 2003, this was supervised by theproject staff but in 2004, the canal’s watermaster took charge. In both cases, those incharge were instructed to regularly note anyinterruptions or failures in the irrigation scheduleas well as measure and record the flow rates atthe head of the distributary and its offtakesthree times a day.

Finally, in the eighth phase - upon completionof the crop season, both in 2003 and 2004, aseries of wrap-up meetings were held by eachWUG to (i) discuss the outcomes of the newmethodology used, (ii) measure overall users’satisfaction, and (iii) refine any furtherarrangements for the next growing season.Following this, another survey was conducted tofollow up users’ perceptions.

Users’ Responses Before and AfterIntervention

The survey (table 3, columns titled “before”)clearly suggested that water users had been quiteunhappy with existing water distribution practices,in terms of equity, reliability and timeliness,longing for better performance. The follow-upsurvey of the same respondents conducted in2004 revealed that the share of those dissatisfiedwith various aspects of the irrigation serviceperformance had considerably decreased, whilethose satisfied increased (table 3, columns titled“after”). Likewise, those sampled also reportedthat the time expended to get their irrigation turnhad also decreased. Table 3 also shows that,overall, the water distribution rated as bad or very

TABLE 3.Irrigation service performance as ranked by survey respondents.

Rankings by

Respondents Reliability Timeliness Equity Overall Performance

[% of total] Before1 After2 Before After Before After Before After

1 – Very bad 9% 4% 12% 5% 40% 14% 5% 10%

2 – Bad 44% 30% 48% 26% 56% 22% 32% 24%

3 - Moderate 31% 18% 40% 15% 4% 17% 38% 15%

4 - Good 12% 37% 0 51% 0 38% 19% 42%

5 - Excellent 4% 11% 0 3% 0 9% 6% 9%

Total No. of 100% 100% 100% 100% 100% 100% 100% 100%

Respondents N=80 N=80 N=80 N=80 N=80 N=80 N=80 N=80

Notes: Created from experimental observations and data from the Sokolok Canal.1 Before implementation of rotational distribution2 After implementation of rotational distribution

12

bad both before and after the intervention wassimilar. This can be partially explained byrelatively poor services for some tailenderscompared to those at the canal head.

Number and Nature of Water Disputes

Disputes between farmers over water distributionare direct indicators of the irrigation serviceefficiency. The number of disputes as well astheir nature can help diagnose water-relatedproblems. In many cases, water disputesbetween different offtakes of the distributary canalare prone not to be registered and handledinternally by the water users. The WUA staff isnormally approached only if parties in a disputefail to settle their grievances on their own. Giventhis, registered water disputes represent only thetip of the iceberg. Nevertheless, patterns in theincidence of such disputes can be quite indicativeof how water users react to any particularchanges in water distribution practices. Thedisputes registered by and large belonged to thefollowing two types: (i) over water volumes,normally occurring between WUA staff (canalmaster) and water users, when the latter complainthat the volume of water supplied was not enoughto grow their crops; and (ii) over irrigation turns,most prevalent between water users along thesame watercourse (prior to the rotational waterdistribution, there were disputes over waterrotation between the canal water users).

In the year preceding the intervention, therewere 26 disputes registered with half of them (13)being over water volumes and another half over

irrigation turns (table 4). Thus, water users wereequally in dispute both with one another and withthe WUA. In the first year after rotational waterdistribution was introduced, the total number ofdisputes declined to 18 or by a third. The secondyear of practicing with rotational water distributionwitnessed further decline down to 14. Mostdisputes (83 percent) occurring in the first yearwere over water volumes, with those over irrigationturns amounting to only 17 percent. The sametrend was also observed in 2004. Thus, waterdistribution improved due to clear water distributionschedules reducing the total number of disputes,especially those between water users.

Water Service Fee Collection

Since 1996, it has been mandatory to pay forirrigation water in Kyrgyzstan. The Kyrgyz WUAsare subject to two Irrigation Service Fee (ISF)tariffs as set by the Kyrgyz government - one forthe summer crop season (“vegetatsionniy period”in Russian) set at $0.73 for each 1000 m3 ofwater withdrawals, and one for the winter cropseason (“mezhvegetatsionniy period” in Russian)set at $0.24. A markup to this amount isadditionally charged by the WUAs to cover theiroperation and maintenance costs. Thus, the totalwater fee for the summer crop season in thestudy WUA “Zhapalak” in 2003-2004 was set at$0.98 per 1000 m3 of water withdrawn. Accordingto national water regulation, at least 70 percent ofISF should be paid in cash with the rest in kind.However, given high poverty and poorly developedlocal markets, the ISF is usually paid in kind. In

TABLE 4.Type and number of water disputes in Sokolok Canal.

Time Period Total No. of Disputes over Disputes over

Disputes Water Volumes Irrigation Turns

Prior to intervention, 2002 26 13 13

After intervention, 2003 18 15 3

After intervention, 2004 14 10 4


13

addition, the volumetric charges, althoughstatutory, are hardly implemented by KyrgyzWUAs due to the lack of accurate watermeasurement. So ISFs are normally charged onan area basis ($/ha). Thus, in WUA “Zhapalak”there was a flat rate of about $11 per hectare,regardless of the crop type or location. Feecollection for the selected Sokolok Canal wasamongst the highest in this WUA even beforerotational distribution was introduced. Total feescollected from half of its sub-commands weremore than on average for the entire WUA - $6.6/ha, although farmers from the offtakes 5, 7, 10,11, 12 and 13 paid on average less than $4/ha inwater fees (figure 3). As for the offtake 14, thelocal water users were exempted from paying ISFdue to the prevalence of kitchen gardens hereand availability of alternative water sources.

It was also assumed that there was a directrelationship between ISF collection and thequality of irrigation service. Although, no otherspecial measures were undertaken by the WUAmanagement when first introducing the rotationalwater distribution, the ISF collection dramaticallyimproved. During the first year of intervention(2003), the number of tertiary commands that hadaverage or above average collection ratesincreased. However, no changes were found inthe case of the downstream offtakes when

FIGURE 3.Water service fee collection by outlets of Sokolok Canal.

compared to 2002. In 2004, when rotational waterdistribution was managed by water usersthemselves, 12 out of 13 offtakes along theSokolok Canal had higher water fee collectionrates than on average for the entire WUA.

Thus, improvements in revenue collectionwere mainly the result of the newly adopted waterdistribution method as well as the empowermentof the local WUGs to take local managementdecisions on their own and collectively managethe process. Following the second year of therotational water distribution experiences, 12 out of14 outlets along the Sokolok Canal increased ISFcollections by 50 percent, having the highest ISFcollection rates among other canals of WUAZhapalak.

Among the main drivers for the farmers ofSokolok Canal to opt for rotational waterdistribution was high water losses coupled withlarge numbers of water users located along thecanal, which made local water distribution aregular nightmare. Despite being an integral partof an irrigation system with quite a secure watersupply, the farmers in this particular canal hadnot normally received their water on time.Therefore, once the method was applied, most ofthe farmers were quite happy to pay for even lesswater, provided it was distributed equitably and ontime.


14

Wheat Yields

An accelerated process of land redistribution inKyrgyzstan has brought about both positive andnegative changes. The restructuring of agriculturehas led to increased land and water productivity,adoption of innovative approaches andconservation of costly inputs by new farmers. Atthe same time, the reforms have resulted in theemergence of a predominantly subsistencefarming system in Kyrgyzstan. Under thissystem, growing wheat and securing sufficientgrain reserves have become one of the mostimportant coping strategies for most subsistencefarmers to survive through the winter. In the studyarea, those at the tail end of canals experiencedmuch lower yields due to a lack of irrigationwater. Therefore, it was much hoped that therotational distribution of water would improve thissituation. Prior to the intervention, the wheatyields in the study offtakes 4, 5, 7, 8, 9, 10, 11,12 and 13 were 1.5-2 times lower than onaverage in the WUA, amounting to 1 t/ha(figure 4).

Improved water distribution had a positiveimpact on wheat yields at all study offtakes. Thenumber of sub-commands that had yields higherthan 2.1 t/ha (i.e., WUA average) increased from4 in 2002 to 10 in 2003. This was especially the

case with the downstream offtakes, where wheatyields in 2003 and 2004 doubled and tripled,respectively. Offtakes 4, 5 and 7, featuring one ofthe lowest yields in the WUA prior to theintervention were still yielding lower than onaverage in the WUA, suggesting that there mighthave been some other problems that were notwater-related. Such other problems could well bepoor seed quality, under-application of fertilizer aswell as poor insect and pest control.

Assessing Rotational WaterDistribution: Equity, Reliability andSensitivity

Equity of water distribution

The equity of water distribution along theSokolok Distributary was assessed bycomparing actual water withdrawals against theinitially planned targets. WUA data for 2002, theyear preceding the intervention, were cross-checked with those from the canal master’swater records. In 2001, the ratio of actual-to-planned water withdrawals for the Sokolok Canalwas 200.5 percent, which is higher than onaverage for the entire WUA “Zhapalak” (123percent). This suggests that water withdrawalsby both the study canal and the entire WUA

FIGURE 4.Wheat yields by outlets of Sokolok Canal


15

TABLE 5.Ratio of actual-to-planned water withdrawals by the offtakes of the Sokolok Distributary (% of planned withdrawals).

Distributary Offtakes Year 2002 Year 2003 Year 2004

No. 1 112.0 67.5 53.2

No. 2 705.9 51.7 60.7

No. 3 468.8 72.0 59.3

No. 4 210.5 62.0 63.3

No. 5 37.0 225.0 54.0

No. 6 71.4 30.0 52.5

No. 7 62.5 28.0 64.0

No. 8 381.8 72.5 66.0

No. 9 91.2 72.5 68.0

No. 10 34.5 26.0 58.0

No. 11 71.4 32.9 56.0

No. 12 41.7 34.0 60.0

No. 13 45.5 24.0 44.0

No. 14 228.0 62.0 87.5

Average, Sokolok Canal 200.5 46.2 60.3

Average, WUA “Zhapalak” 123.0 43.0 47.5

Difference between the sums of ratios for 4 most 1,304.1 136.3 18.5

upstream and 4 most downstream offtakes

were higher than planned. However, in no waydid it mean that all the offtakes along the studycanal received water equally or in excess ofwhat was planned. Thus, the first four upstreamsub-commands were extremely water-abundant,while those in the downstream (offtakes 9through 13) had far less water than planned(table 5). This can be clearly seen from the hugedifference between the maximum (705.9 percent)and minimum (34.5 percent) values for the ratioof actual-to-planned water withdrawals by thestudy offtakes. To show the overall water equityin the study canal, the sums of the actual-to-planned ratios for two groups of watercourseswere compared. The first group comprised of thefour most upstream watercourses while thesecond represents the four most downstreamofftakes. In the year preceding the adoption ofthe new method, the difference between thesetwo groups amounted to an incredible 1304.1percent, i.e., the four most upstream offtakeswere collectively received by 1304.1 percentmore water than the four further downstream! At

the same time, water withdrawals by the tail-most offtake (no. 14) amounted to 228.0 percentof what was initially planned due to thecontinuous flow status it enjoyed for serving theneeds of residential kitchen gardens. In 2003,the year when the new rotational distributionmethod was first introduced the gap between themaximum (225 percent) and minimum values (24percent) of actual-to-planned ratios for waterwithdrawals decreased. It should also be notedthat, in 2003, the entire water use situation bothin the whole WUA and in the study canaldramatically changed. Only 46.2 percent of theinitially planned water was actually withdrawn bythe canal, and 43 percent, in overall, by theWUA. This can be explained by the low qualityof the water planning and high precipitation(almost double of the long-term average) ratesthat occurred in the year 2003.

In 2003, the difference between the sum ofactual-to-planned ratios for water withdrawals bythe four most upstream offtakes and that for thefour most downstream sub-commands along the


16

Sokolok Distributary amounted to only 136.3percent, resulting in far more equitable waterdistribution among the offtakes. In 2004 withwater withdrawals, in overall, for the WUA being47.5 percent of what was planned, diversionsmade by the Sokolok Canal stood at 60.3 percentof the initially planned water. The differencebetween the maximum (87.5 percent) andminimum (44 percent) values of actual-to-plannedratios for water withdrawals in 2004 furtherdecreased, thus suggesting that water distributionwas more equitable.

The reason for far lesser actual waterwithdrawals against the initially planned targets in2003 and 2004 was not that there was little waterin the main canal. In 2003 and 2004, wateravailability in the Aravan-Akbura Main Canal thatsupplies water to the study WUA was 85-90percent. This suggests that the main reasons forsuch, much lower actual water withdrawals shouldbe sought inside the WUA and the study canal. Itcan, rather, be explained by well-above averagerainfalls in those two years (precipitation in 2003and 2004 was 145 and 140 percent, respectively,of the long-term average).

Reliability of water distribution

The overall reliability is the indicator, whichreflects both adequacy and timeliness of waterdistribution for the irrigated area. Bos et al. (1994)have suggested the following formula forcalculating the overall reliability of the irrigationsystem:

Overall = (Actual Volume DeliveredReliability /Planned Volume) x(OR) (Actual supply duration/

Target supply duration) (2)

The optimal value of overall reliability of thecanal (system) is one, when the irrigation canal(system) delivers planned volumes of theirrigation water for the planned durations. ORcould be measured for each outlet of the tertiarycanal. However, under rotational water distribution

only one outlet receives water at a time.Therefore, OR of the entire canal (Sokolok) doesreflect the reliability situation for each outlet. Onthe other hand, OR is a function of the operationsof higher level canals (secondary and primary),while the interventions in the study area involvedonly a tertiary canal. Nevertheless, the ORanalysis helps realizing that even with good watermanagement practices, well applied at the lowestlevels, the problems will still persist due tomalfunctions at the higher levels. Fullyacceptable water service reliability can only beachieved through water managementimprovements at all hierarchical levels.

Overall reliability for Sokolok Canal wasdetermined by comparing the expected (planned)discharge values, and irrigation supply durationagainst the actual ones. The overall reliability ofwater distribution was assessed only for 2004.For this, water discharges and duration of thewater supply in the Sokolok Canal were measuredby the field staff on a daily basis, usingcalibrated standard flumes starting from May 01,2004 through September 10, 2004.

The overall reliability of the Sokolok Canalbeing very low (0.1-0.50) for the period of Aprilthrough mid-July (figure 5), then drasticallyincreasing to 4.00 suggests that water wasdelivered in excess of what was demanded andfor longer periods than planned.

The OR once again declines at the end of thecrop season (September) down to 0.5-0.6, due tothe season ending. The OR analysis for waterservices in the pilot canal clearly suggests that itwas not optimal or high enough to be fullyacceptable. As noted above, the OR of thetertiary canal depends, among other things, onthe higher-level canals, suggesting that there isneed to improve water management at higherlevels, such as secondary canals (WUA-managed) and primary canals (managed by WaterManagement Organizations). Interventionstargeting only tertiary canals will not result indramatic improvements in irrigation waterreliability.

17

Flow stability during water distribution in theSokolok Canal

Stability analysis indicates how changes in thehigher-level canal are reflected on waterdistribution in the lower level. For example, 100percent stability means that dischargefluctuations in a canal are equally distributedamong all its offtakes. This was assessed bycomparing changes in the discharge rate of thestudy distributary canal with variations in that ofthe offtakes, within two time intervals in a dayfrom 8 am to 2 pm and from 2 pm to 8 pm.Ideally, for the rotational water distribution to beproperly implemented the flow stability should be100 percent. This would occur if the entire flow isdiverted to only one watercourse at a time.However, in 2004, when farmers managed theirwater distribution, water was at times delivered to3-4 offtakes simultaneously. As a result, the flowstability during rotational water distributiondecreased.

To perform sensitivity analysis, data on thehead discharges of the study distributary canaland those of its offtakes for the second 10-dayperiod in July 2004 were compared. The resultsof this analysis are presented in figure 6.Throughout the monitoring period, all the offtakes

showed very good daily stability against changesin the head of the distributary canal. Forexample, on July 11, 2004 the head discharge inthe Sokolok Canal at 8 am was 55 l/s, at 2 pm58 l/s and at 8 pm 57 l/s, which translates into a3 l/s or 5 percent increase between 8 am and 2pm and a one l/s or 2 percent decrease between2 pm and 8 pm. At the same time the offtakes2, 7 and 8 that were simultaneously openedduring that day showed the following changes intheir discharges: in the offtake no. 2 thedischarge from 8 am to 2 pm changed from 24l/s to 26 l/s (8 percent increase) and from 2 pmto 8 pm changed from 26 l/s to 25 l/s (4 percentdecrease). The discharge rate at offtake no. 7throughout the day remained unchanged (12 l/s)while discharges in the offtake no. 8 changedfrom 10 l/s to 11 l/s (8 am – 2 pm) and from11 l/s back to 10 l/s (2 pm – 8 pm). Thesepatterns remained consistent throughout theentire monitoring period. Most of the time, theflow stability during rotational water distributionwas very high for the offtakes with highdischarge rates, and lower for those with lowdischarges. To address this issue, the wateruser representatives made the decision not todiscriminate against the offtakes with lower

FIGURE 5.Overall reliability of Sokolok Tertiary Canal for vegetation season of year 2004.


18

Not

e: C

reat

ed f

rom

exp

erim

enta

l ob

serv

atio

ns a

nd d

ata

from

the

Sok

olok

Can

al.

FIG

UR

E 6

.A

sses

sing

sen

sitiv

ity o

f w

ater

dis

trib

utio

n in

Sok

olok

Can

al (

11/0

7/20

04 -

20/

07/2

004)

.

19

discharges. Thus, any changes in the headdischarge of the distributary canal weredistributed among the offtakes with higherdischarges. Overall, the observed flow stabilityin the study canal was quite good. For the firstfive days of the period (July 11, 2004 – July 15,2004) when the discharge rate in the distributarycanal was mainly declining, three of the fiveoperational offtakes followed suit while theremaining two had the opposite trends. Later, onJuly 15, 2004 through July 17, 2004, thedischarge trend in the distributary canal wasincreasing. However, no reaction was observedin the discharge rates of the then operationalofftakes, suggesting very low sensitivity. FromJuly 17, 2004 to July 18, 2004 with thedistributary canal flow declined, the two offtakesshowed very high stability. From July 18, 2004to July 19, 2004, the discharge trend in thedistributary canal was again on the rise, with theofftakes 3 and 14 perfectly following the pattern,thus showing very good stability. Overall, during

the 10 days of monitoring the distributary, thecanal discharge trend changed five times. Out of13 offtakes that rotated water turns during thisperiod, four offtakes were quite sensitive to anyfluctuations in the flow of the distributary canalwhile the remaining nine were quite good.

Despite relatively high water withdrawals fromthe main canal (13,000 m³/ha), water deliveries tothe field level, in most cases, amounted to lessthan 300 mm (millimeters), representing a fractionof the crop water demand. However, this amountmight seem fairly high in comparison with otherirrigated regions, such as in Australia. The majorreason for such a low water supply to the fieldsis water losses amounting to 70 percent of thetotal water withdrawals. Most of these losses,around 70 percent of them, occur due to seepagein the delivery system at all levels, while theremaining 30 percent is lost to leakages in thecanals. Overall, most water deliveries occurduring the crop season and no soil leaching orwinter irrigation are practiced in the area of study.

Conclusion and Recommendations

The preceding discussion clearly suggests thatrotational distribution results in greater equityand transparency in water supply to farmers andmost likely results in: reduced number of water-related disputes; time savings for farmers whenirrigating; and improved ISF collection from arising number of individual farmers. Besides,there is evidence that crop yields also improved.These results indicate that rotational waterdistribution could be usefully applied in a largepart of post-Soviet Central Asia, where most ofthe secondary and tertiary irrigation systemslack flow regulation structures and are poorlymaintained. While these are the explicit benefitsof the proposed method, there are also somethat are implicit.

The intervention employing rotational waterdistribution allowed improving the overall watersituation in the study Sokolok Canal. For

instance, the difference between the sum ofactual-to-planned ratios for water withdrawals bythe four most upstream offtakes and that for thefour most downstream sub-commands along theSokolok Distributary dramatically decreased from1304 percent in the year preceding theintervention to only 136 percent in the year after,resulting in far more equitable water distributionamong the offtakes. Throughout the monitoringperiod, all the offtakes showed very good dailystability against changes in the head of thedistributary canal. The water users located at thetail of the Sokolok Canal enjoyed more equitablewater distribution which allowed them to growrelatively more water-intensive crops, such aswheat. The number of those who were fully oralmost satisfied with their water distribution underthe rotational method more than doubledcompared to the previous management regime.

20

This was largely due to across-the-boardimprovements in the quality of water deliveries(reliability, timeliness and equity), especially, forthose at the tail ends of the canals. Suchimprovements come from a mixture of positivechanges that occurred in the study area, asbetter communication between the farmers, betterscheduling of the water deliveries, improved canalcondition (due to regular cleaning and simpletechnical solutions) as well as transparency andcompliance with agreed water delivery schedules.However, analysis of Overall Reliability (OR) ofthe irrigation services clearly indicated that it wasnot optimal or high enough to be fully acceptable.As noted above, the OR of the tertiary canaldepends among other things on the higher-levelcanals, suggesting that there is the need toimprove water management at higher levels, suchas secondary canals (WUA-managed) and primarycanals (managed by Water ManagementOrganizations). Interventions targeting onlytertiary canals will not result in dramaticimprovements in irrigation water reliability.

Rooted deeply in the legacy of the FSU top-down management paradigm, it is publicparticipation that is the weakest element ofirrigation management in the post-Soviet transitioneconomies, especially, in Central Asia (Ul Hassanet al. 2004). This results in a poor sense ofownership, poor cost and revenue recovery andoverall poor sustainability of local irrigationsystems. Decisions are normally made andimposed on communities using technocraticapproaches with no regard to the socio-technicalnature of irrigation systems. The findings of thisaction research suggest that water distributioncan be efficiently improved through concertedefforts and methods that put “people first.” In anirrigation management context, such efforts andmethods are most effective when they addressissues not only technically (equity, reliability,

sensitivity, timeliness, crop yields, etc.) but alsoemploy equally important social and othersometimes intangible dimensions, such asproviding effective mechanisms for buildingsustainable community-based institutions,nurturing local initiative, collaboration andcollective action, while effectively minimizing andmanaging conflict. Two years of experience haveshown that once the farmers have realized all thebenefits from an intervention, they tend to furtherrefine it more to suit their conditions. As hasbeen evident from the experience elsewhere inSouth Asia, as soon as the users come to gripswith the rotational distribution, there is potentialfor it to be gradually converted from being apurely technical solution to a localized andsustainable institution, which not only ensureseffective water distribution among communitymembers, but also improves the maintenance oftheir common property irrigation infrastructure dueto a more responsible and coordinated collectivebehavior.

This clearly suggests that it is the needs andconcerns of the end users that provide a goodentry point to pragmatically understand andanalyze the situation, from where appropriateremedial strategies and methods can be furtherdevised and employed. This is also a goodstarting point to initiate farmers’ debates anddiscussions on public participation, which shouldultimately lead to and end up in a truly farmer-owned process and action.

In summary, the experiment hasdemonstrated that interventions aiming to improvewater distribution at the tertiary level cansucceed, if carefully planned and implemented inclose consultation with and full participation of thebeneficiaries themselves from the very beginning,by addressing their real needs and problems andnurturing their own initiative, collective action andinstitutions.

21

Literature Cited

Abernethy, C. L. 1988. The concept of flexibility in irrigation systems. Paper presented at the IrrigationManagement Conference, Wuhan, China.

Bandaragoda, D. J.; and Ur Rehman, S. 1995. Warabandi in Pakistan’s canal irrigation systems: Widening gapbetween theory and practice. Pakistan Country Paper No. 7. Colombo, Sri Lanka. International IrrigationManagement Institute.

Bartold, V. V. 1970. Irrigation of the Central Asia. Nauka. Moscow. 235 pp.

Bastiaanssen, W.; and Bos, M. G. 1999. Irrigation performance indicators based on remotely sensed data: A reviewof literature. Irrigation and drainage systems. 13: 291-311.

Bos, M. G.; and Nugteren, J. 1974. On irrigation efficiencies. Publication no.19. Wageningen, The Netherlands.ILRI (International Institute for Land Reclamation and Improvement). 138 pp.

Bos, M. G.; Burton, M. A.; and Molden, D. J. 2005. Irrigation and drainage performance assessment: Practicalguidelines. CABI Publishing. 158 pp.

Bos, M. G.; Murray-Rust, D. H.; Merrey, D. J.; Johnson, H. G.; and Snellen, W. B. 1994. Methodologies for assessingperformance of irrigation and drainage management. Irrigation and drainage systems. 7: 231-261.

Burt, C. M.; Clemmens, A. J.; Strelkoff, T. S.; Solomon, K. H.; Bliesner, R. D.; Hardy, L. A.; Howell, T. A.; andEisenhauer, D. E. 1997. Irrigation performance measures: efficiency and uniformity. Journal of irrigation anddrainage engineering. 123(6): 423-442.

Chaudhry, M. A.; and Young, R. A. 1989. Economic impacts of alternative irrigation water allocation institutions:Pakistan’s warabandi system. In: R. K. Sampath and R. A. Young (eds). Social, economic and institutionalissues in Third World irrigation management. Papers in Water Policy and Management, No. 15. Westview Press.

Horst, L. 1987. Choice of irrigation structures: The paradox of operational flexibility. In: Irrigation design formanagement. Proceedings of a Workshop held at Kandy, Sri Lanka. HR Wallingford and International WaterManagement Institute.

Horst, L. 1990. Interactions between technical infrastructure and management. Network paper 90/3b, IrrigationManagement Network. London: Overseas Development Institute.

Horst, L. 1998. The dilemmas of water division: Considerations and criteria for irrigation system design. Colombo,Sri Lanka. International Water Management Institute. 123 pp.

IFPRI (International Food Policy Research Institute). 1994. Water resources allocation: Agricultural productivity andenvironmental impacts (mimeo).

IWMI (International Water Management Institute). 2004. Assessment of water use and management in Water UsersAssociation “Zhapalak”. Project report. Integrated Water Resources Management in Ferghana Valley. Tashkent.23 pp.

Jensen, M. E. 1977. Water conservation and irrigation systems. Climate-Technology Seminar Proceedings,Columbia, Missouri.

Latif, M.; and Sarwar, S. 1994. Proposals for equitable water allocation for rotational irrigation in Pakistan. Journalof irrigation and drainage Systems. 8: 35-48.

Levine, G. 1982. Relative water supply: an explanatory variable for irrigation systems. Technical report no. 6.Ithaca, New York, USA. Cornell University.

Lowdermilk, M. K.; Clyma, W.; and Early, A. C. 1975. Physical and socio-economic dynamics of watercourses inPakistan’s Punjab: System constraints and farmer’s responses. Water Management Technical Report No. 42.Fort Collins: Colorado State University.

22

Makin, I. W. 1987. Warabandi in Punjab, India. Oxfordshire, UK. Hydraulic Research Limited, Wallingford.

Merrey, D. J. 1990. Warabandi in Pakistan. Chap. 28 in Management of farm irrigation Systems. St. Joseph,Michigan: American Society of Agricultural Engineers.

Molden, D. J. 1997. Accounting for water use and productivity. SWIM paper 1. Colombo, Sri Lanka. InternationalWater Management Institute. 16 pp.

Mukhamedjanov, M. 1986. History of the irrigation of the Samarqand oasis. Fan. Tashkent. 231 pp.

Mukhamedjanov, S. Sh.; Galustian, A.; and Nerozin, S. 2004. The results of improving irrigation water use. SICICWC Project report - Integrated Water Resources Management in Ferghana Valley. Tashkent. 58 pp.

Murray-Rust, D. H.; and Snellen, W. B. 1993. Irrigation system performance assessment and diagnosis. Joint IIMI(International Irrigation Management Institute), ILRI (International Institute for Land Reclamation andImprovement) and IHE (International Institute for Hydraulic and Environment Engineering) publication.Colombo, Sri Lanka. International Irrigation Management Institute. 148 pp.

Qureshi, K. S.; Hussain, Z.; and Zeb-un-Nisa, H. 1994. An assessment of warabandi (irrigation rotation) inPakistan: A preliminary analysis. Islamabad, Pakistan. Pakistan Institute of Development Economists.

SIC ICWC (Scientific-Information Center of the Interstate Coordination Water Commission of the Central Asia).2003. Report on water productivity studies of IWRM-Ferghana project. SIC ICWC Project report - IntegratedWater Resources Management in Ferghana Valley. Tashkent. 27 pp.

Singh, K. K. ed. 1981. Warabandi for irrigated agriculture in India. Publication no. 146. New Delhi: Central Boardof Irrigation and Power, India.

Small, L. E.; and Svendsen, M. 1990. A framework for assessing irrigation performance. Irrigation and drainagesystems. 4: 283-312.

Taylor, P. 2001. Some principles for development and implementation of water-allocation schemes. In: Brennan,D. ed. Water Policy reforms: Lessons from Asia and Australia. Proceedings of an international workshop heldin Bangkok, Thailand, 8-9 June 2001. pp. 62-74.

Thurman, M. 1998. The history of irrigation in Ferghana Valley. Draft of paper. 123 pp (unpublished).

Ul Hassan, M.; Starkloff, R.; and Nizamedinkhodjaeva, N. 2004. Inadequacies in the water reforms in the KyrgyzRepublic: An institutional analysis. IWMI Research Report 81. Colombo, Sri Lanka. International WaterManagement Institute.

Uphoff, N. 1986. Improving international irrigation management with farmer participation: Getting the process right.Boulder, CO. Westview Press.

Vehmeyer, P. W. 1992. Irrigation management strategies at the farm level in a warabandi schedule. UnpublishedM.Sc. Thesis, University of Delft, the Netherlands.

Wolters, W. 1992. Influences on the efficiency of irrigation water use. Ph.D. Thesis, Delft University of Technology.150 pp.

87. Economics and Politics of Water Resources Development: Uda Walawe IrrigationProject, Sri Lanka. François Molle and Mary Renwick. 2005.

88. "Bright Spots" in Uzbekistan, Reversing Land and Water Degradation WhileImproving Livelihoods: Key Developments and Sustaining Ingredients forTransition Economies of the former Soviet Union. Andrew Noble, Mohammed ulHassan and Jusipbek Kazbekov. 2005.

89. Planning for Environmental Water Allocations: An Example of Hydrology-basedAssessment in the East Rapti River, Nepal. V. U. Smakhtin and R. L. Shilpakar.2005.

90. Working Wetlands: Classifying Wetland Potential for Agriculture. Matthew P.McCartney, Mutsa Masiyandima and Helen A. Houghton-Carr. 2005.

91. When “Conservation” Leads to Land Degradation: Lessons from Ban Lak Sip,Laos. Guillaume Lestrelin, Mark Giordano and Bounmy Keohavong. 2005.

92. How Pro-Poor are Participatory Watershed Management Projects?—An IndianCase Study. Mathew Kurian and Ton Dietz. 2005.

93. Adoption and Impacts of Microirrigation Technologies: Empirical Results fromSelected Localities of Maharashtra and Gujarat States of India. Regassa E.Namara, Bhawana Upadhyay and R. K. Nagar. 2005.

94. Balancing Irrigation and Hydropower: A Case Study from Southern Sri Lanka.François Molle, Priyantha Jayakody, Ranjith Ariyaratne and H.S. Somatilake.2005.

95. Irrigation and Water Policies in the Mekong Region: Current Discourses andPractices. François Molle. 2005.

96. Locating the Poor: Spatially Disaggregated Poverty Maps for Sri Lanka. Upali A.Amarasinghe, Madar Samad and Markandu Anputhas. 2006.

97. Strategies to Mitigate Secondary Salinization in the Indus Basin of Pakistan: ASelective Review. M. Aslam and S. A. Prathapar. (not published yet) 2006.

98. Multiple-Use Water Services to Advance the Millennium Development Goals.Barbara van Koppen, Patrick Moriarty and Eline Boelee. 2006.

99. Irrigation and Schistosomiasis in Africa: Ecological Aspects. Eline Boelee andHenry Madsen. 2006.

100. The Reliability Improvement in Irrigation Services: Application of Rotational WaterDistribution to Tertiary Canals in Central Asia. Iskandar Abdullaev, Mehmood UlHassan, Herath Manthrithilake and Murat Yakubov. 2006.

Research Reports

SM

IWMI is a Future Harvest Centersupported by the CGIAR

Postal AddressP O Box 2075ColomboSri Lanka

Location127, Sunil MawathaPelawattaBattaramullaSri Lanka

Telephone+94-11-2787404

Fax+94-11-2786854

[email protected]

Websitehttp://www.iwmi.org

I n t e r n a t i o n a lWater ManagementI n s t i t u t e

ISSN 1026-0862ISBN 92-9090-639-1

978-92-9090-639-1ISBN

The Reliability Improvement in Irrigation Services: Application of Rotational Water Distribution to Tertiary Canals in Central Asia 100 RESEARCH REPORT

Documents