Role of sediment transport in operation and maintenance of supply and demand based irrigation channels : application to Machai Maira Branch canals

Like in many emerging and least developed countries, agriculture is vital for Pakistan’snational economy. It contributes 21% to the annual gross domestic product (GDP),engages 44% of total labour force and contributes 60% to the national export. Pakistanhas a total area of 80 Mha (million hectares) with 22 Mha arable land, out of which 17Mha is under irrigation, mostly under canal irrigation. Due to the arid to semi-aridclimate, the irrigation is predominantly necessary for successful crop husbandry inPakistan.The development of modern irrigation in Indo-Pakistan started in 1859 with theconstruction of the Upper Bari Doab Canal on Ravi River and with the passage of timethe irrigation system of Pakistan grew up to the world’s largest contiguous gravity flowirrigation system, known as the Indus Basin Irrigation System (IBIS). In the IBIS almostall irrigation canals are directly fed from rivers, while river flows carry heavy sedimentloads. Irrigation canals receiving such flows get massive amounts of sediments, whichare then deposited in the irrigation canals depending upon the hydrodynamic conditionsof the canals. Sediment deposition in irrigation canals causes serious operation andmaintenance problems. Studies reveal that silt reduces up to 40% of the availabledischarge in irrigation canals.Researchers have been striving since long to manage this problem in a sustainableway and a number of approaches have been introduced in this connection. As a first stepsediments are controlled at river intakes by silt excluders and ejectors. Then a canaldesign approach is adopted for keeping sediments in suspension and to distribute themas much as possible on the irrigated fields. Even then sediments tend to deposit inirrigation canals and become a serious problem in canal operation and maintenance,which then requires frequent desilting campaigns to keep water in the canals running. Itcauses a continuous burden on the national economy. In emerging and least developedcountries, adequate and timely availability of funds for operation and maintenance isgenerally a problem. It causes delays in canal maintenance, which affects their hydraulicperformance. Water is then delivered inadequately and inequitably to the water users.The story becomes further complicated when it comes to downstream controlleddemand based irrigation canals under flexible operation. In fixed supply based operation,canals always run at full supply discharge and such operation, generally, does not allowsediment deposition in the canal prism due to sufficient velocities. Whereas in demandbased flexible operation the canals cannot run always at full supply discharge but insteadthe discharge is changing depending upon the crop water requirement in the canalcommand area. Such type of canal operation is not always favourable to sedimenttransport as under low discharges, flow velocities fall quite low and hence sedimentdeposition may occur in the canal prism. The questions arise here what sort ofhydrodynamic relationships prevent sediment deposition in downstream controlledirrigation canals and how these relationships can be adopted, while catering crop waterrequirements of the command area? How the maintenance needs can be minimized bymanaging sediment transport through better canal operation?This study has been designed to investigate such type of relationships and practicesin order to manage sediment transport in downstream controlled demand based irrigationcanals and to attain maximum hydraulic efficiency with minimum maintenance needs.The hypothesis of the study states that in demand based irrigation canals the volume ofsilt deposition can be minimized and even the sediments which deposit during low cropwater requirement periods can be re-entrained during peak water requirement periods. Inthis way a balance can be maintained in sediment deposition and re-entrainment byadequate canal operation.Two computer models have been used in this study, namely, Simulation ofIrrigation Canals (SIC) and SEdiment TRansport in Irrigation Canals (SETRIC). Bothmodels are one-dimensional and are capable of simulating steady and unsteady stateflows (SETRIC only steady state flows) and non equilibrium sediment transport inirrigation canals. The SIC model has the capability to simulate sediment transport underunsteady flow conditions and can assess the effect of sediment deposition on hydraulicperformance of irrigation canals. Whereas the SETRIC model has the advantage oftaking into account the development of bed forms and their effect on resistance to flow,which is the critical factor in irrigation canal design and management. In the SETRICmodel, a new module regarding sediment transport simulations in downstreamcontrolled irrigation canals has been incorporated.The study has been conducted on the Upper Swat Canal – Pehure High Level Canal(USC-PHLC) Irrigation System, which consists of three canals, Machai Branch Canal,PHLC and Maira Branch Canal. The Machai Branch Canal has upstream controlledsupply based operation and the two other canals have downstream controlled demandbased operation respectively. These canals are interconnected. The PHLC and MachaiBranch canals feed Maira Branch Canal as well having their own irrigation systems.PHLC receives water from Tarbela Reservoir and Machai Branch Canal from the SwatRiver through USC. Water from Tarbela Reservoir, at present, is sediment free, whereasthe water from Swat River is sediment laden. However, various studies have indicatedthat soon Tarbela Reservoir will be filled with sediments and will behave as run of theriver system. Then PHLC will also receive sediment laden flows. The design dischargesof Machai, PHLC and Maira Branch canals are 65, 28 and 27 m3/s respectively. Thecommand area of the USC-PHLC Irrigation System is 115,800 ha.The USC-PHLC Irrigation System has been remodelled recently and waterallowance has been increased from 0.34 l/s/h to 0.67 l/s/h. The upper USC system, fromMachai Branch head to RD 242 (a control structure from where the downstream controlsystem starts), was remodelled in 1995, whereas the system downstream of RD 242 wasremodelled in 2003. The upper part of Machai Branch Canal up to an abscissa of about74,000 m is under fixed supply based operation, whereas the lower part of MachaiBranch Canal, Maira Branch Canal and the PHLC are under semi-demand based flexibleoperation. The semi-demand based system is operated according to crop waterrequirements and follows a Crop Based Irrigation Operations (CBIO) schedule. Whenthe crop water demand falls below 80% of the full supply discharge, a rotation system isintroduced among the secondary offtakes. During very low crop water requirementperiods the supplies are not reduced beyond a minimum limit of 50% of the full supplydischarge because of the canal operation rule.The study consisted of fieldwork of two years in which daily canal operation data,monthly sediment inflow data in low sediment periods and weekly sediment data in peakconcentration periods were collected. Three mass balance studies were conducted inwhich all the water and sediment inflows and outflows were measured with suspendedsediment sampling at selected locations along the canal and boil sampling at theofftaking canals, immediately downstream of the head regulators. Further in the fourmonths during the peak sediment season June, July, August and September, massbalance studies were conducted by boil sediment sampling in order to estimate water andsediment inflow to and outflow from the system. To determine the effect of sedimenttransport on the canals’ morphology, five cross-sectional surveys were conducted andchanges in bed levels were measured. On the basis of these field data the two computermodels, used in this study, were calibrated and validated for flow and sediment transportsimulations.The downstream control component of the system is controlled automatically andthe PHLC has been equipped with the Supervisory Control and Data Acquisition(SCADA) system at the headworks. Any discharge withdrawal or refusal by WaterUsers Associations (WUA) through offtaking secondary canals, or any dischargevariation in the inflow from Machai Branch Canal is automatically adjusted by theSCADA system at Gandaf Outlet, the PHLC headworks. The SCADA system hasProportional Integral (PI) discharge controllers. The study found that the existing PIcoefficients led to delay in discharge releases and resulted in a long time to achieve flowstability. The discharge releases showed an oscillatory behaviour which affected thefunctioning of hydro-mechanically operated downstream control “Aval Orifice” (AVIO)and “Aval Surface” (AVIS) gates. After calibration and validation of the model the PIcontrollers were fine-tuned and proposed for improved canal operation, which wouldhelp in system sustainability and in improved operational efficiency of the canals.Field data show that during the study period sedimentation in the studied irrigationcanals remained within control limits. The incoming sediment loads were, generally,lower than the sediment transport capacities of the studied irrigation canals. Hence thisincoming sediment load was transported by the main canals and distributed to theofftaking canals. The sediment transport capacities of the studied irrigation canals werecomputed at steady and unsteady state conditions. The canal operation data showed thatthe system was operated on Supply Based Operation (SBO) approach rather than CBIO.The morphological data revealed that there was no significant deposition in the studiedcanals. Therefore there was no particular effect on the canal operation and the hydraulicefficiencies, attributed to sediment transport.As mentioned earlier, the Tarbela Reservoir will soon be filled with sediments andconsequently PHLC will get sediment laden flows from the reservoir. Various studieshave been taken into account to project the time when sediment laden flows will flowinto the PHLC and what will be the characteristics and concentrations of the incomingsediments to the PHLC from the reservoir. The studies project that the sediment inflowfrom the Tarbela Reservoir will be much higher than the sediment transport capacities ofthe PHLC and Maira Branch Canal under full supply discharge conditions. This scenariowill create sediment transport problems in downstream controlled canals, particularlywhen they will be operated under CBIO.Various management options have been simulated and are presented in order tobetter manage sediments in the studied canals under the scenario of sediment inflowfrom Tarbela Reservoir. The hydraulic performance of downstream controlled canalswill be affected under this scenario and frequent maintenance and repair will be requiredto maintain the canals. Various options have been analysed to deal with the problem.The study presents a sediment management plan for downstream controlled irrigationcanals by improvements in canal design and operation in combination with the need ofsettling ponds at the canal headworks.Currently sedimentation in the irrigation canals under study is not a big issue forcanal operation and maintenance (O&M). However, it would emerge as a major problemwhen sediment discharge from the Tarbela Reservoir starts. The canals’ maintenancecosts will soar and the hydrodynamic performance of these canals will also be affected.In this study, a number of ways have been evaluated and proposed to deal with theapproaching problem of sediment transport in these irrigation canals in order to keeptheir hydraulic performance at desired levels and to minimize the maintenance costs.The first and the foremost effect of sediment deposition will be reduction in canals’ flowconveyance capacities, which will result in raise of water levels. The raise of waterlevels will cause a reduction in water supply to the canals due to automatic flow releases.It can be dealt with by a temporary and limited raise in target water levels dependingupon the maximum headloss at the downstream AVIS/AVIO cross regulator. Further, tominimize the effect of water level raise on discharge through the AVIS/AVIO gates, thedecrement in such canals can be kept relatively small, in order to make the gates lesssensitive to water level changes. Further, for efficient withdrawal of sediment to thesecondary canals, it is needed to locate the secondary offtakes close to AVIS/AVIOcross regulators on the downstream side. More sediment will be discharged because theturbulent mixing of sediment at the downstream side of the control structures keepsmore sediment in suspension. In addition, during the peak sediment concentrationperiods, the canals need to be operated at supply based operations, in order to minimizethe deposition.Sediment transport in general and in irrigation canals in particular, is one of themost studied and discussed topic in the field of fluid mechanics all over the world. Italso has been studied extensively in Indus Basin in order to design and manage irrigationcanals receiving sediment laden flows. The outcome of Lacey’s regime theory and thesubsequent work are the result of these studies. In addition to regime method variousother methods like permissible velocity method, tractive force method and the rationalmethods, etc., have been developed for stable canal design. Anyhow, as a matter of fact,the management of sediment transport in irrigation canals is still a challenging task evenafter all these investigations and studies. Because most of the knowledge on sedimenttransport is empirical in nature, most sediment transport formulae have inbuiltrandomness, which makes predictions difficult, when conditions are changed. It needs alot of care while applying a sediment transport formula, developed under one set ofconditions, to other situations. Therefore, it becomes extremely important to understandthe origin of the development of the formulae and the limitations associated with thembefore applying some sediment transport formulae to different conditions andcircumstances. The introduction of numerical modelling made it comparatively easy totest and shape the sediment transport relationships to some local conditions by running avariety of simulations and calibrating the formula in light of the field measurements. Thesediment transport predictions can be made reliable in this way and can be used forfurther analysis.