SIMULATION STUDY OF AUTOMATIC REGULATION OF CANAL

Abstract

Supervisory control is a control system in which system information is periodically monitored and globally analyzed to decide the control action. In a multi-reach, multi-offtake canal system with changing lateral flow conditions, such control system coordinates the gate movements and provides more stable control. It produces simultaneous and instantaneous gate response to flow changes. Such gate response stabilizes the water levels within a shorter time than that in local control methods. Supervisory control involves comprehensive detailed analysis of canal system for which necessary software need to be developed. In this study a supervisory control regulation methodology (SCRM) has been proposed and the necessary algorithms have been developed. The method provides water supply on demand without any condition on advance information about the time and amount of lateral flow discharges. The method is based on multiple water level measurements only, without discharge measurements. The typical canal system under consideration is the storage backed, prismatic, straight, multi-reach multi-offtake irrigation canal with vertical control gates. Also, this study explores the canal design aspects for automation and the usefulness of the proposed simulation algorithms in design by simulation. The necessary algorithms for the proposed Supervisory Canal egulation Methodology (SCRM) are; i)the INITIAL CONDITION ALGORITHM; Ithe algorithm 'COMPARATOR'; iii)the algorithm 'CONTROLLER'; (1) iv)Unsteady Flow Simulation Algorithm; and v)Daily Demand Simulation Algorithm. Initial Condition Algorithm : This algorithm is developed for i)Updating gate discharge coefficients and gate discharges as per the sensed water depths and gate openings; ii)Calculating current pool volumes as a function of the sensed water depths; iii)Keeping all information up-to date and available for use in other algorithms; and iv)Activating the algorithm 'COMPARATOR' for further analysis. Location of water level sensors are designed to provide accurate estimation of the pool volumes in the presence of lateral canals along the pools. The study shows that the sensor locations should be located at pool ends as well as at the computational sections upstream and downstream of every lateral canal offtaking point. Algorithm 'COMPARATOR' : The primary function of this algorithm is to compare the existing pool volumes (calculated by the Initial Condition Algorithm) with the target, maximum, and minimum pool volumes. Output of the algorithm are the deviation errors. These deviations are required by the algorithm 'CONTROLLER' for calculating the control actions (gate movements). This function of the comparator is repeated every time step 'AT' during on-line operation. The 'COMPARATOR' also calculates, every 'AT", the current limiting feasible gate movements as well as pool volume loss due to seepage and evaporation to be utilized by the 'CONTROLLER'. Target pool volume (setpoint) and deadband are calculated and updated periodically to ensure that water will always be available for lateral canals on-demand. On fortnight basis, based on daily demand simulation and expected delay time, the algorithm 'COMPARATOR' calculates the wedge storage requirements and then the target pool volumes and deadbands. The algorithm 'COMPARATOR' calls the Daily Demand Simulation Algorithm for simulating the lateral demands. Algorithm 'CONTROLLER' : This control algorithm is the heart of the control process. It is designed to perform the following functions; 1. Activating the Emergency Condition Algorithm to detect and deal with the emergency conditions; 2. Computing new gate discharges based on the above computed pool volume deviation errors, satisfying requirements of; i) maintaining the target pool volumes; ii) maintaining the canal safety against overtopping; iii) maintaining the water levels above the minimum required water levels (MRLs) at offtake locations, by maintaining the target pool volumes; iv) maintaining minimum hydraulic transient, i.e. minimum stabilizing time of water levels, pool volumes, and gate movements, by ensuring instantaneous and simultaneous response of control gates; v) maintaining gate speeds and openings within specified practical limits; and vi) avoiding overcorrection by including anti-hunting and deadband considerations. 3. Updating the gate discharge coefficients and computing new gate openings accordingly, based on the current water depths and the required gate discharges. 4. Examining and filtering the calculated gate openings for practicability, based on the limiting gate speeds (maximum and minimum) as well as the current sensed gate openings and water levels. Unsteady Flow Simulation Algorithm : This algorithm is developed to simulate the unsteady flow behavior in multi-reach, multi-offtake, irrigation canals. St. Venant's unsteady flow equations are solved by using Direct Implicit Finite Difference Method. Four-point implicit scheme is used. Newton's iterative procedure is used to linearize the resulting set of equations. The solution algorithm integrates the coefficient matrices of the different pools in one global matrix for the entire multi-reach canal. The solution algorithm is oriented to assign the matrix elements close to the diagonal, thus obtaining a diagonally banded matrix and sparse. The matrix solver subroutine is developed to make use of this matrix property in reducing the computational time. Boundary conditions are developed for three modes of simulation; 1) free wave propagation; ii) response of water levels to lateral flow changes by using the control gate discharges as boundary conditions; and iii) response of water levels to lateral flow changes by using the control gate openings as boundary conditions.....