1 River Basin Water Allocation Planning: Simulation--Optimization Approach Simulation Mojtaba Shourian Water and Power University of Technology, Water Research Institute, Ministry of Energy, Tehran, Iran m_shourian@yahoo.com 2 Outlines 1- Introduction 2- MODSIM, River Basin Simulation Model 3- PSO, Heuristic Optimization Algorithm 4- Water Allocation Planning Using PSO--MODSIM Model, Results PSO 5- Conclusion 3 Introduction 4 Introduction • The task of IWRM IWRM:: development of possible water resource system designs or management plans and evaluation of their environmental, economic and social impacts • IWRM evaluation is complex because of: • Large Scale Problems • Dynamic Variables • Surface & Groundwater Resources Quantity & Quality Issues • Multi Multi--Objective Purposes • Socio Socio--Economic Issues 5 Introduction • To do so, there is no way than modeling the performance of water resources systems • Simulation is perhaps the most widely used method for evaluating alternative water resources systems with defined set of design and operational policies • e.g e.g.: .: HEC 5, HEC ResSim, ResSim, RIBASIM RIBASIM,, MikeBasin MikeBasin,, WEAP,, RiverWare and MODSIM WEAP • Simulation models are not able to identify optimal design and operation of the components of a river 6 basin system Introduction • Simulation models are not able to identify optimal design and operation of the components of a river basin system Need of Optimization Models Simulation--Based Optimization is an alternative • Simulation method in solving large scale river basin problems • MODSIM (Labadie, 1970) 1970), a generic river basin simulation model (DSS) under the .NET framework with the ability of customization embedded in the Particle Swarm Optimization (PSO) algorithm (Kennedy & Eberhart, 1995 1995)), as the optimization 7 procedure Introduction • The resulting PSO PSO--MODSIM model has been used in analyzing water resources development plans in the Sirvan river basin In Iran • The model’s decision variables are the sizes of planned dams, transfer and pumping systems as well as the relative priorities of reservoirs target storage volumes • The model solves the problem of optimum water allocations between competing demands inside and outside of the basin 8 MODSIM, River Basin Simulation Model 9 MODSIM • MODSIM, a generalized river basin network model, was originally developed by Dr. John Labadie of Colorado State University in the midmid-1970 • MODSIM employs a statestate-ofof-the the--art network flow programming algorithm • Water is allocated according to physical, hydrological, and institutional aspects of river basin management • MODSIM has been successfully applied to a number of complex river basin systems 10 MODSIM Custom Coding Graphical User Interface (GUI) Generating Reports from Outputs 11 MODSIM • Minimum Cost Network Flow Program (NFP): Priority--Based Solution Priority • Network Structure in MODSIM: 12 Particle Swarm Optimization 13 PSO • The PSO algorithm facilitates simple rules simulating bird flocking and serves as an optimizer for nonlinear functions • formed by a set of particles, representing potential solutions for a given problem • particles fly through the problem space by following its own experience and the best experience attained by the swarm as a whole 14 PSO • Each particle has a position vector xi, a velocity vector vi and the position of the best candidate solution encountered by the particle pi. The model also maintains the overall best found point pg, which also attracts PSO particles • Changing Location of a Particle 15 Integrating PSO & MODSIM: the PSOPSO-MODSIM model 16 PSO-MODSIM PSO Algorithm: Optimizing decision variables like reservoir volumes, water transfer systems capacities MODSIM: Simulating the river basin with the set of variables proposed by PSO Calculating the objective function: Net benefit of water quantity supplied to different demand sites PSO--NFP Algorithm PSO 17 Water Allocation Planning Using PSOPSO-MODSIM • Case Study Simulating in MODSIM • Objective function: net benefit of design and operation = benefit of supplying water for different demands – cost of construction and operation of plans 18 Sirvan river basin modeled in MODSIM 19 PSO--MODSIM: Results PSO 20 PSO--MODSIM : Results PSO Variation of the best objective function Ravansar-Gavoshan transfer capacity Convergence of the objective function Convergence of Gerdalan reservoir volume 21 FLow (MCM) 80 70 Demand 60 Supply PSO--MODSIM : Results PSO 50 40 In-Basin annually average water supply 30 20 10 120 0 2 3 4 5 6 7 Month 8 9 10 11 12 450 400 350 FLow (MCM) Demand 100 FLow (MCM) 1 Supply 80 60 40 300 20 250 200 0 1 150 2 3 4 5 6 7 8 9 10 11 12 Month 100 Water Supply for Karkheh basin 50 0 1 2 3 4 5 6 7 Month 8 9 10 11 12 Basin Out Flow 7 Discharge (cms) 6 5 4 3 2 1 0 0 10 20 30 40 50 60 70 80 90 100 Variation of storage volume in Gerdalan reservoir Exceeding Probability (%) Duration curve of flow in Banidar tunnel 22 Conclusion • A simulationsimulation-based optimization model is developed to find efficient development plans and optimum water allocations at basin scale • The general river basin network flow model, MODSIM, as the simulation engine, was embedded in the PSO, which is an evolutionary optimization algorithm • The resulting PSOPSO-MODSIM model was used to propose the optimum sizes of the planned water storage and transfer systems in the basin considering the coordinated system operations 23 Conclusion • The proposed methodology presents various useful capabilities in decision supporting systems for decision making in large scale river basin management problems • Use of abilities of various models in a platform linking them 24 Thank You For Your Attention 25