A Scalable Mixed-Integer Programming Model for Multi-Period Hydropower Planning

Fatemeh Karamyar

In this talk, I present a tractable yet realistic optimization framework for multi-period, long-term water release scheduling that optimizes hydropower production in reservoirs. The model integrates a nonlinear hydropower production function with key operational constraints and decision-informed policies. The approach discretizes water levels, simplifies nonconvex dynamics, and prescribes easyto-interpret operational to address uncertainty and make trade-offs between power generation, flood control, and ecological needs. Key features include a head-dependent hydropower function, upstream flow travel time, smoothing constraints on water release, and a turbine switch-off policy. The model is evaluated using real data from the Lower Mekong Basin.

Short Bio

Fatemeh Karamyar holds a Ph.D. in Industrial Engineering from Arizona State University. Her expertise lies in optimization, broadly defined, with a focus on challenges at the intersection of scheduling, network optimization, and integer programming. Her work applies advanced optimization techniques to real-world systems in production planning, water and energy systems, and sustainable infrastructure design in public policy and urban planning. Her research is interdisciplinary with peer-reviewed publications and submitted papers co-authored alongside researchers in hydrology, transportation, public health, data science, and economics. Her current research focuses on the design of resilient systems that sustain critical operations under adverse uncertain circumstances. Specifically, she is developing models for multi-reservoir hydropower planning, electric bus scheduling and charging, and is coauthoring a comprehensive review of dual-resource scheduling problems. In addition to her research, Fatemeh has supported instruction in operations research and statistics, and she is currently teaching engineering economics in Industrial Engineering department at Sharif University of Technology.

Venue

Friday, October 17, 2025, 4.00 pm

Online