Mixed-integer Programming of Stochastic Hydro Self-scheduling Problem in Joint Energy and Reserves Markets

This article presents a stochastic model for self-scheduling of a purely hydro generating company participating in the day-ahead joint energy and reverse market aimed at maximizing the profit. In the proposed framework, the mixed-integer non-linear programming of the hydro self-scheduling problem is converted into mixed-integer programming. The hydro self-scheduling problem with uncertainties, such as outages of generation units and volatile market prices, is modeled as a stochastic mixed-integer programming problem. Random 24-hr scenarios are generated using Roulette wheel mechanism and lattice Monte Carlo simulation methods. Afterward, the proposed stochastic model is decomposed into deterministic optimization sub scenarios. For great accuracy, multi-performance curves are employed in the modeling stage. Moreover, wide-ranging parameters are taken that result in an absolutely real outcome, e.g., head dependent reservoirs, start-up cost, water discharge limits, initial and final volume, water balance constraints, spinning and non-spinning reserve, spillage of reservoir, and a cascade hydro plant. Finally, for a test system with eight hydro plants, the bilateral contracts are evaluated in terms of performance indices to demonstrate the effectiveness of proposed model.