A hydropower plant scheduling model considering the stochastic correlation between runoff and electricity prices

Abstract

Evaluating and addressing the risks posed by runoff and electricity prices in hydropower plants involved in medium-to long-term scheduling is a pressing issue. To address this， a scheduling model is proposed that aims to maximize expected net revenue while minimizing medium- to long-term operational risks. First， using the conditional value at risk （CVaR） theory， marginal distribution functions are utilized to characterize the risks of runoff uncertainty and electricity price volatility （EPV）， enabling accurate risk assessment for market-operated hydropower plants. Second， the least-squares cross validation （LSCV） is applied to determine the optimal bandwidth parameter in kernel density estimation （KDE）， ensuring a good fit for discrete runoff and electricity price data. Next， a Copula-Monte Carlo simulation method is used to model the joint risks of runoff and electricity prices， with Latin hypercube sampling （LHS） employed to enhance computational precision. Finally， case simulation and analysis are conducted to validate the effectiveness of the proposed model.

Keywords

• hydropower plant
• optimal scheduling
• runoff
• cvar
• kde