Generation Scheduling of a Hydrodominated Provincial System Considering Forecast Errors of Wind and Solar Power

Abstract

The integration of large-scale uncontrollable renewable power greatly affects the security and reliability of power system operations. An effective way of coping with this challenge is to employ traditional dispatchable generation to buffer the uncontrollable power. This study focuses on the coordinated operation of a hydro-wind-solar system. The Yunnan power grid with 62.4 GW hydropower capacity and 10.8 GW wind and solar power capacity is selected as an example. A methodology for the generation scheduling of a hydrodominated provincial system with large-scale wind and solar power is developed. This methodology introduces additional positive reserve, negative reserve, and ramping response constraints to handle the forecast errors of wind and solar power. Moreover, the methodology merges renewable power into the original loads to determine an equivalent load curve for hydropower plants. An optimization model with a peak-shaving objective is formulated and solved by a two-phase approach. The first phase uses knowledge rules to check and adjust the generation schedules of hydropower plants with predetermined schedules or dispatching modes. The second phase presents a multidimensional search method to optimize other major plants to respond to peak loads. The methodology is demonstrated by day-ahead scheduling of the given provincial system, which includes 397 plants. The obtained generation schedules provide a reasonable reserve capacity for buffering wind and solar power fluctuations and meet the peak demands of the power grid. A detailed sensitivity analysis for different dates is presented to illustrate the seasonal characteristics of renewable power.