Data-Driven Optimization of Energy Storage Allocation for Stable Outward Delivery in Wind and Solar Enrichment Region-A Case Study of Provincial Power Grid

Abstract

The volatility and uncertainty of renewable energy sources aggravate the difficulties of power consumption balance in the new power system. One practical approach is to equip the grid with sufficient energy storage to reduce the operation risk. This paper focuses on a provincial grid with a heavy power outward delivery burden and its storage allocation problem. First, the optimization model with minimum storage allocation is established. Scenarios are constructed based on the typical set and the validation set. Second, a data-driven method is applied to build the typical set by k-means++. To describe the tolerance of wind and solar curtailed and delivery shedding, a chance-constrained approach is established and converted to linear constraints by big-M. The case study shows the power system needs considerable energy storage to ensure renewable energy consumption and delivery. With the lower tolerance, the allocated storage capacity will increase, and the lowest feasible tolerance is limited by scenarios with poor renewable output.