Data-Driven Scheduling of Electric Boiler with Thermal Storage for Providing Power Balancing Service

TL;DR

This paper proposes a data-driven stochastic optimization method for electric boilers with thermal storage, improving power balancing and reducing costs amid increasing renewable energy variability.

Contribution

It introduces a novel approach combining Copula theory, scenario generation via clustering, and stochastic optimization for EBTS operation under temperature uncertainty.

Findings

Cost savings compared to deterministic models

Effective handling of outdoor temperature uncertainty

Enhanced power balancing capabilities

Abstract

The rapid development of renewable energy has increased the peak to valley difference of the netload, making the netload following being a new challenge to the power system. Electric boiler with thermal storage (EBTS) occupies a non-negligible part of the load in the winter season in Northern China. EBTS operation optimization can not only save its own energy cost but also reduce the peak shaving and valley filling pressure of the system. To this end, the operation optimization of EBTS for providing the power balancing service is studied in this paper, which mainly includes three parts: First, the joint probability distribution between the predicted and actual temperatures is built by utilizing the Copula theory; Secondly, the actual temperatures are sampled based on the predicted temperatures of the next day, and the scenario set is generated by clustering these samples, where K-means clustering method are used; Thirdly, the stochastic operation optimization model of EBTS considering the uncertainty of outdoor temperature is constructed. Through the case study, it is found that the proposed method can save the total operation cost of the EBTS compared with the deterministic EBTS operation optimization model.