Estimating the impact of variable renewable energy on base-load cycling in the GB power system. (15th March 2020)

• Record Type: Journal Article
• Authors: de Mars, Patrick; O'Sullivan, Aidan; Keppo, Ilkka
• Is Part Of: Energy. Volume 195(2020)
• Publication Date: 2020-03-15
• Subjects: Power plant cycling -- Renewable energy -- Data analysis -- Regression; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2020.117041
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Between 2009 and 2017 the share of wind and solar energy sources in the GB electricity generation mix increased from 2.5% to 17%. Due to the variable nature of these renewable sources, large thermal power stations designed for constant base-load operation have been required to operate more flexibly to compensate for fluctuations in renewable generation. This flexible operation results in increased thermal stress and reduced efficiency causing increased operation, maintenance and fuel costs for these assets. In this paper we present the results of what is, to the best of our knowledge, the first empirical study on the impact of renewables generation on startups, ramping and part-loading (collectively, 'cycling') of base-load generators. We develop regression models using half-hourly generation data from 2009 to 2017 that capture the impact of increased renewable penetration while taking into account confounding factors including seasonality and demand. We find that with 2009-levels of renewable generation, cycling in 2017 would have been less severe, with 20% fewer startups. We also present estimates for cycling under National Grid Future Energy Scenarios to 2030 with implications for investment in generation assets. Additionally, the dataset derived in this research is made available and comprises the first open-access dataset on cycling. Highlights: A dataset for base-load cycling in the GB power system from 2009 to 2017 is presented. Base-load cycling is shown to have become stronger during the period 2009–2017. Regression models for cycling variables are estimated using power system variables. The models are used to predict cycling under Future Energy Scenarios to 2030.