Drought and climate change impacts on cooling water shortages and electricity prices in Great Britain

Reliable and affordable electricity is crucial for modern economies. Disruptions in electricity supply, even without physical damage to the grid, can lead to significant economic impacts, particularly in liberalized markets where more expensive plants must compensate for unavailable capacity. Large-scale meteorological events like droughts can simultaneously affect multiple power plants, causing substantial welfare losses. Past droughts in Brazil and California have demonstrated the significant cost increases associated with reduced hydropower availability and increased reliance on thermal power. The electricity sector's dependence on water resources, particularly for thermoelectric power plants (using water for cooling), is increasingly recognized as a major climate risk. While some studies have explored the impacts of reduced cooling water availability, few have employed probabilistic methods and risk assessment across large spatial domains to quantify the economic consequences for energy markets and consumers. This study addresses this gap by combining national-scale, risk-based water resources planning approaches with a model of power plant availability and electricity market supply prices.

Previous research highlights the economic losses from business disruption caused by power outages, particularly in conjunction with other hazards like flooding. Studies on droughts in Brazil and California have estimated significant additional costs in electricity prices due to reduced hydropower capacity and increased thermal power usage. Existing literature emphasizes the electricity sector’s dependence on reliable water resources for both hydropower generation and thermoelectric plant cooling. However, most studies lack probabilistic methods and comprehensive spatial risk assessments. A key limitation is the lack of probabilistic modeling combined with detailed spatial assessment of the effects of low flows on power plants. This study seeks to improve on these limitations.

The research utilizes a coupled hydroclimate and electricity supply-demand model framework. This framework combines a large ensemble of climate and hydrological model simulations to predict water availability at power plants, accounting for environmental flow requirements. The Weather@Home climate simulation system (W@H) generates 100 unique 30-year projections for three time slices: a historical baseline (1975-2004) and two future scenarios under climate change (near future and far future). The DECIPHER hydrological model simulates river flows based on the W@H climate projections. Power plant availability is calculated daily by comparing simulated flows with environmental flow requirements. A bespoke short-run marginal supply curve is created, incorporating data from 893 power plants. A machine learning model (gradient boosting regression trees) estimates daily electricity demand, using weather variables as covariates. By removing impacted capacity from the supply curve, the model calculates the impact on the electricity strike price and quantifies the resulting welfare impacts. Sensitivity analyses are conducted to assess the influence of renewables production and fuel prices.

The study reveals substantial impacts of cooling water shortages on electricity prices in Great Britain. In the baseline scenario, median capacity unavailability at individual plants ranges from 3.4% to 4.2%, increasing to 5.5–6.9% and 5.8–11.2% under the near-future and far-future climate change scenarios, respectively. On extreme days (p99), almost 50% (7 GW) of freshwater thermal capacity is unavailable in the baseline scenario, increasing to 46% and 52% under future climate scenarios. Annualized cumulative costs on electricity prices range from £29–66 million per year in the baseline, rising to exceed £100 million per year under climate change. The single-year impacts of a 1-in-25-year event exceed £200 million. These costs are higher in late summer and autumn. The study further demonstrates that the high sensitivity of electricity costs to renewables production and the effects of fuel prices. Higher-than-average renewable production may lessen the impact, but the study finds that the effects of drought accumulate even in relatively benign conditions. This points to the need for improved mitigation strategies.

The findings highlight the significant vulnerability of Britain's electricity system to drought and climate change, even with the current mix of power sources. While extreme droughts are unlikely to cause blackouts, the curtailments in thermal plant production lead to substantial costs. The spatial heterogeneity of drought risk is revealed, with some power plants experiencing far greater impacts than others. The results underscore the importance of considering climate uncertainty in risk assessment and highlight the need for investments to mitigate the risks associated with cooling water shortages. The interplay between renewables production and the effects of drought on the strike price is crucial, suggesting the potential to offset the costs through increased investment in renewables capacity.

This research provides a robust quantification of the economic impacts of drought on Britain's electricity sector, emphasizing the significant costs and spatial variability of this risk. The findings highlight the need for strategic investments to mitigate these climate-related risks, considering the interaction of drought impacts with the increasing penetration of variable renewables. Future research should investigate the effectiveness of different adaptation strategies and the long-term economic implications of climate change on the energy sector.

The study focuses solely on drought's impact on freshwater-dependent thermoelectric plants, neglecting other water users and electricity producers. The model assumes static generation capacity and demand, not accounting for long-term changes due to decommissioning, new plants, population growth, or technological change. Also, the impacts of other weather conditions were not considered, only drought conditions.