Water Stress on U.S. Power Production at Decadal Time Horizons

TL;DR

This study assesses the impact of climate-driven water scarcity and rising stream temperatures on U.S. thermoelectric power production over the next few decades, highlighting regional vulnerabilities and the challenges of decadal climate risk assessment.

Contribution

It presents a novel approach combining climate models, observations, and population data to evaluate power production risks due to water scarcity and temperature increases at decadal scales.

Findings

Over 200 counties may face water scarcity in the next three decades.

Significant increases in stream temperatures exceeding EPA thresholds are observed in some counties by 2030s and 2040s.

Power plants in South Carolina, Louisiana, and Texas are particularly vulnerable.

Abstract

Thermoelectric power production at risk, owing to current and projected water scarcity and rising stream temperatures, is assessed for the contiguous United States at decadal scales. Regional water scarcity is driven by climate variability and change, as well as by multi-sector water demand. While a planning horizon of zero to about thirty years is occasionally prescribed by stakeholders, the challenges to risk assessment at these scales include the difficulty in delineating decadal climate trends from intrinsic natural or multiple model variability. Current generation global climate or earth system models are not credible at the spatial resolutions of power plants, especially for surface water quantity and stream temperatures, which further exacerbates the assessment challenge. Population changes, which are difficult to project, cannot serve as adequate proxies for changes in the water demand across sectors. The hypothesis that robust assessments of power production at risk are possible, despite the uncertainties, has been examined as a proof of concept. An approach is presented for delineating water scarcity and temperature from climate models, observations and population storylines, as well as for assessing power production at risk by examining geospatial correlations of power plant locations within regions where the usable water supply for energy production happens to be scarcer and warmer. Our analyses showed that in the near term, more than 200 counties are likely to be exposed to water scarcity in the next three decades. Further, we noticed that stream gauges in more than five counties in the 2030s and ten counties in the 2040s showed a significant increase in water temperature, which exceeded the power plant effluent temperature threshold set by the EPA. Power plants in South Carolina, Louisiana, and Texas are likely to be vulnerable owing to climate-driven water stresses.