A Probabilistic Framework for Uncovering Water-Energy Infrastructure Dependencies and Assessing Systemic Consequences

The coupling of water and energy systems is critical for ensuring resilient infrastructure, particularly as water plays an essential role in the safe and reliable operation of energy sources. The dependency on water becomes increasingly complex when energy systems are integrated with facilities that require not only vast amounts of power but also vast amounts of water for cooling, such as industrial plants and data centers. However, the simultaneous demand for electric power and cooling water underscores gaps in comprehensive vulnerability assessments for water-energy resilience as the source of electric power and the users of that power (e.g., industrial plants, data centers, etc.) both require water.
A specific example of the interconnection between simultaneous use of water for energy generation and meeting other facility demands is in the expected growth of nuclear power for data centers. Nuclear power, which has long been a cornerstone of reliable energy generation at GW scale, has recently expanded its role with partnerships between major technology companies and nuclear energy providers to meet growing energy demands for data centers. For instance, Meta signed a 20-year agreement with Constellation Energy Corp. for electricity from Clinton nuclear facility in Illinois, Google has partnered with Kairos Power for energy from advanced reactors for its data centers, Amazon is collaborating with X-Energy for nuclear energy solutions for its data centers, and Microsoft has signed a 20 year power purchase agreement from the restart of Three Mile Island, Unit-1. While these developments highlight the importance of advanced nuclear technologies, such as small modular reactors to support energy-intensive applications like data centers, they also highlight the expected simultaneous demand for vast amounts of power and cooling water.

To address these challenges, it is crucial to develop a robust framework for vulnerability and consequence assessments that analyze water and infrastructure interdependencies across both natural (e.g., heatwaves, wildfires, hurricanes, ocean acidification) and manufactured (e.g., cyber-attacks, physical attacks) threats. Such capabilities could significantly mitigate outage times for both power and water systems, ensuring resilience in the face of disruptions. This framework should include water dependency for various power sources and evaluate secondary applications like industrial facilities and data centers while accounting for other demands on scarce water resources. By creating comprehensive assessments and mitigation strategies, infrastructure managers can maintain the continuous operation of critical facilities, such as data centers, even amid unforeseen scenarios and disruptions. These steps are essential for enhancing water and energy resilience in a rapidly evolving and interconnected world.

Probabilistic modeling of vulnerability scenarios requires computationally intensive capabilities to comprehensively capture the system and intersystem dependencies in water-energy infrastructures. ORNL's supercomputer resources, including hybrid infrastructure of classical and quantum computers, can enable multi-loop dependency and uncertainties assessment of what-if scenarios of micro-grid threats. A vulnerability assessment framework for micro-grids and grids (including hydro power plants, nuclear reactors, data centers, and other critical facilities) using hybrid computing capabilities can analyze thousands of scenarios, uncover dependencies, identify potential vulnerabilities, and ultimately help reduce power and water outages by 2030.