Communities and water utilities are grappling with the increasing need for infrastructure to support digital service demands. The American Water Works Association (AWWA) has released a white paper, titled Cooling the Cloud, which provides guidance for assessing the new challenges and opportunities arising from data centers.
Data centers are high-impact customers for water utilities and require more adaptive planning. This report examines key issues including:
- Utilities must engage proactively, conduct thoughtful analysis, and collaborate effectively to mitigate risks and support long-term system sustainability.
- The chosen cooling technologies significantly impact both water and energy consumption at data centers.
- Examples from utilities in Virginia and Colorado illustrate successful adaptation through proactive planning, infrastructure coordination, and policy development.
Eight planning priorities are identified in the report that can guide utilities through this development:
- Regional Coordination: Coordinate regionally on capacity limits; involve utilities early to ensure design changes (e.g., non-evaporative cooling) are feasible.
- Source Capacity: Assess all source capacity limits (regulatory, physical, rights) and account for future growth and permitted allocations.
- Treatment Capacity: Evaluate drinking water and reuse treatment limits; adding capacity is costly and a long-term capital investment.
- Transmission & Distribution: Identify distribution areas that can support demand; be prepared for limits in less ideal, affordable locations.
- Customer Impacts: Analyze the effects on existing customers; ensure data center needs do not limit future growth or impose new costs.
- Financials: Determine how to cover costs; upgrades for single/cluster customers may mean the financial burden falls on that smaller group.
- Alternatives: Consider different approaches such as requiring low/zero water-use cooling, use diverse sources like reuse water, and encourage internal reuse.
- Contingencies: Plan for disruptions like drought or power outages; establish procedures for when key usage data is unknown until late in the process.
Posted in Uncategorized