Energy and water dynamics in data center cooling: Insights from a modeling study in hot-arid climates

The rapid growth of data centers (DCs) has led to increased energy and water consumption for cooling, which is heavily influenced by climate conditions. Optimizing DC cooling performance is crucial to reducing environmental impacts, lowering costs, and ensuring sustainable operations. This study developed novel 8760-energy and water models to estimate hourly energy and water use in Phoenix, AZ DCs, and examined three cooling systems (air-cooled chillers, water-cooled chillers, and evaporative cooling) with a particular emphasis on seasonal effects. The models estimate year-round energy and water savings in DCs using air-side and water-side economizers (ASE and WSE). This study fills a critical gap in the literature by providing a high-resolution, dynamic assessment of DC cooling performance that captures hourly, seasonal, and annual variations in energy and onsite and source water use. Results indicate that air-cooled chillers consume the most energy and source water, while water-cooled chillers have the largest annual onsite water use. ASE integration achieves about 22% energy and source water savings for air-cooled chillers and 29% onsite water savings for water-cooled chillers. Precooling in air-cooled chillers results in 11% annual energy savings. While chillers can operate year-around, evaporative systems with ASE provide cooling for only 64% of the year. With computer room cold aisle containment (CAC), ASE can cool DCs for 35% of the year, while WSE can provide cooling for 12% of the year. These findings offer actionable insights to optimize energy and water usage in DC cooling, minimizing environmental impacts and enhancing DC operation and sustainability.