TA-23-C019 – Rationalizing Data Center Whitespace Airflow and Energy-Flow Metrics PDF

Whitespace airflow and energy-flow metrics provide insights which assist with the design and operation of robust and resource-efficient data centers – and have their greatest utility in CFD. However, some existing metrics were developed more than two decades ago when CFD was a research tool – at a time when data center architectures, best-practices, and CFD were evolving rapidly. Today, airflow and energy-flow metrics are incorporated into data center CFD tools which are routinely used by large numbers of non-experts to design and operate facilities. Standardization and guidance are needed for CFD users, developers, and the broader data-center community to maximize the potential value of the metrics. This work is intended as progress towards that goal. We first rationalize airflow and energy-flow metrics by removing redundancies, filling logical gaps, and creating a consistent and intuitive naming convention. We then show that analogously-defined airflow and energy-flow metrics are approximately equal anyway for many practical applications – which provides flexibility in how they are computed in CFD or determined from measured temperatures in a real facility. Finally, we state the advantages of airflow-based metrics in data center CFD and show how we have implemented them to provide design insights at the IT, cooler, and room levels.