Dive Brief:
- ASHRAE, the National Electrical Manufacturers Association and the Pacific Northwest National Laboratory on Wednesday released an AI data center performance framework that they say will provide facility managers and others expert guidance on designing and operating energy-efficient data centers. The framework is hosted on ASHRAE’s website.
- “This guide brings together the most comprehensive industry expertise on data centers in a single resource,” said Bing Liu, PNNL director of buildings and industrial programs. “It's a dynamic online resource that can be updated, remain relevant and stay accessible to anyone involved in developing a data center.”
- The framework can help facility managers translate complex technical challenges into clear, actionable strategies that help operators enhance performance, control costs and make more effective use of energy, while strengthening reliability at both the facility and grid level, ASHRAE said.
Dive Insight:
Artificial intelligence and high-performance computing are driving demands on facility systems and energy infrastructure, according to the organizations. The United States has more than 3,000 data centers and an additional 1,500 are in development. More growth is expected, the groups said.
The framework provides strategies for addressing the full lifecycle of data centers, from planning and design to commissioning, retrofitting and operations.
Thermal management, integrated system performance, energy and water use and facility reliability are among the operations issues for which the framework offers best practices. The recommendations are tailored to account for load densities, operating conditions and the climate in which the facility is located, the organizations said.
In the operations and maintenance section, the framework recommends facility managers improve system performance and uptime by deploying real-time monitoring and anomaly detection for critical systems.
The framework also recommends facility managers design continuous protection of critical infrastructure from cyber threats and physical threats, establish continuous training programs that align with industry certifications and align operational setpoints and control strategies to ASHRAE TC.9.9 guidance.
“AI data centers require operations and maintenance programs that balance analytics with disciplined facilities practices,” ASHRAE said in the guide.
Although AI and machine learning tools can provide continuous monitoring of building systems with alerts for action, their accuracy relies on the quality of inputs and other factors that facility managers play a role in managing, ASHRAE said. “Facilities personnel retain accountability for interpreting results, authorizing actions, and executing maintenance activities safely and correctly,” it said.
ASHRAE led development of the guidance related to HVAC systems, thermal management and facility performance based on its existing body of work, the group says. It drew from resources developed through its Technical Committee 9.9, which sets standards for data centers and other mission-critical facilities.
Other committees were involved in the work, including those that oversee ASHRAE’s Standard 90.4, Energy Standard for Data Centers, and Standard 127, which governs testing of air conditioning units serving data centers and IT equipment.
NEMA contributed expertise for the sections on electrical systems, equipment and safety. PNNL used its expertise in energy systems research to help the working group.
“As AI continues to drive rapid changes in load density, system design and operational expectations, this Framework brings together the collective expertise of ASHRAE, PNNL and NEMA,” ASHRAE said.
ASHRAE will host the 2027 Data Center and AI Integration Conference on March 3-5 in Dallas as part of its support for industry collaboration and knowledge sharing, the organization said.
