Cooling data centers: Keeping up with emerging technologies

In today’s digital and hyperscale computing landscape, the need for data centers has grown exponentially. According to the International Data Corporation, the global data sphere is expected to grow to 175 zettabytes by 2025, up from 33 zettabytes in 2018, with no signs of abating. This growth in data center infrastructure has led to a significant increase in heat generation, making its management crucial for business efficiency. Data centers can produce heat densities ranging from 50 watts to as high as 300 watts per square foot. Therefore, cooling has become an important and costly issue, accounting for up to 40 percent of total energy consumption in data centers.

This has propelled the development of new and emerging cooling technologies, as traditional cooling methods such as air cooling, refrigeration and “free” cooling—although effective to a degree—have presented challenges such as high energy consumption, limited scalability, significant environmental impact and high operational cost. To that end, trends in the evolution of cooling technology have emerged to combat these challenges, including direct-to-chip cooling, liquid cooling and reliance on renewable source energy. In particular, liquid immersion cooling, which involves the submergence of electronic components (such as servers and other IT hardware) into a thermally conductive but electrically insulating liquid, continues to gain traction in the industry. Each of these technologies provide specific benefits and may be better suited to different data centers depending on the facility’s characteristics. With these advances in cooling technologies, data center operators may find themselves in a conundrum when navigating between the various options. 

This article outlines some key considerations for data center operators to consider when determining how to cool their data center.

1. Assess the business drivers and data center’s operational characteristics 

When considering cooling technologies, it is essential to first assess the specific business drivers and operational characteristics of the data center facility, whether existing or greenfield. A comprehensive review of the financial, operational and other critical business drivers should help an operator identify the essential characteristics required of any potential cooling technology option. For example, if retrofitting an existing data center, it will be important to consider the existing cooling mechanisms and their compatibility with alternate cooling methods in tandem with the broader, strategic business goals. 

On the other hand, a new facility should incorporate a design that can scale over time, not just regarding cooling but for all critical equipment, including power and security. This approach could help to deliver both a first mover advantage and competitive differentiation, and, in turn, position a business to stay ahead of changes to regulatory requirements. For example, the American Innovation & Manufacturing (AIM) Act, enacted in December 2020, mandates an ongoing gradual phase-down of the production and consumption of hydrofluorocarbons (HFCs), which are widely used as refrigerants in cooling systems. This has resulted in increased costs for maintenance and refrigerant replacement as HFC supplies diminish and prices rise. By proactively implementing cutting-edge cooling solutions as part of its ongoing operational strategy, operators can ensure compliance with future regulations, thereby reducing the risk of potential disruptions and compliance costs, and set itself apart as an industry leader in innovation and environmental responsibility.

2. Conduct an energy and cost analysis

Another key consideration is to perform an energy and cost analysis to understand the actual costs of a particular cooling technology. Typically, data center operators will find tension between the long-term sustainability and efficiency of modern cooling technologies, and the significant initial capital expenditure to acquire and install them. Costs associated with cooling technologies often extend well beyond the initial acquisition and installation to long-term operational, maintenance and repair costs. For example, liquid immersion technology involves high acquisition costs, complex installation and large floor space requirements, all of which may not be suitable for small, high-density data centers with more pronounced space and budget constraints. For remote facilities, experimental cooling technologies may not be appropriate, as the cost of building and maintaining infrastructure may be both logistically challenging and prohibitively costly. Water-based cooling technologies may also not be appropriate for facilities located in desert regions. However, the abundance of sunlight in those locations may be leveraged to generate solar power—traditional energy may prove to be more effective in such areas. 

3. Engage in collaborative design for new data centers

Given the aforementioned tensions, it is important for operators to adopt a collaborative design approach in determining how cooling considerations may be integrated into the design of the facility.  Greenfield developments have a distinct advantage here. Raised floors, vents and different floor configurations allow for airflow optimization. The use of reflective materials or coatings on the roof of the facility may also significantly reduce heat absorption. Free cooling is significantly more effective in warmer regions. Thus, for data centers in the desert, it may be helpful to include shading devices, insulation, vents, windows and other natural ventilation mechanisms in the design. This process may also result in hybrid cooling systems by using multiple cooling technologies, which can create a robust system with backups if one cooling system malfunctions. 

4. Evaluate vendors

At the commercial level, data center operators should also conduct a thorough vendor evaluation as part of selecting their cooling technology. Given the immense business impact, the choice of a vendor should not be based solely on cost. Perhaps, more importantly, operators should focus on reliability, sustainability and innovation. This may be done through a competitive tender process that clearly outlines all the energy, scalability and other critical business requirements specific to the operation.

5. Utilize advanced technologies

Increasingly, the use of systems that allow an operator to leverage data and artificial intelligence to optimize and operate cooling technology for their facilities has become an important part of the overall commercial consideration. For example, the implementation of real-time monitoring systems provides insight into the actual cooling needs of the facility, thereby providing valuable data for an operator to evaluate the efficiency of different cooling methods. These information systems can also provide up-to-date and accurate information on minor temperature changes, alerting an operator to a system malfunction or an environmental anomaly, which can prove to be vital to long-term maintenance and repair decisions.

Evaluating the benefits and challenges of various cooling technologies in relation to the data center’s specific business and operational demands is the most effective approach to selecting the optimal cooling technology. Emerging technologies continue to develop in this area, as evidenced by experimental proposals on building data centers underwater and in space. In resolving most challenges, there is no one-size-fits-all approach, and operators should focus on what is most appropriate for their specific priorities. Regardless of what they may be, scalability, sustainability and security will likely continue to be the key principles driving decisions on this issue.

References:

_{1. "Data Age 2025: The Digitization of the World From Edge to Core", The International Data Corporation, Nov 2018.
2. Alphabet Q4 2023 Earnings Call. Transcript
3. "HVAC cooling systems for data centers” (no date) A. Bhatia, CED Engineering PDH Course. 
4. "DOE Announces $40 million for More Efficient Cooling for Data Centers". Department of Energy, May 9, 2023}. 

_{White & Case means the international legal practice comprising White & Case LLP, a New York State registered limited liability partnership, White & Case LLP, a limited liability partnership incorporated under English law and all other affiliated partnerships, companies and entities.}

_{This article is prepared for the general information of interested persons. It is not, and does not attempt to be, comprehensive in nature. Due to the general nature of its content, it should not be regarded as legal advice.}

_{© 2025 White & Case LLP}