The rapid growth of artificial intelligence is reshaping data center infrastructure, particularly when it comes to cooling. As high-performance chips generate more heat than traditional systems can handle, liquid cooling technologies are emerging as a critical solution, and with them, a new set of water treatment challenges.
A recent article published by Global Water Intelligence explores how this shift is driving increased demand for advanced water systems. The piece also features insights from Carollo’s John Rydzewski on the complexities of managing high-purity water in these environments.
Liquid Cooling Is Transforming Data Center Water Demand
As AI workloads expand, conventional air cooling is being replaced by direct-to-chip liquid cooling systems. These systems bring water into much closer contact with sensitive equipment, significantly increasing the importance of water quality and system control.
According to the article, liquid cooling could grow from just 3% of global data center capacity today to 28% by 2030. This shift introduces new requirements for water treatment, particularly during system startup and commissioning, where high-purity water is used to clean and fill closed-loop systems.
“For the high-purity water, treatment is preferably done on-site to get feedwater up to the specified quality,” Rydzewski explained, highlighting the growing importance of localized treatment systems.
High-Purity Water Brings New Engineering Challenges
While high-purity water is essential for effective cooling, it also introduces unique risks. Unlike traditional systems, direct-to-chip cooling loops require tight control over water chemistry to prevent damage to sensitive components.
“Using high-purity water in technical loops is challenging because copper cold plates, while great heat conductors, can corrode in overly pure water,” Rydzewski noted. “Striking the right balance is crucial.”
This balance between purity and stability represents a new frontier for water treatment in data centers, where even small deviations can impact performance, reliability, and equipment lifespan.
Monitoring and Control Are Critical for Industrial Water System Performance
Beyond initial treatment, maintaining water quality over time is equally important. Closed-loop systems operate under tight tolerances, with high recirculation rates and minimal room for error.
As Rydzewski pointed out, operators are increasingly focused on monitoring not just chemical conditions, but also biological and particulate factors.
“People are starting to look at bacteria and particle counts in that water because it needs to be clean enough to not plug the cold plate microchannels,” he said.
These requirements are driving demand for advanced monitoring systems, filtration technologies, and real-time controls that can maintain fluid integrity without interrupting operations.
A Growing Role for Water Expertise in Data Centers
The transition to liquid cooling is introducing a level of process engineering complexity that is new to many data center operators. As a result, water treatment providers and engineers are playing a more central role in system design, commissioning, and long-term operation.
From high-purity water production to corrosion control and biological management, water is becoming a critical component of data center performance and reliability.
As the industry continues to evolve, these systems will require closer integration between cooling technologies and water treatment expertise to support the next generation of AI infrastructure.
To learn more about how liquid cooling is reshaping water use in data centers and the role of high-purity water systems, read the full article.
