Modern data centres depend on stable cooling systems to maintain equipment performance and availability. As high-density computing increases, liquid cooling is becoming an important solution for removing heat from servers and other demanding workloads. However, maintaining the quality of the coolant is equally important. Changes in pH, conductivity, ORP or turbidity can indicate developing issues within the cooling circuit.
For a Tier IV data centre operated by a European colocation provider in Bavaria, Germany, Rika Sensor implemented a liquid cooling water quality monitoring solution integrated with the facility’s existing Building Management System (BMS).
The project was designed for a cold plate liquid cooling system using a water-glycol coolant. It provided continuous monitoring across 12 racks while also generating data that could support operational management and compliance reporting.
Project Background
The customer operates a high-availability data centre environment supporting Industry 4.0 workloads. Such applications can require reliable computing infrastructure and stable environmental conditions, making cooling-system performance an important part of overall facility operation.
The cooling system uses a water-glycol mixture to transfer heat away from high-density equipment. Although glycol-based coolants can be suitable for liquid cooling applications, maintaining appropriate fluid conditions is essential for long-term system reliability.
The facility was also subject to EU compliance audits, increasing the importance of reliable monitoring records. The customer therefore required more than occasional manual measurements. It needed a monitoring system capable of continuously recording relevant water-quality parameters and providing usable historical data.
Challenge
The customer had previously experienced problems associated with conductivity and pH drift, which had contributed to pipe corrosion. These changes demonstrated the importance of detecting coolant-quality trends before they developed into more significant infrastructure problems.
However, the existing monitoring approach relied largely on monthly laboratory testing. Laboratory analysis can provide detailed information about a sample, but monthly testing leaves significant gaps between measurements. Gradual changes or short-term deviations could occur without being captured immediately.
Another requirement was automated reporting. With the data centre operating in a regulated environment, manually compiling water-quality information would increase the workload of the operations team and make consistent reporting more difficult.
The new solution therefore needed to provide continuous measurement, integrate reliably with the existing BMS and create records that could be used for compliance documentation.
Monitoring Requirement
The customer required a data center coolant water quality sensor system capable of monitoring four key parameters:
● pH: To identify changes in coolant acidity or alkalinity that could affect the cooling circuit.
● Conductivity: To monitor changes in the ionic characteristics of the coolant and identify potential changes in fluid condition.
● ORP: To provide additional information about oxidation-reduction conditions within the cooling loop.
● Turbidity: To detect suspended particles and changes in coolant clarity.
These measurements were required at multiple locations to provide better visibility into the condition of the supply and return branches. The system also needed to communicate with the existing BMS without creating communication conflicts with other building-management equipment.
Solution – Deployment & Commissioning
Rika Sensor installed 24 monitoring units across the cooling infrastructure serving 12 racks. The units measured pH, conductivity, ORP and turbidity and were positioned on the supply and return branches.
Monitoring both sides of the cooling circuit provided the customer with a more complete view of changes occurring as coolant moved through the system. This approach can help operators compare incoming and returning fluid conditions and identify developing trends.
For system integration, the monitoring units used 4–20 mA hardwired signals to connect directly with the existing BMS. This was selected specifically to avoid potential bus conflicts and provide a straightforward connection with the facility’s established monitoring architecture.
During commissioning, two monitoring points produced abnormal readings. The engineering investigation found that these sensors were positioned too close to chemical injection ports. Localised changes in coolant composition around injection points can temporarily affect measurements and may not accurately represent the overall condition of the cooling loop.
The two affected monitoring points were therefore relocated to more suitable positions. After relocation, the monitoring arrangement provided more representative measurement conditions.
The alarm system was configured as push-only, meaning the monitoring system could generate alerts but did not automatically initiate control actions. This allowed the BMS to receive water-quality information while keeping operational decisions under the control of the facility team.
Daily monitoring records were also configured for automatic CSV export. Customers could use these files to review historical trends and generate documentation required for compliance reporting.
Rated Products
The system can incorporate dedicated Rika Sensor products for liquid cooling applications, including conductivity sensors for coolant monitoring, coolant pH sensors, ORP sensors and turbidity sensors.
These dedicated monitoring devices provide individual measurements that can be combined into a broader coolant water quality sensor solution. The appropriate configuration can be selected according to the cooling system, installation points and required communication architecture.
Results
The project provided the Bavaria data centre with continuous visibility into important coolant-quality parameters across 12 racks. Instead of depending primarily on monthly laboratory tests, the facility could monitor changes in pH, conductivity, ORP and turbidity throughout normal operation.
Integration through 4–20 mA signals also allowed the monitoring system to work with the existing BMS without introducing bus conflicts. Following commissioning adjustments, the relocated sensors provided more representative measurements by avoiding the localised effects of chemical injection points.
Automated daily CSV records further simplified the collection of historical water-quality information. These records could be used by the customer to review coolant conditions, identify trends and prepare compliance documentation.
The project demonstrates how a CDU water quality sensor and broader liquid coolant quality monitoring strategy can support preventive maintenance, BMS integration and operational documentation in high-availability data centres. For facilities using water-glycol cooling, continuous monitoring can provide valuable insight into coolant condition while reducing dependence on infrequent manual or laboratory testing.
