Rika Sensor is a weather sensor manufacturer and environmental monitoring solution provider since 2010
A coolant distribution unit (CDU) can report stable temperature and flow without telling you whether its coolant chemistry is changing.
For data center operators, CDU manufacturers and system integrators, this creates an important distinction: thermal performance and coolant condition are related, but they are not the same measurement problem.
CDU water quality monitoring uses fluid-appropriate measurements, such as conductivity, pH, turbidity and, where justified, oxidation-reduction potential (ORP), to track changes in coolant condition. These measurements complement temperature, pressure, flow and leak detection. They do not replace coolant specifications or laboratory testing.
THE RIKA SENSOR PERSPECTIVE
At RIKA SENSOR, our recommendation is straightforward: define the coolant-quality question before choosing the sensor.
This article focuses on single-phase, water-based direct-to-chip cooling, including suitable water-glycol formulations. Its recommendations should not be transferred automatically to dielectric immersion fluids.
In a liquid-to-liquid CDU, the facility water system (FWS) and technology cooling system (TCS) exchange heat while keeping their fluid circuits separate. The TCS circulates coolant through the IT cooling infrastructure, including cold plates.
These circuits should not automatically share the same water-quality limits. The coolant formulation, wetted materials and equipment requirements must inform the monitoring plan. OCP’s CDU guidance specifically calls for material compatibility and vendor-recommended conductivity and pH ranges.
Before specifying sensors, answer four questions:
“Keep conductivity as low as possible” and “keep pH neutral” are not sufficient specifications.
A useful monitoring plan defines what acceptable coolant condition means for the actual loop.
Each parameter answers a different question. None provides a complete assessment of coolant health on its own.
CDU coolant monitoring parameters and their limitations
Conductivity is useful for tracking changes in the ionic characteristics of a coolant. However, treatment chemicals can affect the reading, and conductivity is temperature-dependent.
OCP’s water-based fluid guidance recommends establishing baselines around treatment and investigating changes during operation. This supports a trend-based approach rather than interpreting every increase as proof of contamination.
For meaningful comparisons, record the coolant formulation, measurement temperature and temperature-compensation method.
Select the conductivity range and accuracy around the expected operating values and alarm requirements. A wide measuring range alone does not demonstrate suitability for detecting small changes in a low-conductivity loop.
pH monitoring helps operators identify changes in the coolant’s acid-base condition. The relevant question is whether the reading remains within the approved operating window, not whether it stays close to pH 7.
Measurement conditions also matter. Low-conductivity water can produce unstable or slow pH readings with unsuitable electrodes. Sensor selection and sampling arrangements therefore deserve particular attention in these applications.
For an operating team, a sustained pH shift is a reason to investigate. It is not, by itself, a diagnosis of a particular corrosion mechanism.
Turbidity adds a different perspective from electrochemical measurements. It can help flag changes associated with suspended material, but interpretation requires care.
Bubbles, fluid color and the condition of optical surfaces can influence readings. The USGS identifies these as important sources of measurement interference.
If turbidity rises, check the installation and operating context before concluding that corrosion debris or another contaminant is present.
Where particle size or composition determines the maintenance response, additional particle analysis or laboratory testing is needed.
ORP reflects the combined influence of redox-active substances in a liquid. It does not measure the concentration of one specific chemical, and its interpretation can be affected by pH and temperature.
For a CDU cooling loop, ORP should therefore have a defined purpose. Include it when the coolant-management strategy identifies a useful redox trend and an appropriate response.
RIKA SENSOR’s position is that ORP should be selected for its application value, rather than treated as a mandatory channel in every installation.
Sensor placement should produce representative measurements while allowing inspection, verification and servicing.
Depending on the system design, locations worth evaluating include:
These are design options, not universal installation instructions. A common-header reading may not reveal a localized branch issue, while an inadequately flowing side stream may not represent current loop conditions.
RIKA SENSOR recommends checking the proposed installation against the sensor’s requirements for flow, pressure, temperature, orientation and bubble avoidance. Also allow room for removal and servicing.
The best location is not simply the nearest available threaded connection. It is the location that provides useful data under repeatable measurement conditions.
A dashboard becomes more valuable when the operating team knows what to do with a change.
We recommend organizing the monitoring plan around four steps.
Record the approved coolant identity, concentration, supplier limits and initial measurements after the specified commissioning procedure.
Keep commissioning acceptance records separate from routine operating trends. OCP’s TCS pre-commissioning guidance includes sampling and testing during flushing and after coolant filling, reinforcing the importance of documented starting conditions.
Compare measurements with the baseline and approved limits.
Record coolant additions, filter changes, maintenance work and temperature changes alongside the trend data. This context helps distinguish an unexplained shift from a change associated with a known event.
Warning levels, persistence rules and escalation thresholds should be agreed with the relevant equipment and coolant specialists.
When an unexpected reading appears, check sensor condition, sampling flow, bubbles and recent maintenance activity.
Use a suitable reference measurement or laboratory sample where necessary. Online measurements provide visibility between sampling events; laboratory analysis can answer more specific questions about fluid composition.
OCP’s CDU guidance describes periodic chemical analysis for evaluating changes such as inhibitor depletion and material leaching.
Every meaningful alarm should have an owner and a documented response.
Depending on the findings, that response may involve further sampling, inspection or supplier-directed corrective work. Do not assume that a single pH, conductivity or ORP alarm justifies automatic chemical dosing.
The objective is a defensible maintenance decision, not simply a return to a green dashboard.
RIKA SENSOR’s liquid-cooling sensor portfolio provides separate instruments for different monitoring priorities.
| RIKA SENSOR model | Parameter | Listed measuring range |
|---|---|---|
| RK500-12LC | pH | 0–14 pH |
| RK500-13LC | Conductivity | Options including 0–20, 0–200, 0–2,000 and 0–5,000 µS/cm |
| RK500-07LC | Turbidity | 0–10 NTU or 0–100 NTU |
| RK500-06LC | ORP | −1,500 to +1,500 mV |
On smaller screens, scroll the table horizontally to view all specifications.
These are instrument measuring ranges, not recommended coolant operating limits.
The specifications list simultaneous 4–20 mA and RS485 outputs, an operating temperature range of 0–60°C, and pressure resistance of 1 MPa (10 bar). The listed media include deionized water, PG25 and EG25.
Final selection should confirm the exact coolant formulation, required accuracy, temperature-pressure conditions, wetted-material compatibility and process connection. A listed coolant category does not establish suitability for every commercial formulation.
For higher-temperature cooling designs, the 0–60°C operating specification requires particular attention.
The aim is to select a measurement package that fits the loop, rather than adapt the loop’s requirements to a standard package.
No. The selection should follow equipment requirements, coolant chemistry and the risks being monitored. Additional parameters are useful when they provide information that changes an operating or maintenance decision.
No. Conductivity must be interpreted against the approved coolant formulation and operating limits. Treatment chemicals can affect the baseline, so different fluids should not be judged against one universal target.
Not directly. ORP measures oxidation-reduction potential. Assessing corrosion may require dedicated corrosion measurements, material examination or laboratory analysis, depending on the investigation.
No. Online sensors provide ongoing measurements of selected parameters. Laboratory testing remains important for questions that those sensors do not directly answer, including specific chemical composition and contamination.
Effective CDU water quality monitoring starts with a clear understanding of the fluid and ends with a clear response to meaningful change.
For RIKA SENSOR, the priority is not the largest number of measurement channels. It is the right combination of sensors, representative installation conditions and a practical interpretation plan.
Discuss Your CDU Monitoring Requirements with RIKA SENSOR
If you are developing a CDU or reviewing a data center coolant-monitoring strategy, share your coolant specification, operating temperature and pressure, target parameters and installation requirements with RIKA SENSOR.
LEAVE A MESSAGE