Data Centre OCM Program

A data centre depends on a continuous path for removing heat from IT equipment. Liquid-cooled installations add fluid circuits, heat exchangers, pumps, filters and wetted materials to that path. Their condition needs to be understood alongside temperatures, flow, pressure and the maintenance history of the installation.

In a direct-to-chip system, coolant circulates through cold plates and associated distribution equipment. In an immersion system, compatible hardware operates in a dielectric fluid. Facility water may carry heat away through a separate circuit. These fluids serve different purposes: a technology-loop coolant, an immersion fluid and cooling-tower water cannot be assessed against one common specification.

A Data Centre OCM Program applies the discipline of condition monitoring to this fluid population. Cooling fluids are the central focus, with generator oil, engine coolant, stored diesel and ancillary lubricants added where relevant. The programme connects repeat laboratory measurements with a stable asset register, the correct fluid baseline and the operating events that can explain a change.

The starting point also depends on the site's stage. A new installation needs evidence around fill quality and commissioning. An established facility needs continuity of trends. A retrofit, coolant change or expansion needs a documented transition so old-fluid history is not mistaken for the baseline of the new system.

The programme follows the actual installation. Establish what each fluid contacts, how it is circulated and treated, where heat is exchanged, and which specification applies at each boundary.

Online readings and visual checks provide valuable operating information. Laboratory analysis adds selected evidence about composition, contamination and property changes that may not be explained by those readings alone.

Chem-Tech develops a parameter matrix for each fluid and purpose. Commissioning, routine monitoring and an abnormal-result investigation may require different combinations. The tables below explain the selection logic; the final method, capability, sample requirements and accreditation status are confirmed for the agreed scope.

Direct-to-chip coolant parameters

For water-based or inhibited-glycol DTC fluids, the programme considers fluid composition, chemical condition and contamination together. Keep technology-loop samples separate from make-up water and facility-side samples.

Immersion cooling-fluid parameters

Start with the dielectric product, operating phase and the hardware/fluid-supplier requirements. A compatible single-phase oil-based fluid may use the routine families below; electrical and extended properties are included according to the agreed question and monitoring plan.

The fluid register defines coverage. The main cooling-fluid scopes are supported by the relevant water, fuel, coolant and lubricant checks below; equipment that is not installed is not added to the programme.

These are programme-planning stages, with parameters and deliverables agreed for the site. They are not a fixed package applied to every fluid.

Chem-Tech works with the customer to turn the equipment population and maintenance questions into a practical laboratory monitoring plan. Development establishes what will be sampled, what will be measured, how results will be reviewed and who will respond.

1. 01

   Understand the site and the decisions to support

   Chem-Tech begins with the cooling arrangement, site stage, maintenance concerns and the equipment population. The site team identifies critical systems, operating constraints, past incidents and the decisions for which laboratory evidence is needed.

   Programme output: Agreed programme brief, scope boundaries and technical contacts.

2. 02

   Build the fluid and asset register

   Map each CDU/technology loop, immersion tank, facility circuit, generator and relevant auxiliary asset. Record product identity, concentration where relevant, fluid volume if known, wetted materials, filtration and an unambiguous sample-point identifier.

   Programme output: Fluid/asset register and sampling-point schedule with separate histories for separate circuits.

3. 03

   Develop the test and interpretation matrix

   Select condition, contamination and investigation parameters for each fluid. Chem-Tech reviews method suitability and current laboratory capability; the customer supplies the applicable coolant, hardware and OEM specifications. Agree units, reporting limits, test conditions and the basis for reviewing exceptions.

   Programme output: Parameter/method matrix, scope confirmation and documented sources for interpretation criteria.

4. 04

   Define baselines, intervals and responsibilities

   Plan unused-fluid and commissioning references where practical, then select a consistent risk-based sampling interval. Set event-driven checks, sample handling, review contacts and responsibilities. Include a pilot stage where sample quality or the starting condition needs to be established.

   Programme output: Baseline plan, sampling instructions, review/escalation workflow and pilot acceptance points.

Implementation converts the agreed plan into a repeatable site workflow. The pilot or baseline stage checks that the sample points, records and selected methods produce information that can be compared over time.

1. 01

   Prepare the sampling team and records

   Agree containers, required quantities, handling and dispatch for the chosen tests. Align labels and submission information with the asset register. Establish who collects, checks and submits each sample; coordinate sampling guidance or training within the agreed programme.

   Programme output: A repeatable collection process with named site responsibilities.

2. 02

   Run a baseline or pilot sampling round

   Collect samples from the agreed points, retaining the correct unused-fluid reference where available. Review identity, sample condition, collection consistency and the completeness of operating records before treating the first round as a monitoring baseline.

   Programme output: Initial result set, baseline observations and any adjustments to the sampling or testing plan.

3. 03

   Implement repeat sampling and laboratory testing

   Roll out the agreed schedule by loop, tank and component. Track sample receipt and the selected testing scope through the reporting workflow. Record missed or changed points, top-ups, fluid changes and maintenance so the resulting history remains comparable.

   Programme output: An organised sample history with consistent methods, units and fluid identities.

4. 04

   Interpret exceptions with the site team

   Review laboratory changes with supplier limits, previous results, temperatures, flow, pressure, filtration and maintenance evidence. Agree whether a repeat sample, targeted analysis or site/OEM inspection is the appropriate next step. The responsible site engineer retains operational decisions.

   Programme output: A prioritised exception and follow-up record, with the reason for each requested investigation.

5. 05

   Verify actions and review the programme

   After an intervention, resample at the agreed point and document the new trend phase. Review sampling completeness, recurring exceptions and the usefulness of the selected tests. Revisit the programme after expansions, product changes, revised OEM requirements or sustained changes in operating risk.

   Programme output: Documented follow-up results and a monitoring plan updated to the site's current condition.

Consistency starts at the sampling point. Define the safe procedure, container and handling requirements for each selected test, and keep the point and operating context comparable from round to round.

• Identify the site, asset, loop or tank, sampling point, fluid manufacturer/product, batch, glycol type and concentration where relevant. Keep facility water and technology cooling loops separate.
• Agree the bottle, quantity, cleaning and handling requirements with the laboratory before collection. Moisture, particles and microbiology can require different containers or handling; one bottle is not always suitable for every test.
• Use an approved representative sampling point and site isolation procedure. Do not open a hot, pressurised or energised system to collect a sample. Follow the fluid safety data sheet and trained site procedures.
• Record fluid temperature, service hours or age, recent top-ups, filtration, maintenance, fluid replacement and abnormal observations. Take a fresh-fluid reference from the same product when practical.
• Avoid mixing samples from loops or tanks. Seal and label promptly, follow agreed transport/storage instructions and record the collection date and time.

Information to submit with each sample

• Site, hall/building, asset and loop/tank/component identifiers, plus the exact sampling point.
• Cooling technology, fluid manufacturer/product, batch and glycol type/concentration where applicable.
• Collection date/time, fluid temperature, service age or hours and relevant operating condition.
• Top-up product/quantity, fluid replacements, flushing, filtration and maintenance since the previous sample.
• Symptoms, alarms, changes in flow/pressure/temperature, filter observations and the investigation objective.
• Applicable supplier/OEM specification and an unused-fluid reference when needed for comparison.

Examples below show how a laboratory signal can lead to a focused review. Check sample quality, the correct baseline, test conditions and independent site evidence before assigning a cause.

The agreed programme can include the records below. Laboratory reports and repeat-sample histories use Chem-Tech's reporting workflow; the programme brief defines the review scope, responsible contacts and any additional trend or management review required.

Review methods, units and specifications consistently. Annotate top-ups, product changes and interventions, and identify when a new baseline is required. The result history supports the responsible engineer's assessment; it does not provide a stand-alone remaining-life or system-reliability prediction.

Use the relevant supplier/OEM specifications with the agreed laboratory methods. The external references below provide industry context and do not imply an endorsement or a blanket laboratory accreditation claim.

• Open Compute Project — Cold Plate project and cooling-loop resources
• Open Compute Project — Base Specification for Immersion Fluids
• Dow — Data-centre cooling fluid guidance
• Intel and Castrol — Single-phase immersion cooling fluid study