Know what changed.

Coolant and oil testing for liquid-cooled data centers.

Laboratory analyst drawing a clear data-center coolant sample beside stainless analytical equipment
Laboratory scopeData Center Fluid TestingRegional focusAsia

Faster starts

Speed starts before dispatch.

Scope the work before the bottle moves, so time is not lost deciding methods after arrival. We confirm the tests, sample requirements, receiving details and quoted timing before collection.

  • Test panelMatched to the fluid and question
  • Sample requirementsContainer, volume and handling
  • Receiving routeLocation and dispatch instructions
  • Quoted timingAgreed before the sample moves

From sample to operating decision.

A result is useful only when it stays connected to the fluid, asset, sample point, operating state and baseline.

See how we protect quality
  1. Brief

    Name the fluid, asset, sample point and question that needs an answer.

  2. Collect

    Use the agreed container, volume and collection instructions.

  3. Test

    Apply methods and quality checks suited to the actual fluid.

  4. Review

    Compare the results with baselines, history and operating context.

  5. Decide

    Make the finding, uncertainty and recommended next step clear.

Four fluid systems, one brief

Coolants, dielectric fluids and oils in one testing brief.

Coordinate cooling, power and mechanical samples without forcing different fluids into the same test panel or acceptance limits.

Analyst pipetting fluid samples into a controlled laboratory test array

A better model for fluid testing

Faster service. AI-enabled chemists. Clearer decisions.

Faster service starts with fewer manual handoffs. Our direction is AI assistance that helps chemists organize evidence, compare history and focus their review. Laboratory results should connect with available operating and sensor context, with a human reviewer accountable for the interpretation and next step.

  • Reduce avoidable intake and reporting handoffs
  • Support chemists with traceable AI assistance
  • Connect laboratory, sensor and maintenance evidence
  • Explain the next step, not just the result
Read our approach to actionable insights

PG25 technical focus

25% glycol does not define fluid condition.

A useful PG25 assessment begins with the exact product, concentration basis, inhibitor chemistry, wetted materials and applicable OEM limits. Only then can chemistry, contamination and metals be interpreted together.

Explore PG25 testing
How we assess PG25Representative methods
Five parts of the assessment
Condition areaRepresentative method
Identity + concentrationRefractometry and matched-reference FTIR
D3321 / E1252
ChemistrypH, reserve alkalinity and product-specific UV-Vis indicators
D1287 / D1121 / validated assay
Ions + degradationChloride, sulfate and selected degradation markers
D5827 / validated route
Wetted materialsElements selected from the verified system materials
D6130
CleanlinessAppearance, sediment and particle context where the route is suitable
Method confirmed by fluid

Method applicability and acceptance criteria are confirmed for the named fluid and system. Generic limits are not substituted for product or manufacturer guidance.

Liquid cooling training in Asia

Train the people who own the loop.

Practical training in coolant management, CDU commissioning and secondary-loop reliability, delivered through Reliability Engine for teams across Asia.

Explore training
Operations

Liquid Cooling Operations Fundamentals

Map the FWS, CDU, TCS, rack and responsibility boundary.

Fluid management

Sampling and Interpretation

Build baselines, collect representative samples and set clear review triggers.

Commissioning

Secondary Loop Reliability

Connect cleanliness, filling, filtration and handover evidence.

For multi-site teams

One brief. Comparable records across sites.

Common identities, comparable units, defined method routing and clear exception handling across facilities, countries and operating teams.

Sample integrity

Structured briefs, collection context and traceable identity.

Data integrity

Raw results remain distinct from normalized values and interpretation.

Review accountability

Software organizes signals; a human reviewer owns the interpretation.

Start with the operating question

Make the next sample count.

Tell us what you need tested. We will confirm the recommended tests, collection instructions and expected timing before you send the sample.