In a data centre, a single overheating connection doesn’t just cause a fault. It causes an incident report, an SLA breach, and an uncomfortable call with a client whose infrastructure just went down.
Electrical thermography is one of the few inspection methods that can catch a failing connection, an overloaded busbar, or a degrading breaker before it takes anything offline, without powering anything down or interrupting live operations. For mission-critical environments, that makes it a core part of planned preventive maintenance (PPM), not an optional extra bolted on after something has already gone wrong.
Why data centres are different
Most electrical assets get some slack: a warehouse can tolerate a breaker tripping for an afternoon, a retail unit can work around a dead circuit until Monday. Data centres are built around the opposite assumption. Redundancy (N+1, 2N) protects against a single component failing outright, but it does very little against the failure mode thermography is built to catch: a connection that is quietly degrading under load, getting hotter on every cycle, for weeks or months before it finally lets go, often outside a redundant path, and often at the worst possible moment.
That’s the gap electrical thermography closes. It doesn’t wait for a component to fail; it measures the heat signature that shows a fault is developing, while there’s still time to plan a fix instead of reacting to an outage.
What a routine data centre thermal survey actually finds
It’s rarely dramatic. On a typical PPM visit, the findings that come up again and again are quiet ones:
A termination that was torqued correctly on install but has loosened slightly over time, invisible on a visual walk-round, clearly visible on a thermal one. A breaker carrying more than its rated share of load because a circuit was rebalanced months ago and never rechecked. A connection running a few degrees warmer than its neighbours under identical load, nothing that trips anything yet, but a clear trend if it isn’t addressed.
None of these show up on a visual inspection. All of them show up on a thermal one, graded against a documented baseline so the facilities team knows whether it’s a “note it and recheck in six months” finding or a “get someone in this week” one.
How we grade and document findings
Every fault we identify is graded against a standardised scale, Minor, Important, Serious, Critical, based on documented temperature-rise (ΔT) bands, not a subjective read of a colour palette. Grading is done by ITC Level 3 certified thermographers, the highest level of thermographer certification, which means the report explains not just that a fault exists but why it’s graded the way it is, how load conditions at the time of inspection affected the reading, and what it means for the asset going forward.
Inspections are aligned to ISO 18436-7, BS7671 (with automatic load correction built into our reporting), and NFPA 70B’s documentation requirements for corrective measures, so the report holds up whether it’s being reviewed internally, by an insurer, or by an auditor. Our equipment fleet is camera-agnostic across FLIR, Fotric, Workswell, Hikmicro, Guide Sensmart and InfraTec, so the survey isn’t dictated by whichever camera happens to be available.
PPM that follows condition, not the calendar
An annual thermal survey is a reasonable default. It’s not the right answer for every asset. Inspection frequency should follow condition: an asset in good condition might genuinely only need checking once a year, while one with a documented fault trending upward, even a minor one, needs a six-month recheck, not a twelve-month one, until it’s resolved. That’s how we structure PPM recommendations for data centre clients, and it’s a small shift in thinking that catches problems a purely calendar-based programme misses.
Who this is for
We work with UK and UAE/GCC data centre and critical facilities teams who need electrical thermography built into planned maintenance as standard, not scheduled reactively after a near-miss. If electrical thermography isn’t already part of your data centre’s PPM schedule, that’s worth a conversation before it becomes a reactive one.
Get in touch to discuss a data centre electrical thermography survey for your site.