Live electricity is not enough. For an electrical thermographic inspection to actually detect the faults it is designed to find, the equipment must be under meaningful operational load. Here is why the 40 percent rule exists, what the physics behind it actually is, and what happens when inspections are conducted under it.
There is a misconception in electrical maintenance that any thermographic inspection conducted on energised equipment will reveal the faults that exist. As long as the system is live, the camera will see what needs to be seen.
This is not true. An inspection conducted on energised equipment that is carrying minimal load will produce thermal images that show very little. Loose connections will be barely warmer than their neighbours. Overloaded circuits will not yet be overloaded. Phase imbalances will be invisible. The thermographer will leave site believing the equipment is healthy, when in reality the inspection has simply not stressed the system enough to make the existing faults thermally visible.
This is why NFPA 70B 2023 Section 7.4.5 requires electrical equipment to be operating at no less than 40 percent of normal circuit loading during thermographic inspection. It is also why The Snell Group's electrical thermography course material teaches the 40 percent rule as one of the fundamental requirements of valid IR survey work.
This article explains the physics behind the rule, why operating load matters more than most facilities teams realise, what gets missed when inspections are conducted under low load, and how structured load correction makes findings comparable across inspection cycles even when load conditions vary.
The Physics: Why Load Determines What Thermography Can See
Electrical thermography works because faulty connections, components, and conductors generate more heat than healthy ones. That heat differential is what the thermal imaging camera detects. No heat differential, no visible fault.
The heat generated at any point in an electrical circuit follows a fundamental physics relationship known as Joule heating, also called ohmic heating or resistive heating. The formula is:
Power dissipated = Current² × Resistance
Written more compactly as P = I²R. This formula tells you exactly how much heat a given resistance produces under a given current. And the critical word in that formula is the squared exponent on current.
A connection carrying 100 amps with a resistance of 50 microohms dissipates 0.5 watts at the joint. The same connection carrying 200 amps dissipates 2 watts. The same connection carrying 400 amps dissipates 8 watts. Doubling the current does not double the heat. It quadruples it.
This exponential relationship is why load matters so much for thermography. A loose busbar joint that produces a barely detectable 1°C rise above ambient at 20 percent load will produce a clearly visible 6 to 8°C rise at 80 percent load. Same fault, same connection, same equipment. The only difference is whether the system is actually carrying enough current to make the fault thermally visible.
The implication: Thermography does not test how hot equipment is. It tests how much faster heat develops at faulty points compared to healthy ones, under conditions where current is high enough to generate detectable differentials. Below a certain load threshold, that differential collapses to zero and the inspection effectively becomes blind.
The 40 Percent Rule: What NFPA 70B Actually Requires
The NFPA 70B 2023 standard specifies the minimum load requirement for thermographic inspections in Section 7.4.5: where normal circuit loading cannot be achieved during inspection, the equipment must operate at no less than 40 percent of normal load. Section 7.4.6 then requires that the actual load conditions during the inspection be documented and retained for future reference.
These two requirements work together. The 40 percent threshold establishes the minimum load at which thermography can reliably detect significant faults. The documentation requirement ensures that any inspection conducted below full load can be correctly interpreted later, particularly when comparing readings across multiple inspection cycles.
It is worth being precise about what the rule does and does not say:
- It does not say 40 percent is sufficient. It says 40 percent is the minimum that should be accepted. Many low-grade anomalies are still below the detection threshold at 40 percent load and will only become visible at higher loading.
- It does not say maximum load is required. It says maximum load is ideal where it can be safely achieved. For most data centre, hospital, and continuous process operations, inspections are conducted at whatever the normal operating load happens to be, which is often well above 40 percent.
- It does not exempt low-load inspections from documentation. If the inspection is conducted at less than 100 percent load, the actual load percentage must be recorded for every finding, so future comparisons remain valid.
What Gets Missed When Inspections Are Conducted Under Low Load
A thermographic inspection conducted under the 40 percent threshold produces what is technically known as a false negative result. The faults are still there. The equipment is still degrading. But the inspection produces thermal images that do not show the problem, and the report concludes that the equipment is healthy. This is in many ways worse than no inspection at all, because it produces documented evidence of an issue-free system when issues exist.
Loose Connections Stay Hidden
A loose busbar bolt or cable termination produces heat in proportion to the square of the current passing through it. At 25 percent load, that heat may be only 1 to 2°C above ambient and indistinguishable from background thermal noise. At 75 percent load, the same connection might be 8 to 12°C above its neighbours and immediately obvious.
Phase Imbalance Disappears
Phase imbalance is detectable as a temperature difference between the three phases. At low load, all three phases are barely above ambient and the imbalance is invisible. At meaningful load, the overloaded phase pulls noticeably hotter than the others and the diagnostic is immediate.
Cable Heating Becomes Undetectable
Conductors heating along their length, indicating undersized cabling or developing insulation degradation, only show up clearly under load. At low duty, the entire cable run reads close to ambient and there is nothing to flag.
Overloaded Circuits Are Not Yet Overloaded
A circuit running at 110 percent of its design rating is generating measurable extra heat. The same circuit running at 30 percent is not yet stressed and the overload condition is invisible. This is especially relevant in installations where load growth has occurred since the original design and the actual circuit utilisation is now significantly higher than the design assumed.
Transformer Losses Stay Buried
Transformers under low load dissipate very little heat. Their thermal signature is dominated by no-load losses, which are relatively small. Under heavier load, copper losses dominate and any winding degradation, core hotspots, or cooling system inadequacy becomes thermally visible. A transformer running 10°C above its nameplate temperature loses approximately 50 percent of its expected service life, which makes detection of thermal anomalies under load particularly consequential.
Load Correction: Making Findings Comparable Across Inspections
Even with the 40 percent rule respected, real-world inspections rarely happen at exactly 100 percent load. A data centre might be at 78 percent during the inspection window. A commercial building might be at 55 percent. A factory between shifts might be at 65 percent. Comparing raw temperatures across these conditions, or comparing them year-over-year for trending purposes, is meaningless without normalisation.
This is what load correction does. The load correction formula takes the measured temperature, the ambient temperature, the actual load in amps, and the component rating in amps, and calculates what the temperature would be if the component were running at 100 percent of its rated load.
Load-corrected temperature = (Measured temperature - Ambient) × (Component rating / Actual load)² + Ambient
This is the ITC load correction formula referenced in BS7671 and used throughout UK and international electrical thermography practice. It uses the same I²R physics that makes load matter in the first place, but in reverse: it projects forward from the measured operating state to the equivalent fully loaded state, so findings can be graded against the appropriate reference temperature.
Done manually with a calculator, load correction adds significant time to every fault entry. A typical electrical thermography report with 25 to 30 findings might add an hour of manual calculation if every reading needs to be corrected by hand. Done automatically, the calculation is instantaneous and consistent across every finding in every inspection.
Scheduling Implications: When to Inspect
Once you understand that load matters this much, inspection scheduling becomes a strategic decision rather than an administrative one.
Inspect at Peak Operational Load Where Possible
Peak demand periods produce the strongest thermal signatures and the highest likelihood of catching low-grade anomalies that would otherwise be missed. For data centres, this is typically during peak compute hours. For commercial buildings, peak afternoon air-conditioning load. For factories, peak production. Schedule inspections to coincide with these windows where it is safely possible.
Inspect Under Comparable Conditions Year-over-Year
For trending to be meaningful, year-over-year inspections should be conducted under similar load conditions. An inspection in February at 40 percent load is not directly comparable to an inspection in August at 85 percent load, even with load correction applied. Consistent timing reduces the variables and improves the quality of multi-year trending data.
Avoid Low-Load Windows
Inspections conducted during planned maintenance windows when the protected load has been transferred to bypass or backup, during overnight low-demand periods, or during seasonal shutdowns will produce poor-quality data. If these are the only windows available, document the load conditions clearly and acknowledge the limitation in the report.
Use Multiple Visits Where Necessary
For installations with widely varying load profiles, multiple inspection visits at different times may produce a more complete picture than a single visit. This is particularly relevant for industrial facilities with intermittent loads, commercial buildings with strong day-night cycles, and any installation where peak demand differs significantly from average demand.
What the Report Must Document
NFPA 70B 2023 Section 7.4.6 requires circuit loading characteristics to be documented and retained. This is not just a compliance requirement. It is a practical necessity for any trending or multi-year analysis. Every thermographic finding in a properly compliant report should include:
- Measured temperature at the component
- Ambient temperature at the time of measurement
- Actual load in amps at the time of measurement
- Component rating in amps from the nameplate or design documentation
- Load-corrected temperature calculated from the above
- Fault grade assigned against the configured reference temperature
- Methodology notes recording the inspection load conditions for the equipment as a whole
Without these data points, the inspection cannot be defended in an audit, the findings cannot be trended forward, and the report has limited value beyond a point-in-time observation.
How SnapCor Handles Load Correction Automatically
The reporting workflow in SnapCor is designed around the load correction requirement. For every thermographic finding, the thermographer enters the measured temperature, ambient temperature, actual load in amps, and component rating in amps. The app immediately calculates the load-corrected temperature using the BS7671 formula, compares it against the configured reference temperature (typically 75°C for cable terminations), and assigns a fault grade.
Every data point is stored in the inspection record and carried forward into the generated PDF report. The methodology section of the report documents the overall inspection load profile. The individual fault pages show every value used in the calculation, making the analysis transparent and auditable. And the structured data is available for trending against subsequent inspections of the same asset, with consistent normalisation applied across every cycle.
This combination of mandatory data capture, automatic correction, and consistent application across inspections is what turns a series of individual reports into a defensible long-term thermography programme. The compound value of structured load-corrected data over multiple years is covered in our article on what thermal trending reveals over 3+ years of inspections.
Frequently Asked Questions
Why is the threshold 40 percent and not, say, 50 percent or 30 percent?
The 40 percent figure is based on the relationship between load percentage and detectable heat differential under the I²R formula. Below 40 percent, heat dissipated at faulty points is typically too small to produce reliably detectable thermal differentials against background noise. Above 40 percent, the differential grows exponentially and detection becomes increasingly reliable. The Snell Group analysis notes that even 40 percent may be insufficient for very low-grade anomalies, but it is the practical minimum at which most significant faults are reliably visible.
What if my facility never runs above 40 percent load?
Some installations, particularly oversized backup distribution and reserve capacity systems, run at low utilisation by design. Inspection of these systems should be conducted under the highest available load, the actual load percentage should be documented, and the thermographer should explicitly note in the report that low-load inspection may miss low-grade anomalies. For high-criticality reserve systems, periodic load-bank testing combined with thermography during the load test is a practical solution.
Does load correction work for all electrical equipment?
Load correction using the I²R formula is most directly applicable to current-carrying connections: busbars, cable terminations, breaker stabs, fuse clips, contactor terminals. For components with more complex thermal behaviour, including motors, transformers, and capacitors, additional correction factors may apply. The thermographer's judgement and the manufacturer's guidance take precedence over the standard formula in these cases.
How do I know if a previous inspection respected the 40 percent rule?
A properly documented inspection report will state the load conditions either in the methodology section or on each individual fault page. If the report does not document load conditions, the inspection should be treated as low-confidence data and a new survey conducted under properly documented load conditions for trending purposes.
Does SnapCor enforce the 40 percent rule?
SnapCor does not block low-load inspections, but it does capture and document the load percentage for every finding, in line with NFPA 70B Section 7.4.6 requirements. If the inspection is conducted below 40 percent load, the thermographer can record this in the methodology notes and the documentation supports any subsequent analysis or audit. See the first inspection walkthrough for the full workflow.
Load Is the Variable That Makes the Inspection Real
Energised is not the same as loaded. Live is not the same as detectable. The 40 percent rule exists because the physics of electrical heat generation do not care whether the equipment is on or off. They only care how much current is flowing, and the heat produced grows with the square of that current.
For mission-critical infrastructure, the practical implication is unambiguous. Schedule inspections under representative operating load. Document the load conditions for every finding. Apply load correction to make the data comparable across cycles. And use a reporting platform that captures all of this automatically so the thermographer can focus on the inspection, not the spreadsheet.
Try SnapCor Free for 14 Days >>
For UK enterprise thermography services across data centres, hospitals, and multi-site portfolios, contact the TI Thermal Imaging team. For UAE and GCC enquiries, see Thermal Imaging UAE. New to SnapCor? Start with the installation guide.
SnapCor is a thermographic inspection reporting platform built by TI Thermal Imaging. Reports are aligned to ISO 18436-7 and informed by BS7671 reference temperatures. The 40 percent minimum load requirement referenced in this article is sourced from NFPA 70B 2023 Section 7.4.5 and 7.4.6. Always combine software outputs with qualified thermographer judgement and applicable site-specific safety procedures.