Thermal Imaging for Manufacturing: Electrical & Mechanical PdM
By James Smith on August 7, 2026
Every electrical connection with rising resistance generates detectable heat before it fails, and every mechanical component losing lubrication or developing misalignment produces a measurable temperature differential before it seizes. Infrared thermography turns that physics into an inspection method: a qualitative temperature-difference measurement, commonly called Delta T (ΔT) criteria, defined as the temperature rise of a component above a reference — typically ambient temperature, a comparable component under the same load, or the maximum allowable temperature the component is rated for. This isn't a new or experimental technique; it's governed by established standards from NETA, NFPA, and ASTM, and as of 2023, NFPA 70B shifted from a recommended practice to an enforceable standard requiring annual infrared inspection of electrical equipment. That shift changes what a thermography program actually needs to look like — documentation and consistency now matter as much as the scan itself. See how iFactory tracks thermal scan history and ΔT trending against your specific equipment baselines and reporting requirements.
Condition Monitoring · Thermal Imaging
Thermal Imaging for Manufacturing: Electrical and Mechanical PdM
Scan procedures, ΔT temperature thresholds, and reporting standards for a thermography program governed by NETA, NFPA 70B, and ASTM E1934 — not internal guesswork.
A Temperature Rise Above a Defined Reference, Not an Absolute Number
ΔT criteria measure an increase in temperature above a reference value, not a fixed absolute temperature threshold applied uniformly. That reference can be ambient air temperature, the temperature of a comparable component operating under the same load conditions, or the maximum allowable temperature the component is rated for by its manufacturer. This matters practically: a connection running at 60°C might be perfectly normal in a hot mechanical room and seriously abnormal on an otherwise-cool panel where every comparable connection sits at 35°C — the ΔT against the right reference is what actually indicates a developing problem, not the raw number in isolation. Some thermographers classify severity based on their own field experience rather than a published standard, but doing so consistently and documenting the basis for that judgment matters just as much as following NETA or NFPA tables directly.
The Governing Standards
NETA, NFPA 70B, and ASTM E1934
Three organizations govern most infrared thermography work in industrial and electrical applications, each covering a different piece of a complete program — from procedure and equipment use to maintenance frequency requirements.
Standard
Governs
2026 Status
NETA (InterNational Electrical Testing Association)
ΔT severity classification tables widely used for electrical component testing
Industry-standard reference for classifying severity once a ΔT is measured
NFPA 70B
Electrical equipment maintenance practices, including infrared inspection frequency
Enforceable standard as of the 2023 edition — annual IR inspection of all electrical equipment is now a requirement, not a recommendation
ASTM E1934
Standard guide for examining electrical and mechanical equipment with infrared thermography
Reference methodology for inspection procedure and equipment use
The shift of NFPA 70B from recommended practice to enforceable standard is the single most significant recent change in this space — it moves annual infrared inspection of electrical equipment from a best-practice recommendation to a compliance requirement, with the practical consequence that a program without documented, regular IR inspection is now a compliance gap, not just a missed opportunity.
Scan Procedure
Getting a Reliable Reading, Not Just a Colorful Image
A thermal image looks impressive on a report regardless of whether the underlying scan technique was actually sound — these three practices are what separate a genuinely reliable reading from one that only looks authoritative.
Load Conditions Matter as Much as the Camera
Electrical thermal defects are load-dependent — a loose connection generating excess heat under 80% load may show a much smaller signature at 20% load. Scanning during representative, ideally peak, load conditions is what makes a ΔT reading meaningful rather than misleading.
Correct for Emissivity and Reflected Temperature
Different surface materials emit infrared radiation differently, and reflective surfaces can show a reflected temperature rather than the component's actual temperature — a scan that doesn't account for this can produce a reading that looks alarming or reassuring for entirely the wrong reason.
Compare Against the Right Reference, Every Time
Consistently documenting which reference — ambient, a comparable component, or rated maximum — a given ΔT reading is measured against is what makes the severity classification defensible later, during trending or during an audit of the program itself.
The consistent structure across this four-tier scale is what makes findings comparable across different components, different scan cycles, and even different thermographers on the same team — everyone is classifying against the same conceptual framework even when the exact degree thresholds differ by standard or component type. What breaks that comparability isn't disagreement about where exactly a threshold sits; it's a program where severity gets classified inconsistently, sometimes against a documented standard and sometimes against gut feel, with no record of which method applied to which finding.
A Number Without a Reference Is Meaningless
60°C Is Normal in One Context and a Critical Finding in Another
iFactory tracks every scan's reference basis alongside the reading — so ΔT severity classification stays consistent and defensible across your entire program.
The Same Physics, Two Different Failure Signatures
A single thermography program frequently covers both applications, but treating them as identical inspections with identical expected patterns is a mistake — the underlying ΔT logic is shared, while the specific things to look for differ meaningfully.
Electrical — Connection and Load Issues
Loose connections, overloaded circuits, unbalanced phases, and failing breakers all generate excess resistive heat that shows up as a localized hot spot on a panel, terminal, or bus bar — this is the application NFPA 70B's annual inspection requirement specifically targets.
Mechanical — Friction and Lubrication Issues
Bearing housings, couplings, and other rotating equipment generate excess heat from degrading lubrication, misalignment, or developing mechanical wear — the same ΔT-against-reference logic applies, but the reference and expected thermal pattern differ meaningfully from an electrical connection.
Reporting Standards
What a Defensible Scan Record Actually Needs
These four elements are what turn a thermal image into a genuinely defensible maintenance and compliance record, not just an interesting picture with a temperature number attached.
01
Document the Reference Basis for Every ΔT Reading
Whether the ΔT was measured against ambient, a comparable component, or rated maximum needs to be recorded alongside the reading itself — a ΔT number with no stated reference cannot be reliably classified or compared later.
02
Record Load Conditions at Time of Scan
Since electrical thermal signatures are load-dependent, the load percentage at scan time belongs in the record — without it, a follow-up comparison scan at a different load level can produce a misleading trend.
03
Classify Severity Consistently Against a Named Standard
Whether using NETA's severity tables or an internally documented equivalent, the classification method used for a given finding should be stated explicitly and applied consistently, not chosen case by case based on how alarming a finding appears.
04
Maintain Scan History for Trending, Not Just Point-in-Time Findings
A single scan shows a snapshot; a documented history across multiple scan cycles is what actually reveals whether a component's thermal signature is stable, slowly worsening, or has jumped severity tiers since the last inspection.
Field Perspective
“
The mistake I see most in thermography programs isn't bad camera technique — it's treating a ΔT number as meaningful on its own, without recording what it was actually measured against. Sixty degrees means something completely different on a panel where everything else reads thirty-five than it does in a mechanical room running hot across the board. Once NFPA 70B became enforceable, that documentation gap stopped being a minor sloppiness issue and became a real compliance exposure. The programs that hold up under scrutiny are the ones where every finding has its reference basis, its load condition, and its severity classification written down at the time of the scan — not reconstructed from memory when someone asks for it later. That reconstruction gap is exactly where a program looks fine on paper until someone actually audits it.
Callum Osei-Meraz
Reliability Engineer & Certified Thermographer · 14 years running electrical and mechanical infrared thermography programs across industrial manufacturing
Common Questions
Frequently Asked Questions
What's the difference between ΔT criteria and an absolute temperature threshold?
ΔT criteria measure the increase in temperature of a component above a defined reference value — ambient temperature, a comparable component under similar conditions, or the component's maximum allowable rated temperature — rather than applying a single fixed temperature number to every component regardless of context. This matters because the same absolute temperature reading can indicate a serious problem in one context and be entirely normal in another, depending on what the surrounding conditions and comparable equipment actually look like. Consistently documenting and using the correct reference for each ΔT reading is what makes thermography a reliable diagnostic method rather than a source of false alarms or missed findings. Book a thermography program review to assess whether your current program is applying ΔT criteria correctly.
Is annual infrared inspection of electrical equipment actually required now, or still just a best practice?
As of the 2023 edition, NFPA 70B transitioned from a recommended practice to an enforceable standard, and a key update in that transition specifically requires annual infrared thermography inspections of all electrical equipment. This is a meaningful shift for any facility that previously treated thermography as a discretionary best practice — a program without documented, regular IR inspection now represents a compliance gap under the current standard, not simply a missed opportunity for early fault detection, and facilities should confirm their current program actually meets this requirement rather than assuming legacy practices are sufficient. The shift also raises the bar on documentation, since a compliance-relevant inspection needs a defensible record, not just a scan that technically happened.
Why does the load condition at the time of a thermal scan matter so much for electrical inspections?
Electrical thermal defects, such as a loose connection generating excess resistive heat, are load-dependent — the same developing fault can produce a small, easy-to-miss thermal signature under light load and a much more pronounced signature under heavy load, since resistive heating scales with current. Scanning during representative or peak load conditions is what makes a ΔT reading genuinely diagnostic, and comparing scans taken under meaningfully different load conditions without accounting for that difference can produce a misleading trend — a reading that looks like improvement might simply reflect a scan taken at lower load, not an actual repair or improvement in the component's condition. Talk to solutions engineering about recording and tracking load conditions alongside your thermal scan data.
Does infrared thermography apply the same way to mechanical equipment as it does to electrical systems?
The underlying physics and the ΔT-against-reference logic are the same, but the specific failure signatures and expected thermal patterns differ meaningfully between the two applications. Electrical thermography typically targets connection resistance, overloaded circuits, and phase imbalance producing localized hot spots at panels and terminals, while mechanical thermography targets friction and lubrication issues in bearings, couplings, and rotating equipment, which tend to show a different thermal pattern shaped by the mechanical component's geometry and normal operating temperature rather than a connection point. A thermographer experienced primarily in electrical applications benefits from specific additional training before confidently interpreting mechanical thermal patterns, and vice versa.
What documentation does a thermal scan record need to be considered defensible during an audit or compliance review?
A defensible scan record needs the ΔT reading itself along with its stated reference basis, the load condition at the time of the scan, the severity classification applied and which standard or internal equivalent that classification was based on, and a maintained history across scan cycles rather than only the most recent finding in isolation. Missing any of these elements makes a finding difficult to defend or act on consistently — a ΔT reading with no documented reference, for instance, cannot be reliably reclassified or compared to a future scan of the same component, which undermines the entire point of running a structured thermography program in the first place.
A Program That Holds Up to Scrutiny
ΔT Tracking, Reference Documentation, and Severity Trending in One System
iFactory tracks ΔT readings, reference basis, load conditions, and severity classification across your full thermal scan history — so your program meets NFPA 70B's enforceable annual inspection requirement with a defensible, auditable record.
