7 Infrared Thermographic Tips Every Buyer Should Know

A useful thermography inspection starts with a written scope, not a camera setting. ISO 18434-1:2008 outlines procedures for thermographic condition monitoring, including data collection, analysis, and reporting. List the assets, operating states, and components to inspect. A motor terminal box under load tells a different story from one photographed after shutdown. Record load, ambient conditions, viewing distance, and access limits. Small details matter.

Set acceptance criteria before images are taken. ISO guidance supports consistent inspection and interpretation, but it does not make every warm spot a fault. Compare similar components under similar loads, and note emissivity or reflected heat that could distort readings. That sounds obvious. It is easy to skip. The U.S. Department of Energy’s Operations & Maintenance Best Practices guide reports 8–12% savings for predictive maintenance programs compared with preventive maintenance. That figure applies to programs broadly, not thermography alone. A thermal survey is one input, not a repair order. Buyers should ask for annotated images, temperature measurements, findings ranked by urgency, and limitations that explain what the survey could not verify. A neat report can still miss a hidden surface. Keep that possibility in view.

NETD measures the smallest temperature difference a thermal camera can resolve above its own noise. A rating of 50 mK equals 0.05°C under specified test conditions. Lower values can help reveal subtle patterns, such as a faint warm strip along a window frame. But NETD is not the same as temperature accuracy. A camera can have low NETD and still report temperatures inaccurately. ASTM E1934-99a provides guidance for infrared examination of electrical and mechanical equipment, while ISO 18434-1:2008 describes thermography procedures for condition monitoring. Neither makes 50 mK a universal buying rule.

For building checks, prioritize sensitivity when you need to spot small contrast changes across walls, seals, or pipes. Then compare spatial resolution and focus performance: a sensitive sensor cannot recover details blurred by distance or poor focus. Check sample images taken on similar materials and at similar distances. A lower NETD helps, but it does not do the inspection for you.

An infrared camera’s accuracy line deserves more scrutiny than its headline resolution. Common specifications quote ±2°C or ±2% of reading, but check whether the manufacturer means “or” or “whichever is greater.” The difference can affect a decision. At a displayed 50°C, 2% is 1°C; at 100°C, it is 2°C. Confirm how the percentage is defined and whether the figure applies across the camera’s full temperature range. A neat number can still mislead.

Measurement conditions matter, too. ISO 18434-1:2008 identifies emissivity and reflected apparent temperature as important considerations in thermographic measurement. ASTM E1933-14(2022) covers measuring and compensating for emissivity with infrared imaging radiometers. A glossy pipe and a matte painted panel may therefore show different readings, even at the same actual temperature. Ask what calibration reference supports the stated accuracy, and whether the specification assumes a particular distance, target size, or ambient condition. That matters. A small target filling only a few pixels can produce an unreliable reading, despite an impressive accuracy claim. I would compare the stated tolerance with the temperatures and surfaces I actually need to inspect, then leave room for uncertainty.

A thermographic camera does not read temperature directly; it converts infrared radiation into a temperature estimate. Emissivity is the setting that describes how efficiently a surface emits radiation. Many nonmetallic materials are near 0.95, but that is a starting point, not a universal constant. Published engineering reference tables put concrete around 0.94 and ordinary glass around 0.92–0.94. Coatings, moisture, dirt, and viewing angle can shift the result. Small differences matter.

I still double-check.

Reference values can tempt buyers to skip verification, and that is an easy mistake. Keep distance, angle, reflected surroundings, and surface condition consistent when comparing readings.

A thermal camera cannot measure a tiny component accurately just because it appears on the screen. Its instantaneous field of view, or IFOV, describes the angle covered by one detector pixel. At a greater viewing distance, that pixel represents a larger area on the target.

Check the camera’s IFOV and calculate the approximate spot width: distance multiplied by IFOV in radians. Then compare that width with the target’s dimensions.

A pixel is not always enough. For a dependable reading, the target should fill the camera’s stated measurement spot, which may span several pixels. Check the instrument’s spot-size specifications, not just the screen image or a general distance-to-spot ratio. Keep the camera square to the surface, focus carefully, and move closer when practical.

Small targets demand care. A warm screw beside a cool panel can blend into one reading if the measured area includes both. I have also found that a crisp-looking image can still produce a poor temperature measurement. Shiny surfaces add another complication: reflected heat may distort the apparent temperature, so compare readings from a suitable, consistent surface area and note the conditions.

An infrared camera estimates surface temperature from detected radiation. It does not simply read the temperature beneath the lens. On shiny or low-emissivity surfaces, reflected heat from nearby objects can distort the result. A cool-looking metal panel may reflect a cold window, while a warm one may mirror a person or heater. Reflections can mislead.

For a more reliable assessment, use the camera’s reflected-temperature setting if available, and follow its measurement guidance. A small piece of matte, high-emissivity tape can provide a useful comparison point when appropriate for the surface. Let it reach the surface temperature before measuring. Don’t assume the tape tells the whole story; its placement and contact matter. Record the viewing angle, distance, and nearby conditions, especially when checking the same area later. Readings remain estimates, so confirm concerning results with another suitable method.

An infrared image is useful only when its context is recorded. When commissioning an inspection, request a report aligned with ASTM E1934 guidance for examining electrical and mechanical equipment. Ask the inspector to identify assets, inspection conditions, measurement locations, and access limitations. The report should pair thermal images with visible-light images and explain notable temperature patterns in plain language. A color image alone proves little.

Thermography detects surface temperature patterns; it does not reveal every internal defect. Treat an anomaly as a prompt for follow-up testing, not a diagnosis by itself. ASTM E1934 guidance offers a reporting framework, not a guarantee that every fault will be found. If a conclusion seems unclear, ask what evidence supports it and what could not be inspected.