An industrial oven was reading 400°C on the controller, but the product came out under-baked. The operator checked with a handheld thermometer — the actual temperature was 370°C. The controller was reading 30°C high. The customer thought the controller was bad. It wasn’t. The thermocouple had drifted.
Why thermocouples drift
A Type K thermocouple is two wires of different alloys welded at the junction. The millivolt output depends on the alloy composition at the junction. Over time, the junction oxidizes, the alloys diffuse, and the millivolt output changes. At 400°C, a Type K thermocouple drifts by about 1-2°C per year of continuous use. After 10 years, that’s 20°C. The oven’s thermocouple was 12 years old. The reading was 30°C high — a bit more than expected, but the sign and scale match drift.
Type K is also vulnerable to a specific failure: the “green rot” oxidation of the positive (chromel) leg in the temperature range 250-600°C. The chromel oxidizes and the output drops. The thermocouple reads low by 20-30°C. In this case the reading was high, which is the more common manifestation — the junction alloy changes and the calibration curve shifts. Either way, the thermocouple is no longer within its ±2.2°C tolerance.
The fix and the rule
The fix was simple: replace the thermocouple. A new Type K junction cost $15. The oven now reads 401°C when the handheld reads 400°C. The calibration check confirmed the new thermocouple is within 1°C.
The rule: thermocouples are consumables, not permanent sensors. Replace them on a schedule. For an oven running at 400°C continuously, replace every 2 years. For intermittent use, every 4 years. The $15 thermocouple is cheap insurance against a batch of under-baked product that costs $2000 to rework.
And when you replace a thermocouple, don’t just swap the probe — check the extension wire too. The extension wire is a second thermocouple junction where it connects to the controller. If the extension wire has been overheated, or the wrong type was used (Type K extension wire is color-coded green/yellow), the reading shifts by the difference between the junction temperature and the reference. The oven’s extension wire ran past the heater and had been cooked for 12 years. Its insulation was brittle. The new wire cost $5 per meter and the reading became stable.
The calibration check that catches drift
You don’t need a calibration lab to catch thermocouple drift. A boiling water check (100°C) catches gross errors. For the 400°C range, use an ice bath (0°C) and a known reference temperature (a second, new thermocouple in the same location, read simultaneously). The two readings should match within 2°C. If they differ by more than 5°C, the installed thermocouple has drifted. Replace it. The check takes 10 minutes and should be done every 6 months on a critical oven.
The controller’s built-in cold junction compensation is another drift source. The controller assumes the terminal block is at a known temperature. If the enclosure is hot (the controller was mounted above the oven), the terminal temperature is higher than the assumed 25°C. The reading shifts by the difference. The fix: keep the controller and terminal block away from heat, or use a thermocouple with a remote cold junction.
The industrial thermocouple selection
For ovens above 200°C, Type K is the default but not always the best. Type N is more stable than Type K in the 300-800°C range — it resists the green rot failure. If the oven runs continuously at 400°C, Type N costs 2x but lasts 4x longer. For a production oven, the Type N upgrade pays for itself in fewer replacements and more consistent readings. Type J is cheaper but drifts faster at 400°C. Don’t use Type J above 300°C.
Thermocouples drift. Replace them on schedule, check the extension wire, and do a six-month calibration check against a new junction. The 30°C high reading wasn’t a controller fault — it was a 12-year-old junction that had quietly gone out of tolerance. Fifteen dollars of new wire and probe fixed a $2000 rework problem.
