Challenges in Industrial Instrument Calibration and Long-Term Stability in Harsh Environments

Back when I was still working in a factory, we had a calibration interval chart that we used to determine initial calibration intervals. It had several criteria to determine the interval:

• environment - class A, B, and C. A was harsh (like very hot) conditions; B was medium (like warm) conditions; and C was non-harsh (like ambient outside air) conditions.
• purpose - class A, B, and C. A was safety purpose; B was control; and C was for monitoring.
• utilization - class A, B, and C. A was continuous (24/7/365) use; B was 2/3 use (up to 9 months per year); and C was 1/3 use (up to 4 months per year).

Then we made a chart that had intervals in multiples of 3 months. For example: if an instrument was AAA, the interval was 3 months; if the instrument was AAB, the interval was 6 months; if it was AAC, the interval was 9 months.

The result was a chart that allowed us to make a calibration interval for the instrument. The maximum interval was 24 months (for test weights).

There was a lot of discussion before we came up with a chart that satisfied everyone. Some wanted the environment classification to have a fixed interval of 3 months if it was Class A, meaning it didn't matter if it was a Class ACC (harsh environment, monitoring, and used for only ≤4 months per year), it should still be 3 months interval.

These intervals were only initial, however. After several calibrations, we would have a record of the calibration drift and we could increase or decrease the interval based on how fast it was drifting. Of course, we needed to establish the tolerance for each instrument that would trigger a change in the interval.

Regarding selection of materials, we had materials guidelines. If the environment was, say, water, we could use this material. If it was acidic (also depends on the type of acid), we use a different material. Sometimes, we consulted with the manufacturer.

Smart diagnostics? By the time I left, smart diagnostics just started to appear in instruments and they were very simple. A parameter would just say that the instrument was near failure and needed replacing. I haven't been keeping up to date so I don't know how capable they are these days. I usually relied on process readings. For example: a particular control valve position is usually 50% to achieve, say, 150°C. Over time, it became 70% to achieve the same position. That tells me it's time to check the valve, the positioner, or the process (yes, sometimes, it's the process' fault).

regards,

Vulcan