I'm looking to purchase some thermistor-based probes to
> monitor the performance of the HVAC system (return and
> supply duct temperatures along with liquid and suction
> lines. Maybe the temperature of the compressor, etc.)
>
> As part of my research, I look at the problems people
> (consumers) report with the units used in their systems.
>
> I would have thought NONE of these would fail: they are
> out of the way so not likely to be physically disturbed;
> often encased in a protective sheath (e.g., probes);
> shouldn't be operating beyond their design constraints
> (unless poorly designed); etc.
>
> Even the pigtails from such devices should be reasonably
> safe from disturbance!
>
> Yet, I see folks replacing them and proud that they
> did so without calling in the HVAC contractor!
>
> Are there other "wear" factors that come into play?
> What sort of service life (in a conservative design)
> should I expect?
About the only failure mode that I'm aware of is drift - thermistors are sintered metal oxides, so there are gaps in the material and stuff could diffuse in.
The good ones are enclosed in glass beads - Yellow Springs were the first to introduce them - and those drift a lot less than their predecessors. I had my nose rubbed in this back when I was working at Kent Instruments in England in the early 1970's, and I've used them from time to time since then.