High Velocity Temperature Measurement


Thread Starter


What is the minimum insertion depth for high velocity flows requiring reduced length but without losing accuracy?

The process dept wants it longer than the TW codes allow.
Thermocouples and modern Pt RTDS are very tip sensitive. I can't see any reason to put a thermowell more than a couple of inches into a liquid line, perhaps a bit further for gas.

If you make them too long, you need to do a Wake Frequency Calculation. A Process Engineers sole aim is to make it impossible for the instrument guys (just joking).
It is not just about the measurement on such projects...indeed

In this particular case, the customer was dealing with condensate slugging events every few seconds and flange mounted wells.

The 19.3TW standard strictly applied did not pass resonance criteria even if the well was shortened to insert only a few 2-3mm of the tip into the flow.

We got pass this using flow profile arguments and para 6-10-8 and further shortening of the wells, but the supplier was uncomfortable as resonance was greater than the 80% limit.

Interested in how others deal with this.
Yes, you didn't mention slugging.

I can think of another application with a small line at very high pressure. Where instead of hanging a huge 900# flange on the piping, we simply measure the pipe wall temperature. This avoids a huge heat sink and possibility of a leak.
For pipe stress relieving post welding, they weld thermocouples direct to the pipe, one wire on one side of the weld with the other wire on the other side so the pipe material is part of the thermocouple (law of intermediate metal).

The wires are spot welded to the pipe using a capacitor discharge welder (no pre-heating required).
The sensor weld tabs were mentioned, but were not acceptable to the customer standards at the time. Although some package equipment did rely on skin temps as part of an ESD.

Insulated, thick walled NPS18, high pressure.

Fortunately there was enough flow adjacent to the nozzle opening to insure decent response time and tip error. The slugging events created the possibility of severe resonance stress with traditional insertion lengths.