Hi All,

We’ve had several RTD’s fail in the same way, whereby the sheath has split (time frame in process ranges from 1 - 5 years). The RTD’s sit in a thermowell with no chance of mechanical impact on the thermowell.

My current theory is that the thermowell has been compromised from time, vibration and being sat in an abrasive fluid. This has allowed salt and steam from the process to enter the well (this can be seen on visual inspection). Furthermore, the thermowell is likely to be made out of mild steel and the RTD is made out of 316SS. Could galvanic corrosion be possible and cause the sheath to split? I aim to get the thermowell material tested to confirm whether it is mild steel.

Has anyone experienced this failure? Could there be an alternative reason? Any help is much appreciated.

 

Your observations of the failure of thermowell seem to point in the right direction. Even 316 SS is not that corrosion resistant for hot salt water. Consider Monel 400 or Inconel 825 for thermowell material.

The ID of the bore in the thermowell creates a "slip fit" for OD the element so there isn't a lot loose play . How did you get the RTD sheath out of the thermowell when the RTD had split open like that?

 

Witnessed failed Pt RTDS that slip fit within the well, due to excess vibration. RTD's are limited to thermowell tip accelerations exceeding 10 G's . Conventional T/C can take 20-30 G's before failure. Thermowell cracking due to fatigue failure is another issue.
With key dimensions of the TW, and nominal fluid conditions, the service accelerations are readily evaluated.

RTD's are actually quite fragile. Typically the element is filled with extremely fine Al2O3 nanoparticles that is rather loose.

Good luck

 

Your observations of the failure of thermowell seem to point in the right direction. Even 316 SS is not that corrosion resistant for hot salt water. Consider Monel 400 or Inconel 825 for thermowell material.

The ID of the bore in the thermowell creates a "slip fit" for OD the element so there isn't a lot loose play . How did you get the RTD sheath out of the thermowell when the RTD had split open like that?

I agree that an alternative thermowell material is required, thanks for your suggestions.

The thermowell has a nominal bore of 0.75” and the RTD sheath has a diameter of 0.236”, as per the original design documents. I’m afraid a ‘slip fit’ was never present.

 

Witnessed failed Pt RTDS that slip fit within the well, due to excess vibration. RTD's are limited to thermowell tip accelerations exceeding 10 G's . Conventional T/C can take 20-30 G's before failure. Thermowell cracking due to fatigue failure is another issue.
With key dimensions of the TW, and nominal fluid conditions, the service accelerations are readily evaluated.

RTD's are actually quite fragile. Typically the element is filled with extremely fine Al2O3 nanoparticles that is rather loose.

Good luck

Thanks for your response. Would you suggest considering a thermocouple as an alternative?

 

First determine if flow induced resonance conditions are likely, just to identify possible cause and future risk you might be dealing with. All you need are nominal dimensions of your thermowell, and nominal fluid velocity, presumed liquid.

Some applications require RTD measurements, especially where traceable calibrations are required. That decision is made when your equipment is designed, so continue with the RTD's.

 

What makes a Pt100 RTD subject to vibration failure is the relative delicate nature of the very, very thin platinum wire being shaken around by vibration.

The result of vibration is a broken wire, an open circuit and no means of making a temperature measurement and getting a valid process variable.

Your photo has the RTD sheath looking like Godzilla chewed on it and then spit it out which is more in line with the results of severe corrosion, not a broken delicate little wire inside a protective sheath.

Yes, thermocouple wires in MgO elements are heavier and a welded hot junction is far more vibration resistant than Pt100 Pt wire, but I'm skeptical in your case that vibration is the killer, not corrosion.

 

The sheath thickness of Pt RTDS'a is less than 0.1mm and when damaged your first impression is how could it be so damaged.

When flow induced resonances are involved, the tip acceleration can easily become 50-100 G's high frequency vibration. The sensors don't last long at that level.

 

Dave, no argument that vibration causes element failures. The vibration failures I've met have been failures several days after installation on forge hammers (whomp, whomp, whomp where the ground shakes every whomp) where the element failed but the sheath was in fine shape. Conversion to thermcouples worked well.

I've never seen a 316SS sheath torn open like the OP's photo. That can happen from vibration?

 

The first time I saw such damage, it looked like it was intentionally hammered on an anvil, not unlike the op's image