Thermocouple troubleshooting

[jrcagle]

There are three ungrounded, 1/8" OD type J thermocouples in a single pipe well operating at ~690 degF. One of them is reading ~15 degF lower than the other two. We're being asked to troubleshoot the coolest thermocouple to see if the temperature deviation is real or not. There is no indication the pipe well is compromised but we're not going to attempt replacement or removal of the thermocouples until the vessel is out of service due to high pressure.

We tested the mV reading from the thermocouple and it matches the temperature indicated to DCS. We simulated a mV reading to DCS and it read correctly. So far so good.

I also asked our techs to check insulation resistance between the lead wires and the sheath for all three thermocouples. All three read ~110 ohms. I expected a much, much higher resistance since they're ungrounded. These were checked with a battery power multi-meter rather than a megger. Do I need to megger test to get accurate insulation resistance?

Any other troubleshooting steps I can do?

 

[IRstuff]

Considering that the reference design (

http://cds.linear.com/docs/en/application-note/an28f.pdf)

for a Type-J thermocouple uses a 1Mohm amplifier feedback resistor, anything remotely close to 110 ohm is problematic. The current load from the circuit in Fig9 would be about 3.5 uA. Your 110 ohm would cause about 0.2 uA additional load, resulting in an apparent 6% error, which would actually be more like 40F error, but there may be mitigating effects, depending on where all the parasitic currents actually go. The bottom line is that the minimum resistance to anything else needs to be at least 1 kohm to ensure that the measurement is correct.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com:

http://www.engineering.com/AskForum/aff/32.aspx

 

[Compositepro]

Your thermocouples are essentially grounded. This may cause a number of problems. It depends on your electronics and if you have any ground loops.

 

[danw2]

6-7 couple years ago I looked at DRY MI, a non hygroscopic mineral insulation for T/Cs that reportedly has much higher resistance than the run-of-the-mill MgO. The sales guy claimed that ordinary MgO has distinctly lower resistance than the dry MI and I was skeptical, so I cut a new thermocouple sheath in two and was quite surprised that the sales guy was right, MgO had resistance on the order of 100K ohms or thereabouts (depends on how far apart the probe points are), whereas the dry MI was megohms of resistance.

Apparently water vapor gets into the MgO (presumably through the transition) and lowers the MgO resistance. The lower MgO resistance better enables chemical pollution of the junction at high temperatures. Not that such applies to your case, but it does tell you that something's wrong with only 100 odd ohms between element and sheath, probably like the MI is soaking wet?

The resistance of a polluted thermocouple junction increases as the junction is less and less 'pure' alloy. A polluted junction no longer tracks the accepted thermocouple tables and associated polynomials that do temperature conversion. You get mV, but the mV aren't valid for the given temperatures for that type thermocouple. The thermocouple is a polluted 'bastard' thermocouple, with no standard table. The altered mV readings are 'drift' from the pure alloy tables.

Honeywell's temperature controllers and PAC T/C input boards measure the T/C loop resistance (probably with the current they use for burn-out detection, I'm speculating) and can alarm a 'warning' at 80 ohms resistance and alarm 'failure imminent' at about 160 ohms. A new, unpolluted isothermal T/C (both ends at the same temperature) will have junction resistance of less than 1 ohm. Extension wire has considerable resistance compared to copper wire, so if you use an ohmmeter, you need to take that into account.

There'll be an error when the T/C EMF bucks the ohmmeter current, too. But it looks like you've got a couple elements to compare against.

But I really suspect mechanical damage where the sheath is crushed and one of the legs is separated from the sheath by a couple grains of wet MgO.

 

[roydm]

Have you considered the possibility of moisture in the well.
If it's orientated vertically the moisture can flash off, cooling the tip condense higher up then run back down to do it over.
Since it's so hot if you vent the well for a few minutes it will drive any moisture out very quickly.