RTD Terminatation

[himmyhat]

We recently ran temporary extension wire for RTD sensors on our Benzene and Toluene columns. The electricians used crimp-on connectors to terminate the connections. On each Column there is a set of RTDs to measure temperature differential across the appropriate trays. Since this unhappy ocurrence, we have not had satisfactory results in our temperature differential measurement.Before this change,the temperature differentials were what they were supposed to be. The electricians used 18 gauge twisted and shielded Belden to replace 8 gauge solid core copper wire.I am looking for any input I can get as to why this may be ocurring.

 

[TD2K]

Signal degeneration maybe? You are using a lot thinner wire than previous and the crimp-on connections may be giving you additional resistance.

What sort of signal strength should the RTDs be developing for the dTs you expect versus possible instrumentation line losses? Have you talked to the RTD vendor to find their wiring recommendations or ideas on this? Have you an instrumentation engineer to talk to about this as this is something I've no experience with to help you.

 

[hacksaw]

The problem is not in your extension wires.

Make sure that the rtd's are properly installed (3-wire vs 4-wire vs 2-wire). Also pay attention to the possibility of ground loops. Soe rtd's are grounded (believe it or not)and can cause all kinds of wiring fiascos.

Fundamentally you can check the rtd with a good multimeter to confirm that every thing is okay with the sensor and work your way back to the controller input or rtd transducer/transmitter.

 

[TD2K]

Thanks Hacksaw, be interesting to hear back from the OP if he found the problem.

 

[himmyhat]

Thanks to TD2K and hacksaw for your input. We have checked the RTDs on the Benzene Column (2-wire 100 ohm platinum) and they both read the same resistance; therefore the same temperature. This translates to a delta T of 0 across the fifth and fifteenth trays of said tower. From my own experience, plus the screaming and hollering of the operators, this scenario appears to be highly unlikely. However, if there are any ChE.s out there who could explain how this could be taking place in a distillation column with trays,I am all ears. The situation in the Toluene Column is essentially the same. From the research I was able to do this weekend, 4-wire RTDs would improve this situation greatly.Again, any feedback on this situation would be greatly appreciated. Thank You.

 

[hacksaw]

I agree with the operators, you are not going to get "identical readings" from two rtd's in a column. Even factory matched rtd's have detectable differences.

Having faced this same issue in the past, about the only recourse you have is to put a test resistor(or test rtd) in the circuit at the connection head.

Without demonstrable evidence that it is not the sensors, then what do you want the operators to do?

Back to the tower: about the only thing I've seen along the lines of inexplicable readings given that all flows were normal, was collapsed or lifted trays. We had to get the gamma/xray ray people out to sort that mess out.

Good luck,

 

[209larry]

himmyhat,

The best way to detect temperature differential is to use two thermocouple elements wired in series without cold junction.

The reason why the temperatures on both trays are the same on RTD is that the difference is negligible with respect to the ranges of tx's especially when you have a 4-20 ma transmitters or converters.



Regards,

209larry

 

[himmyhat]

Thanks to hacksaw and 209larry for your observations. I am not familiar with the test resistor/test RTD concept. If you could provide additional information I would appreciate it. After thinking about this all weekend, I am wondering if the extension wire is acting like one long RTD since it is just tie-wrapped to the side of the tower. The tower is insulated, but the insulation is very old. The problem has to involve the new wire, because there was no problem before the new wire was swapped with the old. Again, I appreciate everybody's interest and participation. As they say, keep those cards and letters coming.

 

[hacksaw]

The two wire rtd's are limited to 100-200 feet of wire because your wire resistance is added to the measured resitance. In a pinch you can zero out the wire resistance, but it won't be perfect. Use of 8 ga wire sounds excessive. You can accomplish the same result be paralleling conductors to get the loop resistance of the wiring down.

If you have rtd transducers located near the connection heads, then the problem is lessened. Ditto 3-,4-wire configurations.

A test rtd is a spare sensor pulled from stock. you can also use the existing rtd if you pull it an hold it at known temperature so you can zero your instrument loop.

Some instrument shops have a test box (decade box or a couple of 1% temperature compensated resistors to simulate the temperatures being measured and calibrate the loop.

Transmation is one supplier of such devices.
the rtd is 100 ohms at 0 deg C, it is pretty much linear with temperature 0.00385 ohms/deg c being a common calibration (DIN standard). An accurate low current ohm meter can read the rtd temperature. by shorting the rtd at the connection head you can also read you wiring loop resistance and identify the equivalent temperature error.

sounds like fun, in less you're the instrument engr. in that case you're the designated target...good luck. Look at it this way, "it builds character".

 

[scott1971]

Himmyhat:

Following up on Hacksaw's comments, I pretty sure he's right on about adding extensions to RTD leads. Can you use local signal conditioners (RTD to analog 4-20 mA) and change your respective inputs on your controller/recorder or whatever it is you're using. This would eliminate the extensions. Let me know.