Dear Team members,
I am facing a very peculiar situation related to Multi point RTD operation. Here is the scenario.
This multi point RTD( is being used in Hydrogen Generation via Methanol and water reaction. Temperature goes to around 270~280 deg C. There are various region inside the reactor where reaction takes place and this reaction releases the heat. This temperature is measured by the Multi point RTD. In our case, there are 6 RTDs in this Multi point RTD. This RTD is connected from the bottom without any thermowell.
After one of our SDM, this RTD started malfunctioning. This RTD was showing less than normal as compared to before. Following things done after identifying instrument malfunction.
1. RTD connection to DCS removed and a rheostat was connected to each point. No problem found with the DCS card or cable. DCS value was corresponding to the resistance of rheostat.
2. RTD connection removed at field and resistance was measured. We found that the resistance corresponds to the temperature. Its only when connected with DCS, the temperature drops for all 6 RTDs.
3. Just to confirm, we connected the Multi point RTD with a Temperature recorder. Then also the temperature drops for all 6.
We are not able to identify the problem. Discussion with the vendor also did not solve the problem. So i was hoping to get some insights from this group

 

What is an SDM ?

Check for continuity from either RTD lead to the sheath. If one of the RTDs is no longer insulated from the sheath and its elemnet is touching the protection sheath and the protection sheath is at a different potential than the current source return, then there will be a ground loop current generated that affects measurement.

Did the SDM physically damage the RTD ?

 

What is an SDM ?

Check for continuity from either RTD lead to the sheath. If one of the RTDs is no longer insulated from the sheath and its elemnet is touching the protection sheath and the protection sheath is at a different potential than the current source return, then there will be a ground loop current generated that affects measurement.

Did the SDM physically damage the RTD ?

Sorry David, it's my bad. Thanks for the reply. SDM is shutdown maintenance.
If you are talking about checking continuity, then yes we did that and found the there is none between element and sheath. I will share the details in my next post

 

what style of construction is this multi RTD?

Six individual RTD's each of a different length so that temperature is measured along the axis of the RTD?

One set of wires (conductors), (2 or 3?) for each RTD? So there are 6 sets of wires coming out the end of the RTD assembly?

Connection is one set of wires to one RTD input, requiring 6 RTD inputs?

 

what style of construction is this multi RTD?

Six individual RTD's each of a different length so that temperature is measured along the axis of the RTD?

One set of wires (conductors), (2 or 3?) for each RTD? So there are 6 sets of wires coming out the end of the RTD assembly?

Connection is one set of wires to one RTD input, requiring 6 RTD inputs?

Hello David,
Yes that is absolutely correct. I will send the GA drawing tomorrow for better clarification
Thanks for the reply

 

Somehow the measured voltage drop from the receiver AI's constant current source is lower than expected for two different receivers, the DCS and the temperature recorder. It is highly improbable that two receivers develop faulty constant current sources on their multi-input RTD AI's at the same time.

The lower resistance creating a lower voltage drop resulting in a lower calculated resistance and a lower reported temperature must be internal to the RTD assembly.

How is the transition - where the conductors enter/exit the sheath - sealed? Has the seal been broken? Has water or liquid entered the RTD assembly?

Could water (or any liquid) have entered the internals of the RTD protection sheath through either a hole in the sheath or more likely, through the transition? Conductive water inside the sheath could create all sorts of alternative paths for what should be a constant current in a given circuit or it could affect the 'resistance' that the current sees.

 

Somehow the measured voltage drop from the receiver AI's constant current source is lower than expected for two different receivers, the DCS and the temperature recorder. It is highly improbable that two receivers develop faulty constant current sources on their multi-input RTD AI's at the same time.

The lower resistance creating a lower voltage drop resulting in a lower calculated resistance and a lower reported temperature must be internal to the RTD assembly.

How is the transition - where the conductors enter/exit the sheath - sealed? Has the seal been broken? Has water or liquid entered the RTD assembly?

Could water (or any liquid) have entered the internals of the RTD protection sheath through either a hole in the sheath or more likely, through the transition? Conductive water inside the sheath could create all sorts of alternative paths for what should be a constant current in a given circuit or it could affect the 'resistance' that the current sees.

Hello David,

Thanks for the reply.
Attached with this post is the GA drawing of the multi point RTD.
About the issue with the RTD assembly, we conducted Dye Penetration test on the probe which is actually a 1/2" schedule 80 pipe. Inside the pipe are 7 nos RTD probe. After conducting Dye test, we did not find any breakage or leak mark on the pipe.
Apart from that, we checked the insulation resistance of each RTD. Resistance of each RTD cable was checked with respect to body and we did not find any problem. Also each RTD assembly is embedded in Aluminum Oxide powder for insulation against each other
About your query that water could have entered the internals , that's a possibility. For that we have to remove the probes from the pipe. Please take a look into the RTD GAD attached with this post.

 

1. parasitic parallel resistance path

The RTD troubleshooting table at the link below is titled 'common' RTD problems and you'll note that there's no category for "lower than expected readings" because most RTD errors are "higher than expected" problems.
https://www.instrumentationtoolbox.com/2014/07/common-rtd-application-problems-and.html

Lower than expected readings require a lower resistance (Pt100's have a positive temperature coefficients) So what can lower all eight RTD element resistances at the same time? (not an aging, drift issue)

When there is a resistance in parallel with the element, the combined parallel resistance is lower than either individual resistance (Ohms' Law).

The mineral insulation (MgO, the white powder) used to isolate one element from another and to prevent the element from banging around the in the protection tube can absorb water. I cut a new RTD sheath in half with a grinder and measured the resistance of the MgO with a regular DVM. I left one half open to atmosphere and then measured its resistance every week for a month and noted how the resistance of the MgO dropped noticeably as the MgO absorbed humidity from the air. The other half I left submerged in water for a day and that resistance dropped orders of magnitude lower than the one left open to the atmosphere.

If water (or any conductive liquid) has gotten into the MgO the water could provide a parasitic current path in parallel to the RTD element.

Even if RTD's in the protection pipe have SS sheaths, the transition where the lead wires enter the sheath could allow water entry. Only the vendor can tell you what the transition construction and materials of construction are.

2. Flooded Intermediate junction box

If your DCS/recorder test involved an intermediate junction box between the terminal box on the probe assembly and the DCS/recorder AI's then that junction box could be wet or even filled with water. I'm sure your electricians can tell you stories about junction boxes filling with water/condensation, it is not an uncommon problem. Wet terminals in a junction box cause ground loops and can provide an alternative path for the RTD source current.

 

definitely dealing with a field wiring issue.

 

I personally have very bad experiences using RTDs in industrial processes and prefer the ruggedness and more immunity to cabling and noise of a well calibrated thermocouple, especially for these kinds of temperatures.
Is there a specific reason why RTDs are being used and not thermocouples for this specific application?

 

I personally have very bad experiences using RTDs in industrial processes and prefer the ruggedness and more immunity to cabling and noise of a well calibrated thermocouple, especially for these kinds of temperatures.
Is there a specific reason why RTDs are being used and not thermocouples for this specific application?

Dear patrickduis

Selection of RTD was based on vendor's recommendation. Vendor had provided this RTD everywhere else, but did not find any issue with the instrument. Here only, we are facing this issue.
David above suggested some check points which will be definitely be checked in the next opportunity.

 

I'm not a fan of using RTDs for this temperature range in particular, or even in general as their wiring is often misunderstood. EVERY vendor is going to say the same thing, "We have never had problems with [this] anywhere else."

Given the problems only reportedly started after the SDM (Shut Down Maintenance) it's extremely possible the device was damaged during removal or re-installation or the wiring was damaged or re-connected improperly, etc. I agree with patrickduis that--based on the information provided--this is most likely a field wiring or a device issue since tests have shown the montioring equipment is good.

It's presumed the device can't be changed now that the SDM is over, or not easily or not without another SD (shutdown). At this point the best course of action, barring price concerns, would probably be to just replace the device since a SD is going to likely be costly and another one to eventually replace the device would add to the lost revenue/product.

With the suspect device removed and--hopefully--the new device working properly, the suspect device can be examined in detail to try to determine the cause of the reading errors.

One thing that's NOT clear based on the information provided is whether the test instruments were connected directly to the monitor (after removing the interconnecting field wiring) or at the wiring at the location of the field device (after disconnecting the field device). This would rule out interconnecting field wiring as the source of the error (or poor terminations or moisture in intermediate JBs, etc.).

Please write back to let us know what you find!

 

I'm not a fan of using RTDs for this temperature range in particular, or even in general as their wiring is often misunderstood. EVERY vendor is going to say the same thing, "We have never had problems with [this] anywhere else."

Given the problems only reportedly started after the SDM (Shut Down Maintenance) it's extremely possible the device was damaged during removal or re-installation or the wiring was damaged or re-connected improperly, etc. I agree with patrickduis that--based on the information provided--this is most likely a field wiring or a device issue since tests have shown the montioring equipment is good.

It's presumed the device can't be changed now that the SDM is over, or not easily or not without another SD (shutdown). At this point the best course of action, barring price concerns, would probably be to just replace the device since a SD is going to likely be costly and another one to eventually replace the device would add to the lost revenue/product.

With the suspect device removed and--hopefully--the new device working properly, the suspect device can be examined in detail to try to determine the cause of the reading errors.

One thing that's NOT clear based on the information provided is whether the test instruments were connected directly to the monitor (after removing the interconnecting field wiring) or at the wiring at the location of the field device (after disconnecting the field device). This would rule out interconnecting field wiring as the source of the error (or poor terminations or moisture in intermediate JBs, etc.).

Please write back to let us know what you find!

Dear CSA,

This is what we did to confirm the problem.

1. We removed the RTD connection in the field Junction Box and checked the resistance. Based on Resistance Vs Temperature chart , we found that the temperature based on the chart was as per the chart value.
2. We removed the RTD connection in the field Junction Box and connected the Rheostat. We were getting the temperature on DCS based on the resistance we feeded from the Rheostat. Based on this , we concluded that the cables and temperature card were OK.
3. We checked the insulation resistance of each cable of RTD and found it to be ok.
4. Then we connected a temperature recorder in the field and connected the RTD with the recorder and found the temperature shown on the recorder was less than the desired value mentioned in the chart.

 

1. > We removed the RTD connection in the field Junction Box and checked the resistance.

With a battery powered DVM, correct? A battery powered DVM is floating, it is not connected to a ground, whereas the DCS and temperature recorder analog input circuits are grounded at some point. The DVM test does not rule out a ground loop or parisitic current between the DCS and the RTD assembly.

2. >We removed the RTD connection in the field Junction Box and connected the Rheostat. We were getting the temperature on DCS based on the resistance we feeded from the Rheostat. Based on this , we concluded that the cables and temperature card were OK.

This tells you that a floating, ungrounded rheostat, probably a decade resistance box, produces correct values on the DCS and it probably rules out an intermediate junction box problem. But this test likewise does not rule a ground loop or parastic currents between the RTD assembly and the AC powered DCS.

3. >We checked the insulation resistance of each cable of RTD and found it to be ok

Please describe what the test is, and exactly how it is performed, and with what instrumentation.

4. Then we connected a temperature recorder in the field and connected the RTD with the recorder and found the temperature shown on the recorder was less than the desired value mentioned in the chart.

The temperature recorder is most likely AC power and its analog inputs are similar to the DCS analog inputs and eventually see a ground.

A. Question - what magnitude of error do the DCS and temperature recorder report? 1°C low? 5°C low? 10°C low?

Do all channels show about the same deviation or is there a wide spread of error?

B. Question - can you substitute a battery powered electronic RTD calibrator that reports in temperature units in place of the temperature recorder on one channel and see whether the reported calibrator temperature reads low compared to the DVM resistance reading from the chart?

 

at the connection head (disconnected from the DCS wiring) what is the resistance to the thermowell?