PT100 sensor reading going slowly up

[Kataeb]

Have you seen a PT100 sensor reading going slowly up, even after heating power turned off ?! and it keeps rising for several minutes ...
It looks like the sensor is not directly getting the actual temperature of the material to be heated.
This happens only once every one or two month !

 

[drbitboy]

We don't know your process, but I am sure the answer is yes.

My father, who measured the heat rate and efficiency of large steam turbine processes to a an accuracy of quarter of a percent or better, once told me "a thermocouple always measures the temperature of (not at) the tip of the sensor." I am sure the same is true for RTDs.

It is odd that it is a rare event, but as noted above, we do not know your process.

If the sensor is in a thermowell, and the thermowell is in contact with the material being heater, then the sensor should have a mechanical device (spring?) to keep it in contact with the wall of the thermowell. Is it possible that device is not always doing its job?

 

[Kataeb]

The process is hotmelt glue. The sensor is inserted in a hole, with closed end, made on the body of a heated metallic tank, containing the melted glue.
Do you mean if the sensor has some little air space, between its head and the tank, it will read like that ?

 

[drbitboy]

The process is hotmelt glue. The sensor is inserted in a hole, with closed end, made on the body of a heated metallic tank, containing the melted glue.
Do you mean if the sensor has some little air space, between its head and the tank, it will read like that ?

Yes, that is one possibility that will cause a delay in the reading by the PT100 temperature sensing element. That may also cause the PT100 device to read low, even if that low reading still increases after the heating stops, because the thermal resistance to the ambient thermal environment, relative to the thermal resistance to the material, is lower than if the element was touching the side of the thermowell (the hole with a closed end).

Even so, any process will have a time delay between the actual temperature of the material and the measured temperature at the PT100 sensing element. During heating, the material closer to the heating element will be hotter than the material next to the thermowell and sensor, so there will be a temperature gradient through the material, with heat "flowing" "down" the gradient i.e. flowing from the higher temperature material near the heater to the cooler material near the thermowell. At the moment that the heating stops, the thermal gradient still exists, so heat will continue to "flow" toward the cooler material, which flattens out the gradient over time. The actual time-dependent behavior depends on several factors, such has heating rate, mass of material, thermal conductivity and capacitance of the material, ambient conditions, etc.

Mixing reduces the delay, but it always exists, unless the sensor is at the heating element, in which case the sensor might give a reading higher than the mean of the material.

 

[Kataeb]

Strangely, the problem is happening only once every one or two months, then disappearing after a few hours by itself.
We have some vibrations on the floor, due to running fans and pumps, so we also suspect a loose connection or a minor defect inside the sensor.

 

[drbitboy]

Is it related to any of the following?

• Ambient temperature
• Day of week (e.g. Monday morning, after a weekend of inactivity)
• How long the heating element(s) is(are) on

There is always going to be some temperature rise at the sensor after the heating element(s) turns(turn) off; the only variable would be the actual dynamics.

As far as the sensor not touching the inside of the thermowell, I am not sure if that would make this issue more or less severe.

Are there other temperature measurements?

Is this temperature measurement used to control the heating element(s)?

 

[janner_10]

Cheap as chips - replace and work back from there

 

[Kataeb]

Is it related to any of the following?

• Ambient temperature
• Day of week (e.g. Monday morning, after a weekend of inactivity)
• How long the heating element(s) is(are) on

There is always going to be some temperature rise at the sensor after the heating element(s) turns(turn) off; the only variable would be the actual dynamics.

As far as the sensor not touching the inside of the thermowell, I am not sure if that would make this issue more or less severe.

Are there other temperature measurements?

Is this temperature measurement used to control the heating element(s)?

It happens randomly.
The hotmelt unit has several heating zones, each with its sensor and heater powered by an ssr. But only this zone is causing trouble.
All the sensors and ssr are wired to B&R plc modules.
.

 

[danw]

Temperature is a reflection of the heat transfer going on.

As others have pointed out, there is always a lag time in heat transfer from the hotter mass to the cooler mass (law of thermodynamics), reflected as a lag in the temperature increasing or decreasing.

There is a heat transfer paste (some grease doped with silver particles) that is used for sensors in thermowells to displace the air at tip and replace it with a gel paste that helps transfer the heat.

An increase in resistance in either leg of the RTD will appear as a temperature increase. Wiring resistances can be created by
- a loose terminal screw,
- corrosion on a terminal screw,
- a bad crimp on terminal connector or pin that allows internal broken strands to connect or open intermittently changing the wire resistance
- water in a junction box

 

[lfe]

It happens randomly.
The hotmelt unit has several heating zones, each with its sensor and heater powered by an ssr. But only this zone is causing trouble.
All the sensors and ssr are wired to B&R plc modules.
.

Well, probably that heating zone has a problem with its heater, or the circuit that feeds it.
You see that occasionally the temperature rises slowly, but that does not unequivocally mean that the sensor does not measure well

From what you explain, it seems to me that when this problem occurs the heater in that zone is not working and it heats up slowly because it does it thanks to the surrounding zones

 

[Peter Nachtwey]

Borrow a thermal camera and learn what is really happening.

 

[drbitboy]

• Is there thermal crosstalk between the zones?
  • That is, if a heater failed on one zone, would that zone still get heated, although perhaps not to setpoint, by the zones next to it?
• Do all the zones have the same setpoint?
• Are all of the temperature sensors reading correctly, as verified by a separate measurement?
  • See @PeterN's post suggesting a thermal camera
• What ambient or other conditions are unique to the few times this happens per month?

It may be that the problem is with a zone next to the one that shows the problem, and that next zone is what is providing heat to this zone, and that is what causes the temperature to continue to rise after its heater turns off.