Motor Thermal Capacity

[philly]

I'm trying to figure out exactly how a motor protection relay defines motor thermal capacity and how rtd's in a motor play a role in determining this thermal capacity. We have a multilin 269 relay which will occassionly give us a thermal capacity alarm when the motor is operated near its FLA. I belive the relay is set up to alarm when motor is above 90% thermal capacity.

My question is, how thermal capacity in a motor is defined with a relay. Is thermal capacity strictly a function of the protection curve set up in the relay? Do RTD's add to the thermal capacity when they detect extra heat in the motor? Do they add to the capacity when the motor is operating both above and below FLA?

I was always understanding that thermal capacity was determined by what percent the motor current was operating below the relay overload curve. So if the motor was operating at 90% of FLA then the thermal capacity would roughly be around 90%? If the motor had additional heating for some reason the rtd's would add to this 90%?

 

[Jraef]

I'm trying to figure out exactly how a motor protection relay defines motor thermal capacity and how rtd's in a motor play a role in determining this thermal capacity. We have a multilin 269 relay which will occassionly give us a thermal capacity alarm when the motor is operated near its FLA. I belive the relay is set up to alarm when motor is above 90% thermal capacity.

My question is, how thermal capacity in a motor is defined with a relay. Is thermal capacity strictly a function of the protection curve set up in the relay? Do RTD's add to the thermal capacity when they detect extra heat in the motor? Do they add to the capacity when the motor is operating both above and below FLA?

I was always understanding that thermal capacity was determined by what percent the motor current was operating below the relay overload curve. So if the motor was operating at 90% of FLA then the thermal capacity would roughly be around 90%? If the motor had additional heating for some reason the rtd's would add to this 90%?

It's somewhat complex in MPRs like the 269, even more so in more recent versions like the 369 and 460. But to answer your specific questions, yes, the RTDs can have an effect on the thermal model if programmed to do so. In the 269, you had to have the RTD add-on module. In the setup for that module, you have the option called "Bias the Thermal Model" which means if selected, the current based thermal model will be altered (biased) by another algorithm coming from the RTD readings. It isn't as simple as just reading them, it averages and uses "voting" (if selected) to toss out the readings of any adjacent RTDs that don't agree with each other. That prevents a single faulty RTD from giving false trips or preventing one that should have happened.

Another critical issue that will bias the thermal model is imbalance of the phase currents. Imbalanced current will cause what are called "negative sequence currents" to flow in the motor windings. These negative sequence current creat negative torque, essentially braking torque, in the rotor and thereby increase the rotor temperature disproportionately faster than normal, but because the stator current will remain below "normal" levels, the thermal model will not see that temperature increase. So when the relay detects a current imbalance, even if not enough t trigger a trip on imbalance, it will still bias the thermal image t account for the added rotor heating. Keep in mind, because this heating is happening in the rotor, any RTDs will not see it until long after the rotor damge curve has bee exceeded. I can't remember if you can disable that in the 269, you can in later versions.

 

[benaround]

Jraef,

You are alright, man. that's all I have to say.

 

[philly]

Jraef

Thanks for the great response. So it looks like the RTD's should only bias or add to the totoal thermal capacity.

There are reports that a partiular 269 relay was getting high RTD alarms. I dont think that the RTD should be in alarm (alarm set at 135 deg c) unless there is something wrong with the motor. Even near FLA the RTD's should not be in alarm or close to it. Maybe like you said if there is something wrong with a winding, or there is negative sequence currents this could lead to additional heating and cause alarms.

So aside from the RTD's biasing the thermal capacity, is the capacity strictly a function of where the motor is operating under the relay curve for any given current and time value?

 

[weressl]

It's somewhat complex in MPRs like the 269, even more so in more recent versions like the 369 and 460. But to answer your specific questions, yes, the RTDs can have an effect on the thermal model if programmed to do so. In the 269, you had to have the RTD add-on module. In the setup for that module, you have the option called "Bias the Thermal Model" which means if selected, the current based thermal model will be altered (biased) by another algorithm coming from the RTD readings. It isn't as simple as just reading them, it averages and uses "voting" (if selected) to toss out the readings of any adjacent RTDs that don't agree with each other. That prevents a single faulty RTD from giving false trips or preventing one that should have happened.

Another critical issue that will bias the thermal model is imbalance of the phase currents. Imbalanced current will cause what are called "negative sequence currents" to flow in the motor windings. These negative sequence current creat negative torque, essentially braking torque, in the rotor and thereby increase the rotor temperature disproportionately faster than normal, but because the stator current will remain below "normal" levels, the thermal model will not see that temperature increase. So when the relay detects a current imbalance, even if not enough t trigger a trip on imbalance, it will still bias the thermal image t account for the added rotor heating. Keep in mind, because this heating is happening in the rotor, any RTDs will not see it until long after the rotor damge curve has bee exceeded. I can't remember if you can disable that in the 269, you can in later versions.

All the MPR's have a thermal mass algorithm model for typical motors. It's the Holy Grail of the protection. Since this is a broad model when actual temperature is available from the winding along with current data the model can be adjusted. The adjustment allows for a better heat-up/cool-down model so the motor can be operated at a wider margin than with current data only, quicker restart times and superior single phasing protection.

 

[philly]

To be a little more specific about the problem I'm having, we have recently been seeing alot of RTD temp alarms on a 4.16kV 2300hp motor. The RTDs in the multilin 269 are set to alarm at 135deg C and trip at 140deg C.

I also discovered today that the PFC caps for this motor are installed on the line side of the relay CT's. We had previoulsy thought this PFC was installed on the load side of the CT and adjusted the multilin FLA setpoint to account for this current from the cap. Now that its discovered that the cap is on the line side, we need to set the relay FLA setting to the FLA setting of the motor.

I still cannont seem to figure out why we are getting RTD alarms though. I would have thought that do the the fact we were limiting the current below the motors FLA that we would be keeping the thermal heat in the motor lower due to us not letting the motor run up to near full capacity.

Any ideas what can be causing these RTD alarms? It seems to happen on different stator RTD's each time. The only thinng that comes to mind is current imbalance as mentioned above however the relay is set to alarm at 10% imbalance and trip at 15% imbalance. Would values less than 10% casue signifigant heating in the stator? Hopefully tomorrow we will be able to run the motor again, and I'll be able to verify exactly what amount of inbalance if any there is.

The one strange thing I was was that the multilin showed its last trip as being an RTD trip. It showed the RTD pre-trip value of 74 deg C, which is much lower than the RTD trip point. Also the pre-trip current was shown as being 0A. This would imply that the motor was idle. This did not make sense.

 

[Jraef]

The "rule of thumb" I've always heard is that every 1% of current imbalance increases the winding temperature rise by 10%. So a 5% current imbalance can result in a 50% increase in temperature rise. Depending on how close you are to the limits already, this can be a significant risk to your motor even when the current imbalance is well below 10%.

 

[Jraef]

...

The one strange thing I was was that the multilin showed its last trip as being an RTD trip. It showed the RTD pre-trip value of 74 deg C, which is much lower than the RTD trip point. Also the pre-trip current was shown as being 0A. This would imply that the motor was idle. This did not make sense.

That can be a bad RTD, a bad conductor or a bad transducer inside the RTD module. In general, the A-D converts tend to go "High" on loss of signal or failure of the converter. That means the register value goes to 65536 and gets interpreted as something like 10,000 degrees C. If something is flickering on and off, the microprocessor averages those signals and it ends up with a falsely high value. Noise on the RTD conductors can do that as well. Check for a broken shield, an improper shield connection (the shield must only be grounded at one end) or someone having routed an MV cable differently, etc. etc.

 

[philly]

The "rule of thumb" I've always heard is that every 1% of current imbalance increases the winding temperature rise by 10%. So a 5% current imbalance can result in a 50% increase in temperature rise. Depending on how close you are to the limits already, this can be a significant risk to your motor even when the current imbalance is well below 10%.

That is great information. Do you know where I can find a reference for this. I have seen some other information that states that the temp rise can be calculated from voltage imbalance by the equation:

% temp rise = 2(% Voltage unbalance)^2

I'm going to look at current imbalance and voltage unbalance once we start the motor again. For voltage unbalance I can measure them on the secondary of a PT so I'm assuming I'm looking at the L-L voltages when calculating this unbalance?

We checked the RTD connections at both the multilin and the motor, and they seem to be fine. We thought about the possibilty of replacing the multiln if we didn't find a smoking gun. There was also a reading in the multilin that showed the max temperature since the RTD temps were last viewed, and it showed the same RTD mentioned above has having had a temp of 171deg at one point. All of the RTD's seemed to be reading the same however today while the motor was sitting idle.

With the PFC's on the line side of the CT's is it possible that a faulty capacity could cause unbalanced currents on the motor?

 

[jamieb]

with regards to faulty capacitance, it will definetly cause current imbalance at your motor, i've had some strange readings on our multi-lins ( r.t.d. temps. ) at our plant and they seem to come and go although i have downloaded some new upgrades ( which you can get on-line ) and that seems to have solved some of my problems

 

[philly]

with regards to faulty capacitance, it will definetly cause current imbalance at your motor, i've had some strange readings on our multi-lins ( r.t.d. temps. ) at our plant and they seem to come and go although i have downloaded some new upgrades ( which you can get on-line ) and that seems to have solved some of my problems

How would the capacitors being on the line side of the CT's cause a currnent unbalance in the motor? If one of the caps was damaged wouldn't that reactive current then come from the line on that phase and still provide equal current to the motor?

 

[philly]

We re-started this motor this morning and I was able to look at some values.

The motor was running at about 265A (average) and the RTD's were reading about 110deg C. The full load on this motor is 312 although currently is is set at 275 due to the fact that it was believed PFC caps were after CT's. I need to re-adjust FLA in multilin but want to watch the motor first.

This motor runs a crushing mill so the load looks like its jumping all over the place, and it was hard to get steady current readings. When the current valus were changing, they did not all change in unicen, and they would jump around at all different values. I took a few recordings as it was jumping around in an attempt to calculate current imbalance and came up with an average of about 2.5-3% unbalance. I would think that these values should all change together, but one would change while the other stayed put, and then another would change etc.... with this 3% imbalance its not large, but I guess it may be large enough to produce additional heating?

All RTD's looked about the same, so I dont believe there is an RTD problem.

The thermal capacity when the motor was running was about 28%, so I guess this value does not corrospond to the overload curve, since we are operating up near full load I would expect this to be higher?

 

[Jraef]

Currents would not necessarily change in unison, but if the disparity is extended (i.e. more than a second) then I would be concerned about a problem in your motor's rotor cage. A cracked rotor bar, either in the bar itself or in the end ring connection, can cause the phenomenon you describe, especially on constantly changing loads such as a rock crusher. Have you had the motor rebuilt recently? The traditional test to find that kind of thing involves removing the rotor and putting it on a "growler", something not easily done on a large crusher motor, so you may need to wait until it needs other PM services. There are some new tests that have been discussed however, here is a paper on the subject.

Another

and another

These types of tests would need to be performed by an expert with the right test equipment, something that may cost as much as removing it and sending it to a reputable motor shop. That's why I mentioned waiting for a PM shutdown anyway. If it's a critical path production machine you should have a spare motor available anyway; I would consider putting the spare into service and having this one thoroughly checked out.

 

[weressl]

We re-started this motor this morning and I was able to look at some values.

The motor was running at about 265A (average) and the RTD's were reading about 110deg C. The full load on this motor is 312 although currently is is set at 275 due to the fact that it was believed PFC caps were after CT's. I need to re-adjust FLA in multilin but want to watch the motor first.

This motor runs a crushing mill so the load looks like its jumping all over the place, and it was hard to get steady current readings. When the current valus were changing, they did not all change in unicen, and they would jump around at all different values. I took a few recordings as it was jumping around in an attempt to calculate current imbalance and came up with an average of about 2.5-3% unbalance. I would think that these values should all change together, but one would change while the other stayed put, and then another would change etc.... with this 3% imbalance its not large, but I guess it may be large enough to produce additional heating?

All RTD's looked about the same, so I dont believe there is an RTD problem.

The thermal capacity when the motor was running was about 28%, so I guess this value does not corrospond to the overload curve, since we are operating up near full load I would expect this to be higher?

The type of load you are describing is the most difficult to model. Heat is produced by I Squared t and the irregurarly pulsating current time integer is the issue. Since thermal overloads are analog mimicing the thermal profile it was not a problem, so it is only a problem with digital protection. RTD is a great help to make the adjustment and correct anomalies.