interpreting thermography

[malachi constant]

Hi all,

I'm a licensed engineer, trying to make sense of a thermography report dropped on my desk. It is a new installation (standard 480V & 208V breaker panels and transformers). Owner asked the contractor to provide infrared scans of the equipment to look for potential issues. I'm not sure how to read these.
* Are there certain rise-over-ambient temperatures that are red flags? (Not sure where to get ambient from the data, but I am assuming it is roughly 72F.)
* Panels: The images draw your eyes to the hot spots, but then when you look over to the scale you see the white hot "hot spot" is only 82F, so no big deal, right? What if this 82F hot spot is three phase wires coming out of a breaker, and they stay "hot" all the way out of the panel?
* Out of the 100 or so panels I only see a couple that have any spots over 90F, and only one that approaches 100F. The 100F probably merits attention - in a 100A 208V panel, looks like it is at a neutral or ground bar (though they didn't provide a visible light photo of that equipment so I can't see very clearly what I'm looking at).
* Transformer connections look good, but a bar of some sort near the top of the transformers is frequently up in the 130-150F range. I'm assuming this is not an issue as it appears in nearly every transformer.
* Motor disconnects: most look fine, but on a makeup air disconnect there's a component of some sort that is pushing 170F. A compressor disconnect hits 145F.

Normally I can learn a skill like this by doing some research online, but haven't found the right links that break down the images and state what is advisory vs serious vs critical. I'm sure it is somewhat of an art, but also suspect with a few pointers as a licensed engineer I should be able to read these with a degree of confidence. Any help is appreciated!

 

[Bugman1400]

Typically the temperatures are a rise over ambient. So, make sure the 82F is really what you think it is. It doesn't make sense to show 82F as a hot spot when the ambient is 72F. The cameras also allow you to adjust the colors associated with the rise. So, you can set a white hot spot to be 20 deg over ambient or 150 deg over ambient. Another common mistake I see with thermographers is that they fail to check loading schedule. They will often perform their scan when the circuit is lightly loaded.

 

[big john]

We always include descriptions of what the fault is, what the severity is, and what the likely corrective action would be. I can't imagine charging a customer to just give them a bunch of bare pictures with no interpretation, and if I were in your shoes I'd be questioning how much I paid for that service.

To answer your questions:

Are there certain rise-over-ambient temperatures that are red flags? (Not sure where to get ambient from the data, but I am assuming it is roughly 72F.)

I don't like going by ambient temperature precisely because without the manufacturer data for every component you don't know what an acceptable rise is. E.g., the thermogram for an encapsulated transformer looks like it's about to explode, when in reality it's entirely normal for it to be running 90 degrees hotter than ambient.

Compare to similar equipment operating under the same load conditions and compare different phases within the same piece of equipment, again considering current balance. There are tables from NETA and EPRI that will give recommended rise between similar components. If the ΔT is less than 10 degrees C, we consider that minor.

The images draw your eyes to the hot spots, but then when you look over to the scale you see the white hot "hot spot" is only 82F, so no big deal, right?

Likely correct. Sounds like whoever was taking the photos didn't properly adjust their span or their wouldn't be glaring hotspots for minor temperature increases.

What if this 82F hot spot is three phase wires coming out of a breaker, and they stay "hot" all the way out of the panel?

If it's uniformly hot, then that means the signature is from resistive heating of the conductor. If all conductors are the same size and insulation, then that one is carrying a higher current than the other phases.

Transformer connections look good, but a bar of some sort near the top of the transformers is frequently up in the 130-150F range. I'm assuming this is not an issue as it appears in nearly every transformer.

Might be the iron core sticking up above the insulation? Can you post it?

Motor disconnects: most look fine, but on a makeup air disconnect there's a component of some sort that is pushing 170F. A compressor disconnect hits 145F

Are they combination motor starters with contactors and thermal overloads? These can get very hot during normal operation. But if these are simple knife disconnects, you have a problem. Can you post those?

 

[zog]

We always include descriptions of what the fault is, what the severity is, and what the likely corrective action would be. I can't imagine charging a customer to just give them a bunch of bare pictures with no interpretation, and if I were in your shoes I'd be questioning how much I paid for that service.

That's what low bid gets you

Compare to similar equipment operating under the same load conditions and compare different phases within the same piece of equipment, again considering current balance. There are tables from NETA and EPRI that will give recommended rise between similar components. If the ΔT is less than 10 degrees C, we consider that minor.

Attached

Sounds like whoever was taking the photos didn't properly adjust their span or their wouldn't be glaring hotspots for minor temperature increases.

This is why we got out of IR scanning, can't compete price wise with non certified guys doing work like this and not providing a decent report.