I've been doing some thermography in our Data Center, essentially standard electrical panels, with the exception that the bottoms of the panels are open to the underfloor, where cool air (15C) flows from the AC units, through the floor, and up into the electrical panel. I was suprized how much air flow there is. The panels are usually around 22C. However my question is this: When you have a thermal anomaly that is within a convective flow, essentially an aggressive heat removal system, do you correct or adjust your anomaly severity due to the convective flow?

Yes, but how to do this best is down to opinion. There a some charts that can be used that some feel are good and do help. I do not believe in using them. What I tend to do is if there is a high level of convective cooling, then I just call it a critical problem. My reason is that if the convection stops for any reason the anomaly will be a lot worse.

Thanks Bob for your answer. That sounds like a very practical and reasonable approach.
I was trying to make early identification of carbon build up on the contacts, and trying to judge the criticality of each.
In this particular case, each of the panels were so identical (same panel, same MCCB , same ambient temperature) that I took a thermograph of nearly every breaker from the side, capturing the heat reflected from the back of the breaker off of the galvanized steel back panel.

I also took loads from each breaker(loads varied greatly), of the 19 breakers. Then I subtracted ambient (22c, very well controlled and redundant supplies) and built a scatter plot with Delta T on one side and Load on the other. The outliers were very clear. The reason I went this far is because of the criticality of the equipment (Data Center). In this way I hoped to correct for normal Load heat. As far as correcting for convective cooling, I think your perspective is a good approach. Take a look at the attachment and tell me what you think.

B90_Breaker_Chart.xls (30 Kb, 29 downloads)