Message Board Thread - "what do you think ?"

Back to Threads | Back to Forums

TitleByPosted On
what do you think ? eureka 12/2/2010
Dear Colleagues,

Yesterday i performed infrared inspection on SF6 breaker cubicle and i need advice and also second opinion about my finding below.

The target is 11.8 KV SF6 circuit breaker and carying load 3216 A (nominal current is 3500 A).
Apparent temperature is 35 degree C, distance 0.3 m and the breaker connections is located in the cubicle that some IR windows already installed (CaF2 type IR windows)

although it is hard to measure shiny busbar connections but i believe that because of geometry effect of the connections it can rise the emmisivity. after the transmissivity of the IR windows already compensated i found that the pattern of the picture shows high temperature measurement compare to the adjacent connections.

How do you think about that?

i already notify the operator to safely shutdown the unit because of it is very dangerous to operate the plant with current condition

need your advice
sofan
 
Re:what do you think ? eureka 12/2/2010
I also attached another picture of the connections :-)
 
Re:what do you think ? manuel-thermoimagen 12/2/2010
When i download and review your images, found that you have used emissivity value on 0.38 afecting huge the temp readings.. can you doble check this value and verify that image parameters are ok?.. otherwise, image readings are all wrongs.
roberto.cruz@thermoimagen.com
 
Re:what do you think ? eureka 12/3/2010
manuel-thermoimagen wrote:
download and review your images, found that you have used emissivity value on 0.38 afecting huge the temp readings.. can you doble check this value and verify that image parameters are ok?.. otherwise, image readings are all wrongs.
roberto.cruz@thermoimagen.com
Manuel,

Thank you very much for your feedback... i really appreciate..i want to explain why i wrote low emmisivity value in the picture.

first of all, the connections that i already inspected are made from shiny busbar (tin coated cooper bar) which have very low emmisivity..it will cause the reading have huge error. but because of it has geometry shape, the emmisivity will rise altough a little.

Second,
the connections are enclosed in the panel and i only can inspect the connections trough infrared windows. The infrared windows made from CaF2, that have the best emmisivity value of 0.5 if we use LW camera ( I use E320 camera )

assumed that the emmisivity of the connection (which have a lot of bolts and nuts that create geometry effect) is 0.75 (which i think that it is too high) and i compensate the IR windows transmissivity 0.5 so the results of new emmisivity is 0.375.
if the real emmisivity is lower than 0.75 the new emmisivity will have lower value, and thats mean we will see higher temperature than this

if...the breaker is being shut down and i can check the real comparation to make sure the emmisivity and the transmissivity of the IR windows. i think that i will get higher temperature reading than now

so that in this picture i think that the temperature can be much higher than the camera reading, because it is little bit complicated if we measure shiny busbar trough infrared windows

thanks again for your input
any other suggestion and advice?
 
Re:what do you think ? IRJay 12/3/2010
You have made some proper assumptions. Yes this is a hot connection and needs some repair. But ... when looking through a window such as this with the transmissivty of 50%. You cannot just take 50% of the assumed emissivity of the part. The 50% means that only 50% of the effective radiation from the part is making it through the window. Compensation for a window can be accomplished but you must know an element of inaccuracy is going to made. Not knowing the effective emissivity of the connection is just as large a factor as the window in the way. FLIR software does have a transmissivity adjustment to help in this function. Remember too that the window transmissivity is temperature dependent and that must be included. Trying to come up with an "absolute" temperaure in this scenario is extremely difficult but finding a relative decent decision should be the goal. Note that the temperatures are the coldest this part will be as you have shown it. I also will note that the switch is at virtually full load, which would make this part at it's worst now. This is a factor to consider in that is this was half loaded with the potential to go full load this would be more critical.
if you want, send me the orginal image and I will adjust some of the values to relate a more realistic temperature measurement or verify your existing finding.
Jay.bowen@infraredtraining.com
 
Re:what do you think ? Hill Top TI 12/3/2010
IRJay wrote:
e made some proper assumptions. Yes this is a hot connection and needs some repair. But ... when looking through a window such as this with the transmissivty of 50%. You cannot just take 50% of the assumed emissivity of the part. The 50% means that only 50% of the effective radiation from the part is making it through the window. Compensation for a window can be accomplished but you must know an element of inaccuracy is going to made. Not knowing the effective emissivity of the connection is just as large a factor as the window in the way. FLIR software does have a transmissivity adjustment to help in this function. Remember too that the window transmissivity is temperature dependent and that must be included. Trying to come up with an "absolute" temperaure in this scenario is extremely difficult but finding a relative decent decision should be the goal. Note that the temperatures are the coldest this part will be as you have shown it. I also will note that the switch is at virtually full load, which would make this part at it's worst now. This is a factor to consider in that is this was half loaded with the potential to go full load this would be more critical.
if you want, send me the orginal image and I will adjust some of the values to relate a more realistic temperature measurement or verify your existing finding.
Jay.bowen@infraredtraining.com
Why are their no signs of heat conduction? I see no signs. Looks like it background ambient reflection from some where. Just asking.
 
Re:what do you think ? eureka 12/3/2010
Hill Top TI wrote:
Why are their no signs of heat conduction? I see no signs. Looks like it background ambient reflection from some where. Just asking.
Thanks to everyone that involved in this discussion..

to IRjay..
i will send the original picture so that you can clearly analize the image

to Hill top TI

Actually if you use isotherm you can exactly see the heat conduction pattern ;-)

by the way,

i realize that it is difficult (for me at least :-) )to measure exactly the true temperature that we disscussed above, without testing the true emmisivity of the busbar when offline condition and measure the transmissivity of the IR windows at the same time, the best measurement still can be inaccurate. that's why my recomendation of this measurement to the user is that the temperature reading caught by IR camera is the "minimal value" of the temperature, of course it can be worst than that if we get low emmisivity when measuring in offline condition

i already use thermacam reporter to analyze this picture so the transmissivity correction is already inputed. if i can get an better calculation and method to analize the picture i will really appreciate

 
Re:what do you think ? manuel-thermoimagen 12/6/2010
well, indirect & cualitative thermography criteria turns very important subject.

like IRjay says, whatever the temperature are, you have one cool and one hot busbar conection to make a call, something is wrong with left conection.

same error conditions to both of them, same emiss, same transm, same compensation value..

more than 3000amp?.. if wire insulation supports no more than 90°C, then this device is telling you something important.

 
Re:what do you think ? Hill Top TI 12/8/2010
If you change the emissivity to .05 shiny copper this is what I got
 
Re:what do you think ? TDLIR 12/8/2010
To be as accurate a practically possible, take your temps at areas where there are cavities or gaps such as between bus plates, threads, the space between bots/washers/etc...

These areas will have high e values. Basically it will also be the areas that look hotter.

Adjusting your e too low to match the shiny material will over estimate the temps. Now this may be a good thing as it will encourage the plant to fix it sooner.

Regardless, it is a problem that must be fixed. If they cannot because of production issues, then monitor the best you can until they can fix it.
 


  • Back to Threads
  • Back to Forums

     

  •   Copyright © FLIR Systems, Inc 2012