Why do we use a Contact Resistance Measurement test?

[Manoj Sahu]

I was wondering about contact resistance measurment test. Why do we take such test?

 

[anorlunda]

Because contacts are a frequent point of failure. The resistance of the contacts causes local $I^2R$ heating at the contacts. As they heat up, the resistance can increase more.

 

[Asymptotic]

It isn't a common measurement in end user power equipment of the sort pictured below (infrared scanning is more typically used), but very common for checking high voltage switch gear. If memory serves, 50 micro ohms is the allowable limit, and (at least, for knife switch contacts) it ranges around 20 to 30 micro ohms in a properly maintained switch.

Continuing with @anorlunda's theme, if a switch is passing 400 amps, 31 μohms dissipates about 5 watts, and at the 50 μohms limit 8 watts of heat is produced. One milliohm of contact resistance doesn't sound like much, but it throws off an impressive 160 watts, and chances are the connection won't survive for very long.

Pole 1 fuse overheated due to poor contact in line side fuse clamp. Note the discoloration of both the copper bar, and fuse label.

Another instance where contact resistance measurement is useful is during a powered-down preventative maintenance episode to find switches that are contaminated with oil, or have tarnished contacts. Consider a series circuit of consisting of several normally closed pushbutton switch elements in an emergency stop string. Connect an ohmmeter across it, and resistance ought to be very low (typically no more than an ohm or so, depending upon switch characteristics and connecting wire gauge and length). If resistance isn't low it suggests poor wire connection(s) and/or switch element(s) in the measured circuit. Very often, mechanically perturbing each of the switches (that is, lightly banging them with a suitably heavy object) will track down which one of them are in questionable condition.

 

[Windadct]

We did this frequently in Circuit Breaker Maintenance - to identify areas that would heat up too much in service due to I^2*R losses. Even if the initial conditions do not cause overheating, the cycling of the temperature up-and down, due to daily load changes, will accelerate the aging and tend to make the condition worse over time.

 

[jim hardy]

Back about 1973 or '74 high resistance in a contact caused a relay to de-energize and that tripped a nuke plant and that triggered a blackout from Palm Beach to Miami.

It's how the small things of the Earth confound the mighty.

 

[Manoj Sahu]

I performed this test on Siemens 33kV GIS 8DB 10 panel. The test results were around 210micro ohm whereas the ideal values were 173micro ohm. I asked my Seniors , how come they calculated the ideal value of 173 micro ohm. They told me that it depends on number of contacts and can be easily calculated by a formula. Though they didn't tell me the formula.What may be the formula? @anorlunda @Asymptotic

 

[anorlunda]

Back about 1973 or '74 high resistance in a contact caused a relay to de-energize and that tripped a nuke plant and that triggered a blackout from Palm Beach to Miami.

That brings back memories. A similar incident started my whole career. Does the name Sir Adam Beck #2 ring a bell?

I performed this test on Siemens 33kV GIS 8DB 10 panel. The test results were around 210micro ohm whereas the ideal values were 173micro ohm. I asked my Seniors , how come they calculated the ideal value of 173 micro ohm. They told me that it depends on number of contacts and can be easily calculated by a formula. Though they didn't tell me the formula.What may be the formula? @anorlunda @Asymptotic

A search on "Calculate contact resistance" shows several references. Perhaps this one is most relevant. But I can't say for sure what formula those Seimens engineers used.

https://www.electronics-cooling.com/1997/05/calculating-interface-resistance/#

 

[Asymptotic]

I performed this test on Siemens 33kV GIS 8DB 10 panel. The test results were around 210micro ohm whereas the ideal values were 173micro ohm. I asked my Seniors , how come they calculated the ideal value of 173 micro ohm. They told me that it depends on number of contacts and can be easily calculated by a formula. Though they didn't tell me the formula.What may be the formula? @anorlunda @Asymptotic

How many contacts were involved in this measurement?

 

[Manoj Sahu]

How many contacts were involved in this measurement?

I don't exactly remember. Suppose, it is 13. How will we calculate?

 

[Asymptotic]

Can't say for certain how they arrived at 173 μohm, but one way would be to measure a large enough sample of new switch installations to be able to identify and throw out outliers, then average those in the middle of the distribution.

If this were done, then 210 μohm is 37 μohm higher than baseline. The cause could be anything from a single element with abnormally high resistance, all thirteen slightly above normal, one or more high resistances in the interconnecting wiring, or a mix of all these things.

173 seems high for a single switch element, but if there were 13 in series (and keep in mind one must include wire/bus bar, bolted connection, wire lug, and other connection resistances) it works out to about 13 micro ohms each (173/13).

 

[jim hardy]

Does the name Sir Adam Beck #2 ring a bell?

I had to google the name of that plant, but I sure remember that blackout. Even though i wasn't there.
I was a EE sophomore at a small engineering school in central in Missouri, MSM back then.
A classmate from upstate New York thought it was the start of WW3.

old jim

 

[Windadct]

Can't imagine the "customer" calculated the ideal, but they were probably following the OEM's guidelines. A small contact area with a good amount of heat-sink ( they were occasionally used on connection points) would have one value and a large contact area (more copper = more $) may have a much lower value...

The fact that AC MV breakers were approved for nuclear stations gives me the willies... the pictures still do as well...