So, we all know that an arc flash study, and all the PPE requirements and safe work practices that follow are based on a very big assumption, clearing times.

These clearing times are difficult to verify, if you test your breakers you have to open them, rack them out, then test them, so by the time you record the data the breaker has been cycled a few times, that dried grease has been broken loose, those tired springs got a little workout, etc...

So what about first trip testing? Anyone do it? I have, and the results are often scary. Let's say you are an engineer at a facility, you did an arc flash study, you based your labels on a 60mS clearing time but find out from testing first trip it is really 300mS?? (Those are pretty realistic numbers by the way).

Now what?

Talking about the elephant in the room! Great point to bring up Zog!

70E just says maintain equipment in accordance with manufacturer's instructions or industry consensus standards. (NETA comes to mind) I know when asked, I have had people tell me their equipment has been tested....

.... in 1978 when it was installed.

Testing is not the same as maintained, and I think a lot of places will overlook that one. Proper maintenence means checking adjustments, lubrication (Not spray lubes, that makes things worse), and refirbsment per OEM recomendations. First trip timing tells the real life condition of your breakers, at least the part that is meaningful for arc flash studies.

Wouldn't it be great if there was an easy, inexpensive way to gather first trip data on something like your iPhone????

You are certainly correct about testing vs. maintenance. The iPhone app sounds like a great idea.

Is this where I ask "do you have such a method?"

I can answer that at NETA in New Orleans this spring

Luminant

does it. I just caught one at >1 sec due to someone worrying about current limits on relay contacts and inserting a cheap ice cube relay that had a minimum closing voltage of 85 volts and was not wired to the DC side. We used to use mostly ABB relays but with 10 + week lead times on stock relays like a CO-8 it was time to change. If you take the time to set them up correctly, SEL relays cost less because you only need one relay and the diagnostic stuff is free. We standardized on three models for everything and two of them only differ by an input card.

So what did you do with that >1 sec breaker Paul?

P.S. I have hendreds of CO-8's in stock.

Game Changer! Looking froward to hearing more. Please let us know more when you can.

First thing is that this is a typical mining substation. It has a shunt trip breaker operated by a capacitive trip unit (DC power). The original setup operated this from normally open contacts of an ice cube relay which was in turn operated by normally open contacts of the SEL relay. A separate UV coil (normally closed) energized again via capacitive trip to open the breaker on loss of DC. New scheme is to wire normally closed contacts of SEL relay to the shunt trip directly. An alternative would be to wire another relay to the contacts using normally closed logic.

Not being overly familar with your tripping scheme, why don't you try using an Electro Switch LOR, ER/LOR, or SR/LOR. These aux relays have less then an 8mS trip time and have positive tripping action. We use them everywhere in the utility system. They can have multiple inputs and multiple outputs.

I'm familiar with those. Not sure if anyone ever uses anything else for handles anywhere because that's all I ever see.

The relay was added for isolation (increased current handling) by the gear manufacturer and not for control or isolation. It was totally unnecessary in the first place. When discussing it with the manufacturer their idea was to add ANOTHER relay in parallel using reversed logic (NC) and not simply read specs and get rid of the issue. They work with a lot of different control relays so the knee jerk reaction is that when they use a digital relay they automatically add interposing relays. Same extremely POOR design practice is rife in industrial controls. I can't count the number of PLC IO panels where builders have stuck rows of cheap ice cube relays on every output. Each one just sits there quietly creating transients, failures, downtime, and costs far more than the relay.

I'm not suggesting to go to 100% triac designs and don't use any relays ever. That's not realistic. Use them where they are needed and do not look at the price as the determining factor because they are 1-2 orders of magnitude less reliable than everything else except valves. Learn from what General Motors learned in the 1970's when they went away from 12x8 enclosures full of highly reliable machine tool relays for this exact reason.

How does delayed or extended opening time play into AIC ratings? My understanding is that it requires more bracing to withstand a fault of longer duration, it's not simply based on magnitude. If that is the case, this disaster would exceed the arc flash problem.

This consideration may cause folks to prefer 'simpler' molded case breakers in large sizes for those applications where they can be used.

AIC is a function of the magnetic force during a bolted fault, which is in turn proportional to the square of the peak current. As typically transients developed by inductive sources decrease over time, longer opening times should not impact AIC.