Verifying Absence of Voltage

[big john]

We're developing a plant standard for racking medium voltage breakers. Our position is to completely de-energize the gear we will be working in, but we've hit a hurdle when it comes to testing for voltage.

As I read NFPA 70E it seems like the only acceptable way to check voltage is to put on a 40 cal. suite, and class 2 gloves and go into the energized 4160 volt cubicle with the appropriate tester on a hot-stick; check the bus to confirm tester operation; then de-energize and re-test the bus.

PPE or not, I would really like to avoid going into any energized MV cubicles for any reason.

Would it be considered an effective test to put on appropriate PPE, check the low-voltage side of the potential transformers (120 volts) prior to de-energizing, then de-energize and re-check the PTs?

OR would it be acceptable to use the switchgear metering to determine the presence of voltage? There are multiple meters that could check the same bus, giving us a level of redundancy.

Are there any standards that lay this out in black and white? I'm looking for answers because the irony here is that it seems like the voltage checks designed to ensure safety are in fact on of the most dangerous parts of this operation.

-John

 

[Besoeker]

We're developing a plant standard for racking medium voltage breakers. Our position is to completely de-energize the gear we will be working in, but we've hit a hurdle when it comes to testing for voltage.

As I read NFPA 70E it seems like the only acceptable way to check voltage is to put on a 40 cal. suite, and class 2 gloves and go into the energized 4160 volt cubicle with the appropriate tester on a hot-stick; check the bus to confirm tester operation; then de-energize and re-test the bus.

PPE or not, I would really like to avoid going into any energized MV cubicles for any reason.

Would it be considered an effective test to put on appropriate PPE, check the low-voltage side of the potential transformers (120 volts) prior to de-energizing, then de-energize and re-check the PTs?

OR would it be acceptable to use the switchgear metering to determine the presence of voltage? There are multiple meters that could check the same bus, giving us a level of redundancy.

Are there any standards that lay this out in black and white? I'm looking for answers because the irony here is that it seems like the voltage checks designed to ensure safety are in fact on of the most dangerous parts of this operation.

-John

Much of what we do is for major industrial customers and defence projects.
Prior to the work commencing, a method statement and risk assessments have to have been submitted and approved before anything is done. All our guys who work on such sites must have a current approved safety passport.
This is mandatory independent training and tests passed with at least a 95% score. Many of the customers also have a required induction course followed by a test.


On the practical issues, when we work on our MV (mostly 3.3kV here) or 11kV kit the usual procedure is:

The breaker will be racked out and closed into circuit earth (ground) position by an authorised MV/HV person. This grounds the outgoing cables releases a mechanical interlock key which locks out the breaker and allows us access to our panels either directly or, more commonly, via a key exchange box when multiple keys are required.
At that point the client, or whoever racked out the breaker will issue a permit to work and our nominated engineer

Next, we use a testing stick in our panels to check that the circuits are dead. This is a three-step procedure. First, the testing stick has checker that injects high voltage into it to prove that the stick is operational.
Then the circuit is checked. And finally, the stick is tested again to ensure that it is still operational. Using live equipment as a proving mechanism is not good practice in my opinion.

An earthing stick is then used to ensure that the feeder cables are discharged then they are grounded at the receiving end.

Additionally, we will have our own padlock on the supply breaker to prevent it from being re-closed.

All of this may sound a bit tedious and convoluted, but compromising safety is not an option.
Producing the documentation for method statements and risk assessments is is a one time major exercise. If you plagiarise something done by others and customise it to suit your specific requirements, that makes it a bit easier.

An unusually long post from me.
Take from it what you want.

 

[zog]

We're developing a plant standard for racking medium voltage breakers. Our position is to completely de-energize the gear we will be working in, but we've hit a hurdle when it comes to testing for voltage.

i am confused, are you de-energizing thr gear to rack out breakers (For PM's on the breaker or something) or are you racking out breakers to de-energized the equipment to be worked on? Your 1st statement can be interpreted 2 ways.

As I read NFPA 70E it seems like the only acceptable way to check voltage is to put on a 40 cal. suite, and class 2 gloves and go into the energized 4160 volt cubicle with the appropriate tester on a hot-stick; check the bus to confirm tester operation; then de-energize and re-test the bus.

No, you can verify your tester works on any know live source in that same voltage range, most tester have an accesory available to check the functionality of the tester. The tester should be verified on a know live source before and after checking your equipment as de-energized.

PPE or not, I would really like to avoid going into any energized MV cubicles for any reason.

Me too

Would it be considered an effective test to put on appropriate PPE, check the low-voltage side of the potential transformers (120 volts) prior to de-energizing, then de-energize and re-check the PTs?

No, a failed PT wil give you a false negative.

OR would it be acceptable to use the switchgear metering to determine the presence of voltage? There are multiple meters that could check the same bus, giving us a level of redundancy.

No, same reason as above

Are there any standards that lay this out in black and white? I'm looking for answers because the irony here is that it seems like the voltage checks designed to ensure safety are in fact on of the most dangerous parts of this operation.

-John

Yes, but I need to know what standards you are required to follow. NESC? OSHA? CSA? A state OSHA plan?

 

[weressl]

We're developing a plant standard for racking medium voltage breakers. Our position is to completely de-energize the gear we will be working in, but we've hit a hurdle when it comes to testing for voltage.

As I read NFPA 70E it seems like the only acceptable way to check voltage is to put on a 40 cal. suite, and class 2 gloves and go into the energized 4160 volt cubicle with the appropriate tester on a hot-stick; check the bus to confirm tester operation; then de-energize and re-test the bus.

PPE or not, I would really like to avoid going into any energized MV cubicles for any reason.

Would it be considered an effective test to put on appropriate PPE, check the low-voltage side of the potential transformers (120 volts) prior to de-energizing, then de-energize and re-check the PTs?

OR would it be acceptable to use the switchgear metering to determine the presence of voltage? There are multiple meters that could check the same bus, giving us a level of redundancy.

Are there any standards that lay this out in black and white? I'm looking for answers because the irony here is that it seems like the voltage checks designed to ensure safety are in fact on of the most dangerous parts of this operation.

-John

A visible break in the conductive path is one option OSHA accepted as a valid verification, eg. the fuse is removed.

IMO the same argument can be made for a withdrawable circuit breaker. Utility guys use hotsticks to attach grounds/shorts to avoid backfeed and the hotstick - most of the time - will keep you far enough from the arc potential.

 

[cornbread]

Sound like you are having problems with the verify, test, verify. I know some of our old timers use a flourenscent fixture to verfiy the tester is working.

 

[Besoeker]

A visible break in the conductive path is one option OSHA accepted as a valid verification, eg. the fuse is removed.

IMO the same argument can be made for a withdrawable circuit breaker.

I'd still want to carry out the tests I outlined in post #2.
Typically here 11kV switchgear for a medium sized plant here is contained in a suite of panels in a segregated and locked section of an electrical switch room. The piece of plant a breaker feeds (e.g. transformer, motor or drive) is remote from that. So, OK you can physically see that a breaker is racked out. The tag number should correspond with the piece of plant it is supposed to feed. But can you be certain that really is the correct breaker?
Could the feed have been swapped to a different supply because of a previous problem or for a myriad of other reasons? If it has, it should have been documented and conspicuous warning signs posted. Should.
I wouldn't want to take that risk myself nor expose anyone else to it.

 

[weressl]

I'd still want to carry out the tests I outlined in post #2.
Typically here 11kV switchgear for a medium sized plant here is contained in a suite of panels in a segregated and locked section of an electrical switch room. The piece of plant a breaker feeds (e.g. transformer, motor or drive) is remote from that. So, OK you can physically see that a breaker is racked out. The tag number should correspond with the piece of plant it is supposed to feed. But can you be certain that really is the correct breaker?
Could the feed have been swapped to a different supply because of a previous problem or for a myriad of other reasons? If it has, it should have been documented and conspicuous warning signs posted. Should.
I wouldn't want to take that risk myself nor expose anyone else to it.

Where do you suggest the test to be performed? The other end of the cable is terminated on a motor and there are no live parts accessible. In the breaker compartment there are two sets of stationary contacts and the one that is (still) live is covered with the automatic shutter. The other can only have power on it if it feeds a distribution CB that can be back-feed. In that case test should be performed before the grounding jumpers are installed on all three stabs for the duration of the work. In the US the labeling and circuit identification is mandated by OSHA. Testing to identify a connection is not a foolproof method of identifying the 'right' cable as the circuit could be off. If there is no way to visually trace an unknown cable, or service work like you describe where you are working on an unknown system, after de-energization I would want to send a test pulse to verify that I am dealing with the correct 'other end' of the circuit. If it is a taped up shielded cable splice this would be a much tougher problem.

 

[zog]

If it isnt grounded, it isnt dead.

 

[big john]

Thanks for the responses so far.

To clarify, the policy is to only rack breakers once the switchgear is de-energized. Unfortunately, there are no air-breaks that we could check because the switchgear is fed through OCBs. Withdrawling breakers could isolate parts of the 4160 system, but that presents us with the same problem: We have to rack breakers into and out of hot bus.

We have a stick-mounted tester (I think it actually is called a "Tic Tracer") but there is no self-testing option. It will detect low-voltage and in the past has always been checked against 480 prior to checking 4160. The literature says that's an acceptable means of test, but personally speaking, I would really like some certainty that the tester I'm using will in fact detect the voltage I'm working on.

Besoeker, would you happen to know who makes the "checker" that you use to verify your hot sticks?

Zog, I hear what you're saying regarding failure of a PT or panel-meter. I don't like the panel meter idea simply because they only read in KV so it might be possible to have a couple hundred volts present without it being obvious.

However, my thinking regarding testing the PTs: How are they much different than a tester itself. Yes, PTs can fail, but so can our Tic-Tracer. We check each PT before shutdown, and then we test each PT after shutdown, that seems like a pretty good indicator (in conjunction with other steps). However, I do realise that this method does not allow an effective Check-Test-Check as would normally be done, and that is probably the main hangup.

Maybe the thing to do is get a twelve-foot hot-stick and do voltage checks from a distance.

-John

 

[zog]

To clarify, the policy is to only rack breakers once the switchgear is de-energized. Unfortunately, there are no air-breaks that we could check because the switchgear is fed through OCBs. Withdrawling breakers could isolate parts of the 4160 system, but that presents us with the same problem: We have to rack breakers into and out of hot bus.

Have you looked at remote racking? This is the direction many plants are going towards. See atachment.

We have a stick-mounted tester (I think it actually is called a "Tic Tracer") but there is no self-testing option. It will detect low-voltage and in the past has always been checked against 480 prior to checking 4160. The literature says that's an acceptable means of test, but personally speaking, I would really like some certainty that the tester I'm using will in fact detect the voltage I'm working on.

Look for a device called a pizotest, small hand held device that generates 5kV using pizoelectric crystals, designed specifically for this application.

Zog, I hear what you're saying regarding failure of a PT or panel-meter. I don't like the panel meter idea simply because they only read in KV so it might be possible to have a couple hundred volts present without it being obvious.

However, my thinking regarding testing the PTs: How are they much different than a tester itself. Yes, PTs can fail, but so can our Tic-Tracer.

Not if you do a live-dead-live test, as required by the 70E and OSHA.

 

[Fulthrotl]

We're developing a plant standard for racking medium voltage breakers. Our position is to completely de-energize the gear we will be working in, but we've hit a hurdle when it comes to testing for voltage.

As I read NFPA 70E it seems like the only acceptable way to check voltage is to put on a 40 cal. suite, and class 2 gloves and go into the energized 4160 volt cubicle with the appropriate tester on a hot-stick; check the bus to confirm tester operation; then de-energize and re-test the bus.


Are there any standards that lay this out in black and white? I'm looking for answers because the irony here is that it seems like the voltage checks designed to ensure safety are in fact on of the most dangerous parts of this operation.

-John

it could be worse... you could be working outdoors, and have to do hot,
dead, hot when the next nearest hot to check off of is 3 miles away...

this is a thought, but there are remote racking devices that will rack out
breakers without you being in the room.... might that be a solution that
eliminates the need to test in the first place? that doesn't eliminate the
risk to equipment caused by an arc flash, but it does eliminate the hot
stick dance.

 

[Besoeker]

Where do you suggest the test to be performed? The other end of the cable is terminated on a motor and there are no live parts accessible.

I'd test it at the piece of plant to be worked on. If, for example, we were going to replace a motor, I would test at the terminals before disconnecting any cabling. That is, I'd test to ensure that they weren't live and I'd use a discharge stick to ensure that there was no remaining charge.

In the breaker compartment there are two sets of stationary contacts and the one that is (still) live is covered with the automatic shutter. The other can only have power on it if it feeds a distribution CB that can be back-feed. In that case test should be performed before the grounding jumpers are installed on all three stabs for the duration of the work.

Most or perhaps all of the (11kV) breakers I have had dealings with have a circuit earth position which grounds the outgoing cable. It will remain padlocked locked in this position for the duration of the work. The person responsible for carrying out the work will have will retain that key untlil has completed that work. If there is more than one discipline involved for different parts of the work, each will have a padlock on the breaker. Obviously the breaker can't be re-closed until everyone involved has signed off their permit to work and handed back their key.

In the US the labeling and circuit identification is mandated by OSHA.

Labelling is mandatory here too. It doesn't mean that it is always correct though, TBH, I haven't ever seen it done incorrectly on either 3.3kV or 11kV. Can't say the same for LV though.

Testing to identify a connection is not a foolproof method of identifying the 'right' cable as the circuit could be off.

I quite agree and I wasn't advocating it as a foolproof method. Just an additional test. Peace of mind as much as anything else.

 

[weressl]

I'd test it at the piece of plant to be worked on. If, for example, we were going to replace a motor, I would test at the terminals before disconnecting any cabling. That is, I'd test to ensure that they weren't live and I'd use a discharge stick to ensure that there was no remaining charge.

MV terminals connecting to a single motor are often insulated splices leaving no live parts open to test.

Most or perhaps all of the (11kV) breakers I have had dealings with have a circuit earth position which grounds the outgoing cable. It will remain padlocked locked in this position for the duration of the work. The person responsible for carrying out the work will have will retain that key untlil has completed that work. If there is more than one discipline involved for different parts of the work, each will have a padlock on the breaker. Obviously the breaker can't be re-closed until everyone involved has signed off their permit to work and handed back their key.

We do not have automatic grounding provision on our equipment. (There are some on motor starters, but not standard MV circuit breakers.) The fallacy here is that you're installing the ground on the OUTGOING cable. If you have the wrong breaker open, then you still no proof that the breaker will be reclosed while people at work on the working end. SO I would attach a grounding set on the working end of the DISCONNECTED cable. If the cable end needs to be worked on I would call for 'ringing' out the cable to assure that it is the same piece of copper or aluminum.

Labelling is mandatory here too. It doesn't mean that it is always correct though, TBH, I haven't ever seen it done incorrectly on either 3.3kV or 11kV. Can't say the same for LV though.

Your argument previously raised that as the prime or one of the main points. Now you making it a strawmen.

I quite agree and I wasn't advocating it as a foolproof method. Just an additional test. Peace of mind as much as anything else.

The point is that the peace of mind could be a very dangerous thing and elads to complacency. The more complex the task is the more people tend to believe that it has been throughly thought out leaving no room for error. When preparing for a complex electrical task, especially troubleshooting, the mind already is ahead, devising potential plans for the work. I would venture to say that they could easily overlook the potential nuance differences that could be present in the circuit identification.

..and this is my piece of mind.

 

[zog]

A ground and test device for your breakers would be a problem solver here.

 

[Besoeker]

MV terminals connecting to a single motor are often insulated splices leaving no live parts open to test.

Yes, fair point in such cases.

We do not have automatic grounding provision on our equipment. (There are some on motor starters, but not standard MV circuit breakers.) The fallacy here is that you're installing the ground on the OUTGOING cable. If you have the wrong breaker open, then you still no proof that the breaker will be reclosed while people at work on the working end. SO I would attach a grounding set on the working end of the DISCONNECTED cable.

We do that also as a matter of course. See post #2 where I specifically mentioned it. And you'd probably want to know that the conductors are dead before you attached the grounding set.

If the cable end needs to be worked on I would call for 'ringing' out the cable to assure that it is the same piece of copper or aluminum.

That's fair enough but wouldn't you want to check that it is dead before you try that?

The point is that the peace of mind could be a very dangerous thing and elads to complacency. The more complex the task is the more people tend to believe that it has been throughly thought out leaving no room for error. When preparing for a complex electrical task, especially troubleshooting, the mind already is ahead, devising potential plans for the work. I would venture to say that they could easily overlook the potential nuance differences that could be present in the circuit identification.

I'm sure you realise as well as anyone else that complacency has no place in this sort of work environment - or any other for that matter.
Method statements and risk assessments are prepared and agreed before any work is undertaken to avoid such a mindset.
Accidents still happen. Few human endeavours are entirely risk free.
But we will take all measures available to mitigate that risk.