main breaker tripping

[malachi constant]

We have a client that had a new building (911 center) built two years ago. I know nothing about the original project beyond that.
In the last year their main breaker has tripped twice (including once a couple weeks ago, in the middle of the night). They have asked their neighbors if they have experienced any issues, they said not really - they installed some surge arrestors to take care of some power quality issues, but nothing that tripped a breaker. The utility does not see anything unusual on their end. The EC thinks it may be an issue with the breaker and recommends removing and testing the breaker.

In the absence of any more information I tend to assume the people closest to the situation know it best - so I would tend to say "whatever the EC says". It is a five-hour drive from my office and I don't believe our firm did the original engineering on this. That said, what questions could I toss out there? I've got:
1. Have you consulted with the engineer of record?
2. Does the facility have ground fault protection? Is the main tripping, or is it the GFP?
3. What kind of power quality issues was the neighbor experiencing?
4. What kind of loads within the building could run at night? Was the facility occupied? (I can't imagine what might cause it.)
5. Ask for copy of riser, motor list, original service calcs, and OCPD study.
6. Did the first trip also occur at night?

Anything else?

 

[USMC1302]

You already seem to have the bases covered, but ask if any other building equipment has been having issues? My example is a chiller that had a bad compressor. Didn't always show up depending on load...That has happened to me at several different locations in the last couple years.

 

[hillbilly1]

Where you are at, it could be a bad second stage heat heat strip if they have electric heat, but it looks like you got most of the primary questions covered.

 

[Jraef]

The first thing to assume is that the breaker is doing its job, not that it's defective. Until you know WHAT the breaker tripped on, you do not know anything useful in this situation. "Power quality" is kind of a catch-all diagnosis that means nothing can be found by looking at the surface, but power quality / surges / spikes have next to nothing to do with main breakers tripping. A main breaker trips on current only, unless there is a Ground Fault trip included, in which case it is still current, but it is current to ground rather than phase to phase or phase to neutral. Voltage has no bearing on making a main breaker trip, unless it drops so much that the current increases, and in that case the lights would likely dim and it would be obvious.

Much more likely issues are improper coordination of trip settings with branch circuits, large loads causing a voltage drop when coming on (such as the chiller suggestion earlier), or intermittent ground faults.

If your main CB is electronic, it likely has a record of what caused it to trip. If not, you may need to put in a recording power meter too look at it over time.

 

[ATSman]

Ghost Busting

Ghost Busting

Try to get a single-line diagram from the EC or EEOR.
If it a 480/277V system 1000A or greater then it is required to have GFP on the main breaker only which
makes it susceptible to any intermittent grounds that may occur downstream. As one post mentioned:
a chiller or HVAC motor with a winding problem. If the bldg has 277V lighting then a faulty ballast could
be going through the 20A subpanel breaker back to the main GFP causing the main to trip. We find this
a very common problem when there is no secondary level of GFP on the feeders; everything is seen my
the main GFP.
Then again, the breaker trip unit or GFP relay could be faulty. Having them high current tested is expensive
and may or may not show the problem. Putting a line disturbance analyzer to monitor voltage and current
on the load side of the main will give you more information as to whether it is plant or utility related but is not
cheap, either.
At this point, I would concentrate on finding out the following:
1. Is the main a circuit breaker or bolted pressure switch (Pringle, Boltlock, etc) ?
2. If a breaker then make, model#, trip unit model#, pickup settings?
3. Does the breaker trip unit have integral GF protection? IF the GFP is external then make, model#, current pickup setting?
With all this said, are there any trip indicators on the breaker trip unit or external GFP test panel that show after the trip
occurs. This info is important to steer you in the direction of the fault.
If the cause is nuisance tripping of a faulty component then replacement may be the only solution.

 

[K8MHZ]

We have a client that had a new building (911 center) built two years ago. I know nothing about the original project beyond that.
In the last year their main breaker has tripped twice (including once a couple weeks ago, in the middle of the night). They have asked their neighbors if they have experienced any issues, they said not really - they installed some surge arrestors to take care of some power quality issues, but nothing that tripped a breaker. The utility does not see anything unusual on their end. The EC thinks it may be an issue with the breaker and recommends removing and testing the breaker.

In the absence of any more information I tend to assume the people closest to the situation know it best - so I would tend to say "whatever the EC says". It is a five-hour drive from my office and I don't believe our firm did the original engineering on this. That said, what questions could I toss out there? I've got:
1. Have you consulted with the engineer of record?
2. Does the facility have ground fault protection? Is the main tripping, or is it the GFP?
3. What kind of power quality issues was the neighbor experiencing?
4. What kind of loads within the building could run at night? Was the facility occupied? (I can't imagine what might cause it.)
5. Ask for copy of riser, motor list, original service calcs, and OCPD study.
6. Did the first trip also occur at night?

Anything else?

Does the center have a back up generator that self tests every month, maybe at night?

 

[malachi constant]

Try to get a single-line diagram from the EC or EEOR.
At this point, I would concentrate on finding out the following:
1. Is the main a circuit breaker or bolted pressure switch (Pringle, Boltlock, etc) ?
2. If a breaker then make, model#, trip unit model#, pickup settings?
3. Does the breaker trip unit have integral GF protection? IF the GFP is external then make, model#, current pickup setting?
With all this said, are there any trip indicators on the breaker trip unit or external GFP test panel that show after the trip
occurs. This info is important to steer you in the direction of the fault.
If the cause is nuisance tripping of a faulty component then replacement may be the only solution.

So, I guess we did the engineering on this one in house. I'm the fifth engineer to touch this thing, none of the other four are still here, I've never worked with any of them either. I did round up the riser, motor list, OCPD study, and shop drawings. Here's what I've got:
1. Main is 480V 1600A circuit breaker, GE SS Powerbreak II. I cannot tell from the shops and OCPD study whether this is Power+, MicroVersaTrip-Plus, or MVT-PM. I am not well-versed in electronic circuit breakers -do all three have monitoring capabilities, or just the MVTs? If there is an event record I assume the contractor should be able to extract the pertinent information and send it to me?
2. The original* GE SS settings were: 1600A frame/sensor/plug, 0.9 LTPU, 3 LTD, 9 STPU, Int STD, Out I2T, 15 INST. (*This is from the study done in Feb '11. Two main-GFP-trip events occurred in Nov '11, resulting in a different engineer recommending all settings -- for main and GFP -- be cranked to the max. I have no idea if this was implemented or what the settings are now, but will ask.)
3. I cannot tell for certain from the shops if the GFP is integral to the main or not, as I am pretty ignorant of circuit breakers and the shops aren't spelling it out for me. The OCPD study lists the device as GE GF 200-2000AAS, SS frame; and the settings as: 1600 frame/sensor, GFPU (0.2-0/6 x S) 0.6 (960A), GFD (Min-Max) Max 1^2T Out.

Also:
4. Motor list does not include any chillers. No electric heat except for two 480V 50kVA elec water heaters. Three 20HP AHUs. Six heat pumps of unknown size - maybe 50kva each? 120V exhaust fans. Two data room CRUs, 30kVA ea. Gas CUHs and UHs. Six pumps, all either 15HP or 7.5HP.
5. I don't know that the recent event was a GFP trip, but am guessing it was. I will ask for more information.
6. Facility does have backup paralleled generators. I will ask when the monthly testing of those is scheduled. They wouldn't set up the tests for 2am would they?

Thanks for your help!!

 

[hillbilly1]

I can't remember if the GE uses a separate CT for the neutral, or whether it is built in, or it is the large window type encompassing all phases and neutral, Maybe Zog could chime in on that, but I have come across neutrals terminated on the wrong side of the CT causing tripping because the CT does not "see" the current and assumes there is a ground fault, with the large window CT's, I have seen phase conductors routed around the window, instead of thru it because they cut the conductors too short. Just a few other things to check, especially if they have already turned the settings all the way up, and it's still tripping.

 

[GoldDigger]

I can't remember if the GE uses a separate CT for the neutral, or whether it is built in, or it is the large window type encompassing all phases and neutral, Maybe Zog could chime in on that, but I have come across neutrals terminated on the wrong side of the CT causing tripping because the CT does not "see" the current and assumes there is a ground fault, with the large window CT's, I have seen phase conductors routed around the window, instead of thru it because they cut the conductors too short. Just a few other things to check, especially if they have already turned the settings all the way up, and it's still tripping.

If a run test of the generator is taking place (and 2am for an automated generator exercise while grid power is available is actually common), incorrect configuration of the generator(s) and transfer switch(es) between SDS and non-SDS could easily lead to a redundant ground-neutral bond which could then cause the GFD to trip.

 

[hillbilly1]

If a run test of the generator is taking place (and 2am for an automated generator exercise while grid power is available is actually common), incorrect configuration of the generator(s) and transfer switch(es) between SDS and non-SDS could easily lead to a redundant ground-neutral bond which could then cause the GFD to trip.

True, but I would figure they would have greater frequency of tripping, because most commercial generators are exercised weekly.

 

[GoldDigger]

True, but I would figure they would have greater frequency of tripping, because most commercial generators are exercised weekly.

I was thinking that with the breaker setting cranked to max (do you know when that was done?), whether or not the GF trip was triggered might also depend on the particular loads that were running or starting while on the generator.
The percentage of neutral current that ended up on the EGC could be fairly small depending on the exact wiring geometry. And the result would also depend on what fraction of the load current the neutral was carrying (i.e how close to a balanced condition the loads were running at.)
In the case of an extra ground-neutral bond rather than a phase to ground fault, if all the current were in the phase conductors, there would not be a trip.

 

[hillbilly1]

I was thinking that with the breaker setting cranked to max (do you know when that was done?), whether or not the GF trip was triggered might also depend on the particular loads that were running or starting while on the generator.
The percentage of neutral current that ended up on the EGC could be fairly small depending on the exact wiring geometry. And the result would also depend on what fraction of the load current the neutral was carrying (i.e how close to a balanced condition the loads were running at.)
In the case of an extra ground-neutral bond rather than a phase to ground fault, if all the current were in the phase conductors, there would not be a trip.

The OP wasn't sure if they actually did set them to max, but if they did, I would think they would have a very unbalanced load to have enough neutral current to trip. I think minimum setting is usually 100 amps, which could be possible, but if they have it at max, it would be much more. Lots of possibilities.

 

[rcwilson]

A reversed neutral CT or CT connection will make the neutral current appear to be ground fault current. The backwards connection subtracts the neutral current instead of adding it.

Ideally A+B+C+N=0 except during a ground fault. If the neutral CT is wired backwards, A+B+C-N = -2N. 50 amps unbalance on the neutral now looks like a 100 amp ground fault to the relay. A well balanced load never has a problem until someone shuts off all the lights on A phase or the light dimming system puts out large third harmonics on the neutral.

Cranking up the GF setting will alleviate the tripping problem until the next time an unbalanced load occurs.

My experience in testing and inspecting GF systems is that 30% were installed incorrectly. Reversed or bypassed neutral CT was a common problem.

I would look at the GF first, or have someone who knows what to look for inspect the gear.

 

[hillbilly1]

A reversed neutral CT or CT connection will make the neutral current appear to be ground fault current. The backwards connection subtracts the neutral current instead of adding it.

Ideally A+B+C+N=0 except during a ground fault. If the neutral CT is wired backwards, A+B+C-N = -2N. 50 amps unbalance on the neutral now looks like a 100 amp ground fault to the relay. A well balanced load never has a problem until someone shuts off all the lights on A phase or the light dimming system puts out large third harmonics on the neutral.

Cranking up the GF setting will alleviate the tripping problem until the next time an unbalanced load occurs.

My experience in testing and inspecting GF systems is that 30% were installed incorrectly. Reversed or bypassed neutral CT was a common problem.

I would look at the GF first, or have someone who knows what to look for inspect the gear.

I had that problem with a gear once, the prints called for the main to be on the left hand section, but the owner ordered a right hand main. They swapped the sections around, and put the neutral CT on backwards. I was called in when they were having issues with it tripping, and that's what I found.

 

[ATSman]

These are all good scenarios being discussed here. The important point to note here is that
this is a fairly new installation (1-2 yrs?) and mis-wiring or wrong CT configuration is a good possibility. As the building load increases over time (especially with single phase loads), how the neutral current returns to the XO of the supply transformer becomes critical if (as Hillbilly
mentioned) the neutral to ground bond is down stream of the large window CT (zero sequence config. and GFP is external) or the neutral bus in-line CT (residual config. and GFP is integral to breaker.) The system needs to be shut down, the neutral bus needs to be opened up in the main breaker section and a resistance check to ground should be made using a megger or ohmmeter. The neutral bus should only be grounded upstream of the GF window CT or in-line or small CT around the neutral bus or nuisance tripping can occur because the load currents don't cancel.

 

[techniphase]

had a problem similar to this, panel was original w building in 50's...opened the meter and pulled the covers off

burned bus bar at bus connections loosened over years..

replaced and never tripped again..

(200 amp 3 phase)

 

[techniphase]

oops, sorry just noticed ...new install

anywho....

 

[malachi constant]

Incredible information you guys - thank you all. Here's where I'm at:

* Power quality issues probably not causing the trip. However if there are significant power quality issues one engineer suggested the surge protection device may be contributing to a ground fault. He surmised the MOVs could be sucking all sorts of power to ground. I have not heard of this before but will have the SPD checked, see if it has an event log or an inordinate number of hits.
* We do not have any really large loads. No chillers. A half-dozen ground source pumps is the only thing that jumps out at me.
* I think I can rule out improper coordination with branch circuits, assuming the GFP was cranked up to the max as directed in Nov '11. Max appears to be 960A for this GFP unit.
* Just curious, does ^^^ that ^^^ (960A trip setting) rule out any or all intermittent ground faults? Such as a faulty ballast? Faulty motor? Wouldn't the branch/feeder OCPD catch individual faults before they reached 960A? What could cause a fault of that magnitude?
* I will definitely have someone check the CT wiring. The improperly wired neutral is promising. I won't get my hopes up though.
* There are two parallel generators. I don't know whether the transfer switches are three or four pole but will ask for the shops.
* Related, nor is there a grounding diagram drawn. I spoke to the engineer of record, he apologized to me for never doing a grounding one-line. (He didn't do them at the time but makes regular practice of it now.) He thinks this is a promising place to start though likes the "improperly wired CT" idea too. I'll draw one up, it will be good practice for me.
* I assume it would be pretty easy for a contractor to check grounds are bonded in the proper places, and check a few key places where they might be bonded but shouldn't. I could see improper generator / transfer switch bonding easily being a cause. But what about random mistakes, such as a ground and neutral tied together in a box or panel? Could these be causing it? How would the contractor go about checking?
* Another engineer suggested there could be water in some underground feeders, and maybe megging them would be a good idea. Any thoughts on this?
* As mentioned, the GFP could be faulty, but that is expensive to test. The engineer of record says he considers GE to be on the lower end of the quality scale, and would not be that surprised if this was the issue.
* I will ask for any record of events that they have on the main.

So do these probabilities look about right?
* Faulty GFP - 20%
* Improper GFP/CT wiring - 20%
* Improper transfer switch / generator bonding - 20%
* Something else wrong with generator / transfer switches - 10%
* Improper grounding elsewhere in building - 5%
* Bad utility power / something outside the facility - 5%
* Intermittent ground fault - 5%
* Water in conduits - 5%
* Other (what am I missing?) - 5%

I'll be calling the client this afternoon with an update and some questions. Still haven't talked with the contractor or equipment suppliers. Plenty of research left to do.

Thanks!!

 

[malachi constant]

UPDATE!

To recap, with some new information:
1. Existing 911 call center, occupied about a year and a half.
2. 1200A 480V 3P 4W service.
3. Ground fault protection at ONE location - the main.
4. GFP is integral to the main breaker. GE SS w/ Entelligaurd Trip Unit.
5. A coordination study was performed in Feb 2011 and is believed to have been implemented around that time.
Also (already did all my numbering, forgot these, whoops) --
5A. there are no large motor loads on the building - a handful of ground source heat pumps. Whatever lists of "problem equipment" I found on this site and other places didn't check out with the equipment installed.
5B. Typical building loads are less than 200A/phase. Where a 1000A ground trip comes from is baffling to me. Any links to an article or paper on this are appreciated.

6. Main breaker tripped on a ground fault (we believe twice) in November 2011. The engineer who inherited the project said "crank it to the max". It was then cranked to the max - which is, I believe, 1000A.
7. All was quiet for 14 months until Feb 2013 there was another tripping event. The Owner cleared it without calling the electrician. It is presumed to be another ground fault event but not confirmed. (And may not be able to confirm it now - still working on chasing this one down.)

8. Main breaker was tested (current injected) by contractor this week. It passed.
9. I loved the idea of the CT being backwards - the neutral current inadvertently causing a trip. As the CTs are integral to the breaker and the breaker passed inspection I am not sure if this is a dead end or not. (This is on my list to chase down too.) Is the neutral CT integral to the main breaker, or something installed separately? I had asked for a lot of photographs related to the inside of the switchboard as it pertains to the breaker/GFP/wiring but am afraid they did not get any for me. Feedback?
10. The overcurrent settings were for the most part in line with the "original study except GFP cranked to the max". There were a few minor omissions and discrepancies but I don't think anything worth redoing the study over. Especially as there are no downstream GFPs to coordinate with.
11. We are considering asking the Owner if they are willing to wait and see if it happens again, now that they have the engineers full attention (I was not a part of the arch/eng company that designed this until recently). We would scour the GE PQM II data monthly for any oddities, and if they had another event we would ask that they confirm it was a Ground Fault before clearing it. At that time we would go all out in trying to chase something down - infrared scanning, check earth ground for stray voltages, metering on other panels/equipment, etc. Is taking a wait-and-see approach a bad idea? I do not want to give them advice to sit on their hands if they have a potential life safety issue. I also don't want to leave them no option but to spend money chasing ghosts, esp after having them spend a couple grand testing the breaker this week. Another option would be to lower the GFP to 300A or something, and see if it happens again (the building has a very robust UPS/parallel generators backup system, the building occupants do not notice the power going out). What are the risk factors with "passively" pursuing a resolution? Any thoughts?
12. Is infrared scanning likely to turn up a clue? It was required at the 11-mo warranty walk-through but did not happen. I don't want to hold the contractor's feet to the fire if this is not useful information. At the same time, if it's useful, they would probably do it. (They are good guys, btw - intelligent too - it is sure nice working with a quality contractor on an issue like this.)
13. I found out this week the generators also have GFP on them. I will make sure this is alarm only, not actually something that will cause a trip. I would hate to lose both normal and emergency power while we are working through this.
14. The grounding one-line diagram is a little goofy but the engineers who have looked at who know grounding better than I say nothing that would cause any issues. I'm told by people smarter than me that the issue is likely a combination of an inadvertent neutral-ground bond somewhere, or some equipment failing, or some combination of the two. But I'm not sure how 1000A shows up on the ground without something downstream tripping - even though there are no other GFPs you would think something somewhere would go, wouldn't you? If someone can help explain this I would appreciate it.

That's it. Got to head out the door. Meeting the Owner tomorrow afternoon, so if you have any thoughts between now and then please let me know. Wish I had more time to consolidate my thoughts and/or be a little more thorough, but...well, no time for that this afternoon.

Thanks and have a good Easter weekend!

 

[GoldDigger]

UPDATE!

To recap, with some new information:

9. I loved the idea of the CT being backwards - the neutral current inadvertently causing a trip. As the CTs are integral to the breaker and the breaker passed inspection I am not sure if this is a dead end or not. (This is on my list to chase down too.) Is the neutral CT integral to the main breaker, or something installed separately? I had asked for a lot of photographs related to the inside of the switchboard as it pertains to the breaker/GFP/wiring but am afraid they did not get any for me. Feedback?

Is the neutral as well as the phase wires wired through the breaker (as it does on a 120 volt GFCI in a panel)? Not interrupted or switched, but does the neutral current pass physically from one terminal to another in the breaker? If not then there would have to be terminals on the breaker for the attachment of an external CT.
If only the phase wires pass through the breaker and there is no external CT connection, then the GF detection in the breaker will count all neutral current as a fault.