Building grounding
Building grounding
(OP)
As an electrical engineer, system grounding is not an area I have too much experience with. We had a situation with the user recently, where we had power outages over a weekend for a medical facility, and we have to replace some cables and breakers, and when we turned on the system to test the emergency power, some of the medical equipments were fried. Basically we are looking into what went wrong with the existing system, and see if there is any problem that can be fixed.
The user mentioned that the facility, which was built some 30 years ago, did not have good grounding. He is stating based on his experience as fixing up the place. He said sometimes he noticed that are 2~3 branch circuit sharing only one ground. And nowadays, we always specify each circuit to have individual grounding.
After some painful investigation with the contractor, I have good idea when the surge occurred (although not sure if this is the right term, because it could also be ground fault too). After the power outage, the electrician gradually turned on the building power from substation down to individual panels, and load connection. So I do not think that is where problem is.
The problem occurred when they simulate the emergency "load" test after starting up building power, where building load was "on" and they simulate the "power-down" scenario by:
1. Disconnecting the substation 600A vacuum breaker.
2. Generator detected power down and started.
3. ATS waited for about 3~8 second, then switched from commercial to standby power mode.
4. Then when they are ready to return to normal power, they turned on the substation breaker.
5. Generator detected normal power, and stopped.
6. ATS waited for about 3 second, then switched from Standby power to commercial power
7. At this point, the TVSS sounded the alarm, also some other alarm (fire sprinkler panel) was also set off.
8. Building back to normal power (TVSS still in alarm mode).
Since TVSS alarm occurred in Step 7, I believe that is where the main surge occurred.
According to what user told me, if indeed the facility does not have good grounding, which they said is buried in the crawlspace somewhere since they built this place, but no one knows, so when the power was turned on, and the ATS flipped over from standby to commercial, and since the building is on "some" load already, the surge current may not have anywhere to go. It travels back to the system somewhere. Does that make sense? This is just a theory. I would to like to go and check, but do not know where to start measuring to verify.
Another thing I observe is that there is no clear building ground grid. According to the current set of drawings I have, the substation ground is the only ground I can see. But the substation is some distance away, wouldn’t that cause voltage drop in the ground over distance. That is why we may not have equipotential at all grounding in the building. I could now see the value of a clearly defined building ground system where ground measurement will be the same throughout the facility. But again, this needs to be verified.
Another thing I observed is that there is no surge arrestors at the substation, which has 3 transformers, 2 of them are delta-delta, and the last one which supply this MEDDAC facility is delta-wye, but with nothing. I recommend one to be put in. In Japan, they often use delta-delta system in the past, and their ground system are floating, which is dangerous from NEC perspective.
The user mentioned that the facility, which was built some 30 years ago, did not have good grounding. He is stating based on his experience as fixing up the place. He said sometimes he noticed that are 2~3 branch circuit sharing only one ground. And nowadays, we always specify each circuit to have individual grounding.
After some painful investigation with the contractor, I have good idea when the surge occurred (although not sure if this is the right term, because it could also be ground fault too). After the power outage, the electrician gradually turned on the building power from substation down to individual panels, and load connection. So I do not think that is where problem is.
The problem occurred when they simulate the emergency "load" test after starting up building power, where building load was "on" and they simulate the "power-down" scenario by:
1. Disconnecting the substation 600A vacuum breaker.
2. Generator detected power down and started.
3. ATS waited for about 3~8 second, then switched from commercial to standby power mode.
4. Then when they are ready to return to normal power, they turned on the substation breaker.
5. Generator detected normal power, and stopped.
6. ATS waited for about 3 second, then switched from Standby power to commercial power
7. At this point, the TVSS sounded the alarm, also some other alarm (fire sprinkler panel) was also set off.
8. Building back to normal power (TVSS still in alarm mode).
Since TVSS alarm occurred in Step 7, I believe that is where the main surge occurred.
According to what user told me, if indeed the facility does not have good grounding, which they said is buried in the crawlspace somewhere since they built this place, but no one knows, so when the power was turned on, and the ATS flipped over from standby to commercial, and since the building is on "some" load already, the surge current may not have anywhere to go. It travels back to the system somewhere. Does that make sense? This is just a theory. I would to like to go and check, but do not know where to start measuring to verify.
Another thing I observe is that there is no clear building ground grid. According to the current set of drawings I have, the substation ground is the only ground I can see. But the substation is some distance away, wouldn’t that cause voltage drop in the ground over distance. That is why we may not have equipotential at all grounding in the building. I could now see the value of a clearly defined building ground system where ground measurement will be the same throughout the facility. But again, this needs to be verified.
Another thing I observed is that there is no surge arrestors at the substation, which has 3 transformers, 2 of them are delta-delta, and the last one which supply this MEDDAC facility is delta-wye, but with nothing. I recommend one to be put in. In Japan, they often use delta-delta system in the past, and their ground system are floating, which is dangerous from NEC perspective.






RE: Building grounding
1. The delta-detla connected transformers jump out at me as a possible issue. Ungrounded systems, while permitted by the NEC under certain conditions, were recognized long ago as very susceptible to transient overvoltages in the event of ground faults. Do you have ground detection alarms on those systems?
2. How is the generator connected? Is it on the ungrounded or grounded system? Take a close look at the generator grounding and neutral and ground arrangements at the ATS with respect to the commercial source grounding.
3. Make sure the TVSS is the correct type for the grounding arrangement of the system.
4. In step 5, you said the generator stopped and in step 6 you said the ATS switched to commercial power. Is that really the order things happened? That seems wrong.
5. For a hospital, I would expect to see a well documented and maintained grounding system. It seems like the substation ground should be connected to the building ground.
6. Fire alarm systems normally send a supervisory alarm when they loose power. Perhaps that is what you saw there. It shouldn't cause a general evacuation alarm however.
RE: Building grounding
While the the other 2 delta-delta transformers are not in our scope, but you are absolutely correct, these 2 units do not have grounding. They are feeding the HVAC and X-ray respectively. In additon, I also realised the 3rd transformer, which is supplying the building service, does not have surge arrestors.
One side quesiton I want to ask is, do you normally put surge arrestors, ground fault protection, and TVSS together in the distribution, or is there redundancy? For example, NEC required GFP for disocnnect rated 100A or more on solidly grounded wye system 150 V or more. In that case, is TVSS still required at the service?
RE: Building grounding
I'm sure you meant "1000A" in your last note. Since GFP and TVSS are solutions for two different conditions, I would say that if you're trying to limit surges, then yes, TVSS would be a good thing to have at your main switchboard.
Mike
RE: Building grounding
The NEC requires at least two levels of GFCP (e.g. at the service and main feeder breakers) in hospitals.
RE: Building grounding
grounding does not seem to be an issue here..can u shed more light on what caused the first outage and what exactly was repaired?
RE: Building grounding
And the repair (not really repair) involves only relocating feeders (the previous contractor placed too many cables in some of the conduits that exceed the fill ratio), and also replace circuit breakers with correct AIC rating. And therefore the power outage.
One thing the contractor did was perform the insulation test. I don't believe the TVSS was disconnected prior to they perform the test. The megger test might have damaged the TVSS. However, it did not trigger alarm, until step 7 above.
The facility is known to have grounding problem. And I know the scope of this project does not improve grounding at all, just replacing old switchgear with new ones.
RE: Building grounding
Did they do a "megger" test, or was it DC Hi-Pot?
Mike
RE: Building grounding
So the building actually did not loose power in steps 6, 7 and 8. Only TVSS went in alarm and some equipment were found damaged. Is that correct.
If so, I would say most likey culprit is the meggar test whose high voltage damaged sensitive equipment, but it did not have enough power to power an alarm.
RE: Building grounding
So I thought, this may not be the only problem. System grounding and other issues as aleman mentioned earlier might be present as well.
RE: Building grounding
RE: Building grounding
After reading your post again, it sounds like a low voltage system, so you probably meggered. Like rbulsara said, the megger test probably damaged your TVSS device. Also, depending on the TVSS, it may be the "fusible link" type. Does it have MOV's?
Mike
RE: Building grounding
Mike
RE: Building grounding
Thanks for the tips and clarifications.
RE: Building grounding
Re-reading your original post, the delta-delta transformers apparently have nothing to do with this facility?
The TVSS surge voltage rating is 400V L-N, which would be suitable for a 120V L-N (208/120V) solidly grounded system. Is that what you have?
RE: Building grounding
So when they flipped the main breaker at the substation, it is potential for large in-rush, especially substation vaccum breaker is 600A (12.5kA).
Yes, the TVSS you described is what we have.
The DC test for cable insulation is usually 500V, which exceeds the 400V L-N. I fear that might be the cause.