Grounding a big existing building

[bphgravity]

Sorry Dereck, I don't mean to imply the IEEE in whole or anyone associated is intentionally or unknowingly providing misinformation. I don't even know if it is the wrong information - I could be wrong. I know absolutely nothing about systems over 600V.

I always appreciate and learn a great deal from your posts here. Thanks.

 

[lile001]

Grounding a big existing building

I suspected you had 4160, that is why I danced around it.

As for the GES I have more questions than answers. Without seeing and being familiar it is hard to answer. Is the new service part of new construction? If so I would probable use a "Star Point" counterpoise under the the transformer pad. From there try to figure out how to bond to the other GES systems already in service either via sub-terrain or above ground.

Sorry not much of an answer on the GES part, need more info, lot more info.

Obviously you'd need to know a lot more and I don't expect you could do more than suggest some guildelines. The building is existing. No new concrete footings are planned so we don't get a CEE. I have reviewed plans back to the 1960's and grounding (and bonding!) is not addressed.

Another guy is inspecting the bonding (but not the earthing) in the building, and soon it will all be tied together nicely. He finds bonding violations every time he looks for them. None of the small distribution transformer SDSs are properly bonded. Oh yeah, and there are generators - several now and more planned soon. Not little ones either and nobody can tell us how they are bonded/grounded. All the power transformers are on the roof. I know that some of the delta connected services wander around, one leg being higher voltage than the others, which makes me wonder if they are tied to ground at all. In 4160V, you begin to worry about overvoltages if the service can wander around because it is not properly grounded. You also begin to worry that a phase to ground fault might not be detected!

Your point about subbing the lightning protection review to a master label firm is well taken. I'll go down that path instead of trying to do it myself. I've done plenty of lightning protection but not a system this complex.

You can see why I might be nervous about grounding existing buildings given the medium voltage and planned generators. This isn't the only existing building I am working on but it is the most complex.

Dont' know the exact soil resistivity, but it is in a loamy area near the gulf coast, which remains wet all the time and presumably was swamp before it was built upon. Soil resistivity and moisture should be very adequate. Would you go so far as to measure soil resistivity?

Very little steel structure except in the roof, most of the building is masonry or concrete walls and concrete columns. So there is not much there to tie into, nor to bond to. So far we have been running EGC's back to the service equipment for any SDS's such as transformers.

I could ask that they get out a chipping hammer and expose some rebar to bond to down in the lowest basement. I am tempted to do that to make sure I get a lot of sleep at night. There is also an old abandoned service entrance in a pit in the basement - I haven't inspected it to see if they left some handy grounding points there. It is very, very likely that when that service was moved, the new service was NOT earthed at all. The size of that service, the existing bonding violations, and the medium voltage makes me nervous about this.

So I am trying to establish some general guidelines to follow for this kind of job. For example, a target ground resistance.

 

[lile001]

Grounding and grounding

Grounding and grounding

Sorry Dereck, I don't mean to imply the IEEE in whole or anyone associated is intentionally or unknowingly providing misinformation. I don't even know if it is the wrong information - I could be wrong. I know absolutely nothing about systems over 600V.

I always appreciate and learn a great deal from your posts here. Thanks.

Bryan,

Your points pertaining to systems that are 120V to neutral are well taken. Even with low ground resistance, a 120V to ground system just won't generate much fault current if a phase is shorted to earth, and it is the bonding that is most critical, and also bonded surfaces will tend to be the most likely fault path. Bonded Counduit and EGC's can make a lot of fault current, and trip ground fault devices fast. I'll quit worrying about earthing these systems as much, as long as they meet the letter of the law.

I'm a little more worried about 480V systems, but even then, a phase faulted to earth on a 5 ohm ground resistance system could only generate 55 amps, not enough to trip a ground fault main breaker in many cases. Bonding is still the only hope to generate enough ground fault current to do anything. The code says the biggest ground conductor needed is 3/0, which will handle 55 amps all day without breaking a sweat.

A 4160 volt system might generate 800 amps in a phase-to-earth fault, a whole different ball game! A standard 3/0 grounding conductor is not even capable of handling 800 amps for very long. In these systems, you may be inserting resistors to limit potential ground fault current! This might be especially true with large generators and SDS's.

I would still argue that buildings on a continuum between a residential 200A service and a substation should have earthing sytems that follow that continuum in complexity.

 

[dereckbc]

Lawrence check out a company called XIT out of Torrance CA. They make the chem rods, and do soil analysis (you send a soil sample). They might even be able to help with on site testing 4-point testing. I also know of a company in Ohio and Idaho that specialize in grounding that can inspect and tell you what is there. PM me if interested.

Might be a good time to invest is a AEMC ground loop tester or rent one. If you know how to operate one, it will tell you everything you want to know without digging anything up.

 

[George Stolz]

Re: Grounding and grounding

Re: Grounding and grounding

Fellas, thanks for chipping away at my low voltage mindset, it's interesting to reset and try thinking about big stuff for a change. My mental pictures get radically different imagining how the current is moving at high voltage, I too know absolutely squat about it.

A 4160 volt system might generate 800 amps in a phase-to-earth fault, a whole different ball game! A standard 3/0 grounding conductor is not even capable of handling 800 amps for very long. In these systems, you may be inserting resistors to limit potential ground fault current!

I don't mean to detract from your problem for my own selfish desires, but could you describe how the resistors would be connected, and the reasoning behind this?

I would imagine first you would recommend upsizing beyond 3/0, hoping to kick up the capacity of the grounding system to open an overcurrent device, wouldn't you?

What size OCPD would be protecting such a system, would holding the GEC to an 800A maximum open this?

By all means, hush me up until you've answered your original problem, if you'd like. I'm the inappropriate kid tugging on your pants-leg at the dinner table; y'all can have your fill before you pay attention to me, I fully respect that.

 

[Pierre C Belarge]

Lawrence.

"The code says the biggest ground conductor needed is 3/0, which will handle 55 amps all day without breaking a sweat."

You seem to misunderstanding the difference between grounding and bonding.



"A 4160 volt system might generate 800 amps in a phase-to-earth fault, a whole different ball game! A standard 3/0 grounding conductor is not even capable of handling 800 amps for very long. In these systems, you may be inserting resistors to limit potential ground fault current! This might be especially true with large generators and SDS's"


This type of installation is not treated the same as typical premises wiring installations of 600 volts and less. It takes different engineering.

Wiriring 600 volts and less systems, fault current needs to go back to the source (most likely the transformer), meaning the grounding electrode conductor and the grounding electrode play such a small part in the action, that the code does not permit the use of the GEC/GE as the sole part of the circuit, hence single point grounding in premises wiring. Which means we need the connection at the service disconnect to be a good connection to the neutral conductor so the fault current goes back to the source.



From your description of the installation, it sounds like there may be some existing issues of which someone more familar with this type of installation should be hired to consult and help you with it.

 

[don_resqcapt19]

Lawrence,

A 4160 volt system might generate 800 amps in a phase-to-earth fault, a whole different ball game! A standard 3/0 grounding conductor is not even capable of handling 800 amps for very long.

A 3/0 can handle almost 4000 amps for 5 seconds without damage. The clearing time should be way less than that. Also assuming that the 4160 is a solidly grounded system, the grounding electrode system should not be part of the fault clearing path. It should be the EGC. I still say that once you are at the building, the bonding is far more important than is the grounding no matter what the voltage is.
Don

 

[lile001]

Lawrence.



This type of installation is not treated the same as typical premises wiring installations of 600 volts and less. It takes different engineering.

Yeah, that's why I am asking questions here. People keep talking about 600V wiring, I am trying to learn more about medium voltage wiring.

So far, my original question has remained un-addressed. People focus on bonding. Fine, We all agree that bonding is very improtant, and that it should be designed, installed, and inspected carefully, and most of us probably know the rules about bonding pretty well. We can table the bonding discussion, we all agree on it.

But the gist I get from this board is that folks would be fine putting a single ground rod in as the grounding electrode for a football stadium. That doesn't square with what I read about the subject in references written by the experts.

What exactly is "different" about the engineering at medium voltage?

 

[don_resqcapt19]

Lawrence,
I just do not see what benefit a connection to "earth" brings to the safety of the system. How does a "good" earthing system change anything in your large building example? What does a giant Ufer do that a single rod does not do?
Don

 

[PT]

Try burying 2 ground rods at distance from each other, connect them and test for resistance. Add a few more, maybe another 2 and connect them. So you have 4 connected together, test for resistance. If you get a different reading ( resistance should be lower with more rods), then that may be the reason for it. Of course chemical treatment of the ground lowers resistance also.

 

[winnie]

So far, my original question has remained un-addressed. People focus on bonding. Fine, We all agree that bonding is very improtant, and that it should be designed, installed, and inspected carefully, and most of us probably know the rules about bonding pretty well. We can table the bonding discussion, we all agree on it.

But the gist I get from this board is that folks would be fine putting a single ground rod in as the grounding electrode for a football stadium. That doesn't square with what I read about the subject in references written by the experts.

Thinking on this, I think that there might be a definition problem. Just a bit of musing on grounding and bonding:

In the original post, the issue of 'stray voltage' is brought up, with the example of a corner of a building at 50V relative to "local earth".

Whenever you are dealing with a voltage between two conductive items, and you connect a conductor between them to remove this voltage, this is _bonding_, not grounding.

But in the case of 'stray voltage', what you are 'bonding' are two separate patches of earth, and you are doing this with grounding electrodes.

The difference is that in 'grounding' you care about the impedance of the grounding electrode system to "Earth", meaning the _whole_ Earth. But when you are bonding a patch of "earth" for stray voltage, you only care about the impedance between the electrodes in contact with soil and that local patch of soil.

The NEC doesn't say much about 'stray voltage', and I have no knowledge to figure out how much of a life safety issue such 'stray voltage' is. I have seen one report of the sensation of shock between _bonded_ metal and soil on the other side of a house from the GES. (Hose bib bonded via metal pipe and GEC to building electrical system ground, shocks between hose bib and soil outside of house, earth current caused by local distribution transformer. )

A fair assessment is that the NEC considers design for stray voltage to be a 'design issue', except in the case of swimming pools.

-Jon