Commercial Building Ground Ring Design

[rgeisler]

I have a building that I designed the Elctrical Distribution system for, under constructon. I designed my standard ground ring utilizing #4/0 copper, with 5/8" x 10' ground rods spaced 20' OC. qround the building. As I was never privy to a soils report, and as most of the soil is sandy loam in this area, I utilized a ground I wrote a performance specification that the total system ground be at 5 ohms or less. In addition to the copper ring I required a ufer ground at the foundation level as well as the NEC 250 requirements to bond to water service etc. My problem is this:
The building sight sets on a lot of Rock at the foundation level. The contractor cannot drive the rods and if he did limestone is not a good media for conduction any way. I have now been directed by the owner to assist the contractor in determining the minimum number of grounding plates (Harger is the brand I suggested), the quantity of chemical ground rods (if Any) required and any other ground ring modificfations that will meet MY performance specification of 5 ohms or less.
So to what I could use help with:
I am using the Tables and formulas as found in IEEE Std 142-1991 to attempt to calculate various ring topologies resistance and fault current capability.
Does any-one out there have any suggestions other than this methodology? My firm cannot support the cost of the commercial software and I found this sight via my search for information (on H.B Dwight "Calculation of resistance to Ground&quot.
At any rate, judging by previous threads, it appears tha this is a good group to ask.
Thanks.

 

[kcsubeng]

This may or may not be applicable, but...

If you look at IEEE Std 80 - IEEE Guide for Safety in Substation Grounding, it gives eqns. for calculating the upper and lower ground system resistance that can be obtained from your soil resistivity model. It may be that no matter how much copper you put in the ground around the perimeter of the building, you won't get 5 ohms (I've had this happen to me in a substation application).

 

[peebee]

I've recently heard that encasing your ring in concrete can be a great way to reduce resistance to ground. Regular concrete or cement should work well, there's also a special conductive concrete made specifically for this application that should work even better. The concept is somewhat similar to a Ufer ground.

The idea is to dump concrete into your ground ring trench before backfill. The concrete is hydroscopic, which helps, and increases the effective surface area of your perimeter loop conductor, which also helps.

In rocky ground though, no matter what you do, you may never be able to achieve 5 ohms without resorting to extreme measurese, such as 200-foot bored ground wells & 200-foot deep rods, etc.

If I was you I'd try contacting the various ground system mfgr's and getting their take on it. Maybe you can get one of them to agree to meet & guarantee a functional spec.

By the way, you might want to reconsider the criteria for a 5-ohm resistance. That's pretty low even for good soil conditions. It's also probably unnecessary -- do you mind if I ask what kind of facility this is? Telecommunications?

 

[rgeisler]

Forst of all: Thanks so much for your interest and replies.
The building is a corporate headquarters. It has one complete floor dedicated to a testing lab. It also has a small data floor (2000 sf.).
My spec has evolved from Telcom, Data Center and pharmaceutiacl campuses. Of course the Telco and Data floor specific specifications are more sever and owner derived. I just recently started working in the large scale commercial arena.
The 5 ohm requirement is at the Top of the suggested range of 1-5 ohms per the IEEE Std 142-1991 section 4.1.2., for general industrial and commercial applications. In my experience of the last few years all the sites I've worked on have managed to achieve sub 3 ohms. I guess I was just lucky!
I like the suggestion about the concrete but as this is a commercial, I fear that Value Engineering may preclude that option.At any rate my attack now is to set a target of 5 ohms and calc the resistance with just thering at the worst case soil measurement given me. And then I'll add plates and finally chemical rods as required. Does that seem reasonable?

 

[peebee]

Sounds good to me.

Just remember, concrete's cheap. And chemical rods need ongoing replenishment which increases maintenance costs, they may not be as cheap as you think in the long run. And if someone forgets to dump salt in them, they were just a big waste of money.

Also -- and I'm certainly not trying to contradict IEEE here -- but I'd suggest reviewing the criteria behind the specified ground resistance limits. A higher resistance to earth ground will not substantially affect normal day-to-day operation of the equipment; bonding between equipment will have more of an effect on that, you could put all the stuff in an airplane and it would still work. Lightning protection could be an issue, so you might consider running horizontal conductors out from the corners of the building to assist in dissipating lightning strokes. And again, so long as all equipment within the building is solidly bonded to each other as well as to all piping, steel, metal roofing, stacks, etc., and a reasonable intentional path to earth ground exists, lightning damage should be minimized, and decreasing resistance will have little additional effect.

Your plan is not a bad one. Just make sure the owner understands the maintenance implications, and you should understand the rationale behind your design criteria.

 

[rgeisler]

I concur with the maintenance of the rods. At this point I've performed some calcs which indicate that in the worst of the soils (6874 ohms/cm), the average 5/8" rod that can be burried to depth vertically should yield ohms in moderately moist earth. That figre comes to 20.25 ohms. Per the book, if I can achieve 24 electrodes still seems like a bunch) I can achieve 1.84675 ohms for the system. This seems exceptionaly low and yet the issue of how much fault current each electrode can stand without "smoking" the electrode or drying out the soil around it tends to limit per rod ampacity to 70A per foot of rod. If I have calculated properly that yields 16,900 Amps of system fault capacity, for a very short time period (approximately .000093sec). Ihave NOT attempted to apply the corrective curves implied by the by the tables in the "Green" book, to account for soil temp and moisture content. Instead I have chosen to ignore the ring conductor itself, to give the design some "shelter". I understand that the bonding issue inside is of extreme importance, however it's my take that there are also issues here as to the conduction of Ground Fault currents and as mentioned, the lightning protection system down leads which will by code, be required to be bonded to the ring as they cross.
As suggested, I have Harger (the Plate electrode Vendor) reviewing the values I am assigning their materials. Interestingly enough, they were already aware of this project via their distributor.
I really appriciate the advice and will continue to post the resolution of this design problem.

 

[peebee]

Can you clarify your concern with ground fault current issues? I'm assuming that most of your conductors are in grounded metallic raceways, except for the utility service entrance, which won't be affected by your earth grounding system anyway. Do you have extensive customer-owned underground or overhead installations which are not in raceway? If not, I don't see what the concern with ground fault currents is, they'd all be flowing through metallic ground systems back to your transformer neutral.

Don't get me wrong. I'm a big fan of grounding. But I think a lot of systems are overdesigned, especially in the telecommunications industry.

 

[jghrist]

A resistivity of 6874 ohm-cm (not ohm/cm) is not high. I would think that your ground ring in conjunction with your ufer grounds would provide adequate resistance.