MBJ Location

[dereckbc]

Unique installation at radio tower sites. Radio tower sites use Emergency Generators with an ATS. Some are configured as SDS, and most are NOT SDS. Typically the AC service drop is at the Meter Can, and Disconnect is located on the outside wall of the equipment shelter building. Service enters into a hybrid service rated Main Distribution Panel with built-in ATS and TVSS. Located above or adjacent to the service entrance is the Hatch Entry Plate, where all coaxes and other services enter the building. All to create a true Single Point ground where everything entering and leaving the building can be properly bonded to a single point. The site uses a Ground Ring, Chem Rod, Tower Piers, Building steel, CCE, and Radials for the GES. The equipment shelter Ground Electrode Conductor drops straight down to GES. The N-G bond is at the ATS .

However, I have a few tower sites that do not fit that model out in cow pastures and mountain tops. The AC Power is delivered outside the fenced compound on a Utility Pole. The Meter Can and Service Disconnect are located on the utility pole, and they are using the Pole-Butt ground as the GES, which is NOT bonded to the site GES. The pole is 100 cable feet away from the building being served. There is a 4 conductor feeder (Li. L2, N, and G) going to the Equipment Shelter ATS from the utility pole.

I need to treat the Feeder as a Service to bond the Neutral again to create a single Point Ground. Here is the big issue. Tower sites use 48 volt DC power. The 48-volt plants are referenced differently from the AC Equipment ground 100 feet outside the compound. Towers are struck by lightning with about every passing storm. The AC Ground and DC ground come in close proximity inside the rectifiers. The DC Plant is Separately Derived. During a lightning fault, the DC and AC ground are at extreme voltage differences causing major damage.

So what loophole or exception does NEC provide me to do that? I need the N-G bond at the ATS, which is rated as service equipment. Essentially I have 2-Service Disconnects—one out on a pole and another at the ATS.

FWIW the radio tower is owned and operated by the same electric utility supplying electric power.

 

[zbang]

A not exactly cheap way out is to bring the two hots to an isolation transformer and create a separately-derived system at the shelter (and basically forget that the service neutral exists).

 

[don_resqcapt19]

Does the utility require the disconnect at the meter? The code does not.

 

[dereckbc]

Does the utility require the disconnect at the meter?

Yes sir they do. The funny thing is the utility provides the GES via Pole Butt Ground. It is a rural lateral overhead service. On top of the pole is the transformer using Pole Ground to bond overhead neutral, Transfomer secondary CT, and at the AC equipment ground down at the Meter Can with Disconnect. In other words, the utility is supplying the ACEG one hundred cable feet away before entering the building. The Pole butt ground is not bonded to site GES.

I want to disconnect the Green inulated #2 equipment ground wire, and just run 3-wire and treat it like a Service rather than Feeder. I need the MBJ inside the ATS, not upstream located off the property on a foreign earth ground. .

 

[ActionDave]

I agree with don. There is nothing that says the service disconnect needs to be on the pole. Either get rid of it, change it so it's not a service disconnect, or just ignore it and rebond the neutral at the shelter. Before 2005 we could rebond at a remote structure and there is an exception in 250 that allows it for existing installations.

I don't see how moving the neutral bond to the shelter gets rid of any problem with a ground potential difference at the rectifiers though. How can it make any difference?

 

[dereckbc]

A not exactly cheap way out is to bring the two hots to an isolation transformer and create a separately-derived system at the shelter (and basically forget that the service neutral exists).

The thought occurred to me. However, would require a fairly large costly transformer. Most cell tower and radio sites use a standard 240/120 200-Amp service just like a residence. However, they are heavy constant loads. The Battery Plant can easily draw 100-amps requiring another 60-amp HVAC to keep the place from melting down from excessive heat.

 

[dereckbc]

I don't see how moving the neutral bond to the shelter gets rid of any problem with a ground potential difference at the rectifiers though. How can it make any difference?

Glad you asked. The neutral is bonded to earth a considerable distance away from the building. The GES used to bond the service neutral is not bonded to the site GES. Those two points are not at the same potential. Under normal conditions, the voltage potential is not great enough to cause any problems. However, if there is a utility HV fault or lightning strike the voltage potential difference will be in the thousands of volts between the AC neutral and the site ground. Also, known as step-potential-difference. There are only three circuits that operate at 120 volts that require a neutral; Wall outlets with nothing plugged in for test equipment only. lighting normally off, and rectifiers. Lightning flashes over from neutral to any nearby ground.

If the AC service neutral were bonded to the site GES, no potential difference would exist even under fault conditions.

 

[infinity]

Dereck it's been a while welcome back. Hope that all is well.

 

[ActionDave]

Glad you asked. The neutral is bonded to earth a considerable distance away from the building. The GES used to bond the service neutral is not bonded to the site GES. Those two points are not at the same potential. Under normal conditions, the voltage potential is not great enough to cause any problems. However, if there is a utility HV fault or lightning strike the voltage potential difference will be in the thousands of volts between the AC neutral and the site ground. Also, known as step-potential-difference.

Please understand that I am not being argumentative, I am genuinely fascinated by radio and cell tower grounding.

Given the amount of bonding that goes on inside and outside the structure that I've seen on cell sites I imagine it must be very similar at your radio tower. I'm still having trouble understanding where the step potential difference changes just by moving a screw from a box on the pole to a box at the shelter.

There are only three circuits that operate at 120 volts that require a neutral; Wall outlets with nothing plugged in for test equipment only. lighting normally off, and rectifiers. Lightning flashes over from neutral to any nearby ground.

And it's a darn good thing that every cell site I've worked on specs a full size 4/0 copper neutral.

 

[wwhitney]

I'm still having trouble understanding where the step potential difference changes just by moving a screw from a box on the pole to a box at the shelter.

My understanding:

The DC system is earth referenced at the structure. The AC system is earth referenced at the pole, 100 feet away. The grounded DC and grounded AC conductors are in proximity in some equipment. If there's an earth potential difference over that 100 feet, it shows up as a potential difference between those proximate conductors.

Cheers, Wayne

 

[don_resqcapt19]

That is an application where I could see giving you a 90.4 permission to put the main bonding jumper at the second disconnect, isolate the first service disconnect from the neutral and include a supply side bonding jumper from the second disconnect to the first to bond the metal parts and provide a fault clearing path.
I don't think I would permit a 3 wire feeder, as without a main bonding jumper at the first disconnect, there would be no fault clearing path if there is a fault to the metal parts at that disconnect. Yes, I am aware that clearing the fault at that location often means the fault has to burn itself clear, as it is not common for the primary fuse to open.

 

[dereckbc]

Dereck it's been a while welcome back. Hope that all is well.

Thanks, retired for a short period and got dragged back in the game.

 

[wwhitney]

I don't think I would permit a 3 wire feeder, as without a main bonding jumper at the first disconnect, there would be no fault clearing path if there is a fault to the metal parts at that disconnect.

Presumably the upstream disconnect would be bonded to the neutral in that setup.

Just slap a "Not Service Equipment--Emergency Disconnect Only" sticker on the upstream disconnect. : - )

Cheers, Wayne

 

[dereckbc]

Please understand that I am not being argumentative, I am genuinely fascinated by radio and cell tower grounding.

It was not taken that way, and I hope I did not come off sarcastic. I have not been around in a while, so you may not know my personality. I shoot from the hip and mix in a little fun along the way. You know a wise arse.

Perhaps a different approach. Our profession is full of myths, and a big one is ground rods or electrodes are at Zero-Volts potential, presumably with respect to each other. It can get you into trouble by placing yourself in a ground loop between two earth potentials.

There is a voltage gradient across the surface of the earth (dirt). The voltage is a result of current flowing through dirt resistance. There are many sources of the current, but the major contributor is utilities using dirt as a return conductor or neutral. Dirt is cheaper than wire. Utilities multi-ground the neutral on every pole or tower. That multi-ground neutral dirt resistance path is much lower than the overhead wire. For example, 10 miles of aluminum 2/0 conductor will be on the order of a thousand ohms impedance. The dirt is in parallel with neutral with an impedance in the single digits of less than 10-ohms—most of the utility unbalanced current returns via the dirt.

Suppose you drive a 10-foot rod and drive another rod 20-feet away. You will measure an AC voltage between the two rods and see it on an oscilloscope. Additionally, you might see DC voltages driven by cathodic protection systems at airport lightning protection systems, marinas, and pipelines.

Normally the voltage potential under normal operating conditions is very small and causes no problems. The only voltage developed would be the I*R loss on the neutral with normal load current. We like to limit that to 2 or 3 volts. The problem occurs if there is a lightning strike or primary to second utility fault.

The neutral conductor is referenced to dirt 100 feet away using the utility Pole but ground. From that point, the neutral is isolated from the ground all the way to the point of use in the rectifiers, lighting, and HVAC circuits. Under a lightning fault, the voltage difference between the neutral and site ground will be in the 10's of thousands of volts where there should be no voltage difference between N-G.

A single-point ground is now required in NC 250.94. You bring everything to a central entry point to make a Ground Window where every cable has to pass through and be bonded before entering or leaving. In a home or technical facility, you only bond to earth one time and one time. Otherwise, you will place yourself in a Ground Loop between two earth-grounds. If you do that, then you have no ground. All your equipment and wiring is just a bonding jumper bonding two electrodes together with all equalizing currents flowing through your equipment. For current to flow requires a point to enter and a point to exit or a path.

Mike summed it up pretty well on a video. "Bond everything together below grade. Bond everything above dirt together. Then bond the two together at a single point. Neither system can share currents. You never want to put yourself in parallel with earth ground.

Ever hear of stray voltage at dairy farms and heated livestock water tanks. The service is a long distance away from where it is being used. The animal is standing in the mud in a pasture at a heated water tank in winter is at a difference of potential because of the distance. A horse or cow goes for a drink and receives a mild shock. Dairy farms same thing; the milking machine is a distance away from service. Put the cups on the teets, and the cow receives a mild shock and does not deliver milk.

I hope that helps.

 

[dereckbc]

That is an application where I could see giving you a 90.4 permission to put the main bonding jumper at the second disconnect, isolate the first service disconnect from the neutral and include a supply side bonding jumper from the second disconnect to the first to bond the metal parts and provide a fault clearing path.
I don't think I would permit a 3 wire feeder, as without a main bonding jumper at the first disconnect, there would be no fault clearing path if there is a fault to the metal parts at that disconnect. Yes, I am aware that clearing the fault at that location often means the fault has to burn itself clear, as it is not common for the primary fuse to open.

Don thanks for chiming in. I hope all is well. You know my background and what I am trying to do. I have thought of using 90.4 waivers; I have the Stamp top back it up. Ironically this is the utility arguing between themself. The utility supplying the power is the same utility using the power. There is no inspection to deal with. Just two departments that do not understand what the other is doing.

If you were the inspector, this would be a piece of cake. I can't get the distribution dept to understand that utilities do not provide the customers with the GES. No utility in their right mind would do that or allow a customer to bond to the utility GES. Distribution's only concern is that the Meter Can and Disconnect are solidly bond to protect their equipment and employees. That is fine, and I have no problem if they want to bond with the people, but that is not our service entrance.

The problem is they are pounding their chest; the neutral cannot be bonded again. That completely disregards the safety and operation of what is inside the building. So I am looking for a code-compliant answer that will allow me to bond neutral again and treat it like a service. I have no problem carrying four wires if that makes everyone feel better. I want to bond neutral again at the ATS.

So 90.4 is one code loophole. But what about 250.6, Objectional Current? Rearrange ground conductors to prevent objectionable current. Never use to be a problem until code cycle 2005 I think.

 

[ActionDave]

The neutral conductor is referenced to dirt 100 feet away using the utility Pole but ground. From that point, the neutral is isolated from the ground all the way to the point of use in the rectifiers, lighting, and HVAC circuits......

If you were the inspector, this would be a piece of cake. I can't get the distribution dept to understand that utilities do not provide the customers with the GES. No utility in their right mind would do that or allow a customer to bond to the utility GES. Distribution's only concern is that the Meter Can and Disconnect are solidly bond to protect their equipment and employees. That is fine, and I have no problem if they want to bond with the people, but that is not our service entrance.

The problem is they are pounding their chest; the neutral cannot be bonded again. That completely disregards the safety and operation of what is inside the building. So I am looking for a code-compliant answer that will allow me to bond neutral again and treat it like a service. I have no problem carrying four wires if that makes everyone feel better. I want to bond neutral again at the ATS.

Are you saying that you only have three wires running between the utility pole and the radio tower shelter?

 

[don_resqcapt19]

Presumably the upstream disconnect would be bonded to the neutral in that setup.

Just slap a "Not Service Equipment--Emergency Disconnect Only" sticker on the upstream disconnect. : - )

Cheers, Wayne

If you remove the main bonding jumper, the disconnect should not be bonded to the neutral, however it is likely that there is a factory bond between the neutral meter enclosure in the meter can.

 

[don_resqcapt19]

...
So 90.4 is one code loophole. But what about 250.6, Objectional Current? Rearrange ground conductors to prevent objectionable current. Never use to be a problem until code cycle 2005 I think.

Nice to see you back. 250.6(B) might work. I had not thought of that. One of the remedies in 250.6(B)(1) through (4), especially (1) or (4) would work if the powers to be would agree.
I can see you using (1) to remove the MBJ at the first service disconnect, and using (4) to add the a second bonding jumper at the second disconnect. Either of which should accomplish what you need.

 

[dereckbc]

Nice to see you back. 250.6(B) might work.

I think we are on the same page. You know my background is telecom PQ and Protective grounding. The goal is maximum life safety and system performance. So what if I:

Leave the N-G bond at the pole and carry all 4 conductors to the building service entrance. That makes the distribution folks happy. They get an earth reference at the meter and disconnect.

At the building service entrance, treat it like a service. I land both the Feeder Neutral and Green ACEG conductors to the entrance Neutral bus. Bring the site GEC to re-reference ACEG to the sites GES.

Essentially turns the 4-wire into a 3-wire service. Bonding the neutral twice using a green insulated #2 insulated conductor in parallel with neutral makes a nice oversized neutral conductor. In the event of a lightning fault, there will be a significant fault or equalizing current flowing between the two ground systems. The Neutral and ACEG wires become large bonding jumpers between the two GES systems. That would not cause any problems inside the protective area. We are forcing lightning to stay outside and discharge into the earth. Equipment inside would see no difference of potential using a single point ground.

It sounds like 250.6(B) and 90.4 are my loopholes being made by engineering review and design. Much easier to do before 250.32 was changed.

 

[synchro]

... There are only three circuits that operate at 120 volts that require a neutral; Wall outlets with nothing plugged in for test equipment only. lighting normally off, and rectifiers.

I'm curious, how large is the worst case neutral current? Any "objectionable current" should be some fraction of that. And then some limits on the resulting voltage drop over your ACEG wire between two N-G bonding points due to that current could be determined. This is just to see how "objectionable" all this could possibly be in your application. You still need to find a justification within the NEC, but quantifying the relative importance of this issue vs. the safety issues you mentioned from lightning or MV faults could be worthwhile.

Do the rectifiers (very likely switching supplies) draw mostly L-L currents or is there substantial L-N current on the120V rectifier circuit you mentioned above? If the L-N loads draw a relatively small fraction of the total power then using an isolation transformer such as zbang mentioned just for the L-N loads might be more acceptable.