Understanding that mysterious GEC. Purpose and construction strategy.

[Sunking]

Because that is where the inverters might be located. Hundreds of feet away from the main service equipment, to which it interconnects.

Sometimes it is on a rooftop.
Sometimes it is on a remote building.
Sometimes it is in a remote field for a ground-mounted array, yet still tied in to the building electrically.
Etc.

OK you can stop there because each scenario has different requirements. No way am I going to attempt to try to explain each situation. Example a pole mounted array detached from a building is treated as any AC service with no ground connection between them except the grounded circuit conductor.

[carultch]

Wait a minute, you just changed the whole picture in my head. Why are you having to run hundreds of feet?

Because that is where the inverters might be located. Hundreds of feet away from the main service equipment, to which it interconnects.

Sometimes it is on a rooftop.
Sometimes it is on a remote building.
Sometimes it is in a remote field for a ground-mounted array, yet still tied in to the building electrically.
Etc.

Maybe because there is no space in the main electrical room.
Maybe because I'm strategically meeting the contentious 690.12 with inverters on the roof, immediately adjacent to the array.
Maybe because the owner doesn't want the noisy inverters in the main building.
Maybe because 480 Volts three phase is more efficient at carrying power long distance than 420 Vdc.
Numerous reasons why we don't always have the luxury of putting everything right by the main service equipment and premises wiring electrode.

[Sunking]

Not so well if I gotta run hundreds of feet to get from the inverters to the electrode. Maybe across open land, along the rooftop, thru occupied spaces, or other environments where everything is best kept in a raceway.

Wait a minute, you just changed the whole picture in my head. Why are you having to run hundreds of feet?

[carultch]

Most try to avoid a raceway altogether. Once you get up to #6 AWG and larger can just use cable clamps. Most GEC are very short from the ground up and no conductors to run them with.

That works well when my rod/plate/or pipe is right there in the inverter room, and only authorized personnel can enter the room.

Not so well if I gotta run hundreds of feet to get from the inverters to the electrode. Maybe across open land, along the rooftop, thru occupied spaces, or other environments where everything is best kept in a raceway.

[Sunking]

Whether or not EGC only systems will cause problems eventually is beyond the scope of my knowledge, but I see some experienced people here making a good engineering case that including a GEC would be a better design, at least if lightning is a common occurrence.

Just a tip. No good installer, sparky, or train driver should ever be caught dead without a copy of Soar's Book on Grounding & Bonding. Buy a copy every 3 years with your NEC handbook. Buy them together and get a discount.

[Sunking]

You are telling me that it is a bad practice to put in in the same pipe as other conductors, as convenient as it may be. So this means it is best for its performance if it is put in a dedicated raceway.

Most try to avoid a raceway altogether. Once you get up to #6 AWG and larger can just use cable clamps. Most GEC are very short from the ground up and no conductors to run them with.

If there were ever an event that the GEC is actually used like a lightning strike or utility fault the current rise time is extremely fast with significant magnitude. If ran with circuit conductors some of the energy is going to be magnetically coupled into those conductors just like a transformer.

[carultch]

There is no restriction on putting the GEC in a raceway with other conductors other than the bonding that is required in 250.64 for ferrous raceways. Is that clear enough? Having said that it is a bad practice.

Yep. That seems to be clear enough.

Put it in any plastic raceway with or without the power conductors. No raceway grounding, because you can't ground plastic.

Put it in any non-ferrous metal raceway with or without the power conductors. Standard raceway grounding practices apply.

Put it in a steel raceway with the power conductors, and connect it to a bonding bushing at each end or other fittings that are suitable to connect a ground wire.

You are telling me that it is a bad practice to put in in the same pipe as other conductors, as convenient as it may be. So this means that if it is your choice, you would run a dedicated conduit for the GEC. And you'd recommend I do the same on future projects, where I have that ability.

[Sunking]

Also, I have yet to receive a straight answer about the raceway for the GEC. Does such a raceway have to be dedicated exclusively to the GEC ...yes or no?

There is no restriction on putting the GEC in a raceway with other conductors other than the bonding that is required in 250.64 for ferrous raceways. Is that clear enough? Having said that it is a bad practice.

[carultch]

Look you are making this more complicated than it really is.

It would be a lot less complicated if the people who write the NEC could make up their mind about what we are required to do, with something so rudimentary as what grounding electrode we should use. I thought using the facility grounding electrode system as a GEC destination is good enough, and the best solution you could do. Apparently, I am wrong.

This is a case where overbuilding a system out of ignorance, can actually make it less safe.

Most of the time, overbuilding a system either has no effect on safety, or improves safety, and the only drawback is that it is slightly more expensive. I'm not sure if this needs to be labeled or not...OK, I'm just going to label it. I'm not sure if this is "subject to physical damage" or not...OK, I'm simply going to build it to meet the rules for being "subject to physical damage".

Also, I have yet to receive a straight answer about the raceway for the GEC. Does such a raceway have to be dedicated exclusively to the GEC ...yes or no?

[sensij]

Are you saying that if I build the PV system's GEC under the same rules that I used in 2011 (and hopefully will be using again in 2017), then I can completely forget about 690.47(D) in NEC2014?

I'm in Massachusetts, and the new NEC's always apply for permits initiated on January 1st of the year of the NEC edition's name.

From what I understand, the 2011 cycle that didn't require the rod for the array was kind of a mistake. The 2014 cycle added the rod back in but doesn't require it to be bonded, maybe intentionally, but now by consensus, also a mistake. My uninformed guess is that 2017 will probably be back to 2008, with the rod required and bonding required. In the meantime, you've got a whole bunch of transformerless systems going up in CA and other 2011 NEC states using only an EGC. Maybe it wasn't such a big deal a couple years ago before transformerless inverters were so common and the GEC was required at least to the inverter to ground the DC-.

Whether or not EGC only systems will cause problems eventually is beyond the scope of my knowledge, but I see some experienced people here making a good engineering case that including a GEC would be a better design, at least if lightning is a common occurrence.

[Sunking]

Are you saying that if I build the PV system's GEC under the same rules that I used in 2011 (and hopefully will be using again in 2017), then I can completely forget about 690.47(D) in NEC2014?

Look you are making this more complicated than it really is.

690.47 (D) states you must add additional electrodes for the array, period end of story. It allows you the OPTION of NOT having to bond those electrodes to the existing electrodes. You can leave them isolated or bond them. Your choice either way complies. What I am telling you is comply with the code, and exercise your option to bond the new electrodes to the existing GES.

The hot button for Engineers. Electricians, and other professionals is two or more isolated ground systems permitted in the 2014 code cycle. The whole purpose is to save significant money in installation and material cost at the expense of safety. Solar Installation is a cut throat business, and if you allow Installation contractors to cut corners, they are going to do it because they really do not care if it is safe or not. It is not their problem nor will they be held liable when someone gets killed or home burnt to the ground. They followed the code, thus not responsible even if they know it is unsafe. Its all about money and promoting solar at any cost including loss of human life.

[carultch]

690.47 permits two separate ground electrodes. It does not mean or say shall be two separate ground electrodes. Do you understand what that means? It means you can do it either way. Those in the know do it the right correct way using a the facility ground electrode. The whole reason John put that in there is to save money on installation and labor cost to support his agenda of turning the world into solar energy without regard to life safety and property damage. For those jurisdictions who are on code cycle 2014, what few there are, are making an amendment to recall or write out 690.47 (D)

Are you saying that if I build the PV system's GEC under the same rules that I used in 2011 (and hopefully will be using again in 2017), then I can completely forget about 690.47(D) in NEC2014?

I'm in Massachusetts, and the new NEC's always apply for permits initiated on January 1st of the year of the NEC edition's name.

[Sunking]

Regarding 690.47, section 690.47(D) in the 2014 NEC very specifically requires the use use of an auxiliary (useless) electrode for array grounding. It does not just permit it.

Now there are a few exceptions to this, including a very badly worded one which probably intends to say that you do not need to install the auxiliary electrode if there is already an building electrode within 6 feet of where you would be putting the auxiliary electrode and you connect to that instead. What it actually says is that if you put the auxiliary electrode within six feet of the building electrode, it is no longer required to be there. Catch-22.

Dave part of your quote was lost here. You are correct the array shall be grounded to an electrode. Well Duh! But nothing prohibits you from bonding those electrodes together. I promise this is going away. Unless your jurisdiction is on code cycle 2014, it is a moot point. Those jurisdictions that have adopted 2014 are adding an amendment removing the requirements as written because anyone who knows and understand earthing knows it is extremely dangerous.

How John Wiley ever got this through is beyond me. I am no longer in the NEC. According to Mike he BS'd FD's and Insurance Companies, and got manufactures to go along with it to increase sales volume. This is why John Wiley is a Teacher and not an Engineer. He has no application or practical experience, He has an agenda as most professors have. In this case to promote solar at the expense of Life and Property safety.

There are two ways to properly bond an array on a roof top.

1. Ideally, but expensive is to install Strike terminals above the panels to put the array in the Sphere of Protection. and bond the panel frames to EGC. Strike Terminal are under NFPA 760.

2. Bond Panels with down conductors to facility GES using GEC. Lightning strike goes to earth, DC fault down and back up through Inverter GEC to clear fault. This is the idea of John Wiley, but did not require the bond between the two GE. That is what the fuss is about. Even if bonded is a piss poor way of doing it because if you do have a strike still allows lightning current to go right through the house electrical wiring and going right through the expensive Inverter which causes the Magic Black Smoke to escape.

[inetdog]

Mike Holt, and many others take exactly that position, namely that all of the ground electrodes are connected into a Ground Electrode System, GES, and all of the electrical parts inside the building are tied together by a network of Equipment Grounding Conductors, EGCs, and the two meshes are connected together at exactly one point, right at the place where the electrical service comes in.
That way an external condition that causes currents to flow through the earth (and therefore potentially through the GES) will not be able to cause currents to flow in the EGC network.
Connecting the two at more than one point is allowed by the NEC, but defeats this basic principle.

Regarding 690.47, section 690.47(D) in the 2014 NEC very specifically requires the use use of an auxiliary (useless) electrode for array grounding. It does not just permit it.

(D) Additional Auxiliary Electrodes for Array Grounding. A grounding electrode shall be installed...

Now there are a few exceptions to this, including a very badly worded one which probably intends to say that you do not need to install the auxiliary electrode if there is already an building electrode within 6 feet of where you would be putting the auxiliary electrode and you connect to that instead. What it actually says is that if you put the auxiliary electrode within six feet of the building electrode, it is no longer required to be there. Catch-22.

[Sunking]

Hahaha! Put in the superfluous electrode just to make a picky and ignorant inspector leave you alone, and then rip it out to make the system safe.

Hold the bus there my friend. Now you know why I refrain from answering grounding questions in a DIY forum.

690.47 permits two separate ground electrodes. It does not mean or say shall be two separate ground electrodes. Do you understand what that means? It means you can do it either way. Those in the know do it the right correct way using a the facility ground electrode. The whole reason John put that in there is to save money on installation and labor cost to support his agenda of turning the world into solar energy without regard to life safety and property damage. For those jurisdictions who are on code cycle 2014, what few there are, are making an amendment to recall or write out 690.47 (D)

Look all a GEC is a conductor where one end is terminated to the GES, and the other end terminated to something that needs a direct path to earth. What is that something? Well there is a long list. Allow me to draw a picture in your mind. Imagine a house where CATV and Telco come in on the north side, commercial AC power on the south side, TV antenna on the east side, and a GTI on the west side. Got that picture in your mind? The place is surrounded with services. Builder did a horrible job.

Sparky the electrical contractor shows up to do his rough in, sees what the moron GC did and says WTF? I am going to file a job change order and install a Ground Ring around the house. So Sparky digs a 18 inch deep trench around the house perimeter, lays a #4 AWG in the trench, thermal welds the ends together to make the Ground Ring. With me so far?

Then on all 4 sides where the services enter, Sparky thermal welds a # 6 AWG to the Ground Ring at each one of the 4 location, routes it up through a 1/2 PVC pipe, rolls up the ends to be terminated when the house is finished and Sparky returns to finish. Those 4 conductors are GEC's connected to the Ground Ring.

Sparky just did what Telco's, Data Centers, Communications facilities, Radio towers, commercial broadcast, and sensitive electronic facilities use.

If a fault comes via CATV or Telco, it is shunted to earth, rather than go inside the house looking for a ground inside. Primary falls on your AC service, fault is shunted to earth where your AC service enters rather than going inside through your expense TV to find ground via the CATV ground.

EGC's dos not route anything to earth. Earth connection has no function with an EGC. The EGC routes faults currents directly back to the power source. In the case of commercial AC power, it is routed directly to the Neutral buss in your service entrance via the Main Bonding Jumper that is required. Otherwise no circuit breaker would operate if it does not have a planned path.

GEC are for crap outside, and EGC is for crap inside. They are normally only common at one point at the service entrance.