Grounding

Discussion about the Power System PV Combiner (PSPV)

Moderators: OutBack Moderator Team, OutBack

sparky
Forum Emperor
Posts: 1216
Joined: Fri May 20, 2005 8:24 am
My RE system: OFFGRID SOLAR "we go where power lines don't" Design, consulting, and installation advice on 300+ Offgrid homes.
Location: Sierra Mountains near Mariposa Ca. Webpage on next line
Contact:

Post by sparky » Tue Sep 26, 2006 5:02 pm

DavidB only connect DC negative to ground at the DC side of the inverter where it is bonded to the ground rod or the power center where it is bonded to the ground rod or ground array.

Help to understand this is the basis of most all electronic systems and in PC board design it is called a star ground. Everything is tied to one point and that point goes to gound or chassis ground. The potential differences are therefore all the same. It works for RF real well also. There are exceptions and for conducted RF and induced RF but that is more magic at times. It can hurt your head to think about it so accept it.

The solar grounding has evolved over time so you can be legal at the time the system was installed but better techniques are available now so
accept it and the last point is there are tradeoffs in having a safe NEC system and having lightning protected system that has one design parameter that is above safety. That is my opinion only !

provo
Forum Czar
Posts: 525
Joined: Thu Jul 27, 2006 4:34 pm
My RE system: Sixteen Evergreen EC-120
(4 strings, total 1920W)
Eight Rolls S-550 (2 strings, total ~800Ah @ 24V)
One FM80
One VFXR3524A
Hub 10
Mate3s
FNDC and Trimetric
Location: Sierra foothills

Post by provo » Tue Sep 26, 2006 7:50 pm

I get what Sparky is saying about the star ground, but what about an
additional ground for the pole and module frames out at a distant
array site, just for lightning protection. Is that actually a BAD idea,
a GOOD idea, or maybe just unhelpful?

-K

Jim L.
Forum Guru
Posts: 298
Joined: Wed Jun 15, 2005 12:14 pm
Location: Texas

Post by Jim L. » Tue Sep 26, 2006 8:18 pm

The answer is that it depends. It's situation dependent.

How much money or how critical is the array? How far away? What is the incidence of lightning?

A good place to start here is the thread "Lightning Arrestors" in off-grid forum from July 2006, which has tutorials listed.

A pole is more susceptible to lightning. Normally, an air terminal is placed above the pole. The air terminal is then connected to a conductor which in turn goes to a ground electrode subsystem. It is not the same as the electronic ground, and is not the same as the power ground.

Best regards
Jim L.

Kent Osterberg
Forum Emperor
Posts: 1523
Joined: Mon Jun 09, 2003 11:01 pm
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules
Location: Cove, Oregon
Contact:

Post by Kent Osterberg » Tue Sep 26, 2006 9:46 pm

Provo,

When establishing a grounding system for lightning, I think the code requires it to be connected to the rest of the grounding system. Making that connection in a manner that will allow the energy of a lightning strike to dissipate into the grounding system without going through and damaging valuable items is a trade secret, bad magic, well established time-proven techniques, wishful thinking, dumb luck, and/or a combination of all of the above.

DavidB
Forum Guru
Posts: 246
Joined: Sun May 19, 2002 11:01 pm
Location: Minnesota

Post by DavidB » Fri Sep 29, 2006 5:21 pm

Sorry for not getting it but for a distant array it just doesn't make sense to me why I would want to run a ground wire 200ft+ from an array all the way back to the load center.. The ground at the PSPV is grounding the pole, the PSPV case, and the array frame and PV panel frames. Nothing here that is electrically connected to anything else. The DC negative would still be connected at only one point (back at the load center by the inverter). When lightning hits it will go directly to the ground rod at the array and not have an easy path right back to your house.

Perhaps the code was changed to simply make the code more concise and easier to follow and match what had been followed with AC systems? (to have single ground point).

If you did run a ground wire back to the load center, it would seem prudent to me to use a bare copper cable buried OUTSIDE of any conduit. That way you would have all that extra contact with the earth to make it that much better of a grounding system.

provo
Forum Czar
Posts: 525
Joined: Thu Jul 27, 2006 4:34 pm
My RE system: Sixteen Evergreen EC-120
(4 strings, total 1920W)
Eight Rolls S-550 (2 strings, total ~800Ah @ 24V)
One FM80
One VFXR3524A
Hub 10
Mate3s
FNDC and Trimetric
Location: Sierra foothills

Post by provo » Fri Sep 29, 2006 6:43 pm

I'm not sure, but I think the code is not always rational :-) Running the
ground wire 200 ft back to the electronics room is EXACTLY what the
building inspector just told me he wanted, and that the ground wire
should be INSIDE the conduit (insulated or not -- he doesn't care).

He also thinks a remote frame ground (pole, frames, box) wouldn't be
a bad idea for lightning, but he's not sure if it meets code.
It seems there are lots of different opinions....

-K

sparky
Forum Emperor
Posts: 1216
Joined: Fri May 20, 2005 8:24 am
My RE system: OFFGRID SOLAR "we go where power lines don't" Design, consulting, and installation advice on 300+ Offgrid homes.
Location: Sierra Mountains near Mariposa Ca. Webpage on next line
Contact:

Post by sparky » Fri Sep 29, 2006 8:45 pm

Provo,

I do not get your point unless you are being sarcastic. It seems you have a building inspector who knows the code and that is rare for solar.

What is your point? Is it not what was said in the previous posts? If I can
say one last thing here and that is the NEC is only the bare minimum and it is slanted for safety of human beings not equipment. If you are out in the hills and little kids waiting for the scholl bus will not be leaning on your array do what you want after your inspection!

How is the house going? Got heat yet?

sparky
Forum Emperor
Posts: 1216
Joined: Fri May 20, 2005 8:24 am
My RE system: OFFGRID SOLAR "we go where power lines don't" Design, consulting, and installation advice on 300+ Offgrid homes.
Location: Sierra Mountains near Mariposa Ca. Webpage on next line
Contact:

Post by sparky » Fri Sep 29, 2006 8:46 pm

Provo,

I do not get your point unless you are being sarcastic. It seems you have a building inspector who knows the code and that is rare for solar.

What is your point? Is it not what was said in the previous posts? If I can
say one last thing here and that is the NEC is only the bare minimum and it is slanted for safety of human beings not equipment. If you are out in the hills and little kids waiting for the scholl bus will not be leaning on your array do what you want after your inspection!

How is the house going? Got heat yet?

Kent Osterberg
Forum Emperor
Posts: 1523
Joined: Mon Jun 09, 2003 11:01 pm
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules
Location: Cove, Oregon
Contact:

Post by Kent Osterberg » Fri Sep 29, 2006 8:50 pm

David,
The earth shall not be considered as an effective ground path. NEC 250.4 (A)(5) and NEC 250.5(B)(4)
This is not a new notion in the code. With the PV array and structure connected to its own grounding system and not bonded to the main grounding system there are some serious safety issues.
  • A ground fault (positive wire to ground) of the PV array will not be cleared by breakers in the combiner box. That's because the fault current flows through the earth to the system ground. A path that is rarely low enough in resistance to allow a current that the breakers would clear. A ground fault protection system probably wouldn't even be able to sense the fault.

    Someone working in the combiner box could receive a nasty shock from the negative, normally grounded wire. This could happen even if all the PV array breakers are open. Imagine walking up to the PV combiner, switching all of the breakers off, then getting shocked when you grab one of the PV negative wires.
Nothing here that is electrically connected to anything else.
Systems must be designed to handle abnormal conditions too. Does the safety equipment work as it is supposed to? No, it doesn't when you have a separate grounding system for the PV array.

I agree with you that a bare ground wire running outside the conduit will greatly improve the grounding system. It's probably one of the most effective ways to reduce damage from lightning. However, the inspector may still require a ground wire in the conduit too.

sparky
Forum Emperor
Posts: 1216
Joined: Fri May 20, 2005 8:24 am
My RE system: OFFGRID SOLAR "we go where power lines don't" Design, consulting, and installation advice on 300+ Offgrid homes.
Location: Sierra Mountains near Mariposa Ca. Webpage on next line
Contact:

Post by sparky » Fri Sep 29, 2006 9:10 pm

DavidB,

I was asking this same question of Kent 6 months ago about the 250' remote grounds. Here is what helped me.

#1 It is code and you will fail an inspection if the inspector is in the know!

#2 It is similar in the code to the way remote buildings are inspected and how cables are run so the inspector will want ground in the conduit and in my area they want it insulated. Ask for your area!

#3 Kent sighted places where this (200' foot ground) might be good in stopping corrosion of metals. In my own experience on Cruising Sailboats it does make sense that all metals are bonded to a common point that is tied to a Zinc or less noble metal that is sacrificed to the god of the sea.

#4 This was added by the Guru Supremo John Niles about 6 years ago to his code check reference. Keep in mind that he once had a phone call from his wife telling him that Ball Lightning was dancing around his inverter in the living room. We all learn

#5 Kent went into alot of detail about the ground or dirt. That is where he lost me and I kept seeing the scene in Gone with the Wind with Scarlett talking about Tara and the good earth.................

So?

Jim L.
Forum Guru
Posts: 298
Joined: Wed Jun 15, 2005 12:14 pm
Location: Texas

Post by Jim L. » Fri Sep 29, 2006 9:53 pm

Sparky and Kent giving good advice.

One thing that muddies the water is the functions of "ground wire".

Kent's point about abnormal and out of normal circumstances. The "ground wire" has one function among many which is to detect and to clear faults. NEC wants it kept in same conduit as its conductors to minimize magnetic effects, and as an accountibility that there is a visual verification that the ground wire is present. In this function, the ground wire should not be in contact with earth.

Unless NFPA 780 is mandated by your jurisdiction, then the inspector will not be involved in lightning protection, other than what is in the NEC.

As Sparky and Kent have pointed out, the best approach is to connect each subsystem at one point only, for the purpose of reference (equipotential).

A rule of thumb is that components within 20 feet of each other share a common earth electrode system. Between 20 and 200 feet, either a common EES or the EES's connected with bare copper minimum of 2 1/0's. Over 200 feet, then separate EES's and not worry about connecting EES's. The EES's are mainly for lightning.

Then, one component would have a power ground, DC ground, signal ground, and lightning ground, tied together at one point. 200 feet away, the tower equipment would have a separate lightning ground tied into an EES. But the power or DC ground(s) would still run back to the main site for the reasons above.

Also, if lightning strikes the tower, some energy will go down the lightning protection system, but some will also run down the cables to the main site. The energy divides in ratio to the conductance paths.

One of the practical problems with earth grounding is that the connections to earth usually are not to spec. Wrong EES for the conditions and lack of maintenance make earth grounding dicey.

You will get different opinions because of different earth conditions, incidence of lightning, distances, impedances, and all the other details that should be considered.

Best regards
Jim L.

provo
Forum Czar
Posts: 525
Joined: Thu Jul 27, 2006 4:34 pm
My RE system: Sixteen Evergreen EC-120
(4 strings, total 1920W)
Eight Rolls S-550 (2 strings, total ~800Ah @ 24V)
One FM80
One VFXR3524A
Hub 10
Mate3s
FNDC and Trimetric
Location: Sierra foothills

Post by provo » Sat Sep 30, 2006 7:55 am

Jim L. wrote:
.....if lightning strikes the tower, some energy will go down the lightning
protection system, but some will also run down the cables to the main site.
The energy divides in ratio to the conductance paths.
Not to beat a dead horse, but this is the part that concerns me. With that
ground connection going back to the house, it's hard to imagine a ground
at the array that could take enough of a lightning strike to help much in
protecting the electronics.

-K

Jim L.
Forum Guru
Posts: 298
Joined: Wed Jun 15, 2005 12:14 pm
Location: Texas

Post by Jim L. » Sat Sep 30, 2006 10:44 am

Energy is going to go down every conductor, not just the "ground" wire.

That's why the main site has to have a good grounding scheme, even if the strike is expected at the tower. Every penetration of the main site has to be protected -- telephone, pump wires, satellite dish, etc.

Can look at it as a layered defense. NFPA 780 for the tower itself. Protection at the entrance to the main site. Protection on the electronics both at the tower and at main site. Knock-down protection on any paths into the main site. Low impedance EES.

Best regards
Jim L.

Jim L.
Forum Guru
Posts: 298
Joined: Wed Jun 15, 2005 12:14 pm
Location: Texas

Post by Jim L. » Sat Sep 30, 2006 10:46 am

Energy is going to go down every conductor, not just the "ground" wire.

That's why the main site has to have a good grounding scheme, even if the strike is expected at the tower. Every penetration of the main site has to be protected -- telephone, pump wires, satellite dish, etc.

Can look at it as a layered defense. NFPA 780 for the tower itself. Protection at the entrance to the main site. Protection on the electronics both at the tower and at main site. Knock-down protection on any paths into the main site. Low impedance EES.

Best regards
Jim L.

DavidB
Forum Guru
Posts: 246
Joined: Sun May 19, 2002 11:01 pm
Location: Minnesota

Post by DavidB » Sat Sep 30, 2006 12:29 pm

Kent Osterberg wrote:A ground fault (positive wire to ground) of the PV array will not be cleared by breakers in the combiner box. That's because the fault current flows through the earth to the system ground. A path that is rarely low enough in resistance to allow a current that the breakers would clear. A ground fault protection system probably wouldn't even be able to sense the fault.

Someone working in the combiner box could receive a nasty shock from the negative, normally grounded wire. This could happen even if all the PV array breakers are open. Imagine walking up to the PV combiner, switching all of the breakers off, then getting shocked when you grab one of the PV negative wires.
If I follow correctly the main issue is whether or not you can rely on the ground point at the array and the code says no.

Otherwise your scenario says that touching the negative from the load center that comes into the PSPV and any part of the metal on the grounded array/PSPV at the same time would complete the circuit with you in the middle if there is a short of the positive conductor touching the array frame and the ground point at the array is bad.

By tieing the array grounding to the load center grounding would provide a low resitance parallel path (the other path if you are touching the array frame, through you to directly to the ground below you which should be very high resistance).

If I am still following correctly then this would be why you need to size your ground wire correctly from the PSPV back to the load center; so that the resistance in that ground wire is not too high to make an alternate grounding path less resistant then the path it should be taking.

DavidB
Forum Guru
Posts: 246
Joined: Sun May 19, 2002 11:01 pm
Location: Minnesota

Post by DavidB » Sat Sep 30, 2006 12:30 pm

deleted
Last edited by DavidB on Sat Sep 30, 2006 12:35 pm, edited 1 time in total.

DavidB
Forum Guru
Posts: 246
Joined: Sun May 19, 2002 11:01 pm
Location: Minnesota

Post by DavidB » Sat Sep 30, 2006 12:30 pm

deleted
Last edited by DavidB on Sat Sep 30, 2006 12:34 pm, edited 1 time in total.

DavidB
Forum Guru
Posts: 246
Joined: Sun May 19, 2002 11:01 pm
Location: Minnesota

Post by DavidB » Sat Sep 30, 2006 12:31 pm

deleted
Last edited by DavidB on Sat Sep 30, 2006 12:34 pm, edited 1 time in total.

DavidB
Forum Guru
Posts: 246
Joined: Sun May 19, 2002 11:01 pm
Location: Minnesota

Post by DavidB » Sat Sep 30, 2006 12:31 pm

something is seriously broken with this message board!

sparky
Forum Emperor
Posts: 1216
Joined: Fri May 20, 2005 8:24 am
My RE system: OFFGRID SOLAR "we go where power lines don't" Design, consulting, and installation advice on 300+ Offgrid homes.
Location: Sierra Mountains near Mariposa Ca. Webpage on next line
Contact:

Post by sparky » Sat Sep 30, 2006 2:00 pm

Yes there is a problem with the Outback Site!

One thing that you have to keep in mind David is what happens if your local array ground system gets run over by the school bus or the tractor guy there will still be a path to ground.

The other thing is if you have GCFI the small ground wire back to the house will still keep the stray currents below the trip threshold and keep the circuit enabled to detect a fault.

I am with you in your thought that the 200 foot ground should be sized substantially bigger than code requirements. It is not required however!

DavidB
Forum Guru
Posts: 246
Joined: Sun May 19, 2002 11:01 pm
Location: Minnesota

Post by DavidB » Sun Oct 01, 2006 11:50 am

sparky wrote:#4 This was added by the Guru Supremo John Niles about 6 years ago to his code check reference. Keep in mind that he once had a phone call from his wife telling him that Ball Lightning was dancing around his inverter in the living room. We all learn
Here is that article by John Wiles describing what happened to his installation. It is not completely clear but it sounded to me like he DID have a ground wire from the array back to his load center and that is what caused the problem. He describes how he solved potential future lightning problems by disconnecting the ground wire that ran back to the load center and grounding the array directly at the array to ground rods he added. Yes this was written in 99.

In many parts of the country lightning would be far more of a safety issue than whether someone is going to open up the PSPV and touch the negative when a fault has occured. In this case the code could be changed to simply require a sticker or something to warn anyone opening a combiner box that they should check for current between ground and negative before touching.

The code for grounding at the array could also be more specific if people are worried about a bad ground... such as using sectional ground rods (my ground rod at the array is 24ft straight down), using heavier gauge ground wire (I used #2) and attaching the ground wire to the ground rod with an exothermic weld(SureShot).

Just my thoughts since 3 years ago we had a number of trees struck by lightning within 300 ft of the current array location, so lightning is of concern to me.

Jim L.
Forum Guru
Posts: 298
Joined: Wed Jun 15, 2005 12:14 pm
Location: Texas

Post by Jim L. » Sun Oct 01, 2006 1:57 pm

Just in case a novice is reading this thread.

A trained electrician will not grab a neutral or ground wire. Or even negative DC. This goes back to the abnormal and out of normal conditions.

One should never assume that "ground" is done correctly and correctly maintained, because experience is that they are not. You don't bet your life on it.

Neutral and neg DC are not grounds. They carry current which can be fatal.

You will never see an old electrician grab a ground rod. Fault or unbalanced current can be flowing into the ground rod. The potential between the electrican's feet and the potential on a ground rod often is not zero.

Best regards
Jim L.

Kent Osterberg
Forum Emperor
Posts: 1523
Joined: Mon Jun 09, 2003 11:01 pm
My RE system: SMA Sunny Boy 2500 with eight Mobil Solar Ra 180 modules
Location: Cove, Oregon
Contact:

Post by Kent Osterberg » Sun Oct 01, 2006 5:17 pm

David,


Here's the revelent paragraph from John Wiles Flash Kaboom article. Not only did he add a new grounding system for the PV array, he bonded it to the original main grounding system. The PV array should never be grounded to a grounding system that isn't connected to the main grounding system.
I drove three 8-foot ground rods directly under the PV modules. The frame of each PV module was connected by a number 6 AWG bare copper conductor directly to the nearest ground rod with appropriate clamps on each end. Each of these new ground rods was connected (bonded) to the original main ground rod with a number 2 AWG bare copper wire buried two feet in the ground. The equipment grounding conductor in the UF cable was disconnected at both ends. These changes allowed any direct lightning hits or induced currents in the PV module frames to be directly shunted to the earth. By disconnecting the equipment grounding conductor in the UF cable, surges previously traveling along it could no longer be induced into the adjacent current-carrying conductors. The NEC requires that the ground rods be bonded and this bonding forms a far more effective grounding system than a single rod. The rods should be at least six feet apart to meet the requirements of Section 250-84 of the NEC.

Post Reply