PV System Grounding

[cdcengineer]

Anyone have a clear understanding of how to apply the NEC requirements of article 690.45 thru 690.47?

We have a ground mount 112kW array which consists of (10) rack mounted strings. The strings are (12) 235w panels, mounted (4) strings per rack. So essentially, we have 48 panels per rack, times (10) racks or (480) total modules.

The code requires that we run an EGC. This conductor must be #6 if not run in a raceway and subject to physical damage (NEC 690.46).

We are also required to have grounding electrode (or is it electrodes) for the array. It's my opinion that the multiple (48) module racks are only one array as they tie into a single 100kW inverter at 480V-Y.

It's always a point of contention with contractors and/or installers regarding what they have to install. I'm hoping the 2011 NEC will address this in more detail, but I haven't ordered my new code book just yet.

It's my understanding that we are required to provide an additional grounding electrode per NEC 690.47(D) as the array is not w/in 6' of the premises electrode. This #6 AWG GEC will tie in at the inverter which will be bonded to the AC side EGC (EGC & Bond can be one & the same per 690.47(C)(3)). I also believe that this #6 AWG GEC shall be connected to each array frame (which in this case I interpret this as each rack), and then tie to the additional electrode. I am proposing that the contractor installs a #2 ground ring per 250.52(4).

I am on or off track here folks? Does the #2 ground ring need to be buried? Any thoughts and input are greatly appreciated.

I've corresponded with installers and engineers throughout the Rocky Mtns, CA & Oregon, and each person has a differing opinion.

Input is greatly appreciated.

 

[Smart $]

Anyone have a clear understanding of how to apply the NEC requirements of article 690.45 thru 690.47?

Can't say I have a clear understainding, but I'm willing to participate in this discussion

We have a ground mount 112kW array which consists of (10) rack mounted strings. The strings are (12) 235w panels, mounted (4) strings per rack. So essentially, we have 48 panels per rack, times (10) racks or (480) total modules.

Before we get into discussing electrode(s) required, please describe in detail, the ground mounting structure. Include foundation structuring and conductive elements therein. The purpose is to determine if the foundation and supporting means qualify as an electrode(s).

The code requires that we run an EGC. This conductor must be #6 if not run in a raceway and subject to physical damage (NEC 690.46).

Confirmed.

We are also required to have grounding electrode (or is it electrodes) for the array. It's my opinion that the multiple (48) module racks are only one array as they tie into a single 100kW inverter at 480V-Y.

Unless I can get a better understanding of your perspective, I have to disagree. Part of the definition of an array is that it be a mechanically integrated assembly. I get the impression, from your description of racks, there is mechanical separation from each rack to other racks.

It's my understanding that we are required to provide an additional grounding electrode per NEC 690.47(D) as the array is not w/in 6' of the premises electrode. ...

Confirmed.

...This #6 AWG GEC will tie in at the inverter which will be bonded to the AC side EGC (EGC & Bond can be one & the same per 690.47(C)(3)). I also believe that this #6 AWG GEC shall be connected to each array frame (which in this case I interpret this as each rack), and then tie to the additional electrode.

You are mixing issues. A EGC is not a GEC, and vice versa... and on that note, how did you determine the #6 AWG GEC size [perhaps because you are looking at the electrode(s) being ground rods?]

I am proposing that the contractor installs a #2 ground ring per 250.52(4).

I am on or off track here folks? Does the #2 ground ring need to be buried?

For now, without knowing all the details and making an educated guess, a #2 ground ring would suffice... and yes it has to be buried :roll:

 

[cdcengineer]

Before we get into discussing electrode(s) required, please describe in detail, the ground mounting structure. Include foundation structuring and conductive elements therein. The purpose is to determine if the foundation and supporting means qualify as an electrode(s). :roll:

I understood the definition of array to be a mechanically integrated assembly of modules with support structure and foundation per NEC. So perhaps the rack configurations do constitute different arrays. I've been confused since some people define an array as a single DC power producing unit which to me means the full assembly of panels, inverter, etc.

In regards to the mounting structure, I'll try and explain what I know based on what the contractor has told me. The racks have 16"h x 12"w poured concrete bases supporting poles. There is no single structural member longer than 10' in contact with the earth (NEC 250.52(A)(2), and no conductor was poured into the sonotubes.

Regarding the #AWG EGC, I misspoke, there wil be a #6 AWG EGC on both the DC & AC side as the OCP dictates this size per 250.122. There will also be a bond between AC & DC per 690.47. However, the GEC & EGC are not one and the same.

I'm trying to upload some photos. The one attached is a mock 3D rendering. The actual frame rack photo file and .pdf of the site plan is too large to upload.

We can continue this after Turkey day - enjoy the day off.

Thanks

 

[Smart $]

I understood the definition of array to be a mechanically integrated assembly of modules with support structure and foundation per NEC. So perhaps the rack configurations do constitute different arrays. I've been confused since some people define an array as a single DC power producing unit which to me means the full assembly of panels, inverter, etc.

In regards to the mounting structure, I'll try and explain what I know based on what the contractor has told me. The racks have 16"h x 12"w poured concrete bases supporting poles. There is no single structural member longer than 10' in contact with the earth (NEC 250.52(A)(2), and no conductor was poured into the sonotubes.

Regarding the #AWG EGC, I misspoke, there wil be a #6 AWG EGC on both the DC & AC side as the OCP dictates this size per 250.122. There will also be a bond between AC & DC per 690.47. However, the GEC & EGC are not one and the same.

I'm trying to upload some photos. The one attached is a mock 3D rendering. The actual frame rack photo file and .pdf of the site plan is too large to upload.

We can continue this after Turkey day - enjoy the day off.

Thanks

So you have something to reply to come Friday...

From the rendering, I am even more certain that each rack, as you refer to them, is considered an individual array... regardless of being electrically connected to a single inverter... as there is no mechanical connection of the racks to each other (I don't believe the earth counts as a mechanical connection ).

While you are considering using a #2 ground ring, also consider that the ground ring must encircle each rack and be buried at a deph not less than 30". IMO, this would provide perhaps the best earth grounding without going beyond Code requirements. However, if you are looking at it from the perspective of "value" engineering, you could get away with two rod, pipe or plate electrodes for the lot and run #6 GEC's to each rack... but don't take this as my recommedation

Have a happy and safe turkey day yourself

 

[cdcengineer]

Interesting, I had a peer review, and they didn't point out that we were dealing w/ multiple arrays, but you've got me convinced. The issue is that these racks live on the top of underground cisterns which are used to contain sewage (i.e. wastewater treatment plant). Anyway, between 24" - 30" down live the concrete cisterns.

I'm going to have to think about this some more, but perhaps we could consider setting an 8' ground rod for each rack buried horizontally at 24"? as they cannot be driven 8' vertically. It would've been easy if the contractor had poured copper into the footers and created a UFER ground.

Let me chew on this while I digest the feast. Thanks for the responses and interesting discussion.

Hope all had a great holiday. Gotta love Thanksgiving. It is right there w/ the 4th of July as top honor holiday in my book.

 

[Smart $]

Interesting, I had a peer review, and they didn't point out that we were dealing w/ multiple arrays, but you've got me convinced. The issue is that these racks live on the top of underground cisterns which are used to contain sewage (i.e. wastewater treatment plant). Anyway, between 24" - 30" down live the concrete cisterns.

I'm going to have to think about this some more, but perhaps we could consider setting an 8' ground rod for each rack buried horizontally at 24"? as they cannot be driven 8' vertically. It would've been easy if the contractor had poured copper into the footers and created a UFER ground.

Let me chew on this while I digest the feast. Thanks for the responses and interesting discussion.

Hope all had a great holiday. Gotta love Thanksgiving. It is right there w/ the 4th of July as top honor holiday in my book.

If you look at 250.53, you'll see electrodes must be at least 30" deep, including the lower end of a trenched ground rod. Rather than trenching, I would power-hammer drive it in at roughly 18? downward from level, starting slightly below grade, such that the top is below grade when it is driven almost all the way in.

Also, dont forget about 250.56...

 

[cdcengineer]

Smart $ - thanks for the tips. These arrays live 24" above underground "Digesters" which are essentially giant concrete cisterns. I believe that our best bet is to use the 1/2" diameter rebar which was poured into the footers as the electrode(s) (NEC 250.52(3)). There's about 27' of rebar which is connected by steel tie wire at every footer. We'll measure the resistance and see what we get. Because there's not true undisturbed ground below our feet, it has me worried that we won't measure 25ohms (although this is not a requirement per NEC 250.56). I will recommend that we connect every array/rack structure with a #6 AWG CU to one of it's footers, and then each of these footers will tie with a #6 AWG CU to the inverter (which is the point of DC system interconnection). From there I am recommending we run a #6 AWG CU back to the premises electrode (as a supplemental electrode) to ensure that we have a quality ground as the 1200A MDS has building steel, water pipe and ground rod as it's grounding electrode.

The above GEC will not substitute for a equipment ground which will be run from each array/module through the combiners to the inverter. The AC side will also have and EGC run with the phase conductors.

Thanks alot for the input.

This project keeps getting more interesting as it appears the entire facility is supplied by an optional stand-by generator. We'll speak to the inverter mfg. to discuss possible issues. The grid would never see a back-feed potential in the event of a utility power loss due to the isolation provided by the xfer switch, however, the inverter might try and continue feeding the MDS as the source sensing circuit will only see a momentary loss of power until the gen-set fires and xfer switch flops over. The generator issue is a completely different topic, but just part of the usual "It'll be a piece of cake" according to my client. All projects are the same, they feel they don't need to bring in an engineer, than they attempt to slash the fee and trivialize the scope of work in attempt to lull you into a false sense of simplicity, and finally the entire team realizes that there's is almost no such thing as a slam dunk when it comes to construction.

It's what keeps em' all interesting.

Have a great weekend.

 

[Smart $]

... I believe that our best bet is to use the 1/2" diameter rebar which was poured into the footers as the electrode(s) (NEC 250.52(3)). There's about 27' of rebar which is connected by steel tie wire at every footer. We'll measure the resistance and see what we get. Because there's not true undisturbed ground below our feet, it has me worried that we won't measure 25ohms (although this is not a requirement per NEC 250.56). I will recommend that we connect every array/rack structure with a #6 AWG CU to one of it's footers, and then each of these footers will tie with a #6 AWG CU to the inverter (which is the point of DC system interconnection). From there I am recommending we run a #6 AWG CU back to the premises electrode (as a supplemental electrode) to ensure that we have a quality ground as the 1200A MDS has building steel, water pipe and ground rod as it's grounding electrode.

...

The CEE approach sounds good. This is why I asked about foundation details back in my first post. One aspect with using the rebar in the footers is that there is no depth requirement... only that the concrete is in direct contact with the earth.