PV Array grounding and lightning protection

[ryangittens]

Fellow PV Professionals,

According to 690.43(A), all exposed metal pars must be connected to an EGC. So on a PV array, all the metal parts (frames and rails) are bonded together and connected to and EGC which goes to the inverter. If there is a fault, it's cleared at the inverter. Correct?

I'm currently having a debate as to whether or not a GEC coming off an array and going directly to the existing grounding electrode helps with lightning protection and unnecessary inverter tripping.

I have seen instances where the array is not connected to an EGC run back to the inverter and only this GEC to existing electrode is used. I am of the opinion that this is definitely incorrect because should a fault occur on the array, it might not be interrupted by the DC GFPD.

While I am no expert on lightning protection, I also don't believe that the GEC would help that much during a direct strike as the current would pass through the house anyway.

Am I right to say that removing the GEC and bonding the array to the EGC is the correct thing to do?


Thanks in advance!

 

[jaggedben]

Fellow PV Professionals,

According to 690.43(A), all exposed metal pars must be connected to an EGC. So on a PV array, all the metal parts (frames and rails) are bonded together and connected to and EGC which goes to the inverter. If there is a fault, it's cleared at the inverter. Correct?

A fault is not necessarily 'cleared' at the inverter but if a fault is properly detected then the inverter will shut down. This should both reduce the chances of people getting shocked as well as the chances of a fire starting from a second fault, which are both important. In systems with optimizers or similar features they will likely also shut down and kill dangerous voltages from the array.

I'm currently having a debate as to whether or not a GEC coming off an array and going directly to the existing grounding electrode helps with lightning protection and unnecessary inverter tripping.

It does not do either of those things. "Helping with unnecessary inverter tripping" sounds to me like "masking dangers from faulty wiring."

I have seen instances where the array is not connected to an EGC run back to the inverter and only this GEC to existing electrode is used. I am of the opinion that this is definitely incorrect because should a fault occur on the array, it might not be interrupted by the DC GFPD.

You are correct.

While I am no expert on lightning protection, I also don't believe that the GEC would help that much during a direct strike as the current would pass through the house anyway.

Depending on how it's installed some say it could make it the danger from a nearby strike worse.

I have seen some people opine on other forums that if you have a ground mounted array that is separate from a house it serves, then the NEC requirements that the array structure have a grounding electrode and that it be bonded to the house grounding electrode system increase the danger of lightning energy taking a path into the house. While I can understand that argument, and I'm no expert on the likelihood of that actually causing damage, I would note that the same problem would seemingly apply to NEC requirements for supplying detached buildings without solar arrays, and that doesn't seem to be controversial. At any rate, the issue does not apply to arrays mounted on the same structure they serve.

Some points to round out the subject:
- it is a myth that having a solar array on top of a structure increases the likelihood of lightning hitting the top of the structure because there's a lot of metal there. (If such were true, then no barns in the midwest would have metal roofs.)
- Actual lightning protection systems are a whole other animal from installing a GEC to NEC standards. See NFPA 780, etc..

Am I right to say that removing the GEC and bonding the array to the EGC is the correct thing to do?

In a nutshell, yes.

 

[ryangittens]

A fault is not necessarily 'cleared' at the inverter but if a fault is properly detected then the inverter will shut down. This should both reduce the chances of people getting shocked as well as the chances of a fire starting from a second fault, which are both important. In systems with optimizers or similar features they will likely also shut down and kill dangerous voltages from the array.



It does not do either of those things. "Helping with unnecessary inverter tripping" sounds to me like "masking dangers from faulty wiring."



You are correct.



Depending on how it's installed some say it could make it the danger from a nearby strike worse.

I have seen some people opine on other forums that if you have a ground mounted array that is separate from a house it serves, then the NEC requirements that the array structure have a grounding electrode and that it be bonded to the house grounding electrode system increase the danger of lightning energy taking a path into the house. While I can understand that argument, and I'm no expert on the likelihood of that actually causing damage, I would note that the same problem would seemingly apply to NEC requirements for supplying detached buildings without solar arrays, and that doesn't seem to be controversial. At any rate, the issue does not apply to arrays mounted on the same structure they serve.

Some points to round out the subject:
- it is a myth that having a solar array on top of a structure increases the likelihood of lightning hitting the top of the structure because there's a lot of metal there. (If such were true, then no barns in the midwest would have metal roofs.)
- Actual lightning protection systems are a whole other animal from installing a GEC to NEC standards. See NFPA 780, etc..



In a nutshell, yes.

Thank you for helping me put this to rest sir! I really appreciate it.

 

[pv_n00b]

There is no protection for a direct strike to the PV array or structure, it's likely toast. Lightning protection is about making sure that direct strikes hit a purpose-built lightning protection system with aerial spikes and grounding. At best any protection that is provided to the PV system is to protect it from the effects of a nearby strike or a strike on the power lines that is it interconnected to.
The NEC does not address lightening protection directly, that is covered in NFPA 780.

 

[ggunn]

There is no protection for a direct strike to the PV array or structure, it's likely toast. Lightning protection is about making sure that direct strikes hit a purpose-built lightning protection system with aerial spikes and grounding.

I am not an expert on lightning protection, but all that I have read on the subject indicates that the purpose of a lightning rod is not to attract lightning to itself but to prevent strikes from happening through charge dissipation. As I understand it, in the presence of a charge differential between earth and sky, the sharp point on a lightning rod squeezes the equipotential planes so close together that the dielectric strength of the air cannot withstand the differential and breaks down, allowing the charge to dissipate before the differential gets high enough to provoke a strike. This charge dissipation can cause the air molecules to ionize and glow, which is the cause of what is commonly called "St. Elmo's Fire".

Additionally, the wires I have seen connecting lightning rods to ground looked to me to be far too small to carry the full current of a lightning strike.

 

[pv_n00b]

I am not an expert on lightning protection, but all that I have read on the subject indicates that the purpose of a lightning rod is not to attract lightning to itself but to prevent strikes from happening through charge dissipation. As I understand it, in the presence of a charge differential between earth and sky, the sharp point on a lightning rod squeezes the equipotential planes so close together that the dielectric strength of the air cannot withstand the differential and breaks down, allowing the charge to dissipate before the differential gets high enough to provoke a strike. This charge dissipation can cause the air molecules to ionize and glow, which is the cause of what is commonly called "St. Elmo's Fire".

Additionally, the wires I have seen connecting lightning rods to ground looked to me to be far too small to carry the full current of a lightning strike.

This is from the Lightning Protection Institute (http://lightning.org/learn-more/faq/), so probably fairly authoritative:

"When a lightning protection grounding network is in place, the strike is intercepted and directed to ground without impact to the structure, occupants or contents."

You are probably thinking of the charge dissipators that look like a ball with a bunch of wire spikes coming out or a whisk broom. I've seen those used on sailboat masts and aircraft. They are supposed to prevent charge buildup that may attract lightning.

 

[ggunn]

This is from the Lightning Protection Institute (http://lightning.org/learn-more/faq/), so probably fairly authoritative:

"When a lightning protection grounding network is in place, the strike is intercepted and directed to ground without impact to the structure, occupants or contents."

You are probably thinking of the charge dissipators that look like a ball with a bunch of wire spikes coming out or a whisk broom. I've seen those used on sailboat masts and aircraft. They are supposed to prevent charge buildup that may attract lightning.

According to http://www.aharfield.co.uk/lightning-protection-services/about-lightning:

"An average lightning strike discharges about 30,000 amperes (20,000 amperes in the UK). The current in a lightning strike typically ranges from 5,000 to 50,000 amperes depending on the strength of storm. NASA has recorded strikes of 100,000 amperes and there are other reports of strikes over 200,000 amperes."

What's going to happen to that #6 bare copper wire if it's asked to carry 30,000A to ground?

 

[jaggedben]

...

What's going to happen to that #6 bare copper wire if it's asked to carry 30,000A to ground?

What are you worried about? It's not continuous! :lol:

In all seriousness, this nonsense about an 8 or 6 awg GEC protecting a solar array from lightning damage is exactly that: nonsense.

 

[ggunn]

What are you worried about? It's not continuous! :lol:

In all seriousness, this nonsense about an 8 or 6 awg GEC protecting a solar array from lightning damage is exactly that: nonsense.

I don't believe that a lightning rod is supposed to "attract lightning"; who in his right mind would want to do that? The explanations I have seen (most of them, anyway) speak of the breakdown of the dielectric capacity of air so that charge is dissipated before it reaches the point where a strike occurs.

 

[jaggedben]

I don't believe that a lightning rod is supposed to "attract lightning"; who in his right mind would want to do that? The explanations I have seen (most of them, anyway) speak of the breakdown of the dielectric capacity of air so that charge is dissipated before it reaches the point where a strike occurs.

Ok here is my understanding, for what it's worth...

Lightning rods are intended to be the point where lightning strikes a building if lightning is going to strike that building. A large metal conductor(s) are intended to carry the lightning current to ground so that it doesn't travel through parts of the building that are more likely to explode when such current travels through it. Lightning tends to strike the highest object in the immediate vicinity of the charged field, or a least one of the higher ones. That's because the dielectric capacity of air breaks down more readily when you have less distance through that air (duh), and this, along with the (far far less predictable) location of the highest charge differential, is the primary thing which determines where lightning strikes. If you have a building that is the tallest object around, you want to protect it by putting a lightning rod on top with a beefy conductor to the ground. 'Beefy' seems to mean 2awg or larger, from what I read, and bigger for taller buildings. The fact that the lightning rod is metal has essentially nothing to do with whether lightning will strike it. It is the rod's location, not its material, that 'attracts' lightning, but it is the material's ability to carry the current non-destructively that provides protection from damage, or at least mitigation thereof. I don't really know, but I suppose that because the duration of the strike is so short, conductors can carry a much higher current than for normal uses. I further suppose that charge dissipation like you describe may also prevent strikes from occurring. To my knowledge, there's no reason it can't be both/and.

Now with respect to PV arrays, there is nothing about a PV array being a PV array that typically makes it more likely to be struck by lightning than any other part of a building it's on. Adding a PV array to a building does not inherently constitute a reason to add a lightning protection system where there wasn't already one. Upsizing the bonding (or grounding) conductor from 10awg to 8awg is basically useless. And grounding the array to earth through more than one point - say, to the premises electrode and to the auxiliary electrode required in the 2014 NEC - actually increases the risk of damage to property and people from gradients caused by nearby strikes. I think the CMP finally recognized both these points last time around and that's why they removed the requirements.

 

[ggunn]

Ok here is my understanding, for what it's worth...
That's because the dielectric capacity of air breaks down more readily when you have less distance through that air (duh), and this, along with the (far far less predictable) location of the highest charge differential, is the primary thing which determines where lightning strikes.

To expand a little bit from what I have read on the subject...

The difference in path length between that from the cloud to the array or building and that from the cloud to the lightning rod is minuscule compared to the distance from the cloud to either, so I don't think that makes the difference with respect to the dielectric strength of the air. What makes the difference (again, from what I have read) is the fact that lightning rods are sharpened to a very fine point. The rod is at ground potential and the planes of equal voltage potential between ground and the cloud are squeezed very close together as they traverse the point of the rod, so close together that in the presence of a high voltage differential between ground and cloud, the dielectric of the air is broken down to the extent that it conducts. This results in a relief of the voltage differential to the point where a strike doesn't happen.

I reiterate that I am a layperson when it comes to lightning protection, but this is an explanation I have read in several places, and it makes sense to me.

 

[jaggedben]

To expand a little bit from what I have read on the subject...

The difference in path length between that from the cloud to the array or building and that from the cloud to the lightning rod is minuscule compared to the distance from the cloud to either, so I don't think that makes the difference with respect to the dielectric strength of the air. What makes the difference (again, from what I have read) is the fact that lightning rods are sharpened to a very fine point. The rod is at ground potential and the planes of equal voltage potential between ground and the cloud are squeezed very close together as they traverse the point of the rod, so close together that in the presence of a high voltage differential between ground and cloud, the dielectric of the air is broken down to the extent that it conducts. This results in a relief of the voltage differential to the point where a strike doesn't happen.

I reiterate that I am a layperson when it comes to lightning protection, but this is an explanation I have read in several places, and it makes sense to me.

I'm a lay person too, but what you're saying seems to be a controversial point, at the least.
Just a couple quick google results.

 

[ggunn]

I'm a lay person too, but what you're saying seems to be a controversial point, at the least.
Just a couple quick google results.

I don't think anyone truly and completely understands lightning.