Equipment Grounding For Solar Critter Guard

Status
Not open for further replies.

RQuinn

Member
Location
Albany, NY
Hello Everyone - first time posting, long time viewer here. Thanks to all in this forum for great insight over the years.

I'm looking to install an aluminum critter guard around the perimeter of some rooftop PV systems, attached to the module frames in a non-bonded manner, and I'm trying to sort out equipment grounding requirements. I was surprised to see this conversation hasn't started on this forum, as I know these solutions have been installed for a few years now, though maybe I haven't looked hard enough. When looking at NEC 250.110:

“exposed normally non-current-carrying metal parts of fixed equipment supplied by or enclosing conductors or components likely to become energized shall be connected to an equipment-grounding conductor (EGC)…”

How does this group feel about the likelihood of an aluminum barrier (similar to a gutter guard/fence) being energized?

How would you answer if the critter guard was coated with PVC, or some other non-conductive material?

Has anyone had a code enforcement official weigh in on this yet?

Seeking your valuable opinions on this topic before moving forward with the solution.

Thanks!
 

SolarPro

Senior Member
Location
Austin, TX
Good question. Under what scenario is that aluminum likely to become energized?

I just installed the SolaTrim around a SunPower array on a new garage apartment and home office we are building. We have a standing seam metal roof and potential problems with pigeons, squirrels and leaf litter. Having that protective barrier significantly reduces the likelihood of an inadvertent fault under the array. Also, we deployed the system with SolarEdge dc-to-dc converters, so the the array output drops to 1 volt per module in the event of a fault, rapid shutdown or loss of ac power.

I'm a worrier by nature but it seems extraordinarily unlikely that a piece of SolaTrim would become inadvertently energized in this scenario. Each piece is only a few feet long and is electrically isolated by the adhesive and the anodized coating. If I was an AHJ, I'd be much more worried about what might happen if a squirrel moved in under the array. It's a bad idea to make the perfect an enemy of the good.
 

Attachments

  • 26631716336_60eede282d_z.jpg
    26631716336_60eede282d_z.jpg
    86.4 KB · Views: 0

SolarPro

Senior Member
Location
Austin, TX
This sidebar to an article by Rebekah Hren and Brian Mehalic, Grounding Compendium for PV Systems, is also relevant:

To Bond or Not to Bond?

Many metal components in a PV system must be bonded to the EGC. According to NEC Section 690.43(A), equipment grounding is specifically required for “metal parts of PV module frames, electrical equipment, and conductor enclosures.” Furthermore, Section 690.43(B) (which refers to Section 250.110) requires an EGC between a PV array and other equipment fastened in place or connected by permanent wiring. Studying these requirements goes a long way toward answering this question: What components do you need to bond?

But what about the corollary: What components do you not need to bond? Do you need to bond metal wire clips? What about metal roof flashing? The answers to questions like these are subject to interpretation based on whether the components are likely to become energized. Fortunately, advances in PV-specific bonding devices and techniques are making the goal of long-lasting low-resistance connections easier to attain. Evolving installation best practices, such as improved wire management, can minimize the likelihood of accessories becoming energized.

If you are a system designer, installer or inspector struggling to decipher which components are subject to the requirements in Sections 690.43 and 250.10, an interpretation developed by the Standards Technical Panel (STP) for UL 2703 may offer some guidance. As part of the ANSI accreditation process, the STP considered whether accessories—like metal flashings, wind deflectors, perimeter screens, module clamps, mounting brackets, piers, pilings and so on—needed evaluation as part of a system under the UL 2703 standard. The STP determined that system accessories do not have to be bonded for equipment-grounding purposes, so long as the PV system meets the following requirements: It is a parallel power production source; the installation manual clearly defines a grounding path; conductors are managed in such a way that if a wire is accidentally cut, it cannot fall, droop or move and make contact with any accessories; and grounding devices and grounded locations are clearly marked as such. Note that the STP assumes that an appropriately rated grounding lug with a copper ground wire is clearly marked, since it is visible on the array.

Jeff Spies is the vice president of business development at Quick Mount PV, a company that manufactures flashed roof attachments, and the secretary of NABCEP. He is also one of the industry subject matter experts serving on the STP for UL 2703. Spies explains, “If you wire your system in a UL 2703–compliant manner, then you will not have to ground every mid-clamp or flashing.”
 

RQuinn

Member
Location
Albany, NY
SolarPro - thanks so much. I'm also installing SolaTrim with Solaredge solution - I agree that the anodized coating and adhesive certainly isolate it from the modules, and drastically reduce the likelihood of any rooftop conductor faulting (with protection from squirrels, birds, etc.). This section certainly gives me something solid to discuss with a multiple Code Enforcement Offices that may ask the grounding question:

The STP determined that system accessories do not have to be bonded for equipment-grounding purposes, so long as the PV system meets the following requirements: It is a parallel power production source; the installation manual clearly defines a grounding path; conductors are managed in such a way that if a wire is accidentally cut, it cannot fall, droop or move and make contact with any accessories; and grounding devices and grounded locations are clearly marked as such

I'm finding it difficult to imagine a scenario where the PV source/output conductors will fault and energize the barrier, and I'm glad you agree.

I'm curious, how do you feel about the longevity of the adhesive connection? I spoke with a few installers and customers who have it installed, and only issue I found was improper pressure applied during install which led to the barrier coming loose.

Best of luck with your beautiful new system. You certainly have the deluxe setup with SunPower/Solaredge. Are you planning for StorEdge + Powerwall addition in the future?
 

SolarPro

Senior Member
Location
Austin, TX
I'm pretty confident that the 3M adhesive will hold up here in Texas. I've seen a buddy of mine do pull ups on a sign he attached at to a door at his shop with this adhesive. It will be interesting to see if the weather takes a toll over time. Our challenge here is heat, obviously. Snow seems like a bigger challenge. If the adhesive gets brittle when cold, the downward forces of a snow damn could easily pry the guard loose. I'm not sure what critter guard solution would hold up to a record breaking winter.

While we do not have storage now, we should be able to accommodate storage in the future. Right now there isn't any economic justification for adding energy storage in Austin, outside of a pilot program at the Mueller development known as the Pecan Street Project. However, I did intentionally oversize our SolarEdge inverter relative to the array capacity (4.1 kW). If it ever makes sense to add energy storage in the future, we already have the storage-compatible 7.6 kW SolarEdge inverter. I've also built the service in a way that will make it easy to relocate circuits to a back up loads panel. The biggest challenge is probably wall space. I have a tiny shop that might be able to house the battery; otherwise, it would have to live in the carport.
 

pv_n00b

Senior Member
Location
CA, USA
I generally have a hard time believing any adhesive system outside of a 2 part epoxy applied to a properly prepared surface on a joint protected from sunlight will last even 2 years exposed outdoors. Peel and stick on a bare module frame is strong on day one, but I can't see it lasting. Show me some data on HALT testing.
 

SolarPro

Senior Member
Location
Austin, TX
This is the data sheet for the adhesive:

http://multimedia.3m.com/mws/media/10330O/3mtm-acrylic-foam-tape-4229p.pdf?fn=4229.pdf

Automotive applications are one of the few applications that actually experience more extreme environmental conditions than a rooftop PV system.

Personally, I'm not worried about the longevity on our project, in large part because this is a single project in my own back yard. I have a standing seam roof anchor permanently attached to the north slope of the roof for future O&M activities. If the adhesive fails, I can always come up with a new attachment system. Plus, I have seven leftover pieces of SolaTrim.

If I was the O&M manager for a solar installation company, the calculus would be very different.
 

pv_n00b

Senior Member
Location
CA, USA
Well, last time I checked car manufacturers were not building a product designed to last 20 to 30 years.

Luckily when it comes loose it won't kill anyone, it's just an inconvenience to have to go up and fix it. Just get to it before the squirrels settle in.
 

SolarPro

Senior Member
Location
Austin, TX
The labels on the inverter are to meet 690.51 [dc characteristics], 690.17(E) [potentially energized when disco open] and 690.5(C) [ground-fault]. Technically, the plaque next to the inverter meets the rapid shutdown identification in 690.56(C).

In this case, our inverter faces the main service panel and utility meter. If the inverter and main panel weren't co-located, I would totally add an additional label. Since you can see the rapid shutdown placard from the meter, an addition label at the inverter provides no real benefit. I might add a label at the main panel that basically says "PV system with rapid shutdown on site; turn around to shutdown PV system."

What is interesting (and now well known to Austin solar contractors) is that the you currently have to trigger rapid shutdown for SolarEdge via the inverter-integrated dc disconnect. If you open an ac disconnect or pull the main meter, the array will eventually settle out at the safety voltage (roughly 1 volt per dc-to-dc converter). But this won't happen within 10 seconds. While future SolarEdge products will allow for rapid shutdown initiation on the ac side of the system, current systems require dc initiation.

AC-initiated rapid shutdown is really the right end goal. If there ever is an emergency, you don't want first responders to have to do anything extra or different. That's only true if they can shutdown a PV system simply by shutting down ac power to the site.
 

pv_n00b

Senior Member
Location
CA, USA
SolarEdge has a kit you can install that will enable RSD with AC disconnect. It makes a software mod and adds a discharge resister to the DC input. From what I have seen it's pretty slick.
 

Carultch

Senior Member
Location
Massachusetts
SolarEdge has a kit you can install that will enable RSD with AC disconnect. It makes a software mod and adds a discharge resister to the DC input. From what I have seen it's pretty slick.

Does that discharge resistor hinder your system's performance by making current flow through an additional parallel path during normal operation?
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
Does that discharge resistor hinder your system's performance by making current flow through an additional parallel path during normal operation?

I've wondered about this a bit myself. It seems to be a pretty ordinary looking small resistor, my guess is it could only dissipate maybe 5 watts at most, which you're really not going to notice.

Someone else can probably give a faster guestimate than I regarding what the inverter internal resistance would have to be for that to work.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
The current is small since the capacitance is not large either. I agree that the lost power is negligible.
An "ordinary resistor" in electronic terms is 1/2 watt or smaller.
 
Status
Not open for further replies.
Top