690.12 Rapid Shutdown Exception No.3 PI

[letgomywago]

Can't imagine my soaking wet firefighter gloves provides any voltage protection in the real world.

If your panels are in my way, might just fire up the 2000 GPM straight tip on the tower and wash them off the roof

If that's how they're designed for quick removal that'd be pretty cool

 

[tortuga]

Can't imagine my soaking wet firefighter gloves provides any voltage protection in the real world.

If your interested in some further reading of the research on the topic:

Hazard Analysis of Firefighter Interactions with Photovoltaic Arrays. (Conference) | OSTI.GOV

The U.S. Department of Energy's Office of Scientific and Technical Information

www.osti.gov

 

[don_resqcapt19]

If your interested in some further reading of the research on the topic:

Hazard Analysis of Firefighter Interactions with Photovoltaic Arrays. (Conference) | OSTI.GOV

The U.S. Department of Energy's Office of Scientific and Technical Information

www.osti.gov

Are most systems functionally grounded, and at what current level does the ground fault protection open the ground connection?

 

[mtnelect]

Greetings group I am kicking around an idea for a exception #3 to 690.12 for small 'corner case' PV systems:

Exception No.3 Where a PV system does not exceed 144 Square Feet in area, 2000 Watts DC and 250 Volts Open Circuit Voltage in any string rapid shutdown shall not be required.
This exception shall be limited to one PV system per building.


By limiting the maximum area to 12X12 and voltage to 250V such systems would be a very low risk to emergency responders.
This would carve out an exception for:
Smaller PV systems.
A PV system that is part on a carport and part on a roof.
Older PV systems that get updated or relocated (or like re-roofing).
Older systems that stay in place when new systems are installed.

Any input on how to best word it?

When installing a HVAC unit(s) on a roof, I always feed them from below, because it is a hazard to responding emergency personnel tripping over all kinds of obstacles. From PV systems, refrigerant lines, condensate line, power lines, Satellite cables, etc. And it is getting worse, the AHJ is not enforcing this anymore.

 

[don_resqcapt19]

If that's how they're designed for quick removal that'd be pretty cool

Doubt that they are designed for that, but also doubt they are designed to withstand a directed force of 8 tons of water per minute leaving the nozzle at about 100 PSI. Never seen anyone try it, and probably never will, hence the

 

[tortuga]

Are most systems functionally grounded, and at what current level does the ground fault protection open the ground connection?

I think most are ungrounded now.
Here is an old article from back when changes started to happen:

Ungrounded Electrical Systems! Ungrounded photovoltaic (PV) systems? What is the world coming to?

iaeimagazine.org

Edit I meant I think most are 'functionally' grounded. In that they are not solidly grounded.

 

[don_resqcapt19]

I think most are ungrounded now.
Here is an old article from back when changes started to happen:

Ungrounded Electrical Systems! Ungrounded photovoltaic (PV) systems? What is the world coming to?

iaeimagazine.org

Edit I meant I think most are 'functionally' grounded. In that they are not solidly grounded.

But I see the shock hazard as being the same for solidly vs functionally, unless the ground fault protective device will open the connection to ground at the current value that would flow though a person.

 

[jaggedben]

Are most systems functionally grounded, and at what current level does the ground fault protection open the ground connection?

Functionally grounded systems have been out of fashion for inverters rated less than about 50kW or so, for about a decade now. I would assume that means that the majority of arrays on buildings these days are ungrounded, although I wouldn't hazard a percentage guess much more precise than that. (Many larger rooftop systems use multiple inverters of less than 50kW.)

Functionally grounded systems typically relied on a 1A fuse for smaller systems. Some larger systems used a 5A fuse (I assume this was to avoid nusiance tripping from capacitive leakage on large arrays.) In at least one pretty well known case the 5A fuse was shown to be insufficient and I believe that contributed in no small part to the industry transition to ungrounded/non-isolated systems.

It's unclear to me if the Sandia paper cited above actually did a good job of simulating an older functionally grounded system.

 

[tortuga]

It's unclear to me if the Sandia paper cited above actually did a good job of simulating an older functionally grounded system.

They definitely did not but they discussed weather or not to. The reason to exclude it was double ground faults, where even small grounding resistances can open up a significant blind spot.
They looked at both the ungrounded and grounded cases specifically because the ungrounded case reverts to the grounded case when a conductor becomes grounded.

In theory if you take my proposal and make a equipment list for a system with any combination of equipment that meets the exception, then run it thru UL 3741 using the part 2 method in 12.2.3 you would come up with a very low risk category, becasue of the small size most of the things in Table 9 would be discarded. This is all inside the 144 SQ FT array boundary.

I should add they did a lot of good research and I feel like it all went into UL 3741 and none of it benefited 690. So even if it gets rejected, which it probably will, hopefully it will prompt a discussion.

 

[brycenesbitt]

Rapid shutdown has had a huge impact on reliability of installed PV. A significant fraction of the failures are in the rapid shutdown equipment itself. It's unclear if the vendors are at fault here, or if putting active electronics on roofs is just a bridge too far for the reliability engineering.

 

[jaggedben]

Rapid shutdown has had a huge impact on reliability of installed PV. A significant fraction of the failures are in the rapid shutdown equipment itself. It's unclear if the vendors are at fault here, or if putting active electronics on roofs is just a bridge too far for the reliability engineering.

By experience doesn't back this up. Have you actually collected data on this?

 

[don_resqcapt19]

The task group meetings that discuss the PI are open to the public and you can sit in on them if you ask for an invite. I have been sitting in on a few of them.
Go to nfpa.org/70, click on next edition and then view the "NEC First Draft Task Group Meeting Schedule". There is an email address where you can request an invite to the meetings. In some case the Task Group chairs will permit limited participation by guests who are not members of the task group.
Of course doing this is a lot easier when you are an old retired guy like me

 

[tortuga]

By experience doesn't back this up. Have you actually collected data on this?

There was a different proposal not anything to do with me or my exception proposal here to completely remove RS from the 2026 NEC, the submitter was on behalf of the Oregon State Fire Marshall and it cited at least one fire caused by RS.

 

[jaggedben]

There was a different proposal not anything to do with me or my exception proposal here to completely remove RS from the 2026 NEC, the submitter was on behalf of the Oregon State Fire Marshall and it cited at least one fire caused by RS.

Ok, but I was asking after the statement about reliability.

 

[tortuga]

Ok, but I was asking after the statement about reliability.

Problems with RSD according to one report are second to Microcracks:

https://solargrade.io/wp-content/uploads/2023/06/The-SolarGrade-PV-Health-Report_First-Edition.pdf

PV System Fires, Rapid Shutdown Devices, and Unintended Consequences -

Solar PV system fires are rare incidents with severe consequences. But are the regulations for reducing the risk of these events effective?

solargrade.io

I think other documented problems were less serious related to a type of system that used WIFI that was not working reliably.

 

[jaggedben]

Still really confused if we're talking about safety or reliability, and how it relates to MLPE.

 

[tortuga]

Well the arguments against RSD cite both.

 

[brycenesbitt]

Still really confused if we're talking about safety or reliability, and how it relates to MLPE.

They are the same thing, since any flaw can become high temperature.


According to the authors they are experts in post-failure analysis of commercial solar systems.
"We analyzed more than 60,000 PV system health datapoints to find out. Read the SolarGrade PV Health Report for a comprehensive, ground-up assessment of PV safety and reliability in the U.S. and Puerto Rico."
"The HelioVolta team has assessed the health of hundreds of distributed generation (DG)
PV systems as an independent provider of third-party inspection and quality control
services. Our clients span Fortune 50 companies, asset owner-operators, investment
firms, and insurers."

 

[jaggedben]

They are the same thing, since any flaw can become high temperature.
...

Whether it *can* happen is beside the point if it usually doesn't.

 

[brycenesbitt]

Whether it *can* happen is beside the point if it usually doesn't.

The quoted study are from failure consultants. The NEC is quite conservative: very rare events cause code restrictions.
Why not for RS devices?