705.12(D)(3) GFP and Supply Side Connections

Status
Not open for further replies.

Jeff S

Member
Location
St. Louis, MO
2011 NEC 705.12(D)(3) Exception states, "Connection shall be permitted to be made to the load side of ground-fault protection provided that there is ground-fault protection for equipment from all ground-fault current sources. Ground-fault protection devices used with supplies connected to the load-side terminals shall be identified and listed as suitable for backfeeding." I'm referring of course to those services requiring GFP per NEC 230.95 which is pretty much all 480/277VAC 3 ph systems of 1000A or more.
Comments and questions:
What is, "provided that there is ground-fault protection for equipment from all ground-fault current sources." referring to? Apparently not the GFP in an inverter since that's on the DC side. Mike Holt's book on PV basically says that if there's GFP on the main then connect to the supply side. This can be very difficult in a large building. John Wiles has previously stated that he wasn't aware of any GFP breakers that were rated for backfeeding. Was he referring to branch circuit breakers or main breakers? When I called G.E. tech support about their Spectra Series, I was told that the Power Break 1 or 2 frames are all rated for backfeeding. I believe this system uses a Ground Fault Relay. Is that the key- if the GFP is separate from the breaker?
Are all you installers out there connecting to the supply side or is this rule being ignored?
Is it enough to verify that the main breaker doesn't specifically state "line" & "load" terminals to verify that it can be backfed?
Any help clarifying this would be greatly appreciated.
J.S.
 

Jeff S

Member
Location
St. Louis, MO
Update

Update

With regards to, "When I called G.E. tech support about their Spectra Series, I was told that the Power Break 1 or 2 frames are all rated for backfeeding. I believe this system uses a Ground Fault Relay. Is that the key- if the GFP is separate from the breaker?" , when I spoke with them today, they retracted the o.k. because the breakers can be fed from either direction by moving a jumper. They aren't meant for more than one source. They are looking into it further to see what affect the solar inverter might have on their GFP.
J.S.
 

Jeff S

Member
Location
St. Louis, MO
Like some feedback

Like some feedback

"Are all you installers out there connecting to the supply side or is this rule being ignored?"
I'd really appreciate some feedback to how others are handling this?
Have you run across any 208 3ph systems with GFP?
Thanks,
J.S.
 

SolarPro

Senior Member
Location
Austin, TX
Yes, this is an issue in commercial/industrial settings that PV system integrators deal with all the time. This article might be helpful:

http://www.iaei.org/magazine/2010/11/backfeeding-ground-fault-circuit-breakers/

In light commercial applications, most main breakers are "standard types" (per the article above) and are suitable for backfeeding. I've never come across a GFP-type main breaker in 3-phase 208 V service. But this is a smart thing to be on the lookout for during a site survey.

If you come across a GFP-type main breaker, look for "line" and "load" markings. If these markings are present, then the device cannot be backfed and you'll need to consider a supply-side connection. If these marking are not present, I'd still contact the manufacturer before backfeeding any GFP breaker.

FWIW: In some cases supply-side connections are quite easily accomplished. For example, you may find that you can land the inverter output circuit conductors on empty lugs or holes in the busbars on the line side of the main service enclosure. If so, Bob's your uncle...
 

Jeff S

Member
Location
St. Louis, MO
Min Wire to Tap Service

Min Wire to Tap Service

When making a line side connection in a large commercial system (for example 3000A) it's my understanding that the tap rules don't apply because this is actually another service. It's also my understanding that NEC doesn't specifically state the min size wire and fused disconnect for this but the consensus is that it should be at least #6 and 60A OCPD/Disconnect. Do you agree or does that need to be larger?
Thanks very much for the link to the article. That's the best discussion that I've seen on the GFP breaker issue.
J.S.
 

SolarPro

Senior Member
Location
Austin, TX
According to Section 705.2 in NEC 2011 PV systems are classified as power production equipment: "The generating source, and all distribution equipment associated with it, that generates electricity from a source other than a utility supplied service." As defined in Article 100, the term service refers to "conductors and equipment for delivering electric energy from the serving utility...." So a PV system is not a service as defined in Article 100. However, Section 230.82(6) allows for the disconnecting means for a PV system to connect to the supply-side of a service disconnecting means.

Unfortunately, the Code does not provide a lot of guidance about how to do this, which understandable if you recall that Section 90.1(C) states that the Code is not intended to be a design guide. Mike Holt and others provide some design guidance. The gist of that advice as I understand it is to build the disconnecting means for the power production equipment (the PV system) to the standards outlined in Article 230 as a best practice. The conductor and disconnect sizes you mention in your last post come from Article 230; see 230.23(B) and 230.79(D).

Mike Holt expands on the best practices for supply-side connections in this article:

http://solarprofessional.com/article/?file=SP5_1_pg14_QA&search=
 

Jeff S

Member
Location
St. Louis, MO
Thanks for your response.

Thanks for your response.

Those have been very informative posts. I forwarded the IAEI article to the Tech Support guy at G.E. for help identifying which type of GFP they have in their Spectra Series.
I really appreciate the very applicable information.
I'm still curious to hear if other installers are just making supply side connections to be sure for larger commercial projects. The engineers designing new buildings aren't taking this into consideration as they usually specify our connection point as being down stream from the main disconnect (and presumably the GFP).
One thing I still didn't get even from the great articles is what the exception means, "....provided that there is ground-fault protection for equipment from all ground-fault current sources."

Thanks,
J.S.
 

Jeff S

Member
Location
St. Louis, MO
Regarding IAEI article

Regarding IAEI article

The IAEI article includes the following:
"MCBs equipped with GFPE are distinctive in that they have a white ?pigtail? wire intended for connection to the neutral bar in the panelboard in which they are installed (see figure 1). Connecting the pigtail wire serves two functions. First, because the branch circuit neutral wire, also called the grounded conductor, must be connected to the circuit breaker rather than to the neutral bar, it completes the branch circuit. Second, it also completes the power supply circuit for the electronic ground-fault detection circuitry.
http://www.iaei.org/magazine/2010/11/backfeeding-ground-fault-circuit-breakers/larsen_fig3/

When correctly connected, a ground-fault will result in the circuit breaker contact(s) opening, cutting power to the trip solenoid. If the circuit breaker is backfed, power to the trip solenoid will not be cut when the circuit breaker contact(s) open, resulting in damage to the trip solenoid, which is not rated for continuous duty. For this reason the terminals on these circuit breakers are identified with ?line? and ?load? markings."

The underlined sentence above makes sense if you were backfeeding a generator or alternate utility source but I don't think applies to inverters. When power is tripped the anti-islanding will shut down the inverter and therefore not feedback continuous power to the trip solenoid. Agree or am I missing something?
Granted that we would probably never need to connect to one of these smaller breakers with GFP. The section of the article on larger breakers is what we're really dealing with.
Thanks,
J.S.
 

BillK-AZ

Senior Member
Location
Mesa Arizona
You might be missing something.

While the inverter will shut down due to various faults and detected islanding, it can take up to 2 seconds to do so. This is long enough to damage the trip coil. Then it will not work for further ground-faults and it is difficult to detect a damaged trip coil.

There is a good description of the various shutdown conditions and allowed trip time in the following Enphase link (the actual requirements are in UL-1741 and IEEE-1547 and these are copyrighted documents and I do not have current versions):

http://enphase.com/wp-uploads/enpha...hite_Paper_Meaning_of_Utility_Interactive.pdf
 

Jeff S

Member
Location
St. Louis, MO
Thanks, I didn't realize that 2 sec was too long and would be considered continuous. My understanding is that continuous means 3 hours or longer.
J.S.
 

PWDickerson

Senior Member
Location
Clinton, WA
Occupation
Solar Contractor
I installed a load side connected PV system at a school recently with a GFP protected 3000 amp main breaker, 277/480. During the site consultation I noted the make and model of the main breaker, called the manufacturer, confirmed that the breaker was suitable for backfeed, and had them forward me the documentation confirming that. I called our AHJ, and he brought my attention to the NEC code section you mentioned. (2011 NEC 705.12(D)(3) Exception states, "Connection shall be permitted to be made to the load side of ground-fault protection provided that there is ground-fault protection for equipment from all ground-fault current sources. Ground-fault protection devices used with supplies connected to the load-side terminals shall be identified and listed as suitable for backfeeding.") The AHJ required the PV backfeed breaker have GFP protection. The manufacturer of the backfeed panel did not make a GFP breaker for that panel style, so we installed a GFP device ahead of the backfeed breaker. The AHJ inspected the system himself and approved it.
 

Jeff S

Member
Location
St. Louis, MO
Thanks for your response. I never thought about that section as requiring GFP for the inverter breaker. What type of external GFP device did you use?
J.S.
 

PWDickerson

Senior Member
Location
Clinton, WA
Occupation
Solar Contractor
Schneider Electric, Multi 9 GFP. It is DIN rail mounted and looks like a 2-pole breaker. It comes in lots of flavors, had a long lead time from our supplier and is somewhat pricey.
 

Jeff S

Member
Location
St. Louis, MO
That was very helpful information. It looks like for the 3000A 480Y277 application you mentioned (almost identical to what I'm looking at), that would require the 4 pole version. Do you recall which tripping range you selected?
Thanks,
J.S.
 

PWDickerson

Senior Member
Location
Clinton, WA
Occupation
Solar Contractor
We used the 30 ma trip limit. Ours was a 2-pole version as the system size was small and we were backfeeding a single pole breaker. It has been operating for about 3 months without any issues.
 

BillK-AZ

Senior Member
Location
Mesa Arizona
WARNING

I sent an inquiry to Schneider Electric about the possible use of the Multi 9TM GFP Ground Fault Protectors and the capability to backfeed these.

I just received the following answer:

Thank you for your recent internet inquiry.

Unfortunately the GFP's can not be backfed in a PV/Solar application. We do not have a device for your application that will give you GF protection and be able to be back fed.


The lack of backfeedable breakers with GFP is a major barrier to installing PV systems when the main service OCPD must have GFP, typically with 1000A+ on 480V 3PH services. The UL standard for circuit breakers makes no provision for testing a GFP breaker for backfeed.
 

G._S._Ohm

Senior Member
Location
DC area
Thanks, I didn't realize that 2 sec was too long and would be considered continuous. My understanding is that continuous means 3 hours or longer.
J.S.
The trip coil and many other things have an I^2 T curve and if you exceed either the I^2 rating or the T rating the device will be damaged.

Fuse wire has a small thermal inertia and can reach melting temperature pretty quickly when overloaded. A chunk of Romex at rated current might take 15 minutes to stabilize at a final temperature.

For insulation, each 9C or 10C increase in temperature halves the life.
 

PWDickerson

Senior Member
Location
Clinton, WA
Occupation
Solar Contractor
The GFP is in line with the inverter breaker, which is backfed, but the GFP is not backfed. The current flows through the GFP in the direction that was intended by the manufacturer. The Inverter is connected to the line side of the GFP, and the backfeed breaker is connected to the load side.
 
Status
Not open for further replies.
Top