What good is XW grid-tie with the 120% backfeed rule??

aj164 · February 2010

With a battery-backed Xantrex XW system on the grid, the output of the system typically goes to a sub-panel that you want backed up in case the grid goes down.

The breakers in the XW distribution box are 60A, but the maximum size breaker I can put on the grid 200A service panel is 40A (due to NEC 120% backfeed rule).

Something doesn't seem right about this. At night, with the grid on and inverter off, it means I'm actually limited to 40A at the sub-panel. (I don't want the inverter supplementing any part of the load from batteries as long as the grid is available.)

Is there another way to do this? Meter load-side tap?

Thanks,

AJ

drees · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

Unless you do a tap on the meter to main breaker line, the only other option is to downsize your main breaker at least 20A. But I don't think there's much available unless you go down to 150A.

aj164 · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

That is a good point. I hadn't thought of that. I'll have to see if 150A is enough total service for the home.

drees wrote: »

Unless you do a tap on the meter to main breaker line, the only other option is to downsize your main breaker at least 20A. But I don't think there's much available unless you go down to 150A.

drees · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

Yeah, obviously if you can get a 180A breaker that would be best. But my quick search didn't turn up much besides 125A, 150A and 200A main breakers.

I think you're supposed to run some sort of load analysis when downsizing the main breaker, but I'm not an Electrician - just done some research for my own PV projects. I'm not sure how it would really matter in the long run - everything is still protected and many old houses already run off 100A at the most even with a few high-power electric appliances like stove, dryer and AC anyway, unless they've been upgraded.

mike95490 · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

drees wrote: »

many old houses already run off 100A at the most even with a few high-power electric appliances like stove, dryer and AC anyway, unless they've been upgraded.

Wow 100A service? When I bought my house, it had 2, 30A fuses, 1 circuit for lights and 1 circuit for outlets. Never blew one either, but maybe there was a penny under the fuse!!

BB. · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

mike90045 wrote: »

Wow 100A service? When I bought my house, it had 2, 30A fuses, 1 circuit for lights and 1 circuit for outlets. Never blew one either, but maybe there was a penny under the fuse!!

That was the same as my parent's home... A separate 4 fuse sub-panel for the 15amp circuits... All was fine until my Dad got a big Sears radial arm saw, a small arc welder (boy was the fun with graphite battery electrodes and melting anything we could)--and things were never the same again.

Fuses got a bit bigger (20-30 amp)--and did you know you could purchase adapters to put larger capacity fuses in "standard" screw-in fuse holders?

-Bill

GreenerPower · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

BB. wrote: »

... boy was the fun with graphite battery electrodes

Arc lamp ? been there as a kid, but using 120VAC in series with a portable electric stove, melted a test pyrex tube and put a hole on mom's kitchen table .... trying to diy movie projector ... but seeing about 10s of 8mm movie :-)

By the way, 240VAC, 40A is still alot of power, I would use larger wire gauge for future expansion but staying with 40A breaker for now unless there're lots of high power loads on that sub-panel.

GP

dsp3930 · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

I ran into the same issue. ( Xantrex XW6048 ) I downsized my main breaker to 150A and purhcased a new old stock Cutler-Hammer main breaker for $80 shipped. (Retail is $300+ new) That now leaves 60A for the Xantrex Unit with 30A of room to spare.

Calcs:
200A busbar * 1.2 == 240A max combined
240A - 150A = 90A (max backfeed allowed)
90A - 60A (Xantrex dedicated breaker) = 30A left over

Note: They do make 175A main breakers, though they are harder to find and rather expensive compared to the 150A/200A models.

newenergy · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

I downsized someone from 200A to 175A on one project. It didn't involve an XW, just a fairly large array with two grid tied inverters. The breaker was close to $200.

Does your XW have a power distribution panel?

The 120% rule says it's based on the first overcurrent device on the output of the inverter. Does the breaker in the distribution panel count? If so, doesn't that mean you have to use 60A for the calculation anyway, even if you put a 40A breaker in the main?

RCinFLA · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

I think the issue is with the primary panel consuming the 20% backfeed plus the full grid input breaker limit so there is 120% current flow in the primary panel.

It would seem that if enough of the loads are moved from the main panel to the subpanel, things should work out okay.

Dropping the main breaker down to ensure the main panel stays within it rating is a bit silly when I look at a GE panel and the only difference between a 150 amp panel and a 200 amp panel is a longer breaker bus bar to allow for more breakers. The bus is not thicker and connection to a main breaker is the same.

mike95490 · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

RCinFLA wrote: »

Dropping the main breaker down to ensure the main panel stays within it rating is a bit silly when I look at a GE panel and the only difference between a 150 amp panel and a 200 amp panel is a longer breaker bus bar to allow for more breakers. The bus is not thicker and connection to a main breaker is the same.

You some kind of troublemaker ?

RCinFLA · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

mike90045 wrote: »

You some kind of troublemaker ?

I understand the NEC is trying to make rules that are idiot proof.

mike95490 · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

RCinFLA wrote: »

I understand the NEC is trying to make rules that are idiot proof.

Great. Even Darwin knows that just makes smarter idiots.

russ · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

When İ was in the Navy we called that 'making it sailor proof'.

Next to impossible!

sub3marathonman · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

aj164 wrote: »

With a battery-backed Xantrex XW system on the grid, the output of the system typically goes to a sub-panel that you want backed up in case the grid goes down.

The breakers in the XW distribution box are 60A, but the maximum size breaker I can put on the grid 200A service panel is 40A (due to NEC 120% backfeed rule).

Something doesn't seem right about this. At night, with the grid on and inverter off, it means I'm actually limited to 40A at the sub-panel. (I don't want the inverter supplementing any part of the load from batteries as long as the grid is available.)

Is there another way to do this? Meter load-side tap?

Thanks,

AJ

I have a very similar setup, with the subpanel and the XW6048. However, I did things a bit different.

The subpanel has a 100 amp manual transfer switch for the main breaker. To this transfer switch the main grid 200 amp panel is connected. When the grid is on, power comes in to the main panel, and goes over to the subpanel. The XW6048 meanwhile is supplying power back to the 200 amp main panel, limited to the 40 amp breaker. However, there shouldn't be any objection to this, since the XW6048 isn't going to be putting out more than 40 amps at 240 volts. The grid is on, 200 amps can come from the power company, and another 40 amps from the XW6048.

Now, when the grid goes off, the XW6048 switches the output to Line 2 (the other output line). There is no power to the 200 amp panel from either the electric company or the XW6048. Then you switch the manual transfer switch on the subpanel. Now the XW6048 has the 60 amp breaker on it, and the 100 amp breaker on the subpanel. So the XW6048 is now set to output the full capacity to the subpanel.

I did things this way before Solar Guppy added some input. If I remember correctly, he stated that yes, you could use the 60 amp breaker on the main 200 amp panel, even though it is supposed to be limited to 40 amp by the NEC 120% rule, because the XW6048 itself is rated correctly so that the 60 amp breaker can be used. I will try to find his posting, or else he may add his input again. His help has truly been invaluable to my project, and is appreciated every sunny day.

crewzer · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

The 120% rule says it's based on the first overcurrent device on the output of the inverter.

I respectfully disagree. Here's the wording (in part) for 2008 NEC 690.64(B)(2):

Bus or Conductor Rating. The sum of the ampere ratings of overcurrent devices in circuits supplying power to a busbar or conductor shall not exceed 120 percent of the rating of the busbar or conductor.

HTH,
Jim / crewzer

crewzer · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

Then you switch the manual transfer switch on the subpanel. Now the XW6048 has the 60 amp breaker on it, and the 100 amp breaker on the subpanel. So the XW6048 is now set to output the full capacity to the subpanel.

I'm not sure that I understand the logic here. The XW6048's full continuous rated output is 6,000 VA, or 240 Vac x 25 Aac.

Jim / crewzer

Brock · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

That is what I was wondering about as well. The unit calls for a 60 amp breaker but could never push back more than 25 amps. So do you calculate it on 25 amps or 60 amps, that is the real question...

BB. · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

On other fly in the ointment to watch out for...

If I remember correctly, Solar Guppy also said that can run up to 70% (don't remember exact percentage) of the rated wattage out one 120 VAC leg for off-grid mode:

6,000 watts * 1/120 VAC * 0.70 = 35 amps

And if you look at the surge current support:

L-N: 105 Arms (7 sec)
L-L: 40 Arms (20 sec)

For 20 second support--I would design the branch circuit for:

40 Amps * 1.25 = 50 amps minimum

And, if the output current is energy limited for Grid Tied Circuits, then you really need to look at minimum voltage of 105 volts (or whatever the VAC low voltage is set to (and the XW Inverter regulates at):

6,000 watts * 1/105 VAC * 0.70 = 40 amps

Add NEC 1.25x rule:

40 amps * 1.25 = 50 amp minimum circuit/breaker

So, if I got any of the above correct--you will have to allow for >25 Amp for the off-grid panel (at least 50 amps).

And, since the sell mode range is 214-260 vac with 3 volt error bars... The maximum 6,000 kW current would be:

6,000 W / 210 VAC = 29 amps
29 amps * 1.25 = 36.25 rated circuit minimum

So, you have to allow for > 25 amp for the GT section of the circuit.

I can see the NEC requirement as not really caring the differences between load vs generation current capabilities of the GT inverter--but instead just looking at the Branch Circuit rating and planning for worst case operation based on branch circuit rating and not the XW's limitations. Somebody could later install a second or larger XW on the same circuit and over power the host CB panel.

If I was doing the wiring inside a dedicated device (such as UL approved UPS--I could probably account for load vs generation differences and get away with it). But not for building wiring which does not usually allow for such subtleties.

-Bill

newenergy · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

2008 NEC 690.64(B)(2)

There's a whole paragraph there and the last sentence is:

In systems with panelboards connected in series, the rating of the first overcurrent device directly connected to the output of a utility interactive inverter(s) shall be used in the calculations for all busbars and conductors.

Maybe I'm interpreting something wrong, but that's what it says.

crewzer · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

My take on the NEC and on Wiles' publications is that the OCPD on an inverter's "AC output" should be based on rated voltage and current. My XW6048 Operation Guide (page A-2, Electrical Specifications) specs the 6048's AC Output Voltage to be 120 Vac +/-3% (L-n) and 240 Vac +/-3% L-L.

The AC Output Current specs are 37.5 A L-n and 25.0 A L-L. Applying the 125% rule, I suppose it's possible that the AC output breaker could be 50 A (37.5 A x 125% = 46.9 A; rounded up) or 35 A (25.0 A x 125% = 31.25 A; rounded up). 35 A is a standard value, but it might be hard to find at Home Depot or Lowes.

It's of course useful for the XW6048 to have its surge- and imbalance capabilities. However, it seems to me that the design phase should try to balance the planned loads across the two phases, especially since an upstream generator may not be so forgiving when called on to power downstream AC loads and the built-in charger.

One of my assignments is to replace my Brand "O" hardware with some XW equipment. Since I have a "48 V" array and a "24 V" battery bank, I'll probably go with an XW4024. It's L-L AC Output current spec is 16.7 Aac, so I'll likely install a two-pole breaker rated at 25 A (16.7 A x 125% = 20.9 A; rounded up to 25 A).

I'll try to remember to look into this a bit further the next time I'm at HQ (Burnaby) or Eng'g (Livermore).

HTH,
Jim / crewzer

crewzer · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

newenergy wrote: »

2008 NEC 690.64(B)(2)

There's a whole paragraph there and the last sentence is:

In systems with panelboards connected in series, the rating of the first overcurrent device directly connected to the output of a utility interactive inverter(s) shall be used in the calculations for all busbars and conductors.

Maybe I'm interpreting something wrong, but that's what it says.

I think I now understand...

For a battery-based grid-interactive inverter, there are two separate AC "output" connections with which to deal. One output is the connection labeled "AC Output" (or equivalent), which is the AC output intended to be connected via a sub-panel to downstream AC loads.

The other "output" is labeled "AC Input". This is a "double-duty" connection for the external power source, typically the grid via the main power panel. The connection allows the inverter to "buy" power from the grid (for charging batteries and "pass thru" powering of downstream AC loads). This same connection is also an AC output that allows the inverter to sell power to the grid via the main power panel.

For an application with a 200 A main power panel with a 200 A main breaker, it's my view that the NEC's 120% rule allows for a maximum 40 A breaker to be connected between the main power panel and the XW6048's "AC Input" if the Grid Support "Sell" feature is enabled (the XW can't sell to the grid if "Sell" is disabled). Allowing for the 125% multiplier, a circuit breaker of this size allows for the XW6048 to sell full rated current (25.0 A) to the grid.

A 40 A breaker also allows for the XW6048 to "buy" up to 32 A (32 A x 125% = 40 A) from the grid via the main power panel when operating in "pass-thru" mode for powering downstream AC loads and/or charging the batteries.

Assuming a 100% Max Chg Rate setting and the default Absorb Voltage setting of 57.6 V, I estimate the XW6048's charger's maximum AC load to be (100 Adc x 57.6 Vdc) / ((89.4% charger efficiency at max rate) x (0.98 PF) x 240 Vac) = 27.4 Aac.

Assuming using a 40 A breaker between the main power panel and the XW's "AC Input" connection, I'd probably set the "AC Menu" AC1 Breaker value at 32 Aac. This the maximum current the XW would draw from the main power panel, and, allowing for the 125% requirement, would be compatible with the 40 A breaker.

It's possible that the combined current load of downstream loads and the XW's charger could exceed the 32 A setting. In this case, the XW would automatically reduce charge current to the batteries to limit total AC current load fro the main power panel to ~32 Aac. As downstream AC loads were dropped, the XW would automatically increase battery charging current as required until the total AC input current reached the 32 A limit.

HTH,
Jim / crewzer

tallgirl · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

Crewzer,

The worse scenario is that a GT inverter is feeding the buss bars it's maximum power, while the utility is doing likewise. That would allow an over current condition in the buss bars themselves.

In the 2008 ROP there was a section about backfeed breakers being on the end of the buss bars opposite the utility connection for the panel. It's my understanding that this was the old rule (I'd have to get out my '05 and read it ...), but that the potential for having more current on the buss bars than their rating won out over the wishful thinking that the amps were going to behave themselves.

What's needed is for panel vendors to oversize the buss bars in smaller panels so people are not forced into installing a 400A panel just because they have a net-zero house. Line side taps are a great alternative, but they just demonstrate that the industry is lacking the hardware needed to do a proper install. The main service disconnect should be THE main service disconnect -- not one of many.

crewzer · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

Julie!

My understanding is that the "old" 120% rule applied to residential applications because there was little/no concern about the busbars being fully loaded. Commercial apps were limited to 100%.

The 2008 NEC was changed to also allow 120% for commercial apps if, as you point out, the PV breakers were mounted at the opposite end of the busbars from the main breakers.

Wiles discussed this change in this Code Corner article.

I hope you're well!

Regards,
Jim / crewzer

tallgirl · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

crewzer wrote: »

Julie!

My understanding is that the "old" 120% rule applied to residential applications because there was little/no concern about the busbars being fully loaded. Commercial apps were limited to 100%.

The 2008 NEC was changed to also allow 120% for commercial apps if, as you point out, the PV breakers were mounted at the opposite end of the busbars from the main breakers.

Wiles discussed this change in this Code Corner article.

I hope you're well!

Regards,
Jim / crewzer

The 2008 ROP rule was much more lenient. Somewhere between the time it was supposed to be done, and it was published, the rules were changed.

A better discussion is in the materials from the NEC, not that I seem to be able to find them.

And yes, I'm well. A bit cold (frozen), but none the worse for wear!

crewzer · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

Yeah, it's NEC 690.64(B)(2) that was changed:

Specifically, 2005 NEC 690.64(B)(2) limited the sum of the breakers supplying power to the busbar rating. A 120% "exception" was made for a "dwelling unit".

The 2008 NEC 690.64(B)(2) just allows for 120%. If the sum of the breakers is more than 100% of the busbar(s) rating, 690.64(B)(7) requires that the inverter connection(s) be made at the opposite end(s) of the busbar(s) from the main input CB.

Did you make it to this week's RE World Conference and Expo in Austin?

Regards,
Jim / crewzer

tallgirl · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

crewzer wrote: »

Yeah, it's NEC 690.64(B)(2) that was changed:

No, I mean they changed the rules between the '08 draft and the '08 final version. I was reviewing the '08 draft back when I was still doing disaster work in New Orleans, because we were hoping to get someone to donate gear on account of we didn't have power everywhere we were working. This was before I ever knew OutBack even existed and spent much more of my time in Chapter 2. The next thing I knew, several very useful rules changed. If you have the ROP draft, look at 690.64 (B) (6) (b).

For example, in the '08 Draft you could backfeed 100% of the buss bar rating =and= still have the main disconnect be rated at 100% of the buss bar rating --

b. Where connected at the opposite (farthest) end of
the busbar from the feeder or service, the sum of
ampere rating of the backfed PV supply
overcurrent/disconnect device(s) shall not exceed
the rating of the busbar or conductor. The following
permanent plaque shall be installed at the PV supply
connection or circuit breaker location:

You tell me that doesn't say I can backfeed 100A into a 100A buss that also has a 100A utility disconnect. 690.64 (B) (2) was completely different and unrelated -- all of the OCPD requirements were in 690.64 (B) (6) (a) and (b). It's like there was a moment of sanity, followed by going back to the '05 rules. While I agree with the '05 and '08 rules, they are also excessively paranoid.

690.64 (B) (7) applied uniformly would lead to all conductors having to be oversized to handle the sum of the supply OCPD current rating and the inverter output current. Consider the conductors between a GVFX 3648 and the main panel. I'd argue that if you have a 60A OCPD at the panel, and 30A of available backfeed current from the inverter, that the maximum continuous fault current is 90A, and the correct conductor is #3 for 75C rated terminations. AND you have to use #3 between the bypass breaker and inverter input lugs.

Did you make it to this week's RE World Conference and Expo in Austin?

Nope. My schedule has been trashed for the past few weeks.

newenergy · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

100% of busbar would be sweet. 120% really often leads to problems, just bumping it to 140% or so would never overload a busbar if you put the breaker on the very opposite end and would eliminate almost all the extra problems.

Otoh, speaking selfishly, we are doing pretty good service upgrade business.

tallgirl · February 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

newenergy wrote: »

100% of busbar would be sweet. 120% really often leads to problems, just bumping it to 140% or so would never overload a busbar if you put the breaker on the very opposite end and would eliminate almost all the extra problems.

Otoh, speaking selfishly, we are doing pretty good service upgrade business.

I have a lead on a system where the client has a circa 1975 Sylvania panel. They've done a bit of work in the panel themselves and I'm wondering how they are going to feel when the electrician gives them a quote to upgrade their overloaded 125A service to 250A. Assuming 250A is even big enough!

aj164 · March 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

If it is acceptable to eliminate the 120% busbar backfeed problem by simply turning off the "grid support", then why couldn't we satisfy the 120% rule by setting the "maximum backfeed current" in those same Inverter Settings?

Is this setting enough of an "overcurrent protection device" for NEC? The existence and implementation of this feature makes me think Xantrex already thought about this problem, and they hoped this would be the solution. ?

Cheers,
Adam

crewzer wrote: »

For an application with a 200 A main power panel with a 200 A main breaker, it's my view that the NEC's 120% rule allows for a maximum 40 A breaker to be connected between the main power panel and the XW6048's "AC Input" if the Grid Support "Sell" feature is enabled (the XW can't sell to the grid if "Sell" is disabled). Allowing for the 125% multiplier, a circuit breaker of this size allows for the XW6048 to sell full rated current (25.0 A) to the grid.

drees · March 2010

Re: What good is XW grid-tie with the 120% backfeed rule??

aj164 wrote: »

Is this setting enough of an "overcurrent protection device" for NEC? The existence and implementation of this feature makes me think Xantrex already thought about this problem, and they hoped this would be the solution. ?

I doubt it as your inverter isn't qualified as an "overcurrent protection device".

What good is XW grid-tie with the 120% backfeed rule??

Comments

Categories