diy solar

diy solar

NEM 2.0 / PG&E... Enphase Grid tied system + off-grid crazy plan.

hex4def6

New Member
Joined
Nov 8, 2022
Messages
28
So I'm barely in the finishing stage of my 9.45kW AC / 8.7kW DC Enphase IQ8+ setup, and I already want to start expanding...

I've got 20 panels up on the roof, should have the remaining 10 up soon.

This is what the setup looks like currently:

1709594244519.png

Here's the wrinkle: NEM 2.0 closed back in April last year (this plan was submitted before then, and got in). Any additions would kick me to NEM 3.0, which sucks (except adding batteries -- that's ok).

Question #1: My understanding is adding <10% / 1kW is doable while staying on NEM 2.0. Anyone actually done this? Did it require an entire new plan set / AHJ / PGE inspection?

Adding batteries
is the obvious other idea (and can be done under NEM 2.0), although I've got plenty of additional roof to add panels on... Assuming I want to add panels, off-grid sounds like the only practical / PGE-legit way of doing it.

If I do add off-grid, that sounds like I then need to have dedicated batteries. It also sounds like I'd have to figure out some way of either semi-permanently segregating some loads to that off-grid panel (EV / hot water maybe?), or figure out some sort of ATS system.

On to my crazy plan:

1709593516427.png


What the heck is going on here?

All the stuff in red is "new".
Basically:
I'd add as many new solar panels as I feel like, plus an off-grid AIO, plus a bunch of cheap LFP cells / EG4 server batteries, etc. This off-grid system would charge up the huge cheap non-UL batteries until they're fully charged. At this point, the ATS would be triggered, and they'd start dumping into the new "off grid" sub panel + Encharge 5P battery until it's charged.
At that point, the ATS goes back into the normal position (IQ battery 5P --> Rest of house), and that discharges, while the Huge DIY battery starts charging from the off grid solar again.

This is janky, I know. ;)

Limitations I can see:

1. Lots of cycles of the IQ Battery
2. Lots of wastage going from Offgrid PV -> 48 V battery -> 120 V -> IQ Battery DC, then when switched over, IQ Battery DC -> AC -> House loads.
3. Cost. Off grid PV (6kW / $2000?) + AIO (6kW / $1300?) + Cheap DIY battery (cheap 15x EVE LF304 , <$2000 15kWish) + IQ Battery ($3300)+ Extra IQ Gateway ($700). Total $10k or so.

Question #2: Do off-grid battery installs in california require them to be UL 9540?

I'm open to other suggestions that are less janky that don't lose my NEM 2.0 status...
 
I didn't read your whole scheme so don't have design feedback.

Question #1: My understanding is adding <10% / 1kW is doable while staying on NEM 2.0. Anyone actually done this? Did it require an entire new plan set / AHJ / PGE inspection?

Yes you'll need to go through the whole process again. This will seriously hamper your ROI, to say the least.

Question #2: Do off-grid battery installs in california require them to be UL 9540?

Yes. The difference in my mind is that on-grid battery installs are easy for them to detect/gatekeep UL9540 on. Off grid you can fly under the radar somewhat more easily. You'll be detected if:

- you pick an AIO that is actually a hybrid and backfeeds grid. these are out there. HAHAHA. It's not that hard to detect backfeed with modern smart meter. If you ever backfeed (could be inadvertent due to self-consume) after dark, and you don't have a UL9540 battery in an updated interconnect agreement, you're playing with fire.
- your AHJ/PG&E starts flying drones that send photos back to the cloud, and the cloud is programmed with address-by-address data of how many W-DC of solar panels are supposed to be on the roof.
- PG&E detects big change in your usage pattern.

Of these, the first bullet is probably the only real risk. The second bullet requires a major cultural change in what's acceptable for the man to do.

Limitations I can see:
1. Lots of cycles of the IQ Battery
Who cares? LFP will last a really long time. You're also unlikely to cycle a battery more than once per day.

One way you can get more credits with UL9540 battery is that you can force discharge during peak (and peak is lucrative in the summer) to bank more credits. That is my plan to offset the pain of UL9540 cost and to get more value out of my fixed set of NEM2 panels.

Along similar lines, I'd say you're extremely wasting the value of the IQ5P's UL9540 markup by using it off grid, actually I don't really understand why it's in your new line diagram. Just import whatever the first AC coupling capable hybrid is from overseas if you care about activating your microinverters when grid is down. Or simpler yet, forget about using your microinverters when grid is down, and just buy the cheapest AIO to drive that EV charger.

If you go to HPWH (500W) there's not that much value from hot water diversion load.
 
Yes. The difference in my mind is that on-grid battery installs are easy for them to detect/gatekeep UL9540 on. Off grid you can fly under the radar somewhat more easily. You'll be detected if:

- you pick an AIO that is actually a hybrid and backfeeds grid. these are out there. HAHAHA. It's not that hard to detect backfeed with modern smart meter. If you ever backfeed (could be inadvertent due to self-consume) after dark, and you don't have a UL9540 battery in an updated interconnect agreement, you're playing with fire.

Damn. That is disappointing. I was hoping UL 9540 was a poco thing, not a AHJ thing.

- your AHJ/PG&E starts flying drones that send photos back to the cloud, and the cloud is programmed with address-by-address data of how many W-DC of solar panels are supposed to be on the roof.

I'm hoping to stay legit-ish. My whole crazy scheme is one that relies on adding panels that are not grid-connected (therefore not violating PG&E agreement). So, AHJ would need to know / sign-off on it, but PG&E could kick rocks.
One way you can get more credits with UL9540 battery is that you can force discharge during peak (and peak is lucrative in the summer) to bank more credits. That is my plan to offset the pain of UL9540 cost and to get more value out of my fixed set of NEM2 panels.

This is probably the most practical / boring solution. Just add the enphase IQ battery, call it a day.

Along similar lines, I'd say you're extremely wasting the value of the IQ5P's UL9540 markup by using it off grid, actually I don't really understand why it's in your new line diagram. Just import whatever the first AC coupling capable hybrid is from overseas if you care about activating your microinverters when grid is down. Or simpler yet, forget about using your microinverters when grid is down, and just buy the cheapest AIO to drive that EV charger.

I actually don't care about grid-down really. In fact, without the system controller (+$1000), I can't do grid down.
The idea is (was) to use the legit UL 9540 IQ Battery as a "bucket brigade" bridge between between the off-grid solar + batteries, and the on-grid system.

Effectively, the off-grid PV + batteries are a "generator" being used to charge the IQ Battery. The grid is never connected directly to the off-grid stuff.
 
How about this:
1) put enphase between main panel and grid.
2) put aio between enphase and main panel.
3) set aio to zero export.

All the enphse can be exported under nem2.
Aio acts like off-grid with grid backup.
Any extraneous export from aio is within 1kw allowable excess.

Only downside is you loose the enphase during grid down. That can be solved with a transfer switch to move enphase to gen input of aio.
 
How about this:
1) put enphase between main panel and grid.
2) put aio between enphase and main panel.
3) set aio to zero export.

All the enphse can be exported under nem2.
Aio acts like off-grid with grid backup.
Any extraneous export from aio is within 1kw allowable excess.

Only downside is you loose the enphase during grid down. That can be solved with a transfer switch to move enphase to gen input of aio.

That's not a bad idea. It's not "legit" from PG&E's POV*, but practically they wouldn't know about it, especially if I go with a legit ESS on the grid-tie side.

*(I'm assuming they wouldn't agree to a zero-export extra system if they knew about it, right?)
 
Effectively, the off-grid PV + batteries are a "generator" being used to charge the IQ Battery. The grid is never connected directly to the off-grid stuff.

The off-grid PV would not be eligible for shifting across seasons via NEM2 credits (IE for winter heating/EV charging). You can only shift the energy generated by your NEM2 panels. (Not sure if there are stronger regulations than that). You can probably size the off-grid to supply house + EV loads during the summer and bank your NEM2 energy for the winter.

You might compare a UL9540 system used in off grid mode with new DC-coupled solar, serving as much house load as you can (challenge here is that inverters will need to be sized for max demand). And grid tie inverters exporting all production (with no timeshifting). Vs putting the UL9540 system on your existing grid-tie inverters and doing timeshifting. You can potentially pirate on some DC coupled solar after getting PTO.

EDIT: You could also do 2x UL9540, one grid tie and the other off-grid. This would allow you to time shift the exports and install off-grid solar, but this is likely terrible ROI.

You could do 1x UL9540 1x non-UL9540 AIO (even in zero export) for a lot cheaper, no need to size AIO for surge, and little detection probability but that will not be code compliant and it will violate interconnection agreement (albeit in a hard to detect way). This is probably what DIYrich suggested above.

FWIW, PG&E can potentially scrape AHJ permit databases in the future to see who is modifying their systems. Those records are open where I am. Dunno if they actually do that.

Enphase charges for access to some APIs.
 
So I'm barely in the finishing stage of my 9.45kW AC / 8.7kW DC Enphase IQ8+ setup, and I already want to start expanding...

I've got 20 panels up on the roof, should have the remaining 10 up soon.

This is what the setup looks like currently:

View attachment 200071

Here's the wrinkle: NEM 2.0 closed back in April last year (this plan was submitted before then, and got in). Any additions would kick me to NEM 3.0, which sucks (except adding batteries -- that's ok).

Question #1: My understanding is adding <10% / 1kW is doable while staying on NEM 2.0. Anyone actually done this? Did it require an entire new plan set / AHJ / PGE inspection?

Adding batteries
is the obvious other idea (and can be done under NEM 2.0), although I've got plenty of additional roof to add panels on... Assuming I want to add panels, off-grid sounds like the only practical / PGE-legit way of doing it.

If I do add off-grid, that sounds like I then need to have dedicated batteries. It also sounds like I'd have to figure out some way of either semi-permanently segregating some loads to that off-grid panel (EV / hot water maybe?), or figure out some sort of ATS system.

On to my crazy plan:

View attachment 200070


What the heck is going on here?

All the stuff in red is "new".
Basically:
I'd add as many new solar panels as I feel like, plus an off-grid AIO, plus a bunch of cheap LFP cells / EG4 server batteries, etc. This off-grid system would charge up the huge cheap non-UL batteries until they're fully charged. At this point, the ATS would be triggered, and they'd start dumping into the new "off grid" sub panel + Encharge 5P battery until it's charged.
At that point, the ATS goes back into the normal position (IQ battery 5P --> Rest of house), and that discharges, while the Huge DIY battery starts charging from the off grid solar again.

This is janky, I know. ;)

Limitations I can see:

1. Lots of cycles of the IQ Battery
2. Lots of wastage going from Offgrid PV -> 48 V battery -> 120 V -> IQ Battery DC, then when switched over, IQ Battery DC -> AC -> House loads.
3. Cost. Off grid PV (6kW / $2000?) + AIO (6kW / $1300?) + Cheap DIY battery (cheap 15x EVE LF304 , <$2000 15kWish) + IQ Battery ($3300)+ Extra IQ Gateway ($700). Total $10k or so.

Question #2: Do off-grid battery installs in california require them to be UL 9540?

I'm open to other suggestions that are less janky that don't lose my NEM 2.0 status...
I recommend looking into the newly defined paths that CALSSA blazed with the 3 IOU which specify how you could have this system size increased as long as the new generation never exceeded the generation of the old system.

My understanding is that this in its most interesting version requires a new PCS certification, which isn't really available yet. However it might be complete by the time you are ready to go forward.

Ideally you add new battery at the same time so you don't lose the generated solar if you don't use it.

This additional PV generation in a non-export capacity is pretty bleeding edge right now, but the details will be fleshed out over the next year or so. I wouldnt try to hack together a hardware solution as you are showing. I would follow an established path, though this may be less DIY.
 
My understanding is that this in its most interesting version requires a new PCS certification, which isn't really available yet. However it might be complete by the time you are ready to go forward.
Which inverter companies actually get PCS certs? I’ve been manually looking and only saw a handful. And it’s super inefficient for me as a DIY scrub to manually look.
 
Which inverter companies actually get PCS certs? I’ve been manually looking and only saw a handful. And it’s super inefficient for me as a DIY scrub to manually look.
All the large manufacturers will go after this market. I expect that Tesla, Enphase, SMA and Solar Edge will all have solutions to address this California Market. I would reach out to those manufacturers about non export PCS certifications for the NEM 2.0 expansion market in California.
 
All the large manufacturers will go after this market. I expect that Tesla, Enphase, SMA and Solar Edge will all have solutions to address this California Market. I would reach out to those manufacturers about non export PCS certifications for the NEM 2.0 expansion market in California.
Got it. I only looked at the popular models on this forum (18kpv, MidNite, XW), and Enphase. I found a PCS for XW but not the PCS I wanted (and I wasn’t super interested in that inverter).

Unfortunately it feels like PCS is more beholden to what the installers need/negotiate with the POCO and PUC, and it may not lead to a usable path for DIY

Any insights on UL9540 DC ESS?
 
Got it. I only looked at the popular models on this forum (18kpv, MidNite, XW), and Enphase. I found a PCS for XW but not the PCS I wanted (and I wasn’t super interested in that inverter).

Unfortunately it feels like PCS is more beholden to what the installers need/negotiate with the POCO and PUC, and it may not lead to a usable path for DIY

Any insights on UL9540 DC ESS?

Over the next few months/years I expect that more manufacturers will release a battery solution that can be paired with PV and be set to not export to the grid. Right now it might be a DIY solution is your only path.

As to the question about UL 9540 DC ESS that is a pretty deep topic. I do have a lot of knowledge about this standard but what do you want to know?

From my perspective, it seems that many of the popular DIY batteries have a questionable listing with regards to UL 9540 and UL9540A. The fire protection community is really concerned about the proliferation of ESS and that many of those systems might not be as safe as they could be or at least not properly listed with all components.

The standard itself is one of the most popular technical committees with UL and so there is a lot of work and changes at the moment. There is a lot of movement to have 9540A test report be rolled into 9540 so that the whole thing has more oversight.

Currently, any test lab can perform the 9540A testing but rightfully some FPO do not trust and approve test reports from smaller or unknown labs. UL 9540 is at least a standard that OSHA accredits, so theoretically they could give approval to test labs that are doing this test according to the standard and to accepted industry norms.
 
As to the question about UL 9540 DC ESS that is a pretty deep topic. I do have a lot of knowledge about this standard but what do you want to know?

My interest in 9540 DC ESS is that it would allow a DC ESS listed battery to legally be reused for multiple inverters, so long as the inverters respect whatever instructions were provided to achieve the DC ESS.

However, I don't know what the manufacturer interest is in this, or what the timelines are.

It seems that some inverter companies like MidNite are probably just waiting for UL9540 DC ESS products (or pushing for them) rather than getting a combined UL9540 AC ESS listing with a battery.

There are several different business strategies here I guess.
 
I recommend looking into the newly defined paths that CALSSA blazed with the 3 IOU which specify how you could have this system size increased as long as the new generation never exceeded the generation of the old system.

My understanding is that this in its most interesting version requires a new PCS certification, which isn't really available yet. However it might be complete by the time you are ready to go forward.

Ideally you add new battery at the same time so you don't lose the generated solar if you don't use it.

This additional PV generation in a non-export capacity is pretty bleeding edge right now, but the details will be fleshed out over the next year or so. I wouldnt try to hack together a hardware solution as you are showing. I would follow an established path, though this may be less DIY.

Interesting. Do you have any links to this that I could read up on?

I did notice in the Enphase Installer Toolkit options for PCS, although I didn't really explore them.

I assume in essence this limits peak power delivery / total kwh delivered in a 24 hr window? If the kW AC rating of my install was PTO'd at 8.7kW, this would effectively allow that value to be set as a limiter, regardless of the actual inverter / panel ratings?

That would be super cool; if I had 20kW AC actual, I could output 8.7kW to the grid and divert the rest to house loads / battery storage, and then switch to self-consumption at night.

They also have various customized grid profiles, which seem to achieve similar things. For example, "CA Rule21 201902 VV VW FW 14.25 kW Export Limit (1.2.1)", and "CA Rule21 201902 VV VW FW, No Export"
 
My interest in 9540 DC ESS is that it would allow a DC ESS listed battery to legally be reused for multiple inverters, so long as the inverters respect whatever instructions were provided to achieve the DC ESS.

However, I don't know what the manufacturer interest is in this, or what the timelines are.

It seems that some inverter companies like MidNite are probably just waiting for UL9540 DC ESS products (or pushing for them) rather than getting a combined UL9540 AC ESS listing with a battery.

There are several different business strategies here I guess.
DC ESS are currently a defined item that you can test as of edition 3 of UL 9540. Technically a battery listed to this would not be compliant with the IRC, but most AHJ aren't that sophisticated. The current code calls out an earlier version of the standard without DC ESS as an option.

All that is required for a DC ESS would be an energy storage function and an energy storage protective function. Here is the paraphrase of the language:

Instructions for a DC ESS shall include specifications for the equipment performing
electric power conditioning or conversion function that is intended to be connected for supplying the DC
ESS, including if it is part of the DC load that the DC ESS is supplying. The
instructions can include the specific equipment or systems to be used or can provide sufficient parameters
required to safely connect to the DC ESS such as, but not limited to electrical ratings, comm.
requirements, environment requirements and short circuit tolerance as applicable to the DC ESS.

It sounds like the path is there for manufacturers who want to follow it, though I am not familiar with any that are.
 
Interesting. Do you have any links to this that I could read up on?

I did notice in the Enphase Installer Toolkit options for PCS, although I didn't really explore them.

I assume in essence this limits peak power delivery / total kwh delivered in a 24 hr window? If the kW AC rating of my install was PTO'd at 8.7kW, this would effectively allow that value to be set as a limiter, regardless of the actual inverter / panel ratings?

That would be super cool; if I had 20kW AC actual, I could output 8.7kW to the grid and divert the rest to house loads / battery storage, and then switch to self-consumption at night.

They also have various customized grid profiles, which seem to achieve similar things. For example, "CA Rule21 201902 VV VW FW 14.25 kW Export Limit (1.2.1)", and "CA Rule21 201902 VV VW FW, No Export"
The webinar was recorded but those are for members of CALSSA. I am sure this information will trickle out to others as the tech is available. You could reach out to CALSSA and see if they might share the info.

On a high level, there are a few paths to keep your NEM 2.0 agreement from what I understand:

1. New PCS control inverter that the combined output never exceeds the allowed backfeed of the original NEM 2.0 system. Any strategy from charging battery, to demand mitigation to solar curtailment may be employed.
2. Use an NGOM to ensure the same as #1
3. Install additional DC capacity on a system that was already inverter-limited (upping the DC/AC ratio with an existing inverter)
4. Dedicated non export relay (I think this is similar to what Solar Edge did in Hawaii but not sure)
 
Upping the DC/AC ratio alone sounds kind of suspicious. There are a couple NEM/NEMSTORAGE rules I remember from skimming PG&E Rule21 that this contradicts.
 
Yeah, I'm pretty sure they consider DC Inverter increases or CEC AC increases to be NEM deal breakers. Hence the "off-grid" addition idea.

What I hadn't realized (based on the NGOM comment), (which brought up this article: https://pv-magazine-usa.com/2020/06...fornia-means-relays-meters-and-complex-costs/) was that you aren't supposed to earn NEM credits from "dirty" energy (eg, charging at night and backfeeding from battery during the day).

Hmm....

I'm now thinking I should try and keep them separated. I get cheap panels with a UL listed hybrid inverter with a small battery for stability.
I then dedicate the off-grid system to doing something boring like heating water / mini-split / EV charging in a sub panel with an ATS. When the sun is up, those can merrily consume solar, and then then the ATS can switch over to grid at night.

This is rather than the initial idea of trying dump the off-grid charged battery into the grid-tied side. This avoids a lot of complication, and given how cheap the solar panel side of the equation is vs battery storage, it's ok if solar gets 'wasted' this way.

This is all based on the assumption that only the AHJ would need to sign off on this, and that PG&E wouldn't / couldn't object.... :D
 
Don’t let your off grid stuff ever parallel with PG&E and my understanding is it is fine.

Momentary parallel is technically not OK either but I rather doubt they will police that.
 
Don’t let your off grid stuff ever parallel with PG&E and my understanding is it is fine.

Momentary parallel is technically not OK either but I rather doubt they will police that.

I'm guessing that would come down to the ATS-- "make before break" vs "break before make". It would be a bit of a nuisance to have the latter, but not insurmountable.
 
I'm guessing that would come down to the ATS-- "make before break" vs "break before make". It would be a bit of a nuisance to have the latter, but not insurmountable.
There are a few tradeoffs between using a SBU style off grid or hybrid inverter, vs an external ATS. A good AIO or hybrid would sync before transfer

The parallel operation/PCS approach will be easier to design and more robust, but also $$$$. To some extent all of those are good for installers (with the relative importance depending on how cynical you are).
 
Back
Top