DIY Self-Install in San Diego - 7.84 kW for < $1.30 / W

[sensij]

A new multi pole switch rated for high voltage DC, is pretty expensive. I picked up some 3 pole used ones on Eb+y,
some with fuses, some not. Do you need 4 poles for lack of a combiner box? Bruce Roe

There are two strings, combined at the inverter into a single input. I might be able to use a 2 pole disconnect with twin wire ferrule on the input, one pole for the combined DC+ and the other for the DC-, and bump up the output wire to 8 awg to handle the 30 A combined current instead of the 15 A per-circuit current.

The more I think about it, since he didn't object to the cabinet underneath the inverter/disconnect, only the height, there is a place near the garage door where the built-in drops to counter height that would probably be an OK place to put it. I'd have to put a junction box in to splice onto the existing DC wires, but at least the AC would only be shortened, and more importantly, the GEC for the AC side could remain unbroken, since it is getting shortened too. Basically, where the baby seat is in this picture. That portion of the wall is more exposed and likely to be warmer than where it is now, but at least there would be better airflow.

garage - plan B.JPG

[sensij]

Some more reading... I think the inspector was wanting to apply 404.8 to the location of the disconnect, which states that the center of the switch in its highest position can not be more than 6 ft 7 in above the floor or platform. However, there is an exception to that which allows that switch to be located higher and accessible by portable means if it is next to the equipment being supplied. So, maybe a small reason for hope.

Edit: Here is another take on it, suggesting that if an AC disconnect is added at the inverter, and also the closest readily-accessible point, it might be ok. That might still not be less expensive or require less work than just moving the inverter, though.

There are a few exceptions, but, in general, Section 404.8(A) requires switches and circuit breakers to be readily accessible and limits the height of their operating handle to no more than 6 feet 7 inches above the platform or working surface from which they are accessed. However, Sections 690.14(D) and 705.70 permit inverters to be installed in non-readily accessible locations. This is a common scenario in residential PV systems where microinverters may be mounted behind modules, or in commercial PV systems where string inverters may be mounted on a parapet wall on a low-sloped roof. In these cases, DC and AC disconnecting means are required either within sight of - defined in Article 100 as "visible and not more than 50 feet away from" - or integrated into the inverter. Since they are equipment disconnects, they must be grouped with the inverter. Depending on the particular string inverter, a combination of internal (such as a built-in DC disconnect, which may or may not also disconnect the AC conductors) and/or external (most commonly, an external AC disconnect switch) is used. With microinverters, the plug connectors on the microinverters must be listed and identified to meet this disconnect requirement (per the Exception to Section 690.17 and the requirements of 690.33). An additional disconnect for the inverter output circuit of roof-mounted (not readily accessible) inverters is required. It must meet the requirements of Sections 690.14(C)(1), 690.17(1) and 705.22, which apply to disconnecting means for PV systems and require them to be readily accessible: installed either on the outside of a building or inside at a readily accessible point, and as close as possible to the entry point of the system

[cebury]

Some more reading... I think the inspector was wanting to apply 404.8 to the location of the disconnect, which states that the center of the switch in its highest position can not be more than 6 ft 7 in above the floor or platform. However, there is an exception to that which allows that switch to be located higher and accessible by portable means if it is next to the equipment being supplied. So, maybe a small reason for hope.

Edit: Here is another take on it, suggesting that if an AC disconnect is added at the inverter, and also the closest readily-accessible point, it might be ok. That might still not be less expensive or require less work than just moving the inverter, though.

Higher than the 6ft 7 was what my installer was balking at. He didn't know for sure, but taking the most cautious approach, considered it from the top most part of the inverter ... considering it was one unit as disconnect?. Good luck.

[J.P.M.]

Some more reading... I think the inspector was wanting to apply 404.8 to the location of the disconnect, which states that the center of the switch in its highest position can not be more than 6 ft 7 in above the floor or platform. However, there is an exception to that which allows that switch to be located higher and accessible by portable means if it is next to the equipment being supplied. So, maybe a small reason for hope.

Edit: Here is another take on it, suggesting that if an AC disconnect is added at the inverter, and also the closest readily-accessible point, it might be ok. That might still not be less expensive or require less work than just moving the inverter, though.

Absolutely NOMB, but I suspect that while you probably don't or maybe won't need to access your inverter quite as much as I access mine, I'd not be surprised if you will still probably be getting near it for a lot of reasons more than most. FWIW, after living w/mine for 3.5 yrs., I'm glad the output screen is close to eye level and access is easy. Moving yours might be something you might be glad you did in the future. Just sayin'.

[sensij]

After review, they concluded that it would need to be moved, or another disconnect at ground level be provided, so I think we've all gotten to the same place on this. There was enough discussion about it at the office that they are going to make it the subject of next week's training meeting

They were clear that only the switch itself needs to comply with the height requirement, not the entire inverter.

I will be glad to move it, if for no other reason than it means I have an excuse to pull out the two conduit pieces I liked least and replace them to get to the new location. It drives my wife nuts when I un-do work I'm not proud of, so I had left it, but now it is out of my control.

[SunEagle]

After review, they concluded that it would need to be moved, or another disconnect at ground level be provided, so I think we've all gotten to the same place on this. There was enough discussion about it at the office that they are going to make it the subject of next week's training meeting

They were clear that only the switch itself needs to comply with the height requirement, not the entire inverter.

I will be glad to move it, if for no other reason than it means I have an excuse to pull out the two conduit pieces I liked least and replace them to get to the new location. It drives my wife nuts when I un-do work I'm not proud of, so I had left it, but now it is out of my control.

Too bad about you needing to relocate the inverter, but in retrospect you might find that the original location was a PIA if you had to do any trouble shooting.

[bcroe]

Dang, I got them from pvlabels.com, and yeah, it looks like their filtering system isn't right. Looks like the more specific requirements are new for 2014... selecting 2014 NEC should probably have returned only red labels, but it doesn't. Thanks! I'll need to double-check the other labels now, too.

I have created and modified some signs, using a cheap on line bumper sticker service. Choose your color. Note the
bottom line of this set of signs I picked up surplus, part of a custom bumper sticker. Bruce Roe

PVsign.JPG

[sensij]

I just passed re-inspection, but barely. Pictures are below. The new inverter location is in a warmer part of the garage, but it is still relatively out of the way.

The inspector wanted to tag me for using a wire nut in the junction box on the ground wire going to the roof. I tried to explain that for the SolarEdge system, both the DC GEC and AC GEC can be satisfied with the single GEC from the inverter to the service panel, but he kept insisting that the EGC from the inverter to the roof also had to meet the unbroken (or irreversibly spliced, etc) requirement of the GEC. He eventually agreed to call a senior inspector / supervisor back at the office, who agreed that what I had done was ok. Close call, it could have gone either way. Re-pulling 85 ft of green wire would not have been fun, although I could probably have found another way to make the connection in that box, or one somewhat bigger.




j-box.JPG

garage - new.JPG

inverter.JPG

[SunEagle]

Good to hear your system was cleared by the AHJ.

Since the inverter had to be moved, If it was me I would have put it lower on the wall below the door track. But I am sure you had your reason to put it there.

[sensij]

Got PTO from SDG&E yesterday. The system is live... PVOutput data is here:

https://pvoutput.org/intraday.jsp?id...07&dt=20170620

The inverter temperature peaked at 143.8 deg today, higher than I was hoping to see. Probably time to start looking at how to get a supplemental fan in there.

I'm using a revenue grade meter (from EKM Metering) for the generated power, and consumption comes from calculating the difference between what is generated and the net that goes though the Poco meter, as transmitted by the Rainforest Automation Eagle. That the consumption yields a relatively stable line throughout the day supports that revenue grade accuracy is really being achieved.

[cebury]

My inverter is reporting 151 right now. https://pvoutput.org/intraday.jsp?id...08&dt=20170620

[sensij]

My inverter is reporting 151 right now. https://pvoutput.org/intraday.jsp?id...08&dt=20170620

Ok, thanks... it is higher than I was used to seeing on my 3kW inverter, but maybe typical for 7.6kW.


On the billing side, I'm very sure that I'm going to be over-producing my consumption and end up with a surplus credit. In that situation, TOU isn't "helpful", in that the multiplier between peak generation and off-peak consumption doesn't accomplish anything. I could switch back onto a tiered plan, but I don't think that gains me anything either. One twist... because non-bypassable charges of $0.02-$0.03 / kWh must be paid on all energy imported from the grid, it is actually to my advantage to charge my EV in the 6pm - sunset time frame, when my array is still producing and I can avoid energy import by self-consuming. Because those "off-peak" prices are only a few cents more than the overnight super off-peak, and the difference in the amount of TOU credit is meaningless between those two periods when I'm going to have a massive banked balance either way, anything that minimizes grid import works in my favor.

I guess that supports why SDG&E wants to extend peak hours to later in the day, but for 5 years, it looks like I'll have a backwards incentive to consume during the time when the grid is already seeing the most demand. The amount of energy available for self-consumption from 6 pm to sunset won't add up to much more than 1 kWh a day for a couple months of the year... but intentionally not charging on a Friday night and waiting until Saturday morning, for example, might make a bigger dent (Edit: even with two EV's, I don't think this adds up to more than $50 / yr)

[J.P.M.]

Got PTO from SDG&E yesterday. The system is live... PVOutput data is here:

https://pvoutput.org/intraday.jsp?id...07&dt=20170620

The inverter temperature peaked at 143.8 deg today, higher than I was hoping to see. Probably time to start looking at how to get a supplemental fan in there.

I'm using a revenue grade meter (from EKM Metering) for the generated power, and consumption comes from calculating the difference between what is generated and the net that goes though the Poco meter, as transmitted by the Rainforest Automation Eagle. That the consumption yields a relatively stable line throughout the day supports that revenue grade accuracy is really being achieved.

FWIW, my inverter temp. peaked today at 122.6 F. @ 1315 P.D.T., w/in ~ 5-10 min. of max. system output. Air inlet to the inverter HX section at that time was 91.4 F. --->>> delta T ~ 31.2 F. or so, air inlet to inverter temp. diff. BTW, a good, and necessary piece of info. for any cooling schemes is the air temp. around the inverter.

The super dooper window fan forced convection system designed by a crack retired HX design slug I know manages to knock about 16 - 20 deg. F. out of the inverter reporting temp. Without the fan, I'd be at ~ 140 F or so for an inverter temp. So, the fan knocks ~~ an additional 1/3 more heat out of the inverter than nat. convection alone. The air going through the HX picks up ~~ 10 F. rise in temp. going through the inverter HX. Temp. performance hasn't changed in 3 1/2 yrs.

More FWIW, est. ave. array temp. at 1310 P.D.T (min. array incidence angle time) was 66.2 C., roof amb. air temp. = 36.7 C, ground level amb. air temp. = 33.1 C. 7 minute ave. wind vel. 2.1 m/sec. WNW. system inst. output 4,176 W.

My inst. max. on the day: 4176/5232 = .798. My best guess is my system output is still ~ +1.4 % over spec. or so.

Your inst. max.: 6,540/8120 = 0.805.

More anecdotal observation that Sunpower performs no better than other quality equipment. Tilt/azimuth/location similar.

[CharlieEscCA]

My inverter is reporting 151 right now. https://pvoutput.org/intraday.jsp?id...08&dt=20170620

Ok, exactly where are you guys finding the inverter temp reported out?

As to non bypassable charges, you may be pleasantly surprised by how SDG&E is calculating that. My first full bill last month showed 6 KWh subject to non bypassable charges ($0.10) yet my consumption meter showed much more imported power for the month (was a net generator overall for the month).

[sensij]

Ok, exactly where are you guys finding the inverter temp reported out?

As to non bypassable charges, you may be pleasantly surprised by how SDG&E is calculating that. My first full bill last month showed 6 KWh subject to non bypassable charges ($0.10) yet my consumption meter showed much more imported power for the month (was a net generator overall for the month).

Interesting on the charges, I'll definitely be looking closely at my NEM-ST bill to adjust my spreadsheets that track these tariffs.

Inverter temp is accessible through PVOutput if you've set up the auto-loader for Solaredge. If not, or if you want to see data from a day that wasn't tracked in PVOutput, you can get at it directly with a call to the technical data API.

https://monitoringapi.solaredge.com/equipment/<site ID>/<inverter serial>/data?api_key=<API key>&startTime=2017-06-20 12:00:00&endTime=2017-06-20 13:00:00