Adding additional kw without a permit

[Jim Burrow]

I am of the same opinion, and I will stick to that opinion until the Joules come home from the pasture.
I do believe in following the building code and using UL approved equipment when feasible.

Me too. That is why I'm going to have a licensed electrician tie my system into my power line from the meter to my controller and onto my circuit breaker panel. - although I'm quite capable of doing it myself, I don't know the NEC code. - lets hope when the joules come home from the pasture, they will be friendly.

 

[Hedges]

The trick would be to go to 3 orientations. For me, east to catch the morning sun, North which is the best main orientation and west to make the most of the afternoon light. An online very smart friend said his triple orientation system was east, west and flat.

If you don't give a flying and in my case with my bootleg solar systems, asses the risk of getting caught is remote in the extreme and the penalties if I do are completely inconsequential and what I have to gain ( save) is worth the risks and you have done your homework,

I also asked him for his insider info on what he knew about the power co and the way it worked. How many inspectors did they have looking at this? What monitoring did the power co do though it's smart meters and how would they tell though analouge meters? What would the penalty be if he was discovered? would he be dismissed, Fined or given a stern talking to? How much would it potentially save him and what would the savings mean to his family?

Turns out they have no inspectors at all looking at this, there is no going through the smart meter data let alone looking at analogue metering other than for very High

I did my homework and looked into the meters being used. To my amazement they don't measure Amps fed back, only voltage. This means I can pump a shipload of power back and they can't even tell! I posted this on a local forum and had a couple of people come back and said they did the self same thing,

I suggest you do your own homework and see what applies if you are prepared to give it a go. Check what monitoring they do, see if anyone has ever been Busted for too much feedback and if anyone is even on the job looking for it.

Here, If I'm caught, I'm pretty confident they can't even cut my power off UNLESS I have been in debt and not made ANY payments for 4 months.

Our rules include erroneous measurements.
If a meter is broken, after being fixed the utility can see what your usage is and bill for 3 months in arrears.
If customer had tampered with or bypassed the meter they can go back 12 months.

So when a stalled gas meter was found and fixed (I tried heating the pool, which required operation 24/7 and raised temperature 1 degree per day so they came around trying to find the apparent gas leak), I kept usage low during the measurement period.

Similar when my easel with a couple panels blew over and cracked gauge on the new one. At first I didn't notice the zero bill because credits had been issued that covered all usage for a while. When I did realize no to reading change I contacted utility to report it. They replied they couldn't confirm my identity or something like that. I had tried to get it fixed in summer but they finally did it in winter. I used electric heat in the subsequent billing period and had minimal back billing.


To qualify for a particular size PV system, it is now historical usage or estimated based on new construction. 15 years ago it was up to 200% of usage (rules of rebate program.) I simply ran pool pump for a month straight to establish the consumption level I wanted.

Multiple angles:
To the first order, I figure geometry gives area exposed to sun. 90 degree angle is square root of 2; two arrays have 1.4x projected area of one, or 0.7x the area of both facing same direction. Acute 60 degree angle is 1.0x
Off-angle light has more reflection, light that doesn't enter panel and cell. Haven't tried to quantify.
Morning and evening the light is filtered by longer path through air.
And then there is summer/winter tilt.
Perhaps one array for winter Noon, one for summer 9:00 or 10:00 AM, one for summer 3:00 or 4:00 PM.

When I first permitted and installed my system but didn't sign up with utility, didn't take long for them to send a letter saying it appeared I was operating a solar generator and had to get their permission. I assume that was visual, during meter reading or other work. I rigged up a motion sensor and relay so when they came to view the meter it would be shut off, not backfeeding. But they became insistent, said I would be disconnected. I was trying to pick date of joining net metering program anniversary when "true-up" of bill occurs, but maybe doesn't matter. Ended up with about March 1st.

The electronic meter installed logged Total and Peak, was read visually, and Off-Peak calculated. So what happened hour by hour didn't matter.
One time I got a $2000 bill, which said I had drawn an amount of power equal to 400A continuous Noon to 6:00 PM 5 days per week, and returned almost that much during off-peak hours. After testing, utility said meter was working correctly. I pointed out it was impossible for the 2/0 connection and 200A breaker to have done that. Total consumption similar to previous month. I suggested they estimate peak vs. off-peak to be same ratio as previous month, which they did. This had been a time of solar storms, perhaps corruption of bits in memory. Otherwise appeared correct for a decade.

Later the infamous Smart Meters were installed. This just record power consumed/produced and are read remotely at intervals. Perhaps they can store snapshots so everybody's 3:00 PM reading is captured even though data transfer is not all simultaneous? With these meters they could tell if someone's export was excessive or a 1 hour, 15 minute, or whatever interval used. The data is at the central office, but whether flagged depends on whether software is written to note anything other than billable amounts.

Only voltage not amps? Pretty sure these are watt-hour meters. volts and amps, with a possible phase shift between them, cause the motor to turn one way or the other. Electronic ones have a better chance of implementing correct or incorrect measurement given various power factor, distorted or phase-shifted currents. Industrial customers are explicitly charged for non-unity power factor. Residential are not, but if software uses digitized readings at too low a sample rate, incorrect calculations, slow analog sensors, could easily have measurement error. Could charge for export same as import, whether intentionally or not.

 

[GXMnow]

My net energy agreement which is under the NEM 2.0 rules with So Cal Edison says I can output a peak current of 16 amps, and a maximum of 900 KWHs total in a month. I am not even close to those numbers. The 16 amps would be my solar power max peak going straight out to the grid with none being used in the house. And my best production days just hit 31 KWHs so I only produce a maximum of 930 KWH's so unless I use no power at all in my home, I won't hit that much export. Before I added the battery, my export topped out at about 13 KWH's during the day, but then I had to buy back all of it and more overnight. With the battery, I am hardly exporting at all, and self consuming most of my power. In these winter months without using any A/C, I export only 400 watts or so from 10 am to 9 pm. That is about 4.4 KWH a day. That works out to 132 KWH's, way less than the 900 I am allowed. I buy that power back and a bit more at the lowest overnight rate. My "Smart Meter" seems to be taking a reading each hour. It may be more, but that is how it appears on the bill. It needs decent time resolution for the "Time of Use" rates. The display shows the energy in and the energy out as separate numbers. Those numbers only go up.

 

[Hedges]

I think your 900 kWh/month limit is 8 hours per day at 16A limit.
The amps limit is fairly restrictive, 3680W.
Sounds like 20A backfeed breaker on 100A panel run at 80% of rating.

Considering 5 hours effective sun, either two orientations of panels (doesn't fit micro inverters well unless smaller panels) or extra inverters and battery storage deferring export until evening peak times seem like the ways maximize what is pushed out.

Another is to pad your consumption during peak PV hours, such as storing heat or cold for later use.

 

[fafrd]

Micros need to have a grid connection to sync to so they would have to be wired to a subpanel or service panel like your charger. Micros cost about $100 but you would need a $500 Envoy plus some current transformers to control the micros for Non Export. That is the most cost effective solution for a few panels.
It is complicated finding a standalone inverter which would run directly off solar panels and at the same time be able to run a charging station. My attempts at that kept having ground fault issues.
My EVA rate for super off peak charging is $0.15 per kWh and I charge two Teslas. I don't think the E-6 rate gets that low but you know your usage and the time periods better than I. Apparently you are grandfathered in E-6 like I am in EVA. I think the new EV rate and other TOU rates have slightly less favorable time periods and we only have until 2022 and then we roll into those less favorable time periods.
The main thing I watch to make sure I don't go over my NEM agreement is my maximum export in kW at any one time. That in my opinion is the the only way PG&E is going to be able to observe that you may have added a lot of panels. I am fortunate that I was able to add some additional panels facing west which do not add to my peak generation of my main NEM system, but do extend the time and thus give me more kWhrs later in the day which means more kWhrs stored on the grid. My hybrid inverter does give me data that I can use to tweak my system. You can see in my signature that I have added 3 kW to a 5.7kW system but because of orientation I have never gone over the 1 kWh max because the main system tapers off before the west facing panels start hitting their peak.
Do you have some loads you can shift. The more loads you can shift out of peak period puts more money in the bank from generation that you can use off peak. In summer I get $0.54 for every kWh I generate and I store that on the grid and use it up at a 3 to 1 or better rate at $0.15 per kWh. In other words Every kWh I bank at peak, I can use 3 kWhs at off peak with zero financial cost.
Take a look at these two EV rates and compare them to what you are paying now.

Electric Vehicles

With the number of electric vehicles in California growing every day, PG&E is committed to low electric vehicle rates, and home charging for electric cars in California.

www.pge.com

Exceeding peak output levels would certainly be easy for a utility to detect, but so would increased average or annual output (unless you self-consume to compensate).

if you replace panels with panels at 150% of the ‘old’ rated power rating, even with the same old inverters clipping during the hottest part of the day, daily/monthly/yearly output is going to increase by close to 50%.

Even if you self-consume an additional 50% to offset, unless you scale that added self-consumption to track seasonal changes in added generation, it’d be pretty visible to the utility...

I’m in much the same situation as the OP (including what PG&E said to me when I inquired) and I think either going with a full-blown hybrid inverter + storage solution or GTIL + storage solution to self-consume additional DC generation is the only way to add generation capacity that is not visible to the grid/utility (and does not violate your agreement with them).

 

[Hedges]

Exceeding peak output levels would certainly be easy for a utility to detect, but so would increased average or annual output (unless you self-consume to compensate).

if you replace panels with panels at 150% of the ‘old’ rated power rating, even with the same old inverters clipping during the hottest part of the day, daily/monthly/yearly output is going to increase by close to 50%.

Even if you self-consume an additional 50% to offset, unless you scale that added self-consumption to track seasonal changes in added generation, it’d be pretty visible to the utility...

I’m in much the same situation as the OP (including what PG&E said to me when I inquired) and I think either going with a full-blown hybrid inverter + storage solution or GTIL + storage solution to self-consume additional DC generation is the only way to add generation capacity that is not visible to the grid/utility (and does not violate your agreement with them).

"Visible" and "Being observed in the sea of Big Data" are two different things.
Easy to flag if someone writes software to look for it.
But for now you're just as gnat on the back of an elephant. More important things for the company to focus limited resources on, like implementing rules to boost profitability that you can try to circumvent.

If a second orientation of panel was wired in parallel with each (micro, in this case) inverter, it will extend hours of production better than just over-paneling with a single orientation; more will be useful output rather than clipping.

 

[Ampster]

Visible" and "Being observed in the sea of Big Data" are two different things.

I agree, if you are reasonably within the 1 kW limit the risk is minimal. I added 3kW of PV but because of orientation, they never produce more than 1 kW of power until after the power of my approved GT system has started decreasing for the day.

 

[GXMnow]

Yes, my system is back feeding through a 20 amp breaker, 16 amps in 80%. So that is my peak limit which is probably being monitored and will set a flag if I exceed it. Then the 900 KWH a month is likely set to throw a flag if I exceed that. But anything I do below those numbers, I bet no one would even look at it. I shifted out most of my back feed now, and just back feed a tiny bit now during the peak rate window. No one has said anything, and my usage graphs are totally different with the battery bank operating. I did actually exceed my 16 amp back feed into my breaker panel for a bit once, but since there was some load in my home, I don't think the meter saw more than 16 amps. That was when the solar and the battery were each on their own 20 amp breakers. Solar was pushing about 12 amps, and I hit about 10 amps out of the battery. So that was 22 amps into my panel. The house draw is about 6 amps when the A/C is not running, so I was right at 16 amps going out.

Once I am able to have the PLC control the inverter, I can do so much better. I have a current sensing switch. I will let the inverter crank out up to 10 amps more when either the A/C or my shop air compressor is running. And I will also be able to have it adjust the charge current. If there is a large draw in the house, reduce charge rate. The other huge change in my system was getting the palm trees cut down. My early evening output has more than doubled. Last year to this year, my daily production went from 10 kwh to 15 kwh. I still got some shadowing after 3 pm in winter, but it's not bad now. 3 days ago, I still made 17 kwh in winter, and the worst panel that get's the shade in the evening still made 80% of the best panel on the upper roof. The more common 15 kwh is when a few clouds are going by during the day.

 

[Micmuse]

I'm stringing the panels finally. I decided to keep (9) of my 16 year old 190 watt Sanyo panels and install (8) new Longi 320 watt. My old Sunny Boy SWR 2500 inverter will not be replaced and these dissimilar panels will be all wired in one series except the Sanyo panels will be in groups of three wired parallel within that series. Attached is a diagram showing what I've come up with. I figured the Sanyos have degraded from the original spec of 3.47 Amps but I don't know how much. So I'm raising them to about 10.41 by wiring groups of three in parallel which hopefully puts the Imp close to the Longis Imp of 9.43.

Any input on this would be appreciated since I'm new to this kind of thing. I'm making unauthorized changes to my system but I don't think it will cause any problems with PG&E since my inverter will clip anything over 2500 watt (my original CEC rating is 2350 and 2000 has been typical, in summer). I should be able to get higher and longer watts all year. I think the inverter will now be about 150% loaded at most. Would this risk damaging the inverter?

This thread has been super informative for me. Thanks for sharing your knowledge.

 

[Ampster]

Any input on this would be appreciated since I'm new to this kind of thing. I'm making unauthorized changes to my system but I don't think it will cause any problems with PG&E since my inverter will clip anything over 2500 watt (my original CEC rating is 2350 and 2000 has been typical, in summer). I should be able to get higher and longer watts all year. I think the inverter will now be about 150% loaded at most. Would this risk damaging the inverter?

I can't help with the details of the MPPT channels of the SMA but I do know with my Solaredge and Skybox inverters there are limits on the spec sheets of those products that can provide the answer. I know the Solaredge can handle 153% of capacity in Watts but it also has string limits that are expressed in volts and Amps. The voltage and Amp limits of the Skybox are the governing numbers. Hope that helps conceptually?
As far as my PTO is concerned the description of the Inverter is what stood out to me the last time I looked. I have 5.7kW of panels going into a 3.8 kW solaredge so I am maxed out anyway at a DC to AC ratio of 1.5 to 1.

 

[GXMnow]

I had to open your image in photos to make it readable. That looks like the Sanyo panels are 52.8 volts each. You will now have 3 of those in series for 158.4 volts, along with 8 of the Longi panels at 33.9 volts each for another 271.2 volts, for a total of 429.6 volts. That is at the maximum power point. The VOC is going to be a chunk higher, and I think you are going to be in trouble. The open circuit voltage limit of your inverter is 600 volts. Going by the Vmp to Voc ratio of most panels, this puts your no load voltage at about 530 volts at 25C. If you have a cold but sunny day, you very well could hit 600 volts. Most people recommend a 20% margin from the 25C Voc. So you are cutting this close.

You input current looks safe as the maximum is 13 amps on that model, but it also lists "recommended max PV power" at 3000 watts. What you are proposing comes out to more like 4,000 watts. That is a 60% over panel. The manual I found did not give a solid over panel limit. It only said 600 volts max, 13 amps max, and recommended PV under 3,000 watts. Anything past that is at your own risk.

 

[Hedges]

Without checking the more important Voc on a freezing day that GXMnow mentioned ...

You're connecting the 9 Sanyo panels 3p3s. More than 2 in parallel should have individual fuses.
I would connect them 3s3p, and use 3 fuses for the three strings.

Voltage - you sketch listed Vmp. You need to get Voc for the panels, and adjust them according to temperature coefficient and your location's record cold. (or multiply Voc by 1.2 if you want a safe easy estimate.)
Voc adjusted for cold must remain below 600V.

My own arrays are designs for < 600 Voc. They are 480 Voc on a typical day and 380 Vmp.

Sunny Boy can handle extra current available from PV, but not extra voltage.
You may have to leave off two of the Longi to be safe.

Better, wire 8 Longi in series. Wire 14 Sanyo in series. Connect those two strings in parallel.
That's 5200W, should keep your 2500W inverter at max output all day long (add a fan!)
Better yet, orient the two brands of panels differently, on a different roof face or just tilted. That'll keep the 2500W output going even more hours.

Especially aimed at afternoon sun, if you have new peak rate times late afternoon to evening.

(Later you can upgrade to a 3300W inverter, within your allowed limits.)

 

[Micmuse]

Ampster says 150% oversized panels works on his rig and I read the same about other inverters but found nothing specific to that with the SMA SWR 2500. There is an article on the SMA site that shows a graph of a way oversized system: 15kw panels on 9kw inverter.

7 Reasons Why You Should Oversize Your PV Array - Sunny. SMA Corporate Blog

Oversizing a PV array, also referred to as undersizing a PV inverter, involves installing a PV array with a rated...

www.sma-sunny.com

It also says their inverters can handle excess current but not volts, as Hedges also mentions.

It's odd that the table GXMnow posted shows 3000 watts max and also says max DC power 2710 watts. I don't know what to make of that. If the volts aren't going to exceed 600 then it can only be current that pushes the power over, but their inverter is apparently tolerant of that and it seems that max watts should be higher.

Looks like I have to take the Voc seriously. The Longi 320 watt spec at 40.9 Voc and the Sanyo 190 watt at a wopping 67.5 Voc. My proposed string would be:
8 x 40.9 = 327.2 Longi
3 x 67.5 = 202.5 Sanyo
TOTAL 529.7 x 1.2 = 635.6 V temperature corrected.

So I should drop one of the Longi panels and I'm fine. Maybe I could leave it up there and switch it so it only becomes part of the array when the sun is low in winter. Or would it be in summer when the panels are hotter?

 

[Micmuse]

Also I don't understand why it would matter how I wire the 9 Sanyo panels, 3s3p or 3p3s. Doesn't the voltage and current come out the same? Is it to make it easier to wire? Also why do I need fuses and what should be their rating?

I don't have a south or west option at this point. I only have room on the roof for 10 sanyo panels and 8 Longi. The idea of having a string of each in parallel doesn't seem to work because the current of one string would be so different from the other and this is a single string inverter.
Thanks,
MM

 

[Hedges]

Also I don't understand why it would matter how I wire the 9 Sanyo panels, 3s3p or 3p3s. Doesn't the voltage and current come out the same? Is it to make it easier to wire? Also why do I need fuses and what should be their rating?

Voltage and current do come out the same. My reason for saying this is to minimize the number of fuses needed.

Also, 3p3s requires six 3-Y connectors but 3s3p only requires two.

If you had just two strings (e.g. 3s2p), a short in one string would let current from the other one string backfeed into it, maximum 1 x Isc in any wire. No fuse needed.
If you have 3 strings (e.g. 3s3p), a short in one string would get 2 x Isc backfeeding from the other two strings. So each of the 3 strings requires a fuse, 3 fuses total.

If you do 3p3s, then any panel which shorts gets the current from two panels in its 3p group, so all three need fuses. And you have three separate 3p groups, each of which needs a fuse, so you need 3 x 3 = 9 fuses. That's why I say to do 3s3p not 3p3s.

Fuse size is given on back of panel. According to this spec sheet:

Sanyo HIP-190BA19 (190W) Solar Panel

"Series Fuse Rating 15A"

With 3 in parallel so max backfeed current is 2 x Isc, there hardly seems to be a need, but code calls for fuses. If more like 5 panels in parallel so 5 x Isc backfeed, the current usually exceeds fuse rating. In your case of "Isc 3.75A" it would seem like 2 x Isc and maybe 3 x Isc would not be excessive, so seems like fuses aren't actually needed. (If trying to get it approved by an inspector, would have to see if NEC wording addresses available current and not just "3 or more strings")

 

[Hedges]

I don't have a south or west option at this point. I only have room on the roof for 10 sanyo panels and 8 Longi. The idea of having a string of each in parallel doesn't seem to work because the current of one string would be so different from the other and this is a single string inverter.

If Sanyo and Longi are in series, Imp should be similar. So your ideal of 3 Sanyo in parallel to match one Longi is reasonable. (note that their voltages don't have to match.)

What I'm suggesting is Sanyo and Longi in parallel. Vmp should be similar. (note that current doesn't have to match.)

Given your area for 10 sanyo and 8 longi, consider shuffling those numbers to get more Sanyo and fewer Longi, until "n" sanyo in series is the same voltage as "m" Longi. Then check that voltage against the MPPT range of your inverter.

 

[Hedges]

Looks like I have to take the Voc seriously. The Longi 320 watt spec at 40.9 Voc and the Sanyo 190 watt at a wopping 67.5 Voc. My proposed string would be:
8 x 40.9 = 327.2 Longi
3 x 67.5 = 202.5 Sanyo
TOTAL 529.7 x 1.2 = 635.6 V temperature corrected.

So I should drop one of the Longi panels and I'm fine. Maybe I could leave it up there and switch it so it only becomes part of the array when the sun is low in winter. Or would it be in summer when the panels are hotter?

Don't count on changing wires manually. Design it so it is protected from damage on record cold days.
(voltage is lower on hot days.)

You can sharpen your pencil. Using 1.2 multiple is quite safe. Get the data sheet for each of your panels, find "Temperature coefficient of Voc". Look up record cold temperature for your location. How many degrees below nominal 25 degrees C? Compute accurate correction factor for each model panel.

My 600Voc max inverters have various combinations of parallel strings of 24x "12V" 120W Astropower, 12x "24V" 165W Sharp, 8x "36V" 327W Sunpower. They are all about 480 Voc on a typical day and 380 Vmp while inverters are operating.

 

[Micmuse]

Thanks, Hedges, for the clarifications. It totally makes sense now about the wiring. I have the 8 Longi and 9 Sanyo on the roof now. Only the Sanyo panels are wired (in series). I am going to use the 5 sanyo I removed for an off grid project in the Santa Cruz mountains. I'm driving there shortly today.

I'll do the math and find the max Voc for my particular panels at given temperatures. But when would a situation come up where the inverter is subject to the open circuit voltage? It seems either the panels are hooked up to the inverter or not, and when they are not that would be the only time there is an open circuit. If the circuit is broken due to a bad panel or a disconnect then there would be no voltage to the inverter. I must be missing something.

I think if I had a panel as a temporary part of the array it would only be a seasonal thing. It never goes below 50 in SF for serveral months of the year. So if it looks like I'm only going to reach 600 volts when it's 32 degrees it would be worth the trouble to switch it twice a year (especially since I already have the panel). Worst case senario would be I forget to switch it and my old inverter is junk. Maybe I could put some kind of fuse in the circuit just in case.

 

[Hedges]

You inverter is grid connected. Any time the grid is down, inverter will draw zero current from the panels and they will rise to Voc.
Even connected to the grid, when you first connect to the panels, or when the sun falls on them in the morning, they will rise to Voc, applying that to input capacitors of inverter. The inverter will wake up but draw no current until after it has monitored the grid for 5 minutes.

SWR 2500 has a value stored in non-volatile memory showing the highest PV voltage ever experienced. They are watching you. Not that its under warranty any more, but you don't want to kill it.

Of course I prefer putting in a bigger inverter, take advantage of the extra 1kW you're allowed to export.
Apparently SolArk supports a programmed maximum export? If you wanted to spend the money, you could have battery backed critical loads as well. Or newer Sunny Boy with Secure Power, a PV-direct AC outlet if the grid is down.

 

[Micmuse]

I found a Vmax calculator here:

Solar panel maximum voltage calculator - How to work out the max voltage

To get the maximum solar panel voltage you should expect from your solar panel, use our solar panel maximum voltage calculator.

gold-coast-solar-power-solutions.com.au

Plugging in values from my spec sheets I get a total of 564.22 for the whole array. I used -2 instead of 2 for record cold here just to be safe. As soon as my extra MC3 cables come for the old panels I'll fire it up. If all goes well with pg&e this year I'll think about upgrading the inverter.