Grid-tie concept question.

[snoobler]

36 285W panels recently acquired from Santan. Stickers intact, so not restricted to off-grid.

Know enough to be dangerous with off-grid and have only a superficial vague understanding of grid-tie at the moment.

South facing roof is messy and won't accommodate all of the panels. Some shading issues too from a roof mounted heat pump pack and other protrusions.

Considering 12kW grid-tie string inverter with up to 53A input (Fromius? locally available on CL).

4 parallel strings:

9 panels on south facing roof
9 panels on south facing roof (yes, same spot, but one will be for where shading will be more of an issue with the other in the clear)
9 panels on east facing roof
9 panels on west facing roof

Max current will be under 40A.

My thinking is that using 9 panels per string sets a very similar voltage for each string with varying current, so they'll play nicely in parallel.

Rough layout, North is up:




The E/W front and back angles are a little different. I'm sure this matters, but I suspect it isn't a huge deal. If it is a huge deal, I could use the other roof over the garage as it's the same as the one in the back.

@Hedges @SolarQueen @??????? plz?

Thanks in advance.

 

[snoobler]

george65

Thanks. I'm a little confused that you think you can get 8-10kWh for each installed kW. Panels in sub-optimal orientation produce less than optimal power based on max irradiance, i.e., every panel that's not at optimal tilt isn't going to produce it's optimal power. If you have your entire array on 2 axis tracking, yes. You can get 10... even close to 12, but you're talking about getting 8-10 hours out of panels you're likely averaging less than 6 on each. You can't add up a bunch of smaller numbers expecting a result greater than their actual sum.

The unit I'm looking at is 9-12kW input, and it's available locally for $1350. I'm using it as a template for this concept. I'm aware of over-paneling. Spec-wise, I could put 15kW+ of panels on my Victron 250/100 SCC even though it's only good for 5800W.

Trust me. If you saw my off-grid system, it would never occur to you to warn me about worrying about perfection... I'm a slap it together and make it work quick kinda guy. There's a reason I haven't posted it in Show and Tell. I'll have to step up my game since things like code compliance are in play.

I'm absolutely certain I'll get the most out of the South facing roof even with some shading issues (will do my best to avoid , so it's getting 50% of the array, 23kWh/day average over the year. East and West will get 10kW each average per day. We get good sun here in AZ.

You didn't say it explicitly. Is this 4 string concept with essentially equal voltages on each string but different currents valid?

 

[snoobler]

Ah. That makes a lot more sense. I no longer think you an insane person.

Fortunately here it's just one standard, NEC 2014, and our local FD doesn't engage in any rooftop operations, so we can get pretty crazy with panel placement. I can't get net metering without applying with the power company, so I have to jump through the hoops. This 10kW system will only supply an average of 65% of my annual usage.

Thanks for confirming the concept and your other input.

 

[SolarQueen]

The Fronius 12k has 2 MPPT inputs. I wish it had 3 for your setup. One input per direction would be ideal. Following along the "good enough" idea, it would certainly work, and being on NEC 2014 (I miss those days), you don't have to worry about Rapid Shut Down. I'd probably put the 2 south strings on one MPPT, maximizing its performance, and then the west and east strings in parallel on the other MPPT, accepting that it will be significantly less than MPPT1. But I could be convinced otherwise.

Optimizers on the east and west strings would greatly improve the output. It would add about $1k to the system, so you'd need to determine if the additional energy is worth the additional money. The south panels can be left without optimizers.

 

[Hedges]

With multiple MPPT, strings of different length can be on each.
On a given MPPT, all strings need to be the same length (or similar Vmp, Voc if different models) but can have different orientations.
If one string is slightly shaded (e.g. I had 9S2P and tested complete shading of a single panel) that isn't a problem. Bypass diodes and Vmp of that configuration is found by MPPT
It is best with multiple strings in parallel to not have one partially shaded to a large extent when the other is not, may lose production of shaded string in that case (but some algorithms better than others.) The IV curve has multiple high points, so some algorithms stop at he highest voltage local peak.

Optimizers - Solar Edge allows 24 panels in series of various orientations. It limits voltage to the inverter and matches current.
Other brands, optimizers should help with partial shading.

Your series strings of slightly different orientations of course push them to same current. To first order, power will be from the current of the lower insolation panel (off-angle), and for the panel with better insolation that current point on its IV curve. So some reduction, but you could use longer/shorter standoffs to make same tilt.

On-grid and permits? Need to see what rules apply. Inverters may be required to have arc-fault. Possibly grid support like California's Rule 21. Possibly module-level shutdown (either a box per panel maybe per 4 panels with string inverter, or micro-inverters.) Are there keep-out rules for roof area? (that's walkways for fireman, and space on both sides of ridge so he can decide where to hack a vent.)

Newer inverters are transformerless, lighter cheaper more efficient. May require fusing both ends of PV strings. May be more susceptible to interference or damage than transformer type.

Some GT inverters can play nice with battery inverters, to add backup later.

I looked at the latest 7.7 kW Sunny Boy, which have 3 MPPT. I wanted to over-panel with multiple directions of my panels but if on separate MPPT would exceed current limits. Looked to me like having two orientations paralleled on an MPPT (or using all MPPT in parallel) was the only way to fit my panels the way I wanted. I got older model 10 kW transformerless with single MPPT instead, now have 5 kW transformer type I'm going to swap in.

I figure with multiple orientations (about 90 degrees between them), could put 20 kW STC of panels, peak current similar to 14 kW of single orientation, on a 10 kW inverter without much clipping. With 5 hours effective sun that would produce around 75 kWh/day average.

 

[SolarQueen]

On-grid and permits? Need to see what rules apply. Inverters may be required to have arc-fault. Possibly grid support like California's Rule 21. Possibly module-level shutdown (either a box per panel maybe per 4 panels with string inverter, or micro-inverters.) Are there keep-out rules for roof area? (that's walkways for fireman, and space on both sides of ridge so he can decide where to hack a vent.)

Newer inverters are transformerless, lighter cheaper more efficient. May require fusing both ends of PV strings. May be more susceptible to interference or damage than transformer type.

On-grid and permits? Need to see what rules apply. Inverters may be required to have arc-fault. Possibly grid support like California's Rule 21. Possibly module-level shutdown (either a box per panel maybe per 4 panels with string inverter, or micro-inverters.) Are there keep-out rules for roof area? (that's walkways for fireman, and space on both sides of ridge so he can decide where to hack a vent.)

Newer inverters are transformerless, lighter cheaper more efficient. May require fusing both ends of PV strings. May be more susceptible to interference or damage than transformer type.

He's on NEC 2014, so doesn't need RSD or arc fault. As long as the sum of the 2 strings doesn't exceed the max string fuse size, and the wire is sized properly, he doesn't need fusing per NEC 690.9(A)(2).

 

[snoobler]

I re-checked. Working with two different county/city codes for two different properties ensures my shite memory lets me down. I got confused.

NEC 2017 applies.

The Fronius 12k has 2 MPPT inputs. I wish it had 3 for your setup. One input per direction would be ideal. Following along the "good enough" idea, it would certainly work, and being on NEC 2014 (I miss those days), you don't have to worry about Rapid Shut Down. I'd probably put the 2 south strings on one MPPT, maximizing its performance, and then the west and east strings in parallel on the other MPPT, accepting that it will be significantly less than MPPT1. But I could be convinced otherwise.

Optimizers on the east and west strings would greatly improve the output. It would add about $1k to the system, so you'd need to determine if the additional energy is worth the additional money. The south panels can be left without optimizers.

Very much stuck on what little I know about off grid shenanigan's, so I'm going to start with that and apply it. I essentially assumed there was 1 MPPT input like on my Victron 250/100. The intent was to have all 4 strings in parallel to achieve the same string voltage. The idea of have two identical strings on the South face was to deal with the fact that I'll get a little shading from the A/C unit, and I'd prefer it just affect half the juicy panels (9) instead of all 18 of them.

With 2 MPPT inputs, I envision 9S2P (south) on 1 and 9S2P (East in parallel with West) on the second.

Is the above practical? Does NEC 2017 prohibit that?

Are optimizers required on an entire string, or can they just be on panels where problems are expected? Do they essentially work by bypassing an underperforming panel? I only paid $70/panel, so another $1K doesn't skeer me.

On-grid and permits? Need to see what rules apply. Inverters may be required to have arc-fault. Possibly grid support like California's Rule 21. Possibly module-level shutdown (either a box per panel maybe per 4 panels with string inverter, or micro-inverters.) Are there keep-out rules for roof area? (that's walkways for fireman, and space on both sides of ridge so he can decide where to hack a vent.)

Newer inverters are transformerless, lighter cheaper more efficient. May require fusing both ends of PV strings. May be more susceptible to interference or damage than transformer type.

Some GT inverters can play nice with battery inverters, to add backup later.

Based on initial research, it's my understanding that for areas in which fire fighters do not use rooftop operations, IFCA allows you to run panels right to the edge with no peak access. That's the case in my area, and I see a LOT of professionally installed panels that run right up to the peak.

AC coupling is in the back of my mind, but our power is absurdly reliable. I've been at this location for over 22 years. The number of times we've been without power for more than an hour is counted on one hand, so it's a secondary consideration. Plus, I have about 15kWh of FLA batteries, a 2kW inverter in the house and 400W of panels that I could use for running the fridges if it really came down to it. I also have 3 gen2 Prii that are great little 900W generators if needed.

Thanks to both of you. I feared my question was premature, and I'm too ignorant at the moment. I was right.

I'll see if I can find a professional for hire in my area to design a system... or maybe I'll get some quotes and equipment lists? Nah... that's dirty.

A 5 year pay back puts me at $8600 budget. I have $6000 left after the panels, so money might be a better spend than my time.

Thanks again.

 

[SolarQueen]

I re-checked. Working with two different county/city codes for two different properties ensures my shite memory lets me down. I got confused.

NEC 2017 applies.

Oooooh, 2017. Now that's a horse of a different color. Bummer. You need rapid shutdown to get inspected, which you need to get on to the grid. But, the good news is that that can fix your shading and 3 direction issues. With Tigo, you can get their optimizer, TS4-A-O ($52 each), for the shaded south and west and east panels, and the shutdown only TS4-A-F ($38 each) for the rest of the south. So you are looking at closer to $2k, but with code compliance and better output.

 

[SolarQueen]

The other alternative with the multiple strings would be to use Multiple Inverters.
I bought a bunch of new 4 Kw inverters that have Dual Trackers. You would then run each string on it's own tracker. Still split the north strings with an east/ west string to give best capacity handling but the input would be a good match.

Don't know about the US but Micro's / optomisers are rarely worth while here. I call it my rule of solar inefficiency is more efficient. Always better to run more panels if you have the room at lower efficiency than to spend the money to make less panels More efficient. More sq inches of panels also means better output on the overcast crappy days which you can't make less panels more efficient.

That's a good idea to do 2 inverters with 2 MPPTs each. Unfortunately, here in the states we have code that requires each panel be able to be remotely shut down in the event of an emergency. Called Rapid Shut Down, stated with 2017 (not enforced until last year). Until they incorporate a chip in the junction box to do it (they are working on it), they have to have device installed on the back of each panel. The cheapest option has just the shut down function, but for only $14 more, you also get optimization. So if he's got the optimizers, he can do one inverter. But I'd be interested in running the numbers to see which option is the best price. Also, solar is MUCH more expensive here than it is down in Australia, so the numbers may be different.

The good news is we aren't required to install flaming DC isolators on our roof like you are, so we've got that going for us.

 

[SolarQueen]

You have to shut down each PANEL? As in individually and you can't just shut down the whole string/ array?
If that's the case, that's Nuts! Is there some rationale behind that?

Yup. Remote shutdown at every panel. It is to protect the firefighters who go on the roof to vent during a fire. Absolutely overkill, between 2014 and 2019 it was just per string. Then the code added $2000 dollars to every install that isn't microinverters or SolarEdge.

I enjoy following along with Australia's solar situation on the Facebook page, Crap Solar. I get great laughs, especially the comments. It's on my bucket list to someday go there, after the world gets normal again (whatever that is).

 

[snoobler]

Crap.

The procured panels have the potential to supply 62% of my annual usage. This is based on a PVWatts simulation of the array and its optimal orientation, so my output will likely be notably less than that due to a lack of roof space and the need to have panels on east and west facings.

I discovered SRP has a solar calculator tool that incorporates your usage data and array and analyzes it on the basis of each plan they offer ($2700 after the fact ).

Due to SRP's shitty lack of fair net metering, the 62% generation is offset by fees resulting in ONLY a 37% reduction in my SRP bill.

Plan busted until the AZ corporation commission gets cleaned out.

So, I have 36 HEAVY 285W panels. I'll probably put a few on the motorhome and a couple on the trailer as supplemental solar for giggles. Might add a second array at HBR as I have a feeling the wife is going to want a few luxuries that may not be accommodated by 6kW of solar.

These panels are voltage compatible with my existing 330W Talesun panels, so I could over-panel the existing 250/100 controller and adopt @Hedges "opportunistic" panel orientation schemes and get a lot more usable out of the existing array.

Might also go the route of those cheap "grid-tie" inverters that plug into sockets - probably just a single 1kW unit.

Thanks to those who contributed, especially @SolarQueen, who provided additional advice via PM.

 

[JoeHam]

Crap.

The procured panels have the potential to supply 62% of my annual usage. This is based on a PVWatts simulation of the array and its optimal orientation, so my output will likely be notably less than that due to a lack of roof space and the need to have panels on east and west facings.

I discovered SRP has a solar calculator tool that incorporates your usage data and array and analyzes it on the basis of each plan they offer ($2700 after the fact ).

Due to SRP's shitty lack of fair net metering, the 62% generation is offset by fees resulting in ONLY a 37% reduction in my SRP bill.

Plan busted until the AZ corporation commission gets cleaned out.

So, I have 36 HEAVY 285W panels. I'll probably put a few on the motorhome and a couple on the trailer as supplemental solar for giggles. Might add a second array at HBR as I have a feeling the wife is going to want a few luxuries that may not be accommodated by 6kW of solar.

These panels are voltage compatible with my existing 330W Talesun panels, so I could over-panel the existing 250/100 controller and adopt @Hedges "opportunistic" panel orientation schemes and get a lot more usable out of the existing array.

Might also go the route of those cheap "grid-tie" inverters that plug into sockets - probably just a single 1kW unit.

Thanks to those who contributed, especially @SolarQueen, who provided additional advice via PM.

That’s why I don’t think I will ever do grid tie. Someone else makes the rules and they generally favor the rule maker.

And I don’t trust that they won’t change the rules once I’m playing their game. Just my opinion.

 

[snoobler]

That’s why I don’t think I will ever do grid tie. Someone else makes the rules and they generally favor the rule maker.

And I don’t trust that they won’t change the rules once I’m playing their game. Just my opinion.

Concerning rule changes, prior to 2015, it was much more favorable with net metering and only modest fees. Owners of systems installed prior to the rule change are grandfathered in, but I believe they are contractual between the power company and the user. The next owner of the home plays by the new rules.

I've reached out to one of the commissioners, who is a big solar advocate to try and get a feel for the future. The recent election did nothing to change the composition of the AZCC, so it looks like it's two years away.

I'm trying to avoid running the numbers of going completely off grid.