Please comment on this off-grid design with gen backup

[Sunking]

Mike,

"Why"? Because I am a beginner! I don't believe I have what you consider a really long run. I've heard you can go 300' or more DC, but I haven't run the numbers. I only have 70'.

This has been answered many time in other threads you have been involved in. It is voltage drop and keeping it to less than 2%. Watts = voltage x current. At low voltage requires high current. Click here for a calculator and run some numbers. Try 1000 watts. Enter 140 feet which is how many loop feet 70 feet is. Then for voltage try different voltages from low to high (18 to 150 for example) and see what happens What you are going to find is it gets real expensive the lower you go in voltage to maintain 2% or less.

[Living Large]

This has been answered many time in other threads you have been involved in. It is voltage drop and keeping it to less than 2%. Watts = voltage x current. At low voltage requires high current. Click here for a calculator and run some numbers. Try 1000 watts. Enter 140 feet which is how many loop feet 70 feet is. Then for voltage try different voltages from low to high (18 to 150 for example) and see what happens What you are going to find is it gets real expensive the lower you go in voltage to maintain 2% or less.

You've been involved in these threads too. I chose a higher voltage to reduce the current thinking I would keep wire size down since I am going between buildings. Then I got slammed for going higher voltage. You can't win on the internet, with many experts afoot.

I am an EE as well.

[Living Large]

a) working on 3rd week now, no sun. Got 14 minutes today, been about 3- 4 hrs daily generator runtime. Larger array would not be of help at the moment. No sun forecast for the next week either !

b)

Yes, if the bottom row 4-8-9-16 is shaded, even parallel with 3 other strings, the controller will only be 25-30% reduced. Any string with shade will be disabled. Even shade from a twig can be enough to shut down a string.

And a item I learned, is that the higher the PV voltage, the more loss (heat) in the controller downconversion. My 140V strings into my batteries, has the cooling fan in the controller screaming all the time. For a 48V bank, EQ at 65V, you don't want to go much over 120Vmp in the strings, unless you have really long runs that you can't afford copper for.

And in winter, or cloudy conditions, you run the genset on day 2, and reserve day 3 for when the weather is so bad, you can't get to the genset. If your batteries are already flat, you are in the dark. As much fuel as I have been using and transferring, even though diesel stinks, it doesn't explode like gasoline can. And gas stinks too.

This post is full of goodies. 14 days of no sun makes me question my whole endeavor. Yeah, I saw were a twig can reduce output, which is scary.

Interesting fact about down conversion. I hadn't come across this. So there is a good reason for the 150V max that many MPPT's are rated for? Sounds like going high voltage is reserved for situations were you really need it. Okay, something else new. Thanks.

So I should really try to get down to 3 panels max in a string is what you are saying. That practically means to do a 16 panel array with two MPPTs, 2 x 4 and 2 x 4? Two Classic 150's, or as I believe Paul said earlier possibly one. Then I'll be at 80A or so between buildings. Guilty as charged - I was thinking of keeping current low all along from a wire size and voltage drop perspective. That sounds ok without taking pencil to paper. Can I run two MPPT output circuits through the same combiner box? With house wiring, you can't have more than one (branch circuit) in a box per the NEC.

Edit - well, Sunking recommends 64V, and preferably 90 to 120V for a 48V bank. My Vmpp is 30V. That would be 3 or 4 panels/string. Thus it appears a 16 panel array isn't optimal. There's a lot going on for a newbie to this game. I don't like dropping a panel, but Mike your comment may be "if there isn't sun it doesn't matter how many you have." Should I consider going to 3 * 5?

[Sunking]

You've been involved in these threads too. I chose a higher voltage to reduce the current thinking I would keep wire size down since I am going between buildings. Then I got slammed for going higher voltage. You can't win on the internet, with many experts afoot.

I am an EE as well.

It is all about economics. Try several design models and see which best fits performance vs cost. The high voltage CC are expensive, but if your runs are long can be less expensive once you factor in material cost is all I am saying.

[Sunking]

Edit - well, Sunking recommends 64V, and preferably 90 to 120V for a 48V bank. My Vmpp is 30V. That would be 3 or 4 panels/string. Thus it appears a 16 panel array isn't optimal.

64 volt is minimum for 48 volt battery because EQ voltage is 62 volts at the batteries. What you really want to know is at what voltage is the controller most efficient combined with wire losses. You want to get to the batteries with 6% of less loss from th ewiring and controller

With any MPPT controller, you have to use specs for the specific controller. All have limitation based on panel wattage vs battery voltage. For example the MNS Classic controller is most efficient at 90 volts Vmp has maximum panel wattage of:

1380 watts @ 12 volts with 96 amps charge current.
2700 watts @ 24 volts with 94 amps charge current.
5022 watts @ 48 volts with 86 amps charge current.

Now go look what happens if you run up to 120 Vmp


Move up to a 120 volt battery using Classic 250VKS and you can run panel voltage up to 200 volts @ 5000 watts and only require 35 amps to the batteries and 25 amps on the panels. Only way you are going to find the best fit is make several models and price them out vs performance.

The scary part which I think you are just now figuring out is off-grid expensive. Just battery cost alone is going to cost you 5 times more than buying from POCO, plus all the other crap and a new Part Time Job the rest of your life with no days off.

Edit Note:

You can run two controllers to a common battery, but the Panels will be two separate systems. Only thing in common is the BATTERY

[Living Large]

This is food for thought. I've now looked at multiple configurations, but not as thoroughly as you suggest. I didn't consider efficiency. I did look at wiring losses over my 70' run, which is why I leaned towards higher V, lower current solutions. Lower current, and the ability to generate enough power, was what I was focused on. Then Mike questioned using a high V system and mentioned heating ("screaming fan").

I can go back and do it again. I'm leaning towards 3 or 4 per string for a 150V or 200V controller (same price), or the 8 per string with the 600V controller (twice the cost of the 150,200V controllers). 2 per string requires wire the size of my ring finger, so that is out.

Based on what you are saying and referencing, I am guessing you (and Mike) are suggesting I add efficiency as a variable. One thing I am reminded of at this point is the XW is 80A, which is marginal. So maybe just going to two (60A+) controllers and tossing the high V XW controller out the window is a start. If I go 3 per string I will add two panels instead of dropping one. I'll look at the curves at your link and see if I can come up with something useful in comparing two 3 x 3 arrays (classic 150) and two 4 x 4's (classic 200). My gut feeling is it may be worth paying for two more panels, even if it is a little less efficient.

I'm not appreciating yet if there is an inherent consequence of a greater difference between array Vmpp and the battery voltage, if there is one. For example, if a 4 string array vs 3 implies a less efficient operation for a 48V battery bank. I'll have more than enough voltage either way. Thanks.

[Amy@altE]

Move up to a 120 volt battery using Classic 250VKS and you can run panel voltage up to 200 volts @ 5000 watts and only require 35 amps to the batteries and 25 amps on the panels. Only way you are going to find the best fit is make several models and price them out vs performance.

Midnite discontinued that 250KS charge controller earlier this year.

The problem with finding the right solution is that there are many right solutions. There's no one silver bullet that will solve everything. Unfortunately, there's a heck of a lot of wrong solutions out there too. This forum is good with helping steer you away from them.

[Living Large]

Midnite discontinued that 250KS charge controller earlier this year.

The problem with finding the right solution is that there are many right solutions. There's no one silver bullet that will solve everything. Unfortunately, there's a heck of a lot of wrong solutions out there too. This forum is good with helping steer you away from them.

Yes, this is what I am finding. Everything you said. I really appreciate this resource.

Here is today's problem. I moved to an 18 panel array, with two 3 x 3's feeding two 150V MPPT's. This was to provide redundancy, and create two lower current feeds across my 70' between the combiner and MPPT. This looked great until I ran a sizing program and was told that using a 6 gauge Cu cable, I lose 40W in each feed, or 1.6%. The sizing program (Conext) declared this excessive. The loss was acceptable at 2 ga. The MPPT says it can handle 6 ga and smaller wire.

Logistical question - Can I run four 2 ga (two for each array) from the combiner across the 70', into a disconnect, and change to 6 ga between the disconnect and the MPPTs, as long as the 6 ga will handle the current?

Note: Regarding Amy's comment - recall I initially had a 2 x 8 array with a 600V MPPT, and was asked why. The main driver at that time was lower current to reduce power loss across the 70'. I moved all the way down to 3 per string and a 150V MPPT. I looked at 4 in a string with a 200V, both a single 4 x 4 and two 2 x 4. The 4 x 4 is more than the power capacity of a MS 200 into a 48V battery bank. Two 2 x 4 arrays into two 200V MPPTs gives redundancy, and lower current. 4 separate strings may reduce intermittent shading losses. I was happy with this solution, but my reservation here is a 200V MPPT is not made by many manufacturers. If I go to two 2 x 4 arrays, the wire size problem above goes away and I can use 6 ga all the way. 2nd question - Would building a 200V MPPT requirement into the system be risky in terms of future maintenance (replacement when it goes bad)? This is an opinion one.

Any replies welcome, which may vary a bit. Thanks.

[thastinger]

I bought a spare MN classis 200 just in case the one in service were to ever need repair.

On your wiring questions, are you pulling a permit to do this install? Question is really related to insuring the dwelling. I had to provide the ins company with a copy of the final inspection for mine. I believe what you want to do is acceptable for a few inch section of wire but going from 2 to 6 might be out outside the guidelines of how much a wire can be stepped down to fit a hub. You'll also need to research the requirements for the junction (hydraulic crimp etc).

[Living Large]

I bought a spare MN classis 200 just in case the one in service were to ever need repair.

On your wiring questions, are you pulling a permit to do this install? Question is really related to insuring the dwelling. I had to provide the ins company with a copy of the final inspection for mine. I believe what you want to do is acceptable for a few inch section of wire but going from 2 to 6 might be out outside the guidelines of how much a wire can be stepped down to fit a hub. You'll also need to research the requirements for the junction (hydraulic crimp etc).

The answer to your question is I am trying to design this as much as possible to get an accurate estimate and feeling the project is feasible, and then pass it to a professional to review/improve. I assume there will be a permit. I am totally off grid, and won't be eligible for NYS rebates. Possibly the federal - but I usually pay no federal taxes. Yes, I saw where the insurance co will be looking for documentation.

To be clear, I am asking if I could terminate at the disconnect with 2 ga, and then go out of the disconnect at 6 ga. I wouldn't be splicing 2 ga to 6 ga - I wasn't sure if that is what you were addressing as far as a crimp. My thinking was, as long at the 6 ga can handle the required current, this transition should be alright. And the 6 ga would be a very short piece and thus drop little voltage.

My fear on the MS 200 was that in 10 years if I need to replace it, 200V controllers may not be available. Possibly irrational, I know. Yes, I have noticed that having a backup for most things is a good idea. I was counseled by an installer to have a backup inverter as well. Ouch.

[Amy@altE]

1.6% voltage drop is fine

Gotta love the Germans, 1.6% drop is not excessive, keeping it under 2%, especially if it is a higher voltage array like yours is fine. I haven't been following all of your posts, I know you've been changing around alot, but if memory serves me, you are looking to do 16 275W 60cell panels. If you are dividing them across 2 Classic 200s, 2 strings of 4 for each, that's about 18A (Imp x 2) at 128V (Vmp x 4). That's less than 1% drop with 6AWG. Here"s a voltage drop calculator I like. http://www.calculator.net/voltage-drop-calculator.html Have you checked out the YouTube video I did on "Wiring Your Solar Electric System"? I think you'd like it, it covers a lot of this.

However, remember that you will go from the combiner box into a DC Load Center that will have a 30A breaker before going into the charge controller. If you use a panel mount breaker, you can crimp a ring lug on the big wire and connect it to the bolt on the breaker. The output wire of the breaker can be smaller, as it is going a short distance to the charge controller.

If you do need to transition from a bigger wire to a smaller one for the last inch or so, Burndy makes a nice terminal that the AHJs seem to like. BIBS43

Amy

[Living Large]

Gotta love the Germans, 1.6% drop is not excessive, keeping it under 2%, especially if it is a higher voltage array like yours is fine. I haven't been following all of your posts, I know you've been changing around alot, but if memory serves me, you are looking to do 16 275W 60cell panels. If you are dividing them across 2 Classic 200s, 2 strings of 4 for each, that's about 18A (Imp x 2) at 128V (Vmp x 4). That's less than 1% drop with 6AWG. Here"s a voltage drop calculator I like. http://www.calculator.net/voltage-drop-calculator.html Have you checked out the YouTube video I did on "Wiring Your Solar Electric System"? I think you'd like it, it covers a lot of this.

However, remember that you will go from the combiner box into a DC Load Center that will have a 30A breaker before going into the charge controller. If you use a panel mount breaker, like the Midnite MNEDC30, you can crimp a ring lug on the big wire and connect it to the bolt on the breaker. The output wire of the breaker can be smaller, as it is going a short distance to the charge controller.

If you do need to transition from a bigger wire to a smaller one for the last inch or so, Burndy makes a nice terminal that the AHJs seem to like. BIBS43

Amy

Thanks for the reply, Amy. I switched around, as I have summarized, for the reasons stated. The condensed version is Mike discouraged me from high V, I was encouraged to think about 2 MPPTs for redundancy, and overall I have tried to keep the current low since I am going at least 70' (maybe more). But yes, I went from 2 x 8 high V (rejected), to 4 x 4 (rejected), and finally the option of two 3 x 3 or two 2 x 4. I would go with either of those.

I appreciate your specific comment regarding the drop, and that Schneider may be conservative. That's another data point for me to be aware of. I mentioned the source of that information just in case someone like you could critique it.

I believe you said that I may do the long run at 2 ga, and use a breaker to transition from 2 ga to 6 ga, with the terminal adapter you specified, (if desired) to reduce line drop. That's good to know, and yes, I checked the lower current with 2 x 4 already, but I won't necessarily use this as a deciding factor in one option vs the other. At the moment, these are two viable options. I am leaning two 3 x 3 for greater power and being able to use the Xantrex MPPT, so the monitoring will be only one box. I am inferring here that a breaker may serve as the "Disconnect" I was told I have to have on the input to the MPPT, and a breaker on the output will be the Disconnect to the batteries? These would both be in the load center, I assume.

Thanks again!

ps Thanks also for specifying a splice I *could* use, suggested by thastinger, if that were in fact to come up at some point.

[Amy@altE]

Yes, if you use an all Schneider XW system with 2 3x3 , at 150V, with their Power Distribution Panel, they use 2 60A panel mount breakers between your combiner box and the charge controllers that can be your DC disconnects from inside. Don't worry if the amperage of the breaker is a little high, because you'll have the properly sized breakers in the combiner box for OCP. The one inside is really a convenient place to land the combined PV Output Circuit wire and provide the disco. You can have 2AWG on the input and 6AWG on the output of the breaker.

I ran the numbers, I see that's where the 1.6% drop came with the 6AWG. Especially since you are now oversizing your array to get the right string sizing, it really is up to you if you are OK with losing 1.6% by using a reasonably sized wire. If you do 5 wires of 2AWG, you'll need to oversize the conduit too to fit it all and comply with NEC.

[Living Large]

Yes, if you use an all Schneider XW system with 2 3x3 , at 150V, with their Power Distribution Panel, they use 2 60A panel mount breakers between your combiner box and the charge controllers that can be your DC disconnects from inside. Don't worry if the amperage of the breaker is a little high, because you'll have the properly sized breakers in the combiner box for OCP. The one inside is really a convenient place to land the combined PV Output Circuit wire and provide the disco. You can have 2AWG on the input and 6AWG on the output of the breaker.

I ran the numbers, I see that's where the 1.6% drop came with the 6AWG. Especially since you are now oversizing your array to get the right string sizing, it really is up to you if you are OK with losing 1.6% by using a reasonably sized wire. If you do 5 wires of 2AWG, you'll need to oversize the conduit too to fit it all and comply with NEC.

OK, more great information. I could also go with 4AWG wire, lose about 1%, and that would fit the 60A breaker terminal, I believe. Could I fit five 4AWG in the standard conduit? Incidentally - silly question, are these five wires the 4 power conductors + a ground conductor?

[Amy@altE]

Yes, 2 combined pairs plus ground. You've brought me to the end of my knowledge, I don't know how to size conduit. Maybe I'll research that tomorrow and do a follow-on video to my wire sizing one.