System design math check needed for 100' distant 24V system concept

[SolarSteve]

Hi all! We have a remote cabin in central Maine that is used (mostly) on weekends from late May through early October. The primary gobbler of electrify is the 1947 Servel gas refrigerator I just electrified (if you have questions about that, I'd be happy to answer elsewhere), which draws roughly 30W/2.5A for about 3-5 minutes (average) with between 3-8 cycles an hour (depends on standard refrigerator variables). Secondary uses are LED lights of various sorts for fairly limited amount of time (we're sticking with propane/lamp oil for primary lighting). And of course those days when everybody has a dozen electronic devices charging despite telling people to leave the damned things at home

Since the big, reasonably constant drain is from the refrigerator, I figured out (roughly) that for a worst case 24 hour period it requires about 1,800Wh (30W x 5 Minutes x 5 times x 24 hours). After watching my family bumble around in an over packed fridge for four days in decent temperatures I think this is a reasonable worst case calculation. Obviously it will be MUCH better between visits as it will be nearly empty and nobody leaving the door opening to find stuff.

The problem I have to deal with is the cabin's location has highly filtered sunlight due to tall pines that can not be removed. There is an excellent spot for unobstructed southern exposure for about 10 hours a day, however it is 100' away. Knowing the distance introduces a lot of complications I initially ruled it out and instead put a test system on the roof of the cabin just to see what would come of it. Long story short, I found out that the size of the system I'd need would be stupidly large. Therefore, I am now back to the 100' distant system.

Here's the system I'm thinking of. Please tell me if I'm on the right track:

1. Four 100W panels wired in series
2. Two ~100AH batteries wired in series
3. PWM controller that can handle at least 70V input and has 24V output
4. 100' of 10 AGW

The 400W system wired in series puts out roughly 64V @ 6A max (I am presuming actual max output per panel is 16V @ 6A). If this is put into 10 AWG and pushed 100' the voltage drops to 62.8V by the time it hits the controller, which then dumps into a 24V battery bank consisting of 2 x 100AH batteries in series. The refrigerator and lights can handle either 12V or 24V as is, so no problems there.

If I'm doing my math correctly, the 400W system in full sun day can produce about 3,800Wh (62.8V x 6A x 10 hours). If my worst case 24 hour period usage is around 2,400Wh (1,800Wh for fridge and 800Wh for everything else) I'm able meet the demand and have roughly 1/3 left over to put back into the batteries. Even on a sub optimal solar day there's also relatively hard usage it seems like I'd be OK. Does that math look right?

For batteries it seems to me that with a full 400W system on a maximum sun day and a moderate use pattern will result in a lot of potential stored energy going to waste even if I have a 100AH battery bank. Yet it doesn't seem likely that I really need more than 100AH available to me because for the most part the system won't be in use more than keeping the idle fridge to temperature. So it doesn't seem like I should bump up the storage capacity, though it also doesn't seem like I should reduce the production capacity because I could run into problems here and there with less than 400W. Any thoughts on this?

Final question is about the controller. A MTTP controller seems to be a bad idea for this application because I want to keep the amp load in the lines down and the voltage up. As I understand it, the MTTP attempts to do exactly the opposite while the PWM does what I need in this application. True?

Does all of this sound right to you all? If I goofed somewhere, what would you recommend to remedy the error?

Thanks!

Steve

[bcroe]

1. Four 100W panels wired in series
2. Two ~100AH batteries wired in series
3. PWM controller that can handle at least 70V input and has 24V output
4. 100' of 10 AGW

Final question is about the controller. A MTTP controller seems to be a bad idea for this application because I want to keep the amp load in the lines down and the voltage up. As I understand it, the MTTP attempts to do exactly the opposite while the PWM does what I need in this application. True? Steve

The MPPT controller will operate your panels more efficiently (by definition), and will also lose
less voltage in the wiring run by operating at somewhat lower current. At the battery it will
deliver about twice as much power as a PWM controller. A PWM is all bad here. Bruce Roe

[organic farmer]

Greetings.
We live a bit North of Bangor. Our photovoltaic panels are about 80' from our house, I used regular MC4 cabling and it seems to work fine.

Are you thinking of turning the refrigerator off when you are not at your camp? I could see where you could have some fully charged batteries waiting for you each time that you arrived at camp. But if there were any ghost currents, they could also drain your batteries to zero SOC while you were gone.

Also what are your plans for keeping the batteries warm? We have four homes here in our township that are on solar power. One of those destroyed his batteries when they were left un-heated and they froze one year.

[ButchDeal]

Final question is about the controller. A MTTP controller seems to be a bad idea for this application because I want to keep the amp load in the lines down and the voltage up. As I understand it, the MTTP attempts to do exactly the opposite while the PWM does what I need in this application. True?

Does all of this sound right to you all? If I goofed somewhere, what would you recommend to remedy the error?

MPPT is the answer. PWM would be nothing but bad.
I am not sure what the confusion is but MPPT is designed for this kind of situation. You can run lower amp, higher voltage on the smaller wire from the modules in series to the MPPT charge controller and higher amp lower voltage from the MPPT CC to the batteries. This low voltage run should be short with the CC near to the batteries.

[sensij]

Also, you are overestimating the production potential of 400 W. It will probably be half what you are thinking, even less in winter. Check out PVWatts for decent modeling, and PVOutput.org for actual data from real systems ( you can search by zip code)

[SunEagle]

....

Final question is about the controller. A MTTP controller seems to be a bad idea for this application because I want to keep the amp load in the lines down and the voltage up. As I understand it, the MTTP attempts to do exactly the opposite while the PWM does what I need in this application. True?

Does all of this sound right to you all? If I goofed somewhere, what would you recommend to remedy the error?

Thanks!

Steve

I doubt you will find a PWM CC that will allow you to wire those 4 x 100w panels in series without exceeding its DCV input max based on the Voc of those panels.

On top of that you will probably lose about 33% of the panel wattage due to the way the PWM works.

Using a MPPT type CC will allow you to use a higher DCV input as well as use closer to 100% of the panel nameplate wattage for charging.

[littleharbor]

If you series wire your panels and run that into a PWM controller you will harvest the equivalent of less than one panel. You will get whatever amperage (less than 6 amps) your panels put out at your batteries voltage so lets say 5.5 amps x 13 volts, for example, = 71.5 watts.
You will never get 10 hours of equivalent sun, at best maybe 5 - 6 hours. Panels need direct sun. Anything other than directly perpendicular to the sun will give less than full output. When the sun starts riding lower in the sky in fall and winter you may get shadowing from tall trees.
Unless you already have the 4 12 volt panels you would save a lot of money by using larger, cheaper 60 or 72 cell panels.
Are you at all concerned that your remotely located panels will be vulnerable to theft?

[Mike90250]

i'd suggest you stop and price out an alternative system:
2 or 3: 200w -300w panels wired in series for about 90V
the price you save on panels would pay for the more expensive MPPT controller, but you will find the increased power very useful. Morningstar and Midnght both make solid, reliable small MPPT controllers. Be sure to match the series panel total Voc when cold and frosty, to your max DC input to controller. Don't exceed the max input voltage

100w 12v panels are not the way to go

[SolarSteve]

Lots of great responses! Thanks! To sum up what is above:

1. OK, MPPT is obviously the consensus way to go Noted! Though just an FYI, the Xantrex PWM controller can handle 125VDC @40A, so in theory it is possible to do.

2. I can have ZERO load on the system 4-5 days a week if I should so choose to. It would come down to a financial consideration. I'm not going to invest a ton more into this system to keep an empty fridge cold while I'm not there. Although it would be nice to not have a cool down period when we arrive I consider it a "cherry on top" feature because by electrifying the old Servel the cool down time went from ~16 hours to 1.5 hours. Which means even if I start up the fridge when we arrive it will cool down fast enough that the stuff we bring in a cooler will remain cool/cold until the fridge is down to temp.

3. The batteries will come home with me over the winter. Yup, Maine winters are harsh. I know, as we live the rest of the year a mere 40 minutes south of our cabin We also live off grid year round.

4. Similarly, because we won't be using the system during the winter the loss due to reduced sun opportunities isn't an issue in this case. Therefore, the system can be sized presuming 8-10 hours of summer sun with a little less at the start and end of our time there.

5. Thanks for the calculator link to PVWatts. Not familiar with that one! According to this a 4kW system as planned will produce about 4.7 to 6.2kWh per day for the applicable time period. That is quite a bit more than I was estimating, not less. And I didn't even go back through and plug in the different angles (my system is 2 axis, though tilt is set each month). I'm going to stick with my more pessimistic figure because that's sensible. Plus, I have the option to let the batteries have 4-5 days of uninterrupted charging.

Now to jump to Mike90250's suggestion in my next post.

Steve

[sensij]

4 kW = 4000 W. You were proposing a 400 W system. You might want to try PVWatts again, using 0.4 as the system size, or just divide your numbers above by 10.

[littleharbor]

Lots of great responses! Thanks! To sum up what is above:

1. OK, MPPT is obviously the consensus way to go Noted! Though just an FYI, the Xantrex PWM controller can handle 125VDC @40A, so in theory it is possible to do.

Steve

A Xantrex C-40 is only rated to 125 volts because it is a 12, 24, and 48 volt controller. That is the maximum open circuit voltage it can tolerate before you damage it. That said it is not designed to run a 12 volt system at 125 volts input voltage. The simple matter is, with a PWM controller you get amps in = amps out AT THE BATTERY VOLTAGE. You are making assumptions about this controller that aren't correct.

[SolarSteve]

I hadn't thought of going with larger capacity panels because I already have a brand new 100W panel with an inefficient PWM controller. I can use that combo for something else, so I'm OK with starting from scratch for the cabin project.

Quickly looking at pricing and voltage output I see:

4 x 100W = $520 with ~76V output
3 x 265W = $550 with ~93V output
3 x 300W = $750 with ~96V output

From a cost standpoint the first two options are a wash, but obviously the cost per W, voltage, and total capacity is radically different. I can see why the 100W panels don't look so good now that I know how to look at it.

OK, so here's a new system proposal to kick around:

3 x 265W panels ($550)
MidNite Solar The Kid ($291)
2 x 105AH Duracel Ultra Deep Cycle batteries ($220)
100' 10AWG two strand direct burial 600V rated wire ($174)

So for about $1300 can have the power and storage we need.

That sound reasonable, or did I go off the rails again?

Steve

[littleharbor]

Where are you located? The price for 300 watt panels seems pretty high. They are less than half that in my area (So. Cal.)

[SolarSteve]

4 kW = 4000 W. You were proposing a 400 W system. You might want to try PVWatts again, using 0.4 as the system size, or just divide your numbers above by 10.

Hmm. There appears to be something wrong with their calculator. I entered 0.4kW and it spits out the same values as 4kW. You'd think they would have accounted for that by either not ignoring the decimal or having the system sized in W instead of kW. Oh well!

Now knowing this, I entered the W output of the new 3x265W option and divided by 10. It comes out to be roughly 470W-650W per day, depending on the month. If my worst case usage for the fridge is accurate (and I deliberately made it inaccurate on the high side) it will consume 1800W on the worst of worst days. This means for the worst of worst days I'll need 2-3 additional days to keep it running. Given that it can be shut off for 4-5 days in a row, that seems to be doable.

The question is if 100AH storage is going to store 2-3 days worth of juice. Any thoughts on that?

As for the Xantrex, issues are duly noted. But not relevant as I'm not going with PWM anyway

Steve

[max2k]

Where are you located? The price for 300 watt panels seems pretty high. They are less than half that in my area (So. Cal.)

$250 / 300W panel seems ok price especially taking into account he's buying only 3 or I'm out of market for too long? He didn't specify brand though.