Please check my knowledge about solar and battery sizing.

MysteriousFoundation
MysteriousFoundation Registered Users Posts: 31 ✭✭
I'm building a camper van, and I'd like to have solar power. I've done hours of reading blogs and forums, and watching YouTube videos and such. I have a decent grasp on what I'm doing, but I still feel so ignorant sometimes. I'd like to have my thoughts checked about the basics of sizing a solar system.

Solar panel sizing:

You have to plan for your worst case scenario of sun availability. If the sun doesn't provide to your panels at least as much energy as you use each day, your batteries won't recover. This seems obvious. So, you size your panels based on whether or not you're going to tilt them, and the latitude you're going to be at, and the time of year.

As an example: let's say I wanted to camp in upstate NY in the winter. (Why the heck would I do that, though, right?) I used pvwatts.nrel.gov and determined that with 350W of 0­­° tilted panels, including 12% system losses, that I should expect 18 kWh of power in the worst month: December. If we ignore the difference in daylight from day to day during the month, that works out to 935 Wh a day. If I want to have a full day of backup power, would I need to use only half of that per day in order to recover? Or does that number from pvwatts already include average cloud cover? It wouldn't be a very useful tool if it didn't. If that's the case, then I don't really need to divide by 2 because it's already the average power provided by the panels including cloud cover. (But not accounting for snow on the panels, I'm sure.)

I did some math to determine my bare minimum power usage for lights, a fridge, computer, phone charger, etc., and it came out to 785 Wh a day, which is more than half of the 935 Wh the sun would provide on average, so it's important to know if I actually need to divide that by 2 to plan for a spare day of power.

This is mostly academic up to this point, because I don't plan on wintering in NY unless I'm visiting friends, in which case I can just plug into shore power (or stay in their home!). But I really want to understand the math.

Battery Sizing:

You have to consider the depth of discharge of your batteries. You don't want to overdo it because it lowers their life. 50% is often cited as the maximum depth of discharge, with maybe 15-20% being ideal for battery life. In addition, depth of discharge is limited by the number of hours of light in a day. If it takes 10 hours to reach the float-charging stage and there are only 9 hours of daylight on Dec 22 in upstate NY, then you're never going to fully charge the battery, you'll sulfate the plates, and it will no longer accept a full charge. More panels don't help this. It's a matter of time.

That being said, I have no idea how to determine the maximum DoD given the number of hours of light in a day! Is there a rule of thumb for this?

Here's what batteryuniversity.com says about time of charge: "During the constant-current charge, the battery charges to about 70 percent in 5-8 hours; the remaining 30 percent is filled with the slower topping charge that lasts another 7-10 hours."

Is it really that bad though? I mean the time to charge.

It doesn't specify from what state of charge. Is it assumed to be 0%? I think it's moot, though, because if just the topping charge can take 7-10 hours, then it sounds like I better make sure I never go above 30% DoD in the winter in NY or I might not recover.

Assuming my bare minimum daily usage is that 785 Wh from earlier, and assuming I want a mere 30% DoD and one extra day of power, then here's my math for battery size:

785 Wh * 2 * 0.30 = 6280 Wh = 436.1 Ah @ 12V.

That's not bad, because I was thinking of buying 4x Trojan T-105 6V batteries which will total 440 Ah in series-parallel. Should I plan for more of a buffer, though? I probably shouldn't even be planning for a worst case scenario I'll never actually expect, but I also bet there are other things I'm not considering. Any tips would be appreciated.

A more realistic daily usage from my numbers would be 1200 Wh a day. With a 50% DoD that should be doable with just 402 Ah. Or a 46% DoD on 440 Ah of batteries. That's assuming I have enough panels to cover that much usage.

I was also looking at the T-1275 battery. Having three of them is almost identical as 4 of the T-105: 108% price, 94% volume, 103% weight, exact same amp hours. The heights are identical. It seems to come down to which footprint I think would work better in the van: 14x21" or 7x39" (ignoring required space around them.) Of course, saving $45 if buying the T-105's isn't nothing. ($540 vs $585). Any other considerations? I can think of one. Deciding to go with a fourth T-1275 is cheaper than two more T-105's, if I only needed another 150 Ah.

I think I've posted enough for now. I still want to post a list of my devices and expected energy usage to get some feedback on that. Maybe I'm under- or overestimating things. When I do the math, it makes it seem like the batteries will hardly last at all!

Comments

  • Photowhit
    Photowhit Solar Expert Posts: 6,002 ✭✭✭✭✭
    determined that with 350W of 0­­° tilted panels, including 12% system losses, that I should expect 18 kWh of power in the worst month: December. 
    I guess for the month? Even that seems preposterous! 

    Solar irradiance for Plattsburg NY for Dec is 1.24. So roughly you could expect 1.24 hours of direct sun light on a given day in Dec.




    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Magnum MS4024, Prosine 1800(now backup) and Exeltech 1100(former backup...lol), 660 ah 24v Forklift battery(now 10 years old). Off grid for 20 years (if I include 8 months on a bicycle).
    - Assorted other systems, pieces and to many panels in the closet to not do more projects.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Just to add a data source, fixed array tilted for optimal winter harvest:
    http://www.solarelectricityhandbook.com/solar-irradiance.html

    Albany
    Average Solar Insolation figures

    Measured in kWh/m2/day onto a solar panel set at a 32° angle from vertical:
    (Optimal winter settings)

    JanFebMarAprMayJun
    3.28
     
    3.97
     
    4.27
     
    4.10
     
    4.13
     
    4.26
     
    JulAugSepOctNovDec
    4.37
     
    4.43
     
    4.35
     
    3.89
     
    2.94
     
    2.91
     

    For a 350 Watt array (fixed, tilted to 32 degrees from vertical), in December, the average harvest would be:
    • 350 Watts * 0.52 end to end off grid battery AC inverter system eff * 2.91 hours of sun = 530 Watt*Hours per day (0.53 kWH per day)
    • 350 Watts * 0.52 end to end off grid battery AC inverter system eff * 2.91 hours of sun * 30 days per month = 1,059 WH (1.06 kWH) per month
    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mcgivor
    mcgivor Solar Expert Posts: 3,854 ✭✭✭✭✭✭
    350 Watts * 0.52 end to end off grid battery AC inverter system eff * 2.91 hours of sun * 30 days per month = 1,059 WH (1.06 kWH) per month

    Using the above figures my calculation came to 15,888 Wh (15.9 Kwh)

    This is for seasonal adjusted for optimal harvest,  horizontally mounted panels would only be a fraction of this, a small fraction.
    1500W, 6× Schutten 250W Poly panels , Schneider MPPT 60 150 CC, Schneider SW 2524 inverter, 400Ah LFP 24V nominal battery with Battery Bodyguard BMS 
    Second system 1890W  3 × 300W No name brand poly, 3×330 Sunsolar Poly panels, Morningstar TS 60 PWM controller, no name 2000W inverter 400Ah LFP 24V nominal battery with Daly BMS, used for water pumping and day time air conditioning.  
    5Kw Yanmar clone single cylinder air cooled diesel generator for rare emergency charging and welding.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Thank you McGivor... Some reason I multiplied by 2 days per month vs 30 days in a month...

    -Bill  :/
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • MysteriousFoundation
    MysteriousFoundation Registered Users Posts: 31 ✭✭
    Thanks for the replies! 
    Photowhit said:
    I guess for the month? Even that seems preposterous!
    Yes, for the whole month. After entering 0.35 kW of panels and all of the other options, that's the number it came up with. However, I just redid it, and it came up with 11. I don't know what I did differently before.
    BB. said:
    Just to add a data source, fixed array tilted for optimal winter harvest:
    http://www.solarelectricityhandbook.com/solar-irradiance.html
    Thanks for the additional data source! It's pretty close to what PVWatts was using. 1.18 (handbook) vs 1.29 (PVWatts) for non-tilted panels Dec in Rochester, NY. Also, thank you for the formula.

    Just for the sake of checking the numbers on PVWatts, using it's own data, here's the same math done with the formula from BB, except not using the 0.52 efficiency number.

    I also don't think it's entirely appropriate to my case, because I'm building an RV and most of the devices are DC. I'll only be using an inverter for a couple of things. I'm assuming I can at least get 17.6% more out of my system when just using DC devices (1.176 being the inverse of an 85% efficient inverter that I won't necessarily be using.)

    Again, here's the math using PVWatts' numbers:

    0.350 (kW) * 0.8592 (efficiency) * 1.29 (insolation) * 31 (days) = 12 kWh.

    That's pretty close to the 11 the site gave me. (I don't where the other 1 kWh went.) 18 wasn't that far off. Whether that number is accurate or useful, I'm not saying. It comes down to the real efficiency, but at least I can get close using more-or-less the same math.

    This incredibly detailed article by a physicist at UCSD shows his system having an end-to-end efficiency of 0.62. https://dothemath.ucsd.edu/2012/09/blow-by-blow-pv-system-efficiency/ That's a bit better than 0.52, but in the same ballpark. It's certainly not 0.86 like PVWatts seems to think. Where did you get the 0.52 number?

    I don't plan on leaving the inverter connected when I'm not using it. His inverter's base drain was 8 of the 38% losses he experienced. In my math, I'm assuming I only have it connected 1/2 of the day. That's some real savings.

    If I take that physicist's 0.62 number and add back .04 it's 0.66. Maybe I'll use that number in my math. Regardless of the exact efficiency I use, it sounds like I need to at least double my panels if I truly think I'll spend any time in NY in the winter, off-grid, without tilting them.

    In this example, I'll also use the lower insolation number (from the handbook.)

    0.700 (kW) * 0.66 (efficiency) * 1.18 (insolation) = 545 Wh for the average day in Dec.

    I revised my minimum power usage and got it down to 513 Wh per day. That seems livable, but I need to buy 4x 175W panels now, or reconsider not tilting my panels. Tilting 350W gives me 499 Wh. Tilting 700W gives me 998 Wh. Not tilting 700W is 545 Wh (the formula above.)

    That was all incredibly illuminating! But what about the second half of my post? Any thoughts on batteries, charge time, depth of discharge, specific battery models, and all that?

  • Photowhit
    Photowhit Solar Expert Posts: 6,002 ✭✭✭✭✭
    This incredibly detailed article by a physicist at UCSD shows his system having an end-to-end efficiency of 0.62. https://dothemath.ucsd.edu/2012/09/blow-by-blow-pv-system-efficiency/ That's a bit better than 0.52, but in the same ballpark. It's certainly not 0.86 like PVWatts seems to think. Where did you get the 0.52 number?
    Under system Efficiency,  Your Physicist clearly states;
     
    Tom Murphy said:
    But let’s start our accounting where the wires meet the charge controller
    ...but you are starting with the Standard Test Conditions (STC) rating of the panel at 350 watts.

    The trick is these STCs are NOT Normal Operating Cell Temperature (NOCT) values. A 350 watt panel will normally only produce about 75% of the STC rating. Here is a link to Silfab's 350 watt panel;

    https://www.solar-electric.com/lib/wind-sun/Silfab_SLG-M_350_specifications.pdf

    You will find the STC values next to the NOCT values in a handy chart;


    You will not that normally This 350 watt panel will produce 264 watts or (264/350=) 75% of it's STC

    If you apply this to Tom Murphy's 62% starting at the charge controller input, you will end up around 46-47%. I like 50% as a nice round number.

    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Magnum MS4024, Prosine 1800(now backup) and Exeltech 1100(former backup...lol), 660 ah 24v Forklift battery(now 10 years old). Off grid for 20 years (if I include 8 months on a bicycle).
    - Assorted other systems, pieces and to many panels in the closet to not do more projects.
  • Photowhit
    Photowhit Solar Expert Posts: 6,002 ✭✭✭✭✭
    It's certainly not 0.86 like PVWatts seems to think
    PV Watts does allow you to put in your own system losses.

    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Magnum MS4024, Prosine 1800(now backup) and Exeltech 1100(former backup...lol), 660 ah 24v Forklift battery(now 10 years old). Off grid for 20 years (if I include 8 months on a bicycle).
    - Assorted other systems, pieces and to many panels in the closet to not do more projects.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    If you lay the panels flat, you are going to lose a fair amount of harvested energy in the winter (plus possible covering with snow if you are not there to brush it off). Being up in northern region with flat to roof mounted panels (winter camping/power), it is a killer:

    Albany
    Average Solar Insolation figures

    Measured in kWh/m2/day onto a horizontal surface:

    JanFebMarAprMayJun
    1.74
     
    2.60
     
    3.57
     
    4.34
     
    5.04
     
    5.64
     
    JulAugSepOctNovDec
    5.55
     
    4.97
     
    3.95
     
    2.80
     
    1.76
     
    1.47
     
    Where do I get my numbers:
    • Warm weather (hot panels from solar heating) 0.81 derating (Vmp falls when solar cells get warm/hot). Also a derating if using PWM charge controller (panels in sub freezing weather can generate 10-15% or so more power with MPPT charge controllers: P=Vmp*Imp with Vmp higher in freezing weather). PWM controllers are not affected by "elevated" Vmp-cold. Also includes a couple of percent losses for "dirty/dusty" panels
    • Efficiency of MPPT charge controller ~0.95 (also used for additional PWM controller "efficiency").
    • Flooded cell lead acid battery cycling efficiency ~0.80 (range from 0.80 to 0.90 efficiency)
    • AGM lead acid battery efficiency ~0.90 (0.90 to 0.99 efficiency)
    • AC inverter efficiency ~0.85 (ranges from 0.80 or less to ~0.95)
    For off grid end to end efficiency:
    • End to end DC system eff (nominal/conservative number) = 0.81 panel derate * 0.95 charger eff * 0.80 FLA batt eff = 0.62 eff (62% efficiency)
    • AC off grid system eff = 0.81 panel * 0.95 charger * 0.80 FLA * 0.85 AC inverter eff = 0.52 (52%) end to end AC off grid solar system efficiency.
    PVWatts defaults to a Grid Tied system (solar panels + Grid Tied inverter). No charge controller, no battery bank, "optimistic production numbers".

    Battery wise... LiFePO4 (Lithium Iron Phosphate) batteries are just about the best you can do for an RV system (efficient--close to 100%, lightweight, high surge and charging current, very little self discharge, no gassing, no float charging needed). However they are very expensive and they should have a BMS (battery management system) to keep cells "balanced") and to prevent over/under discharging (Li Ion batteries are very easy to damage/ruin if over/under discharged. Tend to be on the "bleeding edge" of battery technology--But many high end bus/rv installations are using them these days (finding a good vendor, the larger cell AH capacities, building a battery bank, etc. are all part of the learning experience).

    AGM batteries are very nice lead acid batteries. No gassing (under normal usage, no electrolyte levels to check, can be mounted on side), high surge current support. They are probably 2x as similar quality flooded cell lead acid batteries. And susceptible to damage/ruination from overcharging. Probably a year or two less life vs FLA.

    Flooded cell lead acid batteries are cheap, relatively rugged and forgiving. Easier to check state of charge (using a hydrometer to measure electrolyte specific gravity). Messy (gas, electrolyte mist, more cable corrosion issues), some people smell the sulfur from battery when charging.

    Personally, I like to keep to a single string of batteries. You can parallel 2 or 3 strings of batteries, but with additional care/maintenance issues.

    If they meet your needs (voltage/Amp*Hour capacity)... "Golf cart" batteries (usually 6 volt and ~200 to 220 Amp*Hour) flooded cell batteries are a great first bank. Cheap, fairly rugged and forgiving. And if you ruin your first bank (most of us over/under charge, over discharge, insufficient solar charging current, or don't check electrolyte levels), they are cheap to replace.

    If you need to more than one battery--I suggest that (for example) you use 2x 6 volt batteries in series for a 12 volt battery bank. It is easier to take a digital multimeter and measure the voltage on each battery (check battery health). Two 12 volt batteries in parallel--If you have one bad battery, the other battery in parallel "hides" the weak/failing battery.

    Ideally, for Lead Acid Batteries (both FLA and AGM), they take a fairly long time to charge (may take 3 hours to recharge from 50% to 80% state of charge (10% rate of charge), but it will take another ~4-6 hours to "absorb" charge them... That adds up to 7-10 hours of charging per day.

    It works out better if you size your battery bank, instead of 1 day of storage and 50% max discharge (2x daily load), to use 2 days of storage and 50% discharge (4x daily load). If you are nominally discharging/recharging only 25% of your battery daily, then there is usually enough hours of sun per day to get the battery bank "fully charged" (>~90% state of charge--Don't go for 100% charge every day--That is very hard on the battery bank).

    However, for RV's, many times, you don't have the room for a 4x daily load battery bank... And you have to live with 2x (and more genset time during bad weather/winter usage). Use the genset in the morning to recharge from 50% to 80% State of Charge, and the solar to finish charging.

    Generally for an off grid system, we suggest 50% maximum discharge for longer battery life... But for an RV, discharging to 20% state of charge is not a huge issue (other than the many hours of sun/genset to recharge). Most RV batteries probably "age out" (3-5 years for a "cheap" golf cart battery set before they cycle out (most people do not live for months at a time in their RV). Or they "murder" the battery bank (leave loads on when RV is parked, cycle the bank "dead", etc....

    Also, a lead acid battery bank should have around 5% to 13% rate of charge. 5% is good for weekend/sunny seasonal usage. 10%+ is usually good for full time off grid. Notice that means you may need more solar panels for a 4x size battery bank.

    Regarding loads, solar power is expensive (and difficult to harvest in winter, especially in the "far north" and north east US). Generally that means that for RV use, a propane refrigerator (with all of its faults) is usually a "better" solution. If you use a computer--Desktop systems can really suck the energy usage. A small/efficient laptop or pad computer will use less power. LED lighting, no fans, or small efficient fan, and an RV water pump can run nicely from a smaller solar power system (500-1,000 WH per day usually is a good starting point for planning).

    Of course, this depends on your needs (how long dry camping before you go back to town for supplies and propane/gasoline, space for batteries and solar panels).

    Electricity usage is a highly personal set of choices. What may work for me may not fit your needs. Please continue to ask questions. And do several paper designs first. Buying equipment before you know what will need wastes time and money. Buying batteries before you need them in the RV means you have to keep them charged and maintained (batteries age from the day they are made/are filled with electrolyte).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    What you're really concerned about, especially in a mobile application like a camper van, is what life is like day to day. Knowing that Dec has X Kwh on average in Y location isn't particularly helpful if Y has five days of straight gloomy weather when you're there. You will almost certainly need an alternate source of power (eg generator) even maxing out available space for pv on the van.

    The pvwatts monthly numbers are pretty much meaningless for a small off-grid application, because on sunny days you end up with more potential power than you can use or store, and runs of gloomy weather where loads use more than can be practically stored. The monthly numbers are more applicable to a grid-tied system in which the grid is effectively an unlimited capacity battery.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • MysteriousFoundation
    MysteriousFoundation Registered Users Posts: 31 ✭✭
    edited June 2018 #11
    BB, I appreciate you taking the time to reply.
    It works out better if you size your battery bank... to use 2 days of storage and 50% discharge (4x daily load).
    That's exactly what I had in mind. 450 Ah was what I was planning around. I'll have no problem fitting them in the van. It's a 158" wheelbase Sprinter. I mentioned the Trojan T-105 in my original post and the T-1275. (6V vs 12V.) You said you prefer a single string, but also like to avoid having batteries in parallel. It doesn't seem like I can avoid that if I think I need that many Ah.

    Would you rather have 4x T-105's in series-parallel, or 3x T-1275's in parallel?

    I figure you can always disconnect them to get an individual voltage reading, but a hydrometer is the best test, anyway.

    3 hours to recharge from 50% to 80%...  another ~4-6 hours to "absorb" charge them... 7-10 hours of charging per day.
    The link in my original post said, "the remaining 30 percent is filled with the slower topping charge that lasts another 7-10 hours."

    Is that not accurate? Should I only expect 4-6 hours and not 7-10? That's good news if so. My main question was whether the short duration of sunlight in winter would limit your depth of discharge regardless of how many panels you had, because it might simply take too long to recover from a 50% DoD due to the charge time.

    >~90% state of charge. Don't go for 100% charge every day--That is very hard on the battery bank
    This is actually the first I'm hearing of this. Can you explain the theory behind this? Everything I've read says it's important to make sure the batteries are fully charged as often as possible (every day) to avoid damaging them. But you're saying that "fully charged" shouldn't really be fully charged?

    How do you limit this with your charge controller?

    But for an RV, discharging to 20% state of charge is not a huge issue (other than the many hours of sun/genset to recharge)
    Right, it's the charge time that concerns me the most. Assuming you have enough solar panels and sunlight to give back all of the energy you use per day, including all of the inefficiencies, how deep can you realistically expect to take the batteries and still achieve a full charge? (Given that the rate of charge is limited.)

    You mentioned a 13% charge rate. I'm not used to seeing charge rates expressed as a percentage. Usually it's in terms of C/N, so that'd be about C/7.7. Is that what the 13% means? The general recommendation for bulk charging is between C/5 and C/10, so that sounds about right. But I want to make sure I'm on the same page as you. [EDIT: I see in the spec for a Trojan L16P that the maximum charging current is "13% of the C/20 rate". What exactly does that mean? That seems really slow! It's a 420 Ah battery, so C/20 is 21, and 13% of that is a mere 2.73 amps? How can that be right?]

    The slower you charge, the more time it takes, and that's what concerns me. Even at the maximum rate, if you discharge the batteries enough, there might not be enough hours of sunlight in the day to recharge them, regardless of how many panels you have. Is there a rule of thumb to match actual hours of sunlight to charge rate and to depth of discharge? I don't want to plan to use a generator every day. It seems like with more batteries you can avoid a deep discharge, and with enough panels you can always recover because the charge rate won't be a limiting factor. When does it become the limiting factor, though?

    Maximizing the battery life is still a consideration. I'm going to be full-time RV'ing for a year or more. It's the "more" I'm thinking about. I'm not sure how long this will last. Might be permanent. It'd be a shame to have to replace all of my batteries two years down the road (and buy twice as many) just because I didn't buy twice as many in the first place.

    a propane refrigerator (with all of its faults) is usually a "better" solution.
    I was looking at those, or specifically AC/DC/LPG ones. I'm going to be using propane anyway for cooking, and possibly heat, so I'll have it. I also hear that the propane fridges are really efficient. It's hard to find exact numbers, though.

  • MysteriousFoundation
    MysteriousFoundation Registered Users Posts: 31 ✭✭
    Estragon said:
    You will almost certainly need an alternate source of power (eg generator) even maxing out available space for pv on the van.
    I appreciate the dose of reality. I do agree, and I am looking to also get a generator. I just want my build to handle optimal situations when they arise. If it's possible, with the right build, to last through two weeks of sun in December without using the generator, then that's great! However, if that means I can't run anything but a single LED lamp, then I'd like to know that. That's why I want to crunch the numbers.

    I only have room for about 700W of panels on the roof, while still leaving room for a vent fan. Maybe 875W, but that's really packing them in. Hey, maybe I'll make an awning out of solar panels! Shade and power in one! :smiley:
  • Photowhit
    Photowhit Solar Expert Posts: 6,002 ✭✭✭✭✭
    I would actually suggest going with a cheaper set of golf cart batteries for your first set. People tend to screw up their first set of batteries for assorted reason. Trojan's do have a bit longer life in general but a set of Sam's Club/ Costco Golfcart batteries at $85 each plus core might be more cost effect. They both are reported to take about anything for a core. I've used lawn tractor batteries at Sam's Club.


    You can find other deep cycle batteries in the same range of amp hour capacity to avoid running batteries or strings in parallel. Here's a Trojan 6 volt 420 amp hour battery;

    Image result for trojan l16p-ac

    https://www.solar-electric.com/trl16vo225ah.html

    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Magnum MS4024, Prosine 1800(now backup) and Exeltech 1100(former backup...lol), 660 ah 24v Forklift battery(now 10 years old). Off grid for 20 years (if I include 8 months on a bicycle).
    - Assorted other systems, pieces and to many panels in the closet to not do more projects.
  • MysteriousFoundation
    MysteriousFoundation Registered Users Posts: 31 ✭✭
    Thanks! I was aware of the L16's. I'm just limited by what I can get locally, so I figured I'd stick to golf cart batteries from a golf cart dealer. Freight shipping for a battery bought online is completely prohibitive.

    I just called a local golf cart shop, and they can do custom orders, so I'm having him call Trojan to quote me a price on one of these. Thanks for pushing me to ask!

    He already got back to me. $375 including core charge. Two of them would be $750. Why is that so much more than the T-105's or T-1275? I can get a T-105 for $110 plus core charge. Or a T-1275 for $170 plus core. Assuming the cores are only $25, then that's only $540 or $585 for 4x T-105 or 3x T-1275, respectively. And that gets me 450 Ah either way. However, I would have to use more cables, and those definitely aren't free, but they wouldn't bring me to $750.
  • Photowhit
    Photowhit Solar Expert Posts: 6,002 ✭✭✭✭✭
     I can get a T-105 for $110 plus core charge. 
    That's pretty darn cheap, I think I'd go that way vs the Sam's Club/Costco. Might ask if they handle the cores by weight or size. Might be able to use car batteries for cores. Sometimes they are just looking for a count and a car battery would be fine. Might save a few bucks picking them up at repair shops.

    They just make so many golf cart batteries, it's the economy of scale. Far fewer 'floor scrubbers' the typical use for L16's size batteries.
    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Magnum MS4024, Prosine 1800(now backup) and Exeltech 1100(former backup...lol), 660 ah 24v Forklift battery(now 10 years old). Off grid for 20 years (if I include 8 months on a bicycle).
    - Assorted other systems, pieces and to many panels in the closet to not do more projects.
  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    edited June 2018 #16
    May not be a big deal, but 3 x 12v batteries in parallel is 6 x 3 = 18 cells to water and maintain. 4 x 6v in series/parallel is 4 x 3 = 12 cells. Also, with 3 in parallel, each battery should be fused to prevent current from 2 feeding a faulty one.

    I think the Trojan L16 charging spec may be a typo; instead of "13% of the C/20 *rate*", it should read "13% of the *amp-hour capacity* at the C/20 rate".

    The charge to full thing is complicated, and I'm not aware of much really good, practical, unbiased research on the subject. In theory, Bill is right in saying getting to "full" isn't needed daily, and can be hard on batteries. The first problem is defining "full". After initial break-in, all batteries lose capacity, so they never really get "full" thereafter. We make choose abritrary definition of full, based on terminal absorb current, specific gravity, a timer, etc. Continuing a high voltage ("high" also being somewhat arbitrary) beyond "full" does lead to added wear on batteries.

    The flip side is undercharging, the main issue with which being hardening of sulfation. This takes time, especially if battery self-discharge is low in low ambient temps, and if the battery SOC is relatively high to begin with. A battery at a fairly high initial SOC in cool temp won't sulfate as quickly as a battery at a low SOC in high temps.

    I suspect more banks die prematurely from chronic undercharging than overcharging, but IMHO, getting to "full" weekly isn't a bad objective. Off-grid has uncertainties with weather etc though, and it's nice to have a bank "full" ahead of the gloomy days.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    "MysteriousFoundation said:
    BB, I appreciate you taking the time to reply.
    It works out better if you size your battery bank... to use 2 days of storage and 50% discharge (4x daily load).
    That's exactly what I had in mind. 450 Ah was what I was planning around. I'll have no problem fitting them in the van. It's a 158" wheelbase Sprinter. I mentioned the Trojan T-105 in my original post and the T-1275. (6V vs 12V.) You said you prefer a single string, but also like to avoid having batteries in parallel. It doesn't seem like I can avoid that if I think I need that many Ah.

    Would you rather have 4x T-105's in series-parallel, or 3x T-1275's in parallel?

    I prefer the 2x series by 2x parallel T-105 batteries. The 3x T-1275 in parallel is 3 parallel strings, which I like less. And the 3x batteries are in parallel--So cannot easily do a quick health/sanity check of each battery with a voltmeter (unless you disconnect the batteries).
    There are DC current clamp meters (really AC/DC clamp DMM) that make measuring DC (and AC) current very easy and safe. One of the handy tools to have around when doing your own electrical repairs/monitoring.

    I figure you can always disconnect them to get an individual voltage reading, but a hydrometer is the best test, anyway.

    3 hours to recharge from 50% to 80%...  another ~4-6 hours to "absorb" charge them... 7-10 hours of charging per day.
    The link in my original post said, "the remaining 30 percent is filled with the slower topping charge that lasts another 7-10 hours."

    Is that not accurate? Should I only expect 4-6 hours and not 7-10? That's good news if so. My main question was whether the short duration of sunlight in winter would limit your depth of discharge regardless of how many panels you had, because it might simply take too long to recover from a 50% DoD due to the charge time.

    Ideally, the 4x daily battery capacity only discharges (on a normal day) to 75% of capacity--Much quicker to recharge (less bulk time, and less absorb time needed).

    To be honest, if you read up on lead acid batteries--You will convince yourself they will not work with solar. Obviously they do and do it well. However, if you can follow our rules of thumbs as a starting point, you should be a lot happier.

    >~90% state of charge. Don't go for 100% charge every day--That is very hard on the battery bank
    This is actually the first I'm hearing of this. Can you explain the theory behind this? Everything I've read says it's important to make sure the batteries are fully charged as often as possible (every day) to avoid damaging them. But you're saying that "fully charged" shouldn't really be fully charged?

    The last 5% or so charging for lead acid batteries is when most of the "gassing" takes place. At this point the gassing begins to errode the plates, and the Oxygen which forms on the positive plates causes positive grid corrosion.

    FLA batteries, if they are allowed to sit at ~80% state of charge (for more than ~1 day), the soft "fluffy" Lead sulfate (sulfide?) converts to a hard black crystalline form. At this point the lead and sulfur are "removed" from the normal battery charging/discharging cycle--Permanently reducing battery capacity and life. This is where you hear about "desulfators" that convert the crystals back into active battery capacity. I have not seen any proof that these work, but there are a few folks that believe that they do as advertised--Of the few threads that have been locked here, desulfators are a significant number of them.

    Note that when lead acid batteries are cycling daily (discharge/charge cycle), sulfates do not harden. There are some who recomend daily cycling the batteries between ~50-80% state of charge on a daily basis, and recharge >90% SoC once a week.

    Equalization charging is an "overcharging" of the FLA battery... Some cells slowly drift (through self discharge) from the other SOC of the rest of the cells. By overcharging (~15.0 volts or above and ~2.5% to 5% rate of charge) for a period of time, this brings of the SOC of the "low cells" (monitor with hydrometer every 30-60 minutes until all cells "stop" increasing SG--EQ is complete and log SG as the new "100%" SoC levels). Batteries can "overheat" with many hours of EQ, and the overcharging causes lots of gassing (erosion and corrosion) and loss of electrolyte (gassing and misting). Usually done when needed or ~once per month.

    How do you limit this with your charge controller?

    More or less, this is done by limiting "absorb" charging time. Either by littlerly programming 2/4/6 hours of absorb, or for some higher end charge controllers, by programming them to stop charging when the battery charging current (tail of absorb cycle) is around 1%-2% of battery capacity (100 AH battery, 1-2 amps of "end charging" current).
    But for an RV, discharging to 20% state of charge is not a huge issue (other than the many hours of sun/genset to recharge)
    Right, it's the charge time that concerns me the most. Assuming you have enough solar panels and sunlight to give back all of the energy you use per day, including all of the inefficiencies, how deep can you realistically expect to take the batteries and still achieve a full charge? (Given that the rate of charge is limited.)

    I speak to the issue "in theory" (I am/was an engineer, others here have a lot more practical RV+battery experience they can add). RVs are usually weekend or a couple of week trips. And mostly during "sunny weather"--When there is lots of "sun". For longer trips and in poor sun conditions (stormy, winter, shading at campsite, etc.), then you have to pay more attention to loads, amount of solar recharging, and possible genset runtime to make up the charging deficits.

    You mentioned a 13% charge rate. I'm not used to seeing charge rates expressed as a percentage. Usually it's in terms of C/N, so that'd be about C/7.7. Is that what the 13% means? The general recommendation for bulk charging is between C/5 and C/10, so that sounds about right. But I want to make sure I'm on the same page as you. [EDIT: I see in the spec for a Trojan L16P that the maximum charging current is "13% of the C/20 rate". What exactly does that mean? That seems really slow! It's a 420 Ah battery, so C/20 is 21, and 13% of that is a mere 2.73 amps? How can that be right?]

    C/20 (or C20) is the 20 hour discharge rate of a battery. If you have a 100 AH battery and discharge it at 5 amps, it will take 20 hours for the battery to go from 100% to 0% capacity (5 amps * 20 hours = 100 Amp*Hours). There are different discharge rates and ratings (C/5, C/8, C/24, C/100 etc.). The slower the discharge the more "apparent capacity" the battery has (a 100 AH C20 rate battery may have a 120 AH with a C100 rate). For our rules of thumb and basic calculations, the C20 discharge rate is "close enough" for solar work.

    The 13 rate of charge is also the 1/8 rate of charge (8 hour charge rate. 1/8 = 0.125 = 12.5% ~13%). We use the C20 battery capacity for our rules of thumb (100 AH * 13% rate of charge = 13 Amp charging current).

    The slower you charge, the more time it takes, and that's what concerns me. Even at the maximum rate, if you discharge the batteries enough, there might not be enough hours of sunlight in the day to recharge them, regardless of how many panels you have. Is there a rule of thumb to match actual hours of sunlight to charge rate and to depth of discharge? I don't want to plan to use a generator every day. It seems like with more batteries you can avoid a deep discharge, and with enough panels you can always recover because the charge rate won't be a limiting factor. When does it become the limiting factor, though?

    Normally, we suggest using 25% of capacity for a single day of bad weather. If you have 2x days of bad weather, it is going to take you a few days to bring the bank back to "full charge". Using tracking solar arrays does give you more "hours per day" plus more harvest--But is rarely done on an RV.

    Maximizing the battery life is still a consideration. I'm going to be full-time RV'ing for a year or more. It's the "more" I'm thinking about. I'm not sure how long this will last. Might be permanent. It'd be a shame to have to replace all of my batteries two years down the road (and buy twice as many) just because I didn't buy twice as many in the first place.

    Say your batteries have a 2,000 cycle life to 75% state of charge, and 1,000 cycle life to 50% SoC. If you "camp" for 6 month a year or ~180 Days, even a 1,000 cycle life battery will give you (divided by 180 days) 5.6 years of life. And for "cheap" golf cart batteries, 3-5 years is typically a "good life" (note that "hot batteries" >room temperature age faster than cold batteries below room temperature--Basically +10C/+18F cuts battery life by 1/2. -10C/-18F increases battery aging life by 2x).


    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    a propane refrigerator (with all of its faults) is usually a "better" solution.
    I was looking at those, or specifically AC/DC/LPG ones. I'm going to be using propane anyway for cooking, and possibly heat, so I'll have it. I also hear that the propane fridges are really efficient. It's hard to find exact numbers, though.

    Depending on lots of things, a propane fridge may use 1/2 to 1 lb of propane per day (hot weather, more propane). Some folks have added small low power computer fans inside fridge and outside fridge to increase air circulation to improve efficiency. You have to decide if upwards of 1 lb of propane per day is acceptable or not.

    Note that 120 VAC or 12 VDC power for propane refrigerators is horribly inefficient and will not usually work with solar power systems. If you use solar electricity for fridge--They need to be pump based, not absorption/ammonia based refrigerators.

    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • myocardia
    myocardia Solar Expert Posts: 118 ✭✭✭
    edited June 2018 #19
    BB. said:
    a propane refrigerator (with all of its faults) is usually a "better" solution.

    Note that 120 VAC or 12 VDC power for propane refrigerators is horribly inefficient and will not usually work with solar power systems. If you use solar electricity for fridge--They need to be pump based, not absorption/ammonia based refrigerators.

    Indeed. I have not heard of anyone using anything other than a Dometic-type fridge in their vans, and being happy with it. Dometic is the best-selling brand, and they are supposed to be of the highest quality (at least compared to their competition, mind you). Here is a link to the various sizes that they sell here in the US: https://www.amazon.com/Dometic-Electric-Powered-Cooler-Freezer/dp/B075J1NYT6/ref=sr_1_4?ie=UTF8&qid=1529696621&sr=8-4&keywords=dc+refrigerator+freezer+12v&th=1 Most people use the 35-50 liter sizes. Remember that you will be needing to add no ice at all, so a 35 liter size will have considerably more space than you would have with a 50 quart cooler, to which you are having to add ice.

    edit: These are supposed to use along the lines of 150ish watt hours per day (for the smaller models), assuming they aren't left in a closed-up van in Phoenix during July.
    The Dometic page says the 40 liter model uses 40 watts while running, for what that's worth: https://www.dometic.com/en-au/au/products/food-and-beverage/portable-refrigeration/fridge-freezers-and-coolers

    edit #2: Admittedly, if  you can find the space for them, you'd very likely use half the watts with one of these DC refrigerators that our host sells. It has 4.33" of insulation on all 6 sides. https://www.solar-electric.com/sundanzer50l.html

    DoD= depth of discharge= amount removed from that battery   SoC= state of charge= amount remaining in that battery
    So, 0% DoD= 100% SoC, 25% DoD= 75% SoC, 50% DoD= 50% SoC, 75% DoD= 25% SoC, 100% DoD= 0% SoC
    A/C= air conditioning AC= alternating current (what comes from the outlets in your home) DC= direct current (what batteries & solar panels use)
  • MysteriousFoundation
    MysteriousFoundation Registered Users Posts: 31 ✭✭
    Estragon said:
     Also, with 3 in parallel, each battery should be fused to prevent current from 2 feeding a faulty one.
    What would this look like? How large of a fuse? How does a fuse prevent this situation, but not blow during normal use or charging? Couldn't this also be an issue with 4 6V batteries in series-parallel?

    I think the Trojan L16 charging spec may be a typo; instead of "13% of the C/20 *rate*", it should read "13% of the *amp-hour capacity* at the C/20 rate".

    That certainly makes a lot more sense!

  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    It would look like a fuse on the positive terminal of each parallel connection.

    It would be sized to protect the smallest conductor in the circuit. Shouldn't blow in normal use, because in normal use, no conductors would be carrying enough current to blow the fuse. It should only blow in a (short) fault condition. My understanding is L.A. batteries usually fail with reduced capacity from sulfation, or open circuit from grid corrosion, but can also fail shorted (eg loose corroded material creates a short).

    It would be less of an issue with 4x6v because with 2 strings, the maximum current available is the 1 non-shorted string feeding a shorted one. With 3 12v batteries in parallel, you potentially have 2x current through the faulted string (battery), with 2 strings feeding the short in parallel.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • MysteriousFoundation
    MysteriousFoundation Registered Users Posts: 31 ✭✭
    edited July 2018 #22
    Estragon said:
    It would look like a fuse on the positive terminal of each parallel connection.

    It would be less of an issue with 4x6v because with 2 strings...
    I'm pretty sure I'm going to go with 4x Trojan T-105 6V batteries unless someone says it's really worth the money to go down to two large batteries. I just can't beat the price, even though they are technically more capacity than I need. I would prefer, say, 2x Trojan J305P-AC which are 330 Ah 6V, but where I can buy them locally, they are $299 with a $44 core charge each. That's $686. I can get the T-105's for $110 + $27 core charge each. Four of them is $548. If I add in $20 for three additional cables to wire them in series-parallel and $20 for two extra fuse holders, that's still only $588. $100 cheaper.

    The downsides are having to test more cells, take up a bit more space in the van, more weight, more connections...

    I dunno. It's only $100 and I'd have a much simpler setup. It's not like having four cheaper batteries makes it any cheaper if one fails a year later. I'd still have to replace all four, right? Because they'd all need to be the same age.

    At any rate, I came here with a fuse question that I still want to ask. You said, you'd fuse the positive terminal of each parallel connection. Do you mean each parallel string gets one fuse? I don't see exactly how that would work given the most common wiring diagrams I've found on the internet where the two string share one negative and one positive terminal. There simply isn't a way to add two fuses where one isn't shared.

    Here's one example:



    Since the positive end of the parallel strings share one terminal, you can't fuse each positive connection separately. You can only add the one at the blue dot between the two strings, and one at the green dot that's shared. Is this sufficient for protecting against issues from a short? I would think the amp value of the blue fuse would only need to be half of the green's value, since it's only taking one string's supply, and you want every fuse to be no larger than needed. The green fuse is shared by both strings, and I'd fuse there anyway, but it would be for the entire current load, not just half.

    The other way I've seen to wire this is to share a common post somewhere off the battery:


    In this case, I can see why it would make sense to use fuses at the blue dots, like you are suggesting, and then I don't even think I'd need another fuse since the two fuses in parallel will add up to the current limit I'd want. But this introduces an additional post and more connections.

    Is there something else I'm missing? Which way would you wire it? And what size fuses would you put where?

    Maybe the first example is exactly what you meant, and "each parallel connection" would mean the wire between strings. I was coming at it originally thinking you meant "each parallel string".

    Also, you don't worry about fusing the batteries in series?




  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    With two strings, string fusing is optional. With three or more, each series string should be fused so a short in one string doesn't get fed by current from multiple good strings. You'd still want a fuse (preferably a breaker) between the bank and attached devices (ie inverter, charge controller, etc), whether you had individual strings fused or not. Sizing of wire and breakers should be per device makers' specs.

    The second diagram is a better way of wiring in parallel IMHO, as it makes it more likely the resistance will close/identical between the two strings. Having multiple strings in parallel increases the risk of differing rates of charge/discharge in each, which left unchecked can lead to premature failure. It does introduce another connection, but the connection is shared, keeping whatever added resistance balanced.

    Personally, I'd probably go for the two strings of GC batteries. Two strings shouldn't be too bad to get and keep in balance. A bit more time to check SG and water more cells, but has the advantage that when a battery dies, you might be able to get by on one string for a bit. GCs are also probably easier to find in far flung locations than L16s should the need arise.

    If one battery died within a year and the others were fine (ie the one was likely defective), I'd just replace the one. If the others were suffering too (ie bad charging regime or whatever) and one just happened to die first, I'd replace the lot. Much more than a year though, it's probably better to replace all.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • mcgivor
    mcgivor Solar Expert Posts: 3,854 ✭✭✭✭✭✭
    Having a 2 parallel bank is generally not a problem, I use the first diagram setup with one main fuse and a circuit breaker for disconnect to inverter, having to water more cells encourages discipline, which results in a more closely monitored system, it's often complacency which comes to bite one in the ass, so to speak.
    1500W, 6× Schutten 250W Poly panels , Schneider MPPT 60 150 CC, Schneider SW 2524 inverter, 400Ah LFP 24V nominal battery with Battery Bodyguard BMS 
    Second system 1890W  3 × 300W No name brand poly, 3×330 Sunsolar Poly panels, Morningstar TS 60 PWM controller, no name 2000W inverter 400Ah LFP 24V nominal battery with Daly BMS, used for water pumping and day time air conditioning.  
    5Kw Yanmar clone single cylinder air cooled diesel generator for rare emergency charging and welding.