Battery bank sizing

Hairfarm
Hairfarm Solar Expert Posts: 225 ✭✭✭
Greetings,

I can't remember the last time I posted a question here. It's been awhile and now that I'm actually ready install a battery system I've suddenly realized that all of my perishable PV knowledge has, well, perished. Was hoping someone might lend a hand to help me fill in the blanks and keep my math on track.

Hardware for my 12 volt system:
Six 135 watt Kyocera Panels, 810 watts total / Maximum Power Current (lmpp): 7.63A / Maximum Power Voltage (Vmpp): 17.7V
Outback Flexmax 80 MPPT
Batteries might likely be Trojan T105 - 20 HR Rate 225 AH

For the panels, I'm going to wire three pairs of two in series, then parallel the three series/sets together to make 17.2 x 2 = 35.4 volts @ 22.89 a...is this right?

So, to figure my AH:
6 X 135W panels = 810 watts total
For a 12 volt battery system. Each panel puts out 7.63 amps normally.
6 x 7.63 = 45 amps
45 amps X 20 (extra safety buffer) = 900 amp hours of batteries

Or, maybe another way to arrive at the same point:
810W / 17.7v = 46A
46A x 20 (safety buffer) = 920

Pretty much the same right? Do the batteries HAVE to match the solar array Ah's?

My question is how many batteries (Trojan T105 - 20 HR Rate 225 AH) should I buy to make sure that I get maximum efficiency with no wasted PV panel power? Is that the right way to pose that question?

I'm thinking that I need 8 250ah batteries in four series/sets... and those four series/sets are paralleled. Each set of two is 250 amp hours at 12 volts. Thus I'll need 4 sets of two - eight batts @ 250ah x 4 = 1000ah (just rounded up)

This sound about right?

Also, what are the downsides to purchasing two larger batteries to make up my Ah's? For example, wiith the Surrette S-600's (6 Volts, 450 Amp-hours) I would only need 2 big batteries instead of 8 smaller Trojans T105's (20 HR Rate 225 AH).
http://www.solar-electric.com/suba6vo450am.html

The price for two Surrette S-600's (6 Volts, 450 Amp-hours) are $840 for 900 Ah and would conserve space.

The price for 8 Trojan T105 batteries would be $1208 for 1000ah.

Is there a reason to use a higher number of batteries to get my Ah than a smaller amount of larger batteries?

Ok, I officially have a headache now ;)

thanks,

Hairfarm
«1

Comments

  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Battery bank sizing
    Hairfarm wrote: »
    Greetings,

    I can't remember the last time I posted a question here. It's been awhile and now that I'm actually ready install a battery system I've suddenly realized that all of my perishable PV knowledge has, well, perished. Was hoping someone might lend a hand to help me fill in the blanks and keep my math on track.
    Is there a reason to use a higher number of batteries to get my Ah than a smaller amount of larger batteries?

    Ok, I officially have a headache now ;)

    thanks,

    Hairfarm

    Here is some aspirin:

    You do not want to put batteries or strings of batteries in parallel if you can help it. Unless you are super concerned about robustness to keep your system working with one or more bad batteries, your best bet will be to get the four Surrettes and put them in series. (Notice please that two 6 volt 450 AH batteries make a 12 volt 450 AH system, not 900HA. Double your battery cost to match.) That way you will still be putting batteries in parallel, but only two strings. This assumes that you have a good way to lift the four giant batteries into place and do not plan to move them by yourself.
    If you need to use smaller batteries or you want to avoid putting any in parallel at all, then get either three 4 volt batteries or six 2 volt batteries instead of four 6 volt.

    But a more serious question is why the 12 volt system? I see that you did not mention an inverter.
    If you will be using pure DC and will not have any heavy loads, then 12 volts may make sense, but if you have the opportunity to make a 24 volt system instead, now would be the time to do it.
    That will let you use smaller wire everywhere in the system, and will let you use a larger number of smaller batteries to make up the 24 volt 450 AH system.
    SMA SB 3000, old BP panels.
  • waynefromnscanada
    waynefromnscanada Solar Expert Posts: 3,009 ✭✭✭✭
    Re: Battery bank sizing
    inetdog wrote: »
    But a more serious question is why the 12 volt system? I see that you did not mention an inverter.
    If you will be using pure DC and will not have any heavy loads, then 12 volts may make sense, but if you have the opportunity to make a 24 volt system instead, now would be the time to do it.
    That will let you use smaller wire everywhere in the system.
    Have to agree. Early on I painted myself into a corner with 12 volts. I get along OK with it, but have to use very heavy, hard to handle cables, fuses etc. Wishing now I had gone 24 volt, but at the time, never dreamed how big my system would eventually become. I'm now at the very max for a 12 volt system, any expansion would be going backwards unless it was a second, separate 12 volt system with dedicated loads, which would be very inefficient cost and maintenance wise, compared to going 24 volts to begin with.
    Definitely something to consider.
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing
    (Notice please that two 6 volt 450 AH batteries make a 12 volt 450 AH system, not 900HA. Double your battery cost to match.)

    Oh yeah, you're right. My bad. Thanks for the suggestions too.

    I have a 12v Xantrex Prosine 2.0 that I basically got for free. I can't afford a second 24 volt inverter so that's why I'm going 12 volt with my battery system, for now anyway. It's just a vacation cabin, my panels are roof mounted, and my batteries are only three feet from my inverter and CC.

    A question:
    I realize that, theoretically, I could use just a single 12 volt battery and my Outback Flexmax80 CC would just "bleed" off the extra charging power from my 810 watts of panels and keep that single battery happily charged. But what is the maximum amount Trojan 225Ah batts that I could add? In other words is there a formula for how many batteries to add to a 12 volt system? Should the batteries Ah at least match the Ah of the panels, with 25-20% extra thrown in for a safety bufffer?

    I realize that question has a lot to do with my power "needs". But what is a good number to start with if I'm using the Trojan 225Ah batteries?

    Can I get away with 8 of them? And does anyone else overestimate their Ah needs by adding 15-20% buffer for rainy days etc? Too many questions, so little time.

    Hope this makes sense. The battery stuff just ain't clicking mentally for me right now.

    Hairfarm



    With my Outback Flexman80 CC, my batteries should be safe from overcharging right.
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Battery bank sizing
    Hairfarm wrote: »
    Oh yeah, you're right. My bad. Thanks for the suggestions too.

    I have a 12v Xantrex Prosine 2.0 that I basically got for free. I can't afford a second 24 volt inverter so that's why I'm going 12 volt with my battery system, for now anyway. It's just a vacation cabin, my panels are roof mounted, and my batteries are only three feet from my inverter and CC.

    A question:
    I realize that, theoretically, I could use just a single 12 volt battery and my Outback Flexmax80 CC would just "bleed" off the extra charging power from my 810 watts of panels and keep that single battery happily charged. But what is the maximum amount Trojan 225Ah batts that I could add? In other words is there a formula for how many batteries to add to a 12 volt system? Should the batteries Ah at least match the Ah of the panels, with 25-20% extra thrown in for a safety bufffer?

    With my Outback Flexman80 CC, my batteries should be safe from overcharging right.

    At least four different calculations enter into the balance between batteries and panels:

    1. Your panels and CC should be able to deliver at least a C/12 charge rate to FLA batteries. If it cannot do that you may not be able to get good gassing to prevent electrolyte stratification, and you will not be able to charge the batteries quickly enough to make up for a few days without sun. The rule of thumb that goes with this is that the panel watts going into an MPPT controller should roughly match the battery bank AH number for a 12 volt system.
    2. If your panels are considerably oversized for your battery bank (which can help to get full power of a longer time during the day, perform better in poor weather and recharge faster after a deep discharge, you still need to be able to limit the battery charging current to C/8 for FLA batteries. As long as your CC allows you to set a limit on the bulk amps, rather than trying to deliver full rated maximum current, (which the Outback can do), then you will not overcharge your batteries.
    3. The final thing is that your panel array has to be large enough to fully charge your batteries all the way into Float on a good day with the batteries discharged a normal amount. If you do not meet that, you will end up using your generator at least once a week to do a supplemental charge. This is most often done by using the generator early to get at least part way through Bulk and then let the panels deliver the lower current needed to finish. If you follow the watts = AH rule of thumb and get at least 2 solar hours per day in the worst months, you will probably be OK on this one.
    4. If you have loads which you can choose to run during the day when the panels are producing, you can justify a higher ratio of panel to battery than you would use if only charging the batteries for use later. The prime opportunity load is water pumping to fill a storage tank, but things like doing laundry and running the microwave can fall in this category too.
    Some folks with limited battery capacity try to supercool their refrigerator and freezer during the peak solar hours of the day to reduce the load on the batteries the rest of the day.
    SMA SB 3000, old BP panels.
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing

    Thank you inetdog for all of the great, specific information. It's basically what I was looking for.

    I'd like to ask a couple more questions of you:
    inetdog wrote: »
    Your panels and CC should be able to deliver at least a C/12 charge rate to FLA batteries.

    I'm not familiar with the C rating system of charging. Based on my specs below can you tell me if I meet the C/12 requirement that you stated?
    The rule of thumb that goes with this is that the panel watts going into an MPPT controller should roughly match the battery bank AH number for a 12 volt system.

    This is my calculation.
    6 X 135W (17.7v @ 7.63a) panels = 810 watts total
    For a 12 volt battery system. Each panel puts out 7.63 amps normally.
    6 x 7.63 = 45 amps
    45 amps X 20 (extra safety buffer for cloudy days) = 900 amp hours of batteries

    So, I would need roughly 900 amp hours of batteries, right? Is that what you meant by watts = AH rule of thumb?

    My charge controller is an Outback Flexmax80, I'm looking at 8 Trojan T-105 batts @ 225ah - (http://www.trojanbattery.com/Products/T-1056V.aspx)
    4 series/sets to get up to 12v then parallel to get my ah up to 900. But someone on this forum also told me that I should NOT parallel my batteries after putting them in series unless I wanted a robust system that would handle a bad battery in the system. But I wouldn't I need to parallel them to get my 900ah?

    Is this a sound set up?


    thanks so much!

    Hairfarm
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing

    Current you can expect from 810 Watts on an MPPT controller:

    810 * 0.77 (typical efficiency) = 623 / 12 Volt minimum battery Voltage = 52 Amps.
    That's enough for 520 Amp hours @ 12 Volt. A larger battery bank will suffer from lower peak charge current; it will not charge quickly and may not finish a complete charge cycle even on a good day.

    Check this against the panel harvest on a "4 hour" day:
    810 * 4 * 0.52 over-all system efficiency = 1684 Watt hours.

    1684 Watt hours on 12 Volts is roughly 140 Amp hours, which is 25% of 561 Amp hours which fits nicely the aforementioned 520 Amp hour bank (you want to be able to 'harvest' more power from the panels than is stored in 25% of the battery bank because the sun doesn't always shine).

    If you are using a PWM controller the maximum current available for charging is the cumulative Imp of the panels. The application for battery bank sizing is the same: 10% of capacity is a good target number to keep charging within the desired range.
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing
    Current you can expect from 810 Watts on an MPPT controller:

    810 * 0.77 (typical efficiency) = 623 / 12 Volt minimum battery Voltage = 52 Amps.
    That's enough for 520 Amp hours @ 12 Volt. A larger battery bank will suffer from lower peak charge current; it will not charge quickly and may not finish a complete charge cycle even on a good day.

    Check this against the panel harvest on a "4 hour" day:
    810 * 4 * 0.52 over-all system efficiency = 1684 Watt hours.

    1684 Watt hours on 12 Volts is roughly 140 Amp hours, which is 25% of 561 Amp hours which fits nicely the aforementioned 520 Amp hour bank (you want to be able to 'harvest' more power from the panels than is stored in 25% of the battery bank because the sun doesn't always shine).

    If you are using a PWM controller the maximum current available for charging is the cumulative Imp of the panels. The application for battery bank sizing is the same: 10% of capacity is a good target number to keep charging within the desired range.

    Ok. I'm going to just take your word for it. It seems to me that you also implied that I shouldn't drain my batteries more than 25%. I thought I could go to 40 -50%.

    I can't come up with a specific number of 225ah 6v (Trojan 105) batteries that will meet the 520ah battery bank need. I live in the Mojave desert in Zone 1 (http://www.wholesalesolar.com/Information-SolarFolder/SunHoursUSMap.html) with 6 hours of sun a day. I'd have to use at least 6 225ah 6v batts for 675 Ah. But you said that anything over 520 would be too much for my panels to come up with, right?

    thanks Cariboocoot
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing

    Adjusting the numbers a bit is not instant death. :D
    So you design a system to the middle-of-the-road numbers, and then hope you have enough margin to have a balanced system.

    This is another reason why you should start with the load calculations ad fit the battery bank to them, then come up with an array that will recharge them. Starting with the array limits the amount of battery you can charge and thus the amount of load you can supply.

    So working backwards, that 52 Amps from the array could support a maximum of 1040 Amp hours, for which it would be 5%. Note the big problem here; loads drawing while charging will reduce that rate, taking away from the charge and slowing the process even deeper. If it is slowed too much the batteries will not recharge in a reasonable amount of time and may stay at too low of a SOC for too long, resulting in early sulphation.

    Still with me?

    Thus your "ideal" battery bank may be 520 Amp hours (or slightly smaller) but if local conditions are good you could get away with up to 2X that. Not recommended, however, as it moves the spec close to the edge of functionality. Not enough sun = not enough charging = early battery death. If you could run 450 Amp hours (dual golf cart battery strings) you'd have a big margin.

    You want to average a 25% DOD as again this is a good balance between life cycles and power delivered. if occasionally you go 50% that's not a problem. if you average 30% it's not a problem. If you steadily go 50% you will shorten battery life even if you can fully recharge every day. it's not the end of the world, it just accelerates time to replacement.

    Now here's another problem; your location is downright hot. This will reduce panel Voltage (lower current with an MPPT) and battery capacity/life. So you're working with a double disability here; less panel power and less battery power.

    Is this making sense?
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Battery bank sizing
    Hairfarm wrote: »
    But someone on this forum also told me that I should NOT parallel my batteries after putting them in series unless I wanted a robust system that would handle a bad battery in the system. But I wouldn't I need to parallel them to get my 900ah?

    read why parallel batteries cause problems: http://forum.solar-electric.com/showthread.php?14674

    There are solutions if you really need that much battery capacity. One solution is to use a higher system voltage... eight of those T-105 in series (48 volt) holds as much energy as eight of them in a series-parallel arrangement (12 or 24 volt). The other solution is to use different batteries with higher ah capacities per cell.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing

    Hi Cariboocoot,

    First, I realize that working backwards was not ideal, I totally get that. But I bought all 6 of the Kyocera 135 watt panels for $200 each at an auction here in the desert and got the Xantrex Prosine 2.0 for $50 from a guy on Craigs list who thought it was broken. It wasn't. I hooked it up to a 12 volt deep cell marine battery and everything checked out. It was too good of a deal to pass up so I grabbed them while I could. But anyway, that's why I had the panels first. And you're correct. It's forcing me to work backwards.
    Is this making sense?

    Mostly. I just had to read it 13 1/2 times ;) You have given me a lot of good info. Here's my takeaway from what you suggested. And please let me know if I'm on track here in a general sort of way.

    It seems to me that using your recommended 450 Amp hours would almost ensure that my batteries would always stay charged using my panel configuration. In other words, limiting myself to 450ah would sort of keep me honest, in a way, by making sure enough power was always incoming from the panels to the batteries. Too many battery amp-hours would strain the charging ability of the panels.

    That said, some of my options?:

    1) I would need to, roughly, double the amount of panels I have now (which is 810 watt) to effectively use the 1040 amp hours that I originally calculated.

    2) Or, I could just install 1040 amp hours of batteries and keep an eye on my charge controller to make sure that the panels keep up with the batts and maybe I'll get lucky and not use as much power as I thought. (Is that the sound of laughter I hear?) The only real heavy duty appliance I have is an AC 1/2hp water pump, and that will be converted to a DC pump someday. But it only runs for 2-4 minutes at a time @ 4-5 times per day to fill up my 80-gal pressure tank. Also, I can control when to fill it up by keeping the AC pump wired to a power strip and only turning it on when all other loads are off. My cooking and refrigeration will be propane, my heating will be a wood stove (not pellet), and my evaporative cooling in the summer is a 12vdc powered "swamp cooler" that only uses about 8 amps. I posted a vid here of that project: https://www.youtube.com/watch?v=gcWIlm1CZQA (Sorry for the shaky video)

    3) Or, if I find that my batteries are not being fully charged by my panels, I can supplement my 6 panels with my Honda EU200 generator, correct? My Prosine 2.0 inverter is a charging inverter so this is an option for me.

    Yes, I do have lot's of sun and, yes, it does get up to 112-115, but that's only for about 14 weeks. peak, then it's not bad. I live in the "high" desert so we're not like Palms Springs or other areas which get to 120+.

    Someone on youtube gave me the good idea of using a broken refrigerator to hold my batteries - http://www.youtube.com/watch?v=dzufK2q9N-c
    A larger refrigerator should hold roughly 8 (or less) Trojan 225 ah (T-105) batts. It would be insulated from extreme desert heat, water proof, lockable and in the shade. I can use a hole saw to cut vents in the back for charging ventilation and to route cables to my inverter and CC. I would also stack hay bales around it for added insulation, much like a miniature "Earth Ship" hay-bale enclosure. The temp inside would then much lower than ambient. I think this should work for a hot climate battery box. We'll see.

    As far as the panels themselves getting hot and decreasing output there's not much I can do about that except wait out the summer, get more panels, or supplement with the generator. Or maybe I can put them in the shade ;)

    Questions for you:

    Do some of those options sound ok?

    How long does it take a 2000 watt generator to charge 1040 ah of batteries? Assuming the batteries are close to 50% discharge. Is there a formula for that? Do you use your generator to charge your batteries ever?
    So working backwards, that 52 Amps from the array could support a maximum of 1040 Amp hours, for which it would be 5%.

    What is the 5% part of that statement referring to? 5% of what exactly?
    read why parallel batteries cause problems: http://forum.solar-electric.com/showthread.php?14674

    There are solutions if you really need that much battery capacity. One solution is to use a higher system voltage... eight of those T-105 in series (48 volt) holds as much energy as eight of them in a series-parallel arrangement (12 or 24 volt). The other solution is to use different batteries with higher ah capacities per cell.

    --vtMaps


    I will read that link later today. Thanks for the link vtMaps.

    I know I have lot's of questions, but my old high school band teacher once said, "Never be afraid to ask stupid questions". He would be so proud of me right now ;)


    Thanks for the help.

    Hairfarm.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing

    A couple of points first. If you live at high elevation your panels will work a little better than expected. The temperature is still an issue, as both panels and batteries 'generate' heat. They both work best at around 20-25 Celsius, so the higher the ambient the higher the operating temperature. Hyper-insulating the batteries may even work against you, trapping the heat produced by current flow. Sometimes it's tough to get the right balance of ventilation and insulation (just like with a house).

    Another thing: the 450 Amp hours @ 12 Volts is roughly the same stored capacity I have (232 Amp hours @ 24 Volts) which gives me 2.5 kW hours on a good day. My array is slightly smaller @ 700 Watts. I shift the water and septic pump to come on only during daylight when the batteries are full. This allows a bit more power harvest than if it was all on battery-demand. Otherwise some of the power potential from the panels is lost because there's simply no place for it to go once batteries are charged.

    I mention this because it calls into question whether or not you need that very large battery bank.

    How long does it take a 2kW gen to charge 12kW hours of battery? I wish it were as simple as 12/2 but it isn't. The first bit goes quickly. The last bit not so much so. Because the efficiency drops off drastically, the usual practice is to use the gen to bulk charge first thing if necessary and then let the panels finish the job if they can. You want to avoid running a generator hours on end with a small load on it just to complete a full charging cycle. But sometimes it is necessary.

    The 5% charge rate is the bottom of the range recommended for most batteries: 5% of the 20-hour Amp hour rating. It is often expressed as C/20 where 'C' is the capacity. So 1040 Amp hours has a minimum peak charge rate of 52 Amps. A couple of issue with that: it's net, not gross, so any loads drawing while charging detract from the rate and it can take an inordinately long time to recharge at a slow rate. If it did work out as simple math (which it doesn't) and you were trying to replace half the bank capacity at that steady current you'd see it would be over 10 hours to do it. Pretty hard to accomplish in a day with only 5 or 6 hours of good sun in it.

    If I were you I'd be giving serious thought to the load side of things before buying more batteries, panels, whatever. Expansion isn't easy, but at the same time you don't want to spend more $ than you need to.
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing

    Greetings.

    A couple of points first. If you live at high elevation your panels will work a little better than expected. The temperature is still an issue, as both panels and batteries 'generate' heat. They both work best at around 20-25 Celsius, so the higher the ambient the higher the operating temperature. Hyper-insulating the batteries may even work against you, trapping the heat produced by current flow. Sometimes it's tough to get the right balance of ventilation and insulation (just like with a house).

    I'll get a thermometer to monitor the heat in the fridge. If it gets too hot, I guess I can continue to "vent" it with a larger hole/s and use hardware cloth to cover it to keep out the rodents. The fridge seemed like a good idea because the batteries can NOT live inside the cabin. I'm not comfortable with that. A neighbor of mine built a house made of hay bails and clay and it stays within a pretty good temp range in both the winter and summer. In the summer he doesn't need AC. His place uses convection cooling too. He also recommended that I stack hay bales around the fridge as well as my 40' container to insulate in the summer as well. The fridge will live on the shady side of the cabin too. In fact it will be situated directly up against the outside of the cabin wall. On the opposite side of that same cabin wall, inside the cabin utility closet, are where my inverter and charge controller will be installed. My batteries should be no more than 2-3 feet from my inverter and charge controller through the cabin wall. In any case, I'll make it a point to keep checking the thermometer during the summer and try to find the "sweet spot" temperature for the batteries. Might even consider a fan that comes on when charging is taking place.

    How long does it take a 2kW gen to charge 12kW hours of battery?

    What is the formula for calculating kW hours from my system?
    The 5% charge rate is the bottom of the range recommended for most batteries: 5% of the 20-hour Amp hour rating. It is often expressed as C/20 where 'C' is the capacity. So 1040 Amp hours has a minimum peak charge rate of 52 Amps.

    I'm struggling with that part. Are you saying that with my panel array @ 810 watts is only charging at a 5%? What should my charge rate be ideally? Am I thinking about this correctly? With my panel configuration, 22.89 amps is coming out at 35.4 v.

    Attachment not found.
    If I were you I'd be giving serious thought to the load side of things before buying more batteries, panels, whatever. Expansion isn't easy, but at the same time you don't want to spend more $ than you need to.

    Cariboocoot - basically, you've convinced me to pull out my Kill-o-watt and try to get an idea of what I might use at the cabin and test devices here in my home. I know it might be hard to predict 100% what my needs are but it will still give me a better idea of what to expect. Maybe I can try to stop working, as you said earlier...backwards;)

    Also, there's another whole set of issues to consider. During the hot weeks, the cabin might be empty for a few days at a time so I'll need to set up timers to keep some loads on the batteries occasionally, because I know that batteries need loads and shouldn't stay "dormat" for long.
    Read why parallel batteries cause problems: http://forum.solar-electric.com/showthread.php?14674

    There are solutions if you really need that much battery capacity. One solution is to use a higher system voltage... eight of those T-105 in series (48 volt) holds as much energy as eight of them in a series-parallel arrangement (12 or 24 volt). The other solution is to use different batteries with higher ah capacities per cell.

    --vtMaps

    In regard to vtMaps link that he sent me. Is a string when two batteries are wired in series? If so, then six 225ah batteries wired in three groups of two (12volts) then wired in parallel has three strings in it, right? And from the research I've done it seems that three strings should be the max. And should each string be wired together so that they have a common positive and a common negative wiring point. Instead of a "leap frogged" connection. Will this allow all of the individual batteries to be charged and dis-charged the same. Is that a concern?
    And, you should, with parallel strings, have a fuse/breaker per string to prevent the other strings from feeding excessive current into a shorted string.

    What would a diagram of that look like?

    thanks,

    Hairfarm
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing
    Hairfarm wrote: »
    What is the formula for calculating kW hours from my system?

    Amp hours * Volts = Watt hours. In the case of calculating expected usable AC Watt hours you have to adjust the battery capacity to the desired Depth Of Discharge or its maximum. In this case take 1040 down to 520 for 50% DOD, multiply by system Voltage to get the stored Watt hour capacity 520 * 12 = 6240. Then factor in the inverter's consumption and efficiency ratio to get a number that will approximate the usable AC Watts hours. This is typically 85%: 5.3 kW hours AC. I round a lot when giving short answers because we're not dealing with specifics and the actual system will vary, inevitably going down in capacity over time. It's best to anticipate this.


    I'm struggling with that part. Are you saying that with my panel array @ 810 watts is only charging at a 5%? What should my charge rate be ideally? Am I thinking about this correctly? With my panel configuration, 22.89 amps is coming out at 35.4 v.

    Simple formula for determining the 5% charge rate: Current * 100 / Amp hour capacity.
    So 22.89 * 100 / 1040 = 2.2%
    But if you use those panels with an MPPT controller (which you should because their Vmp is much higher than the system Voltage) you will be able to convert much of the 'extra' Voltage (that which is above battery Voltage after system losses) into additional charge current. It looks something like this:
    810 Watts * 77% typical efficiency / 12 Volts system nominal = 52 Amps And so:
    5200 / 1040 = 5%
    Understand that high ambient temperatures increase panel operating temperatures (by reducing the transfer of heat from panel to air) and decrease Voltage so your peak current may be lower. Likewise cold temps will increase Voltage. Changes in other atmospheric conditions will raise or lower panel output. If you try to count absolutely on 5% based on that formula you will likely be disappointed. For one thing loads drawing off current reduce the actual charge rate.
    Cariboocoot - basically, you've convinced me to pull out my Kill-o-watt and try to get an idea of what I might use at the cabin and test devices here in my home. I know it might be hard to predict 100% what my needs are but it will still give me a better idea of what to expect. Maybe I can try to stop working, as you said earlier...backwards;)

    This is a very good idea. Even if it's not the identical equipment it will be a lot closer than just building a system and hoping it meets needs.
    Also, there's another whole set of issues to consider. During the hot weeks, the cabin might be empty for a few days at a time so I'll need to set up timers to keep some loads on the batteries occasionally, because I know that batteries need loads and shouldn't stay "dormat" for long.

    Probably not an issue. Mine are dormant through Winter; no loads, panels & controller active. They survive fine. But they are also cool then, which slows down the chemical reaction. Like putting them to sleep cryogenically. :D
    In regard to vtMaps link that he sent me. Is a string when two batteries are wired in series? If so, then six 225ah batteries wired in three groups of two (12volts) then wired in parallel has three strings in it, right? And from the research I've done it seems that three strings should be the max. And should each string be wired together so that they have a common positive and a common negative wiring point. Instead of a "leap frogged" connection. Will this allow all of the individual batteries to be charged and dis-charged the same. Is that a concern?

    Yes, a 'string' whether of batteries or panels is two or more in series. Four parallel strings is really the limit, and at that point you're asking for trouble as the resistance of any two is unlikely to ever be the same.

    Smart Gauge's wiring diagrams: http://www.smartgauge.co.uk/batt_con.html (Shows individual batteries; think of a string as a higher Voltage battery in the same spot.)
    Method #1; No, no.
    Method #2; works best with just two strings.
    Method #3; my preference for 3 or four strings.
    Method #4; technically feasible, but difficult for most to understand or execute.

    About those panels + breakers. You'd be making panel strings like this:

    (-)PANEL(+)---(-)PANEL(+)---BREAKER---(+)

    And then the parallel connections are between all the first panel (-)'s and the breaker (+)'s.
  • bmet
    bmet Solar Expert Posts: 630 ✭✭
    Re: Battery bank sizing

    Cariboocoot,

    you wrote:

    "810 * 0.77 (typical efficiency) = 623 / 12 Volt minimum battery Voltage = 52 Amps."

    Since this is a battery-based system, wouldn't the 0.52 multiple be appropriate?

    810 * 0.52 = 421 / 12V = 35.1 Amps. Does the 0.52 multiple include real-world panel output (instead of STC)?
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing
    bmet wrote: »
    Cariboocoot,

    you wrote:

    "810 * 0.77 (typical efficiency) = 623 / 12 Volt minimum battery Voltage = 52 Amps."

    Since this is a battery-based system, wouldn't the 0.52 multiple be appropriate?

    810 * 0.52 = 421 / 12V = 35.1 Amps. Does the 0.52 multiple include real-world panel output (instead of STC)?

    It's the difference between looking at panel output for charging batteries (77% efficient conversion from nameplate Watts to output from charge controller) and looking at the over-all system's efficiency at delivering usable AC Watt hours (nameplate Watts * hours of good sun * 52% efficiency end-to-end).

    As always, both deratings are merely 'typical' guidelines, not absolute values. On any given system some days are better and some worse.
  • bmet
    bmet Solar Expert Posts: 630 ✭✭
    Re: Battery bank sizing

    Thanks for making that distinction. The rule of thumb guides are always easier for me to remember, the specific segments get lost till I have to look them up again :confused:
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing

    I should have added that the decrease in over-all efficiency is due to have to store at least some of the power in batteries (losses due to needing more power to bring them up to full than you can get back out) and inverter conversion of DC to AC (usually 90% +/-). Hence 77% efficiency from panels to charge controller output becomes 52% efficiency "at the outlet".

    With a standard GT system, the output in AC Watts is about the same 77% efficiency because there are no batteries taking their 20% toll. :roll:
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing

    Whew. I feel like I owe you a consulting fee Cariboocoot ;) I've actually been printing out copies of these posts for my installation in two weeks. I really do appreciate all of the help you, and others, have given. I never really had "math" brain, so it takes while to for this stuff to sink in. Sometimes I wonder if I'm nuts for spending so much money and time to attempt installing this system myself. I know a solar consultant nearby, but he want $75 an hour. And he's kind of an a**. That said, your helpful posts and tons or internet research, and two dvd's have helped.
    Simple formula for determining the 5% charge rate: Current * 100 / Amp hour capacity.
    So 22.89 * 100 / 1040 = 2.2%
    But if you use those panels with an MPPT controller (which you should because their Vmp is much higher than the system Voltage) you will be able to convert much of the 'extra' Voltage (that which is above battery Voltage after system losses) into additional charge current. It looks something like this:
    810 Watts * 77% typical efficiency / 12 Volts system nominal = 52 Amps And so:
    5200 / 1040 = 5%

    I follow all of that except for this: "5200 / 1040 = 5%". Where did you come up with 5200. Did you simply add two zeros to 52 amps from the previous calc?

    Smart Gauge's wiring diagrams: http://www.smartgauge.co.uk/batt_con.html (Shows individual batteries; think of a string as a higher Voltage battery in the same spot.)
    Method #1; No, no.
    Method #2; works best with just two strings.
    Method #3; my preference for 3 or four strings.
    Method #4; technically feasible, but difficult for most to understand or execute.

    I actually was reading that link last night. I found it on another thread discussing why not to parallel more than three strings. I was originally planning to use #2, but from your post it sounds like #3 would be better.
    About those panels + breakers. You'd be making panel strings like this:

    (-)PANEL(+)---(-)PANEL(+)---BREAKER---(+)

    And then the parallel connections are between all the first panel (-)'s and the breaker (+)'s.

    Ok. I understand the diagram. And apparently all of my wire needs to be the same length and size too, correct?

    I've deiced to downsize my intended battery system, and use 6 Trojan 225ah (675ah) instead 8 like I originally planned. I'll see how that setup works and it will save me more $$ too. They will be wired: 2-in-series, 3 strings, then paralleled which meets the "no-more-than-three-paralleled-strings Rule of Thumb".

    Big question: what type of connectors would I need to collect all three positives and all three negative wire using the #3 wiring example in the link you sent me? I've provided a picture and added red circles around the two connection points I'm referring to. In other words, how do I convert the two wires (pos and neg) from the charge controller into 6 wires - 3 for positive and 3 for negative as shown in the image below? Seems like the wires should be installed after the positive fuse for the CC. I can't seem to find any pictures (not diagrams) on line showing a connection like this. Can you please provide a link to a product, if any? Hope that makes sense.

    Attachment not found.

    Wife says I have to get off of the computer now and fix the fence.

    :)

    Hairfarm
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing

    I found this today:
    Attachment not found.

    I couldn't say if it's helpful or not. The numbers used in the example are not my numbers, but for example only. One would have to enter their own numbers.

    Ok, I GOTTA get that fence fixed now.

    Hairfarm
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing
    Hairfarm wrote: »
    I follow all of that except for this: "5200 / 1040 = 5%". Where did you come up with 5200. Did you simply add two zeros to 52 amps from the previous calc?

    The 52 Amp potential peak output of the charge controller multiplied by 100 so that the end result of the calculation is not expressed as a decimal: 52 * 100 = 5200 / 1040 = 5 You can leave the 100 factor out and get "0.05" as an answer; you just have to remember that as a percentage it is 5, not "0.05%"!
    I actually was reading that link last night. I found it on another thread discussing why not to parallel more than three strings. I was originally planning to use #2, but from your post it sounds like #3 would be better.

    I think it works better that way. One day I shall draw out battery wiring expressed as resistance and post it ... and confuse anyone who looks at it I suppose. :p
    Ok. I understand the diagram. And apparently all of my wire needs to be the same length and size too, correct?

    Actually for panel wiring the length is not so critical as with batteries. The Voltage is higher and they are a current source, and so the current will not differ from one to another regardless. However it is technically possible to have so much wire on a panel that it does reduce the Voltage below usable level. Just not very likely. With panels the more critical wiring is from the combiner to the charge controller as that is usually the longest run and will be carrying the most current.
    I've deiced to downsize my intended battery system, and use 6 Trojan 225ah (675ah) instead 8 like I originally planned. I'll see how that setup works and it will save me more $$ too. They will be wired: 2-in-series, 3 strings, then paralleled which meets the "no-more-than-three-paralleled-strings Rule of Thumb".

    That would give a better charge rate (about 7%). Just be sure you have enough battery for the loads (about 4kW hours stored there).
    Big question: what type of connectors would I need to collect all three positives and all three negative wire using the #3 wiring example in the link you sent me? I've provided a picture and added red circles around the two connection points I'm referring to. In other words, how do I convert the two wires (pos and neg) from the charge controller into 6 wires - 3 for positive and 3 for negative as shown in the image below? Seems like the wires should be installed after the positive fuse for the CC. I can't seem to find any pictures (not diagrams) on line showing a connection like this. Can you please provide a link to a product, if any? Hope that makes sense.

    You run wires from the batteries to common connection points/bus bars. The charge controller and inverter connect to these points. The poor man's way is insulated bolts where the wires come together (mounted firmly in wood, use stainless bolts). It is important to have proper circuit protection on all these circuits. This would ideally be three Blue Sea battery post fuses to protect the runs from each battery string to the common point, a fuse/breaker for the wire from there to the charge controller, and another for the wire to the inverter.

    Fix that fence so the tigers don't get in! :D
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing

    Oh and don't fall in to the "days of autonomy" trap; you can end up building too much of a system for your needs. Unless you have particularly cloudy conditions, go for the 25% DOD and know you've got another 25% available for "day two". Day three of no sun = start the generator.

    Otherwise you may buy more battery and panel than you need. Generators are much cheaper occasional recharge sources than a 3X system size!
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing
    That would give a better charge rate (about 7%). Just be sure you have enough battery for the loads (about 4kW hours stored there).
    Are you saying that 675ah (my total ah) translates into 4kW hours?

    Well, back to that kill-a-watt device again. I do know we already average roughly 16.03 kWh a day...ugh. But then I have two massive UPS units and other various phantom loads, three desktop computers, etc. Time to really pare down if we're going to try living off grid. My wife and I finally convinced our respective employers to let us both work from home full time. They've finally agreed after months of asking (whining). I put together a proposal for them to show that it could be done. So we plan to live in the cabin during the cooler Fall/Winter/Spring months, and migrate to Big Bear mountain for the 4 months of warmer weather when it's 110 in the desert. The nice thing is that Big Bear town and the location of our cabin are only 1 hour apart from each other, so it's not a problem for me go check up on the PV system once a week or so. Down the mountain, up the mountain.
    You run wires from the batteries to common connection points/bus bars. The charge controller and inverter connect to these points. The poor man's way is insulated bolts where the wires come together (mounted firmly in wood, use stainless bolts). It is important to have proper circuit protection on all these circuits. This would ideally be three Blue Sea battery post fuses to protect the runs from each battery string to the common point, a fuse/breaker for the wire from there to the charge controller, and another for the wire to the inverter.

    So a charge controller and inverter can share the same buss, huh? I remember I saw a diagram someone had posted who had used buss bars. I suspect I would still need the 4/0 wire to connect my batteries to the buss bar instead of directly to my inverter and then from the buss bar to my inverter I would need two more 4/0 wires, all under 8 feet max (according to NAWS) I presume?

    Btw, how does the charge controller know to only feed solar panel power into the batteries (and not the inverter) if they're all three (batts, inverter, and cc) hooked up to a common bus? The buss is just a dumb piece of metal. Why not just go straight from the batteries to the inverter with a 4/0 wire and use the buss to only feed the batteries from the CC? Does that make sense?

    I found these updated documents showing the use of a buss bar. I'm pretty sure these have circulated on the forum before.
    Attachment not found.Attachment not found.

    It's headache time again.:confused:

    Just a warning, I'm probably going to keep asking questions until the answers stop coming. But look on the bight side, I'll be the guy that helps push your posting number to 10,000!:D
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Battery bank sizing
    Hairfarm wrote: »
    Are you saying that 675ah (my total ah) translates into 4kW hours?

    I am a little lost--But if I have the numbers right, it would look like:

    12 volts * 675 AmpHours * 0.50 maximum battery discharge (long life) = 4,050 Watt*Hours = 4.05 kWH
    Well, back to that kill-a-watt device again. I do know we already average roughly 16.03 kWh a day...ugh. But then I have two massive UPS units and other various phantom loads, three desktop computers, etc. Time to really pare down if we're going to try living off grid. My wife and I finally convinced our respective employers to let us both work from home full time. They've finally agreed after months of asking (whining). I put together a proposal for them to show that it could be done. So we plan to live in the cabin during the cooler Fall/Winter/Spring months, and migrate to Big Bear mountain for the 4 months of warmer weather when it's 110 in the desert. The nice thing is that Big Bear town and the location of our cabin are only 1 hour apart from each other, so it's not a problem for me go check up on the PV system once a week or so. Down the mountain, up the mountain.

    A good sized desk top computer running at full speed and a nice monitor can run nearly 300 watt load. Run for 10 hours per day * 3 computers:
    300 watts * 3 computers * 10 hours = 9,000 WH = 9.0 kWH per day

    Add losses for UPS systems, printer, network, etc.---It would not be hard to believe 8-10 kWH per day for your computer systems/network.

    Plus another 1.2 kWH for a fridge, and another 2-4 kWH for lights and fans/radio/TV/Sat Receiver/DVR/Water pumping/washer/etc. And you are pretty close to 16 kWH per day.

    I suggest going after the big stuff (at least at the beginning). Unplugging cell phone chargers and even avoiding microwave is not where you are going to save power at the levels you need.

    Computers, either laptop at 20-30 watts (includes display and battery backup--own UPS system--really high end laptops can consume upward of 60 Watts). You have already cut your potential computer energy usage to 1/10th, and your overall energy usage from ~16 kWH to less than 9 kWH per day.

    It is, frequently the smaller stuff than run for many hours per day vs the big stuff you only run for a few tens of minutes per day (computers/refrigerators at 100-300 watts vs microwave or water pumping at 1kW or more).
    So a charge controller and inverter can share the same buss, huh? I remember I saw a diagram someone had posted who had used buss bars. I suspect I would still need the 4/0 wire to connect my batteries to the buss bar instead of directly to my inverter and then from the buss bar to my inverter I would need two more 4/0 wires, all under 8 feet max (according to NAWS) I presume?

    Yep--But we design to loads. So the exact wire diameter and maximum lengths really depends on your actual designed power needs and voltage of the battery bank (you have been talking about a 12 volt system, but at these levels, you probably should be looking at a 48 volt system--1/4 the DC amperage for smaller wires and less money spent on charge controllers, plus less voltage drop so easier to run longer DC wiring when needed).
    Btw, how does the charge controller know to only feed solar panel power into the batteries (and not the inverter) if they're all three (batts, inverter, and cc) hooked up to a common bus? The buss is just a dumb piece of metal. Why not just go straight from the batteries to the inverter with a 4/0 wire and use the buss to only feed the batteries from the CC? Does that make sense?

    In theory, it does not care where the current flows. The Charge Controller simply dumps all available current from the solar array to the battery bank until the battery reaches the "absorb" set point (say 14.5 volts). Then the charge controller holds the battery voltage at 14.5 volts until the battery is "full" (could be end charging amps, absorb timer, or combination). Once the battery is "full" the charge controller drops back to "float" at ~13.6 volts. The battery draws very little current and the loads draw whatever current is required (and the charge controller releases enough current to keep the system voltage at 13.6 volts).

    The details do get a bit more muddy--The "end amps" setting (1-3% of battery bank AH capacity--Say 100 AH, that would be ~1-3 amps ending current). There are either no or one charge controller setup that has the ability to measure the battery current (using a battery monitor shunt). The standard charge controller only knows how much current it is outputting--And does not "know" if the load or the battery is consuming the current--So way there is usually an Absorb Timer to stop charging after 2-6 hours.

    Your cabin power system is not much different than your car's charging/power system. The alternator holds the charging voltage while the engine is running, and the battery supplies surge current and power when the engine is not running.
    I found these updated documents showing the use of a buss bar. I'm pretty sure these have circulated on the forum before.
    Attachment not found.Attachment not found.

    It's headache time again.:confused:

    Just a warning, I'm probably going to keep asking questions until the answers stop coming. But look on the bight side, I'll be the guy that helps push your posting number to 10,000!:D

    The bus bars are nothing special. Generally it is just a large metal plate (or two plates for +/- connections) where it is easy to attach all of the various power cables (large diameter for battery bank and inverter, medium for charge controllers, and smaller gauge for light loads for LED lighting, etc.).

    A bus bar may be a big copper plate with holes, or a brass bolt and nut where you attach all of the ring lugs for the wiring, or it may even be the bolts on the +/- battery terminal.

    -Bill "my post count is already >> 10,000" B. :p
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing

    Thanks for the feedback BB. I will add it to my ever growing information file on pv systems.
    Yep--But we design to loads. So the exact wire diameter and maximum lengths really depends on your actual designed power needs and voltage of the battery bank (you have been talking about a 12 volt system, but at these levels, you probably should be looking at a 48 volt system--1/4 the DC amperage for smaller wires and less money spent on charge controllers, plus less voltage drop so easier to run longer DC wiring when needed).

    How would I power a custom made 12 volt swamp cooler and LED's? And are there 24-48 shallow well pumps available? If I can still power the 12 volt swamp cooler and LED's with a 24 volt system then I'd sell my 12 volt Xantrex inverter on ebay and go 24 or 48 volt in a second. I know that would be ideal. My swamp cooler uses a 12 volt automotive radiator fan and 12 volt 400gph sump pump. I would only need this in the summer though. But I would still need a way to get 12 volts DC to it in the summer. What are my options for DC 24/48 volt shallow well pumps? I do know they're pricey.

    If I did do a 24 or 48 volt system, the only way I know of getting 12 volts to my swamp cooler would be a device like this: http://www.amazon.com/Kendrick-Converter-Output-Supply-Controller/dp/B000B7A1W4/ref=pd_sxp_f_pt

    Seems like a crappy way to make 12 volts though. DC>AC>DC again. I guess I would have to set up an isolated 12 volt system. One that would run my 12Vdc swamp cooler and LED lighting. Seems like I'm just back to square one with that plan.
    The bus bars are nothing special. Generally it is just a large metal plate (or two plates for +/- connections) where it is easy to attach all of the various power cables (large diameter for battery bank and inverter, medium for charge controllers, and smaller gauge for light loads for LED lighting, etc.).

    Are the buss bars one size fits all?

    thanks so much!

    Hairfarm
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Battery bank sizing
    Hairfarm wrote: »
    How would I power a custom made 12 volt swamp cooler and LED's? And are there 24-48 shallow well pumps available? If I can still power the 12 volt swamp cooler and LED's with a 24 volt system then I'd sell my 12 volt Xantrex inverter on ebay and go 24 or 48 volt in a second. I know that would be ideal. My swamp cooler uses a 12 volt automotive radiator fan and 12 volt 400gph sump pump. I would only need this in the summer though. But I would still need a way to get 12 volts DC to it in the summer. What are my options for DC 24/48 volt shallow well pumps? I do know they're pricey.

    It is a big problem. Folks start "small" with a 12 volt system and keep growing it over time. At some point, a choice needs to be made to keep the 12 volt system or scrap and do a 24 or 48 volt system. This is one reason I suggest that people user 120 VAC and an AC inverter instead of DC Direct connections which lock you into a specific voltage.

    Obviously, there is the old 24/48 volt system with a down converter to 12 VDC. The little 15 amp 12/24 volt Morning Star MPPT controller has been used by a few people here to down convert from a 24/48 volt battery and charge a small 12 volt battery bank to operate those DC loads.

    You could leave the system as is and run your 12 VDC devices (battery bank/solar array/charge controllers/etc.) and start a new system to support larger needs--And convert any loads from your 12 VDC system to AC power.

    Or--you maintain your existing system and realize its limitations. If it works well--why not. Note that the largest 12 volt MPPT charge controllers are around 80 amps or so maximum--So when your solar array goes over ~800-1,000 watts--you are looking at needed a new charge controller. Etc...

    If I did do a 24 or 48 volt system, the only way I know of getting 12 volts to my swamp cooler would be a device like this: http://www.amazon.com/Kendrick-Converter-Output-Supply-Controller/dp/B000B7A1W4/ref=pd_sxp_f_pt
    Seems like a crappy way to make 12 volts though. DC>AC>DC again. I guess I would have to set up an isolated 12 volt system. One that would run my 12Vdc swamp cooler and LED lighting. Seems like I'm just back to square one with that plan.

    Yea... Your best best is to spend some time designing and pricing out several options (two systems, two batteries with one charged by the new/larger system, converting to higher voltage 120 VAC, etc.) and see what works best for you.
    Are the buss bars one size fits all?

    Bus bars are whatever size you need... They can be home made, purchased, or even take some 3/4" or 1" copper water pipe, flatten it, and drill holes for bolts and nuts. Since they tend to be copper or brass--Bus bars are expensive.

    One or two people have made bus bars out of aluminum and been very happy--But I am a bit leary--Aluminum froms an oxide within seconds of being exposed to air. Using Star Washers and grease to keep out air can be used to make longer term reliable connections--But I would avoid aluminum if I could.

    Poster 2manytoyz has a website where he has documented various aspects of his DIY solar power system... Lots of pictures:

    http://2manytoyz.com/altpowerboard.html

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing
    From inetdog: At least four different calculations enter into the balance between batteries and panels:

    1. Your panels and CC should be able to deliver at least a C/12 charge rate to FLA batteries. If it cannot do that you may not be able to get good gassing to prevent electrolyte stratification, and you will not be able to charge the batteries quickly enough to make up for a few days without sun. The rule of thumb that goes with this is that the panel watts going into an MPPT controller should roughly match the battery bank AH number for a 12 volt system.

    Can someone please tell me if I'm interpreting this correctly? For a FLA battery, the maximum charge rate should be C/8 and the minimum should be C/12 (according to the above quote.) I thought the minimum is 10% of a C/20 charge rate?

    So now, using C/12 as a minimum it looks like my panel array should provide adequate charging power for battery bank. My battery bank is 675ah.

    My calculation: 675/12 (C/12) = 56.25. So 56.25*120 (C/12 rate) = 6750 / 671 = 10%

    This comes from the AWAS battery FAQ: Example: "The C/8 rating is the battery capacity at the 20-hour rate divided by C/8". So I simply replaced C/8 with C/12 for my calculation based on inetdog's C/12 minimum.

    However, another member on this forum calculated my charge rate this way:
    810 Watts * 77% typical efficiency / 12 Volts system nominal = 52 Amps And so: 5200 / 675 = 7.7%

    So I'm confused a bit. If I use a C/12 rate then I'm in the ballpark for recharging my batts. Can I get by with a C/12 charge rate or do I use the C/20 @ 10% minimum rate? And why is .77 used as a overall efficiency rating? Is that just an average?

    Btw, does C/12 convert to roughly 8 percent? What is the formula for converting a C/ rating to a percentage? I know I shouldn't be so confused over this.:confused::confused:

    FYI, I live in an area that is the sunniest in the US. Zone 1. So this should increase my power theoretically I'm assuming:
    Attachment not found.

    Hope that makes sense.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing

    Minimum recommended charge rate is typically 5% of the battery's capacity rated as the "20 hour" rate.
    That equates to C/20. This should be the net charge rate too; allowances for load draw while charging need to be made.
    10% (C/10) is good "middle-of-the-road" target for planning a system as it usually results in a net charge rate between 5% and 10% even with loads. It will also usually allow the batteries to be fully charged on a good sunny day before the sun runs out providing the DOD is around 25% and you have at least 4 hours of equivalent good sun.
    13% (C/8 ) is usually the upper practical limit, beyond which you tend to be spending more on recharging capacity than you need to. This is not true in all instances of course; these are just guidelines to get the design process in the right "ballpark".

    The "Watts * 0.77 efficiency / nominal Voltage" formula is for predicting the probable peak output current using an MPPT controller at minimum battery Voltage (where the most current is needed). This is in contrast to the use of a PWM controller where the maximum current is limited by the combined Imp of the panels.

    C/'X' to "%" is a matter of reciprocals: 100/20 = 5(%), 100/10 = 10(%), 100/8 = 12.5(%) and so on.
    Just as current * 100 / battery capacity = charge rate percent.

    Ain't math fun? :p
  • Hairfarm
    Hairfarm Solar Expert Posts: 225 ✭✭✭
    Re: Battery bank sizing
    C/'X' to "%" is a matter of reciprocals: 100/20 = 5(%), 100/10 = 10(%), 100/8 = 12.5(%) and so on.
    Just as current * 100 / battery capacity = charge rate percent.

    Cool. Everything just clicked!
    Just as current * 100 / battery capacity = charge rate percent.

    675 / 12 = 56.25, 56.25 * 100 = 5625, 5625 / 675 = 8.3%. (* This also happens to be exactly a C/12 charge rate too. 56.25 *100 = 5625 / 675 = 8.3%. That's weird.)

    I realize that I didn't factor in the .77 efficiency rating though. In that case:
    810 Watts * 77% typical efficiency / 12 Volts system nominal = 52 Amps And so: 5200 / 675 = 7.7%

    In either case it's going to have to work for now. I'm out of budget and roof space:grr

    Do you think a C/12 rate is too low a charge rate then?

    thanks Cariboocoot
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Battery bank sizing

    C/12 i.e. 8% is probably a-okay.
    Keep in mind this is a potential maximum rate; it does not mean this is the rate you will see every day or perhaps even ever, as the SOC of the batteries is a large determining factor for just how much current actually flows. Another factor is the loads, which tend not to be constant. So as long as you can achieve an average charge rate above the minimum it should all work.

    It's so easy to run out of budget, isn't it? :blush:
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: Battery bank sizing

    Eeesh, here we go again with the "minimum charge rate" rule of thumb. To give you some background of where this "rule" came from, as I understand it, it can be stated more completely as:

    As a general rule of thumb, when you're designing an off-grid solar system almost anywhere in the world, using the average of all weather everywhere, for year round full time use, winter through to summer, and you happen to have chosen a battery bank size that gives you 2 days of storage, then choosing an array that provides a 5% charge rate will usually charge your batteries fast enough to avoid serious sulfation issues.

    So you can see that there are a few generalisation there, the most important in your case are the usage period and the location. If you're only planning on using the system in spring/summer and it will be located in a sunny location then those are two very important factors that influence the size of the array and in my opinion once you're down to this level of detailed design then it's time to put the rules of thumb away and start working with more accurate numbers. I'd design it as follows (you can plug in your values where I'm off the mark, but the formula stays the same):

    - Derive your daily energy usage during the worst month (this is the month where the ratio of solar harvest/energy used is the worst). Let's say you can knock your usage down to 10kWh/day. Add 10% to make up for inverter losses: 11kWh/day.

    - Size the battery bank for 2 days storage of this amount and a depth of discharge of 50%: 11 000Wh/48V = 230Ah x 2 days = 460Ah / 0.5 depth of discharge = 920Ah 48V battery.

    - To size the array, on average you need to supply 11kWh/day + 20% to account for battery charging inefficiency + 10% to account for losses in the charge controller (assuming it's an MPPT), cabling etc, so that's about 15kWh/day on average.

    - But some days won't be average days, you'll have bad days where the previous day had poor solar radiation and your batteries are at 70% state of charge and the array needs to both supply your current loads and recharge the batteries. There will be even worse days when the previous 2 days have produced very little energy, your batteries are at 50% and need recharging. Now you have some leeway in deciding how you're going to handle these bad days, either supplement charging with a generator or increase the array size. You'll have a better idea of your local weather to make this call. If there are very few days where you have rain for 2 days straight, then it doesn't make economic sense to size the array for these rare events, better to run the generator. But if they're fairly frequent then you'd want to increase the array size to help recharge the batteries from a 50% state of charge and supply your 11kWh daily load.

    - Now you can consult a tool like PVWatts http://gisatnrel.nrel.gov/PVWatts_Viewer/index.html to determine how much installed PV you need in your worst month to meet that demand. Using southern california as a location with a static array with optimal angle and orientation (you can change this in the tool), it shows that a 1kW array will provide 140kWh/month in May =~ 4.5kWh/day. You need 15kWh/day on average, so that's a 3.3kW array for an average day in May in Southern California.