Basic Sizing for Off Grid Battery Systems

Windsun · February 2012

These are some general rules of thumb that we use for most sunbelt type areas.

1. Figure out how many killwatt-hours you need over a 24 hour period. If not sure then best guess.
2. Add 30% to that for panel, wiring, controller, battery losses, etc.
3. Add 25-35% to that for recharging batteries after a no-charge (ie, rainy day).
4. That, divided by the full sun hours you get, will give you the amount - in watts - of solar panel you will need.
5. Battery bank capacity should equal about 4x the number in #3 for kilowatt-hours of storage.
a. For example, if figure #3 was 12,500 watt-hours, and you are using a 48 volt battery bank, divide 50,000/48 = approximately 1040 amp-hours of battery.

This is a little bit of a change from what we used a few years ago, because batteries have gotten a lot more expensive and solar panels a lot cheaper.

If you are in someplace like the Oregon rain forest you will need a lot more panel and probably a backup generator.

65DegN · February 2012

Re: Basic Sizing for Off Grid Battery Systems

I think that a good time of year to design for is the equinox. It is a median between the extremes of short winter days and long summer days.

solar_dave · February 2012

Re: Basic Sizing for Off Grid Battery Systems

65DegN wrote: »

I think that a good time of year to design for is the equinox. It is a median between the extremes of short winter days and long summer days.

I think you need to design around your annual load profile, if your loads are higher in winter for operating heating systems then your battery sizing should reflect that. So possibly the soltice should come into play more than the equinox. Here the summer soltice and the months following it for AC cooling would be the driver but perhaps the sizing could be a 2 day reserve because of summer solar conditions. Really loads drive sizing. If you grossly undersized because of equinox solar availability then the generator runs a lot to keep the battery levels up. From that panel sizing can be determined.

Windsun · February 2012

Re: Basic Sizing for Off Grid Battery Systems

solar_dave wrote: »

...if your loads are higher in winter for operating heating systems then your battery sizing should reflect that....

The biggest single problem we see and hear about for off-grid systems is undersized panel arrays and/or oversized battery banks (same thing, really). If you have more battery, you need more panel wattage. Essentially, you need your panels to supply 100% of your daily load plus around 20-40% extra for battery recharging after a bad solar day.

solar_dave · February 2012

Re: Basic Sizing for Off Grid Battery Systems

Windsun wrote: »

The biggest single problem we see and hear about for off-grid systems is undersized panel arrays and/or oversized battery banks (same thing, really). If you have more battery, you need more panel wattage. Essentially, you need your panels to supply 100% of your daily load plus around 20-40% extra for battery recharging after a bad solar day.

I agree, battery sizing should go with loads and panel sizing should go to support that battery.

niel · February 2012

Re: Basic Sizing for Off Grid Battery Systems

Windsun wrote: »

These are some general rules of thumb that we use for most sunbelt type areas.

1. Figure out how many killwatt-hours you need over a 24 hour period. If not sure then best guess.
2. Add 30% to that for panel, wiring, controller, battery losses, etc.
3. Add 25-35% to that for recharging batteries after a no-charge (ie, rainy day).
4. That, divided by the full sun hours you get, will give you the amount - in watts - of solar panel you will need.
5. Battery bank capacity should equal about 4x the number in #3 for kilowatt-hours of storage.
a. For example, if figure #3 was 12,500 watt-hours, and you are using a 48 volt battery bank, divide 12,500/48 = approximately 260 amp-hours of battery.

This is a little bit of a change from what we used a few years ago, because batteries have gotten a lot more expensive and solar panels a lot cheaper.

If you are in someplace like the Oregon rain forest you will need a lot more panel and probably a backup generator.

windsun,
correct me on any of my thinking here if i'm not properly following what you are saying.

i get that you are advising to go for a 25% dod point in battery capacity designs, but your example isn't quite right from what i see.

you stated that the watts overall with losses in #3 to be 12500w and that the battery capacity for a 25% dod should be 4x this or 50000wh battery capacity. this is 50000wh/48v=1041.7ah

now as in #4 the original wattage of #3 divided by the hours of sun gives the pv wattage which translates to 12500w/5hrs(top end example)=2500w in pv.

remembering this is desired to be a 48v output to match the battery bank voltage and can roughly be said to be 2500w/48v=52.08a from the pvs. this would represent 260ah divide by 52.08a = 4.992% charge rate and is too low in my view. many forget that going with higher capacity in a battery bank does mean more in pv to properly charge it and not just to sustain loads. had this been a design for 50% dod it would've been fine in the area of 9.99% charge rate.

Windsun · February 2012

Re: Basic Sizing for Off Grid Battery Systems

niel wrote: »

windsun,
correct me on any of my thinking here if i'm not properly following what you are saying.

i get that you are advising to go for a 25% dod point in battery capacity designs, but your example isn't quite right from what i see.

you stated that the watts overall with losses in #3 to be 12500w and that the battery capacity for a 25% dod should be 4x this or 50000wh battery capacity. this is 50000wh/48v=1041.7ah

now as in #4 the original wattage of #3 divided by the hours of sun gives the pv wattage which translates to 12500w/5hrs(top end example)=2500w in pv.

remembering this is desired to be a 48v output to match the battery bank voltage and can roughly be said to be 2500w/48v=52.08a from the pvs. this would represent 260ah divide by 52.08a = 4.992% charge rate and is too low in my view. many forget that going with higher capacity in a battery bank does mean more in pv to properly charge it and not just to sustain loads. had this been a design for 50% dod it would've been fine in the area of 9.99% charge rate.

In #1, assume you use 10 KWH per day
That would make #2 10 x 1.3, or 13 KWH
#3 would be 13 x ~1.25, or 16.25 KWH of production required per day

#4 Assume 5 hours of sun, so 16.25KWH/5 hoiurs = 3.25 KW total panel required.

#3 is 16.25 KWH, the battery should be about 4x that in storage capacity, so Battery should be around 1350 Amp-Hour at 48 volts.

So yes, I made an error in the first example, will correct that in the original.

Thanks for catching that.

niel · February 2012

Re: Basic Sizing for Off Grid Battery Systems

Windsun wrote: »

In #1, assume you use 10 KWH per day
That would make #2 10 x 1.3, or 13 KWH
#3 would be 13 x ~1.25, or 16.25 KWH of production required per day

#4 Assume 5 hours of sun, so 16.25KWH/5 hoiurs = 3.25 KW total panel required.

#3 is 16.25 KWH, the battery should be about 4x that in storage capacity, so Battery should be around 1350 Amp-Hour at 48 volts.

So yes, I made an error in the first example, will correct that in the original.

Thanks for catching that.

you did it again here as 3.25kw in pv at 48v out at best could provide 67.71a. with a battery bank of 1350ah this would be a charge rate of about 5% and although that hits the minimum we recommend for charging it may not be right for the battery to charge at that rate as many manufacturers are saying at about a 10% rate and some even more. where you are going wrong is assuming that the now larger battery bank will be charged fine by the array. the battery bank and solar array must parallel each other to some degree whereas doubling battery capacity will mean a rough doubling of the array to charge it while i acknowledge that some reduction in the array output over doubling it would be possible. minimizing the loads to lessen the impact on the battery bank will only serve as a conservation + in that the battery bank will last longer with a higher cycle life. if you opted for an 85% dod instead of a 75% dod you would more readily see that you have insufficient charge from the pvs at that point to properly charge the batteries with as the charge % will be too low (<5%) in spite of having loads that would dictate keeping the proposed wattage arrays.

basically battery charge requirements can differ from just replacing the load requirement. a large battery bank with an array charging at a good 10% charge rate will have sufficient power to properly bring the batteries up and it would do so in good time. an array at 5% would, if the battery manufacturer says it is ok to use and is still capable of being charged at that rate, still be ok to use even though it is more marginally in the good area. the 5% rate is also a 20hr rate of time and would require all 5 hrs of full array output to replace the 25% power lost and possibly more due to absorb charging times. it isn't quite as simple as you made it due to battery charge requirements.

Windsun · February 2012

Re: Basic Sizing for Off Grid Battery Systems

I know what you are saying about the charge rates, but in many cases it simply is not practical. You would end up over sizing the panel array by as much as 2x what you actually use for power, and most of that would simply go to waste.

The alternative is to lower the size of the battery bank, but then you lose 1-2 days of autonomy. A typical storm might last about 1.5 days, and you want your batteries to get no lower than around 40% DOD if possible. Going below 50% DOD really shortens the battery life, and then you want to get them fully charged back up in 3 days or less.

stephendv · February 2012

Re: Basic Sizing for Off Grid Battery Systems

niel wrote: »

with a battery bank of 1350ah this would be a charge rate of about 5% and although that hits the minimum we recommend for charging it may not be right for the battery to charge at that rate as many manufacturers are saying at about a 10% rate and some even more.

In my opinion the charge rate is not that important and this 5% minimum rate is a bit misleading. Although manufacturers may recommend a 10% rate (mostly for traction batteries) that doesn't mean that a lower rate is going to reduce the life of the batteries. FWIW a minimum charge rate is not specified in any european battery manufacturers literature, nor in any published PV sizing calculation.

IMO, the PV panels should be sized to replace the energy removed from the battery + a margin for charging inefficiency + the loads during the day. Obviously with an increased battery bank size, you need a bit more current to overcome the internal resistance - but this would be much less than the 5% rate.

Taking windsun's example, if instead of 4x storage capacity, let's use 8x battery capacity, so 2700Ah @ 48V but with the loads remaining the same. It doesn't make sense that this system would now require double the PV to charge. The loads are the same, so the depth of discharge is much less - and the amount of Ah to replace into the bank is the same. If anything, the battery bank will last longer because of the lower DoD. Sure when that 5 times a year event occurs and the battery drops to 50% DoD, it will take much longer for the battery to recharge- but will a 5 day recharge instead of 2 day recharge really shorten the life of the battery more than the life gained by the shallow DoD? And remember those events are very infrequent- compared to running at 10% DoD for 95% of the life of the battery.

telljf · February 2012

Re: Basic Sizing for Off Grid Battery Systems

Since the battery supports load only during the night time, the battery bank should be designed based on the night time consumption, and not on the 24hr consumption. Also we should look at AH rather than WH while designing the battery bank as the charging and discharge voltages are different. You need a battery monitor for that. The battery bank should be sized such that the night time consumption is 10-15% of the total bank capacity. For eg. if discharge during the night (6pm – 8am) is 50AH, then the battery bank should be about 400AH total capacity. Now during the day time (8am to 6pm) on an average day, the solar array should be able to fill this 50AH plus an additional 20% (total 60AH) AFTER supporting the daytime load. If you get 4hrs of good sun, 50/4 = 12.5A will be the charge current to the batteries. This is only 3% of the total bank capacity. Even if the panels generate more current, once it reaches the full charge cutoff voltage, the charger will bring it down.

It is really difficult to generalize a common strategy for all solar offgrid battery systems as there are too many variables involved. But this way, with proper care, you can reduce the array size, increase the life of the batteries and get more backup in the event of bad weather.

Windsun · February 2012

Re: Basic Sizing for Off Grid Battery Systems

stephendv wrote: »

Taking windsun's example, if instead of 4x storage capacity, let's use 8x battery capacity, so 2700Ah @ 48V but with the loads remaining the same. It doesn't make sense that this system would now require double the PV to charge. The loads are the same, so the depth of discharge is much less - and the amount of Ah to replace into the bank is the same. If anything, the battery bank will last longer because of the lower DoD. Sure when that 5 times a year event occurs and the battery drops to 50% DoD, it will take much longer for the battery to recharge- but will a 5 day recharge instead of 2 day recharge really shorten the life of the battery more than the life gained by the shallow DoD? And remember those events are very infrequent- compared to running at 10% DoD for 95% of the life of the battery.

What you say about increased batter life is true - but not 100%

Doubling the size of the bank will cost 2x as much, but will probably only give about 1,8x as much more lifespan. As I noted in the original message, batteries nowadays can cost more than the solar panels - the exact opposite of just a few years ago, when many of these "ideal" load vs battery capacity vs panel size scenarios were put forth. The solar array size has to be based on the usage, and more battery may give more storage for the "off days" of bad weather, but battery prices have gone up nearly 250% in real inflation adjusted money since 1999.

niel · February 2012

Re: Basic Sizing for Off Grid Battery Systems

the pvs are cheaper now so why not add more capacity to the array? a quicker recharge is a good thing especially for days with clouds or worse which are inevitable. theoretically, it would've been up to 2x the array size, but i did mention it could be less as in the example you needed 5hrs to charge and maybe 2hrs for absorb so upping it to the point that it would only take 5hrs for both bulk and absorb might mean maybe another 25% in pv. you have to know that batteries will not charge properly when the charge % goes too low so there is a limit to the theory of just replacing the power used plus losses. odds are most everybody will have loads on during the time of charge too that may need to be accounted for because the loads could offset any charging of the batteries.

Windsun · February 2012

Re: Basic Sizing for Off Grid Battery Systems

telljf wrote: »

Since the battery supports load only during the night time, the battery bank should be designed based on the night time consumption, and not on the 24hr consumption.

No, not really. If you have two days of heavy rain, that is almost the same as 48 hours of night as far as the panels are concerned for energy generation. While you may get some small charge even on heavy cloud days, it will be in the 5-15% range usually.

telljf · February 2012

Re: Basic Sizing for Off Grid Battery Systems

Windsun wrote: »

No, not really. If you have two days of heavy rain, that is almost the same as 48 hours of night as far as the panels are concerned for energy generation. While you may get some small charge even on heavy cloud days, it will be in the 5-15% range usually.

I agree. That is why I feel we should always have some sort of backup for charging the batteries, like hybrid inverter that charges from the grid, a gas generator, or maybe a wind power system for those rainy windy days. I have a hybrid inverter that could be switched to mains charging if required. If we design a solar system to meet all worst situations, it would become too expensive and an overkill on most days.

stephendv · February 2012

Re: Basic Sizing for Off Grid Battery Systems

niel wrote: »

...but i did mention it could be less as in the example you needed 5hrs to charge and maybe 2hrs for absorb so upping it to the point that it would only take 5hrs for both bulk and absorb might mean maybe another 25% in pv.

Yep, I agree. 25% more PV for 100% more battery capacity seems more reasonable than 100% more PV if the loads remain the same in both cases. My main point was that the 5% "rule" shouldn't be a rule

The size of the solar array should depend on the loads, not on the size of the battery.

Cariboocoot · February 2012

Re: Basic Sizing for Off Grid Battery Systems

stephendv wrote: »

The size of the solar array should depend on the loads, not on the size of the battery.

I disagree. If the loads are at night, the extra PV's do no good at all. That has to be considered too. Most of us off-gridders take advantage of "excess" panel capacity during the day, but it is not always possible.

The rules-of-thumb system design we discuss around here always, always have to be adapted to the particular site install. There are so many things that affect system efficiency that you can not just take the ballpark figures and run with them.

This is why I repeatedly recommend that for a baseline you use a 10% charging rate, not 5%, and limit DOD to 25% if at all possible. No one ever regretted having more than enough battery capacity or more than enough panel capacity. But a lot of people come to this forum because they have exactly the opposite.

stephendv · February 2012

Re: Basic Sizing for Off Grid Battery Systems

Cariboocoot wrote: »

I disagree. If the loads are at night, the extra PV's do no good at all.

By loads, I mean the loads over a 24 hour period, i.e. Wh/day. Here is a simple spreadsheet that does the calcs. The user needs to supply the following:
1. What are the loads in Wh over a 24 hour period?
2. How many full sun hours are there in the design month?
3. How many days of autonomy do you want from the battery bank?
4. How low do you want the battery DoD to go?

The answers to questions 1 and 2 give you the size of the PV panels.
The answers to 3 and 4 give you the size of the battery bank.
and the two sizes can be independent

It's the same approach windsun posted right at the start - with the difference that the user can choose their DoD and days of autonomy.

Cariboocoot · February 2012

Re: Basic Sizing for Off Grid Battery Systems

"Days of autonomy" is a trap best avoided. It requires lots of battery capacity which necessitates lots of PV to recharge once the sun does come out and leads to the possibility that the batteries may be spending too much time in a low state of charge, shortening their lifespan.

I prefer to use the 25% DOD rule-of-thumb knowing there's another 25% available if needed. Day 3 = start the generator. This is a good compromise against large battery banks and PV arrays that would not be needed most of the time. Local weather is one of those things you must take into consideration when planning. PV Watts can be helpful with this.

Answering the questions above without understanding how they relate to system performance could be problematic. I'll stick to my methods, thank you. So far they have always worked.

niel · February 2012

Re: Basic Sizing for Off Grid Battery Systems

charging a battery with below a 5% charge rate can be iffy and may even depend on the battery type and manufacturer. i will not say that there's no battery that will not charge up at lower than 5%, but this is the old school thinking of more battery and less pv to charge with because pvs cost too much with batteries costing so little and many people had gotten battery woes as a result of this as we've seen here so often. agm batteries are especially good due to their extra efficiency when below 5%.

also, just because manufacturers in the eu don't say it doesn't mean it isn't so. they probably figure they got you for another set of batteries earlier, but still beyond any of their warranties. battery parameters weren't put into place just as a ploy to get you to buy more in pv or bigger chargers as these did come from battery manufacturers primarilly from the physics of the batteries even if only stated here in the states. batteries do have limits of charge and discharge unlike the general population thinking they are little magic electrical boxes made up of chemicals that can do anything and be subjected to any abuse as this is certainly not the case. crewzer and myself did come up with the percentages you see here on naws and we have stated it is not written in stone, but it is a good guideline in most cases for good reasons.

also noteworthy is that the example given is within the 5% to 13% parameters. there's no guarantee that going outside of these parameters is a guarantee of failure either as i've said, but some battery manufacturers all but come straight out and tell you that some rates for their batteries have to be higher making even 5% no good. would any of you arguing that charge % doesn't matter be willing to guarantee anybody they won't have a failure because of too low of a charge rate % to the batteries?

as was already said, that was old school thinking with more battery and less pv due to the pricing on them way back when. there isn't any valid reasons for not upping the charge rates from pv if you can as it may help to lengthen battery life with being they are so much more expensive now and pvs cheaper. many days of autonomy are great to have, but you have to be able to charge them properly regardless of how much you use from them. going with more than 2 days of autonomy will necessitate more pvs without a doubt with 2 days being borderline as i've shown to be so due to extra time for absorb charging all in comparison to 50% dod being 1 day of autonomy. 25% dod 2 days. 12.5% 4 days. and so on. a pv system design for the 1 day does not work for any number of batteries needed down the line for larger autonomies with 2 days as marginal.

if you don't agree then you don't agree and we'll agree to disagree.

65DegN · February 2012

Re: Basic Sizing for Off Grid Battery Systems

Ted, the problem with sizing a system for the winter solstice around here is that the days are only about 5 hours long with maybe 3 hours of decent direct sun at ~70% of rated array output. The sun barely makes it above the horizon here. For many people with average sized 500 KWH monthly loads that puts a solar system out of reach for that time of year. It would take about a 9KW array to produce one days worth of power. Thats why I said, and maybe I should have qualified it by mentioning latitude, that the equinox is a good design median.

65DegN · February 2012

Re: Basic Sizing for Off Grid Battery Systems

Administrator, and it seems that the idea of buying less battery and using the savings to buy more panels is taking hold. I read in this months issue of 'Solar Industry' magazine, that Rob Shappel of Northwest Energy Storage (distributor of Hup Solar One) warns of chronic undercharging of batteries. He is advising that a slightly smaller battery bank be purchased, and with money saved buy more RE generation capacity.
Coming from a battery distributor you know it's sincere.

vtmaps · February 2012

Re: Basic Sizing for Off Grid Battery Systems

65DegN wrote: »

.... and it seems that the idea of buying less battery and using the savings to buy more panels is taking hold.....

a recent thread on this subject:
http://forum.solar-electric.com/showthread.php?14524-Could-cheap-panels-alter-the-panel-battery-ratio
The thread morphed into a discussion of how to protect the battery from too much current during bulk charging.
--vtMaps

Windsun · February 2012

Re: Basic Sizing for Off Grid Battery Systems

65DegN wrote: »

They may regret putting their money on a big battery bank if they can't keep them charged without continuous expenditures on gas.
In the other hand I can't imagine regretting buying an oversized array.

That is the single most common problem we hear about, both here and from people that call or email us. It seems that a lot more batteries are killed by sulfation than by cycling. While using a backup generator sounds good, the last 5% or 10% of the battery charge can mean some pretty long run times, which costs fuel. I looked up an old quote from 1999 for a 5kw off grid system. The batteries were about 15% of the total cost, the panels were about 60%. Roughly translating the same system to today's costs, it was battery 35%, panels 30% (including mounts but not wiring). The rest was BOS stuff.

Mangas · February 2012

Re: Basic Sizing for Off Grid Battery Systems

My two cents.

From someone who has been living "real world" with a sizeable off grid system in all seasonal weather conditions having a reliable standby generator sized properly to the battery bank is indispensable.

Too many folks try to operate without or, one that is incorrectly sized making off grid living very uncomfortable during winter and inclement weather. In my opinion, the genny is as important investment as the batteries and shouldn't be approached as a "band aid" or an afterthought.

Even with mine, unforeseen weather conditions can overwhelm a engineered system. Without a well maintained generator to back it up, 24-48 hour inclement weather can make life very uninteresting.

That said, I fully agree if you can keep a properly sized battery bank over, say, 80% by the end of the day in poor weather conditions using more panels versus adding expensive batteries that's the way to go. Compared to the other system components, batteries have become hideously expensive.

Basic Sizing for Off Grid Battery Systems

Comments

Categories