My battery puzzle

sunbunny
sunbunny Solar Expert Posts: 59 ✭✭✭✭
in Off Grid Solar & Battery Systems #1
I'm a lucky girl ! (sometimes...) A friend gave me 10 Concorde AGMs 1040 12 volt -used but with good charge showing.

I would really prefer to stick with 24 volts because I don't need/want a fancy inverter/charger. I'm powering a stationary travel trailer. Have 3 generators for power tools.

Is there a reason that I can't use one of those high amp marine switches to switch between charging/drawing on 2 battery banks ?
Then I could have a bank of 2 pairs and a bank of 3 pairs and manually switch in the one that needs charging. Would this be better than having the 5 pairs in one bank ?

I have 8 250 watt panels.
Yes, I could buy one more panel and put 3 to a string and have room on my Flexmax 80 for lots more panels in a 48 volt bank system but I'd rather not.

Thanks
· Share on Twitter

Comments

  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    #2
    Re: My battery puzzle
    sunbunny wrote: »
    Is there a reason that I can't use one of those high amp marine switches to switch between charging/drawing on 2 battery banks ?
    Then I could have a bank of 2 pairs and a bank of 3 pairs and manually switch in the one that needs charging. Would this be better than having the 5 pairs in one bank ?

    A switch makes sense and many folks here do just that. There is the issue of whether to use a make-before-break switch vs a break-before-make switch. I would recommend that you break-before-make, and also that you reduce your loads before making the switch.

    Another option is to put all five strings in parallel. This goes against our usual recommendations because it is inherently unstable. However it can be done... forum member Chris Olson does it with 6 parallel strings. The key to making it work is to be constantly monitoring the individual strings for differences in current flow (use a DC clamp ammeter). Chris Olsen regularly disassembles his battery bank, load tests the individual batteries, and reassembles the bank with batteries in different positions.

    You can monitor your bank and disconnect batteries that are outliers, but you can't rearrange them because each string of your batteries is a single 12 volt battery.

    --vtMaps

    EDIT: just realized that you have a 24 volt system. Therefore you can rearrange your batteries (move a battery from one string to another).
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
    · Share on Twitter
  • sunbunny
    sunbunny Solar Expert Posts: 59 ✭✭✭✭
    #3
    Re: My battery puzzle

    Thanks, vtMaps. If the 5 strings of batteries are an unstable situation and need constant attention, I'll probably go with the switch.
    · Share on Twitter
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    #4
    Re: My battery puzzle
    sunbunny wrote: »
    Thanks, vtMaps. If the 5 strings of batteries are an unstable situation and need constant attention, I'll probably go with the switch.

    Well actually even two strings is unstable. read: http://forum.solar-electric.com/showthread.php?14674

    Anytime you have parallel strings you should be monitoring them for differences. When one cell in one battery goes bad (and one will), the other parallel strings mask it until the whole bank is damaged (you won't notice anything wrong until the lights go out).

    Face it, being off grid means being constantly vigilant. The issue with parallel strings is that the problems are magnified, in other words: when you finally realize that something is wrong you have damaged all of your strings.

    Another way to think of this is probability. Consider a single engine aircraft: There is a certain probability that an engine will fail on the next flight. If you have a two-engine aircraft, the probability that an engine will fail on the next flight is doubled. So it is with batteries. When one battery fails it will take down the entire bank (unless you catch it early through constant monitoring).

    If you have five parallel strings the probability of a battery failing in the next week is five time the probability of having a single string failing.

    The switch is a good idea... when one string fails, just switch it out and go on with your life.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
    · Share on Twitter
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    #5
    Re: My battery puzzle
    vtmaps wrote: »
    Consider a single engine aircraft: There is a certain probability that an engine will fail on the next flight. If you have a two-engine aircraft, the probability that an engine will fail on the next flight is doubled. So it is with batteries. When one battery fails it will take down the entire bank (unless you catch it early through constant monitoring).

    On the other hand, if you have two-engine plain and one engine fails you can land it on one engine. If you have only one engine, you're doomed. Same with batteries, you can disconnect one string and live with other(s) until you fix the problem.
    · Share on Twitter
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    #6
    Re: My battery puzzle
    NorthGuy wrote: »
    On the other hand, if you have two-engine plain and one engine fails you can land it on one engine. If you have only one engine, you're doomed. Same with batteries, you can disconnect one string and live with other(s) until you fix the problem.
    Or, to push the analogy way too far, you might fail to notice the bad engine and keep on flying, including more takeoffs and landings, and overstress the remaining engine so that it will fail much sooner than it otherwise would have.

    Old groaner:
    Pilot comes on intercom and says: "I am sorry to report that we are having problems with one engine. Since we have a four engine plane, there is no reason to be afraid, but we will arrive about 20 minutes late."
    Soon: "This is your pilot again, we have lost another engine, and are still perfectly safe, but we will be delayed about an hour."
    Next: "Folka, we are having trouble with a third engine, and will be landing about 3 hours late."
    Man to guy in seat next to him: "I sure hope we don't lose the fourth engine or we'll be up here all night."

    And so it goes with batteries.
    SMA SB 3000, old BP panels.
    · Share on Twitter
  • Ralph Day
    Ralph Day Solar Expert Posts: 1,022 ✭✭✭✭
    #7
    Re: My battery puzzle

    Google "Gimli glider". An engine out does not mean disaster, most planes glide. Pilots are trained for forced landings. Probably most engine fail crashes occur when someone has an engine out on takeoff and tries to turn around and land on the strip instead of taking their lumps and landing straight ahead.

    Ralph
    · Share on Twitter
  • YehoshuaAgapao
    YehoshuaAgapao Solar Expert Posts: 280 ✭✭
    #8
    Re: My battery puzzle
    NorthGuy wrote: »
    On the other hand, if you have two-engine plain and one engine fails you can land it on one engine. If you have only one engine, you're doomed. Same with batteries, you can disconnect one string and live with other(s) until you fix the problem.

    Losing one engine on a twin-engine turboprop (like a King Air) can be disastrous, worse than losing both, because of the torque imbalance is much worse and the pilots not always being trained for it. Well trained (even self-trained on X-plane), non-panicky pilots do very well with some or all engines out.
    · Share on Twitter
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    #9
    Re: My battery puzzle
    I'm a lucky girl ! (sometimes...) A friend gave me 10 Concorde AGMs 1040 12 volt -used but with good charge showing.

    The other question is : How will you keep all that battery charged? If you want to do it with solar, at the very least you'd need about 8000 watts of PV and that's being optimistic...

    ( i'm assuming you mean 10 x 1040AH 12 volt batteries = 10400 AH total!)
    · Share on Twitter
  • sunbunny
    sunbunny Solar Expert Posts: 59 ✭✭✭✭
    #10
    Re: My battery puzzle

    No, 1040 is the model number. 104 amp hours each.
    I am a little concerned about too much current from the 2000 watts PV array trying to charge either 200 or 300 AH battery banks....
    · Share on Twitter
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    #11
    Re: My battery puzzle
    sunbunny wrote: »
    No, 1040 is the model number. 104 amp hours each.
    I am a little concerned about too much current from the 2000 watts PV array trying to charge either 200 or 300 AH battery banks....

    Many of the full-featured controllers (midnite, outback, etc) allow you to limit the charge to the batteries. As PV panels become cheaper and batteries more expensive, more folks are "over paneling" and limiting the charge current.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
    · Share on Twitter
  • Coach Dad
    Coach Dad Solar Expert Posts: 154 ✭✭
    #12
    Re: My battery puzzle
    sunbunny wrote: »
    No, 1040 is the model number. 104 amp hours each.
    I am a little concerned about too much current from the 2000 watts PV array trying to charge either 200 or 300 AH battery banks....

    You don't have to worry about too much current with Concorde Batteries. They will take current levels of up to 5C (500A for a 100ah battery).

    See page 19 of the technical manual..
    http://www.sunxtender.com/pdfs/Sun_Xtender_Battery_Technical_Manual.pdf
    · Share on Twitter
  • sunbunny
    sunbunny Solar Expert Posts: 59 ✭✭✭✭
    #13
    Re: My battery puzzle

    Thanks, vtMaps and Coach Dad.
    Along those lines.... it's recomended to fuse each battery pair... what size fuse/CB would I need ?
    · Share on Twitter
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    #14
    Re: My battery puzzle
    sunbunny wrote: »
    Thanks, vtMaps and Coach Dad.
    Along those lines.... it's recommended to fuse each battery pair... what size fuse/CB would I need ?

    Good question... I am unsure how to answer. It depends on your peak loads and also on the purpose of fuse/CB.

    Note: if you use fuses rather than CB, you should also have a switch in each string. It is very dangerous to remove or replace a DC fuse in a live circuit. Removing the loads from a battery bank does NOT mean there is no current flowing... There may be current flowing from one string into another. That is probably the main reason for having a CB/fuse in each string... if a shorted cell develops in one string, there will be current flowing from the good strings into the bad string.

    Either way (fuses or CB) you will find them very desirable for servicing and testing your battery bank. Instead of having a battery bank selector switch, you could put all five strings in parallel and use the CBs or switches to create a battery bank of 1 to 5 strings.

    The CB/fuse should be large enough that it doesn't trip/blow during your peak loads. If one trips/blows, they will all trip/blow. (I recommend CB over fuses).

    Which ever way you do it (CB with a CB panel, or fuses and switches) you add considerably to the cost and complexity of wiring up your battery bank. That is another reason we usually recommend just a single string of batteries.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
    · Share on Twitter
  • mtdoc
    mtdoc Solar Expert Posts: 600 ✭✭
    #15
    Re: My battery puzzle
    sunbunny wrote: »
    No, 1040 is the model number. 104 amp hours each. ....
    Doh!. Could of sworn i saw an AH in there - but of course 1040AH 12v would be way too big for an AGM and weight over 1000 lbs! (forklift batterry maybe)

    That's what i get for trying to read the NAWS forum late at night. :blush:

    (Sulks and wanders off to find a cup of coffee.....)
    · Share on Twitter
  • sunbunny
    sunbunny Solar Expert Posts: 59 ✭✭✭✭
    #16
    Re: My battery puzzle
    vtmaps wrote: »
    .. if a shorted cell develops in one string, there will be current flowing from the good strings into the bad string.

    Either way (fuses or CB) you will find them very desirable for servicing and testing your battery bank. Instead of having a battery bank selector switch, you could put all five strings in parallel and use the CBs or switches to create a battery bank of 1 to 5 strings.

    The CB/fuse should be large enough that it doesn't trip/blow during your peak loads. If one trips/blows, they will all trip/blow. (I recommend CB over fuses).

    --vtMaps

    Ya, that's what I can't figure out. Peak load would be on a 1100 watt inverter, for now, on a 200/300AH bank. I don't know how to use the battery specs to guess what might occur for current on a shorted cell... any help with that ??

    These are the circuit breakers I like.... but I can't guess which ampereage....
    http://www.solar-electric.com/mr60ampdccib.html
    · Share on Twitter
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    #17
    Re: My battery puzzle

    Fuse/breaker size is based on expected continuous current draw just like wiring. The wires should always be able to handle more than the fuse (weakest link in the chain).

    So if you have an 1100 Watt inverter on a 12 Volt system you can expect either: 1100/12 = 92 Amps or the "full-blown treatment" of maximum surge power at minimum input Voltage; 2200/10.5 = 210 Amps. Quite a difference a you can see. This is why the best answer comes from the people who made the inverter: they should provide specs on wiring and circuit protection. Can't do better than their word.

    I personally stick with the continuous Wattage rating divided by nominal Voltage as I don't expect the inverter to be run close to its maximum, much less over it, long enough to trip CP and I don't like allowing the Voltage under load to drop below nominal (assuring 50% SOC is maintained). So I would be using a 100 Amp fuse on this. But you may find that blows too often and have to go up in size. Not a problem with a breaker, but a real pain to have to replace fuses continuously.

    Remember that the wiring has to be large enough not only to handle the current but also to prevent significant Voltage drop at the inverter under full current. Size for current first, then check the V-drop across the distance for that gauge wire and increase as needed.
    · Share on Twitter
  • sunbunny
    sunbunny Solar Expert Posts: 59 ✭✭✭✭
    #18
    Re: My battery puzzle

    Thanks 'Coot. So if I'm doing 24 volts then a 60 amp CB ??

    Is my PV array capable of more amps than that ( 8 250 watt, Imp 7.9) ?

    I plan on oversize wiring so it might be reusable in an upgrade.
    · Share on Twitter
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    #19
    Re: My battery puzzle
    sunbunny wrote: »
    Thanks 'Coot. So if I'm doing 24 volts then a 60 amp CB ??

    60 Amps * 24 Volts = 1440 Watts, so it should handle an 1100 Watt inverter under most circumstances.
    Is my PV array capable of more amps than that ( 8 250 watt, Imp 7.9) ?

    Eight 250 Watt panels on a 24 Volt system would be about 64 Amps peak charge current using an MPPT controller. However, the battery charging circuit is not the same as the inverter power circuit. Usually the inverter circuit is the higher current one, but it does not matter: even if the array and controller can put out 100 Amps it won't have any affect on what the inverter draws.
    I plan on oversize wiring so it might be reusable in an upgrade.

    Always a good idea. :D
    · Share on Twitter
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    #20
    Re: My battery puzzle
    Cariboocoot wrote: »
    So if you have an 1100 Watt inverter on a 12 Volt system you can expect either: 1100/12 = 92 Amps or the "full-blown treatment" of maximum surge power at minimum input Voltage; 2200/10.5 = 210 Amps. Quite a difference a you can see. This is why the best answer comes from the people who made the inverter: they should provide specs on wiring and circuit protection. Can't do better than their word.

    The folks who made the inverter specify a CB/fuse size for the battery bank as a whole... I thought her question was about what size CB/fuse to use on each individual string of a parallel battery bank.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
    · Share on Twitter
  • sunbunny
    sunbunny Solar Expert Posts: 59 ✭✭✭✭
    #21
    Re: My battery puzzle

    But if the array and controller are putting out 100 amps, how many amps are going to each battery pair ? That won't affect the CB ?

    Yes, I want to know how to fuse each battery string.
    · Share on Twitter
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    #22
    Re: My battery puzzle

    No circuit = no current.
    Current is determined by the circuit as per Ohms law (see my signature line).
    A power source may be capable of 100 Amps, but that does not mean that is how much will flow. That is determined by the load placed on it (resistance).

    This is why I often tell people to think of the system in terms of three or four separate but interacting circuits:
    1). PV to charge controller
    2). Charge controller to batteries
    3). Batteries to inverter/DC loads
    4). AC wiring from inverter

    The amount of load placed on the charge controller output will vary the amount of current it supplies. As far as it is concerned, charging batteries is a load as is powering an inverter. If the power available from the controller exceeds the demands of both battery charging and inverter draw no further current is used even if it is available.

    An inverter at idle may draw 1 Amp. The batteries fully charged may draw 3. This in the middle of the day when the panels and controller are capable of supplying 64 Amps, but the loads do not demand it so only 4 Amps are drawn. Where does the extra power go? It doesn't exist: panels are a current source. Unlike batteries they don't have an Amp hour potential, only an immediate current potential. if not used when available the power is not realized.
    · Share on Twitter
  • sunbunny
    sunbunny Solar Expert Posts: 59 ✭✭✭✭
    #23
    Re: My battery puzzle
    Cariboocoot wrote: »
    No circuit = no current.

    OK. Got that.

    So if a battery goes bad, is there any way to predict the amps it might draw ? Will the 60 amp breaker disconnect the bad string ?
    · Share on Twitter
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    #24
    Re: My battery puzzle
    sunbunny wrote: »
    OK. Got that.

    So if a battery goes bad, is there any way to predict the amps it might draw ? Will the 60 amp breaker disconnect the bad string ?

    No. A battery going bad as in a shorted cell is unpredictable as to its nature/resistance. What the 60 Amp breaker will protect against is the current on the circuit exceeding the safe limit for the wire, which means it will not overheat and go up in smoke.

    Consider that a battery can dump its full Amp hour rating as an instantaneous current flow (like when you drop a wrench across the terminals). As such it is capable of withstanding far more current than the the circuits attached to it. It is highly unlikely to be the weak link, although not impossible.

    Normally you fuse the battery to inverter circuit according to its expected current and the controller to battery according to its expected current. If anything in either circuit causes the current to go high in that circuit the protection will trip. If that same component is part of the other circuit (i.e. the battery) it will trip the protection in that one as well.

    Batteries usually aren't affected much by high current as they are capable of handling quite a lot. High Voltage on the other hand ... When one cell shorts it is the same as putting "12 Volts to a 10 Volt battery": the remaining cells get pushed up by the 'extra' 2 Volts, and they don't like it much (boils off electrolyte).
    · Share on Twitter
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    #25
    Re: My battery puzzle
    sunbunny wrote: »
    Yes, I want to know how to fuse each battery string.

    Sorry I don't know the correct answer (I'm a single string type of guy), and nobody else seems to understand your question :cry: --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
    · Share on Twitter
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    #26
    Re: My battery puzzle
    sunbunny wrote: »
    OK. Got that.

    So if a battery goes bad, is there any way to predict the amps it might draw ? Will the 60 amp breaker disconnect the bad string ?

    In general a battery will fail by loss of capacity rather than by shorting internally. So a fuse or breaker per battery string will do nothing to protect them and will just, by itself increase any imbalance.
    SMA SB 3000, old BP panels.
    · Share on Twitter
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    #27
    Re: My battery puzzle

    Oh that! :D

    With just two batteries in parallel you can treat it as just one battery: single fuse off the one terminal (use the standard diagonal wiring system aka Smart Gauge method #2).

    Otherwise you should put one fuse on each string capable of handling the full expected current. Some will argue you should divide the total by the number of strings, but that can lead to "cascade failure" if one of the fuses should blow: the total current gets divided down by fewer strings so the fuses blow faster.

    Remember fuses are for "worst case scenarios" and you don't want to short-change the current capacity to the point where you're having to reset breakers or replace fuses all the time.
    · Share on Twitter
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    #28
    Re: My battery puzzle
    Cariboocoot wrote: »
    you should put one fuse on each string capable of handling the full expected current. Some will argue you should divide the total by the number of strings

    I've always thought the best approach was to fuse the individual strings with a value that is somewhere between "handling the full expected current" and "divide the total by the number of strings". But as I mentioned, I am uncertain how to calculate that value. If you use too low a value you do run into the problem of cascading failure (as mentioned in post #14 and #27).

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
    · Share on Twitter
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    #29
    Re: My battery puzzle
    vtmaps wrote: »
    I've always thought the best approach was to fuse the individual strings with a value that is somewhere between "handling the full expected current" and "divide the total by the number of strings". But as I mentioned, I am uncertain how to calculate that value. If you use too low a value you do run into the problem of cascading failure (as mentioned in post #14 and #27).

    --vtMaps

    Yes.
    It's not really necessary for just two in parallel as I mentioned.
    But when you go above that an extra measure of security (mainly against shorting battery posts) can be had by equipping each with a maximum current fuse.
    In any case I must add that such is in addition to the main inverter fuse. Otherwise you can have the circuit pulling less than maximum from each but greater than maximum from all. For example if you have four battery strings each with a 100 Amp fuse on it you can pull 99 from each totally 396 Amps whereas your maximum for the inverter may be 100 Amps. So without that single fuse in-line to protect the inverter circuit you actually have no protection.

    If you have trouble wrapping your head around this, think of it this way: each of those batteries has a set of wires going to the common connection/bus bar points. In that sense you are creating separate circuits (one per battery string) which then interact with the circuit from the bus bars to the inverter. All of these circuits need protection against over-current.

    I know: more schematics I should be drawing. :blush:
    · Share on Twitter

Categories