Series/Parallel battery wiring

2»

Comments

  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Series/Parallel battery wiring
    jonr wrote: »
    Obviously one would not measure the batteries while in an unequal SOC and then use those resistance values to balance. They would all be equally charged first.

    Their internal resistances when they are fully charged doesn't tell you what their resistances are when they are discharged.

    I have a quote for you:
    Faced with the choice between changing one's mind and proving that there is no need to do so, almost everyone gets busy on the proof.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • jonr
    jonr Solar Expert Posts: 1,386 ✭✭✭✭
    Re: Series/Parallel battery wiring
    vtmaps wrote: »
    Their internal resistances when they are fully charged doesn't tell you what their resistances are when they are discharged.

    Perhaps a "proof in the pudding" approach. Measure cumulative amp hours delivered by each battery for a week. Reduce some cable lengths and shuffle batteries to make these AH values the same (or at least closer) across all batteries.

    From what I read, battery ohms are typically within 8% of each other when new but this increases significantly as they age. Say one battery does more work and dies early and then the whole bank is replaced 15% earlier than it might have if balanced. Just trying to put a dollar value on it...

    I am available for custom hardware/firmware development

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Series/Parallel battery wiring

    From (our "own" nsaspook?):

    http://www.physicsforums.com/showthread.php?t=634808
    Values for several different sealed valve regulated types.
    3 milliohm 226Ah: http://www.power-sonic.com/images/po..._11_Jan_12.pdf
    3 milliohm 140Ah: http://www.power-sonic.com/images/po..._11_Feb_21.pdf
    5 milliohm 154Ah: http://www.power-sonic.com/images/po..._12_July_2.pdf

    Those are probably on the "low resistance" side of the battery community.

    You can certainly estimate the resistance of your battery bank

    R= change in voltage/change in current.

    Say you test your car battery. It takes 70 amps to crank the the voltage falls from 12.7 to 11.7 volts (just made up numbers):

    1 volt drop / 75 amps = 0.013 ohms = 13 milliOhms

    Of course, you probably want to run this test several times--You don't want to have a "surface charge" that you "burn off" with the first test (i.e., battery at 13.2 volts surface charge and 12.7 end result after a couple of cranking tests).

    Here is a quick table on mOhm/ft (and metric) for common copper wire resistance values:

    http://en.wikipedia.org/wiki/American_wire_gauge

    For 6 AWG cable, you are looking at ~0.3951 mΩ/ft.

    If you have 5 feet of one way cable or 10 feet round trip, you are looking at ~4 mΩ (0.004 Ohms). Compare that to some typical numbers... For AGM, that would be on the order of the battery resistance (12 volt battery). And significant "ballasting" for AGM type battery.

    For a Flooded Cell battery, with (rough estimate) of 13 mΩ internal resistance, then you are looking at the current flow more affected by battery resistance than cabling resistance.

    In a typical off grid power system, heavy/short copper cable is used for wiring up a battery bank... And variation in batteries is going to be the major factor in current steering (assuming solid/clean electrical connections).

    In general engineering terms, "differences" of 2x (or 1/2) or less, are significant effects. And >10x differences can usually be ignored. (i.e., 1.3 vs 13 mΩ could be ignored, 3 vs 6 mΩ have significant effects).

    Also remember that (roughly) battery resting voltage is related to the specific gravity of the cells:

    http://forum.solar-electric.com/showthread.php?17895-Battery-Voltage-and-Specific-Gravity-Question
    Specific gravity = single-cell open-circuit voltage - 0.845 (example: 2.13v – 0.845 = 1.285)
    Or
    Single-cell open circuit voltage = specific gravity + 0.845.

    So, if you have a cell that is ~0.030 different than the others, you have an ~0.03 volt resting voltage difference between cells. Or if two batteries in parallel at 12 volts (6 cell), ~0.18 volts difference in resting voltage.

    In terms of 6 AWG cable (ignore if does not make sense--Trying to use real numbers to see/show effect of unbalanced cable lengths):

    V=I/R
    I = V/R = 0.18 volts diff / 0.0003951 ohms (ref=1 foot) = 456 Amps "differential" through 1 foot of cable

    Obviously, cannot push 456 amps through 1 foot of 6 AWG cable... but if you had a 10 foot difference, you would then need ~45.6 amps of current to get a 0.18 volt differential.

    I would suggest that 0.05 to 0.10 volt maximum difference between battery charging (usually it is the accuracy for a battery charger to measure battery charging voltage for a 12 volt battery)... So:

    I = 0.05 volts / 0.0003951 ohms = 126 Amps through 1 foot of 6 AWG cable

    What does this tell us? A one foot difference between parallel strings and 60 amps of charging current through a 6 AWG cable:

    V=60 amps / 0.0003951 ohms = 0.02 volts drop

    And this is on the scale of a 0.030 difference in one cell of a battery that needs equalizing (0.030 due to specific gravity difference vs 0.02 volt drop due to wiring difference).... Would suggest that these are on the edge of being similar effects in current steering--And are probably on the border of being "significant" to the overall health of the battery bank.

    But because we are comparing voltage effects (resting voltage) vs fixed effects (resistance of wiring, battery resistance), I would be suggesting some detailed testing to see how this really affects the real world.

    Another number--As a battery gets hot, a cell's charging voltages (lead acid) should be reduced by -5mVolts / cell per degree C. Or -0.030 volts per Degree C per 12 volt battery. If you have a difference of 5C between a "hot" and "cool" 12 volt batteries in a bank (due to ventilation, battery health, etc.), that would be a difference of 0.150 volts in "desired" charging voltages. Another "issue" to think about in the analysis. And another data point to indicate what is the system's "sensitivity" to changes in parameters.

    Oh well--Just trying to hang some real world numbers on the math. Not sure this helps with understanding or not. :blush:

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Series/Parallel battery wiring
    jonr wrote: »
    Measure cumulative amp hours delivered by each battery for a week. Reduce some cable lengths and shuffle batteries to make these AH values the same (or at least closer) across all batteries.

    Like I said in post #11:
    vtMaps wrote:
    Try it... monitor what you are doing with an hydrometer.
    <snip>
    With an hydrometer you can figure out what works for your batteries in your system in your environment.

    Let us know how it goes. There's more than one way to skin a cat or charge a battery.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • sanitariu
    sanitariu Solar Expert Posts: 33 ✭✭
    Re: Series/Parallel battery wiring

    I use that scheme
    http://forum.solar-electric.com//attachment.php?attachmentid=5995&d=1381918294
    After 1 year it is still OK for me.
    I plan to shuffle the batteries that summer.
    4 X 240watt Luxor, Victron Energy Multiplus Compact 24 volt 1200VA, electrodacus solar controller, 8 x WB-LYP160AHA LiFeYPO4 3.2volt
  • jonr
    jonr Solar Expert Posts: 1,386 ✭✭✭✭
    Re: Series/Parallel battery wiring

    If variations in individual battery internal resistances are the "major factor", my guess is that some simple/uneven wiring layout plus one of the suggested measurement methods plus arranging the batteries accordingly will typically provide better balancing than the balanced cabling methods described on the SmartGauge site. And if you are willing to shorten and perhaps lengthen some cables, you might get nearly perfect balancing (say over week+ periods).

    I am available for custom hardware/firmware development