Series Lead Acid Balancing

jonr
jonr Solar Expert Posts: 1,386 ✭✭✭✭
This is worth watching for anyone interested in battery pack balance: http://www.eetimes.com/video.asp?section_id=124&doc_id=1326440

I am available for custom hardware/firmware development

Comments

  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    Or see the original video here: http://www.linear.com/solutions/5673

    While a given bank cells might have a near spread at rest, if youve ever measured your series cells during mid absorb, you might be horrified to see the variance in cell voltage.

    My question is how will that kind of solution stand up to the kind of current draws the inverters require. The mosfets can only pass so much, so i assume that it does its thing in the quiet times between current events.
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • CALLD
    CALLD Solar Expert Posts: 230 ✭✭
    I'd like to know just how important this is? I have 2x 12v 260AGM batteries in series for my 24v system.
    Under load the 2 batteries show a voltage difference of around 0.02v per 12v battery. During bulk charge and early absorb the difference is up to 0.05v. However towards the end of absorb the difference between the 2 12v batteries can be up to 0.50v. 14.90 vs 14.40v is the worst variance I've seen. It is of concern because it could be worse at a cell to cell level. But what can I do? There is no way of checking individual cells without butchering the battery! It must be said though that this only happens once the charge acceptance drops below 2 Amps which is less than 0.01C...
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Basically, at less than 2 amps into a 260 AH battery bank is getting pretty near full charge.

    To "balance" or equalize AGM batteries, a couple of mfg. have recommended just holding absorb voltage for a few extra hours... But as you can see with series cells, that is not easy to do. AGM's should not be charged at a high voltage as they will wear out the catalyst and eventually vent (gassing).

    One quick fix you can try--Get a 12 volt brake light bulb (old single filament brake lamp bulb is ~2 amp = 25 watts @ 12 volts) and put it across the battery with the higher voltage... The 12 volt bulb will "bypass" current around the "full battery" and allow the lower voltage battery to equalize (without damaging the higher voltage battery).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • jonr
    jonr Solar Expert Posts: 1,386 ✭✭✭✭
    Looks like the demo board is $150 although I suspect that it could be made for a lot less. I like the fact that it allows one to tap 12V off of a 48V string without causing short-term unbalance. But maybe using a DC-DC converter is better.

    The mosfets are rated at 60A but the fuse is 7A. But note that the inverter and charging currents don't have to flow through them - only the balancing current.

    I am available for custom hardware/firmware development

  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    CALLD wrote: »
    I'd like to know just how important this is? I have 2x 12v 260AGM batteries in series for my 24v system.

    And this is exactly why AGM batteries dont last as long as flooded. Lead in general is "self balancing", ie its somewhat tolerant of overcharging the high cells to get the low cells up. (Interestinglt they are starting to say this about some of the newer more exotic lithium chemistrys like INR.)

    Overcharging is required as the lessor of two evils. Positive plate grid corosion is sacrificed to prevent negative plate sulphation.

    But overcharging, there's costs.
    - As Bill said you have the recombinant catylyst that assists with getting the oxygen and hydrogen that is produced during advanced states of charge converted back to water. Once the catylyst is gone its gone. Its usually made of exotic read expensive metals.
    - Then youve got grid corosion, particularly of the postive grid. Grid corosion is a natural part of the battery charging process, but every minute spent above float voltage increases the amount of active material shed from the grids. AGM plates are very thin.
    - With AGM you have an additional risk of thermal runaway, which AGMs are particularly succeptible to. Relatively little overcharge leads to temp increases leads to current increases in a cycle. As little as 40 degrees C is enough.
    - However from what ive seen, water dry out is pretty much the primary AGM failure mode. Seal leakage for a start. All and any vent events contribute to this, (depending on how pressurised the particularl cell is) and a plastic encased battery will naturally migrate hydrogen slowly over time, through the plastic. I also think that things like grid corosion and sulphation also lead to dry out. Eg PbSO4 has 4 times more oxygen than PbO.
    jonr wrote: »
    The mosfets are rated at 60A but the fuse is 7A. But note that the inverter and charging currents don't have to flow through them - only the balancing current.

    Traditional cell balancing involves pypassing charge currents around high cells. The linear product in concern works differently, it applies extra charge from the redundant cell across the low cells during both charge and discharge. Hence my concern was around what happens when imbalances happen at a rate that exceeds the ability of the balancer to correct. We dont really know what magnitude of system that chip is intended to work, for. I suppose we will have to read the datasheet.

    Then i supose the obviously question of reliability. Plenty of folks BMS failures have killed lithium banks.
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    zoneblue wrote: »

    Overcharging is required as the lessor of two evils. Positive plate grid corosion is sacrificed to prevent negative plate sulphation.

    But overcharging, there's costs.
    - As Bill said you have the recombinant catylyst that assists with getting the oxygen and hydrogen that is produced during advanced states of charge converted back to water. Once the catylyst is gone its gone. Its usually made of exotic read expensive metals.

    One of the defining characteristics of a catalyst is that it is not consumed during the reaction that it promotes. But it can suffer surface contamination and mechanical damage.
    SMA SB 3000, old BP panels.
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    inetdog wrote: »
    One of the defining characteristics of a catalyst is that it is not consumed during the reaction that it promotes. But it can suffer surface contamination and mechanical damage.

    And I believe that these damages are the result of overheating the catalyst... the catalytic reaction being exothermic. --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    Im no chemist, and my language will thus be imprecise. However i get the impression that in VRLA batteres the oxygen and hydrogen recombine as a routine part of the discharge process, via gas channels in the saturated glass seperator. While hydrogen formation is a core part of the charge reaction, oxygen production is more related to secondary processes, and thus it seems that the purpose of the catylyst is more about reducing the effect that oxygen has on depolarisng the negative plate.

    But coming back to the effects of routine overcharge as a method of cell balancing. I did some reading on grid corision. The basic formula for positve grid corosion is:

    Pb + 2H2O → PbO2 + 4H + 4e

    "Corrosion leads to a passive layer formation between the grid and the
    active mass... It is important to mention that the lead, Pb in this reaction, is coming
    from the grid material and it is different from the spongy lead material, that
    is used as negative active material in the negative electrode. This reaction
    not only decreases the positive grid conductivity but also consumes water."

    source: Modeling Positive Plate Corrosion in Lead-Acid Batteries
    https://www.electrochem.org/dl/inter...wtr09_p057.pdf

    And it seems that grid corosion, in the absence of catylysts, is required to offset the negative plate depolarisation. Negative plate depolarisation is a relatively recentlly discovered failure mechanism for VRLAs.

    Personally I envisage that the high cell overcharge that occurs during late absorb, ought to lead to a nasty spiral of cell degradation. The more damaged the cell becomes the cell becomes, the higher its internal resistance, and the higher its terminal voltage. However we need to define high. In the first instance, by high i mean higher soc than its neighbours. Hence its terminal voltage upon charge will be higher. But a partially sulphated or grid coroded cell will also present with a higher terminal voltage during charge.

    So maybe that actually serves to balance things out? I dunno, its too early in the morning to say.
    Inet you seem to know something about this.
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    zoneblue wrote: »
    Im no chemist, and my language will thus be imprecise. However i get the impression that in VRLA batteres the oxygen and hydrogen recombine as a routine part of the discharge process, via gas channels in the saturated glass seperator. While hydrogen formation is a core part of the charge reaction, oxygen production is more related to secondary processes, and thus it seems that the purpose of the catylyst is more about reducing the effect that oxygen has on depolarisng the negative plate.

    But coming back to the effects of routine overcharge as a method of cell balancing. I did some reading on grid corision. The basic formula for positve grid corosion is:

    Pb + 2H2O → PbO2 + 4H + 4e

    "Corrosion leads to a passive layer formation between the grid and the
    active mass... It is important to mention that the lead, Pb in this reaction, is coming
    from the grid material and it is different from the spongy lead material, that
    is used as negative active material in the negative electrode. This reaction
    not only decreases the positive grid conductivity but also consumes water."

    source: Modeling Positive Plate Corrosion in Lead-Acid Batteries
    https://www.electrochem.org/dl/inter...wtr09_p057.pdf

    And it seems that grid corosion, in the absence of catylysts, is required to offset the negative plate depolarisation. Negative plate depolarisation is a relatively recentlly discovered failure mechanism for VRLAs.

    Personally I envisage that the high cell overcharge that occurs during late absorb, ought to lead to a nasty spiral of cell degradation. The more damaged the cell becomes the cell becomes, the higher its internal resistance, and the higher its terminal voltage. However we need to define high. In the first instance, by high i mean higher soc than its neighbours. Hence its terminal voltage upon charge will be higher. But a partially sulphated or grid coroded cell will also present with a higher terminal voltage during charge.

    So maybe that actually serves to balance things out? I dunno, its too early in the morning to say.
    Inet you seem to know something about this.
    Pb + 2H2O → PbO2 + 4H + 4e
    That is what a chemist would call a half reaction, in that it produces an unbalanced charge (4e). That charge must go somewhere, most commonly through the battery terminals and load or CC circuit to participate in a matching half reaction at the positive plate. And it will involve production of free oxygen. At the same time, in the absence of another local reaction, the 4H will combine rapidly to 2H2.

    The phenomenon of gassing is the result of a sst of reactions which taken together are the equivalent of the simple hydrolyisis of water into hydrogen and oxygen gas. These gasses will recombine through combustion if the temperature gets high enough for ignition, producing a hydrogen flame. Obviously those temperatures (and the resulting pressures) cannot happen inside the battery without catastrophic side effects. The catalyst's purpose is to bring the H2 and O2 together closely enough that they can recombine without requiring a high ignition temperature. But the amount of heat produced is still the same as it would be for simple combustion.
    The charging and therefore gassing current has to be kept low enough that the resulting heat can be carried away to the environment without overheating the catalyst or the entire battery.


    SMA SB 3000, old BP panels.
  • jonr
    jonr Solar Expert Posts: 1,386 ✭✭✭✭
    If there a consensus conclusion? For example, "a balancing device on a series string is most important with VRLA batteries but also reduces the need for overcharging/corrosion/water use in flooded strings"?

    http://www.battcon.com/papersfinal2004/symonspaper2004.pdf

    I am available for custom hardware/firmware development

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Jonr,

    The link did not work for me.

    Regarding series string vs parallel/etc... for battery balancing. Technically all batteries are a series string of cells (if more than 2 volts working voltage for lead acid). If you have a "bad or weak" cell--Trying to get it fixed without damaging the rest of the cells in series is the problem.

    Flooded cell lead acid batteries are fairly forgiving when being "overcharged"--Or letting current through the "good cells" to bring up the state of charge for the "weak" cell(s).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    BB. wrote: »
    Jonr,

    The link did not work for me.
    l

    Edit the link in your browser to point to .pdf instead of .p and it works fine.

    Corrected link: http://www.battcon.com/papersfinal2004/symonspaper2004.pdf
    SMA SB 3000, old BP panels.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    OK, I fixed the link in your first post.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    BB. wrote: »
    Jonr,

    The link did not work for me.

    Regarding series string vs parallel/etc... for battery balancing. Technically all batteries are a series string of cells (if more than 2 volts working voltage for lead acid). If you have a "bad or weak" cell--Trying to get it fixed without damaging the rest of the cells in series is the problem.

    Flooded cell lead acid batteries are fairly forgiving when being "overcharged"--Or letting current through the "good cells" to bring up the state of charge for the "weak" cell(s).

    -Bill


    I say flooded are extremely forgiving and often site my friend outside of La Paz Mx who had to wait 3 days to get to his home because of washed out roads due to a hurricane. When he got home he found his 48V flooded bank had been exposed to over 70 volts for 3 days due to a lightning failed charge controller. He added water and used the bank for another 5 years. I really do not have any problems with getting 10 years out of batteries that I totally understand. I think I will start Friday early and go have a beer with my battery bank. Might even watch the world champions of basket ball go on parade. Cheers!
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    Well yes, a flooded bank has thicker plates, tolerates positive grid corosion, vents away the oxygen that seems to create havoc on the negative plate (negative grid depolarisation issues). In short thats why we recomend FLA banks here, unless there are really compelling reasons to do otherwise.

    Coming back to VRLA. It seems that if you are going to spend 3K on a AGM bank, then investing in a cell monitoring system could provide a degree of comfort. But would i trust some cheap tech to try to balance, probably not. But thats just me, based on storys you hear.

    The key to AGM longevity is having a good catylyst, and how well the battery is sealed. Those would be two good questions to ask sales reps.
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • zoneblue
    zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
    When he got home he found his 48V flooded bank had been exposed to over 70 volts for 3 days due to a lightning failed charge controller.

    Had that have been AGM, he would have got to see first hand what a VRLA bank in thermal runaway looks like.
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • topper
    topper Solar Expert Posts: 113 ✭✭
    Funny. I am currently playing with one of these demo boards from Linear. I am still running AGM's and so far so good. I like it so far but it's only been running for a month.
    My previous solution was a simple circuit. 6.8 volt 5W zener, a 2 watt 4 ohm resistor and another 6.8 Volt 5 Watt zener connected in series.. Attatch in pairs or by 4 depending on system voltage (1 unit across each 12 volt battery) and the bank stays close to balance even during bulk charge.
    I also hate the thought of a thermal runaway when I am away from home.