Announcement

Collapse
No announcement yet.

What is wrong with wiring batteries in parallel?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • What is wrong with wiring batteries in parallel?

    As far as I can see the main issues with wiring batteries in parallel are [LIST=1][*]Making sure that the current is shared equally between the batteries so that the they are all put under equal stress so they all have the same lifespan[*]Making sure that each of the batteries has enough "absorb time" so that all the batteries will be fully charged but not overcharged.[/LIST]
    If both these criteria are fulfilled then I don't think there should be any issue with putting batteries in parallel.

    I think the main problem with criteria 1 is that the most obvious way to wire batteries in parallel is like the diagram below
    From http://www.smartgauge.co.uk/batt_con.html
    From http://www.smartgauge.co.uk/batt_con.html
    I wouldn't be surprised if a large number of installations are wired like this. The problem is that this configuration can lead to the current in the last battery in the chain being around half the current in the first battery! This problem is easily fixed by following the guidelines from Smartgauge.

    If the batteries all have different internal impedance/resistance even if the wiring is done correctly the batteries will not share the current evenly. The battery with the highest impedance/resistance will draw or supply the least current. Is this a problem? Surely the weakest cell should be pampered, if it supplying less of the load I would think it would deteriorate at a slower rate to the other batteries. This would be a self balancing mechanism.

    The second issue is only a problem with LA batteries. All LA batteries need enough "absorb time" to make sure the battery is fully charged to stop sulphation but extended time at high charge voltages reduces their lifespan. This problem cannot be eliminated but can be minimised by making sure that all the batteries being paralleled are the same age (preferably new), have the same history and are the same make and model and preferably from the same manufactured batch. I think the same problem occurs with LA cells or batteries in series.

    Regardless of battery chemistry if there is a mismatch in battery impedance/resistance the cells/batteries with the higher impedance/resistance will charge or discharge at a slower rate which will mean that regardless of battery chemistry unless there is some "absorb time" the cells/batteries will not be at the same SOC. I doubt this will be a big problem with all chemistries except LA.

    Am I missing something?

    Simon

    Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
    BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
    Latronics 4kW Inverter, homemade MPPT controller
    Off-Grid LFP(LiFePO4) system since April 2013

  • #2
    Think about what happens when the internal resistance of parallel batteries is not equal, as a load is applied and removed. The voltage under load is the same for all of the batteries in parallel, of coarse, but the SOC of each is different. When the load is removed, the OCV for each battery would want to be different, so rapid current flow between batteries occurs to keep the voltages equal. That extra current flow you get as they self-balance is not a good thing.

    Even if you set up your LA to be perfectly balanced on day 1, IR will change over time and this problem will creep in.

    With AGM and lithium chemistry, internal resistance is very low, making it that much harder for the external resistance to be balanced at the precision necessary to keep current sufficiently balanced.

    With lithium, IR can vary with SOC, which adds another complication to achieving a balanced state. Think about how important it is to start with cells as identical as possible. Because the voltage curve is so flat, think about how far the SOC can drift under load between packs, and how much current will flow between them when the load is removed.
    CS6P-260P/SE3000 - http://tiny.cc/ed5ozx

    Comment


    • #3
      I can see what you are saying being a problem if there is a gross difference in IR and there is a sustained load or charge but is this likely to happen in an offgrid system in the real world?

      I would have thought that the hysteresis voltage between charge and discharge would have limited the current flow to balance the SOC.

      Simon

      Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
      BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
      Latronics 4kW Inverter, homemade MPPT controller
      Off-Grid LFP(LiFePO4) system since April 2013

      Comment


      • #4
        Originally posted by karrak View Post
        I can see what you are saying being a problem if there is a gross difference in IR and there is a sustained load or charge but is this likely to happen in an offgrid system in the real world?

        I would have thought that the hysteresis voltage between charge and discharge would have limited the current flow to balance the SOC.

        Simon

        Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
        BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
        Latronics 4kW Inverter, homemade MPPT controller
        The chances of different resistance paths in a parallel wired battery bank is high. Those chances increase do to the DIY efforts in the terminations. It does not take much of difference in how a wire terminal is crimped or anchored to change the amount of resistance seen at that point. Multiply that difference by the number of terminations and you will find a wide variation of total resistance for each path. Even a little extra resistance in each path will change the amount of amps going through it. That unbalance will continue to change as the terminations get older and weaken.

        Also a lose terminal connection has a very high chance of causing a fire due to the increase of heat build up.

        While this may not happen to people who understand electricity because they take the time to check and re-tighten each terminal periodically someone else will not pay attention and more than likely a fire can start. That is why I am hard on most people that DIY because they do not really understand what they are doing with electricity and circuit wiring.

        Comment


        • #5
          Originally posted by karrak View Post
          I can see what you are saying being a problem if there is a gross difference in IR and there is a sustained load or charge but is this likely to happen in an offgrid system in the real world?

          I would have thought that the hysteresis voltage between charge and discharge would have limited the current flow to balance the SOC.
          Consider that even a "steady" load from an inverter can have enough ripple to cycle current direction at 120 Hz (or 100 Hz). Each cycle is an opportunity to induce flow between batteries.

          I think it is relatively straightforward to state that putting batteries in parallel will hurt their cycle life. Yes, it would take a well designed model and a careful study to quantify how much that hurt is, but at the level of advice that can be reliably shared here, the best practices are clear. Avoid parallel batteries for lead acid, especially AGM. Build a lithium battery by paralleling identical cells to get the required Ah, and put those groups in series, balanced, to get voltage.

          And yes, parallel batteries create real problem in the real world. Look at the systems that people come to this forum to get help with. Almost universally, those that are trying to figure out where their battery life has gone are using systems built with parallel batteries, and have experienced the consequences of that directly.
          Last edited by sensij; 01-14-2018, 11:47 AM.
          CS6P-260P/SE3000 - http://tiny.cc/ed5ozx

          Comment


          • #6
            Proves again Karrak you are no engineer, or even have a basic grasp and understanding of Ohm's Law. I am not saying it cannot be done because it is done even on a commercial operation. Utilities do it as a matter of fact dues to the Capacity involved. There are no 40,000 AH cells around. However utilities do not cycle batteries. They are used in emergency standby operation and the batteries used are close cousins to SLI batteries and thus have very short cycle lives but very long calendar lives in Float Service. I am not going to bother explaining cycle service vs float service to you.

            There is a correct way to parallel batteries to reach a specific capacity required. Every EV made does it, amd it is dome at the CELL LEVEL, NOT BLOCK LEVEL. Example a 12 volt Pb battery is a BLOCK of 6-cells in series.

            It is impossible to have battery Ri matched, and even more difficult is matching connectors, cable, and surface contact resistances. You have clearly demonstrated you have no basic understanding, and have no biz answering any questions or giving any recommendations. Exact opposite you have been banned twice for giving very dangerous suggestions.
            Last edited by Sunking; 01-14-2018, 02:56 PM.
            MSEE, PE

            Comment


            • #7
              From a KISS point of view and to limit the number of interconnections I definitely think you should try to organise a battery without paralleling cells or blocks of cells. This is not always possible or feasible.

              I would like to see more information and evidence about the effect on lifespan of paralleling LA batteries if they are correctly wired with the interconnection resistance being at least an order of magnitude less than the individual cell or blocks of cells.

              With lithium ion batteries it is just about impossible to not have to parallel cells or blocks of cells because the capacity of the individual manufactured cells is quite small. Different Lithium ion battery manufacturers have taken different approaches as to whether to parallel cells at an individual cell or block of cells in series level. I can see advantages and disadvantages to both approaches but in my opinion both methods if done correctly are just as valid. I would like to see a technical explanation and some evidence to backup [USER="2334"]Sunking[/USER]'s grandiose statements "If you parallel any modular battery (more than 1-cell in series), does not matter what type, you are going to significantly shorten cycle life. Proven fact Jack."

              Simon

              Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
              BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
              Latronics 4kW Inverter, homemade MPPT controller
              Off-Grid LFP(LiFePO4) system since April 2013

              Comment


              • #8
                Originally posted by karrak View Post
                I can see what you are saying being a problem if there is a gross difference in IR and there is a sustained load or charge but is this likely to happen in an offgrid system in the real world?
                The problem is that all batteries have slight differences in their internal impedance. A slightly higher than normal internal impedance will, over time, result in less charge current to that battery, thus resulting in a lower average SoC. This in turn will tend to age the battery faster, increasing its internal impedance.

                You can get around this by charging batteries long enough that all the batteries really _do_ reach the same SoC - but that can take a long time (days.) That's why UPS systems generally don't have this problem. But RE systems, that are cycled every day and often don't sit at 100% SOC for any length of time, can get out of balance quickly.

                Comment


                • #9
                  Originally posted by karrak View Post
                  With lithium ion batteries it is just about impossible to not have to parallel cells or blocks of cells because the capacity of the individual manufactured cells is quite small. Different Lithium ion battery manufacturers have taken different approaches as to whether to parallel cells at an individual cell or block of cells in series level. I can see advantages and disadvantages to both approaches but in my opinion both methods if done correctly are just as valid. I would like to see a technical explanation and some evidence to backup [USER="2334"]Sunking[/USER]'s grandiose statements "If you parallel any modular battery (more than 1-cell in series), does not matter what type, you are going to significantly shorten cycle life. Proven fact Jack."
                  If you were a real engineer like you claim , and had any real experience with batteries, you would understand Ohm' Law and know parallel blocks of batteries resistance cannot be equalized and that it is impossible to do so. You have proven time and time again you do not understand battery Internal resistance, and cable/connection resistances. Ohm's Law is first week electrical 101 every student gets drilled with by doing thousands of equations until they completely understand. If not, they drop out and go to cooking school.

                  As I have told you hundreds of times and now Jflorey is telling you, both of us are real engineers telling you the exact same thing. In a Standby Emergency Standby Battery plant used in UPS and Telecom, you can certainly use parallel battery blocks (more than one cel in series like a 4,6, 8, or 12 volt Pb batteries. Those type of plants are called FLOAT plants that rarely ever get used and the batteries will equalize of a period of DAYS.

                  In a cycle application where the battery is cycled every day you DO NOT HAVE THE LUXURY OF TIME. Solar is even MORE TIME LIMITED to a few brief hours each day, you cannot get away with Parallel Battery Blocks.

                  We all understand it is not always possible, well those of us with experience, to obtain a specific AH capacity with a single cell. However there is a right way and wrong way to parallel batteries and that is on a CELL LEVEL not BLOCK LEVEL. Every commercial EV made does that. Example a Tesla Roadster uses 99S69P or 6,831 18650 cells and uses a 99 channel BMS. All 69 cells are in parallel @ 3.6 volts, not 7.2 or 14.4, or 28.8 volt blocks. The exact same thing can be done with any battery type, even PB but would be done with 2 volt batteries, not 4, 6, 8, or 12 volts.

                  Perhaps your brain can only understand pictures.


                  Last edited by Sunking; 01-16-2018, 05:44 PM.
                  MSEE, PE

                  Comment


                  • #10
                    Originally posted by jflorey2 View Post
                    The problem is that all batteries have slight differences in their internal impedance. A slightly higher than normal internal impedance will, over time, result in less charge current to that battery, thus resulting in a lower average SoC. This in turn will tend to age the battery faster, increasing its internal impedance.

                    You can get around this by charging batteries long enough that all the batteries really _do_ reach the same SoC - but that can take a long time (days.) That's why UPS systems generally don't have this problem. But RE systems, that are cycled every day and often don't sit at 100% SOC for any length of time, can get out of balance quickly.
                    The higher internal impedance will also mean that the load current will be less, do these balance out? The difference in charge efficiency between the cells will have an impact on the SOC imbalance.

                    In the situation where there is enough absorb time to fully charge the weak cells which of course will be overcharging the other cells I would have thought that this plus the decreased stress on the weaker cells due to the lower charge/discharge current will tend to degrade the better cells at a faster rate so that the cells impedance and SOC will tend to balance out over time.

                    I agree that keeping LA batteries equalised/balanced and for that matter lithium ion batteries is important for longer lifespan. In the case of lithium ion batteries, making sure that all the cells are charged to the same voltage is an important factor to maintain equal stress on all the cells.

                    I would have thought that having a BMS with active balancing where charge is electronically shunted between cells could improve the life of LA batteries over just relying on passive balancing as used at the moment.

                    Simon

                    Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
                    BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
                    Latronics 4kW Inverter, homemade MPPT controller
                    Off-Grid LFP(LiFePO4) system since April 2013

                    Comment


                    • #11
                      Originally posted by Sunking View Post

                      If you were a real engineer like you claim , and had any real experience with batteries, you would understand Ohm' Law and know parallel blocks of batteries resistance cannot be equalized and that it is impossible to do so. You have proven time and time again you do not understand battery Internal resistance, and cable/connection resistances. Ohm's Law is first week electrical 101 every student gets drilled with by doing thousands of equations until they completely understand. If not, they drop out and go to cooking school.

                      As I have told you hundreds of times and now Jflorey is telling you, both of us are real engineers telling you the exact same thing. In a Standby Emergency Standby Battery plant used in UPS and Telecom, you can certainly use parallel battery blocks (more than one cel in series like a 4,6, 8, or 12 volt Pb batteries. Those type of plants are called FLOAT plants that rarely ever get used and the batteries will equalize of a period of DAYS.

                      In a cycle application where the battery is cycled every day you DO NOT HAVE THE LUXURY OF TIME. Solar is even MORE TIME LIMITED to a few brief hours each day, you cannot get away with Parallel Battery Blocks.

                      We all understand it is not always possible, well those of us with experience, to obtain a specific AH capacity with a single cell. However there is a right way and wrong way to parallel batteries and that is on a CELL LEVEL not BLOCK LEVEL. Every commercial EV made does that. Example a Tesla Roadster uses 99S69P or 6,831 18650 cells and uses a 99 channel BMS. All 69 cells are in parallel @ 3.6 volts, not 7.2 or 14.4, or 28.8 volt blocks. The exact same thing can be done with any battery type, even PB but would be done with 2 volt batteries, not 4, 6, 8, or 12 volts.

                      Perhaps your brain can only understand pictures.

                      Have you read the Smartguage article? Have you worked out the simultaneous equations necessary to calculate the current flows into the individual cells/batteries or run some circuit simulations using SPICE or some other circuit simulator to see what impact different interconnect resistances and battery impedance has on the flow of current?

                      The long absorb time necessary to properly charge LA batteries can be a real problem with solar systems, another good reason not to use LA batteries in off-grid systems.

                      You still haven't answered the question why it is wrong to parallel at a block level and left one very important thing out of your diagram, the balancing circuitry that balances the individual cells within the block.

                      Simon

                      Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
                      BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
                      Latronics 4kW Inverter, homemade MPPT controller
                      Off-Grid LFP(LiFePO4) system since April 2013

                      Comment


                      • #12
                        Originally posted by karrak View Post
                        ...I would have thought that having a BMS with active balancing where charge is electronically shunted between cells could improve the life of LA batteries over just relying on passive balancing as used at the moment.....
                        LA cells are not valuable enough to bother with the added expense of the active balancing. In most cases cells of the proper capacity can be obtained.

                        Active Balancers (and shunting) add a lot of complexity and parts, with each part added, adding to the failure rate of the Balancer system. As noted before, poorly engineered balancer failures ruin many cells.

                        Powerfab top of pole PV mount (2) | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
                        || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
                        || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

                        solar: http://tinyurl.com/LMR-Solar
                        gen: http://tinyurl.com/LMR-Lister

                        Comment


                        • #13
                          Originally posted by karrak View Post
                          You still haven't answered the question why it is wrong to parallel at a block level and left one very important thing out of your diagram, the balancing circuitry that balances the individual cells within the block.
                          Myself and a dozen other people have told you many times why. OHM'S LAW. Anyone with any electrical knowledge clearly understands that. You are proving you know anything about batteries or electrical basic principles and pretending to be an expert makes you a [B]Fraud. [/B]

                          As Mike, Jflorey, and I have said many times to you there is absolutely no reason or excuse to parallel Pb batteries in a solar system. They come in every size from 10 AH to 6000 AH.

                          200 AH and less Pb batteries come in in 12 and 6 volt blocks. [B]200 to 800 AH come in 6 volt blocks[/B], 600 to 1000 AH come in 4 volt blocks. 800 to 6000 AH come in 2 volt cells with a back breaking 900 pounds for a 2 volt 6000 AH cell.

                          Having said all that based on the largest controller you can buy, 80 to 100 amps, and a 48 volt maximum battery voltage limit to remain under the 50 volt NEC limit for exposed electrical parts, the largest battery you can have is 800 AH. That would take a 4000 to 5000 watt panel system. You would have to be a fool to use 4-strings of 12 volt 200 AH batteries. Do you have any clue what the Wiring and OCPD would cost would be? Not too mention the frequent battery replacement cost would be going parallel? Nope you do not have a clue. Smart money would use something like [B][I][U]Rolls 6CS25P 5000 series[/U][/I][/B] and comes with a real [B][I][U]Battery Manufacture 10 year Warranty[/U][/I][/B] that will actually be honored and last longer than any lithium. The Chi-Com Winston Lithium Batteries you sell and use DO NOT HAVE A WARRANTY. Go to [B][I][U]Winston's website and search for Warranty.[/U][/I][/B] It does not exist.
                          Last edited by Sunking; 01-17-2018, 01:53 PM.
                          MSEE, PE

                          Comment


                          • #14
                            Originally posted by karrak View Post
                            The higher internal impedance will also mean that the load current will be less, do these balance out?
                            The higher impedance means that the cell will not see as much charge _or_ discharge current. In general, lead acid batteries in RE systems die because they are not charged adequately - it is this effect that causes problems on large paralleled systems.

                            Series string lead acid batteries have a mechanism where you can guarantee 100% charge on all batteries by effectively overcharging some cells. Since all cells see the same current, you can guarantee this by providing X coulombs to the battery. Of course, overcharging 'uses up' water, so equalizing charges - or even long absorption charges - can reduce life as well. So that's a tradeoff.

                            Parallel strings do not have this mechanism, and indeed the weak cells see less charge current due to Kirchoff. So they tend to not see 100% charges.

                            Again, in applications where you can float forever, eventually all cells will reach 100%. But that's not how RE systems are used.
                            I would have thought that having a BMS with active balancing where charge is electronically shunted between cells could improve the life of LA batteries over just relying on passive balancing as used at the moment.
                            That assumes that keeping lead acid batteries at the same voltage is the same as keeping them at the same state of charge. That is less true of lead-acid than it is of other chemistries (like lithium ion.) It will prevent damage by preventing cell voltages below (for example) 1.75 volts, but will not guarantee full charge.
                            Last edited by jflorey2; 01-17-2018, 01:45 PM.

                            Comment


                            • #15
                              Originally posted by jflorey2 View Post
                              Series string lead acid batteries have a mechanism where you can guarantee 100% charge on all batteries by effectively overcharging some cells. Since all cells see the same current, you can guarantee this by providing X coulombs to the battery. Of course, overcharging 'uses up' water, so equalizing charges - or even long absorption charges - can reduce life as well. So that's a tradeoff.
                              I wonder if the same issues occur with cells in series as in parallel. With cells in series as you say the same amount of charge has to go through all the cells. Unless all the cells in series have the same capacity and same charge efficiency some will end up being overcharged and some undercharged. I am sure you have seen the situation where some cells in an FLA battery use more water than others and where at the end of life of the battery that it will one or a few of the cells that have a far higher impedance than other cells in the battery.

                              That assumes that keeping lead acid batteries at the same voltage is the same as keeping them at the same state of charge. That is less true of lead-acid than it is of other chemistries (like lithium ion.) It will prevent damage by preventing cell voltages below (for example) 1.75 volts, but will not guarantee full charge.
                              I wasn't thinking about keeping the voltage the same but more about actively shifting charge from cells that have reached the absorb voltage early and during absorb actively pushing extra charge into the cells which are staying at a lower voltage rather than letting the voltage rise in some cells to the point where they outgas. That is why I specified active balancing.

                              These sort of issues don't arise with lithium ion batteries. With lithium ion batteries you still want to make sure that the cell interconnections don't result in current imbalance.

                              I am yet to be convinced that you can't mix lithium ion cells of different ages and impedances in parallel as they will automatically share the current dependant on their individual SOC and impedance. With LA batteries you would not want to mix cells with different characteristics for the reasons we have been discussing.

                              Simon

                              Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
                              BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
                              Latronics 4kW Inverter, homemade MPPT controller
                              Off-Grid LFP(LiFePO4) system since April 2013

                              Comment

                              Working...
                              X