Today’s adventure: a couple of rainy days caused low battery levels, but not too low I thought - still 30% or so; these are lithium batteries and can deep cycle. They are “smart” batteries and if one is full in series, none can charge further - so they should all be at the same charge level all the time. But a couple had gotten out of step somehow and when they reached zero everything shut down.

How to bootstrap it? With no battery output (since a zero battery turned itself off and would not let the battery bank show any voltage!) - there is no way to activate the inverter and let street power run the battery charger. With no battery power, there is no way to turn on the MPPT controller and charge the batteries!

I could rearrange the banks to put four batteries with remaining charge in series because I have an 8-battery system, and get things restarted; but if there had been only four like when I first installed the system - I’d be in trouble.

Another thing that happened. Before I figured out the battery problem, I was trying to switch back to street-power. Because power from the street comes to the inverter first and then the inverter powers the distribution panel, when the batteries are down, I cannot get street power to the distribution panel. I could install a manual bypass, but it is not a commonly needed item and it is a large amp switch. So I removed the inputs and outputs at the inverter and bypassed manually. That worked fine. But in the process of disconnecting or reconnecting, I must have loosened the neutral connection to the inverter. So when the inverter was working again and I checked voltage, I only checked across the two hot legs - yay, 240V. I did not check that each leg was 120v from neutral! They were not: one leg was at zero and the other was at 240. I found this discrepancy fairly quickly after only destroying an outlet strip, the oven control electronics from a very old stove we were wanting to replace, and the controls of an old Sharp microwave oven with the 4, 7, 9 and Clear buttons not working. Woohoo we have a new stove out of the deal!

  • cerement@slrpnk.net
    link
    fedilink
    arrow-up
    0
    ·
    1 year ago

    half joking – but was imagining something along the lines of RAID 5 – setup of 4 batteries, but only have 3 batteries worth of power at any time, but if one dies, it would get swapped out with no loss of power

    • fkn@lemmy.world
      link
      fedilink
      arrow-up
      1
      ·
      1 year ago

      First, there are always power losses. Second, batteries shouldn’t just fail. Any pack where a battery is failing is in much worse condition than a raid like solution would solve (or in ops case all of the batteries are over drained and a single fresh one doesn’t necessarily solve the problem). Third, a smart battery balancer can detect when a cell is going bad long before it fails. Fourth, building this system adds complexity that wastes energy, capacity and time…

      That said, this would be reasonably simple to setup with an Arduino, some relays and some voltage monitors.

      • CadeJohnson@slrpnk.netOPM
        link
        fedilink
        arrow-up
        1
        ·
        1 year ago

        When I purchased the first four batteries, I bought a “balancer” from the vendor (KiloVault via AltEStore), but after a year it melted and they said a)it was no longer sold, b) it was not necessary, and c)I could have my money back. So that was fine. But now I am thinking it is time to look into this balancer concept again! I dabbled with an Arduino one time, but I hope there will be an off-the-shelf solution.

        • fkn@lemmy.world
          link
          fedilink
          arrow-up
          1
          ·
          1 year ago

          Balancers are well understood now thanks to the proliferation of electric cars. There are two kinds of balancers, passive and active. Passive balancers work by creating a resistive load across any batteries that are more charged than the lowest charged battery. This is exclusively a power drain, but as long as the batteries are all the same size and have similar capacities this won’t waste too much power. Active balancers are usually much more complex and expensive but they actively transfer power from charged batteries to low batteries through capacitors. There are still some losses but overall they are significantly more efficient than passive balancers.

          If you get a balancer, regardless of the type, you need to make sure that it is capable of supporting the battery size you have. Balancers that are too small are going to burn up really fast. You should also try to get a balancer with cell health information. Again this is more expensive, but will usually be part of a higher quality balancer.

          A couple of notes. I am assuming you have lifepo4 batteries and I assume you have 12v batteries (4s 12v = 48 volts). I also assume you got “smart” batteries since you said it shut off (you should never get non-smart lifepo4 packs for this kind of application, you are just asking to burn your house down). The “smart” part of the lifepo4 battery is that the battery has a built in balancer for it’s internal cells (usually called a bms). These balancers are usually resistive unless the pack was super expensive and/or custom. What this means is your 48v system is actually a 16s 3v lifepo4 pack, but you are only balancing chunks of 4, but not those chunks of 4. This is where a balancer can be helpful. However, it is also possible that your battery packs have a “smart overcharge protection” where it effectively prevents the individual 12v battery from overcharge and they basically have created their own passive balancer system with this mechanic and an additional passive balancer will only reduce your total ability to charge. This works because each battery monitors its own internal voltage and shunts any excess voltage through a big ass resistor… Which is exactly what a passive balancer does. At that point you need to decide if the money should just be spent on a cell health monitoring system or an active balancer… And usually the price is not worth it for most people. Usually it’s just more cost effective to buy a second set of identical batteries and wire them in parallel (one for one) with the existing batteries. Note, you don’t have to match the existing batteries, you just need the new batteries to match each other, and to charge every battery (old and new) to the exact same voltage before wiring them together. (These big batteries, a single tenth of a volt difference can be 50-200amp discharge).

          • CadeJohnson@slrpnk.netOPM
            link
            fedilink
            arrow-up
            1
            ·
            1 year ago

            Thanks! Great info. You are right, I have lifepo4’s -two sets of 4x12V paralleled. There is an app with them to check the internal cell voltages, overall voltage of each unit, state of charge and number of cycles.

            You mention the big current that can arise from paralleling batteries - are you suggesting a series-parallel wiring? I have had a lead-acid bank like that previously with cross-connections between series groups at each voltage level. Those cables were usually carrying small currents when all the batteries were in good condition. In a healthy lifepo4 bank with voltages even, could I set that up and would it help? Would a load resistor help (it normally would only see small voltage drops on the order of 0.1 or 0.2 volts, so a load resistor of a few ohms would pass very little current. I have also asked this question to the battery vendor. Awaiting their reply.

            • fkn@lemmy.world
              link
              fedilink
              arrow-up
              1
              ·
              1 year ago

              You should be able to have a 4s2p setup no problem. It also sounds if you have the individual batteries internal voltages and cycle count you already have the information you need to identify failing batteries… The best indicator is internal resistance of the battery, but cell voltages should with too. Basically, when under load if any individual cells voltage drops more than the other cells by a significant margin you are looking at a battery with issues. The under load part is the critical part… Because just like lead acid batteries the surface charge of a cell can be deceptive. Most good lifepo4 batteries can get at minimum 500 cycles, with some designed for 2000+… And the extra kicker is if you only charge them to 85% you can get another 4-5x that, typically draining then isn’t an issue (as long as you never go below minimum voltage).

              As far as the 4s2p, I think that would actually be preferred, but I would defer to the manufacturer. Just like acid batteries, the parallel configuration allows them to support each other for a more consistent experience.

              The sweet spot for lifepo4s are better 10% and 70% soc. Usually that isn’t quite enough power for cost which is where the 85% number is used. It sounds like you just got into a bad state where you over discharged one and it’s not actually damaged, but having them in a 4s2p should be better than having the two 4s strings in parallel. Again, just make sure the voltages are even. I wouldn’t use a load resistor. The amperages is based on wire gauges and material. But again, do what the navigate suggests.