1 of 5 strings overheating BADLY!

Options
I have a 4 year old, 24V bank of 20 Roll S-530's, charged by 2 stacked Xantrax and one Outback PV charge controller. There are five strings of 4 batts. Fo the last two year I have been charging them as Rolls recommends: charging at 15% of c6 rate which is 296 AH per battery, using their voltage settings. I usually charge the bank at 225ADC split between the two Xantrax, or about 112ADC each. The setting computation in AAC according to Xantrax is 25% for my 24V system, or 27AAC on the panel.

Problem: The generator runs too much. And for some time, batteries seem to be quite hot when charging, but I only have a thermostat in the middle string (and THAT seems to be the culprit!). I can barely get the bank up to 1.200 SG. The charging system was not originally set up for the aggressive charging that Rolls recommends. I suspect they were badly sulfated from undercharging for the first two years, and probably not watered well. I used the recommendation that Rolls has for corrective charging for heavy sulfation: equalize at 5% of c6 (1/3 normal rate) for as long as it takes to reach the charging system's max Volts of 32, then 4 more hours.

It looked good for a while. After 8 hours, the slow charge rate was keeping the batteries at 94F. I thought I could continue the process without the overheating I experience during normal equalization, and clean up the bank. Unfortunately, over the next 8 hrs, the middle string (ONLY the middle string) overheated, badly!

After 8 hours, V=27.2, T=94F, SG=1.146 (avg of 4 test cells)
After 8 more, V=25.1, T=150, SG=1.178.
SG went up in all but the overheated middle string, and overall volts was down.

What's going on?? The water in the other strings has some occasional bubbles, like you might see in a glass of resting soda water and was quite clear, and the strings were a balmy 95F. The middle string was gassing profusely: the color was muddy like when I do a normal equalization.

Why just the middle string? Any explainations?
I can't recondition the bank with the overheating. I'm thinking of removing the middle 4, and trying to recondition the rest.
Am I ever going to get this bank back into shape, or is it junk? Any suggestions?

BTW, I am using an expensive electronic SG gauge, so the reading are accurate.

Comments

  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Options
    Re: 1 of 5 strings overheating BADLY!

    you may have to pull that middle string as there may be a bad cell or cells. after pulling it see if the other strings act better with better bubbling, which is an indicator some good current is passing through them.

    as to the middle string, you may need a dmm to measure the at rest voltage of each to determine if one is off. it almost sounds like a partially shorted condition in that middle string dragging down the other strings. if there is a shorted cell in that string it would cause the whole middle string to cook as it's overcharging. because this string is drawing so much it is pulling charge power away from the other strings.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,491 admin
    Options
    Re: 1 of 5 strings overheating BADLY!

    Get a DC Current Clamp Meter (an inexpensive one here) and see which string is carrying the most current.

    If you have a shorted cell, that one string may carry most of the charging current (and prevent the rest of the bank properly recharging).

    Also, check for dirty connections and excessive voltage drop across each connection and length of cable (use a DMM set to 2 volt maximum scale or 0.200 volts). All wiring and connections should have very low voltage drops (do this with heavy charging and/or discharging currents). Poor connections or internally corroded cables/connectors can cause poor current sharing (it is possible that your "best string" is carrying the most current--and you may damage it from over charging/over use).

    As Niel suggests, you may want to disconnect the "hot string" and see what the other strings will do and if you can get them fully recharged.

    And if the bank is sulfated--heavy equalization may recover some capacity by knocking off old sulfate crystals. But if the plates were exposed--then they may not be recoverable.

    Also, excessive equalization is not good either. It can cause positive grid corrosion (oxygen is formed during equalization at the positive plates). If you see the positive posts being "pushed" up in the case, you may have positive grid corrosion (not recoverable).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Options
    Re: 1 of 5 strings overheating BADLY!

    CallMeChaz, I agree with the advice given above. In order to diagnose and treat your massively parallel battery bank you should understand the nature of parallel battery problems. Bill has just an hour or so ago written a good explanation in another thread:
    http://forum.solar-electric.com/showthread.php?14674
    Not quite as useful as Bill's post is an article in the latest issue of Home Power magazine (#147 Feb-March 2012). The article describes a single string as "Most Desirable". It says "some designers prefer two strings for redundancy -- in case one battery or cell fails, there will still be one functional series string at the correct voltage to rely on". It also says "three parallel strings are considered marginally acceptable". More than three strings are "undesirable".

    The Home Power article does not do a thorough job of explaining why multiple strings are so problematic. Let me add to what Bill has written: Parallel strings of identical batteries with perfect connections are themselves thermally unstable. If one string is a bit warmer than the others (inevitable) it will draw more charging current than the other strings which will make it warmer which will make it draw more current which will make it warmer which will make it draw more current which will _____ (fill in the blank and buy new batteries).
    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • BB.
    BB. Super Moderators, Administrators Posts: 33,491 admin
    Options
    Re: 1 of 5 strings overheating BADLY!

    Thank you vtMaps for the kind words...

    To add what vt was saying about thermally unstable... The hotter a lead acid battery becomes, the lower its charging voltage should be.

    So, a hot battery drops its voltage requirement, the charge controller continues charging or even sees the "voltage drop" as a load/undercharged battery and even ups the current. The hot battery gets even hotter, etc. --Thermal run-a-way.

    Having the remote battery temperature sensor on the center/hot string helps limit thermal run-a-way on the hot string, but the cooler strings still may not get fully charged as the charge controller has ramped down the charging voltage because of the hot string.

    What is your charging current, and can you measure current in different strings yet? (dc current clamp meter)

    We have typed here about the ability to isolate strings using a A-B switch. Not sure that I would recommend normally--but for your case, it may help (or not--damaged batteries usually cannot be recovered to any significant degree).



    Blue Sea Battery Switch 1-2-OFF 350 Amp




    wind-sun_2197_959424
    Blue Sea 9001e 350 Amp Battery Switch Make-before-break contact design allows switching between battery banks without power interruption Ignition protected - Safe for ... mounting Meets American Boat and Yacht Council (ABYC) requirements for battery switches 3/8"-16 tin-plated copper studs for maximum ...




    -Bill

    PS: I will add vtMaps' comment about thermal stability of parallel strings to the other thread--That is one that I really have not addressed (or even thought about) before and is a very pertinent point.
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Options
    Re: 1 of 5 strings overheating BADLY!

    I may be one of the few individuals that believe five parallel strings can be managed. You do have to monitor the current on each string for reasonable matching. I find that some of the strings will supply more current initially then gradually trade off to dominate current to other strings. As long as things level output within a 50% discharge area it will be okay. Same thing happens on recharging.

    I recommend you put current shunt on each and a simple 200 mVdc digital voltmeter to readout current on each shunt. Also consider connecting each string with a 350 amp Anderson connector tied to a master battery line bus. This allows you to break the parallel groups down for maintanance and testing easily.

    I have my parallel bank with each series group on its own heavy duty castor roller rack so I can disconnect and slide out a rack for maintanance.
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Options
    Re: 1 of 5 strings overheating BADLY!

    i agree with rc that it can be done, but it does become more of a challenge maintaining balance with that many strings. even something as simple as having a battery up against another battery will breed inequality due to heat build up leaving batteries on the outer perimeter to be cooler. if that was applied to strings as in an inner string then the inner strings will be hotter. rc's suggestion of being able to monitor individual strings could expose inequalities before it becomes a problem for the batteries.
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Options
    Re: 1 of 5 strings overheating BADLY!
    RCinFLA wrote: »
    I recommend you put current shunt on each and a simple 200 mVdc digital voltmeter to readout current on each shunt. Also consider connecting each string with a 350 amp Anderson connector tied to a master battery line bus.
    Do you have fuses or breakers in each string? I have seen it recommended for 3 or more strings. I don't know much about the Anderson connectors... are they safe to disconnect when carrying DC current? If you had a shorted cell in one string would it be safe to disconnect the Anderson connector?
    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Options
    Re: 1 of 5 strings overheating BADLY!

    it also occurs to me that the currents you list for charging don't add up to the abilities of those controllers. a xantrex cc is 60a and the fm80 would be 80a full power. now xantrex has an 80a model, but it is a high cost high voltage model and i didn't think you opted to go with that one. this adds up to 200a max (10% rate) and is a far cry from the 15% mark of about 300a. if the outback is the fm60 model then you have even less current available maximally for a total of 180a.

    what did you find with that middle string?