One Cell SG low and falling, what happens next?

Dave

Apologies, if this has been discussed and request to be pointed to where I can read about it.

I have a 48 V system, two 6V x 8 battery strings in parallel using Crown CR-235 batteries. System has worked great for almost 4 years. Batteries typically cycle between 80 and 100%, only 4 discharges to between 50 and 60 and one to 50 with triggered auto system cutoff (we left a hose running all night, big well pump cycling and it still ran for 10 hours!). So, these batteries have not been pushed hard routinely and serviced regularly.

I've checked SG periodically and about a year ago I started noticing one cell getting lower. I tried a few equalizes but nothing helped. That cell is down to 1.25 now, adjacent cells in that battery are 1.65 and 1.70 or so and cells in rest of bank typically average 1.75. Battery voltage is 0.05-0.07 V lower than the other batteries in the string.

I am confused why a SG that low is not drawing the voltage down terribly much. And, I need to know what I should expect going forward and when it is necessary to replace the battery to avoid damaging the other batteries.

Technical advice appreciated. Thank you

Photowhit

How did you equalize? 
Were the batteries fully charged before you started an equalizing cycle? Very important!
You should see some gains and continue equalizing until there are no gains for 2 hours running (some say 4 hours running!)
You want to sending appropriate voltage to the battery bank to induce equalizing, for a 48 volt battery bank that is likely to be 62-64 volts.

mcgivor

The lowering of SG means there is sulphur not in solution but locked in sulfation crystals. This is lost capacity once it begins it manifests,  the voltage across the cell begins to drop whilst the others pick up the slack, this masks the problem in that the bank voltage appears normal, it's hard to measure individual cell voltage in monoblock batteries so it often goes undetected until too late. The reversal process can take many hours of EQ without loads, this means removal from service if nesesary. Can some information on voltage set-points used, battery ambient  charge source, charging amps, temperatures and any other relevant details be posted. 

By 1.65 I'm assuming 1.265

MrM1

System has worked great for almost 4 years.  ... I need to know what I should expect going forward and when it is necessary to replace the battery to avoid damaging the other batteries.

If the battery bank is 4 years old,  I don' think you can replace just one battery.    It would be time to replace the whole bank. 

Dave

thx mcgivor, yes, posted too fast, meant the overall battery bank is right around 1.275, bad cell is 1.225, and cells on either side of low cell are 1.260 and 1.270. The cells across the rest of the battery bank are remarkably similar with very little variability 1.275 +0.005. Your explanation makes sense for why it only worsens over time. I think I've been equalizing effectively (at 61.9 V) but don't do it very often because there are fewer than 10 times in a year that the bank doesn't charge fully and go through complete absorption cycle in a year. The fact that really only one cell is low SG indicates the bank is being charged/equalized effectively, doesn't it?

I've been using:

Summer - absorb at 59V for 2 hours

Winter - absorb at 59.5V for 2.5 hours

charge values are temp compensated. Not sure what you mean by "ambient charge source", I have 3.6 kW panels, Morningstar Tristar MPPT 150V controller. 12 panels connected in pairs (series), 6 sets are paralleled in combiner box. Panels are nominal 24V, max voltage is nearly 45V or close to 90V for each pair.

In past attempts I have seen no improvement in hours of equalizing so I was hesitant to put the rest of the batteries under that stress to try to bring the one problem child into line. I don't have another programmable controller so I don't know how I could charge that one battery.

mcgivor

What are the manufacturers recommendations on EQ voltage? 61.9V seems a little low, around 64V is common, absorbtion voltage looks normal perhaps 180 minutes would be better, do you have a remote temperature sensor attached?  Unfortunately equalization is hard on batteries but is a nesesary evil to ballance the cells, they should gas freely whilst in EQ, have you observed that? The numbers are not too far apart but what are the manufacturers SG figures? 

Photowhit

Dave said:

...I think I've been equalizing effectively (at 61.9 V) but don't do it very often because there are fewer than 10 times in a year that the bank doesn't charge fully and go through complete absorption cycle in a year. The fact that really only one cell is low SG indicates the bank is being charged/equalized effectively, doesn't it?

Lots of reasons to equalize, and your manual asks for a minimum of once a month;

I'd like to say I understand the manual, but it appears that they want 61.9 volts (I guess temperature compensated) but for 3 hours...

Extending battery absorb times as they age is a common practice. I use a forklift battery and I just went from 3 to 4 hours as it enters it's 8th year.

Dave

thanks mcgivor,

There is a remote temp sensor and the controller definitely adjusts the charge voltage according to temp. The batteries are considered excellent according to SG by Crown standards except the one with the low cell(s).  I can hear them bubbling away in there any time the V gets above 55 or so.

The absorption and equalization values are both higher than crown recommended although they acknowledged there are no "right" numbers and that the system needs to be set up according to individual conditions. I'd have to dig through my notes to find their recommendations but I compromised between their suggestions and those of NAWS, where I bought the batteries and I've tried to hone them over time by watching charge currents during absorb and float to ensure that full charging was happening. i've also gotten widely different advice from different individuals at Crown so you kind of come away feeling like you've gotta figure it out for yourself. i've experimented quite a bit with the absorb voltages and times and settled in on the values that resulted in near steady state current toward the end of the absorb period. I don't really know how to evaluate equalize times other than to take SGs after various times. When the batteries were new, I could definitely correlate equalize times with Voltage and SG increase but the battery bank as a whole has been really consistent for the past 3.5 years (i.e. equalizing does not increase voltage or SG) other than the one battery and, as I said, I have not been able to increase the SG in that one through probably 10 hours of EQ (not all at once) but maybe I still need to do more?

I appreciate your helping me with this. I'm not sure what I should do next but don't want to damage other batteries by their having to take up the slack from this one.

Do you think a "super" equalize (higher voltage, maybe 64V, longer time?) as an experiment would be advised? If so, how long do you think I should go before I remeasure SG of that battery to look for improvement? And, at what point (how many hours of EQ) do I give up and just replace that bad boy?  I could disconnect the other string so that I'd only be doing this to 8 of the batteries.

Estragon

FWIW, I EQ until SG stops rising on the low cell for 1/2hr. When it stops rising, you've done all you can do. IIRC, EQ voltage is ~63v. I normally start EQ when absorb is nearly done (early afternoon on a good day), and run until either SG stops rising, or sun runs out. If the latter, I repeat the next day.

It's a balance though. I want to keep the whole string going as long as possible. EQ is hard on the good cells, but like myself, I'd prefer it if everything fails at once. Who wants a string to die with one bad cell, or themselves to die with a perfectly good liver

mcgivor

You could try a higher EQ voltage under a watchful eye, taking SG readings every 30 min whilst watching temperatures, terminate if approaching 50°C, if there is no rise in SG that may be the where it will remain. To reduce the stress as well as increase the voltage on the individual battery a load such as a car headlight can be connected over the good batteries, this creates a bypass and focuses on the weak one. Personally I would try the higher voltage to start with, keep in mind though at 4 years old they have gone through 1400+cycles so age is catching up, every cycle even a shallow discharge is still counted as a cycle. One other question, when looking into the cell with a light, are there any deposits on top of the plates?

BB.

Probably getting close to needinga replacement battery. I would suggest that you talk with the person who will be handling handling the warranty replacement (they will probably send you through a series of corrective charging cycles and tests).

One cheap and dirty way to equalize just the one battery... Buy 5x 6 volt filament brake lamps and sucker sockets. Wire up the sockets to a pair of alligator clips.

Clip the lamps to the 5x series "good" batteries, and no lamp on the weak battery. Connect when you have charging pe power (solar or genset).

A typical 6 volt brake lamp will let 2.6 amps around the good batteries and equalize the one battery without the lamp.

Monitor current and voltage... You may not need EQ voltage just absorb setting and only the one battery will be equalize.

Or, go and purchase a 6 volt AC battery charger, and clip it on the volt battery.

Safety would suggest that you disconnect the one string while charging (from solar or genset).

Bill

Dave

thanks you guys for the great tips! i will try an additional EQ at higher V in 30 minute increments. one question, i've always waited a while after EQ or absorption to let the H gas dissipate before opening battery box to check SGs. if you're going in immediately after 30 minutes to recheck SG, do you set up a fan to dissipate or just not worry about it?

Photowhit

I understand that I'm being ignored... That's fine...

You haven't described how/when you are equalizing. Understand that equalizing happens only after the batteries have been fully charged. Also read your manual. Crown suggests a3 hour equalizing, monthly or more often;

https://www.crownbattery.com/hubfs/Brochures/Deep_Cycle_Battery_Product_Suport_Brochure.pdf

littleharbor2

You mentioned you have 24 volt panels, 2 in series, 6 strings. Your MPPT controller really needs to see higher voltage to do its job, as designed. You should rewire your array in a 3s4p arrangement. 

mcgivor

As long as there is adequate ventilation it's not usually a problem, take all preventive measures to prevent sparks, jewelry, uninsulated metal tools etcetera, you know, the regular stuff. Leaving the battery box open will dissapate onto the room, if applicable,  if the battery box has a fan to create positive circulation to the outside leave the box closed, never use synthetic rags around batteries they can build a static charge, cotton is best. 

Dave

sorry photowhit , i missed your 2nd comment. yes, i eq after full absorption cycle if i possibly can. unfortunately, in winter, panels in shade by 2:30 so it's not easy to go thru bulk, 2.5h absorb, and 2h equalize. yes manual sez 3h equalize but I've also reas 2h and as i mentioned, it's quite rare that i don't fully charge each day (full absorb) so it wasn't clear to me (and crown agreed) that i need as frequent eq as typical user. i try to di an eq any time i don't get a full charge cycle for 2 or 3 days. i may not be eqing enough so hopefully the experiment will tell me. i just find it odd that 15 batteries are fine if the overall charging regimen isn't pretty close. maybe more problems will start cropping up as they age and this is just the first.

Dave

littleharbor2 - i don't understand why you say i need to series more panels. as is, i typically hit 87-89V as soon as sun hits its stride and that lasts well through the day.  that should be plenty to achieve the low 60s i need to eq, right? what am i missing?

if i series 3, i'll be right around 135 V at max and that's closer to the 150v max of my controller than i'd like to be. during absorb or eq, i've never noticed that the system is limited by low V.

littleharbor2

I thought about that after I posted. Your 3.6 kw array is 300 watt panels. They are pretty high voltage for 72 cell panels. Your numbers do sound high though. 72 cell panels should be in the 35 - 37 Vmp. range.  which would give you say 74 Vmp. when its not real hot out. Your numbers sound like Voc. numbers and, yes 3 in series would push the Voc. limit of your controller on a cold morning.  60 cell panels are really the best format for a 3s string into a 150 Voc. controller. 
If you're having satisfactory results from your setup as configured I guess you ought to leave it alone. One thing. If you were to equalize at the higher voltages discussed above you could have a hard time reaching those voltages on a hot day with a 2s configuration.

Photowhit

Dave said:

littleharbor2 - i don't understand why you say i need to series more panels. as is, i typically hit 87-89V as soon as sun hits its stride and that lasts well through the day.  that should be plenty to achieve the low 60s i need to eq, right? what am i missing?

I ran back to check to see if you had described your panels before. I didn't find a direct reference to them, I know some higher voltage panels have been made, some 94/96 cell panels. But likely 72 cell panels and you are seeing VOC. Some controller will show this before producing/charging in the morning, reading basically an open circuit or perhaps you are measuring an open circuit with a meter. 

I plugged in numbers for a known 72 cell panel with my local temp boundaries,  into Morningstar's string sizing tool and it gave me No (as in 0) array configurations until I changed the max low temp to positive 10 degrees. Looks like the link includes the info, if you change the max low to -10 it will show no results, The controller doesn't know this and likely just works in PWM mode. No harm no foul. It's not a huge advantage with correctly sized panels.

http://string-calculator.morningstarcorp.com/#manufacturer=1114&module=30202&product=23&vmin=48&vmax=62&tmin=10&tmax=95&tminunits=f&tmaxunits=c

Please note it only recommends 3 panel strings.

BTW - after almost 5 years I had my first PITA glitch with Midnite Classic CC. I added a display and can't get the 2nd controller to consistently come up across the link to the first controller. After reading about people having issues with them not going into equalizing mode, I just start it manually. I had been over the summer anyway, not wanting to equalize on hot days (my battery lives outside) I had actually caught mine equalizing so perhaps I got one of the good ones...lol, but no reason to worry about it, with a forklift battery I'm checking electrolyte pretty regularly.

Dave

wow, thanks Photowhit, a new thing for me to worry about! just kidding. These are 72 cell panels and they did list a Vmp of 37 as I recall.  I'll try to look at some logs and see if I can figure out when and how long those high voltages are occurring. Maybe I should reconfigure into sets of 3. Even at the high Vs I see, I would still be below the 150V max.It does get below 10F here though.  Based on what you said about needing longer absorb as batteries age, I plan to do some more plots of current versus time plots in absorb and float mode and see if my absorb times are still getting me close to steady state. Thanks for mentioning that.

Estragon

It might also be worth checking temps with an infrared thermometer, especially the problem cell, periodically, for anomolies.

mcgivor

Think about this, Vmp or voltage maximum power, an equalization is usually performed after a full charge, this means reduced current which results in increased voltage, this is seen by the observation of the array voltage as the batteries approach full charge. Using my system as an example, when batteries are in a lower state of charge accepting all available current, the array voltage is ~ 70V,  as, the batteries near end of absorption, or 2% of battery capacity, the voltage is in the 96V range, Voc is 102V. It would stand to reason that the Vmp is not a linear value but varies according to current demand, therefore MPPT would then be able to support a higher voltage for equalization. Any thoughts?

Dave

that makes sense to me. and, i think I've mentioned that the system has not had a problem reaching and holding eq voltage as long as there's reasonable solar available.

should have mentioned earlier that batteries are very close (directly below) to controller and i ran 4/0 wire so voltage drop should be minimal.

Photowhit

Dave said:

that makes sense to me. and, i think I've mentioned that the system has not had a problem reaching and holding eq voltage as long as there's reasonable solar available. 

I don't think it's a problem, once the controller reaches absorb or not long after they will often revert to just a standard PWM delivery of electric, and when equalizing, there is only the need for a small amount of current.

What you lose is the little bit extra current when in MPPT mode. I think it may even be able to use MPPT but just not as efficiently. I think The Morningstar actually displays when it's in MPPT mode. I haven't used one but have helped setup a system with one. I don't know how accurate that is or if it just stays on all the time when it's charging.

Photowhit

boB, designer of some Outback and Midnite charge controllers confirms this in this thread. 

http://forum.solar-electric.com/discussion/comment/386802

It would kick back in if the loads were present high enough to require it to maintain voltage. Since there is little 'extra' from 60v to 70v, it's not a huge lose.

Dave

Curious what all your thoughts are if I do decide to replace this battery, which seems probably at this point, I was wondering about going with a Duralast Ultra that has slightly less capacity = 230 Ah. My rationale is that it would better match the 4 y old Crowns that have probably seen some loss from their original 235 Ah. Is this bogus reasoning?

Estragon

I suspect identical batteries could vary by more than that (~2%) when brand new.

Replacement might be feasible if the new one is fairly cheap, and has a similar charging regime as the crowns.

BB.

I suspect that FLA batteries can vary by as much as 20% between in capacity between cells... One reason to avoid going below 20% state of charge... Most rechargable battery chemistries fail if taken below 0% SoC and batteries actually begin to reverse charge.

Bill

Dave

Follow-up - I equalized at 63 V for a half hour, checked SG and then EQd another half hour and checked again. It seemed to make no difference at all so I went ahead and bought a Duracell Ultra 6v 230 Ah from Batteries Plus Bulbs for $111. Thanks for all the help and advice here, I really appreciate it!

BB.

Best wishes for you and yours moving forward.

It sounds like you made the correct decision.

-Bill