Charging large Surrette/Rolls batteries

Vic

Re: Charging large Surrette/Rolls batteries

I am late in reading this thread ... My 2c:

twistedtree owns this bank now, and it represents a large investment, so ...;

I do not racall your location, and wonder about (as mentioned before) the bank temperature, AND the presence of functioning BTSes (Bat Temp Sewnsors).

Highish Asorb voltages should mix the electrolyte well -- much much more than high current-rate Bulk charge, alone.

YES ! The 4KS25s are very tall, indeed.

I would agree that:
1. Cycling this bank to 50% SOC or even lower, and High-Rate recharge at 10% of the C/20 Rate (ie 135-ish Amps) would be very good. You might want to do this several times, depending upon results.

2. Banks which are lightly cycled and lightly charged on every cycle seem to evidence widly varing capacity between the cells in the bank.

3. The ability to charge at high rates on occasion seems to be very important to these EXACT Surrette batteries.

And, in addition, Be certain to temperature-compensate the EQ Voltage when using MX-60 (at the very least, MXes), as this is not done by the MX. Personally, I note the Vasorb on the day of the EQ, and compensate Veq based on the difference of Asorb V reading, and the Asorb setting in the Charger menu. Surrette is very adament on Veq being above 62 V MINIMUM.

AND, WHAT HYDROMETER are you using to measure SGs ? Are you temp-compensating the SG readings (only important if the bank is fairly HOT or COLD)?

These batteries have an astonishing amount of electrolyte, so adding about 4 gallons of distilled water in a six month interval is not alarming to me, but I would, personally, not wait that long to check, EQ, or add water. There appears to be about 100 Lbs of 1.265 SG electrolyte per two-cell battery. There is quite a bit of headroom over the plates !

ENUF for now, will re-read this thread and think about it more.
An additional note: Surrette mentions that these batteries are designed to by (heavily) cycled. This keeps them healthy. I have to force myself to drop the PV input for 4ish days from time to time, and run the large genset to get the charge current up to about 180 Amps or so. It is actually good for the batteries -- Light cycling large batteries ALL the time is like well conditioned athlete laying around for months, only getting exercise by walking to the refer and microwave ... MHO. Good Luck, Vic

niel

Re: Charging large Surrette/Rolls batteries

Dave Sparks wrote: »

I am still puzzeled by your saying that one cell drops voltage very quickly when you load the bank. I say this because you have a great battery, and appear to be not far off of what i expect if you are completing charge near noon. Stay away from AGM's and solve this problem. What is Surrette telling you? It sounds like an intermittent. You can for test purposes run the voltage slightly lower and not use the "bad battery".

dave,
i was only referring to his rate of charge being too low and that i wouldn't recommend that rate even for the more efficient agm batteries. i think being his previous bank had failed at 7yrs with a low charge rate that we might suppose that the same charge rate problem may not have been rectified. i do agree he has a new set now and this problem needs solved for these batteries. theoretically, new agms might make it a tad better, but the low charge rate still exists and is a problem even for an agm's better efficiency. do not read into my post as a recommendation for him to go out and buy new agm batteries.

twistedtree

Re: Charging large Surrette/Rolls batteries

Dave Sparks wrote: »

I am still puzzeled by your saying that one cell drops voltage very quickly when you load the bank. I say this because you have a great battery, and appear to be not far off of what i expect if you are completing charge near noon. Stay away from AGM's and solve this problem. What is Surrette telling you? It sounds like an intermittent. You can for test purposes run the voltage slightly lower and not use the "bad battery".

I'm confident I have a bad batter and/or cell. I think Surrette will replace it under warranty.

I've muddled the issues a bit here. I've found the batteries difficult to get fully charged - that's the primary subject of my post. I now have a dead (presumed) cell, and I'm wondering if it's related to the charging challenges I've had, and the high charge voltages required to reach full charge.

I'll be returning to the site early next week and will pop the lid off the failing battery to figure out which cell (or perhaps confirm both) are bad so I can complete the warranty claim.

I hope to replace the bad battery and get many years of good service. For $14k I darn well should.

twistedtree

Re: Charging large Surrette/Rolls batteries

In response to Vic's comments....

The batteries are in my basement, and their temp ranges from 60 to 70 deg most of the time. The highest I've seen them get is high 70's during EQ. That's measured with a battery temp sensor which controls all the charge sources, and confirmed by the thermometer that's built into my hydrometer.

Speaking of the hydrometer, I've got a really good one (I think). Freiz or Freis or Freil or something like that? It's nice and big so it can be read accurately, and has a built in thermometer. Back when I first started having issues I thought maybe my hydrometer was off, so I bought a good one to rule that out. Both agreed exactly on their readings, btu now I just use the good one 'cause it's easier to read.

All the charging is temp compensated. There is one sensor that attached to the neg terminal and feeds into the Xantrex XW system (6048 and 2 MPPT 60's).

And all the SG readings I've been throwing around are temp-compensated as well.

I'll have to double check the EQ voltage. Maybe I'll jack that up a bit.

As for drawing the batteries down to 50% periodically, that's what I was trying to do with the gen start set to <= 48V for 2hr. I think that represents 50%, but quite frankly different Surrette publications say different things. Maybe I need to drop it lower. But the idea is that when the sun don't shine, I get a 50% draw down followed by a hard charge by the generator. The house is in Vermont, so we get stretches with no sun.

135A charger rate? That will be more difficult, especially in conjunction with a deep discharge. I'd have to run the gen and solar at the same time for bulk. If I can time 3 days of bad weather followed by a sunny day.... Or maybe I need a supplemental charger. I've got plenty of gen power to spare.

Cariboocoot

Re: Charging large Surrette/Rolls batteries

I don't know if you've mentioned it (the longer a thread gets the harder it is to find some specific piece of it) but what has been your criterion for instituting an equalization cycle? I'm looking for reassurance that this hasn't been occurring "every so often" regardless of the batteries' need. Too much EQ can damage a battery too. (As if you didn't have enough to worry about - sorry! ) EQ Voltage on a 48V system is usually no more than 63 Volts.

You could have your gen start as low as 42 Volts, although 46 would probably be more "normal".

twistedtree

Re: Charging large Surrette/Rolls batteries

When do I EQ? It depends, but it's based on the SG readings. I've been finding the SG reads high and steady in the summer, so I check less often (60 days). Then I check more in the winter to be sure things are OK. Sometimes I'll run an EQ for an hour or two just to see if the SG comes up, but it's really settled in pretty well at this point. I checked in Oct and ran 1hr EQ, then I was away all of November. While I was away is when I first started getting alarms.

I guess the simple answer is that I don't have a set pattern either for checking SG or for EQ. It's more based on weather and system performance, and a sense that I should check on things. If any of the checks detect something is amiss, then I monitor much more closely until I'm confident it's back on track again.

I've heard lots of different advice on when to EQ. Some say do it every 30 days. Other say only do it when SG varies by more than X. I guess I follow a bot of a hybrid, looking for variation, and sometimes just running an EQ because it's been a while and i want to see if the SG will come up some more.

So hows that for a non-answer?

As for the gen start voltage, my inverter cuts out at 44V, so I can't go below that. Around 46V or so my 48V to 12V converter cuts out, so I don't want to go below that either. I'd also rather not go below 50% DOD since in Nov and Dec it could happen every few days to a week. That's what was happening in Nov when I was away. I'm more comfortable at 47V as a minimum, but I'll have to recheck the Surrette literature and figure out which I believe represents 50% DOD.

Dave Angelini

Re: Charging large Surrette/Rolls batteries

FYI Surrette has always ( I installed my first set for HP in 1978 ) recommended a preventive EQ every 90 days.

Vic

Re: Charging large Surrette/Rolls batteries

Hi Peter,

Thanks for the additional info on your system.

I, too, use the Freas hydrometer, altho the one I got about 5 years ago, that had the thermo built-in had too little volume in the bulb to sufficiently fill the tube, so now use the Freas without the thermo, and use a separate thermometer. I think that this is as good as you can get with a glass float hydro.

Your batteries are in a much cozier environment than mine, so seems that your temp variations are fairly small.

I do believe that Surrette specs 62 VDC Minimum for eq. Monitor your lowest SG cell, and when the SG stops rising, further EQ (short of the Corrective EQ) does little good. And, I agree with what you have said regarding a Warranty replacement cell -- if Surrette will replace one or perhaps two cells under warranty, then I would try for that.

I have one home made DC charger which will charge individual cells up to about a 70 Amp rate. Have used this in the past to EQ a single cell which was the only laggard in a bank, but this is not available to most folks here. BUT, this approach would allow a corrective EQ on a single cell in your bank, while not stressing any other of your cells.

Regarding deeply discharging your bank and aggressively recharging: Personally, I would leave your GenStart parameters where they are, and do the discharge/recharge manually. This is worth a try, but again if you can get any suspect cells replaced at little or no cost, it is probably worth considering.

My inverter/chargers and do 150 Amps, and to go above that for any period of time, I need to add the PV/MX-60. 100 Amps should be better than 60. The goal is to expose fresh lead to the electrolyte.

Surrette recommends an EQ Voltage range of 61.9-64.1 Volts.

I'm sure that you know this info, but just in case there is something that you have not read, Surrette Bulletin #605:
http://www.rollsbattery.com/content/preventive-maintenance-charging-and-equalization-605

Every EQ charge takes a bit out of the battery bank -- Positive Plate erosion, BUT, it is much better to EQ a bit too much than EQ too infrequently. Tardy EQ can result in sulphation that is difficult to EQ away.

Surrette, and other battery manufacturers say that EQ should be performed when the spread of SG readings between the highest vs lowest cell is more than XX points of SG. Think that Surrette mentions something like 25 points -- lets say 1.245 to 1.270 or more.

Your battery bank IS large. When you make changes to charge settings (voltage and time), it may take quite some time for them to appear in SG readings etc. And, it appears that you are away from the site in the Winter, which makes tweaking the settings and monitoring the resluls a bit more difficult.

To me, it is interesting that your Summer charging seems quite adequate, but there is more trouble in the winter, and in Vermont, seems that there might be weeks of solar challenge.

I hope that the XW's chargers work even better than my SW+'s ... I've never EQed from the inverters, and seldom am happy with the manner in which the SW+s end their charge cycle. They seem to have a mind of their own, and the temperature compensation of the charger is far too coarse, but they can do 150 Amps of bulk, and that is what I use them for.

I do wonder what parameter you use to end your charge cycle from the generator. Guess that there is Bulk, and Asorb for dome few hours, and the generator is stopped after X hours of Asorb ?

My bank's temperature varies more than your Root Cellar type environment. Today Bank A was at 14 degrees C, and Bank B was 9 degrees C, and this in in Sunny CA !

Good Luck with the bank, hope that you will keep us informed about your findings. You are paying attention to your valuable bank. Seems that most of my neighbors pay attention to their batteries only when the lights go out. That, IS the time for maintenance ! Vic

blackswan555

Re: Charging large Surrette/Rolls batteries

How long does EQ usually take you ? I look after several sets of Rolls from 683 ah @ 24v > 2430 ah ah @ 48v (including a 24v set same as yours) The best I get (after a 6 hr absorb) is about 3 1/2 > 4 hrs till all sg`s stop rising for 1 hr,
If you do not get the fully "bulked" before, (And as mentioned, these are meaty batteries, they take a lot of "pushing" up the hill, Once the roll over the top, not so difficult) Just normal Eq can take an awful long time, 16 hrs + is not out the question,,, If you can keep the temp down (yes I have been there)

Tim
Ps Rolls should be able to tell you what went wrong when they get the cells back

Edited to add, EQ v around 61.5, I try to EQ about every 100 X 50% DOD cycles,

Ralph Day

Re: Charging large Surrette/Rolls batteries

Mike,
I don't use a voltage to decide your 50%, use your SG reading. Voltage readings don't give you the SOC (in FLA batteries) as accurately as SG does. I have the CS 8 volt units from Surrette, and have had a similar problem with 2 cells...both on positive terminals. After much email back and forth and a heavy eq charge where their sg's did not rise much at all I was told that I would have warantee replacement of those cells...pro rated of course. The bank is 6 years old. They don't want the old cells shipped back to them, hmm.

I've seen the KS in the 2 volt units, the 4 volt units must weigh a heck of a lot! My CS cells are about 110lbs each so I'm working out the logistics of removal and replacement. I'm thinking the chain hoist is in order.

Luckily I have utility power now so long slow charging or long eq charges don't have my 10kw diesel genset banging away with little current being used. Monthly cost for utility is $40 before any usage charges. Overall cost is more than diesel fuel used in a year, but easier to use and the maintenance is done for me;)

Ralph

twistedtree

Re: Charging large Surrette/Rolls batteries

Regarding length of EQ, it varies. Way back when I discovered I had low SG readings (since corrected), it took 24hrs of EQ to get the SG up where it belonged and until it stopped rising. The last one I did was 3 hrs. I think I'm going to run an extended EQ (probably starting around 6 hrs) next week just to see what happens.

Regarding local vs remote operation/monitoring, I do use the house year round, but I'm not there 100% of the time. Being away a full month is unusual, but I went to antarctica and it's a long trip. Of course that's when things started acting up.

What determines the end of absorb? That's a really good question. Most literature, including Surrette's, says it's when acceptance current drops to C/50 (2%). And they further say that takes 2-4hrs or so. The challenge of course is that most chargers can only see how much current they are putting out, not how much is actually making it to the batteries, so ending absorb when current drops to C/50 is problematic when you have an inverter drawing significant current. That's why most chargers are timer based.

This is where I think Xantrex got it right (or at least more right than anyone has gotten it so far) with the XW system. Because the components talk to each other (wow, what a concept!), the various chargers can figure out how much is actually coming and going at the battery. It's not perfect because you can still have DC loads that are not visible to the XW components, but in my case that's about 1A so it's insignificant relative to the inverter load. So I think it works quite well. As a result, absorb ends when accept current drops to 2% of C. There also is a timer, and if that pops absorb also ends. I've got my timer set high. I don't recall exactly how high, but it's probably 5 hrs and is really just a safety net. Now the 2% cutoff has some issues, but most can be worked around. This is where there is still room for improvement in my view. The 2% number is fixed - that's the problem. The workaround is to lie to the charger about the size of the battery bank. I've got mine set to 800Ah which causes Absorb to end at 16A which is 1.2% of C. It takes around 3 hrs to reach that current. These setting apply to the inverter/charger as well as both MPPT60s.

You really should be able to adjust the 2% cutoff rather than screw with the battery capacity. Messing with battery capacity messes up the battery level gauge, but the gauge is pretty useless anyway. So far that's the only side effect that I'm aware of.

All in all I REALLY like the XW system. It's the most cleanly integrated and well though out system, in my opinion. The physical packaging is clean and attractive and goes together well. And it's the only package that really starts to act as a system rather than a pile of components. I've used Trace and it leapfrogged the competition in it's time. Outback then took it to the next level. The XW package is another leapfrog. That said, the XW had, and still has a few brain farts in it. Not having adjustable absorb cut off is a minor one. The most brain dead, which they have fixed due to my yelling and I'm sure others was a "redefinition" of Absorb and Bulk when they first release. "Bulk" ran until the float voltage was reached (not the bulk voltage), then it switched to "Absorb" but continued to output max current until the bulk voltage was reached (but they called it "absorb" anyway), then it would voltage regulate until acceptance current dropped to 2%. What? What? Electrically it was doing a 3-stage charge (float was handled normally), but for some reason they decided to call the bulk charge between float voltage and bulk voltage "Absorb". It was just a labeling thing, but talk about confusing. What genius came up with that?

The other thing that's totally broken and essentially useless is the Communications Gateway. For example, very day it produces a Daily Harvest Report. You'd expect that to be a report of how much solar power was harvested, wouldn't you? But it's not. It's how much power the inverter put out. So it's really a load report. I attribute it to it's legacy as a GT Inverter device where inverter output and solar harvest are indeed the same thing. But not in a hybrid system. It also puts out weekly and monthly reports that report different daily output (by 20% or more) than the daily reports. Someone can't count. Its SMTP email is completely flawed as well not supporting authentication to an SMTP mail server. Pretty much every SMTP server requires a login before you can send mail through it (it's part of your email setup) so people can't spam. But the gateway can't support that, so it's unable to email through your internet providers SMTP mail server. Instead you have to send the emails to a hidden relay server that Xantrex has that doesn't require authentication. That server was down for over a year and just started working again maybe 6 months ago. All this remains unfixed after 2.5 years. But hey, other people can't do much of anything in this respect, so it's still better.

blackswan555

Re: Charging large Surrette/Rolls batteries

I think your 2% end of charge amps may be a problem, 3 of my sites use an Outback FNDC, It uses end of charge amps (dial in the number) as a gen stop parameter, I do not have the numbers in front of me, but as far as I remember they are all nearer 1.4 > 1.5%, (All reasonably new Rolls but "run in"),
The best way to determine you "number" watch the amps, keep charging till they stop falling significantly, I use watt-plot, But a clamp meter and write it down every ten mins till you see it "flatline" works

Tim

twistedtree

Re: Charging large Surrette/Rolls batteries

blackswan555 wrote: »

I think your 2% end of charge amps may be a problem, 3 of my sites use an Outback FNDC, It uses end of charge amps (dial in the number) as a gen stop parameter, I do not have the numbers in front of me, but as far as I remember they are all nearer 1.4 > 1.5%, (All reasonably new Rolls but "run in"),
The best way to determine you "number" watch the amps, keep charging till they stop falling significantly, I use watt-plot, But a clamp meter and write it down every ten mins till you see it "flatline" works

Tim

2% of C is when the XW gear shuts off, but I've changed the battery capacity setting from it's actual value of 1350Ah to a fictitious value of 800Ah to fool it into running to 1.2%. 2% of 800Ah is 16A. 16A is 1.2% of 1350Ah. When I go lower, it never stops charging, so I think I'm at the bottom of the curve.

I've started another thread over in the equipment review section asking for recommendations on a supplemental charger. I've got plenty of generator capacity, so I'm considering a parallel charger to bring me up to 150-200A bulk current. Recommendations would be welcome.

Dave Angelini

Re: Charging large Surrette/Rolls batteries

Peter,

If you set the XWCC or system battery capacity to zero or some value between that and your actual capacity the end amps can be adjusted. I set my customers to zero and adjust the absorb time for 1.265. Some of the folks I know, play around with it but this does not have to be complicated.

Get your new cells, lower your useage, and do the 90 day EQ's. This is not suppose to be difficult with good equipment.

Vic

Re: Charging large Surrette/Rolls batteries

Morning twistedtree,

I agree with you and blackswan RE End Amps (Return Current). In my system, I run 17 Amps on the MX-60. This is where it needs to be to maintain SGs. The Asorb V is 57.5, currently.

None of these things are perfect, however, as if one has not reached Float the previous day, it takes a longer Asorb time to reach a specific End current setting. And some have recommended using the observed time to reach a specific return current to set the Asorb timer, instead of using current, due to downstream loads changing the reading.

Perhaps the emerging Midnite Classic CC will have more intelligence on this issue and others as well.

And YES, I got the XW SCC-150 early last year. To me it was a GIANT step backwards from the MX ... I stopped using the XW POS CC, for several reasons including that one that it did not even know what charge state it was in, and Xantrex stated that this BUG was a Xantrex INNOVATION ! What crap, it's a BUG !

However, generally I do like the X SW+ Inverters -- very solid. The charger in the inverter has a very gross regualtion scheme, which can upset smaller gensets, like the Honda EU inverter gensets.

And, YES, someone needs to make a PFC Charger (only) for off grid battery charging. 50-to-100 Amps, please. I called Magnum, and suggested it to them, as they have a PFC Charger in their invertter line, but recently, some here have noted problems with some of the Magnum inverter line.

And Ralph, the 4KS25s are a bit heavy, but they are tall, which allows one to have a lot of leverage when moving them. I just moved my second bank of 4KS25s by myself using a hand truck to push them up an 8-foot ramp into a pickup. Worked fine, and took only a couple of hours to load, move and unload them. But this is not out of a root cellar. Folks with the best battery environments may have a bit more trouble.

BTW, I've needed to pull one cell from a 4KS25, because one jar-top was cracked at the weld in shipment from Rolls. A very ROUGH RIDE on Roadway Express. I used a come-along, with small Nylon straps used by mountain climbers to loop through the terminals of the cell -- worked fine. One needs to possibly add electrolyte to the remaining cell, as the jars of these Surreettes bulge more when the sister cell is removed. So if the pair is "divorced" very long the remaining good cell could have possible exposed plates. I did this, by making sure that the compensated SG on the good cell was 1.265, and added some new 1.265 SG fluid to allow a bit more time to do the operation ... and so on ... Too many words! HH GL Vic

twistedtree

Re: Charging large Surrette/Rolls batteries

Vic, I'm interested to see that you have the same battery string as essentially the same solar capacity that I do. The biggest difference in systems that I see is that you appear to have stacked SW5548s which I suspect gives you higher gen charge current than I have. My XW6048 tops out at 100A.

What solar charge settings are you using? You mention absorb of 57.5V which is much lower than I have. Your "Return Current" is 17A which is essentially the same as mine. What about Bulk voltage? I'm guessing you get around 60A out of your panels when charging? And are your generator charge settings the same or different? And what is the bulk charge current? And I'm assuming your charge protocol takes you back to full SG?

Thanks,

Peter

Vic

Re: Charging large Surrette/Rolls batteries

Hi Peter,

Couple things:

1. My max PV Charge Current is only in the mid 40s.

2. Asorb Voltage presently is set at 57.5 V only on the MX-60. I manually set the Asorb V on the SWs to where it needs to be given the Battery Temp, as the SW temp comp is very coarse.

3. I have given up on the XW POS CC, but when I was running it, I set the V Bulk at something 62 V, as I saw NO benefit in limiting V Bulk.

4. For generator charging, I run 150 Amps from the SWs, and usually run 30 or so more Amps from PV.

But, really, my bank is usually lightly cycled, and need to shut off the PV input for about 4 days in order to get the bank down to where the high charge current duration will be 45 minutes minimum.

The return current is basically fixed at about 17 Amps, and have been adjusting the V Asorb to a value that just maintains the SG of the Pilot Cell, which was chosen because it has had the lowest SG.

The BTSes are stuck onto the side of the same battery in the center of the pack, with a 1-inch thick chunk of Styrene foam insulation strapped over the top of them. Realize that the XW system has probably all gone to Flag Terminal mounts.

One final point, I do have the GSM for the generator, but I do not have the guts to allow the genset to run in my absence, seems like it is an opportunity for heartache. Part of this is that the im my absence, the power demands are only 4 or max of 5 KWh per day, and the PV can keep up. Alto, at times it will take three or four days to catch up in very stormy weather.

And so on. You are doing the most important thing for your battery bank's health -- Paying close attention to it.

Nice to see you here on this Forum, Peter. Vic

twistedtree

Re: Charging large Surrette/Rolls batteries

Thanks. It sounds like we use the same power when the house is not occupied. Mine recovers similarly in the summer, but winters here in VT are a bit different than CA. It's still interesting that at 57.5V you can maintain SG. At that voltage mine would drift down over time. Maybe a year ago I backed my voltage down to see what would happen, and sure enough the SG started dropping.

Anyway, at this point I'm going to work on getting replacements for any defective cell(s), then keep doing what I've been doing. It doesn't sound like I'm committing any atrocities, but it sure makes me nervous when I know I'm running high charge voltages and then a cell or two go bad. That's ultimately what caused me to start the thread. I'll just consider it a coincident for now and keep my fingers crossed.

By the way, the stupid Bulk/Absorb definition thing in the XW MPPT charger has been fixed. I guess enough of us yelled at Xantrex about it. I know I sure did. To me, the real value in the XW components is when they are used together as a system. Mixed and matched (as in your case with SW inverters) it's just another MPPT charger. I think the biggest thing I like is that the display (SCP) and all the charging control is based on net current in/out of the battery, not current in/out of the individual component without regard for what else is in the system. As far as I know, nobody else does that. Oh, and everything is QUIET. The fans in my old SW4024 were horrendous when charging via the generator, and the Outback MX60 is no library mouse either. I can't even hear the fan in the XW6048 even when charging at full tilt - just the normal 60hz hum - and I don't think the MPPT chargers even have fans, just big heat sinks. That scored big points with my wife (and me too).

stephendv

Re: Charging large Surrette/Rolls batteries

twistedtree wrote: »

To me, the real value in the XW components is when they are used together as a system. Mixed and matched (as in your case with SW inverters) it's just another MPPT charger. I think the biggest thing I like is that the display (SCP) and all the charging control is based on net current in/out of the battery, not current in/out of the individual component without regard for what else is in the system. As far as I know, nobody else does that.

The Sunny Island also takes that approach, except that there's no external display, it's all built into the sunny island itself. Works the same for parallel inverters, you only program the master and all the slaves follow those settings.

twistedtree

Re: Charging large Surrette/Rolls batteries

Here's a quick update. I'm back on site and ran reproduced the problem this morning. One cells voltage drops to zero, then actually reverses voltage. So the dead cell is identified and I'm working a replacement with Surrette (they wanted me to confirm which of the two cells was bad).

In the mean time, I've bypassed the bad cell and adjusted all my charge voltages accordingly. That's the good part about a 48V system with accessible individual cells.

There are two other battery/cells that look a little suspect, showing lower voltage and lower SG than all the others, so I'm continuing the load with my solar and generator shut off to see if any other cells drop out prematurely. I'd really like to wait 8 weeks for replacements and swap out 300 lb batteries only once.