Help me Fix "Sleepy" Cells in my New Battery Bank

[Matrix]

I have been carefully monitoring my new Trojan L16H-AC 4 battery, battery bank the past few weeks since installed Oct 16, 2017. At first all cells at fully charged were at the same SG.... 1.277 +. But Now after 4+ weeks I have 3 batteries that have lagged behind, and one that has leaped ahead, with about a 15 point SG gap between batteries.

3 batteries have fallen behind the other 1 by about 15 points at full charge of the bank. If the highest battery is at 1277 (100%) the others are around 1262-ish.  I EQed them for 2 hours 2 weeks ago testing at the 1 hour point. This did not bring all cells into equal range. 3-4 cells still lagged 15 or so points although now every cell was at or above 1.277.   So I let it ride for the past  10 days. 

So now yesterday and today,   I again had 3 batteries behind the 1.  So I did an EQ after fully charging with an AC charger,  and got 3 out of 4 batteries all within 5 points of each other. 

But 1 battery ... all 3 cells .... is still showing 10-15 points lower than the others.

So I have EQ'ed ... and EQ'ed ... and EQ'ed. The highest my charger will set the EQ is 32v and 60 min fixed setting. Trojan says for the L16H-AC the EQ should be 32.4, but I cannot set my charger that hi.  I have been monitoring the Highest and lowest cells. After 4 EQs it seems that all batteries have stopped going up in SG.

Charger is the Conext SW 4024 AC Charger: 32v 60 min Fixed EQ charge option highest setting for EQ. This is the highest volt setting that I have available with this charger.

Have done a carefully monitored EQ 4 times until SG appears to stop rising on all batteries  monitoring SG each hour. Temps have not exceed 75* F (cool morning). SG has appeared to stop rising. Battery samples are as follows

20 min after last EQ
- Battery 1 - 6.88v: 1299 taken from Center cell (positive to inverter)
- * Battery 2 - 6.88v: Lowest - 1295 / 1291 / 1290 All cells (This is still the "sleepy" battery)
- Battery 3 - 6.90v: Highest - 1305 taken from cell closest to negative post
- Battery 4 - 6.88v: 1302 taken from center cell (negative to inverter)

I know this is not what Trojan would call "out of range" and "In need of EQ".  Their spec for that is 30 points, but it also appears these 3 cells (all in the same battery) are NOT going to come up to the level of the other battery's cells.

Trojans charging guide says u should EQ if cells differ by .030 in any cell. I don't have that large of a gap....
- So is 10 point gap in one battery all cells a problem or typical for a few cells?
- How do I correct this?   Individual Charge or Discharge of the batteries?
- I EQed, and it didn't help completely. So do I EQ longer or just let it ride?

AND if I should just "let it ride"... Which cell's SG should I use to charge to? To 100% (1.277 SG) of the highest cell or the lowest cell? What is considered fully charged when cells are not perfectly in balance?

Any ideas?

[Westbranch]

Last Q. use the highest cells for Ending Absorb... otherwise you run the risk of damaging the good ones by overcharging.
2. keep EQing weekly if you are seeing improvement otherwise stop.
3. have you done 50 cycles yet?  That is one makers recommendation of the breakin period...
4. have you done a 2 or 3 day depletion ie 50% yet?  Staying in the > 85% SoC area does not 'exercise' the plates very deeply...

[Matrix]

I probably do not have 50 cycles yet ... and have never depleted more than 35% ... so I guess that would help.  I will do that in the next few days and see what happens. 

Thanks.

[Westbranch]

How many days (Hrs.?) to get to 35% dod?   If 2 that would be a good level to repeat at, I don't know your avg sun right now so Just want to make sure you can recover to 100% SoC in1 day, 2 max.

[Matrix]

How many days (Hrs.?) to get to 35% dod?   If 2 that would be a good level to repeat at, I don't know your avg sun right now so Just want to make sure you can recover to 100% SoC in1 day, 2 max.

When you say "get to 35% dod" are you meaning I used 65%?  That has never happened.  I have only every used 35 % and the battery was still at 65% when solar charging started again.  (sorry if I am misunderstanding the terms)

It takes me 12 hrs to discharge 35% of the battery (or down to 65%).  And recharge is no problem.  I can use the AC line charger in a few hours  if there is no sun. 

[ClassicCrazy]

I think he means to discharge the battery down to 35% SOC so you may have to just turn off your PV for a day to do that .

Larry

[Matrix]

Thanks Larry. I will try that

[Westbranch]

Terms SoC and DoD.... 

SoC is the % of charge remaining in a battery, stated as a % of the total capacity.
DoD is the % (amount) of the total capacity that has been used from a battery.

They are the inverse of each other if you think about it, one is what you use and the other what you have left .
In your case you have a daily DoD of 35% and a SoC of 65%. Add them an they should = 100%

What I suggested is that you go closer to 50%, mid-point of DoD/ SoC, to give the bank more of a workout and observe any improvement in the SG readings at the end of a week..

cheers

[Matrix]

What I suggested is that you go closer to 50%, mid-point of DoD/ SoC, to give the bank more of a workout and observe any improvement in the SG readings at the end of a week..

cheers

Would  you say do that everyday for a week or just a couple of times? Right now today I have managed to discharge down to 48% SOC On the classic 150.

I still have a hefty 15-point Gap between the three low batteries and the one high battery. The lowest cell is reading 1175 and the highest cell (different battery) is reading 1190. It seems all cells are fairly balanced within an individual battery.

I will be firing up the charger here in the next hour or so and letting it run back to a full state of charge specific gravity level

[Westbranch]

AS I know you like to be fastidious about your procedures so as to be able to learn the effect of a change , I would shoot for the same DoD everyday before recharging. With the WBjr you can be more exacting and use the Ahrs of use that are rercorded.

Your advantage over Off Grid is having the grid for recharging the bank at will... and not having to  rely on the appearance of the Sun to finish off the charge.... and not having to listen to the generator drone on  for hours. 

[Matrix]

All True.   

BTW ... Trojan Wrote me back.  They seem to say let it ride and see if it works itself out. 

Hi Michael,

I would give it some more time…daily use and cycling the batteries as normal will likely cause the gap to decrease.  A range of 10-15 points is not alarming, especially considering that the batteries are still in the part of their cycle life where they are completing the formation process.  Give it a month or so, then check the electrolyte level after they are fully charged, or even after an EQ cycle.  By then it may be time to add water which may level the gravities to some degree.

[RossW]

One of our group who's quite active in IRC has a mixed bank (ok, not best practice, we all know that - but some times you have to deal with what you've got).

In order to work around the problem of different charge requirements for one half of his bank, we added a balancing load resistor across the "good" half of his bank in order to get more charge into the "bad" half. I think initially he was using a couple of car headlight lamps (two in series) across the "top 24V" of his 48V bank. While he was charging, the load came on, resulting in an extra 5A or so through the "bad" cells. Over time, we completely overcame his chronic charge imbalance and the whole bank is performing really well now.

Once the bank got back to "close", the additional current required to finish the job was in the order of 1A.

I hasten to add, we had monitoring on each cell to see exactly what was going on, and later on we added a raspberry pi with relay board, we could switch in and out loads across most cells to bring them all up to ideal SoC. This might be overkill, but the underlying methodology - resistance in the order of 2-ohms-per-cell (6 ohms for a 3-cell pack) will shunt about an amp of current around that cell. Doing it to "all except the bad one" might give you the extra kick you need?

[Matrix]

One of our group who's quite active in IRC has a mixed bank (ok, not best practice, we all know that - but some times you have to deal with what you've got).

In order to work around the problem of different charge requirements for one half of his bank, we added a balancing load resistor across the "good" half of his bank in order to get more charge into the "bad" half. I think initially he was using a couple of car headlight lamps (two in series) across the "top 24V" of his 48V bank. While he was charging, the load came on, resulting in an extra 5A or so through the "bad" cells. Over time, we completely overcame his chronic charge imbalance and the whole bank is performing really well now.

Once the bank got back to "close", the additional current required to finish the job was in the order of 1A.

I hasten to add, we had monitoring on each cell to see exactly what was going on, and later on we added a raspberry pi with relay board, we could switch in and out loads across most cells to bring them all up to ideal SoC. This might be overkill, but the underlying methodology - resistance in the order of 2-ohms-per-cell (6 ohms for a 3-cell pack) will shunt about an amp of current around that cell. Doing it to "all except the bad one" might give you the extra kick you need?

Cool.  Got a wiring diagram for that?

[russ_drinkwater]

The L16's we have here had a spread in the sg's for quite a while.
I did nurse them as I was new to the solar charging of cells.
They would be worked during the day with the house on stand alone and then at 6pm we switched to grid tie.
As a consequence they never got below 85-90% soc.
It took a while for the sg's in individual cells to settle down.
Now that the bank is at my sons house and is totally standalone they get a lot harder working life.
When running the a/c unit the twin arrays and 150's dump in 100amps at peak load.
The sg's are stable at around 1280-85 and water usage is very low compared to the rolls cells here at my
house. The bank is 870 amp hours is the difference.
I would say give it a couple of months with monitoring.
Do not forget to clean your hydrometer properly to prevent contamination and internal buildup.
Sticking float gives lower sg readings.
Also draw and discharge electrolyte (back into cell of course) a few times before drawing your test sample to
ensure it is better mixed.

[CDN-VT]

Mike /Matrix

Get the WD & learn , But stay the course on the Trogen reply .

Adjust the Conext amount of charge

VT