SG cell difference in new Rolls battery bank

Riley

Hi – I have a new 48V battery bank with Rolls S6-L16-HC batteries (8 X LC16 batteries in series for 48V system). Gear is Schneider MPPT60 and XW6848-01 (plus AGS and Insight communication gear). For solar gear, I have about 2400W of solar output (maximum charge current gets to 33A (low house load) on solar but much higher when on generator output).

Since the batteries are new I am concerned that there’s a 0.025 to 0.030 difference in SG between some of the cells. Most get charged to 1.275 or 1.280 (maybe even 1.285 sometimes) but some others are down at 1.245 or 1.250. For a new bank, I don’t think there should be that much difference between the cells.

My charge Voltage settings are both Bulk and Absorb: 60V (value in the newer version of the Rolls manual – rev6).
End Absorb is at 8.4 amps and 4 hours but the 4 hours is never reached. I assume the MPPT kicks to float due to the End Absorb current at 8.4 amps.
The system seems to work fine but I don’t like the SG delta in the cells

I have EQ'd about 2 times since owning the batteries but I am reluctant to do any more EQ without getting more advice.
After reading the battery manual for the 100th time, I decided to call Rolls technical support.

I was told that I need to spend longer on Absorb and reduce the Bulk & Absorb voltage to 59.4 (maybe 59.6). The message was that I am hitting the batteries too hard at the 60V (even though that’s in the manual) and the cells would even out over time if I charged at a lower voltage for longer (so lower voltage but longer Absorb time). With the charging voltage at 59.4V that I would need to run absorb for approximately 6 hours.

To do that, I think I will tell the MPPT charger to ignore return amps and just use time = 6 hours (360 minutes). Note: I have never been confident in how the Schneider MPPT handles return (or end) amps anyway but at 60V, many cells were getting to 100% full charge OR even over charged.

I was also told to change the temperature compensation on the Schneider gear to -120mV/degree (not the default for Schneider).

The above changes make sense to me and I wish that I had tried this change sooner (bank is 5 months old).
Today is the first day with these new settings and I will watch the SG and charger states today closely and see how this works.
Any thoughts on this issue and the overall strategy of reduced charging voltage but longer time?
Will that help to bring all the cells up to the same SG?
And any thoughts on how much damage I would have done to the bank?

Thanks in advance.
Riley

Vic

Hi Riley,

If you were to take a number of SG measurments, in a row, on a single cell,  how repeatable are those readings?

Do you ever cycle this bank down to 70-ish % State Of Charge (SOC)?

Surrette has been recommending 60 V Absorb, for several years, now.

If cycling the bank to 65 - 70% SOC, does not help even the SG readings,  then, would recommend more EQ.

You are using ONLY real Distilled Water as make-up water for the battery, right?
Approximately how much water does this battery bank use?

As you know,  these batteries almost always ship, with 1.265 SG electrolyte.

Under heavy charge or discharge, have you measured the voltage at the TERMINALS of each battery?
You can measure the voltage difference, between batt terminals and the lug attached to that terminal.

Did  you follow the Surrette Flood Battery Commissioning process, when the batts were just delivered?

Lotsa questions,  more later, thanks for answers,  Vic

Photowhit

I would almost assume this is a standby system and not in general use, or perhaps just runs a single high wattage load, maybe a water pump? A 2400 watt array and a 6000 watt inverter doesn't make much sense.

As Vic asked, did you go through Rolls Setup process? I think they suggest checking SG and doing a commissioning equalizing.

My understanding is that batteries begin life not fully formed and it takes 40-50 cycles to get to their full capacity. If your system is a standby system they might not have gone through enough charge discharge cycles for the surface to fully form. (whatever the heck that means, but I suspect how the surface handles sulfates in charge and discharge cycles)

Vic

YES,  the proper temp compensation for your 48 V battery, is -120 mV per degree C.  Schneider hardware is not bright enough to multiply 24 times -0.005, or look it up in a table.

What is the nominal temperature of your battery, these days?

Last I looked, the Mppt 60, does not have a setting, to temp compensate the EQ voltage.  So,  IMO,  you can,  and probably should do this compensation manually, just before initiating an EQ, based on the nominal battery temperature.

EQs should be done, after Absorb is completed.

Depending on how deeply-cycled your battery has been cycled,  the 8.4 Amps as End Amps, may be a bit too high.  This IS probably about 2% of 20 Hr Capacity.  The Ideal EA setting, varies, with the Depth Of Discharge (DOD) that the battery experienced in the previous discharge.  To see what your battery really needs (at any one Absorb stage), is to set a long Absorb time,  and watch the actual battery charge current.  When this current stops diminishing, this is in a good ballpark range, for your EA.

For Charge Controllers (CC) that do NOT have battery voltage sense leads,  largish loads can change this battery current measurement, due to voltage drop in the cable from the CC, to the battery, and voltage drop in the battery breaker of the CC.  This can disrupt the CC's decision on when to end Absorb, and interfere with your decision on when the battery charge current has stopped diminishing.

Generally the EA recommendations from battery manufacturers for Flooded batts, are for batteries that have been discharged to 0% SOC,  and are being charged using a grid charger, using IUI charging.  So,  this difference can be a bit misleading, IMO.

The Surrette banks here usually need an EA of about 1.25% of actual battery Capacity  --  about 15.5 A on a 1280 Ah bank.

And, so on,   FWIW, Vic

Riley

Vic said:

Hi Riley,

If you were to take a number of SG measurements, in a row, on a single cell,  how repeatable are those readings?

Answer: Pretty repeatable. Sometimes with the low cells, I pull the measurement a few times and the gauge seems to show a little bit higher with 2nd and 3rd readings. 

Do you ever cycle this bank down to 70-ish % State Of Charge (SOC)?

Answer: Rarely but maybe on the lower cells. I baby the bank due to how we live our lives. The deep power pulls are limited to times of high solar output (water pumping into house and the dishwasher). The remainder of the time my load is about 300W with short demands of 1KW (water pump for pressuring house). 75% SOC is probably as low as I go.  
Edit from the following morning: I measured the lowest cell and it's at 1.210 (temp compensated measurement). According to the table in the Rolls manual, that's about 60% SOC (lower than I thought). The highest cell is 1.230 (about 70% SOC??)

Surrette has been recommending 60 V Absorb, for several years, now.

Answer: Agreed. That's why I used the 60V setting. 

If cycling the bank to 65 - 70% SOC, does not help even the SG readings,  then, would recommend more EQ.

Answer: So sounds like I should work the battery bank harder. 
Question: When I pull down the SOC, should the higher cells drop more than the lower cells? Or in other words, even things out between the cells? 

You are using ONLY real Distilled Water as make-up water for the battery, right?
Approximately how much water does this battery bank use?

Answer: Yes, distilled water. Bank doesn't use much water I have added water once or twice in 5 months. 

As you know,  these batteries almost always ship, with 1.265 SG electrolyte.

Answer: I don't think I saw that but not 100% sure due to lots going on at the time. The manual also says it does take a number of cycles to get cells evened out. 

Under heavy charge or discharge, have you measured the voltage at the TERMINALS of each battery?
You can measure the voltage difference, between batt terminals and the lug attached to that terminal.

Answer: I haven't done that. Good idea, thanks. The connections look good but I will measure the voltage drop across the connection. 

Did  you follow the Surrette Flood Battery Commissioning process, when the batts were just delivered?

Answer: I will reread the commissioning process and check. 
Edit - I reread the commissioning section of the manual. i basically followed it except I didn't do a long EQ on the first charge. I did EQ for 1 hour (default on my CC gear). I have been hesitant to do multi-hour EQ cycles, I just need to know a few hours of EQ isn't going to hurt the batteries due to stress.  

Lotsa questions,  more later, thanks for answers,  Vic

Thanks to all for your responses. 
Sounds like I should be concerned with this delta of .0030 to .0020 SG. 

Perhaps I should pull some larger loads on the battery? I have always been reluctant to run down the bank for no real need but if it helps the cell difference, it's worth it. 

Riley

Photowhit said:

I would almost assume this is a standby system and not in general use, or perhaps just runs a single high wattage load, maybe a water pump? A 2400 watt array and a 6000 watt inverter doesn't make much sense.

Answer: My largest typical peak load is 2KW but I have conservative design for the inverter/charger. I suppose if I really pushed things, I could pull 3KW from the bank for a brief period but I try to avoid that in how I live our lives here. This system is for a modern off-grid summer home and we are careful with the loads. Draw is typically 250 -> 350W 24/7 and the 2KW loads are brief and really only when I have solar available and for brief durations (15min) to bulk pump water into a holding tank in the house. I don't use the battery bank to drive the large bulk water pump. Battery bank is used to run the smaller house pump which pressurizes the plumbing in the home (runs once and awhile when water is used 24/7 (ex: flushing toilet). That results in about 1KW total load for about 30 seconds. 

As Vic asked, did you go through Rolls Setup process? I think they suggest checking SG and doing a commissioning equalizing.

Answer: Never did commissioning EQ. I did check the SG and they were out. After a week or so, I tried an EQ but only for 1 hour. 

My understanding is that batteries begin life not fully formed and it takes 40-50 cycles to get to their full capacity. If your system is a standby system they might not have gone through enough charge discharge cycles for the surface to fully form. (whatever the heck that means, but I suspect how the surface handles sulfates in charge and discharge cycles)

Answer: This statement in the manual about the plates fully forming isn't clear to me. I think you might have something here in that my system is run somewhat like a standby system with light loads. By morning I am only down to 75% SOC or so. Maybe I need to work the system harder... ? 

Riley

Vic said:

YES,  the proper temp compensation for your 48 V battery, is -120 mV per degree C.  Schneider hardware is not bright enough to multiply 24 times -0.005, or look it up in a table.

Answer: Noted. I wish I had known that... ugh. My fault I guess. 

What is the nominal temperature of your battery, these days?

Answer: in the early evening after charging all day: 32.5C (ie: right now, sun is down). My notes from charging today show 28.6C. Seems odd that it's warmer now but I need to double check on tomorrow's charge cycle.  

Last I looked, the Mppt 60, does not have a setting, to temp compensate the EQ voltage.  So,  IMO,  you can,  and probably should do this compensation manually, just before initiating an EQ, based on the nominal battery temperature.

Answer: I think you are correct, no EQ compensation for temp. Good suggestion, thanks. I will do that next time. 

EQs should be done, after Absorb is completed.

Answer: Maybe I should go a week with these new settings before doing another EQ? 

Depending on how deeply-cycled your battery has been cycled,  the 8.4 Amps as End Amps, may be a bit too high.  This IS probably about 2% of 20 Hr Capacity.  The Ideal EA setting, varies, with the Depth Of Discharge (DOD) that the battery experienced in the previous discharge.  To see what your battery really needs (at any one Absorb stage), is to set a long Absorb time,  and watch the actual battery charge current.  When this current stops diminishing, this is in a good ballpark range, for your EA.

Answer: Noted. Today I ran at the charge at 59.4V for both bulk and absorb. I got down to about 6amps absorb current before the clouds rolled over and sun behind trees. Now that I am in August, my days are starting to get shorter but I think most days I can get a longer absorb time and thus lower absorb current if needed. 

For Charge Controllers (CC) that do NOT have battery voltage sense leads,  largish loads can change this battery current measurement, due to voltage drop in the cable from the CC, to the battery, and voltage drop in the battery breaker of the CC.  This can disrupt the CC's decision on when to end Absorb, and interfere with your decision on when the battery charge current has stopped diminishing.

Answer: I have observed the EA and switching to float and it does seem to be around 8A. That aligns with the CC setting of capacity around 400AH (ie: 2% of AH capacity). 

Generally the EA recommendations from battery manufacturers for Flooded batts, are for batteries that have been discharged to 0% SOC,  and are being charged using a grid charger, using IUI charging.  So,  this difference can be a bit misleading, IMO.

Response: I don't know the term IUO charging but clearly I am not on a grid charger. In good summer weather I try to avoid using generator but I do if needed to get the batteries fully charged on crappy days or shorter days in fall/winter/spring. 

The Surrette banks here usually need an EA of about 1.25% of actual battery Capacity  --  about 15.5 A on a 1280 Ah bank.

Answer: So for my 420AH bank, that would be 5.25 amps. I think that means to enter 262AH capacity into the CC so that 0.02 x 262AH = 5.25 amps. From my trial today, that might be about 6 hours..... a bit hard to do as the days shorten but ok, I can do that. 

And, so on,   FWIW, Vic

I guess I should have been more vocal with Rolls or my local battery dealer in the beginning... . 
Hopefully I can get this corrected. 

Riley

Hi guys - Anymore thoughts given my information/responses above? 

My first take away is that I will need to do an EQ session and don't be scared to go 2 or 3 hours. 

I think I also need to cycle the bank a bit deeper and use these new relaxed voltage charging specs... this will hopefully pull those hot cells down and then run a relaxed but long charging absorb cycle (I can bulk at 15% 20-H capacity to stir things up. 

The tech at Rolls told me to run these for a week, I guess I will do that before EQ.  

BB.

From my limited experience... Yes, discharge to/below 75% state of charge... EQ at ~60 volts and watch the bubbling (light fizing, not rolling boil)--Adjust EQ voltage up or down as needed. And monitor battery bank temperature. If battery bank gets too hot, then stop EQ.

Ideally, fully charge bank then EQ and monitor each cell SG every 30-60 minutes...EQ is done when all cells stop rising between measurements.

Rolls/Surrette batteries may tank multiple discharge/EQ cycles to stabilize (i.e., 5 or more EQ cycles pretty common???).

-Bill

Photowhit

BB. said:

From my limited experience... Yes, discharge to/below 75% state of charge... EQ at ~60 volts and watch the bubbling (light fizing, not rolling boil)--Adjust EQ voltage up or down as needed. And monitor battery bank temperature. If battery bank gets too hot, then stop EQ.

I think Rolls equalizing specs are higher, I would follow their manual.

Agree on watching the temps, I think the Conext controller will do this, might even have a setable temp.

Dave Angelini

Hey Photowit, the controller is only going to monitor one cell and so he really needs a temp gun to monitor all the cells. You are right that EQ is over 62vdc.

Steve at Surrette has told me a few times to not worry about it for a few months. Let them age in. They can need 3 to 5 hours
of Absorb at the correct temp comp. A 50% discharge every month as they age in. Start EQ in a couple months if needed.

Riley

Dave Angelini said:

Steve at Surrette has told me a few times to not worry about it for a few months. Let them age in. They can need 3 to 5 hours
of Absorb at the correct temp comp. A 50% discharge every month as they age in. Start EQ in a couple months if needed.

Hi Dave - just looking for clarification on the above statememt "not to worry about it for a few months". 
Do (did Steve at Rolls) mean don't worry after a few months after EQ OR since when new bank? 

I am now at 5 months from new so I am thinking I should go ahead and do a full corrective EQ in the next week or so. 

Riley

All - 

Also if when fully charged (after absorb cycle), the cells might vary from 1.280 to 1.250 (highest to lowest). 
Will doing an EQ cycle damage the cells that are at 1.280 (ie: fully charged)?

That's what concerns me but perhaps all corrective EQ cycles are in this situation. 
 

BB.

As I understand (no battery expert)--This is weeks or a few months of cycling and the weekly (or so) EQ cycling...

For EQ current... I would be aiming at 5% to 2.5% of battery 20 hour capacity (i.e., 100 AH * 5% = 5 amps of EQ current). Note that even 2.5% EQ current can heat the batteries to their limits over 8-10+ hours (depending on ambient temperatures, cooling airflow, etc.).

Yes--EQ is "controlled" over charging of your 1.280 SG cells to bring up the cells that are -0.030 SG or lower from "full" charge.

EQ of the fully charged cells is hard on them... And why you only do EQ when needed (for tall cased batteries--Usually suggested EQ once a month to "mix" electrolyte and prevent stratification--"heavy" electrolyte at bottom of cell and "light" electrolyte at top).

Something like 50% of the energy goes into "bringing up" weak cells and gassing... The other ~50% goes directly to heating the battery bank.

Gassing does errode plates to a degree, and the oxygen forming at the positive terminals plates, causes corrosion of the plate grid wires and plate material... And does "knock off" lead sulfate (the hard back sulfate crystals that form over time, add resistance to the plates/no charge/discharge reaction) which is "good" for the overall battery operation.

There is a common engineering rule of thumb... Basically for every 10C (18F) over 25C (77F) "room temperature", the "thing" (batteries in this case) age 2x faster. And conversely, every 10C below 25C, the "thing" ages 1/2 as fast.

So, running your batteries to (for example) 52C:
https://www.rollsbattery.com/wp-content/uploads/2018/01/Rolls_Battery_Manual.pdf (random Rolls manual):

4. Place batteries on charge. Please see Table 2 (a & 2 (b Flooded Charging Parameters for required charging parameters. Do not let the cell temperature exceed 52°C (125°F). If the temperature becomes excessive or the cells begin to gas vigorously, reduce the rate of charge. Continue charging until all cells reach the specific gravity of the filling acid. All cell specific gravities should be even (1.260-1.280) when resting at full charge.

• 52C batt temp - 25C STD temp = 27C diff (rise)
• 27C/10C = 2.7 faster aging
• (2x faster aging (per 10C))^2.7 = 6.5 faster aging
• i.e., 1 day @ 52C = 6.5 days of "room temperature life"

Not to say to never run at 52C... One 24 hour @ 52 EQ is knocking 6.5 day off bank aging life--Once a month--You are not going to be able to see your bank die 1 week sooner because of that time at elevated temperatures.

If, however, the bank runs for weeks/months at @52C, yea, that is going to noticeably shorten battery bank life.

Between plate erosion, corrosion, and faster aging when hot--All good reasons to only perform EQ when needed (couple hours a month, or maybe 10 hours when "corrective EQ" is needed).

Use the battery bank--That is why you have it... Don't baby the bank (i.e., discharging to 90% and recharging every day). That is not really going to give you much improved performance (charging from 90% to 100% every day--gassing, heating, etc.--All good reasons to minimize that sort of operation).

I recall (?) that some solar charger vendors (Midnite and a couple others?) actually have a program value that can be set to only run a full charging cycle if the battery bank falls below 75% SoC.

And there are other ways the battery banks can be cycled... One is to cycle (daily) the bank from 80% to 50% and back to 80% daily--And charge to >90% once a week... Sulfation will not occur (or at least not at an accelerated rate) when doing this. And the battery bank will run cooler too (Lead Acid batteries generate more heat as they approach 90% to 100% SoC).

Again, I am no Oracle on batteries. Others here have a lot more experience than I (which not that much)... But the above are the general recommendations I have seen over the years.

Anyone that has more information to add, or needs to correct my observations--Please feel free. All are here to learn.

-Bill

Riley

Thanks Bill, that's really good information. Once i get these cell balanced, I will see how much they are really getting cycled. It's hard to tell now because of the cell mismatch. I think I am running down to 70% SOC on average. I can try to do more

Photowhit

I haven't read Rolls info on Equalizing, but do recall the split equalizing into 2 groups, Maintenance and corrective. As I have done for years.

Bill splits them up pretty well above. It has been engrained in me to do corrective equalizing for 1-2 hours then check the cells, when you see no change in the low cell(s) over a hour period, go 1 more hour if no change that's your new best for that cell. I've heard of longer being useful, and I was thinking equalizing for Rolls was 63 or 64 volts for a 48 volt bank, but I'm not reading the manual when you have great guidance. I say this because I used 62 volt equivalent for my forklift battery (31 volts on a 24 volt) and was surprised Rolls wanted even higher. That might have come and gone since, or been a directed number from someone having issues with Rolls 8-10 years ago when I think they had some glitch. Typically, pretty much the best Flooded Lead acid battery for daily cycling.

Dave Angelini

Riley,

What is the capacity you have programmed into the chargers?

Riley

Dave Angelini said:

Riley,

What is the capacity you have programmed into the chargers?

For the first 5 months of the battery life my setting were: 

Capacity = 380AH to give a Return Amps of about 7.6A 
Bulk & Absorb charge voltage = 60V  (note: this value is that is defined in the manual). 
Absorb time was 4 hours but likely never went that long as RA was kicking the CC out of absorb and into Float. 

Note very light cycling of the bank never going below 70% SOC (and a number of weeks of very low drain effectively on standby when the cabin not occupied). 

Also the temp compensation was incorrect for Schneider gear at -100mV? ( I think the default). Note that the temp compensation never kicked in so I was clearly over charging the batteries.  

After speaking to Rolls tech support, he recommended a lower charge voltage to 59.4 (possibly 59.6) and pushing the absorb time up to 6 hours. So in effect a reduced charge voltage and much longer absorb time. Also corrected the temp compensation to -120mV/C 

For the record, my new values as of 5 days ago are: 

Bulk/Absorb voltage: 59.4V (actually runs lower due to temp compensation). Absorb voltage tends to be around 58.8 volts I think. 
Capacity: 260AH should give Return Amps of 5.2A (and I observe this however the days are getting shorter and at this point, I am losing PV output. 
Absorb timeout of 6 hours. 
Temp comp: -120mV/C

So I think the logic is that the short absorb times were not long enough to bring up the cells that are low in SG. 
Rolls Tech suggested running like this for a week or so and then (I think he said) do an corrective EQ process. 

One morning I will also pull down the SOC by running my big lake water pump to pull the lower cells clearly down to 50% SOC and at the same time, see if the generator AGS kicks on automatically (never got low enough to trigger it so haven't tested it yet). 

Note that I have never done a corrective EQ cycle during the lifetime (5 months) of these batteries. 
I should have jumped on this issue sooner. I was disappointed when I saw the wide delta of SG when I first installed them. 

I have also thought about just setting Capacity to 0AH and just use the 6 hour absorb timeout to limit absorb phase.  

Any and all good advice is welcomed. I want to get this corrected. 

Edit - Just to clarify as I read over this thread. I haven't ever done a corrective EQ cycle on these batteries. I just did 1 hour cycle a few weeks ago and before that maybe 2 hour cycle. I was being cautious and afraid to 'go hard'. On the 2 hour EQ, I did start to see the lower cells come up in SG. Clearly didn't go long enough. 

Maybe that's a good thing because I didn't have a good temp sensor. This week I will buy a IR thermometer to monitor the individual cell temperatures.

Riley

Update: 
I have been following the advice from Rolls tech support and running with the lower Bulk/Absorb voltage of 59.4 and a very long absorb time (& lower End Amps). Basically running absorb until the absorb current is about 5.2A (just tails off and sits around here). I have run with these long absorb charges for a number of days now. (9 days). 

Result: My SG cell difference seems to be getting smaller (.020 maximum instead of .030). I still plan to run an EQ cycle next week when I have my IR thermometer and some good sunny days. Will use the Generator to get a running start on the day.  

I think Rolls suggested the lower charge voltage (Lower than the value in the manual) until I can get the bank evened out and fully charged and then perhaps moving the charge voltage up towards the 60V that's spec'd in the manual (depending on the charge required).