L16's low capacity

706jim · March 2019

I purchased 8 L16's about 3 years ago to replace 12 GC's that I had used for 19 seasons with 430 watts of solar charging them. The L16's are being charged with more panels than previous and a realistic afternoon maximum amp reading is about 30 at 24 volts. My daily loads are modest, about 1 Kw-hr for the fridge and about half that for a TV. Water pump might run 10-15 minutes a day with 1/4 hp induction motor. The amount of power I can retrieve from these relatively new batteries is disappointing. It doesn't seem to take much to knock them down to 24 volts or less and it also doesn't take a lot to bring them back up to float. I'm using Trojan's recommended 28.8 volt bulk setting and float at 26.4. I never checked my previous GC bank for specific gravity (after all they lasted a long time) but I've checked the Trojans and the best I get is reading in the middle of the white. Never get SG into the green. Trojan recommends between 5 and 13% of the bank rating (in my case 740 a-hr) so I'm close to that with 30 or so amps with my 1 kw panel set.

I'm planning on scrapping my ancient Arco panels and doubling the new bank to 2 Kw, but I'm wondering if this is going to make a difference. One other consideration is that I will then have 3 panels facing south and 3 panels facing west meaning in practice that the max current will be that 30 amps just for a longer time. At this time, it is impractical for me to use an MPPT controller as the panels will not see sun all at the same time.

Am I alone here or has someone else seen this sort of quandary?

mcgivor · March 2019

Classic symptoms of cronic undercharging leading to sulfation resulting in capacity loss. For a daily cycled bank the charging current should be closer to the 13% of rated capacity or 96A, 30A is ~4%. This deficit will manifest over time often going unnoticed, at this point the only possible cure is equalization, until SG readings rise, it would be best not to use the bank at all during the procedure, which may take days to complete given the rather small array.

706jim · March 2019

One thing I will try is to charge 1/2 of the bank at a time. This is easy to accomplish as I have a marine switch allowing either string to be disconnected. It would bring the charging current closer to 8% which should help. Will report back here when I can try this which will be late May.

Estragon · March 2019

1 kw pv will produce ~750w x 3-5hrs full sun equivalent depending on location and season, so roughly 2-3kwh/day on average. If your loads are in that same ballpark daily, it's going to be tough to catch up after a gloomy day or two.

740ah@24v is nearly 18kwh. If you're getting to 24v (sustained, not sagging with a surge load), that would be roughly 50%SOC, so a 9kw deficit. That's going to take 3-5days on average to catch up with no loads, and may never catch up with your typical loads.

My suggestion would be to bulk charge the bank with a generator and use the solar to charge at absorb and equalize voltages for as long as needed to get SG as high as possible. If they aren't too far gone, it may take several days of EQing to get to a point where SG stops rising. If they stop rising at a low level after an hour or so of EQing, they may have irreversible loss of capacity.

If this can't be done until late May (after ice breakup?), I'd be thinking about a replacement bank, and adequate charging for it.

MrM1 · March 2019

If those are Trojan L16 batteries u are under charging them. Recommended Bulk / Absorb from Trojan is 29.6v and I have found it takes even more to reach fully charged to 1.277 using solar. Closer to 30.5v and when I asked Trojan they said even that was no problem.

Estragon · March 2019

If the bank is cool/cold, there could also be a problem if voltage isn't temperature compensated with a RTS.

706jim · March 2019

https://forum.solar-electric.com/discussion/comment/396146#Comment_396146

Good to know. I've been able to put more current into the batteries than I use but the controller would just go to float with my 28.8 bulk voltage. There has never been a problem with them using any water which also points to a need for more charging. Guessing you agree with their recommended 26.4 volts float? I used 27.0 on my previous GC bank of 12 batteries with good results and only lowered float to 26.4 based on the Trojan booklet.

mcgivor · March 2019

Charging voltages recommended by manufacturers are often based on grid charging where time is not a limitations, as it is in solar charging, so increased voltage is often required to get the job done within the window of opportunity.

706jim · March 2019

That makes sense. What are your thoughts on the 26.4 volt float voltage? Should that be increased as well?

MrM1 · March 2019

Too low. Again trojan spec sheet says float for L16 is 27v. I usually float at 27.8v. For my system this voltage allows for my loads to be supplied from the solar array and not battery when the sun is out but the bank is in float.

There are some forum participants (on another forum) who recommend when Solar charging - especially in winter or more further north /south of equator - that bulk/absorb/float should all be set to the same voltage.... The manufacturer absorb voltage or higher because there really is not enough sun to fully charge even all day.

This is what has been referred to as "maximum smoke" if I understand the phrase correctly.

mike95490 · March 2019

if your SG's are low, increase your absorb time 30 min, till they stop rising, then the battery is ready for an EQ cycle, it might take several days to dial in the proper absorb time, and on EQ day, start the AM with the generator to bulk, and by 9

706jim · March 2019

I just checked Trojan's online specs and they agree with what has been posted here: Bulk 29.6 volts (I'm at 28.8) and float at 27.0 (I'm at 26.4).

My values were taken from a booklet provided by Trojan and obviously they have now changed the specs.

I'm generating more power (even at 4% C) than I ever use so likely the problem has been too low voltages when charging. Time will tell if the batteries were damaged (I can't check until May) but hopefully they are salvageable.

If not, I will replace them with half as many batteries as my daily loads are relatively modest maybe 2Kwh at most.

Thanks for all of you who have taken time to respond.

MichaelK · April 2019

The first suggestion I would give you is to buy a good temperature compensating battery hydrometer for checking your state of charge. I bought this one, which works well.

https://www.amazon.com/East-Penn-Temperature-Correcting-Hydrometer/dp/B0002KKTYU/ref=asc_df_B0002KKTYU/?tag=hyprod-20&linkCode=df0&hvadid=312280085431&hvpos=1o1&hvnetw=g&hvrand=16774424684650610275&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9007781&hvtargid=pla-569461448083&psc=1

I would disagree with the power levels you are claiming to be using and the amount of power you are producing. In shopping myself for an off-grid refrigerator, I'm finding small 17-18 cubic foot models in the 1.5kwh range. Regular size frigs start around 2.0kwh/day. Running that, and a TV, and well pump would easily push you well past 2kwh per day that you think you consume.

Your panel output is sadly low. Even doubling it to 2000 watts is only about half of what I would calculate you need.

(Amphour/8) X (charging voltage) X (loss conversion) X (number of strings)

(370/8) X (29.6V) X (0.75/1) X (2) =3650 watts of panels

In my own case, I'm running a 48V system with eight L-16 batteries, the same capacity as yours, and I'm charging them with 15 300watt Renogy panels. Charging my system at 59.3V, I routinely reach 100% charge (1.290 density) even in the winter. Technically, my system is overpaneled (on paper), but in the real world, I keep one array pointed SE, while another is SW, so the maximum amps produced doesn't get dangerously high. I'm running a 1.0hp 240VAC well pump, and this system allows me 6+ hours of continous pumping per day in the summer.

Rather than just trashing the batteries you already have, I'd start with many more watts, and try to first nurse them back to health first. Panels are getting cheaper all the time. I just picked up four 250W panels off ebay two weeks ago (local pickup, no shipping) for 392$. That's the "with sales tax", out the door price. I bought these to position in a due west orientation for extra pumping time around 5pm.

t00ls · April 2019

"It doesn't seem to take much to knock them down to 24 volts or less and it also doesn't take a lot to bring them back up to float."

this is a classic sign of sulphation on the plates....if you look inside, the plates will be a grey color.....healthy batteries are a chocolate color

while you can add more panels ...this wont solve your battery issue

most of the time when you get new batteries (flooded lead acid) you can look inside and the plates will be grey in color....thats sulphation.....healthy batteries have a chocolate brown color on the plates....one reason they say not to mix old and new batteries is because the new ones will have more sulphation buildup on the plates and will lower the overall amphours

my suggestion is to equalize charge the batteries every other day untill the batteries take longer to charge....when batteries take longer to charge it means there is more resistance inside of them meaning there isnt and sulphation shorting anywhere or blocking the chemical reaction inside

I have 3 banks of 6volt batteries in 48 volt configuration....I used to only have 2

I thought by buying the new ones it would increase my capacity...it didnt...it actually lowered it

one day I was servicing the older 2 and left the new ones on solar charge by themselves through an equalize charge for 2 days...when I hooked the other 2 back in, I noticed I didnt have to run the generator at all through the night, the next day, I found out that is was taking longer than usual to get the up to 57 volts ...and again that night, didnt have to run the generator

so I did 2 more equalize charges on the new bank again by themselves.....and now I havent run the generator hardly at all...even on cloudy days and heavy loads

Estragon · April 2019

@t00ls - With flooded batteries, measuring and recording specific gravity of electrolyte regularly is key to minimizing sulfation while also minimizing redundant wear on batteries.

EQ is a controlled overcharge, causing heat, gassing, etc. It's good to do if SG is chronically low, or there are relatively large differences between cells, but only to the extent it gets SG back to where it should be and as consistent as possible. EQing beyond this point is just aging the bank needlessly IMHO.

Reading SG is the only way to know in a quantifiable way what's going on chemically in the cell, and is the best way to manage charging regime for flooded batteries.

706jim · April 2019

In my case after reading the various posts here, I think its a combination of undercharging due to the lower float and bulk voltages I was using and the fact that there is a small but constant drain from the fridge. When I set up these batteries, I installed a marine switch allowing me to run or charge each string individually. If I still cant get the SG higher with the whole bank on charge I can isolate either string to bring them up. In this case, charge current would be at least 30 amps for a 370 ah string which is well within the 5%-13% recommended by Trojan.

During the time I've had these batteries, they have used virtually NO water and I'm hoping this is a good sign. I'll be able to adjust charge voltage at the end of May.

Estragon · April 2019

When my L16s were new, they also used very little water (~1gal in the 1st 6mos for 12 batteries), but not none. Using NO water suggests undercharging to me. IMHO, proper absorb voltage and timing should result in a bit of mild fizzing and gassing, which would use a bit of water.

Five years on, they're using maybe a bit under twice as much water. I'm also EQing a bit more and for a bit longer though, which would account for some of the increase.

Photowhit · April 2019

@706jim "which is well within the 5%-13% recommended by Trojan"

Trojan recommends a 10-13% charge rate...

706jim · June 2019

Remedial actions that have woken up those L16's: Bulk voltage increased from 28.8 to 29.8. Float voltage increased from 26.4 to 27.6.
Panels doubled with 1kw facing southeast and 1kw facing southwest. Equalized at 31.4 volts for 2 hours. SG close to 1.25. My daily loads 2kw-hr when I'm there and 1kw-hr when I'm not. Overnight voltage drop decreased now goes to about 24.6. Can (but haven't yet) charged the two strings of 4 L16's separately; will if SG fluctuates. Thanks for the suggestions given here.

paulcheung · June 2019

You need to get a good charge controller. if the charge controller is the one on you signature you using to charge this battery bank. you are wasting a lot of panel power. get a true MPPT charge controller like Midnite Classic 150 or Outback Flex 60 or 80.

706jim · June 2019

MPPT controller will be a next season upgrade now that I have a matched set of panels. The old Arcos will be retired and just the new Canadian solar panels connected. This system will never be increased in power so a modest MPPT controller would work fine. Two sets of three panels in series would give me 9 + 9 amps at an open circuit voltage of about 135 so a controller capable of handling this current would be all that I require. I'll check out your suggestions.

Photowhit · July 2019

706jim said:

Two sets of three panels in series would give me 9 + 9 amps at an open circuit voltage of about 135 so a controller capable of handling this current would be all that I require. I'll check out your suggestions.

Charge controllers are rated at output current, So a 2000 watt array at charging voltage for a 24 volt bank would be 2000÷28.8 = 69 amps, realistically about 80% of that for NOCT. But a 60+ amp charge controller. I like Midnite Classics, but all of the name brand by Outback, Morningstar, Schneider are good units.

Estragon · July 2019

For sizing controllers, I'd use voltage in bulk at low-ish SOC. Say 22v. 2000÷22v =91a. Yeah, that's conservative given NOCT output, but...

L16's low capacity

Comments

Categories