diy solar

diy solar

all off grid is a gamble

12V 200Ah wired 4s3p for 48V 600Ah
Cycling 25% DoD would be 150 Ah, 7.5 kWh

The AGM brand I use is supposed to give a bit over 2000 cycles to 25% DoD. Some brands might be half that, some FLA much longer life.

You said the loss of capacity was observed a few months after lightning strike. Previously, had you observed the battery still had full capacity by occasionally cycling it deeply? Or did you measure rest voltage after having drawn 7.5 kWh overnight, and observed that it still looked like 75% SoC? Because AGM, you couldn't measure SG like you would for FLA.

If you weren't aware while it was happening, could be capacity was continually degrading. Initially 600 Ah so your 150 Ah draw was just 25%, but later 50% and 80%, eventually drained at 4:00 AM.
Once capacity dropped to the point where cycling was deep, they would degrade more rapidly. Once a cell was imbalanced, it would remain undercharged and degrade faster.
Some AGM vendors do recommend equalization when capacity loss is observed. I have some automotive AGM which had low resting voltage and this process brought it up a bit. Not sure if they actually equalized, or if the batteries just weren't being properly charged by alternator or battery charger.

Alternatively, could lightning have affected charge controller? Have you confirmed that charge voltages are correct? If it stopped charging to the correct absorption voltage, then battery wouldn't be maintained as well.

4x TS MPPT 60
If any under-charged the battery, others should still deliver full charge. If any overcharged the battery it would be overcharged.


Settings ... communication from Chinese supplier ...
Some charge controllers are more sophisticated than others. Some work by voltage alone, while others track amp-hours as well. Certainly the settings (volts, absorption time) need to be correct for the batteries. Default VRLA settings of my inverter/charger were close, but I adjusted them slightly for what battery vendor recommends.

Batteries were $1200 (each?, 200 Ah at 12V)
What brand/model? Link to data sheet?
My SunXtender are 400 Ah at 6V, so same capacity. L16 size, 120 lbs each. They cost $600 each. I have 8 of them, 8s for 48v.


When you switch to lithium, you'll no longer have the problem that less than full charge causes degradation. So long as you set max/min charge correctly, SoC won't harm the battery.
What temperature will the batteries live at? If LiFePO4 is kept at moderate temperatures, it can accept full charge current.
How big a back are you putting in? Likely only half as big as the AGM or less, since they can be cycled more deeply.
Maybe 16, 208 Ah cells for 13,400 Wh?

You've got 40 x 200 = 8000W PV panels.
That was about 0.25C for your AGM. Hopefully within recommended charge rate, which does vary by brand/type.
That would be around 0.5C of 13,400 Wh LiFePO4 bank, which should be OK at room temperature.
If battery gets close to zero degrees C but not quite to low-temperature disconnect, acceptable charge current is probably around 0.1C, 1340W, 28 A. If you exceed that while LiFePO4 is very cold, would shorten life.
Probably easiest way to avoid that is keep battery in a warm location, and set low-temperature (charging) disconnect at a temperature where 0.5C is acceptable.
the batteries that were 1200 apiece here in japan would probably be half of that in the US, remember different cost vs locations. they look similar to my current agm batteries, but i bought them from Rakuten Japan. and yes that was not a typo there were at the time ¥120,000 per battery at last weeks ¥/$ or 107=1 it is just a bit shy of $1200.00 per battery. that's what made e go looking at Alibaba ad tried to find batteries with the same spec. the ones from Alibaba were listed as Green Tree energy 200 a/h 12 volt AGM batteries. they weighed the same were the same size and had similar charge and discharge profiles. regardless they are on their way out and my CALB 200 a/h batteries just cleared customs and we shall see how they do. I bought enough of them to make a 2p16s battery pack which should give me 19,200 w/h.
 
1000 cycles is almost three years, so depending on the quality of the batteries and how much they where used, I am not sure I would be super disappointed.

Can you share details on your current setup?
My AGM's lasted a lot longer than that, daily for over 5 years.
 
My AGM's lasted a lot longer than that, daily for over 5 years.
Yeah batteries are a crap shoot, but obviously there are better brands that last longer. I had a Optima Red Top in my old S-10 that lasted 13 years in Texas heat (it might still be going as I sold the truck). I am just saying that if I got 1000 cycles out of mediocre quality batteries, I wouldn't be devastated.
 
No question LFP is the way to go these days. But when it comes to AGM, you can make it less of a gamble with knowledge about float and its problems for daily-cycling especially with solar!

To actually charge an agm to full, to avoid premature capacity walk-down by never getting that last 1% charged, one needs at least 8-12 hours of non-maintenance level float to actually FINISH the charge. You can't hurry it. Just finishing absorb is not enough! It takes 8-12 hours float to finish.

You can see for daily cycling, whether it is a rapid-turnaround utility vehicle (forklift whatever doing back to back shifts) - or for many solar users, this is nigh impossible to achieve. But there is a strategy to help overcome this.

If you don't actually reach CV and transition to "absorb" where tail current is falling, or spend too little time in it, well that's pretty bad undercharge. Cycle by cycle you walk down capacity due to that last 1% hard sulfating.

OR, if you do reach CV/Absorb, but the SCC timer is short enough that by late afternoon you don't spend much time in it and drop to float prematurely, another bummer.

This problem was well known to early EV users and pure-lead agm manufacturers too where rapid turnaround without adequate float time just killed agm's prematurely. Never mind the deep-cycle aspect. Lack of FULL recharge is a major part of the problem.

One solution the manufacturers came up with was to advise users to simply give their EV's a "day off" to float on the weekends. Eh, ok, but still not practical enough.

OR, you can use what's known as the CC/CV/CC charge (aka IUI) method. This is knife edge charging, and is such that no consumer knows how to do it properly or has off-the-shelf chargers to do so. So special industrialized chargers were created to do this right for the forklift /industrial crowd.

So what is a solar daily-cycler to do??

You adopt strategies in the race to float! You have to go beyond just normal methods mentioned in manuals, since they assume you are charging with a plentiful steady-state charger, not unreliable solar.

Because the sun goes down every night, the hacks presented here are not as harmful as one would think. Daily cyclers only please! And you are using "at the battery terminal" temperature compensation probes right?

Simplest method:
BULK = ABSORB = FLOAT
(Just set it ALL to one single CV voltage, like 14.6v)

Second common method if your SCC drops to float prematurely before the sun sets:
Change your float value to a relatively high value, like 13.8v

Third and most costly method:
Run your genny during the bulk phase as hard as you can - say in the early morning - and let solar finish the job.

Still, even with these hacks, these may not result in a truly full daily charge. Time is not on your side with AGM as a daily cyclist!

Boaters used to run into the same - at least once a week or biweekly as a PM measure, if they couldn't get to float and stay in it, that meant a lot of genny time. If you didn't want to walk down your expensive Concorde's, it was worth it.

Assuming your bank is relatively healthy, and not unbalanced, you may want to investigate giving some of these workarounds a try. Panel up if you can to get in as much float at the end as you can. See the strategy here?
 
No question LFP is the way to go these days. But when it comes to AGM, you can make it less of a gamble with knowledge about float and its problems for daily-cycling especially with solar!

To actually charge an agm to full, to avoid premature capacity walk-down by never getting that last 1% charged, one needs at least 8-12 hours of non-maintenance level float to actually FINISH the charge. You can't hurry it. Just finishing absorb is not enough! It takes 8-12 hours float to finish.

You can see for daily cycling, whether it is a rapid-turnaround utility vehicle (forklift whatever doing back to back shifts) - or for many solar users, this is nigh impossible to achieve. But there is a strategy to help overcome this.

If you don't actually reach CV and transition to "absorb" where tail current is falling, or spend too little time in it, well that's pretty bad undercharge. Cycle by cycle you walk down capacity due to that last 1% hard sulfating.

OR, if you do reach CV/Absorb, but the SCC timer is short enough that by late afternoon you don't spend much time in it and drop to float prematurely, another bummer.

This problem was well known to early EV users and pure-lead agm manufacturers too where rapid turnaround without adequate float time just killed agm's prematurely. Never mind the deep-cycle aspect. Lack of FULL recharge is a major part of the problem.

One solution the manufacturers came up with was to advise users to simply give their EV's a "day off" to float on the weekends. Eh, ok, but still not practical enough.

OR, you can use what's known as the CC/CV/CC charge (aka IUI) method. This is knife edge charging, and is such that no consumer knows how to do it properly or has off-the-shelf chargers to do so. So special industrialized chargers were created to do this right for the forklift /industrial crowd.

So what is a solar daily-cycler to do??

You adopt strategies in the race to float! You have to go beyond just normal methods mentioned in manuals, since they assume you are charging with a plentiful steady-state charger, not unreliable solar.

Because the sun goes down every night, the hacks presented here are not as harmful as one would think. Daily cyclers only please! And you are using "at the battery terminal" temperature compensation probes right?

Simplest method:
BULK = ABSORB = FLOAT
(Just set it ALL to one single CV voltage, like 14.6v)

Second common method if your SCC drops to float prematurely before the sun sets:
Change your float value to a relatively high value, like 13.8v

Third and most costly method:
Run your genny during the bulk phase as hard as you can - say in the early morning - and let solar finish the job.

Still, even with these hacks, these may not result in a truly full daily charge. Time is not on your side with AGM as a daily cyclist!

Boaters used to run into the same - at least once a week or biweekly as a PM measure, if they couldn't get to float and stay in it, that meant a lot of genny time. If you didn't want to walk down your expensive Concorde's, it was worth it.

Assuming your bank is relatively healthy, and not unbalanced, you may want to investigate giving some of these workarounds a try. Panel up if you can to get in as much float at the end as you can. See the strategy here?
I like this answer, as it does make a lot of sense, currently this was a weekend only vacation house so on the weekends it would see use and then through the week it would get 5 days to full float recharge etc. I would guess that probably during last winter we were not getting enough float. hard to know at this point. regardless the LiFePo4's are busy top balancing so its off to the races again.
 
after 10 years offgrid it seems to me that all offgrid is a gamble I run magnum inverters, morning star inverters and sharp panels, and ali babba batteries this is a 4k setup that I am now expanding to 8k as of this post). at the same time I am also converting to lithium (LiFePo4's). my agm's lasted to about 1000 cycles, still trying to figure out why they died so quick, i had a lightening strike, and the possibility that the charging algorithm is not correct (though in line with the battery sellers instructions). Though I am almost sure its due to the low quality of ali babba/alipresss battery's or my settings. One last try with CALB cells are now in the works. a friend in thailand spent 2.5 years on these calb cells and is still working not sure now much more I can stand as a hobby. i can get all all the power needed from my solar setup, and when the sun is strong I can pump power like you would not believe out of the magnum inverter... it depressing at times. (no sponsorship no buyin's of any sort). (I am dyslexic so f you if you are a spelling or grammar Nazi.)


ken
Lots of interesting and thoughtful comments in this thread?. Looks like a lot of experienced users weighing in, always the best source for information. Personally I (we) have been using solar/battery power exclusively for 37 years and have been offgrid longer than that. Solar power can be costly but when I balance that against never paying an electrical bill in 37 years, no contest. Again speaking for myself, since batteries are the most costly and perishable aspect of an offgrid system, I have been diligent to baby them. Never using more than 20% of capacity to keep the cycles shallow. In the beginning I was getting four and a half years out of a set of RV batteries. When I finally graduated to individual 2 volt deep cycle cells I got 10 years plus. I just upgraded to my first set of Gel batteries and with the thought of keeping the cycles shallow I deliberately oversized my battery bank. The main reason I went with Gel is to get away from the gassing of flooded batteries . Even though I have a vented battery enclosure I didn’t like listening to the fan, which only ran when the sun was shining. I am hoping the gel batteries will live up to the manufacturers specifications of 7000 cycles at 20% depth of discharge. I realize not everyone can be this frugal with their batteries but we have adjusted our lifestyle to suit our needs. There are a lot of different parameters for those who live on battery power 24/7 and those who use solar/ batteries as a hobby or occasional weekend, or for RV’ing.
 
I understand Gel has more stringent limits on charge current and requires a long absorption. Also different voltages from other LA batteries.

Does your system regulate charge current? Most DC coupled systems deliver all available DC current from PV, and inverter uses any DC current needed. That leaves a variable charge current for battery. Too big a PV array and SCC means excessive charge current. A few DC coupled systems do measure battery current and seek to keep it at target value.
 
That's right - gel does have more stringent limits on the CV voltage. The problem is that in the past decades, marketing tries to equate them with AGM, because the consumer has no knowledge about differing battery types. Sometimes it is so bad, that the wrong cv values are even printed on the case, or cut-n-pasted from some other manufacturer's spec-sheets.

A modern analogy would be "Hey, I hear you are using a lithium battery bank! Is it LifeP04?"

"Dunno, I just set my charge controller to "lithium" - aren't they all the same? The guy at the store said that it's just a bigger version of my cellphone battery and just go with that". :)
 
I've been off grid for 20 years. My last forklift battery rated for 1500 cycles (when used in a forklift) lasted 10 years. I'm not sure how really to calculate cycles for off grid. I think folks typically just use fuzzy math and say a year has 300-365 cycles. Who knows. I think people who get super long lives out of LFA don't actually use the batteries much. I would drag my forklift battery to 60-70% normally, sometimes to 50%, and once in a while to 20%. It was $3000 new (same unit is $4000 now) so thats $3000/10 or $300 bucks a year. I've since replaced it with LifeP04 and so far would never go back to LFA. The LifeP04 are different beasts and you more or less have to cast aside your traditional LFA thought process. So far everything about them is better. I'll know how they really compare in ten years.
 
I've since replaced it with LifeP04 and so far would never go back to LFA. The LifeP04 are different beasts and you more or less have to cast aside your traditional LFA thought process. So far everything about them is better. I'll know how they really compare in ten years.

I think charge rate and temperature are a key to life (along with voltage limits). What is the max C rate your PV/SCC can deliver to the pack? How cold can it get?
 
I think charge rate and temperature are a key to life (along with voltage limits). What is the max C rate your PV/SCC can deliver to the pack? How cold can it get?
Batteries are in a heated building which stays above 50f. Max C rate is 15.
 
Sounds good.

0.15C, I suggest setting low-temperature cutout to prevent charging if the place ever drops to 10 degrees C (instead of 0 degrees C.)
Or another temperature if you have a good one for your batteries. So long as the heat stays on this would never happen, but prevents charging when cold.

In the following link, of the various brands the most restrictive limit I see is 0.12C below 7 degrees C.

 
Sounds good.

0.15C, I suggest setting low-temperature cutout to prevent charging if the place ever drops to 10 degrees C (instead of 0 degrees C.)
Or another temperature if you have a good one for your batteries. So long as the heat stays on this would never happen, but prevents charging when cold.

In the following link, of the various brands the most restrictive limit I see is 0.12C below 7 degrees C.

I have the cutout set at 40f.

thx
 
I have the cutout set at 40f.

thx
40F is 4C.
Some battery brands in the tables pasted by Dzl say 0.1C or 0.12C at that temperature, other brands say 0.2C

Your expected 50F (10C) minimum temperature allows much more current according to those tables.

Just thinking of a situation where the building gets cold, either fuel runs out or you're away. Then the sun comes out and charge rate might be enough to degrade battery faster. What I don't know is what conditions cause damage if it occurs once, vs. reduced capacity/cycle life if it is done repeatedly.
 
40F is 4C.
Some battery brands in the tables pasted by Dzl say 0.1C or 0.12C at that temperature, other brands say 0.2C

Your expected 50F (10C) minimum temperature allows much more current according to those tables.

Just thinking of a situation where the building gets cold, either fuel runs out or you're away. Then the sun comes out and charge rate might be enough to degrade battery faster. What I don't know is what conditions cause damage if it occurs once, vs. reduced capacity/cycle life if it is done repeatedly.
Based on the eve data, looks like I need to turn up the heat to 15c or 60f. That isn't a problem. Heat is provided by a non electric propane heater with a backup electric heater on a thermostat.

thx
 
Back
Top