50% discharge on AGM batteries - old wife's tail

[Zwy]

all this talk about 'cycles', but what does that really mean?

is it simply whenever the battery gets discharged/recharged? Is there a time interval?

A cycle would be one full cycle of the battery usable capacity.

If you use 25% of the usable capacity per day, then 4 days would be a full cycle.

 

[TimE]

I can't see my lifePO4 making 6000 cycles. That will take me 83 years. Calender aging will get them long before that. You can get discount on cells that have been sat 18 months doing nothing. Presumably as they have already suffered some calender aging. If I get 15 years out of my battery, I will be delighted. About 1100 cycles

 

[SamDeleted]

So your deal is to insult people disguising it as a nice statement.
Can you say passive, aggressive?

You might very well think that, I couldn't possibly comment

 

[Tomthumb62]

Two methods actually.

I didn't have any experience with the Rovers setup / programming details or how it did battery SOC when I started the project, so I added on the same SOC monitor that I always use, a Bogart Trimetic with a shunt. This is my default monitor method for all projects. It isn't perfect but it is quite good. This provides an independent method of measuring what is going on for voltage and current. I use the WiFi module add on to make it easy to look at the info from a tablet pretty far away.

__________

In addition, I have an Extech 380950 ac / dc clamp meter. I think that it has been replaced by a newer model, but I did a search on line and they are still for sale at much less than I paid in the past.

I actually have 2 clamp meters, one for higher currents and this one for measuring lower currents in small wires such as 10 awg and smaller.

Basically the clamp / circle opens like a pliers and it goes around the wires to take the current measurements, so I did this measurement on the wire that runs from the solar charge controller - through the breaker and then on to the bus bar. If you take the measurements at night, this will tell you how much current is coming out of the batteries to power the solar charge controller.

This works well enough for my use, but there are more precise units out on the market. If you are going to own a solar system, owning a clamp meter is very handy.

It is possible that the one we had was defective and using more power than it was supposed to. Since I didn't buy it, he was communicating with renogy on trying to get a replacement but didn't succeed.

I spoke to them some about how they do the set points and SOC for a 48 volt LiFe system and some of the things that they said could be changed - we were not able to access. For AGMs we could access it, but not for Li batteries.

______

There are videos on youtube about how to use a clamp meter and real pros at using them if you dig deeper with a search engine.

Thank you, that’sa really helpful and detailed reply.

Agreed, the SOC reading on the Rover is horrible, but I don’t believe any more or less accurate than many budget charge controllers out there.

One of these days I’ll justify the expense of a clamp meter. I especially like the lower current one you linked to, as I think that would be very useful.

My new lithium battery measures amps in and out (as a sum total, not individual readings) which is useful but isn’t sensitive enough to capture lower current like under 1 amp. This is good enough for now, for a camp trailer. If it was for a full time residence, I’d have a proper shunt and two clamp meters.

However, you’ve given me an idea to measure current draw of the SCC at night using my multimeter which can measure DC amps up to 20A. I’ll report back when I have a chance to measure it.

 

[Hedges]

I can't see my lifePO4 making 6000 cycles. That will take me 83 years. Calender aging will get them long before that. You can get discount on cells that have been sat 18 months doing nothing. Presumably as they have already suffered some calender aging. If I get 15 years out of my battery, I will be delighted. About 1100 cycles

That is why I think best use of lithium is to size it so it is cycled nearly 100% every night. Wear it out over about 16 years.

Having 3 days autonomy without sun would be nice, but then battery cycles 15% most nights and may die of old age at a fraction of its cycle life.
Oversize PV and have backup generator? Or is price good enough to do as you do, buy excess capacity?

FLA forklift battery similarly sized may last 20 years, cycling 15% most of the time. And it has the capacity to provide those 3 days. Assuming it isn't murdered in that time. Then you get the maintenance hassles and loss from inefficiency.

Now that LiFePO4 can be had for similar purchase price (per kWh of capacity) it is very attractive.
Lead-acid prices do seem to vary by location. I think people in Australia put AGM around $0.50/kWh, while my SunXtender AGM were $0.25/kWh. I paid 2x what FLA costs and expect half the life, but wanted maintenance free for a relatively small bank.
I considered buying a bank of Rolls Surette FLA and saving them in dry state for the future, but the price was quite high.

 

[Tomthumb62]

That is why I think best use of lithium is to size it so it is cycled nearly 100% every night. Wear it out over about 16 years.

I think it’s all about use case. Like you said, if you want reserve battery for cloudy days, you’re never going to get your advertised cycle counts.

But besides cycle counts, there’s other reasons to choose lithium over lead acid. Mine is ease of charging. With limited solar capacity on our small camp trailer and not currently interested in the expense and complexity of adding DC-DC charging, the flux of available solar wattage is perfect for lithium. No long absorb time and it doesn’t care if you charge at 3A or max C rate. With lead acid, it’s far more picky on how it’s happy with charging and I’m very glad to be done with that. Our camp trailer solar setup just isn’t big enough to properly charge a lead acid system and not severely shorten their lifespan.

 

[timselectric]

What about calendar aging?

It's the only issue.

What about temperature?

LiFePO4's like the same environment as humans.
If you are comfortable, so are they.

6000 cycles is quoted at 25degC.

That's being used in an EV. Which is what they are made for. Solar storage is a lot less stressful for them. Like an Elephant carrying a golf ball.

We are in the Tropics with batteries constantly around 30degC.

You should be cooling them down.

Lots of people on boats get 8 years out of 6v Trojan T105s. A friend has 6v batteries installed on his house in Antigua for the last 12 years. They are off grid. They are in a high ambient condition but are still going strong.

That's very impressive. They must be doing great maintenance.

 

[Supervstech]

I think it’s all about use case. Like you said, if you want reserve battery for cloudy days, you’re never going to get your advertised cycle counts.

But besides cycle counts, there’s other reasons to choose lithium over lead acid. Mine is ease of charging. With limited solar capacity on our small camp trailer and not currently interested in the expense and complexity of adding DC-DC charging, the flux of available solar wattage is perfect for lithium. No long absorb time and it doesn’t care if you charge at 3A or max C rate. With lead acid, it’s far more picky on how it’s happy with charging and I’m very glad to be done with that. Our camp trailer solar setup just isn’t big enough to properly charge a lead acid system and not severely shorten their lifespan.

Absolutely agree.
Fund issues say otherwise for many though.

 

[TimE]

That is why I think best use of lithium is to size it so it is cycled nearly 100% every night. Wear it out over about 16 years.

Having 3 days autonomy without sun would be nice, but then battery cycles 15% most nights and may die of old age at a fraction of its cycle life.
Oversize PV and have backup generator? Or is price good enough to do as you do, buy excess capacity?

FLA forklift battery similarly sized may last 20 years, cycling 15% most of the time. And it has the capacity to provide those 3 days. Assuming it isn't murdered in that time. Then you get the maintenance hassles and loss from inefficiency.

Now that LiFePO4 can be had for similar purchase price (per kWh of capacity) it is very attractive.
Lead-acid prices do seem to vary by location. I think people in Australia put AGM around $0.50/kWh, while my SunXtender AGM were $0.25/kWh. I paid 2x what FLA costs and expect half the life, but wanted maintenance free for a relatively small bank.
I considered buying a bank of Rolls Surette FLA and saving them in dry state for the future, but the price was quite high.

The biggest benefit for me isn't the 6000 cycles. It's the fact that I can heat hot water, run the 1.8kw watermaker motor and make 200 litres per hour and cook with the induction hob. I couldn't do this with my lead bank. The solar now makes 20% more output as it never goes off bulk phase. We don't need to get the lithium to 100% each day to ensure good battery life. The improvement with lithium over lead, in many ways are quality of life issues on a boat. We don't use gas to cook on apart from using the oven about once a week. The diesel generator has run 3 times in 2 months just to see if it still runs. We are basically using almost zero gas and diesel whilst at anchor

 

[wattmatters]

A cycle would be one full cycle of the battery usable capacity.

It depends on how manufacturers define it. There is no standard.

e.g. often for a LiFePO₄ BMS a cycle count occurs after a combined discharge of 80% of the the rated capacity, unless the manufacturer provides a specific DOD value. That's how my LiFePO₄ battery BMSs calculate cycle count.

On the other hand my SLA batteries have this chart showing the cycle count vs DOD using the specific charge and discharge protocols. It shows a typical hyperbolic-like curve:



As I said earlier, claimed cycle counts need to be accompanied with the specific protocols they apply to.

They really are only a general guide.

All I know is, based on the chart above, my SLA will last a long time doing what they are currently doing, with a daily DOD of no more than 5-10%, mostly kept at float and ready to provide supplemental outage backup power if needed. At $US 6.5c/Wh second hand, they are a really good option for this job. I see little point wasting money on LiFePO₄ at US$0.32/Wh to sit there doing SFA most of the time. Instead my LiFePO₄ is sized to deeply cycle each day.

 

[Hedges]

I couldn't do this with my lead bank. The solar now makes 20% more output as it never goes off bulk phase. We don't need to get the lithium to 100% each day to ensure good battery life.

The fact you can run SoC up and down in the middle does make management better.

What I would like is a linearly variable dump load. Good for using excess power during lead-acid absorption cycle. For you, could also use surplus. But with lithium you might just micro-cycle instead. Turn on water heater full power when ever battery gets full.

I could imagine a Variac with servo control, expensive way to feed resistive load. A PV direct system can accomplish MPPT into a heating element with just a transistor, no inductor needed. But that is separate from your charging system. To run off AC, either use light dimmer (bad PF), or switcher with inductor.

Variable speed pump was what I had planned on. But poor PF of VFD upset my inverter. If I add PF correction front-end, then it could be my variable dump-load after all.

 

[Sabre36]

not just odyssey, also Trojan, Concorde, and Rolls AGM.

And the agm DoD vs lifespan graph he showed is very similar to other manufacturers you can find online

The trick here is those graphs are not derived from testing but rather a mathematical assumption. As one of the first marine electrical businesses in the country offering capacity testing of AGM & Gel batteries I can assure you going to 80% DoD murders AGM batteries including Odyssey & Northstar. Heck just 30 PSOC cycles to 11.7 is enough to kill an East Penn AGM. When Lifeline first hit the market they told everyone 80% DoD was ok, it was not and within two years they revised to to 50%.. Their energy throughput graph still shows 80% DoD as the same as 50% but it’s not. this is because these graphs are not actually derived through physical testing but rather a mathematical assumption. D’oh…

 

[TimE]

It depends on how manufacturers define it. There is no standard.

e.g. often for a LiFePO₄ BMS a cycle count occurs after a combined discharge of 80% of the the rated capacity, unless the manufacturer provides a specific DOD value. That's how my LiFePO₄ battery BMSs calculate cycle count.

On the other hand my SLA batteries have this chart showing the cycle count vs DOD using the specific charge and discharge protocols. It shows a typical hyperbolic-like curve:

View attachment 154523

As I said earlier, claimed cycle counts need to be accompanied with the specific protocols they apply to.

They really are only a general guide.

All I know is, based on the chart above, my SLA will last a long time doing what they are currently doing, with a daily DOD of no more than 5-10%, mostly kept at float and ready to provide supplemental outage backup power if needed. At $US 6.5c/Wh second hand, they are a really good option for this job. I see little point wasting money on LiFePO₄ at US$0.32/Wh to sit there doing SFA most of the time. Instead my LiFePO₄ is sized to deeply cycle each day.

I think if you have a back up only system, then lead can be a good solution. For cycling batteries on a daily basis, it's hard to beat lifepo4

 

[althemusicwizard]

I fitted my system in Dec 2022. I have 32 x sealed lead-acid VARTA LFD230 batteries, a kind of hybrid starter / habitation battery. They weigh 56kg (123.5 lbs) each. I have two tonne of lead in my garage. I sized it so big (c. 90kWh) so that I would only use around 30% of it over 2-3 days (my average consumption being about 15kWh per day). These were 'used' batteries and cost £100 each. The bank cost £3200. I had ordered 4 x SOK Rack batteries (5kWh) for £8800, but cancelled the order. I can now buy 5 x SOK batteries and a free rack for £7400 :

Fully Loaded 25kWh SOK 48v Rack Battery Kit | SOK Battery Europe

EU - Pre-order. Will be in stock in mid-May. All pre-order can enjoy - €100 discount. Contact cusotmer service to get the refund.UK - Out of stock.Kit Includes: 5x SOK 48v 100ah LiFePO4 Server Rack Batteries 1x 5 Slot Rack 1 set of 3m RS232-RJ11 communication cables FREE shipping to all EU/UK area

www.europe.sokbattery.com

Many 48V batteries coming in now, like Seplos Mason, Gopel etc (all being reviewed on Any's 'off-grid garage' Youtube), you can buy 15kWh for £2500 from Fogstar if you accept B Grade prismatic cells and building the kit yourself.

After 9 months of using the VARTA bank, I noticed after September when the solar production falls sharply, the bank was dropping to 47V around 6pm. I get four hours cheap electricity between 00.30 - 4.30 so need the bank to survive until then, so following this ROLLS document,

https://rollsbattery.com/wp-content/uploads/2018/01/Rolls_Battery_Manual.pdf

I've increased the charging now to 60V and the absorption rate to 3 hours. (I also dismantled the bank and charged / tested individual batteries to try and ascertain what was happening). Lots of people (ie. stuff I've read in the past) think this will destroy the batteries. I've been quoted £20 for a scrap price (but interestingly lead's scrap price puts the batteries lead content around £70 per battery...Is this how the scrap-yards make their money?). Anyhow, as it turns out, the voltage is now holding up much better. I've come to think of the battery in any install simply as the buffer between the inverters and the solar panels.
Now that LiFePO4 batteries have come down in price in the UK, my next purchase wouldn't be lead acid. My back thanks me for that.

I also added an Orange Pi and 'Solar Assistant' software to my setup and that has been a revelation. It has simplified and augmented the data I can now collect and instantly see the effect a certain item has on my install. I would highly recommend it to anyone that has an inverter setup that does not have its own app or data collection.

 

[Crowz]

I ran FLA's back at the old house and never actually killed them but I usually discharged to a bit under 50% every day. The first bank of them "died" after 6 years. I said "died" because they still worked but the runtimes were getting pretty bad by then. The headaches alone with checking fluids, worry about going to low on discharge and general worry all the time made Lifepo's a real game changer for me.

I only have the single lifepo4 on the new system and use it as a battery backup right now but I have run it some for testing and it performs awesome for me. Not having to worry about it and the knowledge I won't have to worry about the future batteries is a game changer in my opinion. It having such a flat voltage curve compared to lead is a big plus too. I won't be going back but I haven't made much progress in ramping up the size via buying more since it cost so much per battery. It's more of a hobby this time around so I'm going slower than last time.

That said I do have 4 deka intimidator agm batteries in my 48v mower and it has been going strong for 3 years now cutting 3 to 5 acres worth a go during grass cutting season. Still has full runtime and gets used hard. So they have performed nicely and I won't be changing them out to lifepo4's till they finally die but there is always that worry about discharging them to far that bugs me while using the mower.

 

[SamDeleted]

my next purchase wouldn't be lead acid. My back thanks me for that.

???