With lead acids you just have to make sure you don't go below 40% charge. They don't tend to like that. The deep cycle ones are a bit better but I've experienced them deteriorating quickly too once you go too low. The other disadvantage is that they are quite heavy and might end up taking up lots of space. I'm personally still using lead acid, but with plans to go LiFePo4.

If you can't go below 40%, it might make sense to reduce the capacity by 40% accordingly when making comparison calculations against tech that can do that no problem. How does bang for buck look then with SLA versus LiFePO4?

I don't have the exact numbers but even then it's cheaper. The only issue is - can you trust yourself (or your planning) to ensure you don't go below 40%? Real life is hard to plan.

What happens if you do let it go below 40% (either as a brief dip to high 30s, or letting it go to 0)? Is it closer to a case of "oops you hit 39%, this battery is now worthless" or "oops you went down to 0%, better not do that every day as statistically the battery will last 0.001% less long"?

edit: I pasted my comment (prefaced with "context = lead-acid batteries") into ChatGPT and got this answer, maybe someone who knows could, instead of having to explain it all to me, just confirm GPT-4 got it right, or if not the correct it / add detail?

> If a lead-acid battery is allowed to go below 40% state of charge (SOC), it can cause some issues, but the effects are not necessarily immediate or catastrophic. The depth of discharge (DoD) and frequency of these deep discharges can impact the battery's overall lifespan and performance.

> 1. Brief dip to high 30s: If the battery's SOC occasionally dips into the high 30s, it may not cause significant harm. However, consistently discharging to that level can gradually decrease the battery's overall lifespan.

> 2. Discharge to 0%: If a lead-acid battery is allowed to discharge to 0%, it can cause more serious damage. Repeatedly discharging the battery to this level will dramatically shorten its lifespan and could potentially cause it to fail prematurely.

> The relationship between DoD and battery life is not linear. For example, discharging a battery to 50% DoD might give you around 1,000 to 1,200 cycles, while discharging it to 80% DoD might give you only 300 to 500 cycles. It is generally recommended to keep lead-acid batteries at a higher state of charge to maximize their lifespan and maintain optimal performance.

> In summary, allowing a lead-acid battery to occasionally dip below 40% SOC may not cause immediate damage, but consistently doing so can reduce its lifespan. Discharging the battery to 0% can cause even more significant harm, especially if done repeatedly. It is best to keep lead-acid batteries at a higher state of charge to ensure their longevity and maintain optimal performance.

ChatGPT is not wrong... but in my personal experience, once you go below 40% it's all downhill from there. At least if you've done it 2-3 times. Batteries are not always exactly the same, and temperature conditions vary as well. I've been nursing batteries for years, and this is just my anecdotal comments on that.

One thing that we can all agree on though is that you must never ever use car batteries. Lead acid need to be the deep cycle kind.

More deets from ChatGPT

> please provide an example data to show non-linar relationship between DoD and lifespan in 5% increments.

Depth of Discharge (DoD) Approximate Number of Cycles 05% 3,500 - 4,000 10% 3,000 - 3,500 15% 2,500 - 3,000 20% 2,000 - 2,500 25% 1,800 - 2,200 30% 1,600 - 2,000 35% 1,400 - 1,800 40% 1,200 - 1,600 45% 1,100 - 1,400 50% 1,000 - 1,200 55% 900 - 1,100 60% 800 - 1,000 65% 700 - 900 70% 600 - 800 75% 500 - 700 80% 300 - 500 85% 200 - 400 90% 100 - 300

> thinking through the following steps please create an asci bar chart for this table, dod given in 5% increments:

1. Calculate the number of cycles for the first increment 2. Extract the maximum number 3. Calculate the number of stars with 1 start = 500 cycles 4. Put number of starts into a table 5. iterate over the table

Once the table has been completed create the asci bar chart plotting DoD vs Cycles

Example: 1. 5% DoD = 3,500 - 4,000 2. 4000 3. 8 stars

DoD Cycles

05% ********

10% *******

15% ******

20% *****

25% ****

30% ****

35% ***

40% ***

45% **

50% **

55% **

60% **

65% *

70% *

75% *

80% *

85% *

90% *

Lol, the way this came out to sit perfectly with OP's intuition. Even down to the drop from 2-stars to 1-star!

Yeah, there's def the upfront cost, which not everyone can justify. Though, at the same time, a lower-amp-hour battery can provide as much actually-usable energy as a higher-amp-hour-rated SLA battery (so you could purchase a LiFePo4 battery with a lower amp-hour rating than the equivalent SLA battery but still get the same run time). I wouldn't know how much lower you can go though offhand.

48v 100ah LifePO4 is under 1000$ in China

It is but you have to make very sure that you actually get what you ordered and not some substitute with lower grade cells than what's advertised.

Are there easy(/cheap) ways to verify this if one does buy, as an individual, a cheap battery like that? Or do you need to either rely on brand reputation or have access to expensive equipment for testing?

I think vendor reputation is almost as important (or maybe even more important) as brand reputation. In many cases a brand is just a sticker, and your vendor may well have a longer relationship with their customer.

For instance, telling a brand name cell from a fake cell can be extremely hard, but your vendor knows exactly where they bought them. I buy cells from nkon.nl, who are a high rep local vendor and I've yet to be disappointed but buying a pack that is sealed from a low rep vendor may well incur a lot of risk. You'd almost have to split open the shrink wrap to make sure you got what you paid for. One quick and dirty way to check is to weigh the pack and to compare the weight to the manufacturers spec for the cells. It's obviously not perfect but many fake cells have weights that are wildly different from the cells they pretend to copy (usually due to much lower actual capacity).

You can do a capacity test quite easily, that will only tell if the capacity is the rated one. You will have to check for physical signs of bad cells, they enlarge (bloat), but that means you need to open the case (sometimes it will void any warranty) and check each cell.

Discharge & record the curve. All it takes is one poor-performing cell & your whole pack is going to be hurting bad. Discharge down to like 2.85V * number of cells. The curve should be nice & even. If there are places where the voltage drops before the big drop at the end, it's a sign that one cell isn't performing.

If your has equalization, try to discharge & equalize charge it 2-3 times first.

If you can actually get access to the cells, measuring cell resistance is not expensive. Anomolies will show themselves very quickly. Everything should be pretty clumped together in internal-cell-resistance. If you have a pack equalizer, it can be used to help you read this out. My <$100 charger has a cell resistance test mode.

> All it takes is one poor-performing cell & your whole pack is going to be hurting bad.

Yes, that can't be stressed enough. And depending on the fault and how the BMS handles it a more dangerous condition could easily develop. For instance an undetected short would cause the other batteries to be overcharged to the point where they are much more likely to die themselves, and some of those may do so in pretty violent ways.

Cheap / missing BMS are a serious problem with cheap Lithium-Ion packs, they are the first line of defense and bad cells can even happen when supplied by good brands (but the chances are a lot higher with no-name stuff).

Another good test: charge the pack, balance it, measure the pack voltage and the individual cell voltages, then let it rest for a day or two and measure again. Any kind of deviation between cells is a sure sign something is not working well and is the earliest warning sign for trouble you can get. The rate of self discharge should be the same for all cells and it should be very low, too low to measure (and make sure you disable the balancer after the initial charge!). After a month or so it would be fine to see the pack drop a little bit, but not a lot.

Yet another quick test: use an IR camera to look at the pack both during charging and at rest. Any cells that are warmer than those around them are suspect.