The key issue is that lead acid at these low capacities may be able to take one or two amps above 50% SoC. Especially at 80%+ if the panel can do 4 amps but you can charge only with 2, you are wasting precious sunlight.
Lead-acid can charge at 0.3C but only when empty until 50% or so. In the absorption phase, current drops (dramatically).
Also lead-acid means 50% usable capacity for longevity so the 12v 12Ah lithium is 12v 24Ah lead-acid (minimum). Which also means you can never charge with 0.3C. And even then is longevity of lifepo4 so much better it’s not funny.
Please spend the extra money on lifepo4, save yourself some headaches.
> Also lead-acid means 50% usable capacity for longevity so the 12v 12Ah lithium is 12v 24Ah lead-acid (minimum).
That's not a concern, is it? We're talking about battery backup, meaning the "expected use case" is that the Lead-Acid battery is sitting at near 100% capacity for almost its entire life, without significant amounts of discharge. Sure, you've got to discharge when the sun goes down but I don't expect a properly-sized battery to be deeply discharged (especially when we're aiming for multiple days worth of "worst-case" power, like multiple days of clouds/rain to reduce our power-collection).
Lets just be dumb with napkin math for a second: 12-hours of charge, 12-hours of discharge on a "typical" day, with a 3-day / 72-hour period of worst-case charge. Where does this leave us?
12-hours of typical discharge / 72-hours of worst-case discharge == DoD of 100% to 83% on your typical day-to-day basis.
Hardly anything that considers a "deep discharge". And if we really need deep-discharges, most Lead-Acids can survive something on the order of 300x to 500x deep discharges right? So its not like the 5 to 10 times a year where you need all 72-hours of battery are significantly hampering your lifespan?
> And even then is longevity of lifepo4 so much better it’s not funny.
Measured in deep cycle-counts (ie: cell phones or laptop usage), sure. (like 3000 cycles on LiFePo4 vs 500 on Lead Acid). But years of sitting on a float-charge, I don't think its that much better than Lead-Acid.
----------
That being said, LiFePo4 is a "superior" battery for sure, and its much cheaper than I remember. But Lead Acid remains significantly lower-priced than LiFePo4 as far as I can tell (2-to-1 or less)
I think you have a point about Lead-Acid charging slower when its near full (and keeping it full is key to keeping Lead-Acid long-lasting). So I admit to not recognizing that (LiFePo4 also charges slower when near full, its just that it charges like 3x to 4x faster than Lead Acid so its less of a concern)
> Please spend the extra money on lifepo4, save yourself some headaches.
We're at the hobbyist level where the original post decided that MPPT solar charge controllers weren't needed (at least, in an earlier version of this project)
Clearly no one here is talking about the strictly most optimal design, but are instead talking about simpler and cheaper solutions.
--------
Simplicity here is key. LiFePo4 does _NOT_ like topping charges or staying at 100% charge for long periods of time. Lead Acid however, prefers 100% charge and UPS use Lead Acids for this reason.
LiFePo4, in UPS/Standby like systems, need to stop charging, and wait for the battery to be used before engaging in a new charge cycle.
Lead-Acid on the other hand, can just have a constant voltage applied (ex: 13.5V or so) and you can just hold that indefinitely, for incredibly simple standby circuits.
I may be full of crap, it's been over 15 years since I worked on this stuff. You have been warned.
The systems I worked on used multiple deep discharge, cold weather hardened lead acid batteries.
Deep discharge batteries have an important distinction from automotive batteries. They tolerate significantly more deep discharge cycles but can't produce as many amps. If you're just trickling a more or less steady amount of power into some electronics, you don't need massive amounts of cranking amps that an auto battery provide.
Solar PV and battery tech have definitely improved quite a bit since then, so were I now building these same installations, I might do something entirely different. But "back in the day" your advice of "buy another battery" was more or less exactly how we handled charge rate limitations.
LifePO4 was about the same as lead acid in terms of lifetime costs (i.e. cycles * usable capacity / $) several years ago and have gotten significantly cheaper.
Pretty much the only reason to buy lead acid for a solar generator that faces daily partial discharge is because you can't afford sufficient LifePO4 capacity in the short term but don't care that you'll pay more long term.
> As far as I know, Lead-Acid still is very popular in UPS, which is basically what this SolarPi is.
No, SolarPi is a solar powered server. The application here is a "solar generator" type setup that has no grid connection and is very different from a UPS style setup.
A 3W Rasp. Pi with 12V 20Ah lead-acid backup with a 500 cycle lifespan should last 1500 days at a minimum. Because it takes 3+ days for a discharge cycle to occur, so that's easily 4+ years of worst-case durability.
And I'm willing to bet that if you design the solar panel CORRECTLY, you ain't gonna be doing full discharges every 3 days (IE: enough solar panels to have 99%+ uptime, meaning you've overbuilt solar panels to handle the worst case scenarios and try to stay at 100% power as much as possible).
> No, SolarPi is a solar powered server. The application here is a "solar generator" type setup that has no grid connection and is very different from a UPS style setup.
If you're hoping for 90% or 99% uptime, this solar application is going to look closer-and-closer like a UPS. 90% means you're down at most 36-days of the year (pretty crap reliability, all else considered).
------------
> Grid backup
Also, the SolarPi guy is looking into Grid-backup systems. So yes, we're back to UPS
Your durability calculations use made-up numbers and don't reflect a great understanding of lead acid and it's limitations. Assuming 12hrs of solar charge is ridiculous and a 12v 20ah lead acid will provide ~120wh (since you can only use 50%) which seems like about half the capacity you would need for your 72 hr "worst case".
While you can get 500 "deep" (50% usage) cycles out of a lead acid battery that is specifically designed for that, you'll have lost almost half your capacity by the end. You also ignore the extra wear from partial recharges which will be pretty common since the top 20% capacity charges so slowly
You have to buy 2x as much lead acid capacity, it lasts less than half as long and you get worse charging efficiency. The only time lead acid makes sense is when you need a lot of capacity but that capacity is only used infrequently and you have plenty of time to recharge between usage events. That is not the use case here.
> If you're hoping for 90% or 99% uptime, this solar application is going to look closer-and-closer like a UPS.
No, it is still a solar generator with a solar generator usage cycle. You still have the paired problems of reduced lead acid charging efficiency at the top 20% combined with extra battery wear for every day that you don't get the batteries topped up despite that efficiency. In order to reduce this, you have to have enough extra batteries that they can soak in your solar output even when in the slower charging of the top 20% while also having enough solar capability to get back up to 100% on partially sunny days. These issues significantly hamper the viability of lead acid for solar generators and are why LifePO4 has completely dominated Lead Acid in that market.
> Also, the SolarPi guy is looking into Grid-backup systems. So yes, we're back to UPS
The goal here is to switch to grid only when the battery runs out. That is the opposite of a UPS and is still a solar generator.
I think you're forgetting that this is a $30 Rasp. Pi battery solution that I'm talking about.
When your battery wears out, recycle the old Lead Acid battery and buy a new one. Its not that big of a deal. The battery I spec'd out is a 15lb battery (~7 kg). Heavier than LiFePo4 for sure, but no one is going to have issues moving, replacing, or buying new batteries.
--------
Anyway, I recognize that LiFePo4 is superior durability by a long shot. Thousands of cycles instead of hundreds, yes. My overall point however, is that "hundreds of cycles" is still years in this use case, and I don't think you've challenged the math too much from that perspective.
Obviously, I'd prefer to keep on the cheaper-side of Lead-Acid batteries. But the more expensive Trojan AES batteries or "Gel" batteries have far better durability (rather than the cheapest AGMs that I've been talking about)
I'm talking 1400 cycles at 100% DoD without any precharge or low-charge penalties for Trojan AES (though this is a very new battery only coming out this year). Trojan is working on making Lead Acid competitive vs Lithium Ion and have the cream-of-the-crop Lead Acid chemistries.
But even "Gel" (more expensive than AGMs) type Lead Acids have specifications for 100% DoD cycles, 300 to 500 depending on brand.
---------------
Anyway, this has gone on for many posts. But here's my arguments to finalize.
1. Float-charging "feature" for Lead Acid grossly simplifies circuit design. You possibly don't need a charge controller at all anymore (feasible design with just diodes), although if you want decent charge speeds you'll need to build your own charge controller. (though this seems doable with uCs and/or OpAmps to make CC-CV circuits).
2. Very low explosion risk, so this is appropriate for hobbyists (unlike Li-ion which even LiFePo4 have had some worrying fires). So DYI controllers for LiFePo4 is an issue of trusting the battery-controller and the brand you buy, details I don't have to worry about with Lead Acid.
3. Lead Acid has less durability, "hundreds" depending on brand (and yes, I'm talking about 100% DoD specs, not 50% specs like you wanna talk about). By my estimation, this provides years of service though this obviously depends on various assumptions.
4. Lead Acid's linear voltage is an advantage when determining DoD when building your own controllers (ex: 13V OCV means you're likely full, and 11.6V OCV means you're likely empty, and 12.3V means you're at 50%). LiFePo4 has "constant voltage", which is better for efficiency and power but it makes state-of-charge far more difficult to perform if you're building your own charge circuitry.
----------
LiFePo4 is a "superior" battery from a durability perspective. But no one seems to want to float-charge this thing for extended periods of time, and I certainly don't want to have an explosion risk while I test that out.
And "downsides" (like Lead Acid's linear voltage drop, leading to lower efficiencies) are advantages with regards to circuit simplicity and circuit design. A *simpler* design is more suited for hobbyists, even if its got worse technical specs than a more sophisticated chemistry.
> But no one seems to want to float-charge this thing for extended periods of time, and I certainly don't want to have an explosion risk while I test that out.
You still seem to be missing the use case here, which still does not involve any extended periods of floating.
You seem fixated on building your own UPS with your own controller. Go for it, lead acid sounds like it might be the right choice for you.
However, for a solar generator like this one, LifePO4 is the cost effective solution.
Fancy VLRA batteries like the Trojan do suffer fewer deficiencies around DoD and partial charging, but they cost just about as much per capacity as LifePO4 and still offer significantly fewer cycles.
So, sure, if you don't care about lifetime costs/cycles, leave battery charged most of the time, or want to build your own charge controller, then lead acid could be a better choice.
However, for a solar generator, LifePO4 is better in almost every way.
Lead-acid can charge at 0.3C but only when empty until 50% or so. In the absorption phase, current drops (dramatically).
Also lead-acid means 50% usable capacity for longevity so the 12v 12Ah lithium is 12v 24Ah lead-acid (minimum). Which also means you can never charge with 0.3C. And even then is longevity of lifepo4 so much better it’s not funny.
Please spend the extra money on lifepo4, save yourself some headaches.