Different Battery Chemistries

I thought I would make a new thread here for discussing and comparing different battery chemistries since I think @Vassen’s member introduction thread is probably not the best place for it.

I just found this interesting article:

Here is a short excerpt:
Under strict test conditions, commercially available lithium cells of both types were repeatedly discharged and charged from 0% to 100%. The result? According to the paper, “The LFP cells exhibit substantially longer cycle life spans under the examined conditions.”

This is rather interesting since we know one popular new brand is using NMC cells and is claiming similar performance to LiFePO4 cells.


Whereas lead-acid shouldn’t be charged until it’s depleted to 20% battery capacity, Lithium-ion batteries thrive on what it calls opportunity charging

I thought you shouldn’t discharge LA to below 50%?

Very interesting read.

Oh and thanks for not detailing my thread. :grinning_face_with_smiling_eyes:

Thank you for posting this link. What I take out of this is the amount a Lead Acid should be charged and discharged. I had a view that it should not be used below 50 percent and then charge to 100 percent.

The one thing that stuck in my mind was that with batteries, do NOT let them die of old age.

Article references forklift batteries, I had Trojan’s … the forklift.golf cart owners said they last 5 year, used or not.

Mine lasted 5 years, and then died not long after I sold them.
Should have used them as suggested.

Lifepo4, I’m USING them this round. NOT abusing …

I thought so too, but reading the article they seem to be talking mostly about use in forklifts, which I guess would be different. What I have seen in forklifts that are used constantly is that they usually have a battery on charge permanently. The driver will use the battery in his forklift for his entire shift and then the battery is removed and replaced with the fully charged battery, while the used battery is put on charge again. I guess it’s difficult to know how deeply the battery is discharged during each shift though. In daytime only operations, the forklift is usually just plugged into a charger over-night so it starts with a full battery the next morning. Also difficult to guess how much is used during the day.

A big advantage of using lithium batteries in forklifts is what they call opportunity charging. Since lithium batteries can charge so much faster than lead-acid batteries, the driver can plug the forklift into the charger any time he has a few minutes, i.e. tea, lunch or bathroom break and that can put enough back into the battery that they can use it indefinitely without the need to swap batteries or leave it on charge over-night.

If you take the cycle life and multiply it by the amount of watt-hours you can get out of the battery, that is you calculate a lifetime watt-hour figure for the battery, this tends to be the same number regardless of how deeply you discharge them (although one would probably expect that linear correlation to break down at the extremes, which it probably does). I didn’t look into this scientifically, I just recall back in the day doing the math for one particular battery and realising it doesn’t really matter.

So there isn’t really a hard-and-fast rule of how deep you should go. You basically decide how often you want to replace them, and then you size them according to that. Many cheaper batteries have quite a respectable cycle-life at 20% DoD, but when you multiply it out to account for the fact that you are only getting 20% of the capacity on each cycle, the watt-hour lifetime isn’t that much better than it is at 50% DoD or 80% DoD.

The 50% DoD point is simply a good rule of thumb for solar power applications.

T105 or T105RE’s, forklift/golf cart batteries, had 4000 cycles at 20% DOD or 1600 cycles at 50% DOD.

The choice between working them or babying them, dying of “old age” having hardly been “used” vs them dying prematurely having been abused, that decision was made by one’s budget.

1 Like

That’s a good example. Technically that actually works out to the same thing.

Also People often forget that the battery won’t just die on the 4001 or 1601 cycle. It still has some life in it to carry on. The quoted cycles are the minimum guarantee…. Which again is based on projections.

Most of these new batteries haven’t been around for the claimed warranty period anyway so how does one know how they will perform in 5 or 10 years.

Personally, that’s the reason I would choose / recommend a well known brand. If I wanted to take a chance, I would rather go the diy route then I know my risks upfront.

1 Like


Technically, if you really think about it, the battery can do a 100% DoD discharge exactly once. The next time, it will only be able to handle 99.something% (of whatever the first cycle was), and from there it will deteriorate.

When you specify the battery, you basically decide how long you want it to last before it can no longer deliver the energy required of it.

For a car, that point is when it can no longer start the car on a cold morning. By that point the battery has less than 10% of its original capacity. Took it three years to get there though…

Now I have to add that for LiFePO4 batteries, it works a bit different. The Manufacturer picks an EOL (end of life) capacity, and the industry standard is 80%. When the battery has lost 20% of its capacity, it is considered EOL.

So standard testing of a battery will test it at 1C, 100% DoD, and they’d get 2500 cycles before the battery has lost 20% of its capacity.

But then the manufacturers employ some tricks. BlueNova for example tests at C/10, and they advertise 7000 cycles. Pylontech says recommended discharge is C/2 and EOL is at 60% remaining capacity. Doing this gets them 6000 cycles.

You see the apples and oranges stuff going on? But it doesn’t mean any of these batteries are bad batteries… they are just hard to compare because the marketing departments got involved :slight_smile:

And for every single battery mentioned, none of them were completely dead at the end. Even that car battery with the 10% left can still power an old radio casette player for a few weeks before it is really truly dead… and that is precisely what we used them for in high school!


When the brand names give warranties, wot, like 3500 cycles or 10 years, whatever comes first, it makes me wonder when that Company has not even been around for 10 years.

My ‘wonder’ stems from three considerations:

  1. Stats: The stats on the total companies actually lasting 10 years.
  2. Experience: 20-year warranty on Tenesol panels … then the Cpt factory closes.
  3. Experience: Duratherm geyser, 10-year warranty, the warranty was changed after 9.5 years I needed to claim.

So I did the sums on what a large capacity brand name bank would cost vs DIY bank.
The difference runs into thousands of rands.

So for me, and others, the “savings” one makes on a large capacity DIY bank, we see those thousands of rands not paid as our “warranty”, a “risk”, if you want, that we “bet” on not having to spend vs buying a brand name that has not been around for 10 years, that may close, or change the warranty. BOTH are “risks” today, and this Covid fiasco, it showed us all how fragile businesses really are.

This was my case. We used to have 4H30 load-shedding cycles and the batteries were bought to handle this (+5H) without any issues. Two years later and I struggled to get 4H30 out of them but I still managed with careful load management. Selling them at that point was my only option and then went the LiFePO4 route as I was also going Solar charging and not just Backup so I wanted daily cycles.

Two months later Murphy gave me a slap in the face as our 4H30 cycles changed to 2 x 2H30 cycles. Dammit!!! They could then lasted another two years probably but Hey, things happen as they happen.

Thinking back I initially bought the LA batteries with the idea that something ‘new’ might come along in a few years but never did. The SuperCap stuff surfaced and died but the LiFePO4 prices dropped at least. It will be a while before I get rid of my LiFePO4 battery - it is working hard every day and I couldn’t be happier.

1 Like

There is a learning curve.
I am very electricity savvy, and batteries are more akin to chemical engineering than electrical.
Especially Lead - Acid.
That said.
LA’s should be resigned to rural electric fences and the like. They do have a place.
LiFePo is the way to go if you’re serious.
That said.
Option 1: Expensive BMS inclusive options, most seem to work well, but market research/user references is the task.
Option 2: Diy options, which involve tinkering, but have the potential to work well at around half the price ( if you’re lucky). This option also involves market research and a lot of technical nous as well.

In hindsight, if I hadn’t of cut my teeth on LA’s, I would’ve probably of made some far more expensive mistakes going DIY lithium first. I would have loved to have saved that money, but the truth be known, if I hadn’t spent it then, I would have spent more later.
And it took time… years of insight.

I don’t like having protocols and proprietary stuff that just happens, that’s me.
That my problem with Pylontechs and their ilk. I want to be in control, I want to understand what is happening. I want to refine and tailor to my needs.

So a DIY bank will be cheaper money-wise for me, but that is only because I have such a background and I have paid the school fees. I don’t kid myself about that.

If I was advising extended family, I would recommend: Pay the extra money and get a proven off the shelf product like Pylontech or similar and have it professionally and reputedly installed. Get the after-sales service and realize it will cost.

If anyone thinks they can circumvent either route, (knowledge or cost) it will end in tears.
You will revisit this exercise from the beginning.
In other words, pay someone who has spent the school fees, or spend them yourself.

Of course, everyone knows better, that is human nature, so off you go…


Great points you make here! (Pity they aren’t in the ‘correct’ category but it’s the nature of these forums)
I concur fully with the time and energy it takes to get to grips with a technology. And it can be done!
The learning is never ending however.
I have recently tried to understand the NMH battery technology. It confounded me that one can’t measure the charge of a cell other than guesswork or when it starts heating up. That doesn’t mean it’s inferior and for safe charging and using in portable devices they have their place.
However the moving target of Li-ion batteries is a bigger challenge…
e.g. why do you opt for LiFePo? Why not Li-Po (or any other type)?

1 Like

Easy. LiFePo4 is safer. In a home you don’t need higher density energy as you tend to have more space. But safety is a premium.


The blog article title is a bit overstated though I think (seems a bit more like marketing hype).

With that said it is actually interesting to look at the actual original research article. If excluding things like possible criticism of sample size and general research methodology there is a specific graph that for me is noteworthy (I added the horizontal red line).

In this graph the data indicates by “end of Life” (80% original capacity) of the cells the total energy obtained from the cells, to allow for the difference in Ah capacity and voltage of the various cells as opposed to looking at just total cycles regardless what you get form those cycles.

NMC cells used in a 40-60% state of charge profile (discharged to not lower than 40% and not charged to higher than 60%) could likely outperform LiFePO4 cells - quite dramatically so at a high C rate.

If using cells in a 20-80% state of charge profile NMC and LiFePO4 cells ended up very similar - even at a high 3C rate (high C rate being one of the “my NMC battery is better than your LFP battery” claims).

If cycling cells in a 0-100% type profile LFP appears to have the upper hand and NMC fairs quite a bit worse (even at higher C rates and temperature).

BUT keep in mind the actual reasearchers’ statement:
“These differences suggest that lifetime prognostics based on a particular cell from a particular manufacturer cannot be broadly extrapolated, even to other cells with the same standard form factor, chemistry, and capacity. Subtle variations in materials, such as electrolyte composition, can substantially impact battery lifetime” (Yuliya et al, 2020).

Very nice link from the original research article btw: BatteryArchive.org

1 Like

That is quite interesting, but in all other cases the LFP cells performed as well as or better than the NMC cells. Also limiting yourself to only using 20% of the battery capacity in the 40-60% SOC profile is quite restrictive and you would need a battery that has 4 times the capacity than if you used 20-100% for example.
It’s also interesting that they didn’t do any tests from 20 or 30% to 100%, it was only 0-100% or 20-80%, but that’s not how most people use their batteries.

Interesting, but for my purposes I throw anything that is testing at more than 0.5C out the window. That is not how a solar system at home should be designed.

I don’t even run my batteries (7kWh) more than 250W (AC side) at the moment. Reason being that the days are short and I want to get them down to 40% over the span of a 14.5 hour “night”. So from 17:30ish until 08:00ish they drain constantly at a rate of 250W, using 3.625kWh, and if you take account of 10% DC to AC losses, ~4kWh = 57% DoD. This is only about 0.04C (if I didn’t mess up somewhere). I see no point in running them down to 40% in less time than the sun would take to come up again. I’m still just going to save the same amount of energy (actually a bit less due to heat caused by high current discharge).

The only time my batteries will be discharged faster (or potentially deeper) is during loadshedding, and only if that loadshedding is at night and we use the kitchen appliances, like the toaster or kettle. I’m keeping the 40% reserve in my discharge simply because I’m incredibly risk averse and don’t want to be caught with my pants down during loadshedding.

If I understand that data correctly, LFP is much more resistant to temperature. If you bring NMC into SA where it realistically could run for extended periods at 35 degrees, you seem to be fairly screwed.

I’ll give you one… :stuck_out_tongue:

The inverter itself needs around 30W or so to run. If you discharge at 250W, your overheard is more than 10%. If you rather discharge at 600W, your overhead (and therefore efficiency) becomes much better. For myself, I would rather run off the grid during peak hours, and then run a slightly higher rate from midnight to the next morning. The aim is still the same (hit 50% or so at sunrise), but the efficiency is a little better.

At least, that is what I did before I had a pool pump. Now I can just start the pool pump a little earlier to make space, and I have enough loads later in the day to consume the rest. So I don’t have to cycle to 50% anymore.

Haha point taken on the overhead, but won’t I have to run the inverter regardless or does it go into a type of standby mode after the battery hits min SoC? If so, then I have a bit of a conundrum… I hate hitting min SoC, because then the system wants to charge it up to min SoC + 3% again because using PV for consumption again… And that 3% is “wasted” because I’m never getting the power I purchased from the grid back…

Maybe the solution is just to purchase another battery and put the discharge higher while still avoiding hitting min SoC.