LiFeP04 Batteries for Solar & BMS

[Sunking]

PN not certain what cell phone manufactures do, but some of the EV manufactures underrate battery capacity. Tesla is one of them and Chevy Volt. Not certain on the Tesla other than I know they limit charge and discharge capacity. I know for fact Chevy Volt only charges to 80% SOC, and Cut-Off at 20% SOC so only 60% of the actual capacity is accessible. That can certainly be applied to any application if you can afford it. Example if you want 100 AH per cycle, then use 160 AH cells. What I cannot tell you with any certainty because I have not ran a cost model is if you can justify the extra expense of larger cells to get longer cycle life. We always have to remember lithium and all batteries have a self life that over laps cycle life.

[PNjunction]

LL what PN is saying is the less usable range you use, the less important a BMS becomes...

< ..snip .. >

Bottom line or take away here is if you want to go to or near 100% SOC you have to use a BMS to TOP BALANCE and make sacrifices.

I couldn't have said it better! I get tongue-tied in all the logic.

Evaluate your application and choose what works best for you. Some ways are better than others, but may be impractical from your own maintenance capabilities, involve more or less risk etc.

As for cycle life and staying away from the top and bottom knees - again evaluate your application and environment to see how far you want to take it.

In *general* terms, the closest example is your own lithium cellphone battery, which of course is NOT lifepo4, but serves a useful demonstration.

They are charged and shipped somewhere around 50%. Aside from shipping regulations, keeping a lithium based battery fully charged aggravates "lithium plating" secondary reactions. Likewise, fully discharged they chemically decompose quickly. The hope here is that during retail storage stored at 50% or so, you'll get to them before reaching decomposition, and not being shipped and stored fully charged, lithium plating won't be occuring.

Normally you'll only receive around 350 or so "full cycles" of total charge and total discharge. If you operate them conservatively, then they will last far longer because statistically, you are cutting down on the secondary reactions by staying close to the middle, but yes there is also an aging factor like there is with all batteries. The additional life is not just from shallow-cycling, but avoiding secondary reactions the closer you get to the top or bottom over time. Ideally, never fully charge your cellphone battery. But will that be practical? Only you will know the line between useful and obsessive.

I know this is a blatant oversimplification and really don't want to get into cellphone batteries. However it may serve as a convenient example close to nearly everybody. If you are seeking the max cycle life of any lithium-based battery chemistry, being conservative will help get you there if your application allows such a luxury.

[Legit]

Many thanks SK, will try to get more info on the system.

[Sunking]

Would you agree that the storage of energy is a big cost of a Solar System?

Absolutely, plus inefficiencies and utilization.

- on this I would value your thoughts on the link below, not are what to make of this or how the setup works.

OK it is a Battery Management System or BMS and guessing it is a very expensive system. I take issue when he claims 8 year payback. For one the batteries will not likely last 8 years. Cannot really offer much as the article does not offer any meaningful details. All they give you is it lithium batteries using a BMS.

[Legit]

Hello SK, have been checking out a few of your posts, really appreciate the time, effort and knowledge you give - I would like to go off the grid and have noted and agree with your comments on the costs involved - would however like do the sums on this as our electricity costs are high due to the politics, in short the Have's support the Have Not's in our country, this senario does warrant a peep at going off the grid - our government does not entertain the grid-tie option at present as they stand to loose revenue.

Would you agree that the storage of energy is a big cost of a Solar System? - on this I would value your thoughts on the link below, not are what to make of this or how the setup works.

http://www.pri.org/stories/2015-04-06/how-do-you-catch-sun-make-electricity-night-german-inventor-has-answer

[Living Large]

LL what PN is saying is the less usable range you use, the less important a BMS becomes.

Duh! I missed that totally. I knew there was a big picture I was missing, but I couldn't see it. Thanks!!!

I don't think about the factors such as balancing and LVD we have discussed, because in my mind we resolved how to address them for an off-grid system. Easy for someone with no hardware to do. I appreciate these are very real concerns and affect users of LFP.

[Sunking]

LL what PN is saying is the less usable range you use, the less important a BMS becomes.

Example say you have a 100 AH cell and only want to operate on 50% of the capacity or 50 AH. So you only charge to 75% SOC and discharge to 25% SOC. At that narrow level you don't really need to worry about individual cell voltages. You can operate like FLA working only with total pack voltages So for a 16S aka 48 volt battery you only charge the pack up to 53.6 volts vs 3.65 volts x 16 = 58.4 volts at 100%SOC. Even if the cells are out of balance you are never going to push any cell to 3.65 volts with 3.35 volt average at 75% SOC. Likewise you are never going to discharge a cell down to 2.5 volts if your low voltage cutoff is set to 49 volts 25% SOC or 3.1 volts per cell. Using this method you would want to BOTTOM BALANCE initially and never have to worry about anything. All you gotta do is set the charger to 53.6 volts and LVD to 49 volts.

However the same can be said is you want to charge to 90%, and LVD @ 10% as long as you keep charge current to 1C or less and Bottom Balance initially. Set the charger to 56.8 volts, and LVD to 46.4 volts. This is the method 60% of the DIY EV guys operate. They BOTTOM BALANCE and use no BMS or any battery monitoring other than setting the charger and letting the motor controller operate the LVD. The other 40% use a lot of expensive BMS and have the most reported failures.

As for cycle life if you charge to 100% and discharge to 20% you get around 2000 cycles in a lab. If you limit to 90% or less at charge, and do not go below 10% you get around 3000 cycles. So nothing gained operating 10/90 or 25/75.

One more note here if you operate 10/90 the LFP charge curve is nearly linear under 1C load and less of 3.4 volts @ 90%, and 2.9 volts @ 10%. Makes it very easy to estimate SOC based on voltage.

Bottom line or take away here is if you want to go to or near 100% SOC you have to use a BMS to TOP BALANCE and make sacrifices.

• It is going to hit your wallet because BMS cost money and can get expensive. At a minimum for a passive BMS is $16/cell that does not work very good using Vampire Bleeder Boards, nor gives you any useful iNFORMATION other than an idiot light.
• You will give up to 33% cycle life.
• Run a much higher risk of destroying very expensive LFP cell or cells

[Living Large]

If you are running <1C, preferably <.5C, and are not going into the extreme ends of charge / discharge, they tend to stay in the *state* of whatever balance you last left them in. So it is important to get them close, but if you aren't going to extremes in either current or SOC, you may never see a problem.

I think we forget why we (or at least for me) chose lifepo4 in the first place. Cycle life wasn't the main criteria.

Unlike EV or RC, where one is usually concerned with total miles / flight per pack, and the associated issues of weight, with solar, we have the luxury of running in a sliding-window of SOC.

Just like Pb, I recommend designing capacity for no more than 50% discharge, but for different reasons.

With lifepo4, you are not penalized for partial-state-of-charge operations, and are actually rewarded for it. Makes it a 'natch for solar. Since we are not dealing with Peukert to any extent, it also makes calculating your capacity needs much easier. I have found that simply de-rating the stated capacity to only 80%, and with no Peukert to deal with, I can get very close mentally with static loads.

But what if you over-estimate your bank capacity? While nobody promotes "growing into" a battery capacity, if you do make a mistake and make it larger than needed, the hit in your wallet will be compensated somewhat by running narrow PSOC with extended cycle life beyond the norm. Yes, nobody I know has taken these things beyond 10 years in this mode, but at least that's what the manufacturer's say. Take that with a grain of salt, but chemically that is supposed to be beneficial by staying away from the top and bottom edges of capacity.

Example: if you design for no more than 50% discharge, you can easily run 100/50, 90/40, 80/30, or 70/20 SOC windows and are never penalized for doing so. In addition, NO corrective topping-up or in the case of flooded, EQ is necessary. You may have a genny on hand, but if you do your homework right, even with sliding windows of SOC, as long as you have ample capacity to run your loads, the fuel in your genny might go stale before ever having to use it. Of course voltage is a poor indicator of capacity in the flat part of the curve, so one will want to count / calculate coulombs.

In the EV / RC world, nobody in their right mind would over-size their bank to take advantage of this PSOC - incurring an unnecessary weight and size penalty to stay away from the edges. We have that luxury be it intended or a goof.

What seems to get overlooked are the *practical* aspect advantages of running lifepo4, and how simplifying operations by choosing decent capacity with sliding windows of psoc may alleviate the daily worry of pin-point precision balancing needed with EV / RC. If you design it right, you'll never reach top or bottom regularly and increase cell life further.

Other than the hit on your wallet, over-purchasing lifepo4 won't take a toll on the cells, unlike Pb where you MUST reach 100% SOC on at least a somewhat regular basis. In addition, leaving your lifepo4 in a partially discharged state, provided you don't have any parasitic loads, won't hurt you either.

So balancing / bms discussions are fine - but don't forget some of the practical reasons that differ a LOT from EV/RC. Before spending a dime on a huge permanent setup, get a small bank (like a simple 12v 4S battery of 40ah or so) and witness it yourself. This is no different than getting a "learner" battery for first-time Pb solar operations.

I am a little confused as to your main point. I'll explain my thinking for my application, and perhaps that will allow you to tailor a followup.

I chose LFP because it is relatively lightweight, relatively small, is sealed, charges relatively quickly. The cycle life is good in comparison to others, but honestly cost is not my main concern so I really haven't put a lot of thought into lifetime cost.

I was going to undersize my bank in the sense I was targeting using 10/90 over three days of autonomy. Why 80%? I'll admit I didn't want to spend more than I had to up front on capacity, but am I taking more of a hit on cycles than you think I should? Are you saying that your belief is a 50% window in the long run will be more economical due to the increase in cycles the increase in minimum SOC results in?

[PNjunction]

What I have been hearing and seeing in videos of users is it tends to stay in balance for extended periods of time (months).

If you are running <1C, preferably <.5C, and are not going into the extreme ends of charge / discharge, they tend to stay in the *state* of whatever balance you last left them in. So it is important to get them close, but if you aren't going to extremes in either current or SOC, you may never see a problem.

I think we forget why we (or at least for me) chose lifepo4 in the first place. Cycle life wasn't the main criteria.

Unlike EV or RC, where one is usually concerned with total miles / flight per pack, and the associated issues of weight, with solar, we have the luxury of running in a sliding-window of SOC.

Just like Pb, I recommend designing capacity for no more than 50% discharge, but for different reasons.

With lifepo4, you are not penalized for partial-state-of-charge operations, and are actually rewarded for it. Makes it a 'natch for solar. Since we are not dealing with Peukert to any extent, it also makes calculating your capacity needs much easier. I have found that simply de-rating the stated capacity to only 80%, and with no Peukert to deal with, I can get very close mentally with static loads.

But what if you over-estimate your bank capacity? While nobody promotes "growing into" a battery capacity, if you do make a mistake and make it larger than needed, the hit in your wallet will be compensated somewhat by running narrow PSOC with extended cycle life beyond the norm. Yes, nobody I know has taken these things beyond 10 years in this mode, but at least that's what the manufacturer's say. Take that with a grain of salt, but chemically that is supposed to be beneficial by staying away from the top and bottom edges of capacity.

Example: if you design for no more than 50% discharge, you can easily run 100/50, 90/40, 80/30, or 70/20 SOC windows and are never penalized for doing so. In addition, NO corrective topping-up or in the case of flooded, EQ is necessary. You may have a genny on hand, but if you do your homework right, even with sliding windows of SOC, as long as you have ample capacity to run your loads, the fuel in your genny might go stale before ever having to use it. Of course voltage is a poor indicator of capacity in the flat part of the curve, so one will want to count / calculate coulombs.

In the EV / RC world, nobody in their right mind would over-size their bank to take advantage of this PSOC - incurring an unnecessary weight and size penalty to stay away from the edges. We have that luxury be it intended or a goof.

What seems to get overlooked are the *practical* aspect advantages of running lifepo4, and how simplifying operations by choosing decent capacity with sliding windows of psoc may alleviate the daily worry of pin-point precision balancing needed with EV / RC. If you design it right, you'll never reach top or bottom regularly and increase cell life further.

Other than the hit on your wallet, over-purchasing lifepo4 won't take a toll on the cells, unlike Pb where you MUST reach 100% SOC on at least a somewhat regular basis. In addition, leaving your lifepo4 in a partially discharged state, provided you don't have any parasitic loads, won't hurt you either.

So balancing / bms discussions are fine - but don't forget some of the practical reasons that differ a LOT from EV/RC. Before spending a dime on a huge permanent setup, get a small bank (like a simple 12v 4S battery of 40ah or so) and witness it yourself. This is no different than getting a "learner" battery for first-time Pb solar operations.

[codo1982]

Alliance Renewable Energy, Inc

Please help, I'm looking to buy some LifeP04 Batteries and I found this place. Is it any good? Is any of you buy LifeP04 Batteries from here? Any suggestion where to buy them. Thank You for your help.

[Living Large]

Getting back to the topic of this thread, I came across a really informative write up on an LiFePo4 based system. This is by a fellow that uses his system in a marine environment, but there are many similarities to using LFP as a house bank. He gives really good info about precautions, the bms, cell balancing, etc... Really great writeup!!! There are 3 pages.

http://www.pbase.com/mainecruising/l...n_boats&page=1

I quickly browsed that article, and there is a lot there. Thanks for the link.

True confessions - I have to admit I have to go back to before this thread left the LFP and BMS discussion - Dereck was helping me to understand how to integrate a BMS into my proposed system, and I didn't read his last few contributions several days ago. This article and the previous discussion here should keep me busy...

[northerner]

Getting back to the topic of this thread, I came across a really informative write up on an LiFePo4 based system. This is by a fellow that uses his system in a marine environment, but there are many similarities to using LFP as a house bank. He gives really good info about precautions, the bms, cell balancing, etc... Really great writeup!!! There are 3 pages.

http://www.pbase.com/mainecruising/lifepo4_on_boats&page=1

[Living Large]

You use 700ah a day, what sort of equipment would use that much energy daily. {snip}

No, I didn't mean that, if that is how it was interpreted. A 48 volt 700 Ah battery bank, with 2 to 3 days of autonomy and 50% to 60 DOD. So, 700 Ah * 0.6/3, Or about 140 Ah a day (at 48V).

This was for an FLA bank, but we have strayed significantly from LFP and BMS and for the sake of comparing 12V, 24V and 48V systems, I don't believe it matters too much.

[Sunking]

You use 700ah a day, what sort of equipment would use that much energy daily.

That is completely meaningless information. Amp hours do not mean squat unless a voltage is specified. Watt hours is the measure of electric power consumed, not Amp Hours. You have not made that connection or understand that. 700 AH at 12 volts = 8400 watt hours. 175 Amp Hours at 48 volt = the same 8400 watt hours.

Maybe this will help. Let's say you go out and buy 4 12 volt 100 AH batteries. You have 3 possible ways you can configure them of 12, 24, and 48 volts

4 in parallel is 12 volts @ 400 AH = 4800 watt hours. We connect a 1000 watt load and it will draw 84 amps
2 in series in parallel with 2 in series (2 x 2) gives us 24 volts @ 200 AH = 4800 watt hours. We connect a 1000 watt load and it will draw 42 amps
All in series is 48 volts @ 100 AH = 4800 watt hours. We connect a 1000 watt load and it will draw 21 amps.

The 48 volt configuration is much less expensive to install, and far safer than 12 volts. There is no way you can win the argument using lower voltage is less expensive or safer. Sorry but if you think otherwise you are delusional.

[Mike90250]

It does not matter what equipment is running, if a persons household load is 15Kwh, then a 30Kwh battery (planned usage capacity 20% - 85%) is needed, at minimum, to reduce generator runtime.

Let's compute in Watt Hours, since everyone is dealing with different voltages (12, 24, 48) just stating amp hours is confusing.

Example, I average about 7.5 Kwh usage daily in winter. To be able to skip a charging day, and NOT be totally flat after 48 hours, I need to have 3 days (72 hours) 23Kwh of USEABLE battery. 20-85% middle of the battery capacity. 35% un-useable because of damage risk to cells. So that would require I buy a 31Kwh battery so I have time to start the generator on day #3 of cloudy weather. This would be in a 48V based system, so that would need 645ah of battery.
here's a 200ah cell that looks reasonable.
batteryspace.com/lifepo4-prismatic-module-3-2v-200-ah-10c-rate-640-wh-un38-3-passed-dgr.aspx $370

I'd need 3 parallel strings, to get to 600ah, and that would be 48 cells, @ $24K With >2,000 cycles, 1 yr warranty. So I get maybe 3,000 cycles out of them, 8 years. $3K yearly battery cost.

So, I'd need to tweak the bottom and top %'s a little bit to get really accurate, but I think it's close to start with. And if I'm way off, tell me, I'd like to have some realistic numbers to plug into here.