diy solar

diy solar

Downsides to LFP's? Sooo many questions and no bandwidth for videos.

Rednecktek

Solar Wizard
Joined
Sep 8, 2021
Messages
5,444
Location
On a boat usually.
So as a complete newbie to building batteries, as in I haven't even tried building a set yet, I keep seeing all these complications and "Gotcha's" when building a battery bank. I usually use FLA's and AGM's because my primary usage case is a camp that is A)unattended for 6-8 months at a time and B) gets well below freezing for weeks on end and C) I work for a living and don't have a lot of money to spend at once.

So I've read about things that can screw up your bank like:
Compression, what's that about?
Flakey or temperamental BMS's?
Getting the RIGHT BMS and how to spec for that?
Having to reset a battery if it gets below freezing?**
Thermal runaway?
Charge profiles (lots of good guides for that one at least)?
Cells going in and out of balance? Balancers in general?
Specialized battery chargers?
Top and bottom balancing?
If a battery does a low temp cutoff, doesn't that hurt the MPPT controller feeding it?

**The cold one is of particular concern as I worry that WHEN my bank gets below freezing, are the batteries just going to sit there being drained until I can get back and do a reset? Will I have to worry about resetting them every night and just not having power when it gets cold out in the evening because they shut themselves off? It seems like there's a lot of opportunity to kill a very expensive battery there.

So, what are the other "Gotcha's" that come with building LFP packs and outside of quitting my job to become a permanent resident at my camp can they be avoided?

My first project will most likely be a 24v pack for my garden shed as I'm USUALLY home for the few days it gets super cold at the main house, but it still gets below freezing for a week or so even there.

Thanks gang!

Oh, long text answers preferred. I'm out on the ocean and I only get dial-up or slower except for the 1 day every 2 weeks I can get 5Mb internet for a little while. I'd have done more YouTube scrubbing but even the forums take a few tries to load, and video is pretty much a 2x-a-month-thing.
 
Last edited:
So as a complete newbie to building batteries, as in I haven't even tried building a set yet, I keep seeing all these complications and "Gotcha's" when building a battery bank. I usually use FLA's and AGM's because my primary usage case is a camp that is A)unattended for 6-8 months at a time and B) gets well below freezing for weeks on end and C) I work for a living and don't have a lot of money to spend at once.
if you insulate the batteries and heat them in a way that is independent of the BMS and higher power than the outward heat flow, then there’s a decent chance of keeping LFP cells happy.
basically, some people talk about setting up a small circuit that would draw 5-15 watt from solar panels directly whenever the cells are <5C or something. this way battery get heated priority over charging. also, BMS low temperature cutoff should prevent significant charging under freezing.
but of course, lead acid might be better due to resilience even if the battery material gets cold. and it’s more massive per energy stored so insulating it is generally harder than an equivalent amount of LiFePO4 storage.
So I've read about things that can screw up your bank like:
Compression, what's that about?
For LiFePO4 cells, the internal changes going from fully discharged to fully charged and back, become more significant as get to 250Ah,275Ah,300Ah,310Ah capacities. in fact many of those capacity cells with aluminum shells, “breathe” about 0.5mm per cell from totally empty to totally full. with e.g. 100Ah cells this sort of compression is generally unnecessary. has to do with strength of walls, volume to surface area ratio. basically for the big cells applying about 10 pounds per square inch across the broad face seems to be enough to reduce the flextime on the cell edges which fatigue and break sometimes. like folding a bit of metal back and forth until it breaks due to fatigue. but each charge cycle is a bend. i’m gonna use m12 bolts and stainless steel plates
Flakey or temperamental BMS's?
daly=temperamental
JBD=works great for me in multiple installs
Getting the RIGHT BMS and how to spec for that?
daly gives an amp rating that you must divide to be safe. jbd says 100A can do 100A. idk much about other BMS. if you need really high amps, be sure to check out gigavac because they have sick contactors they can pass 200-400A continuous and only use 1-2W while active. eg MX12 contactor.
Having to reset a battery if it gets below freezing?**
Assuming the battery doesn’t drain all the way, the BMS low temperature protection should simply disable once the thermometer indicates a high enough temperature.
Thermal runaway?
Some may say differently but thermal runaway is one or two orders of magnitude less of a phenomenon with LFP. basically if they are gonna die it seems like bloating due to overcharge is a somewhat common demise. but that’s not thermal runaway. of course lots of heat can be made in a short but that’s also not quite thermal runaway. lifepo4 chemistry just doesn’t have the same breakaway coefficient of self heating as other chems.
Charge profiles (lots of good guides for that one at least)?
for LFP, keep every cell above 3.000V and below 3.600V and you’ll probably have a good time! i usually stay below 3.55 or so. everyone has their preferred trade off based on some variables such as how much capacity wanted and whether MPPT should supplement power and stuff. avoid “equalization charge” for lead acid with LFP since it’ll probably be too high voltage. i charge to 3.550V per cell with LFP so that has me at 14.2V for 4S. it has worked well for me.
Cells going in and out of balance? Balancers in general?
i rely on the slow resistor based balancer on JBD BMS. i configured it to activate above 3.400V per cell. since that’s where the “knee” is. so far so good with balance. it balances for a little while after initial build and then stable delta V. new cells mainly.
Specialized battery chargers?
float charge for LFP is not really needed but it’s useful for power assist to prevent a lot of cycling on battery. any charger that can have the voltage be configured should be enough.
Top and bottom balancing?
there’s some reasons to do either; if you expect to hit total empty regularly maybe bottom balance. if you expect to rarely go all to empty then maybe top balance. def check around a bit for this one.
If a battery does a low temp cutoff, doesn't that hurt the MPPT controller feeding it?
potentially. victron MPPT seem to not mind this. bluesolar has no bluetooth and that’s what my experience is with. some people even go as far as putting a relay on the PV input and totally disconnect it if temperature too low. maybe could have a dedicated circuit that would disconnect PV from MPPT when too cold and flip to other pole which is battery heater. then power go directly from PV to heating battery without needing to do full round trip so to speak.
**The cold one is of particular concern as I worry that WHEN my bank gets below freezing, are the batteries just going to sit there being drained until I can get back and do a reset? Will I have to worry about resetting them every night and just not having power when it gets cold out in the evening because they shut themselves off? It seems like there's a lot of opportunity to kill a very expensive battery there.
if they are insulated with a decent amount of XPS foam or insulation of choice, even a 5-10W heater should be enough. but of course experimenting is crucial here. i am designing to handle below freezing ambient temperatures but don’t expect to see them much. essentially the trick is to insulate to reduce outward heat flow from battery pack to outside world to a few watts and then it won’t need much to stay warm enough.
So, what are the other "Gotcha's" that come with building LFP packs and outside of quitting my job to become a permanent resident at my camp can they be avoided?
having a dedicated heater circuit that doesn’t rely on the battery being charged seems to be a nice trick to get around the low ambient temperature charging “gotcha”
My first project will most likely be a 24v pack for my garden shed as I'm USUALLY home for the few days it gets super cold at the main house, but it still gets below freezing for a week or so even there.

Thanks gang!

Oh, long text answers preferred. I'm out on the ocean and I only get dial-up or slower except for the 1 day every 2 weeks I can get 5Mb internet for a little while. I'd have done more YouTube scrubbing but even the forums take a few tries to load, and video is pretty much a 2x-a-month-thing.
good luck and i hope my rambling was helpful in some way :) please feel free to ask more questions. i’m still learning too.
 
-45c charge

Winston LiFeYPO4(LiFePO4) 3.2V 100Ah Cell​

 
if theres budget for LiFeYPO4 cells, they are indeed much better suited to operating in freezing conditions.
 
Reading your post I am a bit concerned about LiFePo4 batteries being left unattended for 7 months in freezing cold weather. What happens if your panels are covered in snow for two weeks? Given the nature of your situation and budget restraints I would stick to lead acid.
 

more reading, it’s not the most exciting but it’s an interesting case study to me anyways.

1634718365823.png
lotta numberz. definitely do not want to charge LFP at 0.5C when frozen. while away, presumably loads will be low. LFP has very low self discharge so the charge rate would also be low. for a 1kWh pack, maybe worth looking into the specifics of temperature, cell internal resistance dependent on temperature.

just tossing more info out into the wind :)

given that freezing and unattended, that is indeed a pretty common case where lead acid can shine!
 
it’s a bit derpy but yea!

I hope to ACHIEVE derpy someday. :)

Thanks gang, I'll be doing some reading at night. A lot of the numbers are above my head but hopefully I'll wrap around those soon enough.

So as for the insulation and heating, would a small heater under an aluminum plate at the bottom of a Yeti (or Coleman, because $$$) be an adequate enclosure? It would just be a matter of finding one that can physically fit the cells and BMS without being too cramped, right?

I'd rather go with a few larger cells than a bunch of smaller cells being a 24v system, so compression sounds like a must have. 24v = 8 cells which requires an 8s BMS, correct? If I wanted to upgrade in the future would I have to replace the 8s with a 16s?

Fortunately the only draw on the batteries is whatever power my topping MPPT controller and heaters would draw. I think this would be one of the few places that the "Load" terminals on my MPPT would actually be useful, no?

My topping system, used to just keep the batteries topped up when I'm gone, is going to be a 2s/2p of 100w panels about 8ft off the ground. The big array (200w 3s/3p) will be disconnected when I leave, along with the inverter/AIO, buck transformer, 12v bus, etc.

Fortunately it's only freezing for 2-3 months, but I deploy for 6+ months at a time to the ships and it's a rrreeeaaalllyyy long swim to go check on them.

So what kind of maintenance is required for these things? The FLA's I normally use I have to top off the water occasionally, but my plan to use AGM's some day was because it was all the cold resistance of FLA without the maintenance.

Thanks gang!
 
I dont think anyone mentioned this gotcha yet:

With all that expansion/contraction you can put a lot of stress on the cell terminals (bad). People get creative to get around this with compression/braces, playing with tolerances, flexible bus bars, etc.
 
One idea may be to consider making a large portable battery instead. People have videos of basically suitcase batteries, but you could stuff batteries in a lot of things. The 24v system that could be safely transported on wheels means it could be in your shed and at this other place. And then you don't worry about freezing for weeks on end. And you can use it where ever, and it fulfills some emergency needs as well.

It is amazing the number of devices that work on DC now. Walmart had a whole new section of Hart cordless tools since I was there three years ago... drills, but also fans, lights, air compressors, etc, etc. That portable battery will just keep getting more and more useful with the proper electronics.

Lifepo4 become much less complicated if you keep them at temperatures humans find comfortable. The common answer is to heat it, and you could probably get clever in how you do it, but that is a lot of charging and heating (wear and tear on expensive stuff) and you won't even be there? Seems like a waste.
 
Those Winstons sound like a possible solution. I'd like to see some real world data on them.

Big problem with heat is reliability of the heat system. I use propane but have an electric back up. All thermostatically controlled. But I only go away a week or two at a time and can have people stop in and check on stuff. If both fail, then cross fingers the BMS does it's job. I have my BMS cut out a bit higher then specs say is lowest temp for battery.

With potential for snow, you won't have reliable power coming in. Snow could sit on the panels for two months.

Portable battery sounds like a good solution.
 
Some may say differently but thermal runaway is one or two orders of magnitude less of a phenomenon with LFP. basically if they are gonna die it seems like bloating due to overcharge is a somewhat common demise. but that’s not thermal runaway. of course lots of heat can be made in a short but that’s also not quite thermal runaway. lifepo4 chemistry just doesn’t have the same breakaway coefficient of self heating as other chems.

----

potentially. victron MPPT seem to not mind this. bluesolar has no bluetooth and that’s what my experience is with. some people even go as far as putting a relay on the PV input and totally disconnect it if temperature too low. maybe could have a dedicated circuit that would disconnect PV from MPPT when too cold and flip to other pole which is battery heater. then power go directly from PV to heating battery without needing to do full round trip so to speak.

If OP would like to read more about, and see some charts showing the different chemistries and thermal runaway characteristics, can check out this page, which as you can see, thermal runaway is virtually non-issue with LFP:


LFP:
Thermal runaway270°C (518°F) Very safe battery even if fully charged

(compare it to the other cobalt chemistries hehe)...

-----

Also, regarding the statement on the Victron MPPT controllers.

Just an FYI since I bought a 250|100 for my RV. I learned that the Victron Smart Solar do have an internal temp sensor and have their own low-temp cutoff built in, but the internal temp probe setup only allows it to take a temperature sampling once per day.

If you want the Victron to do realtime monitoring of temp all hours of the day so the low-temp cutoff works all the time, you need to purchase an external temp monitor for it.

I bought one like this in link below, so it allows it to pull temp readings right off the cells (wherever you mount the sensor unit) in realtime, and then Victron can be a protection and cut off charging when low-temp trigger kicks in (could set Victron a couple degrees higher than the BMS low-temp cutoff if you wanted).

 
Last edited:
One idea may be to consider making a large portable battery instead.

Sadly that's not really an option as the family will some times head up while I'm overseas so the battery and system has to be in place when they get there.

The more I read about what it takes to keep LFP batteries happy the more I'm thinking I'm best off with some good 220Ah+ AGM golf cart batteries and just let it ride. The battery shelf is 2ft x 4ft so I can fit quite a few Ah of AGM on there.
 
I agree on the LFP in subzero temps for long times. Seems like a real accident waiting to happen and then those could just be junk when one thing happens.

Have you looked at LTO, Lithium Titanate? They are not cheap, but they are somewhat like an investment with 20,000-30,000 cycles (they say). I believe they completely get rid of this below freezing issue. They seem to handle quick charging and discharging well. It seems words apart from AGM or lead acid...

I'd maybe consider costs of Lifepo4 cells with all the necessary heating and insulation, and compare THAT to the cost of Lithium titanate. Consider they may likely be working the rest of our lives. If you just do a "cost per amp hour" calculation vs LFP or AGM you'll throw up, but if you properly consider how each would function, perhaps paying more of the actual cells is the best answer. Not sure. I do know more and more LTO pouch batteries are showing up on places like ebay. Some Toshiba cells have 10 year warranties, so they must have good evidence they last. Need to research them a bit more, as I need a new car battery for my tesla model 3 and it kills me to pay $100+ for a lead acid battery that will constantly charge and die in two years like my current one... And LFP options are $450 and I can't imagine them working well below zero. There is a relatively local guy selling 2.3v LTO 10ah cells for $11. Not nearly as expensive as the cylindrical cells. They are $5-6 in bulk from alibaba.

I certainly could be glossing over some (a lot?) negatives. I thought early on people were thinking the charge and discharge efficiency was shockingly low (~70%) but then others have mentioned upper 90s. It has lower energy density. Different bms, but don't seem super expensive.
 
If OP would like to read more about, and see some charts showing the different chemistries and thermal runaway characteristics, can check out this page, which as you can see, thermal runaway is virtually non-issue with LFP:


LFP:
Thermal runaway270°C (518°F) Very safe battery even if fully charged

(compare it to the other cobalt chemistries hehe)...

-----

Also, regarding the statement on the Victron MPPT controllers.

Just an FYI since I bought a 250|100 for my RV. I learned that the Victron Smart Solar do have an internal temp sensor and have their own low-temp cutoff built in, but the internal temp probe setup only allows it to take a temperature sampling once per day.

If you want the Victron to do realtime monitoring of temp all hours of the day so the low-temp cutoff works all the time, you need to purchase an external temp monitor for it.

I bought one like this in link below, so it allows it to pull temp readings right off the cells (wherever you mount the sensor unit) in realtime, and then Victron can be a protection and cut off charging when low-temp trigger kicks in (could set Victron a couple degrees higher than the BMS low-temp cutoff if you wanted).

Can confirm I use this inside my insulated box right up against the cells to get more appropriate temp readings.
 
.... if you need really high amps, be sure to check out gigavac because they have sick contactors they can pass 200-400A continuous and only use 1-2W while active. eg MX12 contactor....
Gigavac also makes latching relays. I have, IIRC, a ~250-300amp, GXL14, that was $5 on BH. The only small catch is that you need to build a "switchboard" for it, a little flip-flop circuit does the trick, you need about $2 of components from eBay to make one. I would imagine they have the prefabricated PCBs to control latching relays as well.

It uses almost no power to switch, so little that I can't actually measure, in watts, how much power the relay uses to reverse. IDK why other BMS's don't use latching relays too.
 
Gigavac also makes latching relays. I have, IIRC, a ~250-300amp, GXL14, that was $5 on BH. The only small catch is that you need to build a "switchboard" for it, a little flip-flop circuit does the trick, you need about $2 of components from eBay to make one. I would imagine they have the prefabricated PCBs to control latching relays as well.

It uses almost no power to switch, so little that I can't actually measure, in watts, how much power the relay uses to reverse. IDK why other BMS's don't use latching relays too.
Because a mechanically latching relay needs power to turn off.

This can potentially be a point of failure if for some reason that power isn't available.

It's *probably* a minor concern but the programming would need to be changed as well as the bms design to allow it to still pull that power from the battery rather than just turning off an output along with most everything else.

You also have an increased risk that it fails in the on position, where a regular relay just opens up under springs when the coil turns off.
 
Because a mechanically latching relay needs power to turn off.

This can potentially be a point of failure if for some reason that power isn't available.

It's *probably* a minor concern but the programming would need to be changed as well as the bms design to allow it to still pull that power from the battery rather than just turning off an output along with most everything else.
I see why that is an issue. Even if the power for the coil was pulled directly from the cells, a blown fuse or short on the positive input wire for the contactor would lock it open. Sounds good in theory though.
 
Back
Top