Getting ready to dive into LiFePO4

Announcement - Help us fight the BOTS! Please report all spam including stuff in your inbox!

Dr. Dickie

Member
Joined
Sep 23, 2020
Messages
218
I am preparing to get into lithium iron phosphate (which was my ultimate goal). I will probably just put together a 12V system to replace my lead-acid system that runs my HAM radio, a small back-up fridge, and the solar garden pump. More info than you need.
My question is: I see many people are apposed to compression of the prismatic cells, even when the manufacture calls for compression. Now I don't plan to squeeze the heck of the cells, but what is the downside to setting the cells up in a manner where they cannot swell easily (or significantly, I guess). Just wondering so I can make my own decision about whether to or how to compress.
Thanks
 

OffGridInTheCity

Active member
Joined
Dec 15, 2018
Messages
1,746
Just wondering so I can make my own decision about whether to or how to compress.
@LithiumSolar just did a youtube discussing this...
View: https://youtu.be/uZe6BDvWPfI

Sometimes (reading between the lines) I get the impression that modest DOD should not create very much bulge and if you have cells with noticeable bulge the are degraded. Thicker cased cells such as CALB may make the decision for you - e.g. no action needed. But some manufactures say you can extend cycle life with some compression on thinner cased cells.

The algorithm of how to approach it is still fuzzy to me. For example
1) If new + specs say compression will extend life - then yes
2) If new + specs don't say + you plan modest DOD - then less compression? or none?
3) If used/bulging - then cells are not new - does compression help at this point?
4) Does too much compression lead to shorts?
etc....
 
Last edited:

jdeadman

Member
Joined
Mar 28, 2017
Messages
952
I agree with @OffGridInTheCity and I would say that you will want to fully test the cells and see how much they grow during a standard discharge. Clamping down I do disagree with but adding a light pressure when unbuldged but if they start to grow there will be some resistance to that growth.

Also I do agree with @lithiumsolar that the cells do need to be isolated when in series
 

lucasd05

New member
Joined
Oct 1, 2021
Messages
14
Not prismatic cells LiFePO4 cells but A123 20Ah LiFePO4 pouch cells from BCH seem to be the best value of cells that are already in the USA. Coming in at ~$86 per kWh + shipping. https://secondlifestorage.com/index.php?threads/a123-lifepo4-pouch-cells-20ah-compacts.11280/ . These would require compression. https://www.batteryclearinghouse.co...uch-cell-3-2v-8ah?_pos=1&_sid=5fb30e195&_ss=r
later floyd
These really do look like a good value, even including shipping. A question, however - do folks recommend cell level fusing with these when building banks that employ parallel cells? The commercial modules I've seen that are constructed from tabbed pouches don't seem to employ this safety mechanism (please tell me if I'm misunderstanding).
Example: https://batteryhookup.com/products/36v-48ah-1776wh-spim08hp-36v-power-module
 

floydR

Member
Joined
Aug 23, 2017
Messages
1,314
Not a real easy way to make cell level fusing work with pouch cells that I have seen. Being LiFePo4 pouch cells which tend to be safer than LiCo/NMC cells. Probably not needed.
later floyd
 

jdeadman

Member
Joined
Mar 28, 2017
Messages
952
This is my biggest issue with the Large Format Cells. But with my power being in use for over 2 years and I have only ever had one cell fuse break but that was due to handling. Providing you test the cells and monitor them correctly there should be very little issues
 

Redpacket

Member
Joined
Feb 28, 2018
Messages
1,309
@LithiumSolar just did a youtube discussing this...
View: https://youtu.be/uZe6BDvWPfI

Sometimes (reading between the lines) I get the impression that modest DOD should not create very much bulge and if you have cells with noticeable bulge the are degraded. Thicker cased cells such as CALB may make the decision for you - e.g. no action needed. But some manufactures say you can extend cycle life with some compression on thinner cased cells.

The algorithm of how to approach it is still fuzzy to me. For example
1) If new + specs say compression will extend life - then yes
2) If new + specs don't say + you plan modest DOD - then less compression? or none?
3) If used/bulging - then cells are not new - does compression help at this point?
4) Does too much compression lead to shorts?
etc....
Yeah I agree, visible bulge = damaged cells.
My thoughts are cells bulging is caused by gassing when the cell voltage is taken off the "flat" part of the curve & possibly also at high charge &/or discharge rates. I'm pretty sure charging to the manufacturers 3.65V is a bad idea long term.
When I first built my system, I found my LiFePo4 banks would quickly go out of balance if I charged them above about 3.45V/cell. For me this indicates I've started to climb the "J" curve at the upper end. So I set the CV point of the chargers to 3.41V/cell or so & the packs have stayed nicely balanced since. lower end I've set to 2.9V cut off, again bottom end J curve downwards starts about here.
My banks are 480AHr and only see max about 80A charge & occasional <5mins of 100A discharge, mostly 10-20A range.

I don't believe you need compression on cells at all if operated as below:
a) not charged above 3.45V/cell
b) not charged/discharged at high rates
c) DoD in the 10% to 95-100% range
d) top of charge by current tail off at CV 3.45V max
e) temperature range 0C to 40C.
f) 3mm foam between cells to allow normal expansion/contraction of chemistry & by temp
 

OffGridInTheCity

Active member
Joined
Dec 15, 2018
Messages
1,746
When I first built my system, I found my LiFePo4 banks would quickly go out of balance if I charged them above about 3.45V/cell. For me this indicates I've started to climb the "J" curve at the upper end. So I set the CV point of the chargers to 3.41V/cell or so & the packs have stayed nicely balanced since. lower end I've set to 2.9V cut off, again bottom end J curve downwards starts about here.
I read about the 'flat curve' so much, I'm curious about the equipment/methods you use to do the max (3.45v/cell) and min (2.9v cutt-off). I presume charge controller for max but what do you do for min / was it sensitive enough voltage wise or what steps did you do to calibrate this?

Bottom line - I would presume you would agree that one can manage a LifePo4 powerwall by voltage, I keep asking this because I keep reading that it's not a good approach and I wonder if I'm missing anything - but you know, one has to make their own decisions for DIY :)
 
Last edited:

Dr. Dickie

Member
Joined
Sep 23, 2020
Messages
218
Yeah I agree, visible bulge = damaged cells.

f) 3mm foam between cells to allow normal expansion/contraction of chemistry & by temp
Yeah, large bulge means either grade B cells or to high of charging is what I have seen (on the tube)..
With the foam for expansion, you then have to account for stress strain on bus bars if they are not flexiable (which for me, not a problem, the way LiFePO4 are typically connected).
Thanks for all the replies
 

Redpacket

Member
Joined
Feb 28, 2018
Messages
1,309
I read about the 'flat curve' so much, I'm curious about the equipment/methods you use to do the max (3.45v/cell) and min (2.9v cutt-off). I presume charge controller for max but what do you do for min / was it sensitive enough voltage wise or what steps did you do to calibrate this?

Bottom line - I would presume you would agree that one can manage a LifePo4 powerwall by voltage, I keep asking this because I keep reading that it's not a good approach and I wonder if I'm missing anything - but you know, one has to make their own decisions for DIY :)
It seems best practise is to manage batteries by SoC (with Coulomb counting) & cell voltage.

My gear is mostly Victron, 1x Multiplus Inverter (where I set the cutoff), 3x Blue Solar 150/60 chargers, each set to 54.4V CV point, & a BMV712 to monitor SoC with shunt (does the Coulomb SoC tracking) . The battery bank is 4 strings of 16 LiFePo4s 120AHr each. I have a Batrium system with 64 mons.
Overall the battery voltage does move on a slope, just not as much as Li-Ion (18650's) more like 50 - 54.4V for DoD 10ish to 100%.

In practice when I know the bank is really low <10%, I watch & cut it over to grid supply before any cells voltages start dropping (<2.9V).
If we're not home, even on heavy cloud days, the system seems to get enough from the panels to just cover our "base load".

Unfortunately, I can't trust the batrium system to trigger cutoff due to a random "low voltage" problem when balancing (house = disco not allowed here!). Others have seen this issue too & I haven't managed to fix it here - have to get around to trying some wiring changes....
 

Redpacket

Member
Joined
Feb 28, 2018
Messages
1,309
Yeah, large bulge means either grade B cells or to high of charging is what I have seen (on the tube)..
With the foam for expansion, you then have to account for stress strain on bus bars if they are not flexiable (which for me, not a problem, the way LiFePO4 are typically connected).
Thanks for all the replies
I saw that too, but thought that terminals are centered on cells not at the edges so how would the "expansion" add up like mentioned to press on the links when you have the foam? Agree a minor pressure change on foam but still no serious issue.
Ie each cell can get thicker/thinner by say 1 mm but who cares, the distance terminal to terminal across the whole string doesn't change....
 
Last edited:

AntaresAdroit

New member
Joined
Mar 13, 2021
Messages
4
I found an A123 guide from 2014 that has a lot of useful information including guidelines for cell compression. Here's what they say about compression:
1633802895817.png


They also have some details about cell expansion (30.4kPa is about 4.4psi):

1633803521758.png



I'm not sure how widely circulated the document is, so I've also attached a pdf with the full document.

I got a shipment of about 7kWh of those Battery Clearinghouse cells last night and I'm working on a design for a 48V pack (probably with two 24V modules, though I might split it up a little bit more).
 

Attachments

  • a123 Battery Pack Design Guide.pdf
    1.9 MB · Views: 35

floydR

Member
Joined
Aug 23, 2017
Messages
1,314
The 20Ah cells?
From what i have read the cells are testing right at 20 Ah.
am debating on getting some myself.

later floyd
ps thanks for the A123 battery pack designpdf
 
Last edited:

floydR

Member
Joined
Aug 23, 2017
Messages
1,314
The PDF was very interesting reading. I ordered a few cells.
Now I need an ultrasonic welder. :) like that will ever happen. maybe an ultrasonic plastic welder could be adapted to weld .0005mm copper foil and aluminum foil using a multiple layer approach.
Later floyd
 

Redpacket

Member
Joined
Feb 28, 2018
Messages
1,309
+1 the pdf is a comprehensive & interesting read.
The A123 brand seems to be focused on the need for compression - could this be because
a) their cells need it more (tend to produce more gas)?
b) they anticipate more "cell abuse" in their target market, hence a)?
c) the specs listed in the pdf do seem to cater to higher current & temp range scenarios.
Re "cell abuse" situations - in home energy storage systems hopefully we operate our cells fairly differently, eg lower charge/discharge rates, less DoD range + controlled max & min cell voltages.
 

jgilly69

New member
Joined
Aug 11, 2019
Messages
15
BCH has some other larger A123 20ah cells coming in the next few weeks These other ones are a little thinner.
 

floydR

Member
Joined
Aug 23, 2017
Messages
1,314
I ordered some of the 20Ah cells. I spent this months hobby budget and probably next months already.
Have you figured out how you will connect the cells?
Later floyd
 

AntaresAdroit

New member
Joined
Mar 13, 2021
Messages
4
I ordered some of the 20Ah cells. I spent this months hobby budget and probably next months already.
Have you figured out how you will connect the cells?
Later floyd
I feel like we're hijacking the thread a little, but I think I'm going to use some edge card connectors for temporarily connecting the cells for testing and I'll probably make some aluminum bus bars with screws to clamp the battery tabs to make a pack. I got the last of these edge card connectors, but the battery tabs are a millimeter or two too wide to fit these specific ones (a 22 pin version would be fine). For the similar "name brand" ones on DigiKey, a single pin is rated at 3A, so 5 or 6A over 20 pins seems fine. Since I already have them, I'll probably just modify the connectors a bit. I was planning on having to make some shims anyways (the tabs are only about 0.5mm thick, but the slot on the edge connector cards need something about 1-1.5mm thick to make decent contact).

BTW, I suspect the A123 cells might be the ultraphosphate ones rather than the nanophosphate ones that pdf I posted is for, but the clamping information seems like it might be similar enough.
 
Top