270ah DIY LiFePO4 build

[Kelpie]

Kelpie, doesn't work, or rather you need a B2B to do that.
what I've done is I still have the halfdead Trojans online and in parallel to the lifepo. So if something goes wrong and lifepo disconnect, I still have batteries online to take the load (say alternator) and everything's fine.
I wouldn't dare have service bank lifepo only tbh.

Yes I would have needed to have a B2B in there, I wasn't talking about going full hybrid.

It was an idea predicated on the idea that lithium is much more expensive per Ah than lead acid. Use cheap lead acid for c.80% of the total capacity, with a small LFP alongside. Discharge the LFP overnight to extend the life of the lead acid.

I never went ahead because my assumptions turned out to be way off. I built my LFP for less than the equivalent lead acid would have cost. When you also factor in the lifespan, LFP is an absolute no brainer.

 

[geem]

Being able to run power tools and a 1kw wet and dry vac was a game changer for me when working aboard. We have 1.2kw of PV and look a bit like a small floating power station

@geem if you build an experimental LFP pack you'd be in a good position to try out a little idea I had: charge the LFP rapidly during the day, then use it to trickle charge your lead acid batteries overnight to get them up to 100%, prolonging their lives.

I investigated this plan but when I came to decide how to split the Ah between LFP and lead acid, I found I could get LFP cheaper anyway so just kept it simple. I think prices have risen in the two years since then, though.

Our domestic bank are normally fully charged by midday. The engine batteries (2x110Ah @24v) have their own 40w solar panel. It takes about 3.5hrs to get them to float since they are generally close to fully charged. I suspect I could run on lead batteries during the day and use lifepo4 batteries to run the fridges at night. Definitely overkill but I have plenty of battery box space. There used to be 4x8D truck batteries in the box ( 250Ah each at 12v) plus 2x100Ah engine start and another for the genset. I could lie down in it

 

[Kelpie]

I wouldn't dare have service bank lifepo only tbh.

Agree with this. My instruments and autopilot can run from either LFP or lead acid. I just have two standard isolators with one key between them. Turn off one bank, remove the key, use it to turn on the other bank. Takes a couple of seconds.

 

[vas]

Agree with this. My instruments and autopilot can run from either LFP or lead acid. I just have two standard isolators with one key between them. Turn off one bank, remove the key, use it to turn on the other bank. Takes a couple of seconds.

OK, could never do that, don't like spikes, so should first turn off inverter and a sh!tload of things and then turn the other one on. much easier to have them in parallel all working all the time and if lifepo want to take a rest, BMS turns it off (either internally or via a BlueSea el. isolator)

 

[Kelpie]

OK, could never do that, don't like spikes, so should first turn off inverter and a sh!tload of things and then turn the other one on. much easier to have them in parallel all working all the time and if lifepo want to take a rest, BMS turns it off (either internally or via a BlueSea el. isolator)

Sorry I wasn't clear, it's just the instrument circuit that gets turned off and then powered up again from a different bank. Other loads unaffected.

 

[GHA]

So what are yiu guys using for bulk/float?
Having a play here with victron smartshunt/smartsolar, trying 13.5v end of bulk with zero absorption time so goes straight to float, set to 13.4v. Bulk back on if voltage 0.1v under float for 1m so when fridge goes on, bulk comes back for a little while. I had to turn off & on the panels to get the smartsolar state to cycle to "off" to get settings to take effect.

Will leave it like this for a while & see what happens, would be nice to sit up towards 80% full each day automatically though might be asking bit much with the voltage output being so flat against SOC with lifepo4.

What's yours?

 

[Poey50]

So what are yiu guys using for bulk/float?
Having a play here with victron smartshunt/smartsolar, trying 13.5v end of bulk with zero absorption time so goes straight to float, set to 13.4v. Bulk back on if voltage 0.1v under float for 1m so when fridge goes on, bulk comes back for a little while. I had to turn off & on the panels to get the smartsolar state to cycle to "off" to get settings to take effect.

Will leave it like this for a while & see what happens, would be nice to sit up towards 80% full each day automatically though might be asking bit much with the voltage output being so flat against SOC with lifepo4.

What's yours?

General use: bulk 13.8 volts, absorption nil or shortest possible time (depending on charger), float 13.3 (effectively no float)

Storage settings on Victron Smart Solar: 13.3 bulk and 13.3 float

 

[GHA]

General use: bulk 13.8 volts, absorption nil or shortest possible time (depending on charger), float 13.3 (effectively no float)

Float seems so far (like watching for a few minutes ) to behave quite well to keep up with daytime loads & keep SOC quite high, hopefully that can be tuned a bit to just let the regulators do their own thing but always stay below about 80% SOC

 

[Alan S]

GHA, I think there is a possibility that you may in the long term overcharge one or more of your cells using these settings if there is little or no load.

 

[GHA]

GHA, I think there is a possibility that you may in the long term overcharge one or more of your cells using these settings if there is little or no load.

We´ll find out, measuring the real world is so much fun
3 cells exact to the mV today with one cell 1mV lower than the others. Nothing to worry about just yet.....

Any links to the physics describing just how a cell would go out of balance so much? ta

 

[Alan S]

Before I built my LiFePO4 pack I found a lot of useful and I believe accurate information on marinhowto.com. It's American and not up to date but still valid.
The conclusion I came to was that any float voltage over the cell's resting voltage will slowly continue to charge them.
I think the best way to use LiFePO4 is to cycle ie charge, discharge,charge with no float keeping within about 20 to 60% soc unless high demand is anticipated in which case charge to 90 or 100%. The cells will probably slowly drift apart so keep an eye on individual cell voltages. Rather than rely on auto cell balancing I prefer to manually bottom balance my cells using a 3.3V power supply. Only felt it needed this once in 18 months.

 

[Poey50]

Before I built my LiFePO4 pack I found a lot of useful and I believe accurate information on marinhowto.com. It's American and not up to date but still valid.
The conclusion I came to was that any float voltage over the cell's resting voltage will slowly continue to charge them.
I think the best way to use LiFePO4 is to cycle ie charge, discharge,charge with no float keeping within about 20 to 60% soc unless high demand is anticipated in which case charge to 90 or 100%. The cells will probably slowly drift apart so keep an eye on individual cell voltages. Rather than rely on auto cell balancing I prefer to manually bottom balance my cells using a 3.3V power supply. Only felt it needed this once in 18 months.

The marine how to article is referenced on the first page of this thread. It is accurate and Rod Collins updates from time-to-time.. He has also done a new article on drop-ins. 3.4 volts per cell tends to be the pivot point for charging. Less than this the charging stops, more than this and the pack eventually will charge to full. Bottom balancing is unusual - top balancing is the more usual way since this helps to avoid one cell hitting high voltage disconnect too much before others.

 

[GHA]

I think the best way to use LiFePO4 is to cycle ie charge, discharge,charge with no float keeping within about 20 to 60% soc unless high demand is anticipated in which case charge to 90 or 100%.

Not convinced.....
From the data here "float" seems to work quite well in just supplying loads , not really float like in a lead acid though. In the screen grab it´s the fridge that is the load, when it runs it will pull the voltage of the battery down a little, the regulators wake up & provide some more power, then back off the current again when the load stops & the voltage rises again. Though it seems to overshoot a bit.
So it looks like using a low "float" can be useful in reducing cycling each day down to very little. Need to run it & watch for a few weeks now, & check for cell voltage drift. All within 1mV of each other just now. 3.351V / 3.352V

The smartsolar looks like it takes a minute or so to pump up the current when the load turns on.

And of course like everyone else in the world the smartshunt capacity is of some mythical battery based on the parameter settings , not the real world - but hopefully close enough to be useful.

I think there is a possibility that you may in the long term overcharge one or more of your cells using these settings if there is little or no load.

Can´t find anything regarding this on google or google scholar - anyone?

 

[Alan S]

GHA and Poey, You could be right. I just don't like the idea of keeping cell voltage above its natural resting voltage. Most "experts" seem to agree that it is not ideal to float LiFePO4 but they are probably thinking of higher voltages.
My situation is a bit different from yours, mine is a 48V battery used for propulsion on inland waterways, so spends all its time either discharging, being charged or resting at about 3.27V.

 

[Alan S]

Poey, about bottom balancing, the reason I do this is that I very rarely fully charge my battery so even if I had automatic cell balancing it could probably not be effective. I am more concerned about low volts cut off than high so I prefer them to reach my 3.0V alarm point at about the same time. I use an Elerix Cell Performance Monitor and a Victron VBM700 watch cell voltages and state of charge etc.
Different situation from yours of course.

 

[Poey50]

Poey, about bottom balancing, the reason I do this is that I very rarely fully charge my battery so even if I had automatic cell balancing it could probably not be effective. I am more concerned about low volts cut off than high so I prefer them to reach my 3.0V alarm point at about the same time. I use an Elerix Cell Performance Monitor and a Victron VBM700 watch cell voltages and state of charge etc.
Different situation from yours of course.

Yes, that makes complete sense for your situation. I only mentioned that this is unusual as others are reading over our shoulders and I don't want people to get the wrong impression of what is standard practice.

 

[Zing]

Not convinced.....
From the data here "float" seems to work quite well in just supplying loads , not really float like in a lead acid though. In the screen grab it´s the fridge that is the load, when it runs it will pull the voltage of the battery down a little, the regulators wake up & provide some more power, then back off the current again when the load stops & the voltage rises again. Though it seems to overshoot a bit.
So it looks like using a low "float" can be useful in reducing cycling each day down to very little. Need to run it & watch for a few weeks now, & check for cell voltage drift. All within 1mV of each other just now. 3.351V / 3.352V

The smartsolar looks like it takes a minute or so to pump up the current when the load turns on.

And of course like everyone else in the world the smartshunt capacity is of some mythical battery based on the parameter settings , not the real world - but hopefully close enough to be useful.

Can´t find anything regarding this on google or google scholar - anyone?

I agree with Alan on keeping the charged capacity down somewhat unless you need full capacity simply to preserve battery life. High and low states of charge accelerates their life loss. I aim to run them at 30% to 70%.

Resting settled discharge volts should be about 3.31v per cell at 80%, which I think you said you want to achieve, but your resting volts looks like it corresponds to about 100% charge.

 

[geem]

I agree with Alan on keeping the charged capacity down somewhat unless you need full capacity simply to preserve battery life. High and low states of charge accelerates their life loss. I aim to run them at 30% to 70%.

Resting settled discharge volts should be about 3.31v per cell at 80%, which I think you said you want to achieve, but your resting volts looks like it corresponds to about 100% charge.

I fully understand where you are coming from and it makes a lot of sense to maximise life. Its real information from those using lifePO4. The reality is that you are only using 40% of the capacity of the battery in much the same way as those with lead keep their batteries between 50 and 100% for similar reasons.
None the less, I will be ordering some cells this month to build a battery. I little project over the summer.
It's not out of necessity as my lead installation is still going strong. I just fancy building stuff and doing some testing. I plan to sail home next year. We should then be into liquid cooled lifepo4 on my boat and ready for the next adventure

 

[Poey50]

Once you get over the lead-acid mind set of trying to get the batteries as full as quickly and as often as possible, if you aren't a high user of power then hanging around in the mid-range just becomes a natural thing to do. Then if some bigger demands are ahead charging fuller just takes a small amount of planning. I was surprised how often I've switched off chargers - including solar chargers to avoid running too full. That was unthinkable in lead acid days, of course. I'm not really trying to lengthen life - the mid-range thing just happens as part of everyday usage.

 

[Poey50]

I fully understand where you are coming from and it makes a lot of sense to maximise life. Its real information from those using lifePO4. The reality is that you are only using 40% of the capacity of the battery in much the same way as those with lead keep their batteries between 50 and 100% for similar reasons.
None the less, I will be ordering some cells this month to build a battery. I little project over the summer.
It's not out of necessity as my lead installation is still going strong. I just fancy building stuff and doing some testing. I plan to sail home next year. We should then be into liquid cooled lifepo4 on my boat and ready for the next adventure

The difference with using a smaller range of charge with LFP is that you always have the choice of using the whole range. Looking forward to hearing about your build.