Lifepo4 for solar battery?

yopappamon

yopappamon

10 kW
Joined
Mar 31, 2010
Messages
825
Location
Most dangerous city in the USA, Flint, MI
I have 1.3kwh of lipo on the way from HK :twisted: that will free up 28 - 20ah thundersky cells.

I'm wondering if they could be used for batteries in the solar power system I'm piecing together. Is that a good or bad idea?

If I parallel 7 cells each and use a 4 cell BMS?
 
They can certainly be used to store power from any source; the main reason many people use lead-acid for solar is because they're really cheap (often free if recycling), and also because they don't have to be moved around, so weight isn't a concern. Often there is very little cycling of the storage system, too, if it's primarily used at night and the daytime power is used "directly", as it were, depending on the loads of the system and the weather.

I'd expect the main issue with Lithium in a constantly-charging system like solar is preventing overcharge, so some sort of BMS (even one that just disconnects the pack once full charge is reached) is helpful. After that, preventing overdischarge would be a good second reason for a BMS.

Does your solar charge controller already output the voltage needed to max charge the 4S TS cells directly? Or would you be running the solar output to an inverter powering the regular charger for them (less efficient, but potentially "safer", if it is designed specifically for 4S of these cells).
 
amberwolf said:
Does your solar charge controller already output the voltage needed to max charge the 4S TS cells directly? Or would you be running the solar output to an inverter powering the regular charger for them (less efficient, but potentially "safer", if it is designed specifically for 4S of these cells).
Click to expand...

Right now I only have a cheapy solar charge controller that I'm sure is set up for lead acid. It's ratings are:
Stop charging at 14.0v
Low cut off 10.8v
Resume 12.6v
10 amps

I would expect to charge the batteries directly off the solar, not through the inverter.

I'll probably need to get a different charge controller.
 
LiFePO4 is a better fit to solar charge than lead-acid. Lead-acid wants recharge as fast as possible, requiring a lot of panels, and then a long absorption period where a single panel would be more than enough. If not fully charged they lose a little capacity every night through sulfation.

Lithium actually has a longer life at half charge, and not much energy is stored in a wasteful absorption phase. You could tie a 100 watt panel to an empty LiFePO4 pack and just let it run all day without any charge controller. The bigger the pack the better. A MPPT or buck controller would boost the charge rate, but it might be more cost-effective to use that money for extra cells. There would be a substantial off-grid storage market for heavy low power high capacity LiFePO4 batteries, if they could be made cheaply enough.
 
dak664 said:
You could tie a 100 watt panel to an empty LiFePO4 pack and just let it run all day without any charge controller.
Click to expand...

Wouldn't that over volt the cells? 18v/4 = 4.5v per cell. The panel would need to be regulated down to around 14.8v to get 3.7v per cell, right?

The other things I'm not totally clear on is can I do a 4p7c?
Will the 7 cells across stay balanced with each other?
Would I only need a single 4 cell bms to keep them balanced?
Would I need a large balancing current capacity since it would be balancing 140ah?
 
yopappamon said:
Wouldn't that over volt the cells? 18v/4 = 4.5v per cell. The panel would need to be regulated down to around 14.8v to get 3.7v per cell, right?
Click to expand...
A photovoltaic cell is (nearly) a constant current source up to its maximum voltage output. Light kicks a certain number of electrons across an internal (nearly) fixed voltage gradient. If when you short the output you get a certain number of amps, then if you hook it up to a battery you will get nearly the same number of amps. If that's a safe charging rate for the battery then the battery voltage won't increase, and any excess voltage from the panel is lost in the connecting wires. So no regulation is needed till the battery nears full charge, which is guaranteed to take days with a 100 watt panel charging a couple kWh battery. You could alternate battery packs each day and size everything for a typical ride and never need a charge controller. That said, a high voltage cutoff allows oversizing the panel to accomodate cloudy days, and an expensive MPPT or more modest buck converter charger can trade off extra voltage for ~20% more current.

If you want to top-balance I'd say that is best done with a mains charger. Low-current balancing would waste most of the PV output. But you could maybe mid-balance at a few amps with ~10ma shunts across high cells; they would still get the bulk of the charge, just not as much as the low cells. Of course you would want to disconnect those shunts when charging stopped completely, e.g. at night.

Charge controllers would make more sense with two or more attached to the PV panel, one could be doing low current float or equalizing while the main power goes through another for bulk charging.
 
To me it doesn't make sense to use a relatively light compact expensive battery like this for solar.. if you're going to dish out big cash you may as well go for the NiFe ones txpyro & LFP pointed out in another thread.. I guess mebbbe if you've got them already and aren't using them for other things..
 
@dak664 - So you are saying as long as I don't let the batteries top off, no charge regulation would be needed. I have a 120w solar panel, to charge 100ah @ 7 amps would take 14+ hours or around 2 days. As long as I use the batteries down to more that the panel can supply in one day, the batteries will never top off and not over volt?

This is the solar panels specs.
CHARACTERISTICS:
Power at STC ( pm ): 120 Watt
Maximum Power voltage ( Vpm ): 17.3 Volts
Maximum power cusrrent ( Ipm ): 7.0 Amp
Open circuit voltage ( Voc ): 21.9 Volts
Short circuit current ( Isc ): 7.72 Amp
Tolerance: +- 5%
Click to expand...

@vi - I agree, but don't have another use for them at the moment.
 
yopappamon said:
@dak664 - So you are saying as long as I don't let the batteries top off, no charge regulation would be needed. I have a 120w solar panel, to charge 100ah @ 7 amps would take 14+ hours or around 2 days. As long as I use the batteries down to more that the panel can supply in one day, the batteries will never top off and not over volt?
Click to expand...
Right, and with lithium you don't care if the cells never fill, in fact for long life you don't want them to fill. With a vehicle battery you probably want to start out with a full charge, but for home power with lithium it doesn't matter. Expensive charger money is better spent on another cell or two, which will increase your storage capacity and at the same time the life of the other cells (because they don't get charged as fully). With lead-acid batteries it's another story, you want to baby those with absorption charge and equalizations.

CHARACTERISTICS:
Power at STC ( pm ): 120 Watt
Maximum Power voltage ( Vpm ): 17.3 Volts
Maximum power cusrrent ( Ipm ): 7.0 Amp
Open circuit voltage ( Voc ): 21.9 Volts
Short circuit current ( Isc ): 7.72 Amp
Tolerance: +- 5%
Click to expand...
In real life you will get 25-35 amp hours out of that on a sunny day. In summer the voltage and power will be lower but the amp-hours will be more (longer days). In winter the voltage and power will be higher but the amp-hours less, unless you have a mppt or buck charger to convert the higher voltage from colder panels to more current.

Nickel-iron cells have relatively poor charge efficiency of ~75% so waste PV capacity if you are using them to shift loads to nighttime. Their self discharge rate is also very high, ~25% in a month, that's another loss in useful PV output. LiFePO4 would be perfect for home power if it got cheaper. PV, home power, and electric car with swappable battery packs could make a very attractive package.
 
dak664 said:
Right, and with lithium you don't care if the cells never fill, in fact for long life you don't want them to fill. With a vehicle battery you probably want to start out with a full charge, but for home power with lithium it doesn't matter. Expensive charger money is better spent on another cell or two, which will increase your storage capacity and at the same time the life of the other cells (because they don't get charged as fully). With lead-acid batteries it's another story, you want to baby those with absorption charge and equalization.
Click to expand...

Thanks for the info. This is really going to simplify my setup. :mrgreen:

So as far as having 7 cells parallel, will they stay in balanced with each other?
 
Cells connected in parallel will certainly have the same voltage at all times, but what exactly do you mean by balanced?
 
dak664 said:
Cells connected in parallel will certainly have the same voltage at all times, but what exactly do you mean by balanced?
Click to expand...

Exactly that. That they will discharge at the same rate and reach the lvc or hvc together. I guess it's a stupid question when I think about it, there is no way for them to have a different voltage from each other. I suppose they could have a different state of charge while being at the same voltage.

I guess I'm just making sure it's ok to wire up lifepo4's this way.
 
Discharging each cell at at the same rate would mean the same resistance to the load. All the recommendations I've seen argue for the same internal resistance, but I suppose you could compensate through external resistance to the parallel connection, maybe how hard you tighten down a connecting screw. Capacity shouldn't matter, a 9 amp-hour cell in parallel with a 10 amp-hour cell should do fine, total resistance being equal. If one cell has a higher possible voltage you will lose storage capacity by only charging to the lowest maximum voltage cell, but all cells may last longer as a result. I don't see how any damage could be done by parallelling cells. Overdischarge dissolves copper from the anode in other lithium chemistries where there is an excess of lithium, but not properly built LiFePO4 cells, and all would be at that same dangerously low voltage anyway. Overvoltage breaks down the electrolyte, to the same degree in all cells if that's any consolation. I suppose on overvoltage a one amp hour cell would fail sooner than a 10 amp hour cell, given similar engineering parameters. Best not to over- or under- voltage cells methinks, unless you like buying shiny new ones
 
I used copper plated strapping for bussing, foam core board to space the cells so the holes line up, PVC sheeting strapped around the cells.

df4c96b9.jpg
 
yopappamon said:
What would be a good way to switch the high current dc to the inverter? I don't want to use a relay because of the current draw. Thinking solid state. I have a few 4110 fets, could I make a 200 amp, 12v switch from them?
Click to expand...
No reason it wouldn't work, but remember that FETs tend to fail on (shorted), if they do fail. If you think there might be voltage spikes near the max FET VDS, or high-power RF noise on there for any reason, put something in the circuit to snub that. ;)
 
Should work as long as the cells are solidly clamped together and can't move/expand any during charge/discharge. I made some (have to still make the rest) braided flexible interconnects on mine since they'll be in a moving vehicle, but you don't need that for a stationary project. :)

Not sure what metal they are made of; is it really pure copper, or were they adapted from something made for another purpose? Copper alloys may not be nearly as conductive as the pure stuff.

EDIT: I see you say "copper-plated strapping", so I don't know what the base metal is. Running current thru it probably wont' be as effective at high currents as with solid copper, and I don't know to what degree that will matter.

I also am not sure if you will end up with electrolytic actions where the plating meets the underlying metal, at points like the cell terminal bolts where pressure might deform the strapping's plating (depending on how it was plated and how thick it is, how clean the original metal was, etc).

Best way to find out is to test it, though. ;)
 
A123 says you can continuously float charge their cells a 3.45V Very close to your 14V charger output. Adjust the charger a bit or add some diodes in series and you are there.

For balancing I would use a switched capacitor (NOT one of those "capacity" or "capacitance" balancer that people here so often whine about) balancer on the pack that runs continuously. It should have no problem keeping the pack happy. See http://endless-sphere.com/forums/viewtopic.php?f=14&t=20864&start=0&hilit=switched+capacitor for some discussion on this type of balancer.
 
yopappamon said:
I used copper plated strapping for bussing, foam core board to space the cells so the holes line up, PVC sheeting strapped around the cells.
Click to expand...

That may not be a good way of doing the pack. Those cells want to be strapped together tightly with metal plates on the ends to keep them from swelling. Foam core spacers are not good for maintaing a tight pack. I would buy some thin-ish copper sheet and cut/drill some custom bus bars out of it. Gets rid of those little jumper straps too.
 
Any meter can easily measure the millivoltage drop from one end of the bus bar to the other while under a 1 or 10 amp load. Measure both ways in case your cheap meter has an offset. 100 millivots at 1 amp is 0.1 ohm. Solid copper would read zero methinks.
Same thing across jumper straps or from bus bar to battery cap. My headway screw connections to copper flashing busbars read around 2 milliohms.
 
texaspyro said:
That may not be a good way of doing the pack. Those cells want to be strapped together tightly with metal plates on the ends to keep them from swelling. Foam core spacers are not good for maintaing a tight pack. I would buy some thin-ish copper sheet and cut/drill some custom bus bars out of it. Gets rid of those little jumper straps too.
Click to expand...

Is that really important with lifepo4? I never had any more than tie wraps around it when they were on the bike. The foam core is a lot sturdier than you might think. I had to stand on the end of the pack to just gain a 1/32" to get the last screw in.

I agree on the buss bar. For $7 I can get a 6 foot x 1" x 1/8" copper bar from the industrial metals place around the corner.

Although the straps seem to work fine right now. I ran a blow drier off the inverter, showed 100 amp draw and nothing smoked or smelled hot. I'll try the voltmeter reading in a minute.

bde77e63.jpg


0d8080be.jpg
 
You must log in or register to reply here.

Similar threads

B
20S lifepo4 solar battery
Replies
0
Views
314
Bacon-Matte
B
T
Voliomart 1500 watt hub motor- How much can I over volt.
Replies
8
Views
530
hillslayer@
H
tmho
12V LiFePO4 Battery Project - Suggestions Wanted
Replies
10
Views
301
tmho
tmho
A
Is it possible to replace segways 73.6 lifepo4 battery with Li-Ion cells ?
Replies
2
Views
484
amberwolf
amberwolf
O
Lifepo4 cylinder 32650 cells simplified stationary design idea?
Replies
7
Views
1,807
jonescg
jonescg
Back
Top