LFP battery chatter ( AKA LiFePo4, Lithium Iron Phosphate )

[northerner]

OK, there was an intervening post asking about paralleling different chemistries, and that blew my mind. For my needs, keep it simple - same chemistry, same exact thresholds.

I can understand the desire to keep the system simple and you made a good point and such a proposal may not be for you.

The way I see it is that batteries are the single biggest expense in an off grid system, and that is the bottom line. Anything one can do to minimize this expense is to their advantage. Sodium ion batteries have the potential to offer a relatively low overall cost option for storage. They have a few shortcomings that can possibly be dealt with. It does increase some complexity to the system, but not in a big way.

Think of the LFP batteries as just there to give the sodium ion batteries a boost when needed. Otherwise they would be completely separate from the system. Of course it would be best to have the LFP battery pack on a separate charge controller, and still have a low voltage disconnect to protect from damage.

It would not be difficult to control the cut in and out of the LFP pack. Essentially, by examining the voltage and current flows from the sodium ion pack, a decision could be made. Would be a simple circuit that could be fine tuned with further logic possibly.

[Living Large]

I can understand the desire to keep the system simple and you made a good point and such a proposal may not be for you.

The way I see it is that batteries are the single biggest expense in an off grid system, and that is the bottom line. Anything one can do to minimize this expense is to their advantage. Sodium ion batteries have the potential to offer a relatively low overall cost option for storage. They have a few shortcomings that can possibly be dealt with. It does increase some complexity to the system, but not in a big way.

Think of the LFP batteries as just there to give the sodium ion batteries a boost when needed. Otherwise they would be completely separate from the system. Of course it would be best to have the LFP battery pack on a separate charge controller, and still have a low voltage disconnect to protect from damage.

It would not be difficult to control the cut in and out of the LFP pack. Essentially, by examining the voltage and current flows from the sodium ion pack, a decision could be made. Would be a simple circuit that could be fine tuned with further logic possibly.

Just using LFP for solar is living on the edge, from what I have seen. You mention complexity - and you are increasing it. Everything is relative, but you are taking something with built-in risks and making it more complex. You say you want to keep overall costs low, and you mentioned keeping the inverter from kicking out on heavy loads. This is fairly new to me, but LFP is expensive even if you undersize the LFP bank, and it has great current capabilities. For me personally, I believe my use will lend itself to using an undersized bank for the worst months. The fact LFP is happy not being fully charged, I can get 70-80% DOD and I can charge it faster than FLA are good qualities - if it weren't for these I wouldn't be considering LFP.

Others can give a more educated response, but I don't see the route. I'm not a person for innovation though - I'm the guy asking about paralleling the equivalent of a C and AA cell - for all I know you may have the next best thing. Every application is unique, and maybe yours lends itself to what you suggest.

[northerner]

Just using LFP for solar is living on the edge, from what I have seen. You mention complexity - and you are increasing it. Everything is relative, but you are taking something with built-in risks and making it more complex. You say you want to keep overall costs low, and you mentioned keeping the inverter from kicking out on heavy loads. This is fairly new to me, but LFP is expensive even if you undersize the LFP bank, and it has great current capabilities. For me personally, I believe my use will lend itself to using an undersized bank for the worst months. The fact LFP is happy not being fully charged, I can get 70-80% DOD and I can charge it faster than FLA are good qualities - if it weren't for these I wouldn't be considering LFP.

Others can give a more educated response, but I don't see the route. I'm not a person for innovation though - I'm the guy asking about paralleling the equivalent of a C and AA cell - for all I know you may have the next best thing. Every application is unique, and maybe yours lends itself to what you suggest.

Using sodium ion batteries would decrease the complexity of the system, as there would be fewer worries with cell balancing. And paying Mr Peukert is not too bad with them if your loads are relatively low.

Referring to a chart that Aquion has released about the S20 battery stack, they have plotted round trip efficiencies vs current from one S20 stack.

Current draw at 2 amps yields an RTE (round trip efficiency) of ~ 90%

Current draw at 4 amps yields an RTE of ~ 85%

Current draw at 6 amps yields an RTE of ~80%

Current draw at 10 amps yields an RTE of ~70%

As you can see, losses due to heat become higher with bigger loads, and overall efficiency suffers. I know with my system that most of the time, the loads are relatively light. However, at times loads can be heavy. This is where a boost would come in handy, and LFP is very suited to handling the bigger loads with less loss, and would bring up the voltage of the bank, preventing it from potentially dropping out the inverter. If the heavier loads are few and far between, the extra LFP bank may not even be necessary!


The overall economics of using sodium ion batteries could be good, and managing a smaller LFP bank would involve lower economic risk (ie if things did go bad).

[Sunking]

You cannot put LFP in parallel with any other battery. LFP Ri is so low no other battery would ever see a load.

[northerner]

You cannot put LFP in parallel with any other battery. LFP Ri is so low no other battery would ever see a load.

According to this study, it can be done, although technically they are a lithium metal polymer battery (not specified exactly) and not likely LFP:

http://www.battcon.com/PapersFinal2005/NgPaper2005.pdf

The batteries would only be together for heavy load handling, not for any lengthy duration, so even if they did take most of the load, they would accomplish the mission set out for them. Of course you would have to size the LFP bank to be able to handle the loads you task them.

[Sunking]

Not with my money. But I will be more than happy to spend yours.

[northerner]

Not with my money. But I will be more than happy to spend yours.

Sometimes you have to spend a bit of money to save in the long run.

[paulcheung]

Just using LFP for solar is living on the edge, from what I have seen. You mention complexity - and you are increasing it. Everything is relative, but you are taking something with built-in risks and making it more complex. You say you want to keep overall costs low, and you mentioned keeping the inverter from kicking out on heavy loads. This is fairly new to me, but LFP is expensive even if you undersize the LFP bank, and it has great current capabilities. For me personally, I believe my use will lend itself to using an undersized bank for the worst months. The fact LFP is happy not being fully charged, I can get 70-80% DOD and I can charge it faster than FLA are good qualities - if it weren't for these I wouldn't be considering LFP.

Others can give a more educated response, but I don't see the route. I'm not a person for innovation though - I'm the guy asking about paralleling the equivalent of a C and AA cell - for all I know you may have the next best thing. Every application is unique, and maybe yours lends itself to what you suggest.

I have read this thread with great interest, personally I don't like the flatness of this battery, one simple mistake the battery are dead, I will prefer the FLA batteries. with your load if it were me I just get one set of the Trojan IND13-6V or the 2volt, 4 volts with the same capacity.
These Trojan battery are claim to be able to do PSOC and only require once a while for full SOC to battle the sulfation.

[northerner]

I have read this thread with great interest, personally I don't like the flatness of this battery, one simple mistake the battery are dead, I will prefer the FLA batteries. with your load if it were me I just get one set of the Trojan IND13-6V or the 2volt, 4 volts with the same capacity.
These Trojan battery are claim to be able to do PSOC and only require once a while for full SOC to battle the sulfation.

I am actually leaning towards the sodium ion batteries. Better bang for your buck in the long run, and prices are expected to come down further by the end of this year. As mentioned earlier in this thread, they do have shortcomings that can be dealt with if need be.

I did a cost comparison with the IND13-6V and it comes in around 25 cents per kwh. Aquion's S20 stack comes in at about 16 cents per kwh. But it may not be suited to everyone's needs!

[Living Large]

I have read this thread with great interest, personally I don't like the flatness of this battery, one simple mistake the battery are dead, I will prefer the FLA batteries. with your load if it were me I just get one set of the Trojan IND13-6V or the 2volt, 4 volts with the same capacity.
These Trojan battery are claim to be able to do PSOC and only require once a while for full SOC to battle the sulfation.

That's a nice battery at a glance. I don't believe I saw it previously. I'll check into that some more.

It's a personal preference thing to an extent, Paul. I agree that if you make the wrong mistake with LFP, it is a darned expensive one - one likely not to be repeated.

[Sunking]

I have read this thread with great interest, personally I don't like the flatness of this battery, one simple mistake the battery are dead,

Paul do not let the fear of the unknown taint your judgement. Flat discharge curves and low internal resistance are two things you really want out of your battery.

The flat discharge curve and no Peukert Effect make determine the SOC a cinch and reliable even when under load. With very little Peukert Effect you can use Coulomb Counting with great accuracy. All that means is counting energy units in and out of a battery. Not Amp Hours, never count Amp Hours as that don't mean chit because Amp Hours in, does not equal Amp Hours Out. Oh so you want to know why? Good question. Amp Hours go in at higher voltage than they go out, thus meaningless and why we do not deisgn systems using amp hour. Coulomb Counting is Watt Hours because it measures both voltage and amp hours.

Secondly with LFP flat discharge curve SOC is nearly linear from 3.3 volts to 2.9 volt meaning for each .004 volts = 1%.

Lastly the never cross line for LFP is 2.5 volts. The LVD is set for 2.9 volts way before you fall off the cliff.

[paulcheung]

Ok, Dereck, What is the SOC at 2.9 volts per cell? how long does it take to go down to 2.5 volts assume the load is between C/10 to C/20 rate? Some Inverter took a minute or two to cut off the power after it reach the preset voltage to cut off. They don't want to cut off the power prematurely.

Thank you.

[Sunking]

Ok, Dereck, What is the SOC at 2.9 volts per cell? how long does it take to go down to 2.5 volts assume the load is between C/10 to C/20 rate? Some Inverter took a minute or two to cut off the power after it reach the preset voltage to cut off. They don't want to cut off the power prematurely.

Paul at 2.9 volts you are in the last 7 to 10% left in the tanks. All LFP LVD's have a timer you can select to prevent false trips and if they ever hit 2.6 volts trip instantaneously.

[Sunking]

Trojan IND13-6V or the 2volt, 4 volts with the same capacity.
These Trojan battery are claim to be able to do PSOC and only require once a while for full SOC to battle the sulfation.

Paul both The RE series and IND series use the same T2 carbon technology. I would use a trusted name like Trojan before a newcomer.

The RE warranty is 24/60 and Industrial is 36/96. The Industrial line is built like Fork Lift batteries. No where is the kicker and eye opener. Aquion battery warranty is 24 months period. For an additional charge they will sell you an optional 24/60 warranty. That ought to tell you something.

[paulcheung]

Thank you Dereck, I know my next bank battery is the Trojan IND17-6V. I just hope when I am ready to replace the batteries I have enough fund to buy them.