LiFePO4 - The future for off-grid battery banks?

[Coolrio]

Greetings from New Zealand too

Hi Steve g:

Am planning to build an off grid earthship in Hawkes Bay, New Zealand. Would appreciate connecting with you to find your experiences with the LifePO4 cells in your solar system.

Warm regards,

Riaz

[T1 Terry]

Haven't been on here for quite while, how is your system holding up Steve? My 720Ah Winston cell Li system has been struggling through the depths of winter, hovered around 20% down to 0% each day for around a mth but they are clambering back now. A very large gum tree in our front yard hasn't helped things at all with the sun hiding behind it from around 11am to 3.30pm but now it's starting to climb a little higher in the sky things are looking better.
I am trialling an idea that I've kicked around for the last 6 mths or more, running the solar direct and load shedding into the water heater if needed. The idea revolves around the MPPV of 12v panels being between 17.5v cold and 15v hot and a 5 Li cell pack having a capable voltage range of 14v (2.8v per cell) to 20V (4v per cell) The 20v end is past the point the panels make any current so that end is reasonably safe although high cell cut (HVC) is still need as a back up in case the cells go out of balance and a low voltage cut (LVC) to protect the cells from a very deep discharge. This 5 cell pack is used to charge the operating house power 4 cell pack via a solid state relay and heavy cabling.
It has worked well for a couple of days but unfortunately I got side tracked on helping a mate with is project and didn't have an operational LVC, one cell hit 0.456v by the time I got home to check it. Over charging isn't the only thing that causes these cells to bulge, hopefully I'll be able to recover it as it was a new cell.
Basically, the idea is to store the added available solar output rather than loosing it to voltage mismatch through PWM control or heat through MPPT control. The price of a truely worthwhile MPPT unit buys a lot of batteries and simple on/off circuits to drive solid state relays.
The extra Wh harvested this way is quite impressive as is how much quicker the panels produce an input to the batteries compared to MPPT, well compared to the 4 units I've trialled anyway, no one has given me an Outback MX 80 to test, yet The cost of one of those units buys over 250Ah on lithium @ 12v so it's worth the effort if it works as the added storage keeps the main battery bank full well after sun down. Sort of a really big nearly super capacitor set up.
Once I've ironed the bug out and have a better operating control system I'll post more on it, or if it's a complete diaster I'll post that as well to save anyone else going down that path.

T1 Terry

[inetdog]

It has worked well for a couple of days but unfortunately I got side tracked on helping a mate with is project and didn't have an operational LVC, one cell hit 0.456v by the time I got home to check it. Over charging isn't the only thing that causes these cells to bulge, hopefully I'll be able to recover it as it was a new cell.

T1 Terry

If this is like the other common Li chemistries, as found in Li-ion batteries for computers, tools, etc. There is a point (voltage) in the discharge cycle at which the chemical reaction shifts from reversible to irreversible. When you go past this point, the cell becomes useless.

Many people think that the reason a tool using an Li battery stops suddenly without warning at the end of its charge is that the discharge curve drops suddenly. That is not the case. Instead the protective circuitry in the battery pack has cut off the output to protect the cells. That is also why, unlike NiCad or NiMh packs it does not recover at least a little after being left idle.

If one cell actually got to .456 volts, it is gone forever.

[T1 Terry]

These are the lithium ferrous phosphate cells (LiFeP04) they are damaged if pulled down below 2v and held there apparently as the electrolyte heating happens at both ends of the voltage range, hence the cell case swelling. I will try recompressing the cell and test how much capacity I've lost, I did a similar thing at the high voltage end to anoter 2 cells so I'll build a 12v battery out of them with one good cell and measure capacity down to 2.8v per cell and see if the one good cell is still at a high voltage at the end and if all 90ah are still there. Maybe (hopefully) it only effects their charge and discharge rate, that's not as critical in a house power set up as it is in an EV set up.

T1 Terry

[billvon]

I am trialling an idea that I've kicked around for the last 6 mths or more, running the solar direct and load shedding into the water heater if needed. The idea revolves around the MPPV of 12v panels being between 17.5v cold and 15v hot and a 5 Li cell pack having a capable voltage range of 14v (2.8v per cell) to 20V (4v per cell) The 20v end is past the point the panels make any current so that end is reasonably safe although high cell cut (HVC) is still need as a back up in case the cells go out of balance and a low voltage cut (LVC) to protect the cells from a very deep discharge. This 5 cell pack is used to charge the operating house power 4 cell pack via a solid state relay and heavy cabling.

Yikes. A few issues there:

LiFePO4 is more rugged than LiCo, but are nowhere near as rugged as lead acid. A bang-bang (PWM) controller used with a diversion load is going to tend to hurt them since you see momentary overvoltages with them. Also they're not very tolerant of voltages over about 3.8 volts.

Also you can't rely on "the 20v end is past the point the panels make any current." The time you run into that is on a cold day when you're not there and the panel has all day to take the battery up to 24V. You absolutely need a good charge controller - and for a lithium chemistry it's probably going to have to be switchmode.

Sort of a really big nearly super capacitor set up.

Batteries are much more finicky than supercaps. Supercaps are indeed a lot more rugged, provided they are balanced and kept under their voltage limits.

[T1 Terry]

actually there are very close similarities between lithium ferrous chemistry and super capacitor chemistry, same electrolytes, very similar coatings but a much thinner plate coating on a super capacitor which enables it to to charge and discharge at such high currents, the down side is bugger all storage capacity.
Lithium ferrous are far more rugged than lead acid could ever hope to be, what lead acid battery would recover from a 100% discharge repeatedly without loss of capacity or sit for 3 yrs at 50% state o charge and still have the same 50% when the come out of the box? 4.5V is the true danger no return point for lithium ferrous cells, a simple HVC circuit using the Junsi cell logger stops the over charge of any cell, switch the solar output to a water heater element while the battery voltage settles, then back to charging or simply switch the element on if the current draw is greater than the solar input potential. Here is graph of one in action on a 24v battery pack
Brian's heater in action.jpg

The load is greater than the solar input so the cell voltage drops, the fact that there is no capacity over 3.4v is shown here by the near vertical drop of cell 7 & 8 voltage from the high 3's to below 3.4v.

T1 Terry

[billvon]

Lithium ferrous are far more rugged than lead acid could ever hope to be

That hasn't been our experience, but good luck with your experiments.

Here is graph of one in action on a 24v battery pack

That sort of charge profile was rejected by at least two LiFePO4 cell manufacturers that we talked to. They were worried about pulse charging, especially the peak voltages it entailed. It would be interesting to see the entire graph without the high voltages cut off.

[T1 Terry]

Sounds like you have had some bad experiences with lithium ferrous and talked to some people who weren't particularly cluey about lithium ferrous cells, certainly not manufacturers. Pulse charging is actually the preffered method of ultra fast charging but there are very few charging systems capable of doing it, these cell are tested at 20C pulse charge rates so pulse charging certainly isn't a problem for lithium ferrous cells.
There are no secrets or mysteries wit the chart, the cut offs are 3.6v, I'll do a reprint
Brian's battery test day one.jpg

T1 Terry

[deadbeat]

Any Thoughts About These 12V 1000AH LifePO4 Batteries?

http://www.evtradinpost.com/item/556

The discription says that they are "ideal for solar installations, wind and other energy storage applications". Also reads that they include the BMS.

[billvon]

Sounds like you have had some bad experiences with lithium ferrous and talked to some people who weren't particularly cluey about lithium ferrous cells, certainly not manufacturers.

We worked with two manufacturers, actually, to get a better spread of experience.

Pulse charging is actually the preffered method of ultra fast charging but there are very few charging systems capable of doing it, these cell are tested at 20C pulse charge rates so pulse charging certainly isn't a problem for lithium ferrous cells.

After the pulse charging disasters of the 1990's no one is willing to try that charge regime any more.

Like I said, glad it's working for you. We find that there is a big difference between "it looks like it will work" and "we can guarantee its operation over X cycles when charged to Y volts at Z temperatures."

[russ]

http://www.evtradinpost.com/item/556

The discription says that they are "ideal for solar installations, wind and other energy storage applications". Also reads that they include the BMS.

Sorry but that site sounds one or more steps lower than fleabay

[Fatawan]

I am working on a 6. 5 kw system and are considering lithium nmc or iron phosphate batteries. I am currently doing research and have found Corvus energy that offers these for actual solar systems. I am waiting for more info and pricing from the manufacturer. Will get back when I hear more.

Northernflyer

Did you ever get any more info on these batteries? I have a call into the company but no response yet

[zak650]

lifepo4 compatable inverter/charger

Hi,

Has anyone run across an inverter/charger designed to charge lifepo4 batteries?

Thanks,
Zak

[Sunking]

Hi,

Has anyone run across an inverter/charger designed to charge lifepo4 batteries?

Nope no demand for them. Only ones I know about are ones made for RC toys. You would have to custom build one.

[inetdog]

Nope no demand for them. Only ones I know about are ones made for RC toys. You would have to custom build one.

Or find an LiFePO4 battery system with internal BMS which has been designed to work with a conventional inverter/charger. More than one way to skin a cat, although not what the OP asked for.