LiFePO4 fast charging paper (4C)

[curiouscarbon]

Hey all,

Been doing research about safe and fast charging/discharging of LFP cells.
Ran across a paper titled "Fast charging technique for high power LiFePO4 batteries: A mechanistic analysis of aging" [researchgate link]

They discuss charging and discharging LiFePO4 cells at 4C in a constant ambient temperature of 23°C/73.4°F.

Full re- charges were achieved within ~20 min and 17% of capacity was lost after 4500 full charge/discharge cycles.

Pretty promising results. This suggests to me that running a peltier cooler or compressor type cooler to keep the cells at ~74F could drastically improve operation life, at the cost of some lost power to the temperature regulation apparatus. edit: The paper mentions cylindrical cells, so thermal conductivity of the pack is definitely on my mind.

Anyone have thoughts on this research?

 

[curiouscarbon]

Another paper, "Experimental Measurements of LiFePO4 Battery Thermal Characteristics" [pdf link] has some good stuff.. from page 32

In more experimentally focused studies, researchers increase temperature data resolution by adding more thermocouples. In [44], an experimental study was carried out on LiFePO4/C prismatic 8 Ah batteries. Surface temperatures were measured at 10 locations around the batteries. Within an incubator set to 25 °C, a single battery was charged and discharged at rates between 1C and 25C. The results were analyzed to determine the maximum temperature difference between locations of the battery (ΔTmax ), and the average rate of temperature increase ( dT/dt average) for all charging and discharging conditions. The relationships between these values and charge-discharge rate were found to be linear, indicating empirically that discharge rate is the main driving force for heat generation. Results from this experiment are compiled in Table 2.4.

10C charge and discharge tests on 8Ah prismatic cell in 25°C/77°F environment.

 

[curiouscarbon]

The first paper has some more specifics..

The multistage charging method consists of three charging steps, at 4C, 1C and a fixed 5-min length CV step. The fast charging technique allows the cell to recharge in 1/3rd of the standard time. More details of this method can be found in our previous work [32]. [..] The cells were always cycled within the voltage limits recommended by the manufacturer (i.e., 2.0V-3.6V).

And their previous work[32] is: "Fast charging technique for high power lithium iron phosphate batteries: A cycle life analysis" [pdf link] which has this to say about the fast charging profile:

The proposed multistage fast charging technique profile is shown in Fig. 2. The charging process is split into three different stages, referred as CC-I, CC-II and CV-I. The first stage (CC-I) starts with a constant current at 4 C to the charging cut-off voltage (3.6 V). The second stage (CC-II) is a constant current charge at 1 C. Since the current in CC-II is lower than in CC-I, the cell voltage drops below 3.6 V (see Fig. 2 (b)) allowing the charge to be extended, until the cell reaches again the charging cut-off voltage. The last stage (CV-I) is performed at a constant voltage of 3.6 V for a duration of 5 min.

 

[svetz]

Thanks for that reference, interesting looking paper.

My first reaction though is that batteries on PV systems don't generally run at high C rates.

If you did have a high discharge rate, the batteries would exhaust pretty quickly, so you'd probably be better off investing in more batteries to both decrease the rate and increase the operational time.

So, what about charging? Let's look at the math... let's say you have a poor winter insolation of 2. Then for each 10 kWh of battery you'd need 5 kW of solar to recharge per day, or a charging rate of 1/2 C at solar noon and much less at other points of the day. In summer if the insolation went to 5, then the panels would generate 25 kWh per day, but they could still only ever generate a peak of 5 kW, so still only 1/2 C but for a longer period.

 

[curiouscarbon]

Most people don't run high C rates for good reason, but this thread is about high C rates.

Charging a car is a useful situation, even though the house battery might draw down relatively quickly.

e.g.

Victron batteries are really expensive, so I'm trying to spur a conversation on how to safely achieve high C rates when needed.

 

[curiouscarbon]

Admittedly, charging at high C rate from solar does sound uncommon to me too, or a symptom of underspecified battery capacity.

Some batteries are likely more amenable to air ventilation than others. Packs where the spacing between the outside wall of the cells is very narrow would certainly heat up more in the center. fortune/overkillsolar cells have a 3-4mm gap between the cells due to the plastic holder, so cool air could be pushed through there to keep things happy and safe.

 

[John Frum]

Can a lifepo4 battery actually draw 4c with a voltage differential of .5 volts?

 

[curiouscarbon]

Can a lifepo4 battery actually draw 4c with a voltage differential of .5 volts?

Could you help me better understand what you mean by voltage differential of 0.5 volts?

This sinopoly battery test video shows a 200Ah battery delivering ~1400 A continuously, which is 7C discharge with no cooling. It's.. very unhappy about this. Definitely avoid 7C discharging.

You can tell it's the 200Ah cell because there are 11 notches on the outside along the narrow side. the one in the video has 11 as well...

 

[John Frum]

3.65 volts is full to the brim for lifepo4.
So if the charger voltage is 3.65 volts and the cell voltage is 3.6 volts the differential is .5 volts.
For a 100 ah battery
ohms = volts / amps
0.00125 ohms = 0.5 volts / 400 amps
I don't think its practical to get the circuit resistance that low.

 

[John Frum]

Actually maybe with voltage sensing leads it might be doable.
I think the internal resistance of a lifepo4 cell is in that neighborhood.

 

[John Frum]

Dang, nevermind my math is off by 100.
The potential difference between 3.6 and 3.65 is .05 volts.

 

[curiouscarbon]

@smoothJoey the fast charge sequence is: 4C -> 1C -> constant voltage 3.6V for 5 minutes

So in answer, they don't charge at 4C near the top of the charge. But they get really close and then go constant voltage for five minutes and consider it full after that. Hope that helps ?

 

[curiouscarbon]

The fast charge algorithm seems to be, in plain words:

Start at low/zero state of charge. Begin charging at 4C until voltage hits 3.6V. Reduce to 1C rate, continue until voltage hits 3.6V again (it will dip because of lower charge rate). After hitting 3.6V for the second time, hold at 3.6V for five minutes. Fully charged.

Check out these graphs, and notice that the voltage stays below 3.6 and dips after the constant current phases:



Hope this helps clarify how the proposed approach works. There's no need to push large amps at small voltage difference.

 

[curiouscarbon]

Some of these experiments were carried out with A123 nanophosphate cells. Definitely wouldn't assume any old cell can do 4C charging, but if it can, this algorithm seems pretty good.

 

[HRTKD]

Interesting approach. I don't think the way I use my batteries would call for such a high charge rate and my charge source that is used 99.99% of the time probably wouldn't provide that much "C" anyhow.

It seems like if they aimed for a slightly lower voltage threshold, they could reduce the temperature impact.

 

[curiouscarbon]

It seems like if they aimed for a slightly lower voltage threshold, they could reduce the temperature impact.

They used some type of air conditioner to keep the chamber at 74F and the batteries reached 77F.

A Memmert environmental chamber was used to maintain a constant ambient temperature of 23 C throughout the test period.

The paper doesn't seem to mention how much power that apparatus uses, so that's also a factor in efficiency for anyone who wants safe fast charging.

Personally, I am still going to try to stick to 0.5C charge and 1.0C discharge as good system design guides. Daydreaming of using two of those overkillsolar 100Ah 3.2V cells as power pack for a solar RC car.... would want that to recharge quickly from grid when needed. Mini supercharger!

Apparently the KillaCycle was made with LFP cells.

The wiki article says it could recharge in 10 minutes.

 

[HRTKD]

The last RC vehicles I played with were the truggies. Any LiFePO4 cells would make them really top heavy. But a low center of gravity race car would be cool.