Backup Battery Testing LiFePO4 vs. AGM

[rv6ejguy]

I spent most of the day making up a test rig to load test my fairly new backup Shorai LiFePO4 battery and for comparison, a 6 year old PC680 AGM.

I made up an array of seven 20 watt Halogen bulbs, which gave a measured current draw of around 11 amps which is close to what a 4 cylinder SDS EFI system draws with one pump running, the ECU, coils and injectors at 2500 rpm.

The Shorai is rated at 18 AH and the PC680 at 16AH.

The first test on the Shorai showed it was pretty much done at the 20 minute mark, getting close to the 12.9V threshold (unloaded)- the point where it's not recommended to discharge below for cell longevity.

The old PC680 soldiered on for over 35 minutes before dropping down into the low 10V range under load (low 11V range unloaded).

10 volts is getting low enough to affect the operation of some components. In particular, the surge current capacity to charge the ignition coils and fire the injectors is diminished. The engine will probably still run down to 9 volts in some fashion though and the ECU doesn't sign off until a bit below 7.5V.

The lower floor level of the hangar was at around 8C during the test to make this a tougher scenario so this may have affected the Shorai more than the Odyssey. I couldn't find any info on how temperature affects capacity on the Shorai but it also affects the PC680 to the tune of 20-30% from the 25C rating.

I want to repeat the Shorai test again when I have time but from this first test, if you are relying on a LiFePO4 for backup power in cooler conditions, you may want to select something larger than 18AH, especially if you have a 6 cylinder system where current draw is higher. In truly cold conditions, you may want to consider an AGM battery instead.

I shot some video which I'll publish at some point when I get time.



Typical LiFePO4 discharge curve



Shorai capacity vs. voltage

As far as our EI systems (CPI/ CPI-2) go, they draw far less current that the full EFI/EI systems so they will run much longer on the same size batteries.

As always, be aware that even colder temperatures, say colder than -20C, severely affect the ability of LiFePO4 batteries to deliver high current for long periods. If you're at altitude on a winter day and the battery is in a location where it's seeing ambient temperatures, don't count on it to last very long if the alternator takes a dump. Also, be aware that batteries lose capacity with the number of cycles they experience. Don't rely on an old, cold battery to keep your engine running.

This test was a bit of an eye opener for me but I'm glad I did it in cool conditions. The lesson is clear here- don't rely on battery AH ratings solely, different chemistries have different characteristics.

[svyolo]

Ross;
Since the discharge rate is fairly high, I wonder if you could have run the lithium battery to a much lower voltage.

I know you don't want to ruin a brand new battery. I wouldn't either. But in an emergency, the battery only has to get me home one time. I would thank the battery, and dispose of it.

Looking at the lithium battery discharge curve, it looks like it should have lasted about an hour. I have seen similar charts before.

[rv6ejguy]

In the last few minutes, the lithium voltage dropped quite rapidly. At 12.9V, there is only 20% remaining so it might have run for another 3-4 minutes at most. You can see the rapid drop off on the curve above. The AGM has a more gradual drop off.

You're correct, since there is no BMS on these, you can run it until dead and I would in a real emergency but I wasn't about to destroy a new battery for the sake of science.

Most battery ratings are done at 25C and they perform slightly better above this temperature and somewhat worse below it. The LiFePO4 seems to be affected relatively more than the AGM in my test and Shorai lists some techniques to improve cold starting so it's obviously a concern.

I had a customer with another brand of LiFePO4 who was stranded twice in Alaska when it wouldn't turn the engine over. He managed to hand prop a cold 400 inch EFI engine to life. Must have had Arnold arms to do that! I guess you do what you have to when you're stranded in the bush.

[Waiex-guy]

What is the significance of the note ?resting voltage? when you are conducting your test under load?

[Waiex-guy]

Also I was under the impression that these batteries warm up to a normal internal temperature quite quickly when under load. I have read the recommendation to operate the landing lights for a minute before cranking in cold weather. Is this not actually effective?

[rocketman1988]

"...The first test on the Shorai showed it was pretty much done at the 20 minute mark,..."

I would be suspect of that battery if the test is accurate...it is performing nowhere near it's rating...

[rv6ejguy]

Remember the ratings are established at 25C. At 11 amps draw, I'd expect it would warm up in a couple of minutes since these are very low mass. This battery weighs only 2.2 pounds.

Experiments often show us things that straight specs don't, hence their value.

I want to repeat the test and see if I get similar results. From the first test, the Shorai didn't perform as well as I expected from the specs.

Your results may vary with other brands but if you're electrically dependent with only one alternator, I suggest you do similar tests at cooler temperatures to verify performance.

I looked at voltage under load as well as with load removed at 5 minute intervals.

The Shorai spins my engine over very well at these temps, maybe 80-120 amps for a few seconds. It seems not to perform so well pulling 11 amps for many minutes however.

[gasman]

With the PC680 considered in the test, maybe the test should be duration per pound.

But then it gets very expensive........

[skylor]

Originally Posted by rv6ejguy

Remember the ratings are established at 25C. At 11 amps draw, I'd expect it would warm up in a couple of minutes since these are very low mass. This battery weighs only 2.2 pounds.

Experiments often show us things that straight specs don't, hence their value.

I want to repeat the test and see if I get similar results. From the first test, the Shorai didn't perform as well as I expected from the specs.

Your results may vary with other brands but if you're electrically dependent with only one alternator, I suggest you do similar tests at cooler temperatures to verify performance.

I looked at voltage under load as well as with load removed at 5 minute intervals.

The Shorai spins my engine over very well at these temps, maybe 80-120 amps for a few seconds. It seems not to perform so well pulling 11 amps for many minutes however.

Ross,

Thank you for running these tests and sharing the data. As I’m sure you are well aware, over the last 4 years there have been a few incidents with Reno racers that involved alternator failures and dead engines. In all of these cases the duration of operation under the backup battery was far less than was expected and resulted in engine-outages of electrically dependent engines. One aircraft made 2 semi-successful dead-stick landings in the same race weekend, another made a precautionary landing on an Indian Reservation road due to rapidly falling voltage, and the third made a hard landing on a runway that resulted in a totaled aircraft but fortunately the pilot walked away. All of these occurred in summer conditions. I think they all involved lithium batteries.

The point here is that someone planning an electrical system for an electrically dependent engine needs to absolutely ensure that they have more than sufficient backup battery capacity and the systems need to be tested. Also, the lithium batteries seem to fall a little short of their rated capacities under EFI electrical loads.

Skylor

[Mike S]

Sure makes me glad I chose dual Odyssey 545s for the new project.