Tenergy batteries

[karrak]

From the research I have done, as far as I can see at low charge and discharge rates the DOD that you cycle LFP batteries at makes little difference to the total amount of energy that you can cycle through the battery during its lifetime.

The following graph from this research paper shows this.
LFPCapacityFade.jpg

Capacity Fade as a function of cycled DOD
Days is analogous to Total Ah throughput

Simon

[extrafu]

Maybe someday a battery will be able to live for 4000 cycles using 80% of its nameplate charge but as far as I have heard no one has made a battery yet to live that long under "real life" conditions.

Also, Sony's Fortelion cells (LiFePO4) are supposed to keep 80% of their capacity after 6,000 cycles (http://download.solarshop.net/englis...10-08-2012.pdf). I don't know if they are full of it but this is rather impressive.

[extrafu]

From the research I have done, as far as I can see at low charge and discharge rates the DOD that you cycle LFP batteries at makes little difference to the total amount of energy that you can cycle through the battery during its lifetime.

Not sure I grok this. If the battery holds pretty much the same capacity no matter what the DOD you used between cycles, you obviously got a lot more energy from the battery at 80% DOD than at 10, no?

[karrak]

I am becoming more and more confident that a lifespan of 10 years for LFP batteries in off-grid/grid-tied systems where the charge and discharge currents are less than C/2 is more than likely.

steveg who posts on this forum has had his LFP battery for nearly six years with no change in performance. If you are considering using an LFP battery i would recommend reading his posts.
Barba is another forum poster who has been using LFP batteries for more than six years.

In Australia there are a number of multinational and smaller companies offering LFP batteries for off-grid and grid tied systems with 10 year warranties.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger https://github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

[karrak]

Not sure I grok this. If the battery holds pretty much the same capacity no matter what the DOD you used between cycles, you obviously got a lot more energy from the battery at 80% DOD than at 10, no?

The total amount of energy cycled trough the battery is equal to the amount taken on each cycle times the number of cycles. So you might get 1,100 cycles out of an LFP battery if you cycled 90% of its rated capacity whereas you will get 10,000 cycles if you cycled 10% of its rated capacity.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger https://github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

---

[inetdog]

The total amount of energy cycled trough the battery is equal to the amount taken on each cycle times the number of cycles. So you might get 1,100 cycles out of an LFP battery if you cycled 90% of its rated capacity whereas you will get 10,000 cycles if you cycled 10% of its rated capacity.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger https://github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

---

+1
There is a lot of discussion and not as much good data on "microcycles" and how they affect the life of a battery, both Lead and Lithium based.
For lead acid it is complicated by the fact that you cannot leave a cell in a partial state of charge for any length of time without degradation.
For lithium chemistry it seems reasonable to limit the number of small cycles by not recharging every day, but that is counterproductive when charging from solar energy.

My best guess is that a "cycle" on the life test graphs corresponds exactly to the charge discharge regime specified in the test, and that one long cycle every four days is NOT as bad as four smaller cycles every day. And that many partial charge/discharge cycles during the course of a continuous day of charging but intermittent heavy day loads is worse than the equivalent energy transfer in one daily cycle (corresponding to most loads happening during the night.)

[extrafu]

My best guess is that a "cycle" on the life test graphs corresponds exactly to the charge discharge regime specified in the test, and that one long cycle every four days is NOT as bad as four smaller cycles every day. And that many partial charge/discharge cycles during the course of a continuous day of charging but intermittent heavy day loads is worse than the equivalent energy transfer in one daily cycle (corresponding to most loads happening during the night.)

How would you manage this with solar?

[SunEagle]

How would you manage this with solar?

The only way to reduce the number of cycles in any time period would be to oversize your battery system so you only use a small % each day and then charge it back to 100% a few days later. IMO while that may extend the life of the battery system your up front costs will be much higher (this includes pv wattage & CC sizing) and harder to economically justify.

Why pay for a $100,000 system that lasts 10 years or a $40,000 system that last 5 years? Doesn't make dollars and cents to me.

[extrafu]

The only way to reduce the number of cycles in any time period would be to oversize your battery system so you only use a small % each day and then charge it back to 100% a few days later. IMO while that may extend the life of the battery system your up front costs will be much higher (this includes pv wattage & CC sizing) and harder to economically justify.

Yep, that I know - I should have been clearer for my question. My question was more around the automation of that. Say it takes 3 days to go down to 80% DOD. The fourth day, the batteries get used and fully charged. Should the charge controllers simply be disconnected from the battery bank for 3 days or can this be automated? I don't think the HVD would help here, since when charging the batteries, the HDV would have to be set at a higher value.

Thanks!

[karrak]

I think the graph that I posted earlier is saying that it makes very little difference if you do micro cycles or large cycles to the lifespan of the battery. All the test results in that graph were obtained by charging and discharging individual batteries at the same current but to different DODs on a continuous basis. That is why the bottom axis is measured in days. All the batteries had exactly the same amount of energy cycled through them and all suffered nearly the same loss in capacity regardless of how big each cycle was.

From research I have done, my experience and the experience of others I think the best way to charge an LFP battery off solar is to charge it at a maximum charge voltage of around 3.45V per cell until the charging current drops to less than C/50 and then to float them at around 3.35V per cell for the rest of the day. If you have enough sun during the day this will result in a battery being nearly 100% full at the end of the day and the majority of your load being powered by the solar panels thus minimising the energy cycled through the battery.

If you need to run a generator, wait until the battery drops to 10%-20% SOC and then add around 10%-30% from the generator so the sun will do most of the work. The generator does not have to be very large as LFP batteries can be charged at any rate.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger https://github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

[SunEagle]

I think the graph that I posted earlier is saying that it makes very little difference if you do micro cycles or large cycles to the lifespan of the battery. All the test results in that graph were obtained by charging and discharging individual batteries at the same current but to different DODs on a continuous basis. That is why the bottom axis is measured in days. All the batteries had exactly the same amount of energy cycled through them and all suffered nearly the same loss in capacity regardless of how big each cycle was.

From research I have done, my experience and the experience of others I think the best way to charge an LFP battery off solar is to charge it at a maximum charge voltage of around 3.45V per cell until the charging current drops to less than C/50 and then to float them at around 3.35V per cell for the rest of the day. If you have enough sun during the day this will result in a battery being nearly 100% full at the end of the day and the majority of your load being powered by the solar panels thus minimising the energy cycled through the battery.

If you need to run a generator, wait until the battery drops to 10%-20% SOC and then add around 10%-30% from the generator so the sun will do most of the work. The generator does not have to be very large as LFP batteries can be charged at any rate.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger https://github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controller

So where can someone get these LiFeYPO4 batteries?

[Barba]

This test is impressive guys!!!
Not really in topic but impressive

[karrak]

Thanks Barba, This is impressive. The one important point is that they cooled the battery to keep its temperature at 25C.

[SunEagle]

This test is impressive guys!!!
Not really in topic but impressive

Ok so some German research lab conducted tests on a battery system. Who makes that battery and how much is it?

[extrafu]

So where can someone get these LiFeYPO4 batteries?

IIRC, Thundersky/Winston batteries are LiFeYPO4 ones. They are a bit hard to find in the US, impossible in Canada, easy elsewhere.

[SunEagle]

Ok. They do not look cheap. What is a rough cost for their LYP 700Ah or 1000Ah battery?