Have spent the last night reading everything I can about this topic. I have been saying for a few years that calendar aging will degrade our lifepo4 batteries before degradation caused by cycling ever would. I have also found some issues with how manufacturers do their cycle life testing. Would like to share my thoughts on this.
Offgrid Solar Application LiFePO4 Cycling:
Let us consider how a solar lifepo4 battery is used on a daily basis:
What causes degradation from cycling, does not cause degradation for calendar aging. They are independent processes.
Many companies in the solar industry will give cycle life recommendations based on cycle testing alone. Typically, at 1C rates, charging from 100 to 0 to 100, over and over. Sometimes they will modify the cycling threshold SOC, such as a DOD of 20%. This is not a good way to model degradation for solar batteries. The higher C rate will cause significantly more degradation over time (more than a solar battery would ever experience). For most solar systems, the highest C rate they will experience is .2C. This will not generate nearly the same amount of heat, and will reduce the rate of the aging processes significantly.
People will use these cycle life degradation studies to estimate how long their battery will last in years. This is not logical. These cycle life tests are done quickly and at higher internal cell temperatures. And different cycle depths. The degradation from cycling alone in batteries used for solar, will be much lower than any of these studies.
Now let us discuss the rate at which lifepo4 fades whether they are used or not. This is called calendar aging.
This is a fantastic study showing the degradation rate of lifepo4 at various SOC and temperatures: https://www.mdpi.com/1996-1073/14/6/1732
The cells were held at a specific SOC and temperature, and were NOT cycled. They were capacity tested once a month.
This created a measurable capacity fade over time:
This was a 4 year study, so increasing the temperature was a useful way to speed up the calendar aging process. In a solar system that is at a lower temperature, the degradation rate will be slower, but it will still follow the rate that was observed here. And you can still model and predict long term calendar aging using these data points.
Now consider that this capacity fade will occur whether we are cycling the cells or not. Just the cells simply existing at temperatures commonly found on our planet, will create a measurable capacity fade over time.
They then used this data to create a model that can predict long term calendar aging (there seems to be an error for the yellow corner. Not sure what happened there):
Now with a solar system, we are hanging around 50% state of charge. So take a look at the graph where it shows 50% SOC and 25 degrees C. The model estimates that you will get 23.8 years till degradation to 80%.
Currently, batteries are rated in cycle life to 80% degradation. Let us use the power pro battery cycle life estimate of 8000 cycles at .5C with a DOD of 80%. If your solar system is sized properly (back up power for at minimum one day), you will never do a 80% DOD, and you will never charge at .5C rate. That means the cycle life degradation processes alone should be less than what they are predicting. You should easily get many more cycles than 8000 if you are using your system for solar.
Unfortunately, we do not have lower C rate cycling studies with 100% DOD. This is an area that we need more research on. Given the difference in C rate and how much heat generation can occur, this can change the cycle life degradation rate significantly. And in the field, most systems' degradation typically follow the calendar aging models.
The temperature that these cells are operating at can change the degradation rate significantly. And running our systems at a low c rate, with shallow cycles daily, means that cycle degradation will be extremely low. In most systems that I have run, it is not measurable. And the degradation rate follows calendar aging.
Okra Solar Pty Ltd. Actually has a fantastic blog post covering this:
.png)
www.okrasolar.com
www.okrasolar.com
www.okrasolar.com
www.okrasolar.com
www.okrasolar.com
Source:
www.okrasolar.com
Next up, I found it interesting that we always suggest to hang around 50% state of charge for maximum life. And that is true when it comes to degradation from cycling, but not true if you consider degradation from calendar aging. When you have a cell at 90% SOC for a prolonged duration, there will be a thickening of the SEI layer, and this will prevent some aging processes from occurring as fast. So the curve of degradation will actually flatten and over a few years, will be less than keeping your pack at 50%! That is crazy!
Check it out:
This effect will occur even faster at 100% SOC. Here is an explanation from the study:
Seems like going to a high SOC with a solar battery is not so bad! Unless you are doing fast charging and discharging at a high SOC. Then it will increase the cycling degradation. But it reduces the calendar aging degradation rate over time. Which in the long run, will be a larger factor to consider. Especially in higher temp environments.
Now this study states this in plain english:
Going by predictive models of cycle stability, using a lifepo4 battery for solar is not going to be a "limiting factor" here. And they also state that the calendar aging in these systems with high cycle count at low rates is the dominant contributor to degradation.
Read about it here: https://www.sciencedirect.com/science/article/abs/pii/S2352152X18300665
Now there was a recent study that mentioned cycling degradation processes being higher when cycled at high SOC. But like we found out a minute ago, calendar aging is reduced when battery is held at high SOC:
So for cycling degradation at higher rates, which this study is for ev's, not solar, you want to avoid high SOC. But for a solar system that's primary degradation process comes from calendar aging, and calendar aging being less severe when battery is held at high SOC, I don't think it really matters. Even going from 0-100% should do moderate cycling damage in this study.
Now what can be concluded from all of these studies is that the hotter the battery is while it is being stored or cycle, the faster the degradation from cycling or storage. Both degradation processes are independent of each other, but still are affected by heat.
If you want your battery to last a very long time, you need to keep it in a cool environment. But going by the long term cycling degradation rates and cycling threshold data, I do not see any reason to modify your cycling thresholds.
What we need next in the future is a study showing the degradation of solar batteries at 25 degrees C, being cycled at .1 C, with partial daily cycles. Just like a normal solar system experiences.
Here is another useful article to reference:
Covering a multitide of factors for long life with lifepo4:
But again, some of this article goes against what is found in the calendar aging studies. So this seems to focus on cycling related degradation processes.
Conclusion:
I think my recommendation of charging to 100% for balancing and to use the total capacity of the battery, avoid extreme temperatures and install batteries indoors, and to not be afraid to DOD to 0% when necessary because cycling degradation is minimal, is a good recommendation for lifepo4 when used for solar.
If anyone would like to share studies that disagree, please share them below. Thank you
Offgrid Solar Application LiFePO4 Cycling:
Let us consider how a solar lifepo4 battery is used on a daily basis:
- Most battery banks are being cycled at low C rates. Typically .1-.2C
- If the battery bank has any amount of back up power (or "days of autonomy"), the battery will not complete full DOD cycles daily. It will do a partial cycle, typically 50% for an offgrid system, and sometimes much less than that. The depth of discharge also changes depending on season and total load consumption.
- The temperature of these batteries is relatively cool. Typically, solar batteries are cycled indoors or in a shaded area (most outdoor listed batteries recommend this, and you will get over temperature disconnect if overheated). This changes the cycling degradation rate significantly. Less heat = less degradation.
- A solar system is not hanging out at excessively high or low SOC for very long. Typically it hangs out around 50%. This can cause problems with balancing, which can detrimentally affect cell degradation rates across the pack. And limit performance and capacity if imbalance is excessive.
- Cycled at high C rates from 1C to 4C. Capacity then measured at this rate. This higher C rate will create more heat in the cell, and faster degradation.
- Cycled nonstop between various SOC thresholds. 100% to 0%. 100% to 20% and so on.
- The ambient temperature in the lab is controlled, but there is no way to assess internal cell temp. Models have been created, but this is very tricky to do. Considering the rates of charge and discharge for these tests, there should be an effort to keep the cell's internal temp regulated.
- The cycles are rapid and continuous. There is no moment that the pack is hanging out at a specific SOC. It is either charging or discharging. Because the cycle tests hit 100%, there is no issue with balancing. Each cycle causes a small amount of balancing to occur that keeps the cell drift at a minimum.
- Cells are not being cycled. They are held at a specific SOC and temperature, and then capacity tested once a month. The higher the temperature, the faster the calendar aging degradation. Different SOC can have an effect, but that is more complex. Will come back to this later.
Reference study: https://www.mdpi.com/1996-1073/14/6/1732
What causes degradation from cycling, does not cause degradation for calendar aging. They are independent processes.
Many companies in the solar industry will give cycle life recommendations based on cycle testing alone. Typically, at 1C rates, charging from 100 to 0 to 100, over and over. Sometimes they will modify the cycling threshold SOC, such as a DOD of 20%. This is not a good way to model degradation for solar batteries. The higher C rate will cause significantly more degradation over time (more than a solar battery would ever experience). For most solar systems, the highest C rate they will experience is .2C. This will not generate nearly the same amount of heat, and will reduce the rate of the aging processes significantly.
People will use these cycle life degradation studies to estimate how long their battery will last in years. This is not logical. These cycle life tests are done quickly and at higher internal cell temperatures. And different cycle depths. The degradation from cycling alone in batteries used for solar, will be much lower than any of these studies.
Now let us discuss the rate at which lifepo4 fades whether they are used or not. This is called calendar aging.
This is a fantastic study showing the degradation rate of lifepo4 at various SOC and temperatures: https://www.mdpi.com/1996-1073/14/6/1732
The cells were held at a specific SOC and temperature, and were NOT cycled. They were capacity tested once a month.
This created a measurable capacity fade over time:
This was a 4 year study, so increasing the temperature was a useful way to speed up the calendar aging process. In a solar system that is at a lower temperature, the degradation rate will be slower, but it will still follow the rate that was observed here. And you can still model and predict long term calendar aging using these data points.
Now consider that this capacity fade will occur whether we are cycling the cells or not. Just the cells simply existing at temperatures commonly found on our planet, will create a measurable capacity fade over time.
They then used this data to create a model that can predict long term calendar aging (there seems to be an error for the yellow corner. Not sure what happened there):
Now with a solar system, we are hanging around 50% state of charge. So take a look at the graph where it shows 50% SOC and 25 degrees C. The model estimates that you will get 23.8 years till degradation to 80%.
Currently, batteries are rated in cycle life to 80% degradation. Let us use the power pro battery cycle life estimate of 8000 cycles at .5C with a DOD of 80%. If your solar system is sized properly (back up power for at minimum one day), you will never do a 80% DOD, and you will never charge at .5C rate. That means the cycle life degradation processes alone should be less than what they are predicting. You should easily get many more cycles than 8000 if you are using your system for solar.
Unfortunately, we do not have lower C rate cycling studies with 100% DOD. This is an area that we need more research on. Given the difference in C rate and how much heat generation can occur, this can change the cycle life degradation rate significantly. And in the field, most systems' degradation typically follow the calendar aging models.
The temperature that these cells are operating at can change the degradation rate significantly. And running our systems at a low c rate, with shallow cycles daily, means that cycle degradation will be extremely low. In most systems that I have run, it is not measurable. And the degradation rate follows calendar aging.
Okra Solar Pty Ltd. Actually has a fantastic blog post covering this:
A Better Way to Estimate Battery Lifetime
www.okrasolar.com
A Better Way to Estimate Battery Lifetime
www.okrasolar.com
A Better Way to Estimate Battery Lifetime
www.okrasolar.com
A Better Way to Estimate Battery Lifetime
www.okrasolar.com
A Better Way to Estimate Battery Lifetime
www.okrasolar.com
Source:
A Better Way to Estimate Battery Lifetime
www.okrasolar.com
Next up, I found it interesting that we always suggest to hang around 50% state of charge for maximum life. And that is true when it comes to degradation from cycling, but not true if you consider degradation from calendar aging. When you have a cell at 90% SOC for a prolonged duration, there will be a thickening of the SEI layer, and this will prevent some aging processes from occurring as fast. So the curve of degradation will actually flatten and over a few years, will be less than keeping your pack at 50%! That is crazy!
Check it out:
This effect will occur even faster at 100% SOC. Here is an explanation from the study:
Seems like going to a high SOC with a solar battery is not so bad! Unless you are doing fast charging and discharging at a high SOC. Then it will increase the cycling degradation. But it reduces the calendar aging degradation rate over time. Which in the long run, will be a larger factor to consider. Especially in higher temp environments.
Now this study states this in plain english:
Going by predictive models of cycle stability, using a lifepo4 battery for solar is not going to be a "limiting factor" here. And they also state that the calendar aging in these systems with high cycle count at low rates is the dominant contributor to degradation.
Read about it here: https://www.sciencedirect.com/science/article/abs/pii/S2352152X18300665
Now there was a recent study that mentioned cycling degradation processes being higher when cycled at high SOC. But like we found out a minute ago, calendar aging is reduced when battery is held at high SOC:
So for cycling degradation at higher rates, which this study is for ev's, not solar, you want to avoid high SOC. But for a solar system that's primary degradation process comes from calendar aging, and calendar aging being less severe when battery is held at high SOC, I don't think it really matters. Even going from 0-100% should do moderate cycling damage in this study.
Now what can be concluded from all of these studies is that the hotter the battery is while it is being stored or cycle, the faster the degradation from cycling or storage. Both degradation processes are independent of each other, but still are affected by heat.
If you want your battery to last a very long time, you need to keep it in a cool environment. But going by the long term cycling degradation rates and cycling threshold data, I do not see any reason to modify your cycling thresholds.
What we need next in the future is a study showing the degradation of solar batteries at 25 degrees C, being cycled at .1 C, with partial daily cycles. Just like a normal solar system experiences.
Here is another useful article to reference:
Covering a multitide of factors for long life with lifepo4:
But again, some of this article goes against what is found in the calendar aging studies. So this seems to focus on cycling related degradation processes.
Conclusion:
I think my recommendation of charging to 100% for balancing and to use the total capacity of the battery, avoid extreme temperatures and install batteries indoors, and to not be afraid to DOD to 0% when necessary because cycling degradation is minimal, is a good recommendation for lifepo4 when used for solar.
If anyone would like to share studies that disagree, please share them below. Thank you
Attachments
-
FireShot Capture 510 - The Degradation Behavior of LiFePO4_C Batteries during Long-Term Cale_ ...png -
FireShot Capture 503 - A Better Way to Estimate Battery Lifetime - www.okrasolar.com.png -
FireShot Capture 498 - The Degradation Behavior of LiFePO4_C Batteries during Long-Term Cale_ ...png -
FireShot Capture 500 - A Better Way to Estimate Battery Lifetime - www.okrasolar.com.png
Last edited:
I need to get down into the evil bottom discharge level 
