Calendar vs Cycle Battery Degradation

I was hoping to get a better understanding of how the SAM integrates the two types of battery degradation, cyclic and calendar. Is the overall degradation taken to be the maximum degradation level between the two, or is there some weight applied to each? I have run simulations with standalone batteries where only cycle degradation seems to be a factor (typically in cases with regular daily cycling and high DoD). Conversely, I have run simulations with almost no regular cycling and the degradation there seems to primarily follow the calendar degradation regime. There are in-between cases with low-DoD daily cycling where the overall degradation appears to switch between a linear (cyclic) and exponential (calendar) degradation regime (see attached screenshot). I would like to understand better how the model determines which type of degradation to apply at a given timestep.

Thank you,

Marc

Hi Marc,

SAM reports a "battery capacity percent for lifetime" value in each time step of the simulation. This value shows how the battery capacity degrades over time and is restored when the batteries are replaced.

There are two options for modeling battery capacity degradation on the Battery Storage input page: The cycle degradation model reduces the available capacity as the number of charge and discharge cycles increases. This is the default option and is always enabled. (You can effectively disable it by modifying the Capacity vs Depth-of-discharge table so that the capacity is always 100%.) The calendar degradation model reduces available capacity as the age of the battery increases, and offers an option for lithium-ion batteries that accounts for the effect of temperature on battery life.

When you enable the calendar degradation model, for each time step of the simulation, both the cycle and calendar models apply, and SAM chooses the minimum "battery capacity percent for lifetime" value so that worst-case degradation estimate dominates.

If you are interested, you can see where that happens in the C++ source code on GitHub.com -- it's in the runLifeTimeModels() function around Line 788: github.com/NREL/ssc/blob/develop/shared/lib_battery.cpp . That function is called in github.com/NREL/ssc/blob/develop/ssc/cmod_battery.cpp .

Best regards,
Paul.

When you enable the calendar degradation model, for each time step of the simulation, both the cycle and calendar models apply, and SAM chooses the minimum "battery capacity percent for lifetime" value so that worst-case degradation estimate dominates.

Hi Paul, it's been a while since the original post but I would like to continue this discussion a bit further.

First, I would like to confirm my general understanding that - battery degradation is affected simultaneously by both charge/discharge cycles and battery age. So I would like to know why the custom 'Cycle and Calendar Degradation' mode under "Battery>Battery Life>Battery Life Options" section is resulting in minimum of what either of those models results in.

Simulating an LTO battery using 'Li-ion LMO/LTO' model under the same section, results in the degradation being close to the product of the individual degradations.

I have attached simulation results of LFP chemistry with default cycle and a custom calendar degradation to highlight my point. And LTO chemistry with built-in LTO model. I look forward to hearing from you and hoping for some clarity.

Hi Yashwanth,

The original implementation of SAM's Li-ion battery life model is described in Smith, K.; Saxon, A.; Keyser, M.; Lundstrom, B.; Cao, Z.; Roc, A. (2017). Life Prediction Model for Grid-connected Li-ion Battery Energy Storage System. Presented at 2017 American Control Conference. ( PDF 1.4 KB ). We have made some updates to the model since the original implementation, but the general theory of the model as described in the paper is still relevant. You can find references to other papers documenting the battery model on the Battery Publications page: sam.nrel.gov/battery-storage/battery-publications.html .

Section III-A of Smith (2017) describes "calendar life capacity fade." The mechanisms that cause capacity fade when the battery is in storage, or not being cycled, are different from the mechanisms when the battery is experiencing charge and discharge cycles. The two types of capacity fade are not necessarily happening simultaneously.

When you choose the "Cycle and calendar degradation" option on the Battery Life page SAM calculates capacity fade caused by both, and as a conservative assumption, applies the smaller percentage. In the "LFP Chemstry" case that you shared, SAM applies capacity fade due to battery cycling at the beginning of the battery's life. As the battery ages, calendar life capacity fade applies:

 

For the LMO/LTO battery, I would recommend using the "Li-ion LMO/LTO" battery life option on the Battery Life page instead of the "Cycle and calendar degradation."

Best regards,
Paul.

Hi Paul,

Thanks for the quick reply. I understand your point in regards to LFP case.

For LTO, I am already using the "Li-ion LMO/LTO" model in battery life options. The screenshot I have posted in my earlier post depicts the simulated result. The overall degradation (coloured blue) appears to be the product of the individual cycle degradation and calendar degradation curves. These series are selectable in Time Series section of the results, even though there are no user inputs pertaining to them when using the LMO/LTO model.

Thanks & Regards,
Yashwanth

Hi Yashwanth,

When you choose the LMO/LTO battery life option:

 

The time series results should show capacity fade due to cycling at a constant 100% so that the battery capacity fade is represented by the calendar capacity fade:

 

If you are seeing different results, please attach a copy of your .sam file so I can investigate.

Best regards,
Paul.