Battery Cycle Counting

Hello,
 I am modeling a PV + BESS project using the manufacturer's degradation curve.

System details: 400 MWh capacity / 100 MW PCS.

The Issue: I am seeing a large discrepancy in cycle counts. My system throughput is roughly 100 MWh/day, which implies ~0.25 Equivalent Full Cycles (EFC) per day. However, SAM reports approx. 45 cycles for the first month (approx. 1.5 cycles/day).

Question: I understand SAM uses the Rainflow cycle counting algorithm. Is it normal for the Rainflow method to produce a count this much higher than a standard throughput calculation?
 Appreciate your advice in advance.

Hi Tanatip,

Yes, the rainflow algorithm is counting cycles correctly, and the higher cycle count than standard throughput is how we expect the algorithm to work.

We’ve circled a few cases in your plot that show more than one cycle per day:

 

This raises the question of how best to implement a manufacturer’s degradation curve in SAM's cycle degradation algorithm. We’d expect a rainflow degradation number to have 4-6x the number of cycles relative to full equivalent.

Do you have a battery specification sheet and .sam file that you would be willing to share so we can investigate further?

We've had some internal discussions about transitioning the battery model to full equivalent cycles, so this is an area of interest to use for improving SAM's battery life model.

Best regards,
Paul.

Hi Paul,

Thank you for the confirmation.
To support your investigation, I have provide a degradation data as belows:CycleSOH (%)100096.0200092.5300090.5400088.0500085.5600083.0700081.0800077.5850075.0900072.0950067.01000060.5
Note: Based on 100% DoD FEC, 0.25 C-rate, 25deg.

As you noted, SAM’s rainflow algorithm creates a significantly higher number of cycles than the FEC method. This makes it difficult to input the table above directly into SAM.
Do you have any suggestion on how we should implement it?

And when does the SAM team plan to change the battery model to full equivalent cycles?

Thank you.

Best regards,
Tanatip.

Hi Tanatip,

The data you shared is hard to read. Could you reformat it and explain it more?

Best regards,
Paul.

Hi Paul,

My apologies for the confusion. It looks like there was a formatting issue when I pasted the table from Excel.
Please find the corrected version below.

Cycle: 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000
SOH (%): 96.2, 92.5, 90.5, 88.0, 85.5, 83.0, 81.0, 77.5, 75.0, 71.5, 67.0, 60.0

Note: Based on 100% DoD FEC, 0.25 C-rate, 25deg, 1 cycle / day

Thank you.
Best regards,

Hi Tanatip,

Thank you for the additional information.

I created a feature request on the SAM GitHub repository. We do not have the resources to add this to our current work plan, but may in the future:  github.com/NatLabRockies/SAM/issues/2181 .

Best regards,
Paul.