Battery (Table-based) - Aging and SoH Estimation

I have a question regarding the aging model of the table-based battery.

I'm using a matrix to define the temperature-based aging of the battery as in the picture below.

Is therefore the defined SoH absolute and will take the value as defined by the matrix? For instance, I'm cycling the battery at 0°C and the fade is -4% after 100 cycles. When the temperature of the ambient changes now to 30°C, will the fade jump to -8% (although it has been aging at 0°C), since its defined as in dAH? Or will the fade stay at -4% and the battery now fades as respectively to the 30°C?

I'm currently working on a battery aging model and am kinda struggling to understand the tabel-based model.

Thanks in advance.

[dpb]

Run a couple of tests and see...but one would presume the table lookup is returned for the instant in the simulation so it's starting point is from the previous timestep solution.

Answer 1

[Lorenzo]

Hello Vishal,

Quick answer is: the temperature could in the current implementation revert fading mechanism (i.e. you could reobtain capacity). To understand how this work I habe provided a model (attached) that I explain below.

The block you are using does not show the source code. I would advise you to use the Battery Equivalent Circuit Block which is equivalent but gives you the additional possibility of visualizing the source code (so you could get an answer of your question quicker).

To answer your question I have built a model as shown in the figure below:

As you can see the cell is not being charged nor discharged, so we can exclude any influence from current on the fade. With the temperature source I will gradually increase the temperature of the cell. The fade is parametrized in such a way, that only the temperature impacts on the capacity (so that we can exclude any other factor). My parametrization is shown below:

This means that at a temperaure of 298.15 we assume 0% change of the capacity while at 315.15 we assume 10%. Outside of these ranges we will not interpolate but use the nearest value among the breakpoints. The results of the simulation are shown in the figure below. The cell starts at 293.15 and I represented the temperature and SOC over time (please again note we are just heating the cell, no current flow).

So let's see what happens:

1) For temperature below 298.15 there is no fade as we use the nearest point (which is 298.15 and assume a 0% fade)

2) For temperature between 298.15 and 313.15 the fade is activated we see that the SOC increases

3) For temperatures above 313.15 the fade does not change as we use the nearest point which is 313.15 and a constant capacity loss of 10% (does not get higher than that)

So why does the SOC change? Because the SOC is here defined as the charge divided by the "aged charge". The charge of the cell does not change (cell is not being discharged) but the aged charge decreases by up to 10% following the breakpoints we assigned. In fact, the final value of the SOC is 1.1111 which is exactly 1/0.9 (as in the end I have only 90% of capacity left).

If you want it would be possible using the Battery Equivalent Circuit to go to the source code and implement your own aging model!

Best,

Lorenzo