BMS balance wire resistance as proxy for cell impedance and battery life...

[OffgridSystems]

Has anybody seen studies of what the curve of internal cell resistance looks like over the life of a Lithium Ion and Lithium Iron Phosphate (LFP) cell?

According to most papers I'm reading, the end of life comes at 80% of initial capacity value, or internal resistance of 200% of initial value. Also AC impedance can be loosely inferred from DC resistance, that is when one is higher, it's highly likely the other is also.

JK and some of the other BMS makers include the feature of measuring DC resistance (DCR) of the voltage sense wires. If the DCR is plotted over time, will that help to predict a loose approximation of lifespan for Lithium Ion?, Lithium Iron Phosphate?

It is understood that the model for a lithium cell is complex and has many components, inductive, capacitive and resistive, thanks to Keysight Technology:



Here's a plot of lithium cells vs their ac and DC resistance taken from Davide Andrea of Elithium:


So, whatdayathink? Is there a way to get a simple, maybe not super-accurate, but thumbnail view of battery life, already built-in to our BMS boards?

Tim

 

[OffgridSystems]

Quick thought, of course we are trying to pin all the variables to the same value.
I.E. Temperature and state of charge are variables, but for the sake of the argument, we will hold those to the same values, i.e. 65F and 90%.

t

 

[OffgridSystems]

Ok I see a lot of views on the subject, so maybe I should start it off with an example from the JK BMS I'm testing.
On the main screen of the Bluetooth app you get a reading of the "Cells Wire Resistance" along with the cell voltages...



To calculate this value, the JK BMS turns on it's current source, and then measures cell voltage. The cell resistance is R = V/I, where the current flows from the chip on the BMS, through the connectors, the sense wires, and the cell itself. Say we measured a cell's DC resistance when new at 100 milliohms. For our example, let's say the JK reports the measurement at 311 milliohms, we could deduce that .211 ohms is the circuit resistance, and .100 ohm is the cell resistance.


If this value is taken at say 65F and at 100% state of charge, can we then use it for monitoring the life of the cell, by taking yearly measurements at those same parameters? We can't neglect resistance changes to the wiring, but if done properly, and good components are used and kept within their safe operating areas, we might be able to say that at some future resistance value, the cells are at 200% of initial value. We could even do a capacity test to check reality, because the JK BMS has a coulomb counter, and can come close to giving you actual capacity on a 100% to 0% discharge test. Reading a resistance value is a lot easier than doing a capacity test, so if we can use the cell wire resistance reading as a predicter, that would be a time saver.

One way to prove this whole idea is for those of us who have BMS's with the capability, to do cell wire resistance vs. capacity tests over time. It might prove to be a good tool!!

Anybody have thoughts on this? Anyone already done the testing?

Tim

 

[toms]

I use the REC BMS. The calculated cell resistance does vary quite significantly depending on the load on the battery.

I occasionally check the number, more to see if any cells are a long way from average.

 

[OffgridSystems]

Thanks Tom, maybe take some numbers down? What the cells resistance are with no load and some normal connected load, over time? I've got an 8S lithium ion bank I can cycle 100% of capacity daily for a year. That might give some usable data. Yours would have to be over a longer time to make a difference, LFP have thousands of cycles for their cycle life. Does your BMS count cycles? The JK BMS does have a cycle counter, not sure how it works exactly...


t