Self Discharge Rate Testing

[IamIan]

Here is some data / information for those who might be interested.

For those who don't already know.

One of my hobbies for many years now has been rechargeable batteries.

For the last 3 years or so I have focused mainly on the NiMH batteries used in HEVs... with a particular focus on the NiMH in my I1 ... as that is my own personal vehicle.

I have about 3 or 4 more months of long term SD rate testing to do. Then I think I might go back and do some additional follow up testing on the early days of the SD rate.

The trend in the long term seems to have leveled out and stabilized... So I am not anticipating significant change on that end.

The bellow graph shows the Daily SD rate in mAh.

It is at its highest at those days earliest after charging ... the rate of SD seems to decay logarithmically over time... of course even though the rate continues to slowly reduce over time ... the cumulative SD continues to add up.

The data I have collected thus far seems to yield a predictability of about 82%... I suspect that the largest gains in improving this predictability will be from filling in additional data points in the earlier days of the testing... that is where the current data points are the most chaotic.

--------------------

Methodology Overview:

Perform cycle testing to verify what the mAh of storage capacity for a specific 6 cell NiMH sub-pack stick is.

From an initially discharged state fully charge a sub-stick.

Wait a set period of time with the sub-pack stick in a consistent temperature location.

Discharge the sub-pack stick after the waiting period is complete. Record the Ah given out from that sub-pack stick to a discharged state.

--------------------------

Areas of potential further investigation / interest:

#1>
Increasing the accuracy of the expected daily SD rate predictability.

#2>
Compare these SD Rate tests to those from an initially partially charged state.

#3>
Although temperature effects are well documents to effect the SD rate... quantifying those effects on these specific batteries might be desirable.

[3-Wheeler]

IamIan,

I can add some input to some of your self discharge data.

My car sat the entire summer while working on improved under-body smoothing panels. I have charged the HV battery pack twice in this time.

The first time was about two months ago, where the battery pack was down to 152 volts DC and the battery was charged back up to a maximum of 165 volts according to one of my DVM's. The average trickle rate was close to 200 mA or so and then tapering off to 100 mA when the charge rate stopped dropping. I used a variac charger which had to be adjusted often.

The second time, the battery was one day ago, at 148 volts after not seeing any action for two months, and again trickle charged at the same 200mA rate until about 160 volts or so, then reduced the charge to 100mA or so.

Hope this helps, Jim.

[IamIan]

Jim, Thanks for the input... Just remember, what you put in is not necessarily what the battery has lost... that's why my testing method is to test it on the discharge side instead... the battery can't give me more than 100% of what it can store ... but the charger can apply / pump out more than 100% of what the battery can store or than what was lost.

--------------

I did also want to mention that the SD rate graph will probably end up being updated as I gather and add additional details.

Some will come sooner ... as some of the testing is already underway and closer to completion ... some might take a while longer.

I will also eventually be posting more details about the other battery testing I've been doing that are also not yet complete.

Yes... I expect maybe 1% will find these things interesting or useful... But I'm in that 1% and I haven't seen it posted yet ... so there you go.

[retepsnikrep]

Ian

I'm sure quite a few of us are interested, I am, and more will become so especially as our NIMH packs continue to age.

Any extra data on out packs is worth having.

Thanks

Peter

[E27006]

I notice that weak sticks can not only self-discharge at an unsatisfactory high rate but they can also display a drastic reduction in capacity after a few weeks of use.

I.e after pack rebalance, a weak stick which, after reconditioning, was giving a reliable 7000 mah capacity on the Robitronics charger had dropped to well below 3000 mAh after eight weeks use in the car.

The original 7000 Mah capacity of the stick returned after a second session of cycling on a Robitronics charger.

Some bench testing by trickle charging the pack with a grid charger ( 100 to 300 milliamps ) followed by a 10 amp discharge / 8 amp recharge test on the weak sticks has given some hopeful results with regard to stabilising the capacity drop phenomenon.

I will be refitting the pack, having adapted the junction board to permit either in-car trickle charging of the whole pack or high power cycling (10 amps) of the weak sticks.

I look forward to road testing the pack during the coming weeks

[IamIan]

While it is very true that there are a variety of issues concerning the batteries ... I thought it would end up getting confusing for people to follow if I started putting it all in one thread... I guess the other side of it is that if it were one thread it would all be on one place... but I still think that might end up being a bit more confusing to follow.

I look forward to the community having a detailed understanding on file of the batteries even if / as people replace them with other batteries ... SD , Internal Resistance, Impedance, Voltage depression, Reconditioning cycles, Over charging, Over Discharging, Cell Reversals, Cycle life, Service Life, Catastrophic failure, etc... sure some of us know the general NiMH issues and effects of these things ... but having some better quantified results might be interesting / useful.

[Hybrid-Battery-Repair]

Originally Posted by E27006

I notice that weak sticks can not only self-discharge at an unsatisfactory high rate but they can also display a drastic reduction in capacity after a few weeks of use.

I.e after pack rebalance, a weak stick which, after reconditioning, was giving a reliable 7000 mah capacity on the Robitronics charger had dropped to well below 3000 mAh after eight weeks use in the car.

The original 7000 Mah capacity of the stick returned after a second session of cycling on a Robitronics charger.

Unfortunately, you have two variables there. You are driving the car, so that stick might be low because of SD or IR or other reasons. In order to know, you need to test SELF discharge, not the induced discharge that you are giving it. That means charge it and let it sit. I perform a 7 day "sit" test on all the packs I work on. I top up all the batteries and then let them "sit" for a week. After that I top them up again. Now and then I'll find a stick that has discharged 50% or even 100% in that week. As Ian points out, this is not an accurate measurement of the loss, but a reasonable approximation. If I see 1400, 1500, 1270, 1350, 3500, 1600, 1350, 1500 I know that I've got a serious problem with the odd stick. Likewise: 1400, 1500, 1270, 1350, 650, 1600, 1350, 1500 is the opposite problem (high IR - probably).

You might want to try it to see what your results are.