Replaced 1 bad cell #9

You should also rebalance all of the modules (using a charger/rebalancer) and set all of the modules to the same voltage before buttoning up the battery if you want your repair to last a while.

JeffD

 

I just serviced a car that had a DIY repair job done by the prior owner/shade tree battery mechanic before it was traded in.

They did not follow instructions.

They tried to separate the felt from the plastic on the side trim panels, rather than removing the three bolts first and then unclipping it.

They removed the battery floorboard cover without removing the chrome luggage tie-downs -- i.e they broke the corners off.

The modules on the passenger side were not lined up correctly, meaning that the pressure reinforcements were not in the right places. The bus bars were bent, and many nuts had been cross threaded. See below.

The middle modules were very low capacity, as low at 700 mAh. This is the chronic overheating failure mode common in hot climates and/or clogged hv battery fan, where a module in the middle blocks (5-10) fails. The module that failed was in the block pair adjacent (#8) to the pair with the newly installed module (#9).

Battery repair is straight-forward, but it is not as simple as replacing 1 module.

 

I hope it works for you. My post was a cautionary tale about what can happen, in hope that I don't see another one of these in my shop.

 

please correct me if I am wrong -- that a typical recharger a DIY'r would have at home can handle one module at a time ? 28 cycles at a day a module = a month of balancing work.

Is this right ? That is a whole lot of patience, particularly if the car is a daily driver before it conked.

 

Yes, it did require me to patiently rebalance my 27 modules with my Supermate DC6 charger/rebalancer. Of course another $35 (for a second DC6 and 12v supply) cuts the time in half and I did this on my old battery after installing a ReInVolt in my 2004 Prius at 195k miles.

Seilerts (a battery re-builder in Santa Fe NM) has 14 Supermate DC6 units so that he can do a HV battery in about 3 days (assuming all of the modules are good).

My plan upon another battery failure (if my 2004 Prius lasts longer than the ReInVolt) would be to get a low mileage, salvaged 2010 battery (or newer - about $800 including a couple of spare modules) and rebuild that one since a Prius is still drivable for a while after symptoms appear and then swap in the rebuilt battery. That would yield me a large number of modules (up to 27) that can be rebalanced and sold on eBay for about $25 each (clearing over $600). Then the total cost of the repair would be about $300 (including $100 for 3 more DC6 units to save time).

JeffD

 

Good evening guys

got a question

I have a 2006 Prius with 149.000 miles, check engine came up saying Replace Hybrid Battery pack, so i tested the cells 27 out of 28 show 8.00-8.01 V and i believe 24th cell shows 6.75 V, is that considered a bad cell?

Thank you in advance

George

 

yes.

 

i guess time to get a new cell

What are the sings that battery ECU has failed?

Thx

 

You would get some code if the battery ECU was not communicating over the network. If the communication is working, you should be able to read battery parameters over the CAN bus, using for instance priidash software, and many others. If they look to be reasonable, then the ECU is OK. The fact that the ECU gave you a battery code, and there is a bad battery, says it is probably fine.

 

Congrats!

But...

Why do folks think that as a battery ages to the point that one cell flat out fails, you can simply replace that cell? If one cell has failed, aren't the other cells almost worn out to the point that they too will be failing?

We replace headlights in pairs because when one burns out, we expect the other to soon follow. But, folks will replace bad battery cells and think they're good to go.

What am I missing?

 

It is likely that some of the other modules have cell imbalance as well and will tend to have a cell reversal (the cause of the "burn out") and fail in a few months. As long as the battery has not been subject to overheating, the remaining modules can be re-balanced and will have a reasonably long life after the repair.

JeffD

 

http://www.cobasys.com/pdf/tutorial/Considerations-for-NiMH-in-Stationary-Apps-battcon05paper.pdf

states that "NiMH batteries exhibit certain failure mechanisms... Some of these.. can be catastrophic... while others are a natural part of use and aging of the materials in the cell."

Unless the failure of the cell is a random, non-wear related event, I would think that replacing a cell and continuing to use the remaining used cells would only be a stop-gap until the others wear out.

Is it false economy to think you can repair a traction battery just by replacing and rebalancing cells?

 

As part of the rebalance process you measure the capacity of each module. This capacity improves as you repeat the discharge/charge cycle by getting the weaker cells more fully charged. When i did this on the 27 good modules in my 2004 battery at 195k miles, all of them ended up with over 6.5 amp-hours of capacity at the low discharge rate used in the process. The other parameter to measure is the module series resistance (measured by the voltage drop when supplying a current). If all of your modules have the same low series resistance, high capacity and State of Charge, you can be confident that the refurbished battery will last a while in use.

JeffD

 

So the assertion is that by testing the total energy storage and the peak power capability of a cell permits one to fully characterize the state of that given cell.

Is there any way to verify whether this is indeed the case or not?

I can see that a cell can have a catastrophic failure if it is reverse biased because it's weaker - that makes sense.

I also understand that one can rebalance the cells once the battery is apart.

Does the battery computer make any attempt to keep the cells balanced? How well do the cells stay balanced once they are initially balanced? Does the system tend back towards equilibrium, or does it, once perturbed, tend towards increasing imbalance?

I guess I'm asking whether the balance of multiple series cells in a rechargeable battery is inherently a stable system or an unstable system, and what the time constant of the divergence is if it is unstable.

Similarly, if it is an unstable system, what would it take by Toyota to make the battery ecu route a bit more current into cells with lower voltages when the battery is charging to inherently make the system stable?

I'm familiar with high reliability rechargeable battery systems, and they're VERY careful to maintain cell balance (as well as NEVER stressing the cells).

 

1) Do certain modes of use (short trips or maybe high battery use) cause these problems to occur earlier or make them more likely?

2) Would there be anything a user might be able to do to keep the cells healthy longer, like long drives on a weekend, or taking the long way to work more often?

How might a user be able to keep the battery happy and give it a long life?

 

We are testing the module while rebalancing to detect any modules (6 cells in series) are a mismatch to the others. This extends the useful life of the reassembled battery.

We have to wait to see how long the rebuilt batteries last in service.

Actually you only need to remove the bus bars from one side of the battery to do the rebalancing of each of the 28 modules.

No, that would require adding a full BMS system (electronics that limit current going through each cell when it reaches full charge). This would be prohibitively expensive for a NiMh battery system (too many cells due to low cell voltage)

Toyota extends the life of the battery by not fully discharging it. You have to assume that at some point the weakest cell will be reversed and the computer will then detect it.

It is the degree of matching between all of the cells in series (28 * 6) that determines how many charge/discharge cycles the battery can handle before a cell reversal occurs (it is inherently an unstable system). Panasonic/Toyota (or a good battery rebuilder) make an effort to ensure a good match.

As stated above, it would be too expensive to monitor and manage the individual cells. The Battery ECU monitors 14 Module pairs (12 cells in series for each pair). They could have tried to route currents at that level, but that would be only a marginal improvement. It is more reasonable to manage individual cells in a LiOn battery system (3.4 volts/cell versus 1.2 volts/cell in NiMh)

Toyota has successfully built a hybrid battery system that will last long enough so that battery maintenance cost is about 2 cents per mile (less if you rebuild the batteries) which makes it about the same per mile cost as tires or brakes in a conventional car. That is an impressive engineering result.

JeffD

 

Keep the interior of the car cool and avoid turning off the car when the battery is fully charged.

 

The only thing we can do to extend battery life is to make certain that the battery never runs too hot. Heat is the enemy. The Battery ECU does report battery temperature (two sensors) on the OBD2 bus. and running the AC provides cooled air to the battery air intake to help keep it cool. If I lived in a hot climate or drove in mountains regularly, I would look at installing a solid state cooling system in the battery air intake vent.

JeffD