BMS Bottom Balancing HV Battery?

Others will correct me if I'm wrong, but I don't think the car does anything to balance the battery for a combination of reasons.
One reason is that it keeps the state of charge between 23% and 80% (if I remember the numbers correctly - something like that, anyway).
Combine that with the fact that all the terminals are tied together and it can't charge or discharge one cell without doing the same to all of them, and you come up with one reason for HV battery failures - the cells get too far out of balance and you eventually have one at 80% and one at 23%. That's why Hybrid Automotive sells so many grid chargers; to do what the car does not.
As I understand it, the BMS just managed the overall SOC as your driving pulls power from the battery or replaces it.

 

I can't really think of a good explanation for what you observed other than something was using power from the traction battery. Since you had been driving in EV and it was hot (I assume you also were running the air conditioner), the battery cooling fan would probably be running faster than usual.

As far as I know, the car is always aware of the state of charge unless you change the SOC with the car off, such as when using the HA grid charger. I don't think the system was correcting its data; I think the data itself was changing, but don't know why.

As for shutting it off and going to ACC mode, I agree that that was probably not a good thing to do, unless it was just for a moment to open or close windows or something, but I don't know what effect that could have had on the traction battery's SOC. Maybe someone else know more about that aspect.

In the meantime, I would not be concerned unless you are seeing rapid changes in SOC such as swinging between green and purple bars in just a couple minutes. That would be bad. With a 2007, though, you should be maintaining it with a grid charger to try to ward off that eventuality.

 

Ah! When I saw "temps were 23 - 27 degrees celsius," I misread that to be ambient temperature and thought you must be having quite a heat wave.

Glad that it's working well.

 

Both of those things started happening to me after my old HV battery had lost a significant amount of capacity. It didn't spell a sudden death of the HV battery or anything, it kept going for quite a few more years after that, but it did eventually get worse and the HV battery die.

What you are seeing is the Prius battery monitoring getting out of step with the actual battery charge level due to reduced battery capacity.

When the battery SOC gets very high then the BMS will recognize that it's full by seeing the module voltages are relatively high, and the green bars will come up pretty quickly when that happens. Then as you use up that charge the BMS coulomb counts the charge out and the SOC only drops fairly gradually. If however the battery A-hr capacity is low, then it takes less coulombs (A-hrs) to discharge it than the BMS expects, so some time when it gets to about 4 bars it will suddenly figure it out (due to low module voltages) and the SOC will quickly recalibrate to about 1 or 2 bars.

Conversely, when the battery is charging up after being very low, then the BMS will coulomb count the charge coming in and initially the SOC will only rise slowly. Again the BMS will over estimate the battery's A-hr capacity until it gets to about 6 (nearly 7) bars, when it will suddenly figure out that it's actually full (due to high module voltages) and those last two bars (7 and 8) will come up unexpectedly quickly.

If you don't run the battery down to low SOC levels (4 bars or under) or regen to high levels (7 bars or more), but just stay in the middle SOC region, then you might not see this behavior and it might still coulomb count quite happily despite the reduced battery capacity. But when you get extremes of SOC (in either direction) then you will get a subsequent recalibration of SOC some time after that.

With mine it stayed like that for a few years and gradually got worse. Eventually it would "max out" very soon after hitting 6 bars SOC on the way up, and bottom out soon after hitting 4 or 5 bars on the way down. Then one day it (the HV battery) died completely and I replaced it. As soon as I got the new HV battery all of that funny business with the SOC stopped happening.

 

Good word picture of a complex system, @uart. I like to picture vertical bars like in a bar chart, each one at a different level reflecting its own state of charge. They all move up and down together. If one cell is radically lower or higher than the other, it goofs up the whole system. If you could get all the cells showing individually that way on one screen on Torque or Engine Link I think it would be revealing. Maybe a little mind boggling, too.

 

Yeah Jerry, when there are cells out of balance like that then it definitely reduces the apparent capacity of the pack. The true capacity of the cells also reduces over time, and the weakest cells are the most likely to get punished when you get those large SOC excursions.

The SOC bars on the energy display also vary in a non uniform way, with some representing larger amounts of charge than others as shown in the following chart. I suspect that this is mainly for the benefit of the driver, so that they see what they want to see - as in a display that mostly shows nice stable levels at around 6 or 7 bars. This is why people usually notice any anomalous drop in SOC at the low end, but it's not always noticed so much at the top end where the very "broad" 7th bar tends to hide it.

 

Nice graph!!! Really illustrates why the monitor spends so much time on bars 6 & 7 in a healthy battery.

 

You lost me there. I don't know anything about the "Prius Doctor," but it's probably showing voltage in blocks. But there is more than one cell per block. Your Prius has 168 cells altogether. Since the cells are grouped, the BMS only knows the voltage for blocks, not individual modules or cells. And there is one positive and one negative wire going from the inverter to the battery, so they all get the same voltage when regeneration takes place.

 

Well I'm embarrassed to admit that I didn't do any monitoring or maintenance of the battery modules at all. We just drove it until the battery died (at 314,000 km) and then put a new one.

When I first got the Prius I dreamed that the HV battery would go forever, but over time I accepted the reality that at some time I would probably need to replace it. Ah, dreams vs reality.