Sudden massive battery charge drain

Was driving the car earlier today. I had just gotten home and parked with the usual 75-80% charge in the SOC gauge. 10 minutes later I back out. I backed out, then was cruising down the street for no more than 200 meters at an easy 15mph. The ICE then kicked in as there was a hill up ahead. Next thing I knew, I felt a slight give in the gas pedal, heard and felt the engine rev up. I glanced at the dash and noticed that the SOC had all of a sudden dropped to 2 bars, all within a span of less than a minute. This has never happened before. I understand that the EV can significantly deplete the hybrid battery, but never this severely and quickly. The opposite had happened about 10 months ago, when the ICE hyper-revved to over 3k RPM while I was going no more than 25MPH, but the SOC never dropped.

There was a posting of something similar in a Gen 2 a few years ago but I wasn't able to find this occurring in another Gen 3 user. Any clues would be greatly appreciated.

Maybe you didn't shut it OFF during those 10 minutes. I wouldn't worry about it unless it recurs.

That would likely explain the event except I definitely did turn the car off when I went in the house. I know this for sure because I remember hearing the alarm disarming and the doors unlocking from the Smart Key activating when I grabbed the outside door handle

I must not understand the operation of the accelerator pedal correctly, because I don't understand how anything having to do with the charge of your battery would have any effect at all on the spring in your accelerator pedal.

May I ask how many miles are on your car?

I notice a big/faster drain when I begin a trip; I think it has to do with the first minute of propulsion being solely from the traction battery (even though the engine is operating - at idle). You can feel the engine "kick in" to regular operation after that first minute.

Could it be related to this?

-Iain

It was just a strange feeling I have never felt before, as if the pedal sank ever so slightly even though I was applying gentle pressure to maintain a ~15mph speed. It coincided with the ICE's sudden high revving. This was the only reason that I quickly glanced at the SOC and noticed the unexpected drop. I had also thought for a split second that maybe I was inadvertently in PWR mode even though I rarely use it as the jerk felt a little bit like accelerating in that mode, but the dash did confirm that I was in my usual normal(non-EV/ECO/PWR) mode.

About 10k and have had the car 13 months now

I realize that there is a significant drain when in EV mode. But the car is supposed to be able to go about a mile in electric. I had driven the car no more than 1/8 of a mile when the SOC dropped more than 50%. This was definitely not normal behavior for my car. And everything was off at the time(lights, stereo, HVAC).

It could just be a random "hiccup". If it happens again I will submit another post.

Depends quite a bit on the conditions. Were you measuring SoC with just the dashboard display, or do you have a ScanGauge or equivalent hooked up? Without a ScanGauge, it's very hard to accurately judge SoC. This graph is for the Gen II, but look at the image in this post; I believe it's very similar for the Gen III. If you started at 6 bars and dropped to 2, you may have used less than 10% of the total battery capacity (or 25% of the usable range).

A couple weeks ago, I was house-sitting for someone in the mountains - my daily commute to their home included a highway trip up the mountains and most of the way back down (which would charge the battery fully to 80%), then I'd get off the highway and take a local road which went very steeply up over about a quarter mile, and then back down steeply where I could regen again. To be able to capture the most possible regen in that last area, I would EV up the hill (when not being followed, anyway), starting at 80% SoC, or 8 bars. It was drawing 80-105 Amps the whole way, and was drained down to about 55% (about 5 bars) by the top, a quarter mile later. On the other hand, if I had started at the mid-point of 7 bars ("1 bar less"), that amount of charge would have dragged me down to just 2 bars ("3 bars less", or a span of 5 bars instead of 2 bars), because the bars are different sizes. It also might have kicked me out of EV, as I was well into the PWR bar (even though only going 15-20 mph), which it won't let you do as the battery SoC gets lower.

Summit does have a good point, too - when you first turn on the car, it will almost exclusively use electric power for about the first minute. Accelerating uses quite a bit of power, and will very easily drop you several bars. After that minute, the engine will suddenly speed up if you are pressing on the accelerator, as it replaces battery power with the now-available engine power. It may also decide to run the engine much faster than normal to re-fill the battery, especially when around 2 bars or so.

there should not be this feeling, as the "gas" pedal has no force feedback, there is just a spring and pedal position electronic circuit. The only potential option is that the pedal spring is out of its positon, but in this case the situation will persist pernamently. Let try to push pedal by hand slowly (with engine off) and listen some unusual spring noise/rattle.

Hi Summit,

Are you reading SOC from the graphic display ? My experience in my G2 Prius is that the graphic is usually pretty accurate but once in a while it can be way wrong, or take time to update. I see this anomaly most commonly when I go for long glides: the picture shows 4 or 5 bars and I think I can glide forever, but then the ICE kicks in and the graphic drops in one fall swoop to 2 purple bars.

In general, my advice is to not ride the traction battery.

There is another explanation to a change in pedal feel: perception. Humans are very good at anticipating response from an input. For example, have you ever grabbed a milk container thinking it was full, only to find it nearly empty, and in the process nearly toss it through the top of the refrigerator? We become accustomed to the characteristics of a system when exposed repeatedly to the same inputs and responses or reactions.

Just as with the milk carton, when driving a Prius we expect a certain amount of acceleration when we press the go pedal a given amount. When it takes more pedal application to produce the expected acceleration, it feels like the pedal "sinks" a bit. The pedal action is normal, but the car's response to the action isn't. This messes with our perception of how the pedal works.

This same situation works with Prius brakes when they are running in the normal brake-by-wire mode. The brake pedal will always offer the same amount of resistance, but braking action will vary.

Tom

I was using the SOC bar indicator on the car's innate display. I do have a Scan Gauge but don't use that for the battery's SOC, only temps, RPM, and main battery voltage.

That chart was quite useful. I didn't realize there was not much of a difference in the SOC between 6 and 2 bars.

There didn't seem to be anything out of the ordinary with the pedal before or after the incident. It could very well have been my own perception. It's quite possible that the ICE kicking in caused the jerk in the car itself that made me passively push my gas foot, making me think it was the pedal changing characteristics. Even if that were the case, the jerk was stronger than normal.

There is no "sink" since the accelerator linkage is totally fly by wire.

I have a 2005/80kmiles and over the time I have observed the same thing. It's my understanding that this is part of the process of "conditioning" the traction battery so it maintains its cycle depth capability. Keep in mind that the traction battery operates over only about the top 20~25% of its capacity. I'm not sure about the newer models.

Measuring SOC is tricky. On old lead acid batteries, remember how you could estimate charge by sampling the specific gravity of the electrolyte? There's no analog for a NiMH battery. The Prius uses a combination of two approaches.

In principle, you want to know how many Coulombs of charge are in the battery. There's a maximum, and you want to know what % of that max are in there.

There is a charge counter, basically an ammeter. That measures the flow in and out, and they track it. If you're at 60%, then take out 10%, then out 10% back in, you should be back at 60%. The problem is knowing where you started, and errors in the measurement (which would build up over time).

So, they then do periodic corrections. For a given SOC, at a given temperature, and a given charge/discharge rate (amps), the battery will supply a certain voltage. You could just back calculate the SOC from the above, knowing the complicated equation, but there are so many variables, and the equation would change over time, so that's not 100% reliable either.

So, the Prius does some of both. For second-to-second SOC estimates, it uses charge counting. Then once in a while, it looks at voltage, temperature, and maybe current, and does a correction. If the correction is big enough, there can be a jump in the SOC estimate.

I don't know exactly how the whole algorithm works, maybe it does the correction only at 0 amps in/out, and above a certain battery temperature. Who knows. I'm sure it's very proprietary.

"...remember how you could estimate charge by sampling the specific gravity of the electrolyte?" - no, I tried very hard to forget that after school!

I still have a hydrometer somewhere. I also used to have a small one with little colored balls; it was used to check antifreeze concentrations.

Tom