Re-gen efficiency in Prime

It would be interesting to know when friction braking cuts in. Is there really a way to see that on the display? If so, exactly how do you get that information?

In the Prius 3 they had a recall update to the firmware because of the issue where regen cut out (I think) if the car hits a bump during braking, causing people to feel that the car lurched forward. The firmware seemed to only reduce the amount of regen and not fix the bug.

 

Hi Kris,
It sounds to me like you are looking for something like this:
Evaluation of the 2010 Toyota Prius Hybrid Synergy Drive System (Technical Report) | SciTech Connect
(See pages 62-69)
That is for the gen3 Prius. As far as I know, that information is not yet publicly available for the Prime, but the gen4 Prius and Prime efficiency curves should be quite similar to that of the Gen3.

There are also many other factors to be taken into account to get a precise estimation of the overall regeneration efficiency. A few of the most important of which are:

• Inverter efficiency
• Battery efficiency
• Overhead electricity consumption
• Drivetrain friction and rolling and air resistance

The inverter efficiency curve is similar to the Motor/Generator (MG) curve. In the linked study above, you can find a curve for the combined efficiency of the MG and inverter. They tend to be most efficient in the mid to upper power range.

When you subtract the electrochemically stored energy from the equation, the battery acts somewhat like a simple electrical resistor. Consequently, battery losses increase roughly with the square of the current, and the peak overall efficiency is shifted toward the lower part of the power range.

The constant overhead consumption of electricity to power all the computers, pumps, etc., implies that at extremely low regeneration power levels, the overall efficiency drops to zero and below.
.

 

The plots show the opposite of that. For example, this combined motor-inverter efficiency plot shows the peak efficiency at around 60Nm while the peak torque is 200Nm. That's 30% of peak torque for peak efficiency, and that doesn't include battery discharge efficiency.

 

Hi Lee Jay, I think you might have overlooked that I wrote "... mid to upper power range ...", not "mid to upper torque range".

Remember that power is the product of force and speed. (power = torque x rpm)

For example, all of the following combinations of torque and rpm produce the same amount of power:
50 Nm x 8000 rpm
100 Nm x 4000 rpm
200 Nm x 2000 rpm

 

No, I didn't miss that. The peak is at 6000RPM out of over 12,000 maximum. That's the center. The lowest power corner of the 95% sector is at 20kW, which is well below half-rated.

 

... and the highest power corner of the 95% sector is at 44kW, which is well above half-rated. That sector ranges roughly through the middle third of the rated power of 60 kW for the gen3, tending slightly higher.

And the 94% sector borders directly on the "Peak Torque/Power curve" from about 4500 to 8000 rpm, but doesn't even get close to the extremely low (<6 kW) power range.

That is what I meant by "They tend to be most efficient in the mid to upper power range."

 

Smooth and steady and try to stay off the friction brakes.

 

First of all, it's generally still most efficient to reduce braking of any kind by easing off the accelerator as soon as it is possible to do so without annoying anyone.

Then, when braking, I agree with the first answer in this thread on the best technique for optimizing regen:

Yes

 

I agree with all but the last. The computer can only generalize and it is a best fit for an average drive. Everyone's situation is different, so there isn't a one size fits all. For some, this is great but there are ways to do better than the computer in the real world depending on your situation at the moment.

 

Yes, I totally agree with all of that. And cheers to you.

 

As far as drive modes go, I agree with all of that too, especially this:

Also, I always like to say; first learn what the Prius likes to do, then you can help it do its thing.

But specifically, as far as braking goes, I am not aware of any mode that will increase the energy efficiency of regenerative braking.

... On second thought, Eco mode can reduce the electrical overhead by reducing power to the air conditioner. Taken to the extreme, the air conditioning pump could even be switched off every time you brake.

At least in theory, it is more efficient to use electricity for the air conditioning directly from the plug-in battery or from the inverter and generator driven directly from the ICE, than indirectly through the inverter and motor, to transmission, to tires, to road, back through tires, back through transmission, and back through generator and inverter.

I, personally, am not about to start fiddling with drive modes or air conditioning every time I brake. But it's an interesting thought experiment, and if it's not already automatically implemented in some form in the climate electronics, it might be in the future.

 

This is for EV only operation. If applied to the Primes gear ratio's I have the 95%+ area as 30 to 40 mph with MG2 spinning 5000 to 7000 rpm. Torque increase would be 12.31x. MG1 would be spinning counter clockwise 3500 to 4500 rpm.
Optimum acceleration torque? There is no data I have seen on the efficiency of MG1 or 2 as generators.

 

That's right, so we are just making an educated guess, based on the gen3 motor mode tests. Typically, the generator mode of a synchronous motor has an efficiency curve similar to the motor mode. Until we find real data for gen4, we are just speculating based on past experience.

Optimum acceleration torque as measured at MG2 (assuming similar to gen3 efficiency contours) would be about 40 Nm at low speeds, ramp up to about 60 Nm at medium speeds, and then taper off to about 40 Nm at higher speeds, if optimum acceleration torque depended only on MG2 efficiency. But internal battery resistance and other factors skew that considerably downwards when calculating overall efficiency.

On the other hand, the larger battery of the prime should reduce battery losses somewhat.

Complicating things even more, is the one way planet gear carrier lockup in EV mode, allowing MG1 to also contribute power. I personally have not yet come across any information on whether or not MG1 is also used for regenerbraking in the Prime.


The OP was hoping to make useful inferences from just the regenerative efficiency curve of the MG. That turns out to be a bit complicated. I think the best way for those of us without detailed computer models or access to special sources, still is to gather data from test drives, like one of the Prius hacking pioneers, Attila Vass, did a long time ago. See here: The new 2004 Toyota Prius : How to brake


PS While searching for the Attila test, I came across an old post of mine on a related subject: Pulse and Glide in EV Mode

 

For example, when driving down a long, gentle slope, where extremely light regenerbraking would be necessary to remain within the speed limit, it might be more efficient to glide (no arrows) and periodically regenerbrake moderately lightly, rather than to regenerbrake extremely lightly the whole way (traffic permitting of course). And once again, to be perfectly clear, this question has little practical value for most people, but it is a technically interesting question for hypermilers.

Actually, I was trying not to make recommendations, but trying to emphasize that I am only making an educated guess. Although B mode does: absolutely, nothing, whatsoever, to increase the efficiency of regenerbraking, it may be causing people to brake more smooth and steady, rather than gliding or coasting and then slamming the brakes.

But yes there is a resemblance to my thinking on extremely low power regeneration being less efficient than slightly higher power regeneration.

With the gen2 Prius, Attila Vass's test showed maximum regeneration efficiency at around two to three times the deceleration rate of a feet-off-the-pedals regenerative coasting. I have not actually compared the data on B mode deceleration, but according to my buttometer, B mode is more aggressive than that. Again, just an educated guess, now getting closer to pure speculation.

Since the Prime has a larger battery, and more powerful power electronics and motors, I would think the maximum regen efficiency is at a somewhat higher power level than all previous Prii, but probably not quite as high as the power level of B mode. And with that, I have really gone out on a limb, and there is a good chance that someone will acquire some hard data to prove me wrong.

 

From experience, I am in the "same as D and brake camp". I've noticed no advantage to B mode, except that you don't have to use the brake much. I do use it a lot when approaching a stop point if no one is behind me.