How efficiently does the Prius turn on the ICE while driving?

You heard wrong. It takes very little power to start the engine. Consider the mass of the car relative to the mass of the pistons and other moving parts of the engine. Also consider the low amount of drag you feel in a conventional car with the engine running at low RPMs as you coast. The power needed is moderate, but for such a short time that it is pretty insignificant relative to the size of the battery.

 

I have no idea how much power goes into the motor to start, but on cold starts you are likely in the right ballpark of 1%, which is on the order of 0.01 kwh. I do not really know. Someone may no the exacts specs. Remember there are 33.7 kwh of power in a gallon of gasoline though, so this is not a high number. Once the engine is warm, it takes less energy. That and getting the cat warm for pollution control are all part of the warm up routine. If the vehicle is in motion, it can use that motion to start the car, or if accelerating while starting more power will be used to accelerate. If Toyota did the software correctly, the energy saved by having the ice off should be greater than the starting energy. Its not worth worrying about.

Pulse and glide is another matter. You can pulse in the efficient range, then not use gasoline energy on the glide. This can make things more efficient. The gen III can glide at speeds up to 46 mph.

 

In 2005-06, I did some initial measurements using a custom setup:


I remember the engine startup is less than 250 ms but the energy needed to start and later stop the engine is modest compared to what it takes to move the car:

• 2,700 lbs - weight of NHW11
• 27 lbs - estimated weight of engine crank, pistons, cams, and chain

Now you could run an experiment if you have an SOC meter. Just cycle the car from READY to OFF over and over again and map the decrease in SOC. As soon as the ICE begins running, turn off the car. Plot the SOC values over time.

Bob Wilson

 

That sounds very fast for a cold engine. Is less than 1/4 second for start up after it has gone through its warm up routines?

Do you have mg1 current and start up times for a cold engine on a cold day? I would have guessed this was around 5 wh of power but it could be orders of magnitude lower.

I know mazda was playing with stopping the ice at a particular place, then injecting fuel and combusting it, to start without a motor. I don't think you can pull that trick without DI, but does the prius engine have other tricks to lower start up power requirements?

 

I did a fairly extensive set of cold weather and cold engine tests and have some data to share later this evening. From memory, the engine spins for about 3 seconds to get the oil presure up before fuel is added. I can hear the transition in our 1.5L, 2003 Prius. Then until the cats light-off, there is a 45-55 second interval where the control laws do everything they can to keep the ICE unloaded including accelerating the car in EV mode. But once the cats light off, ICE start and stop are so fast my Graham scanner is lucky to get any data points. You have to use other means.

Bob Wilson

 

Thanks,

I got home; fed the wife and dogs, and; took a nap. Here is my earlier report on cold weather warm-up and a more technical report with detailed metrics. For example:


Bob Wilson

 

So a rough estimate from that graph, 270 Volts * 20 Amps * 5 seconds = 7.5 watt hours. So 7.5 Wh / 1.4 kWh is about 0.5% SoC - and that's only from a cold start. On a warm start, it's probably closer to 0.1%.

 

I would probably scale it based upon our best understanding of the times:

• .250 / 5 ~= 5% :: warm-start time over cold start time
• 7.5 Wh * 5% ~= 0.375 Wh :: estimated warm-start energy
• 0.5% SOC * 5% ~= 0.025% SOC :: estimated warm-start traction battery load

Regardless, we're not looking at a whole lot of energy, especially from the traction battery that dwarfs ordinary 12V lead-acid batteries.

Bob Wilson

 

Actually the traction battery voltage is plotted ... check the legend. But what you really need is the original data.

I have to run an errand and when I get back, I'll see if I can find it.

Bob Wilson

 

Actually if the Prius is already moving energy will be put into the traction battery/MG2 rather than removed.

When the car is moving MG1 will be free-spinning backwards with no current to allow the ICE to be stationery. To start the ICE from this condition the speed of MG1 needs to be reduced which will be done by taking power from it and onto the system bus - from there it will go to either MG2 or the battery (probably MG2 if the reason for the ICE start-up was a request for more power).

kevin

 

Only if you take away momentum. In this house, we obey the laws of thermodynamics!

(If the motor is spinning freely at a given speed, it has to be powered to slow it down, thus spinning the ICE up. Even though the motor RPMs are dropping, you're sending power to it. If you want to send power to the battery too, then you've gotta generate power to do both, using the other motor to reduce your momentum.)

 

"Even though the motor RPMs are dropping, you're sending power to it."

That's not true.

To slow down MG1 requires extracting power from it not sending power to it. The energy comes from the cars momentum. As MG1 slows down it causes the ICE to spin-up and start.

The electrical power from MG1 will go to the system bus and will be sent to MG2 if you are trying to accelerate (probably added to power from the battery). In other scenarios it might be sent to the battery.

kevin

 

Let's say you both are a little right and a little wrong.

If you are engine off, and the car changes to engine on, you are likely requesting additional power. Until the engine turns on, this additional power comes from the battery. Turning on the ice, will slow down the car, so power is supplied to MG2 from the battery, even in conditions that MG1 is generating. The net result is battery power will be used when turning on the ice. This power can be quickly added back if the software requests it. The key is power needed to start is low once the ice goes through its warm up stages, and the ice should not shut down when moving until it is warmed up.

 

I concur. Unlike other cars, the Prius does not have a 'bump start.' Rather, MG2 can be used as a generator which we normally call 'regenerative braking.' But if MG1 does not get any power (or generate enough torque to turn the engine,) nothing is going to turn the engine over from the drive train.

What is a hoot is to see use of "B" when descending a hill at high speeds.

Bob Wilson

 

Bob,

Actually I would disagree in that the Prius can do a "bump start" if the car is already moving and it wishes to start the engine.

If the car is moving but the ICE is not rotating then MG1 will be rotating backwards, not giving any torque to the ICE, all that is required to start the engine is to stop MG1 rotating (even shorting its windings would do that, no battery required).

To start the ICE the system will pass current into MG1 to generate the torque but because of the direction of rotation it will actually be causing MG1 to slow down and thus be a generator - it is the momentum of the car that will cause the ICE to rotate and start - pretty much the same as in a conventional car where you release the clutch in-gear with the car already moving.

If acceleration is being demanded the ICE will then run at a relatively high RPM, MG1 will reverse direction but with the torque also reversing polarity so that it still generates power which then is fed to MG2.

As you have shown in many of your excellent graphs MG1 can be run in all 4 quadrants either generating or consuming power with either direction of rotation.

kevin

 

But think about this for a second. The turning ice plus generating with MG1 will cause a load to the system, even supplying the electricity from mg1 to mg2 will result in deceleration. To counter this force, more power from the battery will be added to mg2. Now toyota could have put software in there to bump start, but that would result in a poor driver's interface. Recharging the battery, is likely more efficient and a better driving experience, then decelerating the car when the driver is not calling for it then having the driver call for additional acceleration. I am unsure what speeds and acceleration would cause mg1 to generate electricity versus consume it, but no matter what the mode the car needs more power from the battery to do the work of starting the engine.

What ever power is being demanded, the ice will run at that power setting, power will determine rpm. It is likely in acceleration mg1 is generating, mg2 is using power. Software based on power and soc determines charge or discharge after the engine is started.

 

Agreed - but what happens under the situation where the ICE is not rotating but the car is at < 41 mph and then the car increases in speed to >41MPH (or 34MPH if in EV mode), I frequently get this as an early part of my commute is downhill. The ICE is only being started to avoid exceeding MG1 RPM. In this case there is no need to give extra power to MG2 as there is no acceleration demand - I must admit I don't remember whether the car loses momentum - the engine starts and then immediately stop again in the EV case as it disengages EV mode.