Hi All, As promised, I just published a second, more detailed study on fuel economy with the Prius. You can find it at: La Prius à Côté My goal was to identify the factors that affect fuel economy, and to quantify their effect. Beware - it is a long read. There are 25 figures and lots of interesting results. I hope you like. Feel free to comment, suggest improvements to the study, etc. I am very open to suggestions.
Very detailed report- you obviously put a lot of effort into it! Not really too surprised by the conclusions that tire pressure and ambient temperature effect fuel economy that drastically. For years we've had the mantra of "tire pressure, tire pressure, tire pressure" drummed into our heads. Not too much we can do about seasonal temperature changes- other than grill blocking- which I see you're planning to investigate. Didn't see anything about the engine air filter? Maybe I overlooked it.. but have always been told that a partially clogged engine air filter also kills fuel economy. So, in reality- if you keep your tire pressure up and don't drive like an A-hole (gunning it at green lights and skidding up the the red lights) you should see very good results from the Prius without jumping through hoops to chase it. Thanks for the report SCote!
Great study. I was surprised by the minimal contribution of speed to the variance. I think you would see more contribution at speeds higher than you tested (ie >60 km/h). BTW, what kind of commute is this? City, interstate hwy, or rural hwy? You have to study tire pressure next, please!
For traditional engines, this is just common mythology. The air flow restriction from a clogged filter is really no different than a partially closed throttle plate, so for a driver who never floors it, there should be no loss of fuel economy. Loss of top end power is another matter. But this particular view doesn't necessarily apply to most hybrid engines, which do want to run at low RPM with the throttle fairly wide open. So it may be worth more investigation for the Prius.
+1 It's a myth like the need for 3000 mile oil changes. At one time there was a reason for it but not now. Back in the olden daze of carbureted engines it was true because a restricted air filter enriched the mixture and there was no way for the engine to compensate. With modern fuel injection systems that run closed loop mixture control most of the time, the mixture is adjusted to compensate for reasonable amounts of filter restriction.
Aerodynamic drag goes up by the square of speed. The power required to overcome aerodynamic drag goes up by the cube. Aerodynamics make little or no difference at low speeds, but are the dominate factor at high speed. Tom
And in terms of energy required to push the air out of the way per distance, also the ratio of the speeds, squared. Work = Force*distance The distances are equal, the force (drag) as Tom wrote. I think.
Thanks all, for your comments! Thanks! Yes, indeed... Note that the references to "tires" in the figures is related with the type of tires, not pressure. I tried to maintain constant tire pressure, but did not estimate its effect of fuel consumption. Thanks! The ratio highway:city driving was about 75:25. Yes, I was surprised too. Now speed is a bit tricky, especially in a hybrid. I mentioned it briefly in the report, one problem is the "average speed" is a single value for an entire trip. It does not say anything about the variations of speed within the trip. This is particularly important, as at low speed, the engine may not run, which lowers fuel consumption. A trip at constant, medium speed, might produce the same average speed than another trip at varying velocity, but both would produce very different fuel consumption. In other words, I think that a single, "average" speed value is not sufficient to explain variance, which is why it does not seem to contribute much. And yes, tire pressure is definitely something that I should study. However, there are existing studies on the subject, that quantify the contribution of badly inflated tires on fuel consumption. I might check those first.
Stephane, another fine report. You talked a bit about AC fan speed, and its low contribution to FC variation. Ditto with the ECO button. You didn't have many datapoints for it, but for the next study )) perhaps you can look at the two together and see how effective the ECO switch is at controlling the AC and its effect on FC. That may be of interest to folks who have to use their Prii in warm climates. I suspect there is some contribution of error in the weight calculation technique contributing to the anomalous uphill result -your next study will fix, that, I'm sure. Anecdotally, I can tell you that hauling around 400+ lbs of extra passengers makes a big hit in FC, especially uphill.
Thank you very much! I totally agree. Another study could be designed, controlling all variables fixed except those 2. One problem is I dont have to use AC that much up here, so the number of possible data points throughout the year is limited (Canada is cold... ) Thanks for suggesting it though, I will add it to the list of potential future studies. Very interesting idea. I agree. Variance is still high - I should calculate the error to see how sensitive the slope is in both up and downhill cases. From the start, weight was not controlled sufficiently well - I did not see it as an important variable when I started the study, I only realized later that it could be interesting. So my data is unreliable and noisy. The effect of weight should certainly be studied with more care. I'll add it to the list too. Thanks for your interesting comments!
beyond warm driving temperature, over inflated tire pressure, a top/average speed under 35mph, not having the ac/defroster on, not having to stop, driving longer than 30 minutes, and not carry extra payload/weight... going downhill and having a tail wind helps too...
Indeed! I have not collected the wind data, but it is available on Canada's National Climate Data and Information Archive, which can be accessed online. I was very much wondering how I could add that variable to my study for 2 reasons: Wind data is a vector, not a scalar, which makes correlation studies more complex My trip is approximately 25% westbound, 50% southbound, and 25% eastbound. Wind effect might not be very clear on consumption data. But I just found a way to incorporate the data, and shall do so before the end of the month. I'll post my findings. I did collect that data, but could not find a clear tendency. I will look again at the data and come back to you.
