Intro:
Earlier this year, Wayne worked with the folks at
Pulstar (
www.pulstarplug.com) to arrange for a CleanMPG “Pulse Plug” test fleet, of which my 2007 Prius is a member. The goal is to put some plugs in the hands of experienced hypermilers across the country to determine whether or not the product lives up to their manufacturer’s claims. Specifically, they claim “pulse circuit…more peak power…additional cylinder pressure” blah blah blah “with fuel consumption and associated green house gases reduced by as much as 10%.” In the case of a Prius like mine, the manufacturer claims to have recorded a fuel economy gain of nearly 7% (OK, 6.72% if you want a silly number of decimal places) in controlled testing. But is the claim real enough to justify the purchase price of nearly $25
per plug? Good question!
Testing Methods:
It has been claimed, and I tend to agree, that the “real world” is a lousy place to test anything. Both man and nature can produce variables that could dwarf a 7% fuel economy change, including the following:
-Temperature swings
-Rain
-Wind
-Traffic
-Route changes
-The “nut behind the wheel”
Having spent more than two years dedicating my time behind the wheel to improving my own fuel economy, please believe me when I tell you that these factors are not trivial. My own car has delivered fuel economy in a range between 40 and 75mpg depending on the conditions and my behavior. The real range for all similar cars is more like 30mpg to 120mpg. (For 30mpg, we’re probably talking a caffeine addict in a Fargo winter driving three miles a day. For 120mpg, that’s basically what you get when the fuel economy professionals drive in big circles under ideal conditions for a very long time…slowly.) Suffice it to say that this is where “your mileage may vary” came from.
At this point, the smart consumer asks with a furrowed eyebrow “Then why bother?” Clearly, it’s going to be tough to see 6 or 7% difference when the possible range for a normal vehicle is that large. But the good news is that we can narrow the range by being very careful about how the test is conducted. We can also improve our ability to find meaning in the results by recording the things that commonly send a test sideways. So here’s a breakdown of how I plan to accommodate the variables that I think are most important:
Route: Obviously, if you drive up and down mountains one day and putter around a dead-flat town the next, your fuel economy will be different. This is why recording a single number for each tank of gas and declaring victory is a bad idea. One tank of gas for me could mean commuting for two weeks or taking a road trip to Atlanta. Or the difference could be as mundane as making more extra trips to the store for one tank than the other, which is more than enough to skew a result. To take care of this, I will only record data on my regular commute which is the exact same 21 mile (each direction) route every single work day. So every morning, I reset the trip computer before I light the engine and record the result when I park. Same thing on the way back. And if I make a stop or a detour, the trip doesn’t count. Works for me.
Driving habits: Maximum fuel economy! Experienced hypermilers can get ridiculously regimented in their daily routine, to the point of timing accelerations and decelerations down to a specific crack in the road. On my daily grind, that’s pretty much where I am. That doesn’t mean that I can make every trip perfectly identical, but repeatability is quite reasonable. (This will be supported or refuted by the data when it is available.)
Temperature: This is a huge one. The ambient temperature can make the difference between a high fuel economy result and a new personal best. Worse, it changes constantly! So the only thing I can do is record it. At the beginning and end of each recorded segment, the temperature will be recorded along with the fuel economy. I suspect we will see a visible relationship between FE and temperature that will be useful in analyzing the results.
Vehicle set-up: This is another big one. Tire pressure is an obvious variable, and I mitigate it by checking and adjusting once weekly. Checks will be made early in the morning, which should keep things reasonably consistent throughout the trial period. There are other things that could change over a longer period of time (like oil condition) so I’m not going to do just one round of on-off testing with the pulse plugs. Expect to see me do at least two full rounds, four weeks each (for a new 16 week test) without any changes or modifications to my equipment except for the spark plugs.
Weather flukes: There isn’t much that can be done about the weather except record it, so that’s what I’ll do. If the weather is anything but dry and calm I’ll make a note.
Traffic flukes: As with weather, observe and record. Anything off-normal for a segment can be noted, and the trip data thrown out if necessary.
So is something missing? Probably. I’ve already admitted that there’s no perfect “real-world” test and there isn’t much that can change that. But I think that I’m doing as well as anyone could do to produce a pretty good one. Time will tell if I’ve pulled it off!
UPDATE:
Preliminary results are in! I’ve recorded four weeks’ worth of data for each set of plugs. I have to say that it didn’t turn out quite the way I expected. Here’s a simple plot of morning segment fuel economy:
If you look at the blue dots (FE and starting temperature on stock plugs) it’s pretty clear that FE and temperature are well correlated. That FE goes up with temperature is no surprise to anybody around here. And overall the same is true of the pulse plug data. But what’s with all the noise? I started to get odd FE numbers as soon as the Pulse plugs went in and after a month things never evened out. The same thing happened with the afternoon segment:
By now I’m sure you’ve had a look at the charts and noticed something more interesting: FE with the Pulse plugs got noisy, but it
did not increase. Two or three segments in warm weather were a bit better than the others but the distribution indicates, if anything, worse fuel economy overall. So to me the conclusion is very simple: The
Pulstar “Pulse Plugs” do not live up to the claims. At best, they deliver the same fuel economy as stock spark plugs. But they cost $25 each ($100 for a humble 4-cylinder) and are supposed to last only 50,000mi (stock irridium plugs are good for 100,000mi.)
So what’s next? If I take any more data it will have to be morning segment only since we’re getting into serious A/C season here in SC. Some afternoons I have to run climate control, sometimes I don’t, and any numbers from those conditions would be pretty worthless. I will be paying attention to see if the FE stabilizes, though. The most irritating part of this experiment was having FE start to vary for no discernable reason. It makes me wonder just what these spark plugs
really do.
Linear Mode
