Nobrainer’s Think Tank

I said I’d test it, and I’ve finally gotten my chance. I’ve got about 290 miles of interstate to cover en route to Huntersville, NC. The game plan is simple. I’ll stop at a local hardware store, pick up some “pure” acetone, then head to the gas station where I will top off the tank of my test vehicle, a ‘95 Grand Prix with ~120,000 miles and some body damage, with gasoline and about 4.8 fluid ounces of acetone. The tank is about 16.5 gallons, so that should give me a ratio of about 3 ounces per 10 gallons, which is ideal according to Louis LaPointe. Once I get to Huntersville, I will top off again. Hopefully I will get receipts to verify my location, time, speed, and distance traveled.

Typically my mileage is about 27-28 mpg on the interstate. I’ll be mildly impressed by anything greater than 30. Of course if acetone is as good as is claimed, then I should get an easy 35 mpg.

On my way home, there should be between 1.5 and 2.3 fl. oz. of acetone left in my tank. That means that my trip home will provide a nice little secondary test.

I’ve written the above at about 3 PM on April 19th, 2007. I hope to be at the hardware store around 4PM and to leave the gas station by 4:45 PM. This should be achievable. [Note: I am not publishing my plan ahead of time because I really don’t want to announce that I’m not going to be at my house all weekend.]

THE RESULTS:

I went to Lowe’s, got some acetone and a 1 gallon gas can (for easy mixing) and went to the gas station. There I realized that the acetone can had a lid that had to be pried off. And my test plan changed.

The new plan: stop for gas every ~95 miles. Do two segments without acetone to determine a bit of a baseline, then add it and run with it for the rest of the trip. For every segment, record mileage, average speed, and gas usage.

Much to my surprise, my mileage definitely improved after I added the acetone. The net effect, based on this data (28.4 mpg pre-acetone & 32.7 mpg with acetone), is an increase of about 15%. The most significant result here is my mileage getting over 30 mpg for an extended time. I am not willing to say with confidence that 15% was the true improvement; I reckon it could be between about 9 and 18%. Nor am I ready to say that acetone works as advertised. However, I will say that it definitely appears there was good improvement due to the acetone. There is room for further investigation. In particular, I want to test without acetone ASAP (even though I pretty much can’t), to see if the improvement stays or disappears.

Notes: There were a number of variables which weren’t controlled for. For example, the weather for the first half the trip was cool and very rainy. The other half was warm and sunny. Of course traffic was also not constant. There were several brands of gas. There were several pumps and no guarantee they all operate the same. Plus the legs aren’t equal. There are probably other variables.

UPDATE 6/4/2007:

Over the last few days, I went on a road trip that took me from Charlottesville to Huntersville, beyond, and back. I covered the same miles and stopped at most of the same gas stations. This time the weather going out was warm and clear, while the weather coming back was rainy and cool. Additionally, the car also had to deal with extra weight and moderate (compared to virtually zero the first trip). The weather and heavier traffic led to some stop-and-go driving on the way back, as well as overall slower average speeds. This time, there was no acetone added to the tank, although there should be some residual acetone leftover from the last test. That amount should be quite low, less than 0.07oz/10gal of gasoline.

The mileage comparisons are interesting, but I honestly am not sure of the best way to compare the new data to the old data.

The mileage per segment was 27.93, 26.83, 34.65, 24.95, 33.61, 29.56.

The mileage per segment compared to the first test was -1.13, -1.04, +1.72, -6.02, -2.40, -2.09. Basically, five out of six segments were worse than before.

In the first test, the first two legs had gas with no acetone. Comparing those first two legs, the mileage dropped from 28.44 to 27.35, or about 4%. For the last 4 legs, the mileage dropped from 32.67 to 30.20, or about 8%.

In calculating the results for the first test, I compared the average mileage of the last 4 segments with the average mileage of the first 2. That seemed to show about a 15% improvement. Doing the same for the latest test shows a 10% improvement.

So by this data, the acetone may have accounted for an 8% improvement. It’s not shabby, but it is nowhere near the 30 or 35 percent that some people claim to find. And even that 8% is questionable. On the second trip the car carried more weight, which should reduce mileage anyway. But I went slower which should increase mileage. Although it wasn’t a constant average because it included some stops due to wrecks. And the trip out was in May and the trip back was in June. June 1st is the day to switch to summer blend gasoline. So there may have been more ethanol in the tank on the way back which could also have reduced mileage.

The most important thing that can be taken from this is how hard it is to accurately measure the change caused by a variation of just one of the many factors that affect fuel mileage. There are just too many uncontrollable factors involved in road-testing to be definitive.