The Fuel Truth

story by per schroeder • photos by perry bennett

There are a lot of myths surrounding the various types of gasoline we use in our street and race cars. Which one will help us get the most power? Which one is the best bang for the buck? Do we really need race gas? If so, which type? There’s precious little data to back up many of the claims we’ve seen touted as the gospel truth, so we figured it was time to shine the light of knowledge and science on those guesses and half-truths. After all, we use enough faith and prayer when we’re 10 yards past the braking zone. Why waste any on something that can be tested in a controlled environment before we hit the track—or tire wall?

The Experiment

A straightforward question like “What is the best fuel?” can become complicated in a hurry. To minimize variables, we set out to make this test as simple as possible.

We chose a common and widely used car and engine package, we tested each fuel on the dynamometer, and we tuned the car for best power and torque for each particular fuel. Sure, some fuels allow for a higher compression ratio, but opening up the engine just wasn’t feasible for this test.

We also tried to stick with one race fuel brand for our experiment. Our desire wasn’t to create a brand comparison test, but rather a comparison of the different types of fuel available.

All ignition timing was done to find Mean Best Torque—the point where torque output was maximized. On most mild engines, including the one we used, this is way below the point where knock occurs.

Fortunately, the load-bearing Dynapack dyno used during this test made that step painless: We could simply increase the timing until it stopped making a difference in torque output. Increasing timing beyond that point will, in fact, cause power to drop; knock will happen soon after.

We tested several popular fuels: 93 octane E10, 87 octane E10, 93 octane E0, E85, Shell URT 100, Shell URT 105 and VP’s M1 Racing Methanol.. The two 87 and 93 octane E10 fuels were sourced from the QuikTrip station in Suwanee, Georgia; the remaining fuels came from Hyperfuels, a dealer for Shell, VP and Sunoco race fuels.

Our Nerds

Smart folks know that creating a complex and ingenious device requires a full crew of even smarter folks. NASA didn’t put Curiosity on Mars with the work of just one pencil-necked geek. They had an entire nerd patrol that got the job done. Managed correctly, a group of engineers and scientists can solve just about any problem—from the mundane to the complicated.

Building and tuning fuel-injection systems definitely qualifies as complicated. Jerry Hoffmann is one smart guy, and he’s got his own team of geeks helping out at his business, DIYAutoTune.

The firm, based in Suwanee, Georgia, specializes in building MegaSquirt programmable fuel-injection systems. They assemble circuit boards, program computers, and tune cars for a living—you know, nerd stuff. Jerry’s tech team consists of Matt Cramer, Ben Berusch, Russ Patrick and Kevin Boswell.

Test Subject

Kevin Boswell’s 1990 Miata was built for SCCA’s C Street Prepared autocross class. Thanks to the generous update and backdate allowances in the rule book, the lightweight first-generation Miata runs a later 2001 long block fitted with a 1999 intake manifold and fuel rail.

To make this FrankenMiata run, Kevin installed a full DIYAutoTune MegaSquirt Plug and Play system designed for the original 1.6-liter harness. It’s got an additional VVTuner module to control the later long block’s variable valve timing. The MegaSquirt is fully tuned for typical 93 octane E10 pump gas, and it routinely puts out mid-130s on DIYAutoTune’s Dynapack wheel dynamometer.

We anticipated that the tuning for the different types of gasoline would require more fuel than the factory fuel injectors could handle. The solution was to get a set of four 550cc injectors from RC Engineering. We knew injectors of this size could deliver plenty of fuel—of any type—for the normally aspirated 1.8-liter engine. A bonus is that RC carefully balances their injectors to deliver exactly the same amount of fuel—a detail that can really pay off for racers trying to get the most out of their rule-book builds.


Our day of testing at DIYAutoTune was a carefully orchestrated dance of fuel jugs, nerds and computer cables. The company’s Russ Patrick handled the tuning using TunerStudio MegaSquirt software, while Kevin Boswell changed out the fuel jugs and flushed the fuel lines before each fuel was run. The flushing would eliminate any contamination between test fuels, as leftover fuel in the line could affect the numbers for several pulls.

Each dynamometer run took place under similar conditions: a coolant temperature between 180 and 186 degrees Fahrenheit and an intake air temp between 102 and 107 degrees. There was one exception, though: The 87 octane E10 gasoline’s intake air temp ran at 112 degrees. Russ felt that the ECU’s IAT/CLT temp correction factors did not have a major effect on the results, as we tuned for max output of each fuel on the spot. The numbers listed for power and torque are without any correction factors—they are the raw data that the Dynapack wheel-mounted dynamometer churned out.

Russ tuned on 93 octane E10 pump gas first, then used that tune as a baseline for comparing all the other fuels. Each fuel type took three or four pulls while Russ adjusted the fuel and timing to maximize power. The exception to this was the VP M1 Racing Methanol, which took six pulls due to Russ’s unfamiliarity with it.

After the tuning was completed—but before the final power pull—Russ set the dyno at various rpm and spot-checked the air/fuel and power in 1000-rpm increments. He also checked the manifold pressure range to be sure that the air/fuel ratio transition was correct between low and high throttle positions. The resulting ECU tune was one that could be used at full throttle as well as around town or in the paddock.

QuikTrip 93 E10

(prices variable)
POWER: 136 horsepower
TORQUE: 124 lb.-ft.

Our baseline fuel for Kevin’s C Street Prepared Miata was purchased just down the street from DIYAutoTune at the QuikTrip, a local stop-’n’-rob that sees a lot of traffic. Its popularity guaranteed a fresh supply of fuel.

The E10 designation means that the gasoline is mixed with 10 percent denatured ethanol (EtOH). It’s what you get on nearly every street corner in America—and frankly, it’s not too bad. We’ve been using it in our daily drivers for five years and have had zero issues with the stuff. We’ve heard horror stories about it dissolving some older fuel-injection seals. We’ve also heard tales of the ethanol separating from the gasoline if left still for months on end. However, we’ve experienced neither of these problems.

TAKE-HOME MESSAGE: A solid baseline fuel.

QuikTrip 87 E10

(prices variable)
POWER: 135 horsepower
TORQUE: 117 lb.-ft.

This part of our test questioned one really old myth: Does a car make more power on lower-octane gas because, by its nature, it burns more quickly? On our Miata, the answer was a resounding “No.” The 87 octane gas required that we pull the ignition timing back 1 degree and dial back the fuel 2 percent. The result was 1 less horsepower and a solid 7-lb.-ft. loss in torque.

TAKE-HOME MESSAGE: Don’t cheap out—go for the good stuff if you’re tuned for it.

Hyperfuels 93 E0

POWER: 134 horsepower
TORQUE: 122 lb.-ft.

Here’s the old-fashioned, non-ethanol-laced fuel we had in the U.S. prior to 2007. It’s available locally from a few non-chain gas stations, typically for use in boats and lawn equipment that could have issues with ethanol separation during storage.

The 93 E0 required the same 2-percent reduction in fuel trims that 87 octane liked, but it could handle the same timing curve that the 93 E10 did. The result was a loss of 2-horsepower and 2 lb.-ft. of torque.

Take-home message: This stuff is great if you store a race car for years at a time, but bad if you want to win.

Hyperfuels E85

POWER: 143 horsepower
TORQUE: 128 lb.-ft.

Hyperfuels E85 is 85 percent corn-based ethanol mixed with 15 percent traditional gasoline. It’s available in 5- to 55-gallon quantities and ships from Hyperfuels’s Houston, Texas, facility. Ethanol-based fuels are highly hygroscopic, meaning they absorb moisture from the atmosphere. As a result, they must be kept in closed containers and used quickly once decanted.Gasoline has a 14.7:1 stoichiometric ratio—that’s 14.7 pounds of air burned for every 1 pound of gasoline. E85, by comparison, has a 9.76:1 stoichiometric ratio. Methanol, which we’ll get to later, has a 6.4:1 ratio.

The E85 is a highly oxygenated fuel. A considerably greater amount is required for it to run correctly at the optimal air/fuel ratio. How much greater? Try an average of 63 percent more fuel at the same rpm and load. All that extra fuel did allow us to bump the timing by 2 degrees. We saw an additional 7 horsepower over baseline, with a corresponding gain of 4 points in torque.

TAKE-HOME MESSAGE: Possibly one of the least expensive horsepower gains per dollar we’ve seen, but only if you can tune for it and use your car regularly.

Shell URT 100

POWER: 137 horsepower
TORQUE: 123 lb.-ft.

So what happens when you pour the good stuff into your tank at the race track? You know, the go-juice that smells wonderful, even when burnt in the guts of a LeMons car? We grabbed some Shell URT 100 to see just how much of an improvement it would make in this nearly stock Miata drivetrain. It’s available at most race tracks and can be shipped from Hyperfuels in 5- or 55-gallon containers.

Shell’s URT 100 is considered a highly oxygenated unleaded fuel. Oxygenated fuels are supplemented with extra oxygen-bearing compounds, like methanol (MeOH), ethanol (EtOH) or n-butanol (BuOH), in addition to the normal gasoline hydrocarbons. These additives can help reduce carbon monoxide emissions, but the extra oxygen is also helpful for yielding more power via their faster burn rates.

The URT 100 required 2 percent less fuel but could handle an additional degree of ignition timing. The net result was an additional 1 horsepower, with a corresponding 1-lb.-ft. loss in torque.

The fact that less URT 100 was required surprised us. After all, the addition of all that oxygen would make the mixture leaner for a given amount of atmospheric air entering the engine. We double-checked our logs, and this was indeed the case for our Miata.

TAKE-HOME MESSAGE: Probably not worth the extra coin on a car with a nearly stock engine and a relatively low compression ratio. However, if you’ve got custom high-dome pistons or you’re throwing boost at the equation, it could be worth it.

Shell URT 105

POWER: 137 horsepower
TORQUE: 124 lb.-ft.

Shell’s URT 105 is a traditional non-oxygenated, high-octane, unleaded race fuel. It’s the gold standard for engines with super-high compression ratios or forced induction.

While we could advance the timing curve on the Miata by 3 additional degrees, we used 4 percent less fuel because URT 105’s higher density equates to more energy per gallon. Frankly, we were pretty surprised to see only a 1-horsepower improvement over baseline and no change in torque readings.

TAKE-HOME MESSAGE: If we couldn’t use 100 octane, we certainly couldn’t properly take advantage of 105. Unless you’re compressing a lot of fuel and air into your combustion chamber, there’s no need to spend the money.

VP M1 Racing Methanol

POWER: 160 horsepower
TORQUE: 140 lb.-ft.

While many sanctioning bodies require gasoline-based pump fuels, some racers may use methanol as a fuel. Although methanol has a low energy concentration—similar to that of ethanol—it can improve power. This is thanks to its higher octane rating, high heat of vaporization and ability to handle a lot of ignition timing. Unfortunately, it’s highly corrosive and can’t be used in a daily driver—it eats rubber seals and aluminum like nobody’s business.

The Miata required almost double (92 percent more) the amount of methanol compared to the pump-sourced 93, but it required the least amount of ignition timing to find Mean Best Torque. Despite the softer ignition curve, the methanol-fueled Miata gained an impressive 24 horsepower and with 16 lb.-ft. of torque.

TAKE-HOME MESSAGE: If your rules allow for it, it’s hard to beat methanol for a race car. It simply delivers massive amounts of power while requiring relatively few changes.

Out the Tailpipe

In the end, a given engine only handles so much fancy fuel and timing. It’s like multivitamins: Taking 20 won’t make you healthier; they’ll just make your pee smell funny.

Picking the right fuel for your application is better than worrying about getting the highest octane rating. If you’ve got a stock engine, try tuning on E85 for a healthy bump. If you’ve decked the head, installed domed pistons or upped the boost, then you can really take advantage of the additional octane in pure racing fuels.

Not a subscriber? Each magazine contains all the latest amateur motorsports stories. Subscribe to Grassroots Motorsports now.



(281) 447-7200

RC Injection
(310) 320-2277

Shell Racing Solutions
(888) GO-SHELL

Join Free Join our community to easily find more Mazda articles.
View comments on the GRM forums
Antihero New Reader
3/24/16 3:45 p.m.

Very interesting

Vigo PowerDork
3/25/16 11:58 a.m.

So, basically, if you don't have high compression or forced induction, you don't get much from race gas. Hard not to be snide at this point..

I was interested to see the methanol and e85 results, at least. I guess the whole thing is more relevant to people who have a really hard time gaining 4 hp due to class rules. The only thing limiting the power of everything i own is my own laziness, pretty much.

What i'm curious to know is whether i'm going to gain power going from race gas to e85 on a turbo minivan that's maxed out the airflow of the turbo. Hopefully time will tell.

rslifkin HalfDork
3/25/16 12:01 p.m.

Pretty much what I'd expect. More octane only helps if you're knock limited and having to pull timing on the lower octane fuel. And oxygenated fuels need to be fed in larger quantity, but make more power (hence E10, E85 and methanol making more power).

Durty New Reader
3/25/16 1:07 p.m.

In a race if you're running E85 of M1 aren't you going to have to carry twice as much weight? At what point is that worth it over a 20-40 minute spec miata sprint race? Carrying 10 gallons vs 20 gallons seems like it can really add up.

iceracer PowerDork
3/25/16 5:19 p.m.

Methanol in particular is a real pain to use. The dirt track racer were using it but they all went back to gas.

codrus Dork
3/25/16 6:47 p.m.

What's missing from the article is a good explanation of why you see the power increase with alcohol. Yes, it increases octane and available timing, but the article states at the beginning that even on the 87 octane cheap gas the engine was still able to reach MBT without inducing knock.

As I understand it, the win comes because you get more energy burning hydrogen atoms than you do from burning carbons. Since otto-cycle engines are ultimately limited in power by how much atmospheric oxygen they can intake, the fuel with the highest H/C ratio will give the most power. Methanol gives you 4:1, ethanol 3:1, and gasoline something around 2.5:1 (it's a blend of a bunch of stuff). Pure hydrogen would be even better, but it's got some challenges in adapting the fuel system. :)

The oxygenation in the alcohols is a red herring, while it increases the number of oxygen atoms in the combustion chamber that doesn't net you any more power because the oxygen came in already 'burnt' (that is, already bonded to a C on one end and an H on the other). You had to put in much as much energy to break those bonds and free the oxygen as you get out of reforming them, so it's a zero factor.

secretariata Dork
3/25/16 9:38 p.m.

Man, when I saw it was an article by Per I thought he was back on staff and did a little "Whoop". Not to dis the current staff 'cause I love the magazines, but I miss Per's work being there to enhance the mags that extra bit.

iceracer PowerDork
3/26/16 11:07 a.m.

In race cars, one advantage of alcohol is that it tends to run cooler and the anti-knock quality allows higher compression or boost.

An engine needs to be built to run alcohol.

chiodos Dork
3/26/16 3:39 p.m.

In reply to codrus:

Thanks for explaining that indepth and all sciency but not too egg head so i could still understand!

noddaz Dork
3/27/16 12:01 p.m.

Interesting story... Thank you for writing it. And maybe I even learned something...

bmwpc New Reader
4/7/16 4:09 p.m.

I read this story in December when it was published and was very impressed with the test. However, after reading it a second time I think there is a major failure in this report, not the stats. The test should clearly indicate that the results are only applicable when the vehicle can be tuned for each fuel. So if you are applying this to your street car or any other vehicle for which the mixture, timing, fuel pressure, etc. cannot be individually adjusted, the results are mostly not applicable. The perfect example is the first 2 fuels tested the 93 and 87 octane E10. If your car comes from the factory tuned for 91 or 93 and then you run 87 (regular) your vehicle MAY compensate for it and not knock but that likely means it will cut the timing and some power. Now if any of my M cars which call for 91 WOULD only loose 1 per 100 horsepower and the fuel cost 10-15% less in todays market, I'd be running the 87 all the time. But in most cars not quite as sophisticated or highly tuned as my M's but calling for 91 may not adjust easily to 87 and run poorly or loose more power. That's why the most important part of this story is not necessarily the actual power differences BUT that if you have the ability to adjust all the engine setting, some fuels will produce varying horsepower, torque and consumption. I think that this is mostly missed by the way the "results" are presented. Looking at the results why would anybody ever buy premium fuel if it only cost them less than 1% in power but saved them 10-15% in price. The reason why they don't is that most of us drive our cars everywhere and the results can be much graver that a 1% horsepower loss. So if you have a race car who's fuel system is fully tunable, this report is very helpful. If you don't have a race car like that or are thinking that this applies to your daily driver then the article has clearly confused your understanding of the test.

z31maniac MegaDork
4/7/16 4:24 p.m.

^Because that's that specific car. A Miata with relatively low compression compared to today's vehicles.

I can promise my BRZ which runs 12.5:1 compression stock, will lose more than 1% on 87 from the computer pulling timing.

iceracer PowerDork
4/7/16 4:49 p.m.

I once had a Chrysler that was designed to run on high test (93 today) ,I found that I could run it on regular (87) as long as I kept a light foot. Too much throttle and I would get the marbles in a can rattle It was just an experiment.

rslifkin HalfDork
4/7/16 4:52 p.m.

Pulling timing to run on lower octane fuel has an MPG penalty on top of the power penalty. So in the end, it may or may not even be cheaper.

The only reason any cars are sold at all that can run on 87 is consumer perception. Very few modern engines can't benefit from 89 or 91 with proper tuning. Consumers think that paying a few cents more for 93 is ridiculous and means the car is more expensive to feed, even if it gets better mpg.

Scargod New Reader
4/7/16 5:44 p.m.

I really wished there had been a test of a fuel additive like Torco. I understand this engine didn't "need" more that 93 octane. While I need just a tad more than 93 offers I cannot get and don't want to carry round large quantities of E85. My solution is adding 12 ounces of Torco to five gallons of premium pump gas. I wonder how many others do something similar and how it squares with pricey 100 octane racing fuel? This test doesn't help much (it seems) for us with turbos that are cranked up.

Rupert Dork
4/8/16 9:04 a.m.

In reply to Scargod:

I agree. I run a supercharger on my Miata. Like you, since I put a lot of street miles on my car it isn't real practical for me to run super fuel.

GRASSROOTS & CLASSIC: Please do a follow up test with 93 and additives!

mad_machine MegaDork
4/8/16 11:23 a.m.

pretty much what I expected to happen.. except for the 93octane non-e. I did not expect that to perform like the 87 octane stuff.

I got into a few internet Arguments about using 93 octane on the street. While I had no numbers to back up HP and torque, I had solid, above EPA estimates, MPG numbers to to show that 93 octane gave my stock, 1999 Hyundai Tiburon a good 3 to 4 empeegees above the 87. That alone told me that the higher octane stuff was making more power by keeping the computer from pulling the timing

Happy Carmore
Happy Carmore MegaDork
4/9/16 7:10 a.m.

Nice to see a scientific study show that moonshiners weren't crazy in running white lightning in their car's tank as well as their own.

bsaney None
4/14/16 11:03 p.m.

In all of my Honda cars , 87 octane wasted money and 91 was too expensive for the return, so normally 89 did the trick. And a 3 to 1 mix of 89 to 91 was the best for power and economy all round. You could feel the difference. Just my 2 cents.

frenchyd Reader
4/15/16 5:17 a.m.

In reply to the staff of Motorsport Marketing: Here's one reason why I buy your magazine at the newsstand.. I too like Alcohol fuels but you should have mentioned how toxic Methanol is compared to Ethanol. While it is the cheapest per gallon fuel the nearly double requirement raises it's cost up there.. Yes it does make more power than ethanol but the ease and low price of E85 more than offsets it's power and lower cost.. I treat my E 85 to a 5% upgrade with race gas. It increases the octane by a significant amount while adding nominally to cost. Pure Ethanol is 114 octane. Imagine how low octane the gas blended with ethanol is to drop it to pump numbers!!!!!! A little 114 Octane seems to make a real difference..

I'd love to see you do a test of that same engine Turbo/supercharged, where the real advantage of alcohol would come out..

foxtrapper UltimaDork
4/15/16 6:22 a.m.
bmwpc wrote: ...The test should clearly indicate that the results are only applicable when the vehicle can be tuned for each fuel. So if you are applying this to your street car or any other vehicle for which the mixture, timing, fuel pressure, etc. cannot be individually adjusted, the results are mostly not applicable.

I rather agree with you. The article almost wasn't on fuel, but more on engine tuning for a particular fuel.

Kinda thought the article made much ado about essentially nothing with regards to the tiny differences in base gasoline fuels. Especially in light of the substantial difference methanol gave.

Our Preferred Partners