Estimating Fuel Mileage

Last week at New Hampshire, Kevin Harvick easily had the most dominant car, but failed to win the race. They ran out of fuel with three laps to go and finished 21st. The #4 team wasn’t the only team that gambled on gas, but they were probably the team that lost the most. Now they’re in a must-win situation this week at Dover.

Team owner Gene Haas said that the team made a “simple miscalculation” about how far they could go on the final tank of gas. As reported by Jared Turner on FoxSports.com. Haas said,

“Somewhere we were off in our calculations. That’s something we’ll be talking about tomorrow and the next couple days real hard.”

A lot of people wrote this week that Childers and Harvick have been their own worst enemies during this year’s Chase. Can a wrong calculation cost them a shot at a second title.

https://www.wylio.com/credits/flickr/3288067069How Much Fuel Gets in the Car?

NASCAR race cars have approximately an 18-gallon fuel cell. Each fuel can (at right) holds about 12 gallons.

Before a pit stop, team members carefully weigh each fuel can and record the weight. After the pit stop, the fuel cans are weighed again.

The weight of fuel that went into the car is the difference between the two weights.

To get from weight to volume, you have to use the fuel’s density.

EQ_Density

On their website, Sunoco gives the density of the E-15 fuel that NASCAR specifies for all cars and all races as 6.2 lbs per gallon. That’s accurate enough that we can suss out that the weight of the fuel inside a 12-gallon fuel can is (12 gallons x 6.2 lbs per gallon =) 74.4 lbs.  The fuel plus can weighs about 94 lbs total, which is why the gas man is usually right up there with the jackman in terms of pit crew member muscles. It takes about 10 seconds to fully empty a can of fuel into the car.

So let’s say that we put 105 pounds of fuel in the car on the last pitstop. Then we can back calculate how many gallons that is:

BSPEED_FuelMileage_ExampleCalculation

This seems like a pretty coarse calculation, but Sunoco actually provides a much more specific value for the density to the teams each day. Why would it change? Temperature and humidity can affect the density of any gasoline – which we talked about the last time we were in Dover when Kurt Busch’s team was told they weren’t allowed to try to cool down the fuel during an unseasonably hot race.

This is how regular gasoline changes density as a function of temperature. When the temperature rises, the molecules in the gasoline like to have slightly more space between them, which means that there are fewer molecules in a given volume. If you had a row of seats that normally sat six and you insisted on having an empty seat on either side of a person, you’d end up with only three people in the row instead of six.

BSPEED_ChangeofGasolineDensitywithTemperature_Gasoline

The E15 fuel that NASCAR uses has a similar change in density with temperature. Humidity can also impact the density because ethanol likes to absorb water. So each morning, Sunoco distributes a very precise (measured) number to the teams and that is the number they use. It’s an incredibly precise (and simple) calculation to make.

Theoretically, you’d have to make a pretty massive mistake to screw up that number.

But the big uncertainty in fuel mileage calculation is not in the calculation – it’s in the assumptions that go into the calculation.  There’s really two levels of assumptions, too:

  1. How much fuel got into the car and
  2. How fast is it being used?

And, as the diagram below shows… there are a lot of places to screw things up.
BSPEED_FuelMileage_CalculationDiagram2

Gas In

Spillage and Leakage: We used to have an outlet on the gas tank so that any overflow would go into a catch can. That amount was put back into the gas can and weighed. The new models of gas cans are supposed to prevent spilling, but if you don’t get the can coupled to the gas inlet correctly, or if there’s a malfunction with the gas can, you can spill gasoline on the ground or on yourself. That’s gas that doesn’t get accounted for when the cans are weighed. And since everything happens quickly, it’s possible no one can tell how much was spilled well enough to make a good estimate. It’s also possible that there’s a leak in the fuel line or the fuel cell, although that’s a rather more serious issue because that goes to safety.

Miscalculation: It is surprisingly easy to screw up a calculation when you’re in the rush and crazyness of a race weekend, especially if you’re short-handed, new or a lot of things are changing – or even if you have a new baby and you’re not sleeping much. Someone reads the scale wrong and reports the wrong weight. Someone messes up the excel spreadsheet where the calculation is done, or you’re using the wrong density for some reason. These are the most irritating possibilities because, frankly, they are preventable.

Hold that last thought because I’m coming back to it later.

Gas Out

Here’s where things get really tricky. You’ll hear commentators throw about a number like 4.6 miles per gallon under green and half that under caution. In reality, the teams try to make much better estimates of these two numbers because they are absolutely critical.

Measuring Gas Mileage: Most drivers don’t get the same mileage at all tracks. Gasoline usage is tracked carefully during practices. Gas mileage changes with the condition of your tires, so every time the tires are changed, the gas mileage calculation is done anew. Over the course of a couple hours of practice, the crew chief gets a snapshot of how his or her particular driver’s style at that track with that particular set up uses fuel.

But when you’re racing, some of that goes out of your hands. If there’s someone on your tail and you don’t want to give up a position, the driver may be more aggressive, which means he or she is using gas at a faster rate.  The teams get scads (a techincal term) of data from the Electronic Fuel Injection system that they can mine for more information about how their driver and car use fuel at different tracks.

Cautions: Is it me, or has there been a lot more talk in recent years about drivers’ abilities to save fuel under caution? They’ve always played games like flipping off the ignition and coasting during cautions, or driving on the apron to decrease the distance traveled. Now we’re hearing more about getting off the gas a little earlier, getting on a little later. I think a lot of that focus on finesse is due to the EFI date. After a race, you can go back and show your driver how good he or she was at saving gas.

So What Happened to Harvick?

You have to understand that, for an engineer, being accused of screwing up a calculation is like being stabbed through the heart. It’s a simple calculation and you’d have to be pretty far off your game to screw it up.

So I wasn’t all that surprised when Rodney Childers tweeted an explanation of what went wrong and used some choice words in reference to the people calling for his head for costing Harvick a chance at winning the race.

BSPEED_2015_NHMS_ChildersFuelMileageRaceChilders was able to look at the EFI (Electronic Fuel Injection) data and determine that Kevin had saved enough fuel that they should have had six laps worth of fuel left. They should have been able to finish the race with fuel to spare.

But they were short by three laps.

This can only mean that they overestimated the amount of fuel that went into the car.  Childers suggested the crew hadn’t gotten all the gas from the can into the car or that possibly a fuel cell bladder might have “come apart”. It’s also possible someone misread the scale or mis-recorded the weight.

As he points out, these kinds of mistakes are pretty rare (and I add, especially with top-level people like Childers and his crew) – but it couldn’t have come at a worse time.

Ultimately the crew chief takes responsible for everything except driver decisions, so it does fall on Childers’ shoulders – but all the commentaries attacking him for using bad judgement at New Hampshire are just wrong. He made the right decision based on the data he had in front of him. That data was incorrect. The problem isn’t the judgement – it’s the data.

And you can bet they are going over every step of every process this week at the shop because Dover is their last chance to keep their hopes alive for a repeat championship.

Fuel mileage concerns never go away. Just for kicks, I wanted to see how much fuel you needed for each of the tracks in The Chase.

BSPEED_FuelMileage_FuelforOneLapatChaseTracks

Remember that there are 128 ounces of fuel in a gallon, so even at the longest track in the Chase, you’re talking about a little more than a half-gallon of fuel for that last lap. If you think about the times someone’s run out of fuel coming around turn 4, divide those numbers by four and in some cases (looking at you, Martinsville), as little as a quarter of a cup of fuel could lose you a race. This is a sport not only of tenths and hundredths of seconds, but also of quarter and half-cups of fuel.

Disclaimer: Back in 2008, Rodney Childers was one of the crew chiefs for Elliott Sadler and I spent a weekend with the team at Daytona while writing the update for the paperback version of my book. His desire to win coupled with a measured calm and can-do attitude makes him a person you really can’t help but like and admire. So I’m probably biased toward him, but I think my analysis still holds up!

4 thoughts on “Estimating Fuel Mileage”

  1. There is still no excuse for the miscalculation. Especially missing it by 3 laps. That’s huge. I’ve weighed fuel at 3 am to double check the density. If fuel was left in the can they should have caught it when they put the empty can on the scale.

    The fuel guy is well paid. He and the crew chief at the top of their profession. The top people in any field generally do not screw up. Generally.

    1. There generally will always be fuel left in the last can. The tank is 18 gallons and each container is 12 gallons, so even if the tank was totally empty, you’d have fuel left. While it’s possible someone misread the scale, it’s also possible that fuel was spilled. There’s no good way to estimate how much. The problem wasn’t with the calculation – it was with the inputs into the calculation. I guarantee that they’ve been over and over the pit stop to try to figure out where the problem was.

      You’re right that everyone is well paid, but anyone under pressure can (and sometimes does) make mistakes. How many times have we seen a driver lose a win because they sped on pit road? If they found out who screwed up, I suspect that person would’ve been working a different job in Dover.
      Thanks for reading, Richard – your comments are always welcome!

  2. Fascinating article! On Denny Hamlin’s team communication I heard who I assumed was his crew chief telling him to “Push the 4. I don’t think he can make it (on fuel). I think there were about 20 +/- laps to run. Was the crew chief simply motivating Hamlin to run harder or did he know Harvick didn’t have enough fuel, meaning he knew something Rodney Childers didn’t. Either way, interesting.

    1. The teams not only have their own communications, they listen to the other teams’ radio as well. The 11 may have picked up on something said on the 4 radio – and it could have been something as innocuous as “save some fuel for me, bud”. The eavesdropping has led to some teams coming up with their own code so that other teams can’t pick up on their strategy. You’re right that the crew chief may have been trying to motivate Denny as well. I don’t think there’s any way the 11 could’ve known what happened with the 4 in terms of the fuel issues. Thanks for reading!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: