2009 In Review – Fuel Usage Data

Quite a while ago I analysed the fuel consumption in the races over the season so far to see whether the fuel consumption figures quoted by the F1 Yearbook were correct.  At the time, I found that the F1 Yearbook figures were overestimating the stint lengths by about one lap on average.

So, now it’s time to revisit these figures.  This post forms the first of a series of posts reviewing the 2009 season with the other posts on teams and drivers – hopefully I will find the time to write them all before the 2010 season begins ;)

As a bit of background on the calculations, please click this link.  If you trust my numbers however, read on! :D

To eliminate erroneous data, all stints shorter than expected by more than three laps were removed because when I looked at the data, this seemed to be where most of the stops happened.  Anything more than three laps short was deemed to be a strategic stop or a repair stop and was removed from the data.  In addition, any race that was wet or had a safety car before the first pit stop was not considered as wet races lower fuel consumption and safety cars usually force drivers to pit early.

However, even with a clean dataset, when we align the predicted stints with the actual ones from the races, there seems to be a difference, as shown by the graph below:

2009 Fuel Use Race KG

This graph shows the average figure calculated from all the data collected vs. the F1 yearbook data.

The graph below shows the difference between the two datasets, by calculating the difference between the actual data and the F1 Yearbook data:

2009 Fuel Use Difference Laps

The weird thing is that some races were predicted spot on and some were way off.  Suzuka (Japan) and Germany were one of the largest difference.  Fine, you might say, these circuits have not been used since 2006 and 2007 respectively and the engines were not frozen until the end of 2008.  But then Monza (Italy) was, along with all the others, so there must be something more to it.

Incidentally, the range of values mean that the average difference is almost identical to the average F1 Yearbook figure, so I plotted the differences on a histogram using the mean and standard deviations as my ‘bins’.  The mean was 0.05 laps difference and the standard deviation was 1.38 laps difference.  Here is the histogram:

2009 Fuel Use Difference Histogram

So we can see that most (9 out 12) values lie within –1.33 to 1.43 laps difference from the F1 yearbook values, with the rest a bit further out.

What does this tell us?  Well, it means that the F1 Yearbook figures are pretty accurate, because even after a large regulation change which changed all the aerodynamics, included KERS, slick tyres and froze the engines, the figures are still very accurate!

As a part of this, I tried to find out which team was the most frugal and therefore which engine, using the same dataset as before.  For the teams we have the following:

2009 Fuel Use Team KG

And for the engines:

2009 Fuel Use Engine KG

On the teams, it is worth noting that all the customer engines use more fuel than the ‘works’ teams do.  Maybe this is where KERS comes into play – as Ferrari, McLaren and Renault used it while the customer teams didn’t?  Although, personally I am not sure that a power boost device can save fuel, but if you have some thoughts on the matter please do let me know in the comments!

The Renault is the most frugal by about 0.05kg/lap over the Mercedes, which is the same over Ferrari.  Put it this way – with an average stint of 23 laps, this means that an extra lap can be gained with such an advantage.

So with an equal amount of fuel in the car, the BMW would have to pit about 3 laps earlier than the Renault, as shown on this graph:

2009 Fuel Use Extra Laps

Looking at it this way, you can see that Red Bull would have used it to an advantage over Brawn and Mercedes cars over Ferrari too – so in theory if one was close to the other they could use the extra lap as a hotlap and hope to make a big enough gap after they have made their stop.  Sadly, the rules forbid this in 2010 but with frozen engines, you can see that Ferrari are going to have to run bigger tanks (and therefore a heavier car) to be able to fuel up until the end of the race.  Apart from making the laptimes a little slower, it also helps wear the tyres and brakes faster as they have to do more work (and maybe the transmission too?).

Last month James Allen referenced an Auto Motor Und Sport article on his website (here) which used acoustic analysis and GPS to determine which engines were the most powerful.  The article concludes that the Mercedes engine is the most powerful with apparently 755hp and there is a 2.5% spread between the engines, making the least powerful engine 18hp down or about 3 tenths of a second per lap.  This can be enough to knock you out of qualifying so it is important…

Anyway, the balance of power was Mercedes (most powerful), BMW and Ferrari close behind, then Renault and Toyota.  This means that frugality does not necessarily equate to lesser performance – and makes the Mercedes one heck of an engine as it powerful and frugal.  No wonder Christian Horner of Red Bull wanted to swap the Renault engine they use for one!

This would also make the Toyota the worst engine as it is down on power and not frugal – I would suppose that even if Toyota did continue in 2010 that Williams would not have renewed their contract with them.  Ferrari and BMW could get away with wastefulness with the extra power they had in hand too.

So which engine would you choose for 2010 and which will you miss?  How do you think Cosworth will fit in to this data?  As always, let me know in the comments ;)

Note: at the time of writing, only Ferrari, Mercedes and Renault have engines entered into the 2010 championship with a few teams running Cosworth engines, as mentioned in the James Allen and Auto Motor Und Sport articles referenced above.

About these ads
This entry was posted in F1, Qualifying, Race, Summary and tagged , , , , , , , , by rubbergoat. Bookmark the permalink.

About rubbergoat

Hi there! I’m a mad keen F1 fan who has been addicted to the sport for 20 years. I watch every race and follow the sport in every way I can. I have a keen interest in numbers and I would like to analyse the races from a statistical point of view to see if the data shows something we can't see on TV. As always, I’d love to hear what you think and especially if we can discuss my analyses that would be great – but please no nasty stuff!

32 thoughts on “2009 In Review – Fuel Usage Data

  1. Pingback: Tweets that mention 2009 In Review – Fuel Usage Data « Making Up The Numbers -- Topsy.com

  2. I’m gonna be honest, when I saw the formula I was expecting to be bored rigid (and I can imagine it might put people off reading further) but this was actually very revealing, so good work!
    The Avg consumption graphs at the end are most interesting, there isn’t a huge gap but it seems to be enough to make a difference especially with the tight grids we’ve had this year.
    I’m not sure what to make of Cosworth, they say they’ve only taken the bare bones of the engine and have reworked it for the new regs so I guess we’ll just have to wait and see. I can’t see them being all that competitive though.

    • I might remove the formulas then because I have linked to them anyway – thanks for the feedback Pat!

      • No probs!
        The other thing that surprised me was that *every* customer team was worse off than the parent. You’d expect it for some, if not most, but I’d never have thought that it would be true for all of them!

  3. No probs!
    The other thing that surprised me was that *every* customer team was worse off than the parent. You’d expect it for some, if not most, but I’d never have thought that it would be true for all of them!

    Yeah me too. Do you think it might be KERS or are the manufacturers being a bit naughty and detuning the engines somehow?

    • I don’t think it is either, more that the manufacturer teams probably have more engine people involved with the cars and some may keep information back in case the customer changes suppliers next year. Maybe the lubricants also make a difference.

  4. Pingback: 2009 In Review – Fuel Use « Making Up The Numbers Reviews Me

  5. Pingback: Potrosnja goriva | Slike

  6. Pingback: Thursday Thoughts – Engine Parity « Making Up The Numbers

  7. I wonder if part of the reason was that, if I remember correctly, the customer teams had more occasions where they had incidents forcing an early pitstop (and hence become impossible to calculate under the formulae used)?

    • Hmm I am not sure. I managed to get a lot of data for customer teams – in fact, it was better if an engine had more customers because by the law of averages it lessens the chance of getting bad data, no?

  8. Pingback: Ferrari F2010 (F61) Development Update | Unofficial Ferrari F1 Forum News & Fan Discussion

  9. Pingback: 2009 In Review – Driver Classification « Making Up The Numbers

  10. Pingback: 2009 in Review – Driver Qualifying « Making Up The Numbers

  11. I think the testing ban really shoved us how bad teams can go when quessing the fuel consumption. Remember when Massa lost many places at the end of the race coz he almost ran out of fuel.

    I quess F60 was such a aeroDog.

    2010 i dont think they will have such problems. But ofcource they r playing the catch up game, so we will see how they cope with fuel.

  12. “Maybe this is where KERS comes into play – as Ferrari, McLaren and Renault used it while the customer teams didn’t? Although, personally I am not sure that a power boost device can save fuel, but if you have some thoughts on the matter please do let me know in the comments!”

    The KERS should help fuel economy. The energy of braking is stored for later use, thus in principal saving fuel (or extra power to finish the lap sooner and not use fuel as long). Although the KERS cars are heavier, this process should improve fuel economy overall and is the bases for the so called “green cars” using a hybrid system.

  13. Pingback: RED BULL ¿tiene alas? ; puede que si « Bernoulli

  14. Pingback: Renault engine in the RB6 – Good or not? « Gives You Wings

  15. Pingback: Motores Mercedes e Renault podem largar na frente em 2010 « F1 Around

  16. Did you know that Williams published a fuel consumption estimate for every track before each Grand Prix? They used data normalized for 5 kilometers, which is indeed probably more suitable to use for calculating the average consumption figures’ of different engines (not that it matters much).

    Your analysis is very welcome before the 2010 season and now that we have the supporting data, let’s take note that Red Bull and Renault will have about 7 (!!) tenths of a second over Ferrari and 4 tenths over the Mercs at the beginning of a race attributable to their lower weight, since carrying an extra lap of fuel costs about a tenth.

    If overtaking gets somewhat easier in 2010 and the competitive teams stay close, expect to see the Red Bulls fight their way across the leaders and try to retain their lead as their weight advantage slowly disappears throughout the race. We might even see Vettel overtake someone! :)

  17. Thanks for your post.
    KERS help a lot on low fuel consum.

    But aerodinamics of car help too:
    Cars more efficients use less fuel.

    Brawn had a lot of downforce.. but this mean it need more power to push same mass

  18. It IS the KERS that makes the difference. The KERS power boost comes from energy stored from braking, rather than energy produced by burning fuel. That’s going to bring the avg fuel consumption numbers down.

  19. I drive a hybrid car (the one that Toyota have been making for 12 years now) and know a little of how it works.

    KERS can NEVER save much fuel as it can only deliver 400kJ of energy per lap. A 620kg car has 400kJ of kinetic energy is travelling at 36 metres per second – 129 km/h or 80 mph. A lot of the energy is lost in conversion and to drag and friction. However, F1 cars very rarely accelerate from a standstill – from a slow corner like turn 10 (La Caixa) at Barcelona, a 74km/h corner, adding 400kJ gets you to 149km/h. You get diminishing returns – if you’re doing 100km/h when you press the button and it was 100% efficient, you’d get 163km/h.

    Meanwhile your 2.5kg of fuel consumed on the lap contains 111,000 kJ of energy (calculated from 44.4 MJ per kg, data from Wikipedia:Gasoline). Let’s say the engines are 25% efficient, that 25% of the fuel energy ends up at the driveshaft, and that the motor is 100% efficient, so 400kJ of KERS saves 1,600kJ of fuel energy. That’s 1.4% or 0.035 kg – the margins are not big enough.

    The reason that true hybrid cars save fuel is that they have downsized, highly-efficient engines, and that the engines are efficient where they need to be for most on-road driving: low engine speeds with high load (nearly fully-open throttle).

    The electrical system is there for four reasons on a Prius – to provide a highly-efficient continuously-variable gearbox (allowing the car to run the engine at ~1000rpm when cruising at up to 60mph), to provide a boost when accelerating (reducing acceleration to 100km/h from about 15 seconds to about 10 seconds – motors have high torque output), to drive electric-only below 40mph (where the engine’s efficiency drops off due to low load, needing the throttle to close), allowing the fuel consumption to be averaged across engine on/off cycles and run the engine efficiently when it *is* on, and finally (way down the list) to capture and reuse a little energy from braking and descending hills.

    The current Prius can provide up to 27kW from the (39kg) battery and can deliver something like 0.8kWh = 2.8MJ. On braking to a stop from 70mph, at maximum regen rate, I reckon you capture about 500kJ, maybe 400kJ ends up in the battery – a lot is lost to drag, rolling resistance and other types of friction, and some charging losses. Most of the battery’s energy comes from running the engine slightly faster/under more load than the current road demand, not from braking.

    The F1 teams have completely failed to understand this or are deliberately failing to explain it; Toyota F1 didn’t use any part of the Prius system, or indeed any KERS at all, because they knew it was completely unsuitable. (GM, Mercedes and BMW continue to couple large inefficient engines to their road-going hybrids.) Indeed KERS cannot explain why Toyota F1’s fuel consumption was so much lower than Williams’.

    • I drive a hybrid car (the one that Toyota have been making for 12 years now) and know a little of how it works.

      Heh. After reading that comment I would say that you are very modest ;)

      Thanks a lot Mike, that was very informative and brilliant stuff! :D

    • Interesting stuff Mike. As far as I am aware you can use KERS to keep a car in the efficient part of the engine power curve but F1 don’t use it for that at all. I was told by somebody working for Flybrid that when they ran the maths that it was made the cars faster for them to make the cars rev higher than use less fuel.

      • The reason that KERS in F1 doesn’t save that much fuel is simply because the full throttle percentage remains essentially the same – in F1 you use KERS with the engine, not instead of.

        There is an important exception to this rule though… behind the safety car.

  20. Pingback: Five problems F1 designers face in 2010 | F1 Fanatic - The Formula 1 Blog

  21. Pingback: F1chicane » #F1 Most fuel efficent engine for 2010? #F1technical.net & 2009 Fuel Usage Data #MakingUpTheNumbers

  22. Pingback: Thursday Thoughts – On Blogging Itself « Making Up The Numbers

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s