Why refuelling wouldn't improve F1's racing

Refuelling returned to the agenda of Formula 1's chattering classes last week thanks to a combination of FIA president Jean Todt's suggestion that such a move should be analysed and support for this from the Grand Prix Drivers' Association.

But while refuelling is not without its positives, for the drivers in particular it would be a mistake for grand prix racing to go back there. The arguments in favour are that it would allow drivers to push more during races, a laudable aim, and would help reverse the trend of modern GP cars becoming increasingly bloated.

It would also introduce a new strategic variable that, so the argument goes, would make the races more interesting. Cars would indeed spend far more of the race closer to their ultimate lap times and the space saved by not requiring a fuel cell to lug 105kg around should also confer a small car-size gain because the fuel tank would be smaller.

But fuel is not at the root of the cars gaining weight. From 2009 to today, the minimum weight of car plus driver (including their race kit) has increased from 605kg to 743kg. That's an increase of 22.8% in 10 years. But fuel is only part of this equation, and it should be noted that when refuelling was banned for 2010, when cars carried significantly more fuel than today, the increase was just 15kg over '09.

A far bigger contribution to the minimum weight has been made by the introduction of measures such as the halo, the common crash structures, wheel tethers, and the vastly more complex power unit package plus its associated coolers.

The main impact would be on race weight, given the need to carry a race's worth of fuel, and it's true this would ameliorate matters, but not in a way that represents more than fiddling around the edges.

The drivers are right that the weight of the cars is a problem, but less so in terms of lap time than in terms of how the cars look on track. This generation of F1 cars still looks spectacular when on the limit on a qualifying lap but seem lazy compared to the razor-sharp machines of 15 years ago.

It's a reminder that lap time isn't the key factor in whether a car looks quick to the naked eye, something F1 learned in 2017 when it focused all its efforts on regulations designed to make the cars significantly faster in the belief it would cure all ills.

Let's take a theoretical 60-lap race at a track with a 90-second lap time in flat-out trim. Today's cars would start with the maximum 105kg (ignoring the fuel required for the slowdown lap and requisite fuel samples - not to mention any advantage from starting with less than the maximum race allowance) but, if refuelling was allowed, they would start with 52.5kg for an even-split one-stopper, 35kg for a two-stopper and 26.25kg for a three-stopper.

Carrying 105kg of fuel will be worth just over 3.5s in terms of lap time penalty. In the real world, that's compounded by the need to protect tyres, manage engines and other race-situation factors but it's a useful ballpark figure.

Convert that to a one-stop race and the time loss is just under two seconds per lap, for a two-stopper just under 1.4s, and a three-stopper is just under one second.

For a race situation, which strategy works best would be a function of track position, the time required for a pitstop based on track configuration and the permitted refuelling rate and relative pace.

But in this hypothetical situation, the difference in peak lap time between a car starting on a one-stopper versus a three-stopper is about nine tenths using 0.35s/10kg as the loss.

Quantity of passes does not equal quality of racing, so you can have a great race with fewer overtakes. The correlation is strong enough that refuelling mitigates against on-track moves

There are other factors at play though, and it's true that the management of the car, the brakes, and the tyres would be less of a challenge in the early stages of the race than it currently is with a lighter fuel load, so drivers would push more than they currently do. This would have knock-on effects that could be beneficial.

But what it wouldn't do is guarantee flat-out racing. Even in the days of refuelling, lifting and coasting still happened because efficiency is a crucial part of the performance equation.

You always want to carry as little fuel as you can get away with. Those who repeat the 'flat-out' mantra often overlook this, even though it is fair to say that grand prix cars today are not driven close to their ultimate lap time anywhere near enough these days. There was also the problem of races playing out almost entirely based on strategy.

Overtaking was incredibly difficult, even with offsets in fuel load, so the majority of position changes happened in the pits. This meant the real flat-out, on the edge, laps tended to be the out-laps, in-laps and anything in clear air during the pitstop sequences. Today, in these days of the ubiquitous undercut that does sometimes mitigate against varied strategy, that's still the case.

And while the power of the undercut would be negated by refuelling, that's still not a net gain even though F1 supremo Ross Brawn has talked of how much he enjoyed the era where he would order Michael Schumacher to nail a sequence of laps in clear air to defeat a rival.

And F1 doesn't exist solely for those in the cars and on the pitwall. Add to the mix the fact that teams' simulation tools are vastly more powerful and there would inevitably be a convergence of strategy.

The laws of physics mean there will be a quickest theoretical way to the chequered flag and this means, as always, relative car pace is the key factor in defining results. That's largely a function of resources and, as the drivers themselves have pointed out, this is connected to the share of F1's revenue among teams.

It's also too easy to romanticise the last refuelling era by boiling down 272 races into a handful of occasions when the strategy created a memorable race.

That's the equivalent of judging the current season by the closing stages in Austria or the frenetic racing at Silverstone but ignoring what happened at Paul Ricard.

There were some remarkable races from 1994-2009. Take Hungary '98 when Ross Brawn switched Michael Schumacher to a three-stopper that required him to pull 20 seconds on David Coulthard in 19 laps, or France '04, when Schumacher went to a four-stopper to defeat Fernando Alonso.

There are other examples, but plenty where refuelling had no impact or was even a negative. Then there was Suzuka 2000, when Schumacher and Mika Hakkinen went toe-to-toe for the title in a flat-out classic, which was won by the Ferrari driver emerging ahead at the final pitstops.

Again, that's the pure battle that was great for those involved and the most engaged of the fanbase, but not the millions beyond that. For every classic, there are many more mundane races.

F1 has a habit of chasing panaceas that make matters worse, and refuelling would be another

As for Suzuka 2000, how might that race have panned out if jumping ahead in the pits with a longer middle stint was not an option for Schumacher? It's a classic example of the disincentive to pass on track in favour of doing so in the pits.

Refuelling, frustratingly, privileges pace over battling, and strategy usually depends on pace when not close to the rival you are actually fighting. Overtaking statistics in F1 are deeply flawed, but the numbers were poor during the refuelling era.

In the last season of refuelling in 2009, the average per-race figure was 13.2; in '10, with refuelling banned, it rose to 23.8; in '11, with no refuelling and the arrival of Pirelli high-degradation tyres and the DRS, it increased to 43.2.

Quantity of passes does not equal quality of racing, so you can have a great race with fewer overtakes. But the correlation is strong enough that refuelling mitigates against on-track moves, even though adding it to the current rules package would generate more passes than from 1994-2009.

What's more, during that period, the strategic divergence was not so great. On 77% of occasions, the top two made the same number of fuel stops, and 60% of the time the top three had the same strategy.

Yes, there were variations in stint lengths within that, but it wasn't a constant battle of varied strategies.

Still not convinced? The clincher is that during that last refuelling era, the powers that be introduced the rule forcing drivers to qualify on their race-start fuel load to distort the grid because refuelling was not magically creating the desired variety and excitement.

There will be ways to make the cars lighter, to allow drivers to push more and to create a closer competition of exactly the type the drivers are calling for while also guaranteeing that those outside the cars enjoy it as much as those in the cockpits. But it's another one-dimensional 'solution' for a 3D problem.

Refuelling offers some benefits for the drivers - who necessarily have a particular perspective - but overall it would be a net loss for what the majority of those who make up the viewer base want.

Fortunately, the impact of refuelling has been well understood and is very unlikely to return in the short or medium-term - and if it does one day, it needs to be part of a wide-ranging change rather than just a hit-and-hope introduction.

F1 has a habit of chasing panaceas that make matters worse, and refuelling would be another.