Formula 1 2026, Malaysia: Heat Capacity

Sepang has plenty of places where a Formula 1 car can harvest electricity. The question is whether the recovery system will actually be able to use all of them. Because of the weather. 

Formula 1’s return to Sepang promises an intriguing experiment for the 2026 generation of cars. Teams are expecting air temperatures of around 32°C, track temperatures approaching 60°C and humidity above 80%. 

In purely energetic terms, Sepang is considered roughly average by current Formula 1 standards. But two long straights, heavy braking zones and a fast middle sector still demand careful management of electrical energy. 

The current MGU-K can operate at up to 350 kW, while the permitted change in battery energy remains only 4 MJ. So the electrical half of the power unit is not some huge tank that you fill once and gradually empty over the course of a lap. 

Energy is constantly being pushed backwards and forwards. And all that electricity does not travel through Nikola Tesla’s ether. It runs through very real copper windings, busbars, transistors and battery cells. 

Anyone who remembers school physics can probably see where this is going. Some of that electrical energy inevitably becomes heat. 

More power, not more capacity 

F1 engineers usually note something particularly relevant when discussing the 2026 car. The cooling architecture itself has not been completely revolutionised. But the cooling requirement of the Energy Store has increased dramatically. 

Conventional brake-energy recovery is still insufficient because the generator only acts through the rear axle, while a substantial proportion of braking work is done by the front brakes - as every responsible motorsport enthusiast has been telling people for years. 

So the regulations have created an impressively overcomplicated machine. We now have to pass a much greater flow of energy through roughly the same electrical kettle. Eventually, obviously, the kettle starts boiling. 

And whenever energy flow increases, losses become interesting. 

Suppose the entire chain (MGU-K → power electronics → battery) has a fantastically good efficiency of 97%. At 350 kW of recovery, three per cent losses mean: 

350 × 0.03 = 10.5 kW of heat. 

At five per cent losses: 

17.5 kW. 

And this is not one glowing-hot component onto which we can simply bolt a gigantic radiator. Heat is being generated in the electric-machine windings, the converters, the electrical connections and the battery cells themselves. 

Worse still, all those components have different thermal inertia and different temperature limits. 

School physics versus Formula 1 

The temperature of any object is determined not by some magical property called “cooling”, but by the balance between the heat being generated and the heat being removed. The smaller the temperature difference between the hot object and its surroundings, the harder it becomes to get rid of the next watt. 

As a first approximation: 

Q̇ = K(T_system − T_air) 

For a real Formula 1 heat exchanger, of course, K is not constant. It depends on airflow, air density, heat-exchanger design and an entire collection of other complications. But for a first estimate, this will do. 

Imagine a completely hypothetical cooling circuit operating at 50°C. With ambient air at 20°C: 

ΔT = 30°C. 

Now take Malaysia, with 32°C air: 

ΔT = 18°C. 

Everything else being equal, our imaginary heat exchanger can now reject not 100% of its previous heat load, but: 

18 / 30 = 0.6 

Just 60 per cent! 

Naturally, a Formula 1 team will not leave “everything else equal”. They can increase coolant flow, open larger hot-air exits, change duct dimensions, increase airflow through the heat exchanger. 

But cooling is never free. Larger openings and more cooling air mean aerodynamic losses. And so, once again in this season, we have bought electricity with aerodynamic drag. 

A nasty little problem 

Now we can finally drag all this modern Formula 1 nonsense back into school physics. 

Our imaginary hybrid system is recovering: 

350 kW. 

Three per cent of that becomes heat. So the cooling system must remove: 

10.5 kW.

Suppose that with ambient air at 20°C, the radiator could reject 15 kW with a temperature difference of 30°C. 

Now the air temperature rises to 32°C. 

The system itself is already at its maximum permissible working temperature, so we cannot simply allow it to get hotter. The temperature difference falls from 30°C to 18°C. 

Using our deliberately crude model: 

15 × 18 / 30 = 9 kW. 

But the system is generating 10.5. 

Can we keep recovering at 350 kW continuously? 

No! 

The battery might still be nowhere near full. The MGU-K may be working perfectly. The regulations - for once - may not have been changed halfway through the race and may still permit the full 350 kW. 

None of that matters. 

The next joule of electrical energy arrives with heat that the car can no longer get rid of. If we restore thermal equilibrium purely by reducing electrical power, we have to fall to approximately: 

9 / 0.03 = 300 kW. 

There you are. 

Fifty kilowatts of recovery have disappeared without a single component failing. We simply ran out of cooling. 

The thermal trap 

But do not rush to your conclusions, educated readers. If a device operates in pulses, which power level determines its heating? 

Peak power? Average power? Something in between? 

The answer is unpleasant for people who like dividing cars into “shitboxes” and “rocketships”: all of the above are possible. 

It depends on the duration of the power pulse and the thermal inertia of the component concerned. The MGU-K does not spend an entire lap operating continuously at 350 kW. A few seconds of hard regeneration are followed by acceleration - then another braking zone with partial load - another deployment event. 

So it is not enough to say: “Average thermal losses over the lap were X kilowatts.” 

A small component inside the inverter may heat almost to the temperature associated with peak load before that heat even reaches the coolant. The enormous thermal mass of the battery, by contrast, may barely notice one short pulse - while slowly accumulating heat lap after lap. An average loss of 10 kW tells us almost nothing about which one reaches its limit first.

The copper trap 

There is another small complication. The resistance of copper windings increases by roughly 0.4% per degree Celsius. And resistive heating follows: 

P = I²R 

Suppose a winding becomes another 20°C hotter. Its resistance rises by roughly 8%. 

At the same current, copper losses increase by almost the same amount. Which means we now need still more cooling. 

No, this does not accelerate infinitely until Formula 1 begins firing lumps of molten copper into the grandstands. The computers will notice something is wrong long before that and reduce the current. We probably will not see glowing wire. 

A loss of performance, though? Very easily. 

The battery does not have to be “full” 

Battery discussions are particularly good at producing nonsense. 

It is wrong to claim that a lithium-ion cell’s internal resistance always rises whenever its temperature rises. Going from a low temperature towards the normal operating range usually reduces internal resistance. 

The problem begins above the optimum range. 

High temperatures accelerate unwanted chemical reactions and cell degradation, so the battery-management system has to protect the cells. At that point, allowable charge or discharge current may be limited not by how much free capacity remains in the battery - but by temperature. 

On television, all of this is easily reduced to: “the car ran out of energy.” 

Except there may be plenty of energy available around the car. The problem is that you cannot safely force it into the battery. 

Why Sepang? 

Before Sepang’s final corner, the car brakes heavily and then reaches two long straights separated by the first hairpin. Energy harvested during braking therefore has an obvious value almost immediately afterwards. 

At the same time, Sepang is not regarded as an exceptionally energy-starved circuit. Williams Racing Team places it roughly alongside Montreal and Spielberg in terms of energy sensitivity. 

That makes the interesting cars not necessarily those that simply run out of electricity. Look instead for cars whose energy behaviour begins to change with temperature. A long run may become worse than a short one. Clean air may become significantly better than following another car. 

A more open cooling configuration may unexpectedly repay the aerodynamic drag it creates. And a car that works perfectly well in the morning may begin restricting recovery or deployment earlier in the afternoon. 

Humidity is actually the simpler part of the story. 

For the driver, 80% humidity is a serious problem because the human body relies heavily on evaporation of sweat for cooling - hence the need to carry the additional five kilograms associated with mandatory driver cooling. 

For a normal automotive heat exchanger, the much more important issue is the high temperature of the incoming air. Humidity changes the properties of the air to some extent. It is not the main killer of the temperature gradient. 

And that brings us to the real point. 

Sepang may make one fact about the 2026 Formula 1 car particularly obvious: it effectively has two batteries. 

The first is the real one. The electrical battery, permitted four megajoules. 

The second battery does not appear anywhere in the regulations. It is the car’s thermal reserve. How much loss can the battery, inverter, MGU-K and cooling system absorb before the next portion of electrical energy has to be restricted? 

In Malaysia, that second battery may run out before the first one does. 

And incidentally, this leads straight into another excellent subject: the balance between aerodynamics and cooling. 

If we had more time - and if the editor-in-chief were not enjoying sick leave - we would work out exactly where the optimum lies between opening the cooling inlets too far and paying for it in drag, and keeping everything tightly packaged only to lose performance because the recovery system overheats. 

So watch it yourselves. 

And think for yourselves too.

• < Назад
• Вперёд >