2025/2026 Hybrid Powerunit speculation

[wuzak]

There is already multiple times the energy storage than the regulations allow (4MJ) for the cell weight allowed.

Is it future proofing for a possible modification in the regulations later, where more discharge would be allowed from battery ?

It's likely that there will be a reduction in recovery/deployment energy allowed in the future as the 2026 rules are already extremely difficult to recover enough energy.

It would require a massive increase in recovery power for it make sense, but only then if the discharge power is lower.

Otherwise you end up with the 2026 situation, but with more power.

[gruntguru]

[(a) Overall weight will decrease by ~30kg.
(b) If ICE energy sent to the wheels is to fall from 700 units to 300 as you say, the 2026 ICE max power must be less than 300/700 (43%) x 2025 ICE max power. (In energy terms, the delivery for each ICE will be the same when demand is equal or less than the 2026 max power.) At a rough guess, your numbers would put 2026 max power at 35%-40% of 2025 max power! I don't see anyone predicting that! In fact ICE power is predicted to be more than 375 KW - down from 630 KW (so 60% not 35%)

Many are predicting more than 375 KW, putting peak PU output greater than 2025 in a lighter, more slippery car. These cars will accelerate harder and have a higher top speed. High speed cornering will be slower.

My numbers were just heuiristics, but even then, I was talking about 'energy units', not horsepower.

Yes - so was I (as well as power). Perhaps you should read my post more carefully.

The peak horsepower from ICE was 800hp in 2025, the same is going to be 500hp in 2026, as the regulations clearly say (80% share to 50% share). I am not sure how you arrived at 35% (I never claimed any peak HP number) from my 'energy units' numbers. My claim is - because the share of the fuel+charge energy split has removed the bias towards fuel, it's has a double whammy effect - not only is the fuel's overall contribution weakend, it has to spend much more than in 2025, towards recharging the battery - a double whammy, resulting in a much lesser contribution by fuel to the wheels.

This assumes the fuel energy available is the same - which it is not. There is no fuel limit only the instantaneous "fuel rate limit" (up to 3,000 MJ/hr). The 2026 ICE will be operated at or near maximum output for a much greater percentage of a lap resulting in a similar or greater energy output per lap.

[Cold Fussion]

Many are predicting more than 375 KW, putting peak PU output greater than 2025 in a lighter, more slippery car. These cars will accelerate harder and have a higher top speed. High speed cornering will be slower.

Isn't it more likely that top speed will be around ~345 km/h (the mgu-k cut off speed)? To much higher than that they will need to be substantially lower drag to makeup for the ~1/3rd power loss vs 2025 (I don't know what the projections are for the drag reduction in the straight line mode vs 2025 in DRS).

[gruntguru]

Yes.

Top speeds are currently much lower than 345.

[wuzak]

Yes.

Top speeds are currently much lower than 345.

Except at a few tracks, such as Monza, with very low downforce settings, or Mexico City, where the air is thin.

The low downforce mode should have comfortably less drag than the 2025 cars with Monza wings.

[venkyhere]

[(a) Overall weight will decrease by ~30kg.
(b) If ICE energy sent to the wheels is to fall from 700 units to 300 as you say, the 2026 ICE max power must be less than 300/700 (43%) x 2025 ICE max power. (In energy terms, the delivery for each ICE will be the same when demand is equal or less than the 2026 max power.) At a rough guess, your numbers would put 2026 max power at 35%-40% of 2025 max power! I don't see anyone predicting that! In fact ICE power is predicted to be more than 375 KW - down from 630 KW (so 60% not 35%)

Many are predicting more than 375 KW, putting peak PU output greater than 2025 in a lighter, more slippery car. These cars will accelerate harder and have a higher top speed. High speed cornering will be slower.

My numbers were just heuiristics, but even then, I was talking about 'energy units', not horsepower.

Yes - so was I (as well as power). Perhaps you should read my post more carefully.

The peak horsepower from ICE was 800hp in 2025, the same is going to be 500hp in 2026, as the regulations clearly say (80% share to 50% share). I am not sure how you arrived at 35% (I never claimed any peak HP number) from my 'energy units' numbers. My claim is - because the share of the fuel+charge energy split has removed the bias towards fuel, it's has a double whammy effect - not only is the fuel's overall contribution weakend, it has to spend much more than in 2025, towards recharging the battery - a double whammy, resulting in a much lesser contribution by fuel to the wheels.

This assumes the fuel energy available is the same - which it is not. There is no fuel limit only the instantaneous "fuel rate limit" (up to 3,000 MJ/hr). The 2026 ICE will be operated at or near maximum output for a much greater percentage of a lap resulting in a similar or greater energy output per lap.

this (bold) is I think where the 'disconnect' in our perceptions stems from : I think it will be very difficult to have an implementation where the the A-pedal input from driver 'does not linearly (atleast roughly) translate' to the operating point of the ICE. The driver relies on NVH to 'tune' himself to the power delivery of the car, and operating the ICE near peak power when throttle isn't 100% pressed (or even 0 pressed) is going to 'interfere' with the driving. This line of thought is what led me to think that % duration (in a race stint lap) of ICE operating near peak isn't going to change drastically, w.r.t previous hybrid era - instead of 'peaky' differences between deployment/recharge phases within a lap, the energy management would be more 'evenly spread' over a lap, just that the battery would claim a larger chunk of ICE output 'all the time'.

[michl420]

A engine that have 1/3 less power need to run 1/3 longer to have the same output. This mean they must go on a 66% full trottle lap to ~88%. They can do this only when the power demand is between ~200kw and ~600 kw (assuming that full trottle harvesting wasn`t done in the past).
For me it seems possible but maybe just theoretically.

[BorisTheBlade]

So, to get a better grasp of how the Energy Management might look like next year, I did some further digging.
To do this, I took the fastest race lap during Melbourne '25 (Lando Norris in Lap 43). The Telemetry-Data is from this amazing GitHub-Repo https://github.com/TracingInsights-Archive/2025 provided by https://tracinginsights.com.

The lap chart below is unusual in several aspects:
It does not start at the finish line, but rather at the start of the first braking sequence (the straight before is appended at the end). The x-axis is time, not distance. This makes calculationg (electrical) power and ES SoC much easier. Next to Throttle, Acceleration and Speed, it also shows the ES SoC in KWs.
In order to calculate this, the following assumptions were made:

• Under braking, recovery is simply 1:1 to the calculated braking power up to -350 KW.
• Under partial load, the average is -100 KW. This is roughly in line with rule 5.2.5 that regulates the fuel energy flow according to Power output of the PU under partial load. -100 KW might be a bit optimistic, but it won't make a big difference.
• Under full load, it is simply 350 KW for deployment. That is obviously the worst case scenario, as you are allowed to reduce ERS-K output to 200 KW at the start of each full throttle sequence and can reduce it further during that sequence (down to -250 KW).
• The weight used for the analysis was 768 kg.

Braking lasts 11,5 s (13,5% of the laptime) altogether and recovers 3.317 KWs.
Partial load lasts 17,5 s (20,7%) and recovers 1.916 KWs.
Full throttle lasted for 55,7 s (65,8%).

The energy recovered above would only last for 15 s at 350 KW. But even if a car had the full 9.000 KWs allowed per lap at its disposal, this would only last for 25,7 s at 350 KW.
So in a sustainable race mode, teams will likely use much lower numbers to spread the usage.
In order to recover the full 9.000 KWs allowed, it would need 15 s of super-clipping at -250 KW in addition to the recovered energy under braking and partial load.
But the big question is, if this is even the target for optimal pace, or more specifically:
Will the laptime loss of recovering -250 KW on a straight (and sending only 150 KW to the wheels) be outweighed by using that energy on other parts of a full throttle section? With around 85% efficiency of recovery & deployment combined, for each second of -250 KW super-clipping you get around 0,6 seconds of 350 KW Boost.

So from the 40,6 seconds of full-throttle left without super-clipping, they could boost anywhere between 25,7s at 350 KW and all the 40,6 s at 221 KW (on average).

In Qualifying, they should be able to use the full 9.000 KWs without the need for supper-clipping, as they can start their fast laps with a full ES.

If the maximum recovery energy should really be limited to around 4.500 KWs for tracks like Monza, they will be restricted to just 400 KW from the ICE for most of the full throttle sections - pace should tank there considerably.

Interestingly, ES capacity does not seem to be a limiting factor - at least in the Melbourne race.

What do you think about all of this? Is the approach flawed in any way? Would you like some other tracks / laps to be processed this way?

[wuzak]

The allowable recovery per lap is 8.5MJ/lap, but an extra 0.5MJ is allowed when 1s behind another car.

Deployment is unlimited, but will likely match the normal recovery.

I doubt Melbourne will have the full 8.5MJ/lap allowance. Likely be 5MJ/lap.

Braking seems a bit much. Might also include some lift-and-coast.

[BorisTheBlade]

The allowable recovery per lap is 8.5MJ/lap, but an extra 0.5MJ is allowed when 1s behind another car.

Yes, and under certain circumstances, even less is allowed according to 5.2.10. I just had to decide for one singular value for simplicity's sake. And it is not as if 500 KWs would make a hell of a diference.

I doubt Melbourne will have the full 8.5MJ/lap allowance. Likely be 5MJ/lap.

I would rather think that Melbourne is one of the circuits with a better ratio between recovery and full throttle.

Braking seems a bit much. Might also include some lift-and-coast.

Depends on how you define braking vs. lift-and-cost
The Telemetry Data contains a bit-field for when a driver actually hits the brake - I do not consider this to be lift-and-coast, but rather slight braking.
For what it's worth: Braking above -1g is 8,5s.

[wuzak]

Braking seems a bit much. Might also include some lift-and-coast.

Depends on how you define braking vs. lift-and-cost
The Telemetry Data contains a bit-field for when a driver actually hits the brake - I do not consider this to be lift-and-coast, but rather slight braking.
For what it's worth: Braking above -1g is 8,5s.

1g deceleration is roughly what the aerodynamics can achieve at high speed after the driver lifts off.

So, I would think that -1g is mostly lift-and-coast. There may be some areas where they have very slight braking too.

8.5s braking also aligns with what I have seen in some Brembo fact sheets.

I doubt Melbourne will have the full 8.5MJ/lap allowance. Likely be 5MJ/lap.

I would rather think that Melbourne is one of the circuits with a better ratio between recovery and full throttle.

Melbourne has one of the lowest braking time in the calendar.

It does have some opportunity for part load braking.

Top speed on that chart seems to be around 300km/h, so not much time spent at reduced MGUK output, though the speed will likely increase a bit.

[wuzak]

• Under braking, recovery is simply 1:1 to the calculated braking power up to -350 KW.
• Under partial load, the average is -100 KW. This is roughly in line with rule 5.2.5 that regulates the fuel energy flow according to Power output of the PU under partial load. -100 KW might be a bit optimistic, but it won't make a big difference.
• Under full load, it is simply 350 KW for deployment. That is obviously the worst case scenario, as you are allowed to reduce ERS-K output to 200 KW at the start of each full throttle sequence and can reduce it further during that sequence (down to -250 KW).
• The weight used for the analysis was 768 kg.

Braking lasts 11,5 s (13,5% of the laptime) altogether and recovers 3.317 KWs.
Partial load lasts 17,5 s (20,7%) and recovers 1.916 KWs.
Full throttle lasted for 55,7 s (65,8%).

1.9MJ for 17.5s equates to 109.5kW.

Presumably you are only counting the parts where SOC rises?

With the lower ICE output, 2026 part throttle may require more time where the MGUK is deploying, or at 0kW, rather than recovering energy.

[diffuser]

It's silicon-carbon, not carbide. They've been talking about silicon this and that in the anode for many years. It never really came to pass. Best I remember is Tesla adding some silicon.
In any way a proper solid-state battery is better. It has neither silicon nor carbon, just lithium metal.

Are SSB's anywhere near advanced enough to be considered for use in F1? I believe Mercedes, who are the only relevant manu here that has also conducted real life prototyping and testing, say they are at least still 4 years away, from producing EVs using SSB'S. I believe it was developed in-house, in conjunction with HPP.

I think everyone and their mother are developing solid-state batteries. And I think many have tested it as well.

Ultimately they don't need to be mass market viable to be included in F1. Loosening the battery pack limits is easy, if they want better energy density.

On the original Question "Will we see". I'd be surprised if any of the PU manufactures tell us what battery Chemistry they're using.

[diffuser]

The allowable recovery per lap is 8.5MJ/lap, but an extra 0.5MJ is allowed when 1s behind another car.

They should limit the number of laps the extra 0.5MJ is allowed, when following any 1 specific car, to something like 5 laps. This would prevent the DRS trains that we've had in the past, use it to pass or loose it. If you fall out of the 1 second gap for minimum of 1 lap then return to the sub 1 second behind the same car, the trailing car would get an additional 5 laps.

[wuzak]

The allowable recovery per lap is 8.5MJ/lap, but an extra 0.5MJ is allowed when 1s behind another car.

They should limit the number of laps the extra 0.5MJ is allowed, when following any 1 specific car, to something like 5 laps. This would prevent the DRS trains that we've had in the past, use it to pass or loose it. If you fall out of the 1 second gap for minimum of 1 lap then return to the sub 1 second behind the same car, the trailing car would get an additional 5 laps.

As I understand it, that extra energy can be deployed anywhwere.