2025/2026 Hybrid Powerunit speculation

[mzso]

Well, that would depend on what the math says, but that’s my point, I don’t understand how that would mathematically be possible, surely burning petrol to generate electricity is wholly inefficient. That is What I want explaining

Your example would require 100% Efficiency, and that is not possible

Not as inefficient as you seem to expect. And in some circumstances it could be more efficient*, when you use the best time to charge and deploy. Due to the ICE having a narrow efficiency range, near the peak, and the MG having a broader one, as well as being a lot more efficient overall.

*equivalent ICE power vs. hybrid with fuel burn regen.

[Tommy Cookers]

... Due to the ICE having a narrow efficiency range, near the peak
....the MG having a broader one, as well as being a lot more efficient overall.

the ICE is only allowed to run in a way that helps the electrical side (ie the rpm based fuel allocation)
the MG (not being the prime mover) is a parasite whose only justification is its 'efficiency'

we are far short of the heralded 50/50 system but are starting to see the underlying limitations

[karana]

I’m a little confused… Would there ever be a time a driver would want to turn his ICE into a generator to store up electrical power?

Surely this would be wholly inefficient, not only would you have to carry more fuel which would obviously increase your lap time, without the MGU-H, you be wasting the vast majority of its energy as you did so

Perhaps this has already been explained further back, it I’d appreciate it if someone could explain, in simple language

Comparing the additional fuel weight with the additional power is a bit like comparing apples to oranges. It really depends on the car how big the effects are. Here are some interesting numbers showing the lap time impact from different parts of the car from Willem Toet for a "recent F1 car" https://www.linkedin.com/pulse/devious- ... Jy5w%3D%3D (the article is from 2015):

● Grip – from tyres, suspension, etc. 10% grip = 3 seconds lap time
● Vehicle mass 10% mass = 1⋅7seconds lap time
● Engine (powertrain) 10% Power = 1⋅4 seconds lap time
● Aerodynamics 10% downforce = 0.9 seconds lap time

Let's assume (just as an example) our car (700kg) uses 70kg of fuel with an efficiency of 50% and compare it to a car that uses additionaly 30kg of fuel with 33% efficiency. This would lead to an increase in power of 10/35=~28% so a lap time reduction of ~4 seconds. But the mass increase is just 30/770 =~3.9%, which leads to a lap time increase of ~0.7 seconds. But only at the start! The extra weight will burn off through the course of the race.

The efficiency possibly won't even be that bad as pointed out earlier.

Interestingly, running the MGU-K against the ICE at the end of straights should actually save fuel. This allows them to use more energy early on the straight. Assuming that this makes the cars overall faster (otherwise they wouldn't do it), the cars spend less time on the straights and therefore burn less fuel.
To a degree this is also true for partial load recovery, but the additional fuel burnt in the corners probably will outweigh the fuel saved on the straights (but it will still be worth it).

Are you allowed to burn petrol to charge the battery at all?

Yes.

I wonder what that will sound like?

As far as I know this is already done with the current PUs (they can't recover the allowed 2MJ with braking alone). I wouldn't expect that you can really hear this, the rpm behaves exactly like it would when not doing it. The engine might be louder, but could you tell if it is louder than 'necessary'? (But this is just a wild guess.)

[gruntguru]

Part-load harvesting will be relatively efficient because the extra load applied by the MGU-K will move the ICE into a higher efficiency region. Just making up some numbers - lets assume
- Full-load ICE efficiency is 50% giving 417 kW. This requires the max 833 KW fuel flow (3000 MJ/hr)
- Half load efficiency is 45% for 208 kW. This requires a fuel flow of 462 KW

So if the car is cornering with 208 kW power requirement and the MGU-K is generating another 208 KW, the fuel cost is 833 - 462 = 371 kW. This means the energy generated is at a marginal efficiency of 208/371 = 56% (times the electrical efficiency of the MGU - say 97% for a total of 54%)

[BassVirolla]

Part-load harvesting will be relatively efficient because the extra load applied by the MGU-K will move the ICE into a higher efficiency region. Just making up some numbers - lets assume
- Full-load ICE efficiency is 50% giving 417 kW. This requires the max 833 KW fuel flow (3000 MJ/hr)
- Half load efficiency is 45% for 208 kW. This requires a fuel flow of 462 KW

So if the car is cornering with 208 kW power requirement and the MGU-K is generating another 208 KW, the fuel cost is 833 - 462 = 371 kW. This means the energy generated is at a marginal efficiency of 208/371 = 56% (times the electrical efficiency of the MGU - say 97% for a total of 54%)

Even when I had a hard time following your train of thought (not the same "part load" for engine and for PU was giving me a headache ), it just makes total sense.

[venkyhere]

... surely burning petrol to generate electricity is wholly inefficient. That is What I want explaining

F1 is a series of rolling-start drag races
if drag races were energy-limited eg ....
[b/]it would be better to do the first 500 ft at 8000 hp and the second 500 ft at 4000 hp (rather than 1000 ft at 6000 hp)[/b]

burning for electricity, storing, then re-using it is (in isolation) about 90% as efficient as simultaneous 'burning & turning'
but the ES allows F1 to be run as a series of the above hypothetical drag race so there is a time benefit overall
time benefit in cars built to the mandatory minimum weight of course

Just a minor technical error in your imaginary example - since you mention 'drag-race' from rolling start, the second 500 ft would need the 8000hp rather than the first, since higher speed = higher drag

[Tommy Cookers]

Part-load harvesting will be relatively efficient because the extra load applied by the MGU-K will move the ICE into a higher efficiency region. Just making up some numbers - lets assume
- Full-load ICE efficiency is 50% giving 417 kW. This requires the max 833 KW fuel flow (3000 MJ/hr)
- Half load efficiency is 45% for 208 kW. This requires a fuel flow of 462 KW

So if the car is cornering with 208 kW power requirement and the MGU-K is generating another 208 KW, the fuel cost is 833 - 462 = 371 kW. This means the energy generated is at a marginal efficiency of 208/371 = 56% (times the electrical efficiency of the MGU - say 97% for a total of 54%)

Even when I had a hard time following your train of thought (not the same "part load" for engine and for PU was giving me a headache :lol: ), it just makes total sense.

97% does not apply
unless the full route of MGU-K generation of DC and storage and unstorage then MGU-K motor action adds up to 97%

also ...
at half load a non-hybrid car would eg change to 12th gear and 5250 rpm - but our F1 ICE isn't allowed to do this

[saviour stivala]

Ultimately, the most efficient operation is a balance managed over a 'full-lap' to ensure they have enough electrical energy when needed for acceleration and overtaking, without depleting the battery. For the above reason, the 2026 formula 1 power unit is designed to operate at maximum efficiency during braking zones and specific energy managements periods on each lap, rather than at a single period. (optimizing energy recovery and deployment throughout a lap).

[gruntguru]

I was wondering whether to
a) not worry about MGU-K efficiency at all. The storage/retrieval route applies regardless of harvesting method.
b) apply a one-way efficiency only - say 97%
c) apply a two-way efficiency - say 94.09%

The RPM used for 50% power is not really relevant. It's still not going to achieve the same efficiency it does at full power. (The ICE is optimised for full power operation - it spends a large portion of its time there and an even larger portion of total race energy is consumed there.)

The max efficiency point for 50% power for road cars is (I think) at rpm greater than 50% of max rpm. (Peak efficiency is usually at about 50% load, 50% rpm which only gets it to 25% power. Where that point is on a 2026 F1 engine is anybody's guess.

[michl420]

About efficient a point to think about. Cooling. As an example, under full trottle recovery the ICE run full power, MGUK about 200 kw, ES also some work to do. That results in a net output of just 200 kw with almost maximum cooling demand.

[mzso]

... Due to the ICE having a narrow efficiency range, near the peak
....the MG having a broader one, as well as being a lot more efficient overall.

the ICE is only allowed to run in a way that helps the electrical side (ie the rpm based fuel allocation)
the MG (not being the prime mover) is a parasite whose only justification is its 'efficiency'

we are far short of the heralded 50/50 system but are starting to see the underlying limitations

I wouldn't call something a parasite that actually saves a lot of fuel compared to ICE only power. Also its superior performance characteristics also help with driving, such as filling in the turbo-lag.

[mzso]

Yes.

I wonder what that will sound like?

As far as I know this is already done with the current PUs (they can't recover the allowed 2MJ with braking alone). I wouldn't expect that you can really hear this, the rpm behaves exactly like it would when not doing it. The engine might be louder, but could you tell if it is louder than 'necessary'? (But this is just a wild guess.)

Why would they? There's abundant and un-capped energy recover from the MUG-H. They can recover the allowance from that, and that's otherwise wasted energy. So there's no sense in using K for anything other than breaking.

[karana]

I wonder what that will sound like?

As far as I know this is already done with the current PUs (they can't recover the allowed 2MJ with braking alone). I wouldn't expect that you can really hear this, the rpm behaves exactly like it would when not doing it. The engine might be louder, but could you tell if it is louder than 'necessary'? (But this is just a wild guess.)

Why would they? There's abundant and un-capped energy recover from the MUG-H. They can recover the allowance from that, and that's otherwise wasted energy. So there's no sense in using K for anything other than breaking.

While there is no limit on MGU-H recovery, there is only so much energy they can practically recover (I'm not sure about how much it is, but I think it was something around 2-3MJ per lap, depending on the track). The energy you have available per lap is basically (energy recovered by MGU-K + energy recovered by MGU-H). So you want to maximise both. And running the MGU-K against the ICE doesn't harm the MGU-H recovery in any way. In fact it even helps as it helps keeping the turbo spun.

From Honda's website: https://global.honda/en/tech/motorsport ... train_ESS/

Under partial throttle when starting to exit a corner, a different control technology called partial recovery is employed. It is used to generate electricity by diverting excess output to the MGU-K after the engine generates enough output to meet the driver’s needs.

Honda has even introduced a technique to circumvent the 2MJ recovery limit:

Extra Harvest/Extra Deploy is a control technology fully implemented from the middle of the 2018 season. The maximum amount of energy that can be sent directly from the MGU-K to the ES is 2 MJ per lap, so rather than sending any recovered energy over 2 MJ to the ES, that energy is sent to the MGU-H, which has no restrictions on energy exchanged with the ES. After being used for only a moment to assist the MGU-H, the inertial energy from the rotor is then immediately recovered. Through continuous repetition of this assist-to-recovery cycle at a frequency of 20 Hz or less, the recovered energy is sent to the ES. This is known as Extra Harvest.

[haza]

First sound of the 2026 engines courtesy of Honda

Can hear the power drop towards the end reminiscent of d rating at tracks like monza

https://x.com/hondaracingf1/status/1999 ... 38690?s=46

[BassVirolla]

First sound of the 2026 engines courtesy of Honda

Can hear the power drop towards the end reminiscent of d rating at tracks like monza

https://x.com/hondaracingf1/status/1999 ... 38690?s=46

To me, sounds like blipping throttle inputs between downshifts, probably burning as much fuel possible (allowed) in this time to increase recovery.