2026 Hybrid Powerunits

[michl420]

ERS-K recovery from ICE power under partial load
The chart below . . . . . . .

Very informativ post, thank you.
It shows also how the gear selection influence harvesting.

The full trottle deployment will imho selected from the software based from a preset profile and the ES state.

[saviour stivala]

In my opinion the 2026 F1 ERS will not recover energy from the ICE under partial load, as the MGU-H (which recovered energy from the turbocharger) is being removed, and the primary method for energy recovery will be through braking. Instead of using the engine exhaust, drivers may have to use the engine to generate electricity by revving it in slow corners to charge the battery.

[Tommy Cookers]

... the energy harvesting system, which converts kinetic energy into electrical power to be stored in the battery for later use.

it converts kinetic energy and pressure energy into (stored) electrical energy

[wuzak]

In my opinion the 2026 F1 ERS will not recover energy from the ICE under partial load, as the MGU-H (which recovered energy from the turbocharger) is being removed, and the primary method for energy recovery will be through braking.

The MGUH recovered energy from the exhaust, at full throttle and part throttle. Basically, any time the turbine produced more power than the compressor required to produce the desired boost an air flow.

The rules specify fuel flow restrictions for part throttle beyond the standard rpm restrictions, which suggest an attempt to restrict part throttle energy recovery.

The rules also specify conditions in which up to 250kW can be recovered under full throttle.

Whether braking is the primary source of recovery will depend on the track layout and what recovery per lap is allowed (likely to be reduced for tracks like Silverstone).

Instead of using the engine exhaust, drivers may have to use the engine to generate electricity by revving it in slow corners to charge the battery.

Isn't this the parth throttle recovery that you claim won't happen?

The only way that "revving it [engine] in slow corners" is to decouple the engine from the drive. I.e. disconnect the clutch.

This is undesirable, as it has to be manually controlled by the driver, and cuts drive to the rear wheels.

The cars will, likely, be using lower gears to negotiate many coners, to maximise rpm and recovery potential at part throttle.

Some corners are already taken in the lowest gear, so (much) higher rpm will not be feasible.

[karana]

There are lots of ambiguous or poorly worded sections in the rules. Here is another example:

"the driver maximum power demand must not be reduced at any greater than the rates defined below"

Pretty sure the intent of this wording is "during driver maximum demand, the PU output must not be reduced at any greater than the rates defined below"

I think the wording is fine in this case. The maximum power demand is the power the PU needs to output when the driver presses the accelerator pedal fully. This is a well-defined value regardless of what the driver does. If for example the driver has the accelerator pedal 50% pressed, this will lead to different PU outputs depending on what the maximal power demand at that moment is (at least in my understanding).

...

Aside from having it gotten totally wrong, questions that I could not answer myself until now:

• Is power demand only derived from the throttle pedal or also from the braking pedal?
• If it is only the throttle pedal, is -50 kW the minimum value which stands for minimum Engine Braking? Or could it be possible to have the throttle pedal mapped from -350 kW to750 kW, making it a bit like "One Pedal Driving" in a BEV? Coincidently, there was a rule in an earlier revision of the Technical Regulations prohibiting certain points (like for example the 0 kW point on such a wide band) to be marked (felt by the driver). This rule seems to have vanished.
• If both, how does it get aggregated? Adding both, ignoring the throttle pedal as soon as one hits the brake pedal, something else?

...

There is this rule:

C5.12.3 At any given engine speed, the minimum torque in the driver torque demand map must be a
negative value greater than the minimum curve defined as Torque (Nm) = −0.0027 * engine speed
(rpm) − 30

This corresponds to a minimum power higher than -100kW in the rpm ranges we are likely to see.
I think this can only meaningfully apply to the case when the driver is off throttle and NOT braking, considering they will surely be allowed to run the PU lower than that when braking. (Btw, I think this rule also discourages lift&coast for harvesting reasons, you're better off to just run the ICE gainst the MGUK.)

The rule you mentioned was not removed, just moved to a different section:

C8.6.4 Accelerator Pedal
The only means by which the driver may control acceleration torque to the driven wheels is via a
single foot (accelerator) pedal mounted inside the Survival Cell.
The design or installation of any component, surface finish, or feature, the purpose or effect of
which allow specific points between the 0% and 100% calibrated references of the accelerator
pedal travel range to be identified by the driver or to assist the driver to hold a position are not
permitted.

So I believe that what the rules call the driver power demand only depends on the accelerator pedal, and when braking the part of the requested braking torque not sent to the friction brakes just gets added to the driver torque demand requested through the accelerator pedal. I might be wrong of course.

[gruntguru]

There are lots of ambiguous or poorly worded sections in the rules. Here is another example:

"the driver maximum power demand must not be reduced at any greater than the rates defined below"

Pretty sure the intent of this wording is "during driver maximum demand, the PU output must not be reduced at any greater than the rates defined below"

I think the wording is fine in this case. The maximum power demand is the power the PU needs to output when the driver presses the accelerator pedal fully. This is a well-defined value regardless of what the driver does. If for example the driver has the accelerator pedal 50% pressed, this will lead to different PU outputs depending on what the maximal power demand at that moment is (at least in my understanding).

Disagree.
- "Driver Demand" is the accelerator pedal position chosen by the driver - not the PU output.
- "PU Output" is the power delivered to the wheels.

The relationship between these two is not fixed but can be overridden by the control system - for the purposes of harvesting energy.

If you take the rule wording literally, either:
- the driver is not permitted to reduce the accelerator position at more than the specified rate or
- the control system is not permitted to respond to rapid reductions in accelerator position at more than the specified rate.

Either of these is clearly absurd. The "rate" the rule intends to regulate is the rate at which the ECU can reduce PU output and override the Driver Demand in order to harvest additional energy.

[saviour stivala]

In my opinion the 2026 F1 ERS will not recover energy from the ICE under partial load, as the MGU-H (which recovered energy from the turbocharger) is being removed, and the primary method for energy recovery will be through braking.

The MGUH recovered energy from the exhaust, at full throttle and part throttle. Basically, any time the turbine produced more power than the compressor required to produce the desired boost an air flow.

The rules specify fuel flow restrictions for part throttle beyond the standard rpm restrictions, which suggest an attempt to restrict part throttle energy recovery.

The rules also specify conditions in which up to 250kW can be recovered under full throttle.

Whether braking is the primary source of recovery will depend on the track layout and what recovery per lap is allowed (likely to be reduced for tracks like Silverstone).

Instead of using the engine exhaust, drivers may have to use the engine to generate electricity by revving it in slow corners to charge the battery.

Isn't this the parth throttle recovery that you claim won't happen?

The only way that "revving it [engine] in slow corners" is to decouple the engine from the drive. I.e. disconnect the clutch.

This is undesirable, as it has to be manually controlled by the driver, and cuts drive to the rear wheels.

The cars will, likely, be using lower gears to negotiate many coners, to maximise rpm and recovery potential at part throttle.

Some corners are already taken in the lowest gear, so (much) higher rpm will not be feasible.

''The rules specify conditions in which up to 250 kw can be recovered under full throttle''. No, The rule you refer to is about 'deploying' up to 350 kw of energy to overtake, not to recovering it. The rule is called manual overdrive. You have to understand that the 2026 ERS will not be allowed to recover energy when driver is at full throttle. This boost is applied under full throttle by following driver when they are close to the car ahead to help gain on their opponent. To help harvesting energy in slow corners by revving the engine, the revving is achieved by using a lower gear.

[BorisTheBlade]

Or possibly it's door 3 (as I understand it):
This is the rate, the ECU is allowed to reduce the PU output while the driver stays fully on the throttle pedal.

[venkyhere]

The chart below is a refinement of the one in the post linked above, showing the potential of what could be harvested under partial load.
https://i.imgur.com/HSKChtW.png

• The dark green area is the theoretically harvestable power from the ICE - which is derived from C5.2.5 assuming a (IMHO still challenging) TE of 48 %.
• The blue dots are the needed min. RPM (x1000) according to C5.2.4 - again assuming 48 % TE. This of course means, that you need the right gear for the desired speed in a certain corner.
• The 50 kW base ICE output at/below -50 kW power demand might be intended to just keep the ICE from stalling - so there will not be much to harvest.
• So between 0 kW and 300 kW power demand something between 100 kW and 170 kW could be harvestable - not nothing, but quite a bit less and on a narrower power demand band than I had initially thought.

Kindly help a layman understand this informative graph (thank you Boris), by pointing mistakes in the below summary :

-ve number on torque demand = brake pedal pressed
+ve number on torque demand = throttle pedal pressed
0 on the torque demand = both pedals not pressed

To the left of -350KW demand = brake pedal pressed beyond a critical pressure, asking for maximum retardation, dark green shows highest electrical loading by the stator windings of the MGU-K, extracting maximum possible energy from the crank, towards recharging the ES. The fuel supply to ICE is cut off and the ICE is working like an air-pump/vacuum pump and is in 'max-ICE-braking' mode. The 'red' is of course, the physical deployment of carbon brakes on the rear disc (front brakes are physically on all the time towards the left of 0 on the x-axis anyway).

To the right of 400W demand = ICE has maxed out, it's only MGU-K deployment that is scaling up in-situ to the throttle press.
To the right of 0 demand, until 400W demand = this is the part throttle recharging of the ES via MGU-K that is being talked about over the last several pages of this thread, where energy recovery happens WITHOUT BRAKE PEDAL BEING PRESSED.

Next is the confusing bit for me :
demand 0 to demand -50 (brake pedal pressed)
demand -50 to demand -350 (brake pedal pressed)
What's going on here, when both dark green and light green are present simultaneously ? Which colour signifies 'recharging of ES' and what does the other colour signify in that case ? what is the state of the ICE when both colours are active ? is fuel supplied to it between 0 and -350 demand points ?

[karana]

There are lots of ambiguous or poorly worded sections in the rules. Here is another example:



Pretty sure the intent of this wording is "during driver maximum demand, the PU output must not be reduced at any greater than the rates defined below"

I think the wording is fine in this case. The maximum power demand is the power the PU needs to output when the driver presses the accelerator pedal fully. This is a well-defined value regardless of what the driver does. If for example the driver has the accelerator pedal 50% pressed, this will lead to different PU outputs depending on what the maximal power demand at that moment is (at least in my understanding).

Disagree.
- "Driver Demand" is the accelerator pedal position chosen by the driver - not the PU output.
- "PU Output" is the power delivered to the wheels.

The relationship between these two is not fixed but can be overridden by the control system - for the purposes of harvesting energy.

If you take the rule wording literally, either:
- the driver is not permitted to reduce the accelerator position at more than the specified rate or
- the control system is not permitted to respond to rapid reductions in accelerator position at more than the specified rate.

Either of these is clearly absurd. The "rate" the rule intends to regulate is the rate at which the ECU can reduce PU output and override the Driver Demand in order to harvest additional energy.

So, in your opinion, is there a fixed relationship between pedal position and power demand in the sense that a fixed pedal position corresponds to a fixed power demand, only additionaly depending on engine speed?
If the PU output is reduced to 400kW and the driver lifts, what happens? If 400kW corresponds to pedal position e.g. x=60%, does the PU output stay at 400kW for x=100% to 60%? This seems undesirable from a safety viewpoint.


I would have guessed that the driver demand does not only depend on the pedal position (and engine speed), but also on an additional parameter, which the rules call 'maximum driver demand' (but it's basically just a name for that parameter). So here is what I think is happening: Coming out of the corner, this parameter will be e.g. 750kW. The (actual) driver demand is then (simplified, just as an example) x*750kw where x is the pedal position in percent. This gives the driver complete control over the entire PU power range. As long as the driver is at full throttle (x=100%) this parameter can only be reduced at the rate in C5.12.6. So, if this parameter is 400kW, the driver demand is x*400kW. If the driver now lifts to x=90%, the driver demand will be 360kW. Basically, this gives the driver full control over the PU power that is available at any given moment.

Anyway, this rule is slightly redundant after they introduced C5.12.7, which requires the downward rate for the MGUK in any circumstance (with obvious exceptions).

The relationship between these two is not fixed but can be overridden by the control system - for the purposes of harvesting energy.

I really don't think this is true except for when the driver actually uses the brakes. According to C.5.13.1 the PU output must match the power demand within 50ms.

[mzso]

The cars will, likely, be using lower gears to negotiate many coners, to maximise rpm and recovery potential at part throttle.

Some corners are already taken in the lowest gear, so (much) higher rpm will not be feasible.

Unless there's some standardization I didn't hear about, I expect them to redesign their gearboxes and shifts so that the engine would be at higher revs than nowadays, wherever its desirable. So the first gear might result in higher rpm for the Monaco hairpin.

[BorisTheBlade]

Kindly help a layman understand this informative graph (thank you Boris), by pointing mistakes in the below summary :

-ve number on torque demand = brake pedal pressed
+ve number on torque demand = throttle pedal pressed
0 on the torque demand = both pedals not pressed

To the left of -350KW demand = brake pedal pressed beyond a critical pressure, asking for maximum retardation, dark green shows highest electrical loading by the stator windings of the MGU-K, extracting maximum possible energy from the crank, towards recharging the ES. The fuel supply to ICE is cut off and the ICE is working like an air-pump/vacuum pump and is in 'max-ICE-braking' mode. The 'red' is of course, the physical deployment of carbon brakes on the rear disc (front brakes are physically on all the time towards the left of 0 on the x-axis anyway).

To the right of 400W demand = ICE has maxed out, it's only MGU-K deployment that is scaling up in-situ to the throttle press.
To the right of 0 demand, until 400W demand = this is the part throttle recharging of the ES via MGU-K that is being talked about over the last several pages of this thread, where energy recovery happens WITHOUT BRAKE PEDAL BEING PRESSED.

First of all, my understanding/interpretation of that topic is still very volatile and on a day-to-day-base, but at the time of writing that was what I meant.
But what is really important is, that the max. 750 kW output on the right side is only one option how the car can act at full throttle. It is allowed to act anywhere in between 750 kW (400 kW ICE + 350 ERS-K) and 150 kW (400 kW ICE - 250 ERS-K) as far as I understand - that is my chart with the "corridor".

Next is the confusing bit for me :
demand 0 to demand -50 (brake pedal pressed)
demand -50 to demand -350 (brake pedal pressed)
What's going on here, when both dark green and light green are present simultaneously ? Which colour signifies 'recharging of ES' and what does the other colour signify in that case ? what is the state of the ICE when both colours are active ? is fuel supplied to it between 0 and -350 demand points ?

Dark green is the amount of power that the ERS-K might recover from the ICE, which still generates some power (the 50 kW, which might actually rather be anti-stall).
Light green on the left side is what the ERS-K is recovering while acting as a brake - sorry for not picking a different colour than for the ERS-K while acting as a motor.

But when karana quoted C5.12.3 again, that might have changed a bit.
This is what the rule looks like as a chart:


/edit: This looks like a rather complex way of stating: You shall not go below -30 Nm (and dare you use your pretty ERS-K in this situation)!

So IMHO you can recover up to 350 kW under braking with no laptime impact, followed by up to 250 kW under full throttle with severe laptime impact, followed by 100-150 kW under partial throttle with no laptime impact - but nothing when just lifting the throttle pedal.

And it also means, that the min. value of the driver torque demand - the x-axis on my chart - can change depending on such factors as RPM, maybe speed or whatever. And it does not include braking, only the range from no throttle pedal usage to full throttle. So to comply with the rules, I would have to split the chart

[BorisTheBlade]

I would have guessed that the driver demand does not only depend on the pedal position (and engine speed), but also on an additional parameter, which the rules call 'maximum driver demand' (but it's basically just a name for that parameter). So here is what I think is happening: Coming out of the corner, this parameter will be e.g. 750kW. The (actual) driver demand is then (simplified, just as an example) x*750kw where x is the pedal position in percent. This gives the driver complete control over the entire PU power range. As long as the driver is at full throttle (x=100%) this parameter can only be reduced at the rate in C5.12.6. So, if this parameter is 400kW, the driver demand is x*400kW. If the driver now lifts to x=90%, the driver demand will be 360kW. Basically, this gives the driver full control over the PU power that is available at any given moment.

This absolutely makes sense to me - yay, another parameter to consider

[gruntguru]

OK, I am starting to re-think the definition of "driver maximum power demand". The term only appears two times in the tech regs and there is no definition offered. I agree with @Karana that this refers to a Control System parameter - specifically a power/torque level that is the maximum available at any given engine speed. So - if the driver floors the accelerator, "driver maximum power demand" is the power (mapped to rpm) that will be available at that instant.

• In all cases it will at that instant be equal to the max ICE power (mapped to rpm) available in the current mode setting, plus at least 200 KW from the MGU-K.

• The control system may then reduce the value during this full throttle period (by reducing the MGU-K output) at the rate limited by the regs. This reduction is entirely to accomplish 1. Electrical energy saving and 2. (at MGU-K output below zero) harvesting.

• The control system may not increase the value during this full throttle period.

[wuzak]

OK, I am starting to re-think the definition of "driver maximum power demand". The term only appears two times in the tech regs and there is no definition offered. I agree with @Karana that this refers to a Control System parameter - specifically a power/torque level that is the maximum available at any given engine speed. So - if the driver floors the accelerator, "driver maximum power demand" is the power (mapped to rpm) that will be available at that instant.

• In all cases it will at that instant be equal to the max ICE power (mapped to rpm) available in the current mode setting, plus at least 200 KW from the MGU-K.

• The control system may then reduce the value during this full throttle period (by reducing the MGU-K output) at the rate limited by the regs. This reduction is entirely to accomplish 1. Electrical energy saving and 2. (at MGU-K output below zero) harvesting.

Driver maximum power demand to me means that the driver has the accelerator pedal at 100%. Demanding all that the PU can give them.

The reduction of MGUK output is defined the way it is to prevent it suddenly being shut off, and the driver going from 750kW to 400kW in an instant.

The reduction is also to prevent the cars from achieving speeds too high for the tracks.

• The control system may not increase the value during this full throttle period.

Except where override is selected by the driver.