KERS anatomy

With KERS being revived and expected to race again next year, let’s just recap what hardware’s involved and how its packaging affects the car design for 2011.

KERS (Kinetic Energy Recovery System) is a hybrid drive system that the FIA allowed to be raced as part of the 2009 major rules rewrite. It allows energy to be harvested under braking and stored, then that energy can be released to provide a power boost for around 6s per lap.
In 2009 most engine manufacturers developed their own KERS system, while Williams were the sole team developing a system independently. Albeit not every team raced with KERS and some teams dropped KERS at various races. For 2010 FOTA agreed to drop KERS, albeit it was still legal with in the rules, as a cost cutting measure it was best not to run or develop KERS any further. Even within 2009 season KERS was not a huge success, the system had a FIA cap on the amount of energy that could be re-used, only 400kJ could be stored, which when used for 6.7s per lap, the car gained some 80hp. Thus although a 0.3s boost to laptimes, the system was ultimately limited in its potential to improve laptimes. Thus no team could create a competitive advantage from a more powerful system. Then the weight of the system created issues, At a time when the wider front slick tyres demanded an extreme weight distribution of up to 49% weight on the front axle, the 25+Kg of a KERS system mounted behind the CofG handicapped teams being able to push weight forwards. Most teams dropping or not racing their system cited weight as the main reason for its loss.

What is a KERS system?
In essence a KERS systems is simple, you need a component for generating the power, one for storing it and another to control it all. Thus KERS systems have three main components: The MGU, the PCU and the batteries. They are simply laid out as in the diagram below:

In detail

MGU (Motor Generator unit)

Marelli MGU as used by Ferrari and Renault

Mounted to the front of the engine, this is driven off a gear at the front of the crankshaft. Working in two modes, the MGU both creates the power for the batteries when the car is braking, then return the power from the batteries to add power directly to the engine, when the KERS button is deployed. Running high RPM and generating a significant Dc current the unit run very hot, so teams typically oil or water cool the MGU.


McLaren Mercedes Battery Pack complete with water cooling system

During the 2009 season only electrical batteries were used, although at least two flywheel systems were in development, but unraced. We will focus on the arrays of lithium-ion batteries that were raced. Made up of around 40 individual cells, these batteries would last two races before being recycled. In McLaren’s case these were mounted to the floor in the sidepods beneath the radiators. Other teams mounted them in a false bottom to the fuel tank area for safety in the event of a crash. Being charged and discharged repeatedly during a lap, the batteries would run very hot and needed cooling, this mainly took the form of oil or water cooling, and again McLarens example had them pack water cooled with its own pump and radiator.

PCU (Power Control Unit)

McLaren Mercedes PCU

Typically mounted in the sidepod this black box of electronics served two purposes, firstly to invert & control the switching of current from the batteries to the MGU and secondly to monitor the status of the individual cells with the battery. Managing the battery is critical as the efficiency of a pack of Li-ion cells will drop if one cell starts to fail. A failing cell can overheat rapidly and cause safety issues. As with all KERS components the PCU needs cooling

Marelli prototype PCU


Aside from these main components the KERS system also integrates with the FIA SECU in order to control and monitor the PCU. KERS has to be driver activated; this is achieved from a steering wheel button. Although the drive has to initiate the KERS boost, the teams set the system up such that the driver knows to engage the system out of specific corners, the system then delivers the predetermined amount of boost specific to the demands of that section of track. In practice the KERS systems is being charged and discharged to this preset map of activations. Which smoothes the balance between charging and discharging, so the system does not overcharge above the regulatory limit. Again the SECU ensures no more than the capped amount of energy is delivered each lap.

KERS in 2011
With KERS return to F1 next year, the designers are faced with the same operating requirements in terms of energy storage and discharge. But the packaging requirements of the systems have changed in the two years since its introduction. Firstly the penalty of weight distribution has eased. With narrower tyres and the move to a fixed weight distribution for the Pirelli tyres, means that 25kg KERS system no longer tip the weight balance the wrong way. Plus there will be a higher minimum weight limit for next year.
But challenging the designers will be the amount of space to package the hardware. With the ban on refuelling, teams have enlarged the fuel tank into the sidepods to create sufficient capacity, already the sidepods are full of longer narrower radiators and the gearbox oil coolers have been moved to above the gearbox to save space in the sidepods. Then the aerodynamically undercut shape of the sidepods robs yet more volume.
Given the success of McLarens sidepod mounted solution in 2009 and the safety concerns that dogged the systems introduction, means that teams will probably opt for sidepod mounting of the Batteries and PCU. Especially as to expand the fuel tank area to mount the batteries as in 2009 will create a huge bulky rear to the monocoque. There will no doubt be an aerodynamic penalty to the slightly bulkier sidepods to house the hardware and additional cooling. This needs to be less than 0.3s laptime penalty in order to offset the gain from the power boost. Even with the gains and losses in laptime with a KERS system, teams may opt to run the system simply to use it for an overtaking aid in the race. Off the line and onto long straight the boost might be enough to overtake a rival.

Hydraulic KERS

A filament wound carbon fibre Hydraulic Accumulator

A further alternative to the generation and storage of energy is to use hydraulics. This system has some limitations, but with the capped energy storage mandated within the rules the system could see a short term application. Separate to the cars other hydraulic systems, a hydraulic KERS would use a pump in place of the MGU and an accumulator in place of the batteries. Simple valving would route the fluid into the accumulator or to the pump to either generate or reapply the stored power. Hydraulic accumulators are already used in heavy industry to provide back up in the event of failure to conventional pumped systems.
Using filament wound carbon fibre casing, an accumulator of sufficient capacity could be made light enough to fit into the car (see They might be capped in terms of practical storage with in the confines of an F1 sized system, but McLaren had prepared just such an energy recovery system back on the late 90s, but it was banned before it could race. With the relatively low FIA cap on energy storage, just such a system could be easily packaged, the hydraulic MGU would be sited in the conventional front-of-engine position and the accumulator, given proper crash protection fitted to the sidepod\fuel tank area. Saving space would be minimal control system (equivalent to the PCU) as the valving to control the system could be controlled by the cars main electro hydraulic system. McLaren have recently been quoted as saying the 2011 KERS would be more hydraulic and less electronic. Giving rise to speculation that a hydraulic storage system could be used.


A Flybrid Flywheel system, similar to that intended for the 09 Honda F1 car

As Li-ion batteries are still an expensive emerging technology, plus they have associated risks, recycling and transport problems. The attraction of flywheel KERS is obvious, however no team have raced such a system in F1. Flywheels can effectively replace the Li-ion batteries with in a typical KERS system, the flywheel being mated to a second MGU to convert the power generated by the primary MGU on the engine into the kinetic to be stored in the flywheel. Williams are believed to have just such a system. However the simper flywheel solution is connect the flywheel system via a clutched and geared mechanism. HondaF1 had developed this solution for their 2010 car. This solution was dropped as Honda pulled out and the renamed Brawn team need to focus development resources into the new car and its conversion to Mercedes power. The Flybrid made system would have sat with in the fuel tank area coupled to the crankshaft. This created a system simpler system of equal weight to a typical KERS. The main components being the flybrid flywheel and Torotrak system, plus a relatively small ECU, no additional cooling would be required. Thus the system could be far easier packaged into the chassis robbing just 13l of fuel tank space. A proposal was made to the FIA for a supply of this system to every team on the grid as a cost cutting measure. It seems the FIA did not take up this offer. So it seems this technology may be resigned to lower Formulae or non Motorsport applications.


Honda F1 Electric KERS

Although they also had a flywheel system in place, it transpires Honda also had an electric system.  ITs described in some detail in this link.

Linked images copyright
Generic KERS diagram – Craig Scarborough
McLaren KERS –
Marelli KERS

Follow ScarbsF1 on Twitter

28 thoughts on “KERS anatomy

    • I never heard about Toyota running a flywheel. I do know several teams approached Flybrid to assess the flywheel technology, I heard Mercedes (MHPE) were one of those who approached flybrid, may be Toyota were too.

      I saw the Toyota KERS system at Portimao back in February, it was clearly a battery system, which was since confirmed in Race engine Technology magazines analysis of the Toyota 2009 powertrain.

      It was Jon Hilton of Flybrid that let on to me that Honda were the intended debutante of the flywheel system. This of course was dropped when the team departed the sport leaving the Brawn team behind.

      • Oh, I thought I remembered reading it in 2008 when there were a lot of articles about KERS introduction for 2009.
        Either article was inaccurate or I got Honda and Toyota confused :/

      • My recollection – although this may have been from the BBC – was that Toyota were going to run a Magneti Marelli system. This is the company that has popularised the hybrid car and has now sold over two million of them since 1997. Does the idea that they were going to buy in KERS, and decided not to race it at all, tell us something about the usefulness of KERS?

        The fact is that recovering energy from braking, even on the front brakes, provides relatively little benefit. So much energy is lost to air drag and rolling resistance when decelerating that only a small fraction is captured, and a lot of captured energy is wasted to drag and rolling resistance when accelerating again. I’d be happy not to have it on my Prius, it causes an awkward switch from regen to friction braking if you hit a bump, which was the reason for the recall on the third-generation, as the level of friction braking was often noticeably less than the regen before the switch (accelerator sticking, the other major Toyota recall last year, did not affect the Prius). On my car (Gen 2) it’s limited to taking about 25kW of power (the battery pack fuse is 120A, the pack voltage typically 250-270 volts), which means that braking from high speed needs some friction braking unless you start braking very early indeed.

        The Prius gets its ‘extra-urban’ numbers through an efficient engine coupled to a pretty efficient continously-variable transmission. The engine uses an Atkinson-Miller cycle: very late intake valve closing, between 1/3rd and 2/3rds of the way through the compression stroke. With a high nominal compression ratio, this means that the air/fuel mix is compressed about 10:1 as normal for 91 RON fuel, but has room to expand 13:1, so more useful energy is extracted rather than disappearing down the exhaust pipe. The cost is relatively lower torque and poor power-to-weight ratio. The electrics are there largely to compensate for the low torque and peak power.

        At urban speeds, up to 45mph on current models, the car can turn the engine off and cruise on electrics only. When the battery is depleted a bit, it starts the engine and runs it slightly harder than road demand, which is generally more efficient in terms of energy out vs energy in, though typically worse in MPG terms because more fuel is flowing.

        The transmission looks nothing like a regular belt-and-cone CVT or a conventional manual ‘box. It’s a single planetary gearset with two motor/generators, one of which provides torque to the wheels, and the other controls the relative speed of the engine. This works out more efficient than the belt-and-cone CVT despite double conversion of kinetic energy to electrical and back again. It’s not as efficient as a manual transmission, but the ability to finely control the engine speed generally makes up for it.

        It’s interesting that the third-generation Prius gained 13.5% over its predecessor on the extra-urban part of the EU fuel consumption test, and 28.1% on the urban part, despite not changing the battery’s design and only changing the (air) cooling slightly to get 27kW rather than 25kW maximum charge/discharge rate, and increasing the engine’s nominal capacity from 1.5 to 1.8 litres. That new engine delivers better efficiency overall and at lower revs, partly through external exhaust gas recirculation, and the urban part benefits from exhaust heat capture (pumping coolant through a chamber around the exhaust pipe, to recover heat) warming the engine up faster. The car’s also a bit more aerodynamic, and the speed that the engine *must* run at goes up from 42mph to 45, which is slightly faster than the max speed in the urban part of the test.

        Sorry, this turned into a bit of a Prius advert, but I wanted to emphasise how completely wrong-headed KERS is.

  1. Pingback: KERS anatomy « Scarbsf1′s Blog « szykana

    • Only the pressing of the button out of a turn to request the boost is under the drivers control. The amount and duration of the boost is team controlled via presets in the SECU code.

  2. Why do they limit the usefulness of KERS? This is one of very few F1 technologies that is very applicable to road cars. All the teams can develop it, so why limit the stored energy or timescales. If its safety, then set a maximum stored energy, but let them release it as often as they like.

  3. Ever since its introduction, I’ve wondered why the KERS implementation was a “Fast and the Furious” style boost button. I presume it’s mandated via the FIA rules, but is it a technical limitation of the systems that guided the rules in this direction?

    To my simple mind, a continuous stream of additional power from the system would be far more beneficial, and road-relevant, than a boost-button … or am I missing something?

    • You’re right I think the FIAs vision was like a push to pass system. I’m sure the systems would work better as you describe, but if you recall teams did use the 6s boost in small doses around the lap, only really monza sae long duration boosts.

  4. Thanks Scarbs, another great article.
    Something to add to the wish list would be an article on turbos. I’d love to hear your thoughts on the new (proposed?) engine formula and the engineering implications of such – cooling, packaging, minimising lag etc.

    Julian F

  5. Variable Valve Timing, alternate injection systems, direct etc. are the technology that even hybrids rely upon to improve milage today. Yet these most important developments are banned? or dont exist. Aero development wars in F1 are not relevant to much of anything, movable surfaces will invite flutter accidents and mass balancing and actuator problems that F-ducts rather cleverly avoided. An engine war seems most benficial for manufacturers and earth. Variable Valve timing can allow direct deceleration pumping from the engine to direct air pumping accelertion in the engine, economy modes on and on. If a new turbo era allows variable valve timeing and is an open engine war, that will be great. Indy needs to go back to engine wars too. In short, a severe fueling limit, open engines, severely limited aero…

  6. Hi Craig, any news on how the new, Cosworth teams are going to approach KERS? I believe Xtrac are linked in with the Torotrak/Jaguar flywheel project or has Xtrac’s problems with their gearbox/hydraulic lost them that supply opportunity?

  7. I thought that KERS was within the rules for this year, although declined by FOTA? Again I thought the energy per lap was double (800kj) for this season and would have expected that level for next season? I can’t believe the energy levels are still at 2009 levels, there was just not sufficient advantage to use KERS as Brawn aptly demonstrated.

    They really need to add some additional rules to make KERS work if they are going that way. Many do not understand that the energy derived for KERS comes from the engine when the car is decelerating, not from the back wheels under braking. It’s just not practical to generate the energy from the back wheels unless they allow some method of torque control on the back wheels in this situation.

    We saw last season with KERS how it penalised Kubica more than any other driver. It would be far fairer to set a far higher weight limit at the heaviest diver weight + 5kg, with the proviso that the 5kg + the difference for the lighter driver has to be fitted around the seating area, such no advantage comes from COG change due to driver weight. Otherwise you are just favouring camel jockeys.

    The other aspect with all cost restraints, everything seems to be trying to increase costs without need. Just does not make sense as a cohesive policy.

  8. Pingback: Mercedes AMG: KERS development | Scarbsf1's Blog

  9. Pingback: has anyone made a diesel electric buggy before? - Page 3

  10. Pingback: Renault: European GP Alternator Failures | Scarbsf1's Blog

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s