ETH Zurich - Home
unit title

Research Database - Projects list - Project 21111
printversion e-mail  

Project view

Project title: Pneumatic Hybridization of Combustion Engines

Record ID: 21111  
Project status: Ongoing (01.01.2008)

Project Leader(s): Onder, Christopher (
Guzzella, Lino (

Participating researchers:

Internal Researcher(s): Voser, Christoph (
Zsiga, Norbert (
Organizational unit: Chair: Guzzella, Lino <> (03286)
Institute: Institut für Dynamische Systeme und Regelungstechnik (02619)
Department: Departement Maschinenbau und Verfahrenstechnik (02130)

Short Summary:

Pneumatic hybridization of combustion engines offers a large potential for fuel consumption reduction while guaranteeing excellent driveability and a cost effective system, especially if compared to hybrid electric systems.

Research fields (keywords):

Selected keywords are marked red

Engineering Sciences

Project description:

Click to collapse


In spite all of its well-known problems, the internal combustion engine (ICE) continues to be the most important means of automobile propulsion and is expected to remain so for several decades. However, its fuel consumption and consequent CO2 emissions are symptomatic of major system-inherent problems. First, the vehicle's kinetic energy cannot be transformed back into chemical or other usable energy when braking. Second, consumer demand for high maximum power leads to increased engine displacement. The mean efficiency of the propulsion system is consequently decreased, since internal combustion engines exhibit a low efficiency at part load. Third, a fast and convenient engine startup with a conventional starter after a short vehicle standstill is not possible. This prevents shutting off the engine beforehand and thus leads to additional fuel consumption during idling.

Automobile manufacturers are currently attempting to tackle these problems by focusing on hybrid electric vehicles (HEVs). Some research groups, however, have been investigating another alternative, namely the pneumatic hybridization of ICEs.

The basic idea of pneumatic hybridization is to use the ICE not only for combustion, but also as a pump and a pneumatic motor. All other known hybrid propulsion systems require additional separate devices for energy conversion, a fact that drives up cost and weight.

For pneumatic hybridization, each cylinder of the ICE is connected via a fully variable charge valve to a shared air pressure tank, see Figure 1. In vehicle braking phases with fuel cut-off, the engine can intake air and pump it into the pressure tank. Without the injection of fuel, the pressurized air can then be used again for starting or driving the vehicle. Shifting the operating point of the engine is also possible, using at least one cylinder in a conventional combustion manner at a high load and using at least one cylinder to pump air into the pressure tank. The pressurized air can also be used to boost the conventional engine combustion mode, thereby overcoming the turbo lag in supercharged engines.

The pneumatic hybridization of an internal combustion engine is possible for both spark ignition (SI) and compression ignition (CI) engines.

The hybrid pneumatic engine (HPE) was long thought to be a system in which all engine valves had to be actuated in a fully variable manner to allow for two-stroke pneumatic modes and still enable the four-stroke combustion mode. However, this setup adds complexity and cost. During this project, the authors patented the idea to stick to four-stroke cycles for both combustion and pneumatic operations. The intake and exhaust valves remain camshaft-driven and only the charge valve, which connects the cylinders to the pressure tank, is fully variable. An overview for the valve actuation is given in Figure 2.


The fuel consumption calculations all use the quasi-static simulation technique. The supervisory control deciding on which engine mode to use is based on the deterministic dynamic programming (DP) technique. It yields the optimal control sequence for a given drive cycle while guaranteeing charge sustenance. Due to its optimality, this technique provides a benchmark for the hybrid propulsion system, because it ensures comparability.

All fuel consumption calculations were carried out for the MVEG-95 driving cycle. To ensure comparability between gasoline and diesel engines, the amount of CO2 savings relative to the 2.0l naturally aspirated (NA) SI engine is chosen as a measure instead of fuel savings. The results are shown in Figure 3. Several things should be noted:
- Downsizing a gasoline engine can save as much fuel (and CO2) as 25%. It is the most important contribution to the overall fuel gain. Note that without the HPE concept, the "turbo lag" would be an issue.
- The pneumatic hybridization for an engine system based on fixed camshafts for intake and exhaust valves in combination with downsizing the gasoline engine yields overall fuel savings of 32%. Note that the "hybridization gain" of 7 percentage points is just a small share of the overall fuel gain. Part of this hybridization gain (approx. 4%) comes from using the pneumatic motor mode instead of the conventional mode, using energy from the tank that was stored during recuperation or recharge. The remaining 3 percentage points come from avoiding engine idling by utilizing engine stops with subsequent pneumatic starts. The fuel savings resulting from downsizing and applying start/stop is also shown in Figure 3.
- An overall amount of fuel savings of 34% is achieved when considering a downsized camless gasoline engine with pneumatic hybridization and cylinder deactivation.
- The conventional diesel engine produces less CO2 than the baseline gasoline engine with the same maximum power. However, for the hybrid pneumatic diesel engines, the overall amount of CO2 savings is lower than for downsized hybrid gasoline engines. The main explanation for that is that the diesel engine cannot be downsized any further.

The most important result of this investigation is that the effect of downsizing the engine is much more important than the hybridization itself.


The HPE concept features the supercharged mode, offering the option to inject pressurized air from the tank into the cylinder during the compression phase.
This enables the injection of additional fuel for both PFI and DI engines, providing the fastest torque response possible using the air path of the ICE. This mode is always available as long as the tank pressure is high enough (avoid mixture flow to the tank). Using regenerative braking or the recharge modes, a high enough tank pressure can be guaranteed for almost all driving conditions.

The supercharged mode is especially helpful for overcoming the so-called "turbo lag" in turbocharged engines, since they typically respond slowly to a demanded torque step at low engine loads and speeds. Once the torque increases by activating the supercharged mode, the engine exhaust enthalpy is increased, thereby propelling the turbine of the turbocharger. Since the accelerating compressor of the turbocharger leads to an increased intake pressure, the additional air from the tank is needed only for a very short time.


The HPE concept is assumed to not affect emissions and the added cost is rather low. The advantages of the concept are the excellent driveability and the improved fuel consumption, especially when considering the downsizing effect.

The HPE concept is especially effective when it is combined with strong downsizing and turbocharging of SI engines. The synergy effects of these two concepts are evident.


For this project, a downsized and turbocharged gasoline engine was chosen to be modified into an HPE. It is a two-cylinder multi-purpose engine (MPE750) with a displaced volume of 0.75 liter. In its original form, the engine was designed by Wenko Swissauto AG, manufactured by Weber Automotive GmbH. The engine has a compression ratio of 9.0, and it uses port fuel injection. The engine is equipped with a GT12 turbocharger. The engine modifications for the pneumatic hybridization were made in cooperation with Wenko Swissauto AG.

For the engine modification, the fixed camshaft configuration was chosen. The fully variable valve actuation for the charge valves was realized using the Bosch electro-hydraulic valve system (EHVS). The chosen engine normally has two intake and two exhaust valves per cylinder. It was decided to replace one exhaust valve per cylinder by a charge valve that is actuated by the EHVS. A picture of the modified cylinder head is shown in Figure 4. A picture of the turbocharged engine system is shown in Figure 5. The testbench setup is shown schematically in Figure 6.


The engine was modified successfully and all conventional and new engine modes could be tested. The most important result is the switch from the conventional mode to the supercharged mode. The resulting instantaneous torque step is shown in Figure 7, where the cylinder pressure traces are shown. The supercharged mode is used to overcome the engine’s turbo lag. Measurements comparing the system’s torque response with and without the supercharged mode is shown in Figure 8. The pneumatic start of the engine was also realized. This rapid start is fast enough to justify the elimination of engine idling and thus enables further fuel savings. A measurement of such an engine start is shown in Figure 9.


To determine the actual fuel saving potential of this modified engine, vehicle emulation tests were carried out. For these tests, the engine is connected to an engine dynamometer that behaves like an actual vehicle in a drive cycle. The engine control unit reacts to any velocity error in the drive cycle by adjusting the control signals that are sent to the engine. The supervisory control uses the aforementioned dynamic programming method to determine the optimal engine mode to be used. The fuel consumption is measured using a state-of-the-art fuel scale. To enable a fair comparison, the modified engine is considered for a series production vehicle. The measured fuel consumption is then compared to that of the same series production vehicle equipped with the standard engine of the same rated power.
The measurement results of such a vehicle emulation experiment are shown in Figure 10. It shows the different driver modes, engine modes, the air tank pressure trajectory, and the velocity deviation for a VW Polo in the New European Drive Cycle.
Such vehicle emulation experiments were carried out for several vehicles, the measured fuel consumption reductions compared to the vehicles’ standard engines are shown in Figure 11.


Thus far, fully variable camless valve-trains have been investigated for the charge valve actuation. These systems are expensive, complex and are not available in series production, which prevented a breakthrough of the HPE technology. Hence it was evaluated whether camshaft driven systems currently available can be used to control the air exchange. Such systems are simpler and more cost-effective, which makes them more attractive for a realization. The reduced complexity and cost of the valvetrain come at a loss of variability which can be overcome by control strategies. Hence, the design of the valve-train system of the charge valve becomes crucial.


As mentioned earlier, the fuel consumption saving due to strong downsizing is very large. The boost mode enables strong downsizing, since it solves the driveability problems that arise at low engine torques and speeds (i.e. turbo lag).
In order to reduce complexity and costs of the system even further, a camshaft driven version of the charge valve was analyzed for the application in the boost mode. A design framework was established which relates the most important valve parameters to each other without violating existing limits.
Two control strategies for the boost mode were developed in simulation. By emulating the behavior of the camshaft driven charge valve by the EHVS, both control strategies were experimentally verified.
The authors patented the idea of realizing the boost mode with a camshaft driven valve.


For the pneumatic start a design framework was established, too. For a desired start performance, values for the most important valve parameters such as the diameter, valve profile and tank pressure can be found.
By emulating the behavior of the camshaft driven charge valve with the EHVS, the start performance was experimentally verified.


Due to the promising simulation and emulation results a new prototype engine was developed which consists of cam-actuated charge valves. In addition, the new engine has several actuators for the control of the gas exchange.
The current research focus is on the implementation of the various engine modes. Furthermore, it is investigated how the additional actuators need to be controlled to achieve the best system performance, i.e. torque response, fuel consumption, mode transition.
Popular description:

Click to collapse

The hybrid pneumatic engine concept developed at IDSC uses compressed air as energy buffer and allows for substantial downsizing of engines. Both measures lead to an increased efficiency and an excellent driveability of the automotive propulsion system. Hybrid pneumatic vehicles are expected to be much cheaper than hybrid electric vehicles since no batteries or electric motors are needed.

The project team has realized the world's first fully functional hybrid pneumatic engine. The researchers have proven that this new engine concept saves around 30% fuel compared to conventional gasoline engines with the same power.

Graphics 1: Figure 1: Schematic depiction of a downsized
and turbocharged four-stroke hybrid pneumatic
Graphics 2: Figure 2: (Rough) valve actuation overview
for all engine modes (IV: intake valve, EV:
exhaust valve, CV: charge valve, TC: top dead
Graphics 3: Figure 3: CO2 saving relative to the 2 liter
NA SI baseline engine for the MVEG-95 driving
Graphics 4: Figure 4: Modified cylinder head with EHVS
Graphics 5: Figure 5: Testbench setup of turbocharged
hybrid pneumatic engine
Graphics 6: Figure 6: Testbench structure  
Graphics 7: Figure 7: measurement: torque step using the
supercharged mode at constant intake
Graphics 8: Figure 8: Overcoming the turbo lag using the
supercharged mode
Graphics 9: Figure 9: Rapid pneumatic engine start and
switch to combustion
Graphics 10: Figure 10: Vehicle emulation of a VW Polo
(New European Drive Cycle): measurement
Graphics 11: Figure 11: Vehicle emulation results:
measured fuel consumption reduction for
series production vehicles

Publication 1: 42760, Dönitz C., Vasile I., Onder, C., Guzzella, L., Higelin P., Charlet A., Chamaillard Y., Pneumatic hybrid internal combustion engine on the basis of fixed camshafts  
Publication 2: 42755, Dönitz, C.; Vasile, I.; Onder, C. H.; Guzzella, L., Modelling and optimizing two- and four-stroke hybrid pneumatic engines  
Publication 3: 51607, Dönitz C., Voser C., Onder C., Guzzella L., Turboaufgeladene Hubkolbenkraftmaschine mit angeschlossenem Drucktank zur Turbolochüberbrückung und Verfahren zum Betrieb derselben  
Publication 4: 42758, Dönitz, Christian; Vasile, Iulian; Onder, Christopher; Guzzella, Lino, Dynamic Programming for Hybrid Pneumatic Vehicles  
Publication 5: 42757, Dönitz, Christian; Vasile, Iulian; Onder, Christopher H.; Guzzella, Lino, Realizing a Concept for High Efficiency and Excellent Driveability  
Publication 6: 85610, Dönitz, Christian; Voser, Christoph; Vasile, Iulian; Onder, Christopher; Guzzella, Lino, Validation of the Fuel Saving Potential of Downsized and Supercharged Hybrid Pneumatic Engines Using Vehicle Emulation Experiments  
Publication 7: 51289, Guzzella, Lino; Onder, Christopher; Dönitz, Christian; Voser, Christoph; Vasile, Iulian, Das Downsizing-Boost-Konzept auf Basis der pneumatischen Hybridisierung von Ottomotoren  
Publication 8: 124708, Moser, Michael M.; Voser, Christoph; Onder, Christopher H.; Guzzella, Lino, Design Methodology of Camshaft Driven Charge Valves for Pneumatic Engine Starts  
Publication 9: 108957, Moser, Michael; Voser, Christoph; Onder, Christopher; Guzzella, Lino, Design Methodology of Camshaft Driven Charge Valves for Pneumatic Engine Starts  
Publication 10: 51261, Vasile I., Dönitz C., Voser C., Vetterli J., Onder C., Guzzella L., Rapid Start of Hybrid Pneumatic Engines  
Publication 11: 116000, Voser, Christoph, System design of a directly air assisted turbocharged SI engine using a camshaft driven valve  
Publication 12: 106150, Voser, Christoph; Dönitz, Christian; Ochsner, Gregor; Onder, Christopher; Guzzella, Lino, In-cylinder boosting of turbocharged spark-ignited engines. Part 1: Model-based design of the charge valve  
Publication 13: 117156, Voser, Christoph; Onder, Christopher; Guzzella, Lino, System Design and Analysis of a Directly Air-Assisted Turbocharged SI Engine with Camshaft Driven Valves  
Publication 14: 109755, Voser, Christoph; Onder, Christopher H.; Guzzella, Lino, System Design for a Direct-Boost Turbocharged SI Engine Using a Camshaft Driven Valve  
Publication 15: 109754, Voser, Christoph; Ott, Tobias; Dönitz, Christian; Onder, Christopher H.; Guzzella, Lino, In-cylinder boosting of turbocharged spark-ignited engines. Part 2: Control and experimental verification  
Publication 16: 116128, Zsiga, Norbert; Voser, Christoph; Onder, Christopher; Guzzella, Lino, Intake Manifold Boosting of Turbocharged Spark-Ignited Engines  
Links to important web pages:

Link 1:  
Link 2:  
Funding sources:

Funding source 1: Public institutions (e. g. federal offices)  
Funding source 2: Industry  

Click here to download a complete documentation of this project as a PDF file.

© ETH Zürich | Imprint | Release vom 24.06.2013 | v6.11