MTZ worldwide

, Volume 73, Issue 5, pp 44–49 | Cite as

Potentials and Limits of CO2 Emissions of Gasoline Engines

Part 2: New Combustion Processes
  • Rudolf Flierl
  • Frederic Lauer
  • Stephan Schmitt
  • Ulrich Spicher
Development Gasoline Engines

In MTZ 4 the Technical University of Kaiserslautern and Karlsruhe Institute of Technology (KIT) have presented the mechanical possibilities to further develop homogeneously and stoichiometrically operated gasoline engines. This second part of the article describes new combustion processes to increase engine efficiency. For this purpose stoichiometric and stratified lean operation mode as well as future combustion processes like HCCI and their potentials are presented. The described potentials and limits of both articles are combined and contrasted with each other.


Already in [1] the possibilities and limits of thermodynamic approaches to optimise process on homogeneously, stoichiometrically operated gasoline engines via valve train variability were evaluated. Therefore the application of cam phasers, discretely staged valve lift shifters and fully variable systems were separately examined. Furthermore the potential to reduce fuel consumption with cylinder deactivation in throttle free engine operation was demonstrated; as well as the technology combination of fully variable compression ratio and mechanically fully variable valve train on inlet and outlet sides. This present article extends these statements by possibilities and limits to further develop combustion process.


Fuel consumption of gasoline engines is reduced when operated homogenous lean, respectively with excess air. This homogeneous-lean mode is adjusted through higher inlet valve lifts which effectively decrease the charge-cycle-work losses compared to stoichiometric throttle-free operation. Homogeneous leaning results in charge dilution which reduces the residual gas compatibility of the engine. Thus, full charge cycle potential in lean operation mode cannot be exploited. For this purpose smaller inlet and outlet spreads would be necessary which are not applicable through the reduced residual gas capability. On a 3.0-l engine the effective fuel consumption is reduced by approximately 7.5 g/kWh or 2.5 % at n = 2000 rpm, BMEP = 2 bar through homogenous-lean engine operation mode with throttle-free load control, .

Effective specific fuel consumption through homogenous leaning

During homogenous-lean operation mode, efficient exhaust gas after treatment by three-way catalyst is no longer useable to reduce NOx emissions. Thus, additional extensive exhaust gas after treatment is necessary, e.g. NOx storage catalyst or SCR. From today’s perspective the transfer into mass production is irrelevant due to the small advantages in fuel consumption and the costly emission control.


The idea of lambda-split process is to run a part of the engine’s cylinders stoichiometrically with λ = 1 while the remaining cylinders are operated in homogenous lean mode with λ > 1. Applying lambda-split process to the subsequently analysed six-cylinder engine means, three cylinders are run stoichiometrically whilst three cylinders are operated in homogenous-lean mode. Lean operation mode is applied through reducing the fuel injection into these cylinders while keeping the amount of air constant. In so doing, the load on these cylinders decreases. In order to maintain torque output of the engine, load and hence efficiency of the three cylinders which run stoichiometrically increases. Within the leaned cylinders, on the one hand efficiency increases through homogeneous lean burning, but efficiency decreases through shifting the load point to a lower indicated load. In total, the advantages of the load point and efficiency shifting overweigh, so that savings in fuel consumption of 2.5 % could be gained, .

Lambda-split process and its advantages in fuel consumption

At higher load points with increased residual gas compatibility, fuel consumption advantages of up to 8 % could be reached [2]. Thus, this technique is particularly suited in combination with throttle-free load control to improve fuel consumption at higher loads (customer fuel consumption). However, after treatment of exhaust emissions also has to be adapted to this engine operation mode.


The fuel consumption which can be achieved today by spray-guided gasoline direct injection is described by Waltner et al. in [3] using a 3.5-l naturally aspirated engine with multi-injection through piezo injectors. The determined consumptions as well as the potentials compared to homogeneous stoichiometric engine operation are displayed in . When relating these results to a 2.0-l naturally aspirated engine, the calculated outcome is approximately 300 g/kWh at an engine operating point of n = 2000 rpm, BMEP = 2 bar compared to the 3.5-l naturally aspirated engine with 290 g/kWh. The reason is the increased power consumption of the auxiliaries. In a 2.0-l naturally aspirated engine, this operating point corresponds approximately to the engine performance of NEDC’s current average speed.

Spray-guided direct injection and its advantages in fuel consumption [3]

Since the average driving speed actually driven on the street is approximately double this amount, a higher average engine performance is required in every day vehicle operation. The engine operating point most applicable to this situation is at n = 2000 rpm, BMEP = 5 bar. In a naturally aspirated engine, the specific fuel consumption at this particular operating point lies at 255 g/kWh which corresponds to a specific efficiency of 33 %. When relating this efficiency to the actual efficiencies between 25 and 30 % achieved today in everyday vehicle operation with homogenous gasoline direct injection and charging, an efficiency increase of 10 up to over 30 % can be attained through the use of spray-guided direct injection with stratified charge and piezo injectors. Through more intensive research and further development of this combustion process, focusing especially on the injection technique used here, but also through improved fuels, efficiencies of more than 40 % may be further attained. To date, spray-guided direct injection opens up the highest potential so far for fuel economy that can be achieved with today’s approved technologies.

Similarly to direct injection with homogenous mixture preparation, further improvements can also be accomplished in this mode through charging. Indeed, initial tests in [4] with especially high injection pressures of up to 1000 bar, using multi-hole diesel-nozzles and charging, have shown that in stratified charge mode, efficiency increases of over 40 % (indicated) and over 35 % (effective) with further fuel consumption savings (bi = 208 g/kWh respectively be ≈ 230 g/kWh) are possible, . In so doing the engine was run at a degree of charge of 1.5. Thus, at an indicated mean pressure of IMEP = 6 bar (BMEP = 5 bar), a very lean air ratio of λ = 3.3 arose.

Specific fuel consumption using stratified spray-guided gasoline direct injection with charging (n = 2000 rpm, BMEP = 6 bar, injection pressure = 1000 bar)

It was also evident that the particle emissions (weight and number) in particular stayed clearly below future limits [5]. demonstrates the soot concentration for the same engine operating point with IMEP = 6 bar and an engine speed of n = 2000 rpm. The engine was run here using stratified injection in naturally aspirated mode and was thus operated without charging. The injection pressure was varied in steps of 100 bar, starting from 200 bar proceeding to a maximal injection pressure of 1000 bar. A clear decrease in soot concentration was observed through increasing the injection pressure. The drop in soot concentration from approximately 3 to 0.1 mg/m3 was a result of the reduced number of particles with large diameters. All in all, the number of particles decreases by 88 %, while the particle weight was even reduced by 97 % through an increase of the injection pressure. displays the overall mass of particles for an engine load variation in stratified mode at an engine speed of n = 2000 rpm for injection pressures of 200 bar and 1000 bar utilising the two injection nozzles applied in [5].

Soot concentration when varying the injection pressure, using stratified spray-guided gasoline direct injection with charging (n = 2000 rpm, pmi = 6 bar) [5]

Overall mass of particles using stratified guided gasoline direct injection and load variation (n = 2000 rpm) [5]

It is evident that the overall mass of emitted particles at an injection pressure of 1000 bar is considerably lower than that at an injection pressure of 200 bar. In fact, up to a mean pressure of IMEP = 6 bar, the particle mass remains below 0.05 mg/m3 at an injection pressure of 1000 bar, and only starts to increase at IMEP = 7 bar to 0.5 mg/m3 at IMEP = 8 bar. Given today’s current practice of maximum injection pressures of 200 bar, such a result can only be achieved at an indicated mean pressure of 3 bar.

In so doing a reduction of NOx emissions could likewise be achieved compared to an operation with injection pressures of 200 bar, although further testing is here required. In particular, exhaust gas retention and exhaust gas recirculation strategies are relevant here. In light of this aspect, a combination of direct injection and fully variable valve control on inlet and exhaust side is promising. In fact, tests on stratified exhaust gas recirculation [16] have shown that NOx emissions can be reduced by far more than 90 %, in some cases even up to 99 %.


A mechanically fully variable valve train on the inlet and exhaust side and a variable compression ratio are the best preconditions for homogeneous charge compression ignition (HCCI) operation in a wide engine map range. The conflict between high effective compression ratio at part load and reachable high mean pressures can thereby also be resolved. According to Gottschalk [7], a consumption of 270 g/kWh at part load can be achieved when injecting 8 mg fuel in a turbo charged engine with a geometrical compression ratio of ε = 13:1 in lean engine mode, .

Comparison of efficiency effects in combustion process and degrees of freedom in valve train at the part load point mFuel = 8 mg, n = 2000 rpm [7]

The research described in [8], which examined basic relations and phenomena between injection, mixture preparation, self ignition and combustion in premixed fuel/air preparations, showed that through homogenous charge compression ignition not only fuel consumption can be reduced, but furthermore NOx emissions can also be drastically decreased. shows the results of engine performance at loads between IMEP = 4 bar and IMEP = 6 bar at an engine speed of n = 2000 rpm as well as for an operating point at an engine speed of n = 1500 rpm and an indicated mean pressure of IMEP = 6 bar. The engine was operated here in such a way that the NOx emissions always remained below 10 ppm (0.07 g/kWh), which results in an air ratio of between λ = 1.65 and 1.8. The engine was additionally charged with an externally serviced compressor. Intake pressure and exhaust gas back pressure were similarly modulated in order to adjust the engine operating points in a precise manner. The time at which 50 % of the fuel mass is burned (MFB = 0.5) was set to be at 8 °CA after top dead centre. The variance of the cyclic deviation in indicated mean pressure demonstrated values of approximately 1 %. The midrange maximum pressure gradient fluctuated between 5.5 and 7 bar/° CA. The specific fuel consumption was compared to a similar engine with charging, variable valve control and gasoline direct injection (TVDI). At each of the investigated engine operating points, the advantage in fuel consumption in HCCI mode was between approximately 6 and 7 %. As well as showing a clear reduction of exhaust emissions, in particular of NOx emissions, this also shows that in comparison to today’s current combustion processes, the HCCI combustion technique offers an advantage in terms of fuel consumption. This advantage is not so clearly noticeable compared to stratified charge operation with direct injection and charging.

Air ratio, improvement in fuel consumption, variation coefficient of mean pressure deviation and pressure gradient in HCCI mode with pilot injection [8]

Nevertheless, this combustion technique remains of interest for the further development of gasoline engines. Particularly the extremely low levels of exhaust emissions might make extensive after treatments of exhaust gas (particle filters; NOx catalysts) unnecessary. In this context it is by all means also worth considering combining direct fuel injection with stratified charge and HCCI mode, depending on the operating point in the engine map.


This paper shows that there is still a high potential to reduce the fuel consumption and CO2 emissions in gasoline engines. When examining future combustion technologies it is noticeable that development departs from conventional gasoline combustion. A combination of different combustion methods may allow the lowest possible fuel consumption and exhaust emissions in the overall engine load map.

Some new combustion technologies also make demands on engine mechanics, e.g. fully variable valve trains for advanced residual gas control respectively variable compression ratio. To investigate the limits of such potentials in the combustion of gasoline engines, thermodynamic and mechanical approaches have to be further advanced together in joint research projects, and .

Comparison of the potential to reduce fuel consumption of various technologies

Comparison of the potential to reduce fuel consumption of various technologies

When comparing the different measures aimed at improving engine efficiency through deliberate development of injection systems, charging, valve trains, friction and auxiliaries as well as factors in basic engine design (geometry of combustion chamber, charge motion, compression), drive train efficiencies of 40 % and beyond are expected to be attained. In relation to the NEDC, according to which today’s new vehicles of the B-segment are consuming on average approximately 6.5 l fuel per 100 km (146 g CO2/km) [9], 40 % of efficiency results in 4.15 l per 100 km (93 g CO2/km). It is therefore the case that the required 95 g CO2/km by the year 2020 for a vehicle of the B segment (e.g. VW Golf) can be achieved simply through a consequent enhancement of gasoline engines, using additional funds that are still lower than the costs of diesel engines and far below the costs of new driving systems (full hybrid, plug-in hybrid, electric vehicles, fuel cell).


  1. [1]
    Flierl, R.; Lauer, F.; Schmitt, S.; Spicher, U.: Potentials and Limits of CO2 emissions of gasoline engines — Part I: Mechanic Techniques. In: MTZ 73 (2012), No. 4Google Scholar
  2. [2]
    Werth, C.: Lambda-Split. University of Kaiserslautern, Diploma Thesis, 2010Google Scholar
  3. [3]
    Waltner, A.; Lückert, P.; Doll, G.; Kemmler, R.: New 3.5-l V6 Gasoline Engine with Direct Injection from Mercedes-Benz. In: MTZ 71 (2010), No. 9Google Scholar
  4. [4]
    Spicher, U.; Sarikoc, F.: Engine Downsizing by Gasoline Direct Injection and Turbocharging. ISROMAC-13, International Symposium on Transport Phenomena and Dynamics of Rotary Machinery, Hawaii, 2010Google Scholar
  5. [5]
    Schumann, F.; Buri, St.; Kubach, H.; Spicher, U.; Hall, M. J.: Untersuchung zur Partikelemission bei Benzin-Direkteinspritzung mit Einspritzdrücken bis zu 1000 bar. Conference 19th Aachen Colloquium Automobile and Engine Technology, 2010Google Scholar
  6. [6]
    Sarikoc, F.: Untersuchungen zur Reduzierung der Stickoxidemissionen bei modernen Brennverfahren für Motoren mit Benzin-Direkteinspritzung. Karlsruhe Institute of Technology, Doctoral Thesis, 2009Google Scholar
  7. [7]
    Gottschalk, W.: Wirkungsgrad- und Emissionspotentiale zukünftiger Brennverfahren von Pkw-Motoren. Ilmenau University of Technology, Habilitation Treatise, 2010Google Scholar
  8. [8]
    Hübner, W.; Gehrke, U.; Boulouchos, K.; Peters, N.; Schulz, C.; Spicher, U.: Influence of pilot injection on controlled auto-ignition combustion — A theoretical and experimental study. SIA International Conference The spark ignition engine of the future, Strasbourg, 2009Google Scholar
  9. [9]
    Spicher, U.: Analyse und Effizienz zukünftiger Antriebssysteme für die individuelle Mobilität — Vergleich und Grenzen. MTZ Conference Der Antrieb von morgen, Wolfsburg, 2012Google Scholar

Copyright information

© Springer Fachmedien Wiesbaden 2012

Authors and Affiliations

  • Rudolf Flierl
    • 1
  • Frederic Lauer
    • 1
  • Stephan Schmitt
    • 1
  • Ulrich Spicher
    • 2
  1. 1.University of KaiserslauternKaiserslauternGermany
  2. 2.Karlsruhe Institute of Technology (KIT)KarlsruheGermany

Personalised recommendations