In the Cluster of Excellence Tailor-made Fuels from Biomass (TMFB) at RWTH Aachen University two alternative fuels were identified as promising fuel candidates to be applied in highly boosted gasoline engines. In thermodynamic investigations the extraordinary properties of the two fuels 2-methylfuran and 2-butanone in comparison to conventional RON95 gasoline and ethanol were demonstrated.

Molecular Structure of the Fuel as Degree of Freedom

The legislative pressure to decrease CO2 emissions in the future caused by road transportation is expected to lead to a broader diversification of the applied energy carriers in the transport sector. Hereby, the transition from fossil fuels to biomass-based energy carriers offers the possibility to define new fuel property requirements which possible alternative fuels have to incorporate. In this context, the approach in the Cluster of Excellence “Tailor-Made Fuels from Biomass (TMFB)” at RWTH Aachen University considers the molecular structure of the fuel as a potential additional degree of freedom for an optimized sustainable production and complete usage of the thermodynamic potential of internal combustion engines. Two possible candidates have been identified and were applied in highly boosted gasoline engines: 2-butanone (methyl ethyl ketone) as well as 2-methylfuran. For both fuels, possible synthesis and production routes have been identified [1–5].

Fuel Properties

Due to its high knock resistance and heat of vaporization, ethanol is known to be very suitable for an application in gasoline engines with high specific loads. By this, a significant efficiency increase in comparison to conventional gasoline fuel can be achieved [6–9]. Hence, ethanol was defined as a base fuel for comparisons in the here presented investigations. Additionally, also conventional gasoline with a research octane number (RON) of 95 (EN228, EN51626-1 respectively) was chosen as a second base fuel as it is the far-most spread fuel for spark ignition (SI) engines in Europe. All relevant fuel properties are summarized in Table 1 [7, 10–15]. In comparison to conventional fuels the three alternative fuels feature an increased oxygen content, leading to a decrease of the mass-based heating value of 36 % in the case of ethanol, 28 % in the case of 2-methylfuran and 25 % in the case of 2-butanone.

TABLE 1
figure 4

Relevant properties of the investigated fuels (© vka)

Full size image

When considering the volumetric heating value, the order of the biofuels changes since a decrease in oxygen content is accompanied by an increase in density. Conventional gasoline has the lowest density. This leads — with regard to conventional gasoline — to a reduction of the volumetric heating value of 32 % in the case of ethanol, 11 % for 2-methylfuran and 18 % for 2-butanone. Due to their high oxygen content the octane number for the investigated biofuels was determined according to DIN51756-7 [16, 17]. The highest RON was measured for 2-butanone, Table 1.

The RON of 2-methylfuran is lower than that of ethanol and 2-butanone, but still exceeds conventional gasoline by five units. The higher heat of vaporization of ethanol leads to a more pronounced cooling of the mixture, especially at higher loads. Thereby, the level of knock resistance as described by the RON procedure is increased, especially in direct injection engines.

Disadvantageously, the higher heat of vaporization leads to challenges with regard to cold start operation. For ethanol, the combination of a low vapour pressure and a higher injected fuel mass necessary at lower stoichiometric air requirements leads to a deterioration of the cold-start behaviour [6]. In comparison to ethanol, 2-methylfuran and 2-butanone feature moderate heats of vaporization, higher vapour pressures, higher specific heating values and higher air requirements [6, 7]. This allows for an improved mixture formation especially under critical operation boundary conditions like cold-start or cold intake air temperatures [18].

In addition to the heat of vaporization also the boiling temperature is a key criterion for the selection of fuels for SI-type engines. In this work, a boiling temperature of 100 °C was defined as maximum value to still enable the degassing of fuel from the oil pan at typical warm operation temperatures in the range of 90 to 100 °C [8]. For both, 2-butanone and 2-methylfuran this criterion was met.

Experimental Set-Up

The experiments were conducted on a direct-injection single cylinder research engine [8, 19]. Due to the solvent-type properties of 2-butanone and 2-methylfuran, the resistance of different gasket materials were investigated in advance [20]. The materials ethylene propylene diene monomer rubber (EPDM), fluoroelastomer (FKM) and nitrile butadiene rubber (NBR) showed an unacceptable high degree of swelling when brought into contact with 2-butanone and 2-methylfuran. Instead the materials polytetrafluoroethylene (PTFE) und perfluoroelastomer (FFKM) were used as sealing materials in the fuel system which showed no swelling when in contact with both fuels. The engine is equipped with an external compressor unit to realize a boost pressure of up to 3.5 bar, Table 2.

TABLE 2 Technical details of the research engine (© vka)
Full size table

The different compression ratios (CR) were realized by adapted piston geometries, Figure 1. The engine features a high peak pressure stability as well as separated, symmetrical intake ports to produce a high degree of tumble motion in the cylinder. The spark plug is located between the exhaust valves, while the injector sits between the intake ports, Figure 1. For the thermodynamic investigations two different injectors were used: an outward-opening piezo actuated injector with a hollow cone (cone angle: 90° ± 3°, max needle lift: ∼30 μm) and a six-hole solenoid injector with a spray pattern optimized for lowest wall impingement.

FIGURE 1
figure 1

Research engine layout (© vka)

Full size image

The air was conditioned to reach 25 °C behind the throttle. When operating in throttled mode, the air pressure before the throttle and behind the exhaust port was set to 1013 mbar. In supercharged operation mode the pressure in the exhaust port was raised to the corresponding value in the intake port. The determination of the relative air/fuel ratio (λ) was calculated from the exhaust gas composition based on the formula of Spindt [21] with the extension for oxygenates by Bresenham [22]. The exhaust gas composition was determined from a heated partial flow taken from the exhaust port. The exhaust gas flow sample was analysed using the following systems:

  • — Hydrocarbons (HC): Flame ionization detector (Rosemount NGA 2000)

  • — Oxygen (O2): Paramagnetic oxygen analyser (Rosemount NGA 2000)

  • — Carbon monoxide (CO): Infrared gas analyser (Rosemount NGA 2000)

  • — Carbon dioxide (CO2): Infrared gas analyser (Rosemount NGA 2000)

  • — Nitrogen oxides (NOX): Chemiluminescence analyser (Eco Physics 700 EL ht).

The partial exhaust gas flow to detect particle emissions was separated behind the exhaust gas back pressure valve at a pressure level of 1013 mbar. It was transferred to a Smokemeter (AVL 415s) to detect the Bosch Filter Smoke Number (FSN) and to an Engine Exhaust Particle Sizer (EEPS by TSI) to detect the particle number. In order to avoid possible errors when determining the HC emissions by using a flame ionization detector (FID) for the oxygenate fuels, an assumption to equivalent earlier investigations by Thewes [7] was made that the HC emissions of all three investigated biofuels exclusively consist of the corresponding fuel molecules. Here, a certain degree of uncertainty remains as the exact composition of the HC emissions cannot be detected.

Engine Test Results

Both TMFB-fuels 2-methylfuran and 2-butanone were investigated as pure substances and compared to conventional RON95 gasoline as well as ethanol. In gasoline engines, the highest possible thermal efficiency depends on the compression ratio, which again is limited by the fuels anti-knock properties. In Figure 2 the knock resistance of the investigated fuels based on 50 % mass fraction burned point (MFB 50) in two different load points at an engine speed of 2000 rpm with two different injector configurations (hollow cone and six-hole injector) is shown. Due to their higher knock resistance all investigated biofuels enabled an increase of the engine compression ratio of five units to 13.5 in comparison to conventional RON95 gasoline fuel. A high knock resistance and therefore highest possible thermal efficiency is indicated by a MFB 50 in the optimal range of 7 to 8 °CA after top dead centre (°CA ATDC).

FIGURE 2
figure 2

Experimental results of the investigated fuels at moderate and high load (© vka)

Full size image

In Figure 2 it can be seen that at an engine load of 12 bar indicated mean effective pressure (IMEP) none of the fuels are knock-limited. With all biofuels the efficiency can be increased by more than 10 % due to the higher compression ratio. In addition the efficiency is affected by gasoline’s lower heat of vaporization and the resulting higher wall heat transfer. The differences between the biofuels can be attributed to the differing heats of vaporization as well.

The nitrogen oxide (NOX) emissions are mainly caused by the temperatures during the combustion [23]. In the case of conventional gasoline the lower compression ratio seemed to compensate the higher enthalpy of vaporization of 2-methylfuran. Ethanol showed the lowest NOX emissions due to its high enthalpy of vaporization in combination with the lowest adiabatic flame temperature. The decrease of NOX emissions between the hollow cone and the six-hole injector for ethanol can be attributed to the improved mixture formation and the resulting cooling of the mixture. This effect becomes even more pronounced when the HC emissions are considered. An increase of HC emissions for ethanol with the hollow cone was observed. This effect is assumed to originate from ethanol evaporating from the cylinder wall oil film which was proven by optical investigations [7].

By changing the injector to a multi-hole type with a narrower spray target the cylinder wall wetting was reduced. This lead to a decrease of HC emissions with ethanol reaching the level of the other fuels. An increase of NOX emissions of 2-butanone on the other hand can be explained by its lower heat of vaporization and its higher adiabatic flame temperature in comparison to ethanol. The measured soot emissions clearly showed a decreasing trend for all three investigated biofuels in comparison to conventional gasoline.

When the load is raised to 24 bar IMEP the improvement of thermal efficiency for the biofuels increases even more due to their higher knock resistance. Although operated at a lower compression ratio, the MFB 50 for RON95 gasoline to 26 °CA ATDC, whereas for 2-methylfuran the necessary adjustments only lead to a delay of 18 °CA ATDC. This results in an efficiency increase of approximately 18 % in comparison to conventional gasoline. Regardless of the above mentioned challenges for the combination of ethanol and the hollow cone injector, a MFB 50 of 10 °CA ATDC was reached for this combination due to the high cooling effect and the high RON. This results in an efficiency increase of 21 % in comparison to conventional gasoline.

When using the six-hole injector, ethanol is no longer limited by knocking combustion so an optimum MFB 50 could be adjusted. The higher indicated thermal efficiency of 42.4 % of ethanol with the six-hole injector (compared to 41.1 % with ethanol and the hollow cone injector) can additionally be attributed to the notable reduction of HC emissions which result from the improved mixture formation. Although the heat of vaporization of 2-butanone is lower than that of ethanol, also here no retardation of the MFB 50 was necessary. This can be explained by the very high ignition delay time of 2-butanone which compensates for the higher mixture temperature in comparison to ethanol [18]. The maximum indicated efficiency of 2-butanone of 41.8 % was in the range of ethanol.

In addition to the load sweep measurements at 2000 rpm also catalyst heating operation was investigated with all four fuels in the representative operating point 1200 rpm, 3 bar IMEP to compare them under critical operating conditions. For the engine oil temperature and the cooling water temperature 40 °C and 30 °C were chosen. The investigations were carried out with the six-hole injector and the high compression ratio of 13.5. For all fuels an equal spark timing (ST) and start of injection (SOI) as well as an ignition-coupled injection with constant injection timing (ti2) were chosen, Figure 3.

FIGURE 3
figure 3

Experimental results under catalyst heating operating conditions (© vka)

Full size image

The burn duration of 2-methylfuran which was calculated by a three pressure analysis (TPA) using the software GT-Power results in a good combustion stability, which is also expressed by a low standard deviation of the IMEP [6, 7]. For the other fuels RON95 gasoline, ethanol and 2-butanone comparable combustion stability was observed. In comparison to ethanol benefits with regard to oil dilution could be observed. The oil dilution depicted in Figure 3 was calculated from the relative air/fuel ratio based on exhaust gas components [21,22] and the relative air/fuel ratio determined from the fuel and air flow measurements.

For ethanol — regardless of the optimized six-hole injector spray pattern — it is assumed that a high degree of cylinder wall wetting occurs which is due to ethanol’s high heat of vaporization, low vapour pressure, elevated fuel flow requirement and higher air requirement. This leads to a higher oil dilution and an increase in HC emissions caused by the following partial desorption of the ethanol from the oil film in the gas exchange phase. The HC emissions of 2-butanone are comparably low as those of 2-methylfuran. Additionally, with all three biofuels a notable reduction in particulate emissions was observed.

Conclusions and Outlook

The two fuels 2-methylfuran and 2-butanone were thermodynamically investigated as pure components in a direct injection single cylinder research engine. It could be proven that both fuels feature extraordinary properties when compared to standard RON95 gasoline and ethanol. For both 2-methylfuran as well as 2-butanone lower HC emissions and oil dilution could be observed at low loads and cold boundary conditions in comparison to ethanol.

Under these critical operating conditions especially 2-methylfuran shows an excellent combustion stability behaviour. At the same time, an increase of the compression ratio and hence an efficiency increase of 18 % at high loads could be achieved due to the higher knock resistance in comparison to standard RON95 gasoline. Most notably, 2-butanone combines a high combustion stability at critical (cold) operation conditions with the highest knock resistance at the same level as ethanol which allows for an efficiency increase well above 20 % in comparison to RON95 gasoline in high load points. It can be stated that, although the compression ratio was increased to 13.5, no knock-limited engine operation area was reached. One disadvantage of 2-butanone and 2-methylfuran in comparison to ethanol are the observed elevated NOX emissions which only play a role for lean operated gasoline engines. For all three biofuels a notable reduction of the measured particulate emissions can be reported.