- Email: [email protected]
CO) in a spark-ignition direct-injection engine. Part 1: Combustion, performance and emissions comparison with CNG
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Syngas (H2/CO) in a spark-ignition direct-injection engine. Part 1: Combustion, performance and emissions comparison with CNG Ftwi Yohaness Hagos a,*, A. Rashid A. Aziz b,c, Shaharin A. Sulaiman c a
Faculty of Mechanical Engineering, Universiti Malaysia Pahang, 26600 Pekan, Pahang, Malaysia Centre for Automotive Research and Electric Mobility (CAREM), Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, Malaysia c Department of Mechanical Engineering, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, Malaysia b
article info
abstract
Article history:
The combustion, performance, and emissions of syngas (H2/CO) in a four-stroke, direct-
Received 31 May 2014
injection, spark-ignition engine were experimentally investigated. The engine was oper-
Accepted 28 August 2014
ated at various speeds, ranging from 1500 to 2400 rev/min, with the throttle being held in
Available online 23 September 2014
the wide-open position. The start of fuel injection was fixed at 180 before the top dead center, and the ignition advance was set at the maximal brake torque. The air/fuel ratio
Keywords:
was varied from the technically possible lowest excess air ratio (l) to lean operation limits.
Syngas
The results indicated that a wider air/fuel operating ratio is possible with syngas with a
Direct-injection
very low coefficient of variation. The syngas produced a higher in-cylinder peak pressure
Spark-ignition
and heat-release rate peak and faster combustion than for CNG. However, CNG produced a
Combustion
higher brake thermal efficiency (BTE) and lower brake specific fuel consumption (BSFC).
Performance
The BTE and BSFC of the syngas were on par to those of CNG at higher speeds. For the
Emission
syngas, the total hydrocarbon emission was negligible at all load conditions, and the carbon monoxide emission was negligible at higher loads and increased under lower load conditions. However, the emission of nitrogen oxides was higher at higher loads with syngas. Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Introduction There is growing concern regarding the use of fossil-derived fuels in the transportation sector, primarily due to their emissions and energy-sustainability issues. Furthermore,
advancements in conversion technologies and the abundant availability of solid fuels, along with advancements in the technology of gas engines and their fueling systems, have caused a revival of interest in syngas. Additionally, syngas is a potentially carbon-neutral and renewable fuel. Over the past 30 years, there have been many efforts to bring back the
* Corresponding author. E-mail addresses: [email protected], [email protected], [email protected] (F.Y. Hagos), [email protected] (A.R.A. Aziz), [email protected] (S.A. Sulaiman). http://dx.doi.org/10.1016/j.ijhydene.2014.08.141 0360-3199/Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
technology of syngas in internal combustion engines (ICEs), both in spark-ignition (SI) [1e5] and compression-ignition (CI) engines [6e9]. However, the application of syngas in SI engines was limited to carbureted and port-injection engines, and such fueling systems are now obsolete due to their poor performance. On the other hand, the use of syngas in CI engines was restricted to dual fueling with diesel as a pilot fuel to assist in the ignition. In this arrangement, syngas is not a standalone fuel, and it cannot completely replace diesel. There have been several works reported in the literature related to syngas fueling in carbureted and port-injection SI engines. Bika [10] studied various ratios of H2/CO syngas in a port-injection SI engine for their combustion characteristics and knock limit. The fuel was injected 4 cm from the intake port. The fuels investigated included pure H2, 75% H2/25% CO and 50% H2/50% CO. The study was limited to an equivalence ratio of (f) ¼ 0.6e0.8 and compression ratios of 6:1 to 10:1. It was reported that an increase in the content of CO increased the knock limit of the syngas. The study also indicated that the increase in the CO content advanced the ignition timing of the maximum brake torque (MBT). The overall conclusion of the study was that there is an increase in the combustion duration with an increase in the CO content. A maximal indicated thermal efficiency of 32% was reported with 50% H2/ 50% CO at f ¼ 0.6 and a compression ratio of 10:1 [10]. The study was limited to combustion only. The performance and emissions of these fuels were not investigated. However, the study highlighted the extension of the knock limit with an increase in the CO content in syngas. Mustafi et al. [3] investigated the performance and emission of power gas in a variable-compression-ratio SI engine. The molar ratios of the fuels investigated were 0.52, 0.44 and 0.04 for CO, H2 and N2, respectively. The lower heating value and stoichiometric air/fuel ratio were reported to be 15.3 MJ/kg and 4.2, respectively. The power output of the power gas, gasoline and compressed natural gas (CNG) were compared at a constant speed of 1500 rev/min. It was reported that the brake torque of the power gas was 30% and 23% lower than those of gasoline and CNG, respectively. The power-gas consumed brake specific fuel consumption (BSFC) () 2.7 and 3.4 times in kg/kW-h more than those of gasoline and CNG, respectively. Reduced emissions of total hydrocarbons (THCs) and CO were reported with power-gas compared to results with both gasoline and CNG. However, the CO2 and NOx emissions were higher [3]. To the knowledge of the authors, no study to date has investigated syngas in direct-injection (DI) SI engines. Neither its combustion behavior nor its performance and emission characteristics are known for the broad range of syngas types produced by gasification technologies. Moreover, there is no suitable predictive tool for the effect of the properties of syngas on the optimization of the combustion, performance and emissions, except for some recent trials [2,11,12]. Therefore, advancing the previous studies and targeting the latest engine technologies in the investigation of these fuels in ICEs is important. The results concerning the performance of DISI engines using other gaseous fuels (CNG, H2 and a mixture of both) have been presented by other authors [13e16]. Therefore, the objective of this study is to investigate the use of syngas in a DISI engine as a potential replacement for
17885
fossil-derived fuels. This article is part of an ongoing research project on fueling with syngas with a 50% H2:50% CO composition in a DISI engine. The scope of this article is limited to an experimental study of the combustion, performance and emissions characteristics of syngas in DISI engines, compared to baseline data for CNG. The effect of the injection timing on the combustion, performance and emissions of syngas in the same engine has also been investigated, and the results are presented in part 2, an accompanying article awaiting publication. In the current study, the start of injection timing (SOI) was set to 180 before top dead center (BTDC) for both fuels.
Experimental setup and methods Experimental methods The study was conducted in a four-stroke, single-cylinder, DISI research engine developed by Orbital with a compression ratio of 14:1; its schematic diagram is shown in Fig. 1, and its specifications are listed in Table 1. The study was conducted in the Center for Automotive Research and Electric Mobility (CAREM), Universiti Teknologi PETRONAS. The experiments were conducted in accordance with the Society of Automotive Engineers (SAE) standards of Engine Power Test Code [17]. The engine was coupled to an eddy-current electric dynamometer to measure the brake torque and further to motor the engine at times of no combustion. At early fuel-injection times, the charge has enough time to mix with air before the onset of ignition. As a result, the shape of the piston head has less significance in the charge-mixing process. However, there is insufficient time for charge mixing in late-injection operations, resulting in an uneven air/fuel ratio in the cylinder. The shape of the piston head and the SOI at late injection influences the distribution of charge in the chamber [18]. In this work, a large piston bowl that creates fuel stratification in the chamber was selected. In this arrangement, the bowl position was away from the center aligned with the injector and spark plug position. As a result, the fuel is deflected back from the piston, creating a richer mixture near the spark plug. Such fuel stratification reduces the combustion instability and increases the mixture-distribution quality in the cylinder at a lean air/fuel ratio. The arrangement of the piston, injector and spark plug is depicted in Fig. 2. The injection system was a central DI in which the injector was placed with its axis aligned with the center of the piston bowl and had a 6-mm offset to the spark plug. An air injector was used for the study without any modification. The fuel injection was kept at 18 bar.
Fuels and metering In this work, a premixed syngas was used to overcome limitations in the real syngas produced from gasification. The imitated, premixed gas was prepared in a factory by MoxLinde Gaseswith certification of 50% H2 and 50% CO. This syngas was taken to be representative of the family of various ratios of H2/ CO syngas. It was supplied in a gas bottle at a pressure of 160 bar. CNG was obtained from a local NGV station
17886
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Fig. 1 e Experimental setup schematics.
pressurized in a bottle at 200 bar. Table 2 shows the properties of the syngas and CNG used in this study. The fuel rail of the CNG and syngas was maintained at atmospheric temperature. A coriolis flowmeter was used to measure the mass flow rate of CNG with a sensitivity of 0.001 g/s and a flow accuracy of ±0.05% of the flow rate. The CNG fueling system was equipped with subsequent pressure regulators and a compressed-pressure line that maintains the pressure along the fuel rail to obtain a small variability in the fuel-injection rate. Additionally, the injector was calibrated to measure the CNG mass flow rate throughout the injection duration. An A560 series Concoa 150-mm flowmeter calibrated for syngas was used to measure and control the flow rate of the syngas. A specially designed injector holder was used to support two fuel lines, thereby simplifying the fuel switchover at the time of engine warm-up. The injector holder was first used elsewhere for the in situ mixing of hydrogen and CNG [19]. The engine warm-up and stabilization were performed with CNG due to higher price of the factory-produced imitated syngas. However, the price does not reflect the price of the real syngas produced from gasification. After the engine attained stable operation, fuel switchover was performed and the engine was operated with syngas for more than five minutes before taking the first measurement, to avoid fuel contamination.
Instrumentation The ambient conditions, such as the room temperature, pressure, and relative humidity near the air induction of the engine, were recorded using an Oragon Scientific weather
station Higbo 433 MHz. The engine-inlet manifold air flow rate was measured by Bosch MAF with accuracy 3% fitted to the engine-inlet manifold. The temperature of the coolant, engine oil, induction air and exhaust gas were recorded from the engine-control display measured by thermocouple attachments. The combustion analysis of the engine was performed with the help of pressure readings from the engine cylinder in this experiment. A Kistler piezoelectric pressure transducer with a sensitivity of z25 pC/bar was installed in the cylinder to record the in-cylinder pressure. The transducer accuracy is severely affected by the decompression of the crystal that resulted from the variation of the temperature in the casing; it was equipped with a water-cooling passage to ensure the correct pressure reading. The transducer generated an electric charge proportional to the in-cylinder pressure. The data capturing was synchronized with the crank-angle encoder
Table 1 e Engine specification. Engine properties
Specification
Number of cylinders Displacement volume Cylinder bore Cylinder stroke Compression ratio Number of valves Inlet valve open (IVO) Inlet valve closed (IVC) Exhaust valve open (EVO) Exhaust valve closed (EVC)
1 399.25 cm3 76.0 mm 88.0 mm 14 (Geometric) 4 BTDC 12 ABDC 48 BBDC 45 ATDC 10
17887
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Fig. 2 e Arrangement of large bowl piston, injector and spark plug for stratified mixture formation.
that determined the angular position for each pressure reading. The LabVIEW program displayed and stored the pressure against the engine crank angle. The concentration of the exhaust-gas species was measured by a GASMET CX series gas analyzer working on the Fourier transform infrared (FTIR) principle. It was capable of measuring up to 50 species of gases. An ENOTEC OXITEC 5000 oxygen analyzer that uses a ZrO2 oxygen in situ sensor was used in this experiment. Its output was a voltage proportional to the oxygen concentration in the exhaust gas. This equipment was also used to meter l of the air/fuel mixture of the charge.
Combustion, performance and emissions analysis A comparison of the combustion parameters of the syngas and CNG in the DISI engine was performed by monitoring the in-cylinder pressure data with the help of the pressure sensor. A pressure reading of up to 100 power cycles was recorded for a single run. The mean effective pressure (MEP) was used in the selection of the most representative cycle. The indicated mean effective pressure (IMEP) was calculated from the pressure data collected by the pressure sensor in a range of crank-angle rotation of the engine cycle based on Eq. (1) as follows: n2 Dq X dVðiÞ ; pðiÞ IMEP ¼ Vd i¼n dq
(1)
1
where p(i) is the in-cylinder pressure at the crank angle i, V(i) is the cylinder volume at the crank angle i, and n1 and n2 represent two successive BDC crank-angle positions [20]. The IMEPavg was calculated from the reading of up to 100 power cycles based on Eq. (2) as follows: IMEPavg ¼
n 1X IMEPi ; n i
(2)
where n is the number of power cycles recorded. The analysis of IMEP is sensitive to the crank-angle phasing errors and the thermal shocking of the pressure transducer. A cycle
representative of most cycles was selected by comparing the IMEP of each cycle with the average. Other parameters, such as combustion pressure and the mass fraction burn (MFB) were calculated from the selected cycle. In any crank-angle interval, the actual pressure change DP ðPjþ1 Pj Þ can be assumed to be made up a pressure rise due to combustion (DPC) and a pressure change due to a volume change (DPV) according to Rassweiler and Withrow as cited in Wiseman [21]. The pressure change due to the combustion at the end of the interval Dq is modeled as: Vj n ; DPc ðj þ 1Þ ¼ Pjþ1 Pj Vjþ1
(3)
where n is the polytropic index, quantified to be 1.33 for the current engine by fitting a regression curve of the pressure-
Table 2 e Properties of syngas and its comparison with CNG. Properties Composition, weight %
Syngas Carbon Hydrogen Oxygen Nitrogen
Molecular weight (g/mol) Density at 0 C and 1 atm (kg/m3) Specific gravity at 0 C and 1 atm Stoichiometric air-fuel ratio Molar basis Mass basis Stoichiometric volume occupation in cylinder, % Lower calorific value MJ/Nm3 MJ/kg Mixture Stoichiometric mixture aspirated energy density (MJ/Nm3) Air aspirated Flammability limit, % volume Lower of fuel in air Higher Laminar flame velocity (cm/s) Adiabatic flame temperature, K Auto ignition temperature, K
CNG
40.0 6.67 53.33 0.0 15.0 0.67 0.52 2.38 4.58 29.6
75.0 25.0 0.0 0.0 16.04 0.75 0.58 9.70 17.2 9.35
11.65 17.54 3.3
38.0 47.13 2.9
4.45 6.06 74.2 180.0 2385 873e923
3.6 5.3 15.0 30.0 2233 755e905
17888
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
volume reading during the motoring cycle. Rassweiler and Withrow also assumed that the combustion-pressure rise is proportional to the mass of the charge burned [22]. The MFB at the end of interval Dq is given by: Pj 0 DPC : Xq ðjÞ ¼ PEOC DPC 0
(4)
The comparison of the performance parameters between the syngas and CNG was addressed using operating parameters such as the brake power, brake mean effective pressure (BMEP), BSFC and brake thermal efficiency (BTE). These were calculated from the fuel properties, fuel rate and torque reading from engine-control unit (ECU) based on the equations presented in Heywood [22]. One of the reasons for utilizing syngas as a fuel in the DISI engine is because of the higher exhaust emissions from fossilbased hydrocarbon fuels. With the introduction of increasing emission regulations worldwide, it has become increasingly difficult to attain the specified threshold while utilizing the existing fuels in an ICE. Among the various emissions, there is particular interest in primary pollutants, such as NOx, CO, and THC. The concentrations of gaseous emissions were measured in parts per million or percent by volume. A comparison of the emission levels in the current study was made by brake specific emissions (g/kWh). The procedure followed for the calculation of brake specific emissions was explained in detail elsewhere [23].
Results and discussions In this article, the combustion, performance, and emissions of syngas (50% H2/50% CO) were compared with those of CNG in a DISI CNG engine. This fuel is representative of the family of H2/CO syngases with various ratios. Pressure data from the pressure transducer, data from the ECU, the fuel rate, the atmospheric conditions, the emissions data from the gas analyzer, and the excess air ratio (l) from the oxygen analyzer were input for the analysis. The SOI was set to 180 BTDC, and the ignition advance was set to MBT. The engine speeds 1500 and 2100 rev/min were taken as representative of the operation speed range. The l value was varied from stoichiometric to the maximum lean conditions for CNG. The ECU preloaded software of the engine was optimized based on the performance mapping of CNG. CNG is
a high calorific gaseous fuel and thus requires a narrow injector pulse width. As a result, the injector pulse width in the current engine setup is limited to a maximum of 9 ms by the ECU software. Thus, the syngas operation was restricted to a minimum l of 1.5 due to the limit on the pulse width of the injector. The results and discussions of this study are presented in three subsections.
Combustion Fig. 3 shows the variation of l with the IMEP of the syngas and CNG at 1500 and 2100 rev/min. Syngas was shown to operate under a wider operation load and l than CNG at both engine speeds, extending the lower load end limit of CNG. However, the upper load of syngas was limited by the pulse width of the injector. Both CNG and syngas experienced the same limit on the sensitivity of IMEP with l. A comparison of the injection duration of the two fuels at SOI ¼ 180 BTDC and various engine speeds with their respective maximum IMEP is presented in Fig. 4. A longer duration was observed at the higher engine speed with syngas, and the shortest at a lower engine speed with CNG. This was attributed to the difference in the calorific value of the two fuels (11.65 MJ/Nm3 for syngas and 38.0 MJ/Nm3 for CNG). At the SOI under investigation, the fuel injection of CNG was almost completed before the inlet valve close (IVC). However, the majority of the fuel was injected after the IVC with syngas. As a result, the ignition onset of the syngas was delayed compared to that of CNG for better mixing of the fuel and air. The fast flame-propagation nature of the H2 species in syngas resulted in a short delay period, also leading to the retardation of the ignition onset of the syngas. Fig. 5 presents the coefficient of variation of IMEP (COV) versus IMEP at engine speeds 1500 and 2100 rev/min for the two fuels. Syngas was shown to operate at a wide operation load with a lower COV. A similar observation was reported with other types of syngases with different compositions of the gaseous species in carbureted SI engines by various authors [24,25]. This was attributed to the combined effect of the higher flame-propagation nature of H2 and the flamesuppressive nature of CO; the latter played a combustioncontrolling role. The wider flammability limit of syngas (6.06e74.2%, compared to 5.3e16% for CNG) also contributed to the reduced COV. The COV was shown to increase with a decrease in the load at all engine speeds. The higher COV at
Fig. 3 e Excess air ratios versus IMEP for syngas and CNG.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
17889
Fig. 4 e Comparison of injection duration of fuels for different speeds at maximum IMEP.
lower end loads was attributed to misfiring due to the lower mixture calorific value at a lean charge. The COV of CNG was shown to be more sensitive to the IMEP for all engine speeds. The variation in the pressure, heat release rate and MFB with the crank angle is presented in Fig. 6 for the two types of fuels at an engine speed of 1500 rev/min and an IMEP ¼ 7.9 bar. The syngas was shown to have the highest peak cylinder pressure at 61.44 bar and 14 after top dead center (ATDC), the highest peak heat release rate with 0.034 kJ/ CA at 9 ATDC and the fastest MFB curve with an overall combustion duration (0e90% MFB) of 29.5 crank angle ( CA). On the other hand, CNG was shown to have a peak pressure of 49.93 bar at 12.5 ATDC, a peak heat release rate of 0.019 at kJ/ CA at 13.5 ATDC and MFB with an overall combustion duration of 75.5 CA. This was due to the higher laminar flame velocity of the H2 content of the syngas, prompting the fast flame propagation of the syngas combustion. Similarly, the variation in the pressure, heat release rate and MFB with the crank angle is presented in Fig. 7 for the two fuels at an engine speed of 2100 rev/min and an IMEP ¼ 6.4 bar. The syngas was shown to have a maximal peak cylinder pressure and maximal peak heat release rate of 45.46 bar at 11 ATDC and 0.0164 kJ/ CA at 9.5 ATDC, respectively. However, the peak pressure of 39.13 bar at 10 ATDC, peak heat release rate of 0.015 at kJ/ CA at 13.5 ATDC and MFB with an overall combustion duration of 62.5 CA was exhibited with CNG.
Table 3 shows a summary of the combustion characteristics of the two fuels at speeds of 1500 and 2100 rev/min. The effect of the engine speed on the combustion of the two fuels could not be compared because the operation loads of the two speeds were different. The ignition advance for MBT was retarded with syngas compared to CNG. This was because of the longer injection duration with syngas compared to that with CNG, as shown in Fig. 4. The fast flame-propagation nature of the H2 species in the syngas resulted in a shorter ignition delay and also contributed in the retardation of the ignition onset of the syngas.
Performance characteristics Fig. 8 shows the variation in the maximum brake torque with the engine speed for syngas and CNG. For CNG operating at stoichiometric conditions, the minimum l of the syngas was restricted to 1.5 due to the upper limit of the injector pulse width. The brake torque was shown to increase with engine speed for both fuels. A reduction of 14.2e19.6% in the brake torque was observed with syngas compared to CNG. Reductions of 23% at a compression ratio of 8:1 and 18% at a compression ratio of 11:1 were reported for gas with a similar content, as compared to CNG at 1500 rev/min by Mustafi et al. [3]. The l in the previous study was set to near stoichiometric (1.02e1.08) for CNG and approximately 1.23 for power gas. The
Fig. 5 e COV of IMEP versus IMEP for syngas and CNG.
17890
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Fig. 6 e Combustion characteristics versus crank angle degree for syngas and CNG at 1500 rev/min and IMEP ¼ 7.9 bar a) pressure and heat release rate b) MFB.
small improvement in the current study was due to the directinjection fueling system. However, the attributes stated in the previous study also contributed to the brake torque drop with syngas in the current study.
The variation in the thermal efficiency with BMEP for the syngas and CNG are presented at speeds of 1500 and 2100 rev/ min in Fig. 9. At 1500 rev/min, a maximum brake thermal efficiency of 28.1% was reported with CNG at BMEP ¼ 2.89 bar. For
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
17891
Fig. 7 e Combustion characteristics versus crank angle degree for three different fuels at 2100 rev/min and IMEP ¼ 6.4 bar a) pressure and heat release rate b) MFB.
17892
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Table 3 e Excess air ratio, ignition advance and combustion duration of different fuels at 1500 rev/min at IMEP ¼ 7.9 bar and 2100 rev/min at 6.4 bar. Speed Fuel l Ignition advance for MBT (BTDC) Flame development stage duration ( CA) Rapid burn stage duration ( CA) Overall combustion duration ( CA)
1500 rev/min
2100 rev/min
CNG 1.2 29.0
Syngas 1.5 10.5
CNG 1.4 27.0
Syngas 1.97 16.4
28.0
11.25
27.0
15.0
47.5
18.25
35.5
26.5
75.5
29.5
62.5
41.5
partial direct injection, for which part of the fuel is injected before the inlet valve closes. As a result, the medium calorific value syngas occupies the volume, displacing the air. However, with CNG, it has higher energy density, and less fuel is injected than for the syngas. The amount of air displaced by CNG was lower than the amount that could be displaced by syngas. The variation in the BSFC with BMEP for both the syngas and CNG are presented in Fig. 10 for engine speeds of 1500 and 2100 rev/min. The BSFC of CNG was shown to be less sensitive to BMEP, ranging from 240 to 286 g/kWh for all loads and speeds. A similar observation was reported in the literature [3]. The BSFC of the syngas was highly sensitive to both the engine load and speed. At 1500 rev/min, a minimum BSFC of 874 g/kWhr for syngas was observed at BMEP ¼ 2.48 bar, which is almost threefold that of CNG. Mustafi et al. [3] reported a 3.4 times increase in the BSFC of power gas with CNG at the same speed with a carbureted engine. However, their study was at a compression ratio of 8:1, and the CNG was operated under stoichiometric conditions, whereas the power gas was operated under lean conditions. At 2100 rev/min, a minimum BSFC of 702.5 g/kWhr was observed for syngas at the highest load.
Emission characteristics
Fig. 8 e Maximum brake torque versus engine speed for syngas and CNG.
syngas, a maximum brake thermal efficiency of 23.4% was observed at BMEP ¼ 2.48 bar. There was a reduction of 16.7% in the brake thermal efficiency with syngas at this speed. However, the efficiency of the syngas was improved. A maximum brake thermal efficiency of 31.66% was reported with CNG at BMEP ¼ 3.62 bar at an engine speed of 2100 rev/min. Similarly, the maximum brake thermal efficiency for syngas was reported as 29.23% at BMEP ¼ 3.34 bar. The difference in the maximum brake thermal efficiency of the two fuels was reduced to 7.66% at a speed of 2100 rev/min. The reason for the lower brake thermal efficiency of syngas was due to the volumetric efficiency penalty. At SOI ¼ 180 BTDC, the injection is a
The emissions of CO, NOx and THC versus BMEP are presented at the engine speeds 1500 and 2100 rev/min for syngas and CNG in Figs. 11e13. Because the two fuels have different energy densities, the emissions were computed as brake specific emissions in g/kWh. The CO emission of syngas was observed to significantly increase with a decrease in BMEP at all engine speeds, especially above BMEP ¼ 1.6 bar, as shown in Fig. 11. This was due to the weak mixture calorific value at higher l values, leading to weaker heat release from the combustion. This, in turn, led to the escape of the unburned CO component of the fuel. In contrast to the case of syngas, CNG was observed to have higher CO emission at higher loads (excess air ratio close to unity). The latter was attributed to an incomplete combustion due to a richer mixture at and near stoichiometric conditions. With decreasing load (l > 1.1), the brake specific emission remained at a very low level. There was no such occurrence with syngas because the combustion occurred under lean conditions (l 1.5). A similar shape of the CO-emission curve was reported for CNG by Haung et al. [26]. The emission of NOx versus BMEP is presented in Fig. 12 at the engine speeds 1500 and 2100 rev/min for syngas and CNG. The brake specific emission of NOx was calculated by the
Fig. 9 e Brake thermal efficiency versus BMEP for syngas and CNG.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Fig. 10 e Brake specific fuel consumption versus BMEP for syngas and CNG.
Fig. 11 e Brake specific CO emission versus BMEP for syngas and CNG.
Fig. 12 e Brake specific NOx emissions versus BMEP for syngas and CNG.
Fig. 13 e Brake specific THC emissions versus BMEP for syngas and CNG.
17893
17894
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
summation of individually calculated brake specific emissions of nitrous oxide (N2O), nitric oxide (NO) and nitrogen dioxide (NO2). Syngas exhibited higher emission of NOx at higher loads compared to CNG. For syngas, the maximal NOx emission of 15.5 g/kWhr was reported at BMEP ¼ 3 bar. This was attributed to the rapid combustion and higher peak in-cylinder pressure of syngas reported in Figs. 6 and 7. The variation in the brake specific emission of THC with BMEP is presented in Fig. 13 at the engine speeds 1500 and 2100 rev/min for the two fuels. Similar to the case for NOx, the brake specific emissions of 28 species of hydrocarbons were calculated individually and then summed for the analysis of the brake specific emission of THC. The THC emission was insignificant at all engine loads and engine speeds for syngas. At all loads and all engine speeds, the THC emission of CNG was higher than that of syngas. This was attributed to the absence of heavy hydrocarbons in the syngas. Additionally, syngas is an oxygenated fuel because of its CO species. This might contribute to facilitating combustion. The sharp increase in the THC emission for CNG at both speeds at a lower BMEP was attributed to a weak mixture calorific value.
Conclusions The aim of this experimental investigation was to explore the combustion, performance and emissions of syngas (H2/CO) in a DI SI engine. CNG was taken as a baseline fuel for comparison. Syngas was observed to operate at a wider engine load and l range than CNG, and so the maximal load range could potentially be increased with modification in the injector pulse width. Syngas was shown to have a higher peak cylinder pressure, higher heat release rate and faster combustion duration than CNG at all operation speeds and loads. This was attributed to the fast flame propagation of the hydrogen species in the syngas fuel. The ignition advance for the MBT of syngas was observed to become delayed near TDC. The coefficient of variation of the IMEP of syngas was also smaller than that of CNG at all operation loads and speeds. The brake thermal efficiency of the syngas was reported to be lower than that of CNG, especially at lower engine loads. With increasing speed, the difference was decreased. Similarly, the BSFC of syngas was higher than that of CNG. This was mainly attributed to the low calorific value nature of syngas. Reduced CO emissions were observed at higher loads, whereas they were increased at lower loads compared to CNG at all engine speeds. The NOx emission was higher than that of CNG at all load and speed conditions. The THC emission was also improved significantly with syngas at all engine loads and speeds. Therefore, this fuel proves to be a good substitute for gaseous fossil fuels in DISI engines without requiring major alterations to the engine.
Acknowledgments This work was supported by the STIRF with a fund number: 17/10.11 and Center for Automotive Research and Electric Mobility (CAREM), Universiti Teknologi PETRONAS. The support from Educational Sponsorship Unit, PETRONAS Carigali
Sdn Bhd was massive too. The authors extend their acknowledgements.
references
[1] Sridhar G, Paul PJ, Mukunda HS. Biomass derived producer gas as a reciprocating engine fuel- an experimental analysis. Biomass Bioenergy 2001;21:61e72. ~ o B, Aguillo n J. Numerical simulation of syngas [2] Gamin combustion with a multi-spark ignition system in a diesel engine adapted to work at the otto cycle. Fuel 2010;89:581e91. [3] Mustafi NN, Miraglia YC, Raine RR, Bansal PK, Elder ST. Spark-ignition engine performance with ‘Powergas’ fuel (Mixture of CO/H2): a comparison with gasoline and natural gas. Fuel 2006;85:1605e12. [4] Papagiannakis RG, Rakopoulos CD, Hountalas DT, Giakoumis EG. Study of the performance and exhaust emissions of a spark-ignited engine operating on syngas fuel. Int J Altern Propuls 2007;1:190e215. [5] Shah A, Srinivasan R, To SDF, Columbus EP. Performance and emissions of a spark-ignited engine driven generator on biomass based syngas. Bioresour Technol 2010;101:4656e61. [6] Spaeth CT. Performance characteristics of a diesel fuel piloted syngas compression ignition engine. Master of Applied Science. Department of Mechanical and Materials Engineering, Queen's University; 2012. [7] Konigsson F. Advancing the limits of dual fuel combustion. Licentiate. Department of Machine Design, Royal Institute of Technology (KTH); 2012. [8] Ahrenfeldt J, Foged EV, Strand R, Henriksen UB. Development and test of a new concept for biomass producer gass engines. Roskilde: Risø National Laboratory for Sustainable Energy, Technical University of Denmark; 2010, ISBN 978-87-5503811-0. [9] Boehman AL, Corre OL. Combustion of syngas in internal combustion engines. Combust Sci Technol 2008;180:1193e206. [10] Bika AS. Synthesis gas use in internal combustion engines. PhD, Faculty of Graduate School. Minnesota: University of Minnesota; 2010. [11] Centeno F, Mahkamov K, Lora EES, Andrado RV. Theoretical and experimental investigations of a downdraft biomass gasifier- spark ignition engine power system. Renew Energy 2012;37:97e108. [12] Gonzalez FOC, Mahkamov K, Lora EES, Andrade RV, Jaen RL. Prediction by mathematical modeling of the behavior of an internal combustion engine to be fed with gas from biomass, in comparison to the same engine fueled with gasoline or methane. Renew Energy 2013;60:427e32. [13] Firmansyah. Optimization of injection parameters in compressed natural gas direct injection spark ignition engine. Master of Science. Tronoh: Department of Mechanical Engineering, Universiti Teknologi PETRONAS; 2007. [14] Mohammed SE, Baharom MB, Aziz ARA, Firmansyah. The effects of fuel-injection timing at medium injection pressure on the engine characteristics and emissions of a CNG-DI engine fueled by a small amount of hydrogen in CNG. Int J Hydrogen Energy 2011;36:11997e2006. [15] Mohammed SEL, Baharom MB, Aziz ARA. Analysis of engine characteristics and emissions fueled by in-situ mixing of small amount of hydrogen in CNG. Int J Hydrogen Energy 2011;36:4029e37. [16] Abdullah MA, Aziz ARA. An experimental investigation of hydrogen combustion in a direct injection spark ignition natural gas engine. SAE Paper 2009-01-1499.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
[17] Standard SISV. Engine power test code-spark ignition and compression ignition-gross power rating. In: SAE International Surface Vehicle Standard, editor. [18] Mohamad TI. Compressed natural gas direct injection (spark plug fuel injector). In: Potocnik P, editor. Natural gas. Intech Open; 2010. p. 289e306. [19] Mohammed SE. Analysis of engine characterstics and emissions fueled by in-situ mixing of small amount of hydrogen in CNG. MSc, Mechanical Engineering Department. Universiti Teknologi PETRONAS; 2011. [20] Brunt MF, Emtage AL. Evaluation of IMEP routines and analysis errors. SAE Pap 1996-0609. [21] Wiseman MW. Spark ignition engine combustion process analysis. PhD, Department of Mechanical. Materials and Manufacturing Engineering, University of Nottingham; 1990. [22] Heywood JB. Internal combustion engine fundamentals. New York: McGraw Hill International; 1988.
17895
[23] Abdullah MA. Engine performance and combustion characteristics of hydrogen in a direct injection engine. Master of Science, Mechanical Engineering. Universiti Teknologi PETRONAS; 2009. [24] Ando Y, Yoshikawa K, Beck M, Endo H. Research and development of a low-BTU gas-driven engine for waste gasification and power generation. Energy 2005;30:2206e18. [25] Ahrenfeldt J. Characterization of biomass producer gas as fuel for stationary gas engines in combined heat and power production. Degree for Doctor of Philosophy. Department of Mechanical Engineering, Technical University of Denmark; 2007. [26] Huang Z, Liu B, Zeng K, Huang Y, Jiang D, Wang X, et al. Experimental study on engine performance and emissions for an engine fueled with natural gas-hydrogen mixtures. Energy Fuels 2006;20:2131e6.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Syngas (H2/CO) in a spark-ignition direct-injection engine. Part 1: Combustion, performance and emissions comparison with CNG Ftwi Yohaness Hagos a,*, A. Rashid A. Aziz b,c, Shaharin A. Sulaiman c a
Faculty of Mechanical Engineering, Universiti Malaysia Pahang, 26600 Pekan, Pahang, Malaysia Centre for Automotive Research and Electric Mobility (CAREM), Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, Malaysia c Department of Mechanical Engineering, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 31750 Tronoh, Perak, Malaysia b
article info
abstract
Article history:
The combustion, performance, and emissions of syngas (H2/CO) in a four-stroke, direct-
Received 31 May 2014
injection, spark-ignition engine were experimentally investigated. The engine was oper-
Accepted 28 August 2014
ated at various speeds, ranging from 1500 to 2400 rev/min, with the throttle being held in
Available online 23 September 2014
the wide-open position. The start of fuel injection was fixed at 180 before the top dead center, and the ignition advance was set at the maximal brake torque. The air/fuel ratio
Keywords:
was varied from the technically possible lowest excess air ratio (l) to lean operation limits.
Syngas
The results indicated that a wider air/fuel operating ratio is possible with syngas with a
Direct-injection
very low coefficient of variation. The syngas produced a higher in-cylinder peak pressure
Spark-ignition
and heat-release rate peak and faster combustion than for CNG. However, CNG produced a
Combustion
higher brake thermal efficiency (BTE) and lower brake specific fuel consumption (BSFC).
Performance
The BTE and BSFC of the syngas were on par to those of CNG at higher speeds. For the
Emission
syngas, the total hydrocarbon emission was negligible at all load conditions, and the carbon monoxide emission was negligible at higher loads and increased under lower load conditions. However, the emission of nitrogen oxides was higher at higher loads with syngas. Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Introduction There is growing concern regarding the use of fossil-derived fuels in the transportation sector, primarily due to their emissions and energy-sustainability issues. Furthermore,
advancements in conversion technologies and the abundant availability of solid fuels, along with advancements in the technology of gas engines and their fueling systems, have caused a revival of interest in syngas. Additionally, syngas is a potentially carbon-neutral and renewable fuel. Over the past 30 years, there have been many efforts to bring back the
* Corresponding author. E-mail addresses: [email protected], [email protected], [email protected] (F.Y. Hagos), [email protected] (A.R.A. Aziz), [email protected] (S.A. Sulaiman). http://dx.doi.org/10.1016/j.ijhydene.2014.08.141 0360-3199/Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
technology of syngas in internal combustion engines (ICEs), both in spark-ignition (SI) [1e5] and compression-ignition (CI) engines [6e9]. However, the application of syngas in SI engines was limited to carbureted and port-injection engines, and such fueling systems are now obsolete due to their poor performance. On the other hand, the use of syngas in CI engines was restricted to dual fueling with diesel as a pilot fuel to assist in the ignition. In this arrangement, syngas is not a standalone fuel, and it cannot completely replace diesel. There have been several works reported in the literature related to syngas fueling in carbureted and port-injection SI engines. Bika [10] studied various ratios of H2/CO syngas in a port-injection SI engine for their combustion characteristics and knock limit. The fuel was injected 4 cm from the intake port. The fuels investigated included pure H2, 75% H2/25% CO and 50% H2/50% CO. The study was limited to an equivalence ratio of (f) ¼ 0.6e0.8 and compression ratios of 6:1 to 10:1. It was reported that an increase in the content of CO increased the knock limit of the syngas. The study also indicated that the increase in the CO content advanced the ignition timing of the maximum brake torque (MBT). The overall conclusion of the study was that there is an increase in the combustion duration with an increase in the CO content. A maximal indicated thermal efficiency of 32% was reported with 50% H2/ 50% CO at f ¼ 0.6 and a compression ratio of 10:1 [10]. The study was limited to combustion only. The performance and emissions of these fuels were not investigated. However, the study highlighted the extension of the knock limit with an increase in the CO content in syngas. Mustafi et al. [3] investigated the performance and emission of power gas in a variable-compression-ratio SI engine. The molar ratios of the fuels investigated were 0.52, 0.44 and 0.04 for CO, H2 and N2, respectively. The lower heating value and stoichiometric air/fuel ratio were reported to be 15.3 MJ/kg and 4.2, respectively. The power output of the power gas, gasoline and compressed natural gas (CNG) were compared at a constant speed of 1500 rev/min. It was reported that the brake torque of the power gas was 30% and 23% lower than those of gasoline and CNG, respectively. The power-gas consumed brake specific fuel consumption (BSFC) () 2.7 and 3.4 times in kg/kW-h more than those of gasoline and CNG, respectively. Reduced emissions of total hydrocarbons (THCs) and CO were reported with power-gas compared to results with both gasoline and CNG. However, the CO2 and NOx emissions were higher [3]. To the knowledge of the authors, no study to date has investigated syngas in direct-injection (DI) SI engines. Neither its combustion behavior nor its performance and emission characteristics are known for the broad range of syngas types produced by gasification technologies. Moreover, there is no suitable predictive tool for the effect of the properties of syngas on the optimization of the combustion, performance and emissions, except for some recent trials [2,11,12]. Therefore, advancing the previous studies and targeting the latest engine technologies in the investigation of these fuels in ICEs is important. The results concerning the performance of DISI engines using other gaseous fuels (CNG, H2 and a mixture of both) have been presented by other authors [13e16]. Therefore, the objective of this study is to investigate the use of syngas in a DISI engine as a potential replacement for
17885
fossil-derived fuels. This article is part of an ongoing research project on fueling with syngas with a 50% H2:50% CO composition in a DISI engine. The scope of this article is limited to an experimental study of the combustion, performance and emissions characteristics of syngas in DISI engines, compared to baseline data for CNG. The effect of the injection timing on the combustion, performance and emissions of syngas in the same engine has also been investigated, and the results are presented in part 2, an accompanying article awaiting publication. In the current study, the start of injection timing (SOI) was set to 180 before top dead center (BTDC) for both fuels.
Experimental setup and methods Experimental methods The study was conducted in a four-stroke, single-cylinder, DISI research engine developed by Orbital with a compression ratio of 14:1; its schematic diagram is shown in Fig. 1, and its specifications are listed in Table 1. The study was conducted in the Center for Automotive Research and Electric Mobility (CAREM), Universiti Teknologi PETRONAS. The experiments were conducted in accordance with the Society of Automotive Engineers (SAE) standards of Engine Power Test Code [17]. The engine was coupled to an eddy-current electric dynamometer to measure the brake torque and further to motor the engine at times of no combustion. At early fuel-injection times, the charge has enough time to mix with air before the onset of ignition. As a result, the shape of the piston head has less significance in the charge-mixing process. However, there is insufficient time for charge mixing in late-injection operations, resulting in an uneven air/fuel ratio in the cylinder. The shape of the piston head and the SOI at late injection influences the distribution of charge in the chamber [18]. In this work, a large piston bowl that creates fuel stratification in the chamber was selected. In this arrangement, the bowl position was away from the center aligned with the injector and spark plug position. As a result, the fuel is deflected back from the piston, creating a richer mixture near the spark plug. Such fuel stratification reduces the combustion instability and increases the mixture-distribution quality in the cylinder at a lean air/fuel ratio. The arrangement of the piston, injector and spark plug is depicted in Fig. 2. The injection system was a central DI in which the injector was placed with its axis aligned with the center of the piston bowl and had a 6-mm offset to the spark plug. An air injector was used for the study without any modification. The fuel injection was kept at 18 bar.
Fuels and metering In this work, a premixed syngas was used to overcome limitations in the real syngas produced from gasification. The imitated, premixed gas was prepared in a factory by MoxLinde Gaseswith certification of 50% H2 and 50% CO. This syngas was taken to be representative of the family of various ratios of H2/ CO syngas. It was supplied in a gas bottle at a pressure of 160 bar. CNG was obtained from a local NGV station
17886
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Fig. 1 e Experimental setup schematics.
pressurized in a bottle at 200 bar. Table 2 shows the properties of the syngas and CNG used in this study. The fuel rail of the CNG and syngas was maintained at atmospheric temperature. A coriolis flowmeter was used to measure the mass flow rate of CNG with a sensitivity of 0.001 g/s and a flow accuracy of ±0.05% of the flow rate. The CNG fueling system was equipped with subsequent pressure regulators and a compressed-pressure line that maintains the pressure along the fuel rail to obtain a small variability in the fuel-injection rate. Additionally, the injector was calibrated to measure the CNG mass flow rate throughout the injection duration. An A560 series Concoa 150-mm flowmeter calibrated for syngas was used to measure and control the flow rate of the syngas. A specially designed injector holder was used to support two fuel lines, thereby simplifying the fuel switchover at the time of engine warm-up. The injector holder was first used elsewhere for the in situ mixing of hydrogen and CNG [19]. The engine warm-up and stabilization were performed with CNG due to higher price of the factory-produced imitated syngas. However, the price does not reflect the price of the real syngas produced from gasification. After the engine attained stable operation, fuel switchover was performed and the engine was operated with syngas for more than five minutes before taking the first measurement, to avoid fuel contamination.
Instrumentation The ambient conditions, such as the room temperature, pressure, and relative humidity near the air induction of the engine, were recorded using an Oragon Scientific weather
station Higbo 433 MHz. The engine-inlet manifold air flow rate was measured by Bosch MAF with accuracy 3% fitted to the engine-inlet manifold. The temperature of the coolant, engine oil, induction air and exhaust gas were recorded from the engine-control display measured by thermocouple attachments. The combustion analysis of the engine was performed with the help of pressure readings from the engine cylinder in this experiment. A Kistler piezoelectric pressure transducer with a sensitivity of z25 pC/bar was installed in the cylinder to record the in-cylinder pressure. The transducer accuracy is severely affected by the decompression of the crystal that resulted from the variation of the temperature in the casing; it was equipped with a water-cooling passage to ensure the correct pressure reading. The transducer generated an electric charge proportional to the in-cylinder pressure. The data capturing was synchronized with the crank-angle encoder
Table 1 e Engine specification. Engine properties
Specification
Number of cylinders Displacement volume Cylinder bore Cylinder stroke Compression ratio Number of valves Inlet valve open (IVO) Inlet valve closed (IVC) Exhaust valve open (EVO) Exhaust valve closed (EVC)
1 399.25 cm3 76.0 mm 88.0 mm 14 (Geometric) 4 BTDC 12 ABDC 48 BBDC 45 ATDC 10
17887
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Fig. 2 e Arrangement of large bowl piston, injector and spark plug for stratified mixture formation.
that determined the angular position for each pressure reading. The LabVIEW program displayed and stored the pressure against the engine crank angle. The concentration of the exhaust-gas species was measured by a GASMET CX series gas analyzer working on the Fourier transform infrared (FTIR) principle. It was capable of measuring up to 50 species of gases. An ENOTEC OXITEC 5000 oxygen analyzer that uses a ZrO2 oxygen in situ sensor was used in this experiment. Its output was a voltage proportional to the oxygen concentration in the exhaust gas. This equipment was also used to meter l of the air/fuel mixture of the charge.
Combustion, performance and emissions analysis A comparison of the combustion parameters of the syngas and CNG in the DISI engine was performed by monitoring the in-cylinder pressure data with the help of the pressure sensor. A pressure reading of up to 100 power cycles was recorded for a single run. The mean effective pressure (MEP) was used in the selection of the most representative cycle. The indicated mean effective pressure (IMEP) was calculated from the pressure data collected by the pressure sensor in a range of crank-angle rotation of the engine cycle based on Eq. (1) as follows: n2 Dq X dVðiÞ ; pðiÞ IMEP ¼ Vd i¼n dq
(1)
1
where p(i) is the in-cylinder pressure at the crank angle i, V(i) is the cylinder volume at the crank angle i, and n1 and n2 represent two successive BDC crank-angle positions [20]. The IMEPavg was calculated from the reading of up to 100 power cycles based on Eq. (2) as follows: IMEPavg ¼
n 1X IMEPi ; n i
(2)
where n is the number of power cycles recorded. The analysis of IMEP is sensitive to the crank-angle phasing errors and the thermal shocking of the pressure transducer. A cycle
representative of most cycles was selected by comparing the IMEP of each cycle with the average. Other parameters, such as combustion pressure and the mass fraction burn (MFB) were calculated from the selected cycle. In any crank-angle interval, the actual pressure change DP ðPjþ1 Pj Þ can be assumed to be made up a pressure rise due to combustion (DPC) and a pressure change due to a volume change (DPV) according to Rassweiler and Withrow as cited in Wiseman [21]. The pressure change due to the combustion at the end of the interval Dq is modeled as: Vj n ; DPc ðj þ 1Þ ¼ Pjþ1 Pj Vjþ1
(3)
where n is the polytropic index, quantified to be 1.33 for the current engine by fitting a regression curve of the pressure-
Table 2 e Properties of syngas and its comparison with CNG. Properties Composition, weight %
Syngas Carbon Hydrogen Oxygen Nitrogen
Molecular weight (g/mol) Density at 0 C and 1 atm (kg/m3) Specific gravity at 0 C and 1 atm Stoichiometric air-fuel ratio Molar basis Mass basis Stoichiometric volume occupation in cylinder, % Lower calorific value MJ/Nm3 MJ/kg Mixture Stoichiometric mixture aspirated energy density (MJ/Nm3) Air aspirated Flammability limit, % volume Lower of fuel in air Higher Laminar flame velocity (cm/s) Adiabatic flame temperature, K Auto ignition temperature, K
CNG
40.0 6.67 53.33 0.0 15.0 0.67 0.52 2.38 4.58 29.6
75.0 25.0 0.0 0.0 16.04 0.75 0.58 9.70 17.2 9.35
11.65 17.54 3.3
38.0 47.13 2.9
4.45 6.06 74.2 180.0 2385 873e923
3.6 5.3 15.0 30.0 2233 755e905
17888
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
volume reading during the motoring cycle. Rassweiler and Withrow also assumed that the combustion-pressure rise is proportional to the mass of the charge burned [22]. The MFB at the end of interval Dq is given by: Pj 0 DPC : Xq ðjÞ ¼ PEOC DPC 0
(4)
The comparison of the performance parameters between the syngas and CNG was addressed using operating parameters such as the brake power, brake mean effective pressure (BMEP), BSFC and brake thermal efficiency (BTE). These were calculated from the fuel properties, fuel rate and torque reading from engine-control unit (ECU) based on the equations presented in Heywood [22]. One of the reasons for utilizing syngas as a fuel in the DISI engine is because of the higher exhaust emissions from fossilbased hydrocarbon fuels. With the introduction of increasing emission regulations worldwide, it has become increasingly difficult to attain the specified threshold while utilizing the existing fuels in an ICE. Among the various emissions, there is particular interest in primary pollutants, such as NOx, CO, and THC. The concentrations of gaseous emissions were measured in parts per million or percent by volume. A comparison of the emission levels in the current study was made by brake specific emissions (g/kWh). The procedure followed for the calculation of brake specific emissions was explained in detail elsewhere [23].
Results and discussions In this article, the combustion, performance, and emissions of syngas (50% H2/50% CO) were compared with those of CNG in a DISI CNG engine. This fuel is representative of the family of H2/CO syngases with various ratios. Pressure data from the pressure transducer, data from the ECU, the fuel rate, the atmospheric conditions, the emissions data from the gas analyzer, and the excess air ratio (l) from the oxygen analyzer were input for the analysis. The SOI was set to 180 BTDC, and the ignition advance was set to MBT. The engine speeds 1500 and 2100 rev/min were taken as representative of the operation speed range. The l value was varied from stoichiometric to the maximum lean conditions for CNG. The ECU preloaded software of the engine was optimized based on the performance mapping of CNG. CNG is
a high calorific gaseous fuel and thus requires a narrow injector pulse width. As a result, the injector pulse width in the current engine setup is limited to a maximum of 9 ms by the ECU software. Thus, the syngas operation was restricted to a minimum l of 1.5 due to the limit on the pulse width of the injector. The results and discussions of this study are presented in three subsections.
Combustion Fig. 3 shows the variation of l with the IMEP of the syngas and CNG at 1500 and 2100 rev/min. Syngas was shown to operate under a wider operation load and l than CNG at both engine speeds, extending the lower load end limit of CNG. However, the upper load of syngas was limited by the pulse width of the injector. Both CNG and syngas experienced the same limit on the sensitivity of IMEP with l. A comparison of the injection duration of the two fuels at SOI ¼ 180 BTDC and various engine speeds with their respective maximum IMEP is presented in Fig. 4. A longer duration was observed at the higher engine speed with syngas, and the shortest at a lower engine speed with CNG. This was attributed to the difference in the calorific value of the two fuels (11.65 MJ/Nm3 for syngas and 38.0 MJ/Nm3 for CNG). At the SOI under investigation, the fuel injection of CNG was almost completed before the inlet valve close (IVC). However, the majority of the fuel was injected after the IVC with syngas. As a result, the ignition onset of the syngas was delayed compared to that of CNG for better mixing of the fuel and air. The fast flame-propagation nature of the H2 species in syngas resulted in a short delay period, also leading to the retardation of the ignition onset of the syngas. Fig. 5 presents the coefficient of variation of IMEP (COV) versus IMEP at engine speeds 1500 and 2100 rev/min for the two fuels. Syngas was shown to operate at a wide operation load with a lower COV. A similar observation was reported with other types of syngases with different compositions of the gaseous species in carbureted SI engines by various authors [24,25]. This was attributed to the combined effect of the higher flame-propagation nature of H2 and the flamesuppressive nature of CO; the latter played a combustioncontrolling role. The wider flammability limit of syngas (6.06e74.2%, compared to 5.3e16% for CNG) also contributed to the reduced COV. The COV was shown to increase with a decrease in the load at all engine speeds. The higher COV at
Fig. 3 e Excess air ratios versus IMEP for syngas and CNG.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
17889
Fig. 4 e Comparison of injection duration of fuels for different speeds at maximum IMEP.
lower end loads was attributed to misfiring due to the lower mixture calorific value at a lean charge. The COV of CNG was shown to be more sensitive to the IMEP for all engine speeds. The variation in the pressure, heat release rate and MFB with the crank angle is presented in Fig. 6 for the two types of fuels at an engine speed of 1500 rev/min and an IMEP ¼ 7.9 bar. The syngas was shown to have the highest peak cylinder pressure at 61.44 bar and 14 after top dead center (ATDC), the highest peak heat release rate with 0.034 kJ/ CA at 9 ATDC and the fastest MFB curve with an overall combustion duration (0e90% MFB) of 29.5 crank angle ( CA). On the other hand, CNG was shown to have a peak pressure of 49.93 bar at 12.5 ATDC, a peak heat release rate of 0.019 at kJ/ CA at 13.5 ATDC and MFB with an overall combustion duration of 75.5 CA. This was due to the higher laminar flame velocity of the H2 content of the syngas, prompting the fast flame propagation of the syngas combustion. Similarly, the variation in the pressure, heat release rate and MFB with the crank angle is presented in Fig. 7 for the two fuels at an engine speed of 2100 rev/min and an IMEP ¼ 6.4 bar. The syngas was shown to have a maximal peak cylinder pressure and maximal peak heat release rate of 45.46 bar at 11 ATDC and 0.0164 kJ/ CA at 9.5 ATDC, respectively. However, the peak pressure of 39.13 bar at 10 ATDC, peak heat release rate of 0.015 at kJ/ CA at 13.5 ATDC and MFB with an overall combustion duration of 62.5 CA was exhibited with CNG.
Table 3 shows a summary of the combustion characteristics of the two fuels at speeds of 1500 and 2100 rev/min. The effect of the engine speed on the combustion of the two fuels could not be compared because the operation loads of the two speeds were different. The ignition advance for MBT was retarded with syngas compared to CNG. This was because of the longer injection duration with syngas compared to that with CNG, as shown in Fig. 4. The fast flame-propagation nature of the H2 species in the syngas resulted in a shorter ignition delay and also contributed in the retardation of the ignition onset of the syngas.
Performance characteristics Fig. 8 shows the variation in the maximum brake torque with the engine speed for syngas and CNG. For CNG operating at stoichiometric conditions, the minimum l of the syngas was restricted to 1.5 due to the upper limit of the injector pulse width. The brake torque was shown to increase with engine speed for both fuels. A reduction of 14.2e19.6% in the brake torque was observed with syngas compared to CNG. Reductions of 23% at a compression ratio of 8:1 and 18% at a compression ratio of 11:1 were reported for gas with a similar content, as compared to CNG at 1500 rev/min by Mustafi et al. [3]. The l in the previous study was set to near stoichiometric (1.02e1.08) for CNG and approximately 1.23 for power gas. The
Fig. 5 e COV of IMEP versus IMEP for syngas and CNG.
17890
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Fig. 6 e Combustion characteristics versus crank angle degree for syngas and CNG at 1500 rev/min and IMEP ¼ 7.9 bar a) pressure and heat release rate b) MFB.
small improvement in the current study was due to the directinjection fueling system. However, the attributes stated in the previous study also contributed to the brake torque drop with syngas in the current study.
The variation in the thermal efficiency with BMEP for the syngas and CNG are presented at speeds of 1500 and 2100 rev/ min in Fig. 9. At 1500 rev/min, a maximum brake thermal efficiency of 28.1% was reported with CNG at BMEP ¼ 2.89 bar. For
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
17891
Fig. 7 e Combustion characteristics versus crank angle degree for three different fuels at 2100 rev/min and IMEP ¼ 6.4 bar a) pressure and heat release rate b) MFB.
17892
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Table 3 e Excess air ratio, ignition advance and combustion duration of different fuels at 1500 rev/min at IMEP ¼ 7.9 bar and 2100 rev/min at 6.4 bar. Speed Fuel l Ignition advance for MBT (BTDC) Flame development stage duration ( CA) Rapid burn stage duration ( CA) Overall combustion duration ( CA)
1500 rev/min
2100 rev/min
CNG 1.2 29.0
Syngas 1.5 10.5
CNG 1.4 27.0
Syngas 1.97 16.4
28.0
11.25
27.0
15.0
47.5
18.25
35.5
26.5
75.5
29.5
62.5
41.5
partial direct injection, for which part of the fuel is injected before the inlet valve closes. As a result, the medium calorific value syngas occupies the volume, displacing the air. However, with CNG, it has higher energy density, and less fuel is injected than for the syngas. The amount of air displaced by CNG was lower than the amount that could be displaced by syngas. The variation in the BSFC with BMEP for both the syngas and CNG are presented in Fig. 10 for engine speeds of 1500 and 2100 rev/min. The BSFC of CNG was shown to be less sensitive to BMEP, ranging from 240 to 286 g/kWh for all loads and speeds. A similar observation was reported in the literature [3]. The BSFC of the syngas was highly sensitive to both the engine load and speed. At 1500 rev/min, a minimum BSFC of 874 g/kWhr for syngas was observed at BMEP ¼ 2.48 bar, which is almost threefold that of CNG. Mustafi et al. [3] reported a 3.4 times increase in the BSFC of power gas with CNG at the same speed with a carbureted engine. However, their study was at a compression ratio of 8:1, and the CNG was operated under stoichiometric conditions, whereas the power gas was operated under lean conditions. At 2100 rev/min, a minimum BSFC of 702.5 g/kWhr was observed for syngas at the highest load.
Emission characteristics
Fig. 8 e Maximum brake torque versus engine speed for syngas and CNG.
syngas, a maximum brake thermal efficiency of 23.4% was observed at BMEP ¼ 2.48 bar. There was a reduction of 16.7% in the brake thermal efficiency with syngas at this speed. However, the efficiency of the syngas was improved. A maximum brake thermal efficiency of 31.66% was reported with CNG at BMEP ¼ 3.62 bar at an engine speed of 2100 rev/min. Similarly, the maximum brake thermal efficiency for syngas was reported as 29.23% at BMEP ¼ 3.34 bar. The difference in the maximum brake thermal efficiency of the two fuels was reduced to 7.66% at a speed of 2100 rev/min. The reason for the lower brake thermal efficiency of syngas was due to the volumetric efficiency penalty. At SOI ¼ 180 BTDC, the injection is a
The emissions of CO, NOx and THC versus BMEP are presented at the engine speeds 1500 and 2100 rev/min for syngas and CNG in Figs. 11e13. Because the two fuels have different energy densities, the emissions were computed as brake specific emissions in g/kWh. The CO emission of syngas was observed to significantly increase with a decrease in BMEP at all engine speeds, especially above BMEP ¼ 1.6 bar, as shown in Fig. 11. This was due to the weak mixture calorific value at higher l values, leading to weaker heat release from the combustion. This, in turn, led to the escape of the unburned CO component of the fuel. In contrast to the case of syngas, CNG was observed to have higher CO emission at higher loads (excess air ratio close to unity). The latter was attributed to an incomplete combustion due to a richer mixture at and near stoichiometric conditions. With decreasing load (l > 1.1), the brake specific emission remained at a very low level. There was no such occurrence with syngas because the combustion occurred under lean conditions (l 1.5). A similar shape of the CO-emission curve was reported for CNG by Haung et al. [26]. The emission of NOx versus BMEP is presented in Fig. 12 at the engine speeds 1500 and 2100 rev/min for syngas and CNG. The brake specific emission of NOx was calculated by the
Fig. 9 e Brake thermal efficiency versus BMEP for syngas and CNG.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
Fig. 10 e Brake specific fuel consumption versus BMEP for syngas and CNG.
Fig. 11 e Brake specific CO emission versus BMEP for syngas and CNG.
Fig. 12 e Brake specific NOx emissions versus BMEP for syngas and CNG.
Fig. 13 e Brake specific THC emissions versus BMEP for syngas and CNG.
17893
17894
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
summation of individually calculated brake specific emissions of nitrous oxide (N2O), nitric oxide (NO) and nitrogen dioxide (NO2). Syngas exhibited higher emission of NOx at higher loads compared to CNG. For syngas, the maximal NOx emission of 15.5 g/kWhr was reported at BMEP ¼ 3 bar. This was attributed to the rapid combustion and higher peak in-cylinder pressure of syngas reported in Figs. 6 and 7. The variation in the brake specific emission of THC with BMEP is presented in Fig. 13 at the engine speeds 1500 and 2100 rev/min for the two fuels. Similar to the case for NOx, the brake specific emissions of 28 species of hydrocarbons were calculated individually and then summed for the analysis of the brake specific emission of THC. The THC emission was insignificant at all engine loads and engine speeds for syngas. At all loads and all engine speeds, the THC emission of CNG was higher than that of syngas. This was attributed to the absence of heavy hydrocarbons in the syngas. Additionally, syngas is an oxygenated fuel because of its CO species. This might contribute to facilitating combustion. The sharp increase in the THC emission for CNG at both speeds at a lower BMEP was attributed to a weak mixture calorific value.
Conclusions The aim of this experimental investigation was to explore the combustion, performance and emissions of syngas (H2/CO) in a DI SI engine. CNG was taken as a baseline fuel for comparison. Syngas was observed to operate at a wider engine load and l range than CNG, and so the maximal load range could potentially be increased with modification in the injector pulse width. Syngas was shown to have a higher peak cylinder pressure, higher heat release rate and faster combustion duration than CNG at all operation speeds and loads. This was attributed to the fast flame propagation of the hydrogen species in the syngas fuel. The ignition advance for the MBT of syngas was observed to become delayed near TDC. The coefficient of variation of the IMEP of syngas was also smaller than that of CNG at all operation loads and speeds. The brake thermal efficiency of the syngas was reported to be lower than that of CNG, especially at lower engine loads. With increasing speed, the difference was decreased. Similarly, the BSFC of syngas was higher than that of CNG. This was mainly attributed to the low calorific value nature of syngas. Reduced CO emissions were observed at higher loads, whereas they were increased at lower loads compared to CNG at all engine speeds. The NOx emission was higher than that of CNG at all load and speed conditions. The THC emission was also improved significantly with syngas at all engine loads and speeds. Therefore, this fuel proves to be a good substitute for gaseous fossil fuels in DISI engines without requiring major alterations to the engine.
Acknowledgments This work was supported by the STIRF with a fund number: 17/10.11 and Center for Automotive Research and Electric Mobility (CAREM), Universiti Teknologi PETRONAS. The support from Educational Sponsorship Unit, PETRONAS Carigali
Sdn Bhd was massive too. The authors extend their acknowledgements.
references
[1] Sridhar G, Paul PJ, Mukunda HS. Biomass derived producer gas as a reciprocating engine fuel- an experimental analysis. Biomass Bioenergy 2001;21:61e72. ~ o B, Aguillo n J. Numerical simulation of syngas [2] Gamin combustion with a multi-spark ignition system in a diesel engine adapted to work at the otto cycle. Fuel 2010;89:581e91. [3] Mustafi NN, Miraglia YC, Raine RR, Bansal PK, Elder ST. Spark-ignition engine performance with ‘Powergas’ fuel (Mixture of CO/H2): a comparison with gasoline and natural gas. Fuel 2006;85:1605e12. [4] Papagiannakis RG, Rakopoulos CD, Hountalas DT, Giakoumis EG. Study of the performance and exhaust emissions of a spark-ignited engine operating on syngas fuel. Int J Altern Propuls 2007;1:190e215. [5] Shah A, Srinivasan R, To SDF, Columbus EP. Performance and emissions of a spark-ignited engine driven generator on biomass based syngas. Bioresour Technol 2010;101:4656e61. [6] Spaeth CT. Performance characteristics of a diesel fuel piloted syngas compression ignition engine. Master of Applied Science. Department of Mechanical and Materials Engineering, Queen's University; 2012. [7] Konigsson F. Advancing the limits of dual fuel combustion. Licentiate. Department of Machine Design, Royal Institute of Technology (KTH); 2012. [8] Ahrenfeldt J, Foged EV, Strand R, Henriksen UB. Development and test of a new concept for biomass producer gass engines. Roskilde: Risø National Laboratory for Sustainable Energy, Technical University of Denmark; 2010, ISBN 978-87-5503811-0. [9] Boehman AL, Corre OL. Combustion of syngas in internal combustion engines. Combust Sci Technol 2008;180:1193e206. [10] Bika AS. Synthesis gas use in internal combustion engines. PhD, Faculty of Graduate School. Minnesota: University of Minnesota; 2010. [11] Centeno F, Mahkamov K, Lora EES, Andrado RV. Theoretical and experimental investigations of a downdraft biomass gasifier- spark ignition engine power system. Renew Energy 2012;37:97e108. [12] Gonzalez FOC, Mahkamov K, Lora EES, Andrade RV, Jaen RL. Prediction by mathematical modeling of the behavior of an internal combustion engine to be fed with gas from biomass, in comparison to the same engine fueled with gasoline or methane. Renew Energy 2013;60:427e32. [13] Firmansyah. Optimization of injection parameters in compressed natural gas direct injection spark ignition engine. Master of Science. Tronoh: Department of Mechanical Engineering, Universiti Teknologi PETRONAS; 2007. [14] Mohammed SE, Baharom MB, Aziz ARA, Firmansyah. The effects of fuel-injection timing at medium injection pressure on the engine characteristics and emissions of a CNG-DI engine fueled by a small amount of hydrogen in CNG. Int J Hydrogen Energy 2011;36:11997e2006. [15] Mohammed SEL, Baharom MB, Aziz ARA. Analysis of engine characteristics and emissions fueled by in-situ mixing of small amount of hydrogen in CNG. Int J Hydrogen Energy 2011;36:4029e37. [16] Abdullah MA, Aziz ARA. An experimental investigation of hydrogen combustion in a direct injection spark ignition natural gas engine. SAE Paper 2009-01-1499.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 8 8 4 e1 7 8 9 5
[17] Standard SISV. Engine power test code-spark ignition and compression ignition-gross power rating. In: SAE International Surface Vehicle Standard, editor. [18] Mohamad TI. Compressed natural gas direct injection (spark plug fuel injector). In: Potocnik P, editor. Natural gas. Intech Open; 2010. p. 289e306. [19] Mohammed SE. Analysis of engine characterstics and emissions fueled by in-situ mixing of small amount of hydrogen in CNG. MSc, Mechanical Engineering Department. Universiti Teknologi PETRONAS; 2011. [20] Brunt MF, Emtage AL. Evaluation of IMEP routines and analysis errors. SAE Pap 1996-0609. [21] Wiseman MW. Spark ignition engine combustion process analysis. PhD, Department of Mechanical. Materials and Manufacturing Engineering, University of Nottingham; 1990. [22] Heywood JB. Internal combustion engine fundamentals. New York: McGraw Hill International; 1988.
17895
[23] Abdullah MA. Engine performance and combustion characteristics of hydrogen in a direct injection engine. Master of Science, Mechanical Engineering. Universiti Teknologi PETRONAS; 2009. [24] Ando Y, Yoshikawa K, Beck M, Endo H. Research and development of a low-BTU gas-driven engine for waste gasification and power generation. Energy 2005;30:2206e18. [25] Ahrenfeldt J. Characterization of biomass producer gas as fuel for stationary gas engines in combined heat and power production. Degree for Doctor of Philosophy. Department of Mechanical Engineering, Technical University of Denmark; 2007. [26] Huang Z, Liu B, Zeng K, Huang Y, Jiang D, Wang X, et al. Experimental study on engine performance and emissions for an engine fueled with natural gas-hydrogen mixtures. Energy Fuels 2006;20:2131e6.