Advanced Combustion Operation in a Compression Ignition Engine

Advanced Combustion Operation in a Compression Ignition Engine. Gregory K. Lilik, José .... HCCI can be approximated in a CI engine by early fuel inj...
0 downloads 0 Views 272KB Size
Download PDF
Energy & Fuels 2009, 23, 143–150

143

Advanced Combustion Operation in a Compression Ignition Engine Gregory K. Lilik,† Jose´ Martı´n Herreros,‡ and Andre´ L. Boehman*,† The EMS Energy Institute, The PennsylVania State UniVersity, 405 Academic ActiVities Building, UniVersity Park, PennsylVania 16802, and Escuela Te´cnica Superior de Ingenieros Industriales, UniVersidad de Castilla La-Mancha, AVda. Camilo Jose´ Cela s/n, 13071 Ciudad Real, Spain ReceiVed July 14, 2008. ReVised Manuscript ReceiVed NoVember 8, 2008

In this study, advanced combustion operating modes were investigated on a DDC/VM Motori 2.5 L, fourcylinder, turbocharged, common rail, direct-injection light-duty diesel engine, with exhaust emission being the main focus. The engine was operated under a partially premixed charge compression ignition (PCCI) mode, referred to as high-efficiency clean combustion (HECC), in which NOx and PM emissions dramatically decreased while fuel economy was maintained. In comparison to the default baseline operation at the given speed and load, the HECC mode reduced brake-specific NOx emissions by 72.1%, reduced brake-specific NO emissions by 83.2%, reduced brake-specific NO2 emissions by 33.8%, increased brake-specific HC emissions by 73.9%, increased brake-specific CO emissions by 105.6%, increased brake-specific CO2 emissions by 55.9%, reduced brake-specific PM emissions by 80.7%, and reduced brake-specific fuel consumption by 3.6%. The particle size distribution in the vicinity of the HECC operating mode consists almost entirely of an organic aerosol, and the solid phase of the particles becomes vanishingly small.

Introduction Advanced combustion operating regimes or modes, such as homogeneous charge compression ignition (HCCI) and premixed charge compression ignition (PCCI), are currently of interest to reduce diesel emissions, specifically NOx and particulate matter (PM). HCCI and PCCI operations shift combustion toward an increased premixed combustion phase, resulting in a fuel-lean charge and lowered combustion temperature and, thus, resulting in engine operation away from incylinder conditions that favor NOx and PM formation. In this study, a DDC/VM Motori 2.5 L engine was operated on a particular PCCI mode, referred to as high-efficiency clean combustion (HECC) developed by Wagner, Sluder, and coworkers at Oak Ridge National Laboratory.1-6 This study expands upon the previous work on HECC by examining the PM size distribution during HECC operation. * To whom correspondence should be addressed. Telephone: 814-8657839. Fax: 814-863-8892. E-mail: [email protected]. † The Pennsylvania State University. ‡ Universidad de Castilla La-Mancha. (1) Sluder, C. S.; Storey, J. M. E.; Lewis, S. A.; Lewis, L. A. Low temperature urea decomposition and SCR performance. SAE Tech. Pap. 2005-01-1858, 2005. (2) Sluder, C. S.; Storey, J. M. E.; Lewis, S. A.; Wagner, R. M. A thermal conductivity approach for measuring hydrogen in engine exhaust. SAE Tech. Pap. 2004-01-2908, 2004. (3) Sluder, C. S.; Wagner, R. M. An estimate of diesel high-efficiency clean combustion impacts on FTP-75 aftertreatment requirements. SAE Tech. Pap. 2006-01-3311, 2006. (4) Sluder, C. S.; Wagner, R. M.; Storey, J. M. E.; Lewis, S. A. Implications of particulate and precursor compounds formed during highefficiency clean combustion in a diesel engine. SAE Tech. Pap. 2005-013844, 2005. (5) Wagner, R. M.; Green, J. B.; Dam, T. Q.; Edwards, D.; Storey, J. M. Simultaneous low engine-out NOx and particulate matter with highly diluted diesel combustion. SAE Tech. Pap. 2003-01-0262, 2003. (6) Wagner, R. M.; Sluder, S. S.; Lewis, S. A.; Storey, J. J. Combustion mode switching for improved emissions and efficiency in diesel engine. In Proceedings of the 4th Joint Technical Meeting of the U.S. Sections of the Combustion Institute, Philadelphia, PA, 2005.

In conventional diesel operation, the majority of the heat release occurs during the mixing-controlled combustion phase and, thus, most emissions will be created in the mixingcontrolled phase. Dec furthered the understanding of the mixing control combustion phase in a sequence of laser diagnostic studies.7 Dec developed a generalized explanation for the behavior of the combusting diesel jet. He described the structure of the diffusion flame, indicating the layers by equivalence ratio and concentration of soot, which varies throughout the combusting diesel jet. Exhaust gas recirculation (EGR) is a technique used to reduce NOx emissions in conventional compression ignition engines. Exhaust gas recirculation reduces in-cylinder temperatures in three ways. First, CO2 is a major product of combustion and has a high specific heat. The high CO2 content of EGR gas acts as a heat sink to reduce adiabatic flame temperature and, thus, reduces temperature-dependent emissions, such as NOx. This is known as the thermal effect. Second, circulation of EGR into the air intake dilutes the O2 content of air. This reduces combustion temperatures and provides less O2 to combine with N2 to form NOx. However, the reduction in O2 content, moreover, the shift of air-fuel charge to a fuel-rich ratio, increases PM production.8-10 This is known as the dilution effect. Third, the introduction of reactive exhaust gas in the combustion chamber, specifically CO2, results in differing reactions, which can act to suppress the formation of emissions. This is known as the chemical effect. (7) Dec, J. E. A conceptual model of DI diesel combustion based on laser-sheet imaging. SAE Tech. Pap. 1997-97-0873, 1997. (8) Heywood, J. B. Internal Combustion Engine Fundamentals; McGrawHill Book Company: New York, 1988; p 930. (9) Lapuerta, M.; Hernandez, J.; Gimenez, F. Evaluation of exhaust gas recirculation as a technique for reducing diesel engine NOx emissions. Proc. Inst. Mech. Eng., Part D 2000, 214. (10) Zheng, M.; Reader, G. T.; Hawley, J. G. Diesel engine exhaust gas recirculationsA review on advanced and novel concepts. Energy ConVers. Manage. 2004, 45, 883–900.

10.1021/ef800557d CCC: $40.75  2009 American Chemical Society Published on Web 12/15/2008

144 Energy & Fuels, Vol. 23, 2009

Lilik et al.

Exhaust gas recirculation percentage (EGR %) can be quantified on the basis of the volume percent of CO2 in the ambient air, intake air, and exhaust, as given in eq 1.9 EGR %CO2 )

CO2 intake(vol %) - CO2 ambient(vol %) CO2 exhaust(vol %) - CO2 ambient(vol %)

(1)

This definition of EGR, along with the use of CO2-based simulated EGR, is the method used in this study. This process is based on work from Al-Qurashi et al., who conducted fundamental flame studies that showed that the thermal effect of EGR enhances the oxidative reactivity of diesel soot.11 The effects of CO2 on combustion have been well-studied and isolated. In a fundamental counterflow flame, Du et al. isolated the chemical, thermal, and dilution effects of CO2 and found that the chemical effects directly suppress soot inception.12 Axelbaum and Law conducted co-flow diffusion flame studies with CO2.13 They too isolated the effects of CO2 on soot formation, finding that dilution effects are more dominate than thermal effects when small amounts of inert diluents are added. However, Axelbaum and Law found that when large amounts of diluents are added the thermal effects become dominant. Combustion chamber studies in which stratified or homogeneous charge combustion occurred have shown the thermal effect of CO2 to have a dominant influence on the reduction of the combustion temperature.14,15 However, under standard diesel combustion conditions, the dilution effect (i.e., dilution of oxygen in the intake charge) had been shown to be the dominant cause of combustion temperature reduction.16-19 It should be noted that high concentrations of CO2 in the intake charge will have significant thermal effects as shown by Szybist in a model in which EGR mixtures with different specific heats were compared.15 As an alternative to basing the EGR % on CO2 concentrations in the intake, exhaust, and ambient, EGR % can be calculated on the basis of oxygen if the concentration of oxygen is known for both the intake and exhaust, as given in eq 2. EGR%O2 )

O2 intake(vol %) - O2 ambient(vol %) O2 exhaust(vol %) - O2 ambient(vol %)

(2)

EGR can also be examined on the basis of the intake oxygen concentration. Upatnieks and Mueller have demonstrated the (11) Al-Qurashi, K.; Lueking, A. D.; Boehman, A. L. The deconvolution of the thermal, dilution, and chemical effects of exhaust gas recirculation (EGR) on the reactivity of engine and flame soot. Flames Soot 2008, in press. (12) Du, D. X.; Axelbaum, R. L.; Law, C. K. The influence of carbon dioxide and oxygen as additives on soot formation in diffusion flames. In the 23rd International Symposium on Combustion, The Combustion Institute, Pittsburgh, PA, 1990; pp 1501-1507. (13) Axelbaum, R. L.; Law, C. K. Soot formation and inert addition in diffusion flames. In the 23rd International Symposium on Combustion, The Combustion Institute, Pittsburgh, PA, 1990; pp 1517-1523. (14) Moriyoshi, Y.; Morita, M. Effects of fuel and diluents on stratified charge turbulent combustion in simplified conditions. SAE Tech. Pap. 200301-1807, 2003. (15) Szybist, J. P. Fuel-specific effect of exhaust gas residuals on HCCI combustion: A modeling study. SAE Tech. Pap. 2008-01-2402, 2008. (16) Ladommatos, N.; Adbelhalim, S. M.; Zhao, H.; Hu, Z. The dilution, chemical and thermal effects of exhaust gas recirculation on diesel engines emissionssPart 1: Effect of reducing inlet charge oxygen. SAE Tech. Pap. 961165, 1996. (17) Ladommatos, N.; Adbelhalim, S. M.; Zhao, H.; Hu, Z. The dilution, chemical and thermal effects of exhaust gas recirculation on diesel engines emissionssPart 2: Effect of carbon dioxide. SAE Tech. Pap. 961167, 1996. (18) Ladommatos, N.; Adbelhallm, S. M.; Zhao, H.; Hu, Z. The dilution, chemical and thermal effects of exhaust gas recirculation on diesel engines emissionssPart 4: Effects of carbon dioxide and water vapour. SAE Tech. Pap. 971660, 1997. (19) Ropke, S.; Schweimer, G. W.; Strauss, T. S. NOx formation in diesel engine for various fuel and intake gases. SAE Tech. Pap. 950213, 1995.

ability to reduce NOx and PM emission by spanning intake oxygen concentrations between 21 and 6% oxygen concentrations using nitrogen dilution in an optical engine. Oxygen concentrations lower than 12% led to a loss of efficiency.20 Given the examples above, it is often more convenient and representative of the effect caused by EGR to represent EGR as intake dilution of the oxygen concentration rather than EGR %, even though LTC combustion has been classically defined by EGR %. Given the multiple methods of presenting intake dilution work, it is necessary to note the specific method of dilution. EGR decreases engine efficiency. Pumping losses increase as EGR rates increase. The indicated work decreases as incomplete combustion increases in the form of increased CO and HC emissions. Also, indicated work suffers from the reduced cylinder temperatures. EGR is cooled using engine coolant to recover and prevent the loss of volumetric efficiency caused by fumigating the intake charge with excessively hot gases. However, cooling the EGR increases the losses from heat rejection.21 Large rates of EGR increase cylinder-cylinder variations.22 Under high EGR conditions, individual cylinders do not receive uniform charges of EGR. Cylinders in which a larger degree of dilution occurs experience a greater degree of variation, especially under low-load operation. This cylinder-cylinder disparity is due to the short mixing length between the point where the EGR meets the intake and a given cylinder. This will result in varying emissions from cylinder to cylinder.23 HCCI uses advantages associated with spark ignition and compression ignition engines by combining a homogeneous charge with a compression ignition combustion process.24 The homogeneous mixture of HCCI is fuel-lean and/or dilute. Combustion of the charge occurs globally without a propagating flame, resulting in combustion with local hotspots.25 Fuel-lean mixtures produce less PM because of the high rate of oxidization occurring in the locally lean charges. The locally lower temperatures of HCCI produce less NOx. In contrast, the stratified diffusion flame, conventionally used in compression ignition engines, has layers of fuel-rich zones, where PM is created. Also, at the periphery of these fuel-rich zones, pockets of high temperature are present, which generate thermal NOx. HCCI can be approximated in a CI engine by early fuel injection combined with high EGR. The acronym PCCI has been used in advanced combustion literature with multiple meanings. Neely and co-workers used PCCI to refer to premixed controlled compression ignition combustion, having an increased, advanced pilot injection and a retarded main injection.26 Kanda et al.27 and Araki et al.28 refer to PCCI as premixed charge compression ignition, in which diesel fuel is injected early. Sluder and co(20) Upatnieks, A.; Mueller, C. J. Clean, controlled DI diesel combustion using dilute, cool charge gas and a short-ignition-delay, oxygenated fuel. SAE Tech. Pap. 2005-01-0363, 2005. (21) Jacobs, T.; Assanis, D.; Filipi, Z. The impact of exhaust gas recirculation on performance and emissions of a heavy-duty diesel engine. SAE Tech. Pap. 2003-01-1068, 2003. (22) Zheng, M.; Reader, G. Preliminary investigation of cycle to cycle variation in a nonair-breathing diesel engine. J. Energy Resour. Technol. 1995, 117, 24–28. (23) Edwards, K. D.; Wagner, R. M.; Chakravarthy, V. K.; Daw, C. S.; Green, J. B. A hybrid 2-zone/WAVE engine combustion model for simulating combustion instabilities during dilute operation. SAE Tech. Pap. 2005-01-3801, 2005. (24) Yao, M.; Zhang, B.; Zheng, Z.; Chen, Z. Experimental study on homogeneous charge compression ignition combustion with primary reference fuel. Combust. Sci. Technol. 2007, 2539–2559. (25) Szybist, J. P.; Bunting, B. G. Cetane number and engine speed effects on diesel HCCI performance and emissions. SAE Tech. Pap. 200501-3723, 2005.

AdVanced Combustion Operations

workers refer to PCCI as partially premixed charge compression ignition.3 No matter what the PCCI acronym stands for, PCCI commonly refers to an advanced combustion process that allows for a large premixed burn. In PCCI, fuel is injected early into the cylinder, during which an ignition delay occurs until cylinder conditions are right for autoignition. During the ignition delay, atomized diesel fuel mixes with air, creating a locally fuel-lean charge. If injection of diesel fuel continues past the point of autoignition, the burn will transition from a premixed burn to a diffusion burn. The contrast between HCCI and PCCI should be noted. The air-fuel charge in HCCI is homogeneous when it enters the cylinder. In PCCI, advanced injection of fuel leads to an extended premix combustion phase. PCCI can be seen as an intermediate step between conventional diesel combustion and HCCI. The charge in PCCI is not mixed as well; thus, there will be more hot spots. Also, because PCCI injects fuel via the diesel fuel injector, the long ignition delay may result in diesel fuel penetration to the cylinder walls, resulting in incomplete combustion. Similar to HCCI, PCCI suffers from increased HC and CO emissions related to the overly lean combustion conditions. However, PCCI permits a practical route to approximate HCCI because injection timing and EGR level can be used in concert to control ignition timing. HECC was demonstrated at Oak Ridge National Laboratory as a type of PCCI operating mode. HECC is accomplished by a combination of single-pulse injection, EGR (50%), early injection timing, and increased injection pressure. The EGR reduces NOx emissions and increases PM emissions. The early injection allows time for the diesel fuel to mix with air before combustion. Thus, an extended premixed combustion phase occurs, accompanied by a shortened mixing-controlled combustion phase. The premixed air and fuel are locally fuel-lean, thus decreasing PM. As the premixed air-fuel charge is consumed, the combustion transitions to a diffusion burn. Increasing the injection pressure decreases the injection duration, which causes more fuel to be premixed and burned during the premixed combustion phase. The HECC mode provides a decrease in NOx and PM emissions while maintaining or even increasing fuel efficiency. However, the HECC mode results in increased HC and CO emissions, which is common with HCCI-like modes.1-6,29 In a recent study, Wagner, Sluder, and co-workers have shown the HECC mode to be operable at 1500 rpm at 1.0 bar Indicated Mean Effective Pressure (IMEP), 1500 rpm at 2.6 bar IMEP, 2000 rpm at 2.0 bar IMEP, and 2300 rpm at 4.2 bar IMEP. The tests were conducted on a modified Mercedes 1.7 L, directinjection diesel engine with cooled EGR. In all four of the engine conditions, NOx was reduced by more than 80% compared to the baseline. PM decreased between 30 and 50% and was even further decreased from 85 to 100% when the fuel injector nozzles were replaced to further increase atomization. HC levels doubled at the lowest speed and only slightly increased at the highest speed. The CO emissions doubled in three of the four (26) Neely, G. D.; Sasaki, S.; Huang, Y.; Leet, J. A.; Stewart, D. W. New diesel emission control strategy to meet US tier 2 emissions regulations. SAE Tech. Pap. 2005-01-1091, 2005. (27) Kanda, T.; Hakozaki, T.; Uchimoto, T.; Hatano, J.; Kitayama, N.; Sono, H. PCCI operation with early injection of conventional diesel fuel. SAE Tech. Pap. 2005-01-0378, 2005. (28) Araki, M.; Umino, T.; Obokata, T.; Ishima, T.; Shiga, S.; Nakamura, H. Effects of compression ratio on characteristics of PCCI diesel combustion with a hollow cone spray. SAE Tech. Pap. 2005-01-2130, 2005. (29) Wagner, R. M.; Green, J. B.; Storey, J. M.; Daw, C. S. Extending exhaust gas recirculation limits in diesel engines. In Proceedings of the 93rd Air & Waste Management Association (A&WMA) Annual Conference, Salt Lake City, UT, 2005.

Energy & Fuels, Vol. 23, 2009 145 Table 1. DDC 2.5 L Engine Specification DDC 2.5 L TD DI-4V automotive diesel engine bore (mm) stroke (mm) compression ratio connecting rod length (mm) rated power (kW) peak torque (N m) valve train (valves/cylinder) injection system bosch

92 94 17.5 159 103 at 4000 rpm 340 at 1800 rpm 4 electronically controlled common-rail injection system

conditions. The fuel consumption remained the same as the baseline for all four conditions. Experimental Section Engine and Test Facility. A heavily instrumented DDC/VM Motori 2.5 L, four-cylinder, turbocharged, common rail, directinjection light-duty diesel engine was used for steady-state testing. Engine specifications are given in Table 1. A 250 HP Eaton eddy current water-cooled dynamometer was coupled to the 2.5 L DDC engine to generate load. The engine and dynamometer were controlled by a Digalog Testmate control unit. Time-based data acquisition was managed using a customprogrammed National Instruments LabView VI. Analog signals from pressure transducers, thermocouples, mass flow meters, and emissions data were read by a series of National Instruments FieldPoint modules, including a FP-2015, FP-AO-210, FP-DO-403, FP-AI-102, FP-AI-112, and three FP-TC-120 modules. The data collected by the FieldPoint modules were saved every 10 s during 15 min of sampling per test. An unlocked electronic control unit (ECU) was used to modify and control main injection and pilot injection timings, as well as EGR valve position and fuel rail pressure. The unlocked ECU was connected to an ETAS MAC 2 interface via an ETK connection. The MAC 2 interface was connected to a PC running ETAS INCA version 5.0 software. INCA managed the ECU modifications in real time. The DDC 2.5 L engine regulates EGR rates using an ECU map based on engine speed and injection volume. The ECU map dictates the flow rate by varying the amplitude of the signal sent to a proportional pneumatic valve. The stock DDC 2.5 L engine then introduces EGR to the intake manifold via a Y pipe. The Y pipe was modified to include a stainless-steel tube, which extended into the intake manifold of the engine. CO2 emissions were sampled from this tube, thus providing an accurate indication of CO2 levels in the intake manifold charge. It was necessary to aspirate simulated EGR into the air intake of the engine to achieve a well-mixed and high concentration of EGR charge (∼50%). Bone-dry CO2 with a purity of 99.8% was used as simulated EGR. The flow rate of the simulated EGR was monitored and regulated using an array of Matheson model 605 rotameters. The simulated EGR was aspirated after the charge air cooler, as seen in Figure 1. The simulated EGR was dispersed and mixed with the boosted air using a custom-built mixing manifold. The manifold consisted of four porous metal (Hastelloy) filters, customarily used as spargers, placed on the radial of the manifold. The porous metal filters were used to inject simulated EGR. The manifold is seen in Figure 2. Gaseous Emissions. An AVL Combustion Emissions Bench II was used to measure gaseous emissions. The bench was composed of six gas-specific analyzers. Hot exhaust gases were sampled from the exhaust pipe of the engine by head-line filters and then fed through heated lines kept at a constant temperature of 190 °C. NOx and NO were measured using an EcoPhysics chemiluminescence analyzer. NO2 was assumed to be the value of NO subtracted from NOx. Total hydrocarbons and methane were measured using two separate ABB flame ionization detectors. CO and CO2 were measured by two separate Rosemount infrared analyzers, and O2

146 Energy & Fuels, Vol. 23, 2009

Lilik et al.

Figure 1. DDC/VM Motori 2.5 L turbodiesel engine EGR flow diagram.

Figure 2. Intake air manifold aspiration system.

was measured using a Rosemount paramagnetic analyzer. The hot exhaust sample going to the CO, CO2, and O2 analyzers was first chilled to remove moisture. PM Emissions. PM mass emissions were sampled using a Sierra Instruments BG-1 micro-dilution test stand. The samples were taken at a dilution ratio of 10, a total flow rate of 110 slpm, and a sample flow rate of 10 slpm over 5 min. The PM samples were collected on Pallflex 90 mm filters, type EMFAB TX40HI20-WW. The filters were weighed on a Sartorius M5P electronic microbalance before and after sampling. The scale was located in an environmental chamber set at 25 °C, with 45% relative humidity. The filters were placed in the environmental chamber 48 h prior to mass analysis. Five sample filters were taken per mode, and the four sample fitters having the lowest standard deviation were averaged to represent the mass produced at a given mode. A TSI 3936 scanning mobility particle sizer (SMPS) was used to analyze the size distribution of the PM. The SMPS instrument included a TSI series 3080 electrostatic classifier with a differential mobility analyzer (DMA), a series 3776 condensation particle counter (CPC), and a TSI series 3065 thermal denuder. A PC running Aerosol Instrument Manager Software collected and managed the sampled data. The BG-1 was used to draw and dilute samples from the exhaust. The BG-1 drew samples at a dilution ratio of 10, a total flow rate of 108.6 slpm, and a sample flow rate of 100 slpm. The SMPS

drew samples from the BG-1 at a rate of 1.4 slpm. The SMPS measurements were conducted using three different sample methods: passing the samples through the thermal denuder, passing the samples through the thermal denuder at 300 °C, and having the samples bypass the thermal denuder. The sampling methods affected the content of the volatile hydrocarbon present on the PM. A large number of samples (∼6) should have been taken on the SMPS to correct for and average out the inconsistent residence timing of the exhaust samples in the dilution chamber of the BG-1. Cylinder Pressure Trace Analysis. Pressure traces were measured using AVL GU12P pressure transducers, which replaced the glow plug in each of the four cylinders. The pressure trace voltages from the pressure transducers were amplified by a set of Kistler-type 5010 dual-mode amplifiers. The amplified voltages were read by an AVL Indimodul 621 data acquisition system. Needle lift data were collected from a Wolff Controls, Inc. Halleffect needle lift sensor, which was placed on the injector of cylinder 1. The needle lift signal was also collected by the Indimodul, which was triggered by a crank-angle signal from an AVL 365C angle encoder placed on the crankshaft. The pressure traces and needle lift data were recorded at a resolution of 0.1 crank-angle degrees and were averaged over 200 cycles. The real-time Indimodul data were transferred to a PC, which ran AVL Indicom 1.3 and Concerto 3.90 software to calculate the apparent heat-release rate. The apparent rate of heat release for each of the four cylinders was calculated from the volume and pressure trace data.

Results and Discussion Needle Lift. The needle lift data indicated the crank angle at which fuel is injected, as well as the duration of injection and needle lift height of fuel injected. Figure 3 displays the comparison of needle lift for the baseline, LTC, and HECC modes. Predictably, the baseline and LTC modes had similar start of injection and injection durations. This is due to the injection timing being locked at -17.4° after top dead center (ATDC) for the pilot injection and 2.9° ATDC for the main injection timing in both modes. The HECC mode had a single injection that was set at -4° ATDC. The HECC mode also had higher needle lift because the mode used only a single injection and must inject all of the required fuel during this single injection. The area under the injection peaks represents the

AdVanced Combustion Operations

Figure 3. Needle lift comparison for the (s) baseline, (- - -) LTC, and ( · · · ) HECC operating modes.

Figure 4. Pressure trace comparison for the (s) baseline, (- - -) LTC, and ( · · · ) HECC operating modes.

quantity of fuel injected during the actuation of the fuel injector. It is important to note that this HECC mode also used increased rail pressure; thus, a large quantity of fuel was injected for the given injection duration and needle lift height. Cylinder Pressure. The pressure traces indicate the pressure because of cylinder volume reduction from the travel of the piston, as well as the pressure created from hot product gases. The baseline and LTC modes have similar pressure traces because of their similar parameters. In Figure 4, a 2.7% decrease in maximum pressure occurs in the LTC mode compared to the baseline. The reduction in maximum pressure seen in the pressure traces is due to the high level of EGR of the LTC mode. The EGR absorbs released heat, lowering the adiabatic flame temperature.30 An increase in EGR levels also leads to a reduction in oxygen; the oxidizer needs to burn the fuel. Thus, the maximum pressure of the LTC mode is reduced. The maximum pressure produced in the HECC mode was 21% lower than that of the baseline mode, because of the large degree of premixed combustion, which occurred in the HECC mode. The second pressure peak of the HECC mode is due to the transition to mixing-controlled combustion, which is further explained by the apparent heat-release rate. Apparent Heat-Release Rate. Figure 5 displays the apparent heat-release rate of the baseline, LTC, and HECC operating modes. The start of combustion of the pilot injection (-7.9° ATDC) of the LTC mode is delayed compared to the start of combustion of the pilot injection (-8.7° ATDC) for the baseline mode. The delay in start of combustion observed for the LTC mode is due to the high concentration of EGR, which absorbs the heat produced by the in-cylinder compression, requiring further time to achieve ignition. Because its injection timing is different, the start of combustion of the HECC mode (7.1° ATDC) cannot be compared to that of the other modes. (30) Zhang, Y. Effects of biodiesel on engine performance and NOx emissions in a common rail diesel engine. M.S. Thesis, Department of Energy and Geo-Environmental Engineering, The Pennsylvania State University, University Park, PA, 2006.

Energy & Fuels, Vol. 23, 2009 147

Figure 5. Apparent heat-release rate comparison for the (s) baseline, (---) LTC, and (- - -) HECC operating modes.

The apparent heat-release rate profiles of the baseline and LTC modes appear generic, corresponding to the four diesel combustion phases described by Heywood.8 The HECC mode has a unique apparent heat-release rate profile. Diesel fuel is injected into this mode at -4° ATDC, but combustion does not start until 7.1° ATDC. The HECC mode has an 11.1 °C ignition delay. This extended start of combustion is due to the large concentration of EGR (∼50%) used in the mode, as well as an advanced injection timing. This mode is unique because it uses only three of the four diesel combustion phases described by Heywood, including a longer than usual ignition-delay phase. The single injection of the HECC mode ends at 8.7° ATDC, and the start of combustion begins at HECC (7.1° ATDC). By the time combustion begins, almost all of the fuel is injected into the cylinder, thus causing half the injected fuel to be consumed in a premixed combustion phase as evidenced by the calculated mass burn fraction 50% (MBF 50%) given to be 18.5° ATDC. As indicated by the end of injection and start of combustion overlapping, a small quantity of fuel was burned in the mixing-controlled combustion phase. However, the apparent heat-release rate plot, given in Figure 5, does not indicate the presence of a mixing-controlled combustion phase peak. Rather, it indicates the transition of the heat release directly from a premixed combustion phase to a late combustion phase, which is the trailing trace in the apparent heat-release rate plot of the HECC mode. The apparent heat-release rate plot of the HECC mode provides verification that the mode is a PCCI combustion mode. The gaseous emissions, particulate emissions, and brakespecific fuel consumption along with mode-specific parameters are summarized in Table 2. The results are discussed in detail below. Gaseous Emissions. Oxides of Nitrogen. The high levels of EGR used in the advanced combustion mode lowered the combustion temperature, quenching the production of thermal NO and thus reducing NOx. The LTC mode decreased NOx emissions by 89.5% compared to the baseline mode. The NOx reduction of the HECC mode was less than that of the LTC mode, with a 71.2% NOx reduction from the baseline. The HECC mode used on the DDC 2.5 L engine was not optimized to produce low NOx emissions alone but was rather optimized for simultaneously low NOx emissions, low PM emissions, and high thermal efficiency. Rail pressure, injection timing, and EGR % were the variables adjusted in the HECC mode optimization process. Nitric Oxide. The LTC and HECC modes reduced NO emissions by ∼83% compared to the baseline mode. The NOx and NO emissions together indicate that the HECC mode did not decrease NO2.

148 Energy & Fuels, Vol. 23, 2009

Lilik et al.

Table 2. Summary of Operating Mode Specifications and Emissions for the Baseline, LTC, and HECC Operating Modes total EGR %CO2 engine produced EGR %CO2 simulated EGR %CO2 intake O2 concentration EGR %O2 main timing (°ATDC) pilot timing (°ATDC) rail pressure (bar) NOx (g kW-1 h-1) NO (g kW-1 h-1) NO2 (g kW-1 h-1) THC (g kW-1 h-1) CO (g kW-1 h-1) CO2 (g kW-1 h-1) PM (g kW-1 h-1) BSFC (g kW-1 h-1) intake temperature (°C) exhaust temperature (°C)

baseline

LTC

HECC

11 11 none 20 15 2.9 -17.4 450 1.88 1.48 0.40 0.76 2.21 845 0.98 253 59 342

48 16 32 18 25 2.9 -17.4 450 0.72 0.62 0.10 1.06 3.71 1511 1.57 253 70 357

50 16 34 18 28 -4.0 none 490 0.89 0.61 0.28 1.65 7.17 1501 0.42 244 75 336

Nitrogen Dioxide. Table 2 indicates that the LTC mode reduced NO2 emissions by 119.2%, while the HECC mode only reduced NO2 emissions 33.8%, in comparison to the baseline mode. The appearance of a higher level of NO2 emissions produced in the HECC mode over the LTC mode was not expected. While the major pathway to NO2 formation is the oxidation of NO, the LTC and HECC modes both produced similar values of brake-specific NO emissions. Upatnieks, Mueller, and Martin conducted a study on an optically accessible, heavy-duty DI diesel engine in which intake oxygen was diluted via nitrogen as simulated EGR, resulting in an increased NO2/NO ratio. The increase of NO2 and decrease of NO was attributed to an increased quenching of the NO2-to-NO reaction (eq 3) because of decreasing flame temperatures.31 NO2 + O T NO + O2

(3)

The observation reported by Upatnieks, Mueller, and Martin corresponds to the NO and NO2 emissions of LTC and HECC modes of this study, although CO2 was used as the diluent instead of nitrogen.31 The extended premixed combustion phase of the HECC mode can be assumed to have a lower flame temperature than that of the LTC mode, which has mostly a mixing-controlled combustion phase. The lower flame temperatures of the HECC mode enhance quenching of the NO2-toNO reaction, which explains why the HECC mode produced higher NO2 emissions. Hydrocarbons. The advanced combustion modes increased the HC emissions compared to the baseline mode. The HC emissions of the LTC mode increased by 33% compared to the baseline mode, while HC emissions increased by 73% in the HECC mode. The HC emissions increase in the PCCI mode, similar to the HECC mode, which is caused by overly lean combustion conditions. Wagner, Sluder, and co-workers reported increased HC for the HECC mode under low engine speed operation and a decrease in HC emission under high engine speed operation. While the cause for the increase in HCs was not explored by Wagner, Sluder, and co-workers, the increase was attributed to the mixing time scale of the low speed engine operation.3 Carbon Monoxide. The CO emissions of the HECC mode increased by 105% compared to the baseline mode. The CO emissions of the LTC mode increased by 50% compared to the (31) Upatnieks, A.; Mueller, C. J.; Martin, G. C. The influence of chargegas dilution and temperature on DI diesel combustion processes using a short-ignition-delay, oxygenated fuel. SAE Tech. Pap. 2005-01-2088, 2005.

baseline mode. This dramatic increase in CO emissions, specifically in the HECC mode, can be attributed to incomplete combustion. The HECC mode operates in a locally fuel-lean condition. Overly fuel-lean combustion will lead to incomplete combustion.32 High levels of EGR also compounded the degree of incomplete combustion.21 The increased level of CO emissions in the LTC mode was also caused by incomplete combustion because of the high levels of EGR. Carbon Dioxide. CO2 emissions increased in the advanced combustion modes over the baseline modes because of the increased levels of CO2 added to the EGR gas. Furthermore, the CO2 concentration in the air intake for the HECC mode (6.15% CO2) was slightly higher than that of the LTC mode (5.96% CO2). The CO2 emissions of both LTC and HECC modes were within 10 g kW-1 h-1, as can be seen in Table 2. The HECC fuel efficiency is greater than that of the LTC mode and, as such, should have lower CO2 emissions. The 10 g kW-1 h-1 lower CO2 emissions of the HECC mode and a slightly higher dilution concentration of CO2 in the intake air can be the reason for the LTC and HECC modes producing similar CO2 emissions. The higher efficiency of the HECC mode can be attributed to its large premixed combustion phase. The premixed combustion phase releases energy in a shorter period, causing the combustion to occur near a constant volume combustion event. In contrast, the lower fuel efficiency LTC mode has a long mixing-controlled combustion phase, which occurs as a constant pressure combustion event. Particulate Emissions. PM Mass Emissions. PM is mainly created in the diffusion flame of the mixing-controlled combustion phase, as is the case with the baseline operation condition. The PM is formed in fuel-rich zones of the flame, where fuel is pyrolyzed. The LTC mode uses the same injection strategy as the baseline combustion mode; thus, the combustion of the LTC mode is also dominated by the mixing-controlled combustion phase. However, the LTC mode uses ∼50% EGR, which lowers the combustion temperature by absorbing heat. The reduction in the combustion temperature lowers the rate at which the PM is oxidized. Furthermore, the reduction of oxygen with the increase of EGR reduces the oxygen available to oxidize soot formed in the diffusion flame. Thus, the PM in the LTC mode is 46% higher than the baseline mode. The combustion of the HECC mode is dominated by the premixed combustion phase, with the premixed air-fuel charge combusting locally under fuel-lean conditions at lower temperatures. The HECC mode also uses ∼50% EGR, which decreases combustion temperatures further. The low combustion temperatures and the low fuel equivalence ratio of the HECC mode shifts the mode outside the PM formation peninsula of Akihama and co-workers’ local equivalence ratio versus local temperature model.33 The rate of oxidation is reduced by lowered combustion temperatures, but so little PM is formed that the HECC mode yields lower PM than the baseline or LTC modes. Although the NOx-PM tradeoff usually means that efforts to reduce engine-out PM emissions will lead to increased NOx emissions and vice versa, Table 2 indicates simultaneous reductions of NOx, PM, and brake-specific fuel consumption. Particle Size Distribution. A TSI 3936 SMPS was used with a TSI thermal denuder to analyze the size distribution of the PM. Figures 6-9 compare the baseline, LTC, and HECC (32) Dec, J. E.; Sjo¨berg, M. A parametric study of HCCI combustionsThe sources of emissions at low loads and the effects of GDI fuel injection. SAE Tech. Pap. 2003-01-0752, 2003. (33) Akihama, K.; Takatori, Y.; Inaga, K. Mechanism of the smokeless rich diesel combustion by reducing temperature. SAE Tech. Pap. 2001-010655, 2001.

AdVanced Combustion Operations

Figure 6. Particle size distribution from SMPS while bypassing the thermal denuder for engine operation under the (s) baseline, (---) LTC, and (- - -) HECC modes.

Figure 7. Particle size distribution from SMPS with the thermal denuder at 30 °C for engine operation under the (s) baseline, (---) LTC, and (- - -) HECC modes.

Figure 8. Particle size distribution from SMPS with the thermal denuder at 300 °C for engine operation under the (s) baseline, (---) LTC, and (- - -) HECC modes.

operating modes when the exhaust sample bypassed the thermal denuder and flowed through the thermal denuder at 30 and 300 °C. The diluted exhaust sample contained PM made up of a solid carbon fraction (soot) and an organic fraction. The thermal denuder removes the organic fraction of the PM when operated at 300 °C. A comparison of the PM size distribution under these different operating conditions for the thermal denuder shows how much of the PM is soot and how much is an organic aerosol. The organic fraction is composed of unburned hydrocarbons that condensed onto the soot particles and also may condense into 5-30 nm diameter droplets, which the SMPS counts as nanoparticles. The LTC and HECC modes had increased levels of unburned hydrocarbons, which could indicate higher levels of organic fraction on the soot. Figure 6 shows that, when the sample bypasses the thermal denuder, the LTC mode yields a higher concentration of particles than the baseline mode and

Energy & Fuels, Vol. 23, 2009 149

Figure 9. Particle size distribution from SMPS under HECC mode operation at 1800 rpm, 4.2 bmep, and ∼50% EGR with rail pressure at 490 bar with different thermal denuder (TD) temperatures and start of injection timing (SOI): (O) TD at 30 °C and SOI at -2° ATDC, (0) TD at 300 °C and SOI at -2° ATDC, (b) TD at 30 °C and SOI at -4° ATDC, and (9) TD at 300 °C and SOI at -4° ATDC.

that the particle concentrations under the HECC mode are quite reduced. The mass-based PM data shows the same trend. Figure 7 shows data for samples flowing through the thermal denuder at 30 °C. The concentration of particles is significantly reduced, with the sample flowing through the thermal denuder. The concentration of the baseline mode unexpectedly increased over the LTC mode when the sample passed through the 30 °C thermal denuder, which is set at that temperature to avoid removing the organic fraction. The thermal denuder is, however, a cylinder full of activated carbon, which filters and absorbs the organic fraction from the PM. The LTC mode contained a larger organic fraction of PM than that of the baseline mode, thus accounting for the shift in concentration seen in Figure 7. Figure 8 displays the PM concentrations of the three modes tested, with the sample flowing through the thermal denuder at 300 °C. The organic fraction of the PM is completely stripped away, and because of the thermal denuder being set to 300 °C, only soot remains. Note that the concentrations of all three of the modes are dramatically reduced. While the HECC mode was being explored on the DDC 2.5 L engine, the SMPS was used to generate PM data relatively rapidly as compared to traditional filter-based gravimetric methods. In the exploration for the HECC mode, the injection timing was adjusted to locate where the engine operated at simultaneously low NOx and PM emissions while also maintaining fuel economy. In this process, the engine was found to produce large quantities of nanoparticles when the single main injection timing was -2° ATDC. Figure 9 compares the HECC mode used at -4° ATDC to the mode that produced the increased nanoparticle concentration at -2° ATDC. The nanoparticles are entirely composed of the organic fraction. As can be seen in Figure 9, the thermal denuder removes these nanoparticles and an ultralow soot concentration remains. Brake-Specific Fuel Consumption. Brake-specific fuel consumption of the LTC mode compared to the baseline mode increased by 0.2%. The fuel efficiency of the HECC mode, however, increased by 3.6% over the baseline. In a HECC study conducted by Sluder and Wagner, the fuel consumption rate of the HECC mode matched that of their baseline.3 The increase in fuel efficiency seen in this study can be attributed to the use of simulated EGR, which entered the engine at room temperature. Conversely, Sluder and Wagner used actual EGR, which was passed through a heat exchanger and cooled by an engine coolant. The higher temperature of the EGR used in Sluder and Wagner’s study compared to the present work decreased the volumetric efficiency of the engine, thus decreasing fuel

150 Energy & Fuels, Vol. 23, 2009

efficiency. The use of simulated EGR that was aspirated into the engine intake at room temperature artificially increased fuel efficiency when compared to the study of Sluder and Wagner, in which the EGR gas was cooled using an engine coolant in a heat exchanger prior to mixing with the intake air. Furthermore, the parameter selection of EGR %, rail pressure, and injection timing affects the degree of optimization of fuel consumption, NOx emissions, and PM emissions of the HECC mode. Conclusions In this work, a Detroit diesel/VM Motori 2.5 L TD DI-4V laboratory test engine, using a combination of actual and simulated EGR and advanced injection timing, was operated in a low-temperature, advanced combustion mode referred to as HECC. This HECC mode led to significantly reduced PM and NOx emissions, while improving fuel economy and thus avoiding the NOx-PM tradeoff.

Lilik et al.

The SMPS indicated the formation of nanoparticles from the organic fraction of PM, with a particularly high concentration under the LTC operating condition when the exhaust sample bypassed the thermal denuder. The nanoparticles were observed to be almost entirely an organic aerosol during HECC mode operation, with fuel injection timing at -2° ATDC. When the thermal denuder was used to remove the organic fraction of PM during HECC mode operation with fuel injection timing at -2° ATDC, only trace levels of soot remained. These data indicate the possibility of nearly eliminating soot emissions in a further optimized HECC mode. Acknowledgment. The authors thank Robie Lewis and the National Energy Technology Laboratory for their support of this work under Instrument DE-FC25-04FT42233. The authors also thank Patrick Quarles and Asemblon, Inc., Redmond, WA, for their support of this work. EF800557D