Detailed Investigation on Soot Particle Size ... - ACS Publications

ARTICLE pubs.acs.org/IECR

Detailed Investigation on Soot Particle Size Distribution during DPF Regeneration, using Standard and Bio-Diesel Fuels Jose C. Caroca,† Federico Millo,‡ Davide Vezza,‡ Theodoros Vlachos,‡ Andrea De Filippo,§ Samir Bensaid,† Nunzio Russo,† and Debora Fino*,† †

Department of Materials Science and Chemical Engineering, and ‡Department of Energetics, Politecnico di Torino, Corso Duca degli Abruzzi 24, 10129 - Torino, Italy § General Motors Powertrain Europe, C.so Castelfidardo 36, 10138 Torino, Italy ABSTRACT: The particle number and size distribution are important aspects to qualify diesel engine emissions, considering that new limits, in term of particle number, are expected for Euro 6 regulations. In this scenario it is important to study particulate matter (PM) emissions, not only during engine normal operating mode, but also during diesel particulate filter (DPF) regeneration processes. The aim of this work is therefore to analyze PM emissions throughout the whole exhaust system of a small displacement Euro 5 common rail automotive diesel engine, during both normal operating conditions and DPF regeneration mode. Because the test engine was equipped with a close-coupled after-treatment system, featuring a Diesel Oxidation Catalyst (DOC) and a DPF integrated in a single canning, the exhaust gas was sampled at the engine outlet, at the DOC outlet, and at the DPF outlet to fully characterize the PM emissions throughout the exhaust line. After a two-stage dilution, the sampled gas was analyzed by means of a TSI 3080 SMPS, in the range 6 to 225 nm range. The particle number and size distribution were evaluated at part load, both under normal operating conditions and at DPF regeneration mode, to highlight the impact of the different combustion processes on the PM characteristics. Finally, the particle number and size distribution at the engine outlet were also evaluated while fueling the engine with neat fatty acid methyl ester (FAME) to evaluate the impact of alternative fuels on PM characteristics during normal operating conditions. The results have shown that, under normal operating conditions, the engine and DOC outlet particle number and mass size distributions appear to be very similar, while the DPF exhibits high values of filtration efficiency on a particle number basis, even in the nanoparticle range. Regeneration mode caused a particle number increase of 1 order of magnitude, with a substantial shift of the number distribution peaks toward larger diameters. The particle number across the DOC showed a remarkable reduction for particles larger than 40 nm, with a reduction of 1 order of magnitude in their concentrations. Finally, fueling the engine with FAME leads to remarkable reductions in terms of particle number and mass (up to 80% and 90%, respectively) under normal operating mode conditions.

1. INTRODUCTION Although diesel engine emission regulations throughout the world have always been based on gravimetric methods for the PM measurement, there is increasing concern about PM health and environmental effects based on the number of particles. There is increasing alarm that reductions in PM mass, which can be achieved through engine and after-treatment measures, are not always accompanied by decreases in the total particle number and surface area. This has led to the introduction of new PM limits, in terms of particle number,1 and has created interest among researchers about the particle number and size characteristics, especially as far as engines equipped with DPF are concerned.2-6 The aim of this work was therefore to collect further information on the particle number and size for emissions from a small displacement Euro 5 common rail automotive diesel engine, equipped with a close coupled after-treatment system, featuring a DOC and a DPF integrated in a single canning, especially during regeneration events at low engine speeds and loads, which are representative of urban driving conditions. Finally, due to the growing interest in alternative fuels for diesel engines,7-12 the impact of a biofuel on the PM characteristics has r 2010 American Chemical Society

also been evaluated, by fueling the engine with neat fatty acid methyl ester (FAME).

2. EXPERIMENTAL SETUP The experimental tests were carried out at the ICE Advanced Laboratory of the Politecnico di Torino on the test rig shown in Figure 1a, which is equipped with an eddy current dynamometer connected to a passenger car turbocharged Common Rail DI Diesel engine, the main characteristics of which are listed in Table 1. The engine is equipped with a close coupled aftertreatment system, featuring a DOC and a catalyzed DPF integrated in a single canning (see Figure 1b), the main features of which are shown in Table 2. The exhaust gas was sampled at the engine outlet downstream from the turbine, at the DOC outlet,

Special Issue: IMCCRE 2010 Received: March 19, 2010 Accepted: June 4, 2010 Revised: June 3, 2010 Published: June 28, 2010 2650

dx.doi.org/10.1021/ie1006799 | Ind. Eng. Chem. Res. 2011, 50, 2650–2658

Industrial & Engineering Chemistry Research

ARTICLE

Figure 1. (a) Experimental test rig: two stage dilution system. (b) Architecture of the exhaust system used during the experimental tests. (c) Sketch of the dilution system. (d) First dilution stage scheme (adapted from Dekati Ltd.). (e) Distillation curves for neat diesel (B0) and biodiesel (B100) fuels.

Table 1. Main Characteristics of the Test Engine engine type displacement

diesel 4 stroke 1248 cm3

cylinders arrangement

4 in line

max power

55 kW @ 4000 rpm

bore stroke

69.6 mm 82 mm

compression ratio

16.8:1

turbocharger

single-stage with waste gate

fuel injection system

common rail

Table 2. Main Characteristics of the After-Treatment System DOC

DPF

material

cordierite

SiC

cell density [cpsi]

400

300

substrate volume [dm3] substrate length [mm]

1.1 76.2

2.6 177.8

wall thickness [mil]

4.5

10

Table 3. Compositions and Properties of the Tested Fuels and at the DPF outlet to characterize the particle numbers and sizes throughout the whole exhaust system. The experimental tests were performed fueling the engine with the following fuels: B0, typical European diesel fuel compliant with EN590; and B100, pure FAME, complying with EN 14214. The main properties of the tested fuels are listed in Table 3, and their distillation curves are reported in Figure 1e. Figure 1c shows the sampling system scheme and its characteristics. The sampling system consists of two dilution stages connected to a TSI 3080 SMPS. A first sampling pipe, namely Line A, made of a heated and insulated inox pipe, connects the sampling point to the first dilution stage (first dilutor of a DEKATI DI-2000 package), which is heated at 250 °C to avoid nucleation.13 In the first dilution stage, filtered compressed air, heated to 150 °C, flows through an orifice placed on the sample flow axis, as shown in Figure 1d. To obtain a higher dilution ratio, a second dilution stage has been used (second dilutor of a DEKATI DI-2000

properties

B0

B100

carbon content C [w%]; ASTM D5291-09

85.8

79.6

hydrogen content H [w%]; ASTM D5291-09

13.1

12.1

oxygen content O [w%]; ASTM D5291-09 sulfur content S [ppm]; UNI EN ISO 20846:2005

14.5

8.3 4.2 12.95

stoichiometric ratio (A/F)st

14.37

lower Heating Value [MJ/kg]; ASTM D240-09

42.690

37.365

cetane number; UNI EN ISO 5165:2001

53

51.5

836.8

883.2

density at 15 °C [kg/m3]; UNI EN ISO 12185:1999 viscosity at 40 °C [mm2/s]; UNI EN ISO 3104:2000

2.736

4.254

package). The second stage is not heated and it is directly connected to the TSI 3080 SMPS by means of a flexible pipe (line B in Figure 1c). (It should be pointed out that, due to safety reasons and room constrains, the length of the flexible pipe cannot 2651

dx.doi.org/10.1021/ie1006799 |Ind. Eng. Chem. Res. 2011, 50, 2650–2658


ARTICLE

Figure 2. Injection strategies during normal engine operating conditions (top) and during DPF regeneration (bottom).

Figure 4. Particle number size distributions under normal operating conditions (1750 rpm/3.5 bar bmep): log (top) and linear (bottom) y axis scales.

Figure 3. Exhaust gas temperatures under normal operating conditions and during DPF regeneration mode, at 1750 rpm/3.5 bar bmep.

be kept shorter than 750 cm. This slightly affects the measured data by increasing the mean particle size and by reducing the total particle number, as it was shown by preliminary tests. However, the comparisons between different combustion modes or between particle emissions at different locations throughout the exhaust system remain meaningful. A total dilution ratio higher than 60 was achieved, which, according to the literature, should remarkably reduce particle agglomeration.13-15 At the end of the above-described sampling system, the exhaust gases reach the TSI 3080 SMPS (see ref 14 for a detailed

description of the operating principle), which is composed of the following devices: an electrostatic classifier TSI 3080 with a Kr85 bipolar charger using kripton as ion source; a 1035900 inlet impactor (nozzle size 0.071 cm); a differential mobility analyzer TSI 3081; a condensation particle counter TSI 3025A. With the aerosol flow at 1.5 lpm and the sheath flow at 15 lpm, it was possible to cover a broad range of particle diameters from 6 to 225 nm. The size distributions were recorded as dN/dlogDp including mathematical correction for particle multiple charges and diffusion losses, both provided by TSI SMPS software. In the following figures, the displayed particle number size distributions are also corrected for the above said dilution ratio. The exhaust gases were also sampled at the engine outlet by means of a Fisher-Rosemount NGA-2000 gas analyzer, for the measurement of the gaseous emission; furthermore, an AVL 415S smoke meter, for the measurement of the particulate filter smoke number FSN, was also used for repeatability checks. 2652



ARTICLE

Figure 6. DPF filtration efficiency under normal operating conditions (1750 rpm/3.5 bar bmep).

Figure 5. Particle mass size distributions under normal operating conditions (1750 rpm/3.5 bar bmep): log (top) and linear y (bottom) axis scales.

The in-cylinder pressure was acquired by means of piezoresistive pressure transducers integrated in the glow plugs: the pressure signals were sampled by means of a 12-bit high-speed multichannel data acquisition system (National Instruments DAQCard-Al-16E4), coupled to a high-resolution (0.4°) crank angle encoder. Injector command signals were also acquired by means of a Tektronix TCPA300 current probe.

3. PARTICLE SIZE AND NUMBER MEASUREMENTS The particle number and size distribution were evaluated at part load both under normal operating conditions and at the DPF regeneration mode to highlight the impact of the different

Figure 7. Comparison between engine outlet particle number size distributions under normal operating conditions and during DPF regeneration mode (regeneration with and without post injection) (1750 rpm/3.5 bar bmep): log (top) and linear (bottom) y axis scales.

combustion processes on the PM characteristics. A low-speed, low-load engine operating condition (1750 rpm/3.5 bar bmep) was selected, as representative of urban driving conditions, which can be critical as far as DPF regeneration is concerned. During normal operating conditions at low load, the injection pattern 2653



ARTICLE

Figure 10. Exhaust gas temperatures under normal operating conditions, during DPF standard regeneration, and during regeneration without post injection, at 1750 rpm/3.5 bar bmep.

Figure 8. Comparison between engine outlet particle mass size distributions under normal operating conditions and during DPF regeneration mode (regeneration with and without post injection) (1750 rpm/3.5 bar bmep): log (top) and linear (bottom) y axis scales.

Figure 9. Injection strategies during DPF regeneration without post injection.

usually shows one pilot injection, followed by the main injection (see Figure 2 on top), whereas under the DPF regeneration mode, the main injection pulse is delayed and is then closely followed by an after injection, plus an additional post injection, as shown at the bottom of Figure 2.

Figure 11. Unburned hydrocarbon emissions measured at the engine and DOC outlet during normal operating conditions and during regeneration mode (with and without post injection at 1750 rpm/3.5 bar bmep).

While the delayed main injection along with the after injection allows an increase in the engine exhaust temperature up to about 300 °C even at the low-load tested operating point, the post injected fuel is mainly burned across the DOC, as its increased HC conversion efficiency suggests, thus raising the DPF inlet temperature above 600 °C to promote DPF regeneration, as shown in Figure 3. 3.1. Normal Engine Operating Conditions. The particle sizes and numbers measured under normal operating conditions are shown in Figures 4 and 5. Both logarithmic and linear diagrams have been used: the first ones are useful for a better representation of data with different magnitudes (such as, for instance, concentrations upstream and downstream of the DPF), while the second ones allow the distribution shape to be observed more clearly and its maximum value to be located more easily. It should also be pointed out that, although small differences among distributions are emphasized on linear diagrams, the test repeatability can be considered as satisfactory, as shown by several repetition checks which will not be reported here for sake of brevity. Figures 4 and 5 show that no substantial modification in particle size and number can be observed across the DOC, with the number and mass distributions showing peaks at about 50 nm 2654



ARTICLE

Figure 12. Particle number size distributions under DPF regeneration conditions (1750 rpm/3.5 bar bmep): log (top) and linear (bottom) y axis scales.

and 100 nm, respectively. On the contrary, the particle number is reduced by 2 or 3 orders of magnitude across the DPF, as shown in Figure 6. As a result, the DPF filtration efficiency, defined on a particle number basis, i.e.: efficiency ¼ 1 -

N OUT N IN

is extremely high, approaching unity values. 3.2. DPF Regeneration Conditions. The comparisons between engine outlet particle number and size distributions under normal operating conditions and during DPF regeneration mode are shown in Figures 7 and 8. It should be pointed out that all the experimental tests were carried out on a freshly regenerated DPF, i.e., only after completing a regeneration event, even though lower filtration efficiency could be expected due to the absence of a soot cake layer on the DPF walls. This procedure was preferred to fully eliminate any soot load from the particulate trap and to

Figure 13. Particle mass size distributions under DPF regeneration conditions (1750 rpm/3.5 bar bmep): log (top) and linear (bottom) y axis scales.

evaluate the effects due only to the engine regeneration mode operation and not those related to the combustion of the soot accumulated over the DPF. It can be clearly noticed that, during DPF regeneration, the particle numbers increase by about 1 order of magnitude, while the number distribution peaks are shifted toward larger diameters, passing from 50 to 100 nm. The post injected fuel dramatically increases the number of particles, as well as their mass, due to the large amount of unburned hydrocarbons in the exhaust stream. Further tests have also been conducted modifying the regeneration injection strategy, i.e., eliminating the post injection, to investigate the post injection influence on the PM particle number and mass distribution. Figure 9 shows the injection strategy during this particular operating mode. Figure 10 illustrates the engine and DOC outlet temperatures during normal, 2655



ARTICLE

Figure 14. DOC filtration efficiency under DPF regeneration conditions (1750 rpm/3.5 bar bmep).

Figure 15. DPF filtration efficiency under DPF regeneration conditions (1750 rpm/3.5 bar bmep).

“standard” regeneration (i.e., regeneration with post injection) and regeneration without post injection operating conditions. Figure 11 shows the unburned hydrocarbon measured at the engine and DOC outlet during normal operating and regeneration mode (with and without post injection) and the corresponding DOC efficiency. The experimental results shown in Figure 7 highlight that the main differences between regeneration and normal operating mode particle distribution could be attributed to the HC generated by the post injection. Furthermore, it should also be noticed that when the engine is operated in regeneration mode without post injection, the particle number distribution maintains the same trend shown in the normal operating mode, but with a remarkable reduction (up to 50%) of the particle concentration values, due to the different combustion process (and in particular to the after injection, which enhances the incylinder soot burn-out). As far as the effects of the after-treatment system on the particle number and mass distributions are concerned, the experimental results shown in Figures 12 and 13 demonstrate remarkable differences from the normal operating conditions: the particle number, which, under normal operating conditions remained almost unchanged across the DOC (see Figures 4 and 5), during DPF regeneration shows a relevant reduction for particles larger than 40 nm; the number distribution peak shifts down from 100 to 40 nm, with a reduction in concentrations of 1 order of magnitude. A further drop in particle number is produced, as expected, by the DPF. Similar remarks can also be made for the mass distribution. Furthermore, it should also be pointed out that the particle number and mass distribution at the DPF outlet change their shape from mono- to bimodal. Further investigations, by means of a differential mobility analyzer which allows the analysis of

Figure 16. Comparison between the engine outlet particle number (top) and mass (bottom) size distributions under normal operating conditions with diesel fuel (B0) and with neat FAME (B100) (linear y axis scales, 1750 rpm/3.5 bar bmep).

particle diameters from 3 to 62 nm, will be necessary to better clarify this issue. The DOC and DPF filtration efficiencies (defined on a particle number basis) are shown in Figures 14 and 15: it is worth noting that, thanks to the high temperature levels attained during the regeneration event (see Figure 10), the DOC is capable of effectively oxidizing unburned hydrocarbons (see Figure 11), thus significantly contributing to a remarkable reduction in particle number. 3.3. Fuel Effects: Neat FAME Usage. Finally, the particle number and size distribution at the engine outlet were also assessed fueling the engine with neat FAME to evaluate the impact of alternative fuels on PM characteristics. 2656


Industrial & Engineering Chemistry Research The experimental results of this preliminary investigation, reported in Figure 16, show a comparison between the engine outlet particle number and mass size distribution under normal operating conditions with diesel fuel (B0) and with neat biodiesel (B100). The reduction, in terms of particle number, that can be achieved by means of the neat FAME usage is about 80%, with a slight shift of the particle number peak from 50 to 40 nm, as also demonstrated in previous studies.7-12 Neat biodiesel usage allows even more significant benefits to be obtained in terms of mass size distribution, with reductions that can exceed 90%.

4. CONCLUSIONS The particle number and size distribution from a small displacement Euro 5 common rail automotive diesel engine have been investigated by sampling the exhaust gases at the engine outlet, at the DOC outlet, and at the DPF outlet, to fully characterize not only the engine out PM emissions, but also their behavior throughout the whole exhaust system. The particle number and size distributions were evaluated at part load operating conditions which are representative of urban driving, both under normal operating conditions and during the DPF regeneration mode, to highlight the impact of the different combustion processes on the PM characteristics. Under normal operating conditions, the engine and DOC outlet particle number and mass size distributions appeared to be very similar, without any appreciable effect of the DOC, while the DPF exhibited high filtration efficiency values on a particle number basis, even in the nanoparticle range. This has to be expected considering that the DOC only affects the gaseous phase of the exhaust gases, and does not interact with the soot fraction of the PM. Moreover, the temperature reached by the DOC at low engine load is below its light-off value, thus limiting the DOC effect on gaseous hydrocarbons. As far as the behavior of the DPF under normal operating conditions is concerned, the filter reduced the number of emitted particles by 2 or 3 orders of magnitude, regardless of the particle diameter. The regeneration mode caused a particle number increase of 1 order of magnitude, with a substantial shift of the number distribution peaks toward larger diameters, passing from 50 to 100 nm. Moreover, thanks to the higher temperatures attained in the DOC, and to the high hydrocarbon oxidation efficiencies, the particle number, which under normal operating conditions remained almost unchanged across the DOC, showed a remarkable reduction during DPF regeneration for particles larger than 40 nm, with a reduction in concentrations of 1 order of magnitude. Other tests, conducted by modifying the regeneration injection strategy, i.e., by eliminating the post injection, have highlighted that the main differences between the regeneration and normal operating mode particle distribution could be attributed to the HC generated by the post injection. Interestingly, DPF shows a high conversion efficiency, even when it has been freshly regenerated. The DPF efficiency remains extremely high during regeneration, even when only the deep bed filtration mechanism is available. Finally, the impact of alternative fuels on the PM characteristics has also been evaluated by fueling the engine with neat FAME: remarkable reductions, both in terms of particle number and mass, have been achieved (up to 80% and 90%, respectively) under normal operating mode conditions. In conclusion, when using the actual after-treatment configuration, the deep bed filtration of the DPF results to be a suitable

ARTICLE

filtration mechanism, while it would be convenient to increase the HC conversion efficiency of DOC at low temperatures. It has been demonstrated that, under regeneration mode, the DPF still maintains a high filtration efficiency and that the particle number is also reduced across the DOC, due to unburned hydrocarbons oxidation; thus an increase of the HC conversion efficiency of the DOC, when it is operating at low temperatures, would be beneficial for the regeneration mode.

’ AUTHOR INFORMATION Corresponding Author

*Telephone: þ39-011-090-4710. Fax: þ39-011-090-4699. E-mail: debora.fi[email protected].

’ DEFINITIONS bmep = brake mean effective pressure DI = direct injection DOC = diesel oxidation catalyst DPF = diesel particulate filter FAME = fatty acid methyl ester FSN = filter smoke number HC = unburned hydrocarbons Kr = Kripton N = measured number of particles PM = particulate matter rpm = revolutions per minute SMPS = scanning mobility particle sizer WG = waste gate ’ REFERENCES (1) 2008/692/EC. Implementing and amending Regulation (EC) N. 715/2007 of the European Parliament and of the Council on typeapproval of motor vehicles with respect to emissions from light passenger and commercial vehicles (Euro 5 and Euro 6) and on access to vehicle repair and maintenance information; July 18, 2008. (2) Dementhon, J. B.; Martin, B. Influence of Various Diesel Traps on Particulate Size Distribution. SAE Technical Paper 972999, 1997. (3) Abdul-Khalek, I. S.; Kittelson, D. B. Diesel Trap Performance: Particle Size Measurements and Trends. SAE Technical Paper 982599, 1998. (4) Raux, S.; Forti, L.; Barbusse, S.; Plassat, G.; Pierron, L.; Monier, R.; Momique, J. C.; Pain, C.; Dionnet, B.; Zervas, E.; Rouveirolles, P.; Dorlhene, P. French Program on the Impact of Engine Technology on Particulate Emissions, Size Distribution and Composition Heavy Duty Diesel Study. SAE Technical Paper 2005-01-0190, 2005. (5) Giechaskiel, B.; Munoz-Bueno, R.; Rubino, L.; Manfredi, U.; Dilara, P.; De Santi, G. Particle Measurement Programme (PMP): Particle Size and Number Emissions Before, During and After Regeneration Events of a Euro 4 DPF Equipped Light-Duty Diesel Vehicle. SAE Technical Paper 2000-01-1944, 2007. (6) Merkisz, J.; Pielecha, J. Gaseous and Particle Emissions Results from Light Duty Vehicle with Diesel Particle Filter. SAE Technical Paper 2009-01-2630, 2009. (7) Tan, P.; Hu, Z.; Lou, D.; Li, B. Particle Number and Size Distribution from a Diesel Engine with Jatropha Biodiesel Fuel. SAE Technical Paper 2009-01-2726 , 2009. (8) Moon, G.; Lee, Y.; Choi, K.; Jeong, D. Experimental study on characteristics of diesel particulate emissions with diesel, GTL, and blended fuels. SAE Technical Paper 2009-24-0098, 2009. (9) Beatrice, C.; Di Iorio, S.; Guido, C.; Mancaruso, E.; Vaglieco, B. M. Alternative Diesel Fuels Effects on Combustion and Emissions of an Euro4 Automotive Diesel Engine. SAE Technical Paper 2009-240088, 2009. 2657



ARTICLE

(10) Maunula, T.; Kinnunen, T. Effect of Oxygen Containing Biofuels on the Emissions with Exhaust Gas Catalysts. SAE Technical Paper 2009-01-2737, 2009. (11) Bielaczyc, P.; Szczotka, A. The Effect of Pure RME and Biodiesel Blends with High RME Content on Exhaust Emissions from a Light Duty Diesel Engine. SAE Technical Paper 2009-01-2653, 2009. (12) Karavalakis, G.; Stournas, S. Regulated and Unregulated Emissions of a Euro 4 SUV Operated with Diesel and Soy-based Biodiesel Blends. SAE Technical Paper 2009-01-2690, 2009. (13) Abdul-Khalek, I., Kittelson, D. The Influence of Dilution Conditions on Diesel Exhaust Particle Size Distribution Measurements. SAE Technical Paper 1999-01-1142, 1999. (14) Abdul-Khalek, I. S.; Kittelson, D. B.; Graskow, B. R.; Wei, Q. Diesel Exhaust Particle Size: Measurement Issues and Trends. SAE Technical Paper 980525, 1998. (15) L€uders, H.; Kr€uger, M.; Stommel, P.; L€uers, B. The Role of Sampling Conditions in Particle Size Distribution Measurements. SAE Technical Paper 981374, 1998.

2658


Detailed Investigation on Soot Particle Size ... - ACS Publications

Recommend Documents