Vous êtes sur la page 1sur 6

2426

Energy & Fuels 2006, 20, 2426-2431

Interlaboratory Study of the Exhaust Gas Particle Number Measurement Using the Condensation Particle Counter (CPC)
E. Zervas,*, P. Dorlhe ` ne, L. Forti, C. Perrin, J. C. Momique, R. Monier, H. Ing,| and B. Lopez|
Renault, 1, Alle e Cornuel, 91510 Lardy, France, Institut Franc ais du Pe trole (IFP), 1 et 4 AVenue du Bois Pre au, 92500 Rueil-Malmaison, France, PSA Peugeot Citro en, 18, rue des FauVelles, 92250 La Garenne-Colombes, France, and Union Technique de lAutomobile, du Motocycle et du Cycle (UTAC), Autodrome de Linas-Montlhe ry, 91310 Montlhe ry, France ReceiVed April 27, 2006. ReVised Manuscript ReceiVed September 7, 2006

Four laboratories, Institut Franc ais du Pe trole (IFP), PSA Peugeot-Citroe n, Renault, and Union Technique de lAutomobile, du Motocycle et du Cycle (UTAC), have conducted an interlaboratory test to evaluate the performances of the condensation particle counter (CPC). The technical program was based on tests carried out on four passenger cars, one gasoline and three diesel ones, tested on the New European Driving Cycle (NEDC). The regulated pollutants are also measured, as indicators of test repeatability and good working conditions. The 1.96RSD reproducibility values of the total particle number determined by CPC on NEDC are 87, 58, 41, and 220%, respectively, for the Euro4 gasoline, Euro3 diesel, Euro4 diesel, and Euro3 diesel plus DPF (diesel particulate filter) vehicles. The CPC repeatability and reproducibility are poorest at lower particle numbers. The CPC performances, in terms of reproducibility, are comparable but generally not much better than the reproducibility of particulate matter emissions determined by the gravimetric method.

Introduction Currently, there are many methods for the determination of particle number and/or distribution. The most common methods used in the case of vehicle exhaust gas are the electrical lowpressure impactor (ELPI),1-4 scanning mobility particle sizer (SMPS),5 and particle counters,6,7 but many others are also reported. A description of several analytical instruments can be found in the works of Burtscher8,9 and Mohr et al.10 SMPS has a very good particle size resolution; however, the quite long time resolution (some minutes) of this instrument is insufficient for transient particle measurement on the New European Driving Cycle (NEDC). ELPI has a sufficient time
* To whom correspondence should be addressed. Telephone: +331-76-87-84-77. Fax: +331-76-87-82-92. E-mail: efthimios.zervas@ renault.com. Renault. Institut Franc ais du Pe trole (IFP). PSA Peugeot Citro en. | Union Technique de lAutomobile, du Motocycle et du Cycle (UTAC). (1) Keskinen, J.; Pietarinen, K.; Lehtimaki, M. J. Aerosol Sci. 1992, 23, 353-360. (2) Khalek, I. A. SAE Technical Paper Series 2000, 2000-01-2001. (3) Witze, P. O.; Chase, R. E.; Maricq, M. M.; Podsiadlik, D. H.; Xu, N. SAE Technical Paper Series 2004, 2004-01-0964. (4) Zervas, E.; Dorlhe ` ne, P.; Daviau, R.; Dionnet, B. SAE Techical Paper Series 2004, 2004-01-1983. (5) Wang, S. C.; Flagan, R. C. Aerosol Sci. Technol. 1990, 13, 230240. (6) Willeke, K.; Baron, P. A. Aerosol Measurement, Principles, Techniques and Applications; Van Nostrand Reinhold: New York, 1993. (7) Hinds, W. C. Aerosol Technology; J. Wiley and Sons: New York, 1999. (8) Burtscher, H. Tailpipe Particulate Emission Measurement for Diesel Engines. Swiss contribution to GRPE PMP, 2001, http://wanda.fhaargau.ch/ iss/veroeffentlichungen/tailpipe-emissionmeasurement.html. (9) Burtscher, H. J. Aerosol Sci. 2005, 36, 896-932. (10) Mohr, M.; Lehman, U.; Rutter, J. EnViron. Sci. Technol. 2005, 39, 2229-2238.

resolution (1 s), but its size resolution is lower than that of SMPS (ELPI has only 12 stages for the particles from 7 nm to 10.24 m, while SMPS has 128 channels for the particles from 10 to 500 nm). Another, more recent instrument, is the engine exhaust particle sizer spectrometer (EEPS), which has a sufficient time resolution for measurements on cycles and a size resolution between that of ELPI and SMPS. The performances of this instrument are presented elsewhere.11 If the particle-size distribution is ignored, the total exhaust particle number on steady speeds and cycles can be measured by a condensation particle counter (CPC),12-14 which can determine the total particle number but not the size distribution of exhaust particles. The detailed operation principle of CPC is reported in several studies.6,7,9,10,15,16 The present work presents an interlaboratory study for the evaluation of CPC performances. Four laboratories, Institut Franc ais du Pe trole (IFP), PSA Peugeot Citroe n, Renault, and Union Technique de lAutomobile, du Motocycle et du Cycle (UTAC), participated in this study. The repeatability and reproducibility of the total particle number of a gasoline and three diesel passenger cars (PCs), measured by CPC on the NEDC, are evaluated and discussed here. The definitions of repeatability and reproducibility are given in the Experimental Section. The repeatability and reproducibility are examined on the entire NEDC and also at the urban (UDC, Urban Driving
(11) Zervas, E.; Dorlhe ` ne, P. Aerosol Sci. Technol. 2006, 40, 977-984. (12) ACEA program on emissions of fine particles from passenger cars 2, Report, July, 2002, www.acea.be. (13) Schaberg, P. W.; Zarling, D. D.; Waytulonis, R.W.; Kittelson, D. B. SAE Technical Paper Series 2002, 2002-01-2727. (14) Anderson, J.; Preston, H.; Warrens, Ch.; Brett, P. SAE Technical Paper Series 2004, 2004-01-1989. (15) Pui, D. Y. H.; Chen, D. Direct-Reading Instruments for Analyzing Airborne Particles. In Air Sampling Instruments; ACGIH: Cincinnati, OH, 2001; Chapter 15, pp 377-414. (16) Kasper, M. SAE Technical Paper Series 2004, 2004-01-0960.

10.1021/ef060180j CCC: $33.50 2006 American Chemical Society Published on Web 10/17/2006

Exhaust Gas Particle Number Measurement Using CPC

Energy & Fuels, Vol. 20, No. 6, 2006 2427

Figure 1. Experimental setup of this study.


Table 1. Main Characteristics of the Vehicles Useda gasoline vehicle inertia (kg) fuel emission limits after-treatment devices Renault Clio 1020 gasoline Euro4 TWC diesel1 Peugeot 307 1360 diesel Euro3 DOC diesel2 Renault Laguna 1360 diesel Euro4 DOC diesel3 Peugeot 607 1600 diesel Euro3 DOC plus PF

a TWC, three-way catalyst; DOC, diesel oxidation catalyst; DPF, diesel particulate filter.

Cycle) and extra-urban (EUDC, Extra-Urban Driving Cycle) parts of the cycle. More details about NEDC and its parts can be found in the European Directive 70/220.17 ELPI is also used, to allow for a comparison of the results of the present study with the results of the interlaboratory exercise conducted previously to determine the performances of ELPI.18 CPC measurements of several engines or vehicles are reported by other authors, such as Lehman et al.,19 Mohr et al.,20 and Kinsey et al.21 However, the above three authors worked on heavy-duty engines and in one laboratory only; therefore, they do not report reproducibility values. Lehman et al.19 gives some repeatability results; however, because this author worked on heavy-duty engines, his results cannot be directly compared with the results of this study. Experimental Section
Four PCs were used in this study: a Euro4 gasoline PC operating under stoichiometric conditions (called gasoline), a Euro3 diesel PC (called diesel1), a Euro4 diesel PC (called diesel2), and a Euro3 diesel PC equipped with a diesel particulate filter (DPF, called diesel3). All four vehicles are tested in all four laboratories, except the diesel3 vehicle, which, for technical reasons, is tested by only three laboratories. Table 1 shows the main characteristics of these vehicles. Fuels with less than 10 ppm of sulfur were used for this study. The same lubricant, containing <0.4% of sulfur,
(17) Directive 70/220, www.europa.eu.int. (18) Zervas, E.; Dorlhe ` ne, P.; Forti, L.; Perrin, C.; Momique, J. C.; Monier, R.; Ing, H.; Lopez, B. Aerosol Sci. Technol. 2005, 39, 333-346. (19) Lehmann, U.; Mohr, M.; Schweizer, T.; Rutter, J. Atmos. EnViron. 2003, 37, 5247-5259. (20) Mohr, M.; Forss, A. M.; Lehmann, U. EnViron. Sci. Technol. 2006, 40, 2375-2383. (21) Kinsey, G. S.; Mitchell, W. A.; Squier, W. C.; Lina, K.; King, F. G.; Logan, R.; Dong, Y.; Thompson, G. J.; Clark, N. N. J. Aerosol Sci. 2006, 37, 63-87.

was used for all four vehicles. A total of 25 ppm of commercially used Ce-based additive was added in the fuel in the case of the DPF-equipped diesel vehicle to decrease the necessary temperature for the DPF regeneration from about 650 to 550 C. For each vehicle, three tests were performed on the NEDC (cold start) and regulated pollutants and CO2 emissions were measured according to current European regulations.17 For the particulate matter (PM) measurements, a backup filter is also used. The dilution tunnel configuration of each laboratory is presented in another work.18 No high-efficiency particulate air (HEPA) filters were used to filtrate the dilution tunnel air. Because the target of the current work is to determine the reproducibility of CPC under real industrial laboratory conditions, no modification or improvement on the experimental facilities is performed. The particle number was measured using a TSI 3022A CPC,22 calibrated by the manufacturer. A DEKATI ELPI, covering the particle cut size from 7 nm to 10 m, was also used. Two additional DEKATI diluters were used in series to dilute the sample gas. The first diluter was put in front of CPC and ELPI [diluted 10 times and heated at 120 C with hot nitrogen or air (no difference is found between the two diluters)], to avoid nucleation phenomena. The second diluter was put in front of CPC (to achieve a total dilution of 100 times, not heated). Each laboratory employed its own CPC and ELPI; however, the same experimental configuration is used in the four laboratories. Because ELPI performances are presented in a previous work,18 only the CPC repeatability and reproducibility and the relations between the two instruments are presented here. Figure 1 shows the experimental setup of this study. Tunnel background (blanks) were recorded during 20 min before and after each test. The particle number of the tunnel background tests is expressed in km-1 using the same CVS volume and distance as a NEDC test. These values were not subtracted from the particle numbers of vehicle measurements. The repeatability and reproducibility between the four laboratories are calculated following the ISO 5725 standards.23 In this work, the reproducibility and repeatability is expressed as 1.96RSD (relative standard deviation, for a confidence interval of 95%). The repeatability is defined as the precision when the same test is conducted in the same laboratory, by the same operator, using the same method and equipment and during a short time interval. The reproducibility is the precision, when the same test is conducted in different laboratories, by different operators, using the same method but different equipment.23 The calculations used for the determination of repeatability and reproducibility can be found elsewhere.18,23 It must be noted that a poor repeatability express results
(22) TSI, Inc. Model 3022A Condensation Particle Counter. Instruction Manual; TSI, Inc.: St. Paul, MN, 2002. (23) ISO 5725-2. AFNOR, Paris, France, 1994.

2428 Energy & Fuels, Vol. 20, No. 6, 2006

ZerVas et al.

Figure 3. Distribution of exhaust particles measured by ELPI. Average, on the NEDC, value of the diesel1 vehicle measured in lab2 (three NEDC tests presented with bold, normal, and dashed line).

Figure 2. Average emissions of regulated pollutants and CO2 and corresponding repeatability and reproducibility values.

with high dispersion and means that the 1.96RSD value is high and vice versa. Because the object of this work is to evaluate the CPC performances for industrial application, the repeatability and reproducibility values shown in this paper include the variability of the vehicle, the experimental conditions, and the measurement instrument.

Results and Discussion Emissions of Regulated Pollutants and CO2. Figure 2 shows the emission of CO, HC, NOx, PM, and CO2 of the four vehicles tested. For each vehicle, the emissions of regulated pollutants are within the regulatory limits. The 1.96RSD repeatability and reproducibility values of CO, HC, NOx, and PM are generally low, except in the case of PM of the gasoline and the diesel3 PCs (Figure 2). This last result is due to their very low PM emission levels, which are close to the tunnel background results.18 For the same type of vehicle, the repeatability and reproducibility values of these pollutants are of the same order of magnitude with the values reported in the ELPI interlaboratory study.18 Figure 2 shows that the vehicles used have quite repeatable emissions. Nucleation Particles. The target of this work is to measure only solid particles. To achieve it, particular attention is given on the nucleation phenomena, because volatile particles can be formed, altering the measurement of the particle total number measured by CPC. The formation of nucleation particles is controlled by ELPI measurements. Figure 3 shows the average distribution of the diesel1 vehicle measured in lab2, where it can be seen that no nucleation occurs, because the first ELPI stage has low particle numbers. The results of the other vehicles follow the same trends as ELPI. This figure shows that the experimental configuration used in this study prevents nucleation phenomena.

Figure 4. Particle number of the four vehicles on the NEDC, measured by CPC. Average value of all tests.

Particle Number on the NEDC. Figure 4 shows the particle number versus time, measured by CPC and ELPI for the four vehicles used. Except in the beginning of the cycle where very few particles are detected, the diesel3 vehicle has quite a flat curve, because DPF collects all particles emitted. For the first three vehicles, the particle number decreases with time at the urban part of the cycle, especially at idle conditions, as the engine warms up. This is not observed in the case of the diesel3 vehicle, where the particle number is remarkably constant with time. However, this is due to the extremely low emissions, because almost all particles emitted are collected on the DPF. The particle number increases at the extra-urban part of the cycle (EUDC) for the first three vehicles, while DPF collects all emitted particles in the case of the fourth one. The particle number measured by CPC and ELPI on the NEDC has the same shape for all four vehicles used. However, CPC numbers are generally slightly lower than the ELPI ones, expect sometimes at the point of 120 km/h. In the case of the diesel3 vehicle, there is almost an order of magnitude between

Exhaust Gas Particle Number Measurement Using CPC

Energy & Fuels, Vol. 20, No. 6, 2006 2429

Figure 5. Particle number of the four vehicles on the NEDC as a function of speed.

Figure 6. CPC average particle number versus ELPI average particle number measured on the NEDC, for the four vehicles. The dashed bold red lines correspond to y ) x.

the CPC and ELPI numbers after the initial peak. This is probably due to a small nucleation that occurs at ELPI measurements and that disappears at the CPC ones because of the second diluter. Generally, the particle number emitted from the first three vehicles increases with speed (Figure 5); however, the emissions of the DPF-equipped diesel vehicle are independent of speed, because almost all particles are collected on the DPF. The particle number determined by CPC versus the particle number determined by ELPI (average values on NEDC for the four vehicles tested) is shown in Figure 6. It can be seen that there is a linear relationship between these values. The slope of the best fit line is around 0.86-0.99, indicating that, under the experimental configuration used, CPC measures 1-14% lower particle numbers than ELPI. However, the values of 1-14% are within the repeatability limits of CPC and ELPI measurements (Figure 9 and ref 18), indicating that the results

of the two instruments are similar. The points of the diesel3 vehicle form a block because the particle number is very similar at almost all points of the NEDC. Total Particle Number. Figure 7 shows the total particle emissions of the four vehicles tested. The average particle number of these vehicles is of the same order of magnitude of the values already reported in the literature.4,12,18 CPC and ELPI determine quite similar total particle numbers: 4.2 1012 and 5.3 1012 km-1 measured by ELPI and CPC, respectively, in the case of the gasoline vehicle, 1.3 1014 and 9.2 1013 km-1 in the case of the diesel1 one, 1.3 1014 and 7.4 1013 km-1 in the case of the diesel2 one, and 9.1 1010 and 3.4 1010 km-1 in the case of the diesel3 one. It must be noted that the particulate emissions of the diesel2 vehicle are representative of the emissions of the current new European PCs and that the emissions of the diesel3 vehicle are representative of the emissions of the future European PCs. As already reported,4,18 the gasoline vehicle has 1-2 orders of magnitude lower emissions than the diesel1 and diesel2 vehicles and the diesel3 vehicle has 1-2 orders of magnitude lower emissions than the gasoline one. The mean particle number per kilometer emitted by the gasoline vehicle is 2 (measured by ELPI) to 6 (measured by CPC) times higher at the EUDC than at the UDC (Figure 7). The ratio EUDC/UDC is about 0.5-0.7 in the case of the diesel1 and diesel2 vehicles, while it is only about 0.2 in the case of diesel3. The above statements indicate that the gasoline vehicle emits the majority of its particles at high speeds, while this is not the case for the older and current technology diesel vehicles. The diesel3 vehicle has the opposite trend (higher particle number emissions at the UDC), because of the small particle emissions in the beginning of the cycle when the DPF is still empty and the extremely low emissions at the rest of the NEDC. Because the DPF-equipped vehicle has very low emissions, near the background measurements,18 the particle number of the tunnel backgrounds is also measured to determine the CPC reproducibility at very low particle numbers, free of vehicle variations. The particle number of the tunnel backgrounds is very low, from 5 108 to 2 1011 km-1 (Figure 8). The same order of magnitude is reported in a previous paper.18 The particle number of blank measurements is quite a bit higher than the limit of detection of CPC (LOD 2 108 km-1). The LOD calculation is linked to the measurement setup used in this study. CPC Repeatability and Reproducibility. The CPC repeatability and reproducibility is analyzed in this paragraph. Because of the very low particle numbers, the 1.96RSD repeatability and reproducibility values of the blank measurements determind by CPC are quite high and can be even more than 300% (Figure 8). This is due to the very low numbers of the tunnel background tests. Because of the higher particle numbers of the vehicle emissions than the tunnel backgrounds, the repeatability and reproducibility values of the exhaust particle number of all four vehicles is better than the corresponding values of the tunnel background tests (Figure 9). The 1.96RSD repeatability and reproducibility values of the gasoline vehicle are quite high (23 and 87%, Figure 9) because of low particle numbers.4 The corresponding values of the diesel1 and diesel2 vehicles are quite low (10 and 58% for the diesel1 and 6 and 41% for the diesel2, Figure 9) because of the higher particle numbers. The repeatability and reproducibility values of the diesel3 vehicle (67 and 220%, Figure 9) are higher than the values of the previous three vehicles because of the very low particle numbers. It must be noted that repeatability values also depend upon the

2430 Energy & Fuels, Vol. 20, No. 6, 2006

ZerVas et al.

Figure 7. Particle number measured by CPC and ELPI of the four PCs used.

Figure 9. CPC 1.96RSD repeatability and reproducibility values of the four vehicles used. Only three laboratories tested the diesel three PC.

Figure 8. Tunnel background experiments. Particle number determined by CPC and repeatability and reproducibility of these measurements.

tests performed previously, because of the dropping of the deposed particles,24 but this phenomenon must be taken into consideration in the case of industrial measurements. The repeatability and reproducibility values of the gasoline vehicle are higher at the UDC than the entire NEDC, while
(24) Andrews, G. E.; Clarke, A. J.; Rojas, N. Y.; Gregory, D.; Sale, T. SAE Technical Paper Series 2000, 2000-01-0514.

those values are quite similar on the entire NEDC and the urban and extra-urban part of the cycle in the case of the three diesel vehicles (Figure 9). The 1.96RSD repeatability and reproducibility values of ELPI are quite similar to the values reported in the ELPI interlaboratory study.18 Correlations between 1.96RSD Repeatability and Reproducibility Values and the Particle Number. There is a correlation between the 1.96RSD repeatability and reproducibility values and the particle number determined by CPC (Figure 10). It is clear that that repeatability and reproducibility become poor at low particle numbers. The 1.96RSD repeatability

Exhaust Gas Particle Number Measurement Using CPC

Energy & Fuels, Vol. 20, No. 6, 2006 2431

determination of PM is compared with the repeatability of CPC measurements, one can observe that the CPC repeatability is roughly twice better than the repeatability of the gravimetric method in the case of the low particle-emitted vehicles (gasoline and diesel3 ones) and quite similar to 4 times better in the case of the other two. However, the corresponding reproducibility values do not show the same trends: CPC reproducibility is quite similar to the reproducibility of gravimetric PM determination in the case of gasoline and diesel3 vehicles and 2-3 times poorest in the case of diesel1 and diesel2 vehicles. This last point indicates that CPC performances are generally comparable but not much better than the current gravimetric method. A similar trend is observed in the case of ELPI measurements.18 Conclusions Four laboratories, IFP, PSA Peugeot-Citroe n, Renault, and UTAC, conducted an interlaboratory test on the determination of the exhaust particle number using CPC. Four PCs, one gasoline operating under stoichiometric conditions and three diesel ones, are tested on the NEDC. The results of this study show that (1) the reproducibility of the tunnel background particle numbers is quite poor, because of low numbers measured; (2) on the entire NEDC, the 1.96RSD reproducibility values of the total particle number determined by CPC are 87, 58, 41, and 220%, respectively, for the Euro4 gasoline, Euro3 diesel, Euro4 diesel, and Euro3 diesel vehicles equipped with DPF; (3) the CPC repeatability and reproducibility are poorest at lower particle numbers; (4) CPC and ELPI give quite similar spectra of the particle number versus time on the NEDC; and (5) the CPC performances, in terms of reproducibility, are comparable but generally not much better than the reproducibility of PM emissions determined by the gravimetric method.
EF060180J

Figure 10. CPC 1.96RSD repeatability and reproducibility values as a function of the average particle number of all vehicles and tunnel background tests.

and reproducibility values of CPC are higher than 100% for particle numbers around 1 1011 km-1, indicating that repeatability and reproducibility are critical for the particle number measurement of future vehicles. The same trends were found in the case of ELPI repeatability and reproducibility.18 Comparison between the Reproducibility and Repeatability of Regulated Pollutants and Particle Number Measurements Using CPC. Figure 10 shows the reproducibility and repeatability of PM emissions and particle numbers determined by CPC. When the repeatability of the gravimetric

Vous aimerez peut-être aussi