INTRODUCTION EXPERIMENTAL

Transcription

1 Overview of the European "Particulates" Project on the Characterization of Exhaust Particulate Emissions from Road Vehicles: Results for Heavy Duty Engines Neville Thompson CONCAWE Leonidas Ntziachristos, Zissis Samaras Laboratory of Applied Thermodynamics / Aristotle University Thessaloniki Paivi Aakko VTT, Finland Urban Wass Volvo Technology, Sweden Stephan Hausberger Technical University of Graz, Austria Theodor Sams AVL List GmbH, Graz Austria Copyright 2004 SAE International ABSTRACT This paper presents an overview of the results on heavy duty engines collected in the PARTICULATES project, which aimed at the characterization of exhaust particle emissions from road vehicles. The same exhaust gas sampling and measurement system as employed for the measurements on light duty vehicles [1] was used. Measurements were made in three labs to evaluate a wide range of particulate properties with a range of heavy duty engines and fuels. The measured properties included particle number, with focus separately on nucleation mode and solid particles, particle active surface and total mass. The sample consisted of 10 engines, ranging from Euro-I to prototype Euro-V technologies. The same core diesel fuels were used as in the light duty programme, mainly differentiated with respect to their sulphur content. Additional fuels were tested by some partners to extend the knowledge base. Engine tests were mainly conducted over the standard European regulatory cycles (ESC and ETC), although additional steady state conditions, including some offcycle points, were also assessed. All data (both real time and integrated) were collected in a common data base and centrally analyzed, using common formats and methodologies in order to eliminate inconsistencies and optimize comparability. As for light duty vehicles, the results show that particulate emissions from heavy duty engines are markedly reduced by advanced technologies, most notably by the combination of particulate traps and sulphur-free fuels. However, particulate emissions patterns are also shown to be influenced by operating conditions; in particular fuel sulphur effects are most obvious under high temperature operation. The study provides evidence that particulate number measurement offers the potential for greater sensitivity in evaluating particulate emissions. It demonstrates that the PARTICULATES dedicated sampling procedure is capable of delivering repeatable results even in the case of the unstable nucleation mode, also for heavy duty engines. The data provide a first step towards emission factors based on particle size and number for heavy duty engines. However it should not be forgotten that nucleation mode particles are highly dependent on sampling conditions. Further research continues to be needed on the health relevance of measurements of nucleation mode particles, their chemical composition and their fate in the atmosphere.

2 INTRODUCTION For several years, particulate mass (PM) emission levels from light and heavy duty diesel vehicles have been decreasing as a result of the implementation of continuously more stringent emission standards worldwide. In Europe, the regulatory limits on particulate mass (PM) emissions from heavy-duty engines tighten through Euro-III, IV, and V limits [2]. The PM limit will decrease by a further 80 % from the current limit by the year Even further reductions in particulate emissions are under discussion, though the potential for further reductions in mass emissions is diminishing and other particle properties such as size, number and surface area are increasingly under discussion. New engine and exhaust after-treatment technologies are needed to meet the future emissions targets, in particular those for PM and NOx. Sulphur-free fuels are being introduced to enable the widest range of engine and after-treatment technologies to be employed. Those engine and after-treatment technologies considered most likely to be used to meet Euro-IV and Euro-V limits, i.e. SCR/urea, EGR and particle traps, were selected for evaluation in this programme. NOx traps were not yet available for evaluation. New measurement methodologies for particle emissions may be needed for very low emissions levels. Improved mass measurement, as well as new methods to measure particle size and number, are being developed. When discussing particle size, the definitions accumulation and nucleation mode particles are widely used. There is no precise boundary between the nucleation and accumulation mode particles, although sizes between 30 and 50 nm have been used as cutpoints. The larger accumulation mode particles are mainly carbonaceous in nature and account for the majority of the particulate mass, whereas nucleation mode particles are believed to be comprised predominantly of volatile material, sulphate and heavy hydrocarbons and may dominate the total number emissions. The presence of nucleation mode particles has been related to the concentration of carbon and hydrocarbons in the exhaust. Under conditions where carbon emission is reduced, there is a greater tendency to produce nucleation mode particles [3]. The extent of this formation has been shown to be dependent on engine and fuel technology, the use of after-treatment, operating conditions and also strongly linked with sampling and measurement conditions [4]. In this direction, the European Commission launched a three year research program, PARTICULATES, on the characterization of exhaust aerosol from light and heavy duty vehicles. The project initially assessed the existing evidence and methodologies on particulate emissions [5] and concluded that, for this research program, both solid accumulation mode and volatile nucleation mode" particles should be assessed. The nucleation mode presented a particular challenge which extended the methodology development phase. Nevertheless, the project completed in September 2003 and delivered a methodology for detailed sampling and analysis of collected aerosol from all vehicle types, capable of measuring both modes of particles from a minimum size range of ca 7 nm [6]. Based on this methodology and the associated measurement protocol, a significant number of light and heavy duty diesel vehicles/engines and gasoline cars were measured. This paper presents an overview of the information collected from the tests on heavy duty engines. The results on light duty vehicles were reported in [1]. It is expected that findings from this project will contribute to efforts at an international level on the sampling and understanding of exhaust particle emissions. The methodologies and results supplement other work in the field, such as the UK DETR/SMMT/CONCAWE program [4, 7, 8, 9], relevant projects sponsored by CRC in US [3], ACEA particulates projects [10, 11] and the UN-ECE Particle Measurement Program [12]. This paper tries to assist in considerations of next steps in particulate emissions policy development, by providing new experimental data on several aspects of particulate emissions. EXPERIMENTAL ENGINE MATRIX Heavy duty engine testing was performed in 3 individual laboratories: AVL List GmbH, Graz, Austria (Lab a) Volvo, Sweden (Lab b) VTT, Finland (Lab c) The test sample consisted of a range of engines from Euro-I to prototype Euro-IV and Euro-V systems, with advanced after-treatment such as SCR/urea and EGR plus particulate filters. Table 1 presents the details of the engines tested, also identifying the fuels on which they were tested (further details on the fuels are given in the following section). Several heavy duty diesel vehicles were also tested in the PARTICULATES programme (by the Technical University of Graz, Austria) in order to relate the engine data to real world driving. Evaluation of this data is beyond the scope of this paper and will be reported separately. FUEL MATRICES The core test fuels were selected based on the objectives of PARTICULATES to develop input on representative emission factors for current and future vehicle fleets as well as to enhance understanding of fuel effects. Existing knowledge indicated fuel sulphur as a key fuel effect on particle emissions, both in terms of enabling new exhaust after-treatment technology and as a direct effect on sulphate emissions. Consequently, the updated EU Fuels Directive requires the introduction of sulphur-free fuels (10 mg/kg max sulphur) starting in 2005 [13]

3 Table 1: Engine / vehicle characteristics Table 2: Main Diesel Fuel Properties Engine / After-treatment Model year Cubic capacity [dm 3 ] engine speed [kw/rpm] FIE type Lab a: engines EURO-III /1800 unit inj. Fuels D2, D3, D4, D5, D6, D7 D3, D4, D5, D7 D3, D4, D5, D7 prototype + SCR (EURO-V) /1800 unit inj. prototype + CRT /1900 unit (EURO-IV) inj. Lab b: engines EURO-I /1900 DI D1, D4, D5 EURO-III /1800 DI D1, D4, D5 EURO-III + CRT /1800 DI D4, D5 Lab c: engines EURO-II /2000 DI D2, D4, D3, D5 EURO-II + CRT /2000 DI D4, D3 EURO-III /1900 DI, EDC EURO-III + SCR /1900 DI, EDC D2, D4, In view of the above, the test fuels were mainly designed around the sulphur effect, using a base fuel with other properties held as close as possible to average year 2000/05 levels (given the available blending components), but with sulphur content as low as possible. The sulphur levels of this diesel fuel were adjusted to historic, current, 2005 and 2009 levels by doping with di-tertiarybutyl-disulphide (fuels coded: D1-D4). An additional sulphur-free fuel, Swedish Environmental Class 1 (D5), was also tested in order to assess any further potential benefits from extreme changes to other fuel properties. Fuel D1, with 1550 mg/kg sulphur, was only tested in the Euro-1 engine to provide an example of emissions levels that may have been expected from engines and fuels typical of that era. Some additional fuels were also tested by lab a: D6 representing typical pre-2000 diesel fuel and D7 containing 5% of rape seed methyl ester (RME). The analytical data on the test fuels are shown in Table 2. LUBRICANT It was also expected that lubricating oil may have an effect on the characteristics of particles emitted from each vehicle. In order to minimize any lubricant effects, a common batch of lubricant was used for the program. A typical high volume, conventional mineral oil formulation was used for all vehicles, meeting the following specifications: D4 Fuel Code D1 - D4 D5 D6 D7 Description Sulphur Swedish EN590: 5% pre- RME Matrix Class in D4 Cetane Number Cetane Index Density (kg/m 3 ) T50 ( C) T95 ( C) FBP ( C) CFPP ( C) C (mm 2 /s) Poly-aromatics (% m/m) 4.3 < Mono-aromatics (% m/m) Carbon (% m/m) Hydrogen (% m/m) H:C ratio 1.82 : : : : 1 LHV (MJ/kg) HFRR (µm) FAME Nil Nil Nil 5% v/v Sulphur (mg/kg) D D D-3 38 D-4 8 ENGINE TEST PROCEDURES: The test conditions for the heavy-duty engines included the standard EU regulatory test cycles: European Steady-state Cycle (ESC) and European Transient Cycle (ETC), plus selected steady-state points. The ESC and ETC heavy-duty test cycles are described in the European legislation [2]. The ESC is comprised of 13 steady-state loads, and the final result is the average calculated with weighting factors (Figure 1). In addition to idle, the other 12 modes are set at a combination of the three speeds established above and at 25%, 50%, 75% and 100% load. Weighting factors are assigned to each mode. Viscosity type: 15W-40 ACEA Class A3/B3 Sulphur content: 6000 ppm w/w

4 Additional modes determined by certification personnel Six extended steady-state (SS) loads were selected for the heavy-duty engine measurements. Two loads were selected from the ESC (modes 5 and 12) and one load from the former European (ECE R49) test cycle (10% load, intermediate engine speed). In addition, three offcycle loads were selected: road load at speed 50/50 between ESC A/C, 25% load at speed ESC A-10% and 50% load at 50% speed. The road load was not available for all engines as it can be defined only with a known vehicle weight. This paper focuses on the results obtained from the regulatory ESC and ETC tests and only includes selected results from the steady-state tests. DAILY TEST SEQUENCE: Previous work had indicated that consistent engine/exhaust system conditioning is critical to producing repeatable results for particulate emissions. A standard daily test sequence was therefore followed in order to provide comparable information at the various measurement points. In summary, the sequence was: Figure 1: Schematic Figure of the 13-mode ESC Warm-up and condition on the test fuel Dummy ESC Test ESC Test ETC Steady state points PARTICULATES SAMPLING AND MEASUREMENT The main emphasis of the programme was on particulate emissions characterization. Standard regulated PM emissions were compared with other particulate properties using the sampling system developed and applied for the characterization of exhaust aerosol that has been presented in detail elsewhere [6]; its basic outline is shown in Figure 3. One of the aims of the particulates sampling system is to provide sampling conditions which favour nucleation mode formation and this is achieved by using a moderate primary dilution ratio (12.5:1) and cooling of the exhaust with conditioned air at 32 C. Figure 2: Schematic Figure of the ETC The transient ETC is based on certain torque and speed values. Figure 2 shows how these values represent the vehicle speed to give a more realistic view on the test. The ETC is divided into three sub-parts (urban, rural and motorway). However, the limit values for regulated emissions apply to total cycle Silica gel Charcoal Filter MFC CONTROLLERSP UNIT ST ET EP DT CT Exhaust gas CO2 anal. (raw) On/off valve CVS Pressure regulators 3 bar Cooling agent CO2 anal. (dil) ELPI CPC Ejector dilutors Ageing chamber TD SMPS or DMA/CPC Dilution Air Line Sample Line DC DGI Bypass Trap Flowmeter Pump Figure 3: Sampling system used for characterization of exhaust aerosol Measurements of both accumulation mode and nucleation mode particles were achieved by splitting the diluted exhaust gas (porous tube diluter) into two: the wet and dry branches. A thermodenuder was used (250 C) in the dry branch to remove volatile materials.

5 An ageing chamber was used in the wet branch to provide time for nucleation mode growth. Additional diluters (ejector-type) were used in both branches as necessary to help protect equipment and to bring the particle concentration into a convenient, measurable range. The extent of dilution depended on the level of particulate emissions of the engines/vehicles examined (Euro-I to Euro-V, trap-equipped etc ). An Electrical Low Pressure Impactor (ELPI) was used in the dry branch to measure solid particle number count and size distribution. In the wet branch, a Condensation Particle Counter (CPC) was used to measure total number count and a diffusion charger (DC) to measure the active particle surface area. A five stage gravimetric impactor (DGI) was also used in the wet branch to define the mass-weighted size distribution. In the steady-state tests a Scanning Mobility Particle Sizer system (CPC+DMA) was used instead of CPC to measure the number-weighted size distribution from the wet branch. The sampling conditions, parameters, dilution air properties etc. were defined with tight limits, e.g. the primary dilution ratio limit was 12.5±2, even in the transient tests. The selected parameters, settings, tolerances, the round-robin and other measures to validate the method are described in reference [6]. Based on the sampling system developed, Table 3 provides a summary of the aerosol information that is collected over a test. In addition to these, PM and regulated gaseous pollutants were measured during the tests following the reference procedure based on the CVS. Only the particulate emissions are discussed in this paper. For practical reasons, one partner was not able to include the full PARTICULATES measurement suite, but did include additional measurements by Dual Differential Mobility Particle Sizer (DDMPS) at steady state conditions and Transient Differential Mobility Particle Sizer (TrDMPS) at transient conditions. These provide a cross-check on the particle size and number measurements with the PARTICULATES system. DATA HANDLING AND QUALITY ASSURANCE The large number of data collected over a single engine/fuel test together with the multiple laboratories involved, made the data handling and processing a demanding task. In order to eliminate inconsistencies that might arise from using different instrument software versions, calculation parameters and post processing, a central application was developed which was used to store data and make all necessary calculations for all instruments involved in the different labs. A flowchart of the data handling process is presented in Figure 4. The individual files with data from each instrument, the lab instrumentation and the vehicle and fuel specifications are stored in a coded database format. A central application performs all calculations to produce integrated values over a cycle, starting with primary instrument signals and taking into account dilution ratios, flowrates, the necessary synchronization between the instruments and the test cycle, and particle losses in the sampling system. The processed information is stored in an indexed database structure which allows easy extraction of any information necessary. Finally, a post-processing application enables the creation of charts for visualization of the data and statistical processing of the information. Table 3: Instrument Condensation Particle Counter (CPC) Scanning Mobility Particle Sizer (SMPS) Aerosol information collected over a test Property Particle number concentration Particle sizing and concentration Size resolution One channel >7 nm 64 channels per decade nm or nm Temporal resolution 1 s (transients) 90 s (steady states) Original Recordings Instruments Vehicle/Fuel Data Dynamometer CVS Coded database structure Calculation File Particle Concentration Synchronization Losses Correction Electrical Low Pressure Impactor (ELPI) + thermodenuder (TD) Diffusion Charger (DC) Gravimetric Impactor (DGI) Solid particle sizing and concentration Active surface Mass-based particle sizing First 8 channels considered with filter stage 7nm - 1 µm One channel 7nm - 1 µm 5 stages <10 µm 1 s (transients) 1 s (transients) Integral over a test Post Processor Data extraction Charts Statistics Figure 4: Data handling process flowchart Results Database Great care was taken to increase the comparability between the different laboratories involved by setting up similar sampling systems, utilizing an identical sampling protocol and developing a common application to - 5 -

6 process all data. Nevertheless, a certain degree of variability has to be expected in the data collected from different test laboratories. Ntziachristos et al. [6] demonstrated that the expected reproducibility between labs in light duty diesel vehicle tests is in the range of 20-30%. This should be considered as a reference baseline reproducibility. Obviously, there might be individual cases where larger differences between particular instruments and/or laboratories may exist. Such a range of variability should be taken into account when considering the significance of different effects. RESULTS In this paper we try to draw general conclusions concerning the effect of engine technology, fuel and operating conditions on the emissions, rather than focusing on the individual performance of each engine. In this respect we identify the different engine technologies, according to their emissions certification level and after-treatment system, viz: Euro-I, Euro-II, Euro-III, Euro-II+CRT, Euro-III+CRT, Euro-IV+CRT, Euro-III+ SCR and Euro-V+SCR. The fuels are identified according to their codings as described earlier. In order to establish the absolute levels of different properties of particle emissions, the following sections include plots of the particulate measurements of all engines, as listed in Table 2, tested over the standard ESC and ETC tests. Each point on each plot corresponds to the emissions of a single engine over a single test on the particular cycle using the corresponding fuel in the abscissa. The following sections provide the results for each particle property measurement, and attempt to identify the effects of different engine and fuel technologies (over the regulated cycles) in order to compare particle emissions of the different technologies when following the type-approval procedure. REGULATED PM EMISSIONS The standard regulated PM emissions are considered first. Figures 5 and 6 show the ESC and ETC data respectively Effect of vehicle technology Clear progress in the reduction of particulate mass emissions is evident from Figures 5 and 6. Over the ESC, the Euro-I engine produced by far the highest mass emissions, followed by the Euro-II and Euro-III engines which performed similarly. Addition of a PM trap to Euro-II and Euro-III engines reduced PM emissions, especially when operated on 10 mg/kg sulphur fuel. However, the lowest PM emissions were achieved with the specifically designed Euro-IV engine with EGR and CRT. A Euro-V engine designed with an SCR/urea system, but without a particulate filter achieved almost as low PM emissions as the Euro-IV engine with CRT. A Euro-III engine retrofitted with an SCR/urea system for NOx after-treatment showed no change in its PM emissions level. Regulated PM [g kwh -1 ] Figure 5: Regulated PM [g kwh -1 ] Figure 6: ETC EURO I I + CRT I + CRT EURO IV + CRT I + SCR EURO V + SCR D1 D2 D3 D4 D5 D6 D7 Regulated PM over the ESC cycle I ESC + CRT I + CRT EURO IV + CRT I + SCR EURO V + SCR D2 D3 D4 D5 D6 D7 Regulated PM over the ETC cycle Over the ETC, similar trends were observed, although one Euro-III engine produced very high emissions on this cycle. It should however be noted that such Euro-III engines were only certified on the ESC, not on the ETC. Those engines which were designed for the ETC achieved the regulatory limits. Effect of fuel Clear effect of fuel sulphur content (from 1550 mg/kg to 10 mg/kg) is evident in the Euro-I engine results. The effect of sulphur reduction from 300 mg/kg (D2) was also significant in the Euro-III engines

7 especially over the ESC where higher temperatures are reached. The Euro-II engine which was post fitted with a CRT, showed a difference between 50 and 10 mg/kg sulphur fuels (D3 vs. D4) indicating that considerable sulphate formation occurred with this design. On the other hand, the Euro-IV engine with CRT and the Euro-V engine with SCR/urea gave similar PM emissions with both 50 and 10 mg/kg sulphur fuels. Swedish Class 1 fuel (D5) showed an additional benefit over fuel D4 in the Euro-I to Euro-III engine technologies, of the order of 10%. The reduction in PM with fuel D5 cannot be directly related to any single fuel property; compared to fuels D2-D4, D5 has lower density, lighter distillation characteristics, very low aromatics and negligible poly-aromatics content. However, the effect is generally consistent with the findings of EPEFE [14]. In the more advanced Euro-IV and Euro-V prototypes, no additional benefit for Swedish Class 1 over conventional 10 mg/kg sulphur fuel was observed. The addition of 5% RME made no significant difference to the PM emissions. Effect of test cycle There is a clear influence of test cycle on the PM emission level, depending on the engine technology and fuel sulphur content. Fuel sulphur effects are more evident over the ESC due to the higher temperatures experienced on some operating modes. Engine technology effects are also evident; in particular one Euro-III engine produced very high PM emissions on the ETC cycle, although as noted earlier, this engine was not designed for this cycle. Overall - It is evident that Euro-IV and Euro-V engine technologies with advanced after-treatment systems operating on sulphur-free (10 mg/kg sulphur max) fuels should bring dramatic improvements in PM emissions. This represents a much larger step than the steps taken so far from Euro-I to Euro-III. The introduction of a transient test cycle into the certification process also brings an additional level of control versus the former steady state only testing. SOLID PARTICLE NUMBER Changes in regulated particulate mass might be expected to be reflected in changes in solid accumulation mode (carbonaceous) particles. The ELPI, used with a thermodenuder (TD), on the dry branch, provides a direct measure of these carbonaceous particles, so these data are considered next. Results are shown for the ESC and ETC tests in Figures 7 and 8. As a result of the wide spread of data, particle number graphs are shown on a logarithmic scale. Solid particle emission rate [ km -1 ] 1.0E E E+09 EURO I I + CRT I + CRT EURO IV + CRT I + SCR EURO V + SCR D1 D2 D3 D4 D5 D6 D7 ESC Figure 7: Integrated solid particle number emission rate over the ESC cycle Solid particle emission rate [ km -1 ] 1.0E E E+09 I +CRT I+CRT EURO IV+CRT I+SCR EURO V+SCR D2 D3 D4 D5 D6 D7 ETC Figure 8: Integrated solid particle number emission rate over the ETC cycle Total concentration of "solid" particles was recorded with an Electrical Low Pressure Impactor (ELPI) which samples downstream of a thermodenuder (TD). "Solid" should be taken here as a convention for those particles not evaporated at TD temperatures up to 250 C and hence heavy semi-volatiles might also be included. We report here results collected over the first seven impactor stages, corresponding to aerodynamic diameters 30 nm 1 µm. Results have also been corrected for solid particles losses in the TD, although it is recognized that correction for TD losses may in itself generate variability since the losses may vary from engine to engine and - 7 -

8 cycle to cycle. The absolute number concentration of the ELPI should be considered indicative because it also depends on the particle effective density [15] and no corrections have been performed for this. Although correction for the effective density may change the absolute concentration of solid particles reported in the following sections, we decided not to do that at this stage but report emissions assuming unit particle density, as is common in all relevant diesel exhaust characterization studies. SIZE SEGREGATED PARTICULATE MASS Size segregated particulate mass was collected by two of the labs using a gravimetric impactor (DGI) as described earlier. These data were only obtained over the ETC and selected extended steady states, but not over the ESC, due to complexities of switching between the short duration ESC steady state points. Typical examples of the ETC data are shown in Figure 9. Effect of vehicle technology Conventional Euro-I to Euro-III engine technologies generally produced total solid particle number emissions of the order of particles/kwh. One of the Euro-III engines showed total solid particle number emissions almost an order of magnitude lower over the ESC and this still requires further explanation. The best trap-equipped systems can be seen to offer the potential to reduce total solid particle numbers by some 3-4 orders of magnitude when operated on sulphur-free fuels. This is consistent with the results found for light duty vehicles. % mass collected in each stage ETC overall DGI Stage 1 <0.2 µm DGI Stage µm DGI Stage µm Lab c EURO-II Lab c EURO-III Lab b EURO-III DGI Stage µm DGI Stage 5 >2.5 µm The Euro-V system with SCR/urea (but without a particle trap), produced around particles/kwh, ca. 90% less than the typical Euro-III cases. However, this remains some 2 orders of magnitude higher than the best trapequipped Euro-IV engine, despite approaching the trap results for regulated particulate mass. The Euro-III system with a retro-fitted SCR/urea system gave, as expected, total solid particle emissions close to its Euro-III base case. Effect of fuel No consistent fuel effects were seen in the data from the dry carbon particles. Fuel effects seen in particulate mass were not reflected in numbers of carbonaceous particles as measured by the ELPI. Effect of test cycle Total solid particle number emissions were at about the same order of magnitude for both the ESC and ETC tests. Similar trends between engine technologies were apparent. Overall Solid particle emission rates from Euro-I to Euro-III engines are in order of particles/kwh. Particulate trap-equipped engines operating on sulphurfree fuels offer the potential to reduce this by ca. 3 orders of magnitude. Conclusions on vehicle technology effects were generally similar to those for particulate mass, with the exception that the number count showed the capability to discriminate between the Euro-IV trap system and the Euro-V non-trap system. In this regard, ELPI measurement of dry particle number potentially offers additional information over and above the particulate mass method. Figure 9: over ETC DGI Data. Size segregated particulate mass The main message from this chart regards the general mass / size distribution. The largest portion of the total mass collected is in the range below 0.2 µm; 90% of the mass is of a size below 0.5µm. The particulate size distribution profile will be further examined later in the paper based on SMPS data, although it should be acknowledged that the two techniques measure size distribution based on different physical principles of separation. ACTIVE SURFACE AREA In engineering terms, active surface is a surrogate for the specific particle surface area which is available for interaction with the environment, i.e. for gas-to-particle and particle-to-particle interactions. Active surface is recorded with diffusion chargers (DCs) and in this case it corresponds to particles in the range 7 nm 1 µm. Only two partners were able to conduct the active surface area measurements with comparable equipment so a smaller data-base is available. Nevertheless, several conclusions can still be drawn from the data illustrated in Figures 10 and 11. As described earlier, active surface was measured from the wet branch. Effect of vehicle technology Euro-I to Euro-III engines produced values for active surface in the range 10 5 to 10 6 cm²/km. The Euro-III engine with CRT gave 1-2 orders of magnitude reduction, broadly in-line with its performance on ELPI. The Euro-II engine with CRT produced active surface values in the same range as the - 8 -

9 Euro-I to Euro-III conventional engines, indicating that the reduction in solid particle number was counterbalanced by formation of significant numbers of nucleation mode particles. This will be explored further in the following sections. although for the Euro-II engine with CRT, higher emissions were seen over the ESC, indicating a greater tendency to form nucleation mode particles at the higher temperature conditions which occur on certain modes of the ESC test. Active Surface Area [cm 2 kwh -1 ] 1.0E E E E E+02 Figure 10: 1.0E+06 ESC EURO I I + CRT I + CRT I + SCR D1 D2 D3 D4 D5 D6 D7 Integrated active surface area over ESC ETC TOTAL PARTICLE NUMBER Total particle number count, including the volatile particles, is considered next. For all the measurements reported in this section, samples were taken from the wet branch or directly from the CVS. Different measurement techniques were used depending on engine operation: within the PARTICULATES measurement system - SMPS for steady state and CPC for transient. Lab a also made measurements with DDMPS at steady state and TrDMPS over transient operation. Particle number counts may be expected to differ slightly between instruments due to the different size ranges measured. CPC DATA - Figures 12 and 13 show data generated using the CPC over the ESC and ETC cycles Effect of vehicle technology Total particle (CPC) emissions of conventional Euro-I to Euro-III heavy duty diesel engines are in the range to particles/km; somewhat higher than the ELPI data due to the contribution from the volatile, nucleation mode particles. The Euro-II engine showed higher total particle emissions than the Euro-III engines, due to higher nucleation mode particle formation. Active Surface Area [cm 2 kwh -1 ] 1.0E E+04 I + CRT I + CRT 1.0E+03 I + SCR 1.0E+02 D2 D3 D4 D5 D6 D7 Figure 11: Integrated active surface area over ETC Particulate trap-equipped engines operating on low sulphur fuels are shown to have the capability to reduce the total number count by 2-3 orders of magnitude. However, the Euro-II + CRT case showed total CPC emissions similar to the Euro-II engine over the ETC and higher than the Euro-II engine over the ESC, indicating a high level of nucleation mode particle production, particularly at the higher temperatures which occur on certain modes of the ESC test. Effect of fuel The influence of fuel sulphur content can be seen in several of the data-sets. It is particularly striking for engines equipped with CRTs where nucleation mode particle production on higher sulphur fuels can bring total number emissions up to levels equal or above those of non-trap engines. This is consistent with the active surface data described earlier. Effect of fuel The impact of fuel sulphur reduction could be observed as well as some effect from Swedish Class 1. Fuel sulphur effect was particularly evident from the high sulphur Euro-I engine fuel (D1) and on the Euro-II + CRT engine case over the ETC. Effect of test cycle Similar levels of active surface area were generally produced over the ESC and ETC tests, - 9 -

10 1.0E+16 ESC deeper research into fuel and vehicle effects than simple mass measurements. Total particle emission rate [ km -1 ] 1.0E E+10 EURO I I + CRT I + CRT I + SCR D1 D2 D3 D4 D5 D6 D7 Figure 12: Integrated total particle number emission rate over the ESC Total particle emission rate [ km -1 ] 1.0E E E+10 I + CRT I + CRT EURO IV + CRT I + SCR EURO V + SCR ETC D2 D3 D4 D5 D6 D7 Figure 13: Integrated total particle number emission rate over the ETC Effect of test cycle The main effect of test cycle observed here is the higher tendency for nucleation mode particle production at higher temperature conditions, which occur on certain modes of the ESC. Overall Given the known instability of the nucleation mode particles, fairly good repeatability is demonstrated in the CPC data, at least for the conventional engines. The variability in the data with the trap-equipped engines is a concern but is not completely unexpected given knowledge on the instability of the nucleation mode. Nevertheless, the CPC data does offer potential for The comments on total particle number count and the impact of nucleation mode particles can be examined further by examining the particle size distributions, described in the following sections. PARTICLE SIZE DISTRIBUTIONS The SMPS can provide information on the number/size distribution of the total particles emitted during steady state operation. Figure 14 shows the composite SMPS data for the ESC, obtained by lab a, for Euro-III, IV and V engines. These data are split to show the total particle counts below 30 nm and above 30 nm separately. All fuel/engine combinations were tested 3 times in a randomized block test design and statistically analysed. Lab a also carried out additional measurements with a DDMPS, essentially a twin SMPS with an alternative dilution system (same DR with direct dilution of raw exhaust with ambient air) for comparison. These data are shown in Figure 15. The error bars shown on Figures 14 and 15 have been constructed so that nonoverlapping error bars would indicate a difference between those fuel/engine combinations at 95% significance. Mean data were calculated on a geometric basis. In general, trends in engine and fuel effects between the two measurement systems were similar, although the DDMPS system measured a somewhat higher particle count. In part this can be attributed to its ability to measure to a slightly lower particle size. Fuel effects on particles <30nm were greater than those on particles >30nm suggesting that most of the fuel influence is on the nucleation mode. This is consistent with the observed effect of sulphur. The observed trends in the total particle counts are dominated by the trends in small particles (<30nm). Particles >30nm (as with the ELPI) showed little sensitivity to fuel. In the Euro-III engine, the high sulphur fuels, D2 and D6, gave the highest number of particles below 30nm, showing the benefit of sulphur reduction for nucleation mode particles. The Euro-IV engine, equipped with the CRT produced the lowest number of accumulation mode particles. However, the use of after-treatment did not reduce the number of particles below 30nm (nucleation mode)

11 1.0E+15 HD engine test results (ESC) (geometric means) Overall the combination of advanced engines/aftertreatment systems and lower sulphur fuels demonstrated clear benefits over the current European regulatory situation of Euro-III engines and 350 mg/kg max sulphur fuels. SMPS, N, kwh-1 1.0E+10 N < 30nm N 30nm N < 30nm N 30nm N < 30nm N 30nm Euro-III Euro-IV Euro-V Engine Figure 14: SMPS distributions of Euro-III to Euro-IV engines from lab a over the ESC cycle, split into particles <30nm and particles 30 nm D2 D3 D4 D5 D6 D7 Lab a also carried out measurements over the ETC with a TrDMPS. This data is shown in Figure 16. In contrast with the results from the ESC tests, the Euro-IV engine equipped with a CRT reduced the numbers of particles both above and below 30nm. The lower number of smaller particles probably reflects the lower operating temperature of the ETC, as well as perhaps the difference in the dilution system from the standard PARTICULATES set-up, which has been optimized to maximize the potential to see nucleation mode particles. Under these conditions, even the 300 mg/kg sulphur fuels were not differentiated. DDMPS, N, kwh-1 1.0E E+15 HD engine test results (ESC) (geometric means) N < 30nm N 30nm N < 30nm N 30nm N < 30nm N 30nm Euro-III Euro-IV Euro-V Engine Figure 15: DDMPS distributions of Euro-III to Euro-IV engines from lab a over the ESC cycle, split into particles <30nm and particles 30 nm D2 D3 D4 D5 D6 D7 dn/dlogdm (#/kwh) 1E+16 1E+15 1E+14 1E+13 1E+12 1E+11 1E+10 Euro II Euro I Euro IIIa Euro IIIb and IIIc Euro IIIb+DPF Euro II + DPF Proto+DPF 1E Mobility diameter (nm) 50% load speed ESC A Figure 17: SMPS distributions of selected engines/fuels at steady speed, 50% load. Fuel used: D4 TrDMPS, N, kwh-1 1.0E E+10 HD engine test results (ETC) (geometric means) N < 30nm N 30nm N < 30nm N 30nm N < 30nm N 30nm Euro-III Euro-IV Euro-V Engine Figure 16: TrDMPS distributions of Euro-III to Euro-IV engines from lab a over the ETC cycle, split into particles <30nm and particles 30 nm D2 D3 D4 D5 D6 D The SMPS distribution data from other partners showed generally similar trends in technology effects. Two examples of SMPS distribution profiles for all engines (for fuel D4) on two extended steady state points are shown in Figures 17 and 18. These show a fairly consistent size distribution profile for the accumulation mode for the conventional diesel engines. At high load conditions the appearance of a nucleation mode is also evident (even on the low sulphur fuel D4), especially on the trap-equipped Euro-II engine. The overall level of SMPS particle number emissions for the accumulation mode measured by lab a was about an order of magnitude lower that that measured by the other two labs. This has not been fully explained, but is likely to lie in calibration/calculation issues rather than being a real engine difference. Clearly inter-laboratory calibration issues would have to be further addressed if such

12 methodologies were to be used for other than research purposes. dn/dlogdm (#/kwh) 1E+16 1E+15 1E+14 1E+13 1E+12 1E+11 1E+10 Euro II+DPF Euro II Euro I Euro IIIb Euro IIIc Euro IIIa Euro IIIb+DPF Proto+DPF 1E Mobility diameter (nm) 75% load speed ESC C Figure 18: SMPS distributions of selected engines/fuels at steady speed, 75% load. Fuel used: D4 SUMMARY AND CONCLUSIONS The new PARTICULATES measurement protocol, has been successfully applied in 3 different laboratories to the testing of heavy duty diesel engines on a range of fuels, providing information on a wide range of particulate emission properties. The measured properties included particle number, with particular focus on both nucleation and accumulation modes, particle active surface and particle mass-based size distributions, in addition to the current regulated particulate mass. The heavy duty engine sample tested consisted of 10 engines, from Euro-I to advanced prototype Euro-IV and Euro-V systems. A wide range of diesel fuels were evaluated, with particular emphasis on sulphur content in view of the current progress in legislation and relevance for emission factors. The data generated provides a clear step towards one of the main objectives of PARTICULATES which was to extend emissions factors beyond particulate mass into other properties of particulates. Data in this paper gives clear indications of the levels of particulate numbers and active surface that may be expected with various levels of engine technologies and fuels. The effect of fuel sulphur was greatest under high speed/temperature operation. Under these conditions, lower sulphur fuels reduced both particle mass and number emissions. A heavy duty prototype Euro-V engine equipped with SCR/urea, but without a particle trap, produced very low particulate mass, within the Euro-V limits, but its particle number emissions remained considerably higher than the trap-equipped option. In Euro-III engine technology, two 300 mg/kg sulphur fuels with widely different fuel compositions gave similar particulate emissions. In the more advanced engine technologies, fuel effects (other than sulphur) on particulate emissions were small in absolute terms. The addition of 5% RME to a 10 mg/kg sulphur diesel fuel made little difference to particulate emissions levels. Particle mass measurement is capable of distinguishing between engine technology levels up to trap-equipped systems. Its continued use in regulation has the advantage of providing continuity with previous data. Particle number measurement techniques offer the potential for greater measurement sensitivity and discrimination, and are of particular value for further research and development into cleaner vehicles and fuels. There is some evidence that the number of solid particles emitted does not always correlate with PM mass. However, further methodology development, including definition of suitable instrument calibration procedures and standards and multi-lab validation exercises would be required prior to use of solid particle number standards in regulation. In this programme, both solid accumulation mode and volatile nucleation mode particles have been successfully measured under laboratory operation. However, it must be remembered that nucleation mode particles are highly dependent on sampling conditions. Further research continues to be needed on the health relevance of measurements of nucleation mode particles, their chemical composition and their fate in the atmosphere. ACKNOWLEDGMENTS This work has been conducted in the framework of the EC DG TREN "PARTICULATES" project (Contract No RD.11091). The authors would like to thank all partners for a constructive collaboration and significant contribution in all stages of the project development. Heavy duty diesel engines equipped with particulate traps produced very low particulate mass emissions, low numbers of carbonaceous particles and low total numbers of particles when operating on low sulphur fuels

13 REFERENCES 1. Ntziachristos, L.; Mamakos. A.; Samaras, Z.; Mathis, U.; Mohr, M.; Thompson, N.; Stradling, R.; Forti, L.; de Serves, C. Overview of the EU PARTICULATES Project on the Characterisation of Exhaust Particulate Emissions from Road Vehicles. Results for light duty vehicles. SAE Paper Directive 1999/96/EC of the European Parliament and of the Council of 13 December 1999 on the approximation of the laws of the Member States relating to measures to be taken against the emission of gaseous and particulate pollutants from compression ignition engines for use in vehicles, and amending Council Directive 88/77/EEC. Official Journal of the European Communities No. L044, Kittelson, D.; Watts, W.; Johnson, J. Diesel Aerosol Sampling Methodology CRC E-43. Final report. University of Minnesota. Department of Mechanical Engineering. 8/19/ Andersson, J.D.; Wedekind B.G.A. Final report on the UK DETR/SMMT/CONCAWE Particulate measurement programme. Ricardo. May Marjamäki, M.; Keskinen, J Vehicle exhaust particulates characterization, properties, instrumentation and sampling requirements. Deliverable 2 of the "PARTICULATES" project, Internet reference at particulates. 6. Ntziachristos, L.; Giechaskiel, B.; Pistikopoulos, P.; Samaras, Z.; Mathis, U.; Mohr, M.; Ristimäki, J.; Keskinen, J.; Mikkanen, P.; Casati, R.; Scheer, V.; Vogt, R. Performance evaluation of a novel sampling and measurement system for exhaust particle characterization. SAE Paper Andersson, J.D.; Wedekind, B.G.A.; Hall, D., Stradling, R.; Barnes, C.; Wilson, G DETR/SMMT/CONCAWE Particle Research Programme: Sampling and Measurement Experiences. SAE Paper Andersson, J.D.; Wedekind, B.G.A; Hall, D.; Stradling, R.; Barnes, C.; Wilson, G DETR/SMMT/CONCAWE Particle Research Programme: Heavy Duty Results. SAE Paper Andersson, J.D.; Wedekind, B.G.A.; Hall, D.; Stradling, R.; Wilson, G DETR/SMMT/CONCAWE Particle Research Programme: Light Duty Results. SAE Paper ACEA ACEA Programme on Emissions of Fine Particles from Passenger Cars, December 1999, Brussels, Belgium. 11. ACEA ACEA Programme on Emissions of Fine Particles from Passenger Cars [2], July 2002, Brussels, Belgium. 12. UNECE Programme Overview: GRPE Particle Measurement Programme. UNECE WP29/GRPE Group, GRPE 42 nd, 29 May 1 June European Parliament and Council, Directive 2003/17/EC of the European Parliament and of the Council of 3 March 2003 amending Directive 98/70/EC relating to the quality of petrol and diesel fuels, Official Journal, L 076, pp ACEA and EUROPIA, European Programme on Emissions, Fuels and Engine Technologies (EPEFE). Final Report, Brussels, Belgium. 15. Ristimäki, J.; Virtanen, A.; Marjamäki, M.; Rostedt, A.; Keskinen, J On-line measurement of size distribution and effective density of submicron aerosol particles, Journal of Aerosol Science, Vol. 33, pp CONTACT Zissis Samaras, Professor Laboratory of Applied Thermodynamics Aristotle University GR Thessaloniki Greece zisis@auth.gr Web: ADDITIONAL SOURCES Additional information on the results and technical details of the PARTICULATES project may be found at DEFINITIONS, ACRONYMS, ABBREVIATIONS CRC CPC CRT CVS DC DDMPS DGI DPF DR EGR ELPI ESC ETC FIE LHV OEM PM PMP RME SCR SMPS TrDMPS TD Coordinating Research Council Condensation Particle Counter Continuously Regenerative Trap Constant Volume Sampling Diffusion Charger Dual Differential Mobility Particle Sizer Gravimetric Impactor Diesel Particle Filter (equipped vehicles) Dilution Ratio Exhaust Gas Recirculation Electrical Low pressure Impactor European Steady-state Cycle European Transient Cycle Fuel Injection Equipment Lower Heating Value Original Equipment Manufacturer Particulate mass collected in the CVS Particle Measurement Programme Rapeseed Methyl Ester Selective Catalytic Reduction (with urea) Scanning Mobility Particle Sizer Transient Differential Mobility Particle Sizer Thermodenuder