On-Road and Chassis Dynamometer Evaluations of Emissions from

INTRODUCTIONEuro 6, the next stage of European emissions regulations for light-duty vehicles, becomes mandatory from 1 September 2014 for the Type Approval of new passenger cars and the smaller light commercial vehicles (categories M and N1 Class 1), and 1 year later for all other commercial vehicles not

exceeding 3.5 tonnes maximum mass (categories N1 Class II and N1 Class III). One year after these dates, Euro 6 becomes mandatory for all new registrations.

A number of papers and reports have raised concerns over the emissions performance - especially diesel NOx emissions - of current and future light duty-vehicles in real, and in particular

On-Road and Chassis Dynamometer Evaluations of Emissions from Two Euro 6 Diesel Vehicles

Jon AnderssonRicardo UK

John May, Cecile Favre, and Dirk BosteelsAECC

Simon de Vries, Matthew Heaney, Matthew Keenan, and Jonathon MansellRicardo UK

ABSTRACTThe exhaust emissions of two Euro 6 diesel cars with different emissions control systems have been evaluated both on the road and over various chassis dynamometer test cycles. European emissions limits are currently set using the New European Driving Cycle (NEDC), but the European Commission is preparing additional test procedures to ensure that emissions are well controlled both in real-world use and over the legislative test cycle. The main focus of this work on ‘Real Driving Emissions’ (RDE) is on measurements using Portable Emissions Measurement Systems (PEMS) in truly representative, on-road, driving. A key focus of the test programme, undertaken as a collaboration between AECC (the Association for Emissions Control by Catalyst) and Ricardo UK, was therefore the use of PEMS systems to measure on-road emissions of both gaseous pollutants and particulate matter. This included measurement of particle number emissions with a new candidate system for this type of measurement.

The results from this testing are compared with emissions measured over four different chassis dynamometer test cycles - the current legislative cycle (New European Driving Cycle, NEDC); the Common Artemis suite of test cycles (CADC) that is widely used in emissions modelling; the new Worldwide Light-duty Test Cycle (WLTC) defined by the UN Working Party on Pollution and Energy (GRPE) as part of the development of the Worldwide harmonised Light vehicles Test Procedure (WLTP); and a set of cycles produced by a Random Cycle Generator based on ‘short trip’ segments from the EU database used to construct WLTC. This Random Cycle (RC) approach was originally considered by the European Commission and Member States as an alternative to PEMS measurements, but remains an option for particle number measurement should PEMS instrumentation not be fully available by the time such measurements are required by legislation.

The aim of this test programme was to evaluate the emissions performance of two different vehicles when using PEMS systems in real-life driving, and to identify and help understand the differences in emissions that may arise between the various test procedures. The test results show the differing vehicle emissions performances that may be encountered in real-driving for two vehicles that both meet the Euro 6 limits in Type-Approval laboratory testing. The results show that in some cases the on-road PEMS-measured emissions for complete test routes can exceed Type-Approval limits, and indicate that real-world conditions require more robust control strategies over those needed to satisfy existing legislation.

CITATION: Andersson, J., May, J., Favre, C., Bosteels, D. et al., "On-Road and Chassis Dynamometer Evaluations of Emissions from Two Euro 6 Diesel Vehicles," SAE Int. J. Fuels Lubr. 7(3):2014, doi:10.4271/2014-01-2826.

2014-01-2826Published 10/13/2014

Copyright © 2014 SAE Internationaldoi:10.4271/2014-01-2826

saefuel.saejournals.org

Downloaded from SAE International by Cecile Favre, Monday, October 13, 2014

http://dx.doi.org/10.4271/2014-01-2826

http://saefuel.saejournals.org/

urban, driving conditions, compared to the emissions results obtained in the legislative test procedure on a chassis dynamometer [1,2,3]. The European Commission has therefore announced in its ‘CARS 2020 Action Plan’ [4] and its Clean Air Programme for Europe [5] its intention to include an additional test for ‘Real Driving Emissions of Light-Duty Vehicles (RDE-LDV) from the start of Euro 6. The Commission has indicated [6] that this will take the form of on-road emissions measurements using Portable Emissions Measurement Systems (PEMS) and the EU's Joint Research Centre (JRC) has been leading the development of appropriate procedures [7,8]. It is planned that test procedures for RDE-LDV will be implemented in the Euro 6 Regulation from 2014, with the introduction of ‘Conformity Factors’ proposed from the Euro 6c stage starting in 2017.

AECC has previously conducted chassis dynamometer tests emissions over various chassis dynamometer test cycles [9, 10], but these tests did not include the RDE-LDV proposals. To fill this knowledge gap, AECC and Ricardo collaborated on a project conducting emissions tests on two modern light-duty diesel cars. The main aims of the work were to help identify and understand the differences in emissions that may arise between the various test procedures including real driving emissions measurements using PEMS, and to understand the practical and experimental aspects of using these measurement systems. Both of the cars were Type Approved to the Euro 6b emissions requirements and incorporated a Diesel Particulate Filter for PM and PN control. To assist in understanding how different vehicle systems might respond, one of the vehicles used a urea-SCR system for NOx control, selected as being typical of most current Euro 6 vehicles. The other vehicle provided a view of alternative emission control architecture by its use of a dual EGR to avoid using NOx-specific aftertreatment.

WORK SCOPEThe programme of work proceeded in a number of stages:

• Two vehicles were identified and located in continental Europe and shipped to the Ricardo UK Shoreham Technical Centre.

• Following PEMS manufacturer-supported installation of their system on the SCR vehicle, a correlation between the on-board and lab-based emissions measurement systems was undertaken.

• After investigation and resolution of any PEMS measurement issues, the on-road measurement campaign was conducted.

• The PEMS was then switched to the dual-EGR vehicle and a second measurement campaign, on the same on-road routes, was conducted.

• In parallel, the SCR vehicle's exhaust system was instrumented and it was then subjected to chassis dyno tests.

• To complete the scheduled test programme, chassis dyno tests were also conducted on the dual-EGR vehicle

• The instrumented exhaust systems were then removed, replaced with identical new OEM systems, and the vehicles returned to the market.

Test Vehicles and Emissions Control SystemsThe two light-duty diesel test vehicles employed in this work were sourced from the European market, with both certified to Euro 6b standards. The two vehicles used alternative approaches for meeting the regulatory NOx targets: the SCR vehicle was selected as an early application of Selective Catalytic Reduction (SCR) into the European market; the dual-EGR vehicle identified as employing combustion and engine hardware (both long-route and short-route Exhaust Gas Recirculation (EGR)) measures for NOx reduction, and both employed Diesel Particulate Filters (DPFs) for Particulate Mass (PM) and Particle Number (PN) control. A brief classification of the test vehicles is given in Table 1.

The SCR vehicle was equipped with a close-coupled Diesel Oxidation Catalyst (DOC) and DPF. The urea injection system was placed close to the outlet of the Diesel Particulate Filter (DPF), with the SCR placed underfloor. The SCR system was split into two parts as shown in Figure 1.

Figure 1. Configuration of the SCR Vehicle's SCR system

The dual-EGR vehicle also included a close-coupled DOC-DPF system with the low pressure EGR loop drawing from a location just downstream of the DPF-exit (Figure 2).

Andersson et al / SAE Int. J. Fuels Lubr. / Volume 7, Issue 3 (November 2014)


Figure 2. Configuration of the dual-EGR vehicle's exhaust system

Table 1. Test Vehicles' Data

For operation on UK highways both vehicles were equipped with trade registration plates and driven by Ricardo employees with advanced driving qualifications and experience.

Fuel and LubricantA batch of <10 ppm sulfur UK pump-grade diesel fuel, compliant with EN590:2009, was set-aside and employed for both chassis dynamometer and on-road PEMS testing in this work. The fuel specification is provided as Appendix 1.

Both vehicles were tested with as-received lubricants. Due to low mileage (<12 000 km) of these vehicles it is expected that these lubricants were the manufacturer's factory-fill product.

PEMS Testing

PEMS System, Analytes and MethodsThe PEMS used in this work was an ECOSTAR unit supplied by Sensors Europe GmbH. The ECOSTAR modules enabled the measurement of CO, CO2, THC, NOx, NO2, filter-based PM and a PN metric based on particle mobility. With the exception of NO and NO2, gaseous emissions were measured using the analytical approaches prescribed for certification testing in the emissions laboratory, and most commonly used. NO/NO2 were measured using non-dispersive ultra-violet (NDUVR) analysis, which has recently been permitted for certification testing in Europe, rather than the chemiluminescence analysis (CLA)

technique subject to longer-term prescription in European legislation. The PEMS employs a pitot flow meter to determine raw exhaust flow, and a downstream proportional dilution system enables a representative PM sample to be collected on-board and this also allows PN to be measured from diluted exhaust - as in a Constant Volume Sampler (CVS) in the laboratory. The PN system incorporated a feature to limit the lower particle size measured to ∼23 nm, thus mimicking the performance of regulatory PN equipment. Measurement techniques are summarised in Table 2. However, it did not incorporate a system for removal of volatile particles.

Table 2. ECOSTAR Emissions Measurements

InstallationThe PEMS was initially installed on the SCR Vehicle by Sensors' technicians, witnessed by Ricardo personnel. After testing, the PEMS was then uninstalled by Ricardo and installed on the dual-EGR vehicle. Similar installation layouts were used for both vehicles. Span gases were employed to verify analyzer performance. These gases were not placed on the vehicle - cylinders were situated in the garage and spans performed prior to vehicle departure and immediately on return.

The complete PEMS installation included requirements for combining exhaust flow from dual exhaust, a cycle rack to mount the exhaust dilution system and flow meter on the vehicle tailgate, and warning stickers indicating flammable gases, hot surfaces and a mobile hazard. Tubes for routeing sample gases to in-car analysers ran from the tailgate to the rear windows and were anchored using clamps and adhesive tape. In both vehicles the PEMS system was installed inside the void created with the rear seats down. Even so, the majority of the space was used. The front of the analysers faced the back of the driver's seat to receive the sample transfer lines entering through the rear window. Two sample lines entered - one for gases, one for PM and PN. Large battery packs, able to power the FID and multiple heated lines, were installed rear-most in the vehicle and secured. The H2/He fuel for the FID (HC measurement) was procured in small, easy to handle bottles and secured to the analysers.

Following testing, and prior to final uninstall of the PEMS equipment, the SCR vehicle was weighed. After uninstallation the vehicle was weighed again, and it was determined that the entire weight of the PEMS equipment, batteries, PCs and necessary ancillaries was ∼345 kg - equivalent to 3-4 passengers. The measured weight of the PEMS is applicable to both SCR vehicle and Dual-EGR vehicle's on-road tests.



Risk AssessmentDetailed risk assessments addressing chemical & heat releases, road safety and driver competence were performed prior to on-road testing and relevant documentation carried on-board the vehicles during testing.

On-Board Data AcquisitionIn addition to the regulated emissions measurements of the PEMS, other data were logged either into the PEMS software, or via independent data loggers.

• Exhaust temperatures were measured using 3 mm K-type thermocouples.

• Exhaust pressures were measured using Druck ±200 mbar differential transducers and Druck −1 to +4 bar gauge pressure transducers.

• Both thermocouples and pressure transducers were connected to a CSM dualscan module and logged by ETAS INCA version 5.4.1

OBD information was logged by different software for each vehicle:

• SCR vehicle - Gendan HEX+CAN OBD interface and other specialist software.

• Dual-EGR vehicle - Gendan EOBD interface and Gendan Engine Check Pro v3.1 software

Real-World Drive CyclesTwo PEMS Routes were developed by Ricardo, based upon criteria developed with AECC, with one route to be 60% urban, the other to be 60% motorway and both routes to include urban, rural and motorway segments. To differentiate the vehicle operating effects on the emissions control systems the routes were to be broadly similar: commencing and completing at the sea-level Ricardo site, but with one route commencing with urban driving, the other with motorway driving. The minimum duration of the route was to be 60 minutes.

Parts of existing drive cycles around the urban and city localities of Shoreham-by-Sea, Hove and Brighton were employed, with the close proximity of dual-carriageways and the Brighton to London motorway exploited. Two 3-segment routes (urban, rural, motorway) were adopted for use in the test work:

• Route 1 commenced with motorway operation, moved to the city and then completed on rural roads

• Route 2 commenced with city operation, then motorway, but again completed with rural road operation.

Testing OverviewTo mirror chassis dynamometer measurements, each route commenced following a ≥ 6h soak. Emissions were logged continuously including key-on. Key-on was preceded by analyser warm-up, span and operational safety checks.

Three repeats of each PEMS route were undertaken on each vehicle. For example, for the SCR vehicle, speed vs time drive traces for Route 1 are shown in Figure 3, with similar traces for Route 2 in Figure 4.

Excepting one occasion when asked to adopt a more aggressive style, the driver was instructed to follow each PEMS route using his normal style of driving. The driver had advanced driving qualifications and chassis dynamometer experience, so given the variability in traffic flow, especially in the urban section of each route, it was considered that driver consistency would have a minor impact on test-to-test ‘repeatability’.

Figure 3. Speed v time traces, 3 repeats PEMS Route 1

Figure 4. Speed v time traces, 3 repeats PEMS Route 2



Data Processing and AnalysisPEMS data were processed using EMROAD Version 5.60 Build 2, an Excel-based post-processing software tool produced by the European Commission's Joint Research Centre (JRC). EMROAD is supplied in the form of a macro based “add-in” for MS Excel and contains templates suitable for Light-duty, Heavy-duty and Non-Road Mobile Machinery (NRMM).

The basis of EMROAD processing is to use a Moving Average Window technique (MAW) and is comprised of four stages

• Segmentation of the data (identification and exclusion of non-standard operation).

• Averaging emissions data by sub-sets (Window creation). • Identification of normal testing conditions based on CO2

curve. • Sensitivity of the emissions for different driving conditions

(speed-based bins).

For the segmentation task, cold start emissions (based upon coolant temperatures below 70°C) and DPF regenerations (elevated temperature events highlighted by inspection of data from on-board logging) were identified as possible exclusions, based on current European Commission discussions.

Following identification and elimination of exclusions, the remaining data is broken down into subset “windows” which are then averaged to produce a data value. Each window is based on a reference quantity, which based on JRC experience was set to be 1 kg of CO2. When the vehicle has travelled sufficient distance to produce 1 kg of CO2 all data within that window is averaged to produce one MAW value. Several windows can overlap and data can contribute to more than one MAW window, as long as data is valid for the whole window based on the exclusion conditions. Selection of the reference quantity and value is a trade-off between retaining a small enough window so as not to lose detail but retaining sufficiently long duration to smooth large instantaneous changes arising from transient conditions. Once averaged the MAW approach can be used to identify normal driving conditions by comparison with those obtained from the same test vehicle driving the WLTC on the chassis dynamometer. The sensitivity of a vehicle's emissions to different road speeds can then be assessed. This approach allows a comparison of the emissions behaviour of different vehicles on different drive cycles on-road, and is a candidate for future RDE regulations.

EMROAD also produces tables of emissions results for a whole PEMS route, with calculated emissions including concentration (ppm), total mass (g/test), distance-specific mass emissions (g/km) and fuel specific emissions output. In this work the g/km emissions from each PEMS route, with and without exclusions, were extracted from EMROAD for analysis and comparison with chassis dynamometer test results.

Chassis Dynamometer Testing

Drive CyclesSix different drive cycles were tested on both vehicles. These included the New European Drive Cycle (NEDC), Common Artemis Drive Cycle (CADC), Worldwide Light-duty Test Cycle (WLTC) and 3 ‘Random cycles’ created using a specific cycle generator. This random cycle generator produces emissions cycles based on ‘short trip’ segments from the EU database used to construct WLTC. NEDC, CADC and WLTC cycles were each tested in triplicate, while for the Random Cycles, it was decided to test three different cycles singly, to assess the degree of variability seen and to increase the database for comparative purposes.

On each vehicle the chassis dynamometer test programme lasted 6 days. The NEDC and WLTC were tested as ‘cold-start’ cycles following an overnight soak, with the CADC and random cycles tested as ‘hot-starts’ following a standardized warm-up procedure that incorporated prior tests in a well-defined daily test sequence (Table 3).

Table 3. Sequence of Chassis Dynamometer Testing for each Vehicle

Changes to the procedures for setting vehicle test inertia between NEDC and WLTP mean that test masses for the WLTC will often be higher than for the NEDC. In this work test masses for the WLTC and random cycles were calculated according to WLTP (TMH), and those for the NEDC and CADC came from current European regulations. In addition, a one-off test on both vehicles was performed with the CADC at WLTP inertia. Preconditioning for each test cycle was undertaken at the inertia required for the test, as shown in Table 4.



Table 4. Preconditioning Approaches for Emissions Tests

Emissions MeasurementsA variety of regulated emissions, emissions considered for future regulated emissions as part of the WLTP, and other exhaust components/properties of interest, such as particle size distributions, were measured during chassis dynamometer testing. Where possible emissions were measured both in real-time and as a cumulative sample from each cycle. Measurement techniques were selected to be consistent with those currently being considered by the Additional Pollutants Working Groups of the WLTP.

Regulated EmissionsRegulated gases and particle metrics quantified during this work are listed, along with analytical approaches, in Table 5.

Table 5. Regulated Emissions Measurements

Unregulated EmissionsIndividual oxides of nitrogen plus ammonia, methane, elemental carbon and non-acidic carbonyl compounds were also measured in the test programme, as shown in Table 6.

Additionally, and as complementary measurements to the regulatory PM and >23nm PN, other particle properties in the exhaust aerosol were measured, as shown in Table 7.

Table 6. Unregulated Emissions Measurements

Table 7. Additional Particle Properties Measured

Other MeasurementsEach vehicle's CAN logged data that was later employed to assist in the processing of data through EMROAD, or to help in the interpretation of emissions from the test vehicles (Table 8).

Table 8. Data Logged via Vehicles' CAN

RESULTS

Practical Use of PEMSFollowing installation of the PEMS system, a few days were required for familiarization with the operation of the analysers. Initial commissioning issues with the system were mainly attributable to immaturity of the software, so both on-site and remote support was provided by the instrument supplier. Issues powering the PEMS modules were also encountered. The manufacturer had never run all the PEMS modules, including the FID and PN, together and eventually it was necessary to use a 12V lead-acid car battery for the data logger, two 115 Ah AGM 12V lead-acid supplied by Sensors for the NDIR, NDUV



and PM/PN systems and a 155 Ah 12V lead-acid battery was used to power the FID system, via a 240V inverter, and heated lines. However, due to the broad similarity of operating this equipment to that of lab-based analysers, and the development and implementation of written processes with defined steps, it proved relatively straightforward for a test facility experienced in emissions measurements to perform PEMS testing.

Correlation of PEMS Gaseous Emissions and PN with Lab-Based AnalysersCorrelations of regulated gases measured by the PEMS analysis modules and the lab-based Horiba MEXA 7000 analyser indicated that real-time emissions events were reported representatively by the PEMS, but that some absolute differences were apparent:

• CO emissions were similar except, initially, at <20 ppm levels due to an offset of ∼15 ppm in the PEMS. This was traced to a software issue inhibiting a forced zero, and after correction the correlations were improved to within a few percent

• CO2 emissions were similar to within 2% of measured levels

• NOx emissions agreed within ∼10% across a wide range of values for the first 2-3h of instrument operation, though substantial downward drift of the PEMS NDUV analyser was observed after this, possibly due to increasing water condensation in the measurement cell. PEMS tests seldom exceeded this duration

• THC measurements using the PEMS FID closely tracked those from the lab-based FID, but the PEMS typically reported levels ∼50% lower. It was not possible to access the calibration of the system to correct this

• Raw exhaust PN measurements from the developmental, pre-calibration, on-board diffusion charging system showed good real-time correlation with a Horiba SPCS system at higher particle levels; this also indicating a dominance of solid particles. After correction of PN backgrounds from the full-flow dilution system, and subtraction of pre-test noise in the Diffusion Charger (DC), acceptable agreement was also achieved between the two systems at lower PN levels. PN data collected on-road was subjected to the baseline noise subtraction prior to calculation of PN levels.

In general, after the evaluations summarised above, the results from the PEMS were considered to be directly comparable with lab-based measurements, excepting the PN measurement where baseline corrections were made and the THC, which consistently reported ∼50% low.

Emissions Results: Chassis Dyno Cycles v PEMSThe following section presents emissions data from chassis dyno cycles and PEMS tests on both test vehicles. Emissions are reported from entire chassis dyno test cycles, and on-road routes, with each datum representing the mean of

three repeat tests, and the error bars indicating one standard deviation. PEMS route data includes emissions from ambient cold start (∼15°C), NEDC and WLTC from ∼25°C starts, while other cycles were hot starts. PEMS data are shown without any exclusions.

Oxides of NitrogenNOx emissions of the SCR and dual-EGR vehicles are given in Figure 5. Both vehicles showed mean NOx emissions from the NEDC very close to the 80 mg/km Euro 6 limit, but the emissions levels on other cycles and PEMS routes ranged from 2 to almost 6 times these levels.

Figure 5. NOx Emissions from Chassis Dyno and PEMS Tests

Highest emissions from the SCR-equipped vehicle were seen from the WLTC and PEMS routes, with maximum NOx emissions at <4x the NEDC levels. Mean PEMS emissions from this vehicle were higher from Route 2, which started with urban driving, than Route 1: indicating a delay in optimum NOx control as the SCR climbs to its operational temperature of ∼200°C. Lowest emissions were seen from the hot-start tests, where SCR operational temperature was achieved most rapidly.

In comparison with the SCR-equipped vehicle, the mean NOx emissions of the dual-EGR vehicle were higher from the long-duration CADC (>3000 s) and PEMS routes, where NOx emissions ranged from ∼320 mg/km on the CADC to >500 mg/km on PEMS Route 1, but lower (180-200 mg/km) from the shorter (∼1800 s) WLTC and random cycles. The calibration of the d-EGR vehicle appeared to offer better NOx control on the WLTC than that of the SCR vehicle, but this was not the case for real urban driving operation. The dual-EGR vehicle showed significantly higher emissions on both PEMS routes than the SCR-vehicle, with emissions from the initial high-speed PEMS Route 1 highest of all. This indicating the advantages of SCR-based NOx control on longer cycles, once the operational temperature is achieved. The results on the Urban Driving Cycle (UDC) part of the NEDC indicate some potential benefit from the use of EGR before the SCR system reaches operating



temperature, with emissions of 128 mg/km for the d-EGR vehicle compared to 213 mg/km for the SCR vehicle, albeit both still exceeding the level of 80 mg/km required for the full NEDC. It appears likely that a combination of EGR and SCR would offer the best opportunity for low NOx emissions in the full range of real-life driving, with EGR able to provide some benefit at start-up and SCR able to offer better overall control once it reaches operating temperature.

Ammonia emissions were low from both vehicles, with a maximum of 4 ppm seen from the SCR vehicle and lower levels from the d-EGR vehicle.

Emissions of NO2 from the two vehicles (Figure 6) were broadly similar for several cycles (NEDC, both CADCs and PEMS Route 2). However, emissions over the WLTC and random cycles were higher from the SCR vehicle than from the d-EGR vehicle. An active regeneration during one of the PEMS Route 1 tests elevated the mean NO2 from the d-EGR vehicle, otherwise this route's mean emissions would also have been similar to the SCR vehicle's result. NO2 emissions from the d-EGR vehicle averaged 28 mg/km on the NEDC but rose to 170 mg/km on the PEMS tests (including the one on which a regeneration occurred - 161 mg/km excluding it). The SCR vehicle gave very similar NO2 results (at 24 mg/km) to the d-EGR vehicle on the NEDC, but somewhat lower average results on the PEMS tests (140 mg/km). Averaged across all cycles and routes, NO2 emissions were 101 mg/km from the d-EGR vehicle and 110 mg/km from the SCR vehicle. This effect is indicative of the need for the DOC to generate additional NO2 to support the fast SCR reaction [11] for NOx reduction, which results in overall NOx emissions being significantly lower in real driving. For any vehicle, the NO2 emissions (both absolute and as a proportion of NOx) will be dependent upon the system architecture and operating conditions. The generation of NO2 within the emissions control system is needed for DPF regeneration as well as for optimum SCR operation, but the advantage of an SCR system following the DPF is that this controls both NO2 and total NOx.

Figure 6. Nitrogen Dioxide (NO2) Emissions from Chassis Dyno and PEMS Tests

It is worth noting that NO2 is being considered by the European Commission as a possible future regulated emission in its own right. As this develops, the historical regulatory approach of converting NO to NO2 and then quantifying total NOx emitted by the vehicle as total NO2 detected by the analyser will have to be revisited: quantifying NO (mass ∼30) and NO2 (mass ∼46) separately and adding the masses will report a lower total NOx level than quantifying both as mass 46.

Particulates and Particle NumbersParticulate mass emissions were below 1mg/km from all chassis dynamometer and PEMS tests (Figure 7), excepting one test, a CADC at WLTC inertia on the SCR vehicle, where an active DPF regeneration was observed. In the regeneration test, the PM mass approached 7 mg/km, but with regeneration periodicity taken into account, PM emissions were well below the 4.5 mg/km Euro 6 limit. It is worth noting that visual inspection of the PM filter paper revealed no obvious signs of elemental carbon, so this additional mass is presumed to be volatile.

On-road (PEMS) gravimetric PM emissions were approximately an order of magnitude below chassis dyno levels, at <0.3 mg/km. The collected filter masses from emissions tests were broadly similar (10 μg to 30 μg) from both chassis dyno and PEMS tests. The similar masses may be as a result of the collection and removal of semi-volatile materials during the longer test, so the much lower PM emissions are likely a simple function of dividing the same mass by a longer distance driven (NEDC ∼ 11 km, PEMS Route ∼120 km).

Post-DPF PM has been shown to be dominated by volatiles [12] with as little as 0.5% of the mass measureable as non-volatile particle numbers. As a chassis dyno correlation of PEMS-PN with Horiba SPCS had indicated a strong linear correlation, when measurements of regulatory particle numbers (Figure 8) did not show the same factor-of-ten discrepancy between PEMS and chassis dyno tests that was observed in the PM data, this can be taken as evidence that there were few >23nm volatile particles from the PEMS tests. Chassis dyno tests produced particle number emissions below the Euro 6 limit of 6×1011/km, though the CADC that featured an active regeneration was borderline. PEMS Route PN results were all above the Euro 6 limit, even exceeding 50% above, especially where an active regeneration featured. In both chassis dyno and PEMS tests, active regenerations appeared to increase PN by a factor of ∼10. Further work is required with a fully calibrated PN PEMS system, especially as there may be an effect from the lack of a specific volatile particle remover in the instrument used for this test programme.



Figure 7. PM Emissions from Chassis Dyno and PEMS Tests

Figure 8. PN Emissions from Chassis Dyno and PEMS Tests

In parallel with the regulatory PN measurements, a second particle number measurement system was employed for the chassis dyno tests. This was equipped with 2 particle number counters, giving dual d50 capabilities: 23 nm (TSI 3010D CPC) and 7 nm (TSI 3022A CPC). Results from the 3010D correlated to a very high degree (R2 = >0.999) with the results from the SPCS shown in Figure 8. These dual results were used to study the presence of particles between 7 nm and 23 nm in comparison with particles >23 nm. As Figure 9 shows, >23 nm emissions from non-regenerating tests were always greater than those in the 7 nm to 23 nm range, and if these vehicles are representative of the diesel vehicle parc, there appears to be minimal benefit in modifying the size range measured in the regulatory procedure to a lower d50. From a test that featured a regeneration (shown enclosed in a red box in Figure 9), the 7 nm to 23 nm levels increased substantially - to 4x those of the >23 nm particles, probably due to a temporary reduction in the filtration efficiency of the DPF.

Figure 9. Proportions of 7 nm to 23 nm and >23 nm PN emissions from Chassis Dyno and PEMS Tests [regeneration test in red enclosure]

Total Hydrocarbons and Carbon MonoxideEmissions of total hydrocarbons (Figure 10) and carbon monoxide (Figure 11) showed similar comparative profiles for the two vehicles. All results were <30% of the Euro 6 limits. Highest emissions from the chassis dyno cycles were seen from the NEDC, with next highest, but substantially lower, levels from the WLTC. This is a clear indication of the cold start effect, where THC and CO are elevated pre-DOC light-off, and the proportionally greater impact the cold start has on the much shorter NEDC. The extremely low CADC emissions test (long duration, hot start) results are the corollary of this effect.

Figure 10. THC Emissions from Chassis Dyno and PEMS Tests



Figure 11. CO Emissions from Chassis Dyno and PEMS Tests

Route 1 PEMS results showed CO and HC emissions levels similar to those from chassis dyno cycles. However, PEMS Route 2 CO emissions were generally higher than the emissions from chassis dyno cycles. This contrast with HC control on PEMS Route 2 may stem from the extended low temperature operation at the start of the route, and the ability of the ceramic DOC to trap HC (but not CO) for later post light-off combustion. Given the significance of the cold start event on HC emissions, the effect of cycle length on mass/km emissions and the ease of evaluating cold starts on a chassis dynamometer, it is hard to see a significant benefit in measuring on-road THC for certification purposes for current vehicle technologies.

Greenhouse GasesAs Figure 12 shows, with the red dots indicating type approval values, NEDC g/km CO2 emissions were substantially above certification levels for both SCR and d-EGR vehicles. The differential was particularly large for the d-EGR vehicle. However, there was minimal impact of WLTC inertia on CO2 emissions with either vehicle. Nevertheless, WLTC CO2 figures are identical to (d-EGR) or lower than (SCR) the NEDC values, but higher than Type Approval data. Apparent variability in the repeat NEDC CO2 data can be partially attributed to DPF regenerations during prior preconditions and tests on other emissions cycles, leading to differing backpressure.

PEMS Route CO2 emissions were similar to chassis dyno levels, despite carrying an extra ∼345 kg. In the real-time gaseous emissions data there appeared to be minimal impacts of accelerations on emissions. In addition, the lack of climbs and steep gradients in the PEMS routes further moderated the on-road testing. To fully study the impacts of vehicle weight, these influences would need to be exaggerated within the test cycle.

Figure 12. CO2 Emissions from Chassis Dyno and PEMS Tests

Methane emissions (Figure 13) were highest from cold start cycles, and higher from the d-EGR vehicle than from the SCR vehicle. At peak emissions (NEDC) of ∼9 mg/km (d-EGR) and ∼6 mg/km (SCR) these emissions equate to an equivalent 100 yr Global Warming Potential (GWP) (21 times CO2) of ∼0.22 g/km and 0.15 g/km of CO2 respectively.

Figure 13. Methane (CH4) Emissions from Chassis Dyno Tests

Emissions of nitrous oxide (Figure 14) were at similar mg/km levels to methane, with highest levels, independent of cycle, from the SCR vehicle. Partial oxidation of ammonia to nitrous oxide, across the ammonia slip catalyst, may explain the small increases observed relative to the d-EGR vehicle. Since N2O has a larger GWP (310 times CO2) than methane, the peak emissions levels observed from the two vehicles result in impacts equivalent to 1.5 to 3.5 g/km of CO2.



Figure 14. Nitrous Oxide (N2O) Emissions from Chassis Dyno Tests

Interesting ObservationsReal-time tailpipe NOx data from the PEMS Routes were compared with acceleration data from both vehicles (Figure 15). This results in a near-Gaussian distribution of data around the origin, and strongly indicates that highest NOx is seen under steady state conditions. This suggests that transient NOx control is well addressed by both test vehicles. However, it is worth noting that peak NOx (g/s) is almost twice as high from the d-EGR vehicle as from the SCR vehicle.

Figure 15. NOx Emissions v. Acceleration Rates: d-EGR & SCR Vehicles

Plotting engine speed (rpm) v torque (Nm) as shown in Figure 16, indicates that NOx emissions are dominated by high speed, mid to high load events. The d-EGR vehicle appears to offer some benefit at lower load whilst lacking the control of the SCR vehicle in the mid-to-higher load area. The control of NOx by an SCR system will, of course, be dependent on the urea dosing strategy and whether this is based on the full engine map or only on the legislative test cycle.

Figure 16. NOx Emissions v. Engine Speed & Torque: d-EGR and SCR Vehicles

Impacts of EMROAD ExclusionsFigures 17, 18, 19 and 20 show the impacts of applying exclusions to PEMS Route data collected from the two test vehicles. In all four cases the figures show the total percentage mass of emission (g/test) omitted by exclusions.

It is important to understand that when applied, the exclusions eliminate the entire period of time from the test data, not just the incremental emissions “added” by the excluded event.

This effect is particularly noticeable with g/km CO2, where a portion of data with bad fuel economy may be removed from the test, making average trip CO2 appear better.

In this study, exclusions considered were:

• Cold start exclusions - data collected prior to coolant temperature reaching 70°C.

• Regeneration exclusions - manually selected based upon temperature rise with post-injection and subsequent return to ‘normal’ exhaust temperatures.

Figure 17 shows the impacts of cold start, and a relatively short duration ‘mild’ active regeneration on emissions of THC, CO, CO2, NOx (wet & dry), NMHC and NO2 from the SCR Vehicle. On this vehicle the DPF is regenerated relatively frequently so only limited soot is burnt.

Figure 18 shows the impacts of cold start, and a longer duration ‘complete’ active regeneration on emissions of THC, CO, CO2, NOx (wet & dry), NMHC and NO2 from the d-EGR vehicle. On this vehicle the DPF is regenerated infrequently and hence a large soot mass is removed.



Figure 17. Impacts of EMROAD Exclusions: PEMS Route 2, Cold Start and Active Regeneration, SCR Vehicle

Figure 18. Impacts of EMROAD Exclusions: PEMS Route 1, Cold Start and Active Regeneration, d-EGR Vehicle

Figure 19 shows the impacts of cold start alone on emissions from the d-EGR vehicle.

Figure 19. Impacts of EMROAD Exclusions: PEMS Route 2, Cold Start, d-EGR Vehicle

Figure 20 shows the impacts of cold start alone on emissions from the SCR vehicle.

Figure 20. Impacts of EMROAD Exclusions: PEMS Route 1, Cold Start, SCR Vehicle

Considering the four previous figures, on both vehicles:

• Around 3% of the emissions measured on the PEMS route are excluded due to cold start and while cold start produces around 30% of total CO, there appears to be only minimal impact on THC and NMHC. The cold start exclusion results in <2% reduction in g/km NOx on both vehicles.

• DPF regeneration ‘severity’ has a major impact on the significance of the exclusions: there was a slight reduction in CO2 from excluding DPF regeneration on the SCR vehicle ∼ 1 g/km, but a reduction of ∼14 g/km in CO2 from the d-EGR vehicle. The impacts on NOx were less clear, with ∼5% reduction in g/km NOx with the regeneration exclusion from both vehicles. In addition, after the regeneration exclusion was applied both vehicles showed an increase in NO2 emissions, suggesting that the DPF reaction was digesting NO2.

DISCUSSION

PEMS PracticalitiesExperiences of testing the PEMS indicate that the system bulk, weight and power consumption need to be reduced. There was substantial power usage from HC analyser in this work - both the heated line and the FID. Because of the practical difficulties, the European Commission has proposed making HC measurement from the PEMS tests an optional requirement. Integrated logging of data within the PEMS, instead of requiring an external laptop would be helpful, and reduce space and weight requirements.

There was a “development level” feel to the software, and while this was acceptable for use by an experienced calibration engineer, it would be more challenging for a relatively inexperienced end user such as a test driver.



Automation of span and zero processes needs to be done to reduce operator effort. Certification analysers have all gases connected and one click starts whole zero and span process.

PEMS PM and PN analysers were developmental and require further work. In particular, when collected filter masses are a few tens of microgrammes, and there is no visible stain on the filter, validation of correct sampling of PM will be important.

In this work, PEMS instruments measuring CO and CO2 correlated very well with lab-based systems. However, HC and NOx drift occurs after ∼2-3 h. Up until this point, NOx by NDUV correlated within 10% of the lab-based chemiluminescence analyser. The HC correlation was linear, but reported on-board emissions levels were 50% lower than from the lab system.

In retrospect, the data from this work indicates that little value would have been lost if HC and PM had not been measured.

CONCLUSIONSPEMS has been shown to be a viable tool for measuring emissions of gaseous pollutants including NO2. Further work is required to ensure that robust measurements of PN are possible.

Even though relatively moderate in terms of altitude, ambient temperature, driving style and inclines, the PEMS routes tested included a variety of driving elements: cold start, urban, rural, highway and motorway, with DPF regenerations encountered during on-road testing of both vehicles.

Both cold start and DPF regenerations have been seen to occur in moderate European driving and should be considered part of normal driving. These events were shown to have significant impacts on the emissions observed for the test vehicles and excluding them from PEMS route results may result in unrepresentative emissions data for the vehicles.

There are indications in this study that emissions from the test vehicles, and their emissions control systems, were different from the two distinct PEMS routes. These effects were noticeable in the NOx, the CO and perhaps the THC data. The main difference between the PEMS routes was the order of the component phases, with one route commencing with high-speed operation (and thus rapid light-off), but the other commencing with a long urban phase. Further work is required, but the differences observed between these PEMS routes may indicate that as well as defining the driving styles within the mandated PEMS route composition, the order of the component phase should also be prescribed. It may be that initial low speed phases will be required to present the greatest challenge for emissions control.

CO and HC emissions did show some variation between the various cycles and routes, but in all cases were well within the Euro 6 limits. The current (NEDC) test, for which these vehicles were calibrated, ensures good control of these emissions

under cold-start low load conditions and future procedures will need to ensure that this benefit is protected. The difficulties of using FID equipment on-board has, since these tests were conducted, led to the European Commission recommending that HC emissions should not be measured in the PEMS test at this stage.

Cycle NOx emissions from both vehicles were ∼80 mg/km over the NEDC, but increased substantially with all other cycles, and well in excess of a possible 1.5x compliance factor over the regulatory limit. It appears that neither vehicle is well optimised for the WLTC, resulting also in higher NOx emissions over this cycle. For the SCR vehicle, WLTC and PEMS emissions were broadly similar: in the 250 mg/km to 300 mg/km range, with these results higher than any of the other cycles. However, for the d-EGR vehicle, WLTC emissions were lower than from the SCR (∼200 mg/km) indicating better NOx control over this cycle, but emissions from the hot start CADC were between 300 mg/km and 400 mg/km and the PEMS routes emissions were between 400 mg/km and 600 mg/km

Considering the results of chassis dyno emissions tests: there was no overall trend of lower NOx emissions from the d-EGR vehicle than from the SCR vehicle, although the d-EGR approach appears to offer some benefit under low load conditions.

When considering emissions from the two vehicles on PEMS routes: the d-EGR vehicle had highest NOx emissions from Route 1 (initial phase is high speed), while the SCR vehicle had lowest average NOx emissions from both routes.

NO2 emissions were highest from higher average temperature tests through greater periods of operation in the DOC's optimum NO to NO2 conversion window.

CO and HC were highest from cold start cycles; PEMS emissions were similar, on a mg/km basis, to hot start chassis dyno cycles

N2O emissions were increased in the higher speed emissions cycles. Emissions levels from both vehicles were (GWP) equivalent to 1.5 to 3.5 g/km of CO2. The d-EGR vehicle's N2O emissions over the combined cycles were in the range ∼4 mg/km to ∼7 mg/km, substantially less than observed from the SCR vehicle (5 - 12 mg/km) and implying that at least some of that vehicle's N2O emissions are derived from partial oxidation of ammonia (this more likely than reduction of NOx) within the SCR system

Ammonia emissions were low from both vehicles: with a maximum of 4 ppm from the SCR vehicle, lower from the d-EGR vehicle

From both vehicles methane emissions constituted less of a GWP impact than N2O.



CO2 emissions of both vehicles were substantially higher than Type Approval values, but neither WTLP inertia, nor the increased weight of the PEMS system on road had a major impact on the g/km emissions. This indicates minimal impact of accelerations, and lack of climbs and steep gradients in the PEMS routes. To fully study the impacts of vehicle weight, these influences would need to be exaggerated within the PEMS test route.

PM emissions from chassis dyno tests were typically < 1 mg/km. Regeneration impacts on PM (7 mg/km from a test featuring a regeneration) were small when regeneration periodicity was considered. PEMS PM emissions were very low (≪ 1 mg/km). These data are hard to validate, but the lower emissions determined might merely be a function of the effect of volatiles leading to dividing a similar nominal mass by longer cycle distances.

PN from both vehicles were below the Euro 6 limit on all chassis dyno tests. DPF regeneration increases solid PN by a factor of 10 - 20, but the Euro 6 limit would not be challenged on the basis of regeneration frequency. Given the immaturity of PN PEMS, further work is required in this area.

<23 nm PN are present in the exhausts of both vehicles, but including them in the integrated total less than doubles solid PN during normal engine operation. With regenerations <23 nm particles were ∼5 times higher than >23 nm PN. However, there is no evidence in this study to suggest any benefits in reducing the lower threshold of the PN measurement method from the current d50 of 23 nm. Work by the California Air Resources Board (CARB) in conjunction with the European Commission's Joint Research Centre [13] showed a large increase in volatile particles below the 23nm cut-off during regeneration and also indicated that the PM mass increase associated with regeneration appears to be due to semi-volatile particles collected on filter media.

REFERENCES1. Weiss, Bonnel, Kühlwein, Provenza et al, “Will Euro 6 reduce

the NOx emissions of new diesel cars? Insights from on-road tests with Portable Emissions Measurement Systems (PEMS)”, Atmospheric Environment 62657-665, 2012, doi:10.1016/j.atmosenv.2012.08.056.

2. Chen & Borken-Kleefeld, “Real-driving emissions from cars and light commercial vehicles - Results from 13 years remote sensing at Zurich/CH”, Atmospheric Environment 88, 157-164, May 2014, doi:10.1016/j.atmosenv.2014.01.040.

3. Ligterink, Kadijk, van Mench, Hasuberger and Rexeis, “Investigations and real world emission performance of Euro 6 light-duty vehicles. TNO Report 2013 R11891, 5 December 2013, https://www.tno.nl/downloads/investigations_emission_factors_euro_6_ld_vehicles_tno_2013_r11891.pdf

4. “Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. Cars 2020: Action Plan for a competitive and sustainable automotive industry in Europe”, (2012), COM (2012) 636 final, http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2012:0636:FIN:EN:PDF

5. “Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the regions. A Clean Air

Programme for Europe”, (2013), COM(2013) 918 final, http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2013:0918:FIN:EN:PDF.

6. Steininger, “Completing European Emissions Targets with RDE test procedures for light duty vehicles, International Conference on Real Driving Emissions, Bonn, Germany, 2-4 December 2013

7. Weiss, Bonnel, Hummel & Steininger, “A complementary emissions test for light-duty vehicles: Assessing the technical feasibility of candidate procedures”, JRC Scientific and Policy Report EUR 25572 EN, 2013, http://publications.jrc.ec.europa.eu/repository/handle/111111111/27598 or doi:10.2790/65654.

8. Weiss, Bonnel, Riccobono & Vlachos, “Developing an on-road test for light-duty vehicles in Europe”, 24th CRC Real World Emissions Workshop, San Diego, USA, 30 March-2 April 2014.

9. May, Favre & Bosteels, “Emissions from Euro 3 to Euro 6 light-duty vehicles equipped with a range of emissions control technologies”, IMechE Internal Combustion & Engines Conference, London, 27 November 2013

10. Favre, C., Bosteels, D., and May, J., “Exhaust Emissions from European Market-Available Passenger Cars Evaluated on Various Drive Cycles,” SAE Technical Paper 2013-24-0154, 2013, doi:10.4271/2013-24-0154.

11. Cooper, McDonald, Walker & Sanchez, “The Development and On-Road Performance and Durability of the Four-Way Emission Control SCRT System”, US DOE, 9th Diesel Engine Emissions Reduction Conference (DEER), Newport, RI, August 2003, http://www.eere.energy.gov/vehiclesandfuels/pdfs/deer_2003/session8/2003_deer_walker.pdf

12. Andersson, Giechaskiel, Muñoz-Bueno, Sandbach & Dilara, “Particle Measurement Programme (PMP) Light-duty Inter-laboratory Correlation Exercise (ILCE_LD) Final Report”, European Commission Institute for Environment and Sustainability, 2007, EUR 22775 EN, http://ies.jrc.ec.europa.eu/uploads/fileadmin/Documentation/Reports/Emissions_and_Health/EUR_2006-2007/EUR_22775_EN.pdf

13. Dwyer, Ayala, Zhang, Collins, Huai, Herner and Chau, “Emissions from a diesel car during regeneration of an active diesel particulate filter”, Journal of Aerosol Science 41 (6) 541-552, June 2010.

CONTACT [email protected]

[email protected]

ACKNOWLEDGMENTSThe authors wish to thank the AECC members, Ricardo staff and Directors, and Sensors Europe / Inc for their support.

DEFINITIONS/ABBREVIATIONSAECC - Association for Emissions Control by Catalyst

CADC - Common Artemis Driving Cycles

CAN - Controller Area Network

CLA - Chemiluminescence Analyser

CVS - Constant Volume Sampler

d50 - Median particle size diameter

DC - Diffusion Charger

d-EGR - Dual system Exhaust Gas Recirculation

DOC - Diesel Oxidation Catalyst

DPF - Diesel Particulate Filter

DNPH - 2,4-Dinotrophenylhydrazine

EGR - Exhaust Gas Recirculation

EOBD - European On-Board Diagnostics



http://dx.doi.org/10.1016/j.atmosenv.2012.08.056



https://www.tno.nl/downloads/investigations_emission_factors_euro_6_ld_vehicles_tno_2013_r11891.pdf

https://www.tno.nl/downloads/investigations_emission_factors_euro_6_ld_vehicles_tno_2013_r11891.pdf

http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2012:0636:FIN:EN:PDF





http://publications.jrc.ec.europa.eu/repository/handle/111111111/27598

http://publications.jrc.ec.europa.eu/repository/handle/111111111/27598

http://dx.doi.org/10.2790/65654

http://www.sae.org/technical/papers/2013-24-0154

http://dx.doi.org/10.4271/2013-24-0154

http://www.eere.energy.gov/vehiclesandfuels/pdfs/deer_2003/session8/2003_deer_walker.pdf

http://www.eere.energy.gov/vehiclesandfuels/pdfs/deer_2003/session8/2003_deer_walker.pdf

http://ies.jrc.ec.europa.eu/uploads/fileadmin/Documentation/Reports/Emissions_and_Health/EUR_2006-2007/EUR_22775_EN.pdf



EU - European Union

FC - Fuel Consumption

FID - Flame Ionisation Detector

FTIR - Fourier Transform Infra-Red spectrometry

GC - Gas Chromatography

GRPE - United Nations Working Party on Pollution and Energy

GWP - Global Warming Potential

HPLC - High Pressure Liquid Chromatography

JRC - Joint Research Centre (European Commission)

MAW - Moving Average Window

NDIR - Non-Dispersive Infra-Red analyser

NDUVR or NDUV - Non-Dispersive Ultra-Violet Resonance absorption analyser

NEDC - New European Driving Cycle

NMHC - Non-Methane Hydrocarbons

NOx - Oxides of Nitrogen

OBD - On-Board Diagnostics

OEM - Original Equipment Manufacturer

PEMS - Portable Emissions Measurement Systems

PM - Particulate Mass

PN - Particle Number

RC - Random Cycle

RDE - Real Driving Emissions

RDE-LDV - Real Driving Emissions of Light-Duty Vehicles

SCR - Selective Catalytic Reduction

SPCS - Solid Particle Counting System

TGA - ThermoGravimetric Analyser

THC - Total Hydrocarbons

TMH - Test Mass High (WLTP UNECE gtr No 15)

UN - United Nations

UV - Ultra-Violet

VPR - Volatile Particle Remover

WLTC - Worldwide Light-duty Test Cycle

WLTP - Worldwide harmonised Light vehicles Test Procedure



APPENDICES

APPENDIX 1: FUEL SPECIFICATION

APPENDIX 2: TEST INERTIA, MASS AND LOAD TERMS FOR THE TWO TEST VEHICLESWLTC test mass can either be Test Mass Low (TML) or Test Mass High (TMH). For worst case TMH was selected:

• TMH = OMH + 100kg + 0.15(LM − OMH − 100kg)

Where:

• OMH - Mass of empty vehicle with options • LM - Permissible max. Laden load

Since OMH was not readily available information, an approximation was employed:

• The final result would increase the inertia and keep the load terms the same, so inertia class was raised from the base Euro inertia class to the next highest to give the same effect

The load terms selected for the SCR vehicle were drawn from the EPA database. While the dual-EGR vehicle is not quoted by the EPA, a similarly sized vehicle was identified and those EPA terms employed.

Dual-EGR vehicle not weighed; SCR Vehicle 1540kg at weighbridge, * = estimated value

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE International.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE International. The author is solely responsible for the content of the paper.



Documents

On-Road and Chassis Dynamometer Evaluations of Emissions from