Size and composition distributions of particulate matter emissions: Part 2--heavy-duty diesel vehicles.

ABSTRACT

Particulate matter (PM) emissions from heavy-duty diesel vehicles (HDDVs) were collected using a chassis dynamometer/dilution sampling system that employed filter-based samplers, cascade impactors, and scanning mobility particle size (SMPS) measurements. Four diesel vehicles with different engine and emission control technologies were tested using the California Air Resources Board Heavy Heavy-Duty Diesel Truck (HHDDT) 5 mode driving cycle. Vehicles were tested using a simulated inertial weight of either 56,000 or 66,000 lb. Exhaust particles were then analyzed for total carbon, elemental carbon (EC), organic matter (OM), and water-soluble ions. HDDV fine ([less than or equal to]1.8 [micro]m aerodynamic diameter; [PM.sub.1.8]) and ultra-fine (0.056-0.1 [micro]m aerodynamic diameter; [PM.sub.0.1]) PM emission rates ranged from 181-581 mg/km and 25-72 mg/km, respectively, with the highest emission rates in both size fractions associated with the oldest vehicle tested. Older diesel vehicles produced fine and ultrafine exhaust particles with higher EC/OM ratios than newer vehicles. Transient modes produced very high EC/OM ratios whereas idle and creep modes produced very low EC/OM ratios. Calcium was the most abundant water-soluble ion with smaller amounts of magnesium, sodium, ammonium ion, and sulfate also detected. Particle mass distributions emitted during the full 5-mode HDDV tests peaked between 100-180 nm and their shapes were not a function of vehicle age. In contrast, particle mass distributions emitted during the idle and creep driving modes from the newest diesel vehicle had a peak diameter of approximately 70 nm, whereas mass distributions emitted from older vehicles had a peak diameter larger than 100 nm for both the idle and creep modes. Increasing inertial loads reduced the OM emissions, causing the residual EC emissions to shift to smaller sizes. The same HDDV tested at 56,000 and 66,000 lb had higher [PM.sub.0.1] EC emissions (+22%) and lower [PM.sub.0.1] OM emissions (-38%) at the higher load condition.

INTRODUCTION

The characterization of particulate matter (PM) size and composition distributions emitted from in-use diesel vehicles operated under real-world conditions is essential to help quantify the environmental impact of these particles. Diesel PM emissions have been found to be associated with adverse health effects including increased human mortality and morbidity. (1-3) Many carcinogenic and mutagenic compounds have been measured in diesel PM, (4-8) and it has been designated as a toxic air contaminant by the state of California (9) and as a mobile source air toxic by the U.S. Environmental Protection Agency (EPA). (10) Diesel engines are believed to be the major source of elemental carbon (EC) in both urban environments (11) and in the global atmosphere. (12) The optical properties of diesel PM have been shown to reduce regional visibility (13) as well to directly affect the radiative balance of the atmosphere. (14) In all cases, the environmental impact of the diesel exhaust particles is directly influenced by their size and composition.

Diesel engine technology and diesel fuel formulation have evolved over the past decades to reduce air pollution emissions. It is important to characterize the size and composition distribution of particles emitted from contemporary diesel engines and to compare these measurements to older engines to quantify our progress on air pollution problems. Heavy duty diesel vehicles are of special interest because these vehicles comprise only a small fraction of the total vehicles on the road but they contribute significantly to on-road mobile source primary PM emissions. (15) A common method to measure emissions from a vehicle is to operate it on a chassis dynamometer while collecting the emitted particles on filters that can then be analyzed for chemical composition. Heavy-duty diesel vehicles (HDDVs) with a gross vehicle weight greater than 30,000 lb are too large to test on most chassis dynamometers under realistic driving cycles and so simplified cycles are often used. Although these simplified tests provide valuable information, (16-18) they do not fully reveal all the emissions trends that can occur during dynamic driving cycles. Likewise, tunnel studies (19-22) do not fully capture the range of HDDV driving cycles and direct engine dynamometer measurements (23,24) do not fully represent emissions from complete vehicles. Attempts to measure emissions from on-road vehicles using real-time instruments (25,26) study the most realistic vehicle emissions, but real-time measurement techniques are still under development and therefore currently provide an incomplete characterization of particle size and chemical composition. There is no perfect method to characterize emissions from HDDVs, and so a combined weight of evidence approach must be continued for the present term.

The purpose of this study was to report the size and composition distributions of PM released from contemporary HDDVs measured using a chassis dynamometer/dilution sampling system that employs filter-based samplers, cascade impactors, and scanning mobility particle size (SMPS) measurements. The chassis dynamometer used in this study was able to simulate realistic dynamic vehicle driving cycles and inertial loads. The data set includes a range of different vehicle types and emission control technologies. In the present study, particle size and composition distributions in six size fractions between 0.056 and 1.8-[micro]m particle diameters are reported in addition to data from bulk PM with diameters less than 1.8 [micro]m. Ultrafine PM is defined as particles with aerodynamic diameter between 56 and 100 nm ([PM.sub.0.1]; as collected by stage 10 of a Micro Orifice Uniform Deposit Impactor [MOUDI]), and fine PM is defined as particles with aerodynamic diameter less than 1.8 [micro]m ([PM.sub.1.8]; as collected by Reference Ambient Air Quality Sampler [RAAS] filters). These are useful functional definitions because very little of the PM mass collected in this study had aerodynamic particle diameters below 56 nm or above 1.8 [micro]m. Vehicle emissions characteristics as a function of time for different vehicles and driving cycles are also presented.

EXPERIMENTAL METHODS

PM emissions from HDDVs were collected at the West Virginia University (WVU) transportable chassis dynamometer facility at a grocery distribution center in Riverside, CA in June and July of 2003. (27) Four HDDVs were tested (Table 1) spanning a wide range of vehicle age and engine technology; all vehicles tested had stock exhaust systems with no additional emissions reduction technologies installed. The chassis dynamometer used in the tests employed a combination of flywheels and electrical motors to simulate inertial loads of either 56,000 or 66,000 pounds. All vehicles were tested using partial or full California Air Resources Board (CARB) Heavy Heavy-Duty Diesel Truck (HHDDT) cycles consisting of five modes--an idle mode, a creep mode, a transient mode, a low speed cruise, and a high speed cruise. (28) Table 2 presents details on the HHDDT driving cycle.

The diesel fuel used to power HDDVs was 'tank fuel' (the fuel in the tank when the vehicle was procured for testing), and was assumed to be California diesel fuel. Most diesel fuels in California are alternative formulations, which means that even though the nominal fuel specification calls for 10% aromatic content and no more than 500 ppm sulfur, the actual aromatic content tends to be approximately 19-20% (or higher) and the sulfur content approximately 150-200 ppm. (29) Fuel and oil samples were collected from each vehicle and their composition will be reported in future studies.