Ultrafine Particle Number Concentrations from …bholmen/BAHwork/docs/TRB06-2253_VikaraHolmen...Ultrafine Particle Number Concentrations from Hybrid Urban Transit Buses ... ABSTRACT

Vikara and Holmén Page 1 of 27

1

Ultrafine Particle Number Concentrations from Hybrid Urban Transit Buses using On-Board Single-Diameter SMPS Measurements

Derek Vikara Graduate Research Assistant Environmental Engineering Program University of Connecticut 261 Glenbrook Road 412-841-6091 (cell) Storrs, CT 06269 [email protected] Dr. Britt Holmén Associate Professor Department of Civil and Environmental Engineering University of Connecticut 261 Glenbrook Road Storrs, CT 06269 [email protected] 860-486-3941 (Tel.)

860-486-2298 (FAX)

Manuscript Submitted to: 85th TRB Annual Meeting, January 22-26, 2006, Washington, D.C. TRB Transportation and Air Quality Committee (ADC20) Submitted: August 1, 2005 Revised: November 15, 2005

Final Revision: March 27, 2006 Word Count : 6296 + 2 Figures = 6796 words Paper Number 06-2253


2

ABSTRACT Recent studies have focused on mass-based quantification of gas and particulate matter (PM)

transit bus exhaust emissions under laboratory dynamometer testing conditions because transit

buses frequent heavily-populated areas and are major contributors to the ambient fine

particles in urban regions. This study examines the ultrafine particle (Dp < 100 nm) number

concentrations and size distributions for conventional diesel and hybrid-electric diesel transit

buses using on-board exhaust measurements with a TSI Scanning Mobility Particle Sizer

(SMPS) operated in single-diameter mode. The buses were run on three bus routes (freeway

commuter, local and high grade arterial), different fuels (No. 1 diesel and ultralow sulfur

diesel), and with and without diesel particulate filter (DPF) aftertreatment at different points

throughout the study. To our knowledge this is the first urban bus emission study conducted

while driving actual bus routes. Particle number distributions varied by route, but not by bus

type or fuel sulfur content. Particle number concentrations were higher on average for high-

load routes (freeway 65 mph commuter and steep grade). There were no significant

differences in particle number distributions between the 2003 parallel hybrid-electric and the

2002 conventional diesel bus types, likely due to the similar diesel engine specifications and a

hybrid control strategy that was not optimized for particulate emissions benefits. For both bus

types, use of a DPF resulted in 95-99% number concentration reductions for all diameters

sampled (10-130 nm) on all routes. The study results point to diesel particulate filters as the

cost-effective solution for achieving particulate emissions control from diesel transit buses.


3

INTRODUCTION

Particulate matter (PM) is a general term used to describe the mixture of solid particles and

liquid droplets in the air. Primary particles produced from diesel and gasoline engines are of

major concern because of their adverse effects on human and animal health, decreased

atmospheric visibility (haze), and tarnishing of buildings [1]. Vehicle particulate emissions

are important because the number of vehicles on the road and the vehicle miles traveled

(VMT) have been increasing steadily over the years. As a result, transportation-derived PM

has gained attention from researchers and health professionals alike. Ultrafine particles (Dp <

100 nm), which dominate the number concentration in diesel exhaust, have longer

atmospheric residence times and represent a higher human health risk than larger diameter

particles which leave the atmosphere quickly via settling. Approaches being implemented to

meet current and upcoming vehicle PM emissions standards include: cleaner burning fuels,

more efficient engines, and retrofit technologies such as catalytic particulate filters. This

study demonstrates the feasibility of real-world on-board tailpipe particle emissions

measurement and examines three control factors to reduce transit bus emissions: (1) diesel

fuel sulfur content, (2) parallel hybrid technology, and (3) DPF aftertreatment.

Particulate pollution from diesel engines has contributed significantly to air quality

problems in the United States and many other countries. Despite strict heavy-duty highway

engine standards that are to take effect in 2007 [1], diesel engines will still contribute large

quantities of particulate matter (PM) and nitrogen oxides (NOx) which can both lead to

serious human health effects [2]. Heavy-duty engine emission standards are based on the


4

mass of a particular pollutant emitted during dynamometer testing. Dynamometer emissions

measurements test a given vehicle (or engine) under driving conditions (or operating

conditions for heavy-duty engines) dictated by the laboratory equipment despite the fact that

driving conditions experienced in real-world, everyday-driving can be quite different from

those experienced in standardized laboratory tests.

The main purpose of this paper is to compare the ultrafine particle (Dp < 100 nm)

number distributions in the exhaust of two types of urban transit buses, conventional diesel

and hybrid-electric diesel, using an on-board, single-diameter SMPS sampling technique

while driving on three bus routes in Hartford, CT. The ultrafine emissions were quantified

while operating on two diesel fuels (No. 1 and ultralow sulfur diesel (ULSD)) and with and

without a DPF. The focus of this work is on particle number measurements, not volume- or

mass-weighted distributions (that can be derived from particle number distributions). Number

measurements, unlike gravimetric mass measurements, allow comparison between bus types

even when diesel particulate filter (DPF) aftertreatment is employed because of the higher

detection limit for particle counting instruments compared to gravimetric mass measurement.

To our knowledge this is the first on-board, on-road tailpipe particle emissions testing study

of hybrid and diesel urban transit buses.

DIESEL ENGINE-DERIVED PM

Urban bus engine emissions standards are stricter than other heavy-duty diesel engine

standards because buses frequent urban areas more than other heavy-duty vehicles and human

exposure to diesel exhaust is at its highest in urban areas. Diesel-powered buses comprise

almost 80% of the U.S. bus fleet and offer excellent reliability and availability. However,


5

despite recent improvements in diesel-powered bus emissions, (80% less PM and 20% less

NOx than 1993 model year buses), conventional diesel buses currently being manufactured

will not be able to meet 2007 Federal EPA standards unless emissions are further reduced.

The use of retrofit technologies like diesel particulate filters in combination with ULSD can

substantially reduce gas and particle emissions [3].

Modifications to bus designs are another technological solution. Hybrid-electric

vehicles (HEVs) have recently become commercially available for both light and heavy-duty

vehicle applications. Hybrid-electric vehicles employ two power sources to directly drive the

vehicle wheels: an internal combustion engine (ICE) and an electric motor. The electric

motor, powered by an advanced energy storage device (battery), lowers the demand placed on

the combustion engine. Hybrid buses have shown 50-90% PM mass reductions and 25-65%

increases in fuel economy compared to conventional diesel bus types [4,5] because hybrids

allow the ICE to be used less and over a narrow range of engine speeds and loads [6] than in a

non-hybrid design. The ICE can therefore be downsized compared to a conventional diesel

bus, resulting in lower engine weight and higher overall vehicle efficiency. In addition,

regenerative braking allows for further reductions in energy loss by storing the energy used to

slow down or stop a vehicle in the battery for use later [6]. There are two basic HEV

configurations: series and parallel. A series hybrid configuration ICE is used as an electric

generator. The electricity generated is either stored in batteries or sent to an electric motor

which drives the wheels. The series hybrid ICE is therefore only operated when the battery

drops below a minimum charge level. In the parallel hybrid configuration, there are two

power paths so that either the ICE or the electric propulsion system can be used to drive the

wheels. The parallel hybrid design has the ability to turn off or reduce dependence on the ICE


6

and run on the electric motor from the battery pack for short duration in-town driving, which

can lower emissions significantly [7]. For heavy-duty applications, hybrid manufacturers

have focused on urban transit bus demonstration vehicles for various transit operators [8].

Experiences with the hybrid-powered buses have been largely positive, especially with the

buses owned by New York City Transit (NYCT) where 40-foot Gillig low-floor buses are

equipped with the GM/Allison transmission series hybrid drive system and diesel particulate

filter (DPF) aftertreatment [8]. Series hybrids with DPF’s in the NYCT fleet have emission

rates (g/mi) comparable to the NYCT compressed natural gas buses and 10-16 percent better

fuel economy than the conventional diesels in the NYCT fleet [8].

HYBRID-ELECTRIC DIESEL BUS EMISSIONS STUDIES

Particle emissions from NYCT hybrid-electric buses were considerably lower and the

hybrid fuel economy was better than for a standard diesel bus [9]. It is important to note,

however, that the test hybrid buses were equipped with DPFs whereas the diesel bus used for

comparison did not have DPF aftertreatment. The Northeast Advanced Vehicle Consortium

(NAVC) performed gas and PM mass emissions testing on 1998 Orion VI series hybrids and

1998 NovaBUS RTS conventional diesel buses using dynamometer testing. Results indicated

that the hybrid bus PM mass emissions were 50 – 99% lower over four driving cycles than the

conventional diesel bus [9]. It is important to note that the majority of research devoted to

particulate emissions from hybrid vehicles has primarily focused on gravimetric mass PM

measurements with little to no attention directed toward particle number emissions. Several

authors have also noted the importance of acquiring “real-world”, on-road emission

measurements during actual vehicle use [10,11]. Recent developments in on-board emissions


7

instrumentation have been widely recognized as enabling emissions testing from vehicles as

they operate under actual conditions. The main advantage to this approach is that data is

collected under “real-world” driving conditions at any location traveled by the vehicle [10].

On-board data collection consists of equipping a vehicle with instrumentation that collects

data on: operating parameters (i.e., vehicle speed, engine speed, engine load, air-to-fuel ratio,

vehicle location) simultaneously with time-resolved gas and particle exhaust concentrations.

Differences in vehicle emissions as a result of variations in roadway type, vehicle location,

operating parameters, driver, and many other factors can be characterized and explored much

more realistically than with any other emissions measurement method [10]. Uncertainties that

are related to dynamometer tests and roadside experiments are eliminated because multiple

instruments are used to simultaneously measure many different factors that are occurring

under “real-world” driving. The present study fills a gap in real-world urban transit bus

emissions research by using an on-board emissions measurement approach to quantify particle

number concentrations and size distributions.

METHODOLOGY

On-board emissions testing for particulate number concentration emissions from in-

use conventional diesel and hybrid-electric diesel transit buses was conducted using the

Scanning Mobility Particle Sizer (SMPS) after exhaust dilution with an ejector-diluter mini-

dilution tunnel. The buses were run on three different bus routes, with two different fuels

(No. 1 diesel and ultralow sulfur diesel fuel) and with and without an aftertreatment device

(diesel particulate filter, DPF) at different points throughout the study. The SMPS was


8

operated in single-diameter mode and size distributions were generated based on averaged

data collected over multiple sampling days on a given route.

STUDY VEHICLES AND SAMPLING SCHEDULE

Four separate New Flyer low-floor 40-passenger buses (2 diesels and 2 hybrids) from

the in-service Connecticut Transit (CTTRANSIT) bus fleet were sampled between January

and November 2004. Two of the buses, fleet numbers 201 and 202, were 2002 conventional

diesel buses outfitted with 280 hp Detroit Diesel Corporation (DDC) series 40E engines. The

other two buses, fleet numbers H301 and H302, were 2003 hybrid-electric diesels with

Allison EP parallel hybrid transmission and Cummins ISL 280 engines. Both the hybrid and

conventional bus engines tested in this study had identical PM mass emissions certification

standards [12]. Sampling for the study began in January 2004, using No. 1 diesel fuel with a

sulfur content of 230-320 ppm and diesel oxidation catalyst (DOC) aftertreatment. The fuel

was changed to ULSD (8 – 50 ppm sulfur) in July 2004, and the diesel oxidation catalysts

(DOC) were replaced with DPF’s in October 2004 for the last phase of sampling. Hybrid bus

types were equipped with either a dual brick diesel oxidation catalyst (DOC) or a Johnson-

Matthey CRT™ particulate filter, whereas the diesel bus types had either a single brick DOC

or an Engelhard DPX particulate filter.

Routes selected for this study were all in and around the city of Hartford, Connecticut.

The three-route sequence driven each day of sampling consisted of: (1) the Enfield commuter

bus freeway driving route (I-91 between Hartford and Enfield, CT), (2) a local arterial city

stop-and-go driving route complete with simulated bus stops (Farmington Avenue/Asylum

Avenue), and (3) a steep suburban arterial route with a significant grade portion (Route 44,


9

Avon). Prior to running the freeway route each day the bus engines were subjected to 15 min

warm-up run on the freeway and the bus air-conditioning unit was never in operation during

testing.

Freeway Route (Enfield): High speed (typically 60-65 mph) and relatively high

engine load (82%) steady-state cruise dominated the operating mode for this 34 mile

(roundtrip) freeway route.

City Driving Route (Farmington): The Farmington route, six miles long in each

direction, began in downtown Hartford and ended in West Hartford, CT. Farmington Avenue

is a local arterial with significant start and stop driving, both due to traffic signals and bus

stops (the buses stopped for simulated passenger pickup at every third bus stop on the

Farmington route). Frequent accelerations and decelerations dominate the majority of the

Farmington route and speed ranged from zero to 40 mph along the arterial, despite a 35 mph

speed limit.

High/Low Grade Route (Avon): The Avon route, 5.2 miles in each direction,

consisted mostly of travel on local arterials where the average speed was 28 mph with a

posted speed limit of 45 mph. The focal point of this route was the steep upgrade (up to 9%)

and downgrade over Talcott Mountain, in which relatively low vehicle speed (35 mph), high

engine load (95%), and high RPM (1838) dominated the uphill climb and a slightly higher

vehicle speed (37.5 mph), low engine load (1.2%) “coasting” dominated the downhill portion

of the mountain. For data analysis, this route was subdivided into two separate subroutes,

upgrade and downgrade, to evaluate the emissions performance of the vehicles under high and

low grade driving conditions.


10

All engine parameter data for the study was logged from the vehicle diagnostic ports.

The diesel buses communicated using the Society of Automotive Engineers (SAE)

J1587/J1708 protocols, and the hybrid buses used the SAE J1939 protocol. Engine values, for

example engine speed (RPM), engine load (%), and vehicle speed (MPH), were collected on a

second by second basis for both bus types – by Cummins “InSite” software for the hybrid

buses and ProLink 9000 for the diesel buses. An additional scantool, a Vansco USB Data

Link Adapter (DLA), was also used on both the hybrid and diesel bus types from April to

November 2004.

PARTICLE NUMBER CONCENTRATION

Ultrafine particle distributions were quantified using an SMPS, an instrument consisting of an

electrostatic classifier (EC) and a condensation particle counter (CPC). Particles are sorted

based on mobility diameter in the EC and sent to the CPC to be counted. The TSI, Inc. Model

3936 SMPS outfitted with the long DMA and 3025A ultrafine CPC was operated at 1.5 L/min

in the size-selective mode. In size-selective mode, the operator sets the DMA to a specific

voltage corresponding to a given particle mobility diameter, Dp; thus only particles of that Dp

will pass through to the CPC to be counted. Particle diameters of 10, 20, 40, 80, 100 and 130

nm were sampled. The SMPS procedure consisted of sampling a given particle diameter for a

certain amount of time during a particular driving route. Diameter selections were random to

avoid bias in data collection and to acquire data for each diameter at different physical

locations along each route for each sampling day over the entire field study. Generation of

the full ultrafine particle size distribution required data aggregation from multiple days of

sampling on each route.


11

Engine exhaust was sampled through a single-stage, constant dilution ratio mini-

dilution system based on an ejector-diluter as previously described [13]. Incoming exhaust air

was directed to the ejector-diluter (Dekati Ltd.) by a 5/8 inch diameter pipe wrapped with heat

tape to prevent heat loss (temperature > 100oC). The dilution air was applied from an air

compressor located on a trailer behind the bus and passed through a series of water

condensers and water traps to minimize the moisture entering the silica gel dryer tube before

dilution air passed through activated charcoal and a HEPA filter to remove hydrocarbons and

particles, respectively. Orifice meters placed in the exhaust sample line (which entered the

bus through the rear roof hatch) and in the dilution air line near the ejector-diluter allowed

continuous monitoring of sample and dilution air flows. Dilution ratios ranged from 25-32

during the study. The SMPS was located on a shock-reducing mount attached to one of the

rear bus seats and pulled sample from the dilution tunnel attached to the overhead hand rail.

More details about the sampling setup are available in reference [14].

Ten raw CPC counts, collected at 0.1 second resolution, were averaged to obtain one

second particle counts. The one-second CPC count data was converted to particle number

concentration (#/cm3) based on a correction factor determined by particle electric mobility,

aerosol and sheath flow rates and DMA geometry [15]. The final particle concentrations

(dN/dlogDp) were computed as:

( ) DRCFcmQ

QSDpd

dNpD

M

Ccm **min

sec60*sec

301**][

log 310#

3 =

Where S10 is the one second particle count, QC is the CPC inlet flowrate (1.5 lpm), QM is the

SMPS inlet aerosol flowrate (1.5 lpm), CFDp is the correction factor for diameter Dp, and DR


12

is the mini-diluter dilution ratio. It should be noted that the data representing the average and

standard deviations of particle concentrations by diameter on a given route were collected

from multiple sampling dates.

STATISTICAL ANALYSIS

The main focus of the statistical analysis was to identify important predictor variables

that influence particle number concentration by diameter. An Analysis of Variance

(ANOVA) was conducted for two separate fuel/aftertreatment configurations: (1) to compare

emissions data when operating on No. 1 and ULSD fuels (with DOC aftertreatment), and (2)

to compare emissions with and without the DPF when operating on ULSDfuel. The ANOVA

analysis was conducted separately for each of the six sampled particle diameters (10, 20, 40,

80, 100, 130 nm). Particle number concentration was modeled against three independent

variables:

Bus = Diesel or hybrid bus type.

Fuel = No. 1 diesel fuel or ULSD fuel; (ULSD or ULSD + DPF on 2nd ANOVA).

Route = Enfield, Farmington, Avon Upgrade or Avon Downgrade.

All factors were considered “fixed factors” with the exception of “route.” Route was

considered a “random factor” in the ANOVA analysis because traffic volumes and flow

varied from day to day and could not be ideally controlled. The natural log (Ln) of the

particle number concentration (N) was used in the ANOVA to achieve normal sample

distributions. The following null hypothesis was tested at the 95% confidence level for all

bus, fuel and route configurations:


13

Null Hypothesis = HO = RouteXFuelXBusXDpN,,, = RouteYFuelYBusYDpN

,,,

Alternative Hypothesis = HA = RouteXFuelXBusXDpN,,, ≠ RouteYFuelYBusYDpN

,,,

RESULTS

A substantial amount of SMPS data was collected for each diameter on each route and for

each bus with the exception of 10 and 100 nm data for hybrid buses on Avon downgrade and

for 40 and 100 nm on Avon upgrade, both using No. 1 diesel fuel. For these given diameters,

no SMPS data was collected because the uphill and downhill portions of the Avon route were

so short (approximately three minutes each) compared to other routes in the study (each about

30 minutes), and the random selection of classifier voltages resulted in no data for these

diameters on these upgrade/downgrade routes.

Particle number distributions varied depending on the driving route and engine

parameters. In the discussion below, the particle results for operation on No.1 diesel and

ULSD with DOC aftertreatment are discussed first, followed by discussion of the effects of

DPF aftertreatment.

Enfield Route Distributions The Enfield route, characterized by steady-state freeway driving at high (~ 65 mph)

speed, had a mean particle number distribution that peaked around 40 nm for both bus and

fuel types (Figure 1 a,b). For different fuel types (No.1 Diesel and ULSD), the freeway 10

and 20 nm particle concentrations were lower for ULSD operation, but for all particle

diameters >40 nm concentrations were higher when operating on ULSD than on No. 1 diesel


14

for both bus types. These observations may partly reflect differences in ambient temperature

and relative humidity conditions for the two fuel conditions (No.1 Diesel = winter sampling

Jan-June; ULSD = summer July - September) The Enfield number distributions were

similar between the hybrid-electric and diesel buses and, with the exception of 10 nm, mean

particle concentrations were typically lower for the hybrid bus than the diesel bus on the

Enfield route. The distributions in Figure 1 a,b are typical of steady-state engine operation

[16] and reasonable because the Enfield freeway route allows the vehicle to operate under

fairly steady-state conditions (constant speed with constant RPM and high engine load) with

few transient events.

In a parallel hybrid design, the diesel engine provides primary power to the vehicle

with the electric motor assisting only in periods of fast acceleration and hill climbs [7].

Therefore, on the Enfield trip, the diesel engine would provide the majority of the vehicle

power; explaining the similar diesel and hybrid bus number concentrations for all diameters

and fuel types on the Enfield route. The similarly-sized engines in these hybrid-electric and

conventional diesel buses were apparently working equally hard to maintain the 65 mph

cruise speed. The hybrid system contributes very little power to the drive shaft on high speed

routes like Enfield, despite high engine load. It should be noted that series hybrids with

smaller diesel engines would not typically be capable of 65 mph freeway commuter route

operation.

At high vehicle speeds, particulate matter mass emissions increase due to high engine

load, higher exhaust flow and increased exhaust temperatures [17]; consistent with the

observed higher particle concentrations at larger diameters (40 nm to 130 nm) on the Enfield

route.


15

Farmington Route Distributions

The peak for the distribution of the Farmington route, classified by significant start-

and-stop driving with high accelerations and long periods of idle, shifted away from the 40

nm peak seen in the Enfield distribution and towards 10 or 20 nm (Figure 1 c,d). The Enfield

distributions also have a broad second peak at 80 - 100 nm for both bus and fuel types (Figure

1 c,d). Particle concentration values varied greatly for a given diameter on the Farmington

route due to stop-and-go driving, causing the standard deviations to be relatively large. Hybrid

technology initiates when the vehicle operates under short periods of fast acceleration and

grade inclines [7]; most of which occur in the Farmington route, especially the rapid

accelerations from periods of idle at the numerous stops. The parallel hybrid design uses both

power sources during accelerations and the battery source typically dominates during

accelerations from idle in order to achieve lower emissions and better fuel economy. Because

the Enfield route was characterized by higher engine loads than the Farmington route, particle

number concentrations at all ultrafine diameters (Dp < 100 nm) were expected to be greater

for the Enfield than the Farmington route and this was observed for diameters > 10 nm.

Surprisingly, emissions were not significantly lower for the hybrids compared to the

conventional diesels on the Farmington route and for the 10 and 20 nm diameters the hybrid

bus emissions were higher than those for the conventional diesel for both No.1 diesel and

ULSD operation. Apparently, on the Farmington route the hybrid bus diesel engine was

working equally hard as for the diesel bus type with little help provided from the battery or

electric motor. The parallel design or control strategy set by the manufacturer may be the

main reason for no noticeable improvement in particle number concentrations on the

Farmington route between the conventional and hybrid bus types. For example, the fact that


16

the hybrid buses tested were capable of 65 mph freeway cruise speeds and experienced little

to no reduction in performance on significant (9 %) grade (see Avon Upgrade below),

suggests that the hybrid controller was set for optimal vehicle performance, not lowest

emissions. Based on the Farmington results, we speculate that, during the testing period, the

hybrid bus was not programmed to take full advantage of its regenerative braking capabilities

possibly due to manufacturer concern about the real-world in-service performance of the

nickel metal hydride (NiMH) battery type unique to these prototype Ep-40 buses.

Effects of Road Grade

The upgrade and downgrade portions of the Avon route had a similar number

distribution to that seen for Farmington with maxima at 10 nm (Figure 1e) or 20 nm (Figure

1f). The upgrade number concentrations were 30-90% higher than the Farmington route

concentrations for all sampled particle diameters, whereas the Avon downgrade number

concentrations (Figure 1 g,h) were 12-87% lower than measured for the Farmington route.

Engine load is directly proportional to road grade and increased load produces

particles with higher elemental carbon content (soot). High road grade (10-19%) reduces

the fuel economy of most bus types (hybrid electric, conventional diesel and compressed

natural gas) [18]. It is also frequently noted that diesel engine particle number

concentrations increase with increasing engine load [16,18-21] and that operation at high

load, high speed and high road grade increases particle number and mass concentrations [16-

19]. The average particle concentrations were much lower on the downgrade than the

upgrade Avon route, mostly due to differences in engine load. The other diameters had


17

particle number concentrations lower than those for 10 and 20 nm (Figure 1 e,f). On Avon

upgrade with No. 1 diesel fuel, the hybrid bus type emitted much lower particle

concentrations for the smallest nuclei mode particles measured (10 and 20 nm) compared to

the conventional diesel (Figure 1f); whereas the diesel bus 10, 20 and 40 nm particle

concentrations were lower than the hybrid bus type under ULSD fuel operation (Figure 1e).

Grade on the Avon upgrade route was not constant, therefore, vehicle operation, and

corresponding particle emissions, varied as a function of grade change. Therefore, the large

standard deviations in the Avon distributions can be attributed to the variation in the grade

(and corresponding engine load) experienced while going over the mountain. Upgrade

distributions were bimodal with local maxima at 10 nm and over 40 – 80 nm for both bus

types, similar to a previous study [22] that observed peaks at 40 nm and 80 – 90 nm for diesel

bus exhaust emissions in a tunnel. It is difficult to tell if the bimodal distribution presents

itself in Figure 1f,h due to missing hybrid data, although the Avon upgrade distribution for the

diesel bus was weakly bimodal.

The Avon upgrade route is characterized by a relatively constant speed (35 mph) and

high load (95%) on the vehicle. The typical speed for Avon upgrade was considerably lower

than for Enfield, the measured route-average engine load was slightly higher (82% on

Enfield), and the mean engine speed values for these routes were similar (1838 rpm Avon

upgrade, 1779 rpm Enfield). In comparison to Farmington, the Avon upgrade route involved

higher load, engine speed and mean vehicle speed (Farmington = 28 mph) because the Avon

trip involved hill climbing with no stops, whereas Farmington involved many periods of idle

that reduced the average vehicle speed. The overall higher mean engine load explains the

higher Avon upgrade emissions at all diameters compared to Farmington (Figure 1 c-f). The


18

vehicles had similar mean vehicle and engine speeds on Avon upgrade and downgrade, but

the percent load was significantly greater for Avon upgrade, resulting in higher number

concentrations [16-18], as seen in Figure 1 e-h.

The diesel bus Avon downgrade concentrations were 90% -97% lower than for Avon

upgrade for all diameters. In contrast, the hybrid bus type upgrade/downgrade reduction

ranged between 60% - 90%, much lower than that observed for the conventional buses and

similar to a previous diesel bus study showing 66 – 95% reduction [19].

Effects of Fuel Sulfur Content

The concentration of nanoparticles (< 50 nm), which are primarily comprised of

sulfates [16], should decrease with the implementation of ULSD fuel because most of the

sulfur in the fuel to make those smaller nucleation particles is removed. Lower mean

nanoparticle emissions with ULSD was observed only for Enfield and Farmington 10 and 20

nm concentrations for both bus types. However, on the Avon upgrade route, 10 nm particle

concentrations were higher using ULSD than No. 1 diesel fuel for the hybrid bus type (Figure

1e). The fact that these particle reductions occurred only for diameters <40nm was expected

because sulfate particles are expected in the nanoparticle (<50 nm) size range.

In addition, there was an 8-160% increase in the 40 - 130 nm particle concentrations using

ULSD on the Enfield route for both bus types compared to using the No. 1 diesel fuel. This

shift in the distribution towards larger particles with the reduction in fuel sulfur content is

expected for high load, high speed freeway cruise operation that increases the available


19

surface area for semivolatile sulfate and organic compound condensation, thereby reducing

nuclei formation (10 and 20 nm) . The observed similarity between particle emissions on No.

1 diesel and on ULSD fuel by both bus types (see error bars, Figure 1) can be attributed to the

removal of nanoparticle precursor species by the diesel oxidation catalysts on these relatively

new buses and variation in DOC efficiency under different operating conditions between

routes.

Effect of Diesel Particulate Filter Aftertreatment

The addition of the DPF dramatically reduced particle number concentrations at all

diameters compared to using ULSD and an oxidation catalyst (Figure 2). DPF type and make

were different for each bus type, however number concentrations for all diameters regardless

of bus type and route were reduced by 95% - 99%. Measured concentrations with the DPF

were close to ambient levels as raw CPC counts did not exceed three times the HEPA data

counts. The measured reductions in number concentration with DPF aftertreatment are

similar to those observed previously by others for PM gravimetric mass measurements.

Statistical Interpretation

Predicting particle emissions from heavy-duty buses as a function of bus type, fuel type and

route can provide insight as to the benefits of a given fuel in terms of particulate matter

emissions and help assess which bus types offer lower emissions under certain driving

conditions. For the ANOVA comparisons, all of the data for a single diameter for all bus

types, routes and fuel types were tested separately. All interactions between factors were

investigated, but none were significant. The ANOVA results comparing emissions during


20

operation on ULSD vs. No. 1 diesel fuel showed lower 10 and 20 nm emissions when

operating on ULSD compared to No. 1 diesel, implying a benefit in terms of emissions with

the fuel sulfur reduction. However, all diameters larger than 20 nm had higher mean

emissions values on ULSD than on No. 1 diesel and corresponding p-values for “bus” were

not significant.

Results from the No.1 vs. ULSD fuel type ANOVA analysis indicate that factors affecting

particle concentration emissions varied by diameter. Route was always a significant factor

affecting particle concentrations. It is important to note that although the “Route” factor was

determined as random, the variability in route between days cannot be statistically analyzed

properly due to variation in the single-diameter SMPS sampling method – daily changes in

spatially varying operating parameters (% load, vehicle speed or engine speed, etc…) on a

given route likely affected particle emissions. Fuel was a significant factor in determining 10

nm particle concentrations, mostly attributed to the reduction in sulfur content which lowered

nanoparticle concentrations. Particle number concentrations between bus type on the ULSD

vs. ULSD + DPF ANOVA are mostly likely attributed to the dramatic reduction caused by the

DPF (concentrations essentially ~ ambient air) rather than to the effects of bus alone.

CONCLUSIONS and BROADER IMPACTS

Surprisingly, the results indicate no statistically significant differences in particle number

concentration between the parallel design hybrid-electric diesel and the conventional diesel

bus types for all routes, fuels and diameters studied. Particle number concentrations on all

routes resulted in similar distributions by diameter between bus types. The Enfield route

(highway driving) had a particle concentration peak at 40 nm, whereas the Avon upgrade


21

(steep upgrade), Avon downgrade (downhill route) and Farmington (city start and stop

driving) had peaks in the 10 – 20 nm range with secondary maxima occurring around the 80 –

100 nm range, resulting in almost bimodal distributions. Avon downgrade had the lowest

particle concentrations on all diameters for all fuel types and bus types because the engine

demand going downhill was extremely low (engine load approximately 1% on average). The

Avon upgrade route, on average, had the highest 10 and 20 nm concentrations for all fuel

types and bus types compared to other routes, because the engine demand was so high (engine

load approximately > 95%), yet vehicle speed was relatively low (35 mph average).

The use of ULSD fuel reduced the number concentrations of some diameters within the nuclei

mode range (10 nm – 40 nm, which are mostly comprised of sulfates) for both bus types on

some routes, however a major reduction in all nuclei mode particles with the switch to ULSD

was expected but not seen (Figure 1) for any bus/route combination. This result differs from

previous studies and may be explained by the use of diesel oxidation catalysts on both bus

types for No.1 and ULSD operation. The addition of the diesel particulate filters resulted in a

significant reduction in particle concentration in the exhaust for all bus types, diameters and

routes, approximately 95 – 99% (Figure 2).

This study is the first to use the on-board emissions measurement technique to determine

particle number concentration distributions from in-use transit buses. This technique has

advantages over using laboratory testing which fails to account for “real-world” driving

conditions. Results were similar to some previous studies in terms of particle distribution

trends, however these single-diameter SMPS results contradict many other studies which

indicate a significant emissions benefit to hybrid buses, although the majority of previous


22

studies examined series hybrids with smaller diesel engines and often compared hybrids with

a DPF to conventional diesel buses without aftertreatment. The Allison parallel design hybrid

buses, in their prototype control configuration, do not have any significant emission benefits

over the conventional late-model year diesel buses tested, but may have other fuel economy

and maintenance benefits that are not addressed by this study. The results of this study

suggest that transit bus fleet emissions from relatively new diesel buses (2002/2003; Cost ~

$250,000) are most economically reduced by installation of DPF’s (Cost = $5000 to $7,000)

rather than investing in a parallel hybrid bus (Cost ~ $500,000). It is not known whether a

series hybrid design will offer more emissions benefits without sacrificing the other

advantages of the hybrid bus such as lower noise, smoother rides and performance

characteristics comparable to conventional diesel transit buses on freeway commuter routes

and routes with high grade (up to 9% in this study).

ACKNOWLEDGMENTS

This research was sponsored in part by the Joint Highway Research Advisory Council of the

University of Connecticut and the Connecticut Department of Transportation through Project

03-8.


23

WORKS CITED

[1] Environmental Protection Agency. (2001). Control of Air Pollution from new Motor Vehicles: Heavy-duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur Control Requirements; Final Rule. 40 CFR Parts 69, 80 and 86. Jan 18, 2001.

[2] Chalupa, D., Morrow., P., Oberdörster, G., Utell, M., Frampton, M. (2004). Ultrafine Particle Deposition in Subjects with Asthma. Environmental Health Perspectives112, June 2004.

[3] Lanni, T., Chatterjee, S., Windawi, H., Conway, R., Rosenblatt, D., Bush, C., Lowell, D., Evans, J., McLean, R. (2001). Performance and Durability Evaluation of Continuously Regenerating Particulate Filters on Diesel Powered Urban Buses at NY City Transit. SAE Paper 2001-01-0511.

[4] McKain, D.L., N.N. Clark, T.H. Balon, P.J. Moynihan, S.A. Lynch and T.C. Webb (2000). "Characterization of Emissions from Hybrid-Electric and Conventional Transit Buses." SAE Paper 2000-01-2011.

[5] Wayne, W.S., N.N. Clark, R.D. Nine, D. Elefante (2004). "A Comparison of Emissions and Fuel Economy from Hybrid-Electric and Conventional-Drive Transit Buses." Energy & Fuels 18, 257-270.

[6] National Renewable Energy Laboratory. (2004). Hybrid Electric & Fuel Cell Vehicles. [Online]. http://www.nrel.gov/vehiclesandfuels/hev January 17, 2005.

[7] Sullivan, R. (1999). The Technical Background of Hybrid Electric Vehicles. Office of Transportation Technologies. US Department of Energy.

[8] Meyer, N., Rideout, G. (2002). Allison EP Systems Electric Hybrid Test Program Regulated Emissions and Fuel Economy Results. Environmental Technology Centre Emissions Research and Measurements Division. ERMD Report 02-25-1.

[9] Chandler, K., Walkowicz, K., Eudy., L. (2002). New York City Transit Diesel Hybrid Electric Buses: Final Results. DOE/NREL Transit Bus Evaluation Project.

[10] Frey, C., Rouphail, N., Unal, A., Colyar, J. (2001). Measurement of On-Road Tailpipe CO, NOx, and Hydrocarbon Emissions Using a Portable Instrument. Proceedings: Annual Meeting of the Air & Waste Management Association. June 24-28.

[11] Hung, W., Tong, H. (2002). Review of Vehicle Emissions and Field Consumption Modeling Approaches at Signalized Road Network. Hong Kong Polytechnic University. Department of Civil and Structural Engineering.

[12] Environmental Protection Agency. (2005). Federal Register. [Online] May 9, 2005. http://www.epa.gov/fedrgstr/

[13] Holmén, B.A. and A. Ayala (2002) Ultrafine PM Emissions from Natural Gas, Oxidation-Catalyst Diesel and Particle-Trap Diesel Heavy-Duty Transit Buses, Environmental Science & Technology 36, 5041-5050.


24

[14] Holmén, B.A., Z. Chen, A. C. Davila, O. Gao, D. M. Vikara. Particulate matter emissions from hybrid-electric diesel and conventional diesel transit buses: fuel and aftertreatment effects. Final Report. Connecticut Cooperative Highway Research Program, August, 2005. http://www.engr.uconn.edu/ti/Research/crp_completed.html.

[15] TSI, 2001. Model 3936 SMPS (Scanning Mobility Particle Sizer) Instruction Manual. TSI Incorporated. P/N 1933796. Revision F. February 2001.

[16] Kittleson, D. (1998). Engines and Nanoparticles: A review. Journal of Aerosol Science. 29, 575-588.

[17] Kittleson, D., Watts, W.F., Johnson, J.P. (2004). Nanoparticle emissions on Minnesota highways. Atmospheric Environment 38, 9-19.

[18] Dwyer, H., Tang, J., Brodrick, CJ., Khau, L., Becker, C., Wallace, J. (2002). The Influence of Grade on the Operating Characteristics of Conventional and Hybrid Electric Transit Buses. SAE Technol. Pap. 2002-01-3118.

[19] Brown, J., Clayton, M., Harris, D., King Jr., F. (2000). Comparison of the Particle Size Distribution of Heavy-Duty Diesel Exhaust Using a Dilution Tailpipe Sampler and an In-Plume Sampler during On-Road Operation. J. Air & Waste Management Association 50, 1407-1416.

[20] Kean, A., Harley, R., Kendall, G. (2003). Effects of Vehicle Speed and Engine Load on Motor Vehicle Emissions. Environmental Science & Technology 37,3739 – 3746.

[21] Kittleson, D., Watts, W. F., Johnson, J. P. and Drayton, M. K. "Fine particle (Nanoparticle) emissions on Minnesota highways," Center of Diesel Research, Department of Mechanical Engineering, University of Minnesota, 2001.

[22] Jamriska, M., Morawska, L., Thomas, S., He, C. (2004). Diesel Bus Emissions Measured in a Tunnel Study. Environmental Science and Technology. 38, 6701-6709.


25

LIST OF FIGURES

Figure 1: Particle number distributions for all routes on ULSD (a, c, e, g) and No. 1 diesel (b, d, f, h) fuels. Average number concentration (dN/dlogDp) and one standard deviation are plotted for each diameter sampled. Log scale axes are identical for each route plot.

Figure 2: Particle number (dN/dlogDp) distributions plotted by bus type on a given route with one standard deviation error bars. Symbols indicate different fuel/aftertreatment. Note: linear scale x-axis.


26

10 100

105

106

Enfield ULSDa10 100

dc

bEnfield No. 1

10 100

105

106

Farmington ULSD10 100

g h

e f

Farmington No. 1

10 100

105

106

Avon Upgrade ULSD10 100

Avon Upgrade No. 1

10 100

105

106

Avon Downgrade ULSD10 100

Hybrid Diesel

Avon Downgrade No. 1

Diameter (nm)

Part

icle

Con

cent

ratio

n (#

/cm

3 )

Figure 1: Particle number distributions for all routes on ULSD (a, c, e, g) and No. 1 diesel (b, d, f, h) fuels. Average number concentration (dN/dlogDp) and one standard deviation are plotted for each diameter sampled. Log scale axes are identical for each route plot.


27

Figure 2: Particle number (dN/dlogDp) distributions plotted by bus type on a given route with one standard deviation error bars. Symbols indicate different fuel/aftertreatment. Note: linear scale x-axis.

Ultrafine Particle Number Concentrations from …bholmen/BAHwork/docs/TRB06-2253_VikaraHolmen...Ultrafine Particle Number Concentrations from Hybrid Urban Transit Buses ... ABSTRACT

Documents