Using Tracers to Quantify In-Cabin Concentrations of School Bus Exhaust and Crankcase Emissions

Recent attention has been given to concentrations of diesel exhaust particulate matter (PM) inside school buses that are attributable to the buses’ own exhaust. Previously a study of one bus used emission measurements and a fuel-based quantitative tracer to provide bus-specific results (Ireson et al. 2004). In another study, the sequential application of retrofit emission control technologies found, using non-specific optical methods, that crankcase vent emissions may be an appreciable contributor to in-cabin mass concentrations (Hill et al 2005). Since crankcase and exhaust PM emissions differ markedly in size and composition, the ability to differentiate between them is important.

Thomas W. Hesterberg, International Truck and Engine Corp.
Robert G. Ireson, Air Quality Management Consulting
John M. Ondov, University of Maryland
Barbara Zielinska, Desert Research Institute
Christopher S. Weaver, Engine Fuel and Emissions Engineering Inc.
Douglas R. Lawson, USDOE National Renewable Energy Laboratory
Mark E. Davey and L.-J. Sally Liu, University of Washington

Abstract

Recent attention has been given to concentrations of diesel exhaust particulate matter (PM) inside school buses that are attributable to the buses’ own exhaust. Previously a study of one bus used emission measurements and a fuel-based quantitative tracer to provide bus-specific results (Ireson et al. 2004). In another study, the sequential application of retrofit emission control technologies found, using non-specific optical methods, that crankcase vent emissions may be an appreciable contributor to in-cabin mass concentrations (Hill et al 2005). Since crankcase and exhaust PM emissions differ markedly in size and composition, the ability to differentiate between them is important.

The present study developed a new dual tracer method to quantify in-cabin concentrations of school buses’ exhaust and crankcase PM. The new method continues the use of an organo-metallic iridium complex dissolved in the fuel, and adds a high molecular weight deuterated alkane tracer in the lube oil. On-road emission sampling was used to establish exhaust and crankcase emission rates for both fine particles (PM_2.5) and particle-bound iridium and deuterated alkane tracer. Samples collected inside the bus and inside a lead vehicle traveling in front of the bus were analyzed for tracers to allow estimation of the contribution of the two tracer-tagged sources.

For the two buses studied, on-road PM_2.5 exhaust emission rates were typical of buses of this type and age. The first bus’s exhaust emission rate averaged 0.15 g/km over three tests, and the second bus averaged 0.14 g/km. Somewhat surprisingly crankcase PM_2.5 emissions are a significant fraction of exhaust emissions, averaging 0.03 g/km for bus 1 and 0.10 g/km for bus 2. Using in-bus measurements of Ir and d-alkane tracer and the mass ratios of emissions to tracers, the contribution of each bus’s exhaust and crankcase emissions to in-cabin concentrations was estimated. With windows closed, a relatively large fraction of in-bus PM_2.5 is contributed by crankcase emissions, ranging from about 9 to as high as 21 µg/m3. A substantially smaller amount is contributed by exhaust, ranging from about 0.2 to 2 µg/m3. Windows-open testing showed combined exhaust and crankcase contributions to drop to about 0.5 µg/m3or less for bus 1. For bus 2, windows-open crankcase contributions were approximately 2-3 µg/m3, while exhaust contributions were about 5 times lower.

In summary, we utilized a dual tracer method to estimate the in-cabin concentrations of PM_2.5 from school bus exhaust and crankcase vent emissions. This method proved to be sensitive and specific. The in-cabin PM mass attributable to crankcase emissions appears to be consistently higher than that attributable to exhaust, with closed-window conditions leading to higher contributions of both.

Methods

• Tracers intentionally added to:
  • Fuel: iridium (Ir) complex as a definitive tailpipe (exhaust) DPM tracer
  • Lube oil: deuteratedalkane(C36D74) as a definitive crankcase PM tracer
• On-road source testing using the “Ride Along Vehicle Emission Measurement “(RAVEM) system to establish quantitative PM:traceremission ratios
• Calculate in-bus tailpipe DPM and crankcase PM contributions from tracers
• 2 conventional school buses (MY 2000 and 2003)
• Testing on a primarily residential route in Seattle
• In-vehicle sampling for tracers, PM2.5and EC/OC
• Concurrent sampling in lead vehicle (~3 minutes ahead) for over-road background
• On-road source sampling to quantify emission rates of PM_2.5 and tracers from tailpipe and crankcase.

Source Testing with RAVEM

Results

PM_2.5 and Carbon Emission Rates

Bus 1 PM_2.5, Carbon, and Tracer Results

Source Testing Results – PM:Tracer Ratios

Bus 2 PM_2.5, Carbon, and Tracer Results

In-Vehicle Testing – Sensitivity of Tracer for Tailpipe (DPM_tp) and Crankcase (PM_ck)

• DPM_tp:Ir≈1000, PM_ck:d-alkane≈300
• Worst case lead vehicle uncertainties show high sensitivity of tracer method: -Ir±3 pg/m³ ⇒DPM_tp±0.003 µg/m³ -d-alkane±0.13 ng/m³ ⇒PM_ck±0.05 µg/m³
• Tracer ratios differ markedly between sources, allowing definitive differentiation -Ir_tp:d-alkane_tp≈5 -Ir_ck:d-alkane_ck ≈0.0005

PM Inside Bus Cabin

UMd PM_2.5 In-Bus Samplers

Discussion

Previous L.A. Study

• PM emissions typical of this vehicle class — 0.19-0.22 g/mi
• DPM:Ir mass emissions ratio = 662 -COV of 17% (5 runs, 10 samples) -Ir concentration of 1 pg/m³ ⇒ DPM concentration of 0.00066 µg/m³
• DPM and Iremissions consistent for both CSHVR and UDDS cycles

Previous L.A. Study: PM_2.5

Potential Tracer Issues

• DPM suppression by Ir–Possible, but emission tests show bus emission rates in expected range
• Reliability of Ir and d-alkane as proportional tracers
  • Ir and DPM_tp emissions both expected to be roughly proportional to fuel flow
  • d-alkane used is a naturally occurring component of lube oil, and PM_c expected to be primarily condensed lube oil (blowby) with small amount of combustion products and fuel

Previous L.A. Study: DPM_tp

Conclusions

• Dual tracer method is definitive and highly sensitive: DPM_tp±< 3 ng/m³; PM_ck±< 50 ng/m³
• Sampling issues highlight need for special care for paired filter analyses – data refinements continue
• Crankcase PM may be emitted at rates approaching exhaust DPM
• DPM_tp and PM_ck concentrations are highest with closed windows; much lower with ventilation
• In-bus DPM_tp is comparable/slightly higher than L.A. Ir study, but lower than non-specific studies
• In-bus PM_ck concentrations are substantially higher than DPM_tp, despite being emitted at a lower rate
• In-bus PM_2.5 concentrations much lower than the L.A. Ir study and other urban in-vehicle studies probably due to differences in route and regions
• Additional work is being planned