                         Supporting Technical Document
March 1, 2013
MEMORANDUM
SUBJECT:  Estimated Reductions in Particulate Matter Emissions from Light-duty and Medium-duty Gasoline Vehicles through Implementation of Tier 3 Regulations.
FROM:	Darrell Sonntag  -  USEPA/OTAQ/ASD
--------------------------------------------------------------------------------
TO:  EPA Docket # EPA-HQ-OAR-2011-0135
Overview
The NPRM Tier 3 RIA estimated reductions in particulate matter (PM) emissions from light-duty vehicles. The estimated PM reductions were modeled through reductions in particle-phase organic carbon (OC) which was projected to be reduced by 30% through improvements in controlling lubricating oil consumption. Similar reductions were projected for medium-duty gasoline engines that are regulated under Tier 3 standards. No reductions in particulate matter emissions were projected for light-duty diesel engines. The technical analysis underlying the 30% OC reduction assumption was based on an analysis of data collected from the Kansas City Light-Duty Vehicle Emissions Study (KCVES). 
Lubricating Oil Contribution to PM Emissions using the Kansas City Light-Duty Emissions Study
Sonntag et al. (2012) estimated the contribution of lubricating oil to light-duty PM emissions using data from the KCVES[1].  A brief description of the analysis is provided in this technical document. In 2005, the US Environmental Protection Agency (EPA) completed a vehicle emission sampling campaign in the Kansas City Metropolitan Statistical Area (KCMSA) to improve understanding of the contribution of light-duty gasoline vehicles to ambient particulate matter (PM). The KCVES recruited 496 light-duty vehicles in a stratified random sample from the KCMSA. The KCVES tested each vehicle on a chassis dynamometer using the California Unified Cycle. The chemical composition of the PM emissions was chemically analyzed from a subset of 99 vehicles. Measurements of particle-phase organic species were measured from 52 individual and composite PM samples obtained from these 99 vehicles. 
The contribution of lubricating oil to the elemental carbon (EC), organic carbon (OC), total carbon (TC), and total particulate matter (PM) emissions was estimated using hopanes and steranes as markers for the lubricating oil, with three polycyclic aromatic hydrocarbons used as gasoline markers. A fleet-weighted regression model was used to apportion particulate emissions to lubricating oil and gasoline, according to the population of the Kansas City vehicle fleet.
The vehicle samples from pre-1991 model years were analyzed separately from 1991-2004 model year vehicles. The older vehicles contribute disproportionately to the fleet-average PM emissions derived from lubrication oil. While only contributing 13.5% of the vehicle population, the pre-1991 vehicles contribute 70% of the oil-derived PM for the entire fleet, and 33% of the fuel-derived PM. Lubricating oil is a significant contributor to organic carbon emissions, with older vehicles emitting a large percentage of OC emissions from the lubricating oil. Lubricating oil was estimated to contribute 60% of the OC emissions from the pre-1991 vehicles, and only 21% of the OC for the 1991-2004 vehicles. The greater contribution of lubricating oil in the older vehicles is expected due to deterioration and malfunction of older vehicles. Elemental carbon emissions were dominated by the gasoline contribution, with no detectable contribution from lubricating oil. Fleet-average PM emission rates were calculated from weighting the contribution of each set of model year vehicles to their estimated population in the KCMSA. Based on this weighting, the contribution of lubricating oil emissions to PM was estimated to be 25%, which was roughly half the contribution of fuel to PM (47%).
A complete apportionment of the EC, OC, TC, and PM between fuel and lubricating oil was not possible due to varied chemical composition of PM among vehicles, uncertainties in the fuel and oil marker measurements, metals and dust in the exhaust, and PM measurement artifact. The uncertainty was estimated by applying a survey data analysis re-sampling method to the regression estimates. 95% confidence intervals for the lubricating oil contribution to the fleet-average PM emissions range from 13% to 37%. More information regarding the data, analysis steps and model assumptions, used to reach these conclusions are documented in Sonntag et al. (2012)[1]. 
Estimated Reduction in PM emissions by Tier 3 standards
The Tier 2 PM emission standard is 10 mg/mi on the FTP cycle. MOVES2010 estimates that the average PM emissions from 2005 and later light-duty gasoline vehicles are well in compliance with the Tier 2 standards. Using MOVES2010 light-duty PM emission rates to simulate an FTP cycle, the FTP composite emissions for passenger cars and trucks are estimated as 2.72 mg/mi and 3.08 mg/mi, respectively when the vehicles are between 0-3 years old. MOVES2010 predicts that 2005 model year vehicles will eventually exceed the Tier 2 standards when the vehicles are aged to 20 years. MOVES2010 assumes that PM emission rates for 2006 and later model year vehicles begin at the same levels as PM emissions from new 2005 vehicles, and deteriorate at same rate as they age. 
The proposed Tier 3 PM regulations implement tighter FTP and supplemental FTP PM standards (US06 cycle) at a full-useful life of 150,000 miles. The proposed Tier 3 FTP standard is 3 mg/mi, and the US06 standard is 10 mg/mi for passenger cars and 20 mg/mi for light-duty trucks. In order to achieve the proposed Tier 3 standards, the US EPA believes that current technology gasoline vehicles will need to have both lower emissions and improved engine and emission control durability. The Kansas City analysis showed that lubricating oil contributes significantly to PM emissions, and increases as vehicles age and deteriorate (1). Assuming that such deterioration would continue in the absence of Tier 3, to achieve Tier 3 PM standards at full-useful life, the EPA anticipates that vehicle manufactures may implement a variety of strategies, including reducing lubricating oil consumption. 
For quantifying the emissions reductions in the NPRM analysis, only the impact of reduction in lubricating oil is calculated. The Kansas City analysis estimated that 42% of the fleet-average OC emission rates are attributed to lubricating oil emissions, and 25% of the total PM emission rates are attributed to lubricating oil emissions (1). The EPA analysis assumes that a comparable average reduction in OC and PM will be achieved for Tier 3 vehicles over the lifetime of the vehicles. 
The EPA analysis used the fleet-weighted PM contribution from the Kansas City analysis because many of the newer vehicles in the Kansas City study had not reached full deterioration. In the absence of PM data on deteriorated Tier 1 and Tier 2 vehicles, the EPA analysis assumed that the relative contribution of PM emissions from older vehicles observed in the Kansas City study would continue for Tier 2 vehicles. The EPA analysis assumed that all the PM contributed oil emissions were eliminated under the Tier 3 scenario. This was deemed to be a reasonable assumption because the EPA anticipated other emission reduction strategies would be implemented in Tier 3 vehicles which were not accounted for in this analysis. Other PM reduction strategies include reducing over-fueling at cold start and high-load conditions. 
Implementation of Emission Reductions in MOVES
The reduction of lubricating oil-derived PM in MOVES was modeled as a reduction in OC emissions, such that the PM is reduced by 25%. Table 3 shows the MOVES OC needed to achieve the 25% PM reductions from lower OC rates. In MOVES, OC was estimated as the non-EC component of PM from the Kansas City Light-Duty Vehicle Emission Study. As such, the `MOVES OC' rates include the `OC' as well as non-carbon components of PM, metals, and ions. Due the definition of `MOVES OC', to achieve similar reductions in total PM through reductions in OC, the change in `MOVES OC' is less than OC lubricating-oil contribution estimated in the Kansas City analysis.
    Table 3. Estimated PM reduction due to improved Tier 3 control in MOVES
                                       
                                    Source
                                  MOVES OC/PM
                              MOVES OC reductions
                                 PM reductions
                                     start
                                      car
                                      66%
                                      38%
                                      25%
                                     start
                                     truck
                                      68%
                                      37%
                                      25%
                                    running
                                      car
                                      82%
                                      30%
                                      25%
                                    running
                                     truck
                                      93%
                                      27%
                                      25%

The calculated reductions for starts and running MOVES OC emission rates range from 27% to 38%. Due to the uncertainty in the analysis, the EPA chose to represent the Tier 3 control with a 30% reduction in the OC rates in MOVES for both start and running process, and for all ages of Tier 3 vehicles.  
No differences were estimated for the different sources (cars and trucks) or for different emission processes (start and running). The Kansas City analysis estimated the lubricating oil contribution to PM emissions by pooling the passenger cars and trucks together. The 30% reduction is applied equally across all operating modes. The oil analysis was conducted over individual and filter composites of the LA-92 cycle, and the lubricating oil contribution could not be estimated separately for cold start, hot-start or hot-running conditions. 
At the time of this analysis, the EPA had no available data to evaluate the impact of Tier 3 on medium-duty gasoline vehicles that are regulated by the Tier 3 standards. The PM emissions rates in MOVES from medium-duty gasoline vehicles have previously been adjusted from the light-duty PM emission rates. Thus, the same 30% reduction in OC emission rates was also applied to medium-duty gasoline vehicles.
Modeling Tier 3 PM Impacts in the Final Rule Making
A particulate matter test program was underway at the time of this analysis at the U.S. EPA National Vehicle Fuel & Emission Laboratory on Tier 2 vehicles aged to 120,000 miles. The test program was being conducted to assess the feasibility of the Tier 3 standards on current and future technology vehicles. The modeled benefit of the Tier 3 rule will be revisited when the test program is complete.  
References:
   1. Darrell B. Sonntag, Chad R. Bailey, Carl R. Fulper, and Richard W. Baldauf. Contribution of Lubricating Oil to Particulate Matter Emissions from Light-Duty Gasoline Vehicles in Kansas City. Environmental Science & Technology 2012, 46(7), 4191-4199.
