

                                       
                                       
                 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                NATIONAL VEHICLE AND FUEL EMISSIONS LABORATORY
                              2565 PLYMOUTH ROAD
                        ANN ARBOR, MICHIGAN 48105-2498

                                   OFFICE OF
                               AIR AND RADIATION


May 18, 2011

MEMORANDUM


SUBJECT:	Adjusting Combined City/Highway CAFE Fleet Values to Determine Equivalent 5-Cycle Label Values


FROM:	Roberts W. French, Jr.
		Environmental Protection Specialist
		Office of Transportation and Air Quality
		U.S. Environmental Protection Agency


TO: 		U.S. Environmental Protection Agency
		Docket No. EPA-HQ-OAR-2009-0865

As explained in the preamble to the final rule, our methodology for creating a ranking system for fuel economy and greenhouse gases depends on selecting an appropriate mid-point representing the fuel economy of the average vehicle.  Given that Corporate Average Fuel Economy (CAFE) and greenhouse gas emission standards are programs with fleet average standards, these programs will essentially be determining the fuel economy of the average vehicle in any given model year for which there are standards.  Consequently, given a rating system based on MPG, we chose to use the CAFE program to define the average vehicle, and from that vehicle we determined the MPG "bins" that define the 1 to 10 rating system.  

The EPA/NHTSA rulemaking that put greenhouse gas and CAFE standards in place for the 2012 through 2016 model years established the following fleet values (see 75 FR 25331, May 7, 2010). It is these numbers, shown in the following table, that form the basis for our rating system, but with some adjustment.  

                                  Model Year
                                     CAFE
                                     2012
                                     28.7
                                     2013
                                     29.7
                                     2014
                                     30.6
                                     2015
                                     31.5
                                     2016
                                     32.7

The CAFE standards and the label values are determined in different ways. Congress recognized the importance of maintaining a stable method for compliance with standards and essentially fixed the test methods for CAFE such that the test results don't change over time.  However, Congress has also recognized that the label values need to reflect real-world performance, and now requires EPA to periodically evaluate how well the label test and calculation methods are reproducing real-world experience.  This memorandum details the method used to derive label values (sometimes called "5-cycle" values) from the CAFE values above (sometimes called "2-cycle" values).  

In 2006 EPA promulgated a rule that established new test and calculation methods that took effect with the 2008 model year. The new method does not apply a flat, linear adjustment in the way that the previous method did.  The new method applies a higher proportional adjustment to high fuel economy values than to low fuel economy values.  Thus, although the difference could be subtle, the adjustment to the 2012 model year CAFE value of 28.7 MPG will be different than the adjustment applied to the 2016 value of 32.7.  

One other minor issue is that the CAFE values above are Combined City/Highway values, whereas the label values are adjusted separately by test cycle, i.e., the City value is adjusted using one formula, the Highway by another, and the Combined value determined from the adjusted City and Highway values.  Because the CAFE values are fleet average standards based on the Combined City/Highway MPG, there are no separate specific City and Highway values to use to determine the real-world label value, as we would do for an individual vehicle.  If such fleet City and Highway values existed we would simply adjust these by the existing methods to get our desired label values corresponding to the above CAFE values.  

To estimate Combined City/Highway label values that correspond to the above CAFE values, we used fuel economy data from the 2011 model year.  The data includes City, Highway, and Combined in unadjusted (i.e., 2-cycle values comparable to CAFE) and adjusted (i.e., 5-cycle label values).  We simply regressed the unadjusted Combined MPG values against the adjusted Combined values to determine an approximate relationship between the two.  The data and the regression equation are shown in the following figure.


 
As noted earlier, the existing method to adjust test results to get to label values is non-linear, thus a simple linear regression such as this should not be used for exacting purposes.  For our purposes here, however, this approach is a sufficient way to translate each of the CAFE values to average label values, rounded to a whole MPG.  For example, in the 2012 model year when the fleet is required to meet a CAFE standard of 28.7 MPG, we would expect, based on the method in this memorandum, that the similarly sales-weighted average of the label values across the fleet would be 22 MPG.  Thus the values in the "Label Average" column below are used to define the center of the rating system for these model years.  The rating system for future model years will be addressed as CAFE standards for future years become known. 

                                  Model Year
                                     CAFE
                                 Label Average
                                     2012
                                     28.7
                                      22
                                     2013
                                     29.7
                                      23
                                     2014
                                     30.6
                                      23
                                     2015
                                     31.5
                                      24
                                     2016
                                     32.7
                                      25

