BACKGROUNDER: U.S. EPA Fuel Economy Testing

Current EPA requirements for fuel economy testing are inexact in that they provide vehicle manufacturers with wide latitude.   As a result, vehicle manufacturer test results as a whole are imprecise, inconsistent and not repeatable, as noted by consumer advocates ranging from Consumer Reports to the Union of Concerned Scientists.   Additionally, EPA has acknowledged that it has no regulatory pass/fail “acceptance criteria” for its confirmatory testing of vehicle manufacturer certification tests.

Applicable Rules and Guidance

Applicable EPA Regulation – 40 CFR § 86.129-00(e)(1):
“[The] manufacturer shall supply representative road load forces for the vehicle…  The road load force shall represent vehicle operation on a smooth level road, during calm winds, with no precipitation, at an ambient temperature of 20°C (68°F), and atmospheric pressure of 98.21kPa.”

EPA Advisory Circular No. 55C – Alternative Dynamometer Power (December 12, 1986)
Only EPA/SAE document addressing setting vehicle weight for coastdown testing

Dear Manufacturer Letter re: Single Roll Dynamometer Adjustment and Road Force Determination (VPCD-98-16) (December 21, 1998)
“EPA will use SAE J2263 procedure for confirmatory coastdown testing, manufacturers may use any procedure or method that yields equivalent results.”

Dear Manufacturer Letter re: Notice of EPA/Industry Road Force Workshop (CCD-02-01) (January 7, 2002)
EPA in-use testing shows road load energy variance from -5% to +10%
“Inherent test and vehicle variation” identified
“Pass/fail” criteria for in-use road load force not set

EPA Procedure

Testing CriteriaEPA SpecificationHyundai Comments
Track surface“a smooth level road”EPA has not defined “smooth;” a detailed definition would increase accuracy and repeatability of tests as this is the largest source of variance between vehicle manufacturer certification tests and EPA’s confirmatory testing
Temperature41°-95°FSAE J2263 temperature conversion is not accurate at lower temperatures; it yields road load force values at 68°F regardless of actual test temperature within defined range; narrower temperature range would increase accuracy and repeatability of tests
Wind speed22 MPH average/31 MPH gustCoastdown test design establishes road load values with “calm wind;” SAE J2263 has inaccurate wind correction method, with tests at lower wind speeds more accurate than tests at higher end speeds; high wind speed tests within EPA’s parameters can result in elimination of many data points, reducing accuracy of results; EPA’s most recent and most detailed regulations on coastdown testing at 40 CFR 1066.310(b)(1)(i) recommend winds less than 6 MPH for heavy-duty vehicle coastdowns; a lower wind threshold would increase accuracy and repeatability of tests
Tire break-inMinimum 2,175 miles, and preferably 4,039Break-in location/method not specified despite fact that tire break-in procedures impact road load; no maximum mileage; EU and SAE specify maximum tire break-in to limit tire wear to at least 50% of tread-life remaining; test track break-in would result in consistent, repeatable wear patterns
Warm-upAt least 30 minutesBearing and tire friction may not be stabilized after only 30 minutes at cooler temperatures; warm-ups between coastdowns are required to maintain temperatures and stable levels of friction, which lead to increased accuracy and repeatability of tests
Vehicle weightNot specified in EPA coastdown regulations or in SAE J2263EPA’s only published guidance on coastdown vehicle weight is AC No. 55C, which requires weight to be within 100 lbs. +/- of the designed empty weight with 33% options, plus instrumentation, with 40% fuel and no driver
Braking between runsEPA’s use of airport runways may require driver to apply brakes after each coastdown runAmount of variance depends on amount of brake application; testing on a continuous track would increase accuracy and repeatability of tests
Data set selectionMinimum of five sets; no maximum; no prohibitions or recommendations re: set selectionRuns may be affected by outside forces including unnecessary steering inputs, interference from other vehicles on test track and wind gusts; SAE J2263 allows for elimination of data “outliers,” “eliminate[ing] extreme data points,” and “smooth[ing] remaining data points; EPA’s latest and most detailed coastdown requirements specify that runs greater than two standard deviations from the mean values of all sets must be “eliminated” (40 CFR 1066.310(b)(6) EPA’s 2011 HD GHG Regulations); averaging the longest five sets would eliminate sets impacted by outside forces and yield the true road load performance of a vehicle
Single direction run selectionNot specified in EPA regulation; SAE J2263 and good engineering judgment requires use of paired runs in both directions in cases where there is enough wind to lengthen runs in the down-wind directionThis is the source of the isolated data processing error that involved coastdown testing to determine resistance or road load

Over the past 30 years, the EPA has acknowledged the variability of its coastdown testing, and currently is working to develop new guidance for the industry in order to improve its precision, repeatability and accuracy.