1 Automotive CO 2 Emissions Characterization by U.S. Light-Duty Vehicle Platform John DeCicco, * Feng An, † Huiming Gong † Presentation at the TRB Annual Meeting Washington, DC – January 2005 * Environmental Defense † Energy and Transportation Technologies, LLC
2 Overview Objectives (why look at platforms?) What is a platform? Methodology and data sources Platforms in the U.S. auto market CO 2 Emissions Characterizations Explore variability within and across platforms Compare platform efficiency estimates Conclusions
3 Objectives Link CO 2 emissions and related factors to the way production is organized. Proliferation of nameplates and artificiality of the car-truck distinction makes traditional class-based analysis more difficult and less revealing. Foundation for analyzing issues of part-scale production and staggered design change. Provide a basis for assessing costs pertinent to production credits or similar incentives. Why look at platforms?
4 What is a Platform? In general, a collection of manufacturing assets shared among different products. Historically related to common chassis components and "hard points" for an assembly line. Flexible manufacturing long since obviates need for fixed dimensions. Platform ("architecture") now entails sharing of both "soft" and "hard" assets.
5 Platform Strategy as a Balancing Act Minimizing costs through economies of scale Maximizing the market benefits of product differentiation
6 Data and Methodology EPA & NHTSA data for fuel economy, matched to trade (Ward's) platform data Only up to 8,500 lb gvw, even though some platforms also include heavier models Platforms are not always "well defined" MY2002 sales, CY2002 platform production early MY2003 models not counted in sales Nominal, direct CO 2 emissions based on 8.8 kg/gal, 15% fuel economy shortfall Diesel and AFV use assumed negligible (diesel LDV share was only 0.1% in MY2002; estimated FFV credits were backed out)
7 Top Platforms Ranked by U.S. Sales Next 5: Ford Explorer, Honda Accord, Chevy Trailblazer, Chrysler Voyager, Chevy Malibu
8 Platform Distribution by MY2002 Sales
9 Platform Distribution by MY2002 CO 2 Emissions
10 Variability within a Platform Factors: # of engines, # of body styles, weight Examples GMT800 (Silverado, etc.) 7 models, 5 engines, 3 body styles variations: 33% in disp, 26% in wt, 23% in CO 2 Dodge Dakota/Durango 2 models, 4 engine, 2 body styles variations: 75% in disp, 31% in wt, 45% in CO 2 Honda Odyssey / Acura MDX 2 models, 1 engine, 1 body style variations: (0) in disp, 5% in wt, 6% in CO 2 Variation ≡ (Max-Min)/Mean [sales-weighted]
11 Typical Variations within a Platform Weight, in general, varies least: median 17% Greatest variation (26%-35%) in pickup platforms, which include body-on-frame SUVs Engine displacement median variation: 26% Greatest for compact pickups, with I4 - V8 options CO 2 emissions rate median variation: 20% Outlier is VW Jetta, with diesel: 67% variation For others, compact pickups show 45% variation N.B. Drive type was not examined, but other analysis indicates typical 10%-15% CO 2 impact for 4- vs. 2- WD.
12 Variability Across Platforms Comparing platform averages (but remember the significant within-platform variability) Examined: Power, specific power (HP/L) Ton-MPG »Reciprocal of mass-normalized fuel consumption »Isolates non-mass-related aspects of efficiency »A good (but not perfect) index of powertrain efficiency
13 Platform average peak power vs. engine size
14 Ton-MPG Indeces for Selected Platforms (identified here by representative models)
15 Ton-MPG vs. average platform weight No correlation to weight (r = -0.04) Ton-MPG for trucks only 5% lower than cars on average Some, but not all, large variations reflect platform age ("dated"-ness)
16 Conclusions Platform-level data enable analysis linked to how the industry manages production Highest volume platforms contribute, by a modest margin, disproportionately to CO 2 Top 30 ⇛ 69% of sales, 72% of CO 2 (MY2002) Variability within and across platforms can reflect some opportunities for CO 2 reduction Newer platforms generally more "efficient" Provides a baseline and foundation for several types of future analyses