API PUBL 4557-1994

VOLUME I: In 1991, the American Petroleum Institute (API) sponsored an emissions testing program to determine: (a) whether changes in the composition of aromatic compounds in gasoline have a significant impact on exhaust emissions; and (b) the significance of the impact on emissions from changing fuel aromatics composition after other fuel properties (such as distillation factors) are controlled. The study measured both regulated exhaust emissions and the composition of exhaust hydrocarbons from a fleet of ten automobiles representative of 1989 and 1990 model year technology operated under 1975 Federal Test Procedure (FTP) temperatures and conditions. The test program was conducted in three phases. The first phase measured emissions on a base fuel and a set of four additional gasolines which had been systematically varied with respect to their relative compositions of C7, C8, C9, and C10+ aromatics. A second phase measured emissions from vehicles operated on the base fuel and two additional gasolines. These latter two gasolines were specified to have distillation properties which matched those of the base fuel and all other characteristics (including aromatics composition) were designed to be similar to two of the fuels from the first phase. A third test phase was conducted after an inspection revealed that the sulfur levels of the Phase II fuels were uniformly low. The fuels for the second phase were reblended to match the Phase I sulfur levels. The study found that exhaust mass emissions of hydrocarbons, carbon monoxide (CO), and nitrogen oxides (NOx) rise when the fraction of total fuel aromatics contributed by species with higher carbon numbers (i.e., C9, C10) is increased and fuel distillation properties are not controlled. However, when fuel distillation properties are held constant, shifts in the composition of fuel aromatic compounds by carbon number have no statistically significant effect on exhaust emissions of these pollutants.

VOLUME II: Regulated and unregulated exhaust emissions were measured on a fleet of ten model year 1989 and 1990 vehicles operated on a matrix of test fuels which varied with respect to the composition of aromatics and distillation properties. The test program was conducted in three phases. The first phase measured emissions on a base fuel and a set of four additional gasolines which had been systematically varied with respect to their relative compositions of C7, C8, C9, and C10+ aromatics. A second phase measured emissions from vehicles operated on the base fuel and two additional gasolines. These latter two gasolines were similar in aromatics composition to two fuels from the first phase but had distillation properties which had been matched to those of the base fuel. A third test phase was conducted after an error was discovered in the sulfur levels of the fuels used in the second phase. Fuels were similar to those in the second phase except that the sulfur level was adjusted to match Phase I fuels. The study provided data collected under standard test temperatures and conditions for total exhaust hydrocarbons, carbon monoxide, nitrogen oxides, aldehydes and detailed exhaust hydrocarbon species. The data and the test procedures used during the study are presented in this volume of the report. Statistical analyses and interpretations of the data are provided elsewhere.