Relating liquid fuel and headspace vapor composition for California reformulated gasoline samples containing ethanol

Changes in gasoline formulation will be required as use of methyl tert-butyl ether (MTBE) is phased out. Changes in evaporative emissions of volatile organic compounds (VOCs) may be of concern in cases where ethanol is added to gasoline to replace MTBE. Regular, mid-, and premium grade gasoline samples containing ethanol were collected in the San Francisco Bay area during May 1999. The compositions of the liquid fuel samples and their headspace vapors were measured. Ethanol contents ranged from 3.25 to 9.65 wt % in the liquid fuel samples. Four compounds (n-butane, n-pentane, 2-methylbutane, and ethanol) together accounted for >50% of the total headspace vapor mass. The partial pressure of ethanol in headspace vapors increased only modestly, despite a S-fold increase in ethanol in the liquid samples of premium vs regular grade gasoline. Reactivity with respect to ozone formation of the liquid fuel samples was dominated by aromatics, whereas the reactivity of headspace vapors was dominated by alkanes and cycloalkanes. The olefin and sulfur contents of the liquid fuel samples were unusually low. Equilibrium headspace vapor composition was predicted using measured liquid fuel composition, pure liquid vapor pressures of each fuel constituent, and activity coefficients derived from equilibrium P, T, x, and y measurements for ethanol in solution with gasoline hydrocarbons. Weight fractions of the 13 most abundant species in gasoline headspace vapors were predicted to within +/-15% of measured values, except for n-butane which was overpredicted by 22%. Alternate predictions of headspace vapor composition using the universal functional-group activity coefficient (UNIFAC) model were generally less accurate, especially for ethanol.