Temperature dependence of source specific volatility basis sets for motor vehicle exhaust

Abstract

Recent work on emissions testing has focused on developing source specific volatility distributions which could be used to improve emissions inventories. One problem about these volatility profiles is that they are evaluated only at one temperature which is usually 298 K. This study uses a simple statistical model to evaluate the temperature dependence of the source-resolved volatility basis set, considering gasoline and diesel vehicle exhaust. The steps involved (a) fitting a distribution to the emissions data (b) evaluating the goodness of fit using a statistical test (c) updating the volatility bins using the Clausius–Clayperon equation; calculating the heats of vaporization of each volatility class using a regression model (d) assessing how the volatility of different VOC classes-Extremely Low Volatile, Low Volatile, Semi-Volatile, Intermediate Volatile and Volatile Organic Compounds – are affected by temperature. The results indicated that there could be significant changes in gas-particle partitioning of these emissions. For diesel exhaust at 298 K, the fractions are 5.4 × 10−4 (ELVOC), 0.074 (LVOC), 0.76 (SVOC), 0.17 (IVOC) and 10−5 (VOC) respectively. Looking at a window of ∓20 K, the partitioning for 278 K is 3 × 10−3 (ELVOC), 0.26 (LVOC), 0.67 (SVOC), 0.07 (IVOC) with no VOC fraction; while at 318 K it is 1.5 × 10−7 (ELVOC), 9 × 10−3 (LVOC), 0.64 (SVOC), 0.35 (IVOC) and 2 × 10−5 (VOC); demonstrating a significant change with temperature. The parameterizations developed in this work could be used to improve motor vehicle emissions inventory models such as MOVES.