The Design of a Bi-Fuel Engine Which Avoids the Penalties Associated with Natural Gas Operation

An alternative fuel that has demonstrated considerable potential in reducing emissions and crude oil dependence is compressed natural gas (CNG). A dedicated CNG vehicle suffers from the lack of an adequate number of fueling stations and the poor range limited by CNG storage technology. A vehicle capable of operating on either gasoline or natural gas allows alternative fuel usage without sacrificing vehicle range and mobility. Although many such bi-fuel vehicles are in existence, historically they have employed older engine designs and made compromises in engine control parameters that can degrade performance relative to gasoline and increase emissions. A modern production engine, a 1992 Saturn 1.9 liter 16 valve powerplant, is being optimized for operation on each fuel to realize the full potential of CNG in a bi-fuel system. CNG operation in an engine designed for gasoline typically suffers from reduced power, due in part to displacement of air by gaseous fuel. In order to regain that lost power and therefore promote the alternative fuel usage, two methods have been experimentally investigated with the bi-fuel engine: turbocharging while retaining the stock compression ratio, and raising the compression ratio under naturally aspirated operation. Performance differences between the two fuels were considerably reduced with both methods. Emissions of total engine-out hydrocarbons increased with the high compression ratio, attributed to a greater ratio of crevice volume to total volume at TDC. A small efficiency increase was observed with the high compression ratio. Raising the compression ratio for CNG operation takes advantage of the high knock resistance of CNG, but without further modifications this leads to the occurrence of knock during gasoline operation. To aid in the design of a remedy for knock, a numerical model that can predict in-cylinder events is being developed. In order to predict knock in the model, an experimental study of knock was necessary. A method of quantifying knock from the most negative values of the third derivative of the in-cylinder pressure was proven to be valid for this engine, and a threshold for the onset of light knock has been established as values of the third derivative of in-cylinder pressure less than -40 kPa/deg³.