Computing Tetraethyl-Lead Requirements in a Linear-Programming Format

There are some types of nonlinearities that are difficult to incorporate into a linear-programming model. One of these occurs in the scheduling of refinery operations where linear-programming methods are widely used. The specific problem is the one of simultaneous determination of the quantities of various components of gasoline and tetraethyl lead to blend to form regular or premium type gasolines. It is believed the general approach can be used for other nonlinear situations. Linearization is accomplished by viewing the lead requirements of a blend first as a known general function of two variables each of which is a linear function of the quantities in the blend. The general function is then approximated by a convex linear combination of a mesh of representative points.