Many optimisation problems in Chemical and Process Engineering are non
convex resulting in multiple local solutions. A modified interval algo
rithm for obtaining global minimisers is presented. The modifications
accelerate the convergence of the interval algorithm and identify mini
misers lying on constraints to provide a single rigorous globally opti
mal point to a general optimisation problem. The results of applicatio
n to test problems indicate that a substantial performance increase is
obtained when solving multiextremal problems.