The mixed integer nonlinear programming(MINLP) model by Modi et al. was modified for optimized design of multiproduct batch chemical processes to accomodate not only the existence of intermediate storages but also the adjustment of the number for parallel units operating in-or out-of-phase. A heuristic algorithm for solving the model mentioned above was proposed the advantages of which were lower demand for initial values of decision variables; lower storage and less CPU time for computations and relative deviation of less than 1 % in comparison with the results of using mathematical programming software (MINOS 5.0). An example was shown to prove its availabilities.