Given a set of product recipes, the main goal of production scheduling is to attain an optimal solution which satisfies the market demand taking into account a variety of constraints and specific policies. Thus, a schedule is generated capable of accepting certain degree of variability during batch production. But, in most cases an on-line adjustment to the real conditions is required. In this paper, it is presented a reactive scheduling algorithm, which adapts the current schedule to real time disturbances, integrated to a supervisory system capable of handling the new schedule generated thus, filling the existing gap between high production scheduling and low sequential control levels in the Chemical Process Industries. Changes of tasks start times and units reassignment are considered in a decision tree level based on heuristic rules.