We consider the problem of scheduling heterogeneous batch processor under the conditions non-identical job sizes, dynamic job arrivals and incompatible job families. The problem studied in this paper is motivated by heat-treatment operations in the pre-casting stage of steel casting manufacturing. We proposed a few heuristic algorithms with maximizing the average utilization of batch processors as scheduling objective. From the computational experiments carried out, it appears that all the proposed heuristics are yielding very close to the optimum (estimated optimum) average utilization of the batch processors. Further, we carried out a few simple sensitivity analyses. From sensitivity analysis, in general, it appears that there is an influence on the performance of the proposed heuristic algorithms when the input parameter changes. The influence observed is further verified statistically. Finally, it is observed that with respect to the scheduling objective, the choice of the best algorithm(s) do not significantly affected due to changes in the parameters considered.