This paper deals with numerical methods for predicting air flow patterns in large industrial halls. Some major findings of the investigation of the airflow patterns in paper machine hall of Umka Cardboard Mill are presented in the paper. The main reason for the interest in this problem is to find optimal locations for extract air intake connections of the ventilation system connected to the exhausted air waste heat utilization. Previous studies have shown that the amount of heat released from the cardboard machine to the surrounding air in the hall and extracted by the series of ceiling mount axial fans was almost 30% of the total waste heat from the paper machine's drying section. These results have indicated the need for the waste heat utilization, but also for the optimization of the ventilation system. CFD simulation for predicting of air flow patterns was applied. The accuracy of the simulation was evaluated by comparing its results with the results of field measurements. Simulation results served well for qualitative analysis, gave better insight in general air movements inside the hall and indicated the extract air intake locations. By utilizing the waste heat from proposed optimal locations, fuel savings of 5% and reduction of 1140 t/year in CO2 emissions can be achieved. (C) 2011 Elsevier B.V. All rights reserved.