Atmospheric residue hydrodesufurization (ARDS) is an important hydroprocessing unit in Kuwait refineries to upgrade heavy residues into more valuable and environmental friendly products. To enhance the process cost-effectiveness, it is important to understand the interaction between the catalyst system, process operation conditions, and feedstock properties and the effects on the unit performance. This requires extensive programs of pilot plant work, which is very costly. To minimize the costs, efforts, and time required for experimental work, mathematical modeling is a powerful tool to assist in achieving such targets. Performance life study of an industrial ARDS catalyst system identified as the CAT-B system, which contains five different catalysts, is studied using a downflow pilot plant unit. The feedstock used was the atmospheric residue of Kuwait Heavy Crude (KHC-AR) from the blending of Ratawi/Burgan heavy oils. A portion of the pilot plant data of the life test is used in the parameter estimation for the developed model before model prediction is performed. Model simulation is then applied to predict the performance of the pilot plant life test with KHC-AR as the feedstock. Some discrepancies are found between model predictions and pilot plant life test results. These discrepancies may be attributed to the changes in the activities of the HDM catalysts used in CAT-B.