Continuous efforts have been made to improve the performance of conventional hydroprocessing catalysts. For this purpose, catalyst performance has been evaluated using both model compounds and real feeds. However, the beneficial effects of the catalyst modification and/or the novel catalysts on activity observed for model compounds were not confirmed for heavy Maya feed (108). The study of Gray et al. (109) can be used as another example of the different effects observed for model compounds and real feeds. In this case, a series of coked NiMo/Al ₂O ₃ catalysts were used to determine their remaining activity. HDS activity decline with increasing amount of deposited coke was more pronounced for the model compound such as DBT than for bitumen used as the feed. Therefore, the use of heave feeds rather than model compounds and light feeds during catalyst development is the essential requirement. In this case, the effect of various additives, different supports, method of catalyst preparation and pretreatment, etc., has been receiving attention. Novel catalytic phases, differing markedly from those present in conventional catalysts, have also been evaluated. The surface properties of catalysts have been receiving much attention because of the presence of resins, asphaltenes and metals in heavy feeds. As was indicated earlier, without suitable porosity a desirable catalyst life could not be achieved. Models have been developed and used as tools for designing and selecting suitable catalysts to match the properties of heavy feeds with the type of catalytic reactor.