Numerical tuning has shown to be an effective approach to improve computational performance of simulation models, especially complex large-scaled ones. One purpose of numerical tuning is to find a set of numerical parameters that make the simulation run faster with less material balance error. However, in many cases, these two objectives are conflicting with each other. The most common way to tackle this problem is to create an aggregated global objective function, which is the weighted average of the two objectives, and then try to optimize it using a single objective optimizer. In this study, Multiple Objective Particle Swarm Optimizer (MO-PSO) is applied to conduct numerical tuning on a sub-model of a real field SAGD case study. MO-PSO tries to reduce the simulation run time, and the material balance error simultaneously. The results obtained by MO-PSO are compared with that of single objective particle swarm optimizer (SO-PSO). Results showed that MO-PSO is much more effective and efficient in finding Pareto front compared with SO-PSO. MO-PSO eliminates the usage of any weighting factor, which makes it much easier to set up and avoid biased searching results. In addition, MO-PSO converges to optimal solutions much faster than SO-PSO for this study. SO-PSO with well-defined aggregated weighted global objective function can also find a good Pareto front, but inappropriate weighting factors in the aggregated global objective may lead to biased solutions.