Tank truck loading operations in fuel distribution terminals have particular characteristics which are challenging to simulate. Demands for different fuel types are correlated rather than independent. In addition, queuing policies which do not address the current state of the terminal may increase equipment idleness and trucks’ lead time. We propose a decomposition of the joint probability distribution of fuel types and a procedure to simulate it. We also propose a new queuing policy aimed at reducing equipment idleness and trucks’ lead time. The procedures are tested and validated through a discrete-event simulation model of a real fuel distribution terminal. We simulate alternative scenarios in which we increase fuel flow rate as well as the number of loading arms and test a new queuing policy. Results show the effectiveness of the simulation procedure in replicating real performance indicators. Moreover, the proposed idleness-based queuing policy resulted in lower idleness and lead time compared with a first-come-first-served policy.