The popular mathematical and simulation methods both have some shortage in evaluating the stationary operational availability. The former assumes that the item demand is independent of the operating systems, which can introduce a serious underestimation of the operational availability. The latter in order to get acceptable confidence level requires a large of trials under the pre-assigned scenario. This paper addresses the issue of determining the stationary operational availability basing on mathematical models. Not only the system passivation and operation mission, but also other influencing factors, such as the system design parameters, the number of working systems, turn round time and maintenance time, are all considered in the model. In order to achieve the result, a powerful approximation method to the operational availability is given. At last, an example is used on researching the relationship among aforementioned factors. Experiments show that the model works well with Monte Carlo simulation result and so, the feasibility and rationality of the mathematical method are validated.