A nonlinear optimization model is developed in this paper to identify the optimal replacement strategy for military aircraft. In the model, the aircraft operating and maintenance (O&M) costs per available year are estimated as a function of age during the aircraft life cycle. After determining the optimal replacement policy, the model is applied to the CF Long-Range Patrol CP-140A Arcturus fleet. A sensitivity analysis is also carried out to assess the impact of some key model parameters on the result.
INTRODUCTIONThe longer equipment stays in service, the higher is its maintenance cost and the lower is its productivity. When equipment reaches a certain age it may be more economical to replace it. The problem thus reduces to determining the optimal economic age (Taha, 2007). Following this rationale, the Canadian Forces (CF) is looking to determine the optimal replacement policy for its aircraft fleets.Replacing or repairing equipment is a classic issue in dynamic optimization. However, few papers have dealt with the optimal age for replacing military aircraft. The reason for this void is that such assets generate an advantage that is not directly quantifiable (in dollars, for example) and therefore not easily treated using conventional cost-benefit analysis. Among the papers that have addressed a few limited aspects of this question, one finds Schwartz, et al. (1971) who presented a dynamic program to determine optimal repair and replacement policies for aircraft. Chickermane and Gea (1996) presented a methodology that generates optimal mechanical repairs for aging aircraft. Lincoln and Melliere (1999) provided a procedure to determine the economic life of military aircraft. Greenfield and Persselin (2003) developed an intergenerational model to identify the optimal replacement strategy that recognizes cost trade-offs and incorporates age effects. Despite its impeccable theoretical foundation, the applications of this model have often suffered from a large discrepancy between their results and those expected by industry and end-users. As we will demonstrate in this paper, this lack of applicability would lay in the omission of the important concept of operational availability and its relation to operating and maintenance (O&M) costs. Taking into account the operational availability of the fleet, Keating and Dixon (2004) suggested a methodology to decide when to replace an aging system. Castro