SENSITIVITY ANALYSIS OF DEMAND DISTRIBUTION FOR NAVAL AVIATION READINESS-BASED SPARING MODEL

Abstract

Readiness Based Sparing (RBS) models support the life cycle of any system through the optimization of stock allowance levels. Optimal RBS results are essential to maintain fleet readiness at an acceptable cost. Naval Aviation RBS Model (NAVARM) is the tool used by Naval Supply Systems Command to plan the stock allowances for embarked airwings and shore-based aircraft. In order to gain confidence in NAVARM results, it is necessary to validate some modeling assumptions that have not been tested to date. RBS models like NAVARM assume that the distribution of the mean time between failures (MTBF) for any part is exponential. This assumption may not hold in practice for certain parts. Therefore, a question arises as to whether the quality (operational availability by cost) of the solution provided by NAVARM is subject to the effects of this assumption.

This thesis tests the influence of the MTBF distribution on operational availability using the Readiness-Based Sparing Simulation (RBSIM) developed by a former Naval Postgraduate School student. We test the alternate distributions Weibull, gamma, and log-normal, with mean to variance ratios (MTVRs) of 1.5 and 0.5. These MTBF distributions are applied to either all parts or a select subset of parts (based on demand). Initial results on the aviation consolidated allowance list for the USS Carl Vinson (CVN 70) show that both distribution type and MTVR may have a significant effect on operational availability of all weapon systems.