A methodology for reliability forecasting and performance evaluation of a mining system

ABSTRACT One of the major lacunae in almost all mine maintenance and production information systems is that the generated data often remains either completely unused on under utilized without effecting any suggestive inferences on productivity and reliability of a system. Probability distributions of the failure data help in recognising the expected performances of any machine element or elements together. A final distribution so generated will be useful in understanding expected machine behaviour over a time. The paper attempts at deriving a distribution of failure data and a method to propagate element data to a final distribution of system failure. The paper also describes a simple but a very effective means of calculating the retrospective performance vis a vis current state of performance of a system. The feed back information which is necessary for such analyses needs not be accurate and full of details. The method proposed to analyse this data is discussed together with validation exercises. Anothe...