Determining return on investment of maintenance projects using statistical cost modeling

This paper presents a methodology for quantifying the expected benefits of maintenance improvement initiatives using statistical techniques. Multivariable regression models are used to correlate monthly equipment maintenance costs with monthly planned and unplanned downtime. The models permit the expected cost savings for an equipment fleet to be quantified as a function of the percentage reduction in unplanned maintenance downtime likely to result from an improvement initiative. The models also make it possible to quantify expected increases in fleet availability and reliability and to calculate cost factors that indicate how much more expensive it is on average to apply unplanned as opposed to planned maintenance strategies. A case study is presented that demonstrates how the methodology was applied to the mobile equipment fleets in an openpit copper mine in Northern Chile