Cost analysis of mining operations in general shows that 30 to 50 percent of direct mining costs are related to maintenance and losses related to lost production during equipment downtime. To reduce these losses one first needs to improve the equipment reliability and thereafter to reduce the downtime losses through improved maintainability and supportability. The mining operational environment is often harsh and may severely impact all three of these abilities. In this paper we focus on how to improve the estimation of spare parts by taking into account the operating environment in the estimation models. By having better models to predict spare parts needs, one can avoid logistics delays and thereby reduce downtime. In this study we develop an improved statistical-reliability (S-R) analysis approach that take into account the system/ machine operating environment. The analysis approach is multiple regressions based on Cox’s proportional hazards modeling (PHM). Subsequently, in a case study, the management of the spare parts inventory based on the economic order quantity and required performance of the product is addressed.