Reliability is a significant performance index for any electrical equipment in the industry, especially for electric motor driven (EMD) systems. Modelling the reliability of an EMD system is a difficult and challenging task. However, based on field data, the dominant failure modes of some motor system parts can be used to develop a simple and credible reliability model. This study proposes a failure rate prediction model for an EMD system that is currently in use in hydrocarbon industries. In addition, a fault tree model was designed to identify the root causes of the failures and the failure mechanism of the most dominates failure modes in EMD systems. To accomplish the study's objectives, an existing failure databank of over 100 failures of EMD systems was collected and surveyed from the field. The historical information and failure rate models that were collected for the electric motors are investigated and used to model the prediction model of the EMD system. Three competing failure modes were found to dominate motor life: bearing failures, winding failures and shaft failures. A mathematical model for each of these failure modes with dominant variables is developed, and a risk failure model is established.