Predictive Motor Failure Prevention in Refinery Using Cloud Based Motor Current Signature Analysis

As the industrial world is evolving into Industry 4.0, connected products allow for greater visualization and awareness of the status of critical assets. The collection, analysis, and communication of data from these assets allow insight into the status of equipment, along with the level of wear and lifespan of this equipment. Diagnosing the current health of each asset gives stakeholders advanced warning to possible failures, which pose significant safety and financial risks. Internet of things (IOT) connected products allow for real-time monitoring and have the benefit of data processing in the cloud where historical data can be easily used to analyze potential failures. For over the last 10 years, vibration analysis paired with artificial intelligence has become a common method for motor failure identification. Motor Current Signature Analysis (MCSA) is a novel, non-invasive method that leverages recently available cloud processing power to deliver very similar predictive results comparable to vibration analysis. For critical motors, MCSA allows for a non-invasive, easy to install, and cost-effective motor monitoring method. In this paper, we will explore two main topics: 1) The MCSA data journey for critical motors where details are being continuously collected and transferred to a cloud-based analytics platform where they are compared on a data lake with other similar motors across the world to improve prediction and failure detection algorithms over time. 2) The detection and analysis of three different motor failure modes using Motor Current Signature Analysis on low voltage motors at a Refinery.