Improving Equipment Reliability and Availability through Real-time Data

Abstract

Improved plant reliability is one of the major business drivers for any organization in today's competitive environment. The goals of reducing downtime and moving to a more proactive maintenance strategy requires commitment to put into place intelligent maintenance and repair practices in order to identify the root cause of unplanned shutdowns and take the necessary steps to prevent future occurrences. ADNOC Onshore has developed a solution with the combination of subject matter expert analysis and available real-time data from Plant Historian (OSI PI). Rotating equipment's emerging problems can be traced through condition monitoring parameters changes. When these parameters are available for online trending along with historical data, we can perform regression and correlation analysis to find out relations between any two or multiple parameters. This solution works 24X7 on Plant Historian and identify certain conditions scripted on Plant Historian on real time basis and will generate emails to relevant subject matter expert's for further actions. These proactive email notifications cover the information such as, failure mode of the machines and relevant parameter profile/cause and effect with the required actions defined in Reliability Centered Maintenance based philosophy. This solution also includes the integration of Plant Historian with Asset Management System. This helps the operators for timely capturing the START/STOP events generated by the rotating equipment's into Asset Management System and also helps to generate the equipment's Availability & Reliability KPI's. This is one step towards the implementation of Artificial Intelligence (AI) using machine learning techniques based on the available parameter where basically invents/incident/symptoms are developed affecting the equipment/plant production and availability that are captured without human interventions. This has benefited ADNOC Onshore to address various issues on rotating equipment and they have been attended proactively to increase reliability/availability/maintainability of equipment's towards business mission and goals. Purpose of this paper is to show how intelligent diagnostic performed on available dynamic/design data from past, present for operational/condition monitoring parameters for rotating machines will be beneficial to trend and predict the performance deterioration. Identifying any developing abnormal condition before it reaches to alarm/trip condition and bringing it to the relevant expert notice is prime purpose of this paper. Maintenance management is generally evolved as the digital data availability increases with the implementation of digital solutions such for real-time data acquisition and storage. Many companies implement solutions for real-time data acquisition and storage but still maintenance strategy evaluation towards latest philosophies is on a lagging mode. In order to get maximum advantage, both maintenance strategy and digital data usage should go hand in hand. At most of the companies’ Digital Oil Field projects were started with the objective to reduce manual/human interventions for maintenance decision making. Every company tries its best to use these projects out comes at their best. But not all the benefits gets realized due to various reasons. In order to gain the benefits of real-time data, we started to match business objectives of a rotating equipment by analyzing functional failures and how these functional failures can be proactively predicted with the available real-time data. We found many of the equipment's anomalies can be detected well in advance to have a proper maintenance planning and maintenance interventions. This has resulted in reduction of large amount of unplanned jobs.