Unit Health Assessment- Oil & Gas Equipment Probabilistic Case Study

Abstract

Abstract The focus of this case study is the analysis of offshore Oil & Gas facilities recorded downtime data which are classified into gas turbine downtime categories and causes. Each event is then correlated with the maintenance repair records to determine the respective root cause. The key objective of this study is to establish the Critical Success Factors (CSF) for unit health after a gas turbine has been in operation for more than 10 years. The outcome is used to enhance the unit performance, efficiency, maintainability, and operability. As a first step, Content Analysis technique was employed to systematically decipher and organize the downtime causes from collected data. Over 500 data samples collected over a period of 3 years were sorted into relevant categories and causes: comprising a total downtime of 11,410 hours. The downtime data, which is interval scale in nature that is in ‘hours’, is meticulously tabulated against respective downtime categories and causes location by location for the 11 gas turbines sites and correlating this to the repair work. Within scope is downtime related to: Forced Outage Automatic Trip; Failure to Start; Forced Outage Manual Shutdown; and Maintenance Unscheduled while those out of scope are Non-Curtailing and Reserve Shutdown as these are external to gas turbine operational influence. In the second step, descriptive statistics analysis was carried out to understand the key downtime drivers by categories. Pattern recognition is used to identify whether the cause is a “One Time Event”, “Random Event” or “Recurring Event” to confirm data integrity and establish the problem statement. This approach assists in the discovery of erroneous data that could mislead the outcome of statistical analysis. Pattern recognition through data stratification and clustering classifies issue impact as reliability or availability. Simplistic analyses can miss major customer impact issues such as: frequent small shutdowns that do not accumulate a lot of hours per event but cause operational disruption; or infrequent time consuming events resulting from a lack of trained personnel, spares shortages, and difficulty in troubleshooting. In the third step, statistical correlation analysis was applied to establish the relationship between gas turbine downtime and repair works in determining the root causes. Benchmarking these analyses outcome with the actual equipment landscape provides for high probability root cause, thus facilitating solutions for improved site reliability and availability. The study identified CSF in the following areas: personnel training and competency; correct maintenance philosophy and its execution in practice; and life cycle management including obsolescence and spares management. Near term recommendations on changes to site operations or equipment based on OEM guidelines and current available best practices are summarized for each site analyzed.