Marathon recognized early in the reliability process that data would be the driving force behind the company's reliability improvements. As discussed in Part I of our series, many companies struggle with issues surrounding data integrity; but like Marathon, they also recognize that waiting until their data is 100% up-to-snuff to launch a reliability effort translates into millions of dollars in lost production opportunities.
With an enterprise-wide understanding of the critical role equipment assets play in plant availability; and a corporate mandate to increase the reliability of their assets, Marathon set in motion a reliability effort which has yielded significant improvements in asset and plant availability.
One of the first steps towards jumpstarting the company's reliability effort was to implement asset performance management (APM) software, providing reliability engineers with the necessary tools to begin to utilize the pockets of existing good data to schedule inspections, analyze failure data, identify areas for improvement, and develop corrective actions. This first step, resulting in significant reliability improvements, begged the question, "How many more opportunities could we identify with more good data?"
Recognizing the significant production opportunities in having more accurate and complete data for rigorous analysis, one of the company's next steps in their reliability effort was the development of a data integrity process, objectively measuring work order data for completeness and accuracy, and communicating the results throughout the enterprise.
As discussed in Part II of our series, regardless of the industry, accurate data collection is essential to maintaining data integrity. Prior to data collection, the careful selection of data collection instruments, rigorous training and the development of clear, comprehensive and standardized data collection procedures are all essential elements of a successful data collection process. But is this enough to ensure data integrity? Marathon decided the answer was no, and took their commitment to producing quality data one step further.
Despite good procedures and adequate training, you can still end up with bad data. To ensure data integrity you must establish quality control activities before and after data collection. Too often, companies spend money and resources on quality control activities prior to data collection, but fail to implement any quality control measures after data collection.
Work orders are the primary tool for capturing data associated with the costs, labor hours, time and the maintenance history of Marathon's equipment assets. Recognizing the strategic importance of acquiring accurate and complete work order data, Marathon established a consistent, verifiable data integrity process for use after data collection. Each month, a random sampling of work orders are selected at each of Marathon's seven facilities and manually evaluated in the following six areas, resulting in a Data Confidence Index for each facility:
- Was the correct equipment number entered?
- Was there a correct failure coding designation of fail or no fail?
- Was time placed against the correct work order?
- Was failure coding captured on turnarounds?
- Did the craftsman put in the failure code per procedure, or did someone else?
- Did the reliability engineer review the work order prior to going to history?
The Data Confidence Index for each of the seven facilities is one of five metrics shared and compared in the company's quarterly Key Performance Indicator (KPI) Report circulated to corporate and refinery management to direct the company's technical and financial resources.
Marathon has found that what gets measured improves - work order accuracy now stands at 80%. The company's "focus on data" - securing its integrity and then using it for rigorous analysis to drive asset improvement strategies - has yielded notable results, including significant improvements in plant availability, increased production and safety.
Yet while the company has the good fortune of having reliability champions from the plant floor to the top floor, the company realizes that despite all their successes, defect identification and elimination is an ongoing effort requiring constant vigilance. Marathon is confident that their investment in standard practices and asset performance management software will help ensure the stability of their current and future reliability efforts.
Click here to download Moving Marathon Oil to a Reliability Culture and learn more about how Marathon leverages data across the enterprise, driving equipment strategies leading to profitable change.
