Making Your Data Work for You

Summary

Reliability programs focus on the health of assets at industrial facilities and inform business strategies by helping owner/operators minimize process safety risk, maximize production and optimize lifecycle spend. One critical input for a successful reliability program is data. When it comes to data, some believe that more data means better results, and this mindset can be dangerous. More is not always better. Sometimes more is simply…more.
 
We are living in a day and age where we have more data than ever. This includes design, operation, process, asset performance, inspection and risk model data. How do we know that the data is quality data? How do we know if the data we have is valuable to decision-making? What if data is siloed across a myriad of databases and business units, not transformed into a digital format, or institutional knowledge.
 
There are so many “ifs” that can lead to obstacles in decision-making, including analysis paralysis. In situations where owner/operators cannot trust their data or their data exists in an unusable format, how can they make the best decisions quickly and with confidence when it comes to operating their facility?

Here is one example of a problematic situation. Most owner/operators collect empirical data from the field. For example, thickness readings are taken on assets at various time intervals to establish a wall loss rate. This wall loss rate is used to predict the remaining life of an asset and drives decision-making around repairs and replacements. Beyond traditional applications, these thickness readings and measured wall loss rates are utilized in reliability programs like Risk-Based Inspection (RBI). When calculated or measured rates are not available for use, RBI programs rely on estimates from the American Petroleum Institute (API) codes or Subject Matter Experts (SMEs), which tend to be conservative.
 
Further, when inspection data is not readily available, reducing risk in the RBI program due to certainty in the asset condition becomes difficult, and more conservatism is introduced. Without inspection-related data, RBI programs have overwhelming uncertainty, which can lead to complexity in prioritizing inspection tasks since the conservative assumptions are typically that all high-risk assets need inspection ‘immediately.’ This can leave owner/operators with the same question before implementing RBI: what do I need to inspect tomorrow?
 
So, how do owner/operators solve their data woes? There is no “one size fits all” answer. Here are some solutions to consider:

• Start with the end in mind. Do not collect hundreds of data points just because you can. Determine your objectives and focus on collecting the specific data points that allow you to make strategic decisions and achieve success. 
• Consolidate data management systems. Evaluate where site teams rely on the same data housed in multiple locations. Identify or create a single source of truth.
• Bring siloed data into a functional space. Siloed data can take on various forms: no one knows the data exists, no one knows where the data exists, or the data exists in a format that renders it useless for large-scale processing (i.e., statistical analysis and reliability programs).
• Explore automation solutions. Implement reliability software that can auto-update records and analyses from multiple data sources when changes occur or new data is collected from the field. This allows for a real-time, complete picture of asset health which promotes agility in decision-making. It also reduces the time your site personnel need to spend handling data.
• Develop a culture that promotes intentional collaboration between on-site teams: process, operations, materials, safety, maintenance and integrity, to name a few. This multi-disciplinary approach can generate repeatable, reliable systems for collecting high-impact, high-quality data.

Good data management practices lead to effective reliability programs that owner/operators can trust to fuel the right intelligence and assist in making the right business decisions.

Are you working for your data? Or is your data working for you?