Despite risk-based inspection (RBI) programs being utilized for nearly 15 years, companies, unfortunately, are still experiencing significant failures. While the general consensus is that the initial value of an RBI planning program has been realized, there are new opportunities being discovered that can improve the accuracy of an RBI program. Chief among them are advancements in the application of data science; however, without a subject matter expert (SME) to validate the results of such large quantities of data, companies can remain exposed to unknown risks and/or sub-optimal asset strategies in their RBI programs. A blended approach is recommended, combining mechanical integrity and subject matter expertise with new technologies and methodologies in data science to drive better recommendations for any RBI program. But how can companies leverage their SMEs when they’re being pulled in so many directions?
It's no secret that SMEs are in high demand these days, and fewer are available to support RBI programs in ways that were once deemed appropriate. For instance, materials engineers support integrity operating windows (IOWs), damage mechanism reviews (DMRs), and RBI programs. Therefore, it’s critical that when a company is utilizing an SME, it can make the best use of their time while simultaneously extracting the most valuable information for use in advanced analytics. By leaning on one another, SMEs and data science are able to glean information in two ways:
- An SME can direct data science to look for specific things in an inspection or process data to confirm or deny certain situations, or
- Data science can catch inconsistencies that can then be flagged for the SME to review (perhaps they didn’t realize there could be a problem on the horizon, or overlooked a problem). This approach creates efficiencies by learning patterns in large data sets.
In both scenarios, it is working in conjunction with one another that yields better results, thus improving a company’s RBI program and reducing the time needed by an SME.
In addition, multi-disciplinary cross-validation for each risk review, which includes SMEs for Mechanical Integrity, Materials and Corrosion, and Data Science, is recommended. In doing so, companies can review major discrepancies between data analytics and the SME expectation, such as:
- Identifying potential problems based on data or significant gaps in data
- Quantifying rates and uncertainty in data rather than using assumed rates with a subjective level of conservativeness considered
- Identifying areas of concern that were previously not anticipated
As a result of garnering all this data and cross-validating, SMEs can now provide improved asset strategies. They can give better recommendations on inspections, repair or replacement, upgrades, and IOWs.
Example 1:
In this exchanger outlet, the SME identified a few horizontal condensate locations and deadlegs within a corrosion loop. In this case, the SME expected a higher rate of corrosion on the deadlegs than in the straight-run horizontal locations. Before the data science analysis, it might have been expected that the deadlegs had only five years of life remaining. After analysis, it was found that there was no dramatic difference between horizontals and deadlegs, and life expectancy was actually closer to 20 years. Data science confirmed potential accelerated corrosion on horizontal condensate locations and found no evidence of accelerated corrosion or serious issues in the deadleg CML.
Example 2:
In this second example, an SME identified a typical problem – when coming out of a vapor outlet of a furnace with changes in direction from elbows, accelerated corrosion can be visually observed. Often the first elbow is the worst case, and as SMEs go downstream, the corrosion is expected to decrease. The first elbow was identified as a CML family. CML families seem to behave similarly and therefore can be grouped together and analyzed. In this example, the family is across multiple circuits. Data science confirmed that the first elbow was corroding at a higher rate, and the second and third elbows were corroding at an accelerated rate but not as high as the first. Finally, the data science analysis identified accelerated corrosion in the vertical riser location to the right as well, which was initially not anticipated.
By optimizing the scarce SME resource time and focusing their efforts, SMEs can spend time on the things that are most important, such as why a problem occurs, as well as identifying possible solutions to problems.
For the full webinar, watch Four Ways to Get More from Your RBI Program. For additional application examples, watch this webinar on CML Prioritization leveraging LVCs.
Comments and Discussion
There are no comments yet.
Add a Comment
Please log in or register to participate in comments and discussions.