Make 2019 the Year You Improve the “Data Fidelity” of Your Inspection Program

Introduction

2018 was a very busy year for most. The upstream Oil & Gas industry continued to rebound and work smarter to improve margins and get as much insulation as possible from negative cash flow problems that can arise when the price of crude oil slumps or drops. That should be a lesson for all of us, how can we work smarter?  

The price of natural gas is still favorable for petrochemical companies (and homeowners using natural gas). The mid and upstream sectors are coming on-stream with mechanical integrity (MI) practices that downstream facilities have been implementing and improving for many years, especially for their process safety management (PSM) covered facilities. So what’s on the horizon?

Get the Most Out of Available Technologies and Resources

One way to get the most out of available technologies and resources is to make learning a top priority in 2019. I almost said education, but the value isn’t realized until the information is assimilated in a useful context and used to get maximum benefit. In 2018, Inspectioneering provided a truckload of information, with context to help owner-operators understand new technologies and how to better capitalize on existing technologies. Many articles tackled subjects within a fixed equipment reliability program improvement context, like turnarounds, getting better value from new and existing NDE, corrosion under insulation, fitness for service, improving communications, AIM programs, data management, mobility solutions, regulatory changes, damage mechanisms reviews, moving to enterprise asset management platforms, RBI, and, most importantly in my opinion, via case histories. Case histories provide context and an opportunity for our readers to see how others are addressing these challenges. When possible, the practices, limitations, and costs of implementation were covered, as well as the benefits realized. So let’s pick an area for improvement that is often overlooked – data quality.

Data Fidelity

We don’t hear the word fidelity much anymore, especially as sound/stereo technology has improved drastically over the last 15-20 or so years. Remember when the old stereo commercials and promotional materials talked about high fidelity in sound? How true is the sound? One applicable definition from the internet for fidelity reads, “Faithfulness to a person, cause, or belief, demonstrated by continuing loyalty and support.” Loyalty and constancy were two synonyms offered with this definition. Loyalty and constancy are key words in all of this. Our plant initiatives need to be sustainable and used to our best advantage. The consistency of use/practice, employee ownership, and buy-in are essential to constancy and loyalty to the cause or initiative.

We know that a big challenge all corporations face is managing BIG DATA and figuring out how to best mine (leveraging the value directly and indirectly) gold from the data we have paid to capture and manage. The importance of data fidelity and opportunities to leverage the data become apparent as we implement special emphasis programs, many of which use the same data – and nobody wants to enter data twice. For example, RBI uses a lot of the data that comes from a damage mechanisms review and vice versa. This includes operating temperatures and pressures, materials of construction, fluid stream chemistries, etc. For our purposes, the data usually relates to pressure vessels, heat exchangers, tanks, piping circuits, heaters, bundles, pressure relieving devices, etc. It’s common for the number of equipment items and circuits at a site to exceed 6,000 and for a corporation to exceed 100,000 to 200,000. We are talking massive databases with millions of bits of information; some important and some not so important. Some of this data probably hasn’t been updated in a decade or more.

Data Cleansing and Reconciliation

Our databases are likely a mixture of mostly good, stale, questionable, and some missing information (including databases filled with default values that have not been resolved). At a minimum, let’s make 2019 a year for data cleansing and reconciliation (although, this should really be an ongoing practice that requires diligence).

First, no database is perfect, but we at least hope the important data is there and good. How can we do this in a leveraged manner? A first step is taking the time to validate the data when implementing or updating our special emphasis initiatives, such as damage mechanisms reviews or corrosion control documents. Another important step is identifying where overlapping data collection and/or use initiatives exist.

Some miscellaneous practices to consider are asking the EPC (engineering, procurement and construction) company to provide a table or spreadsheet with all desired information for uploading as part of your next capital project. Writing queries to check data is a necessary practice, too. For example, are you checking the data in your RBI and DMR projects against what you already have in your database? Do you have IT tools to help identify the questionable data? What are you doing to ensure the fidelity of the data? Are credible subject matter experts (SMEs) engaged who can ask the right questions about the data and help design IT tools to interrogate, identify and repair or address bad data? This is important to understand the difference between gold and pyrite in our mountain of data. Important decisions about integrity and reliability of pressure equipment depend upon the accuracy of the data.

For example, what if your IDMS (inspection database management system) showed a corrosion rate of 0.010” per year (10 mpy) and the recently completed DMR (damage mechanisms review) arrived at a rate of 0.002” per year (5 mpy)? What is the correct rate? What rate should you be using for RBI and/or other calculations that require a corrosion rate? These corrosion rates, and perhaps the information that was used as the basis for arriving at both rates, need to be validated and the information needs to be correctly revised. What would insurers, regulators, and others think if audited or, even worse, after an incident? Imagine the consequences if an organization was using the 5 mpy rate to determine next inspection and decisions such as repair, replace or run and a vessel failed but the 10 mpy rate was correct. Could a situation like this be considered a “smoking gun”? The data in these initiatives should be in sync.

Conclusion

To summarize, the challenge for 2019 and beyond is to ensure your databases are in a state of continuous improvement. SMEs  can help you better understand what data is critical and what is not. I would like to take this opportunity to encourage all to use technology and leverage work to check and reconcile the data. Take advantage of every reasonable opportunity to properly organize your databases. Validate data during your various initiatives (another leveraged activity). As the information moves toward High Fidelity, you can achieve improved effectiveness and efficiency via your fixed equipment reliability and mechanical integrity programs. Existing and emerging initiatives, software systems and the industrial internet of things are here. The data mountains are growing and so are the opportunities to mine them for gold.