Blog

What Is Predictive Analytics and How Could It Help the Oil and Gas Industry?

What Is Predictive Analytics and How Could It Help the Oil and Gas Industry?

Last updated Jul 29, 2021 | Published Jun 29, 2015 | Petrochemicals & Refining

Discussion of the uses of predictive analytics in the oil and gas industry has been around at least since 2008, though only recently has the conversation gotten more intense.

Publications like E&P, Oil & Gas Journal, Oil & Gas Monitor, and The American Oil & Gas Reporter published major stories on the topic in 2014, and IBM hosted its own free one-hour seminar on predictive analytics in the industry this past November.

What is predictive analytics, and why all the talk? Predictive analytics is essentially the process of using any of one or more statistical techniques to analyze real-time or historical data, with the intent of making some sort of prediction about the future. The subject has become more prominent largely because of growing dialogue about a related topic: big data. As business and research entities around the globe produce increasingly complex data sets, tools for not only organizing and storing but also filtering and analyzing the data have become necessary. Predictive analytics is one of the tools that have emerged from this need.

Many data management experts have suggested ways in which predictive analytics can help in the development and management of upstream and downstream facilities. In a recent seminar, IBM suggested companies could “extend asset life” and “improve quality, operational performance, and profitability” by incorporating more integrated data sources into optimization strategies. And as the workforce shrinks and efficiency needs grow tighter, using a facility’s data streams to better optimize business processes and equipment increasingly makes more sense.

A 2014 study by GE and Accenture found that while 65 percent of energy companies are using operational data for monitoring purposes, only 29 percent of those companies are using that same data for optimization and predictive analysis. While that number is likely to go up, investing in predictive analytics tools and methodologies isn’t enough; gaining the trust of management and end-users, demonstrating how and why these analysis tools work, is also vital. Devon Energy’s data scientist Beau Rollins expressed this idea in an October 2014 interview with E&P Magazine: “You can build a model that is very sophisticated and probably very accurate, but if you can’t tell anybody how to use it, they won’t trust it.”