The Unfinished Revolution: What is Missing From the E&P Industry's Move to "Big Data"

Abstract

Guest editorial One cannot help but be impressed by the inroads that digital oilfield technologies have made into the exploration and production (E&P) industry in the past decade. Today’s production systems can be monitored by “smart” sensors that allow engineers to observe almost any aspect of performance in real time. Our understanding of how reservoirs are behaving has improved considerably since the dawn of this revolution, and the industry has been able to move away from point answers to more holistic “big picture” integrated solutions. Indeed, the industry has already reaped the rewards of many of these kinds of investments. Many billions of dollars of value have been delivered by this heightened awareness of what is going on within our assets and the world around them (Van Den Berg et al. 2010). Underpinning this heightened awareness is data—lots of it. As Beckwith (2011) notes, today’s seismic data centers can easily contain as much as 20 petabytes of information, equivalent to 926 times the size of the US Library of Congress. If this amount of information was copied into books and put on a single continuous bookshelf, it would go around the Earth’s equator six times. And while seismic data sets are notoriously large and cumbersome, practically every part of the E&P ecosystem is generating orders of magnitude more data than it used to. What is more, there is every reason to believe that this trend toward more digital information is just getting warmed up. Martin Hilbert of the University of Southern California estimates that the total amount of digital data in the world—including items such as books, images, e-mails, music, and video—doubles every 3 years or so (Mayer-Schulnberger and Cukier 2013). In light of this sweeping global trend, it is hard to imagine a future in which the E&P industry is not collecting significantly more data than it does at the moment. With such big piles of digital information accumulating around us, it is easy to understand why many in the E&P sector believe that it is solidly on track to reap the benefits of the “Big Data” revolution. But the E&P sector seems to be approaching these rapidly growing piles of data with the same attitudes and analytical techniques that have been with us for years. As Feblowitz (2013) suggests, a lot of potentially valuable digital information harvested from upstream oil and gas assets is barely given a cursory glance, and much of it is simply thrown away. Moreover, in those instances in which data is stored, it is often kept by the service companies responsible for generating it rather than the operator in charge of managing the long-term welfare of the asset. This is not how the Big Data revolution is unfolding in many other industries. Several other sectors—most notably, the health care, financial, retail, and media industries—have come to realize that new and valuable insights are frequently gleaned from using new techniques to analyze massive data sets in ways that were never possible with smaller ones. These insights tend not to be discovered by testing hypotheses between variables whose relationships are well understood; rather, they are found by applying advanced analytical techniques to massive numbers of variables that, at first blush, might seem to be unrelated. FICO, an American analytics company, has discovered a surprisingly tight relationship between aspects of a person’s car ownership records and their propensity to take prescribed medication.