As oil and gas companies continue to adapt to the new normal of lower-for-longer oil prices, there’s been a lot of talk about digitization and its potential to drive better performance. Digital innovations like the Industrial IoT, machine learning, augmented reality, and artificial intelligence are widely expected to play leading roles in the industry’s evolution over the next decade. But when will digital be ready for prime time? And will it really be the change agent that upstream players have been hoping for, to help them claw back profits lost during the downturn?
To explore this question, let’s look at one of the biggest revenue and profit opportunities that producers have—maximizing reservoir recovery.
Maximizing recovery. Since the market decline in the early 2010s dramatically slowed new capital investments, the focus has been primarily on optimization: maximizing recovery from existing wells and reservoirs while lowering production costs, wherever and whenever, possible. This approach is often carried out within two different but largely interdependent domains, which we’ll call the “reservoir loop” and the “production loop.” These can be, and often are, two entirely discrete organizations—especially in the case of major multinational companies. And therein lies
the problem.
The reservoir loop includes the geologists and petrophysicists, who analyze seismic data, develop reservoir models, and help plan drilling activities. Because of the time and computational power required, reservoir models aren’t updated with actual production data on a regular basis. As a result, the model is essentially just a best estimation of the reservoirs’ true geologic structure, with significant variances and uncertainties built into the data.
Over in the production loop, operations and production staff carry out the day-to-day tasks of field operation, working to increase well performance and bring down production costs. They determine where to apply artificial lift, when to schedule maintenance, how to fix problems with active wells, and what’s needed to ensure regulatory compliance.
These operational decisions are made without taking the reservoir model into account, usually because production crews can’t access the model data in a way that allows them to view it in the context of day-to-day events. Conversely, because the reservoir models aren’t updated with real geologic data from the field, things like well placement and trajectory are less accurate and take longer to plan. All of this can have a huge impact on the ability to extract the most product from the field.
Breaking down the silos that separate the reservoir and production loops is one way that digitization has the potential to be a real game-changer.
In this scenario, production decisions would be driven by reservoir models that can be updated quickly with information from the field, instead of probabilistic projections from initial seismic data. Engineers on the surface would be able to access model data anywhere there’s a WiFi or cellular connection. Algorithms and machine learning would boil the flood of information from the field into insight that can guide operations and identify opportunities for optimization. And domain experts would put machine learning outcomes in the appropriate context, apply “sanity checks,” and harmonize the various directions and strategies of various applications (reservoir, reliability, costs, business goals) into the right actions.
It can be done now. This isn’t just some pie-in-the-sky vision of the future. The tools needed to make it happen are available—and the time to use them is now. Why? First, the new reality of $50-to-$60/bbl oil prices has forced the industry to focus on maximizing existing assets. At the same time, more wells are being drilled in shorter time, as is the case with shales. Breakthroughs in cloud computing and better Industrial IoT connectivity have made it possible to link the various necessary systems, and provide integrated views of reservoir data in the field and anywhere else they’re needed. Combined with new modeling technology, models can be updated much faster, so that production decisions are made with the best possible information.
To be sure, the problems involved with implementing integration on such a wide scale aren’t simple ones. Sharing information in the right context is one of the biggest issues. Traditionally, the reservoir experts building the models and the operations team pumping the oil just don’t talk much. This obviously has to change. But of course, change management, in general, is a problem of its own. To take full advantage of the opportunities afoot, organizations are going to need to focus on adapting to the capabilities of the new technology. Ideally, this kind of project would be approached as part of a larger digital transformation strategy, comprised of cultural change, workflow and process changes, and IoT deployment.
Leveraging digital solutions to merge the two historically separate reservoir and production loops will go a long way toward ushering in a new era of maximum recovery for both new and existing reservoirs. The technology is here. It’s up to the operators to commit to carrying out the changes needed to make it all happen. 
About the Authors
Peter Zornio
Emerson Automation Solutions
Peter Zornio is Chief Technology Officer (CTO) for Emerson Automation Solutions and has been with Emerson for 11 years. As CTO, Mr. Zornio has responsibility for overall coordination of technology programs, product and portfolio direction, and industry standards across the Automation Solutions group. This includes Emerson’s digitization and Industrial Internet of Things (IoT) developments, such as the Plantweb™ digital ecosystem. His past roles at Emerson have included leading development and marketing for Emerson’s systems and solutions portfolio. Prior to Emerson, Mr. Zornio spent over 20 years at Honeywell in a variety of positions across the entire automation portfolio. He is based in Austin, Texas, and holds a degree in chemical engineering from the University of New Hampshire.