Edge computing comes to the oil and gas industry

Forum bolsters producers’ confidence in technology-powered future.

By Sidney Hill, Jr. February 1, 2018

The watchword was optimism as several hundred individuals, all interested in the confluence of digitization in the upstream oil & gas industry, gathered in Houston this past December for the TIBCO Energy Forum. This annual event is hosted by TIBCO Software, a company that specializes in helping corporations organize and analyze data for strategic business purposes.

The prevailing story line at the event was that oil & gas producers finally have figured out how to remain profitable, even if prices never return to the highs seen in the early 2000s. The answer, according to most forum attendees, is to use information technology to create business models that can withstand shifting market dynamics.

There also seemed to be a consensus that analytics—which is sometimes equated with Big Data—is the ideal technology for building these new oil & gas business models.

It’s not surprising that a collection of TIBCO users would see analytics as the foundation on which to build structurally sound business models, since most of them already have chosen that path. There is evidence, however, intelligent analytics can, in fact, make almost any business—oil and gas companies included—more efficient, which generally translates into higher profits.

It’s not how much you spend

In his opening remarks at the Energy Forum, TIBCO Software CEO Murray Rode referred to a study recently published in the Harvard Business Review that discovered companies experiencing the following results after investing in analytics:

  • Revenue growth five times better than their peers
  • Profit growth eight times higher than their peers
  • Return to shareholders two times higher than their peers.

Even more interesting, Rode said, is the study found almost no difference in how much is invested by companies deemed leaders in analytics in comparison to those considered industry laggards. The leaders, on average, spend 2.5% of their revenue on analytics while the laggards spend 2.3% of revenue.

"The difference is not in how much you spend, but in how you spend those dollars," Rode said.

Throughout the forum, both TIBCO executives and customers shared stories of how individual oil & gas companies are spending their dollars on analytics, and the benefits they’re reaping from those investments.

Examples ranged from companies using well sensor data to optimize maintenance schedules to other companies changing drilling paths in real-time in response to anomalies spotted on an analytics dashboard. There also were examples of companies developing completely new business models with analytics at the core, as well companies managing what could be considered Industrial Internet of Things (IIoT) platforms.

Shawn Rogers, TIBCO’s senior director of analytic strategy, argues that there is no reason for a company to invest in IIoT technology without a plan to connect it to an analytics platform. "Simply collecting data has no inherent value," Rogers said.

"Analyzing data and taking action based on that analysis is where the payoff comes." Analytics can add value to an oil company’s IIoT infrastructure, according to Rogers, by quickly filtering through all the data collected by sensors and identifying the few specific data points that are signaling potential problems. He also said TIBCO offers solutions that allow for doing two types of analysis, both of which are important to helping oil and gas companies thrive in increasingly volatile markets. 

Which fluctuations are significant?

Analytics platforms have always allowed for bringing sensor data to a central platform, where the data can be assessed for historical trends that can help in adjusting processes to prevent future problems. Recently, however, Rogers said the company has developed tools that allow for taking analytics "to the edge" where operators can take readings from individual devices in the field. In those scenarios, operators can spot potential problems, and adjust to avoid them, in real-time.

"Analytics at the device determine which fluctuations in the data are significant," Rogers said. "If I can analyze data at the edge, and know that I’m only looking at the important data points, that’s a huge payoff."

Michael O’Connell, TIBCO’s chief analytics officer, said ConocoPhillips is one of many companies who have used TIBCO’s Spotfire analytics platform in this fashion.

ConocoPhillips reported a 50% reduction in the time it takes to drill wells after linking the Spotfire platform to the hundreds of sensors it has attached to wells in the Eagle Ford shale basin in South Texas. Spotfire constantly runs comparisons on data from all those sensors. It will automatically adjust settings—such as the weight being placed on a drill bit or the speed at which the bit is moving—if a well starts pumping at a lower rate than other wells that have not shown changes in such settings.

Industry titans like ConocoPhillips are not the only oil & gas producers reaping big benefits from analytics.

Several smaller producers—including Anadarko Petroleum and Apache Corp.—shared their stories at the TIBCO Energy Forum. These companies also offered examples of how analytics can support new business models.

At Apache, analytics is one part of an entire information technology infrastructure the company developed following some corporate soul searching in the wake of the industry downturn that started in 2014. "After the market shakeout, we had a new leadership team, and we looked at what we needed to do to come out of the downturn as a stronger company," said Travis Osborne, Apache’s director of information management. 

Five key systems

From an IT perspective, Osborne said, Apache determined it needed easier ways of getting data to people who make decisions in the field. To accomplish that, Osborne’s group undertook a project that involved first developing a smooth process for managing and moving data across the enterprise. That led to the purchase of a TIBCO product called BusinessWorks, which provides a visual environment for developing business processes, along with a messaging system for packaging and transporting data that support those processes.

"Using BusinessWorks, we were able to build an architecture that allows us to follow data from where it originates to where it lands," Osborne said. "That gave everyone a good insight into how processes were working, and whether any of them needed changing."

Osborne said it took roughly 90 days to build this architecture, which also helped Apache identify five key systems that were most critical to the operation of the business.

The next step was using the TIBCO Enterprise Messaging System (EMS) to build a backbone for seamlessly moving data between those five systems. That took roughly 15 months, which also involved connecting the messaging backbone to the TIBCO Spotfire analytics platform, where data can be viewed and analyzed.

"We already had Spotfire in place, but it had been used primarily as a tool for engineers, allowing them to do some self-generated analytics," Osborne said. "You might call it a bit of early data science."

As part of Apache’s new information architecture, Spotfire has become a mainstream analytical tool available to everyone in the enterprise.

"We’ve gone from it being a tool exclusively for engineers and people solving scientific problem to accountants, field foremen, and pumpers at the well site all having access to Spotfire content," Osborne said. "We’ve pushed it out to the pumper foreman level, so the guys on the front line making production management decisions can get more information on how we manage assets, as well as how we spend money," he said.

Making that data widely available allowed Apache to implement what Osborne referred to as a "ground up" approach to financial management.

"We want everyone looking at the same data—from the people in the field to regional vice presidents and executive vice presidents at the corporate office. The visibility of that data is valuable because people in the field see things first. To see how their actions either increase revenue or lower cost is golden."

Teaming engineers with data scientists

Anadarko took a slightly different approach to pushing analytical data to the field. It hired a group of data scientists and matched each of them with a petroleum engineer to create two-person teams that began exploring ways of using data to improve operations. These teams are part of what Anadarko calls its advanced analytics and emerging technologies group.

One of the teams—consisting of Chad Loesel, a staff drilling engineer and Dingzhou Cao, a staff data scientist—gave a presentation at the TIBCO Energy Forum.

The teams were charged with finding ways of getting data from the field to the corporate office, where it could be analyzed to produce meaningful information that could then be relayed back to the field in quick fashion.

"We always had access to real-time drilling data," Loesel said. "But historically by the time that data reached the office, it was old and static. We wanted to combine real-time drilling data with advanced analytics to see what we could gain in terms of a competitive advantage."

As they examined that question, the teams were all given one mandate: to move quickly in deciding whether a project had the potential to generate a real return on investment. "We were looking for rapid proof of concept that we could make work, or we were moving on," Loesel said.

Roughly a year later, Anadarko has a full-blown real-time data analytics system in place, with four modules already in use and three others in the pipeline.

The modules currently in use are: 

  • Rule-based drilling activity recognition, which constantly monitors drilling activity and provides useful information to the remaining analytics modules
  • Sliding drilling guidance system
  • Key drilling performance indicators
  • Torque and drag
  • Trajectory and recognition. 

A rotational drilling guidance system was set for deployment at the time of the forum event, and modules for analyzing hydraulic models and wellbore trajectory were under development, according to Loesel.

A TIBCO product called Streambase is the backbone of Anadarko’s real-time data analytics system. Streambase is programmed to move the data users request into the Sportfire platform, where it’s analyzed and pushed back to Streambase for delivery to the appropriate users via the Spotfire analytics dashboard.

Cao, the data scientist, said Anadarko was able create the first prototype of an analytic application in three months, largely because the TIBCO tools are built on an open platform that allows for making quick connections between systems that need to share data.

Now Anadarko users have full, real-time visibility of field data. "We have taken computing power to the edge," Loesel said. "We can identify problems and get rigs back on track immediately."

Having ready access to this type of technology is why oil & gas producers are more confident about their ability to prosper, even if prices within the industry remain volatile.

Sidney Hill, Jr. is a graduate of the Medill School of Journalism at Northwestern University. He has been writing about the convergence of business and technology for more than 20 years.


  • Oil & gas companies are spending on analytics in pursuit of unique business models.
  • A reported 50% reduction in the time it takes to drill wells in response to ubiquitous sensing capability.

An architecture provides insight into where data starts, and where it ends.

Original content can be found at Oil and Gas Engineering.