How the oil and gas industry exploits IoT

Like many traditional industries that have long-standing, tried-and-true methods of operation, the oil-and-gas sector hasn’t been the quickest to embrace IoT technology – despite having had instrumentation on drilling rigs, pipelines and refining facilities for decades, the extraction industry has only recently begun to work with modern IoT.

Part of the issue has been interoperability, according to Mark Carrier, oil-and-gas development director for RTI, which produces connectivity software for industrial companies. Energy companies are most comfortable working with the same vendors they’ve worked with before, but that tendency means there isn’t a strong impetus toward sharing data across platforms.

“On a very low level, things are pretty well-connected, at the connectivity to the back-end they’re well-connected, but there’s a huge expense in understanding what that data is,” he said.

Christine Boles, a vice president in Intel’s IoT group, said that the older systems still being used by the industry have been tough to displace.

“The biggest challenge they’re facing is aging infrastructrure, and how they get to a more standardized, interoperable version,” she said.

Changes are coming, however, in part because energy prices have taken a hit in recent years. Oil companies have been looking to cut costs, and one of the easiest places to do that is in integration and automation. On a typical oil well, said Carrier, a driller will have up to 70 different companies’ products working – sensors covering everything from flow rates to temperature and pressure to azimuth and incline, different components of the drill itself – but until fairly recently, these all had to be independently monitored.

An IoT solution that can tie all these various threads of data together, of late, has become an attractive option for companies looking to minimize human error and glean real-time insights from the wide range of instrumentation present on the average oil rig.

Those threads are numerous, with a lot of vertically unique sensor and endpoint types. Mud pulse telemetry uses a module in a drill head to create slight fluctuations in the pressure of drilling fluid to pulse information to a receiver on the surface. Temperature and pressure sensors operating in the extreme environmental conditions of an active borehole might use heavily ruggedized serial cable to push data back aboveground.

Andre Kindness, a principal analyst at Forrester Research, said that the wide range of technologies, manufacturers and standards in use at any given oil-and-gas facility is the product of cutthroat competition among large traditional players in the industry. ABB, Siemens and Rockwell, to name a few, have worked hard to avoid losing their edge through interoperability.

Yet those companies, too, have been moving with the times. While traditional IT companies like HP and Dell have seen comparatively little success in selling to energy companies directly, computing firms have had more luck making whitebox edge compute devices for the Siemenses and Rockwells of the world.

Edge computing is a particularly important technology among oil-and-gas companies, given the remoteness of some installations, potentially insufficient local networking infrastructure in some countries and the consequent difficulty of connectivity backfill. Safety and maintenance applications for oil drilling aren’t tolerant of lag, so a round-trip between a remote rig and the home office is likely to be impractical. Simply put, it’s much easier to do some of the computational work close to the endpoint.

On the network

The main use cases for oil-and-gas IoT are preventive maintenance, centralized control and operational insight. All of these rely on getting information from the myriad of sensors attached to a given drilling rig, refining facility or pipeline into an edge device or back to the cloud. Wi-Fi is a popular medium for connecting endpoints in a refining facility, although low-power WAN options are also used, according to Carrier.

Sensors on a drilling rig, as mentioned, use a wide variety of wired and wireless technologies, some proprietary, to get information into usable form for edge or data center processing, and pipeline technology usually uses wired industrial Ethernet.

It’s complicated, according to Kindness, in part because of the long life-cycle required of equipment in the extraction industry.

“Unlike the IT world, we’re not replacing things every three years,” he said. “A lot of this equipment is in the field for 10 to 20 years, and that’s why you see so many variations of Ethernet being used.”

That said, equipment manufacturers that work with oil and gas companies are increasingly alive to the need for smooth connectivity, according to IDC associate vice president of manufacturing insights Emilie Ditton. Newer equipment is natively instrumented and less dependent on proprietary connectivity standards, which generates more usable data for those companies.

“There are some pockets of extreme innovation, and some areas where processes are still very manual,” she said. “The level of maturity is very variable.”

Moving data from remote locations to the cloud or data center also presents challenges to the oil and gas industry. Where coverage permits, carrier connectivity can be used for wireless backhaul, but that’s often not an option for oceanic oil rigs. Some companies use satellite connectivity for this purpose, while others run underwater fiber-optic cables directly to their offshore facilities.

The embrace of fiber-to-the-platform is a relatively recent phenomenon, according to Jeremy Calac, a product manager and engineer with TE Connectivity, which makes network-connection equipment for industrial customers.

“In the past, the fiber-optic underwater option was seen as unreliable because of temperature and pressure issues,” he said, and that’s changing because of improved reliability among connection cable manufacturers.

Crunching the numbers

However the data makes it to the edge or cloud, the real value is created via automated analysis. Historically, the industry has had access to the type of data needed for this analysis since the 1980s, according to Jim Wang, director of engineering at Corva, a startup that makes a single-dashboard visualization product for oil and gas companies.

That said, the energy industry has until very recently lacked the ability to process that data in a timely way, with the result that a great deal of human error has crept into business processes. Drilling a well, for example, relies on extremely precise measurements from a wide range of sensors, and checking everything “by hand” can be a complex process.

“Humans are not great at making sure all the data is calibrated manually,” Wang said. “Until you have real-time analytics, you wouldn’t know there were errors until much later.”

Using modern machine-learning techniques, however, the oil and gas industry can realize gains in several areas. At the drilling level, of course, smarter systems can synthesize information from all the sensors involved to create a much more accurate and timely picture of whether a well is being dug accurately, how smoothly the machinery involved is functioning and how productive a well is likely to be, both in the short- and long-term. Industrial IoT principles can be easily applied to refining facilities – tracking maintenance data, identifying inefficient processes and potential safety issues, and monitoring pipelines so that problems can be spotted before they become severe.