With the promise of saving hundreds of thousands of dollars for each oil well drilled and preventing costly equipment failures, there are compelling reasons for companies in the oil and gas industry to embrace IoT. Early adopters are already seeing the benefits.
We talked with Derick Jose and Rick Harlow of Flutura, an AI solutions company focused on oil and gas, chemicals, and manufacturing, about what they are seeing in terms of how IoT is being harnessed in the energy sector.
“There is a lot of digitization on the exploration side to minimize costs,” says Jose, Flutura’s chief product officer. “On average, it takes about 18 days to drill a well, and each day of drilling costs about $130,000.”
The perfect drilling day
But, Jose says, that drilling time should ideally be about 15 days, if all the small, often hidden, inefficiencies can be taken out of the process.
“Drillers want to shave off as many days as possible, and there has been a lot of emphasis on using sensor data to gain efficiencies. The ability to look at sensor data and figure out where their micro-inefficiencies are, in real time, that is huge win for them,” Jose says.
Harlow, executive vice president, adds that what drillers really dream of is the “perfect drilling day.” Any percentage of time they can shave off the drilling process is an advantage.
One way to gain an advantage is to leverage sensors in all the drilling and related equipment, to be able to catch the earliest indications that a piece of equipment might fail and bring the process to a halt.
“They (drillers) want insights about the lead-up to a possible failure, so they can plan any maintenance events, instead of facing non-productive downtime,” Harlow says.
IoT and the art of equipment maintenance
With IoT sensor data, oil and gas companies can move from time-based maintenance to conditions-based maintenance. Rather than check out a site just because it has been six months since the last visit, technicians can focus on those sites where the data is indicating that equipment is showing initial signs of failure.
“In some places, you may have to drive a couple of hours to get to one of these wells,” Harlow points. “Better data means that technicians can go directly to the well site that is having issues. If they’re on time-based maintenance, they may not be able to get there for days or even weeks.”
Improving maintenance via IoT is also a matter, he says, of being able to leverage data from historical and maintenance systems along with the data from the equipment, and deriving value from that. When you know about previous maintenance events and when and why failures occurred, you can address the future more effectively.
IoT can even take this a step further by tying in the spare parts inventory and determining the most qualified technician for each issue, by leveraging some measure of artificial intelligence to help make those determinations.
Self-driving drilling and platforms?
Being able to perform autonomous drilling is analogues to the surging interest in self driving vehicles. It’s still in the visionary stage, but the industry is definitely looking in that direction, where the drilling and the oil rig operations are done with a minimal number of workers on-site.
Harlow says one European company is already talking about operating sea-based drilling platforms with as few as 10 workers, a reduction from the 30 to 40 employees typically working there today.
To accomplish that, he emphasizes, will require a heavy dose of artificial intelligence paired with ongoing IoT sensor data and vast amounts of historical data, so that the systems truly “understand” efficient and safe platform operation. But clearly it represents a major path to cost reduction.
The new revenue approach
While you can never go wrong by cutting costs, there is another advantage to IoT and all the data it makes available: embracing a new business model and revenue stream.
With a platform-based business model, operators would collect all the relevant sensor data and either provide raw data to the numerous contractors they work with or handle the analysis and deliver refined insights to them. Jose said he expects this model to take off in the next two to five years.
Overall, the industry is still generally in the early adopter stage, Jose says, due to the various challenges involved with IoT.
“One of the primary challenges is handling the vast amounts of data that get flushed out in the process,” he says. “To set up a network and the analysis and predict lost time is a capability not every electromechanical company has, and they are in the early phases of building this capability.”
However, he adds, more companies are further along when it comes to using IoT data to predict equipment failure. Some of the other applications will follow in due time.