Developing a smarter way to drill for oil with analytics
In recent years, the energy market has been especially tough for oil companies. Oil prices have fluctuated wildly, and new market entrants are encroaching on the market share of established producers.
All of this has put oil companies under increased pressure to streamline their operations and make sure that rigs run as effectively as possible.
At TOTAL, we understand that running an oil well profitably and efficiently is mechanically and technically complex. That’s why we are always looking for new ways to innovate and become more efficient.
Becoming more productive
There are cases in which a significant amount of drilling time ends up being classed as non-productive time (NPT). That’s the time taken to return to an original drilling plan after a deviation necessitated by failure of equipment or any other event that causes drilling operations to stop.
One of the common causes of NPT is a phenomenon that we call “stuck pipe.” It’s quite literally when the drill pipe or string can no longer be moved up or down, or it can’t be rotated within the wellbore to continue normal drilling operations.
Stuck pipe is preventable, but it’s difficult to predict when it may occur. Wells are constantly changing, and the traditional, physics-based models that engineers use to guide drilling operations management and decision-making provide delayed insights on the condition in a well at any given time.
We had a new idea. We decided to use IBM Analytics technology to build a near-real-time picture of the condition of a well. With this insight, we could provide engineers with a tool to spot conditions that could lead to stuck pipe and other events that typically cause NPT.
Convincing the experts
When we told our engineers, who have been working on this same problem for their entire careers, that we had found a potential solution, they were skeptical. Understandably, they wanted proof.
So we put our money where our mouth is and started a project to develop a predictive drilling analytics solution. The idea was to combine our engineers’ expertise with physics-based models of wells to build statistical models to help us predict when stuck pipe might occur.
Using cutting-edge technology from IBM Cloud and the IBM Watson Studio platform helped us get our idea off the ground fast. Choosing a cloud solution meant we avoided spending lots of time and money sizing and deploying infrastructure. It enables us to easily scale up performance and capacity as our models become more sophisticated and data volumes grow.
Now that we have tested our predictive drilling analytics solution on two wells, we have already spotted conditions that can lead to stuck pipe. As we continue to further train our statistical model with the help of IBM, our solution will become even more adept at identifying issues that cause NPT.
Once operational, we believe that our solution will have a truly transformative impact on the oil industry.
Since engineers can better avoid stuck pipe and other significant NPT events, wells can be drilled in shorter time, with reduced cost and increased project profitability. Increased drilling efficiency enabled by use of IBM technology will help companies weather tough market conditions.
Thanks to the IBM team and the expertise and dedication of engineers at TOTAL, we are at the vanguard of next generation oil extraction and have created a tool that could be transformational.