Finding the sweet spot
Dr. Geoff Rideout, mechanical engineering professor and director of first-year engineering in the Faculty of Engineering and Applied Science, researches how to predict and control vibrations. His research, as a member of Memorial’s Drilling Technology Laboratory led by Dr. Steve Butt, can help in the design of efficient and reliable drill strings for the oil and gas industry.
The drill string — the pipe leading from an offshore rig down to and into the ocean floor — consists primarily of steel pipe lengths that are connected end-to-end, sometimes for several kilometres. As the diamond-headed drill bit grinds its way through solid rock at the bottom of the ocean, the steel pipes following the drill bit can deform and vibrate in a variety of ways.
Twisting, bending and other vibrations can sometimes be felt right up to the floor of the rig. These vibrations can damage downhole equipment like the battery-operated sensors, which detect drilling trajectory and help the driller steer the bit during directional drilling, other drilling tools and components in the bottomhole assembly, the drill bit, and even sections of the drill string itself.
“They no longer have to drill straight down, extract whatever oil is there, move the rig, and drill straight down in a new location,” explained Dr. Rideout. “The drill string can now follow a complex, predefined three-dimensional path so that maximum oil production can be achieved for a given field.”
Drilling companies need to predict how the drill string is going to vibrate and avoid drilling in directions that will cause too much vibration. Impacts and stresses from vibrations can lead to failure and premature wear of downhole components. In turn, this could potentially lead to the shutdown of offshore drilling operations, which could be in the range of a million dollars a day.
Dr. Rideout and his team can predict the various types of vibrations and how they may affect downhole equipment by using finite element modelling, a popular type of computer modelling that can capture very fine detail in the physical system.
The only trouble is that finite element modelling can take days to predict a few seconds of vibration. To reduce the time needed, Dr. Rideout’s research uses simpler models, which take far less time but are not always as detailed or accurate as a finite element model.
“We are looking for a proper model that is just complicated enough to predict what you want, but not any more complicated,” he noted. “It’s no good to have a fast model if it’s wrong. It’s a trade-off between accuracy and speed. I’m looking for the sweet spot. Sometimes it’s tempting to add complexity to a model just because you think you know how… but as soon as you add real life complexity to uniform pipes, the math becomes pretty crazy in a hurry. You have to make a lot of assumptions,” he said.
To deal with these challenges, Dr. Rideout and recently-graduated PhD student Mejbah Sarker have come up with a model that divides continuous pipes into shorter chunks joined by 3-D springs. This model acts like the real thing, while allowing Dr. Rideout and Dr. Sarker to insert tool and sensor submodels at any location.
“If you think of the drill string as a bunch of smaller masses that can’t stretch, connected by lightweight springs, then that system is mathematically much easier to deal with and will vibrate like the continuous system if you break it into enough chunks. The computer will crank through the resulting equations much faster,” Dr. Rideout said.
“Now that we have the models, we want to use them to perform forensic analysis on broken components, and to design better ones — or at least figure out where to put them in the drill string to keep them out of trouble.”
In a perfect world, Dr. Rideout’s proper model of masses and springs will properly predict bit vibrations and avoid unnecessary and costly shut downs in the oil and gas industry in the future.
Though his work is complex, Dr. Rideout finds inspiration in the words of one of the greatest scientists.
“As Einstein once said: ‘Everything should be made as simple as possible but not simpler.’”
This project was funded by the Atlantic Canada Opportunities Agency Atlantic Innovation Fund.
This story was first published in The Telegram on Aug. 14, 2017.