The prospect of natural gas in a riser commonly summons up a vision of a deadly, fast-rising bubble exploding on to the floor of a drilling rig. But comments by consultants on an OTC panel argued that it is a poor description of how gas behaves in a deepwater riser.
Research and experience indicate that the movement of a volume of gas is not nearly as fast or unified in a mud-filled deepwater riser as widely assumed. A trip up a mile-long riser can take hours during which the gas disperses like carbonation in a can of cola. For most of the trip it is in a liquid form.
“A major problem is that many people focus on riser gas velocity, rather than behavior caused by gas that may not actually be ‘rising’ at all, but is nonetheless ‘expanding,’” said Paul Sonnemann, vice president of technology for Safekick. While gas is capable of expanding in an almost explosive fashion if it is released into the air, he said this can be “delayed by either slow migration, or by dispersion of the gas in a long fluid column.”
Based on the only reported test in a deepwater well, gas movement is glacial compared to the rates reported in papers stating it can rise at 100s or 1,000s of ft/min in a riser. “From a full-scale field experiment in deepwater, we have a good idea of how fast annular gas can rise. It is less than 20 ft/min,” said Robert Ziegler, president of RZI Deepwater.
Higher numbers are predicted in “hundreds of papers on gas behavior in drilling fluids going back to the 70s.” Those are the basis of simulations predicting what gas will do in a riser, said John Kozicz, technology manager for Transocean.
The predictions are based, in part, on laboratory tests observing fluid movements in relatively small tubes, Ziegler said. What has been learned about gas movements from experience using it for gas lift is that larger tubing sizes quickly reduce the efficiency of a gas-lift system, so a test in a 2-in. tube is not comparable to what will happen in a 20-in. diameter marine riser, he said.
Pressure in the riser also alters the picture. In a deepwater well, natural gas remains in a liquid form until it rises to near the middle of a mile-long riser, and then gradually begins changing state, Ziegler said. If the drilling mud has an oil or synthetic base, the gas comes out of solution only near the top of the riser, and he said that phase change can be arrested by adding a few hundred pounds of backpressure, which a managed pressure system can do.
There are occasions when gas from a riser reaches the open air, expands extremely rapidly, and shocks the crew by spraying drilling mud on the deck. Such unwelcome incidents are a source of estimates concluding gas rises rapidly in risers. Ziegler said those calculations are based on assumptions about the movement of gas up the riser by people who had not observed the problem developing because “otherwise they would have done everything they could to avoid it,” he said.
As evidence against the notion of big, fast-moving gas bubbles, Ziegler described the deepwater well test done in 1986 by Shell, Amoco, and Exxon. It is little known because the results were only reported in an article in the Oil and Gas Journal published 17 July 2000, long after Amoco was acquired by BP, and Exxon added Mobil.
In the test, a series of increasingly large volumes of air were injected below the BOP of an inactive well in more than 3,000 ft of water. The test began with an injection of 10 bbl of air, which acts in a well like methane. The air injections were incrementally increased to 50 bbl.
The injected 10 bbl of air remained in place for 3 hours, until the mud pumps were turned on. Then it took more than 6 hours to reach the surface, and by then it was widely dispersed, Ziegler said.
Increasingly large injections followed with similar results: slow movements of gas dispersed in the fluid. While larger amounts did create gas slugs—the 50-bbl injection at depth which expanded to 13,622 bbl at the surface created seven slugs—they could be handled by surface equipment if circulated out slowly.
The test stands out because the result runs counter to the common wisdom, and because in-well test results are hard to find. More well data are needed to build a realistic view of this little-studied place. “Once you have the ability to isolate the riser at the bottom and the top there are a variety of things we are prepared to do,” Sonnemann said. “We are studying it. It is a totally different animal.”
