Last year, 15 well-control incidents on the Norwegian Continental Shelf (NCS) led to 35 days of nonproductive time (NPT).
To Havtil, the Norwegian Ocean Industry Authority, that translated into 35 days of unplanned activity, Roar Sognnes, principal engineer at Havtil, told JPT.
As part of its effort to reduce or prevent well-control incidents on the NCS, the agency commissioned a report that found better communication of pore-pressure uncertainty could reduce incidents related to using inadequate mud weight for the formation pressure. Well-control incidents include kicks or influxes of gas, oil, or water entering the wellbore, which in the worst-case can develop into blowouts from unrecognized or uncontrolled influxes, as well as shallow-water-flow events.
The report by GeoProvider, published earlier this year, recommended improved methodologies and technology along with clearer communication to help reduce the risk of well-control incidents related to insufficient mud weight due to underestimated pore pressure.
About a decade ago, Sognnes said, the agency examined causal factors for well-control incidents offshore Norway and found the majority had been drilled with mud weight that was too low for the formation’s pressure.
“71% of all incidents over the period from 2013 to 2022 had their origin in this problem of erroneous estimates of pore pressure or too-low estimates of pore pressure,” he said.
Similar incidents around the world were also due to the same cause. “This was quite a global phenomenon, we realized," he said.
These incidents led various governmental agencies to request the International Regulators Forum develop guidelines to help prevent well-control incidents. That work led to the publication of a recommended practice (Report 608) by the International Association of Oil & Gas Producers (IOGP).
“We adopted that as a guideline into our regulations,” Sognnes said.
Havtil tracks nearly two dozen incident types on the NCS, and of those, some, like well-control, have major accident potential. Over the past 2 decades, there has been a major reduction in many types of incidents with major-accident potential, but that has not been the case for well-control incidents, he said. Identifying pore-pressure-prediction uncertainty as a root factor in a significant number of well-control incidents suggested reducing this uncertainty could help prevent future incidents.
“If you prevent a well-control incident, you prevent major risk potential,” Sognnes said.
Not only is this a safety issue, he noted, it causes NPT and increases risk exposure.
Estimates and Uncertainty
He said formations vary, as can the knowledge about them. For example, seismic may be flawed. All of these can introduce error into pore-pressure estimates.
The well-planning process creates opportunities for failure to communicate uncertainty about pore-pressure estimates, he said. “There’s a lot of handovers in between the geology and geophysics departments with the operator, all the way through the well planning.”
What may start off as estimated intervals of pore pressure may morph into an estimated number or line on a graph. “When you send in the drilling program, you have this curve which looks fairly exact, but, in fact, it has a lot of uncertainty,” Sognnes said.
Today, he added, the fact that there is uncertainty in that pore-pressure line may not always be communicated, which goes to the heart of the problem.
Report Recommendations
The GeoProvider report offered a number of operational recommendations that fall into a trio of categories: reducing uncertainty of pore-pressure and stress prediction, following techniques developed for high-end wells such as high-pressure, high-temperature and extended-reach wells, and improving communication with wellsite personnel.
When it comes to best available techniques for high-end wells, the report recommends managed pressure drilling. Other recommendations include early-warning inflow and outflow indicators and mud-gas systems that provide early underbalance warnings.
“There’s quite a few things in this report that are very, very operational,” Sognnes said. “The report includes checklists that are quite comprehensive so that you have a very good quality assurance of the process. You should know the limits of the technologies that you use.”
Havtil-Checklist-Snippet
(Report Appendix/P69)
A better understanding of the techniques, their limitations, and how to combine relevant methods can help generate a more accurate pore-pressure estimate, he said.
Key, though, is better communicating uncertainty through methods like presenting pressure and subsurface-stress predictions in measured depth; using best estimate lower- and upper-bound labels outlined in engineering practices like DNV-RP-C207 rather than probabilistic labels such as P95; clearly stating safe drilling windows for static and dynamic mud weights for each step; and sharing depletion and injection pressure data across licenses to avoid inflicting major hazards to other teams or operators.
The report also recommends industrywide standard terminology, plots, and legends for consistent communication. The report said a pore-pressure and borehole-stress checklist could help communicate uncertainty to the drilling team, highlight the importance of calibration, and be used in dedicated risk management sessions.
Sognnes said the report and its recommendations are not planned for inclusion in Havtil’s regulations but rather to serve as a tool by the industry.
“We hope that we will, in due time, see improvement, that we will see fewer well-control incidents. That is the ultimate goal,” he said.