Multiphase Flow in Concentric and Eccentric Annulus Spaces

On the 20th of April 2010, explosion and fire killed 11 people on the Deepwater Horizon oil rig in the Gulf of Mexico. After two days in flames, the rig sank and left the riser (the pipe going up from the seabed) on the sea floor leaking oil and gas from the well through the gap between the drilling riser and the drill-string (rod holding the drillbit) inside it. The accident is known as the Macondo accident. Immediate concerns were raised regarding the amount of oil being released into the Gulf of Mexico killing fish and seabirds. The size of the oil spill has also played a significant role in the court proceedings where the operator BP at the end accepted to pay a fine of more than 18 bn US dollars. One of many techniques that were applied was based on using simulation software to match the observed frequency of plugs of oil flowing out of the sunken riser. There are, however, serious flaws to this approach. The most important is the lack of reliable models for multiphase flow in annuli, that is, flow of oil and gas in the ring shaped space between the drillstring and the outer pipe. Commercial simulation software packages for flow of oil and gas in pipes lack the ability to correctly predict pressure drop in annulus flow, even when the flow contains only gas or only liquid. When both gas and oil are flowing in the pipe at the same time, which is called multiphase flow, the models are even worse. Well design is thus based on crude formulas with large uncertainty and high error margins. This may lead to inefficient well design and high risk of production problems. ANNULUS has brought a step change in the scientific knowledge in this field and provided data and physical understanding that can be used as a basis for developing more usable and accurate commercial simulation software in the future. Improved models can be used by operators and contribute to significantly safer and more efficient and profitable well operations and production. The project work is now completed and has produced a large amount of unique data that give a vast amount of information about these types of flow and how flows of oil and gas in annuli differ from those in ordinary pipes. We have also developed a new and unique simulation model describing the flow phenomena. We have already given three presentations at scientific conferences. Six scientific journal papers have been published, one more has been submitted to a journal and one is close to being finished.

The project has resulted in two spin-off projects with an oil major, where related but different conditions to the ones in the researcher project were studied. The impact of the work on society is more difficult to assess in the short term, but depends on the results being picked up by software companies as well as operators. The results should be well suited for developing commercial models that can be provided to the operators by the major software providers, but this depends on company priorities. If models are provided to the operators and deployed in their daily use, improved well design and operation should be possible, potentially enhancing safety in drilling and production as well as potentially contributing to increased public revenues from oil production.

"Annulus flow" often occurs in oil and gas wells where oil and gas flow between an inner and an outer pipe in the well. This type of flow is not well understood and has not been studied scientifically except in very simplified cases. Commercial software lacks ability to correctly predict annulus flow, and well design is based on rough models with large uncertainty. This may lead to bad well design, risk of production problems, and inability to model accident scenarios. This project will be led by IFE and include detailed experimental studies and theoretical and numerical modelling in close cooperation between IFE and UiO with one PhD student, one postdoc and three supervisors, and support from Imperial College. The experiments will be run in the medium scale, inclinable Well Flow Loop at IFE. This loop is transparent, allowing flow visualisation, and equipped with advanced instrumentation including a unique X-ray system. Annulus test sections will be built with adjustable eccentricity. The X-ray system will be modified to accommodate annulus geometry. The modelling/simulation part will pursue detailed simulations of two phase annulus flow through a combination of theoretical analysis and detailed CFD simulations using a research code from Imperial College. The models will be compared to the experimental data and are intended to aid development of future engineering models. IFE has more than 30 years of experience in developing models for multiphase pipe flow. ANNULUS will bring a step change in the knowledge in this field and provide data and theoretical understanding than can lead to better engineering models in the longer term. Improved models can be used by operators and contribute to safer and more efficient well operations. There is a strong commercial software industry in Norway for multiphase technology that can benefit from this new knowledge in the longer term, but right now the industry is unable to find funding for such high-risk fundamental research