Abstract

Drilling and producing oil wells safely and with minimum risk of environmental hazards is crucial in Abu Dhabi offshore fields. A significant challenge to achieve this goal is steering in very selective sub-layers without conventional radioactive source porosity tools. Acquiring real-time data is critical, especially when drilling slim holes with long horizontal drains, and a well profile that targets several thin reservoir beds.

To meet this challenge, a nuclear magnetic resonance (NMR) sourceless porosity tool was introduced for the first time in this reservoir in a six-inch hole. Two lateral drains, each more than 3,000 ft., were drilled, targeting different reservoirs. As a part of the geosteering methodology, nearby offset well properties were incorporated to generate a forward model along the planned well trajectory. The forward integrated geosteering model was updated in real time with porosity and gamma ray data, resistivity images and, structure views, etc. Decisions were made from this input to place the trajectory in the optimum target zones, achieving the reservoir objectives.

This paper demonstrates the successful use of an NMR sourceless tool's real-time data acquisition that played a significant part in facilitating effective geosteering and well placement. Real-time NMR total porosity was critical to confirm optimum well placement, especially where lateral facies variations were expected and due to the limited availability data for such a green field. In addition, the available porosity distribution was used to estimate a permeability index. This index, after normalization to offset the well data, was used for selecting formation pressure while drilling and reservoir zonation.

The case study in this paper shows that with current advance technologies, incorporated data for available wells, and real-time technical support, the demands of sourceless drilling can be achieved to place the well in a precise target during slim borehole drilling.

You can access this article if you purchase or spend a download.