The presentation “Real-time Reservoir Fluid Identification from Mud Gas” will be given by Dr. Tao Yang, from Equinor.
Reservoir fluid sampling or downhole fluid analysis are the main technologies to acquire reservoir fluid properties. However, many real-time well decisions are made before or without wireline logging. Logging while drilling (especially density-neutron separation) is the primary tool to identify reservoir oil or gas. However, the fluid typing accuracy from density-neutron separation is not satisfied, and any misinterpretation will lead to large consequences regarding well placement and completion decisions. Therefore, there is a strong business need to provide accurate reservoir oil or gas identification while drilling.
Mud gas data is available near real-time for all the wells. The data is mainly used by drilling as a safety measure and by geochemistry to study the petroleum systems for post well use. Our digital innovation identified a strong relationship between mud gas data and reservoir fluid properties using a machine learning algorithm. The new method turned the underused mud gas data into real-time fluid identification, which is highly valued for real-time well decisions. Field cases are given to demonstrate that the new technology created significant business values.
The success of the technology shows digital innovation solved a long-lasting technical challenge in the oil industry. The in-expensive mud gas data provide accurate and continuous reservoir fluid property prediction in the drilling phase. The new method significantly improves the real-time well decisions. The innovation also reshapes the mud gas service in the industry as a real-time service for much broader user groups beyond drilling and geochemistry, especially for well placement, completion, and production.