Data Mining and Statistical Analysis of Hydraulic Fracture Performance in the Eagle Ford Formation

As the global supply of oil and gas from conventional reservoirs (i.e., porous rock formations) continues to diminish, it becomes increasingly important to produce these fluids from unconventional (“tight”) reservoirs. Hydraulic fracturing is generally required in order to achieve sufficient production rates from these tight reservoirs. Key questions to be addressed in hydraulic fracture design include the following: How much fluid and proppant (sand) should be injected? How many fractures should be created, and at what spacing? How is the effectiveness of the design affected by the depth, thickness, fluid pressure and temperature of the reservoir? This project will take data from over 1000 wells and use neural network (artificial intelligence) techniques to identify patterns between design parameters, reservoir properties and oil production rates. The outputs of this research will enable the partner organization (Baytex Energy) to design more effective hydraulic fracture treatments, hence increasing oil production and/or reducing well completion costs.

Faculty Supervisor:

Christopher Hawkes

Student:

Luisa Porras

Partner:

Baytex Energy Corp

Discipline:

Environmental sciences

Sector:

Mining and quarrying

University:

University of Saskatchewan