A value of information analysis of fracture prediction methods

This thesis applies a multidisciplinary approach to evaluate the technical effectiveness and economic value of a fracture characterization method that uses limited core information (including sidewall cores) to diagnose the presence or absence of open fractures. The potential value of the technique lies primarily in completion decision making. The characterization method known as the degradation (Dg) index (Laubach, 2003) uses petrographic observations of the unfractured rock (i.e., a fracture surrogate) to overcome the inherently incomplete sampling that limits analysis of large fractures. Relying on recent insights into how progressive diagenesis and fracture growth interact, degradation index predicts whether large fractures (aperture > 0.1 mm) are sealed or open by measuring the abundance of rock-mass cement that precipitated after fractures ceased opening. I conducted an accuracy assessment of the predictions using degradation index profiles for wells from the Upper Cretaceous Mesaverde Group Sandstones, Piceance Basin in Colorado; Lower Cretaceous Frontier Formation, Green River Basin, Wyoming; Permian Wolfcamp sandstone, West Texas; and Permian Weber sandstone, Rangely Field, Colorado. Results indicate that the accuracy of the degradation index technique is around 82%. To determine the economic value of insights provided by the technique, we used decision analysis and Bayes’ theorem to perform a Value of Information (VOI) analysis. Since the parameters that go into calculating the VOI are mostly unique to specific locations and depend on the exact design of the well, I built a user-friendly spreadsheet model to calculate VOI using play-specific geologic and cost parameters. Using a range of typical costs for a vertical well drilled into a tight sandstone formation having multiple completion targets, I calculated the VOI at US$ 151,750. In the base case scenario, the VOI of the degradation index is most sensitive to completion costs, accuracy of the degradation index, expected revenue from a successful well, and probabilities we assign to the state and presence of fractures in this order.