Drilling and Completion Optimization in Unconventional Reservoirs with Data-Driven Models Available to Purchase

We shall discuss the following objectives: the use of Multivariant Analysis (MVA) to identify and classify important performance variables implemented in the hydraulic fracture treatment strategies. We shall discuss field optimization workflows initially performed in the Pinedale asset in Wyoming. Data were collated from multiple fluvial sand layers that aggregate across the anticline structure under study. Initial exploratory data analysis and bivariate analyses fell short of a comprehensive appreciation of the multivariate, stochastic and multivariant nature of this complex heterogeneous system. The MVA approach addressed the complexity "inherent in the data's coincident variation in multiple parameters. Data clustering was used to create different models and assess different parameters. Models were able to identify the relative impact of the most significant variables affecting stage production performance, and develop probability distributions for potential outcomes at different categories of production. A neural network was chosen to evaluate both reservoir parameters as well as variables that are controlled by the operator such as proppant volume and flowback methods."¹ 

"Evaluation was conducted on 195 stages of which 49 were identified as candidates for increase in proppant volume. Through this process the authors identified a need to update the model to include the impact of pressure depletion from down spacing. Even in the absence of accounting for pressure depletion the team experienced excellent results."¹ 

The goal is to design as efficient a completions strategy as possible across the anticline as more wellbores are drilled. Capturing the knowledge garnered from the geological parameters to maximize reservoir contact and the proppant volumes deemed appropriate across each stage of the wellbore, it is feasible to implement a function that identifies the values of operational parameters to maximize production. We can also identify which stages are most productive and shut down those stages that are under performers to reduce operational expenditure.