Optimization Strategies for Shale Gas Asset Development

Advisor

Louis Durlofsky

Abstract

With the recent boom in US shale gas production, optimal development of these as-
sets has become a topic of significant interest. In addition to the complex physics
typically associated with shale plays, field development optimization can be a chal-
lenging problem itself when binary, integer and continuous variables are concurrently
present. For complex problems of this nature, it is appropriate to use simulation-based
optimization to search the solution space.

Prior work has shown that shale gas reservoirs can be simplified, through applica-
tion of a history-matching-like tuning procedure, from a locally refined, dual-porosity,
dual-permeability model that considers desorption and non-Darcy effects, to a surro-
gate model without desorption, non-Darcy corrections or refined grids. This surrogate
model is simply a single-porosity, single-permeability model with tuned parameters
in the stimulated reservoir volume (SRV). The two tuning parameters, SRV porosity
and permeability multipliers, are determined through a history matching process that
minimizes the dierence in production between the full and surrogate models. With
an appropriately tuned surrogate model, optimization function evaluations can be
performed at a low computational cost. Here we utilize PSO-MADS (particle swarm
optimization - mesh adaptive direct search), a hybrid global-local optimization algo-
rithm, to find the optimal tuning parameters for the surrogate model and to find an
optimal field development plan. During the asset development optimization we con-
sider five decision variables per well: well location, lateral length, number of fracture
stages, bottomhole pressure, and finally whether or not to drill that well at all.
In this work, we integrate PSO-MADS, coupled-geomechanics, and new decision
variables into the workflow. The results demonstrate the efficiency and utility of using
proxy-based optimization for shale gas asset development. The results also indicate
that geomechanics can have an effect on the optimal development plan and should
be considered during optimization. The overall optimization framework developed in
this study should be applicable for a wide range of shale development projects.