Methods for adaptive optimization of enhanced oil recovery performance under uncertainty

A method for performing an enhanced oil recovery (EOR) project under uncertainty, the method comprising:

• utilizing a predictive physics-based reservoir model to estimate the performance of the EOR project;
• identifying a plurality of input parameters into predictive physics-based reservoir the model as control variables and uncertain variables;
• optimizing the predictive physics-based reservoir model to obtain values of control variables maximizing a mean value of a chosen performance metric under initial uncertainty of formation and fluid properties;
• applying a global sensitivity analysis to quantify and rank contributions from uncertain input parameters to the standard deviation of the optimized values of the performance metric, wherein the sensitivity analysis comprises a first order sensitivity index;
• obtaining additional information using a wellbore measurement tool regarding at least one high-ranking of the uncertain variables in order to reduce uncertainty therein;
• optimizing the predictive physics-based reservoir model with at least one uncertain parameter having reduced uncertainty, to obtain a set of updated set of optimized values of the control variables;
• using the model to provide a performance analysis for the EOR project;
• and
• performing an EOR project on a formation according to the updated set of optimized values of the control variables, wherein performing the EOR project includes altering an operation of a production well or an injection well based upon, at least in part, the updated set of optimized values of the control variables.

A system for adaptive optimization of an enhanced oil recovery (EOR) project for a formation containing fluids, the system comprising:

• a wireline logging system measurement tool configured to perform a measurement on at least one of the formation and a sample from the formation;
• and
• a processing system configured to:
• (i) define a performance metric, a plurality of control variables, and a plurality of uncertain variables of the EOR project;
• (ii) apply the plurality of control variables and the plurality of uncertain variables as input parameters into a predictive physics-based reservoir model;
• (iii) optimize the performance metric under initial uncertainty of the plurality of uncertain variables to obtain a set of optimized values of the control variables that provide an initial optimum of an objective function based on the performance metric;
• (iv) conduct a global sensitivity analysis for the plurality of uncertain variables using the set of optimized values of the control variables, wherein the sensitivity analysis comprises a first order sensitivity index;
• (v) reduce uncertainty in at least one of the plurality of uncertain variables using measurements performed by the wireline logging system measurement tool;
• (vi) optimize the performance metric with reduced uncertainty in the at least one of the plurality of uncertain variables to obtain an updated set of optimized values of the control variables that provide an updated optimum of the objective function based on the performance metric;
• (vii) running the reservoir model with the updated set of optimized values of the control variables in order to determine performance of the EOR project;
• and wherein the wireline logging system measurement tool is used, in part, to perform the EOR project on a formation according to the updated set of optimized values of the control variables,
• wherein performing the EOR project includes altering an operation of a production well or an injection well based upon, at least in part, the updated set of optimized values of the control variables.