Whitmire, Ryan Daniel (2018-05). Applying Uncertainty Quantification and Value-of-Information Concepts in Unconventional Reservoir Development. Master's Thesis. Thesis uri icon

abstract

  • The oil and gas industry has a long history of underperformance relative to forecasts. Underperformance in the industry has been directly linked to poor assessment of uncertainty. Uncertainty is large in the context of unconventional reservoir development. Therefore, reliable assessment of uncertainty is necessary for the optimization of decision making in unconventional reservoir development. Once uncertainty has been reliably assessed, the financial benefit of reducing each uncertainty should be estimated. Not all uncertainties are worth reducing; in fact, the value driven by the reduction of some uncertainties may be less than the cost of acquiring the relevant data - meaning that the data acquisition hurts financial performance. Yet, a well-established method of quantifying financial support for data-acquisition decisions in a multiple-variable context is largely absent from the literature related to unconventional reservoir development. Value-of-information analysis quantifies the financial benefit of reducing the uncertainty of variables within specific decision contexts. In this work, multi-variable value-of-perfect-information analysis was applied to a well-spacing decision model in the context of unconventional reservoir development. The application of multi-variable value-of-perfect-information analysis to an Eagle Ford well-spacing decision context indicated that the parameters for which uncertainty reduction would provide the most value are commodity price, created-fracture propagation, and matrix porosity. This analysis also indicated that reducing the uncertainty related to matrix permeability and natural fracture density would provide little value in the analyzed well-spacing decision context. The effect of biases in uncertainty quantification on multi-variable value-of-information calculations was investigated, and it was demonstrated that biased uncertainty assessment for one variable can skew value-of-perfect-information calculations for all uncertain variables and can change value-of-perfect-information rankings. A rational approach for data-acquisition decisions is achievable through creation of a reliable decision model and multi-variable value-of-information analysis. Widespread awareness of the power of multiple-variable value-of-information analysis to justify data acquisition and focus research efforts could lead to increased application of value-of-information analysis. Increased application should lead to improved decision making and financial performance in unconventional reservoir development.
  • The oil and gas industry has a long history of underperformance relative to forecasts. Underperformance in the industry has been directly linked to poor assessment of uncertainty. Uncertainty is large in the context of unconventional reservoir development. Therefore, reliable assessment of uncertainty is necessary for the optimization of decision making in unconventional reservoir development. Once uncertainty has been reliably assessed, the financial benefit of reducing each uncertainty should be estimated. Not all uncertainties are worth reducing; in fact, the value driven by the reduction of some uncertainties may be less than the cost of acquiring the relevant data - meaning that the data acquisition hurts financial performance. Yet, a well-established method of quantifying financial support for data-acquisition decisions in a multiple-variable context is largely absent from the literature related to unconventional reservoir development. Value-of-information analysis quantifies the financial benefit of reducing the uncertainty of variables within specific decision contexts. In this work, multi-variable value-of-perfect-information analysis was applied to a well-spacing decision model in the context of unconventional reservoir development.

    The application of multi-variable value-of-perfect-information analysis to an Eagle Ford well-spacing decision context indicated that the parameters for which uncertainty reduction would provide the most value are commodity price, created-fracture propagation, and matrix porosity. This analysis also indicated that reducing the uncertainty related to matrix permeability and natural fracture density would provide little value in the analyzed well-spacing decision context.

    The effect of biases in uncertainty quantification on multi-variable value-of-information calculations was investigated, and it was demonstrated that biased uncertainty assessment for one variable can skew value-of-perfect-information calculations for all uncertain variables and can change value-of-perfect-information rankings.

    A rational approach for data-acquisition decisions is achievable through creation of a reliable decision model and multi-variable value-of-information analysis. Widespread awareness of the power of multiple-variable value-of-information analysis to justify data acquisition and focus research efforts could lead to increased application of value-of-information analysis. Increased application should lead to improved decision making and financial performance in unconventional reservoir development.

ETD Chair

publication date

  • May 2018