Applications of Sensitivity Analysis in Petroleum Engineering

Date

2007-05

Authors

Lawal, Azeez Adeyinka

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

A crucial question that may be asked during exploratory reservoir analyses and data gathering is: "What is the relative significance of different reservoir parameters?" A parameter is significant if the knowledge of its exact value results in an appreciable reduction in the uncertainty of model estimates. Sensitivity Analysis (SA) quantifies and apportions the uncertainty in a model's estimates based on the uncertainty in the model's parameters. Thus, SA can be used to determine the relative significance of reservoir parameters. This report starts with the SA of four models using differential analysis, regression, correlation and variance-decomposition methods. The models are the bi-linear, polynomial, Stock Tank Oil Originally in Place, and Panda and Lake equations. These models demonstrate the relative merits of the SA methods. The report culminates in the development of a variance-based Monte Carlo (MC) tank model that is useful in the SA of reservoir production rate, pressure and recovery forecasts. The model assumes the reservoir is homogeneous and undersaturated, and that the reservoir is produced by primary depletion mechanisms. It provides a choice between Latin Hypercube and Sobol sampling. These sampling methods are more efficient than conventional MC or random sampling. In the cases considered, the sensitivity effects of different reservoir parameters vary over the producing period. Parameters with negligible main effects can have significant joint effects.

Description

LCSH Subject Headings

Citation