We used multiple mathematical techniques to develop primary ultimate oil recovery (PRUR), initial waterflooding ultimate oil recovery (IWUR), and infill drilling ultimate oil recovery (IDUR) models for carbonate reservoirs in West Texas. These are (1) non-linear regresion, (2) non-parametric regression (a statistical approach that constructs the functional relationship between dependent and independent variables, without bias towards a particular model), and (3) neural network models. Development of the oil recovery forecast models help understand the relative importance of dominant reservoir characteristics and operational variables, reproduce recoveries for units included in the database, and forecast recoveries for possible new units in similar geological setting.

One of the challenges in this research was to identify the dominant and the optimum number of independent variables. The variables is large including porosity, permeability, water saturation, depth, area, net thickness, gross thicknees, formation volume factor, pressure, viscosity, API gravity, number of wells in initial waterflooding, number of wells for primary recovery, number of infill wells over the initial waterflooding, PRUR, IWUR, and IDUR. The limited number of field data (43 data) points) are inexact and often exhibit uncertain relationships. We used an intelligence software1  that integrates multivariate statistical and neural networks to develop accurate neural network models. Multivariate principal component analysis is used to identify the dominant and the optimum number of independent variables. We compared the results from neural network models with the non-parametric approach. The advantage of the non-parametric regression is easy to use and the disadvantage is retaining a large variance of forecast results for a particular data set.

Develop a neural network model that is an "accurate" representation of data may be and ardous task that require experience with the qualitative effects of the structural parameters of neural network models. The advantage of neural network PRUR, IWUR, and IDUR models are capable of forecasting the oil recovery with less error variance.

You can access this article if you purchase or spend a download.