Spatial Measures for Quantifying Reservoir Connectivity
Open Access
- Author:
- Hoydic, Corey James
- Area of Honors:
- Petroleum and Natural Gas Engineering
- Degree:
- Bachelor of Science
- Document Type:
- Thesis
- Thesis Supervisors:
- Dr. Sanjay Srinivasan, Thesis Supervisor
Dr. Russell Taylor Johns, Thesis Honors Advisor - Keywords:
- Petroleum Engineering
Geostatistics
Reservoir
Image Processing
Petroleum and Natural Gas Engineering - Abstract:
- A suite of statistical tools was developed with the purpose of representing the heterogeneous characteristics of geologic depositional systems. Employing a set of both scanned and synthetic images, data representing spatial heterogeneity were represented through an indicator code corresponding to the color index value at each pixel location in an image. In establishing these color indicators, basic statistical tools were used to establish thresholds that would yield indicators that preserve the connectivity and key characteristics of the images. Multivariate spatial statistics were employed to quantify connectivity in different directions of varying complexity in images, as well as to quantify and optimize thresholds in order to create a thresholded image that is most representative of the original data set. These tools were employed along with an algorithm to smooth noise in the data that arises from scanning the image, yielding a set of spatial connectivity measures that can quantify how well key features are preserved in an image and is able to quantitatively distinguish between different images. These tools will, then, be applied in future studies to create a reservoir indexing scheme that categorizes different reservoir models by their key features.