Semester of Graduation

2020 Fall

Degree

Master of Science in Petroleum Engineering (MSPE)

Department

Petroleum Engineering

Document Type

Thesis

Abstract

Geological folding and/or faulting may create fractured reservoirs containing a semi-parallel system of long, sparsely - spaced fracture corridors separated by exclusion zones. Presently, the best method for detecting and assessing fracture corridor networks requires drilling, logging, and coring from horizontal wells. However, there is no method to evaluate such reservoirs by pressure testing vertical wells. Vertical wells - completed either in highly-conductive corridors (fracture wells) or in the exclusion zone (matrix wells) would respond quite differently to well testing. Therefore, their pressure response patterns can be used to identify well's placement in the corridor system and some other properties such as permeability of the exclusion zone, for example. (The actual permeability of the exclusion zone - due to diffuse fractures - is higher than the rock matrix permeability measured on the core samples.) The objective of this study is to apply the well flow testing analysis technique to estimate well’s location, permeability of the exclusion zone, distance from well to fracture corridor, corridor length, and conductivity.

In this study, pattern recognition technique is used to analyze diagnostic plots of pressure drawdown generated by simulated flow tests with commercial software (CMG). A unique simulation model has been built by combining a local model of fracture well or matrix well with adjacent fracture corridor and a "homogenized" global model of the remaining corridor network. The global model generalizes the corridor network using single-porosity and radial permeability approach. The approach is verified as being sufficiently accurate.

The results show that diagnostic plots of bottom hole pressure response to constant production rate for the matrix and fracture wells clearly indicate the well's location as the plot patterns are quite different. Moreover, for matrix well (completed outside the fracture corridor) permeability of the exclusion zone and well-to-corridor distance can be determined from the initial radial flow regime after removing the wellbore storage effect by β-deconvolution. It is also shown that for fracture well (intercepting fracture corridor) diagnostic plot of the bilinear flow regime provides data for finding the fracture corridor conductivity and fracture corridor length. The corridor length, however, can be estimated with more precision from the pseudosteady-state flow regime plot representing reservoir boundary and reservoir shape factor effects. However, the approach is only practical for production rather than transient flow testing.

This study also employs statistics - the cumulative logit models - to qualify accuracy of two techniques: finding permeability of the exclusion zone and distance from the well to the nearest corridor. The results show that the more distant the well from the corridor and the lower the exclusion zone permeability the more accurate permeability estimation becomes. Also, accuracy of the well-corridor distance estimation improves for longer corridors and lower permeable exclusion zones.

Committee Chair

Wojtanowicz, Andrew

DOI

10.31390/gradschool_theses.5235

Download

Included in

Engineering Commons

Share

COinS