Ailin Assady

Date of Award

January 2022

Document Type


Degree Name

Doctor of Philosophy (PhD)


Petroleum Engineering

First Advisor

Vamegh V. Rasouli


Effective stress exerted on porous rocks can change and alter reservoir permeability accordingly during reservoir development. The permeability evolution under different reservoir statues will impact oil production and EOR design in the Bakken shale porous media. An accurate permeability model can improve capturing the fluid transport mechanism and create a reliable long-term dynamic fluid forecast via reservoir simulation. This research is focused on studying permeability alteration behavior under different pressure circumstances. The reservoir gradually loses its original pore pressure during production, increasing reservoir net effective stress. Therefore, a reduction in reservoir properties such as permeability or porosity can occur in response to net stress change within the pores due to the withdrawal of the fluids from the reservoir. In contrast, a fluid injection can reduce formation pressure drop and maintain pressure during the development process in tight rock reservoirs. However, physical parameters (e.g., permeability) cannot be fully recovered, and back to its initial value, this nature of rock is characterized as stress sensitivity or hysteresis. Stress-dependent properties are hard to model accurately in reservoir simulation because of the uncertainty associated with the stress-dependent coefficients and correlations. The conventional reservoir simulators use the compressibility concept to consider the change of pore volume, where the rock properties are usually assumed to be insensitive to the evolution of the stress state. However, reservoir compaction and stress changes can significantly impact reservoir management and production performance. In this study, a review of different rock characterizations of the Three forks and Bakken core samples to determine stress dependency of permeability and its hysteresis during pressurizing/ depressurizing rock samples is conducted. Core samples from the Middle Bakken formation in North Dakota for further permeability alteration experiments are utilized. This data will be used to evaluate the permeability behavior with respect to critical pressure known as pressure shock. Also, the data analytic approach to model permeability on a larger scale based on several inputs such as depth, different net confining stress, and porosity is performed. Numerical reservoir simulation using Bakken and Three Forks formation is utilized to integrate permeability pressure correlation in simulation modeling and compare several injection scenarios with non-sensitive permeability models. The results indicate that ignoring the effect of slope discontinuity at a critical effective stress using the same equation for a whole range of data is inaccurate. Indeed, developing permeability-stress correlations cause inapplicable mathematical models and, consequently, erroneous permeability damage prediction. Following this concept, modifying the correlation for two Bakken cores shows that considering the critical points on each hysteresis path could improve the final form of the stress-dependent permeability relationship. Also, machine learning modeling using available lab core data can be used as an alternative method to capture Bakken and Three Forks permeability changes under different net confining stress while incorporating the critical pressure effect. Furthermore, to evaluate the several gas injection scenarios, the timely reservoir pressure change is divided into three distinct regions where critical effective pressure impact and miscibility of gas injection vary based on current reservoir statutes. The results demonstrate that gas injection in these formations is a strong function of fracture/matrix permeability damage. Compared to the model without considering stress-dependent permeability, the cumulative production could reduce because the permeability decreases along with reservoir pressure decline. As a result, considering permeability modeling in numerical simulation can help to understand the role of different injection scenarios and enhance the knowledge for controlling and managing reservoir production by proper operation decisions in unconventional reservoirs.