Date of Award

January 2022

Document Type


Degree Name

Doctor of Philosophy (PhD)


Petroleum Engineering

First Advisor

Vamegh Rasouli


Hydrocarbon production from tight oil and/or shale plays requires drilling horizontal wells with very long laterals and hydraulic fracturing operations to stimulate the reservoirs which have generally very low porosity and permeability. A great amount of experience has been accumulated over the past two decades in the North America from production of these types of reservoirs. The Bakken formation in North Dakota, is one of the largest tight oil producers in the U.S.A. and currently ranks as the second largest producer of oil in the nation. The Bakken formation is divided into the upper and lower shale members, which are the source rocks. The middle Bakken which is a fractured silicious-carbonate formation is the main producer. The Bakken formation has a complex mineralogy, composed of mainly quartz and clay in the two shale members and quartz, calcite, dolomite and anhydrite in the middle member, plus several other minerals with lower volume fractions. Past studies have reported the importance of mineral composition in development, completion and production from shales. This research study, in line with the Bakken Production Optimization Program, being led by the Energy & Environmental Research Center (EERC), aims at studying the effect of mineral composition of the three Bakken members on rock elastic and velocity properties. These properties are the main input to drilling, completion and hydraulic fracturing design and production optimization and planning. Rock physics models were developed at macro (core) and log scales from data obtained from Dunn and Mountrail Counties in North Dakota. The minerals’ volume fractions from core XRD data and elemental capture spectroscopy (ECS) log data were used for estimation of rock stiffness and velocities. The calculations were perfomred for the upper and lower Bakken (UB, LB) shales as well as the clastic and carbonate intervals of the middle Bakken (MB). Hill averaging was used to mix the non-clay members. Clays were added using both Kuster-Toksoz (K-T) model and differential effective medium (DEM) theory. The effect of fluid phase (assuming both homogeneous and patchy fluids) was applied using K-T and the Gassmann equation. For shales, at core scale, Backus averaging captured the upper and lower bounds of data, whereas, the K-T model was developed for data parallel and perpendicular to laminations. Collectively, their use allowed for accurate curve fitting to the lab data, by changing the pore’s aspect ratios. While the core scale studies were based on data from certain depths, the log-based modeling estimated the effective stiffness properties as continuous logs. Overall, it was observed that, the UB and LB shales, due to the presence of clay minerals, show lower properties than the MB and composed mainly of microfracture porosity. The MB, is characterized by more intraparticle pores with stiffer properties. The rock physics models at macro and log scale complemented each other and showed reasonable agreements. The estimated stiffness and velocity properties are valuable input data for the development of and production from future wells.

Available for download on Sunday, August 25, 2024