Recap and Integrated Rock Mechanics and Natural Fracture Study on the Bakken Formation, Williston Basin Cosima Theloy, Department of Geology & Geological Engineering Abstract The late Devonian to early Mississippian Bakken Formation in the Williston Basin represents one of the most prolific tight oil plays worldwide. Although the Bakken play is characterized as an unconventional petroleum system, it is not a continuous petroleum accumulation, at least in the economic sense. There are dry wells and sweet spot areas, and it is not yet well understood what are the causes for observed variations in productivity. This research project aims to achieve a better understanding of factors influencing productivity, including natural fracture permeability. In order to gain a measure on whether or not natural fractures have a significant impact on economic production rates in the Bakken play an integrated multi-disciplinary approach was designed. Figure 1 displays the four main components, which may be involved in natural fracture formation, such as hydrocarbon generation-induced fractures created by extremely high pore pressure gradients, variations in rock-mechanical properties and mineralogical and textural characteristics of the reservoir facies, as well as regional-scale fractures in association with structural elements and stress regime. By correlating those potential fracture formation mechanisms with productivity maps (Figure 2), conclusions can be drawn upon whether areas with fracture-enhanced permeability coincide with identified sweet spot areas. However, also the effect of technological advancements in completion design of wells must be taken into account to prevent a skewed perspective on the dataset. Figure 1: Flowchart illustrating factors potentially influencing natural fracture formation (top), while the effect of natural fractures on production and completion design (bottom) can be reversed and used as backtracking tool to infer on the presence of natural fractures.
Figure 2: Initial production (bbl/day) map of all available Bakken producers overlain with structure contours of the Bakken Formation (top), and initial production maps in three time slices (bottom). Same color scale was used for all maps!
The first part of the presented work will highlight findings and results from completed research, while during the second part the rock-mechanical characteristics of the reservoir facies will be discussed. Figure 2 displays initial production maps, including three time slices to illustrate the areal development of the Bakken play and increasing production rates through time as horizontal multi-fracturing stage completion techniques were implemented. These maps served as basis to identify highly productive areas for both the Bakken and the Three Forks formations, which were then correlated with geochemical parameters, pore-overpressure distribution, drilling mud weights, structural elements and lineaments, and numerous other parameters possibly influencing productivity. Already Meissner (1978a) suggested a connection between hydrocarbon generation and pore-overpressure as result of volume expansion during the decomposition of kerogen to oil and gas. Meissner s (1978a) pore-pressure contours were based on five data points and any basinwide correlation based on such a small dataset leaves room for improvement. Hence, an effort was made to create a new pressure map and filtering out unreliable measurements or data points of unknown origin. From an original dataset of 426 pressure measurements only 80 data points, for which the build-up curves suggested good quality measurements, were selected. The new pressure map shows relatively smooth contours and is in good correlation with the combined original TOC contents of the Lower and Upper Bakken shales (Figure 3). The highest accumulations of organic matter coincide with the highest pressure gradients, as well as the bulk of the highly productive areas. In the Sanish / Parshall area, however, the original TOC contents are quite low, suggesting that other factors than hydrocarbon generation-induced overpressuring and fracturing play a role and that this concept is too simplistic. The current working hypothesis is that migration and, in particular, trapping mechanisms play a key role in understanding the occurrence of sweet spots, and in-detail geochemical analysis of oil families are needed to corroborate or disprove this hypothesis. Earlier research results have shown that the facies of the Middle Bakken are laterally continuous, but may possess distinctive facies-internal differences in terms of mineralogical composition. A new rock-mechanical dataset was provided by the Energy and Environmental Research Center (Darren Schmidt, EERC) aiming to investigate the impact of changes in facies, mineralogical composition, and texture on rock-mechanical properties. A total of 48 samples from North Dakota have been analyzed from which 28 samples came from Middle Bakken cores and 20 samples derived from the Three Forks. Three out of 20 wells have been sampled throughout the Middle Bakken and the Three Forks reservoirs (7 to 8 samples per well). Poisson s ratio versus Young s Modulus plots indicate that there is no distinct grouping or clustering of data points with regard to reservoir facies or texture (Figure 4). Similar results were yielded for differences in mineralogical composition, presence / absence of visible fractures in cores in vicinity to the sample location, or increasing burial depths and compaction.
Figure 3: This map shows the distribution of back-calculated original TOC (wt. %) values as sum of both the Upper and Lower Bakken shales, and is overlain by pore-pressure gradient contours (psi/ft). Red outline encircles the bulk of highly productive areas in the McKenzie and Williams counties.
FACIES TEXTURE Figure 4: Variability of Young s Modulus and Poisson s Ratio with respect to reservoir facies and texture of the samples. References: Meissner, F. F., 1978a, Petroleum geology of the Bakken Formation Williston Basin, North Dakota and Montana, in D. Rehrig, ed., 1978 Williston Basin Symposium: Billings, Montana, Montana Geological Society, p. 207 227.