Organic petrology and geochemistry of the Devonian-Mississippian bakken formation, Williston Basin, North Dakota

Abstract

The Devonian–Mississippian (D–M) black shales of the Bakken Formation are of interest as a hydrocarbon source due to their high total organic carbon (TOC; 2.2–17.4%) content. The Upper and Lower Members of the Bakken Fm. are shallow marine (100–150 m) sequences. Thirty samples were selected for maceral identification, kerogen typing, and solid bitumen reflectance (SBR_o) based on TOC content and down-core spacing. The shales contain alginite, bituminite, abundant solid bitumen (SB), and minor amounts of inertinite. Solid bitumen increases in quantity with increasing thermal maturity. Pyrolysis (85 samples) provided S1 (avg. 8.0 mg HC/g rock), S2 (avg. 24.3 mg HC/g rock), hydrogen index (HI; avg. 201 mg HC/g TOC), oxygen index (OI; avg. 7 mg CO₂/g TOC), and R_o (0.60–1.03%) calculated from T_max. Plots of HI vs. OI and HI vs. T_max (°C) were used to assess kerogen type but are not consistently in agreement with the petrographic assessment. Some samples from more thermally mature cores plot as Type III (vitrinite) kerogen instead of Type II (alginite and bituminite) kerogen, the latter confirmed through petrographic observations of lower maturation samples. This is largely due to increased SB in more thermally mature samples (R_o = 0.83–1.03%), as SB is known to have a lower HI content than Type II kerogen. Petrographic data show more alginite and bituminite (19–55%) in the thermally less mature samples (R_o = 0.60–0.83%) compared to more dispersed SB (67–86%) and less alginite and bituminite (<1%) in the more thermally mature samples (R_o = 0.89–1.01%).

Early research on the Bakken Fm. reported lower than expected vitrinite reflectance values and attributed them to vitrinite “suppression”. The scarcity of vitrinite and abundance of SB suggest that early work likely reported reflectance on SB. Recent attempts to assess the thermal maturity of the Bakken black shales have converted SBR_o to vitrinite reflectance equivalence (V_RE). However, there are multiple SB populations present in these shales and it is not always clear which SB populations were included, possibly contributing to error. In the current study, only smooth, homogenous SB was measured (0.68–1.14% SBR_o) and V_RE values calculated (0.54–1.49%) to assess thermal maturity from the basin margin to the depocenter; inclusion of measurements on granular, heterogeneous SB (14–21 vol%), which are ∼53% lower than those for smooth, homogenous SB (3–12 vol%), results in lower mean reflectances, especially in more mature samples. Vitrinite reflectance equivalent data calculated using the D–M New Albany Shale equation of Liu et al. (2019) agrees with liptinite fluorescence and Rock-Eval R_o, whereas V_RE based on the D–M Woodford Shale equation of Cardott and Comer (2021) does not. This suggests the importance of applying V_RE equations from similar formations both in terms of thermal history, as well as kerogen type and age. Results from SBR_o, Rock-Eval R_o, V_RE, and observations of alginite fluorescence indicate that samples from the current study range from the early oil window into the condensate, wet gas zone.