Journal of Petroleum Science and Engineering 195 (2020) 107847 Available online 11 September 2020 0920-4105/© 2020 Elsevier B.V. All rights reserved. Shale rock core analysis using NMR: Effect of bitumen and water content Kaishuo Yang a , Paul R.J. Connolly a , Ming Li a , Scott J. Seltzer b , Douglas K. McCarty b , Mohamed Mahmoud c , Ammar El-Husseiny c , Eric F. May a , Michael L. Johns a, * a School of Engineering, Faculty of Engineering and Mathematical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, WA, 6009, Australia b Chevron Energy Technology Company, Houston, TX, USA c College of Petroleum Engineering and Geosciences, King Fahd University of Petroleum and Minerals, Dhahran, 31261, Saudi Arabia A R T I C L E INFO Keywords: Nuclear magnetic resonance Shale Total organic content Free water content ABSTRACT Hydrocarbon production from unconventional shale reservoirs has gained increasing worldwide focus, spurred largely by improvements in production techniques. Nuclear magnetic resonance (NMR) is increasingly being explored as a means of characterizing such shale rock. Compared to conventional sandstone and carbonate rocks, shales present a much lower porosity and permeability with NMR signal arising from both producible hydro- carbon fuid content as well as semi-solid organic solids and bitumen. Interpretation of the resultant NMR signal from shales is consequently comparatively more complex. To this end we apply combined NMR free induction decay (FID) and Carr-Purcell-Meilboom-Gill (CPMG) measurements and interpret the spliced data using a combined Gaussian and exponential inversion method. This was applied to a range of shale samples with variable moisture content in an attempt to assess the accuracy of such an approach to delineate the inherent predomi- nantly bitumen signal from that of the added water. The analysis was performed at a range of frequencies (20, 40 and 60 MHz), which are applicable to more widely accessible bench-top NMR spectrometers. Consistently, the Gaussian component of the resultant T 2 *|T 2 distribution was found to be independent of moisture content, scaling rather with the total organic content of the respective shale cores. In contrast, the exponential component of the resultant T 2 *|T 2 distribution was found to scale linearly with the moisture content of the cores. 1. Introduction Unconventional reservoirs, predominantly gas and tight oil shales, have gained increasing importance for the world’s energy supplies due primarily to reducing production costs and their potentially much greater hydrocarbon reserves relative to conventional reservoirs (Ratner and Tiemann, 2015; Jia, 2017). Natural gas production from tight rocks/shale gas now contributes in excess of 60% of total US natural gas production and is expected to increase by ~25 trillion cubic feet over the next 30 years to constitute over 90% of total US natural gas production in 2050 (U.S. Energy Information Administration, 2019). Accurate characterization of shales is, however, critical to enable more predictive estimates of both reserves and productivity (Wang et al., 2012; Song and Kausik, 2019; Nikolaev and Kazak, 2019). Unlike conventional hydro- carbon reservoirs, shales feature both comparatively low porosity and permeability, a signifcant solid organic content and complex nano- structures. Consequently, shale characterization is particularly challenging. Nuclear magnetic resonance (NMR) has become a powerful tech- nique to characterize the petro-physical properties of rock due mainly to non-invasive detection and lithology-independent porosity estimation (e.g. Logan et al., 1998; Coates et al., 1999; Kleinberg, 2001). Such NMR applications include estimation of pore size distributions (e.g. Davies and Packer, 1990; Davies et al., 1990; Arns, 2004; Mitchell et al., 2013), identifcation of pore fuids (e.g. Zhao et al., 2011; Hürlimann, 2012; Yuan et al., 2018), determination of wettability (e.g. Odusina et al., 2011; Sulucarnain et al., 2012; Yan et al., 2019) and estimation of permeability (e.g. AlGhamdi et al., 2013). For conventional hydrocar- bon rocks such as sandstone with negligible detectable hydrogen in the ‘solid’ matrix, the 1 H NMR signal originates from pore space fuids, enabling rock porosity to be accurately determined. However, shales are generally very heterogeneous with both signifcant organic content and nano-scale to micro-scale pores in both the organic and inorganic phases (e.g. Minh et al., 2012; Chalmers et al., 2012; Josh et al., 2012). This organic content is composed of both kerogen and bitumen (which are broadly defned as being insoluble and soluble in organic solvents * Corresponding author. E-mail address: michael.johns@uwa.edu.au (M.L. Johns). Contents lists available at ScienceDirect Journal of Petroleum Science and Engineering journal homepage: http://www.elsevier.com/locate/petrol https://doi.org/10.1016/j.petrol.2020.107847 Received 12 May 2020; Received in revised form 21 August 2020; Accepted 26 August 2020