Shale, Clay and Thin Bed Reservoirs: Appropriate Characterization with the Single Model

Abstract

Abstract Thomas-Stieber (TS) model analysis (Thomas et.al, 1975) has been used to define thin bed properties using standard logs in multiple reservoirs. A key model assumption is a mixture of sand and shale formation components, and the latter has been particularly discussed on numerous occasions (Kennedy 2018, Juhasz 1986, Skelt 2014, Moss 2008). Net from Thomas-Stieber analysis models are routinely compared with core or other methods and has validated the use of Vshale in those instances. However, in the net sand fraction the presence of dispersed clay can have a dramatic impact on porosity and permeability, implying that Vclay is a preferred parameter to accurately estimate reservoir quality. Furthermore, Vclay is a quantitative core measurement allowing for a robust core-log calibration, supported by other parameters. This paper will investigate the key assumption of the TS model by directly incorporating clay volume data in the thin bed reservoir formation model. The presented modification accurately leverages the difference between Vclay and Vshale parameters integrating them in a single solution concept. Extended TS formulae (defined as the Thomas-Stieber-Tyurin or T-S-T model) can more accurately assess the impact of dispersed clay presence on thin bed reservoir quality and at the same time treat NTG as a binary parameter, e.g. shale or non-shale laminae. An updated formation hierarchy scheme is presented to accurately reflect the occurrence of shale, sand, and dispersed clay mixes in a formation. Geology-wise we demonstrate the applicability of a single model for sand-shale-silt laminations, further highlighted in the Net-NonNet vs Sand-NonSand fractions difference discussion. A multi-parameter model for permeability that includes dispersed Vclay is also presented. Partial derivative analysis, presented in this paper, highlights the sensitivity and irreducible uncertainty associated with T-S-T model and demonstrates the importance of in-situ corroboration with independent methods to provide robust outputs. Stochastic simulation of the T-S-T model show the uncertainty range of the analysis and alignment with core and NMR to validate the model results. The resulting analysis provides a robust framework to define volumetric ranges, reservoir quality in thin beds, key priorities for future data acquisition and support business critical decisions. We believe that presented workflow addresses continuous conversation in the petrophysical community regarding limitations and applicability of the TS model.