Fracture attribute scaling and connectivity in the Devonian Orcadian Basin with implications for geologically equivalent sub-surface fractured reservoirs

Abstract

Abstract. Fracture attribute scaling and connectivity datasets from analogue systems are widely used to inform sub-surface fractured reservoir models in a range of geological settings. However, significant uncertainties are associated with the determination of reliable scaling parameters in surface outcrops. This has limited our ability to upscale key parameters that control fluid flow at reservoir to basin scales. In this study, we present nine 1D-transect (scanline) fault and fracture attribute datasets from Middle Devonian sandstones in Caithness (Scotland) that are used as an onshore analogue for nearby sub-surface reservoirs such as the Clair field, west of Shetland. By taking account of truncation and censoring effects in individual datasets, our multiscale analysis shows a preference for power-law scaling of fracture length over 8 orders of magnitude (10−4 to 104 m) and kinematic aperture over 4 orders of magnitude (10−6 to 10−2 m). Our assessment of the spatial organization (clustering and topology) provides a new basis for up-scaling fracture attributes collected in outcrop- to regional-scale analogues. We show how these relationships may inform knowledge of geologically equivalent sub-surface fractured reservoirs.

Multiscale lineament analysis and permeability heterogeneity of fractured crystalline basement blocks

“…These parameters are usually adopted for the quantification of the spatial occupancy and of the "fractal" character of fracturelineament networks (fractal dimension D and length distribution scaling laws), to constrain their physical connectivity and spatial organization (spacing and length) and, ultimately, for the quantification of the permeability structure of the host fractured medium (Bonnet et al, 2001;Healy et al, 2017;Nyberg et al, 2018;Peacock and Sanderson, 2018). The statistical tests adopted to constrain the most representative fitting curves and distributions include (i) leastsquare regression (LSR) and maximum likelihood estimation (MLE) coupled with Kolmogorov-Smirnov (KS) statistical tests adopted on cumulative distributions (Dichiarante et al, 2020;Kolyukhin and Torabi, 2013;Rizzo et al, 2017) and (ii) bivariate box-and-whisker plots to evaluate the distribution of fracture spacing heterogeneity parameters and their statistical significance.…”

“…The implications of the adoption of general scaling laws on the upscaling and/or downscaling of fracture network properties, as well as the possible analytical biases and sources of errors in the analytical approach, are then evaluated and discussed. In this work, rather than trying to accurately quantify the scaling parameters, we focused on highlighting and analyzing the weaknesses and uncertainties that invariably accompany this sort of lineament analysis even when very robust statistical approaches and analytical methods are applied (Dichiarante et al, 2020). By integrating this information with existing field structural analyses and modeling of lineament petrophysical properties (Ceccato et al, 2021b, a;Scheiber and Viola, 2018), we provide further constraints on the multi-scale heterogeneity in magnitude, orientation, and spatial distribution of the permeability structure of the studied fractured crystalline basement.…”

“…One way to obtain quantitative constraints upon the scale dependency of fracture and fault network attributes is to perform a multiscale analysis of, for example, their length and spacing distributions. The aim of these multiscale analyses is to obtain scaling laws that can quantify the variability of fracture network properties across scales (Bonnet et al, 2001;Bossennec et al, 2021;Castaing et al, 1996;Chabani et al, 2021;Dichiarante et al, 2020;Gillespie et al, 1993;McCaffrey et al, 2020;Odling, 1997).…”

“…Furthermore, lineament maps provide very large and statistically robust datasets, which are, however, subject to analytical, methodological, and interpretative biases Scheiber et al, 2015). Therefore, accurate statistical analyses need to be performed to evaluate the possible biases affecting each dataset and the extrapolation limits of the observations retrieved from their analyses (Bistacchi et al, 2020;Dichiarante et al, 2020;.…”

Fluvial sedimentation and its reservoir potential at foreland basin margins: A case study of the Puig-reig anticline (South-eastern Pyrenees)

“…Reservoir quality in such settings is mainly controlled by their sedimentary characteristics, e.g., petrographic characteristics and stratigraphic sequence, as well as the diagenesis and fracturing processes that they experience (e.g., Morad et al, 2010;Taylor et al, 2010;Zhang et al, 2011;Watkins et al, 2018). Due to the limited well data and seismic resolution of subsurface reservoirs, outcrop analogues play an important role in improving the accuracy of reservoir prediction in the subsurface (e.g., Howell et al, 2014), by providing reliable geological conceptual models and quantitative attribute information (Dichiarante et al, 2020). It is therefore essential to systematically study outcropping reservoir analogues to fully understand and predict the distribution and reservoir potential of alluvial-fluvial deposits in the subsurface.…”