Domainal Fabric

Key Takeaways

• In geostatistics, the nested variogram model is the quantitative expression of domainal fabric in spatial data: the experimental variogram (the average squared difference between property values at all pairs of data points separated by a given lag distance h) of a domainal fabric property typically shows multiple inflection points or nested structures at different lag distances, each inflection corresponding to the range of spatial correlation for one domain level; a nested variogram model fits this experimental variogram as the sum of individual variogram structures: gamma(h) = nugget + C1 * variogram_model1(h, a1) + C2 * variogram_model2(h, a2) + ..., where each term represents one scale of the domainal fabric with its own partial sill C and range a; the nugget represents the proportion of variance occurring at scales smaller than the sampling interval (micro-scale heterogeneity or measurement error); fitting a nested variogram model to the experimental variogram of core porosity or permeability data from a naturally heterogeneous reservoir provides the spatial statistics needed to generate geologically realistic stochastic property models using sequential Gaussian simulation, multiple-point statistics, or object-based modeling methods that honor the domainal fabric structure of the natural rock.
• Carbonate reservoir domainal fabrics are particularly complex because carbonate sedimentology creates heterogeneity at multiple scales simultaneously: at the pore scale (micrometers), intergranular porosity, intragranular porosity, and vuggy porosity from dissolution create a domainal pore structure that controls permeability at the core plug scale; at the lamina to bed scale (millimeters to centimeters), alternating grainstone and packstone laminae, bioturbation burrows, and early diagenetic nodules create cm-scale domains of different porosity and permeability within a single facies; at the facies scale (meters to tens of meters), shoal bodies, tidal flat cycles, and reef buildups create meter-scale domains of different rock type and reservoir quality; at the sequence scale (tens to hundreds of meters), systems tracts and sequence boundaries create domains of different diagenetic overprint (freshwater diagenesis near subaerial exposure surfaces creating dissolution-enhanced porosity vs. cementation of marine-flooding surfaces); and at the regional scale (kilometers), paleogeographic position, structural tilt, and burial history create kilometer-scale domains of different thermal diagenesis and dolomitization; the practical challenge of modeling carbonate domainal fabrics is that each scale of heterogeneity requires separate characterization data (thin sections for pore scale, whole core for facies scale, logs and seismic for bed to sequence scale) and separate simulation methods.
• Structural domainal fabrics in deformed rocks manifest as alternating ductile and rigid domains that form during progressive deformation: in a rock undergoing folding or shearing, mineral grains with different mechanical properties (mica vs. quartz, for example) respond differently to the applied stress, with the softer minerals accommodating more strain by dynamic recrystallization, grain boundary sliding, and crystal plastic deformation while the harder minerals deform more rigidly; this differential response produces a fabric where mica-rich domains contain well-developed foliation (aligned mica crystals, preferred orientation of clays) alternating with quartz-feldspar-rich microlithon domains with less developed internal foliation, creating the microscale crenulation cleavage seen in many slate and phyllite samples; at a larger scale, the same principle of domainal strain partitioning creates pressure solution seam networks in calcareous sediments (carbonate-rich seams alternating with clay-rich residual domains from which the carbonate has been dissolved and reprecipitated), spaced cleavage in limestone (widely spaced pressure solution seams separated by less deformed inter-seam domains), and S-C fabrics in mylonites (alternating planar (C) shear bands and oblique (S) foliation domains that record the finite strain geometry of the shear zone).
• Reservoir simulation of domainal fabric heterogeneity requires either upscaling the fine-scale geological model to a coarser simulation grid (losing some of the fabric heterogeneity in the upscaling process) or using an explicit hierarchical simulation approach: for simple two-domain fabrics (for example, fractures and matrix in a dual-porosity model), the dual-porosity or dual-permeability simulation framework explicitly represents the two domains as separate continua exchanging fluid through a matrix-fracture transfer function; for more complex multi-domain fabrics (three or more scales of heterogeneity), stochastic realizations of the property model at each scale are generated and upscaled to the simulation grid using arithmetic, harmonic, or power-average upscaling methods that preserve the effective flow properties of the domainal fabric at the scale of the simulation cell; the quantification of how much heterogeneity is lost in the upscaling from the fine-scale domainal fabric model to the coarse simulation grid is a critical uncertainty analysis step that determines whether the simulation will accurately predict sweep efficiency, waterflood breakthrough timing, and recovery factor in the heterogeneous reservoir.
• Field-scale examples of domainal fabrics in petroleum reservoirs include the stylolite-separated flow units in Middle Eastern carbonate reservoirs (where pressure solution stylolites create permeability baffles that divide the reservoir into domainal flow units), the mud-drape-separated sand bodies in tidal deltaic reservoirs (where periodic tidal mud drapes create a domainal fabric of highly permeable sand alternating with nearly impermeable mud that channels flow in unexpected directions during waterflood), and the fracture swarm networks in chalk and tight carbonate reservoirs (where the fracture network has a domainal structure at multiple scales, from the micro-crack to the fault damage zone, that controls flow but is not visible at any single scale of observation); in all these cases, the geologist or reservoir engineer who recognizes the domainal nature of the fabric -- that the reservoir cannot be described by a single characteristic length scale and a single set of petrophysical properties -- is better positioned to design a development strategy that accounts for the multi-scale heterogeneity rather than treating it as homogeneous at the well spacing scale.