Sorting (Sedimentology and Reservoir Quality)

Sorting in sedimentology and reservoir engineering refers to the degree of uniformity in grain size within a sedimentary rock — a well-sorted sediment has grains of nearly uniform diameter (narrow size distribution), while a poorly sorted sediment contains grains ranging from fine clay particles to coarse pebbles (wide size distribution) — with sorting being a fundamental control on reservoir porosity and permeability because well-sorted sands have higher porosity (fewer fine grains filling pore spaces between larger grains) and higher permeability (more uniform pore throats) than poorly sorted sands of equivalent mean grain size, making sorting one of the key parameters in grain-size analysis, facies characterization, and reservoir quality prediction.

What Is Sorting in Sedimentology?

Sedimentary processes are selective — they transport, deposit, and modify sediment particles in ways that depend on the physical properties of both the particles (density, size, shape) and the transporting medium (water, wind, ice, gravity). A wind-blown sand dune accumulates grains in a narrow size range because wind of a given velocity can only lift and carry grains within a specific mass range — grains too heavy fall out immediately, grains too light are carried past the dune. This selectivity by the transporting process creates sorting: a measure of how narrow or wide the resulting grain size distribution is.

Sorting is fundamentally a record of the depositional history of the sediment. Repeated reworking by high-energy processes (waves, wind) progressively removes outlier sizes and narrows the distribution toward well-sorted. Single-event deposition without sorting (mass flows, glacial diamicton) produces the widest distributions — very poorly sorted sediment. The sorting of a sandstone reservoir therefore carries information about the depositional environment that can be used to predict reservoir quality in undrilled areas from the depositional facies model.

In reservoir engineering, sorting is not just a descriptive property — it quantitatively affects the petrophysical properties that control production. The relationship between sorting, porosity, and permeability is a reliable predictor of reservoir quality variation within a formation, enabling reservoir engineers to use facies maps (which distinguish depositional environments with different expected sorting) as proxies for porosity-permeability distribution when core data is sparse.

Sorting in Reservoir Quality Prediction

Porosity prediction from sorting and grain size uses empirical relationships derived from regional datasets of core analysis measurements paired with grain size and sorting data from petrographic thin sections or sieve analysis. These relationships take the form of regression equations or classification trees that predict porosity from depositional setting (aeolian, fluvial, turbidite) and grain size — the depositional setting implicitly encodes sorting. In basin-scale reservoir quality models (such as those built by operators for regional play assessment or by service companies for petrophysical benchmarking), sorting enters as a parameter in the porosity prediction workflow, particularly for pre-drill reservoir quality assessment where no core data is available.

Permeability-porosity transforms (k-phi crossplots) show characteristic scatter that often reflects sorting variation — within a single rock type (same depositional environment, similar diagenetic history), well-sorted samples cluster along a high-slope k-phi trend while poorly sorted samples cluster along a lower trend with lower permeability at the same porosity. Separating core data by sorting class (from petrographic descriptions) before fitting k-phi transforms reduces scatter and improves the accuracy of permeability prediction from porosity in heterogeneous formations where mixed facies (and therefore mixed sorting classes) are present.

Grain size and sorting from log analysis uses calibrated acoustic slowness or gamma ray measurements to estimate grain size trends between core data points. The gamma ray log detects clay content (fine-grained intervals with poor sorting tend to have elevated gamma ray from clay minerals) and the sonic log responds to grain contact framework stiffness (coarser, better-sorted grains provide stiffer grain-to-grain contacts with higher sonic velocity). Neural network or multivariate regression models trained on core-calibrated log datasets can predict grain size and sorting quasi-continuously, providing a basis for reservoir quality mapping in wells without core coverage.

Sorting Across International Jurisdictions

Canada (AER / WCSB): Sorting variation in WCSB reservoirs reflects the diverse depositional environments of the Western Canada Sedimentary Basin: Viking Formation sands (shoreline and estuarine) are well to very well sorted, providing some of the best reservoir quality in WCSB shallow gas plays; Cardium Formation sands range from well-sorted beach deposits in the upper Cardium to poorly sorted conglomerate in the lower Cardium, creating reservoir quality tiers directly correlated with sorting; and Mannville Group point bar and channel sands have moderate sorting that degrades toward the margins of channel systems where fine-grained floodplain deposits mix with channel sand, reducing sorting and reservoir quality in the transition zones. AER core analysis databases for WCSB pools contain grain size and sorting descriptions from thousands of core samples that provide the petrographic basis for formation-specific reservoir quality prediction.

United States (API / BSEE): Permian Basin Wolfcamp and Bone Spring reservoirs are characterized by mixed sorting from their deep-water slope and basin depositional environments — proximal debris flow deposits are very poorly sorted (angular, poorly rounded clasts in a silt-clay matrix), while distal turbidite sheets are moderately sorted (sand-dominated but with silt interlamination). The sorting variation directly drives the bimodal porosity-permeability distribution observed in Wolfcamp core data, with the debris flow facies contributing large volumes but poor quality rock and the turbidite facies providing better reservoir quality for hydraulic fracture stimulation. Gulf of Mexico Miocene turbidite sands (Nansen, Ursa, Mars) are generally well to moderately sorted, reflecting the distal, high-energy turbidite current deposition in the deep-water environment, and provide the excellent reservoir quality (porosity 25-35%, permeability 1-5 Darcy) that makes these sands productive at large rate.

Norway (Sodir / NORSOK): North Sea Brent Group sands (Brent Province fields: Statfjord, Gullfaks, Oseberg, Brent, Cormorant) are typically well to very well sorted in the Tarbert and Etive formations (shoreline-to-shelf sand bodies) and moderately sorted in the Ness Formation (fluvial and deltaic plain deposits) — sorting variation within the Brent Group is a primary control on the reservoir quality heterogeneity that complicates production performance prediction in Brent Province fields. Norwegian Core Depository at Geological Survey of Norway holds physical core samples from NCS wells with documented grain size and sorting data that provides the petrographic basis for regional sorting maps used in Brent Province reservoir characterization.

Middle East (Saudi Aramco): Arab Formation carbonates in Saudi Arabia are not siliciclastic sediments and therefore grain size and sorting in the sedimentological sense refers to carbonate grain types (ooids, peloids, bioclasts, intraclasts) rather than quartz-feldspar framework grains — but the same principles apply: Arab D grainstone facies (well-sorted carbonate grains, high primary porosity) provides the best Arab Formation reservoir quality, while Arab D wackestone and mudstone facies (poorly sorted mixture of carbonate particles in a micritic matrix) have much lower porosity and permeability. Saudi Aramco's Arab Formation reservoir characterization framework uses carbonate rock typing based on depositional texture (Folk classification: grainstone, packstone, wackestone, mudstone) as a proxy for grain size and sorting that controls reservoir quality, with rock types mapped between wells using log responses calibrated to core thin sections.