Distribution

Key Takeaways

• Proppant particle size distribution determines fracture conductivity and how well the proppant pack maintains permeability under closure stress — a narrow, well-graded proppant distribution (where most particles are close to the nominal mesh size) creates a pack with high porosity and low resistance to fluid flow, maximizing fracture conductivity; a wide or poorly controlled distribution containing excessive fines (particles much smaller than the nominal size) fills the voids between the larger grains and dramatically reduces pack porosity and permeability; API RP 19C specifies allowable fines content (particles passing through the finest sieve in the stack) as a quality control parameter precisely because fines are the most damaging component of a poor PSD; when proppant is transported from the mine to the blender through loading, bulk handling, and pneumatic transfer, attrition can generate fines that degrade the size distribution; on-site laser diffraction testing of delivered proppant provides a rapid QC check that the distribution meets specification before it enters the blending equipment and goes downhole, where it cannot be corrected.
• Bridging agent distribution design for completion fluids is an engineering problem that starts with the pore throat size distribution of the target formation — calcium carbonate or other acid-soluble bridging agents must be sized to physically bridge across the largest pore throats in the formation to stop fluid loss, while particles fine enough to plug the smaller throats must also be present; the Abrams rule of thumb (now considered a starting point rather than a definitive specification) states that the D50 of the bridging agent should equal or exceed one-third of the median pore throat diameter; modern completion fluid design goes further, using pore throat size distributions measured from core samples or estimated from formation permeability to engineer a multi-modal bridging agent distribution (blending two or three size grades of CaCO3) that spans the full range of pore throat sizes in the target formation; a perfectly engineered distribution creates a low-permeability filter cake that stops fluid loss without causing irreversible damage, then dissolves completely when the well is acid-cleaned or put on production — the acid-solubility of CaCO3 is the "undo" button that makes it the industry's preferred bridging agent over silica or barite alternatives.
• Well EUR distribution within a play or acreage position has enormous implications for investment risk and capital allocation — in unconventional plays like the Permian Basin or Marcellus Shale, individual well EURs within a single operator's acreage can range from 200,000 BOE to 2 million BOE or more, driven by local geology, completion design variation, and operational execution; the shape of this distribution (lognormal in most plays, with a long tail of exceptional wells and a mode at lower values) determines the probability that any individual well drills at or above the economic threshold; operators who model only the mean EUR without understanding the distribution underestimate the probability of wells that fail to reach payout; investors who evaluate acreage by average type curve without seeing the distribution of actual results can be systematically misled; the distribution is also what makes "high-grading" economically rational — operators who can identify the geological and operational variables that predict which wells land in the upper tail of the EUR distribution can concentrate capital in those locations and improve portfolio economics without drilling more wells.
• Reservoir property distribution described by geostatistics determines how accurately static models represent the heterogeneity that controls flow — porosity, permeability, and net-to-gross ratio are not uniform within a reservoir; they vary spatially in patterns controlled by depositional environment, diagenesis, and structural deformation; conventional deterministic reservoir models interpolate well data across the interwell space using simple kriging, which produces smooth property maps that honor the wells but underestimate the variability between them; geostatistical simulation (sequential Gaussian simulation, sequential indicator simulation) generates multiple equally probable realizations of the property distribution, each honoring the well data and matching the statistical variability (variogram) measured from the data; these multiple realizations capture the range of possible reservoir architectures and, when run through flow simulation, produce a distribution of production forecasts rather than a single deterministic answer — the P10, P50, and P90 of this forecast distribution are what get reported to investors as the range of resource potential.
• Injection profile distribution during hydraulic fracturing determines whether proppant and fluid actually reach all intended perforation clusters or channelize into a subset — in a multi-cluster horizontal well completion, the ideal distribution is equal flow to every cluster; in practice, flow tends to concentrate in the clusters with lowest entry friction and highest local stress contrast, leaving some clusters taking more than their share of the treatment and others receiving almost nothing; the result is a proppant distribution where some clusters are well-packed and productive while others are poorly stimulated or unstimulated; distributed acoustic sensing (DAS) and distributed temperature sensing (DTS) fiber-optic technologies, pumped inside the cemented casing, provide a continuous measurement of injection profile distribution during the frac job, allowing the engineer to see which clusters are taking fluid and which are not; perforation cluster spacing, limited-entry friction design, and diverter placement decisions are all aimed at making the distribution more uniform — because a uniform distribution means every cluster contributes to production and the total well EUR is maximized.