Square Log

Key Takeaways

• The choice of averaging method for log blocking depends on the petrophysical property being blocked and the flow geometry of the reservoir: permeability is averaged using the arithmetic mean when flow is parallel to the layering (horizontal flow in a laterally continuous layered system, where total flow equals the sum of flows in each layer) and the harmonic mean when flow is perpendicular to the layering (vertical flow through layered beds in series, where total resistance equals the sum of resistances of each layer); porosity is typically averaged arithmetically (simple thickness-weighted average) because it is a volumetric property; water saturation is averaged using a pore-volume-weighted arithmetic mean that accounts for both the porosity and the thickness of each interval; net-to-gross is computed as the fraction of the total interval thickness that meets the pay cutoff criteria; applying the wrong averaging method introduces systematic errors in the blocked property — most commonly, arithmetic averaging of permeability in a vertically heterogeneous interval overestimates the effective vertical permeability that controls crossflow and gravity drainage in the simulation model.
• Upscaling from log scale to simulation scale is the most critical application of square log creation in reservoir engineering, because the mismatch between the vertical resolution of wireline logs (centimeter-scale for density and neutron logs, decimeter-scale for sonic and resistivity) and the vertical cell thickness of reservoir simulation models (one to ten meters for sector models, five to twenty meters for full-field models) requires a systematic method for aggregating fine-scale measurements to coarser representative values: the upscaling workflow begins with petrophysical interpretation of the fine-scale logs to produce continuous curves of porosity, water saturation, and permeability at log scale, then assigns each depth sample to a simulation cell based on the cell boundary depths derived from the geological model, and finally applies the appropriate averaging method to compute the cell-average property; the power-averaging method (using an exponent omega between -1 for harmonic and +1 for arithmetic averaging, with omega = 0 corresponding to geometric averaging) is sometimes used to tune the averaging between these end-members for cases where the flow geometry is intermediate between pure parallel and pure series flow; flow-based upscaling methods that simulate single-phase flow on the fine-scale log to compute an effective permeability that honors both the heterogeneity and the flow geometry are the most accurate but computationally expensive method for complex heterogeneous intervals.
• Net pay determination from blocked logs requires the application of petrophysical cutoffs (minimum porosity, maximum water saturation, minimum permeability) to identify which depth intervals within the reservoir contribute to producible hydrocarbon volume: the net-to-gross ratio (N/G) is calculated as the fraction of the gross reservoir interval (defined by the structural or stratigraphic boundaries) that meets all pay cutoffs, and the blocked log assigns a binary value (net pay = 1, non-net = 0) to each sampling interval based on whether the fine-scale log measurements within that interval satisfy the cutoff criteria; the choice of blocking interval significantly affects the computed N/G — if the blocking interval is larger than the thickness of thin pay beds, the averaging process can cause the thin bed's properties to be diluted by adjacent non-pay intervals, producing a blocked value that falls below the pay cutoff even though the thin bed itself is pay quality; the optimal blocking interval for net pay determination is typically the smallest geological unit that is geologically meaningful for the reservoir characterization objective, which may be as fine as the log sampling interval (0.1-0.5 foot) for thin-bedded turbidite reservoirs or as coarse as a major stratigraphic unit for thick homogeneous reservoirs.
• Electrofacies classification using blocked logs is a method for assigning geological facies or rock types to depth intervals based on the values of blocked log measurements, providing a simplified facies model that can be propagated through the inter-well volume using geostatistical methods: the electrofacies classification clusters depth intervals with similar blocked log responses (combinations of gamma ray, neutron porosity, bulk density, sonic transit time, and resistivity) into discrete facies types that correspond to geological rock types (clean sand, shaly sand, silt, shale, carbonate, etc.) using multivariate statistical methods such as cluster analysis, neural networks, or multi-resolution graph-based clustering; the blocked log values within each electrofacies are then described by their statistical distributions (mean, variance, histogram) that are used as the conditioning data for geostatistical simulation (sequential Gaussian simulation, sequential indicator simulation) to populate the inter-well volume with facies-conditional property distributions; the quality of the electrofacies classification depends on the consistency of the log response within each facies type across all wells — if the same geological facies produces different log responses in different wells due to diagenetic differences or log quality issues, the classification will be inconsistent and the geostatistical model will contain artefacts that cannot be reconciled with the geological understanding of the reservoir.
• Dynamic data integration using blocked logs requires matching the simulated production response to the observed well test and production data by adjusting the blocked properties (primarily permeability and skin) within the simulation model: history matching starts with the blocked log properties as the initial model and perturbs the permeability within geologically constrained ranges to minimize the difference between simulated and observed bottomhole pressure, flow rate, and water cut; the blocking interval defines the minimum scale at which permeability can be independently adjusted during history matching — if the blocking interval is five meters, the history matching can only modify permeability at five-meter resolution, and finer-scale heterogeneities that affect the well's transient response cannot be captured; the relationship between the blocked simulation model and the well test interpretation is mediated by the skin factor that captures near-wellbore damage, stimulation, and partial penetration effects that are below the resolution of the simulation grid; integrating well test permeability-thickness (kh) with blocked log permeability requires reconciling the dynamic measurement (which integrates flow over the entire drainage volume of the test) with the static measurement (which samples the formation locally at the wellbore) through the concept of effective permeability that accounts for both the blocking and the near-wellbore effects captured in the skin.