Apparent Anisotropy

Key Takeaways

• The NMO velocity overestimates the true vertical velocity in VTI-anisotropic formations by a factor controlled by the Thomsen η parameter: For a VTI (vertically transversely isotropic) medium — a transversely isotropic rock in which the symmetry axis is vertical, as in a horizontal shale layer — the relationship between the NMO velocity and the vertical velocity for a P-wave reflection from a horizontal reflector at depth H is approximately: V_NMO² ≈ V_V² × (1 + 2η), where η = (ε − δ) / (1 + 2δ). For the Montney siltstone in northeast BC, typical Thomsen parameters measured from VSP data are ε ≈ 0.12 to 0.18, δ ≈ 0.06 to 0.10, giving η ≈ 0.05 to 0.09. At η = 0.07, V_NMO / V_V = √(1 + 2 × 0.07) = √1.14 = 1.068, meaning the NMO stacking velocity overestimates the true vertical velocity by about 6.8 percent. Using V_NMO directly in depth conversion (without anisotropy correction) would place the Montney top approximately 6 to 7 percent too deep, a depth error of 200 to 280 m at a typical Montney depth of 3,200 to 4,000 m — a critical error that would cause the drill bit to miss the target and potentially penetrate an over-pressured formation without adequate mud weight preparation.
• Apparent anisotropy from structural dip is a geometric effect distinct from true rock fabric anisotropy: Even in a perfectly isotropic medium (no true rock anisotropy), a dipping reflector will cause the NMO velocity measured from surface seismic reflection to differ from the true layer velocity. For a layer dipping at angle α, the NMO velocity from a common midpoint (CMP) gather is V_NMO = V_true / cos(α) for up-dip shooting and V_NMO = V_true × cos(α) for the dip direction, depending on whether the receiver spread is in the dip or strike direction. In the Alberta Foothills where structural dips of 10 to 40 degrees are common in fold-thrust belt anticlines, this dip-induced apparent anisotropy can cause NMO velocity errors of 1 to 15 percent (cos(10°) = 0.985, cos(40°) = 0.766) that are geometrically identical in their moveout effect to moderate true VTI anisotropy. Pre-stack depth migration (PSDM) methods effectively correct for dip effects by migrating data before velocity analysis, so that the velocity field extracted from PSDM is a true velocity model rather than a stacking velocity contaminated by dip. However, PSDM is computationally expensive and requires a good initial velocity model, making it a significant processing investment for exploration targets in the foothills.
• Apparent anisotropy is routinely measured by comparing surface seismic stacking velocities to VSP-derived interval velocities: The most reliable method for quantifying the combined apparent anisotropy (including both true η and any processing or geometric contributions) is to acquire a zero-offset VSP (vertical seismic profile) in the same well and formation, which directly measures the vertical propagation time from surface to each depth level using a downhole receiver and surface source. The VSP-derived interval velocity (calculated from the first-arrival times between adjacent receiver depths) is the true V_V free of any anisotropy or dip contamination. Comparing the VSP-derived V_V to the surface seismic NMO velocity at the same depth (using the Dix interval velocity extraction from surface seismic RMS velocities) gives the combined apparent anisotropy correction factor directly: (V_NMO / V_V)² − 1 = 2η_apparent. If the VSP and seismic data are of high quality, this comparison isolates η_apparent to ±0.02, which is adequate for depth conversion at typical WCSB exploration targets. Walkaway VSP data (in which the source is offset from the wellhead) allow the Thomsen parameters to be extracted directly from the near-vertical and wide-angle P-wave travel times, separating the individual contributions of ε, δ, and η for use in anisotropic depth migration of the 3D surface seismic data.
• Uncorrected apparent anisotropy causes systematic depth errors that propagate into structural interpretation and well placement: In exploration projects on the Montney or Duvernay plays where wells cost CAD 6 to 15 million each, a depth conversion error of 3 to 8 percent (from uncorrected apparent anisotropy) that mislocates the structural crest by 100 to 300 m can cause the discovery well to land in the wrong fault block, miss the optimum updip position on an anticlinal trap, or underestimate the vertical depth to the target reservoir, leading to an incorrect casing programme (insufficient intermediate casing to cover the abnormal pressure transition, for example). The economic consequence of a missed well or a wellbore integrity failure from an unexpected overpressure zone is 10 to 50 times the cost of including anisotropy characterisation and correction in the seismic processing workflow. Best practice for WCSB Montney and Duvernay exploration programmes is to acquire at least one VSP in the first exploration well of a new area, characterise the apparent anisotropy in the target formations, and incorporate the anisotropy correction into the depth conversion workflow for all subsequent structural maps and well proposals in the area.
• AVO analysis and pore pressure prediction from seismic velocities are both sensitive to uncorrected apparent anisotropy: Amplitude-versus-offset (AVO) analysis uses the variation of reflection amplitude with source-receiver offset to infer changes in rock properties (porosity, fluid content, lithology) across a reflector. Because AVO analysis is based on the offset-dependent travel time moveout of reflected waves, any systematic difference between the actual moveout (including anisotropy) and the assumed isotropic moveout will leave residual moveout in the offset gathers, causing amplitude versus offset to be contaminated by a moveout-amplitude coupling artifact. In Montney and Duvernay AVO studies, failing to apply anisotropic NMO correction (using V_NMO computed from the anisotropic model rather than the isotropic approximation) before AVO attribute extraction can introduce systematic errors in the extracted intercept and gradient that mimic or mask the lithological and fluid effects that AVO is designed to detect. Pore pressure prediction from seismic velocities uses the Dix interval velocity to identify zones where velocity is anomalously low relative to the compaction trend, inferring undercompaction-induced overpressure from the velocity anomaly. If the interval velocity extracted from Dix analysis using stacking velocities is inflated by 6 to 8 percent due to uncorrected apparent anisotropy, the predicted pore pressure gradient in the anisotropic shale zones will be correspondingly underestimated, reducing the predicted mud weight requirement and potentially leading to underbalanced drilling conditions in the overpressured zone.