Amplitude Variation with Offset

Key Takeaways

• AVO is classified into four standard classes based on the sign and magnitude of the normal-incidence reflectivity (intercept R₀) and the AVO gradient (G), with each class corresponding to a distinct geological and fluid scenario: Class I (positive R₀, negative G) indicates a reservoir with higher impedance than the surrounding shale (hard reflection that dims with offset), Class II (near-zero R₀) is a polarity-sensitive crossover case, Class III (negative R₀, negative G) is the diagnostic bright-spot gas signature in low-impedance sand below shale, and Class IV (negative R₀, positive G) occurs in unusual geological settings where gas sand amplitude decreases with offset despite being a soft reflector: Class I AVO occurs where the reservoir sand (even when gas-saturated) has higher acoustic impedance than the encasing shale, which is typical of deeply buried tight sandstones with significant quartz cementation (Deep Basin Cadomin and Spirit River formations in northwest Alberta, depths 2,500 to 4,500 m). In this case, gas saturation causes the impedance to decrease from the brine-saturated value but not below the shale impedance, so the reflection remains positive-polarity (hard) at normal incidence but decreases with offset as the Vs constraint reduces the gradient. Class III AVO (negative R₀, negative G) is the most prevalent gas sand signature in WCSB shallow exploration (Viking, Cardium, Mannville at 500 to 1,500 m depth), where low confining stress, high porosity, and high gas saturation combine to produce gas-sand impedance well below the shale impedance, generating bright spots that increase in amplitude with offset. The AVO class is not a fixed property of a formation but depends on the specific Vp-Vs-density relationships of both the reservoir and the surrounding shale at the burial depth and compaction state of the target.
• The AVO intercept-gradient crossplot is the primary diagnostic tool for identifying gas sands in pre-stack seismic data, with the background trend (shale-shale reflections plotting near the origin with a negative slope defined by the local mudrock line) providing a reference against which hydrocarbon anomalies plot as outlier clusters in the quadrant corresponding to the expected AVO class for the target formation, enabling semi-automatic identification of anomalous reflectors from hundreds of millions of seismic data points: On an intercept (R₀) versus gradient (G) crossplot, elastic reflections from shale-shale interfaces follow a linear trend (the background or wet trend) with a negative slope reflecting the normal Vp/Vs ratio in the shale sequence. Gas sand reflectors at the top of a Class III gas reservoir plot in the third quadrant (negative R₀, negative G), displaced from the background trend by the anomalously low Vs/Vp ratio of the gas sand relative to the surrounding shale. Brine-saturated sands with the same lithology plot near the background trend on the crossplot, providing the discrimination between gas and brine that normal-incidence amplitude alone cannot provide when the acoustic impedance contrast is similar for both fluid states. In WCSB Viking play evaluation, intercept-gradient crossplot analysis of 3D pre-stack seismic data routinely identifies 15 to 30 anomalous reflectors per 100 km² of 3D seismic coverage, of which 40 to 60% are subsequently ranked as medium to high confidence DHI prospects by AVO class, amplitude conformance to structure, and well calibration consistency.
• Gassmann's equations (1951) provide the rock physics foundation for AVO by predicting the bulk modulus (and therefore Vp) of a porous rock saturated with different fluids, enabling quantitative fluid substitution modelling that calculates the expected AVO response for gas saturation, oil saturation, and brine saturation from well log data, validating that the observed seismic AVO anomaly is consistent with the expected fluid contrast before committing to a drilling decision: Gassmann's relation: K_sat = K_dry + (1 - K_dry/K_mineral)² / (φ/K_fluid + (1-φ)/K_mineral - K_dry/K_mineral²), where K_sat is the saturated-rock bulk modulus, K_dry the dry-frame modulus, K_mineral the mineral grain modulus, K_fluid the pore fluid modulus, and φ the porosity. By substituting K_fluid values for brine (2.2 GPa), oil (0.7 to 1.5 GPa depending on API gravity), and gas (0.02 to 0.10 GPa depending on pressure and composition), the geophysicist can calculate the Vp for each fluid scenario from the same K_dry measured on dry core plugs (or estimated from log Vp and the Biot coefficient), then compute the expected AVO response for each scenario and compare to the observed seismic AVO. For WCSB Viking gas sands, Gassmann modelling using core-plug K_dry values typically predicts Class III AVO with gradient G = -0.04 to -0.09 for gas saturation above 50%, while brine substitution predicts G = -0.01 to +0.02, providing a clear discrimination between gas and brine fluid states on the AVO gradient measurement.
• Azimuthal AVO (AVAZ) analysis uses the variation of AVO gradient as a function of seismic acquisition azimuth to detect seismic anisotropy caused by aligned natural fractures or in-situ stress anisotropy, providing fracture orientation and intensity information from 3D seismic data that guides horizontal well landing orientation and perforation cluster spacing decisions in naturally fractured Montney and Duvernay reservoirs in Alberta: In a naturally fractured reservoir, aligned vertical fractures create seismic velocity anisotropy: P-wave velocity propagating perpendicular to the fractures is lower than velocity propagating parallel to the fractures (because fractures open in the perpendicular direction and reduce the elastic stiffness). This velocity anisotropy translates into azimuth-dependent AVO responses: the AVO gradient is larger (more negative for a soft reflector) in the azimuth perpendicular to the fractures than in the azimuth parallel to the fractures. Fitting the AVO gradient variation with azimuth (using a cos²θ azimuth function) yields the fracture orientation (from the direction of maximum gradient) and the fracture intensity (from the magnitude of the azimuthal gradient variation). In the Montney Formation of northeast British Columbia (Dawson Creek, Pouce Coupe, Encana and ConocoPhillips acreage), AVAZ analysis of wide-azimuth 3D seismic has identified NW-SE to NNW-SSE fracture orientations (consistent with the regional maximum horizontal stress direction) that are used to orient horizontal well laterals perpendicular to the fractures and to select perforation cluster spacings that maximise interaction with the natural fracture network, improving stimulated reservoir volume (SRV) and first-year cumulative production by an estimated 10 to 25% over wells landed and completed without AVAZ guidance.
• AVO analysis requires careful data quality control and processing to preserve the angle-dependent amplitude variations that carry the fluid information, including offset-balancing of near and far trace amplitudes, removal of amplitude distortions from acquisition geometry (source and receiver coupling variations), residual normal moveout correction (to ensure that reflections are properly aligned before AVO extraction), and overburden Q-compensation (to correct for the frequency-dependent attenuation that distorts amplitude differently at near and far offsets), with any uncorrected error in these steps potentially creating false AVO anomalies that are mistaken for hydrocarbon indicators: The most common AVO artefact in WCSB 3D seismic data is near-surface coupling variations: the amplitude of seismic energy recorded at receivers in soft near-surface materials (glacial lake clays, peat bogs, river valley alluvium) is attenuated differently than at receivers in hard near-surface materials (compact till, limestone outcrop), creating receiver-term amplitude variations that appear in the recorded data as offset-dependent amplitude changes (because different offsets use different receiver locations). Uncorrected, these near-surface coupling variations produce false AVO gradients that can be mistaken for lithological or fluid signals. Amplitude-preserving surface-consistent amplitude corrections (SCAC) remove the source and receiver coupling terms from the data in a least-squares fit using all reflections from multiple horizons, enabling AVO gradient extraction with accuracy sufficient to discriminate Class III gas sands from Class IIp background in WCSB shallow gas exploration. High-quality AVO results in the WCSB require raw data fold greater than 60 (typically 80 to 120 for DHI-grade data quality) and signal-to-noise ratio above 8 dB at the target level, setting the minimum 3D seismic survey design specification for AVO-capable datasets.