Detectable Limit

Key Takeaways

• Tuning thickness as the practical resolution limit defines the depth below which further thinning of a layer causes its seismic amplitude to decrease rather than increase: at exactly the tuning thickness (lambda/4, approximately 20 to 50 feet for typical exploration seismic with 50 to 80 Hz dominant frequency and 8,000 to 12,000 ft/s rock velocity), the top and bottom reflections of the layer are separated by exactly one half-period and interfere constructively to produce the maximum possible composite amplitude for that impedance contrast; for layers thicker than tuning thickness, the composite amplitude approximately equals the amplitude expected from the single interface reflection (the top or bottom reflection alone); for layers between the tuning thickness and the detectable limit, the composite amplitude decreases monotonically with decreasing thickness because the top and bottom reflections begin to cancel each other (the partial destructive interference increases as the two-way time through the layer decreases below the half-period); the thickness of a sub-tuning thin bed can be estimated from the composite amplitude using the wedge model calibration (a wedge-shaped model that relates observed amplitude to layer thickness for the known acoustic impedance contrast), provided that the amplitude-thickness relationship for the specific reflection has been calibrated against well data or forward modeling.
• Frequency content of the seismic data is the primary engineering parameter that determines both the tuning thickness and the detectable limit for a specific geologic target, with higher-frequency seismic data providing finer resolution and deeper detectability: for a 30-meter water depth site survey acquiring data with 200 Hz dominant frequency in 1,500 m/s near-surface material, the tuning thickness is 1.5/(4 times 200) = 1.9 meters and the detectable limit is approximately 0.5 meters; for a conventional 3D exploration survey with 50 Hz dominant frequency in 3,000 m/s sandstone, the tuning thickness is 3,000/(4 times 50) = 15 meters (50 feet) and the detectable limit is approximately 5 meters (15 feet); acquiring broadband seismic data with extended high-frequency content (through ghost-free acquisition, longer recording time for lower noise floor, or marine vibroseis swept to higher frequencies) reduces the dominant wavelength and moves the detectable limit to shallower thicknesses, improving the ability to detect thin reservoir sands that are critical for reserve estimation and development planning in stratigraphically complex producing formations.
• Amplitude versus offset (AVO) analysis extends the detectability of thin beds by exploiting the systematic variation of reflection amplitude with source-receiver offset that depends on the angle-dependent acoustic impedance contrast between the thin layer and its bounding rocks: a thin gas-saturated sand below the detectable limit for zero-offset amplitude may be detectable at far offsets where the gas-sand's lower Poisson's ratio creates a stronger AVO response than the brine-saturated equivalent would produce; the Class III AVO response (which shows increasing negative amplitude with increasing offset for gas sands below higher acoustic impedance shale) is detectable at thicknesses well below the zero-offset amplitude detectable limit because the AVO gradient (change in amplitude with sin-squared of the offset angle) is less sensitive to interference between top and bottom reflections than the zero-offset amplitude is; AVO-based detection of thin gas sands in the Paleogene turbidite sequences of the deepwater Gulf of Mexico and West Africa has enabled discovery and development of reservoirs thinner than 5 to 10 feet that would not have been identifiable from conventional amplitude analysis of zero-offset stacked data alone.
• Seismic inversion converts the amplitude information in seismic data into acoustic impedance (or elastic impedance in multi-offset inversions), which can then be used to estimate formation thickness more reliably than direct amplitude analysis in the sub-tuning regime by incorporating a well-based low-frequency model that constrains the impedance distribution at frequencies below the seismic bandwidth: model-based inversion (which starts with the impedance model derived from the well log and iteratively adjusts it to match the seismic amplitude response) can resolve layers at thicknesses of lambda/10 or less when the impedance contrast is high and the well control is adequate to calibrate the low-frequency model; the theoretical detectable limit in a model-based inversion is still determined by the signal-to-noise ratio and impedance contrast, but the incorporation of well-derived low-frequency information extends the effective resolution beyond what the seismic amplitude data alone can provide; the practical limitation is that model-based inversion is only reliable within the spatial range where the well-derived low-frequency model is constrained (typically within 1 to 2 kilometers of the calibration well for lateral property continuity assumed in the model), and becomes less reliable as the inversion extrapolates the well-constrained model into structurally or stratigraphically different areas away from well control.
• Practical detectable limit estimation for a specific exploration or development target combines theoretical wavelength analysis with data quality assessment (signal-to-noise ratio, migration quality, spectral bandwidth) and empirical calibration from wells that have encountered the target formation: the theoretical detectable limit (lambda/30 based on wavelength alone) assumes perfect data quality and infinite signal-to-noise ratio; the practical detectable limit at the actual data quality in a specific survey may be 1.5 to 3 times the theoretical limit if the data is noisy or has been affected by migration artifacts; wells penetrating thin beds of the target formation provide the empirical check, by comparing the known bed thickness from the well log to the observed amplitude on the seismic data at the well location, and establishing the minimum thickness at which the seismic still shows a detectable anomaly above the noise floor; this well-calibrated detectable limit is then applied to the seismic amplitude map to distinguish between areas where the absence of an amplitude anomaly confidently indicates no reservoir (thicker beds would have been detected if present) versus areas where the detectable limit is not thin enough to exclude the possibility of a sub-seismic reservoir that could support production but is below the current seismic detectability threshold.