Acquisition (Seismic)
Seismic acquisition is the process of generating controlled acoustic energy at or near the earth's surface (or seafloor), measuring the reflected and refracted wavefields at arrays of receivers distributed across the survey area, and recording the data in digital form for subsequent processing and interpretation. The objective is to collect seismic traces with sufficient spatial density, offset range, azimuthal diversity, and signal-to-noise ratio to produce, after processing, a seismic image that meets the geological objectives of the project. Acquisition is governed by the fundamental trade-off between survey cost (which scales with the number of source points and receiver stations) and image quality (which improves with fold, offset range, and azimuth uniformity). The choice of acquisition geometry, source type, receiver type, and recording parameters is made during survey design using modelling tools that predict how different configurations will illuminate the target interval and suppress noise. Modern seismic acquisition has evolved from simple 2D single-fold profiles in the 1930s to wide-azimuth, high-density 3D surveys with thousands of receiver channels and multiple simultaneous sources, fundamentally changing the level of geological detail that seismic data can resolve.
Key Takeaways
- Seismic fold is the number of independent source-receiver pairs whose midpoint falls within a given bin (the spatial cell used to sort traces before stacking). Higher fold improves the signal-to-noise ratio of the stacked trace because random noise averages down as the square root of fold while coherent signal sums constructively. A typical land 3D survey in the WCSB targets 60 to 120 fold in the full-offset stack, while marine high-density surveys may exceed 200 fold. Fold is not uniformly distributed across a 3D survey: edges have lower fold (the "fringe zone" or "ramp-up area") and the effective full-fold area is smaller than the total receiver spread. Full-fold area must extend beyond the drilling target to ensure the target is imaged with design fold on all sides; survey design rules of thumb commonly add one migration aperture distance (often 500 to 2,000 metres depending on target depth and complexity) to each edge of the target area to determine the full-fold acquisition footprint.
- The offset range collected in a seismic survey determines both the ability to perform velocity analysis (longer offsets give more moveout curvature for velocity discrimination) and the ability to perform amplitude versus offset (AVO) analysis for lithology and fluid discrimination. Near-offset traces sample near-vertical incidence (close to the zero-offset reflection coefficient) while far-offset traces sample reflection amplitudes at incidence angles of 30 to 45 degrees, where the difference in reflectivity between gas-saturated and brine-saturated sands is measurable. The maximum useful offset is limited by the longest two-way time that the target can sustain without the offset-derived moveout correction (normal moveout, or NMO) stretching the wavelet unacceptably. Maximum useful offset is typically set to approximately the target depth (a one-to-one offset-to-depth ratio) to balance velocity resolution against NMO stretch, though wide-azimuth surveys may use offset-to-depth ratios of 1.5 to 2.
- Source types differ between land and marine surveys. Land surveys most commonly use vibroseis trucks, where large hydraulic vibrators press a base plate against the ground and sweep through a frequency range (typically 8 to 100 Hz over 8 to 20 seconds), generating a controlled, known source wavelet that is cross-correlated with the recorded data to produce a compressed seismic trace. Impulsive sources (dynamite in shallow shot holes, typically 20 to 60 metres deep) are used in areas where vibroseis trucks cannot access (steep terrain, dense vegetation, swamps) or where the near-surface conditions cause vibroseis coupling problems. Marine surveys use pneumatic airguns, which release compressed air (typically at 2,000 psi) from chambers totalling 2,000 to 6,000 cubic inches into the water, generating a broadband acoustic pulse. Marine sources are towed behind the vessel in arrays of multiple airguns fired together or in patterns to shape the source spectrum and reduce ghost reflections from the sea surface.
- The receiver geometry of a 3D land survey is described by the receiver line interval, receiver station interval, source line interval, and source point interval, which together determine the bin size (typically 12.5 to 25 metres in the WCSB), the maximum offset, the cross-line fold, and the azimuth distribution. Orthogonal geometry (receiver lines perpendicular to source lines) is the most common land 3D configuration and provides good azimuth diversity but can create acquisition footprint (systematic amplitude or travel-time variations caused by bin-to-bin fold and offset distribution differences that alias into the interpreted horizons). Wide-azimuth designs (brick or zigzag patterns) improve azimuth distribution at the cost of more complex logistics. Receiver types include planted geophones (single-axis or three-component), dragged arrays (for vibroseis efficiency), and increasingly MEMS (microelectromechanical system) digital sensors in autonomous nodal recorder units that do not require a cable connection to a recording truck.
- Simultaneous source (or blended source) acquisition fires two or more source points at the same time (or in rapid succession with sub-second time gaps) and records the interleaved wavefields together, then uses mathematical separation algorithms in processing to unmix the overlapping signals. Simultaneous source acquisition can double or triple survey productivity because the vessel or vibroseis fleet does not have to wait for the previous shot's energy to decay before firing the next source point. The technique requires careful shot time randomisation or source encoding to ensure that the separated data can be deblended reliably in processing. Simultaneous source methods are now standard in marine acquisition (where they are called simultaneous shooting or blended acquisition) and are increasingly used in land surveys, reducing acquisition cost and environmental footprint by decreasing the number of source-line passes needed for a given fold and offset distribution.
Survey Design: Translating Geological Objectives Into Acquisition Parameters
A seismic survey design begins with the geological objectives: what depth is the target, what is the structural dip, what is the expected size of the smallest trap to be imaged, and is AVO analysis needed? From these objectives, the required frequency content (which determines thin-bed resolution, typically quarter-wavelength at the dominant frequency), the required fold (which determines signal-to-noise ratio), the required offset range (which determines AVO capability), and the required azimuth distribution (which determines the ability to detect anisotropy or directional fracture sets) are derived. The offset and azimuth requirements are translated into the physical layout of source and receiver lines using modelling software that simulates the acquisition geometry over the target and computes fold-offset-azimuth distributions in each bin.
For a shallow Cretaceous target in the WCSB at 1,000 to 1,500 metres depth, standard acquisition uses a bin size of 12.5 to 15 metres, maximum offset of 1,200 to 1,800 metres, and fold of 80 to 120. For a deep Devonian carbonate target at 3,500 to 5,000 metres depth in the Alberta Foothills, the survey requires a larger bin size (25 metres) because the dominant frequency at depth is lower and fine spatial resolution is not achievable, but a much longer maximum offset (3,000 to 5,000 metres) to provide sufficient AVO analysis capability and refraction tomography data for near-surface velocity model building. High-fold, dense spatial sampling surveys designed for imaging beneath the complex thrust structures of the Foothills also require irregular topographic correction and anisotropic velocity analysis that push the limits of both acquisition logistics and processing algorithms.
Noise Types and Mitigation in Seismic Acquisition
Seismic acquisition noise is divided into coherent noise (spatially consistent patterns such as ground roll, direct waves, multiples, and wind noise) and random noise (electrical, wind, traffic, and ambient vibration without spatial pattern). Ground roll (Rayleigh waves travelling along the surface) is the dominant coherent noise on land surveys; it travels more slowly than the reflected P-wave (typically 200 to 800 m/s compared to 3,000 to 5,000 m/s for the formation), arrives at the receiver at a high apparent dip on shot records, and has high amplitude and low frequency (4 to 20 Hz) that can overwhelm the reflected signal on shallow traces. Receiver arrays (groups of multiple geophones planted in a pattern and wired together to a single recording channel) attenuate ground roll by destructive interference within the array, taking advantage of the high apparent wavenumber (spatial frequency) of ground roll to discriminate it from the near-vertical reflections. In marine surveys, sea-state noise from wave-induced cable motion (swell noise) is the primary coherent noise source and is attenuated by multi-sensor (hydrophone plus geophone) recording and by deghosting filters that remove the delayed reflection of the source and receiver ghost from the sea surface.
Fast Facts
The first seismic reflection survey performed for petroleum exploration was conducted in Oklahoma in 1921 by the Geophysical Research Corporation, detecting the Viola Limestone at approximately 450 metres depth. 3D seismic acquisition was first applied commercially in the Gulf of Mexico by Exxon in the late 1970s, and 3D became the dominant acquisition mode for development and exploration by the mid-1980s. Today approximately 150,000 to 200,000 square kilometres of new 3D seismic data are acquired worldwide each year (pre-pandemic peak), with the Canadian Foothills and the WCSB representing significant portions of North American land acquisition activity. The transition from analogue to digital recording (24-bit floating-point systems became standard in the 1990s) and from cable-connected receivers to wireless nodal systems (driven by inaccessible terrain and urban noise in areas like the Foothills and the Montney) has dramatically changed field operations over the past two decades. Wide-azimuth towed streamer (WATS) surveys in the Gulf of Mexico and ocean-bottom node (OBN) surveys over complex deepwater fields represent the current frontier of marine acquisition, providing azimuth diversity previously available only to land 3D surveys. The Alberta Geological Survey and the British Columbia Oil and Gas Commission maintain geophysical databases of historical 2D and 3D seismic data acquired in their respective provinces, accessible to industry for geological studies and play fairway analysis.
Synonyms and Related Terminology
Seismic acquisition is also called seismic data collection, field data acquisition, or simply fieldwork in the context of a seismic program. The programme itself is often called a seismic survey or shoot. Related terms include seismic fold (the number of independent source-receiver pairs contributing a trace to each CMP bin; higher fold improves signal-to-noise ratio after stacking and is the primary quality indicator of 3D seismic acquisition density), vibroseis (the land seismic source system in which a hydraulic vibrator presses a base plate against the ground and sweeps through a frequency range to generate a known swept-frequency waveform; the dominant land seismic source in accessible terrain, replacing dynamite in most WCSB surveys from the 1970s onward), airgun (the marine seismic source that releases compressed air from a pressurised chamber into the water to generate a broadband acoustic pulse; towed in arrays behind the vessel to produce the source waveform used in 2D and 3D marine seismic surveys), common midpoint (CMP, the geographic point halfway between a seismic source and receiver; traces sharing the same CMP from different source-receiver pairs are sorted into a CMP gather and stacked to improve signal-to-noise ratio, the fundamental operation of seismic reflection processing), and nodal seismic acquisition (a wireless land seismic recording system in which autonomous sensor units are planted at receiver stations without cable connections to a central recording truck; enables acquisition in areas where cable deployment is impractical, including mountainous terrain, densely built environments, and inaccessible swamp or muskeg).