three-dimensional seismic data, Oil & Gas Glossary

What Is 3D Seismic Data?

3D seismic data is a densely sampled, three-dimensional volumetric image of the subsurface created by recording seismic reflections from a closely spaced grid of source and receiver positions at surface or seafloor, then processing the reflected acoustic energy into a spatially continuous representation of subsurface reflectivity. Unlike 2D seismic surveys (individual lines with no lateral connectivity between them), 3D surveys illuminate the subsurface from multiple azimuths with small bin sizes (typically 12.5–25 m offshore, 25–55 m onshore), producing a seismic volume that can be sliced in any direction — in-line, crossline, time slice, or arbitrary trajectory — to map geological structures, fault networks, and stratigraphic features. 3D seismic has become the industry standard for exploration and development in all major petroleum-producing basins because its structural imaging accuracy, fault resolution, and attribute extraction capability are fundamentally superior to 2D surveys — leading to better well placement, reduced dry hole rates, and improved reservoir characterisation.

Key Takeaways

3D seismic acquires reflected acoustic energy from a dense surface grid — the resulting volume can be "cut" at any angle to produce seismic sections or attribute maps at any depth or time horizon.
Bin size (the area of each individual sample cell, typically 12.5×12.5 m to 25×25 m offshore) controls horizontal resolution — smaller bins resolve finer structural and stratigraphic detail but require more receiver groups and higher acquisition cost.
3D seismic volume interpretation identifies reservoir structure (traps), faults (seal assessment), stratigraphy (reservoir extent and connectivity), and direct hydrocarbon indicators (DHIs: bright spots, flat spots, AVO anomalies).
Seismic attributes extracted from the 3D volume — amplitude, frequency, coherence, curvature, acoustic impedance inversion — provide quantitative information about porosity, fluid content, and rock type beyond simple structural mapping.
Time-lapse (4D) seismic — repeated 3D surveys over the same area — detects changes in the seismic response caused by fluid displacement during production, enabling direct monitoring of waterflood fronts, gas cap expansion, and pressure changes in producing reservoirs.

Acquisition, Processing, and Interpretation

3D seismic acquisition uses arrays of sources (air guns in marine surveys, vibroseis trucks or explosive shots in land surveys) firing into arrays of receivers (hydrophone streamers towed behind seismic vessels at sea; geophones in surface patches or nodes for land) at a predefined spatial sampling grid. Marine acquisition typically uses 6–12 streamers (each 6–8 km long) towed 50–100 m apart, with the vessel making parallel passes to cover the survey area — a typical deepwater 3D survey covers 1,000–5,000 km² and takes 4–8 weeks. Ocean Bottom Node (OBN) surveys use autonomous sensor units placed on the seafloor instead of towed streamers — OBN records both pressure (hydrophone) and particle velocity (geophone), enabling better imaging below gas clouds, shallow water, and in areas with complex bathymetry. Onshore 3D surveys use explosive or vibroseis sources with geophone receiver patches — logistics are more complex but costs are lower per square kilometre in accessible terrain.

3D seismic processing converts the raw field records (shotgathers) into a subsurface reflectivity image through a sequence of steps: geometry assignment (assigning source-receiver pairs to correct subsurface midpoints), noise attenuation (removing multiples, ground roll, cable noise), surface-consistent deconvolution (removing near-surface effects from wavelet shape), velocity analysis (building the velocity model needed for time-to-depth conversion), migration (repositioning dipping reflectors to their true subsurface position — pre-stack depth migration is the gold standard in complex areas), and post-migration imaging. Full waveform inversion (FWI) is the most advanced processing approach — it iteratively updates the velocity model by minimising the difference between modelled and observed seismic waveforms, producing velocity models with resolution approaching geological detail.

Fast Facts: 3D Seismic Data

Typical bin size: 12.5–25 m (marine), 25–55 m (land) — controls horizontal spatial resolution
Vertical resolution: λ/4 of dominant frequency × interval velocity; typically 5–20 m at reservoir depth
Acquisition cost: $5,000–30,000/km² marine; $50,000–150,000/km² land (depending on logistics)
Key seismic attributes: amplitude, RMS amplitude, frequency, coherence/dip, curvature, impedance inversion
DHIs (direct hydrocarbon indicators): bright spots (gas sands), flat spots (gas-water contacts), AVO anomalies
4D seismic (time-lapse): repeat surveys detect reservoir fluid changes during production — critical for waterflood management
OBN vs streamer: ocean bottom node surveys better for complex subsurface (sub-salt, shallow water, gas clouds)
Key service companies: CGG, TGS, PGS, SLB WesternGeco, BGP (CNPC)

Exploration Tip:

Invest in the best 3D seismic pre-stack depth migration your budget allows before drilling in complex structural settings — subsalt, sub-thrust, or areas with strong lateral velocity variations. Imaging quality is the single largest controllable variable determining whether the drilled well matches the pre-drill depth map. Post-stack time migration (a lower-cost processing approach) can misplace the top of a complex structure by 100–400 m laterally and 50–200 m vertically in areas with significant velocity heterogeneity — and a well drilled 200 m off-target in a 5-km closure finds nothing, while the full field remains undrilled. Anisotropic pre-stack depth migration (PSDM with tilted transverse isotropy modelling — TTI PSDM) is the current standard for subsalt and complex thrust belt imaging. A $5M investment in advanced reprocessing of existing data before a $30M exploration well is almost always justified when the alternative is a dry hole caused by imaging error rather than a truly absent reservoir.

3D seismic data is also referred to as:

3D survey — the acquisition campaign that generates the data; the "survey" is the field operation, the "data" is the processed result
Seismic cube / seismic volume — the three-dimensional data object produced by processing; can be sliced in any direction, manipulated for attribute extraction, or used for depth conversion
4D seismic / time-lapse seismic — a temporal sequence of 3D surveys at the same location, analysed for difference signatures from production-induced fluid changes; the "fourth dimension" is time between surveys
Full waveform inversion (FWI) volume — a 3D seismic volume that has been processed using FWI to produce a high-resolution velocity or acoustic impedance model

Frequently Asked Questions About 3D Seismic Data

What is the difference between 2D and 3D seismic surveys?

2D seismic surveys consist of individual lines acquired with sources and receivers along a single transect — the data from each line represents a vertical slice of the earth beneath that line, with no information between lines. In a 2D survey programme covering a prospect, lines might be spaced 1–5 km apart — meaning subsurface features between lines are inferred by interpolation rather than directly imaged. This can miss faults, pinchouts, or small structural closures that fall between lines. 3D surveys acquire data from a dense grid of source-receiver combinations covering the entire survey area at typical spacings of 12.5–50 m — every point in the subsurface within the survey is illuminated from multiple source-receiver azimuths. This dense sampling allows migration to correctly image dipping reflectors, resolve closely spaced faults, and produce accurate structural maps without interpolation between lines. The quantitative difference: 2D structural maps have depth uncertainties of 50–500 m lateral and 10–50 m vertical in complex areas; 3D structural maps have uncertainties of 5–50 m lateral and 2–10 m vertical. The dry hole rate in acreage covered by 3D seismic is typically 30–50% lower than in acreage covered by 2D seismic alone — an economic improvement that justified the global adoption of 3D surveys as the standard exploration and development tool by the late 1990s.

How does seismic attribute analysis extract reservoir information from 3D data?

Seismic attributes are mathematical transformations of the seismic amplitude volume that extract information about wavelet characteristics, reflection geometry, or rock physics properties. Amplitude attributes (RMS amplitude, maximum amplitude, envelope) highlight anomalous reflections — bright spots from gas sands or flat spots from hydrocarbon-water contacts. Coherence (or dip, curvature) attributes measure the similarity of adjacent seismic traces — chaotic zones of low coherence highlight faults, fracture zones, salt diapir edges, and mass transport deposits that appear as linear discontinuities on coherence maps. Frequency attributes track changes in dominant frequency — anomalously low-frequency zones can indicate hydrocarbon-filled sands (a phenomenon attributed to scattering and attenuation in gas sands). Acoustic impedance inversion converts the seismic amplitude volume into a model of rock acoustic impedance (density × velocity) — the impedance volume can then be tied to well log data to map porosity, lithology, and fluid content away from well control. Spectral decomposition decomposes the seismic volume by frequency, revealing thin-bed tuning signatures that indicate reservoir thickness approaching the tuning threshold (λ/4). Multi-attribute machine learning approaches use all of these attributes simultaneously as input to neural networks trained on well data, predicting lithology and fluid volumes across the 3D volume at seismic resolution — the frontier of quantitative seismic interpretation.

What is 4D seismic and how is it used in waterflood management?

4D (four-dimensional, or time-lapse) seismic repeats a 3D survey over the same area after a period of production to detect changes in the seismic response caused by reservoir fluid changes — the "difference" between surveys (4D signal) reflects areas where fluid displacement has occurred. Water replacing oil or gas changes acoustic impedance: water is more dense and slower than gas (large impedance change gives bright 4D signal from gas-water displacement) and slightly faster than oil (smaller impedance change, subtler 4D signal from oil-water displacement). In waterflood fields, 4D seismic maps the water flood front — showing which parts of the reservoir have been swept by injected water and which areas remain unswept with bypassed oil. This information directs infill drilling into bypassed zones, identifies preferential flow paths (thief zones channelling injection water away from target areas), and confirms reservoir connectivity (or compartmentalisation) between injectors and producers. The Norwegian North Sea has the most mature 4D seismic programme globally — Equinor (formerly Statoil) has conducted over 30 repeat surveys on Gullfaks, Sleipner, and Snorre fields since the 1990s, with documented incremental recovery from 4D-guided infill drilling exceeding 100 million barrels across their operated assets. Equinor's experience established the economic case for 4D seismic that has since been replicated in the Gulf of Mexico, Brazil pre-salt, and North Africa.

Why 3D Seismic Data Matters in Oil and Gas

3D seismic data is the primary subsurface intelligence tool that determines where wells are drilled, what size reservoirs look like, and whether development investments are economically sound. Before 3D seismic became standard in the 1990s, exploration dry hole rates in major basins ranged from 60–80%; 3D seismic coverage reduced dry hole rates to 30–50% by improving structural imaging accuracy and enabling direct hydrocarbon indicator analysis. In development, 3D seismic enables optimal well placement, correct assessment of drainage area per well, identification of compartments requiring additional wells, and fault seal analysis that determines hydrocarbon column heights. For multi-billion-dollar deepwater and LNG projects — where a single well costs $50–200M and a wrong subsurface model can misplace the entire development — the value of 3D seismic data and its continued reprocessing and reinterpretation as new well data arrives is impossible to overstate.