How does seismic data analysis support oil exploration?

Seismic Data Analysis in Oil Exploration

Seismic data analysis converts reflected wavefields into geologically consistent images and rock-property volumes, enabling risked prospect generation, well targeting, and portfolio decisions across the exploration–appraisal interface.

I. High-level purpose and value-chain fit

• I.1 Purpose: Reduce subsurface uncertainty by mapping structure, stratigraphy, and fluid indicators; quantify trap integrity and reservoir quality; and establish well locations with defendable risk profiles.
• I.2 Where it fits: Upstream exploration workflow post-acquisition. Feeds play-based exploration, prospect maturation (LE–ME–HE), and appraisal planning; informs well design and development concept selection.
• I.3 Core outcomes: De-risking (Pg), volumetrics, depth prognosis, sweet-spotting, and value-of-information framing for additional data (e.g., reprocessing, new surveys, VSPs).
• I.4 Link to drilling/production: Guides initial wellbores, geostopping depths, casing points, and hazard avoidance; improves development spacing and recovery factors through reservoir property prediction.

II. Step-by-step process flow

• II.1 Data intake and QC
  • II.1.1 Validate navigation/positioning, geometry, and receiver/source consistency; assess noise, statics, and bandwidth.
  • II.1.2 Key checks: fold distribution, amplitude histograms, frequency spectra, and preliminary common-midpoint (CMP) gathers.
• II.2 Preprocessing (amplitude- and phase-friendly)
  • II.2.1 Debubble/deghost, denoise (random/coherent), surface-consistent deconvolution, true-amplitude recovery, statics (onshore).
  • II.2.2 Preserve relative amplitudes for AVO/AVA by avoiding over-aggressive scaling.
• II.3 Velocity analysis and moveout
  • II.3.1 Semblance picking on CMP/CRP gathers; anisotropic tomography (VTI/TTI) and, where applicable, full-waveform inversion (FWI) for shallow/complex updates.
  • Equations: Normal moveout (flat reflector)

  \( t^2(x) = t_0^2 + \dfrac{x^2}{v_{\mathrm{NMO}}^2} \)

  \( v_{\mathrm{RMS}}^2(T) = \dfrac{\sum_{i=1}^{n} v_i^2\, t_i}{\sum_{i=1}^{n} t_i} \quad;\quad v_{\mathrm{int}}\big|_{T_1\to T_2} = \sqrt{\dfrac{v_{\mathrm{RMS}}^2(T_2)\,T_2 - v_{\mathrm{RMS}}^2(T_1)\,T_1}{T_2 - T_1}} \) (Dix)

• II.4 Multiple and ghost attenuation
  • II.4.1 Apply model- and data-driven de-multiple (e.g., SRME, radon-based, wave-equation); adaptive subtraction and residual notch/ghost removal.
• II.5 Imaging/migration
  • II.5.1 Time migration for benign overburden; depth imaging (Kirchhoff, RTM, LS-RTM) for complex media (salt, basalt, carbonates) with anisotropy (VTI/TTI).
  • II.5.2 Output angle-domain common-image gathers (ADCIGs) for AVO and residual moveout-based model refinement.
• II.6 True-amplitude conditioning
  • II.6.1 Wavelet extraction, Q-compensation, illumination/obliquity correction, offset/azimuth balancing; build stable, phase-consistent volumes.
• II.7 AVO/AVA analysis and inversion
  • II.7.1 Reflectivity fundamentals: acoustic impedance \( Z = \rho V \); normal-incidence reflection \( R_0 = \dfrac{Z_2 - Z_1}{Z_2 + Z_1} \).
  • II.7.2 Shuey two-term approximation (for moderate angles):

  \( R(\theta) \approx A + B \sin^2\theta \), where \( A \approx \dfrac{1}{2}\left(\dfrac{\Delta V_p}{V_p} + \dfrac{\Delta \rho}{\rho}\right), \; B \approx \dfrac{1}{2}\dfrac{\Delta V_p}{V_p} - 2 \left(\dfrac{V_s}{V_p}\right)^2 \left(\dfrac{\Delta V_s}{V_s} + \dfrac{\Delta \rho}{\rho}\right) \)

  • II.7.3 Execute angle-stack AVO classification, fluid/rock discrimination via crossplots (A vs. B, ??–µ?) and perform post-/pre-stack (acoustic/elastic) inversion to derive \( V_p, V_s, \rho \) volumes.
• II.8 Attribute generation and DHI evaluation
  • II.8.1 Structural: coherence, curvature, fault likelihood; Stratigraphic: spectral decomposition, sweetness, RMS amplitude; DHI: flat spots, polarity/phase reversals, amplitude conformance-to-structure.
• II.9 Tie to wells and depth conversion
  • II.9.1 Generate synthetic seismograms from logs, calibrate wavelet, refine time–depth via checkshots/VSP.
  • II.9.2 Depth conversion with multi-scenario velocity models and uncertainty envelopes.

  \( z(t) = \dfrac{1}{2}\int_{0}^{t} v_{\mathrm{int}}(\tau)\, d\tau \)

• II.10 Interpretation and prospect maturation
  • II.10.1 Map traps/seals, charge/fairways, and reservoir bodies; integrate rock physics classes to constrain lithology and fluid scenarios.
  • II.10.2 Volumetrics and risking. Example (oil in place, field units):

  \( \mathrm{STOIIP} = 7{,}758 \times A \times h \times \phi \times (1-S_w) / B_o \)

  • II.10.3 Combine geophysical evidence to assign geophysical Pg; integrate with geological risking (trap, seal, charge, reservoir) for overall PoS.
• II.11 Uncertainty and decision support
  • II.11.1 Ensemble realizations (velocity, wavelet, petrophysics); probabilistic maps (P10–P90) for structure depth, net pay, and properties; VOI for reprocessing/new data.

III. Major equipment/components and functions

• III.1 High-performance compute (HPC)
  • III.1.1 CPU/GPU clusters for imaging (RTM/LS-RTM), FWI, and large inversions; interconnects with high bandwidth for I/O-bound steps.
  • III.1.2 Petabyte-scale storage with fast scratch and tiered archives for multiple processing iterations and what-if models.
• III.2 Seismic processing suites
  • III.2.1 Signal processing, demultiple, anisotropic tomography, imaging, and QC toolkits; batch pipelines and provenance tracking.
• III.3 Interpretation and inversion workstations
  • III.3.1 3D visualization, horizon/fault interpretation, AVO analysis, attribute extraction, and post-/pre-stack inversion environments.
• III.4 Petrophysics and rock physics inputs
  • III.4.1 Well logs (sonic, density, resistivity), core measurements, fluid properties to calibrate inversion and AVO classification.
• III.5 Survey metadata and positioning
  • III.5.1 Source/receiver geometry, tide/velocity statics, navigation QC—critical for amplitude and imaging fidelity.

IV. Key performance drivers

• IV.1 Efficiency
  • IV.1.1 Cycle time from raw field data to prospect-ready volumes; automated QC and model updates; parallelized imaging and inversion runs.
  • IV.1.2 Data management discipline (versioning, metadata) to prevent rework.
• IV.2 Cost
  • IV.2.1 $/km² processed; compute-hour utilization; reprocessing vs. new acquisition trade-off; license and storage costs.
• IV.3 Safety
  • IV.3.1 Better hazard mapping (shallow gas, faults) and precise well targeting reduce drilling exposure and nonproductive time.
• IV.4 Emissions
  • IV.4.1 High-fidelity reprocessing can defer/avoid additional vessel days; optimized drilling reduces unnecessary wells; energy-efficient compute policies lower processing footprint.
• IV.5 Technical fidelity
  • IV.5.1 Bandwidth and signal-to-noise; correct anisotropy; amplitude preservation and illumination compensation; robust well ties.

V. Typical challenges/bottlenecks and mitigation

• V.1 Complex overburden (salt/basalt/carbonates)
  • V.1.1 Mitigation: Anisotropic RTM/TTI, FWI for shallow/high-contrast updates, multi-azimuth illumination and least-squares migration to balance amplitudes.
• V.2 Multiples and ghosts masking DHIs
  • V.2.1 Mitigation: SRME + wave-equation demultiple, adaptive subtraction; deghosting and notch-fill techniques; angle-consistent workflows.
• V.3 Anisotropy and velocity uncertainty
  • V.3.1 Mitigation: Azimuthal/offset tomography, well/PSDM residual moveout constraints, multi-scenario velocity models with Bayesian updates.
• V.4 Thin-bed tuning and limited vertical resolution
  • V.4.1 Vertical resolution limit \( \approx \lambda/4 = \dfrac{V}{4f} \). Use spectral enhancement, wavelet-consistent inversion, and high-frequency attributes to resolve sub-seismic features.
• V.5 Amplitude reliability for AVO
  • V.5.1 Mitigation: True-amplitude processing, illumination/Q corrections, careful scaling, and systematic well-tie validation; avoid mixing vintages without harmonization.
• V.6 Human bias and overinterpretation
  • V.6.1 Mitigation: Blind tests, peer reviews, predefined DHI checklists, probabilistic risking, and decision trees tied to evidence quality.
• V.7 Data governance
  • V.7.1 Mitigation: Rigorous metadata/header QC, audit trails, and single-source-of-truth models to avoid version drift.

VI. Why this activity matters economically/operationally

• VI.1 Higher commercial success
  • VI.1.1 By upgrading Pg (e.g., 0.20 to 0.35 via robust AVO/inversion and depth imaging), expected value rises materially.
  • Estimated example: Single offshore test well (cost = $50 million); discovery NPV (risked to FID) = $600 million; dry-hole cost only if failure.

  Baseline EV: \( \mathrm{EV}_1 = 0.20 \times 600 - 0.80 \times 50 = 120 - 40 = \$80 \) million

  Post-analysis EV: \( \mathrm{EV}_2 = 0.35 \times 600 - 0.65 \times 50 = 210 - 32.5 = \$177.5 \) million

  Gain: \( \Delta \mathrm{EV} \approx \$97.5 \) million (estimated)

• VI.2 Cycle-time and cost reduction
  • VI.2.1 Better images reduce appraisal wells and sidetracks; reprocessing often substitutes for new surveys in early phases.
• VI.3 Improved safety and environmental footprint
  • VI.3.1 Fewer dry holes and optimized trajectories cut well count and emissions; hazard mapping minimizes incident potential.
• VI.4 Strategic optionality
  • VI.4.1 Portfolio ranking, farm-out leverage, and better-timed lease commitments depend on defendable seismic evidence and uncertainty bounds.