How is reservoir simulation used in oil and gas exploration?

Reservoir simulation in exploration is used to quantify subsurface uncertainty, predict pre-drill deliverability and recovery ranges, design exploration well tests, and inform prospect risking and value. It bridges static geoscience interpretation and early development concept screening, guiding drill/no-drill decisions and appraisal planning.

I. High-Level Purpose and Where It Fits in the Value Chain

• I.1 Purpose: Test plausible geological/fluid scenarios dynamically to estimate ranges for rates, pressures, and recovery factors before or immediately after the first exploration well.
• I.2 Value-chain position: Late exploration to early appraisal. Inputs come from seismic interpretation and prospect volumetrics; outputs inform well test design, appraisal data acquisition, and early development concept screening.
• I.3 Key decisions enabled:
  • Risked STOIIP/GIIP to recoverable volumes mapping and P10–P90 forecasts
  • Expected well test signatures and deliverability to de-risk commerciality
  • Drive mechanism and aquifer strength hypotheses to guide appraisal
  • Screening of primary vs. early secondary concepts and facility pre-sizing

II. Step-by-Step Process Flow

• II.1 Frame objectives and metrics
  • Define decisions: drill/no-drill, test/no-test, appraisal design, bid value.
  • Select metrics: P10/P50/P90 rates (q), cumulative recovery (Np), pressure drawdown (?p), recovery factor (RF), and EMV.
• II.2 Assemble and QC inputs
  • Seismic-derived structure, depth conversion, reservoir presence/quality maps.
  • Analog well logs/cores; if new basin, analog fields; if offset discovery, petrophysics and tests.
  • Fluid PVT assumptions (black-oil or compositional EOS ranges) and SCAL analogs.
• II.3 Build static realizations (uncertain geocellular models)
  • Grid framing and layering consistent with seismic resolution.
  • Populate f, k, NtG, Sw via geostatistics; vary fault transmissibility multipliers (FTM) and contacts.
• II.4 Define fluid and rock-fluid models
  • Black-oil tables or EOS tuned to analog PVT envelopes (Bo, Rs, µ, z). [estimated]
  • SCAL curves: kr(S), Pc(S) families for low/medium/high mobility and wettability cases.
• II.5 Set dynamic boundary conditions and well concepts
  • Aquifer models (van Everdingen–Hurst/Fetkovich) with weak/medium/strong support cases.
  • Trial wellbore models (skin s, rw, kh uncertainty) and surface constraints.
• II.6 Generate uncertainty ensembles
  • Latin hypercube or Monte Carlo over key drivers: A, h, f, k, Sw, contacts, kr/Pc, PVT, aquifer, FTM, s.
  • Use upscaled proxy grids and reduced-physics runs for fast screening; escalate to full-physics on representatives.
• II.7 Pre-drill forecasts and test-case simulation
  • Simulate exploration DST programs (flow–buildup–flow) to predict pressure transients and multi-rate deliverability.
  • Produce rate and cumulative envelopes and diagnostic plots (p*, Horner, log–log) expected if hydrocarbons present.
• II.8 Decision integration
  • Map recoverable distributions to EMV and facility pre-sizes; define appraisal tests that best reduce VoI.
  • Establish update plan: assimilate actual mud-gas, LWD, MDT, DST into ensemble (Bayesian update) post-well.
• II.9 Post-well rapid update
  • Constrain models to RFT/MDT pressures, DST rates, fluid samples, and contacts; re-forecast for appraisal.

II.A Representative Equations Used

• Volumetrics (oil): $$N = \\frac{7{,}758\\, A\\, h\\, \\phi\\,(1 - S_{wi})}{B_o}$$
• Volumetrics (gas): $$G = \\frac{43{,}560\\, A\\, h\\, \\phi\\,(1 - S_{wi})}{B_g}$$
• Recovery factor: $$RF = \\frac{N_p}{N} \\quad \\text{(oil)}; \\quad RF = \\frac{G_p}{G} \\quad \\text{(gas)}$$
• Darcy/inflow (radial, steady-state oil): $$q_o = \\frac{2\\pi k h}{\\mu_o B_o\\,[\\ln(r_e/r_w) + s]}\\,(p_e - p_{wf})$$
• Diffusivity (slightly compressible): $$\\frac{\\partial p}{\\partial t} = \\frac{k}{\\phi\\,\\mu\\,c_t}\\,\\nabla^2 p$$
• Bayesian update concept: $$\\text{Posterior}(\\theta\\,|\\,D) \\propto \\text{Likelihood}(D\\,|\\,\\theta)\\times \\text{Prior}(\\theta)$$
• EMV link to simulation outcomes: $$EMV = POS\\cdot E[NPV\\,|\\,\\text{success}] + (1-POS)\\cdot E[NPV\\,|\\,\\text{failure}] - C_{well}$$

III. Major Equipment/Components and Their Functions

• III.1 Subsurface data inputs
  • Seismic interpretation: structure maps, depth models, fault frameworks.
  • Petrophysical logs/cores (offsets or analogs): f–k trends, NtG, Sw models.
  • Pressure data (offsets): regional gradients, contacts; analog DSTs.
• III.2 Laboratory and test support
  • PVT cells and viscometers: Bo, Rs, µo/µg, z-factor; EOS tuning.
  • SCAL core-flood rigs: kr, Pc, wettability characterization.
  • Exploration DST package: downhole gauges, surface test separator, choke manifold for test design/validation.
• III.3 Software stack
  • Static modeling: structural framework, property modeling, upscaling.
  • Dynamic simulators: black-oil for speed; compositional for volatile oil/condensate or gas cycling scenarios.
  • Uncertainty and optimization: ensemble generation, proxy modeling, assisted history matching.
• III.4 Compute infrastructure
  • Engineer workstations for rapid proxies; on-prem or cloud HPC for ensemble runs.
  • Version control and data management for traceable scenario workflows.

IV. Key Performance Drivers (Efficiency, Cost, Safety, Emissions)

• IV.1 Decision-cycle time: Fast iteration via reduced-order models and smart sampling to meet license and drilling windows.
• IV.2 Uncertainty coverage: Properly spanning drivers (A, h, f, k, kr/Pc, PVT, aquifer, FTM, skin) to avoid optimistic bias; P10–P90 stability with added realizations.
• IV.3 Predictive fidelity vs. data scarcity: Choosing the simplest model that preserves physics material to the decision (e.g., black-oil vs. compositional).
• IV.4 Test design effectiveness: Simulated DST that yields diagnostic pressure behavior and mobilities within operational limits, minimizing rig time.
• IV.5 Cost and emissions: Avoiding unnecessary wells by ruling out non-commercial scenarios; optimized test durations reduce flaring and logistics footprint.
• IV.6 Compute efficiency: Grid upscaling, local grid refinement only where needed, and parallelization to control compute cost.

V. Typical Challenges/Bottlenecks and Mitigation Strategies

• V.1 Sparse or no well control
  • Challenge: High non-uniqueness in f, k, contacts, and fluid type.
  • Mitigation: Analog-anchored priors; wide but geologically consistent ranges; ensemble simulation with Bayesian narrowing as data arrives.
• V.2 Fault seal and connectivity risk
  • Challenge: Unknown transmissibility across faults compartmentalizes flow.
  • Mitigation: Multiple FTM scenarios; simulate DST interference across compartments; design tests to probe connectivity.
• V.3 Rock-fluid uncertainty (kr/Pc, wettability)
  • Challenge: Recovery and deliverability highly sensitive to SCAL.
  • Mitigation: Use SCAL families; anchor to analog wettability systems; target cores and early SCAL in appraisal to collapse uncertainty.
• V.4 PVT/fluid phase behavior
  • Challenge: Volatile oils/retrograde condensates require compositional fidelity.
  • Mitigation: Screen with black-oil proxies, then compositional on key realizations; prioritize early PVT sampling in DST plan.
• V.5 Aquifer strength estimation
  • Challenge: Drive mechanism mis-specification skews decline and RF.
  • Mitigation: Bracket via Fetkovich/van Everdingen–Hurst models; test designs to observe water influx signatures.
• V.6 Grid/upscaling trade-offs
  • Challenge: Coarse grids miss heterogeneity; fine grids delay decisions.
  • Mitigation: Flow-based upscaling; local refinement near wells; surrogate models for screening with periodic calibration.
• V.7 Operational test limits
  • Challenge: DST safety/environmental constraints limit flow periods.
  • Mitigation: Simulate shorter multi-rate programs to preserve diagnostic value; optimize choke schedules and separator limits.

VI. Why This Activity Matters Economically or Operationally

• VI.1 Improves capital allocation: Converts static prospect risk into dynamic recoverable and rate distributions that feed EMV and bid strategies.
• VI.2 Reduces dry-hole and non-commercial outcomes: Identifies scenarios where deliverability or connectivity is inadequate before committing rig time.
• VI.3 Optimizes exploration DSTs: Tailors test durations, rates, and gauge placements to maximize information value per hour of rig time and minimize flaring.
• VI.4 Accelerates appraisal and concept select: Early view of development concepts and facility ranges shortens time to first production when successful.
• VI.5 Strengthens stakeholder confidence: Decision traceability via ensembles and clear P10–P90 envelopes supports internal approvals and partner alignment.

VI.A Quick Computation Links to Decisions

• Rates for test design: Using $$q_o = \\frac{2\\pi k h}{\\mu_o B_o\\,[\\ln(r_e/r_w) + s]}(p_e - p_{wf})$$ to size chokes and separator capacity for exploration DSTs.
• Resource to reserves bridge: STOIIP/GIIP via $$N, G$$ with RF scenarios yields recoverable distributions that drive EMV and concept screening.
• Uncertainty framing: Diffusivity and aquifer models indicate how long to flow and build up to distinguish strong vs. weak support—directly impacting rig time and emission plans.