Petroleum Play Fairway Analysis

Five elements.
One prospectivity
surface. Basin-scale risk mapping · reproducible, auditable, quantitative

The method the industry has used for three decades — now in a commercial-grade desktop application. Score reservoir, source, seal, trap, and charge evidence into calibrated [0, 1] surfaces. Propagate with explicit spatial models. Combine into a composite Common Risk Segment map with paired confidence and full provenance audit trail.

0 m1 km 2 km3 km 4 km HC ACCUMULATION SOURCE ROCK (MATURE) CAPROCK / SEAL RESERVOIR STRUCTURAL TRAP CHARGE / MIGRATION BASEMENT
fig. 01 · complete petroleum system reservoir · source · seal · trap · charge
74
GDE library entries
reservoir · source · seal
81
Lithology entries
reservoir quality · seal quality
64
Structural domains
trap · seal integrity · migration
4
Hierarchy levels
play · element · component · proxy
5
Combination methods
geometric mean default
01 The Framework

A petroleum play exists where five independent elements are all present and effective.

The petroleum system requires all five elements to work simultaneously: a reservoir to hold the hydrocarbons, a source to generate them, a seal to prevent escape, a trap to focus accumulation, and a charge pathway to deliver fluids at the right time. Failure of any single element means a dry hole. The PFA application scores each element independently and combines them with non-compensatory methods that respect this logic.

RESERVOIR SOURCE SEAL TRAP CHARGE PROSPECTIVE PLAY

Each element decomposes into components.

Reservoir is not a single number — it decomposes into Presence (is there sand at all?) and Effectiveness (does it have adequate porosity and permeability?). Source decomposes into Presence, Richness, and Maturity. The PFA application supports this hierarchy natively: Play → Element → Component → Proxy, with combination at each level.

The same GDE polygon map can serve as a proxy for Reservoir Presence, Source Presence, and Seal Presence simultaneously — the system reads the appropriate column from the 74-entry GDE Library for each context. Multi-context reuse of data layers is not a workaround; it is the designed workflow.

Element I
Reservoir
presence + effectiveness

Does porous, permeable rock exist at the target horizon? The GDE Library provides reservoir_presence_prob directly as a calibrated [0, 1] value. Effectiveness scored from porosity, N:G, and permeability data.

  • GDE map (reservoir presence)polygon · GDE
  • Porosity (log-derived)points · kriging
  • Net-to-Grosspoints · interp.
  • Lithology (reservoir quality)polygon · lookup
Element II
Source
presence + richness + maturity

Has organic-rich rock been buried deeply enough to generate hydrocarbons? Source potential from GDE Library, richness from TOC measurements, maturity from vitrinite reflectance and Tmax.

  • GDE map (source potential)polygon · GDE
  • TOC (wt%)points · wells
  • Vitrinite reflectance (Ro%)points · wells
  • Tmax (°C)points · wells
Element III
Seal
presence + effectiveness

Is there a laterally continuous, ductile, low-permeability cap above the reservoir? GDE and Lithology Libraries provide seal scoring. Thickness and capillary pressure data quantify effectiveness.

  • GDE map (seal potential)polygon · GDE
  • Lithology (seal quality)polygon · lookup
  • Seal thickness (isopach)raster · seismic
  • Capillary entry pressurepoints · lab
Element IV
Trap
presence + integrity

Is there a structural or stratigraphic geometry that focuses hydrocarbon accumulation? Structural closure maps, four-way dip closure, and fault-seal risk from the Structure Library.

  • Structural closureraster · depth map
  • Fault seal riskstructure lookup
  • Four-way dip closureraster · derived
  • Stratigraphic pinchoutraster · seismic
Element V
Charge
migration + timing

Did hydrocarbons migrate to the trap, and did the trap exist when charge arrived? Kitchen proximity, migration fairway modeling, and burial-history timing constraints.

  • Kitchen proximityraster · distance
  • Migration fairway modelraster · modeled
  • Burial history (timing)raster · 1D model
02 Evidence Layers

From well logs to GDE polygons. Score everything on a common scale.

Petroleum exploration produces the widest variety of evidence types: well data (logs, cores, tests), seismic interpretations (horizons, attributes), geological maps (GDE polygons, lithology polygons, structural domains), and derived surfaces (gravity, magnetics, basin models). The PFA importer handles all formats and the scoring system normalizes everything to [0, 1] favorability.

CON-AF-PFSM-SH-USF DM-TF-CL
polygon · depositional
GDE Map

The workhorse of petroleum PFA. The 74-entry GDE Library auto-populates reservoir_presence_prob, source_potential, and seal_potential. One map, three element contexts.

SHPGeoJSON
0.1 8.4 wt%
point · geochemical
Source Rock TOC

Total Organic Carbon measurements from well cuttings and cores. Scored with a piecewise-linear function: below 0.5 wt% = no source potential; above 2.5 wt% = excellent.

CSVLAS
point · petrophysical
Reservoir Porosity

Log-derived effective porosity at reservoir depth. Scored monotonic-increasing with a minimum cutoff. Combined with N:G as proxies for the Reservoir Effectiveness component.

LASCSV
four-way closure
raster · structural
Structural Closure

Depth-to-top-reservoir map from seismic interpretation. The system computes four-way dip closure and closure area. Scored with sigmoid: larger closure = higher trap probability.

GeoTIFFZMAP
IMMATUREOIL WINDOW GAS
point · thermal
Vitrinite Reflectance

Ro% measurements defining thermal maturity windows: immature (< 0.5%), oil window (0.6–1.3%), gas window (> 1.3%). Scored with bell function peaked at the target hydrocarbon phase.

CSVLAS
10 m30 m 60 m100 m
raster · stratigraphic
Seal Thickness

Caprock isopach from seismic interpretation. Thicker seal = higher containment probability. Scored monotonic-increasing with a minimum-thickness cutoff for adequacy.

GeoTIFFZMAP
raster · potential field
Bouguer Gravity

Regional gravity grid illuminates basement architecture, basin depth, and structural geometry. Gravity lows track deep basins (mature source kitchens); gradients localize buried faults.

GeoTIFFZMAP
KITCHEN
raster · modeled
Kitchen Proximity

Euclidean or fairway-model distance from the mature source kitchen. Scored monotonic-decreasing: closer to the kitchen = higher charge probability. Structural grain controls migration directionality.

GeoTIFFZMAP
03 Hierarchical Scoring

Play → Element → Component → Proxy. Combination at every level.

The SWSS v2.0 scoring system operates on a four-level hierarchy. Each proxy produces a scored raster. Proxies combine into component scores. Components combine into element CRS maps. Elements combine into the play-level CCRS. Combination method and weights are independently configurable at each level, with non-compensatory geometric mean as the default.

PLAY RESERVOIR SOURCE SEAL TRAP CHARGE PRESENCE EFFECTIVE PRES. RICH MATURE GDE map Seis. facies Porosity N:G Perm PLAY ELEMENT COMPONENT PROXY ∏ pᵢwᵢ ∏ wgm

Every level has its own combination.

  • Proxy
    Score one data layer. Apply a scoring function (continuous or categorical) and a spatial influence model. Output: one scored raster [0, 1].
  • Component
    Combine proxies → component. Reservoir Effectiveness = f(porosity, N:G, permeability). Default: weighted geometric mean. Single-proxy components skip combination.
  • Element
    Combine components → element CRS. Reservoir = f(Presence, Effectiveness). Output: the Common Risk Segment map for one petroleum system element.
  • Play
    Combine elements → CCRS. Composite = f(Reservoir, Source, Seal, Trap, Charge). Output: the play-level prospectivity surface with paired confidence.
Multiple plays per project. A basin may host several concurrent plays (e.g., Jurassic Marine, Cretaceous Turbidite, Pre-salt Carbonate). Each play has its own element tree, its own weights, and its own CCRS output. Up to 50 scenarios per project.
04 Composite Prospectivity

Five elements, one surface. Choose the combination that matches the geology.

CCRS — Composite Common Risk Segment
PROSPECT A CCRS = 0.76 LEAD B CCRS = 0.52 0.00.501.0CCRS prospectivity
  • GEOMETRIC MEAN
    P = ∏ páµ¢wáµ¢ · floor 0.01
    Petroleum default. Non-compensatory: if any element scores zero, the composite is near-zero. Weights control relative influence. Floor at 0.01 prevents exact-zero propagation from a single marginal element.
  • MULTIPLICATION
    P = ∏ pᵢ
    Probabilistic AND. Assumes independence. Severely punishing — appropriate when elements are verifiably independent and the analyst wants the strictest possible risk assessment.
  • MINIMUM
    P = min(p₁, p₂, … pₙ)
    Liebig limiting factor. Composite = weakest element. Useful when one element dominates risk (e.g., frontier source presence).
  • OWA
    P = Σ wáµ¢ · p(i)
    Ordered weighted average with adjustable AND-OR continuum. ORness = 0 is pure AND; ORness = 1 is pure OR. For sensitivity exploration.
  • WEIGHTS OF EVIDENCE
    log(P/1−P) = Σ Wᵢ⁺
    Bayesian posterior on log-odds scale, calibrated from known discoveries and dry holes. The gold standard when training data exists.
Petroleum recommendation Default to weighted geometric mean — it is non-compensatory (a zero element kills the composite) but less punishing than pure multiplication for sparse frontier data. In mature basins with discovery/failure calibration, switch to weights-of-evidence. Always run weight-perturbation sensitivity before committing to a prospect ranking.
05 Confidence & Uncertainty

A prospect ranking without confidence is a guess with a grid behind it.

The PFA system generates a paired confidence surface for every CCRS. Confidence is computed via kernel density estimation (Sheather–Jones bandwidth), independent of the interpolation decay. Well-explored fairways show high confidence; frontier blocks show low. The composite can be confidence-weighted so that poorly constrained anomalies cannot outrank well-constrained leads.

Prospectivity
viridis · [0–1]
Confidence (KDE)
grayscale · well density
Alpha-Weighted
prospectivity × confidence
Low-confidence areas fade via alpha blending. The explorer sees one integrated view: prospects are only vivid where data density supports the score.
  • KDE confidenceGaussian kernel, Sheather–Jones bandwidth, Silverman fallback. Bandwidth decoupled from interpolation decay — they answer different questions.
  • Domain-awareKDE bounded by the project AOI polygon. Prevents edge artefacts where coastal wells would project zero confidence offshore into the study area.
  • Weight-perturbation ensembleMonte Carlo mode: ±15% perturbation across N realizations. If a prospect's rank flips, the discrimination is weight-sensitive. The system reports rank-stability per cell.
  • Audit trailEvery parameter — kernel, bandwidth, decay, background, weight — serialized to the H2 project database. The audit report exports to DOCX or Markdown for peer review.
06 Case Study

Frontier basin evaluation — from data ingest to prospect ranking in one session.

A regional explorer opens a new project, imports five evidence layers (TOC, N:G, seal thickness, structural closure, Bouguer gravity), tags each with a play element, and runs the scoring pipeline. In under 15 minutes the application produces a composite CCRS map with confidence transparency, identifying two leads and one prospect — all with full audit trail and sensitivity analysis ready for peer review.

Methodology: Bonham-Carter (1994)
GIS for Geoscientists
Play Fairway Analysis approach
as established by USGS, AAPG, and
industry best practice since the 1990s.
Basin evaluation workflow · Illustrative
PROSPECT ALPHACCRS 0.76 · high conf. LEAD BETACCRS 0.52 · mod. conf. LEAD GAMMACCRS 0.41 · low conf. FRONTIER BASIN · CCRSEPSG:32637 · WGM · background=0.5 N 0.01.0 prospectivity
Composite prospectivity surface from a five-element PFA. Prospect Alpha occupies the convergence of all five favorable elements. Illustrative — methodology-demonstration data.
Scenario Manifest · 14 Layers · 5 Elements
Reservoirw = 0.25
GDE map (res. presence)0.40
Porosity (log-derived)0.35
Net-to-Gross0.25
Sourcew = 0.25
GDE map (source potential)0.30
TOC (wt%)0.40
Vitrinite Ro%0.30
Sealw = 0.20
GDE map (seal potential)0.45
Seal thickness isopach0.55
Trapw = 0.15
Structural closure0.60
Fault seal risk (structure)0.40
Chargew = 0.15
Kitchen proximity0.55
Bouguer gravity (basin depth)0.45
Combination Per-component: weighted geometric mean. Per-element: weighted geometric mean. Across elements: weighted geometric mean (floor 0.01). Weight perturbation: ±15% over 200 runs — Prospect Alpha ranked #1 in 97% of realizations.
07 Technical Specifications

Desktop. Offline. Your basin model stays on your machine.

Platform
Java 21 · JavaFX 25 · Windows, macOS, Linux. No cloud, no telemetry.
Spatial Engine
GeoTools 34.2 · GridCoverage2D · StreamingRenderer · SLD · 6,000+ EPSG CRS · on-the-fly reprojection.
Database
H2 2.x + H2GIS · single-file .pfa project · 317 attribute library entries across 5 tables (GDE 74, Lithology 81, Structure 64, Alteration 52, Hydrogeology 46) · 49 scoring contexts.
Hierarchy
4-level: Play → Element → Component → Proxy. Standard 5-element template + custom elements. Multiple named plays per project. Up to 50 scenarios with A/B comparison.
Scoring
5 continuous functions (monotonic, bell, sigmoid, piecewise-linear, Gaussian) + categorical lookup from 5 attribute libraries. GDE Library auto-populates reservoir, source, seal scores. All mappings per-scenario editable with override audit.
Interpolation
Gaussian decay · IDW · ordinary kriging · thin-plate spline · minimum curvature · anisotropy (azimuth + ratio).
Combination
Weighted geometric mean (default) · multiplication · minimum · OWA · weights-of-evidence. At every hierarchy level independently.
Confidence
KDE (Sheather–Jones) · kriging variance · weight-perturbation Monte Carlo · rank-stability reporting.
Import
Vector: Shapefile, GeoJSON, KML · Raster: GeoTIFF, ZMap+, ASCII, Surfer · Well: LAS 2.0/3.0, CSV · Tabular: CSV, TSV, Excel.
Audit
Full provenance: every scoring function, decay parameter, weight, and combination method serialized to H2 database as JSON. Export to DOCX or Markdown for peer review.
08 Get In Touch

The method is three decades old. The tooling finally catches up.

We are working with exploration teams, new ventures groups, and research partnerships. If you are evaluating a frontier basin, ranking a portfolio of leads, or building a regional prospectivity atlas — the PFA application is the tool that was missing.

Request a demo Download data sheet
Basin evaluation · New ventures · Portfolio ranking