Trap geometry and prospect identification

The anatomy of a trap

A trap is a subsurface configuration that prevents hydrocarbons from escaping to the surface. Traps require three physical elements:

• Reservoir rock: porous and permeable, so it can both HOLD hydrocarbons (porosity) and eventually DELIVER them to a well (permeability). Sandstones and carbonates are the two main classes.
• Seal: an overlying and laterally confining rock with effectively zero permeability to hydrocarbons at geological timescales. Shale is the most common seal; halite (salt) is the most effective; tight carbonates and anhydrite also seal.
• Geometric closure: a configuration such that hydrocarbons accumulated in the reservoir have nowhere to escape UPWARD (because of the seal) and nowhere to leak out to the SIDES (because the reservoir pinches or is terminated by a sealing structure).

Trap geometries are classified by how the closure is achieved:

• Four-way structural closure (dome): the reservoir forms a topographic high. Hydrocarbons fill the top down to a spill point. ANY fault presence or stratigraphic pinchout is not required. This is the lowest-risk trap because only one element (the dome) must be correct. Textbook examples: many Middle East supergiants, most supra-salt drape folds.
• Three-way structural closure: the reservoir topography closes on three sides; a FAULT closes the fourth side. Risk doubles because BOTH the structure AND the fault seal must work. Fault-sealed tilt blocks and horst blocks against extensional faults are classic three-ways. Analysis: see §3.2 fault-seal treatment.
• Stratigraphic trap: the reservoir PINCHES OUT laterally (e.g., sand becoming shale) or is terminated by an unconformity (older reservoir overlain by younger seal at an unconformable surface). No structural closure required, but much harder to identify on seismic. Turbidite fans, shoreline sands, reef buildups — these are the hallmark stratigraphic plays.
• Combination trap: a structural closure combined with stratigraphic elements. Most real-world traps are combinations. A dome whose top is eroded at an unconformity, or a three-way with a sand pinchout on the fourth side.
• Hydrodynamic trap: unusual; flowing aquifer tilts the OWC or GWC off-horizontal, so hydrocarbons are held in structures that would normally be too subtle to trap statically. Rare but real (Pembina, Alberta is a partial example).

In practice, exploration geologists think in terms of combinations and the dominant trap style — the one that carries most of the risk and most of the volume — but the actual trap is usually a blend.

Volumetrics: how much is there?

Once you have a mapped trap, you estimate its hydrocarbon volume using the Stock-Tank Oil In Place (STOIIP) equation:

Where:

• A = trap area at the horizon surface (km² or acres)
• h = average net pay thickness (m or ft)
• NTG = net-to-gross ratio: fraction of the total interval that is actually reservoir-quality rock (0 to 1)
• φ = effective porosity: fraction of the reservoir that is pore space (typically 0.05 to 0.35)
• S_w = water saturation: fraction of the pore space filled with water. Hydrocarbon saturation = 1 - S_w. Typically 0.15 to 0.60 for oil reservoirs.
• B_o = formation volume factor: ratio of oil volume at reservoir conditions (high-pressure, containing dissolved gas) to oil volume at surface stock-tank conditions. Typically 1.05 to 1.6 for black oils.

Each parameter is multiplied, so each is equally leveraged on the result. But each is known with different uncertainty: area from structure maps may be ±20%, porosity from logs may be ±25%, NTG from core/log may be ±30%, water saturation from resistivity may be ±20% depending on petrophysical quality. A full volumetric analysis carries uncertainty through as a distribution (e.g., Monte Carlo with P10, P50, P90 outputs), not as a single deterministic number. The STOIIP number you report is the P50 (median) expectation.

Recoverable reserves = STOIIP × Recovery Factor (RF). RF is the fraction of in-place oil that can be produced economically, depending on drive mechanism, well spacing, and facilities. Typical ranges: 5–15% for primary depletion of oil reservoirs, 25–45% with waterflood, 45–65% with effective gas injection or thermal EOR. Gas reservoirs have higher recovery factors (typically 60–85%) because gas expands as pressure drops.

Exercise — working with the prospect evaluator

• The widget starts with a medium-sized prospect: 20 km² area, 40 m pay, standard reservoir properties. Note the UNRISKED RESERVES and POS at the top. These are the two numbers that matter for ranking.
• Drag the trap area slider up to 100 km². Reserves scale linearly. Now drag it back down to 20 and net pay up to 200 m. Same effect. Volumetrics is linear in all its inputs — a doubling anywhere is a doubling of the result.
• Reset the volumetrics. Now focus on the five RISK sliders. Each starts at 0.65–0.80 — the typical starting point for a prospect of moderate maturity. The POS at the top is the PRODUCT of these five. Note that the running POS drops BELOW all individual risks because probabilities multiply, not add.
• Now drag Charge down to 0.20 (suppose the source rock is unproven). Watch how the POS collapses — from ~0.19 to ~0.06. One weak element kills the prospect even if everything else is strong. This is why exploration companies invest heavily in PROVING source rocks before bidding for blocks.
• Reset all risks to 0.50 (maximum uncertainty for each element independently). POS = 0.5⁵ = 0.03, or only 3%. Five unknown risks collapse to near-zero success. This is the fundamental difficulty of frontier exploration: unless you CAN prove some of the elements, the arithmetic is against you.
• Inspect the cascade plot: each green bar shows the running probability after that risk has been applied. The LOST portion (hatched red) is what the bar chopped off. Compare the cascade between a 0.9/0.9/0.9/0.9/0.9 prospect (POS = 0.59) and a 0.5/0.5/0.5/0.5/0.5 prospect (POS = 0.03). Same average, order-of-magnitude different POS.

The five risks in detail

Exploration risk is conventionally decomposed into five INDEPENDENT elements, each with its own probability of success:

• Structure (trap integrity, P ~0.7–0.95 for well-imaged closures): Does the mapped trap actually CLOSE? Is the depth conversion reliable? Did faults breach the seal vertically? Known as "trap risk" in some conventions. Low when seismic is poor or velocity model is suspect; high when the structure has been drilled nearby or is visible in high-quality 3D.
• Reservoir (P ~0.5–0.95): Is there a rock with sufficient porosity (≥5%) and permeability (≥1 mD) to hold and deliver hydrocarbons? Depends on the depositional system, diagenetic history, depth, and age. Well-calibrated from nearby producing wells; poorly-calibrated in frontier plays.
• Seal (P ~0.6–0.95): Does something prevent the hydrocarbons from leaking out the top? Regional shale seals, salt seals, and tight carbonates are the usual suspects. Seal risk is often HIGH when the seal is thin, fractured, or juxtaposed against permeable units by faulting.
• Charge (P ~0.3–0.9): Is there a mature source rock that has generated hydrocarbons? Has the generation window been reached (thermal maturation)? Is the volume of generated hydrocarbons enough to fill the trap? Charge risk is typically the HIGHEST in frontier basins (no proven source) and the LOWEST in mature basins (many producing fields nearby prove charge works).
• Migration / Timing (P ~0.5–0.95): Did the hydrocarbons migrate FROM the source rock TO the trap? Was there a continuous pathway (carrier bed, fault, or up-dip continuity)? And critically, did the migration happen AFTER the trap was formed (timing)? A trap that formed AFTER migration is dry.

The multiplicative combination is critical:

This is valid under the INDEPENDENCE assumption: the failure of one element does not cause (or correlate with) the failure of another. In practice some correlations exist (e.g., seal failure and structural failure often share a common cause in highly faulted areas), and practitioners sometimes add a correlation adjustment. But for 80–90% of prospects the five-element multiplicative framework is a reasonable first approximation.

Why multiplication kills you: five independent 70% probabilities collapse to 0.7⁵ = 0.168, or 16.8% POS. Five 80% probabilities give 0.8⁵ = 0.328, or 32.8%. Each additional UNCERTAIN element brutally compounds the overall risk. This is why mature basins (where most elements are already proven) win over frontier basins (where many elements are unknown) — the arithmetic of compounding probabilities favors ALREADY DEMONSTRATED petroleum systems.

Prospect ranking and the risk–reward trade-off

A single prospect has two numbers: the unrisked reserves (upside) and the POS (likelihood of realizing any upside at all). A portfolio manager ranks prospects by expected value = unrisked reserves × POS, which is the "risked reserves" in the widget. Two prospects can have the same risked reserves by very different paths:

• Prospect A: 500 MMbbl unrisked, 10% POS → 50 MMbbl risked. A supergiant upside, but 90% chance of dry. Called a "moonshot."
• Prospect B: 100 MMbbl unrisked, 50% POS → 50 MMbbl risked. A more modest upside, but even chance of success. Called a "solid prospect."

Both have the same expected value, but a rational exploration company prefers Prospect B because:

• Variance: Prospect B's outcome is more predictable. Prospect A drills five wells before an expected hit; four of those are expensive failures.
• Information value: a failed Prospect A well teaches little; a successful Prospect B well opens a whole trend.
• Portfolio impact: across 10 prospects, the Prospect-A portfolio has high variance and risks a string of failures; the Prospect-B portfolio has more predictable outcomes.

BUT: Prospect A may be the only way to find a truly giant field that transforms a company, so large exploration companies maintain a mix — several Prospect-B prospects to pay the bills, and a few Prospect-A moonshots for the huge upside. This is called a prospect inventory, and managing it is a key competency of exploration management.

Volumetric and risk pitfalls

• Overestimating area. A common error: using the GRV (gross rock volume) area instead of the net pay area. The NET pay area is smaller — only the part of the trap with reservoir-quality rock AND with hydrocarbons (above the OWC/GWC). Always distinguish GRV area from net pay area.
• Double-counting risk. If reservoir risk is 0.6 and charge risk is 0.6, multiplying gives 0.36 — appropriate IF the two are independent. But if the charge AND reservoir are both controlled by the same depositional event (e.g., an organic-rich source rock and its coeval turbidite reservoir), they are correlated, and the correct POS is higher than the naive product. Adjust for correlation when it exists.
• Not distinguishing geological from commercial risk. POS as discussed here is GEOLOGICAL (will there be hydrocarbons in the trap?). Commercial risk includes project economics (oil price), political risk (fiscal terms), and execution risk (can we actually drill?). Companies track these separately.
• Using deterministic numbers when distributions are warranted. Area, pay, porosity, etc., are all uncertain. A deterministic STOIIP = 250 MMbbl looks precise but misleads; a P50 of 250 with P10 = 800 and P90 = 60 is a more honest characterization. For major prospects, always use Monte Carlo simulation.
• Bias in the POS assessment. Exploration geologists are notoriously overconfident. Studies of post-drill outcomes show that actual success rates are 30–50% LOWER than pre-drill POS estimates across mature basins. Calibrate your POS against historical drill outcomes, not against your confidence in your own mapping.
• Treating stratigraphic traps like structural traps. Stratigraphic trap identification carries MUCH higher risk than structural trap identification because the closure is invisible on conventional seismic. Adjust reservoir + seal + structure risks accordingly (structure risk for a stratigraphic trap can be 0.3–0.5 even when the prospect "looks good").