Reservoir model and uncertainty: closing the QI loop

The STOOIP formula — what each factor depends on

Stock-tank original oil in place, in convenient mixed units:

**

**

Each factor comes from a different upstream input:

• Area: from the structural map (Parts 2–3). Ranges are driven by the spill-point and contact-depth uncertainty.
• NetPay (m): from the probabilistic net-pay cube (§7.4). Sum along depth within the reservoir horizon, weighted by P(net pay) per voxel.
• φ (porosity): from the φ cube (§7.4), volume-averaged over the net-pay voxels.
• (1-Sw) hydrocarbon saturation: from the Sw cube (§7.4), volume-averaged. Most uncertain of the rock-property terms because Sw is the least inversion-robust.
• Bo (formation volume factor): from reservoir-engineering PVT analysis. Fixed per reservoir, ~1.1–1.5 for oil, varies with solution gas.

Each factor is ITSELF uncertain. Treating them as single values (the classic “deterministic volumetrics” approach) hides the uncertainty. Monte Carlo volumetrics samples from the distributions of each factor and runs the STOOIP formula thousands of times, generating the full distribution of possible answers.

Exercise — cycle through the three QI scenarios

• Open the widget in Baseline QI project scenario, Monte Carlo forecast view. You see a histogram of STOOIP values from 8000 Monte Carlo draws. Orange bars inside the P10-P90 band; grey bars outside. Notice the distribution is LOG-NORMAL-LIKE — skewed to the right, with a long upside tail. Most reservoirs are modest; a few are home runs. This is a CHARACTERISTIC shape of volumetric distributions.
• Read the summary line: P50 is the MEDIAN (50% chance STOOIP is at least this). P10 is the CONSERVATIVE estimate (90% chance STOOIP is at LEAST this). P90 is the OPTIMISTIC (only 10% chance STOOIP exceeds this). The P90/P10 ratio quantifies uncertainty: close to 1.0 = tight; 3.0+ = wide.
• Switch to the Frontier exploration scenario. The histogram BROADENS dramatically — P10 drops, P90 climbs, ratio balloons. Same modal value (5 km², 25 m net pay) but the UNCERTAINTY around every input is much wider. This is what it feels like to make drill decisions on limited calibration data.
• Switch to Developed field. The histogram COLLAPSES — tight distribution, P90/P10 ratio approaches 1. This is the reward for years of well drilling + QI calibration: you KNOW where the barrels are.
• Switch to Tornado sensitivity view (still in the Baseline scenario). Four horizontal bars, sorted largest-to-smallest. The TOP bar identifies the parameter whose min-to-max excursion moves STOOIP the most. For baseline QI, this is typically AREA or NET PAY (the big volumetric drivers). Porosity moves STOOIP moderately; hydrocarbon saturation can move it a lot if its range is wide.
• Flip the tornado between scenarios. In frontier, all bars are LONG — every parameter is wildly uncertain. In developed, all bars are SHORT. Compare the RANKING: which parameter stays at the top across all scenarios? That’s usually Area — structural uncertainty tends to dominate even when QI is good, because spill-point depth is ultimately a structural/stratigraphic question, not purely QI.
• Strategic use of the tornado: the TOP bar is where to invest next. If NET PAY dominates, commission more wells to pin down reservoir thickness. If AREA dominates, fund higher-resolution structural imaging. If Sw dominates, invest in pre-stack inversion or additional angle stacks. Tornado diagrams turn vague “uncertainty” into actionable “here’s where to spend the next million dollars.”

From P10/P50/P90 to business decisions

The Monte Carlo STOOIP distribution feeds directly into financial decisions:

• Reserves booking (SEC guidelines): P50 = PROVEN + PROBABLE (2P) reserves. P90 = PROVEN (1P) reserves. P10 = PROVEN + PROBABLE + POSSIBLE (3P). Each category has different tax and financial treatment; the distribution’s shape drives what the company can legally book.
• Field development decision (FID): management compares P50 STOOIP × recovery factor × oil price against the capital cost + operating cost discounted over the field life. Approvals usually require P50 net present value > 0 at a pessimistic oil-price scenario.
• Risk-weighted economics: probability-of-commercial (POC) = P(STOOIP > minimum economic threshold). Minimum economic is usually the size that covers the drilling + facilities cost. The POC times the value-if-success minus the cost-if-failure is the EXPECTED VALUE used for portfolio decisions.
• Partner negotiations: farm-out terms, farm-in bids, and asset trades all reference P10/P50/P90 STOOIP. Your counterparty has their own estimate; the negotiation is partly about whose estimate governs.
• Drilling strategy: the VARIANCE of the STOOIP distribution (not just the mean) drives how many wells to drill up-front. High variance = drill one well first, get more information, then decide scale. Low variance = commit to full-field development immediately.

Why the tornado is the most cost-effective diagnostic

A tornado chart answers the most actionable question: “Where do I spend the NEXT million dollars of my reservoir-uncertainty-reduction budget?” The answer is always: the parameter with the longest bar. A few examples:

• Net pay is the top bar: drill more wells, rerun inversion with more calibration wells, or acquire higher-resolution seismic to beat the tuning thickness.
• Area is the top bar: invest in structural imaging — PSTM/PSDM reprocessing, wide-azimuth acquisition, or additional 3D coverage around the prospect edges.
• Porosity is the top bar: recalibrate the rock-physics template, or augment the calibration with core data that ties porosity to log response more tightly.
• Hydrocarbon saturation is the top bar: run pre-stack inversion (if you only have post-stack), or add angle stacks to a pre-stack inversion to improve Vp/Vs resolution.

The tornado doesn’t just show you uncertainty. It shows you the LEVERAGE your investments have on that uncertainty. Every asset manager should be able to read one and know what to do.

The end-to-end QI value story

Zooming out to Part 7 as a whole, the value chain is:

• Seismic acquisition → wiggly traces over a 3D volume.
• §7.3 inversion → Ip, Is, ρ cubes (elastic properties per voxel, with uncertainty).
• §7.4 transforms → φ, Vsh, Sw cubes (rock properties per voxel, with uncertainty).
• §7.5 probabilistic facies → per-class probability cubes (preserving classification uncertainty).
• §7.6 Monte Carlo volumetrics → STOOIP distribution + tornado diagnostic.
• Business layer (outside QI) → FID, reserves booking, portfolio ranking, drilling program.

Each stage adds value AND inherits uncertainty from the previous stage. The Part 7 discipline is to manage uncertainty all the way through — not to hide it with deterministic shortcuts. A team that propagates uncertainty honestly delivers a MORE TRUSTWORTHY product, even if the headline numbers are less dramatic than a deterministic treatment would produce. And trust is, ultimately, what lets asset teams spend capital.

What comes next: Part 8 — advanced topics

With Part 7 complete, you have the core QI workflow end-to-end. Part 8 (outside the scope of this section but the logical continuation) covers advanced and specialized topics:

• Geomechanics: elastic cubes into stress and fracture prediction.
• Time-lapse (4D) QI: repeat surveys to track production-induced pressure/saturation changes.
• Anisotropy and azimuthal attributes: when the isotropic assumption breaks.
• Full-waveform inversion (FWI): combining velocity model building with QI.
• Machine-learning QI: neural-net inversion, deep-learning classifiers, feature engineering.
• CO2 storage QI: monitoring injected CO2 plumes through changes in Ip, Vp/Vs, and density.

All of them build on the Part 7 foundation: you now speak the language of QI — probabilistic, uncertainty-aware, rock-physics-grounded. Every advanced topic is a specialized extension, not a replacement, of what you’ve learned here.