Strata-intel · Energy module live · April 2026

A reasoning instrument for global commodity disruption.

Click any refinery, terminal, pipeline, project or chokepoint. Strata-intel computes how a disruption propagates through the global supply system — which back-up routes activate, how much demand goes unmet, and which kind of new equilibrium emerges.

Coverage312 entities
Countries scored43
Equilibrium archetypes6
Historical backtests1 of 4 shipped
Confidence frameworkVerified · Estimated · Modeled · Seed
StatusPhase 4 calibration in progress
§ 01
Four views, one engine
Atlas · Markets · Simulate · Engine

The platform is organised around four jobs. Each view interrogates the same dependency graph from a different angle — geographic, country-comparative, scenario-driven, or model-introspective. Click any of the four to enter the platform on that tab.

▸ TAB 01
Atlas
Explore infrastructure
Geographic atlas of refineries, terminals, pipelines, offshore & onshore projects, power plants and chokepoints. Toggle layers, filter by region, click any asset for full detail panel.
▸ TAB 02
Markets
Country & supply scoring
43 countries scored on net energy position, refining self-sufficiency, chokepoint exposure and supplier concentration. Sortable, filterable, with full per-country profile breakdowns.
▸ TAB 03
Simulate
Disruption scenarios
Arm one or many entities, set severity and duration, run. The engine returns active substitution routes, unmet demand, equilibrium archetype, and a full reasoning chain.
▸ TAB 04
Engine
Inspect · score · calibrate
Three sub-views into the model itself. Inspector (logic, parameters), Fragility (Monte Carlo probability), Calibration (engine-vs-reality backtest against historical events).
§ 02
A platform, modular by asset class
Energy → Metals → Critical → Ag

The dependency-graph engine is asset-agnostic by design. Energy ships first because it is the most consequential and best-instrumented. Each subsequent module reuses the engine; only the entity catalogue and substitution profiles change.

Live
Strata-intel · Energy
Oil & gas (offshore, onshore), refineries, LNG terminals, pipelines, power generation, chokepoints. 312 entities, 6 equilibrium archetypes, 43 countries scored.
April 2026
Q1 2027
Strata-intel · Critical Minerals
Lithium, cobalt, nickel, rare earths, graphite, manganese. Refining concentration risk (China's 80%+ share of refining for several REEs is the headline structural feature).
Q1 2027
2027
Strata-intel · Agriculture
Grains (wheat, corn, rice, soy), edible oils, fertilizer inputs (urea, potash, phosphate). Black Sea, Indian Ocean and Strait of Hormuz exposures map cleanly from energy module.
2027
§ 03
How the engine reasons
Three layers · separation by design

Strata-intel's reasoning is grounded in three separable layers. The engine (procedural code — how the system reasons) reads from logic (taxonomies, parameters, lens definitions) which reads from data points (entity declarations — what facts the system knows). All three are inspectable.

01 · Engine
A two-pass propagation algorithm walks a weighted dependency graph. Pass one: forward propagation from the disrupted entity through outputs and consumers. Pass two: reverse traversal from each importing country's portfolio. Outputs: active substitution routes, unmet demand by country, equilibrium archetype.
02 · Logic
Six equilibrium archetypes (price spike · inventory drawdown · freight inflation · regional shortage · Asian crack-spread · correlated compound). Default substitution profiles by entity type. Stakeholder lenses (importer, exporter, transit). Parameter distributions for Monte Carlo. All editable in MODEL_CONFIG.
03 · Data points
Per-entity declarations (capacity, dependencies, replacement options). Five v1.1 deeply-curated anchors: Strait of Hormuz, Suez/SUMED, Abqaiq, Houston Ship Channel, Jurong Island. Confidence tiers attached to every assertion — verified, estimated, modeled, or seed.
Confidence framework
Verified
Operator filings, government releases, audited disclosures. The standard for capacity, location, ownership.
Highest tier
Estimated
Reasoned from public attributes — trade press, industry analyst notes, secondary sources. Most numbers in the system.
Default tier
Modeled
Outputs of the engine itself — substitution profiles, equilibrium predictions, fragility probabilities.
Engine output
Seed
Placeholder values used during development. Visible to users and prioritised for replacement.
Replacement queue
§ 04
Honest about being uncalibrated
What we don't claim

Most engine parameters are currently reasoned-or-estimated, not evidenced. Phase 4 calibration against historical events (2019 Abqaiq, 2021 Suez, 2022 Russia, 2024 Houthi) is partially shipped — see the Calibration sub-view in the Engine tab. The platform's credibility comes from being explicit about what it doesn't know yet, not from claiming accuracy it hasn't earned.

Live data integration (AIS, market prices, weather), the full historical backtest corpus, automated calibration loop, and downloadable institutional reports are on the roadmap. The platform is a working sample, not a finished product. Buyers who value rigour over polish are the right buyers today.

Talk to the team.

Strata-intel is in early access for a small number of energy desks, sovereign wealth funds, and geopolitical risk teams. Tell us what you'd use it for and we will get back within 48 hours.

Direct ·  [email protected]
No tracking. We respond personally within 48 hours.
Strata-intel.
Energy module · v135 · April 2026
02 · COUNTRY SCORING

Energy Independence Atlas

43 countries scored on net energy position, import dependency, and chokepoint exposure. Click any country for full profile breakdown.
Sort by
Tier
#
Country
Region
Score
Independence (0–100)
Tier
Score methodology: weighted blend of net energy export position (oil + gas + electricity), refining self-sufficiency, chokepoint transit dependency, and supplier-concentration risk. Higher = more energy-independent. Sources: IEA WEB, BP Statistical Review, EIA, ENTSO-E, national energy ministries.
Inspector
How the engine reasonsread-only
The platform's reasoning is grounded in three separable layers: engine (procedural code · how the system reasons), logic (model decisions · taxonomies, parameters, lens definitions), and data points (entity declarations · what facts the system knows). This view inspects the logic layer.
Fragility analysis
Probability of equilibrium breachPhase 3
Monte Carlo simulation across the parameter distributions declared in MODEL_CONFIG.fragility_params. Each archetype carries its own distribution shape (μ, σ) from observed historical disruption magnitudes. Output is the probability that the new equilibrium clears within tolerance — and what regime it lands in.
Calibration
Engine vs. realityPhase 4
The engine's reasoning is honest only if it can be checked. This view replays curated historical disruption events through the engine and scores predictions against actual observed outcomes. Each backtest yields a quantitative match score plus actionable calibration insights — over-predictions, under-predictions, and gaps where the engine missed reality.
Layers
Click any entity for details. For country-level analysis, see Country Scoring tab.
Pick layers to choose which assets to disrupt. Click any visible asset → ⊕ Arm scenario. Multi-select supported for compound disruption.
Energy Independence
85+ Major exporter
70-84 Self-sufficient
50-69 Partially dependent
30-49 Highly dependent
<30 Critically dependent
Larger circle = higher score
Pipelines
Oil
Gas
Refined products
HVDC
Dashed = planned/idled
Terminals + STS
LNG export (square)
LNG regas (faded circle)
Crude export (circle)
Product / bunker (diamond)
STS zone (dashed circle)
Sample: ~60 entities. Live atlas backend will populate to 150+ globally. Includes shadow-fleet STS hubs (Laconia, Lome, Vostochny) with sanctions context.
▾ Scenario Builder
Click any asset → ⊕ Arm scenario → set severity + duration → run. Multi-select supported.
Armed for compound
Substitution flow
Unmet demand
Every typed entity is armable. Curated v1.1 entities use evidenced substitution paths; others use default profiles by entity type.
▾ SCENARIO BUILDER
Build any disruption you want to reason about
The platform's reasoning engine works on any asset you arm. Click an asset on the map, set severity and duration sliders, run. Multi-select supported — armed entities run as a compound scenario with congestion-aware substitution allocation and predicted price impact.
or browse pre-built examples
● click any asset to arm
● two sliders per entity
● multi-select for compound
● share via URL
Global Capex Timeline · 2026