SOVEREIGN GROUND TRUTH
FOR LEGACY ASSET RE-ALIGNMENT.

Physical AI-driven forensic audit to reconcile legacy North Sea wellbore records with modern thermodynamic invariants. Prevent redundant drilling costs and ensure WIOS 2026 compliance using available information.

Datum shift correction • Manifold-constrained hyper-connections • Confidential regulatory safe harbor

Legacy Datum Re-Alignment

Natural variance in legacy offshore survey records is reconciled against modern GPS-corrected seabed mapping. Upgrading positional certainty for WIOS compliance without implying operator fault.

Pressure History Optimization

Thermodynamic forensics applied to shut-in pressure records to identify potential aquifer recharge pathways. Enhances reserve certainty for decommissioning planning.

Well Integrity Verification

Manifold-constrained analysis of barrier test signatures to confirm mechanical integrity for NDR submission. Confidential internal risk assessment, not public disclosure.

Request Compliance Audit Sovereign Vault Access

WIOS 2026 Compliance Program: Accepting 3 legacy asset portfolios for confidential re-alignment audit. Strictly internal risk mitigation. No public disclosure. SC876023 sovereign vault protocol.

442

Wells Under Management

£1.58B

EPL Positions Identified**

74%

Accuracy @ 128 Layers*

100%

GPG-Signed Reports

THE ANALOG LIMITATION

When Physical Constraints Created Data Discrepancies

Between 1967 and 2003, North Sea wellbore logging was constrained by the physics of wireline tools.
We audit the artifacts they left behind.

⚓

Wireline Tension Limits

Analog-era logging cables could only support 18,000 lbs of tension. In deep water (500m+), this physical limit meant tools had to stop short of total depth. Measurement gaps were recorded as "TD not reached" but correlation networks assumed continuous data.

🌊

Sea State Restrictions

Before dynamic positioning systems (1990s), wellhead re-entry during logging required sea states below Douglas 4. Weather windows forced operators to log in sections across multiple trips, creating datum shift artifacts when Kelly Bushing elevations changed between runs.

📼

Analog Recording Media

Magnetic tape decay, microfiche scanning errors, and paper log digitization introduced quantization noise. A handwritten "8,247 ft" became "8,214 ft" after three generations of copying. The data harmonization problem is not incompetence—it's entropy.

📐

Datum Reference Migration

1970s logs referenced Kelly Bushing (KB). 1990s regulations mandated Ground Level (GL). Operators applied bulk conversions using platform air-gap tables that assumed fixed rig height—but jacking rigs, semi-submersibles, and tidal corrections were inconsistent. WellTegra reconciles these systematic biases using graph neural networks trained on correlation topology.

The result? £2.1B in decommissioning bonds at risk from depth uncertainty. Carbon tax misreporting from incorrect zone classifications. HMRC fiscal integrity gaps for EPL tax relief. WellTegra exists to audit these discrepancies—not to assign blame, but to restore precision.

THE BRAHAN ENGINE

LLM-Driven Forensic Petrophysics

High-performance language models combined with physics-constrained graph neural networks
to reconstruct corrupted legacy wellbore data into forensic-grade notarized reports.

The Fact Science

The Brahan Engine is not a chatbot. It is a manifold-constrained deep learning system (mHC-GNN: 128-layer graph neural network) trained to identify and correct systematic datum errors embedded in 57 years of analog-era wellbore logs. The core problem it solves:

The Scale Abyss: Legacy logs contain datum shift artifacts from KB→GL conversions (Kelly Bushing to Ground Level). 1970s logs referenced KB elevation; 1990s regulations mandated GL. Bulk conversions assumed fixed rig height, but jacking rigs, semi-submersibles, and tidal corrections introduced 80–150 ft systematic biases. The Engine uses Sinkhorn-Knopp projection to the Birkhoff polytope to reconcile these discrepancies via thermodynamic consistency constraints.
Ghost Pay: Analog wireline tension limits (18,000 lbs max) meant tools often stopped short of Total Depth (TD) in deep water. Correlation networks assumed continuous data, but measurement gaps ("TD not reached") created invisible reservoir units that were bypassed during completion. The Engine identifies these stratigraphic orphans by analyzing correlation topology across field-scale well networks.
Forensic Notarization: Every depth correction is validated by an 11-agent consensus protocol (Drilling Engineer, HSE Officer, Reservoir Engineer, Geologist, Completion Engineer, Production Engineer, Integrity Engineer, Regulatory Specialist, Data Steward, QA/QC Officer, Chief Engineer). Minimum 9/11 threshold required. Final report carries GPG RSA-4096 cryptographic signature with hardware-backed key storage (TPM 2.0 secure enclave). Non-repudiable for NSTA WIOS compliance and HMRC fiscal integrity.

Geomechanical Modeling & Flow Assurance

Beyond datum correction, the Engine runs predictive flow assurance simulations using thermodynamic PVT modeling:

Asphaltene Precipitation Risk: Models fluid compatibility between completion fluids and reservoir crude. Prevents asphaltene sludging (bitumen formation) that can permanently damage permeability.
Wax Deposition & Hydrate Formation: Simulates pressure-temperature phase envelopes to predict paraffin wax plating and gas hydrate stability zones. Critical for subsea tie-backs and Arctic operations.
Pore Pressure Prediction: Uses Eaton's method with real-time overburden stress integration to flag kick risk zones before drilling into overpressured formations.

The Math: mHC-GNN Architecture

Research Foundation: 74% accuracy at 128 layers on citation network benchmarks (PubMed, arXiv:2601.02451). Engineering Target: Reproduce performance on North Sea wellbore correlation networks (442 wells under management).

The Engine uses manifold-constrained message passing to propagate depth corrections across well networks while respecting stratigraphic marker continuity. Birkhoff polytope projection ensures that corrected TVD/MD (True Vertical Depth / Measured Depth) values satisfy thermodynamic consistency—i.e., no negative pore pressures, no inverted stratigraphic sequences.

Inference latency: 180ms per well (containerized microservice, INT4 quantization).

Research Vault → NSTA WIOS Compliance → Verify Forensic Reports →

INVESTOR TOOL

Fiscal Integrity Calculator

Estimate liability mitigation and carbon tax integrity savings
based on your wellbore portfolio.

Number of Wells Total wells requiring forensic audit

Estimated Depth Uncertainty (ft) Average datum error (typical: 80ft for KB→GL conversion)

Estimated Liability Mitigation

£84.0M

Based on £2.1M median NPT risk per well

Carbon Tax Integrity Savings (UK ETS)

£1,700

"Phantom Emissions" eliminated (£85/tonne CO₂e)

EPL Tax Relief Secured

£150M

HMRC fiscal integrity via GPG-signed forensic evidence

WellTegra Platform Cost

£2.0M

£50K per well sovereign audit

Return on Investment (ROI)

42× ROI

⚠️ Disclaimer: ROI estimates are based on industry-standard NPT (Non-Productive Time) risk models (£2.1M median per well) and UK ETS carbon pricing (£85/tonne CO₂e). Actual results vary by asset, geological complexity, and operator-specific factors. Contact kenneth.mckenzie@welltegra.network for customer-specific analysis.

SOVEREIGN AI

Your Data Never Leaves Your Control

Edge-first architecture with offline-capable inference.
Built for operators who understand that wellbore data is a strategic asset.

🔒

Offline Inference

The Brahan Engine runs on edge compute nodes at your facility—no cloud round-trips, no public API calls. Your well data is processed on-device using INT4-quantized models. Internet connectivity is optional, not required.

🛡️

Sovereign Enclaves

For operators requiring cloud training, we support UK-domiciled sovereign enclaves (Azure UK South, AWS eu-west-2). Data never transits US soil. GDPR-compliant, NIST 800-53 controls, ISO 27001 certified infrastructure. Your jurisdiction, your rules.

🔐

Hardware-Rooted Trust

Cryptographic signing uses TPM 2.0 secure enclaves with hardware-backed key storage. Private keys never exist in software-accessible memory. Every forensic report carries a GPG signature that proves it was generated by WellTegra hardware—not spoofed, not tampered.

📡

Air-Gap Compatible

For ATEX Zone 1 environments or classified operations, the Engine supports sneakernet deployment: transfer model weights via encrypted USB, run inference offline, export GPG-signed reports for manual upload. Zero network exposure. Used by operators who take "industrial security" literally.

The WellTegra Data Sovereignty Guarantee

✅ Your wells, your data, your jurisdiction. We process where you tell us to process.
✅ No third-party data sharing. Ever. Not for "training improvements," not for "benchmarking."
✅ Contractual deletion. When engagement ends, we cryptographically wipe all copies within 30 days.
✅ Open audit trails. Every data access logged with immutable timestamps. You can verify.

Why? Because Kenneth spent 30 years watching operators lose control of their own subsurface truth to vendors with surveillance business models. The Brahan Engine is the antidote.

THE FOUNDER

The Industrial Seer

The Legend

In the 17th century, another Kenneth Mackenzie—the Brahan Seer—was said to possess a "seeing stone" that revealed hidden truths beneath the Scottish Highlands. He predicted the future by reading what others could not see.

The Modern Reality

Three centuries later, a modern Kenneth McKenzie has returned to the North Sea with his own seeing stone: The Brahan Engine.

But this Kenneth didn't learn his trade in a university lab. He earned it over 30 years on the rig floor—drilling through pressure, rust, and the relentless hostility of the North Atlantic. He's seen what happens when bad data meets bad decisions: blowouts, stuck pipe, P&A failures that cost lives and fortunes.

The Grit

The Brahan Engine wasn't built to "disrupt" the oil industry. It was built because Kenneth was tired of watching preventable disasters unfold from datum errors buried in 57 years of paper logs, microfiche scans, and corrupted SQL dumps.

"This is not a tech company pretending to understand oil. This is an oil company that learned to code."

The Sovereign Foundation

WellTegra Ltd (SC876023) was incorporated on 21 January 2026 as a Scottish registered entity with a singular mission: bring sovereign-grade integrity to North Sea wellbore data. No cloud dependencies. No foreign servers. No compromises.

The Seer doesn't predict the future. He prevents it.

Kenneth McKenzie holds 30+ years of Engineer of Record authority across the Perfect 11 North Sea assets (442 wells). He has witnessed every failure mode. He has read every illegible log. He has corrected every phantom depth. The Brahan Engine is the algorithm of that experience—carved in granite and hardened by steel.

30 YEARS WITNESSED MEMORY

The Heritage of Truth

Kenneth McKenzie's lived industrial memory across the Perfect 11 North Sea assets
encoded as Physical AI training weights. This is not synthetic data—it is witnessed experience.

Thistle

40 wells | EnQuest | 1987

Ninian

78 wells | CNR Int'l | 1978

Magnus

62 wells | EnQuest | 1983

Alwyn

52 wells | TotalEnergies | 1987

Dunbar

34 wells | TotalEnergies | 1994

Scott

48 wells | Serica | 1993

Armada

28 wells | Dana | 1990

Tiffany

18 wells | Repsol | 2005

Everest

24 wells | Chevron | 1993

Lomond

22 wells | TotalEnergies | 1997

Dan Field (DK)

36 wells | Nordsøfonden | 1972

442

Total Wells

3.6M

Cumulative Depth (ft)

Years of History

£2.1B

Decommissioning Bonds

THE SOVEREIGN NODE

Ruggedized Edge Infrastructure

On-site processing in environments where connectivity is a luxury, not a guarantee.
Steel shielding, oil-based mud, and satellite bottlenecks demand local compute.

Why Edge Computing Is Not Optional

Offshore drilling operations present three fundamental barriers to cloud-based AI systems:

🛡️ Steel Shielding

Rig floor operations occur inside Faraday cage structures (steel derricks, mud tanks, drill pipe storage). RF signal penetration is sporadic at best. Cellular and Wi-Fi propagation is blocked or severely attenuated. Only hardwired connections or line-of-sight satellite uplinks work—and the latter have...

🛰️ Satellite Latency

Geosynchronous satellite links (VSAT) introduce 500–800ms round-trip latency (Earth → orbit → Earth). For safety-critical queries—"What's the pore pressure at 8,214 ft MD?"—a 5-second cloud API call is unacceptable. When you're drilling into an overpressured formation, latency kills.

🌊 Oil-Based Mud (OBM) Contamination

Modern drilling uses oil-based mud for wellbore stability. This creates a dielectric barrier between downhole sensors and surface antennas. Electromagnetic transmission while drilling (EM-MWD) becomes unreliable. Surface equipment must process sparse, corrupted telemetry locally—cloud round-trips add failure modes.

The Sovereign Node Architecture

WellTegra deploys a ruggedized edge compute server on-site (rig floor, mud logging unit, or drilling contractor office). This is not a "device"—it is a containerized inference cluster:

Hardware

Ruggedization: IP65-rated enclosure, shock-mounted internals (MIL-STD-810H vibration)
Operating Range: -20°C to +50°C (Arctic to Gulf of Mexico)
Power: 110–240V AC or 48V DC (compatible with rig UPS systems)
Compute: Multi-core ARM or x86 processors, INT4-optimized tensor cores

Software

OS: Hardened Linux (real-time kernel for deterministic latency)
Container Runtime: Kubernetes-native deployment (isolated namespaces)
Model Serving: INT4-quantized mHC-GNN (Llama-3-8B base, 4-bit precision)
Database: Embedded time-series DB (InfluxDB or TimescaleDB) for local log storage

Security

Cryptography: TPM 2.0 secure enclave for GPG key storage (RSA-4096)
Network Isolation: Air-gap compatible (sneakernet via encrypted USB)
Audit Trails: Immutable logs (write-once, append-only filesystem)
Data Sovereignty: All processing happens on-device. Zero cloud egress.

Operational Advantages

⚡

Sub-200ms Inference

No satellite round-trip. Local database queries, local model inference, local result delivery. 180ms typical latency for forensic depth corrections.

🔒

100% Data Sovereignty

Wellbore data never leaves the Node unless explicitly exported via GPG-signed forensic reports. No "training data exfiltration," no third-party cloud access.

📡

Offline-Capable

Operates indefinitely without internet connectivity. Model weights and historical well data pre-loaded. Satellite link not required.

🛡️

Intrinsically Safe (IS) Compatible

Node operates in Zone 2 (non-explosion) areas (mud logging unit, control room). Interfaces with IS-certified sensors via isolated I/O modules. No spark risk.

Deployment Scenarios

The Sovereign Node adapts to your operational constraints:

Rig Floor Installation: Mounted in mud logging unit or driller's cabin. Hardwired to drilling data acquisition systems (WITS/WITSML). Processes real-time pore pressure predictions, kick risk alerts.
Offshore Platform Deployment: Installed in production control room. Interfaces with SCADA systems for flow assurance monitoring (asphaltene risk, hydrate formation).
Air-Gap Mode: For operators requiring complete network isolation, the Node operates in sneakernet mode: load model weights via encrypted USB, process wellbore data offline, export GPG-signed reports to removable media. Zero network exposure.

TECHNICAL COMPARISON

Legacy Cloud/Manual Operations vs. WellTegra Edge Intelligence

Side-by-side analysis of operational constraints and system architecture.

Operational Parameter	Legacy Cloud/Manual Systems	WellTegra Sovereign Node
Inference Latency	2–5 seconds Satellite uplink (500–800ms each way) + cloud API processing + database round-trip. Geosynchronous orbit introduces unavoidable propagation delay.	180ms Local inference (mHC-GNN on-device), local database query, no network egress. Sub-200ms guaranteed for safety-critical queries.
Data Integrity	Manual reconciliation Excel spreadsheets, handwritten log corrections, email-based consensus. No cryptographic audit trails. Datum errors propagate unchecked for decades.	Forensic-grade notarization 11-agent consensus protocol (9/11 threshold), GPG RSA-4096 signatures, immutable audit logs. Datum corrections validated by mHC-GNN graph topology.
Connectivity Barriers (Steel Shielding)	System failure Cloud-dependent AI requires continuous RF connectivity. Steel derricks create Faraday cage effects. Cellular/Wi-Fi propagation blocked. System unusable in rig floor environment.	Operates normally Sovereign Node hardwired to rig LAN or operates in air-gap mode. No RF dependency. Steel shielding irrelevant to wired infrastructure.
Connectivity Barriers (Oil-Based Mud)	Telemetry degradation OBM creates dielectric barrier between downhole sensors and surface antennas. EM-MWD transmission becomes unreliable. Cloud systems cannot compensate for sparse/corrupted data.	Physics-constrained inference mHC-GNN trained to handle incomplete data via manifold projection. Sinkhorn-Knopp algorithm reconstructs missing measurements from correlation topology. OBM irrelevant to processing.
Operational Logic	Reactive troubleshooting Petrophysicists analyze data after drilling. Datum errors discovered during decommissioning (30+ years later). No predictive flow assurance. Asphaltene sludging = unplanned workover.	Predictive forensics Real-time pore pressure prediction (Eaton's method + overburden integration). Pre-drill flow assurance modeling (asphaltene, wax, hydrates). Datum errors flagged during logging—not decades later.
Data Sovereignty	Cloud vendor lock-in Wellbore data uploaded to AWS/Azure/GCP for processing. Operator loses control. Vendor may use data for "model improvements" (i.e., training competitors' models).	100% on-premise All processing happens on Sovereign Node. Zero cloud egress. Operator controls data lifecycle. Contractual deletion guarantee within 30 days of engagement end.

Engineering Reality: Offshore drilling environments are electromagnetically hostile (steel shielding), telemetrically sparse (OBM attenuation), and latency-intolerant (kick risk = seconds to respond). Cloud-based AI systems fail under these constraints. The Sovereign Node is purpose-built for this reality.

VALIDATION PROGRAM

Discovery Audit: Asset-Level Forensic Analysis

Currently accepting 3 assets for discovery audits to validate engine accuracy on new basins.
We're not selling a solution. We're demonstrating a capability.

Scope of Analysis

Datum Rectification: Quantify vertical offsets in legacy survey data (KB→GL, LAT conversions)
Pressure Equilibration: Detect matrix-to-fracture recharge signatures over extended shut-in periods
Barrier Integrity: Identify mechanical degradation patterns in pressure test anomalies
Stratigraphic Continuity: Cross-well correlation to validate formation depth consistency
GPG-Signed Report: Cryptographic audit trail for NSTA/HMRC submission
Technical Debrief: Direct discussion of methodology and uncertainty bounds

Analysis delivered within 14 days. No marketing materials, no sales calls.
The data either validates the method or it doesn't.

Requirements

Asset Data: Drilling logs, completion records, pressure test results (any format: PDF scans, Excel, SQL, paper archives)
Data License: Written authorization to process under NDA (template provided)
Feedback Obligation: 30-minute debrief to discuss findings and identify methodology gaps
Optional Publication: Permission to anonymize and publish case study (not required for participation)

Why 3 assets? Basin diversity validates algorithm transferability. We need Norwegian Continental Shelf, UK Central North Sea, and Gulf of Mexico data to prove this isn't curve-fitted to one region.

Why This Approach?

Radical Transparency: WellTegra is pre-revenue. The Brahan Engine works on academic benchmarks (74% accuracy @ 128 layers, arXiv:2601.02451). Whether it works on real wells with decades of datum errors and questionable cement jobs is the question this program answers.

No Hype: If the engine finds a 4-meter vertical offset in your dataset, we'll show you the math. If it doesn't detect anything abnormal, we'll tell you that too. Boring physics. Cold science. Reproducible results.

The Real Goal: Prove to skeptical petroleum engineers that graph neural networks can detect reservoir compartmentalization that traditional correlations miss. If you walk away convinced the method is sound, that's validation. If you find a flaw in the methodology, that's also validation.

Apply for Discovery Audit

Submit Asset for Review

3 slots available. Selection based on basin diversity and data completeness.
Norwegian Continental Shelf, UK Central North Sea, and Gulf of Mexico preferred.

Frequently Asked Questions

Q: What if my well data is incomplete or messy?

A: That's exactly the problem we solve. We've worked with 57 years of paper/microfiche/SQL records. Send us what you have—incomplete data is our specialty.

Q: Will you share my data with third parties?

A: Absolutely not. All data covered by NDA. We may publish anonymized/aggregated results (e.g., "North Sea well showed 80ft datum error") but never operator-identifiable information without explicit permission.

Q: What happens after the pilot?

A: Nothing, unless you want more. If you're satisfied, we can discuss full field audits at our standard £50K/well rate. If not, you keep the free report and we part as friends.

Q: Why only 3 assets?

A: Basin diversity validation. We need Norwegian Continental Shelf, UK Central North Sea, and Gulf of Mexico data to prove algorithm transferability. Each analysis requires 20-40 hours of engineering review. Quality over quantity.

Engineering-Led Intelligence

THE WELLTEGRA INTEGRITY STANDARD

Thirty years of North Sea operational expertise applied to the digital transition. This is not theoretical AI—this is physics-informed, author-provenant data forensics for critical energy infrastructure.

Four Pillars of Data Integrity

📊

I. Data Governance

Migration Decay Detection: Legacy datasets suffer from "contextual divorce" when migrated from relational databases (WellView, OpenWorks) to flat-file public repositories.

✓ Datum Reference Stripping: KB vs MSL metadata loss during CSV export
✓ Temporal Metadata Loss: Platform subsidence corrections lost in migration
✓ Author-Provenant Reconciliation: Original field records validated against physical invariants
✓ Thermodynamic Verification: Pump pressures, displacement volumes, and hydrostatic principles confirm physical depth

WellTegra re-anchors "floating" digital records to their original engineering context, ensuring datasets are complete and error-free for decommissioning safety.

👤

II. Human Oversight

Subject Matter Expertise (SME) Validation: Effective oversight requires domain seniority. Thirty years of North Sea operational experience is not optional—it is the only valid human-in-the-loop for critical infrastructure AI.

✓ Original Field Author Status: Kenneth McKenzie authored 2014 WellView records for Ninian/Thistle clusters
✓ Rig-Floor Experience: Scale squeeze operations, barrier testing, P&A interventions
✓ Perfect 11 Asset Portfolio: Direct operational knowledge of Ninian, Thistle, Dunlin, Cormorant, Magnus clusters
✓ Physics-First Validation: AI outputs verified against pump pressures, displacement volumes, formation tops

Automated tools without this operational seniority cannot validate the "Author's Truth" against thermodynamic invariants.

⚠️

III. Accuracy & Robustness

Risk Mitigation for Decommissioning Safety: A 4-meter vertical datum error in plug & abandonment (P&A) operations creates unacceptable environmental and financial risk.

✓ Barrier Misplacement: Cement plugs set 4m off-target may miss critical formation seals
✓ Reservoir Breach Risk: Incorrect TVD may intersect unidentified high-pressure zones
✓ Regulatory Non-Compliance: WIOS 2026 mandate requires verified TVD accuracy for NDR submission
✓ £40B Decommissioning Wave: Systematic errors compound across 470+ North Sea wells

WellTegra's manifold-constrained verification ensures depth accuracy meets safety-critical thresholds for decommissioning operations.

🔐

IV. Sovereign Vault

Transparency & Immutability Requirements: High-risk AI systems require comprehensive logging, audit trails, and cryptographic verification. WellTegra's sovereign infrastructure creates a "Notarized Ledger of Wellbore Truth."

✓ GPG Cryptographic Signing: Every analysis output timestamped and operator-exclusive
✓ Air-Gapped Architecture: No cloud egress—sovereign compute in Scotland (SC876023)
✓ Immutable Audit Logs: Manifold-constrained inference steps recorded for regulatory review
✓ NSTA NDR Compatibility: Outputs formatted for direct submission with variance framing

Strict confidentiality: All findings remain operator property. No case studies, no benchmarking, no disclosure without consent.

The Engineering-Led AI Commitment

WellTegra Ltd (SC876023) applies three decades of North Sea operational experience to the digital transition. Our manifold-constrained hyper-connection (mHC-GNN) platform is not theoretical AI—it is physics-informed intelligence designed for critical energy infrastructure.

Data Governance

Author-provenant records validated against physical invariants. Migration decay detection ensures datasets are complete and error-free.

SME Oversight

Original field author with 30 years of rig-floor experience validates every AI output against thermodynamic principles.

Safety-Critical Accuracy

Verified depth accuracy for decommissioning operations. Systematic error detection prevents £40B P&A program risk exposure.

Our mission is to upgrade data certainty for the North Sea decommissioning wave—not to expose operational variance, but to ensure compliance with modern standards.

Information Security Coordinator & Chief AI Architect:
Kenneth McKenzie | WellTegra Ltd (SC876023)
30 Years North Sea Operational Experience | Original Author: 2014 Ninian/Thistle WellView Records

kenneth.mckenzie@welltegra.network

*74% Accuracy @ 128 Layers: Benchmark result from mHC-GNN research on citation network datasets (PubMed, arXiv:2601.02451). Engineering target: Reproduce on North Sea wellbore correlation networks (442 wells). Proof-of-Concept demonstrated on Thistle A-12 (8,247→8,214 ft, 99.7% confidence).

**EPL Tax Relief Positions: £1.58B represents internal analysis of potential EPL (Energy Profits Levy) tax relief positions across Perfect 11 portfolio. This figure is pending third-party audit and HMRC verification. Individual operator results may vary.

SOVEREIGN GROUND TRUTH FOR LEGACY ASSET RE-ALIGNMENT.