Refined Product Inventories
Monitoring refined product inventories presents a different challenge from crude oil. While crude is predominantly stored in floating roof tanks that can be measured using satellite imagery, refined products are typically stored in fixed roof tanks, making conventional satellite measurement much more difficult and sometimes impossible.
Kpler's methodology uses infrared drone technology to measure storage at selected hubs, delivering direct inventory assessments where traditional satellite techniques are not effective. As new surveys are completed, additional hubs are incorporated into the service, steadily expanding towards global coverage.
The Process: From Imagery to Actionable Insight
Our data processing pipeline is designed for high-frequency, high-precision results, combining advanced drone technology with rigorous human and AI validation.
Weekly Data Collection: Every Thursday, we conduct drone-based infrared imagery flights to capture the current state of fixed-roof tanks in monitored terminals (subject to weather conditions).
Data Processing and Analysis: Raw imagery is processed using human annotation and 3D modeling techniques to accurately identify individual tanks and compute precise fill levels.
Human Validation: To ensure the highest data quality, our expert liquids analyst team reviews all model outputs. They cross-reference the data with cargo tracking intelligence and external market sources.
Data Publication: Finalized datasets are published to the market every Tuesday at 9:00 AM CET.
Data Methodology & Granularity
We distinguish between the physical measurement of a tank and the logical allocation of the product stored within it.
Tank Fill Level Measurement
We monitor both Commercial Tanks (belonging to independent terminals) and Refinery Tanks.
Commercial Tanks: Storage levels are reported at the individual tank level.
Refinery Tanks: Granularity is provided at the facility level.
Product Type Allocation
Determining what is stored in a tank requires a multi-layered approach:
Hard Data: For commercial tanks, we leverage verified market sources to determine the specific product stored.
Refinery Stockage Model: For refinery tanks, where physical product markers are not directly visible, we utilize our proprietary in-house Refineries model. This model estimates product output share per refinery, allowing us to approximate product volume using the formula:
Key Specifications
Coverage: Currently focused on the ARA (Amsterdam-Rotterdam-Antwerp) hub, including main CPP (Clean Petroleum Products) and DPP (Dirty Petroleum Products). The team is actively working on expansion of regional coverage to New York Harbor.
Frequency: Weekly updates, published every Tuesday.
Historical and Upcoming Data:
PoROT weekly data is available from week ending 17 April 2026.
PoAMS weekly data is available from week ending 24 April 2026.
PoANT will kick off from week ending 7 August 2026.
Accessibility: Data is available via API endpoints. The specifications can be found in the developer portal.
Frequently asked questions
How does Kpler ensure data quality?
Quality is maintained through a “human-in-the-loop" approach. While our data pipeline is highly automated, the final step involves a manual review by our expert liquids analysts to verify the measurements before they reach our clients.
What differentiates Kpler’s data from existing market offerings?
Tank Product Allocation Frequency: We provide bi-weekly updates on product storage, enabling near-real-time visibility into tank product switches.
Tank Product Allocation Accuracy: We leverage verified on-ground intelligence to ensure high precision in product allocation.
Publication Timeline: Data collected during our Thursday drone flights is published the following Tuesday - the fastest time-to-market in the industry.
Comprehensive Capacity Coverage: Our rigorous mapping allows us to track terminal capacity more extensively than alternative market solutions.
Cross-referenced Cargo Tracking: By leveraging Kpler’s extensive cargo tracking footprint, we identify flow shifts early and cross-validate storage data against market movements.
What percentage of ARA is monitored?
Kpler monitors fixed-roof tanks under main refined products terminals. The current scope includes main CPP and DPP products. Tanks prone to carrying Chemical and Biofuel/Vegoil are not monitored yet.
Are there other hubs on the roadmap?
Yes! We’re actively working on the addition of NYH tanks. Stay tuned!
How frequently is the data updated?
Once a week, on Tuesdays 9AM CET, pending any weather delays.
How do your updates handle weather delays?
If weather prevents our scheduled Thursday drone flights, users will be promptly notified via email, and the updated values will be made available through the API as soon as data collection and processing are complete.
Does Kpler model its data against public figures (e.g., JODI)?
No. Our ARA Drone Inventory data is calculated entirely independently, providing a proprietary and unfiltered view of the market.
Revision History
29 July 2026: First version of this methodology.