DIGITAL BEHAVIOR TWIN

The most reliable FDP forecast software. And the fastest too.

Relies on past production data and embarks other key inputs (well completion, PVT, permeability…) and all physical equations related to petroleum engineering.

FOROIL sets up a bespoke Digital Behavior Twin, with the sole objective of calculating future production per well, pursuant to the principles of statistical learning theory and constrained by the physics of the reservoir and wells.

This original and patented approach leads to 95% accuracy and high speed (less than a second per run).

Fact: a Reliable Forecast

Genuine modeling reliability is explained by Statistical Learning Theory (Vapnik), and is demonstrated in practice by performing blind tests.

A blind test compares the Digital Behavior Twin calculation based on activities actually conducted on the field on the one hand, with the recorded production data on the other hand, the latter being kept hidden from FOROIL until the calculation is released. FOROIL Digital Behavior Twin achieves a 95%+ accuracy.

Focused on brown fields

It applies to brown fields, as soon as they have accumulated sufficiently rich data (minimum of five years of production, with at least 10 producing wells).

The complexity of the Digital Behavior Twin is tuned in such a way it is comprehensive enough to include all key phenomena in the field, yet remaining simple enough to be calibrated unambiguously (US Patent).

No assumption needs to be made about geological properties, as their impact on production is already included in past data. This maximizes the reliability of the forecast.

A novel approach

FOROIL Digital Behavior Twin is quite different from other solutions available in the industry:

The learning process of past production data is constrained by all the laws of reservoir and well physics.

Physics is coherently embedded at all relevant scales (wells, groups of wells, reservoir, surface), in a systemic approach.

The Digital Behavior Twin is constructed according to a (patented) industrial process, relying on increasingly complex models.

As a result, the forecast accuracy is ensured, even far from current functioning conditions (with the same recovery technique).

(Very) Fast forecast

Owing to its moderate complexity, the Digital Behavior Twin is by construction quick to compute: it calculates any production forecast, well by well, month by month, over five years, in seconds.