Our ML-Based Solution with integrated Heliometric Data guarantees minimum possible errors in the hydrodynamic modelling
HYDRODYNAMIC MODELLING WITH HELIOMETRICS
Surface Heliometric survey can be carried out at various stages of field development and be successfully applied in order to map drained/undrained areas of hydrocarbon-saturated reservoirs and to locate residual oil reserves, which have not yet been involved in field development.
Heliometric data is used together with the latest generation of data-driven INSIM-FT model, which allows to significantly reduce the time of calculating the physical model of the field compared to traditional hydrodynamic simulators.
Our ML-based model allows to automatically adapt hundreds, thousands of physical models, while reducing MAPE - mean average percentage error to 20%.
A more accurate physical model of the field with heliometric data solves the field optimization problem and supports selection of the best decisions on what exactly should be done in the field in order to maintain and increase production, to involve in the development of areas not currently covered.
This service is performed by our sister company Froswell Ltd. www.froswell.com
Solution:
Heliometric Data + Artificial Intelligence
METER PROBLEM
well production history
Heliometric data
improved geological model
better performance
As the base of our Heliometric technology over development field, we use a powerful and exact in accuracy software to model oil reservoirs and flow processes and to combine into one - Helium data, physical models and algorithms of Machine Learning.
After running Areal 3D Heliometric survey over the production field and measuring Helium gas at the wellheads, we refine field data (up to 70% well historical events) with Machine Learning algorithms. Only refined data will be entered into the physical model for automatic adaptation. It means that after adaptation of few thousands of different models our software finds the best model suitable for this certain field with minimum possibility of error.
The physical model will be tuned by the Machine Learning in boundary conditions and presumptions. That makes our calculations clear, quick and targeted.
We can ensure our clients that the measure of accuracy in our final oil field model will be with a margin error of 20-30% where other known existing models do not use Heliometric data have error of 100-200%.
As soon as the field model with an actual knowledge of oil accumulation and distribution values and parameters is ready, we generate the list of recommended procedures for every and each well. From there on, you can be exact in your decision making in regards of water or polymer flooding, side or new drilling, EOR operations, hydrofracturing etc.
With our method we have achieved increases in oil production in the magnitude of 40-70% (in relation to base values) without drilling any additional well and 100% and much more with drilling.
Our strong experience in geo-technological services will help you to increase the capitalization of your company without any large investments, see your field alternately, re-evaluate procedures at the wells, reduce up to 30-40% of operational cost and extend life-time of your field.