Back to results
US 11,906,695 B2Granted
Method and system for generating sponge core data from dielectric logs using machine learning
AssigneeSaudi Arabian Oil Company
InventorsMohammed A. Alghazal, Ghazi Kraishan
Life of this patent
Estimating the term…
Abstract
A method may include obtaining dielectric log data regarding a geological region of interest. The method may further include generating sponge core data of the geological region of interest using a neural network model and the dielectric log data. The neural network model may be trained using various dielectric logs from various well sources. The method may further include determining, using the sponge core data, an amount of hydrocarbons in the geological region of interest.
Claims8
A method, comprising:
- obtaining, using a plurality of sponge coring tools, acquired sponge core data from a plurality of wells in a plurality of formations, wherein at least one sponge coring tool among the plurality of sponge coring tools comprises a core barrel and a sponge sleeve that is made of a polyurethane sponge;
- obtaining, using a dielectric logging tool, dielectric log data regarding a geological region of interest wherein the dielectric logging tool comprises an antenna, at least one transmitter configured to transmit an electromagnetic field, and at least one receiver configured to measure a phase and an amplitude of the electromagnetic field;
- obtaining, by a computer processor, a rock type for the geological region of interest, a reservoir zone type for the geological region of interest, and a porosity value for the geological region of interest using one or more well logs;
- generating, by the computer processor, predicted sponge core data of the geological region of interest using a neural network model, the rock type, the reservoir zone type, the porosity value, and the dielectric log data,
- wherein the rock type, the reservoir zone type, the porosity value, and the dielectric log data are data inputs to the neural network model,
- wherein the predicted sponge core data comprises an oil saturation value that describes an amount of oil found in a core sample acquired during a coring operation in the geological region of interest, and
- wherein the neural network model is trained using a plurality of dielectric logs the acquired sponge core data, and a backpropagation algorithm;
- and
- determining, by the computer processor and using the predicted sponge core data, an amount of hydrocarbons in the geological region of interest.
6 more elements
A system, comprising:
- a plurality of sponge coring tools, wherein at least one sponge coring tool among the plurality of sponge coring tools comprises a core barrel and a sponge sleeve that is made of a polyurethane sponge;
- a logging system coupled to a plurality of logging tools, wherein the plurality of logging tools comprises a dielectric logging tool, and wherein the dielectric logging tool comprises an antenna, at least one transmitter configured to transmit an electromagnetic field, and at least one receiver configured to measure a phase and an amplitude of the electromagnetic field;
- and
- a reservoir simulator comprising a computer processor, wherein the reservoir simulator is coupled to the logging system and comprises functionality for:
- obtaining acquired sponge core data from a plurality of wells in a plurality of formations, wherein the acquired sponge core data is generated using the plurality of sponge coring tools;
- obtaining, using the dielectric logging tool, dielectric log data regarding a geological region of interest using the logging system;
- obtaining a rock type for the geological region of interest, a reservoir zone type for the geological region of interest, and a porosity value for the geological region of interest using one or more well logs;
- generating predicted sponge core data of the geological region of interest using a neural network model, the rock type, the reservoir zone type, the porosity value, and the dielectric log data,
- wherein the rock type, the reservoir zone type, the porosity value, and the dielectric log data are data inputs to the neural network model,
- wherein the predicted sponge core data comprises an oil saturation value that describes an amount of oil found in a core sample acquired during a coring operation in the geological region of interest, and
- wherein the neural network model is trained using a plurality of dielectric logs the acquired sponge core data, and a backpropagation algorithm;
- and
- determining, using the predicted sponge core data, an amount of hydrocarbons in the geological region of interest.
10 more elements
Dependencies are inferred from the claim text.