Back to results
US 10,242,312 B2Granted
Synthetic logging for reservoir stimulation
AssigneeQuantico Energy Solutions LLC
InventorBruce H. Storm, Jr.
Life of this patent
Estimating the term…
Abstract
The present disclosure generally relates to synthetic logging for reservoir stimulation. A synthetic logging method for stimulating a reservoir includes: training a machine learning algorithm using historical or exploratory data; and generating a synthetic elastic property log of the reservoir by supplying the trained machine learning algorithm with data acquired from a production wellbore.
Claims15
A synthetic logging method for stimulating a reservoir, comprising:
- training a machine learning algorithm (MLA) using:
- training inputs comprising historical or exploratory data measured by one or more logging tools selected from a group consisting of:
- passive nuclear logging tools and nonnuclear logging tools;
- and
- training outputs comprising one or more property logs selected from a group consisting of:
- density logs;
- and porosity logs, wherein the one or more property logs are derived from the historical or exploratory data and additional historical or exploratory data measured by one or more tools containing radioactive sources;
- generating a synthetic property log comprising one or more synthesized properties of the reservoir by supplying the trained MLA with inputs comprising data acquired from a production wellbore by one or more logging tools selected from a group consisting of:
- passive nuclear logging tools and nonnuclear logging tools, wherein the data is acquired without using an induced nuclear logging tool, wherein the data comprises measurements of one or more properties different than the one or more synthesized properties, wherein the synthetic property log is generated by the trained MLA as an output based on the inputs, and wherein the synthetic property log comprises either a density log or a porosity log;
- and
- generating a stimulation plan by supplying the synthetic property log to a reservoir model, wherein the stimulation plan is used to stimulate the reservoir.
8 more elements
A synthetic logging method for analyzing a reservoir, comprising:
- supplying a trained machine learning algorithm (MLA) with inputs comprising data acquired from a production wellbore by one or more logging tools selected from a group consisting of:
- passive nuclear logging tools and nonnuclear logging tools, wherein the data is acquired without using an induced nuclear logging tool, and wherein the trained MLA was trained using:
- training inputs comprising historical or exploratory data measured by one or more logging tools selected from a group consisting of:
- passive nuclear logging tools and nonnuclear logging tools;
- and
- training outputs comprising one or more property logs selected from a group consisting of:
- density logs;
- and porosity logs, wherein the one or more property logs are derived from the historical or exploratory data and additional historical or exploratory data measured by one or more tools containing radioactive sources;
- generating a synthetic log comprising one or more synthesized properties of the reservoir using the trained MLA and the supplied data, wherein the supplied data comprises measurements of one or more properties different than the one or more synthesized properties, wherein the synthetic log is selected from a group consisting of:
- a density log and a porosity log, and wherein the synthetic log is generated by the trained MLA as an output based on the inputs;
- and
- generating a stimulation plan by supplying the synthetic log to a reservoir model, wherein the stimulation plan is used to stimulate the reservoir.
9 more elements
A synthetic logging method for determining properties of a formation, comprising:
- training a machine learning algorithm (MLA) using:
- training inputs comprising historical or exploratory data including data measured by one or more logging tools selected from a group consisting of:
- passive nuclear logging tools and nonnuclear logging tools;
- and
- training outputs comprising one or more property logs selected from a group consisting of:
- density logs;
- and porosity logs, wherein the one or more property logs are derived from the historical or exploratory data and additional historical or exploratory data measured by one or more tools containing radioactive sources;
- generating a synthetic property log comprising one or more synthesized properties of the reservoir by supplying the trained MLA with inputs comprising data acquired from a production wellbore by one or more logging tools selected from a group consisting of:
- passive nuclear logging tools and nonnuclear logging tools, wherein the data is acquired without using an induced nuclear logging tool, wherein the data comprises measurements of one or more properties different than the one or more synthesized properties, wherein the synthetic property log is generated by the trained MLA as an output based on the inputs, and wherein the synthetic property log comprises either a density log or a porosity log;
- and
- generating a stimulation plan by supplying the synthetic property log to a reservoir model, wherein the stimulation plan is used to stimulate the reservoir.
8 more elements
Dependencies are inferred from the claim text.