Back to results
US 12,392,241 B2Granted
Determining reservoir fluid properties from downhole fluid analysis data using machine learning
AssigneeSchlumberger Technology Corporation
InventorsShahnawaz Molla, Farshid Mostowfi
Life of this patent
Estimating the term…
Abstract
Methods for determining in situ the value of a formation fluid parameter using a downhole fluid analysis (DFA) tool. The methods utilize advanced statistical learning tools to build a predictive model to estimate a fluid property given a set of input parameters. In one embodiment the fluid saturation pressure parameter is determined by using the DFA tool to obtain the fluid and to obtain weight fractions of at least C1, C6+, and CO2 of the fluid. The weight fractions and a reservoir temperature are input into a trained statistical learning machine to obtain a determination of the saturation pressure of the fluid.
Claims17
A method of determining in situ a value of a parameter of a reservoir fluid taken from a formation traversed by a borehole, comprising:
- performing a first fluid analysis in real-time while a tool is concurrently in the borehole, wherein performing the first fluid analysis comprises:
- drawing a first fluid into the tool, wherein the tool comprises an optical module including a spectrometer;
- shining light through the first fluid in the tool and sensing with the spectrometer a first plurality of resulting signals via one or more channels of the spectrometer;
- determining a first set of inputs based on the first plurality of resulting signals;
- applying, via a controller, the first set of inputs to a trained statistical learning machine;
- taking, via the controller, a first output of the trained statistical learning machine as a determination of the value of the parameter, wherein the trained statistical learning machine is trained utilizing a dataset of fluid properties for fluid samples as input parameters;
- modifying, via the controller, the trained statistical learning machine based at least upon the first output and the first plurality of resulting signals associated with the first output;
- and
- determining, via the controller, a subset of the one or more channels, wherein the subset of the one or more channels are associated with relevant values to determine the value of the parameter;
- performing a second fluid analysis in real-time while the tool is concurrently in the borehole, wherein performing the second fluid analysis comprises:
- drawing a second fluid into the tool;
- shining light through the second fluid in the tool and sensing with the spectrometer a second plurality of resulting signals via the subset of the one or more channels;
- determining, via the controller, a second set of inputs based on the second plurality of resulting signals;
- applying, via the controller, the second set of inputs to the trained statistical learning machine;
- taking, via the controller, a second output of the trained statistical learning machine as the determination of the value of the parameter;
- and
- comparing, via the controller, the first output and the second output to determine a change in the value of the parameter;
- and
- modifying, via the controller, a drawdown pressure of the tool based on the comparison between the first output and the second output.
16 more elements
A method, comprising:
- obtaining samples of reservoir fluid from one or more formations at measured formation temperatures;
- analyzing the saturation pressures of the samples in a laboratory;
- analyzing the samples for optical density as a function of wave-length to obtain indicative information of the different makeups of the samples;
- using the measured formation temperatures, saturation pressures and indicative information as at least training data to train a statistical learning machine, wherein the measured formation temperatures, saturation pressures and indicative information comprises a dataset of fluid properties for training the statistical learning machine and wherein the dataset is split into a training portion for training the statistical learning machine and a testing portion for testing the statistical learning machine;
- performing a first fluid analysis in real-time while a tool is concurrently in a borehole, wherein the first fluid analysis comprises:
- drawing a first fluid into the tool at a first drawdown pressure, wherein the tool comprises an optical module including a spectrometer, and the first fluid is taken from a downhole formation at a measured downhole formation temperature;
- shining light through the first fluid in the tool and sensing with the spectrometer a plurality of resulting signals via one or more channels of the spectrometer;
- applying indications of the resulting signals and the measured downhole formation temperature as inputs to the trained statistical learning machine, wherein the indications of the resulting signals are optical density indications of a plurality of channels of the spectrometer, wherein the optical density indications are the inputs to the trained statistical learning machine;
- taking an output of the trained statistical learning machine as a determination of a saturation pressure of the first fluid;
- modifying the trained statistical learning machine based at least upon the output and the resulting signals associated with the first fluid;
- and
- determining a subset of the one or more channels of the spectrometer, wherein the subset of the one or more channels are associated with relevant values to determine the saturation pressure;
- performing a second fluid analysis in real-time while the tool is concurrently in the borehole, wherein the second fluid analysis comprises:
- drawing a second fluid into the tool;
- shining light through the second fluid in the tool and sensing with the subset of the one or more channels of the spectrometer a plurality of resulting signals related to the second fluid;
- applying indications of the resulting signals related to the second fluid as inputs to the trained statistical learning machine;
- and
- taking an output of the trained statistical learning machine as a determination of a saturation pressure of the second fluid;
- and
- modifying the first drawdown pressure to a second drawdown pressure different than the first drawdown pressure based on the relative values obtained for the saturation pressure of the first fluid and the second fluid.
17 more elements
Dependencies are inferred from the claim text.