Three-dimensional inversion of multi-component electromagnetic measurements using a fast proxy model
Described herein are systems and techniques for monitoring for monitoring and evaluating conditions associated with a wellbore and wellbore operations that use neural operators instead of computationally intensive iterative differential equations. Such systems and techniques allow for determinations to be made as operations associated with a wellbore are performed. Instead of having to wait for computationally intensive tasks to be performed or take risks of proceeding with a wellbore operation without real-time evaluations being performed, these wellbore operations may be continued while determinations are timely made, thus improving operation of computing systems that perform evaluations and that make decisions regarding safely and efficiently performing wellbore operations such as drilling a wellbore, cementing wellbore casings in place, or injecting fluids into formations of the Earth.
1. A method comprising:
controlling at least one of a wellbore drilling and evaluation operation at a wellbore;
collecting data when the at least one of the wellbore drilling and evaluation operation is controlled at the wellbore;
converting first set of one or more time and space domain functions into one or more frequency domain functions based on one or more neural operators, wherein the one or more time and space domain functions are associated with collected data;
converting the one or more frequency domain functions into a second set of one or more time and space domain functions;
identifying a condition associated with the at least one of the wellbore drilling and evaluation operation based on at least one time and space domain function of the second set of one or more time and space domain functions and
an operational rule; and
initiating a change to the at least one of the wellbore drilling and evaluation operation based on the condition according to the operational rule.
2. The method of claim 1 ,
wherein the change includes steering a drilling apparatus to avoid a first material that has a first characteristic.
3. The method of claim 1 , further comprising:
wherein the change includes steering a drilling apparatus to avoid an area of a subterranean formation according to the operational rule.
4. The method of claim 1 , wherein the operation rule is associated with at least one of water located in strata near the wellbore, leakage of carbon dioxide (CO 2 ) from a reservoir near the wellbore, steering a drilling apparatus based on permeabilities of materials near the wellbore, or water intrusion into the wellbore.
5. The method of claim 1 , wherein the wellbore operation corresponds to cementing a casing in the wellbore.
6. The method of claim 1 , further comprising:
filtering data associated with the one or more frequency domain functions.
7. The method of claim 6 , further comprising:
generating an image of materials that surround at least a portion of the wellbore after the data associated with the one more frequency domain functions has been filtered.
8. The method of claim 1 , further comprising:
injecting a substance into a subterranean formation that surrounds the wellbore, wherein the subterranean formation includes a reservoir;
identifying a change in a conductivity or a resistivity of the subterranean formation that surrounds the wellbore;
identifying that the reservoir has reached a full level based on an operational rule associated with the change in the conductivity or the resistivity of the subterranean formation; and
stopping the injection of the substance into the subterranean formation based on the operational rule.
9. The method of claim 8 , wherein the change includes changing a flow rate of the substance injected into the subterranean formation.
10. The method of claim 8 , wherein the substance is at least one of a fracturing fluid, water, a fluid associated with enhanced oil recovery, carbon dioxide (CO 2 ), or steam.
11. A method comprising:
converting a first set of one or more time and space domain functions into one or more frequency domain functions based on one or more neural operators, wherein the first set of one or more time and space domain functions are associated with collected data;
converting the one or more frequency domain functions into a second set of one or more time and space domain functions;
identifying one or more material characteristics of a wellbore based on application at least one of the second set of one or more time and space domain functions;
identifying that a material characteristic of the one or more material characteristics of the wellbore corresponds to an operational plan of the wellbore; and
controlling a flow of a substance provided to the wellbore based on the material characteristics of the one or more material characteristics of the wellbore corresponding to the operational plan of the wellbore.
12. The method of claim 11 , further comprising:
controlling flow of carbon dioxide (CO 2 ) into a portion of the wellbore; and
monitoring the flow of CO 2 into the portion of the wellbore according to a CO 2 sequestering rule.
13. The method of claim 12 , wherein the 12 , The method of claim 11 , wherein the CO 2 sequestering rule controls one or more of:
a CO 2 injection pressure,
a temperature,
a CO 2 stability level, or
boundaries of the portion of the wellbore that form a CO 2 reservoir.
14. The method of claim 12 , further comprising:
controlling the flow of the CO 2 injected into the portion of the wellbore such that no more than a threshold amount of the CO 2 escapes boundaries of the portion of the wellbore that form a CO 2 reservoir.
15. The method of claim 11 , further comprising:
controlling flow of a fracturing fluid into a portion of the wellbore; and
monitoring the flow of the fracturing fluid into the portion of the wellbore according to a hydraulic fracturing rule.
16. A method comprising:
converting a first set of a plurality of time and space domain functions into a plurality of frequency domain functions based on a plurality of neural operators;
converting the plurality of frequency domain functions into a second set of a second set of a plurality of time and space domain functions;
estimating a first material characteristic of a wellbore based on application of at least one time and space domain function of the second set of the plurality of time and space domain functions;
identifying an error associated with the estimate of the first material characteristic of the wellbore;
applying a second time and space domain function of the second set of time and space domain functions based on the error corresponding to an error threshold level; and
updating the first material characteristic based on the application of the second time and space domain function.
17. The method of claim 16 , further comprising:
applying the second set of the plurality of time and space domain functions in parallel;
identifying a neural operator of the plurality of neural operators that results in a minimal error of the first material characteristic; and
identifying the first material characteristic based on application of the neural operator being associated with the minimal error of the first material characteristic.
18. The method of claim 16 , wherein the first material characteristic corresponds to one or more of a permeability, a porosity, a type of material, or distances between the wellbore and the type of material.
19. The method of claim 16 , further comprising:
performing a frequency domain evaluation based on the first set of the plurality of time and space domain functions, wherein conversion of the at least one time and space domain function of the plurality of frequency domain functions is based on the evaluation in the frequency domain.