# US 11,905,817 B2 · Method and system for managing carbon dioxide supplies using machine learning > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,905,817 B2 - Document ID: US11905817B2 - Status: Granted - Kind: B2 - Application number: 17/644,728 - Filed: 2021-12-16 - Published: 2023-06-22 - Granted: 2024-02-20 - Earliest claimed: 2021-12-16 - Assignees: Saudi Arabian Oil Company - Applicants: Saudi Arabian Oil Company - Inventors: Klemens Katterbauer; Abdulaziz S. Al-Qasim; Alberto F. Marsala; Sara Abu Alsaud; Ali Abdallah AlYousif - Web page: https://patentgrep.com/patent/US11905817B2 ## CPC - E21B 43/164: {Injecting CO2 or carbonated water (in combination with organic material C09K8/594)} - E21B 47/00: Survey of boreholes or wells (monitoring pressure or flow of drilling fluid E21B21/08) - E21B 43/166: {Injecting a gaseous medium; Injecting a gaseous medium and a liquid medium (CO2 injection E21B43/164; steam injection E21B43/24)} - G06N 3/084: Backpropagation, e.g. using gradient descent - G06Q 30/018: Certifying business or products - E21B 49/00: Testing the nature of borehole walls; Formation testing; Methods or apparatus for obtaining samples of soil or well fluids, specially adapted to earth drilling or wells - E21B 2200/20: Computer models or simulations, e.g. for reservoirs under production, drill bits - E21B 2200/22: Fuzzy logic, artificial intelligence, neural networks or the like - G06N 3/02: Neural networks - G06N 3/08: Learning methods - G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks - G06N 20/00: Machine learning - G06N 20/10: using kernel methods, e.g. support vector machines [SVM] - G06N 20/20: Ensemble learning - Y02P 30/00: Technologies relating to oil refining and petrochemical industry - E21B: EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR - Y02P: CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS ## Abstract A method may include obtaining reservoir data for a geological region of interest. The method may further include obtaining production data regarding one or more wells coupled to the geological region of interest. The method may further include obtaining carbon emission data for the one or more wells. The method may further include determining predicted carbon emission data and predicted production data using a machine-learning model. The method may further include determining one or more stimulation parameters for a stimulation operation based on the predicted carbon emission data and the predicted production data. The method may further include transmitting a command to a control system coupled to an injection well. The command adjusts an amount of carbon dioxide that is supplied to the injection well based on the one or more stimulation parameters. ## Claims ### Claim 1 A method, comprising: obtaining, by a computer processor, reservoir data for a geological region of interest; obtaining, by the computer processor, acquired production data regarding one or more wells coupled to the geological region of interest; obtaining, by the computer processor, carbon emission data for the one or more wells; determining, by the computer processor, first predicted production data for the one or more wells using the reservoir data and a first machine-learning model; determining, by the computer processor, error data based on a mismatch between the first predicted production data and the acquired production data; updating, by the computer processor, the first machine-learning model to produce an updated machine-learning model, wherein the first machine-learning model is updated using the error data and a machine-learning algorithm, and wherein the machine-learning algorithm is a backward propagation algorithm; determining, by the computer processor, first predicted carbon emission data for the one or more wells and second predicted production data for the one or more wells using the updated machine-learning model; determining, by the computer processor, one or more stimulation parameters for a stimulation operation based on the first predicted carbon emission data and the second predicted production data; transmitting, by the computer processor, a first command to a control system coupled to an injection well; and adjusting, using the control system and in response to the control system receiving the first command, an amount of carbon dioxide that is supplied to the injection well based on the one or more stimulation parameters. ### Claim 2 (depends on 1) The method of claim 1, wherein the first machine-learning model comprises a neural network comprising at least one hidden layer, at least one neuron, and an activation function, wherein the neural network determines a predicted production rate at a predetermined time for the one or more wells, and wherein the one or more stimulation parameters correspond to a carbon dioxide injection based on the predicted production rate. ### Claim 3 (depends on 1) The method of claim 1, wherein the first machine-learning model comprises a neural network comprising at least one hidden layer, at least one neuron, and an activation function, wherein the neural network determines a predetermined amount of carbon dioxide emissions associated with providing electric power to the one or more wells to achieve a predetermined production rate, and wherein the one or more stimulation parameters correspond to a carbon dioxide injection based on the predetermined amount of carbon dioxide emissions. ### Claim 4 (depends on 1) The method of claim 1, wherein the first machine-learning model comprises one or more non-linear autoregressive neural network with an exogenous input (NARX) models. ### Claim 5 (depends on 1) The method of claim 1, further comprising: obtaining, by the computer processor and from a user device, a request to determine the one or more stimulation parameters in response to a user input to the user device. ### Claim 6 (depends on 1) The method of claim 1, wherein the reservoir data comprise sensor data that describes gas content in the geological region of interest based on one or more gas tracers. ### Claim 7 (depends on 1) The method of claim 1, wherein the first predicted production data corresponds to a predetermined amount of oil, a predetermined amount of gas, and a predetermined amount of water that are produced by a production well. ### Claim 8 A method, comprising: obtaining, by a computer processor, reservoir data for a geological region of interest; obtaining, by the computer processor, acquired production data regarding a production well coupled to the geological region of interest; obtaining, by the computer processor, carbon emission data for the production well; determining, by the computer processor, predicted production data for the production well using a first machine-learning model and the acquired production data; determining, using a second machine-learning model, predicted carbon emission data for the production well for a predetermined period of time based on well data for the production well, historical injection data, and historical electrical-power data, wherein the predicted carbon emission data corresponds to electric power requirements for the production well that achieve carbon dioxide neutrality with a stimulation operation for the production well; determining, by the computer processor, one or more stimulation parameters for the stimulation operation based on the predicted carbon emission data and the predicted production data; transmitting, by the computer processor, a command to a control system coupled to an injection well; and adjusting, using the control system and in response to the control system receiving the command, an amount of carbon dioxide that is supplied to the injection well based on the one or more stimulation parameters. ### Claim 9 A system, comprising: a first control system coupled to an injection well; a second control system coupled to a production well; and a carbon dioxide manager comprising a computer processor, wherein the carbon dioxide manager is coupled to the first control system and the second control system, the carbon dioxide manager being configured to: obtain reservoir data for a geological region of interest; obtain acquired production data regarding the production well; determining first predicted production data for the production well using the reservoir data and a first machine-learning model; determining error data based on a mismatch between the first predicted production data and the acquired production data; updating the first machine-learning model to produce an updated machine-learning model, wherein the first machine-learning model is updated using the error data and a machine-learning algorithm, and wherein the machine-learning algorithm is a backward propagation algorithm; obtain carbon emission data for the production well and the injection well; determine predicted carbon emission data for the production well and second predicted production data for the production well using the updated machine-learning model; determine one or more stimulation parameters for a stimulation operation based on the predicted carbon emission data and the second predicted production data; and transmit a command to the first control system, wherein the first control system adjusts an amount of carbon dioxide that is supplied to the injection well based on the one or more stimulation parameters in response to the first control system receiving the command. ### Claim 10 (depends on 9) The system of claim 9, wherein the first machine-learning model comprises a neural network comprising at least one hidden layer, at least one neuron, and an activation function, wherein the neural network determines a predicted production rate at a predetermined time for the production well, and wherein the one or more stimulation parameters comprise a carbon dioxide injection that is based on the predicted production rate. ### Claim 11 (depends on 9) The system of claim 9, wherein the first machine-learning model comprises a neural network comprising at least one hidden layer, at least one neuron, and an activation function, wherein the neural network determines a predetermined amount of carbon dioxide emissions associated with providing electric power to the production well to achieve a predetermined production rate, and wherein the one or more stimulation parameters comprise a carbon dioxide injection that is based on the predetermined amount of carbon dioxide emissions. ### Claim 12 (depends on 9) The system of claim 9, wherein the first machine-learning model comprises one or more non-linear autoregressive neural network with an exogenous input (NARX) models. ### Claim 13 (depends on 9) The system of claim 9, further comprising: a user device configured to transmit, to the carbon dioxide manager, a request to determine the one or more stimulation parameters in response to a user input to the user device. ### Claim 14 (depends on 9) The system of claim 9, wherein the reservoir data comprise sensor data that describes gas content in the geological region of interest based on one or more gas tracers. ### Claim 15 (depends on 9) The system of claim 9, wherein the first predicted production data corresponds to a predetermined amount of oil, a predetermined amount of gas, and a predetermined amount of water that are produced by the production well. ### Claim 16 (depends on 9) The system of claim 9, wherein the carbon dioxide manager is further configured to: determine, using a second machine-learning model, second predicted carbon emission data for a second production well for a predetermined period of time based on well data for the second production well, historical injection data, and historical electric-power data, wherein the second predicted carbon emission data corresponds to electric power requirements for the production well that achieve carbon dioxide neutrality with a stimulation operation for the production well.