US-12619212-B2 - Reducing CO2 production levels in a gas operation network using a predictive model

Abstract

This specification relates to reducing CO2 levels of a gas operation network using a predictive model. The systems and methods described in this specification process an input data set that includes (i) gas flowrate data and (ii) CO2 concentration data. The systems and methods generate the predictive model by processing the input dataset. The systems and methods predict, by the predictive model, a CO2 production target for the gas operation network based on a gas flowrate predicted for each of the plurality of wells by the predictive model. The systems and methods generate control signals to control a respective valve at each well based on the CO2 production target and the predicted gas flowrate for each well. The systems and methods regulate CO2 levels of the gas operation network based on the CO2 production target by controlling the respective valves at each well.

Inventors

• Muhammad Idris
• Emad Abbad M. Alabbad

Assignees

• SAUDI ARABIAN OIL COMPANY

Dates

Publication Date

20260505

Application Date

20220209

Claims (20)

1. 1 . A method for reducing CO 2 levels of a gas operation network using a predictive model, the gas operation network comprising a plurality of wells, the method comprising: processing an input data set comprising (i) gas flowrate data for each of the plurality of wells and (ii) CO 2 concentration data for each of the plurality of wells; generating the predictive model by processing the input dataset in response to applying a reduced gradient algorithm or a linear regression algorithm to data values of the input dataset; predicting, by the predictive model, a CO 2 production target for the gas operation network based on a gas flowrate predicted for each of the plurality of wells by the predictive model; generating one or more control signals to control a respective valve at each of the plurality of wells based on the CO 2 production target and the predicted gas flowrate for each of the plurality of wells; and regulating CO 2 levels of the gas operation network based on the CO 2 production target by controlling the respective valves at each of the plurality of wells using the one or more control signals.
2. 2 . The method of claim 1 , wherein: the input data set comprises historical data from a previous time period; and the predictive model is generated based on a training phase that coincides with the previous time period or overlaps the previous time period.
3. 3 . The method of claim 2 , wherein the previous time period spans at least 100 days.
4. 4 . The method of claim 1 , further comprising: providing a threshold CO 2 level for the plurality of wells that is based on (i) the gas flowrate data for each of the plurality of wells and (ii) the CO 2 concentration data for each of the plurality of wells.
5. 5 . The method of claim 4 , wherein the predictive model is based on the threshold CO 2 level for each of the plurality of wells and a corresponding production factor.
6. 6 . The method of claim 1 , wherein the gas operation network comprises a plurality of gas gathering manifolds and a plurality of slug catchers.
7. 7 . The method of claim 6 , wherein generating the predictive model comprises: determining a total gas production of each of the plurality of gas gathering manifolds based on historical gas flowrate data for each of the plurality of wells connected to a corresponding gas gathering manifold; comparing the total gas production of each of the plurality of gas gathering manifolds to a gas flowrate measured at each of the plurality of slug catchers; and updating the predictive model based on the comparison.
8. 8 . The method of claim 1 , wherein generating the predictive model comprises: minimizing a difference between (i) CO 2 production predictions for the gas operation network and (ii) CO 2 production data for the gas operation network.
9. 9 . The method of claim 8 , further comprising: measuring, by one or more gas flow meters located at each of the plurality of wells, the gas flowrate data; measuring, by one or more separator tests located at each of the plurality of wells, the CO 2 concentration data; and measuring, by a CO 2 meter located at a gas processing plant or downstream of the gas processing plant, the CO 2 production data.
10. 10 . The method of claim 1 , wherein generating the predictive model comprises solving a minimization problem using at least one of a generalized reduced gradient solver and a multiple linear regression solver.
11. 11 . The method of claim 1 , wherein predicting the CO 2 production target of the gas operation network comprises satisfying one or more production conditions.
12. 12 . The method of claim 11 , wherein the one or more production conditions comprise a maximum CO 2 production level of the gas operation network.
13. 13 . The method of claim 12 , wherein the one or more production conditions comprise a minimum condensate production level of the gas operation network.
14. 14 . The method of claim 1 , wherein predicting the CO 2 production target for the gas operation network comprises: predicting a plurality of CO 2 production levels based on a gas flowrate variation for at least one well of the plurality of wells; determining a plurality of CO 2 production differences between each of the plurality of CO 2 production levels and a baseline CO 2 production level; and selecting the CO 2 production target based on the plurality of CO 2 production differences.
15. 15 . The method of claim 14 , wherein the gas flowrate variation represents a variation between a reduction in gas production of the at least one well of the plurality of wells by 50% and an increase in gas production of the at least one well of the plurality of wells by 50%.
16. 16 . The method of claim 14 , wherein selecting the CO 2 production target comprises: predicting a condensate production level for the gas operation network for each of the plurality of CO 2 production levels by multiplying each predicted gas flowrate of the gas flowrate variation with a condensate-gas-ratio for each of the plurality of wells; determining a plurality of condensate production differences between each of the condensate production levels and a baseline condensate production level; and selecting the whole network CO 2 production target based on the plurality of CO 2 production differences and the plurality of condensate production differences.
17. 17 . The method of claim 16 , further comprising measuring, by one or more separator tests located at each well of the plurality of wells, the condensate-gas-ratio for each of the plurality of wells.
18. 18 . A system for reducing CO 2 levels of a gas operation network using a predictive model, the gas operation network comprising a plurality of wells, the system comprising: a plurality of gas flow meters operable to measure a gas flowrate of a gas at each of the plurality of wells; a CO 2 concentration measurement device operable to measure a CO 2 concentration of the gas at the plurality of wells; a CO 2 meter operable to measure a network CO 2 production level of the gas, wherein the CO 2 meter is located at a gas processing plant of the gas operation network or downstream of the gas processing plant; a plurality of valves operable to control a flow of the gas at the plurality of wells; a computer storing instructions that, when executed by a processor of the computer, cause the processor to perform operations comprising: processing an input data set comprising (i) the measured gas flowrate for each of the plurality of wells and (ii) the measured CO 2 concentration for each of the plurality of wells; generating the predictive model by processing the input dataset in response to applying a reduced gradient algorithm or a linear regression algorithm to data values of the input dataset; predicting, by the predictive model, a CO 2 production target for the gas operation network based on a gas flowrate predicted for each of the plurality of wells by the predictive model; generating one or more control signals to control each of the plurality of valves at each of the plurality of wells based on the CO 2 production target and the predicted gas flowrate for each of the plurality of wells; and regulating CO 2 levels of the gas operation network based on the CO 2 production target by controlling each of the plurality of valves at each of the plurality of wells using the one or more control signals.
19. 19 . The system of claim 18 , wherein the gas operation network comprises a plurality of gas gathering manifolds.
20. 20 . The system of claim 19 , wherein the gas operation network comprises a plurality of slug catchers.

Description

TECHNICAL FIELD The present disclosure relates to predictive modeling for reducing carbon dioxide (CO2) production levels in a gas operations network. BACKGROUND Oil and gas wells extract hydrocarbons from underground reservoirs. In some cases, extraction of the hydrocarbons results in the release of an undesirable gas—CO2. The CO2 can be released to the environment or sequestered back into the ground. However, releasing the CO2 to the environment contributes to global warming and sequestering the CO2 into the ground is an expensive process that requires complicated machinery. Thus, efficient methods for reducing overall CO2 production levels are desirable for oil and gas well operations. SUMMARY The systems and methods described in this specification use a predictive model to reduce the amount of CO2 produced from a network of oil and gas wells. The predictive model is trained based on historical data from the wells and a processing plant. The trained predictive model is used to predict how the flowrates of the wells should be changed (for example, increased or decreased) to achieve a desired (or threshold) reduction in the overall CO2 produced from the network of wells. The wells are controlled based on the predicted flow rates to achieve the desired CO2 reduction. In addition to reducing the amount of CO2 produced from the network of oil and gas wells, the specification describes techniques for determining how predicted flow rates affect condensate production and ethane-plus production. In some implementations, the disclosed techniques are used to select predicted flow rates of the wells based on corresponding changes in condensate production and ethane-plus production. Some methods for reducing CO2 levels of a gas operation network using a predictive model include one or more of the following steps. In some implementations, the gas operation network includes a plurality of wells. Some methods include processing an input data set including (i) gas flowrate data for each of the plurality of wells and (ii) CO2 concentration data for each of the plurality of wells. Some methods include generating the predictive model by processing the input dataset in response to applying a reduced gradient algorithm or a linear regression algorithm to data values of the input dataset. Some methods include predicting, by the predictive model, a CO2 production target for the gas operation network based on a gas flowrate predicted for each of the plurality of wells by the predictive model. Some methods include generating one or more control signals to control a respective valve at each of the plurality of wells based on the CO2 production target and the predicted gas flowrate for each of the plurality of wells. Some methods include regulating CO2 levels of the gas operation network based on the CO2 production target by controlling the respective valves at each of the plurality of wells using the one or more control signals. In some examples, the input dataset includes historical data from a previous time period and the predictive model is generated based on a training phase that coincides with the previous time period or overlaps the previous time period. In some examples, the previous time period spans at least 100 days. Some methods include a threshold CO2 level for the plurality of wells based on (i) the gas flowrate data for each of the plurality of wells and (ii) the CO2 concentration data for each of the plurality of wells. In some implementations, the predictive model is based on the threshold CO2 level for each of the plurality of wells and a corresponding production factor. Some gas operation networks include a plurality of gas gathering manifolds and a plurality of slug catchers. In such cases, some methods include generating the predictive model by determining a total gas production of each of the plurality of gas gathering manifolds based on historical gas flowrate data for each of the plurality of wells connected to a corresponding gas gathering manifold; comparing the total gas production of each of the plurality of gas gathering manifolds to a gas flowrate measured at each of the plurality of slug catchers; and updating the predictive model based on the comparison. In some implementations, generating the predictive model includes minimizing a difference between (i) CO2 production predictions for the gas operation network and (ii) CO2 production data for the gas operation network. In some implementations, generating the predictive model includes solving a minimization problem using at least one of a generalized reduced gradient solver and a multiple linear regression solver. In some implementations, predicting the CO2 production target of the gas operation network includes satisfying one or more production conditions. In some examples, the one or more production conditions include a maximum CO2 production level of the gas operation network. In some examples, the one or more production conditions include a minimu