CN-122014179-A - Variable-component flue gas drives coal seam gas and seals up CO2Intelligent regulation system and method

CN122014179ACN 122014179 ACN122014179 ACN 122014179ACN-122014179-A

Abstract

The invention discloses an intelligent regulation and control system and method for driving coal seam gas and sequestering CO 2 by variable component flue gas, and belongs to the technical field of coal seam gas exploitation and CO 2 sequestering. The system comprises a variable component injection module, a multi-parameter distributed sensing network, a pumping and utilizing module, a digital twin model, an AI intelligent regulation engine and a safety interlocking and emergency response module. The variable component injection module is used for adjusting the proportion, pressure, flow, temperature and humidity of CO 2 /N 2 , the sensing network is used for realizing high-frequency panoramic monitoring of the whole injection-extraction process through various sensors, the digital twin model is used for coupling permeability evolution, adsorption and crack diversion models, real-time inversion of key parameters is realized, the AI engine is used for outputting an optimal injection strategy based on LSTM prediction and PPO reinforcement learning algorithm, and the safety interlocking and emergency response module is used for triggering emergency response at the millisecond level when the limit is exceeded. The system realizes the collaborative optimization of the coalbed methane recovery ratio, the CO 2 sequestration rate and the operation energy consumption, and completes the paradigm conversion from 'blind injection' to 'intelligent regulation'.

Inventors

GE ZHAOLONG
JIA YUNZHONG
LU YIYU
TANG JIREN
LU CHAOHUI
LI QIAN
HE YUHUAN
HUA HAO

Assignees

重庆大学

Dates

Publication Date: 20260512
Application Date: 20260106

Claims (10)

1. The intelligent regulation and control system for driving coal seam gas and sequestering CO 2 by using variable component flue gas is characterized by comprising a variable component injection module, a multi-parameter distributed sensing network module, a extraction and utilization module, a digital twin model module, an AI intelligent regulation and control engine module, a safety interlock and an emergency response module; The variable component injection module injects smoke into the well according to a preset initial value, the multi-parameter distributed sensing network module collects well data and transmits the data to the digital twin model module for calculation, and a calculation result is input into the AI intelligent regulation engine module to issue an optimal control instruction; The extraction and utilization module is used for separating, purifying and storing CH 4 gas after extraction; The safety interlocking and emergency response module, the variable component injection module, the multi-parameter distributed sensing network module, the digital twin model module and the AI intelligent regulation engine module run in parallel and independently in the whole course, and are started immediately when any index is detected to break through a preset safety threshold, wherein the index is generated by the AI intelligent regulation engine module.
2. The intelligent regulation and control system for driving coal seam gas and sealing CO 2 by variable component flue gas according to claim 1 is characterized in that the variable component injection module comprises a hot flue gas storage tank (3), a CO 2 gas storage tank (4), an N 2 gas storage tank (5), an intelligent proportional valve group (6), a variable frequency gas injection pump (7), a humidifying and dedusting system (8), a heat exchanger group (9) and a gas injection pipeline (11), wherein the hot flue gas storage tank (3) is communicated with the CO 2 gas storage tank (4), the N 2 gas storage tank (5) and the variable frequency gas injection pump (7) in sequence through the intelligent proportional valve group (6), the variable frequency gas injection pump (7) sequentially inputs gas into the humidifying and dedusting system (8) and the heat exchanger group (9) through the gas injection pipeline (11), the injection well (12) sequentially penetrates through a rock stratum (1) and a coal seam (2), the gas injection pipeline (11) is placed in the injection well (12), the injection well (12) is L-shaped, and a gas injection sealer (13) is arranged between the rock stratum (1) and the coal seam (2).
3. The intelligent regulation and control system for variable component flue gas-driven coal seam gas and sequestration CO 2 , wherein the multi-parameter distributed sensor network module comprises an injection well integrated multi-parameter monitoring device (14), an injection well data transmission line (10), a ground control center (25), an extraction-monitoring combined well integrated multi-parameter monitoring device (16) and an extraction-monitoring combined well data transmission line (15), the injection well integrated multi-parameter monitoring device (14) is arranged in an injection well (12), the injection well integrated multi-parameter monitoring device (14) is collected at a high frequency of 1Hz through the injection well data transmission line (10) and transmitted to the ground control center (25), the extraction-monitoring combined well integrated multi-parameter monitoring device (16) is arranged in an extraction-monitoring combined well (19) and transmits data in an extraction process to the ground control center (25) through the extraction-monitoring combined well data transmission line (15), the extraction-monitoring combined well (19) penetrates rock strata (1) and (2), and a coal seam (17) is arranged between the extraction-monitoring combined well (1) and the coal seam (2).
4. The intelligent regulation and control system for driving coal seam gas and sequestering CO 2 by using variable component flue gas according to claim 1, wherein the extraction and utilization module comprises an extraction pipeline (18), a coalbed methane extraction pump (20), a heat exchanger group (21), a gas filtering device (22), a gas separation device (23) and a CH 4 gas storage tank (24), wherein the extracted gas is cooled from the extraction pipeline (18) through the coalbed methane extraction pump (20) through the heat exchanger group (21), the impurity gas is removed by the gas filtering device (22), and the CH 4 gas is separated by the gas separation device (23) and then enters the CH 4 gas storage tank (24) for subsequent use.
5. The intelligent regulation and control system for driving coal seam gas and sequestering CO 2 by using variable component flue gas according to claim 3, wherein the ground control center (25) is integrated with a digital twin model module and an AI intelligent regulation and control engine module.
6. The intelligent regulation and control system for driving coal seam gas and sequestering CO 2 by using variable component flue gas according to claim 1 or 5, wherein the digital twin model module comprises a permeability evolution model, an adsorption saturation model and a crack diversion model.
7. The intelligent regulation and control system for driving coal seam gas and sequestering CO 2 by using variable component flue gas according to claim 1 or 5, wherein the AI intelligent regulation and control engine module comprises an LSTM prediction algorithm and a PPO reinforcement learning algorithm.
8. The intelligent regulation and control system for driving coal seam gas and sequestering CO 2 by using variable component flue gas according to claim 6, wherein the permeability evolution model has the following expression: In the formula, Is the initial permeability; as an effective stress function; As a function of matrix strain; Is a chemical dissolution correction factor; The expression of the adsorption saturation model is as follows: In the formula, Is composed of Is not limited, adsorption saturation of (2); Is composed of Langmuir constant of (a); Is composed of Is a partial pressure of (2); Is composed of Langmuir constant of (a); Is composed of Is a partial pressure of (2); representing summing the three gases; The expression of the crack flow guiding model is as follows: In the formula, Is the current fracture porosity; is the initial fracture porosity; the current crack width; is the initial crack width; Correction coefficients for shear displacement; As a function of chemical dissolution time.
9. The intelligent regulation and control system for driving coal seam gas and sequestering CO 2 by using variable component flue gas according to claim 7, wherein the LSTM prediction algorithm takes historical underground pressure, temperature, CH 4 /CO 2 concentration and gas production time sequence of not less than 30 days as input, and outputs average pressure distribution, CH 4 output and CO 2 breakthrough time predicted value of 6 hours in the future; The PPO reinforcement learning algorithm comprises a state space consisting of average pressure, CH 4 concentration, CO 2 plume position and accumulated microseismic energy, an action space consisting of smoke injection pressure 0-25 MPa, smoke flow 0.1-10 t/h and CO 2 /N 2 volume ratio 30% -70%, and a reward function as follows: Wherein, the Is the stimulation rate compared with the previous cycle; The CO 2 sequestration rate in a single cycle; Energy consumption for the injection pump and the heating unit; The first weight coefficient, the second weight coefficient and the third weight coefficient.
10. An intelligent regulation and control method for variable component flue gas-driven coal seam gas and sequestration CO 2 is used for realizing the intelligent regulation and control system for variable component flue gas-driven coal seam gas and sequestration CO 2 as set forth in claims 1-9, and is characterized by comprising the following steps: s1, system starting, self-checking and initializing: The system comprises a variable component injection module, a multi-parameter distributed sensing network, a digital twin model, an extraction and utilization module, an AI intelligent regulation engine and a safety interlocking module, a verification intelligent proportional valve group (6), a variable frequency gas injection pump (7), a humidifying and dedusting system (8), a heat exchanger group (9), an injection well integrated multi-parameter monitoring device (14), an extraction-monitoring combined well integrated multi-parameter monitoring device (16), an injection well data transmission line (10), an extraction-monitoring combined well data transmission line (15), an edge computing node and a blockchain audit interface, wherein the integrity of the intelligent proportional valve group, the variable frequency gas injection pump (7), the humidifying and dedusting system (8), the heat exchanger group (9), the injection well integrated multi-parameter monitoring device (14), the extraction-monitoring combined well integrated multi-parameter monitoring device (16), the integrity of the injection well data transmission line (10), the extraction-monitoring combined well data transmission line (15), the edge computing node and the blockchain audit interface is electrified, if abnormal, the integrity is alarmed, and is stopped, and otherwise the system enters S2; s2, data real-time acquisition and data preprocessing; synchronously collecting pressure, temperature, flow, CO 2 /N 2 /CH 4 concentration, AE sound wave, microseismic and geothermal gradient according to high frequency of 1 Hz; The edge is cached for 5min to slide window data, and the ground control center (25) performs data cleaning and outlier rejection on the received real-time data to generate a cleaned data packet D t ; s3, rolling and updating the digital twin model module; Inputting the cleaned data packet D t obtained in the step S2 and historical drilling, logging and fracturing data into a digital twin model, calling a permeability evolution model, an adsorption saturation model and a crack diversion model, inverting and obtaining a current full-field permeability field, a pressure field, CO 2 plume distribution and CH 4 desorption degree, and rolling and updating the twin body state with 1h as a period to serve as an initial field for follow-up prediction and optimization; S4, AI intelligent regulation engine closed-loop decision, including: S4.1, predicting the LSTM model state, namely inputting cleaning data and the latest twin body state which are more than or equal to 30 days in the past, and outputting average pressure distribution, CH 4 yield and CO 2 breakthrough time for 6 hours in the future; s4.2, PPO reinforcement learning optimization: Encoding the key indexes of the twin body at the moment t, namely average pressure, CH 4 concentration, CO 2 plume position and accumulated microseismic energy into a state vector S t to construct a state space; The set action space comprises injection pressure P_ { inj }, injection flow Q_ { inj } and CO 2 /N 2 proportion R_ { CO 2 :N 2 }, which is expressed as action space A t ={P_{inj},Q_{inj},R_{CO 2 :N 2 }, wherein P_ { inj } E [0,25] MPa, Q_ { inj } E [0.1,10] t/h, R_ { CO 2 :N 2 } E [30%,70% ]; calculating instant rewards according to the rewarding function, and outputting optimal action vectors; s4.3, checking safety constraint: If present: the bottom hole pressure dip rate P_ { drop } >1MPa/10min; CO 2 increase >2%/d; The energy released by a single microseismic event E_ { seis } >10 4 J; S5, otherwise, utilizing PPO shearing importance sampling and gradient rising to update network parameters, finishing strategy updating, further issuing instructions, analyzing an optimal motion vector into a specific set value, issuing within 500ms, wherein the issuing objects comprise that an intelligent proportional valve group (6) adjusts the proportion of CO 2 /N 2 , a variable-frequency gas injection pump (7) adjusts P_ { inj }, Q_ { inj }, a humidifying and dedusting system (8) adjusts the moisture content to be 0-20%, a heat exchanger group (9) adjusts the flue gas temperature to be 150-400 ℃, and finally, waiting for a 1h rolling period and returning to S2 to form a closed loop; S5, safety interlocking and emergency response; when any safety constraint triggering condition is established in S4.3 safety constraint verification, immediately starting a safety instruction: ① Closing the injection valve to reduce the flow to be less than or equal to 20% of the maximum flow; ② Switching the air source to a high N 2 mode, wherein the switching conditions comprise that CO 2 :N 2 is less than or equal to 3:7 or less than or equal to 1:9; ③ Forcibly reducing the injection pressure to be less than or equal to 5MPa; ④ Starting a groundwater chemical monitoring unit, and collecting pH, conductivity and dissolved CO 2 data; ⑤ Uploading the operation log to a blockchain audit platform in an encrypted manner, and synchronously sending a risk alarm to a supervision platform; before the instruction is issued, the system performs final security check, including: S5.1, judging whether the optimal strategy parameters solved by the PPO algorithm meet all preset safety constraint conditions or not; S5.2, if the safety constraint is not met, the system does not execute a dangerous instruction, and immediately starts a safety interlocking and emergency response module to switch to a degraded operation mode; S5.3, if the strategy meets all safety constraints, the system issues the generated optimization instruction to the variable component intelligent injection module S6, instruction execution and closed loop feedback; After the variable component intelligent injection module receives the instruction, an intelligent proportional valve group (6) is driven to allocate the proportion of CO 2 to N 2 , a humidifying and dedusting system (8) is driven, a heat exchanger group (9) is used for regulating the smoke to the target temperature and humidity, a variable-frequency gas injection pump (7) is used for executing injection operation according to the set flow and pressure, afterwards, the S1 is returned again, new underground data after injection is read, and accordingly a continuous iterative and self-optimized closed-loop intelligent regulation and control loop is formed, and the loop takes 1 hour as a control period and adapts to dynamic changes of a reservoir in real time.

Description

Intelligent regulation and control system and method for driving coal seam gas and sequestering CO 2 by variable component flue gas Technical Field The invention belongs to the technical field of coalbed methane exploitation and CO 2 sequestration, and particularly relates to an intelligent regulation and control system and method for driving coal bed methane and sequestering CO 2 by variable component flue gas. Background The direct emission of industrial flue gas (the main components are CO 2 and N 2) causes huge pollution to the environment, under the background, the industrial flue gas rich in CO 2 is injected into a hypotonic coal seam, CO 2 can effectively seal and replace CH 4 by virtue of stronger adsorption capacity, N 2 can more effectively promote desorption and migration of CH 4 by partial pressure effect, and coal bed gas (CH 4) exploitation is driven to be facilitated, and the method is regarded as a very promising 'three-in-one' technical path, and is hopeful to synchronously realize reduction of industrial carbon emission, production increase of unconventional natural gas and energy safety guarantee. However, the scale-up and commercial application of this technology is always limited by its inherent complexity and hysteresis of traditional regulatory means. This complexity stems primarily from the strong nonlinear coupling of multiple physicochemical processes. After injection into the coal seam, the industrial flue gas will initiate a series of complex reactions including competitive adsorption, matrix expansion/contraction, dynamic evolution of permeability, and fracture network reconstruction. The traditional fixed component injection mode ignores the dynamic complementarity of CO 2 and N 2, resulting in suboptimal displacement efficiency and sequestration potential. Furthermore, prior art perception of subsurface conditions relies heavily on wellbore point measurements and periodic geophysical prospecting, resulting in the overall system operating under a nearly "black box" condition, making it difficult to capture the CO 2 plume front in real time, accurately evaluate containment safety, and pre-warn of potential leakage risks. The current process control strategy depends on static models and artificial experience, and lacks an intelligent brain capable of fusing real-time data, insight into system states and performing prospective optimization. This results in a system that cannot adaptively respond to rapidly changing reservoir conditions, and is difficult to dynamically trade-off between the objectives of "maximize CH 4 recovery", "maximize CO 2 sequestration," and "minimize operating energy consumption," thus limiting the overall performance of the technology. In order to break through the bottleneck, a novel method is provided, the core innovation is that dynamic component allocation, a multi-parameter distributed sensing network, a digital twin model and an artificial intelligent decision engine are deeply fused, the purpose is to realize the paradigm conversion from 'blind injection' to 'perspective' and from 'experience driving' to 'intelligent optimization', and a novel research idea is provided for the development of the coalbed methane industry. Disclosure of Invention In order to solve the technical problems, the invention provides an intelligent regulation and control system and method for driving coal seam gas and sequestering CO 2 by using variable-component flue gas. To achieve the above technique, it includes: an intelligent regulation and control system for driving coal seam gas and sealing CO 2 by using variable component flue gas comprises a variable component injection module, a multi-parameter distributed sensing network module, a extraction and utilization module, a digital twin model module, an AI intelligent regulation and control engine module and a safety interlocking and emergency response module; The variable component injection module injects smoke into the well according to a preset initial value, the multi-parameter distributed sensing network module collects well data and transmits the data to the digital twin model module for simulation, and a simulation result is input into the AI intelligent regulation engine module to issue an optimal control instruction; The extraction and utilization module is used for separating, purifying and storing CH 4 gas after extraction; the safety interlocking and emergency response module, the variable component injection module, the multi-parameter distributed sensing network module, the digital twin model module and the AI intelligent regulation engine module run in parallel and independently in the whole course, and are immediately started when any index is detected to break through a preset safety threshold, wherein the index is generated by the AI intelligent regulation engine module and comprises the following components: a) The pressure dip rate is >1 MPa/10 min; b) Monitoring the CO 2 concentration day increase >