Deep foundation pit engineering groundwater purification recycling system and method

Abstract

This invention belongs to the field of groundwater purification technology for deep foundation pit engineering, and relates to a system and method for the purification and recycling of groundwater in deep foundation pit engineering. The invention includes six steps: pretreatment via dewatering wells, cyclone sedimentation and coagulation, biofilm purification monitoring, membrane separation and chemical oxidation, sludge dewatering and resource recovery, and intelligent monitoring and control. This invention simultaneously degrades organic matter and ammonia nitrogen under bubble-free aeration conditions, ensuring stable effluent compliance and improving treatment efficiency and operational stability. Through the synergistic effect of a DC electric field and mechanical pressure, this invention disrupts the colloidal structure of sludge and drives the directional migration of water, dewatering the sludge. The dewatered sludge is then sintered with construction waste to produce ceramsite, and the dewatered filtrate is recycled, achieving sludge reduction, resource recovery, and zero filtrate discharge, thus realizing efficient water resource reuse. This invention collects water quality, membrane status, and environmental data in real time, and intelligent control throughout the entire process improves system stability, ensuring continuous effluent compliance and extending membrane cleaning cycles through predictive maintenance.

Description

Technical Field

[0001] This invention belongs to the technical field, specifically relating to a system and method for purifying and recycling groundwater in deep foundation pit engineering. Background Technology

[0002] In deep foundation pit engineering, groundwater control is a crucial aspect of ensuring construction safety and environmental protection. Traditional foundation pit dewatering techniques often involve directly extracting and discharging groundwater. While this can meet construction needs, it has significant drawbacks. Excessive groundwater extraction can easily lead to stress imbalance in the surrounding strata, causing environmental problems such as building settlement and cracking. Furthermore, the direct discharge of untreated groundwater results in water resource waste, which contradicts the concept of green construction.

[0003] Currently, groundwater in deep foundation pit projects is recycled through a water tank recharge system. However, due to limitations in the permeability of the recharge wells and the water purification capacity, physical filtration is usually performed using only quartz sand and non-woven fabric, which cannot remove dissolved organic matter, nitrogen, and phosphorus, resulting in insufficient purification. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a system and method for purifying and recycling groundwater in deep foundation pit engineering.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for purifying and recycling groundwater in deep foundation pit engineering includes the following steps. Step 1: Pre-treatment of groundwater through dewatering wells. Groundwater is pumped out through dewatering wells, and after passing through a screen to intercept large particles of impurities, it enters the collection tank. Step 2: Cyclone sedimentation and coagulation. The incoming water to the collection tank is centrifugally desanded by a cyclone separator. The effluent is introduced into a sedimentation tank to allow the sludge to settle naturally. Coagulant is added and flocculation is enhanced by stirring. The sludge at the bottom of the sedimentation tank is periodically discharged to Step 5 for treatment to obtain pretreated water. Step 3: Biofilm purification monitoring. The pretreated water obtained in Step 2 is fed into the membrane aeration MABR membrane reactor. Under bubble-free aeration conditions, the biofilm simultaneously degrades organic matter and ammonia nitrogen. Water quality parameters and the operating status of MABR membrane purification are monitored in real time by sensors. Step 4: Membrane separation and chemical oxidation. The effluent from the membrane-aerated MABR membrane reactor is subjected to membrane separation and selective chemical oxidation to generate purified water for use. At the same time, the used water is returned to Step 1 for further treatment. Step 5: Sludge dewatering and resource utilization. The sludge from the sedimentation tank and biological purification is dewatered by electro-osmosis and then co-utilized with construction waste. The dewatered water is then transferred back to Step 2 for further treatment. Step Six: Intelligent Monitoring and Control. Based on the sensor datasets of the inlet, membrane module surface and outlet of the membrane aeration MABR membrane reactor, a dynamic water quality prediction model is constructed through LSTM neural network. The model outputs the chemical oxygen demand fluctuation trajectory and membrane fouling trend value, and links the edge computing node to generate control commands.

[0006] The coagulant in step two includes low-aluminum residual PAC and bio-based coagulant.

[0007] The membrane aeration MABR membrane reactor in step three uses hollow fiber membrane modules to form a composite biofilm under bubble-free aeration conditions, utilizing biofilm microorganisms to simultaneously degrade organic matter and ammonia nitrogen.

[0008] Step four involves real-time monitoring of total dissolved solids (TDS) concentration using an online conductivity sensor. Ultrafiltration is activated when TDS is <2000 mg / L; otherwise, reverse osmosis is switched to TDS. The cleaning cycle is calculated using the MFI model formula, triggering ultrasonic rinsing, and the membrane flux recovery rate is ≥95%.

[0009] The selective chemical oxidation in step four includes: Electrochemical oxidation, using boron-doped diamond electrodes, at a current density of 10–50 mA / cm², degrades petroleum hydrocarbon pollutants; Heterogeneous catalytic ozone oxidation is achieved through a honeycomb ceramic catalyst supported on composite metal oxides of iron and manganese, with an ozone dosage of 5–20 mg / L.

[0010] In step four, the refluxed water and the newly added groundwater are mixed at a volume ratio of 1:3 to 5.

[0011] In step five, the sludge undergoes electro-osmotic dewatering with a DC electric field voltage of 10-30V and a mechanical pressure of 0.5-1.5MPa, reducing the sludge moisture content to below 40%.

[0012] In step five, the separated water flows back through a pipeline to the inlet of the hydrocyclone separator, where it is mixed with the newly added groundwater at a volume ratio of 1:2 to 4 before being treated.

[0013] The specific process in step six includes the following steps. S601, Real-time Acquisition of Multi-Source Data Sensors are installed at the inlet of the membrane-aerated MABR membrane reactor, the surface of the membrane module, and the outlet of the membrane-aerated MABR membrane reactor to collect operational data in real time, which is then used as a dataset, including: Turbidity sensors, TOC analyzers, pH / temperature composite sensors, and ammonia nitrogen sensors are deployed at the inlet of the membrane aeration MABR membrane reactor to monitor the raw water pollution load in real time. The membrane module surface is equipped with a laser biofilm thickness probe, a 9-point gridded dissolved oxygen microelectrode array, a transmembrane differential pressure sensor, and a patch-type temperature sensor to dynamically track biofilm activity and fouling status. The outlet is equipped with nitrate and nitrite sensors, an acute biotoxicity meter, a residual chlorine sensor, and a dissolved oxygen sensor to ensure that water quality meets standards and ecological safety. S602, Water Quality Change Trend Prediction Based on the historical operating dataset of the membrane aeration MABR reactor, including influent chemical oxygen demand (COD), ammonia nitrogen concentration, dissolved oxygen gradient on the membrane surface, transmembrane pressure difference, and biofilm thickness, a long short-term memory (LSTM) neural network was trained to capture water quality abrupt changes and output future COD fluctuation trajectories, ammonia nitrogen concentration changes, and quantitative indicators of membrane fouling trends. S603, Dynamic Control Strategy Generation Based on the chemical oxygen demand fluctuation trajectory, ammonia nitrogen concentration changes, and membrane fouling trend quantitative indicators predicted by LSTM neural network, the following controls are implemented: Aeration pressure adjustment: The gas-side pressure P of the membrane aeration MABR membrane reactor is dynamically set according to the formula P=0.02×Cn+0.05, where Cn is the predicted ammonia nitrogen concentration; When the predicted biofilm thickness is >300μm, switch to pulse aeration mode; The method for selecting membrane separation mode is as follows: if the TDS (total dissolved solids) concentration is <2000 mg / L, ultrafiltration separation should be used. If the TDS (Total Dissolved Solids) concentration is ≥2000 mg / L, switch to reverse osmosis separation; When the membrane fouling index is 3.0 to 6.0, sodium hypochlorite solution is injected into the inner cavity of the membrane aeration MABR membrane reactor.

[0014] A groundwater purification and recycling system for deep foundation pit engineering includes: The groundwater pretreatment module enables preliminary filtration and temporary storage of groundwater, intercepting large particulate impurities. The cyclone sedimentation coagulation module removes suspended solids from groundwater pretreated by the dewatering well pretreatment module through centrifugal sand removal and gravity sedimentation, and adds composite coagulant to enhance solid-liquid separation. The sludge resource utilization module dewaters the sludge treated by the cyclone sedimentation and coagulation module, solidifies it, and then co-sintersects it with construction waste to form ceramsite. The filtrate is returned to the cyclone sedimentation module for further treatment. The membrane aeration MABR bioreactor module degrades pollutants in the pretreated water after the swirl sedimentation coagulation module under bubble-free aeration conditions, and collects data of the entire biological treatment process in real time through a three-position sensor network. The membrane separation chemical oxidation module selectively oxidizes the concentrated water and specific pollutants treated by the membrane aeration MABR biological reactor module based on water quality switching membrane separation mode, and the purified water that meets the standards is reused in the project or returned to the pretreatment module. The intelligent control center module uses LSTM model prediction driven by sensor data in the membrane aeration MABR reactor to dynamically optimize the operating parameters of biological reaction, membrane separation and chemical oxidation.

[0015] Beneficial effects: 1. This invention utilizes salt-tolerant microorganisms in a MABR membrane reactor to simultaneously degrade organic matter and ammonia nitrogen under bubble-free aeration conditions, ensuring stable effluent quality. Furthermore, bubble-free aeration improves oxygen utilization and reduces energy consumption. At the same time, the salt tolerance of the biofilm enhances the system's resistance to shocks, thereby improving treatment efficiency and operational stability.

[0016] 2. This invention uses the combined action of a DC electric field and mechanical pressure to destroy the colloidal structure of sludge and drive the directional migration of water to dewater the sludge. The dewatered sludge is then sintered with construction waste to form ceramsite. The dewatered filtrate is then recycled for further treatment, achieving sludge reduction, resource recovery, and zero discharge of filtrate. This forms a sustainable purification system that combines reduction, resource recovery, and recycling, achieving efficient water resource reuse.

[0017] 3. This invention collects water quality, membrane status and environmental data in real time, combines LSTM neural network to predict water quality trends, dynamically optimizes MABR membrane aeration pressure, membrane separation mode and chemical oxidation intensity, and achieves intelligent control throughout the process, which significantly improves system stability, ensures that the effluent continuously meets the standards, and predictive maintenance extends the membrane cleaning cycle and reduces maintenance costs.

[0018] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1: See Figure 1 A method for purifying and recycling groundwater in deep foundation pit engineering includes the following steps: Step 1: Pre-treatment with dewatering wells; groundwater is pumped out through dewatering wells, and after passing through a screen to intercept large particles of impurities, it enters the collection tank; Step 2: Cyclone sedimentation and coagulation; the incoming water to the collection tank is centrifugally desanded using a cyclone separator, and the effluent is introduced into a sedimentation tank for natural sedimentation of sludge and sand. Coagulant is added, and flocculation is enhanced by stirring. The sludge at the bottom of the sedimentation tank is periodically discharged to Step 5 for treatment to obtain pretreated water. Step 3: Biofilm purification monitoring; The pretreated water obtained in Step 2 is fed into the membrane aeration MABR membrane reactor. Under bubble-free aeration conditions, the biofilm simultaneously degrades organic matter and ammonia nitrogen, and the water quality parameters and the operating status of the MABR membrane purification are monitored in real time by sensors. Step 4: Membrane separation and chemical oxidation; The effluent from the membrane-aerated MABR membrane reactor is subjected to membrane separation and selective chemical oxidation to generate purified water for use. At the same time, the used water is returned to Step 1 for further treatment. Step 5: Sludge dewatering and resource utilization; The sludge from the sedimentation tank and biological purification is dewatered by electro-osmosis and then co-utilized with construction waste for resource utilization. The dewatered water is then transferred back to Step 2 for further treatment. Step Six: Intelligent Monitoring and Control; Based on the sensor datasets of the inlet, membrane module surface and outlet of the membrane aeration MABR membrane reactor, a dynamic water quality prediction model is constructed through LSTM neural network, which outputs the chemical oxygen demand fluctuation trajectory and membrane fouling trend value, and links the edge computing node to generate control commands.

[0023] Specifically, in step one, groundwater is extracted through wells or vacuum wells, and gap grids are installed in the pipelines to intercept large particles such as tree branches and plastic bags, so as to initially remove large particles and silt and avoid clogging subsequent treatment equipment (such as hydrocyclones and MABR membrane modules).

[0024] Step two involves applying a centrifugal force field to the incoming water in the collection tank using a hydrocyclone separator to remove denser sand and mineral particles. The effluent from the hydrocyclone is then introduced into a sedimentation tank where suspended sludge and sand naturally settle and separate under gravity. Simultaneously, a coagulant is added and stirred to promote the combination of suspended solids and coagulant, enhancing flocculation and sedimentation efficiency. Finally, the concentrated sludge at the bottom of the sedimentation tank is directionally transported to step five for dewatering treatment, obtaining pretreated water that meets the influent turbidity standards of the subsequent biofilm purification unit, reducing the load on subsequent biological treatment, and mitigating the risk of contamination for subsequent membrane separation.

[0025] In step three, salt-tolerant microbial communities (such as Halomonasspp) within the MABR membrane reactor are used to simultaneously degrade organic matter (chemical oxygen demand) and ammonia nitrogen (NH3-N) under bubble-free aeration conditions. At the same time, online sensors monitor water quality parameters, membrane flux, and aeration pressure in real time, providing data support for subsequent intelligent control. This improves the removal rate of organic matter and ammonia nitrogen, stabilizes the effluent quality, and increases oxygen utilization and reduces energy consumption. Subsequently, the sensor-collected data drives an LSTM neural network to predict water quality trends, ensuring efficient system operation.

[0026] Step four involves water quality-adaptive membrane separation treatment of the effluent from the membrane-aerated MABR membrane reactor. An online conductivity sensor determines the total dissolved solids concentration in real time and intelligently switches between ultrafiltration (UF) and reverse osmosis (RO) modes of the membrane-aerated MABR reactor. Simultaneously, a membrane fouling index model triggers ultrasonic cleaning to maintain membrane flux. Selective chemical oxidation is applied to the concentrated effluent and purified water with excessive levels of specific pollutants. A dual-channel system of electrochemical oxidation and catalytic ozone oxidation is used, dynamically matching the oxidation intensity and reaction path based on the detected target substances such as petroleum hydrocarbons and heavy metals. This generates purified water that meets industrial reuse standards, suitable for dust suppression in deep foundation pits and concrete mixing processes. The wastewater containing newly added pollutants is then returned to Step one to form a closed-loop cycle, achieving water resource recycling and reducing the risk of salinity and pollutant accumulation.

[0027] Step 5 involves electro-osmotic dewatering of the sludge discharged from the sedimentation tank and the residual sludge produced by the MABR membrane reactor with membrane aeration. A DC electric field is applied in conjunction with mechanical pressure to break down the sludge floc structure and efficiently remove bound water, generating a dewatered sludge cake with a moisture content of <40%. The sludge cake and construction waste are then processed into ceramsite to achieve resource reuse. Simultaneously, the filtrate generated during the electro-osmotic dewatering process is collected and returned to Step 2 for cyclone sedimentation and coagulation reprocessing, forming a closed-loop system of solid-phase sludge resource utilization and full liquid-phase recovery, achieving zero pollution discharge and in-situ regeneration of building material raw materials.

[0028] In step six, by collecting water quality parameters (such as chemical oxygen demand and ammonia nitrogen), membrane module status (such as membrane flux and transmembrane pressure difference) and environmental data (such as water temperature and flow rate) in real time, and combining them with LSTM neural network to predict water quality change trends, the MABR membrane aeration pressure, membrane separation mode and chemical oxidation intensity are dynamically adjusted to achieve intelligent control of the entire process. This significantly improves the system's operational stability, ensures that the effluent water quality continuously meets standards, effectively extends the membrane module cleaning cycle, reduces maintenance costs, and forms an efficient and low-cost groundwater recycling system.

[0029] The coagulant in step two includes low-aluminum residual PAC and bio-based coagulant.

[0030] Specifically, low-aluminum residual PAC forms polynuclear hydroxyl complexes through hydrolysis, which rapidly adsorb colloids and microparticles in water, forming flocs and accelerating sedimentation. Bio-based coagulants (such as chitosan derivatives) utilize the bridging effect of natural polymer chains to enhance the structural stability of flocs. At the same time, their biocompatibility avoids inhibiting the subsequent MABR biofilm. The synergy of the two can improve the suspended solids removal rate and enhance the solid-liquid separation effect. Low-aluminum residual PAC reduces the toxic effects of aluminum ions on biofilms, while bio-based coagulants are biodegradable and do not cause secondary pollution, reducing the environmental risks of chemical agents. Together, they reduce sludge production and make the flocs easier to dewater, creating conditions for subsequent sludge resource utilization.

[0031] In step three, the membrane aeration MABR membrane reactor uses hollow fiber membrane modules to form a composite biofilm under bubble-free aeration conditions, utilizing biofilm microorganisms to simultaneously degrade organic matter and ammonia nitrogen.

[0032] Specifically, a high specific surface area hollow fiber membrane module is used to provide sufficient attachment sites for microorganisms. Simultaneously, precise control of the air-side pressure achieves bubble-free aeration, ensuring efficient oxygen transfer to the biofilm interface and maintaining microbial activity. Bubble-free aeration avoids the energy loss and bubble interference of traditional aeration methods. In this embodiment, the hollow fiber membrane module has a specific surface area of ​​200–250 m² / m³, and the air-side pressure is controlled at 0.05–0.2 MPa.

[0033] In step four, the total dissolved solids concentration is monitored in real time using an online conductivity sensor. Ultrafiltration separation is activated when the total dissolved solids concentration (TDS) is less than 2000 mg / L; otherwise, reverse osmosis separation is switched to. The cleaning cycle is calculated using the MFI model formula, and ultrasonic flushing is triggered. The membrane flux recovery rate is ≥95%.

[0034] Specifically, the system monitors the total dissolved solids (TDS) concentration in real time using an online conductivity sensor, intelligently switching between ultrafiltration (UF) and reverse osmosis (RO) modes to provide appropriate deep treatment for water with different salinity levels. It also predicts membrane module fouling trends using the MFI model formula, automatically triggering ultrasonic cleaning to maintain stable membrane flux and ensure that the effluent quality meets reuse standards. This reduces energy consumption for treating high-salinity wastewater. Furthermore, the system extends membrane module lifespan and reduces the amount of chemical reagents used through membrane mode switching and preventative cleaning. Ultrasonic rinsing efficiently restores membrane performance, ensuring continuous and stable water purification operation.

[0035] The MFI model formula is as follows: ; in: The dynamic viscosity of water, Where T is the transmembrane pressure difference (TMP), A is the membrane area, t is the filtration time, and v is the filtrate volume. The above formula is used to calculate the membrane fouling index in real time. When the fouling level reaches the set severity level, the membrane module is automatically subjected to periodic ultrasonic rinsing. After rinsing, the system monitors the membrane flux recovery to ensure it meets the predetermined recovery standard. Simultaneously, this reduces energy consumption for treating high-salinity wastewater. Furthermore, membrane mode switching and preventative cleaning extend the membrane module's lifespan, reduce chemical reagent usage, and ultrasonic rinsing efficiently restores membrane performance, ensuring continuous and stable water purification operation.

[0036] The selective chemical oxidation in step four includes: Electrochemical oxidation, using boron-doped diamond electrodes, at a current density of 10–50 mA / cm², degrades petroleum hydrocarbon pollutants; Heterogeneous catalytic ozone oxidation is achieved through a honeycomb ceramic catalyst supported on composite metal oxides of iron and manganese, with an ozone dosage of 5–20 mg / L.

[0037] Specifically, through electrochemical oxidation (boron-doped diamond electrodes) or heterogeneous catalytic ozone oxidation (iron-manganese composite catalyst), recalcitrant organic matter (such as petroleum hydrocarbons) and heavy metal ions in the concentrated water from membrane separation are specifically degraded, achieving deep removal of pollutants. This ensures stable effluent quality that meets standards. The catalyst and electrode materials significantly improve oxidation reaction efficiency, reduce chemical reagent dosage, and the synergistic effect of selective oxidation and membrane separation extends the membrane module's lifespan, reduces system energy consumption, and avoids the secondary pollution problems of traditional oxidation processes, providing a reliable guarantee for the recycling of groundwater from deep foundation pits.

[0038] In step four, the recirculated water and the newly added groundwater are mixed at a volume ratio of 1:3 to 5.

[0039] Specifically, by diluting and reducing the salinity and residual pollutant concentration of the returned water, balancing water quality fluctuations, and ensuring the stable operation of subsequent steps two and three, the inhibitory effect of high salinity or high concentration of pollutants on biofilms can be effectively mitigated, the degradation efficiency of organic matter and ammonia nitrogen can be improved, and the amount of fresh water extracted can be reduced through water resource recycling, thereby achieving the goal of efficient reuse and low-carbon treatment of groundwater in deep foundation pits.

[0040] In step five, the DC electric field voltage for sludge electro-osmotic dewatering is 10-30V, and the mechanical pressure is 0.5-1.5MPa, reducing the sludge moisture content to below 40%.

[0041] Specifically, by applying a DC electric field and mechanical pressure in synergy, the colloidal structure of the sludge is disrupted and water is driven to migrate in a specific direction. At the same time, free water is further separated by mechanical extrusion, thus achieving efficient dewatering of the sludge.

[0042] In step five, the separated water flows back through a pipeline to the inlet of the hydrocyclone separator, where it is mixed with the newly added groundwater at a volume ratio of 1:2 to 4 before being treated.

[0043] By diluting and reducing the concentration of pollutants (such as residual organic matter, ammonia nitrogen, and heavy metals) in the filtrate, water quality fluctuations are balanced, providing stable influent conditions for the subsequent MABR membrane reactor. The dewatered filtrate is recycled back to step two for further treatment, reducing fresh water consumption, achieving resource utilization, avoiding landfill pollution, and forming a sustainable sludge management system of "reduction, resource utilization, and recycling," thus achieving the goal of efficient recycling and low-carbon treatment of groundwater in deep foundation pits.

[0044] The specific process in step six includes the following steps. S601, Real-time Acquisition of Multi-Source Data Sensors are installed at the inlet of the membrane-aerated MABR membrane reactor, the surface of the membrane module, and the outlet of the membrane-aerated MABR membrane reactor to collect operational data in real time, which is then used as a dataset, including: Turbidity sensors, TOC analyzers, pH / temperature composite sensors, and ammonia nitrogen sensors are deployed at the inlet of the membrane aeration MABR membrane reactor to monitor the raw water pollution load in real time. The membrane module surface is equipped with a laser biofilm thickness probe, a 9-point gridded dissolved oxygen microelectrode array, a transmembrane differential pressure sensor, and a patch-type temperature sensor to dynamically track biofilm activity and fouling status. The outlet is equipped with nitrate and nitrite sensors, an acute biotoxicity meter, a residual chlorine sensor, and a dissolved oxygen sensor to ensure that water quality meets standards and ecological safety. S602, Water Quality Change Trend Prediction Based on the historical operating dataset of the membrane aeration MABR reactor, including influent chemical oxygen demand, ammonia nitrogen concentration, dissolved oxygen gradient on the membrane surface, transmembrane pressure difference and biofilm thickness, a long short-term memory (LSTM) neural network was trained to capture water quality change characteristics and output the chemical oxygen demand fluctuation trajectory, ammonia nitrogen concentration change value and membrane fouling trend quantitative index value for future periods. S603, Dynamic Control Strategy Generation Based on the chemical oxygen demand fluctuation trajectory, ammonia nitrogen concentration changes, and membrane fouling trend quantitative indicators predicted by LSTM neural network, the following controls are implemented: Aeration pressure adjustment: The gas-side pressure P of the membrane aeration MABR membrane reactor is dynamically set according to the formula P=0.02×Cn+0.05, where Cn is the predicted ammonia nitrogen concentration; The method for selecting membrane separation mode is as follows: if the total dissolved solids concentration (TDS) is <2000 mg / L, ultrafiltration separation should be enabled. If the total dissolved solids concentration (TDS) is ≥2000 mg / L, switch to reverse osmosis separation; When the membrane fouling index is 3.0 to 6.0, sodium hypochlorite solution is injected into the inner cavity of the membrane aeration MABR membrane reactor.

[0045] Specifically, the S601 uses turbidity sensors, TOC analyzers, pH / temperature composite sensors, and ammonia nitrogen membrane fouling index sensors deployed at the inlet of the membrane aeration MABR membrane reactor to capture fluctuations in raw water pollution load in real time. Laser biofilm thickness probes, 9-point gridded dissolved oxygen microelectrode arrays, transmembrane pressure differential sensors, and patch-type temperature sensors configured on the membrane module surface dynamically track the evolution of biofilm activity and fouling status. Then, nitrate and nitrite sensors, acute biotoxicity meters, residual chlorine sensors, and dissolved oxygen sensors installed at the outlet verify the treatment effect and ecological safety. The three-position sensor network coordinates monitoring to provide a data foundation for subsequent water quality prediction and dynamic control. Based on historical operating data of the membrane aeration MABR reactor, S602 trains an LSTM neural network model to predict water quality parameters (such as organic matter and ammonia nitrogen concentration) and membrane fouling trends (such as membrane flux decay rate) for the next 30 minutes to 2 hours. This allows for early prediction of water quality changes and membrane fouling risks, providing a forward-looking basis for dynamic control. Furthermore, predictive maintenance reduces equipment downtime, extends membrane module cleaning cycles, and lowers maintenance costs. Based on the water quality change trends predicted by the LSTM model and the membrane fouling data, the S603 automatically adjusts the aeration pressure, membrane separation mode, and chemical oxidation intensity of the MABR membrane reactor to generate the optimal control strategy, dynamically optimize the energy consumption of aeration and membrane separation, achieve low-carbon operation, ensure that the effluent water quality continuously meets the standards, respond quickly to water quality fluctuations, and improve the overall operating efficiency and stability of the system.

[0046] Example 2: A method for purifying and recycling groundwater in deep foundation pit engineering includes the following steps: Step 1: Pre-treatment of groundwater through dewatering wells. Groundwater is pumped out through dewatering wells, and after being filtered by a screen to remove large particles of impurities, it enters the collection tank. Step 2: Cyclone sedimentation and coagulation. Fine sand is removed by passing the mixture through a cyclone separator and sediment is allowed to settle naturally in a sedimentation tank. Low-aluminum residual PAC and bio-based coagulant are added at a mass ratio of 3:1 to enhance flocculation. Step 3: Biofilm purification monitoring. Pretreated water is passed through a hollow fiber membrane module (specific surface area...). (200m² / m³, air-side pressure 0.05MPa), utilizing salt-tolerant bacteria to degrade organic matter and ammonia nitrogen under bubble-free aeration conditions, and monitoring water quality and membrane status in real time through NH3-N sensors, dissolved oxygen sensors, etc., with a sampling frequency of 3 times / minute; Step 4: Membrane separation and chemical oxidation. The total dissolved solids (TDS) concentration is monitored by an online conductivity sensor. When TDS < 2000 mg / L, ultrafiltration is activated; otherwise, reverse osmosis is switched. Electrochemical oxidation is used to degrade petroleum hydrocarbons at a current density of 10 mA / cm². Step 5: Sludge dewatering and resource utilization. Electro-osmotic dewatering is carried out using 10V voltage and 0.5MPa pressure to reduce the sludge moisture content to below 40%. The dewatered sludge is sintered with construction waste to produce ceramsite. The filtrate is returned to Step 2 and mixed with newly added groundwater at a volume ratio of 1:2 for treatment. Step 6: Intelligent monitoring and control. Predict water quality trends for the next 30 minutes using an LSTM neural network, and dynamically adjust aeration pressure, membrane separation mode, and chemical oxidation intensity.

[0047] Example 3: The differences between this embodiment and Implementation Two are as follows: Step 2: Swirl sedimentation coagulation, adding low-aluminum residual PAC and bio-based coagulant at a mass ratio of 2:1; Step 3: Biofilm purification monitoring, using high specific surface area hollow fiber membrane (250m² / m³, gas-side pressure 0.1MPa). Step 4: Membrane separation chemical oxidation, the total dissolved solids concentration (TDS) trigger threshold is increased to 2500 mg / L, switch to reverse osmosis mode, and use heterogeneous catalytic ozone oxidation to treat heavy metals, with an ozone dosage of 20 mg / L; Step 5: Sludge dewatering and resource utilization. Electro-osmotic dewatering is carried out using 30V voltage and 1.5MPa pressure to treat sludge with high water content. The filtrate is returned to Step 2 and mixed with the groundwater pumped from the dewatering well in Step 1 at a volume ratio of 1:4. Step 6: Intelligent monitoring and control. The LSTM neural network predicts water quality changes in the next 2 hours, adjusts aeration pressure and optimizes membrane separation strategy to adapt to high salinity shocks.

[0048] Table 1. Comparison of Working Condition Adaptability between Example 2 and Example 3

[0049] The adaptability of the working condition in Example 2 is as follows: Suitable for conventional groundwater quality (total dissolved solids (TDS) ≤ 2000 mg / L, chemical oxygen demand ≤ 500 mg / L), it achieves efficient flocculation and sedimentation through a 3:1 mixture of low-aluminum PAC and bio-based coagulant. MABR uses standard salt-tolerant bacteria to simultaneously remove nitrogen and carbon under bubble-free aeration. Ultrafiltration separation ensures low-energy operation, electrochemical oxidation (10 mA / cm²) targets and degrades petroleum hydrocarbon pollutants, sludge electro-osmotic dewatering (10 V / 0.5 MPa) matches medium sludge load, and the filtrate is mixed at a 1:2 volume ratio to maintain salinity balance. LSTM neural network predicts water quality trends over 30 minutes to achieve precise adjustment of aeration and separation modes, meeting the general groundwater purification needs of deep foundation pits.

[0050] The specific working condition adaptability of Example 3 is as follows: For high-salinity and highly polluted groundwater (total dissolved solids concentration, TDS ≤ 2500 mg / L, containing heavy metals / recalcitrant organic matter), hollow fiber membranes with high specific surface area enhance the salt tolerance of biofilms, the reverse osmosis mode switching threshold is increased to 2500 mg / L to cope with high salt shocks, catalytic ozone oxidation (20 mg / L) efficiently decomposes heavy metals and recalcitrant substances, high-pressure dewatering of sludge (30V / 1.5MPa) treats highly viscous sludge, the filtrate is highly diluted at a volume ratio of 1:4 to prevent salinity accumulation, and LSTM neural network extends prediction to 2 hours and optimizes high-salinity operating conditions control strategies to ensure system stability under extreme water quality conditions.

[0051] Example 4: A groundwater purification and recycling system for deep foundation pit engineering includes: The groundwater pretreatment module enables preliminary filtration and temporary storage of groundwater, intercepting large particulate impurities. The cyclone sedimentation coagulation module removes suspended solids from groundwater pretreated by the dewatering well pretreatment module through centrifugal sand removal and gravity sedimentation, and adds composite coagulant to enhance solid-liquid separation. The sludge resource utilization module dewaters the sludge treated by the cyclone sedimentation and coagulation module, solidifies it, and then co-sintersects it with construction waste to form ceramsite. The filtrate is returned to the cyclone sedimentation module for further treatment. The membrane aeration MABR bioreactor module degrades pollutants in the pretreated water after the swirl sedimentation coagulation module under bubble-free aeration conditions, and collects data of each process of biological treatment in real time through a three-position sensor network. The membrane separation chemical oxidation module selectively oxidizes the concentrated water and specific pollutants treated by the membrane aeration MABR biological reactor module based on water quality switching membrane separation mode, and the purified water that meets the standards is reused in the project or returned to the rainwater well pretreatment module. The intelligent control center module uses LSTM model prediction driven by sensor data in the membrane aeration MABR reactor to dynamically optimize the operating parameters of biological reaction, membrane separation and chemical oxidation.

[0052] Specifically, the groundwater pretreatment module performs initial filtration and interception of large particles in the groundwater, and then transports the effluent to the cyclone sedimentation and coagulation module for centrifugal sand removal and gravity sedimentation. Simultaneously, low-aluminum residual PAC and bio-based coagulants are added to enhance flocculation. The cyclone sedimentation and coagulation module separates sludge to the sludge resource utilization module, and the pretreated water is then transported to the membrane aeration MABR bioreactor module. Under bubble-free aeration conditions, pollutants are degraded by the biofilm on the surface of the hollow fiber membrane. Meanwhile, a three-position sensor network at the inlet, membrane module surface, and outlet of the membrane aeration MABR bioreactor module collects water quality and biofilm status data in real time. The water at the outlet enters the membrane separator... The chemical oxidation module intelligently switches between ultrafiltration and reverse osmosis separation modes based on real-time monitoring of TDS (Total Dissolved Solids). Simultaneously, it selectively oxidizes the concentrate, and the treated water that meets the standards is reused in engineering projects or returned to the pretreatment module of the rainwater well. Afterward, the sludge resource utilization module dewaters the received sludge through electro-osmosis and then sintersulates it with construction waste into ceramsite. The filtrate is returned to the cyclone sedimentation and coagulation module for further mixing and treatment. The intelligent control center module uses a three-position sensor dataset to drive an LSTM model to predict water quality and membrane fouling trends, dynamically optimizes the aeration pressure, membrane separation and cleaning strategy, and oxidation intensity of the membrane aeration MABR reactor, forming a closed-loop intelligent control of the entire system.

[0053] The membrane aeration MABR bioreactor module in this embodiment uses a membrane aeration MABR membrane reactor.

[0054] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

[0055] Where there is no conflict, those skilled in the art can combine the relevant technical features in the above examples according to the actual situation to achieve the corresponding technical effects. Specific details of the various combinations will not be elaborated here.

[0056] The above description is merely a preferred embodiment of the present invention. The present invention is not limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein. Any simple modifications, equivalent variations, and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for purifying and recycling groundwater in deep foundation pit engineering, characterized in that: Includes the following steps, Step 1: Pre-treatment with dewatering wells; groundwater is pumped out through dewatering wells, and after passing through a screen to intercept large particles of impurities, it enters the collection tank; Step 2: Cyclone sedimentation and coagulation; the incoming water to the collection tank is centrifugally desanded using a cyclone separator, and the effluent is introduced into a sedimentation tank for natural sedimentation of sludge and sand. Coagulant is added, and flocculation is enhanced by stirring. The sludge at the bottom of the sedimentation tank is periodically discharged to Step 5 for treatment to obtain pretreated water. Step 3: Biofilm purification monitoring; The pretreated water obtained in Step 2 is fed into the membrane aeration MABR membrane reactor. Under bubble-free aeration conditions, the biofilm simultaneously degrades organic matter and ammonia nitrogen, and the water quality parameters and the operating status of the MABR membrane purification are monitored in real time by sensors. Step 4: Membrane separation and chemical oxidation; The effluent from the membrane-aerated MABR membrane reactor is subjected to membrane separation and selective chemical oxidation to generate purified water for use. At the same time, the used water is returned to Step 1 for further treatment. Step 5: Sludge dewatering and resource utilization; The sludge from the sedimentation tank and biological purification is dewatered by electro-osmosis and then co-utilized with construction waste for resource utilization. The dewatered water is then transferred back to Step 2 for further treatment. Step Six: Intelligent Monitoring and Control; Based on the sensor datasets of the inlet, membrane module surface and outlet of the membrane aeration MABR membrane reactor, a dynamic water quality prediction model is constructed through LSTM neural network, which outputs the chemical oxygen demand fluctuation trajectory and membrane fouling trend value, and links the edge computing node to generate control commands.

2. The method for purifying and recycling groundwater in deep foundation pit engineering as described in claim 1, characterized in that: The coagulant in step two includes low-aluminum residual PAC and bio-based coagulant.

3. The method for purifying and recycling groundwater in deep foundation pit engineering as described in claim 1, characterized in that: The membrane aeration MABR membrane reactor in step three uses hollow fiber membrane modules to form a composite biofilm under bubble-free aeration conditions, utilizing biofilm microorganisms to simultaneously degrade organic matter and ammonia nitrogen.

4. The method for purifying and recycling groundwater in deep foundation pit engineering as described in claim 1, characterized in that: Step four involves real-time monitoring of the total dissolved solids (TDS) concentration using an online conductivity sensor. When the TDS concentration is less than 2000 mg / L, ultrafiltration separation is activated; otherwise, reverse osmosis separation is switched to TDS. The cleaning cycle is calculated using the MFI model formula, and ultrasonic rinsing is triggered. The membrane flux recovery rate is ≥95%.

5. The method for purifying and recycling groundwater in deep foundation pit engineering as described in claim 1, characterized in that: The selective chemical oxidation in step four includes: Electrochemical oxidation, using boron-doped diamond electrodes, at a current density of 10–50 mA / cm², degrades petroleum hydrocarbon pollutants; Heterogeneous catalytic ozone oxidation is achieved through a honeycomb ceramic catalyst supported on composite metal oxides of iron and manganese, with an ozone dosage of 5–20 mg / L.

6. The method for purifying and recycling groundwater in deep foundation pit engineering as described in claim 1, characterized in that: In step four, the refluxed water and the newly added groundwater are mixed at a volume ratio of 1:3 to 5.

7. The method for purifying and recycling groundwater in deep foundation pit engineering as described in claim 1, characterized in that: In step five, the sludge undergoes electro-osmotic dewatering with a DC electric field voltage of 10-30V and a mechanical pressure of 0.5-1.5MPa, reducing the sludge moisture content to below 40%.

8. The method for purifying and recycling groundwater in deep foundation pit engineering as described in claim 1, characterized in that: In step five, the separated water flows back through a pipeline to the inlet of the hydrocyclone separator, where it is mixed with the newly added groundwater at a volume ratio of 1:2 to 4 before being treated.

9. A method for purifying and recycling groundwater in deep foundation pit engineering as described in claim 1, characterized in that: The specific process in step six includes the following steps. S601, Real-time Acquisition of Multi-Source Data Sensors are installed at the inlet of the membrane-aerated MABR membrane reactor, the surface of the membrane module, and the outlet of the membrane-aerated MABR membrane reactor to collect operational data in real time, which is then used as a dataset, including: Turbidity sensors, TOC analyzers, pH / temperature composite sensors, and ammonia nitrogen sensors are deployed at the inlet of the membrane aeration MABR membrane reactor to monitor the raw water pollution load in real time. The membrane module surface is equipped with a laser biofilm thickness probe, a 9-point gridded dissolved oxygen microelectrode array, a transmembrane differential pressure sensor, and a patch-type temperature sensor to dynamically track biofilm activity and fouling status. The outlet is equipped with nitrate and nitrite sensors, an acute biotoxicity meter, a residual chlorine sensor, and a dissolved oxygen sensor to ensure that water quality meets standards and ecological safety. S602, Water Quality Change Trend Prediction Based on the historical operating dataset of the membrane aeration MABR reactor, which includes parameters such as influent chemical oxygen demand (COD), ammonia nitrogen concentration, dissolved oxygen gradient on the membrane surface, transmembrane pressure difference, and biofilm thickness, a long short-term memory (LSTM) neural network was trained to capture water quality abrupt changes and output the future COD fluctuation trajectory, ammonia nitrogen concentration change value, and membrane fouling trend quantitative indicators. S603, Dynamic Control Strategy Generation Based on the chemical oxygen demand (COD) fluctuation trajectory, ammonia nitrogen concentration changes, and membrane fouling trend quantitative indicators predicted by LSTM neural networks, the following controls were implemented: Aeration pressure adjustment: The gas-side pressure P of the membrane aeration MABR membrane reactor is dynamically set according to the formula P=0.02×Cn+0.05, where Cn is the predicted ammonia nitrogen concentration. When the predicted biofilm thickness is >300μm, switch to pulse aeration mode; The method for selecting membrane separation mode is as follows: if the total dissolved solids concentration (TDS) is <2000 mg / L, ultrafiltration separation should be enabled. If the total dissolved solids concentration (TDS) is ≥2000 mg / L, switch to reverse osmosis separation; When the membrane fouling index is 3.0 to 6.0, sodium hypochlorite solution is injected into the inner cavity of the membrane aeration MABR membrane reactor.

10. A groundwater purification and recycling system for deep foundation pit engineering, characterized in that, include: The groundwater pretreatment module enables preliminary filtration and temporary storage of groundwater, intercepting large particulate impurities. The cyclone sedimentation coagulation module removes suspended solids from groundwater pretreated by the dewatering well pretreatment module through centrifugal sand removal and gravity sedimentation, and adds composite coagulant to enhance solid-liquid separation. The sludge resource utilization module dewaters the sludge treated by the cyclone sedimentation and coagulation module, solidifies it, and then co-sintersects it with construction waste to form ceramsite. The filtrate is returned to the cyclone sedimentation module for further treatment. The membrane aeration MABR bioreactor module degrades pollutants in the pretreated water after the swirl sedimentation coagulation module under bubble-free aeration conditions, and collects data of the entire biological treatment process in real time through a three-position sensor network. The membrane separation chemical oxidation module selectively oxidizes the concentrated water and specific pollutants treated by the membrane aeration MABR biological reactor module based on water quality switching membrane separation mode, and the purified water that meets the standards is reused in the project or returned to the pretreatment module. The intelligent control center module uses LSTM model prediction driven by sensor data in the membrane aeration MABR reactor to dynamically optimize the operating parameters of biological reaction, membrane separation and chemical oxidation.

Images