Process for separating oxygen by three-tower pressure equalization cycle pressure swing adsorption

Abstract

This invention relates to the field of gas separation technology, specifically to a three-tower pressure-equalizing circulating pressure swing adsorption process for oxygen separation. It comprises three adsorption towers, each filled with a bottom layer of high-efficiency activated carbon and an upper layer of molecular sieve separation. A single-tower cycle includes: at the end of adsorption gas production, high-purity oxygen from the top of the tower is introduced counter-currently into the middle section of the tower via an external bypass, performing in-situ self-displacement and pushing nitrogen back to the bottom activated carbon layer via mass transfer; during pressure equalization, a microporous distribution tube inserted into the central axis of the tower is used to inject gas radially into the molecular sieve layer of the booster tower, suppressing gradient disruption; during vacuum desorption, a trace amount of product oxygen is continuously injected at the interface between the activated carbon and the molecular sieve to form a top-down aerodynamic diaphragm, preventing desorbed impurities from the activated carbon from flowing upwards and contaminating the molecular sieve. This invention improves oxygen yield and purity stability and extends the service life of the core adsorbent through synergistic improvements in the flow field topology and cycle sequence.

Description

Technical Field

[0001] This invention relates to the field of gas separation technology, and more specifically, to a process for separating oxygen using a three-tower pressure equalization circulating pressure swing adsorption method. Background Technology

[0002] Pressure swing adsorption (PSA) gas separation technology is widely used in industrial and medical oxygen supply. In existing three-tower PSA oxygen generation processes, three adsorption towers are typically used in a cyclical alternation of adsorption, pressure equalization, desorption, and pressurization processes to achieve continuous oxygen production and improve energy efficiency. Among these processes, the pressure equalization step is crucial for recovering high-pressure oxygen-enriched gas from the bed and reducing compression energy consumption. However, existing three-tower pressure equalization processes still face the following technical bottlenecks in practical operation: Traditional pressure equalization steps typically employ a conventional axial connection method, involving simple end-to-end gas flow between the depressurization tower and the booster tower (e.g., top-to-top or bottom-to-bottom). Initially, due to the significant pressure difference between the two towers, the transient gas flow velocity is high, easily causing substantial axial backmixing within the adsorption tower. This backmixing disrupts the oxygen-nitrogen concentration distribution established during the adsorption stage, causing the nitrogen-rich mass transfer zone front gas in the lower part of the depressurization tower to diffuse towards the oxygen-rich upper region, entering the booster tower with the pressure equalization gas flow. This reduces the purity of the equalization gas and disrupts the concentration gradient within the booster tower.

[0003] To prevent moisture, carbon dioxide, and hydrocarbon impurities in the feed air from poisoning and deactivating the molecular sieve, existing processes often use activated carbon as a protective layer, layered at the bottom of the adsorption tower below the molecular sieve. During the adsorption stage, the feed gas flows from bottom to top, and the activated carbon initially traps moisture and impurities. However, in the subsequent reverse vacuum desorption stage, the entire tower is under negative pressure, and the gas flow is reversed. At this time, moisture and impurities desorbed from the bottom activated carbon protective layer under negative pressure are easily carried away by the reverse airflow, permeating or entraining them and diffusing towards the edge of the molecular sieve layer. This causes the adsorption performance of the molecular sieve to gradually decrease, shortening the effective service life of the adsorbent.

[0004] While some existing technologies attempt to improve oxygen production performance by adjusting operating parameters such as equalization time, cutting pressure, or changing adsorbent loading ratio, these conventional adjustments at the parameter level are insufficient to effectively address axial backmixing during equalization and reverse permeation of impurities during desorption. Therefore, optimizing the process flow and timing to suppress concentration gradient disruption and prevent interlayer contamination of the adsorbent is a pressing technical challenge in this field. Summary of the Invention

[0005] The purpose of this invention is to provide a three-tower pressure equalization circulating pressure swing adsorption process for separating oxygen, in order to solve the technical problems mentioned in the background art of existing three-tower pressure swing adsorption oxygen production processes, such as the disturbance of the mass transfer zone and the destruction of the concentration gradient caused by the impact of the axial high-speed airflow during the pressure equalization stage, and the reverse diffusion and contamination of the molecular sieve by impurities trapped in the bottom protective layer during the reverse vacuum desorption stage.

[0006] To achieve the above objectives, this invention provides a three-tower pressure swing adsorption (PSA) process for separating oxygen. This process operates alternately in a PSA system consisting of three parallel adsorption towers. Each adsorption tower contains, from bottom to top, a high-efficiency activated carbon protective layer and a molecular sieve separation layer. A complete single-tower operating cycle includes the following steps performed in sequence: S1. Adsorption and In-situ Self-Displacement Process: After pretreatment, the raw air is introduced from the bottom inlet of the adsorption tower. Under adsorption pressure, it flows from bottom to top through the high-efficiency activated carbon protective layer and the molecular sieve separation layer. Nitrogen and moisture in the air are selectively adsorbed, and oxygen is discharged from the top outlet and enters the oxygen buffer tank. At the end of the adsorption gas production period, the bottom raw air inlet valve is closed, and the self-circulation bypass located outside the adsorption tower is opened. A portion of the high-purity oxygen-enriched gas output from the top of the adsorption tower is reverse-directed to the middle section of the side wall of the adsorption tower and injected into the tower. The middle section of the side wall is located above the junction of the high-efficiency activated carbon protective layer and the molecular sieve separation layer. The injected high-purity oxygen-enriched gas is used to compress the nitrogen mass transfer band in the lower middle part of the tower towards the bottom, causing it to retreat back into the high-efficiency activated carbon protective layer before breaking through the molecular sieve layer at the forefront of the mass transfer band. S2, Radial Distribution Pressure Equalization and Reduction Process: After completing the in-situ self-displacement process, the adsorption tower acts as a pressure reduction tower, and the corresponding pressure equalization connection valve group is opened to connect with another adsorption tower in the final stage of reverse vacuum desorption and in a low-pressure state as a pressure boosting tower for pressure equalization and connection; the gas discharged from the pressure reduction tower is introduced into the pressure boosting tower through a pipeline, and injected into the molecular sieve separation layer of the pressure boosting tower in the form of radial flow from the center to the surrounding area through the microporous distribution tube inserted at the central axis of the pressure boosting tower, until the internal pressure of the pressure reduction tower and the pressure boosting tower reach equilibrium; S3. Gas-sealed reverse vacuum desorption process: After the pressure equalization and reduction are completed, the pressure equalization connecting valve group between the two towers is closed, and the vacuum valve at the bottom of the adsorption tower is opened. The vacuum pump is used to perform reverse vacuum desorption from top to bottom in the tower, and the nitrogen and impurity components adsorbed on the molecular sieve and high-efficiency activated carbon are discharged through the bottom of the tower, so that the pressure of the entire tower is reduced to the set minimum vacuum desorption pressure. During the entire process of vacuum desorption, a small amount of product oxygen is continuously injected into the interlayer gap between the high-efficiency activated carbon protective layer and the molecular sieve separation layer through the gas-sealed gas injection valve set in the middle section of the tower wall. Under the action of vacuum negative pressure, the small amount of product oxygen forms a pneumatic membrane that flows directionally from top to bottom in the interlayer gap, blocking the reverse migration and diffusion of the impurity components desorbed from the high-efficiency activated carbon protective layer to the molecular sieve separation layer. S4. Pressurization process: Close the vacuum valve and gas seal injection valve, accept radial pressure equalization gas discharged from the other adsorption towers for initial pressurization, then open the top product oxygen inlet valve, and use product oxygen from the product oxygen buffer tank for pressurization until the pressure of the entire tower rises back to the adsorption pressure, ready to enter the adsorption and in-situ self-displacement process of the next cycle.

[0007] Preferably, in the adsorption and in-situ self-displacement process of S1, the transient flow rate of the high-purity oxygen-enriched gas introduced by the self-circulation bypass is 3% to 8% of the current transient gas flow rate of the adsorption tower, and the execution time of the in-situ self-displacement is 5% to 10% of the total time of the entire adsorption stage.

[0008] Preferably, the specific geometric height of the injection tower into the middle section of the side wall of S1 is set at 35% to 45% of the total height of the adsorption bed in the adsorption tower.

[0009] Preferably, in the radial distribution pressure equalization and pressure reduction process of S2, the outer wall of the microporous distribution tube is provided with an array of micropores, the diameter of a single micropore is 0.5mm to 2.0mm, and the axial length of the microporous distribution tube extending inside the molecular sieve separation layer accounts for 60% to 80% of the total height of the molecular sieve separation layer.

[0010] Preferably, in the radial distribution pressure equalization and depressurization process of S2, when the gas discharged from the depressurization tower flows through the microporous distribution pipe into the booster tower, the axial linear velocity of the fluid is converted into radial divergence velocity after entering the microporous distribution pipe, so that the linear velocity of the airflow on the cross section of the molecular sieve separation layer is reduced to 10% to 15% of the conventional axial flow pressure equalization linear velocity.

[0011] Preferably, in the reverse vacuum desorption process of the gas seal isolation in S3, the flow rate of the trace product oxygen injected into the interlayer gap through the gas seal injection valve is controlled to be 1.5% to 3.5% of the transient vacuum exhaust flow rate of the current process.

[0012] Preferably, the high-efficiency activated carbon protective layer uses nitrogen-doped hierarchical porous coconut shell-based modified activated carbon, wherein the volume of micropores with a pore size of less than 2 nm accounts for 75% to 80%, the volume of mesopores with a pore size of 2 nm to 50 nm accounts for 18% to 22%, and the filling height of the high-efficiency activated carbon protective layer accounts for 15% to 25% of the total height of the adsorption bed in the entire tower.

[0013] Preferably, the high-efficiency activated carbon protective layer, when subjected to radially distributed pressure reduction and lateral transient pressure relief, utilizes its mesoporous network as a pressure stress buffer for local transient pressure drop, absorbing the pressure drop impact, and relies on the weak throttling and cooling effect within the activated carbon layer to suppress the accumulation of adsorption heat.

[0014] Preferably, the molecular sieve separation layer is composed of a modified 13X molecular sieve layer and a high-efficiency lithium molecular sieve layer stacked from bottom to top, with the filling thickness ratio of the two being 1:1 to 2:1.

[0015] Preferably, the pressure swing adsorption system consists of three adsorption towers, A, B, and C, connected in parallel. Under the centralized control of a PLC programmable logic controller, the three towers operate alternately and staggered according to their respective working sequences with a phase difference of 120 degrees. The oxygen production pressure of the system pulsates between 0.4 MPa and 0.55 MPa, thereby achieving uninterrupted output of product oxygen with a purity of ≥93%.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: In this three-tower pressure equalization circulating pressure swing adsorption process for separating oxygen, an in-situ self-displacement step is added at the end of the adsorption period. The oxygen-rich gas at the top of the self-circulating tower pushes back the nitrogen mass transfer band, confining the leading edge of the mass transfer band within the protective layer at the bottom before the pressure equalization action occurs. This improves the purity background of the pressure equalization gas in subsequent steps, avoids contamination of the high-purity oxygen area by medium and low-purity gases, and effectively improves the oxygen recovery rate of the system.

[0017] Secondly, a central distribution pipe is used to radiate the high-speed airflow radially in all directions. Due to the significantly enlarged cross-sectional area of ​​the flow channel, the linear velocity of the airflow within the molecular sieve bed is greatly reduced, thereby physically eliminating the mechanical impact of the high-speed gas jet on the adsorbent bed, suppressing axial backmixing, and maintaining a stable residual concentration gradient within the adsorption tower. By introducing product oxygen between the high-efficiency activated carbon protective layer and the molecular sieve separation layer to create a gas seal, a continuous downward directional weak airflow is formed, constituting a dynamic pneumatic diaphragm. This diaphragm blocks the upward diffusion and permeation of impurities such as moisture and carbon dioxide desorbed from the protective layer with the reverse airflow, protecting the molecular sieve separation layer from chronic toxicity and extending the service life of the core adsorbent, ensuring the long-term stability of the product oxygen purity.

[0018] Furthermore, the high-efficiency activated carbon of the present invention, with its hierarchical porous structure, acts as a stress buffer network during lateral transient pressure relief, reducing the risk of adsorbent pulverization. At the same time, it utilizes the weak throttling and cooling effect during the desorption process to alleviate the accumulation of adsorption heat in the bed, thereby improving the robustness of the oxygen production system. Attached Figure Description

[0019] Figure 1 This is an overall flowchart of Embodiment 1 of the present invention. Detailed Implementation

[0020] The technical solutions in 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 some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1: This example uses a parallel three-tower system consisting of towers A, B, and C. The total effective internal adsorption bed height of each adsorption tower is 5.0 meters, and the tower diameter is 1.2 meters. A 2.2-meter-long microporous distribution tube is built into the center of the tower along the axial direction. The outer wall of the microporous distribution tube has perforations with a pore size of 1.0 mm arranged in an array, resulting in an overall porosity of 15%.

[0022] The adsorption tower is filled with the following materials from bottom to top: The bottom layer is a high-efficiency activated carbon protective layer with a filling height of 1.0 meter (accounting for 20% of the total height of the adsorption bed). It uses nitrogen-doped hierarchical porous coconut shell-based modified activated carbon, with micropores (pore size less than 2nm) accounting for 76% of the volume and mesopores (pore size 2nm~50nm) accounting for 20% of the volume. The middle and top layers are molecular sieve separation layers, with a total filling height of 4.0 meters. From bottom to top, it consists of a 2.2-meter-high modified 13X molecular sieve layer and a 1.8-meter-high high-efficiency lithium molecular sieve (Li-LSX) layer stacked together.

[0023] A transverse air inlet is made on the tower wall at a distance of 2.0 meters from the bottom of the adsorption tower (i.e., 40% of the total height of the adsorption bed, 1.0 meter above the junction of the high-efficiency activated carbon protective layer and the molecular sieve separation layer), which is connected to a self-circulation bypass valve and a gas seal injection valve, respectively. This geometric height corresponds to the region where the nitrogen mass transfer front resides at the end of adsorption.

[0024] The entire system is controlled by an integrated PLC logic controller. The operating cycle of a single tower is set to 150 seconds, and the specific process actions are staggered by 120 degrees between towers A, B, and C. Taking a complete cycle of tower A as an example, the specific process steps and valve action logic are as follows: S1. Adsorption and In-situ Self-Displacement Process (Running Time: 0-65 seconds): The raw material air inlet valve at the bottom of Tower A opens, allowing clean raw material air, after preliminary filtration and pressure adjustment to 0.48 MPa (gauge pressure), to enter from the bottom of Tower A. The airflow first passes through the 1.0-meter high-efficiency activated carbon protective layer at the bottom, where most of the moisture, carbon dioxide, and trace amounts of gaseous organic matter in the air are trapped and captured. Subsequently, the airflow rises and passes sequentially through the modified 13X molecular sieve layer and the high-efficiency lithium molecular sieve layer, where nitrogen is selectively adsorbed, and the unadsorbed oxygen is enriched in the gas phase and discharged from the top of Tower A, entering the oxygen buffer tank through the opened gas generation valve. When the operation reaches 58 seconds, the PLC controller closes the raw material air inlet valve at the bottom and the gas generation valve at the top, while simultaneously opening the external self-circulation bypass valve of Tower A. A portion of the high-purity product oxygen (purity ≥94%) at the top of column A is injected laterally into the column from a position 2.0 meters from the bottom (40% of the column height) under the influence of the pressure difference between the early and late stages and the external booster pump. This transient flow, equivalent to 5% of the current normal production flow rate, forms a downward fluid propulsion surface in the middle section of the molecular sieve layer, compressing and pushing the nitrogen mass transfer belt downwards as a whole. This causes the nitrogen's leading edge to retreat back to the bottom 1.0-meter high-efficiency activated carbon protective layer at 65 seconds, completing the in-situ self-displacement of the adsorption bed.

[0025] S2. Radial Distribution Pressure Equalization and Reduction Process (Running Time: 65-80 seconds): After self-displacement, the internal pressure of tower A is maintained at approximately 0.45 MPa. At this time, the PLC controller closes the self-circulation bypass valve and opens the pressure equalization connecting valve group between tower A and tower B, which is in the final stage of vacuum desorption (system low pressure is -0.05 MPa). Tower A acts as a pressure reduction tower, and the gas inside it is gradually released; tower B acts as a pressure boosting tower, receiving the gas. This pressure equalization gas flow is guided through the pipeline to the microporous distribution pipe at the central axis of tower B, and then through the 1.0 mm array micropores on the outer wall, it is diffused radially from the central region of tower B towards the surrounding circular tower walls and injected into the molecular sieve separation layer of tower B in a horizontal radial flow pattern. During this process, the cross-sectional area of ​​the fluid flow suddenly expands from the original ordinary axial circular cross-section to the cylindrical cross-section of the distribution pipe, and the linear velocity of the gas flow on the cross-section of the molecular sieve separation layer decreases to about 12% of the original axial linear velocity. By the end of the 80-second pressure equalization period, the internal pressures of towers A and B reached equilibrium, approximately 0.20 MPa. During this period, the bottom high-efficiency activated carbon protective layer, utilizing its mesoporous network, acted as a pressure stress buffer, absorbing the system's pressure drop impact.

[0026] S3. Gas-Sealed Reverse Vacuum Desorption Process (Running Time: 80-115 seconds): After pressure equalization and reduction, the PLC closes the pressure equalization valve between towers A and B. The vacuum desorption valve at the bottom of tower A is opened, and the connected screw vacuum pump is started to perform reverse vacuuming of tower A from top to bottom. As the gas inside the tower is reversely extracted, the internal pressure of tower A rapidly decreases from 0.20 MPa to the minimum vacuum desorption pressure of -0.05 MPa. Nitrogen adsorbed on the molecular sieve and impurities adsorbed on the high-efficiency activated carbon are desorbed and discharged from the bottom of the tower. During the entire 35-second vacuum desorption process, the PLC simultaneously opens the gas-sealed injection valve located 2.0 meters from the tower wall, continuously injecting product oxygen from the external oxygen buffer tank into the tower. The injection flow rate is locked at 2.5% of the current transient vacuum exhaust flow rate through the regulating valve. Under the negative pressure of the entire tower, this trace amount of oxygen forms a continuous, downward-flowing, directional, weak airflow (dynamic aerodynamic diaphragm) at the interlayer gap, blocking the channel for the trace impurities released from the protective layer to migrate and diffuse in the reverse direction to the molecular sieve separation layer above.

[0027] S4. Pressurization Process (Running Time: 115-150 seconds): After vacuum desorption is completed, the PLC closes the vacuum valve at the bottom and the gas seal injection valve in the middle. Tower A first opens the equalizing valve between itself and Tower C, which is currently in the final stage of adsorption, to receive the central radial equalizing gas discharged from Tower C, performing primary pressurization (115-130 seconds) to raise the tower pressure from -0.05 MPa to approximately 0.20 MPa. Then, the equalizing valve is closed, and the final product pressurization valve at the top is opened, introducing product oxygen from the product oxygen buffer tank to perform a final reverse pressurization of Tower A (130-150 seconds) until the overall tower pressure returns to the initial adsorption pressure of 0.48 MPa. Tower A completes the closed-loop cycle, and the system controls Towers B and C to repeat the above actions in parallel with staggered timing.

[0028] Example 2: In this example, the total height of the effective internal adsorption bed of the adsorption tower is 5.0 meters, and the tower diameter is increased to 2.0 meters. The bottom high-efficiency activated carbon protective layer is filled to a height of 1.25 meters (accounting for 25% of the total height), of which micropores (pore size less than 2nm) account for 79% of the volume and mesopores (pore size 2nm~50nm) account for 19%. The geometric height of the air inlet in the middle section of the side wall is set at 1.8 meters from the bottom of the tower (i.e., accounting for 36% of the total height of the adsorption bed). The axial length of the micropore distribution tube extending inside the molecular sieve separation layer accounts for 75% of the total height of the molecular sieve separation layer, and the diameter of a single micropore is 1.8mm.

[0029] The PLC-controlled single-tower operation cycle is 150 seconds, and the specific action parameters are adjusted accordingly: At the end of the adsorption gas production stage (at the 59th second), the self-circulation bypass is activated, and the transient flow rate of the introduced high-purity oxygen-enriched gas is adjusted to 8% of the current transient gas production flow rate. The in-situ self-displacement execution time is 7 seconds (accounting for 10.7% of the total adsorption stage time). High-flow-rate product oxygen is injected laterally at the 36% height position, forming a downward propulsion surface inside the bed, pushing the nitrogen mass transfer belt back into the bottom 1.25-meter activated carbon protective layer.

[0030] In the radial distribution pressure equalization and depressurization process, the gas from the depressurization tower is injected laterally into the molecular sieve layer of the booster tower through a microporous distribution tube. Due to the increased tower diameter and the extension length of the microporous tube reaching 75%, the fluid linear velocity is reduced to 10% of the linear velocity of the traditional axial flow pressure equalization, thus weakening the axial impact of the airflow on the bed.

[0031] In the reverse vacuum desorption process with gas seal isolation, the injection flow rate of the gas seal isolation oxygen is locked at 3.5% of the transient vacuum exhaust flow rate of the current process. At this flow rate, the product oxygen forms a dynamic aerodynamic diaphragm in the interlayer gap at a height of 1.8 meters, blocking the upward backflow diffusion of desorbed impurities.

[0032] The remaining pressurization and gas generation steps are the same as in Example 1.

[0033] Comparative test with existing technologies: Using the apparatus and process path described in Examples 1 and 2, a pressure swing adsorption (PSA) process system was compared with a system that removed the in-situ self-displacement bypass, the radial microporous distribution pipe (replaced with conventional top and bottom axial ventilation pressure equalization), and the desorption gas seal isolation valve, relying only on the conventional three-tower time / pressure parameter optimization, under the same loading volume and the same raw material air conditions (containing dust, moisture, and trace amounts of volatile organic compounds) in a long-term operation comparison test. The measured operating technical indicators are shown in the table below:

[0034] As shown in the table above, both Embodiments 1 and 2 of this invention, under different filling scales and control boundaries, reduced the concentration gradient backmixing caused by axial impact, improved the initial purity of the gas entering subsequent steps, increased the oxygen recovery and extraction rate from 58.5% to 72.8%~73.2%, and reduced oxygen production power consumption. Simultaneously, due to the dynamic interception protection of the pneumatic diaphragm during the vacuum desorption stage, the performance decay rate of the molecular sieve layer was controlled below 0.9%, extending the effective service life of the core adsorbent system.

[0035] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A process for separating oxygen using a three-tower pressure swing adsorption system, wherein the process operates alternately in a pressure swing adsorption system consisting of three parallel adsorption towers, and each adsorption tower is provided with a high-efficiency activated carbon protective layer and a molecular sieve separation layer from bottom to top; characterized in that, A complete single-tower operating cycle includes the following steps performed in sequence: S1. Adsorption and In-situ Self-displacement Process: Raw air is introduced from the bottom of the adsorption tower for adsorption and oxygen production; at the end of the adsorption and gas production period, the supply of raw air is stopped, and part of the oxygen-rich gas at the top of the adsorption tower is reverse-flowed to the middle section of the side wall of the adsorption tower through an external self-circulation bypass and injected into the tower, which carries the nitrogen gas in the tower to the bottom and compresses it into the high-efficiency activated carbon protective layer; the middle section of the side wall is located above the junction of the high-efficiency activated carbon protective layer and the molecular sieve separation layer; S2, Radial distribution pressure equalization and depressurization process: After self-displacement, the adsorption tower is connected to another pressure boosting tower in a low-pressure state as a pressure reducing tower; the gas discharged from the pressure reducing tower is injected into the molecular sieve separation layer of the pressure boosting tower in the form of radial flow through the microporous distribution tube inserted into the central axis of the pressure boosting tower for pressure equalization; S3, Gas-sealed reverse vacuum desorption process: After pressure equalization, the adsorption tower is subjected to reverse vacuum desorption from top to bottom; during the vacuum desorption process, product oxygen is continuously injected into the interlayer gap between the high-efficiency activated carbon protective layer and the molecular sieve separation layer through the gas-sealed gas injection valve to form a pneumatic membrane flowing from top to bottom in the interlayer gap. S4. Pressurization process: The gas is pressurized by the equalizing gas from the other adsorption towers and the product oxygen to complete the cycle.

2. The process for separating oxygen using a three-tower equal-pressure circulating pressure swing adsorption method according to claim 1, characterized in that, In the adsorption and in-situ self-displacement process of S1, the transient flow rate of the oxygen-enriched gas introduced by the self-circulation bypass is 3% to 8% of the current transient gas flow rate of the adsorption tower, and the execution time of the in-situ self-displacement is 5% to 10% of the total time of the entire adsorption stage.

3. The process for separating oxygen using a three-tower equal-pressure circulating pressure swing adsorption method according to claim 1, characterized in that, The specific geometric height of the injection tower into the middle section of the side wall of S1 is set at 35% to 45% of the total height of the adsorption bed in the adsorption tower.

4. The three-tower pressure equalization circulating pressure swing adsorption process for separating oxygen according to claim 1, characterized in that, In the radial distribution pressure equalization and pressure reduction process of S2, the outer wall of the microporous distribution tube is provided with an array of micropores, the diameter of a single micropore is 0.5mm to 2.0mm, and the axial length of the microporous distribution tube inside the molecular sieve separation layer accounts for 60% to 80% of the total height of the molecular sieve separation layer.

5. The process for separating oxygen using a three-tower equal-pressure circulating pressure swing adsorption method according to claim 1, characterized in that, In the radial distribution pressure equalization and pressure reduction process of S2, when the gas discharged from the pressure reduction tower flows through the microporous distribution pipe into the pressure boosting tower, the axial linear velocity of the fluid is converted into radial divergence velocity after entering the microporous distribution pipe, so that the linear velocity of the airflow on the cross section of the molecular sieve separation layer is reduced to 10% to 15% of the axial flow pressure equalization linear velocity.

6. The process for separating oxygen using a three-tower equal-pressure circulating pressure swing adsorption method according to claim 1, characterized in that, In the reverse vacuum desorption process of the gas seal isolation in S3, the flow rate of product oxygen injected into the interlayer gap through the gas seal injection valve is controlled to be 1.5% to 3.5% of the transient vacuum exhaust flow rate of the current process.

7. The process for separating oxygen using a three-tower equal-pressure circulating pressure swing adsorption method according to claim 1, characterized in that, The high-efficiency activated carbon protective layer uses nitrogen-doped hierarchical porous coconut shell-based modified activated carbon, wherein the volume of micropores with a pore size of less than 2 nm accounts for 75% to 80%, the volume of mesopores with a pore size of 2 nm to 50 nm accounts for 18% to 22%, and the filling height of the high-efficiency activated carbon protective layer accounts for 15% to 25% of the total height of the adsorption bed in the entire tower.

8. The process for separating oxygen using a three-tower equal-pressure circulating pressure swing adsorption method according to claim 1, characterized in that, The molecular sieve separation layer is composed of a modified 13X molecular sieve layer and a high-efficiency lithium molecular sieve layer stacked from bottom to top, with a filling thickness ratio of 1:1 to 2:

1.

9. The process for separating oxygen using a three-tower equal-pressure circulating pressure swing adsorption method according to claim 1, characterized in that, The three adsorption towers connected in parallel are controlled to operate alternately and staggered according to their respective working sequence with a phase difference of 120 degrees, and the oxygen production pressure of the system fluctuates between 0.4MPa and 0.55MPa.

Images