Acid gas emission management system and method

Abstract

The invention discloses an acid gas emission management system and method, and relates to the technical field of acid gas emission management.The system comprises a control module, a heavy metal removal module and an adjusting module, and the heavy metal removal module comprises a three-phase separation tank; the control module is used for controlling the temperature and the pressure of the three-phase separation tank within a first temperature range and a preset pressure respectively, so that hydrogen sulfide in the acid gas reacts with ammonia gas in the three-phase separation tank to generate non-gaseous ammonium hydrosulfide, and heavy metal in the acid gas reacts with sulfide in the system to generate a sulfide solid phase; the three-phase separation tank performs three-phase separation, the separated solid phase and liquid phase are discharged, the separated gas phase is introduced into the acid gas flare head for combustion, the adjusting module adjusts the combustion temperature of the acid gas flare head to be within a second temperature range, and the second temperature range is a temperature range for inhibiting generation of gaseous sulfur. The production of sulfur particles, ammonium sulfate particles and sulfur trioxide can be reduced or avoided.

Description

Technical Field

[0001] This application relates to the field of acid gas emission management technology, and in particular to an acid gas emission management system and method. Background Technology

[0002] Oil refineries and petrochemical plants generate large amounts of gas during operation, some of which must be properly vented to prevent accidents in the event of leaks. For example, the main components of acidic gases are toxic hydrogen sulfide and ammonia (hydrogen sulfide accounts for 30%-90% of acidic gases, and ammonia accounts for 2%-3%). A hydrogen sulfide concentration of 10 ppm can irritate the eyes, causing symptoms of poisoning such as coughing and tearing. An exposure to a hydrogen sulfide concentration of 500 ppm for 15-30 minutes can be fatal. Ammonia is highly soluble in water and, upon inhalation, quickly combines with moisture in the respiratory tract, eyes, and skin, causing alkaline corrosion. Therefore, acidic gases must be properly vented.

[0003] The disposal and emission of acidic gases are generally achieved through acidic gas treatment systems. Acidic gas flares serve as an emergency or auxiliary means. In the event of a malfunction in the acidic gas treatment system or a sudden increase in acidic gas production, the acidic gas can be incinerated and emitted into the atmosphere.

[0004] However, the current method of burning acidic gases through acidic gas flares is outdated and simple. During the combustion process, sulfur particles (yellow) and ammonium sulfate particles (NH4SO4, white) that float in the air are produced. In addition, sulfur trioxide (SO3) is produced during the combustion process. Sulfur trioxide can react with water vapor in the air to produce dilute sulfuric acid (H2SO4) and sulfuric acid mist, which is one of the sources of acid rain, thus causing environmental pollution. Summary of the Invention

[0005] The purpose of this application is to provide an acid gas emission management system and method that can reduce or avoid the generation of sulfur particles, ammonium sulfate particles and sulfur trioxide.

[0006] To achieve the above objectives, this application provides the following solution.

[0007] In a first aspect, this application provides an acid gas emission management system, which includes: a control module and a heavy metal removal module and a regulation module that are communicatively connected to the control module; The heavy metal removal module includes a three-phase separation tank. The control module controls the temperature and pressure of the three-phase separation tank within a first temperature range and at a preset pressure, respectively, so that hydrogen sulfide and ammonia in the acidic gas introduced into the three-phase separation tank react to generate non-gaseous ammonium hydrosulfide. Furthermore, heavy metals in the acidic gas react with sulfides in the system to generate a sulfide solid phase. The three-phase separation tank performs three-phase separation, discharging the separated solid and liquid phases, and introducing the separated gaseous phase into an acidic gas flare head for combustion. The gaseous phase includes hydrogen sulfide and ammonia that did not participate in the reaction in the three-phase separation tank. The adjustment module is used to adjust the combustion temperature of the acid gas torch head to a second temperature range; the second temperature range is the temperature range that suppresses the generation of gaseous sulfur.

[0008] Secondly, this application provides a method for managing acid gas emissions, applied to the aforementioned acid gas emission management system, the method comprising: The temperature and pressure of the three-phase separator are controlled within a first temperature range and at a preset pressure, respectively, so that the hydrogen sulfide in the acid gas introduced into the three-phase separator reacts with ammonia in the three-phase separator to generate non-gaseous ammonium hydrosulfide. In addition, the heavy metals in the acid gas react with the sulfides in the system to generate sulfide solid phase. An adjustment command is issued; the adjustment command is used to cause the adjustment module to adjust the combustion temperature of the acid gas torch head to a second temperature range; the second temperature range is the temperature range that suppresses the generation of gaseous sulfur.

[0009] According to the specific embodiments provided in this application, this application has the following technical effects.

[0010] This application provides an acid gas emission management system and method, including a control module, a heavy metal removal module, and an adjustment module. The heavy metal removal module includes a three-phase separation tank. The control module controls the temperature and pressure of the three-phase separation tank within a first temperature range and at a preset pressure, respectively, so that hydrogen sulfide and ammonia in the acid gas introduced into the three-phase separation tank react to generate non-gaseous ammonium hydrosulfide. Furthermore, the heavy metals in the acid gas react with sulfides in the system to generate sulfide solid phases. The three-phase separation tank is used for three-phase separation, discharging the separated solid and liquid phases, and introducing the separated gas phase into the acid gas flare head for combustion. The gas phase includes hydrogen sulfide and ammonia that did not participate in the reaction in the three-phase separation tank. At this time, the heavy metals in the acid gas are discharged through the three-phase separation tank in the form of sulfide solid phases to minimize their introduction into the acid gas flare head. In the traditional combustion process of an acid gas flare, the residual heat of the exhaust gas can be maintained in the temperature range of 450°C-650°C. This temperature range is precisely within the suitable temperature range for the oxidation of sulfur dioxide (SO2) to sulfur trioxide. Therefore, the sulfur dioxide produced during combustion will be converted into sulfur trioxide within this temperature range and under the catalysis of heavy metals. Sulfur trioxide further reacts with water vapor to generate dilute sulfuric acid and sulfuric acid mist (one of the sources of acid rain). However, in this application, because a three-phase separator is used beforehand to remove heavy metals as much as possible, the gas phase introduced into the acid gas flare head contains as few heavy metals as possible. Therefore, the further conversion of sulfur dioxide into sulfur trioxide within the temperature range of the residual heat of the exhaust gas can be avoided as much as possible. The regulating module is used to adjust the combustion temperature of the acid gas flare head to be within a second temperature range, which is the temperature range that inhibits the formation of gaseous sulfur. At this point, the combustion temperature of the acid gas flare head is within the temperature range that inhibits the formation of gaseous sulfur. In the traditional combustion process of acid gas flares, the generation of sulfur particles and ammonium sulfate particles is due to the excessively high combustion temperature of the acid gas flare head. Hydrogen sulfide reacts with sulfur dioxide to produce gaseous sulfur, which then undergoes cooling and phase change to form solid sulfur (i.e., sulfur particles). In the presence of ammonia, this gaseous sulfur undergoes oxidation, hydration, and neutralization reactions to generate ammonium sulfate particles. However, in this application, by controlling the combustion temperature of the acid gas flare head within a range that inhibits the generation of gaseous sulfur, the reaction of hydrogen sulfide with sulfur dioxide to generate gaseous sulfur is avoided, further preventing the generation of sulfur particles and ammonium sulfate particles, thus eliminating the generation of these particles. In summary, this application can reduce or avoid the generation of sulfur particles, ammonium sulfate particles, and sulfur trioxide, without causing environmental pollution. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of an acid gas emission management system provided in Embodiment 1 of this application.

[0012] Figure 2 This is a schematic diagram of the acid gas torch head provided in Embodiment 1 of this application.

[0013] Figure 3 This is a schematic diagram showing the communication connection between the acid gas emission management system and the traditional acid gas flare head management system provided in Embodiment 1 of this application.

[0014] Figure 4 This is a flowchart illustrating an acid gas emission management method provided in Embodiment 2 of this application.

[0015] Figure 5 This is a schematic diagram of the structure of a computer device provided in Embodiment 3 of this application. Detailed Implementation

[0016] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0017] Example 1.

[0018] During the combustion process of a traditional acid gas flare, sulfur particles and ammonium sulfate particles are produced that float in the air, as well as sulfur trioxide. Sulfur trioxide can react with water vapor in the air to produce dilute sulfuric acid and sulfuric acid mist, which is one of the sources of acid rain, thus causing environmental pollution.

[0019] The following section analyzes the reasons for the generation of sulfur particles, ammonium sulfate particles, and sulfur trioxide during the combustion process of traditional acid gas torches.

[0020] (1) The causes of the formation of sulfur particles and ammonium sulfate particles.

[0021] Currently, the common practice is to use acid gas to co-fire fuel gas. Specifically, fuel gas is introduced into an acid gas flare head and mixed with acid gas, so that acid gas and fuel gas burn together in the acid gas flare head. Mixing high-calorific-value fuel gas with low-calorific-value acid gas can increase the calorific value of the mixed gas, further increasing the combustion temperature of the acid gas flare head and ensuring that the acid gas can be stably and completely burned and decomposed. However, the inventors discovered that if the combustion temperature of the acid gas flare head is too high, gaseous sulfur (mainly S2) will be produced. The gaseous sulfur will further transform into sulfur particles and ammonium sulfate particles: gaseous sulfur is cooled and undergoes phase change to generate solid sulfur (i.e., sulfur particles). In the presence of ammonia, gaseous sulfur undergoes oxidation, hydration, and neutralization reactions to generate ammonium sulfate particles. Sulfur particles and ammonium sulfate particles will float in the air and even fall onto the windowsills and roofs of nearby residents.

[0022] (2) The causes of sulfur trioxide production.

[0023] Acidic gases contain heavy metals (such as nickel and vanadium). During the combustion process of a traditional acidic gas flare, the residual heat of the exhaust gas can be maintained in the temperature range of 450°C-650°C. This temperature range is exactly within the suitable temperature range for sulfur dioxide to be oxidized into sulfur trioxide. Therefore, the sulfur dioxide produced during the combustion of the acidic gas flare head will be converted into sulfur trioxide under the temperature range of 450°C-650°C and the catalytic effect of heavy metals. Sulfur trioxide further reacts with water vapor to generate dilute sulfuric acid and sulfuric acid mist, forming acid rain and causing environmental pollution.

[0024] Based on the clear understanding of the causes of sulfur particles, ammonium sulfate particles, and sulfur trioxide, this embodiment further studies the combustion emission mechanism of acid gas flares and explores whether the generation of sulfur particles, ammonium sulfate particles, and sulfur trioxide can be reduced or avoided.

[0025] (1) Ideas for reducing or avoiding sulfur particles and ammonium sulfate particles.

[0026] Acidic gases contain 35%-90% hydrogen sulfide (H2S), 0.05%-3% ammonia (NH3), 7%-45% carbon dioxide (CO2), and trace amounts of water vapor, hydrocarbons, and heavy metals. See Table 1 below for the proportions and calorific values ​​of the main components of acidic gases.

[0027] Table 1. Proportions and calorific values ​​of the main components of acidic gases

[0028] Acid gas can originate from the stripping tower of acidic water treatment and is typically discharged through an acid gas treatment system (such as a sulfur recovery unit). The chemical mechanism of the sulfur recovery unit utilizes a Claus reaction to break down hydrogen sulfide and ammonia at 1000°C-1400°C, producing gaseous sulfur. This gaseous sulfur is then gradually cooled, solidifying at 60°C, and is subsequently recovered. If the acid gas treatment system malfunctions or there is a sudden increase in acid gas production, the acid gas will be discharged through an acid gas flare. The combustion process in the acid gas flare is essentially a Claus reaction, and its chemical equation is as follows: ① Hydrogen sulfide (H2S) reacts with oxygen (O2) at 900°C-1400°C to produce sulfur dioxide (SO2) and water (H2O): ② The remaining hydrogen sulfide (H2S) reacts further with sulfur dioxide (SO2) to produce gaseous sulfur (S2) and water (H2O): The hydrogen sulfide conversion rate is approximately 60%-70%, and the product is gaseous sulfur. The reaction time is short (approximately 1 second). The heat from the gaseous sulfur is recovered, specifically through heat exchange via a heat exchanger, and gradual cooling is achieved by controlling the outlet temperature to form stable solid sulfur at 60°C, which is then recovered.

[0029] The following conclusions can be drawn from the above chemical reaction equations: The direct reaction of hydrogen sulfide with oxygen does not produce gaseous sulfur. A high-temperature reaction at 1150°C-1300°C will produce gaseous sulfur (mainly S2). A medium-temperature reaction at 140°C-160°C will cause gaseous sulfur to polymerize into liquid sulfur S8. A low-temperature solidification reaction at ≤60°C will cause gaseous sulfur to crystallize into solid sulfur. That is, the gaseous sulfur mentioned above will generate solid sulfur (i.e., sulfur particles) through cooling and phase change. It can only be stabilized into solid sulfur when the temperature drops below 60°C. At the same time, gaseous sulfur will generate ammonium sulfate particles through oxidation, hydration and neutralization reactions in the presence of ammonia. Obviously, both sulfur particles and ammonium sulfate particles will only be generated when gaseous sulfur is produced.

[0030] Concept: If the combustion temperature of the acid gas torch head is controlled below 1150°C, no gaseous sulfur will be generated, and therefore no sulfur particles and ammonium sulfate particles will be generated, thus preventing sulfur particles and ammonium sulfate particles from falling.

[0031] (2) Ideas for reducing or avoiding sulfur trioxide.

[0032] On the one hand, acidic gases contain heavy metals such as nickel and vanadium. Under the catalytic action of these heavy metals, sulfur dioxide is converted into sulfur trioxide within a suitable temperature range of 450°C-650°C. Vanadium, in particular, can catalyze the conversion of sulfur dioxide to sulfur trioxide at a rate of up to 80%. The catalytic effect of heavy metals is the main reason for the formation of sulfur trioxide. On the other hand, within the 450°C-650°C temperature range, impurities carried by the acidic gases (originating from upstream processing) introduce water vapor with a volume fraction of 2% to 20%. The sulfur trioxide generated under the catalytic action of heavy metals reacts with this water vapor to produce dilute sulfuric acid and sulfuric acid mist. Dilute sulfuric acid and sulfuric acid mist are among the main components of acid rain.

[0033] Idea: If heavy metals in acidic gases can be removed and sulfur dioxide can be prevented from converting into sulfur trioxide, then acid rain caused by sulfur trioxide can be further prevented.

[0034] Based on the above concept, this embodiment provides an acid gas emission management system, which includes a control module and a heavy metal removal module and a regulation module that are communicatively connected to the control module.

[0035] The heavy metal removal module includes a three-phase separation tank.

[0036] The aforementioned control module is used to control the temperature and pressure of the three-phase separator within a first temperature range and at a preset pressure, respectively, so that the hydrogen sulfide in the acid gas introduced into the three-phase separator reacts with ammonia in the three-phase separator to generate non-gaseous (solid and / or liquid) ammonium hydrosulfide. Furthermore, the heavy metals in the acid gas react with sulfides in the system to generate sulfide solid phases.

[0037] The three-phase separator is used to perform three-phase separation, discharging the separated solid and liquid phases and passing the separated gas phase into the acid gas flare head for combustion. The gas phase includes hydrogen sulfide and ammonia that did not participate in the reaction in the three-phase separator.

[0038] The adjustment module is used to adjust the combustion temperature of the acid gas torch head to a second temperature range, which is the temperature range that suppresses the generation of gaseous sulfur.

[0039] This embodiment uses a three-phase separator to remove heavy metals as much as possible beforehand, ensuring that the gas phase introduced into the acid gas flare head is as free of heavy metals as possible. Therefore, it minimizes the possibility of sulfur dioxide further converting to sulfur trioxide within the temperature range of the exhaust gas's residual heat. By controlling the combustion temperature of the acid gas flare head within a range that inhibits the formation of gaseous sulfur, the reaction of hydrogen sulfide and sulfur dioxide to form gaseous sulfur is prevented, further avoiding the formation of sulfur particles and ammonium sulfate particles, thus eliminating the production of sulfur particles and ammonium sulfate particles. In summary, this embodiment can reduce or avoid the generation of sulfur particles, ammonium sulfate particles, and sulfur trioxide, preventing environmental pollution.

[0040] like Figure 1 As shown, this is a schematic diagram of the acid gas emission management system used in this embodiment. Figure 1 The meanings of the labels in the table are shown in Table 2 below.

[0041] Table 2 Meaning of Labels

[0042] The following, combined with Figure 1 The acid gas emission management system used in this embodiment will be described in detail.

[0043] (a) Heavy metal removal module.

[0044] The acidic gas contains gaseous hydrogen sulfide and ammonia, liquid water, and solid heavy metals. In this embodiment, a three-phase separation tank Q-101 is set up to quickly separate the three phases. Hydrogen sulfide reacts with ammonia to generate ammonium hydrosulfide, while heavy metals react with sulfides in the system to generate solid sulfide phases. The entire chemical reaction is completed within 1 second.

[0045] The above system refers to the overall reaction environment and the sum of all substances within the three-phase separator Q-101, including all substances participating in the reaction and their physicochemical environment.

[0046] The core components of the system are as follows: Material composition: including the introduced acidic gas (hydrogen sulfide, ammonia, water, heavy metals, etc.) and the reaction products (ammonium hydrosulfide, sulfide solid phase, etc.). Environmental conditions: first temperature range and preset pressure; Phase distribution: It covers the gas phase, liquid phase (water), and solid phase (the generated precipitate) in the three-phase separator Q-101. The three phases together constitute the phase structure of the system.

[0047] In one example, with the temperature in the three-phase separator Q-101 controlled within the range of 118°C-145°C (first temperature range) and the pressure in the three-phase separator Q-101 controlled at 0.1 MPa (preset pressure), hydrogen sulfide (H2S) and ammonia (NH3) in the acid gas react to produce ammonium hydrosulfide (NH4HS): If carbon dioxide or sulfur trioxide is present, ammonium bicarbonate (NH4HCO3) or ammonium hydrogen sulfate (NH4HSO4) will also be generated.

[0048] Within a temperature range of 118°C-145°C and a pressure of 0.1 MPa, heavy metals in the acid gas react with sulfides in the system to form a solid sulfide phase (also known as a solid metal sulfide), which is then discharged by the three-phase separator Q-101. More than 98% of the heavy metals in the acid gas are discharged along with the solid sulfide phase, minimizing the amount of heavy metals entering the acid gas flare head and avoiding the generation of sulfur trioxide as much as possible.

[0049] It should be noted that the sulfides in the system mainly refer to hydrosulfide ions (HS-H2S) generated by the dissociation of hydrogen sulfide (H2S) in the acidic gas. - ) and sulfide ions (S 2- Furthermore, ammonium hydrosulfide (NH4HS) is an ionic compound. When dissolved in a liquid phase (such as water in the system), it first dissociates to produce ammonium ions (NH4). + ) and hydrosulfide ions (HS) - The hydrosulfide ions (HS-H2O) produced by dissociation - It will further undergo secondary dissociation, producing a small amount of sulfide ions (S). 2- The degree of secondary dissociation is affected by system temperature, pressure, and pH, and is usually weak. The hydrosulfide ions (HS-H2O) generated by the above dissociation... - ) and sulfide ions (S 2-It will react with heavy metal ions in acidic gas to form a sulfide solid phase.

[0050] Since sulfide solids are prone to precipitation and accumulation, which can cause blockages in the three-phase separator Q-101 and pipelines, this embodiment strengthens the management of the flowability of the sulfide solids to maintain a certain flowability within the three-phase separator Q-101 and pipelines.

[0051] To enhance the fluidity of the solid phase, the following design is implemented in this embodiment.

[0052] A fully jacketed steam tracing device is installed on the three-phase separator Q-101 to control the outer wall temperature of Q-101 at 150°C, thereby preventing the accumulation of sulfide solid phase within Q-101 without dead zones. Full-jacketed steam tracing (also known as "full-jacketed steam heating") is a method of heat tracing that involves encasing the entire outer wall of the equipment in steam: a sealed jacket is fitted over the equipment / pipeline, and steam is continuously introduced, allowing the latent heat of the steam to be evenly transferred to the internal medium through the metal wall, thus maintaining the material temperature and preventing solidification, crystallization, or increased viscosity.

[0053] A first heat tracing device is installed on the inlet pipe of the three-phase separator Q-101 to control the outer wall temperature of the inlet pipe at 150°C and prevent blockage. The inlet pipe is used to introduce acidic gas into the three-phase separator Q-101.

[0054] A second heat tracing device is installed on the outlet pipe of the three-phase separator Q-101 to control the outer wall temperature of the outlet pipe at 150°C and prevent blockage. The outlet pipe is used to pass the separated gas phase into the acid gas flare head.

[0055] To improve temperature control accuracy, this embodiment employs a control module for PID control, enhancing the temperature control precision of the full-jacketed steam tracing device, the first tracing device, and the second tracing device. This ensures that the outer wall temperatures of the three-phase separator, the inlet pipe, and the outlet pipe are maintained at 150°C, preventing blockage. Furthermore, by installing the full-jacketed steam tracing device, the first tracing device, and the second tracing device, not only can the sulfide solid phase be made fluid to prevent blockage, but the temperature inside the three-phase separator Q-101 can also be kept within the first temperature range.

[0056] That is, in this embodiment, the heavy metal removal module further includes: a fully jacketed steam tracing device installed on the three-phase separator Q-101, a first tracing device installed on the air inlet pipe of the three-phase separator Q-101, a second tracing device installed on the air outlet pipe of the three-phase separator Q-101, and a pressure regulating device installed on the three-phase separator Q-101.

[0057] The control module is communicatively connected to the full-jacketed steam tracing device, the first tracing device, the second tracing device, and the pressure regulating device. The control module is also used to perform PID control on the full-jacketed steam tracing device, the first tracing device, the second tracing device, and the pressure regulating device to control the temperature and pressure of the three-phase separator within a first temperature range and at a preset pressure, respectively.

[0058] The first heat tracing device is an electric heat tracing device, and the second heat tracing device is a steam heat tracing device. Electric heat tracing is used for the inlet pipe because the inlet pipe is long, and the electric heat tracing device can operate stably below 250°C. Steam heat tracing is used for the outlet pipe because the outlet pipe leads to the acid gas flare head, which is at a high altitude, making electric heat tracing unsafe.

[0059] To achieve temperature and pressure control, the following settings can be made.

[0060] A fully jacketed steam tracing temperature transmitter is installed on the outer wall of the three-phase separator Q-101 to detect the temperature of the outer wall of the three-phase separator Q-101. The temperature of the outer wall of the three-phase separator Q-101 is used as the actual temperature. The control module performs PID control on the fully jacketed steam tracing device and adjusts the opening of the fully jacketed steam control valve TV-101 so that the fully jacketed steam tracing device operates at 150°C.

[0061] An electric heat tracing temperature transmitter is installed on the outer wall of the air intake pipe to detect the temperature of the outer wall of the air intake pipe. The temperature of the outer wall of the air intake pipe is used as the actual temperature. The control module performs PID control on the electric heat tracing device so that the electric heat tracing device operates at 150°C.

[0062] A steam tracing temperature transmitter is installed on the outer wall of the outlet pipe to detect the temperature of the outer wall of the outlet pipe. The temperature of the outer wall of the outlet pipe is used as the actual temperature. The control module performs PID control on the steam tracing device and adjusts the opening of the steam control valve TV-102 so that the steam tracing device operates at 150°C.

[0063] A pressure sensor is installed inside the three-phase separator Q-101 to detect the pressure inside the three-phase separator Q-101. The pressure inside the three-phase separator Q-101 is used as the actual pressure. The control module performs PID control on the pressure regulating device so that the pressure regulating device operates at a preset pressure (e.g., 0.1MPa). Specifically, the pressure regulation can be achieved by adjusting the flow rate of acidic gas.

[0064] For example, the above-mentioned electric heat tracing device may include: an electric heat tracing tape (heating body), a power junction box (connecting the electric heat tracing tape to the power supply cable), a thermostat, etc., wherein the thermostat can monitor the temperature and control the power supply to achieve precise temperature control. The thermostat can be equipped with a built-in PID control algorithm to achieve temperature control, or other devices can send control signals (such as 4-20mA or switching signals) to the thermostat to achieve temperature control.

[0065] Of course, in addition to electric heat tracing devices, the first heat tracing device can also be a steam heat tracing device, a hot oil heat tracing device, etc.

[0066] A steam tracing system is a device that uses saturated steam as a heat medium to indirectly heat process pipelines, tanks, or materials inside them. For example, a steam tracing system mainly includes: a steam supply source (boiler room or steam distribution pipe), a main steam pipe (which delivers steam to the area requiring heat tracing), a heat tracing pipe, a pressure regulating valve (which can be installed on the steam supply source or main steam pipe), and a flow regulating valve (which can be installed on the main steam pipe or heat tracing pipe). The heat tracing pipe is tightly attached to (or wrapped around) the outlet pipe, directly exchanging heat with it. In addition, a steam tracing system may also include: steam traps, a condensate recovery pipe, and controls and accessories (stop valves, filters, pressure gauges). The steam traps only allow condensate and non-condensable gases to pass through, preventing steam from escaping. The condensate recovery pipe collects the condensate discharged from the steam traps and returns it to the recovery system. The pressure regulating valve and flow regulating valve are used to adjust the pressure and flow rate, thereby controlling the heat tracing temperature. Steam tracing systems can integrate temperature controllers, which can incorporate PID control algorithms to control pressure and flow regulating valves based on temperature sensor feedback, thereby achieving temperature control. Alternatively, steam tracing systems can also receive control signals from external devices to control the pressure and flow regulating valves.

[0067] In other examples, the second heat tracing device may also be a hot oil heat tracing device.

[0068] The exemplary structure of the three-phase separator Q-101 is described below.

[0069] In this embodiment, the residence time of acidic gas in the three-phase separator Q-101 can be set to 20 seconds. The residence time is achieved by designing the external dimensions of the three-phase separator Q-101, so that the reaction is more complete.

[0070] Please see again Figure 1In this embodiment, the three-phase separator Q-101 has two separation chambers, referred to as the first separation chamber Q-1011 and the second separation chamber Q-1012. The residence time in the three-phase separator Q-101 is the sum of the residence time in the first separation chamber Q-1011 and the residence time in the second separation chamber Q-1012. In actual operation, it was found that having only one separation chamber is unsafe, and since the residence time is short, it is unnecessary to design three separation chambers, which would lead to excessive costs. Therefore, it is more appropriate to set two separation chambers.

[0071] Specifically, the three-phase separator Q-101 is internally equipped with a partition that divides the chamber of the three-phase separator Q-101 into a first separation chamber Q-1011 and a second separation chamber Q-1012. A partition is designed between the first separation chamber Q-1011 and the second separation chamber Q-1012, and there is a gap between the partition and the top of the three-phase separator Q-101. This creates a gas flow space above the partition, connecting the first separation chamber Q-1011 and the second separation chamber Q-1012. The two separation chambers ensure sufficient residence time for the acidic gas, achieving vertical descent of the solid and liquid phases and horizontal flow of the gas phase, greatly increasing the three-phase separation effect. Simultaneously, the reaction is more complete, ensuring that over 98% of the heavy metals are discharged along with the sulfide solid phase, while also preventing blockages in the three-phase separator Q-101.

[0072] The top inlet of the first separation chamber Q-1011 is used to introduce acidic gas into the three-phase separator Q-101. The bottom outlet of the first separation chamber Q-1011 and the bottom outlet of the second separation chamber Q-1012 are used to discharge the separated solid and liquid phases. The top outlet of the second separation chamber Q-1012 is used to discharge the separated gas phase. The top inlet and bottom outlet of the first separation chamber Q-1011 and the top outlet and bottom outlet of the second separation chamber Q-1012 are all vertical cylindrical pipes, and their diameters are designed according to the flow rate to ensure that the inlet and outlet remain in a smooth state. The cylindrical shape of the inlet and outlet makes it easier for the solid and gas phases to flow and be discharged. Furthermore, the vertical installation of the inlet and outlet can utilize gravity acceleration to promote the descent speed of heavy metals and accelerate the discharge of heavy metals with the sulfide solid phase.

[0073] To further enhance fluidity, in this embodiment, the heavy metal removal module also includes: a first liquid level sensor installed at the bottom outlet of the first separation chamber Q-1011, a second liquid level sensor installed at the bottom outlet of the second separation chamber Q-1012, and a conveying component installed on the discharge pipe of the three-phase separation tank Q-101. The discharge pipe is connected to the bottom outlet of the first separation chamber Q-1011 and the bottom outlet of the second separation chamber Q-1012, respectively, and is used to send the separated solid and liquid phases into the waste liquid incinerator.

[0074] The first liquid level sensor is used to detect the first liquid level at the bottom outlet of the first separation chamber Q-1011.

[0075] The second liquid level sensor is used to detect the second liquid level at the bottom outlet of the second separation chamber Q-1012.

[0076] The conveying components are used to send the separated solid and liquid phases into the waste liquid incinerator through the discharge pipe.

[0077] The control module is communicatively connected to the first liquid level sensor, the second liquid level sensor, and the conveying component. The control module is also used to adjust the working parameters of the conveying component based on the first liquid level and the second liquid level, so that the sulfide solid phase is in a flowing state.

[0078] Both the first and second level sensors can be dual-flange level transmitters (also known as dual-flange level gauges).

[0079] The aforementioned conveying component can be a screw pump P101. Its operating parameters, exemplarily, include rotational speed. In this embodiment, a pre-set correspondence between liquid level and rotational speed is provided; the higher the liquid level, the greater the rotational speed. The specific correspondence can be set by the user based on experience. For example, the larger of the first and second liquid levels can be selected as the input. The rotational speed is determined using the correspondence between liquid level and rotational speed. The control module controls the screw pump P101 to operate at the determined rotational speed, thereby automatically adjusting the rotational speed of the screw pump P101 according to the liquid level (frequency conversion control can be used), keeping the sulfide solid phase in a flowing state, and discharging the solid and liquid phases from the three-phase separator Q-101 and sending them to the waste liquid incinerator.

[0080] In other embodiments of this application, the screw pump P101 can be connected to the bottom outlet of the two separation chambers via two independent branches, each branch being equipped with a control valve (i.e., Figure 1 (LV-101 and LV-102 in the model), the opening degree of this control valve can be adjusted.

[0081] After determining the rotational speed of screw pump P101 based on the higher liquid level, the control valve on the branch corresponding to the lower liquid level separation chamber (hereinafter referred to as the "first target control valve") can be coordinated and controlled to ensure that its opening degree is no greater than that of the control valve on the branch corresponding to the higher liquid level separation chamber (hereinafter referred to as the "second target control valve"). This measure aims to prevent the liquid level in the low liquid level separation chamber from dropping sharply due to excessively rapid discharge, thereby maintaining the stability of the internal pressure of the three-phase separator Q-101.

[0082] As an example of a fine-grained control strategy, the opening degree of the control valves on both branches can be controlled proportionally. The specific method is as follows: Assuming the liquid level in the chamber corresponding to the first target control valve is L1 (lower liquid level), and the liquid level in the chamber corresponding to the second target control valve is L2 (higher liquid level), the opening degree of the second target control valve is set to A (A is not greater than its maximum allowable opening degree). Based on the real-time ratio of the two liquid levels, the opening degree B of the first target control valve is dynamically calculated using the formula: B = A × (L1 / L2). B should not be less than its minimum allowable opening degree. Under the conditions that the cross-sectional areas of the two separation chambers are the same, the bottom outlet inner diameter is the same, and the branch pipe inner diameter is the same, this control method can make the instantaneous flow ratio of the two branches approximately equal to their liquid level ratio L1 / L2, thereby achieving a near-synchronous decrease in the liquid levels of the two separation chambers, further optimizing the dynamic balance.

[0083] To further enhance safety, a minimum liquid level threshold can be set (e.g., 5% of the dual-flange liquid level metering range). When the liquid level in any separation chamber falls below this minimum threshold, the proportional control mode described above will automatically exit, and the control valve of the corresponding separation chamber will be switched to a preset fixed small opening state. This safety mechanism effectively prevents the liquid in the chamber from being instantly drained, prevents acidic gas from entering the screw pump P101, and thus eliminates the risk of equipment damage caused by pump dry running and cavitation.

[0084] In a system using the aforementioned heavy metal removal module, a stable operating period was selected. Under the same process conditions (temperature: 118°C-145°C, pressure: 0.1MPa), acid gas samples were simultaneously collected from the inlet and outlet of the three-phase separator Q-101 and sent to the laboratory for analysis. The analysis results are shown in Table 3 below.

[0085] Table 3 Analysis Results

[0086] Based on the inlet and outlet mass flow data in the table, the following can be calculated: Nickel (Ni) removal rate = [(0.0010 - 0.00002) / 0.0010] × 100% = 98%; Vanadium (V) removal rate = [(0.0015-0.00003) / 0.0015]×100%=98%.

[0087] (ii) Adjustment module.

[0088] The adjustment module is used to adjust the combustion temperature of the acid gas torch head to a second temperature range, which is the temperature range that suppresses the generation of gaseous sulfur.

[0089] When the combustion temperature is below 650°C, hydrogen sulfide cannot be completely burned, resulting in a foul odor in the emissions. When the combustion temperature is above 1150°C, a large amount of gaseous sulfur is produced, which further produces sulfur particles and ammonium sulfate particles. Therefore, in this embodiment, the second temperature range is set to 650°C-1150°C.

[0090] Currently, there are multiple ways to adjust the combustion temperature of an acid gas torch head. For example, the combustion temperature can be adjusted by introducing a co-firing substance into the acid gas torch head, or by introducing air into the acid gas torch head. Taking the introduction of a co-firing substance to adjust the combustion temperature as an example, in this embodiment, the adjustment module includes: a detection component and a co-firing flow rate adjustment component that are communicatively connected to the control module.

[0091] The detection component is used to detect the obtained detection data, which includes the flow rate of the gas phase introduced into the acid gas torch head, the concentration of hydrogen sulfide, the concentration of ammonia, and the concentration and flow rate of the co-burning material.

[0092] The control module is also used to calculate, based on the detection data, the target flow rate of the co-firing material required to keep the combustion temperature of the acid gas torch head in the second temperature range.

[0093] The co-firing flow rate regulating component is used to regulate the flow rate of the co-firing material introduced into the acid gas flare head based on the target flow rate of the co-firing material.

[0094] The following section provides a more detailed introduction to the detection components and the co-firing flow rate adjustment components.

[0095] (1) Detection components.

[0096] For example, please see Figure 1 The detection components include a first concentration sensor AI-101, a second concentration sensor AI-102, and a first flow sensor FI-102, which are connected to the control module and installed on the gas outlet pipe of the three-phase separator Q-101. The gas outlet pipe is used to pass the separated gas phase into the acid gas flare head.

[0097] The first concentration sensor, AI-101, is used to detect the concentration of hydrogen sulfide introduced into the acid gas torch.

[0098] The second concentration sensor AI-102 is used to detect the concentration of ammonia gas introduced into the acid gas torch head.

[0099] The first flow sensor FI-102 is used to detect the flow rate of the gas phase introduced into the acid gas torch head.

[0100] The aforementioned co-firing material was mentioned. In this embodiment, the co-firing material can be designed to include propane gas and steam. The detection component also includes a third concentration sensor AI-103, a second flow sensor (i.e., a propane gas flow transmitter), and a third flow sensor (i.e., a steam flow transmitter) that are communicatively connected to the control module.

[0101] The third concentration sensor AI-103 and the second flow sensor are both installed on the propane gas inlet pipe, which is used to introduce propane gas into the acid gas flare head. The third concentration sensor AI-103 is used to detect the concentration of propane gas introduced into the acid gas flare head, and the second flow sensor is used to detect the flow rate of propane gas introduced into the acid gas flare head.

[0102] The third flow sensor is installed on the steam inlet pipe, which is used to introduce steam into the acid gas flare head. The third flow sensor is used to detect the flow rate of the steam introduced into the acid gas flare head.

[0103] Currently, when adjusting the combustion temperature, a gas chromatograph is used to measure the composition of the mixed gas in the acid gas flare head. Based on the concentration of each component in the mixed gas, the calorific value of the mixed gas is calculated, and the combustion temperature is further calculated. The combustion temperature is then adjusted by controlling the co-firing of fuel gas. However, the entire measurement process using a gas chromatograph takes more than 30 minutes, while the chemical reaction time of the acid gas flare head is only 0.3 seconds. This significant measurement lag means that adjusting the combustion temperature by controlling the co-firing of fuel gas also has a significant lag, causing the combustion temperature of the acid gas flare head to fluctuate within a large temperature range of 500°C-1400°C. This generates gaseous sulfur, causing sulfur particles and ammonium sulfate particles to fall off.

[0104] Therefore, this embodiment adopts pipeline in-situ laser component analyzer technology, which can avoid the serious lag in the above measurement. In this case, the first concentration sensor, the second concentration sensor and the third concentration sensor all adopt pipeline laser analyzer, and the first flow sensor, the second flow sensor and the third flow sensor all adopt flow transmitter.

[0105] The control module is also used to calculate the target flow rate of the co-firing material required to keep the combustion temperature of the acid gas torch head in the second temperature range based on the detection data, as follows.

[0106] For hydrogen sulfide, calculate the product of the molar concentration of hydrogen sulfide, the volumetric flow rate of the gas phase introduced into the acid gas flare head, and the molar mass of hydrogen sulfide to obtain the mass flow rate of hydrogen sulfide. Calculate the product of the mass flow rate of hydrogen sulfide and the lower heating value of hydrogen sulfide to obtain the combustion heat release power of hydrogen sulfide. Divide the combustion heat release power of hydrogen sulfide by the average isobaric specific heat capacity of the mixed gas (the mixed gas refers to the gas phase introduced into the acid gas flare head, propane gas, and steam) and the total mass flow rate of the mixed gas to obtain the theoretical adiabatic temperature rise caused by hydrogen sulfide. Calculate the product of the theoretical adiabatic temperature rise caused by hydrogen sulfide and the flare thermal efficiency coefficient to obtain the actual temperature rise caused by hydrogen sulfide.

[0107] For ammonia, calculate the product of the molar concentration of ammonia, the volumetric flow rate of the gas phase introduced into the acid gas torch head, and the molar mass of ammonia to obtain the mass flow rate of ammonia. Calculate the product of the mass flow rate of ammonia and its lower heating value to obtain the combustion heat release power of ammonia. Divide the combustion heat release power of ammonia by the average isobaric specific heat capacity of the mixed gas and the total mass flow rate of the mixed gas to obtain the theoretical adiabatic temperature rise caused by ammonia. Calculate the product of the theoretical adiabatic temperature rise caused by ammonia and the torch thermal efficiency coefficient to obtain the actual temperature rise caused by ammonia.

[0108] For propane gas, calculate the product of propane gas molar concentration, propane gas volumetric flow rate, and propane gas molar mass to obtain the propane gas mass flow rate. Calculate the product of propane gas mass flow rate and propane gas lower heating value to obtain the propane gas combustion heat release power. Divide the propane gas combustion heat release power by the average isobaric specific heat capacity of the mixed gas and the total mass flow rate of the mixed gas to obtain the theoretical adiabatic temperature rise caused by propane gas. Calculate the product of the theoretical adiabatic temperature rise caused by propane gas and the flare thermal efficiency coefficient to obtain the actual temperature rise caused by propane gas.

[0109] For steam, calculate the product of the volumetric flow rate and the steam density to obtain the mass flow rate of the steam. Calculate the difference between the specific enthalpy of the steam at the flame temperature and the specific enthalpy of the steam under the current operating conditions, and calculate the product of the difference and the mass flow rate of the steam to obtain the heat absorption power of the steam. Divide the heat absorption power of the steam by the average isobaric specific heat capacity of the mixed gas and the total mass flow rate of the mixed gas to obtain the theoretical adiabatic temperature drop caused by the steam. Calculate the product of the theoretical adiabatic temperature drop caused by the steam and the flare thermal efficiency coefficient to obtain the actual temperature drop caused by the steam.

[0110] It should be noted that the above molar concentration is calculated from the volume concentration, and the above volume concentration and volume flow rate are the concentration and flow rate obtained by the sensor.

[0111] The predicted temperature of the acid gas flare head is obtained by summing the base temperature, the actual temperature rise caused by hydrogen sulfide, the actual temperature rise caused by ammonia, the actual temperature rise caused by propane, and the actual temperature drop caused by steam (which is a negative value). The base temperature refers to the temperature of the mixture of hydrogen sulfide, ammonia, propane, and steam before it enters the acid gas flare head.

[0112] After calculating the predicted temperature of the acid gas flare head, if the predicted temperature is within the second temperature range, no adjustment is needed. If the predicted temperature is below the lower limit of the second temperature range, the temperature needs to be increased. In this case, the steam flow rate is reduced first. If the steam flow rate is reduced to the lower limit, the propane flow rate is increased. Following this principle, the control module calculates the difference between the predicted temperature and the median of the second temperature range to obtain the error. Based on the error, PID control is performed to calculate the target flow rates of propane and steam required to bring the predicted temperature within the second temperature range. If the predicted temperature of the acid gas flare head is above the upper limit of the second temperature range, the temperature needs to be reduced. In this case, the steam flow rate is increased first. If the steam flow rate is increased to the upper limit, the propane flow rate is reduced. Following this principle, the control module calculates the difference between the predicted temperature and the median of the second temperature range to obtain the error. Based on the error, PID control is performed to calculate the target flow rates of propane and steam required to bring the predicted temperature within the second temperature range.

[0113] (2) Firing flow rate adjustment component.

[0114] As mentioned above, the co-firing material includes propane gas and steam. Therefore, the target flow rate of the co-firing material calculated by the control module may further include the target flow rate of propane gas and the target flow rate of steam. The co-firing flow rate regulating component may further include a propane gas flow control valve FV-101 and a steam flow control valve FV-201 that are communicatively connected to the control module.

[0115] The propane flow control valve FV-101 is installed on the propane gas inlet pipeline. The propane flow control valve FV-101 is used to regulate the flow rate of propane gas introduced into the acid gas torch head based on the target flow rate of propane gas.

[0116] The steam flow control valve FV-201 is installed on the steam inlet pipe. The steam flow control valve FV-201 is used to regulate the flow rate of steam introduced into the acid gas torch head based on the target steam flow rate.

[0117] The control module in this embodiment can further calculate the error between the target flow rate of propane gas and the flow rate of propane gas (obtained by the second flow sensor), perform PID control based on the error, calculate the target opening degree of propane gas flow control valve FV-101, and then control propane gas flow control valve FV-101 based on the target opening degree to adjust its opening degree to the target opening degree. At this time, propane gas flow control valve FV-101 can adjust the flow rate of propane gas introduced into the acid gas torch head to the target flow rate of propane gas.

[0118] The control module in this embodiment can further calculate the error between the target steam flow rate and the steam flow rate (obtained by the third flow sensor), perform PID control based on the error, calculate the target opening degree of the steam flow control valve FV-201, and then control the steam flow control valve FV-201 based on the target opening degree to adjust its opening degree to the target opening degree. At this time, the steam flow control valve FV-201 can adjust the steam flow rate introduced into the acid gas torch head to the target steam flow rate.

[0119] In this embodiment, the flow rates of propane gas and steam are adjusted to ensure that the acid gas is safely and completely combusted after passing through the acid gas flare head, while achieving environmental protection requirements of no sulfur particles, no ammonium sulfate particles falling, and no foul odor.

[0120] (iii) Temperature monitoring module.

[0121] In this embodiment, thermocouples and thermal imagers are used to monitor the combustion temperature of the acid gas torch head. The temperature detected by the thermocouples is compared with the temperature detected by the thermal imager to correct for instability in the temperature field.

[0122] In this embodiment, the acid gas emission management system further includes a temperature monitoring module that is communicatively connected to the control module. The temperature monitoring module includes a first thermocouple TC-301, a second thermocouple TC-302, a third thermocouple TC-303, and a thermal imager TC-401 that are communicatively connected to the control module.

[0123] The first thermocouple TC-301 is installed in the core reaction zone of the acid gas torch head. The first thermocouple TC-301 is used to detect and obtain the first temperature.

[0124] The second thermocouple TC-302 is installed in the main high-temperature zone of the acid gas flare head. The second thermocouple TC-302 is used to detect the second temperature. The main high-temperature zone surrounds the core reaction zone.

[0125] The third thermocouple TC-303 is installed in the diffusion and cooling zone of the acid gas torch head. The third thermocouple TC-303 is used to detect the third temperature. The diffusion and cooling zone surrounds the main high-temperature zone.

[0126] The TC-401 thermal imager is used to detect the temperature field (raw data) of the combustion zone of the acid gas torch. The raw temperature field data matrix is ​​unreadable to the human eye, but a thermal image can be obtained by visualizing the temperature field data.

[0127] The combustion zone of the aforementioned acid gas flare head includes the core reaction zone, the main high-temperature zone, and the diffusion and cooling zone. The distance between the TC-401 thermal imager and the acid gas flare head can be 80-150 meters. The TC-401 thermal imager is equipped with a high-intensity air-cooled or water-cooled system and can clearly distinguish hot spots larger than 0.75 meters and their temperatures. Hot spots refer to areas with abnormally high temperatures in thermal imaging images.

[0128] The control module is also used to correct the temperature field based on the first temperature, the second temperature, and the third temperature to obtain the corrected temperature field, and to issue an alarm when the temperature in the corrected temperature field is greater than a preset temperature threshold.

[0129] For example, the temperature field can be corrected in the following way.

[0130] (1) Data preprocessing.

[0131] (1.1) Thermocouple data preprocessing.

[0132] 1) This can verify whether the temperature data measured by the first to third thermocouples (i.e., the first temperature, the second temperature, and the third temperature) are within a reasonable range. For example, during the combustion of an acid gas torch, the temperature in the core reaction zone is typically 1200°C-1500°C, the temperature in the main high-temperature zone is typically 800°C-1200°C, and the temperature in the diffusion and cooling zone is typically 400°C-800°C. If the temperature data measured by any thermocouple exceeds the temperature range of its corresponding area, it may indicate a sensor malfunction, requiring an alarm. For instance, if the first thermocouple is installed in the core reaction zone of the acid gas torch, it can be verified whether the temperature data measured by the first thermocouple is between 1200°C and 1500°C. If it exceeds this range, an alarm can be triggered, indicating "abnormal sensor data" or "possible sensor malfunction," and the abnormal time and value can be recorded.

[0133] 2) For the temperature data measured by each thermocouple at multiple consecutive times, filter the data (using methods such as moving average filtering or median filtering) to obtain a "smoother" temperature data sequence that is closer to the true temperature.

[0134] (1.2) Temperature field data and thermal imaging image preprocessing.

[0135] Preprocessing of temperature field data can include denoising, such as using Gaussian filtering to remove Gaussian noise and using morphological filtering to remove salt-and-pepper noise.

[0136] Preprocessing of the thermal imaging image corresponding to the denoised temperature field data may include enhancement processing to improve the contrast and sharpness of the thermal imaging image in order to more accurately determine the position of the thermocouple in the thermal imaging image.

[0137] (2) Establishment of spatial correspondence.

[0138] (2.1) Divide the thermal imaging image (after preprocessing) into grids, with each grid corresponding to a pixel, and record the coordinates of each pixel.

[0139] (2.2) Determine the pixel coordinates (i.e. the coordinates of the corresponding pixel points) in the thermal imaging image of the three thermocouples at the same time.

[0140] For example, the pixel coordinates of the thermocouple can be determined by setting markers at each thermocouple installation location and using image processing algorithms (such as template matching) to find these markers in the thermal imaging image.

[0141] Alternatively, the pixel whose temperature value is closest to the temperature data measured by the thermocouple (after preprocessing) can be searched within the thermal imaging image, and the pixel coordinates of that pixel can be used as the pixel coordinates of the corresponding thermocouple.

[0142] (3) Temperature field correction.

[0143] Since the measurement accuracy of thermal imagers may be affected by environmental factors (such as atmospheric attenuation, changes in target emissivity, etc.), the same actual temperature may present different measurement temperatures at different locations. Therefore, the measured temperature field of the thermal imager (after preprocessing) can be corrected by using the measured temperature after preprocessing by three thermocouples as a reference.

[0144] Temperature field data is essentially a matrix of temperature values ​​per pixel. Therefore, it corresponds one-to-one with the pixels of a thermal imaging image. Thus, the target pixel coordinates of the thermocouple in the thermal imaging image and temperature field data can be located through the thermal imaging image.

[0145] In one example, the correction may include the following steps.

[0146] (3.1) For each thermocouple, the target pixel coordinate (x) i y i (Also known as a reference point), calculate its temperature correction using the following formula: E i =T_tc_i-T_img_i; Among them, E iT_tc_i represents the deviation (correction amount) between the thermal imager's measured value and the true value at the known reference point; T_tc_i represents the measured temperature of the i-th thermocouple after preprocessing; T_img_i represents the target pixel coordinates (x, y) of the pixel point corresponding to the i-th thermocouple in the measured temperature field. i y i ( ) The measured temperature.

[0147] Assuming the correction amount varies continuously in space, subsequent adjustments can be made based on the known reference point E. i The correction amount E(x, y) for each pixel in the entire measured temperature field is estimated by spatial interpolation.

[0148] Commonly used interpolation methods include linear interpolation, inverse distance weighted interpolation, and Kriging interpolation.

[0149] (3.2) Perform interpolation based on the selected interpolation method.

[0150] The following describes two interpolation methods.

[0151] 1) Inverse Distance Weighted Interpolation (IDW): The correction amount for the unknown point is calculated by weighting the distance between the known reference point and the unknown point. The closer the distance, the greater the weight.

[0152] The calculation formula is: ; in, Unknown point The estimated correction amount; The number of known reference points (3 in this example); For the unknown point (x, y) to the th Distance between known reference points This is the distance weighting factor, which is a power parameter, usually taken as 2.

[0153] 2) Kriging interpolation: A statistical interpolation method that considers the spatial correlation of data. It uses a semi-variogram to describe the spatial variation structure of the data and then predicts the values ​​of unknown points based on information from known baseline points. This method has high accuracy when dealing with spatially correlated data, but its computational complexity is relatively high.

[0154] In other examples, machine learning models can also be used to correct the temperature field. A large number of pre-processed thermocouple temperatures and measured temperature fields can be collected as samples under different operating conditions (such as different acid gas emission rates, different combustion conditions, etc.) to form training and testing sets. The established correction model is trained using the training set data, and model parameters (such as the weights of the neural network, the coefficients of the regression model, etc.) are adjusted to minimize the error of the correction model on the training set. The trained correction model is validated using the testing set data, and the prediction error of the trained correction model is calculated. If the prediction error is large, the cause needs to be analyzed and the trained correction model optimized.

[0155] After optimizing the trained correction model, the measured temperature and the measured temperature field after thermocouple preprocessing are input into the trained correction model. The trained correction model can output the correction amount of each pixel in the measured temperature field.

[0156] (3.3) Correction.

[0157] After obtaining the correction values ​​for each pixel, the temperature values ​​of the corresponding pixels in the measured temperature field are corrected using the correction values ​​to obtain the corrected temperature field.

[0158] The corrected formula is: T corrected (x, y) = T img (x, y) + E(x, y); Among them, T corrected (x, y) represents the temperature of pixel (x, y) in the corrected temperature field, T img (x, y) represents the temperature of pixel (x, y) in the measured temperature field.

[0159] Wind in the natural environment can significantly affect the shape of a flame: in a windless environment, the flame is axially symmetrical in the shape of a column or cone; under the action of wind, the flame will tilt, stretch, and shake (the greater the wind speed, the more obvious the shape distortion), and may even shift or bifurcate, which will directly interfere with the recognition accuracy of the core reaction area, the main high temperature area and the diffusion and cooling area in the thermal imaging image.

[0160] In other embodiments of this application, the above-mentioned acid gas emission management system may further include an automatic moving system and a wind direction sensor. The wind direction sensor can measure the wind direction data of the current environment in real time (such as the wind direction angle of 0°-360° in the geographic coordinate system). The automatic moving system moves the thermal imager according to the current wind direction so that the thermal imager faces the current wind direction. At this time, from the perspective of the thermal imager, the flame is "facing the lens", and the shape is closest to the symmetrical state when there is no wind (without obvious tilt / stretching). The temperature field distribution is more regular and no additional zoning correction is required.

[0161] For example, the automatic moving system may include: an annular guide rail (or telescopic arm) and a controller installed around the acid gas flare head, wherein the center of the annular guide rail coincides with the center of the acid gas flare head on a horizontal plane, the movement trajectory range of the telescopic arm is based on the center of the acid gas flare head, and the controller is communicatively connected to a wind direction sensor and a thermal imager. The center of the annular guide rail or the center of the movement trajectory range of the telescopic arm coincides with the center of the acid gas flare head. The controller can perform a moving operation upon receiving real-time wind direction data from the wind direction sensor: receiving the real-time wind direction data from the wind direction sensor, determining the target observation position of the thermal imager, and driving the annular guide rail (or telescopic arm) to move the thermal imager to the target observation position. The target observation position is the "windward direction of the current wind" (i.e., the direction of the incoming wind facing the front of the thermal imager). At this time, from the perspective of the thermal imager, the flame shape is closest to the axisymmetric state when there is no wind (no obvious tilting or stretching), and the temperature field distribution is regular. Specifically, the deployment radius of the ring rail / telescopic arm can be 80-150 meters to match the effective observation distance of the thermal imager.

[0162] Once the controller moves to the target observation position, it can send a ready signal to the thermal imager, triggering the thermal imager to start acquiring the temperature field.

[0163] To prevent the thermal imager from moving frequently due to light winds, in other embodiments, the system can also be equipped with a wind speed sensor to collect real-time wind speed data of the current environment. The wind speed sensor can communicate with the controller, and when the wind speed data measured by the wind speed sensor is greater than a preset wind speed threshold (e.g., 1 m / s-3 m / s, which can be adjusted according to actual working conditions), the controller will then perform the aforementioned movement operation.

[0164] The following section provides a more detailed description of the control module's functions in controlling combustion temperature.

[0165] Please see Figure 2 The combustion process of the acid gas torch head is mainly divided into three zones. The three zones diffuse in a step-like manner from the core reaction zone to the main high-temperature zone, then from the main high-temperature zone to the diffusion and cooling zone, and then from the diffusion and cooling zone to the surrounding space. The range of the three zones is given in advance by the torch manufacturer.

[0166] The core reaction zone (internal zone) is located directly above the outlet of the acid gas torch, which is also the ignition area for the mixed fuel gas. Its characteristics are: the initial combustion reaction of combustibles such as hydrogen sulfide and hydrocarbons occurs, releasing a large amount of heat and forming the core framework of the flame. Oxygen is relatively insufficient, which is a fuel-rich combustion.

[0167] If the combustion temperature of the core reaction zone is not precisely controlled at this time, the combustion temperature of the core reaction zone will vary in the range of 1000°C-1400°C. At 1150°C-1300°C, the chemical reaction of the components of the acid gas will produce gaseous sulfur. The inventors discovered that if the combustion temperature of the core reaction zone is controlled between 800°C-1100°C, gaseous sulfur will not be produced.

[0168] The main high-temperature zone (the middle zone of the flame) is located as follows: it diffuses from the core reaction zone to form a high-temperature area with a lower combustion temperature than the core reaction zone. Its characteristics are: the combustion of components such as hydrogen sulfide and ammonia occurs at a high and stable temperature, ensuring a destruction rate of over 99%.

[0169] If the combustion temperature of the main high-temperature zone is not precisely controlled at this time, the combustion temperature of the main high-temperature zone will vary in the range of 900°C-1200°C, and gaseous sulfur will be produced when it is above 1150°C. The inventors discovered that if the combustion temperature of the main high-temperature zone is controlled between 700°C-1050°C, gaseous sulfur will not be produced.

[0170] The diffusion and cooling zone (outer zone) is located at the outermost layer of the flame. Its characteristics are: the combustion reaction is basically completed, and the products (SO2, CO2, H2O, N2) are mixed with the surrounding air and cooled, rapidly decreasing from 650°C to the ambient temperature.

[0171] In this embodiment, the second temperature range may further include a first combustion temperature sub-range (i.e., 800°C-1100°C) for the core reaction zone of the acid gas torch head, and a second combustion temperature sub-range (i.e., 700°C-1050°C) for the main high-temperature zone of the acid gas torch head, with the main high-temperature zone surrounding the core reaction zone.

[0172] Based on the partitioning of the acid gas torch head, and the combustion temperature sub-ranges of the core reaction zone and the main high-temperature zone, the control module can further control the combustion temperature of the flame in the following ways.

[0173] The symbols involved in this method are defined as follows, as shown in Table 4 below.

[0174] Table 4 Symbol Definitions

[0175] Note: Steam does not actively absorb heat in the main high-temperature zone. Instead, it dissipates heat slightly as the flue gas cools down (or only maintains heat transfer without additional heat absorption). Therefore, the heat absorption term of steam in the main high-temperature zone does not need to be included in the heat balance model below.

[0176] The methods for determining each item in Table 4 above are as follows.

[0177] (1) T基 .

[0178] A temperature sensor is installed in the mixing pipe (or mixing chamber) before the inlet of the acid gas flare head to directly collect the real-time temperature of the mixed gas as T. 基 .

[0179] (2) T H2S1 .

[0180] T H2S1 =(ρ H2S ×Q 酸气 ×M H2S ×LHV H2S )×η1 / (C 混1 ×m 总1 ); Where, ρ H2S Molar concentration of H2S in acidic gas, in mol / Nm³. 3 It is calculated based on the volume fraction (i.e., concentration) measured by the first concentration sensor AI-101; Q 酸气 Volumetric flow rate of acidic gas, in Nm³ 3 / s, based on the flow rate measured by the first flow sensor FI-102, that is, the gas phase flow rate measured by the first flow sensor FI-102; it should be noted that Q 酸气 The condition (standard condition or operating condition) must be consistent with ρ H2S The conversion states are strictly consistent: if ρ H2S If it is "standard molar concentration", then Q 酸气 It must be "standard volumetric flow rate"; if ρ H2S If it is "operating condition molar concentration", then Q 酸气 It must be "operating condition volumetric flow rate"; M H2S Molar mass of H2S, in kg / mol; Also: ρ H2S ×Q 酸气 ×M H2S The mass flow rate m of H2S in acidic gas H2S Unit: kg / s; LHV H2S H2S has a low calorific value, measured in kJ / kg, with a fixed value of approximately 9050 kJ / kg. η1: Combustion efficiency in the core reaction zone, dimensionless. Under low load conditions (acid gas flow rate < 30% of rated value), η1 = 0.7-0.75; under high load conditions (acid gas flow rate > 80% of rated value), η1 = 0.8-0.85. η1 can be dynamically fine-tuned (e.g., once per second), with a recommended adjustment step of 0.005-0.01. However, it is necessary to ensure that the value after fine-tuning is within the range determined by the operating conditions. The logic of dynamic fine-tuning is based on the target temperature range (800°C-1100°C) and uses the measured temperature T of the core reaction zone as a reference. 核实 The deviation from the target temperature range is used to correct the combustion efficiency value in reverse: if T 核实 >The upper limit of the target temperature range (1100°C) requires reducing η1 (always not lower than 0.7) to suppress overheating (avoid the formation of gaseous sulfur); if T 核实 If the target temperature range is below the lower limit (800°C), η1 needs to be increased (always not exceeding 0.85) to raise the temperature (ensuring combustion stability); if T 核实 If the temperature is within the target temperature range, the current η1 can be maintained; C 混1 The average isobaric specific heat capacity of the mixed gas in the core reaction zone, in kJ / (kg·℃), refers to the average specific heat capacity of all components (acid gas, propane gas, and steam) entering the core reaction zone within the temperature range (800°C-1100°C) of the core reaction zone. It can be weighted by the real-time mass ratio of each component, which is an existing mature method and will not be elaborated further. m 总1 Total mass flow rate of mixed gas in the core reaction zone, in kg / s, m³. 总1 =m 酸气 +m 丙烷 +m 蒸汽 m 酸气 m 丙烷 m 蒸汽 These represent the real-time mass flow rates of acid gas, propane gas, and steam, respectively, in kg / s.

[0181] The above m 酸气 There are two calculation methods. Method 1 is based on the "standard volumetric flow rate" (i.e., the standard flow rate of the gas phase measured by the first flow sensor FI-102) and the "standard total density" of the acid gas. Method 2 is obtained by "summing the mass flow rates of each component".

[0182] The formula for calculating the standard total density in Method 1 can be: ρ 酸气 (Standard condition total density) = (Standard Condition) + (Standard Condition) + 惰性 ×ρ惰性 (Standard condition); in: The volume fraction of S can be the direct measurement value of the first concentration sensor AI-101 (e.g., 35% = 0.35). ρ H2S (Standard conditions): Standard density of H2S, unit: kg / Nm³ 3 , ρ H2S (Standard conditions) = 1.539 kg / Nm 3 ; The volume fraction can be the direct measurement value of the second concentration sensor AI-102 (e.g., 10% = 0.10). ρ NH3 (Standard conditions): Standard density of NH3, unit: kg / Nm³ 3 , ρ NH3 (Standard conditions) = 0.760 kg / Nm 3 ; 惰性 Volume fraction of inert gases 惰性 =1- (The default is components that do not participate in combustion, such as N2 and CO2); ρ 惰性 (Standard conditions): Standard density of inert gas, unit: kg / Nm³ 3 Based on N2, the estimated value is 1.250 kg / Nm³. 3 (If the inert gas component is known to be mainly CO2, 1.977 kg / Nm³ can be used instead.) 3 ).

[0183] The calculation formula for method two is: m 酸气 =m H2S +m NH3 +m 惰性组分 ; m 惰性组分 =ρ 惰性 (Standard molar concentration) × Q 酸气 (Standard condition) × M 惰性平均 ; Where, m 惰性组分 : Mass flow rate of inert gas component, unit kg / s; ρ 惰性 (Standard condition molar concentration): Molar concentration of the inert gas component, unit: mol / Nm³ 3 ; Q 酸气 (Standard Conditions): Standard volumetric flow rate of acidic gas, unit: Nm³ 3 / s; M惰性平均 The average molar mass of the inert gas component, expressed in kg / mol. Given that the inert gas component is primarily N₂, we can directly let M = N₂. 惰性平均 =0.028 kg / mol; if CO2 is present, the weighted average can be calculated based on volume fraction (for example, assuming the volume fraction of N2 is...). =0.5, CO2 volume fraction =0.05, then M 惰性平均 =0.5×0.028+0.05×0.044=0.0292kg / mol, where 0.028 is the molar mass of N2 and 0.044 is the molar mass of CO2.

[0184] (3) T NH31 .

[0185] T NH31 =(ρ NH3 ×Q 酸气 ×M NH3 ×LHV NH3 )×η1 / (C 混1 ×m 总1 ); Where, ρ NH3 : Molar concentration of NH3 in acidic gas, unit: mol / Nm 3 It can be calculated based on the volume fraction (i.e., concentration) measured by the second concentration sensor AI-102; M NH3 Molar mass of NH3, in kg / mol; Also: ρ NH3 ×Q 酸气 ×M NH3 The mass flow rate m of NH3 in acidic gas NH3 Unit: kg / s; LHV NH3 NH3 has a low calorific value, measured in kJ / kg, with a fixed value of approximately 18600 kJ / kg.

[0186] It should be noted that, similar to the above, Q 酸气 The condition (standard condition or operating condition) must be consistent with ρ NH3 The conversion states are strictly consistent.

[0187] (4) T 丙烷1 .

[0188] T 丙烷1 =(ρ 丙烷 ×Q 丙烷 ×M 丙烷 ×LHV 丙烷)×η1 / (C 混1 ×m 总1 ); Where, ρ 丙烷 Molar concentration of propane gas, in mol / Nm³. 3 It can be calculated based on the volume fraction (i.e., concentration) measured by the third concentration sensor AI-103; Q 丙烷 Volumetric flow rate of propane gas, in Nm³. 3 / s, which can be measured based on the second flow sensor; M 丙烷 Molar mass of propane gas, in kg / mol; also, ρ 丙烷 ×Q 丙烷 ×M 丙烷 The mass flow rate m of propane gas 丙烷 Unit: kg / s; LHV 丙烷 The lower heating value of propane gas is measured in kJ / kg, with a fixed value of approximately 46,400 kJ / kg.

[0189] It should be noted that, similar to the above, Q 丙烷 The condition (standard condition or operating condition) must be consistent with ρ 丙烷 The conversion states are strictly consistent.

[0190] (5) T 蒸汽1 .

[0191] T 蒸汽1 =(ρ 蒸汽 ×Q 蒸汽 ×(h 21 -h1))×η1 / (C 混1 ×m 总1 ); Where, ρ 蒸汽 Steam standard density, unit kg / Nm³ 3 The fixed value is approximately 0.804 kg / Nm³. 3 ; Q 蒸汽 : Volumetric flow rate of steam under standard conditions, in Nm³ 3 / s, measured based on a third flow sensor; additionally, ρ 蒸汽 ×Q 蒸汽 The mass flow rate of steam is m 蒸汽 Unit: kg / s; h 21 : Steam enthalpy at the core reaction zone temperature, in kJ / kg. It can be obtained from a steam enthalpy table based on steam pressure and temperature. Specifically, it can be found in the superheated steam table based on a target core reaction zone temperature of 950℃ and a steam pressure of 0.1MPa. h21 ≈4650kJ / kg; Note: 0.1MPa is the standard value of the flame zone pressure. The core reason is that torch combustion is "open combustion", and the flame zone pressure is approximately equal to the local atmospheric pressure. 0.1MPa is the "standard atmospheric pressure reference value" commonly used in industrial calculations. h1: Initial enthalpy of steam, in kJ / kg, determined by the saturation temperature under initial steam pressure P1. For example, if P1 = 0.9 MPa and initial steam temperature T1 = 200℃, the corresponding superheated steam table would be h1 ≈ 2860 kJ / kg.

[0192] (6) T 散1 .

[0193] T 散1 = (5.67 × 10 -8 ×ε×A1×(T 核实 4 -T 环 4 )+h 对1 ×A1×(T 核实 -T 环 ))×10 -3 / (C 混1 ×m 总1 ); Among them, 5.67×10 -8 σ is the Stefan-Boltzmann constant, in W / (m²). 2 ·K 4 ); ε: Flame emissivity, dimensionless, used to describe the ability of a flame to radiate heat. In engineering applications involving high-temperature flames in acid gas torches, a value of 0.8-0.9 is commonly used, for example, ε=0.85. A1: Heat dissipation area of ​​the core reaction zone, in meters (m²) 2 It is approximately the "cylindrical lateral surface area", calculated from the torch structure parameters: A1=π×d1×h1, where d1 is the diameter of the acid gas torch head and h1 is the height of the core reaction zone, both of which are given by the manufacturer; T 核实 The measured temperature of the core reaction zone is determined by the corrected temperature field and is expressed in Kelvin. The measured temperature of the core reaction zone can be exemplified as the temperature characteristic value of the core reaction zone. The effective area corresponding to the core reaction zone in the corrected temperature field can be used as the representative value of the global temperature of the core reaction zone. The effective area can be delineated using existing mature methods in combination with flare structure parameters (diameter d1 of the acid gas flare head, flame height of the core reaction zone, etc.). The temperature characteristic value can be determined based on the temperature data within the effective area using existing mature methods (such as the weighted average method, the high-temperature zone average method, etc.), which will not be elaborated here. T 核实 4 :T 核实 The fourth power is the core driving factor for the radiative heat of the flame in the core reaction zone. The higher the temperature, the faster the fourth power value increases and the greater the radiative heat dissipation. T 环 Ambient temperature, in Kelvin (K). T 环 4 The fourth power of the ambient temperature represents the background radiation of the ambient air, used to calculate the "radiative temperature difference" between the flame and the environment. The larger the difference, the stronger the radiative heat dissipation. h 对1 : Convective heat transfer coefficient of the core reaction zone, in W / (m²) 2 ·K), h 对1 The empirical formula for flare combustion systems is: h 对1 =18+5v, where v is the wind speed in m / s and 18 is in W / (m²). 2 ·K), the unit of 5 is W / (m 2 The formula ·K·(m / s) is suitable for normal operating conditions with wind speeds of 0.5-10 m / s. If the wind speed is >10 m / s, it can be corrected to h. 对1 =18+5.5v, improving calculation accuracy.

[0194] (7) T 主前 .

[0195] After the flue gas leaves the core reaction zone, it will slightly dissipate heat to the environment during the brief transmission process as it diffuses into the main high-temperature zone (without additional combustion / heat absorption).

[0196] For example, the average temperature at the boundary between the core reaction zone outlet and the main high-temperature zone inlet can be directly extracted using the corrected temperature field data as T. 主前 Specifically, based on the two-zone boundary provided by the torch manufacturer (e.g., the core reaction zone height h1 = 0.6m, the main high-temperature zone starting height h1 + transmission path length = 1.1m), a "boundary region" (vertical 0.6m-1.1m, horizontally the same as the core reaction zone boundary) can be delineated on the corrected temperature field data. Temperature data within this boundary region can then be extracted from the corrected temperature field data, and extreme interference points (such as those higher than T) can be removed. 核实 (The outliers and low-temperature points below 500°C) are identified. Then, the arithmetic mean of the remaining temperature data after removing extreme interference points is calculated, which is T. 主前 (If Kelvin K is required in the thermal equilibrium model, the unit conversion must be performed according to "K=°C+273").

[0197] For example, assuming the remaining temperature data extracted from the boundary area are 895°C, 902°C, 905°C, and 898°C, the arithmetic mean is approximately 900°C.

[0198] (8) T 余1 .

[0199] T 余1 = ( T H2S1 + T NH31 + T 丙烷1 - T 蒸汽1 - T 散1 )×C 混1 ×m 总1 / (C 混2 ×m 总2 ); in, T 余1 Net exothermic heat generated from the core reaction zone that was not absorbed; C 混2 The average isobaric specific heat capacity of the mixed gas in the main high-temperature zone, measured in kJ / (kg·℃), is a mass-weighted average of the specific heat capacity of the flue gas in the core reaction zone plus the specific heat capacity of the entrained air. Its value is close to the specific heat capacity of air (due to the high proportion of entrained air). The calculation formula is: C 混2 =(m 烟1 ×C 烟1 +m 空吸 ×C 空 ) / m 总2 , where m 烟1 =m 总1 C 烟1 =C 混1 C 空 The mass weighted average is approximately 1.07 kJ / (kg·°C), where 0.78 is the mass percentage of nitrogen, 1.08 is the specific heat capacity of nitrogen, 0.21 is the mass percentage of oxygen, 1.05 is the specific heat capacity of oxygen, 0.01 is the mass percentage of other gases, and 0.52 is the specific heat capacity of other gases. m 总2 Total mass flow rate of mixed gas in the main high-temperature zone, unit: kg / s, m³ 总2 =m 总1 +m 空吸 ; Note: The main high-temperature zone is diffusion combustion, which entrains ambient air to replenish oxygen for the subsequent combustion of incompletely burned components. The mass flow rate of the entrained air is expressed as m³. 空吸 Some torch manufacturers will directly provide the m value in the torch's technical manual or design drawings. 空吸 The standard value. If the torch's technical manual or design drawings specify "the mass ratio of air intake to fuel gas (α)", then... 空吸 Then, it can be expressed by formula m. 空吸 =α 空吸 ×m 燃料气 Calculate, where m 燃料气 It is the total mass flow rate of fuel gas in the flare combustion, which includes the sum of the mass flow rates of combustible components (H2S, NH3) in the acid gas and auxiliary fuel gas (such as propane gas). Alternatively, it can be directly calculated from the "total volumetric flow rate of combustible media" of the flare's rated processing capacity combined with the corresponding density.

[0200] (9) T H2S2 .

[0201] T H2S2 =[ρ H2S ×Q 酸气 ×M H2S ×LHV H2S [×(1-η1)]×η2 / (C 混2 ×m 总2 ); Wherein, η2: Combustion efficiency in the main high-temperature zone, with a value range of 0.9-0.95. The initial value (engineering default value) depends on the situation. Under normal operating conditions (sufficient entrained air, stable acid gas load), η2 = 0.92-0.93 (typical range of near-complete combustion); under high-load conditions (acid gas flow rate > 80% of rated value, entrained air volume increases simultaneously), η2 = 0.93-0.95 (sufficient oxygen, more complete combustion); under low-load conditions (acid gas flow rate < 30% of rated value, relatively excessive entrained air), η2 = 0.90-0.92 (sparse fuel, some components may not be completely oxidized); under emergency scenarios (sensor failure, insufficient entrained air), η2 = 0.90 (a conservative value to ensure that the calculated temperature is not lower than the minimum target value). After the initial value is determined, it can be calibrated periodically according to a mature calibration method (e.g., once a month).

[0202] (10) ΔT NH32 .

[0203] ΔT NH32 =[ρ NH3 ×Q 酸气 ×M NH3 ×LHV NH3[×(1-η1)]×η2 / (C 混2 ×m 总2 ); Wherein, η2: Combustion efficiency in the main high-temperature zone.

[0204] (11) T 散2 .

[0205] T 散2 = (5.67 × 10 -8 ×ε×A2×(T 主实 4 -T 环 4 )+h 对2 ×A2×(T 主实 -T 环 ))×10 -3 / (C 混2 ×m 总2 ); Where A2: Heat dissipation area of ​​the main high-temperature zone, in m² 2 It is approximately "cylindrical lateral surface area", which is consistent with the geometric model of the core reaction zone A1 and is adapted to the diffusion pattern of the torch flame. A2 = π × d2 × h2, where d2 is the average diameter of the flame in the main high-temperature zone (determined by the torch manufacturer through combustion tests) and h2 is the flame height in the main high-temperature zone (a typical value given by the manufacturer). T 主实 : Measured temperature values ​​of the main high-temperature areas, in K, determined based on the corrected temperature field; h 对2 : Convective heat transfer coefficient in the main high-temperature zone, h 对2 =18+5v.

[0206] Based on the above description, the input parameters of the thermal balance model constructed by the control module in this embodiment are shown in Table 5 below.

[0207] Table 5 Input parameters of the thermal equilibrium model

[0208] Examples of the output parameters of the thermal equilibrium model are shown in Table 6 below.

[0209] Table 6 Output parameters of the thermal equilibrium model

[0210] The important formula used in the thermal equilibrium model is T 核预 T 主预 The calculation formulas are as follows. These two formulas can be used both to forward predict the combustion temperature of the core reaction zone and the main high-temperature zone, and to reversely solve for the target flow rate. For example, assuming T...核预 The target value is 950°C, T 主预 The target value is 875°C. With other parameters remaining unchanged, the target flow rate of propane gas and / or steam can be calculated in reverse.

[0211] Predicted combustion temperature (T) 核预 T 主预 ) and measured combustion temperature (T) 核实 T 主实 It can be used to control the combustion temperature of the core reaction zone and the main high-temperature zone in the control module. The core control logic is to build a two-layer control system of "prediction-actual feedback-closed-loop correction": use "predicted value" to intervene in advance to prevent risks, use "actual value" to calibrate the accuracy to make up for deviations, and finally adjust the flow rate dynamically through feedback.

[0212] First-level control: based on predicted values ​​(T) 核预 T 主预 Determine the basic direction of regulation (forecasting and judgment): T 核预 and T 主预 It is a predicted value calculated by the heat balance model based on real-time parameters (acid gas flow rate, concentration, current propane / steam flow rate), characterizing the future temperature trend. Its core function is to proactively adjust the flow rate before the measured temperature exceeds the limit, avoiding the situation where the temperature remains constant at T. 核实 / T 主实 If the standard is exceeded, further adjustments will be made (at which point gaseous sulfur has already been generated or the pollutant destruction rate has failed to meet the standard).

[0213] The main controllable objects in this layer include: propane flow rate (prioritized for regulation, as it directly affects the combustion heat release in the core reaction zone and has a significant impact on T). 核预 / T 主预 (Highest sensitivity), steam flow rate (auxiliary control).

[0214] The control rules for this layer (primarily based on the core reaction zone, with indirect linkage to the main high-temperature zones) are shown in Table 7 below.

[0215] Table 7 Regulation Rules

[0216] Second layer: Based on measured temperature values ​​(T) 核实 T 主实 Feedback correction (calibration accuracy): T 核实 and T 主实 It is the actual temperature collected by the corrected temperature field (thermal imaging + thermocouple). Its core function is to verify the effect of predictive regulation, correct model deviations, and avoid excessive or insufficient flow control.

[0217] Feedback corrections are divided into two categories: "real-time fine-tuning" and "periodic calibration".

[0218] (1) Make real-time fine adjustments in the basic control direction (e.g., once every 1 second, targeting the core reaction zone T). 核实 ).

[0219] Examples of real-time fine-tuning logic are shown in Table 8 below.

[0220] Table 8 Real-time Fine-tuning Logic

[0221] (2) Periodic calibration of thermal balance model parameters (exemplary once a month, for the main high-temperature zone T) 主实 ).

[0222] The combustion temperature in the main high-temperature zone is indirectly controlled by the waste heat in the core reaction zone. This response is delayed and does not require real-time fine-tuning; it can be controlled simply by adjusting the T... 主实 Regularly calibrate model parameters (such as η2) to ensure T 主预 The long-term accuracy. Exemplary calibration rules are shown in Table 9 below.

[0223] Table 9 Calibration Rules

[0224] Both of the aforementioned control layers are completed within a single control cycle. This means that each control cycle involves "receiving feedback → prediction → basic regulation + fine-tuning → issuing commands," and only one integrated command is issued per control cycle to avoid frequent actions. The next control cycle then repeats this closed loop. Each control cycle's "prediction" is based on the "feedback data" of the previous control cycle, essentially "correcting the next prediction with the actual results," making the adjustment increasingly precise.

[0225] The following describes the control actions within three consecutive control cycles, with one second as one control cycle.

[0226] 1 second: Receive feedback data: Collect feedback data (e.g., T) after adjustment in the previous control cycle. 核实 =1120°C, T 主实 =1060°C); Prediction: The predicted temperature (T) will be recalculated using a thermal balance model. 核预 and T 主预 For example, calculating T 核预 =1110°C (overheating), T 主预 Temperature not exceeded (<1000°C); Determine the basic direction of regulation and conduct real-time feedback fine-tuning: Based on Table 7, the basic control direction is determined as follows: the propane gas flow rate is reduced by 5% of the rated value in a single operation, while the steam flow rate remains unchanged. Assume the rated propane gas flow rate = 10 Nm³.3 / h, and the current flow rate is 10, then the flow rate after basic adjustment is 9.5; If this is not the first time the temperature has exceeded the limit, then according to Table 8, the real-time feedback fine-tuning should be: continue to reduce the temperature by another 0.5 Nm at a fixed step size. 3 / h, increase steam flow rate by 1Nm 3 / h; then the propane gas flow rate will be reduced by 0.5 from 9.5, becoming 9; Then an instruction was issued: "Propane gas flow rate from 10 → 9 Nm" 3 / h”+Increase steam flow rate by 1Nm 3 The control command is " / h". Alternatively, an instruction can be issued to adjust the target flow rate (e.g., reduce propane flow rate by 1 Nm³). 3 / h, steam flow rate increased by 1Nm 3 The control command is " / h".

[0227] 2nd second: Receive feedback data: Collect feedback data after adjustment in the previous control cycle (e.g., T). 核实 =1080°C, T 主实 =1020°C); Prediction: The predicted temperature (T) will be recalculated using a thermal balance model. 核预 and T 主预 For example, calculating T 核预 =1070°C (still above the limit but improving), T 主预 Temperature not exceeded (<1000°C); Determine the basic direction of regulation and conduct real-time feedback fine-tuning: Based on Table 7, the basic control direction is determined as follows: the propane gas flow rate is reduced by 5% of the rated value in a single operation, while the steam flow rate remains unchanged. Using the example from the previous second, the propane gas flow rate is reduced from 9 to 8.5. If this is not the first time the temperature has exceeded the limit, then according to Table 8, the real-time feedback fine-tuning should be as follows: continue to reduce the temperature by 0.5 Nm in fixed increments. 3 / h, increase steam flow rate by 1Nm 3 / h; then the propane gas flow rate is reduced by 0.5 from 8.5, to 8; Then an instruction was issued: "Propane gas flow rate from 9 to 8 Nm". 3 / h”+Increase steam flow rate by 1Nm 3 The control command is " / h". Alternatively, an instruction can be issued to adjust the target flow rate (e.g., reduce propane flow rate by 1 Nm³). 3 / h, steam flow rate increased by 1Nm 3 The control command is " / h".

[0228] 3 seconds: Receive feedback data: Collect feedback data (e.g., T) after adjustment in the previous control cycle.核实 =1050°C, T 主实 =1000°C); Prediction: The predicted temperature (T) will be recalculated using a thermal balance model. 核预 and T 主预 For example, calculating T 核预 =1045°C (meets standard), T 主预 Temperature not exceeded (<1000°C); Determine the basic direction of regulation and conduct real-time feedback fine-tuning: Since the predicted values ​​meet the standards in both zones, the basic control direction is determined according to Table 7: maintain the current flow rates of propane gas and steam, while the measured values ​​also meet the standards, and T... 核实 With T 核预 If the deviation is 5°C, then according to Table 8, fine-tune η1 = 0.8 → 0.799 (feedback fine-tuning to smooth fluctuations). Then, a control command is issued instructing "fine-tune η1 from 0.8 to 0.799", or a control command is issued instructing "decrease η1 by 0.001". Of course, if T... 核实 With T 核预 If the deviation is 5°C, no fine-tuning is required.

[0229] The two layers mentioned above represent the standard rules. In addition, special scenario linkage controls can be set to prevent the system from going out of control.

[0230] For example, when there is an extreme deviation between the predicted temperature value and the actual temperature value (such as sensor failure or sudden change in operating conditions), it is necessary to break the conventional rules and achieve emergency control through the linkage of "prediction-actual measurement". The following is an example.

[0231] Scene 1: T 核预 Normal, but T 核实 A sudden increase.

[0232] For example, T 核预 =1050°C (meets the standard), but T 核实 A sudden increase in temperature from 1050°C to 1150°C (overheating within 1 second) is usually caused by a sudden increase in the concentration of acidic gas (the model did not collect parameters in time). Linked emergency control actions: not dependent on T 核预 Based directly on T 核实 Implementing an "emergency propane reduction + steam increase" strategy may, for example, involve reducing the propane gas flow rate by 15% of the rated value (e.g., from 8 → 6.8 Nm). 3 / h), increase steam flow rate by 10% of rated value (e.g., from 15→16.5Nm). 3 / h), which simultaneously triggers a concentration sensor fault alarm.

[0233] Scene 2: T核实 Normal, but T 主实 A sudden drop.

[0234] For example, T 核实 =1000°C (meets standards), T 核预 =1020°C (meets the standard), but T 主实 A sudden drop from 950°C to 650°C (low temperature) indicates insufficient air entrainment in the main high-temperature zone (m 空吸 Reduce), and replenish incomplete combustion; Linked emergency control action: Increase propane gas flow rate (e.g., from 7 → 7.5 Nm). 3 / h), to enhance the waste heat ΔT in the core reaction zone. 余1 At the same time, check the torch head suction structure (whether it is blocked) to ensure sufficient oxygen in the main high-temperature zone.

[0235] Scene 3: T 核实 / T 主实 Sensor malfunction (no measured value).

[0236] For example, the thermal imager lens was obscured by smoke, T 核实 / T 主实 Unable to collect data, or thermocouple malfunction; Linked emergency control actions: Switch to "pure predictive control" - based solely on T 核预 / T 主预 Adjust the flow rate, but reduce the adjustment range by half (e.g., reduce the rated value by 3% at a time) to avoid over-adjustment, and trigger a sensor fault alarm until the fault is resolved.

[0237] The flow rate adjustment amounts described in the above embodiments are relatively fixed (generally 0.5, 1, etc.). In other embodiments of this application, the flow rate adjustment amount can also be obtained based on self-learning. Compared with the fixed flow rate adjustment amount, the value of the flow rate adjustment amount obtained by self-learning is more flexible and can be considered as an optimization of the fixed flow rate adjustment amount. In this case, the temperatures of the thermocouple and the thermal imager are compared and correlated, and the self-learning function is used to optimize the flow rate of the propane-blended gas, the steam flow rate, and the flame zone temperature parameters, so that the acid gas can be safely and completely combusted through the acid gas torch head.

[0238] The following describes an exemplary implementation of self-learning.

[0239] First, a sample set for machine learning models is constructed based on the operating condition data.

[0240] For example, each sample includes T 基 T 环 、v、Q 酸气 Q 丙烷 Q 蒸汽 ρ H2Sρ NH3 ρ 丙烷 m 酸气 m 丙烷 m 蒸汽 LHV H2S LHV NH3 LHV 丙烷 T 核实 T 主实 T 主前 η 破 The collected data, η 破 The label represents the pollutant degradation rate and corresponds to the optimal adjustment amount (including the optimal adjustment amount for propane gas and steam). This article will explain how to obtain the label later. Those skilled in the art can add or subtract parameters from the sample as needed; for example, each sample may include v and Q. 酸气 Q 丙烷 Q 蒸汽 ρ H2S ρ NH3 ρ 丙烷 ΔT 核 ΔT 主 η 破 , where ΔT 核 =T 核实 - Target combustion temperature in the core reaction zone (e.g., 950°C), ΔT 主 =T 主实 - Target combustion temperature for the main high-temperature zone (e.g., 875°C).

[0241] Then, the machine learning model is trained using samples from the sample set, so that the trained machine learning model can output the corresponding optimal adjustment amount based on the real-time parameters.

[0242] The following section introduces more specific training and reasoning methods.

[0243] (1) Preprocess the samples (e.g., clean, normalize, label) to obtain the sample set.

[0244] (2) Set constraint targets.

[0245] Constraint objectives may include core constraint objectives, cost optimization objectives, and operational stability objectives.

[0246] The core constraint objectives (highest priority, assigned weight W1) can be set as follows: 800°C≤T 核实 ≤1100°C, 700°C≤T 主实 ≤1050°C, pollutant destruction rate η 破 ≥99%, gaseous sulfur formation = 0; The cost optimization objectives are set as follows (second priority, weighted by W2): The propane consumption intensity (propane consumption per unit of acid gas treatment) is the lowest. Set the stable operation target (lowest priority, weight W3) as follows: The optimal steam consumption (the minimum steam volume required to meet smoke suppression needs, i.e., a steam / acid gas mass ratio ≥ 0.8) and the minimum fluctuation in adjustment (e.g., single-time ΔQ) are achieved. 丙烷 ≤5% of rated value, ΔQ 丙烷 ΔQ refers to the difference in propane regulation between the current control cycle and the previous control cycle. 丙烷 = Current control cycle propane adjustment amount - previous control cycle propane adjustment amount, which is used to quantify the adjacent cycle fluctuation range of propane adjustment amount).

[0247] Where W1>W2>W3=0.1, and W1+W2+W3=1, for example, W1=0.6, W2=0.3, W3=0.1 can be set.

[0248] (3) Set the objective function according to the constraints.

[0249] An example of the objective function J could be: J=W1×(1-δ 约束 )+W2×(Q 丙烷 / Q 酸气 )+W3×(Q 蒸汽 / Q 酸气 +|ΔQ 丙烷 | / Q 丙烷额定流量 ); Where, δ 约束 The constraint satisfaction coefficient (δ when all constraint objectives are satisfied) 约束 =1, if any constraint objective is violated, then δ 约束 =0), the goal is to minimize the value of J, Q. 丙烷额定流量 This is the rated flow rate for propane.

[0250] (4) Use the sample set to iteratively train the machine learning model.

[0251] In one example, a two-stage learning logic of "offline training + online update" can be adopted to balance learning accuracy and real-time performance.

[0252] (1) Offline training phase (monthly calibration).

[0253] Construct offline data samples: Collect at least one month of historical operating condition data as the original sample, and divide the original sample into multiple operating condition clusters according to "load range (low / medium / high) × H2S concentration range × wind speed range"; For example, the H2S concentration range can be divided into three ranges: <60%, 60%-80%, and >80%, and the wind speed range can be divided into three ranges: <3m / s, 3m / s-5m / s, and >5m / s. Thus, the original sample can be divided into 3×3×3=27 operating condition clusters. Preprocess each original sample (e.g., cleaning, normalization, labeling) to obtain labeled samples; among them, the labeled samples in the same working condition cluster can be divided into 70% as training samples, 20% as test samples, and the remaining 10% as validation samples (used to adjust model parameters during training to avoid overfitting); if the sample size is moderate (500-1000 samples), it can be simplified to 8:2 (training set: test set), and cross-validation of the training set can replace the independent validation set; Offline training method: For each working condition cluster, the training samples within the cluster are used separately to train a gradient boosting regression (GBRT) mapping model with the goal of minimizing the objective function. Finally, a total of 27 independent mapping models are obtained (one model for each working condition cluster). Each mapping model is used to output the optimal adjustment amount under the corresponding working condition based on the input data. Model validation: Validate using 20% ​​of the test samples, requiring adjusted temperature deviation ≤ ±5°C, propane consumption reduction ≥ 5%, and constraint satisfaction rate ≥ 99%. If the requirements are not met, retraining is required.

[0254] It should be noted that the sample labels can be obtained through the "optimization-sorting-verification" process. The "optimization-sorting-verification" process is a labeling operation in the preprocessing stage.

[0255] Taking a single-condition cluster as an example, the preprocessing includes: The original samples in the operating condition cluster are cleaned and normalized to obtain the first intermediate sample; From the first intermediate sample, samples that satisfy the aforementioned core constraints are selected for optimal solution seeking (the selected samples that satisfy the aforementioned core constraints can be called the second intermediate sample): For each second intermediate sample, calculate its J value according to the objective function; The flow regulation of the second intermediate sample with the smallest J value is taken as the "optimal regulation" for this operating condition cluster, i.e., the label. Perform label verification: Reproduce the label in the working conditions of similar historical samples in this working condition cluster using the flow adjustment amount corresponding to the label, and ensure that the constraint satisfaction rate is ≥99% after adjustment, the J value is stable and minimal, and there are no abnormalities, then solidify the label; Label the first intermediate sample in the working condition cluster with a fixed label to obtain the labeled sample.

[0256] The reproduction logic is as follows: The label to be verified (e.g., propane-0.6Nm)3 / h, steam +0.4Nm 3 / h) Substitute the working conditions provided by historical similar samples (such as Q) 酸气 =100Nm 3 / h、ρ H2S =70%, wind speed =3m / s, etc.), the simulation system runs according to this verification label; Check whether the results of the simulation meet the core constraint objectives. For example, if the T value of the next control cycle in the simulation... 核实 =950°C、η 破 =99.3% is satisfied, and if the simulation yields T for the next control cycle... 核实 =780°C does not meet the core constraint objective; After reproduction, calculate the constraint satisfaction rate, standard deviation of J value, and mean J value: The constraint satisfaction rate can be calculated as: (Number of historical similar samples that satisfy all core constraints after reproduction / Total number of historical similar samples) × 100%; The standard deviation of the J value and the average J value can be obtained as follows: The simulation results after running each historical similar sample (Q of the next control cycle obtained from the simulation) 丙烷 Q 酸气 Q 蒸汽 ΔQ 丙烷 Q 丙烷额定流量 ), calculate the J value for each historical similar sample, and obtain the J value for each sample. Calculate the standard deviation and mean of the J value based on the J value corresponding to each historical similar sample. 平均1 ); If the standard deviation of the J value is ≤0.02, the J value is considered stable. If J 平均1 If the value is ≤J1 (original cluster label J value) × 1.03, then it is considered to be normal.

[0257] (2) Online phase (real-time inference is performed during this phase, and label updates and online training are conducted to adapt to changes in operating conditions).

[0258] Online training is a "dynamic upgrade of offline training". The core logic is: under safety constraints, the model is gradually iterated from "historical fixed adjustment amount" to "fine optimal adjustment amount adapted to the working conditions" through "small-scale trial supplementation of fine samples + real-time optimization and label update + incremental training to optimize the model", while ensuring that the process does not violate the process constraints (temperature, environmental protection, equipment safety).

[0259] The online phase may include the following parts.

[0260] (1) Data processing.

[0261] 1) Real-time data acquisition and preprocessing.

[0262] Real-time data may include the “real-time acquisition parameters” and “measured feedback parameters” in Table 5.

[0263] One use of real-time data is as input to a thermal equilibrium model. Combined with the output of the thermal equilibrium model, it can be used for real-time inference. In addition, after preprocessing and filtering, it can be used for online training.

[0264] For online training, for example, real-time v and Q values ​​within the same control cycle can be used. 酸气 Q 丙烷 Q 蒸汽 ρ H2S ρ NH3 ρ 丙烷 ΔT 核 ΔT 主 η 破 As operating condition parameters, they are cleaned and normalized to form a real-time sample. The real-time samples that meet the core constraint objectives are selected and can be called the first real-time sample (at this time, they are not yet labeled).

[0265] 2) Matching of detailed working condition clusters.

[0266] According to "load range × H2S concentration range × wind speed range × ΔT" 核 Interval (Note the addition of ΔT) 核 Three temperature ranges: -10℃ to -5℃, -5℃ to +5℃, and +5℃ to +10℃, will be used to match the first real-time sample to the corresponding sub-condition cluster.

[0267] It should be noted that, compared with offline training, the online phase further subdivides the working condition clusters, resulting in a total of 3×3×3×3=81 subdivided working condition clusters, in order to improve the model's relevance.

[0268] (2) Small-scale trial (optional).

[0269] Small-scale trials are a "fine-grained optimization supplement" to online training. Their core purpose is to explore fine-grained optimal solutions (such as 0.4 Nm) that are not covered by historical fixed adjustment values ​​within a safe boundary. 3 / h, 0.6Nm 3 The ` / h` function executes only when all triggering conditions are met, without affecting system stability. It can periodically (synchronously with the control cycle) verify in real time whether all triggering conditions are met.

[0270] Examples of triggering conditions include operating condition matching conditions, temperature deviation conditions, and operating condition stability conditions. These are described in detail below: Operating condition matching conditions: The sub-operating condition cluster corresponding to the current operating condition parameters is a "medium-high frequency cluster". The medium-high frequency cluster refers to the sub-operating condition cluster with a proportion of ≥5% and a sample accumulation of ≥500, to ensure that the trial is meaningful; the proportion here refers to the ratio of the cumulative running time of the sub-operating condition cluster to the total running time within a certain period of time. The total running time is equal to the sum of the cumulative running times of each sub-operating condition cluster within that period of time. Temperature deviation condition: ΔT 核 ∈[-5℃, +5℃] (meets standards or has slight deviations, with no major risks); Stable operating conditions: If the following three sub-conditions are met simultaneously for five consecutive operating cycles, the operating condition stability condition is satisfied; otherwise, the operating condition stability condition is not satisfied: Subcondition 1: Q 酸气 Relative fluctuation ≤ ±3%; Sub-condition 2: Relative fluctuation of H2S concentration ≤ ±3%; Sub-condition 3: Absolute fluctuation of ambient wind speed (v) ≤ 0.5 m / s; A small probe is triggered only when all of the above triggering conditions are met simultaneously.

[0271] The rules for conducting small-scale tests are as follows: Trial stride: 0.1 Nm can be taken as the trial stride. 3 / h; The testing range is limited to ±0.3 Nm of the flow regulation amount obtained from real-time inference (which can be referred to as the current tag regulation amount). 3 Explore within the / h range. For example, assume that in the current label adjustment, the propane gas adjustment is -0.5Nm. 3 If / h, then its testing range is -0.8Nm. 3 / h~-0.2Nm 3 / h, this helps avoid deviating from the safe range; Probing direction: Prioritize exploring energy conservation, including: When the temperature exceeds the limit, the current label adjustment amount of propane gas should be reduced in a trial step (at this time, the current label adjustment amount of propane gas is a decrease, and reducing it will make the absolute value of the decrease larger, thus cooling down faster), while steam should be maintained or the current label adjustment amount should be increased in a trial step (at this time, the current label adjustment amount of steam is an increase, and increasing it will make the absolute value of the increase larger, thus absorbing more heat). At low temperatures, the current label adjustment amount of propane gas is increased in a trial increment (at this time, the current label adjustment amount of propane gas is an increment, and increasing it will make the absolute value of the increment larger, thus heating up faster), while the current label adjustment amount of steam is maintained or increased in a trial increment (at this time, the current label adjustment amount of steam is a decrement, and increasing it will make the absolute value of the decrement larger, thus absorbing less heat). Single test limit: The same sub-condition cluster shall not be tested more than 3 times consecutively, and the interval between two tests shall be ≥10 control cycles to avoid temperature oscillation caused by frequent adjustments; A small trial includes: adjusting the current label adjustment amount upward or downward with a trial step size based on the trial execution rules, determining a unique set of trial adjustment amounts (propane gas + steam) as the trial adjustment amounts for this control cycle, and issuing a control command indicating the trial adjustment amounts.

[0272] For example, assuming overheating, the adjustment amount obtained from real-time inference is propane gas flow rate -1 Nm³. 3 / h, steam flow rate +1Nm 3 / h, then the exploration in the direction of energy saving includes: with a trial step size of 0.1Nm 3 The propane gas flow rate was adjusted to -1.1 Nm³ / h. 3 / h, the steam level is maintained or the steam regulation is increased to +1.1Nm in trial increments. 3 / h.

[0273] Of course, if no small-scale trial is conducted within this control cycle, a control command indicating the normal flow rate adjustment amount will be issued—this normal flow rate adjustment amount is obtained through real-time reasoning, which is also the current tag adjustment amount.

[0274] After a small-scale trial, the effectiveness of the trial will be evaluated and addressed, which may include: Real-time data acquisition: Acquire real-time data (including T) in the next control cycle. 核实 ), calculate ΔT based on real-time data 核 ; If ΔT 核 If the temperature exceeds ±8℃ (constraint boundary), immediately revert to the flow rate adjustment before the test and mark the corresponding test sample as "invalid". If ΔT 核 If the temperature is within ±8℃, further effective trial determination will be performed: if the trial adjustment amount meets the core constraint target and (J2-J 试探 If J / 2×100%≥3%, then the trial sample is marked as a "valid trial sample" and included in the training pool; otherwise, it is discarded. 试探 J1 refers to the J value corresponding to the trial adjustment amount, and J2 refers to the J value corresponding to the current label adjustment amount.

[0275] It should be noted that the above-mentioned test sample (whether valid or invalid) is a structured data record of "closed-loop information of the entire test process". It contains four major categories of key content. Even if it is marked as "invalid", it must be stored in its entirety for subsequent traceability and rule optimization. The specific contents are as follows.

[0276] The first category is background information on the working conditions before the trial, in order to clarify "under what scenario the trial should be initiated".

[0277] Examples include the currently matched sub-family of operating conditions and core process parameters (such as the standard flow rate of acid gas, ρ). H2S (specific value of v), temperature state (e.g., T) 核实 T 主实 and ΔT 核 ΔT 主 ), the optimal label adjustment amount for the current operating condition cluster, and the precise timestamp of the trial initiation (e.g., 2024-05-20 14:32:15).

[0278] The second category: information on actions performed during trial execution, to record "what trial operation was performed".

[0279] Examples include: trial adjustment amount, trial step size, and trial marker (to indicate that the adjustment action is a trial operation rather than a regular adjustment).

[0280] The third category: feedback information on the effects of the trial, to show "what results were produced after the trial".

[0281] Examples include real-time temperature data for the next control cycle after adjustment (T for the next control cycle). 核实 T 主实 and ΔT 核 ΔT 主 The data includes: constraint satisfaction (whether the core constraint objectives are met), equipment operating status (such as the opening degree of the propane regulating valve, whether there is an alarm signal), and J value (calculated using measured data from the next control cycle).

[0282] The fourth category is sample attributes and processing information, which are used to label "sample status and subsequent actions".

[0283] Examples include: validity markers (invalid samples are labeled 0, valid samples are labeled 1), and specific reasons for invalidity (such as ΔT). 核 =-8℃ trigger rollback threshold, rollback operation (such as immediately rolling back to the adjustment amount before the test), and operation subject (such as automatic execution by the system, marked as automatic).

[0284] (3) Dynamic updates of the training pool (periodic updates, synchronized with the control cycle).

[0285] The training pool is dynamically updated, including: Sample accumulation: The "normal adjustment effective samples" and "effective trial samples" are classified and stored in the corresponding training pools according to the sub-operating condition clusters. The maximum capacity of the training pool for each sub-operating condition cluster is 2000 samples (if the capacity is exceeded, the earliest sample is removed according to the "first-in, first-out" method to ensure the timeliness of the data). It should be noted that a valid sample for routine adjustment refers to the first real-time sample corresponding to the "adjustment amount obtained by real-time inference" that was issued without any small-scale trial during the control period and satisfies the core constraint objectives. Sample labeling: Automatically label new samples in the training pool with temporary labels. If the sample is a regular adjustment sample, the temporary label is its own adjustment amount. If the sample is a valid trial sample, the temporary label is the trial adjustment amount.

[0286] (4) Real-time optimization and tag updates.

[0287] When the training pool of a certain sub-condition cluster adds ≥50 valid samples or the cumulative running time of the sub-condition cluster reaches a preset time (e.g., 100 seconds), real-time optimization is triggered to update the label of the cluster.

[0288] Real-time optimization may include the following steps.

[0289] 1) Select effective samples with a working condition similarity of ≥90% from the training pool of this cluster to form the optimization group (at least 50 samples).

[0290] The training pool contains regular adjustment effective samples and effective trial samples, from which core normalization parameters (such as Q) shared by these two types of samples can be extracted. 酸气 ρ H2S ρ NH3 Ambient wind speed v, ΔT 核 (etc.) to form a working condition feature vector with consistent dimensions and uniform units; by calculating the similarity of the working condition feature vectors among each effective sample (e.g., cosine similarity), effective samples with a working condition similarity ≥ 90% can be selected based on the similarity value.

[0291] 2) Calculate the J value for each valid sample in the optimization group according to the aforementioned objective function.

[0292] 3) Select the adjustment amount corresponding to the effective sample with the smallest J value as the "candidate real-time label".

[0293] 4) Use the independent samples in the training pool of this cluster that did not participate in the optimization to form a validation group for candidate real-time label validation.

[0294] The number of independent samples in the verification group should be ≥30, and the independent samples should satisfy the requirement that the similarity between the working condition feature vector and any valid sample in the optimization group is not less than 90%.

[0295] Label verification for candidate real-time labels includes: reproducing the flow adjustment amount corresponding to the candidate real-time label in the operating conditions corresponding to the independent sample, using the reproduction results for verification, and updating the label if the verification is successful.

[0296] If it fails, the adjustment value corresponding to the second smallest effective sample with J value is selected as the "candidate real-time label", and the process is returned to "use the independent samples in the training pool of this cluster that did not participate in the optimization to form a verification group to verify the candidate real-time label" until the verification is successful or the optimization group is empty.

[0297] The reproduction logic is similar to the offline reproduction logic described above: Substitute the candidate real-time label into the operating conditions provided by the independent sample, and the simulation system will run according to the candidate real-time label. Check whether the results after the simulation (i.e. the reproduced results) meet the core constraint objectives; After reproduction, calculate the constraint satisfaction rate, standard deviation of J value, and mean J value: The constraint satisfaction rate can be calculated as: (Number of independent samples that satisfy all core constraints after reproduction / Total number of samples in the validation group) × 100%; The standard deviation of the J value and the average J value can be obtained as follows: Calculate the J value using the results of the simulation run for each independent sample, and obtain the J value corresponding to each independent sample. The standard deviation of the J-values ​​is calculated based on the J-values ​​corresponding to each independent sample, and used as the standard deviation and mean of the validation set (J). 平均2 ); If the standard deviation of the validation set is ≤0.02, then the J value is considered stable; If J 平均2 If the value is ≤J3×1.03, then it is considered to be without anomalies. J3 refers to the J value corresponding to the label currently being used in the subdivision condition cluster.

[0298] 5) Label update: Replace the current real-time label with the verified candidate real-time label to obtain a new real-time label for this sub-condition.

[0299] (5) Incremental training of the model (synchronized with real-time optimization).

[0300] Step A, Training Data Preparation: Taking a single subdivision working condition cluster as an example, the latest 1000 valid samples (including regular samples and valid trial samples) are selected from the training pool of the cluster and split into training set and test set in an 8:2 ratio (full sample cross-validation is used when the sample size is insufficient). Step B: Perform incremental training based on the latest selected valid samples: This incremental training uses the GBRT model obtained from offline training or the previous incremental training, without reconstructing the model structure, only updating the model parameters (e.g., updating the learning rate to 0.005 and the number of iterations to 500 to avoid overfitting). The training objective of this incremental training is to minimize the absolute error between the model's predicted adjustment and the real-time label (the absolute value of the difference between the two). For example, the error... 丙烷 ≤0.05Nm 3 / h, error 蒸汽 ≤0.1Nm 3 / h, Note: Real-time tags refer to the latest tags updated after real-time optimization; The GBRT model obtained after incremental training using the latest valid samples is used as a candidate model. The candidate model is validated using a test set. Requirements: The prediction error of the adjustment amount is ≤0.05Nm. 3 / h (propane), ≤0.1Nm 3 / h (steam); Temperature deviation after adjustment ≤ ±5℃; The constraint satisfaction rate is ≥99%; this means that the proportion of samples that satisfy all core constraint objectives after the adjustment amount predicted by the candidate model is reproduced in the working conditions corresponding to the test set samples is ≥99%; please refer to the above description for how to calculate the constraint satisfaction rate, which will not be repeated here.

[0301] If the validation passes, the candidate model is used to replace the old model for that cluster; if it fails, the old model is retained, and only the training pool samples are updated.

[0302] The flow rate of the gas phase separated by the three-phase separator Q-101 is the main factor causing changes in the combustion temperature of the acid gas flare head. A large flow rate results in a longer and hotter flame and a larger volume in the main high-temperature zone. A small flow rate may cause the flame to detach from the acid gas flare head (extinguish) or the combustion temperature to be insufficient. In this case, controlling the flow rate of propane gas increases the combustion temperature. Injecting steam can promote mixing, smoke elimination, and steam reaction. When the combustion temperature of the core reaction zone and the main high-temperature zone of the acid gas flare head increases, the flow rate of injected steam is controlled to reduce the combustion temperature, so that the combustion temperature of the core reaction zone and the main high-temperature zone of the acid gas flare head reaches the required level, preventing the formation of gaseous sulfur.

[0303] The following describes the structure of the control module.

[0304] like Figure 3As shown in other embodiments of this application, the control module in all the above embodiments includes a newly added acid gas emission management system (AFIS) and a traditional acid gas flare head management system. The traditional acid gas flare head management system can be a DCS (Dispersed / Distributed Control System) or a PLC (Programmable Logic Controller). The newly added acid gas emission management system and the traditional acid gas flare head management system exchange data through an OPC (Open Platform Communications) interface. Compared with the traditional acid gas flare head management system, this embodiment also adds sensors (belonging to the aforementioned detection components) and instruments. For examples, please refer to [link to relevant documentation]. Figure 1 The newly added sensors include a full-jacketed steam tracing temperature transmitter for a full-jacketed steam tracing system, a steam tracing temperature transmitter for a steam tracing system, an electric tracing temperature transmitter for an electric tracing system, a first liquid level sensor, a second liquid level sensor, a first concentration sensor AI-101, a second concentration sensor AI-102, a third concentration sensor AI-103, a first flow sensor FI-102, a second flow sensor, a third flow sensor, a first thermocouple TC-301, a second thermocouple TC-302, a third thermocouple TC-303, and a thermal imager TC-401. The newly added instrumentation includes a three-phase separator Q-101, a full-jacketed steam control valve TV-101 for a full-jacketed steam tracing system, a steam control valve TV-102 for a steam tracing system, a pressure regulating device, conveying components (such as a screw pump P101), a first control valve LV-101, a second control valve LV-102, a propane gas flow control valve FV-101, and a steam flow control valve FV-201.

[0305] The traditional acid gas flare head management system can collect relevant parameters from sensors (including newly added sensors) and send them to the newly added acid gas management system. The newly added acid gas management system issues control commands based on the collected parameters. The equipment controlled by the control commands includes, but is not limited to: full-jacketed steam tracing device, steam tracing device, electric tracing device, pressure regulating device, conveying components, first control valve LV-101, second control valve LV-102, propane gas flow control valve FV-101, and steam flow control valve FV-201. The traditional acid gas flare head management system controls the operation of the newly added instrumentation equipment based on the above control commands.

[0306] The traditional acid gas flare head management system or the newly added acid gas management system can display and analyze the temperature measurement data (obtained by three thermocouples and a thermal imager) of the core reaction zone, main high-temperature zone, and diffusion and cooling zone of the acid gas flare head. Specifically, it can plot the temperature change curve of the temperature measurement data to characterize the temperature change trend. At the same time, it can display the temperature measurement data and temperature change curve on the large screen in the control room and alarm when the combustion temperature exceeds the limit.

[0307] In one example, computer SCADA software can be installed in the newly added acid gas management system. This software can process the collected parameters and generate control commands. The process of processing the collected parameters and generating control commands can be implemented by running Python code.

[0308] In addition, the human-machine interface of the newly added acid gas management system adopts 3D software configuration display, which can clearly see the three-dimensional dynamic picture of the flame of the acid gas torch head, including the changes in different combustion temperatures in each zone. It can intuitively reflect the effect of temperature control. It can also determine the flame diffusion of the acid gas torch head under the influence of wind direction based on the wind direction at the acid gas torch head. Combined with the thermal imaging effect, it can realistically display the combustion temperature trend and alarm when the combustion temperature exceeds the limit.

[0309] The treatment process for acidic gas is as follows: Acidic gas enters the top inlet of the first separation chamber Q-1011 of the three-phase separation tank Q-101, which is equipped with a fully jacketed steam heating device, through the inlet pipe equipped with an electric heating device. A chemical reaction occurs in the three-phase separation tank Q-101. The solid and liquid phases are discharged through the bottom outlet of the first separation chamber Q-1011 and the bottom outlet of the second separation chamber Q-1012. The gas is then discharged to the waste liquid incinerator by the screw pump P101. At this time, heavy metals are discharged in the form of sulfide solid phase. The gas phase (mainly hydrogen sulfide, ammonia, carbon dioxide, and water vapor) is discharged through the top outlet of the second separation chamber Q-1012 and is introduced into the acidic gas flare head through the outlet pipe equipped with a steam heating device. Specifically, the gas phase first enters the water seal tank Q-201. At this time, the gas phase breaks through the water seal and enters the acidic gas flare head from the top of the water seal tank Q-201 for combustion. Among them, the electric heat tracing device, the fully jacketed steam heat tracing device, and the steam heat tracing device operate at a temperature of 150°C to prevent sulfide solid scale buildup and blockage in the pipeline and the three-phase separator Q-101.

[0310] The propane gas and steam are mixed and introduced into the acid gas flare head. The flow rates of the propane gas and steam are controlled to regulate the combustion temperature of the acid gas flare head.

[0311] The combustion temperature of the acid gas torch head is collected by thermocouples and thermal imagers, and an alarm is triggered when the temperature exceeds the limit.

[0312] Water enters the water seal tank, and the liquid level in the water seal tank is monitored, triggering an alarm when the level exceeds the limit.

[0313] The main function of the continuous flame lamp is to ensure that the acid gas torch head can be continuously and stably ignited; therefore, the ignition of the continuous flame lamp is controlled.

[0314] Example 2.

[0315] This embodiment provides an acid gas emission management method, applied to the acid gas emission management system of Embodiment 1, such as... Figure 4 As shown, the acid gas emission management method includes the following steps.

[0316] Step S1: The temperature and pressure of the three-phase separator are controlled within a first temperature range and at a preset pressure, respectively, so that the hydrogen sulfide in the acid gas introduced into the three-phase separator reacts with ammonia in the three-phase separator to generate non-gaseous ammonium hydrosulfide. In addition, the heavy metals in the acid gas react with the sulfides in the system to generate sulfide solid phase.

[0317] Step S2: Issue an adjustment command; the adjustment command is used to cause the adjustment module to adjust the combustion temperature of the acid gas torch head to a second temperature range; the second temperature range is the temperature range that suppresses the generation of gaseous sulfur.

[0318] Example 3.

[0319] In one exemplary embodiment, a computer device is provided that can be used to perform the relevant functions of the control module described above. The computer device includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the acid gas emission management method in Embodiment 2.

[0320] This computer device can be a server or a terminal, and its internal structure diagram can be as follows: Figure 5 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a method for managing acid gas emissions.

[0321] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0322] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0323] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0324] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. An acid gas emission management system, characterized in that, The acid gas emission management system includes: a control module, a heavy metal removal module, and a regulation module that are communicatively connected to the control module; The heavy metal removal module includes a three-phase separation tank. The control module controls the temperature and pressure of the three-phase separation tank within a first temperature range and at a preset pressure, respectively, so that hydrogen sulfide and ammonia in the acidic gas introduced into the three-phase separation tank react to generate non-gaseous ammonium hydrosulfide. Furthermore, heavy metals in the acidic gas react with sulfides in the system to generate a sulfide solid phase. The three-phase separation tank performs three-phase separation, discharging the separated solid and liquid phases, and introducing the separated gaseous phase into an acidic gas flare head for combustion. The gaseous phase includes hydrogen sulfide and ammonia that did not participate in the reaction in the three-phase separation tank. The adjustment module is used to adjust the combustion temperature of the acid gas torch head to a second temperature range; the second temperature range is the temperature range that suppresses the generation of gaseous sulfur.

2. The acid gas emission management system according to claim 1, characterized in that, The heavy metal removal module further includes: a fully jacketed steam tracing device installed on the three-phase separator, a first tracing device installed on the inlet pipe of the three-phase separator, a second tracing device installed on the outlet pipe of the three-phase separator, and a pressure regulating device installed on the three-phase separator; the inlet pipe is used to introduce acidic gas into the three-phase separator, and the outlet pipe is used to introduce the separated gas phase into the acidic gas flare head; The control module is communicatively connected to the full-jacketed steam tracing device, the first tracing device, the second tracing device, and the pressure regulating device, respectively. The control module is also used to perform PID control on the full-jacketed steam tracing device, the first tracing device, the second tracing device, and the pressure regulating device to control the temperature and pressure of the three-phase separator within a first temperature range and at a preset pressure, respectively.

3. The acid gas emission management system according to claim 2, characterized in that, The first heat tracing device is an electric heat tracing device, and the second heat tracing device is a steam heat tracing device.

4. The acid gas emission management system according to claim 2, characterized in that, The first temperature range is 118°C-145°C, and the preset pressure is 0.1 MPa.

5. The acid gas emission management system according to claim 1, characterized in that, The three-phase separator is equipped with a partition inside, which divides the cavity of the three-phase separator into a first separation chamber and a second separation chamber. There is a gap between the partition and the top of the three-phase separator, so that the first separation chamber and the second separation chamber are connected. The top inlet of the first separation chamber is used to introduce acidic gas into the three-phase separation tank, the bottom outlet of the first separation chamber and the bottom outlet of the second separation chamber are used to discharge the separated solid and liquid phases, and the top outlet of the second separation chamber is used to discharge the separated gas phase. The top inlet and bottom outlet of the first separation chamber, as well as the top outlet and bottom outlet of the second separation chamber, are all made of vertical cylindrical tubes.

6. The acid gas emission management system according to claim 5, characterized in that, The heavy metal removal module further includes: a first liquid level sensor installed at the bottom outlet of the first separation chamber, a second liquid level sensor installed at the bottom outlet of the second separation chamber, and a conveying component installed on the discharge pipe of the three-phase separation tank; the discharge pipe is connected to the bottom outlet of the first separation chamber and the bottom outlet of the second separation chamber respectively, and the discharge pipe is used to send the separated solid phase and liquid phase into the waste liquid incinerator. The first liquid level sensor is used to detect the first liquid level at the bottom outlet of the first separation chamber; The second liquid level sensor is used to detect the second liquid level at the bottom outlet of the second separation chamber; The conveying component is used to send the separated solid and liquid phases into the waste liquid incinerator through the discharge pipe; The control module is communicatively connected to the first liquid level sensor, the second liquid level sensor, and the conveying component, respectively; the control module is also used to adjust the operating parameters of the conveying component based on the first liquid level and the second liquid level, so that the sulfide solid phase is in a flowing state.

7. The acid gas emission management system according to claim 1, characterized in that, The adjustment module includes: a detection component and a calcination flow rate adjustment component that are communicatively connected to the control module; The detection component is used to detect the obtained detection data; the detection data includes the flow rate of the gas phase introduced into the acid gas torch head, the concentration of hydrogen sulfide, the concentration of ammonia, and the concentration and flow rate of the co-burning material; The control module is also used to calculate, based on the detection data, the target flow rate of the co-firing material required to keep the combustion temperature of the acid gas torch head in the second temperature range. The co-firing flow rate regulating component is used to regulate the flow rate of the co-firing material introduced into the acid gas torch head based on the target flow rate of the co-firing material.

8. The acid gas emission management system according to claim 7, characterized in that, The detection component includes a first concentration sensor, a second concentration sensor, and a first flow sensor, which are communicatively connected to the control module and installed on the gas outlet pipe of the three-phase separator; the gas outlet pipe is used to pass the separated gas phase into the acid gas flare head; The first concentration sensor is used to detect the concentration of hydrogen sulfide introduced into the acid gas torch head; The second concentration sensor is used to detect the concentration of ammonia gas introduced into the acid gas torch head; The first flow sensor is used to detect the flow rate of the gas phase introduced into the acid gas torch head.

9. The acid gas emission management system according to claim 8, characterized in that, The co-firing material includes propane gas and vapor, and the detection component further includes a third concentration sensor, a second flow sensor, and a third flow sensor that are communicatively connected to the control module. Both the third concentration sensor and the second flow sensor are installed on the propane gas inlet pipe, which is used to introduce propane gas into the acid gas flare head; the third concentration sensor is used to detect the concentration of propane gas introduced into the acid gas flare head, and the second flow sensor is used to detect the flow rate of propane gas introduced into the acid gas flare head. The third flow sensor is installed on the steam inlet pipe, which is used to introduce steam into the acid gas flare head; the third flow sensor is used to detect the flow rate of the steam introduced into the acid gas flare head.

10. The acid gas emission management system according to claim 9, characterized in that, The first concentration sensor, the second concentration sensor, and the third concentration sensor are all pipeline laser analyzers; the first flow sensor, the second flow sensor, and the third flow sensor are all flow transmitters.

11. The acid gas emission management system according to claim 9, characterized in that, The target flow rate of the co-firing material includes the target flow rate of propane gas and the target flow rate of steam. The co-firing flow rate regulating component includes a propane gas flow control valve and a steam flow control valve that are communicatively connected to the control module. The propane gas flow control valve is installed on the propane gas inlet pipeline; the propane gas flow control valve is used to adjust the flow rate of propane gas introduced into the acid gas torch head based on the target flow rate of propane gas. The steam flow control valve is installed on the steam inlet pipe; the steam flow control valve is used to regulate the flow rate of steam introduced into the acid gas torch head based on the target steam flow rate.

12. The acid gas emission management system according to claim 1, characterized in that, The second temperature range is 650°C-1150°C.

13. The acid gas emission management system according to claim 1, characterized in that, The second temperature range includes a first combustion sub-range for the core reaction zone of the acid gas flare head, and a second combustion sub-range for the main high-temperature zone of the acid gas flare head, the main high-temperature zone surrounding the core reaction zone.

14. The acid gas emission management system according to claim 1, characterized in that, The acid gas emission management system further includes a temperature monitoring module that is communicatively connected to the control module; The temperature monitoring module includes: a first thermocouple, a second thermocouple, a third thermocouple, and a thermal imager, which are communicatively connected to the control module; The first thermocouple is installed in the core reaction zone of the acid gas torch head; the first thermocouple is used to detect and obtain the first temperature; The second thermocouple is installed in the main high-temperature zone of the acid gas flare head; the second thermocouple is used to detect the second temperature; the main high-temperature zone surrounds the core reaction zone; The third thermocouple is installed in the diffusion and cooling zone of the acid gas torch head; the third thermocouple is used to detect the third temperature; the diffusion and cooling zone surrounds the main high-temperature zone; The thermal imager is used to detect the temperature field of the combustion zone of the acid gas torch head; the combustion zone includes the core reaction zone, the main high-temperature zone, and the diffusion and cooling zone; The control module is also used to correct the temperature field based on the first temperature, the second temperature, and the third temperature to obtain a corrected temperature field, and to issue an alarm when the temperature in the corrected temperature field is greater than a preset temperature threshold.

15. A method for managing acid gas emissions, applied to the acid gas emission management system according to any one of claims 1-14, characterized in that, The acid gas emission management method includes: The temperature and pressure of the three-phase separator are controlled within a first temperature range and at a preset pressure, respectively, so that the hydrogen sulfide in the acid gas introduced into the three-phase separator reacts with ammonia in the three-phase separator to generate non-gaseous ammonium hydrosulfide. In addition, the heavy metals in the acid gas react with the sulfides in the system to generate sulfide solid phase. An adjustment command is issued; the adjustment command is used to cause the adjustment module to adjust the combustion temperature of the acid gas torch head to a second temperature range; the second temperature range is the temperature range that suppresses the generation of gaseous sulfur.

Images