Powder storage apparatus

Abstract

To provide a powder storage apparatus and its dust explosion prevention system.SOLUTION: A powder storage apparatus 1 comprises a silo container 100 for storing powder 104, and an injection unit 300 that injects at least one kind selected from the group consisting of an inert gas, ions, and mist into the silo container 100.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a powder storage device and a dust explosion prevention system thereof. [Background technology]

[0002] A conventional silo is disclosed, for example, in Japanese Patent Application Laid-Open No. 2018-185264 (Patent Document 1). Patent Document 1 discloses that the silo includes a suction hose hanging down from the top of the silo, a suction hose length adjustment device that adjusts the length of the suction hose so that the suction port of the suction hose is located near the surface of the stored material based on height information of the stored material, and a gas sensor that detects the concentration of a predetermined component in the gas sucked by the suction hose. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2018-185264 Summary of the Invention [Problem to be solved by the invention]

[0004] However, conventional silos have the problem of potential explosions. [Means for solving the problem]

[0005] The powder storage device includes a silo container for storing powder, and an injection unit for injecting at least one selected from the group consisting of an inert gas, ions, and mist into the silo container.

[0006] The powder storage device configured in this manner can suppress dust explosions within the silo container by injecting an inert gas or the like into the silo container.

[0007] Preferably, the apparatus further comprises a control unit for adjusting the injection amount of at least one selected from the group consisting of the inert gas, ions, and mist, and a storage unit for storing a function required for adjusting the injection amount.

[0008] In this case, the injection amount is adjusted in accordance with the function stored in the storage unit, so that the injection amount can be controlled effectively.

[0009] Preferably, the control unit adjusts the injection amount depending on the type of powder and the total volume of the powder, thereby effectively suppressing dust explosions.

[0010] Preferably, the control unit adjusts the injection amount depending on the type of powder and the moisture content of the powder, thereby effectively suppressing dust explosions.

[0011] Preferably, the control unit adjusts the injection amount depending on the type of powder and the particle size of the powder, thereby effectively suppressing dust explosions.

[0012] Preferably, the control unit adjusts the amount of powder to be injected depending on the type of powder and the humidity at the time of injection, thereby effectively preventing dust explosions.

[0013] Preferably, the control unit adjusts the injection amount in accordance with the explosion limit concentration of the powder. Preferably, the control unit adjusts the injection amount according to the statistical minimum ignition energy of the powder. do.

[0014] Preferably, the control unit adjusts the injection amount depending on the limit oxygen concentration of the powder. A dust explosion prevention system according to the present invention is a dust explosion prevention system for a powder storage device that realizes a neural network using an information processing device, wherein the powder storage device comprises a silo container for storing powder, and an injection unit for injecting at least one kind selected from the group consisting of an inert gas, an ion, and a mist into the silo container, and the dust explosion prevention system for the powder storage device comprises an input layer and an output layer, and input data of the input layer is at least one of a plurality of risk factors related to a powder dust explosion when powder is charged into the silo container, and output data of the output layer is a time in the future from the time of charging the powder. The system includes a neural network that determines the probability of a dust explosion occurring in the silo container, a machine learning unit that trains the neural network using actual values ​​of the input data and the output data as training data, an estimation unit that inputs the input data to the neural network trained by the machine learning unit using the current time as a reference time and calculates a future estimate based on the output data for which the current time is the reference time, and a control unit that determines the amount of injection from the injection unit in accordance with the estimate by the estimation unit, wherein the risk factors are the type of powder, the total volume of the powder, the moisture content of the powder, the particle size of the powder, and the humidity at the time of injection.

[0015] The dust explosion prevention system for a powder storage device configured in this manner uses a neural network to obtain future estimates, making it possible to determine the injection amount with high accuracy. [Brief explanation of the drawings]

[0016] [Figure 1] FIG. 1 is a schematic diagram of a powder storage device according to an embodiment. [Figure 2] FIG. 2 is a detailed view of the powder storage device according to the embodiment. [Figure 3] 1 is a graph showing the relationship between the total volume of powder and the amount of gas or the like injected. [Figure 4] 1 is a graph showing the relationship between the moisture content of powder and the amount of gas or the like injected. [Figure 5] 1 is a graph showing the relationship between the particle size of powder and the amount of gas or the like injected. [Figure 6]10 is a graph showing the relationship between humidity at the time of adding powder and the amount of gas or the like injected. [Figure 7] 1 is a graph showing the relationship between the explosion limit concentration of powder and the amount of gas or the like injected. [Figure 8] 1 is a graph showing the relationship between the statistical minimum ignition energy of powder and the amount of gas or the like injected. [Figure 9] 1 is a graph showing the relationship between the limit oxygen concentration of powder and the amount of gas or the like injected. DETAILED DESCRIPTION OF THE INVENTION

[0017] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. (Device configuration) Fig. 1 is a schematic diagram of a powder storage device according to an embodiment. As shown in Fig. 1, the powder storage device 1 according to the embodiment includes a silo container 100 for storing powder, and an injection unit 300 for injecting at least one selected from the group consisting of an inert gas, ions, and a mist into the silo container.

[0018] The powder storage device 1 includes an inlet unit 200 that injects powder into a silo container 100. The powder storage device 1 also includes a controller 400 that controls the flow rate and injection timing of gas or the like injected from the inlet unit 300, and a memory unit 500 that stores data to be used for calculations by the controller 400.

[0019] The silo container 100 is cylindrical, cylindrical, or other cylindrical shape. The inner diameter does not need to be constant and may vary depending on the height. The volume of the silo container 100 is not particularly limited.

[0020] The powder is stored in the silo container 100. The pellets ( Examples of raw materials (powder) include food products such as wheat, barley, rye, corn, soybeans, adzuki beans, and soybean meal, as well as industrial products such as alumina, coal, tire chips, wood chips, and straw.

[0021] The charging unit 200 is provided at the top of the silo container 100. The charging unit 200 drops the powder into the silo container 100, for example, by gravity. However, instead of dropping by gravity, the powder may be charged into the silo container 100 from the charging unit 200 by, for example, a screw. When using power such as a screw, the charging unit 200 does not necessarily have to be provided at the top of the silo container 100, and may be provided at the side or bottom of the silo container 100.

[0022] The injection unit 300 is a device for injecting gas, ions, mist, etc. into the silo container 100. The injection unit 300 may include a nozzle provided inside the silo container 100. The injection unit 300 injects an inert gas such as nitrogen or carbon dioxide into the silo container 100 to suppress dust explosions inside the silo container 100.

[0023] If plants such as fir are present in the silo container 100, the fir will breathe even during storage. This consumes oxygen and produces carbon dioxide. As a result, the air inside the silo container 100 becomes thin, resulting in an oxygen deficiency. When entering the silo container 100 for inspection or other purposes, the oxygen concentration is measured with an oxygen concentration meter. A concentration of 18% or higher is considered safe, but anything below this level could lead to anoxia. To prevent this, oxygen may be supplied from the injection section 300 to maintain the oxygen concentration inside the silo container 100 at 18% or higher. In other words, the injection section 300 may inject not only inert gas to prevent dust explosions, but also oxygen gas or air to allow workers to work inside the silo container 100.

[0024] Ions may be injected into the silo 100 from the injection unit 300. Injecting ions can neutralize the charge on each powder, reducing the risk of dust explosions. A device that generates such ions can be the SJ-E series hybrid ultrafast sensing ionizer manufactured by Keyence Corporation. Injecting ions does not change the gas composition inside the silo 100 as compared to injecting gas as described above, making it ideal for storing powder (such as unhulled rice) whose gas composition needs to be controlled inside the silo 100. Furthermore, both gas and ions may be injected from the injection unit 300.

[0025] Mist may be injected from the injection unit 300. Mist can be produced, for example, by a product called "Mecha Swing Nozzle" series or "Water Mist" manufactured by Gallue Co., Ltd. Mist is preferably used for powders that do not cause problems such as corrosion even when moisture adheres to them.

[0026] The control unit 400 is a device for controlling the flow rate, flow velocity, and injection timing of the gas, ions, mist, etc. injected from the injection unit 300. The control unit 400 includes, for example, a computer.

[0027] The storage unit 500 stores tables used for control by the control unit 400. The storage unit 500 may be a hard disk of a computer that constitutes the control unit 400. The storage unit 500 may be a recording medium that can be detached from the computer. Examples of recording media include DVD-RAM, DVD-ROM, CD-ROM, FD, Examples of media that permanently carry a program include hard disks, magnetic tapes, cassette tapes, optical disks, and semiconductor memories such as EEPROMs and flash ROMs. Recording media are also non-transient media that allow a computer to read the program. The term "program" as used here refers not only to programs that can be directly executed by a CPU, but also to programs in source program format, compressed programs, and encrypted programs. etc.

[0028] Fig. 2 is a detailed view of a powder storage device according to an embodiment. The powder storage device 1 shown in Fig. 2 inerts the inside of a silo container 100. The silo container 100 has tapered inlet 110 and outlet 120. The silo container 100 is cylindrical, and powder 104 is stored therein.

[0029] A plurality of temperature sensors 102 are provided on the outer peripheral surface of the silo container 100. A temperature sensor 105 is also provided in the center of the silo container 100. The temperature sensors 102 and 105 measure the temperature inside the silo container 100, and when the temperature of the silo container 100 reaches or exceeds a predetermined value, the inside of the silo container 100 is cooled. Note that the temperature sensor 102 is not necessarily provided.

[0030] An input section 200 is provided above the silo container 100. Powder is input from the input section 200 to the inlet 110. The input speed of the powder from the input section 200 to the inlet 110 is adjustable.

[0031] The powder 104 is discharged from the outlet 120 of the silo container 100 . An injection section 300 is provided on the peripheral surface of the silo container 100. The injection section 300 has a pipe 107 and a nozzle 103 attached to the pipe 107. The pipe 107 is connected to a control section 400. The flow of gas and the like in the pipe 107 is controlled by the control section 400.

[0032] The control unit 400 includes, for example, a valve and a computer that controls the valve. At least one of a high-pressure vessel group 610, a CE (cold evaporator) tank device 620, and a PSA (pressure swing adsorption) device 630 is connected to the control unit 400. There are.

[0033] The high-pressure container group 610 is made up of a plurality of cylinders 611. The plurality of cylinders 611 are filled with, for example, nitrogen gas. The nitrogen gas is supplied from the cylinders 611 when the nitrogen gas is injected into the silo container 100 from the nozzle 103. The control unit 400 and the high-pressure container group 610 are connected by a pipe 619.

[0034] The CE tank device 620 has a tank 621 that stores liquid nitrogen, and a regulator 622 that receives liquid nitrogen from the tank and vaporizes the liquid nitrogen. The control unit 400 and the CE tank device 620 are connected by a pipe 629.

[0035] The PSA unit 630 is a pressure swing adsorption unit. By utilizing the differences in the adsorption characteristics of the adsorbent with respect to gas, the target gas (nitrogen) is continuously separated by alternately repeating pressurization and depressurization operations. For example, "Bellfine Activated Carbon," a high-performance MSC (molecular sieve carbon) manufactured by Air Water Inc., can be used as the adsorbent. The control unit 400 and the PSA unit 630 are connected by a pipe 639.

[0036] The device that supplies nitrogen gas to the control unit 400 may be a high-purity nitrogen gas generator "V1" (product name) manufactured by Air Water Inc. This device stably generates high-purity nitrogen gas through heat exchange using the cold energy of liquefied nitrogen. In other words, a large amount of heat is lost when liquid nitrogen evaporates. This heat is used to cool the air, liquefying oxygen, carbon dioxide, etc. in the air, and the remaining gaseous nitrogen can be used for a specified purpose. This allows for a stable supply of nitrogen gas at low cost.

[0037] For a dust explosion to occur, there must be combustible dust and an ignition source. The three requirements are necessary: ​​1) The presence of oxygen necessary for explosion. In this invention, the risk of dust explosion is reduced by injecting nitrogen gas, etc. This not only reduces oxygen, but also contributes to reducing the flammability of combustible dust and preventing the generation of ignition sources (static electricity).

[0038] FIG. 3 is a graph showing the relationship between the total volume of powder and the amount of gas or other material injected. As shown in FIG. 3, there is a correlation between the total volume of powder and the risk of a dust explosion, with the risk of a dust explosion increasing as the total volume of powder increases. This is presumably because, as the total volume of powder increases, the number of times the powder particles come into contact with each other when being poured into the silo container 100 increases, generating static electricity between the powder particles, which can trigger a dust explosion. Furthermore, the risk of a dust explosion is higher when the powder is wood pellets than when the powder is grain. This is presumably because wood pellets have a lower moisture content than grains. Data related to the graph in FIG. 3 is stored in the memory unit 500.

[0039] Data on the total volume of the powder can be measured when the powder transported by ship or the like is unloaded. The data on the total volume is input, for example, to memory unit 500. The amount of gas or the like to be injected into silo container 100 is determined based on the input data and the graph of FIG. 3. Then, when the powder is charged from charging unit 200 to inlet 110, gas or the like is injected based on the determined amount.

[0040] Figure 4 is a graph showing the relationship between the moisture content of powder and the amount of gas or other injected material. As shown in Figure 4, there is a correlation between the moisture content of powder and the risk of dust explosions; as the moisture content of powder decreases, the energy required to burn the powder (activation energy) decreases, increasing the risk of dust explosions. This is presumably because as the moisture content of powder decreases, the powder becomes more likely to burn within the silo container 100. Data related to the graph in Figure 4 is stored in the memory unit 500.

[0041] Data on the moisture content of the powder is obtained by extracting a portion of the powder before it is charged into the silo container 100 and measuring its moisture content. The data on the moisture content is input, for example, into the memory unit 500. The amount of gas or other agent to be injected into the silo container 100 is determined based on the input data and the graph in FIG. 4. Then, when the powder is charged from the charging unit 200 to the inlet 110, the gas or other agent is injected based on the determined injection amount.

[0042] Figure 5 is a graph showing the relationship between powder particle size and the amount of gas or other substance injected. As shown in Figure 5, there is a correlation between powder particle size and the risk of a dust explosion, with the risk of a dust explosion increasing as the powder particle size decreases. This is presumably because as the powder particle size decreases, the area of ​​contact between the powder and oxygen increases, making the powder more likely to burn within the silo container 100. Data related to the graph in Figure 5 is stored in the memory unit 500.

[0043] Data on the particle size of the powder can be obtained by extracting a portion of the powder before it is charged into the silo container 100 and measuring its particle size. The particle size can be measured, for example, based on JIS-Z-8825:2013. Furthermore, the powder before it is charged into the silo container can be photographed and the image analyzed to determine the particle size of each particle. The data on particle size is input, for example, into the memory unit 500. The amount of gas or other agent to be injected into the silo container 100 is determined based on the input data and the graph in FIG. 5. Then, when the powder is charged from the charging unit 200 to the inlet 110, the gas or other agent is injected based on the determined injection amount.

[0044] Figure 6 is a graph showing the relationship between humidity when powder is charged and the amount of gas or the like charged. As shown in Figure 6, there is a correlation between humidity when powder is charged and the risk of dust explosion, and the risk of dust explosion increases when the humidity when powder is charged is low. This is because when the humidity is low, moisture does not easily adhere to the surface of the powder, so the energy required to burn the powder is reduced, and the powder burns less easily in the silo container 100. The data relating to the graph of FIG.

[0045] Data regarding the humidity at the time of powder introduction can be obtained from the humidity announced by the Japan Meteorological Agency at the time before introduction into the silo container 100. Furthermore, a hygrometer may be provided inside the silo container 100 and used to measure the humidity.

[0046] FIG. 7 is a graph showing the relationship between the explosive limit concentration of a powder and the amount of gas or other substance injected. As shown in FIG. 7, there is a correlation between the explosive limit concentration of a powder and the risk of a dust explosion, and as the explosive limit concentration of a powder decreases, the risk of a dust explosion increases. This is because a low explosive limit concentration indicates that the powder is prone to explosion at low concentrations. The explosive limit concentration is determined primarily based on the type of powder, taking into account factors such as particle size and moisture content. The explosive limit concentration may be determined solely based on the type of powder transported to the feed section 200, or it may be determined by taking into account the type and particle size of the powder. As shown in FIG. 7, the explosive limit concentration of wood pellets is approximately one-quarter of the explosive limit concentration of grains, indicating that wood pellets are prone to explosion. Data related to the graph in FIG. 7 is stored in the memory unit 500.

[0047] FIG. 8 is a graph showing the relationship between the statistical minimum ignition energy of a powder and the amount of gas or other material injected. As shown in FIG. 8, there is a correlation between the minimum ignition energy of a powder and the risk of a dust explosion. As the minimum ignition energy of a powder decreases, the risk of a dust explosion increases. This is because a smaller minimum ignition energy indicates that a dust explosion is more likely to occur with less energy. The minimum ignition energy is determined primarily based on the type of powder, taking into account factors such as particle size and moisture content. The minimum ignition energy may be determined solely based on the type of powder transported to the feed section 200, or it may be determined by considering the type and particle size of the powder. As shown in FIG. 8, the minimum ignition energy of wood pellets is approximately 1 / 64 of the minimum ignition energy of grains, indicating that wood pellets are more likely to explode. Data related to the graph in FIG. 8 is stored in the memory unit 500.

[0048] FIG. 9 is a graph showing the relationship between the critical oxygen concentration of a powder and the amount of gas or other substance injected. As shown in FIG. 9, there is a correlation between the critical oxygen concentration of a powder and the risk of a dust explosion, and as the critical oxygen concentration of a powder decreases, the risk of a dust explosion increases. This is because a low critical oxygen concentration indicates that a dust explosion can occur at a low oxygen concentration. The critical oxygen concentration is determined primarily based on the type of powder, taking into account factors such as particle size and moisture content. The critical oxygen concentration may be determined based solely on the type of powder transported to the feed section 200, or it may also be determined taking into account the type and particle size of the powder. Data related to the graph of FIG. 9 is stored in the memory unit 500.

[0049] The dust explosion prevention system for a powder storage device 1 is a dust explosion prevention system for a powder storage device 1 that realizes a neural network using an information processing device, and the powder storage device 1 comprises a silo container 100 that stores powder, and an injection unit 300 that injects at least one selected from the group consisting of an inert gas, an ion, and a mist into the silo container 100, and the dust explosion prevention system for the powder storage device 1 comprises an input layer and an output layer, and input data for the input layer is at least one of a plurality of risk factors related to a powder dust explosion when powder is charged into the silo container 100, and output data for the output layer is a machine learning unit that trains the neural network using actual values ​​of the input data and the output data as teacher data; an estimation unit that inputs the input data to the neural network trained by the machine learning unit using the current time as a reference time and obtains an estimated value for the future based on the output data for which the current time is the reference time; and a control unit that determines the amount of injection from the injection unit in accordance with the estimated value of the estimation unit, and the risk factors include the type of powder, the total volume of the powder, the amount of powder to be injected, and the probability of a dust explosion occurring in the silo container in the future from the time the powder is injected. These are the moisture content of the powder, the particle size of the powder, the humidity at the time of adding the powder, the explosive limit concentration of the powder, the statistical minimum ignition energy of the powder, and the limit oxygen concentration of the powder.

[0050] Input and output data for machine learning are obtained through simulation. Data is input into a commercially available simulator while varying the values ​​of various risk factors, and the values ​​output by the simulation are used as actual values ​​to create training data.

[0051] The embodiments disclosed herein are illustrative in all respects and should not be considered limiting. The scope of the present invention is defined by the claims rather than the above-described embodiments, and it is intended to include any modifications within the scope of the claims and meanings equivalent to the claims. [Explanation of symbols]

[0052] 1 Powder storage device, 100 Silo container, 102, 105 Temperature sensor, 103 Nozzle, 104 Powder, 107, 619, 629, 639 Piping, 110 Inlet, 120 Outlet, 200 Feeding section, 300 Injection section, 400 Control section, 500 Memory section, 610 High-pressure vessel group, 611 Cylinder, 620 CE tank device, 621 Tank, 622 Regulator, 630 PSA device.

Claims

1. a silo container for storing powder; and an injection section for injecting at least one selected from the group consisting of an inert gas, an ion, and a mist into the silo container.

2. 2. The powder storage device according to claim 1, further comprising: a control unit that adjusts the injection amount of at least one selected from the group consisting of the inert gas, ions, and mist; and a memory unit that stores a function required for adjusting the injection amount.

3. The powder storage device according to claim 2 , wherein the control unit adjusts the injection amount depending on the type of powder and the total volume of the powder.

4. The powder storage device according to claim 2 or 3, wherein the control unit adjusts the injection amount depending on the type of powder and the moisture content of the powder.

5. The powder storage device according to claim 2 , wherein the control unit adjusts the injection amount depending on the type of powder and the particle size of the powder.

6. The powder storage device according to claim 2 , wherein the control unit adjusts the amount of powder to be poured depending on the type of powder and humidity at the time of pouring.

7. The powder storage device according to claim 2 , wherein the control unit adjusts the injection amount depending on the explosion limit concentration of the powder.

8. The powder storage device according to claim 2 , wherein the control unit adjusts the injection amount in accordance with a statistical minimum ignition energy of the powder.

9. The powder storage device according to claim 2 , wherein the control unit adjusts the injection amount depending on the limit oxygen concentration of the powder.

10. A dust explosion prevention system for a powder storage device that realizes a neural network using an information processing device, The powder storage device is a silo container for storing powder; an injection unit that injects at least one selected from the group consisting of an inert gas, an ion, and a mist into the silo container, The dust explosion prevention system for the powder storage device comprises an input layer and an output layer, and the input data of the input layer is at least one of a plurality of risk factors related to a powder dust explosion when powder is poured into the silo container, and the output data of the output layer is the probability of a dust explosion occurring in the silo container in the future from the time the powder is poured into the silo container; a machine learning unit that trains the neural network using actual values ​​of the input data and the output data as training data; an estimation unit that inputs the input data to the neural network trained by the machine learning unit using the current time as a reference time, and obtains an estimated value for the future based on the output data for which the current time is the reference time; and a control unit that determines the injection amount from the injection unit in accordance with the estimated value of the estimation unit. The risk factors are the type of powder, the total volume of the powder, the moisture content of the powder, the particle size of the powder, and the humidity at the time of adding the powder.

Images