Application and Analysis of Microbial Spray Filtering in Environmental Exhaust Gas Treatment

Green grain storage methods have proliferated recently, including low-temperature storage [1], [2], controlled atmosphere storage [3], inert powder storage [4], [5], and recirculating fumigation storage [6], [7]. The quality of grain storage is superior thanks to this cutting-edge technology, but the expense to improve quality is higher. The majority of grain depots in China use traditional techniques of storage and chemical fumigation to eradicate insects [8]. At grain depots, chemical fumigation residue is a significant source of environmental exhaust gases that contain phosphide. A significant number of phosphides, a characteristic environmental exhaust gas, will be produced even during the strictest fumigation [9]. Sulfides and nitrides, two other sources of grain store exhaust gases, are produced by mildews when grains are exposed to air. These exhaust fumes will contaminate the environment and endanger people if they are released into the air directly.

Dry methods and wet methods are the two main categories for the many technical techniques used for domestic and international grain store exhaust gas degradation. Grain store exhaust gases can be immediately burned [10] by dry techniques or degraded [11] by using solid adsorbents or oxidants [12], [13]. Examples of such practices are the combustion method and the adsorption approach. The dry methods harm the environment despite their straightforward procedure and low cost. The liquid-phase oxidation reaction between exhaust gases and solutions containing oxidants or catalysts is primarily carried out using wet techniques. The plasma method [14], [15] and the biological method [16] are examples of wet methods. The plasma process has a broad variety of applications and a high rate of purification, but it also needs expensive equipment and uses a lot of energy. The biological approach uses straightforward, low-maintenance treatment equipment and has a high clearance efficiency. Additionally, it can prevent secondary contamination. The characteristics and applicability range of each exhaust gas removal technology are listed in Table 1.

The biological technique for exhaust gas treatment outperforms other conventional methods by a wide margin, as demonstrated in Table 1. Utilizing low-cost, straightforward, simple, easy-to-operate, manageable, and maintainable devices, the method can achieve effective treatment without any secondary contamination [17]. Microbial strains, meanwhile, have a lot of benefits. They can, for instance, adapt to a variety of harsh settings and effectively treat exhaust gases. Almost every substance can be broken down by microorganisms. Microbes may grow on a variety of substrates and are both easily produced and self-replicating. Numerous exhaust gases can be treated with a relatively tiny number of microbial strains.

The theory behind the microbial treatment of grain storage exhaust gases is explained in this study. Microbial solution is first injected into the exhaust gas flow channel, followed by COMSOL Multiphysics simulation and optimization. The purpose is to optimize the exhaust gas treatment device, and improve the treatment efficiency.

Under the coexistence of water, microbes, and oxygen, exhaust gasses can be purified through the microbial metabolism, which oxides and decomposes the phosphides, sulfides, and nitrides in the gases. Microbial treatment can be roughly divided into the following three steps:

Step 1. The phosphides, sulfides, and nitrides are effectively exposed to the circulating biological bacteria liquid on the surface of the filler carrier, and fully dissolved in water.

Step 2. The microbes spread to the surface of the phosphides, sulfides, and nitrides dissolved in water, and absorb these substances.

Step 3. The microbes oxide and decompose the phosphides, sulfides, and nitrides. The phosphides are decomposed into PO₄^-₃, the sulfides into S, S₂O_3^2-, and SO_4^2-, and the nitrides into NO₂^-, NH₄^-, and NO₃^-.

The specific mechanism is as follows:

Phosphides; Burkholderia cepacia (aerobic or anaerobic) [18]

Corynebacteria

Sulfides

Bacillus / Thiobacillus (aerobic or anaerobic) [19]

Photosynthetic bacteria (anaerobic, CO₂, and illumination)

Bacillus; Lactic acid bacteria; Saccharomycetes [20], [21]

Nitrides

Pseudomonas, etc.

According to the exhaust gas features, the appropriate working pH of the microbial solution is 4.0 ~ 6.0. Exhaust gases often contain acidic substances. When an exhaust gas in contact with microbial solution directly, the pH value of the solution will be reduced by these substances, which suppresses the efficiency of microbial treatment of the exhaust gas. For this reason, the acidic substances in the exhaust gas must be removed before the gas is in contact with the microbial solution. Then, the pH of the gas will be neutral, when it is exposed to the solution.

Figure 1 shows the principle of the exhaust gas treatment structure. The exhaust gas enters the pre-processing area for deacidification, such that the exhaust gas reaches the neutral state. Then, the near-neutral exhaust gas will reach the treatment area. To mix the microbial solution evenly with the exhaust gas, spray filtering will be adopted to spray the microbial solution into the exhaust gas pipe, using high-speed nozzles. It is an urgent problem to scientifically design the diameter of the exhaust gas pipe, and arrange the nozzles on that pipe.

Based on COMSOL Multiphysics, a segment of exhaust gas treatment pipe was taken for numerical calculation to obtain the optimal pipe diameter and optimal nozzle distribution. The optimization results guide the design of pipe diameter and nozzle arrangement in actual production.

Compared with other simulation software, COMSOL Multiphysics, a simulation software package for multiple physical fields, has high degrees of freedom, and a convenient function input mechanism. The first half of the pipe where the exhaust gas first enters the mixing area was segmented for simulation. In this segment, the gas has just left from the pre-processing area, and features a high pressure and a fast speed. We simulated the nozzle distribution and the mixture between exhaust gas and microbial solution in this segment, which is more representative than other areas.

Given the symmetry of nozzles on the pipe, the pipe and nozzles were illustrated with a simple two-dimensional (2D) model to reduce the computing load. Figure 2 shows the situation of five nozzles arranged at an interval of 400mm. The factory mainly provides 300mm and 450mm pipe diameter; The initial distance of the nozzle cannot be too far away from the exhaust inlet, so 190mm is selected as the initial value; The number and spacing of nozzles should be considered in combination with pipe length and nozzle cost. The pipe length is 2000mm, and other parameters are parameterized, as shown in Table 2.

Because the entire exhaust gas pipe is too long, the first half of the pipe where the exhaust gas first enters the mixing area was segmented for simulation. The inlet and outlet pressures of the segment were 1,100Pa and 1,000Pa, respectively. Considering the small pressure difference between the inlet a12314nd outlet, the exhaust gas in the segment was treated as incompressible. Given their complex compositions, the exhaust gas and microbial solution were respectively simplified as the air and water solution. The periodic velocity inlet was selected for nozzles. The velocity period is shown in Figure 3. The k- turbulence model of physical fields was couped with the horizontal set to simulate the exhaust gas transport, and the distribution of the microbial solution ejected by the nozzles. The mean density and mean viscosity were measured by the mean harmonic volume, and the mean harmonic mass, respectively. The time step was set to 0.05s, and the total time to 2s. The simulation considers the impact of gravity and surface tension on the results.

The simulations aim to disclose the impacts of two pipe diameters (300mm, and 450mm), four initial distances of nozzle (190mm, 290mm, 390mm, and 490mm), three different number of nozzles (3, 4, and 5), and four nozzle intervals (200mm, 300mm, 400mm, and 500mm).

Subgraph (a) and (b) of Figure 4 are the cloud maps for the velocity and microbial solution distribution at 0.4s, respectively, when the pipe diameter was 300mm, the initial distance of nozzle was 290mm, the number of nozzles was 4, and the nozzle interval was 400mm. Subgraph (c) and (d) of Figure 4 are the cloud maps for the velocity and microbial solution distribution at 0.4s, respectively, when the pipe diameter was 450mm, while all the other parameters remained the same.

Comparing subgraph (c) of Figure 4 (pipe diameter: 450mm) with subgraph (a) of Figure 4 (pipe diameter: 300mm), it was observed that the velocity difference between the upper and lower layers in subgraph (c) of Figure 4 was more obvious than that in subgraph (a) of Figure 4. The velocity near the nozzles was much slower than that at the bottom of the pipe. Combined with the microbial solution distribution in subgraph (a) and (d) of Figure 4, it can be derived that the large difference between fast and low velocities disrupt the distribution of microbial solution.

Figure 5 shows the volume fractions of the microbial solution with the two pipe diameters at different moments. It can be observed that, under the periodic velocity inlet condition, the volume fraction of the microbial solution in the pipe slightly declined, after each velocity period. The decline was relatively small in the pipe with the smaller diameter. Thus, the small pipe diameter (300mm) is more favorable than the large pipe diameter (450mm).

Figure 6 shows the simulation results at the pipe diameter of 300mm, with different initial distances of nozzles. The right subgraph shows the velocity distribution in the pipe at 0.4s, while the left subgraph shows the microbial solution distribution at 0.4s. The initial distance a1 of the first nozzle was 190mm. Since the nozzle interval was 100mm, the initial distances b₁, c₁, and d₁ of the second, third, and fourth nozzles were 290mm, 390mm, and 490mm, respectively.

As shown in Figure 6, the closer the nozzle was to the exhaust gas inlet, the faster the velocity in the pipe, and the greater the difference between velocities in the upper and lower layers. Due to the velocity difference between the two layers, the microbial solution mainly concentrates in the upper part of the pipe. As a result, the microbial solution cannot fully mix with the exhaust gas, which suppresses the exhaust gas treatment effect.

When the nozzle was far from the exhaust gas inlet, e.g., d₁, the velocity difference between layers was greatly reduced, but the maximum velocity appeared at the middle of the pipe. Then, only the microbial solution from the first nozzle could easily reach the pipe bottom, while that from the other nozzles tend to concentrate in the upper part of the pipe.

When the nozzle was of moderate distance from the exhaust gas inlet, e.g., b₁ and c₁, the velocity difference between layers was much less significant than that of a₁, as evidenced by the velocity cloud maps. The cloud maps of microbial solution distribution also demonstrate that b₁ and c₁ are superior to d₁.

Figure 7 shows the volume fractions of the microbial solution with different initial distances of nozzle at different moments. It can be observed that, the volume fraction of the microbial solution in the pipe slightly declined, after each velocity period. When the initial distance of nozzle was 190mm, the volume fraction of the microbial solution was the highest initially, and the fastest in decline. When that distance was 490mm, the volume fraction of the microbial solution was the lowest initially, and the slowest in decline. Thus, neither distance is the optimal. There was an insignificant difference between the initial distances of 290mm and 390mm. With these two distances, the volume fractions of the microbial solution fell between those of the distance of 190mm and that of 490mm. Judging by the extremes of the volume fractions in Figure 7, the peak volume fraction at the distance of 290mm was higher than that at 390mm. Overall, the optimal initial distance of nozzle is 290mm.

Figure 8 shows the simulation results with the pipe diameter of 300mm, the initial distance of nozzle of 290mm, nozzle interval of 400mm, and different number of nozzles. The right subgraph shows the velocity distribution in the pipe at 0.4s, while the left subgraph shows the microbial solution distribution at 0.4s. The number of nozzles a₁ was 3, b₁ was 4, and c₁ was 5.

As shown in Figure 8, with the growing number of nozzles, the velocity in the exhaust gas pipe became increasingly ununiform, and the microbial solution did not concentrate in the upper part of the pipe. By integrating the microbial solution in the pipe, the areas of microbial solution under a₁, b₁, and c₁ were determined as 0.0542 m², 0.0573 m², and 0.066 m², respectively. The 5 nozzle condition greatly improved the area from the 3 and 4 nozzle conditions. There was no significant difference between 3 and 4 nozzle conditions.

Figure 9 shows the volume fractions of the microbial solution with different number of nozzles at different moments. It can be observed that, the volume fraction of the microbial solution in the pipe slightly declined, after each velocity period. However, with the growing number of nozzles, the decline became smaller and smaller. Thus, more nozzles led to better performance. Five is clearly the optimal number of nozzles.

Figure 10 shows the simulation results with the pipe diameter of 300mm, the initial distance of nozzle of 290mm, number of nozzles of 5, and different nozzle intervals. The right subgraph shows the velocity distribution in the pipe at 0.4s, while the left subgraph shows the microbial solution distribution at 0.4s. The nozzle distance a₁ was 200mm, b₁ was 300mm, and c₁ was 400mm.

As shown in Figure 10, an obvious velocity difference was observed between the upper and lower parts of the pipe with a₁, b₁, and c₁. In the case of a₁, the peak velocity appeared at the back of the pipe; In the case of b₁, the peak velocity appeared at the middle of the pipe; In the case of c₁, the peak velocity appeared near the outlet.

According to the microbial solution distributions a₂, b₂, and c₂, the microbial solution was mixed well with the exhaust gas, always at the back of the pipe. By integrating the microbial solution in the pipe, the areas of microbial solution under a₂, b₂, and c₂ were determined as 0.0898 m², 0.0802 m², and 0.066 m², respectively. Therefore, the best mixing between microbial solution and exhaust gas was realized at the nozzle interval of 200mm, followed by 300mm. The poorest mixing effect was observed at the nozzle interval of 400mm.

Figure 11 shows the volume fractions of the microbial solution with different nozzle intervals at different moments. It can be observed that, the volume fraction of the microbial solution in the pipe slightly declined, after each velocity period. When the nozzle interval was 200mm, the volume fraction decreased slowly, and remained at the highest possible level in all scenarios. Therefore, the smaller the nozzle interval, the better the mixing between the microbial solution and the exhaust gas. In other words, the nozzle interval of 200mm is better than 300mm and 400mm.

(1) Compared with the traditional way of exhaust gas treatment, microbial treatment of waste gas has obvious advantages, remarkable effect, no secondary pollution.

(2) Through numerical simulation, the influences of pipe diameter, initial distance of nozzles, number of nozzles, spacing of nozzles on the volume fraction of microbial solution in the pipe were analyzed, and the optimal distribution of nozzles was obtained. That is, when the diameter of the pipe is 300mm, the initial distance between the nozzles is 290mm, the number of nozzles is 5, and the distance between the nozzles is 200mm, the mixture of microbial solution and waste gas is the best, and the waste gas treatment effect is the best.