THE FLOW PATTERNS OF BUBBLE PLUME IN AN MBBR*

LI Shi-rong

Key Laboratory of Northwest Water Resources and Environmental Ecology of Ministry of Education, Xi’an University of Technology, Xi’an 710048, China

Oil Production Plant of No. 4 Changqing Oilfield Company, China National Petroleum Corporation，Jing Bian 718500, China, E-mail: lsr_cq@petrochina.com.cn

CHENG Wen, WANG Meng, CHEN Chen

Key Laboratory of Northwest Water Resources and Environmental Ecology of Ministry of Education, Xi’an University of Technology, Xi’an 710048, China

THE FLOW PATTERNS OF BUBBLE PLUME IN AN MBBR*

LI Shi-rong

Key Laboratory of Northwest Water Resources and Environmental Ecology of Ministry of Education, Xi’an University of Technology, Xi’an 710048, China

Oil Production Plant of No. 4 Changqing Oilfield Company, China National Petroleum Corporation，Jing Bian 718500, China, E-mail: lsr_cq@petrochina.com.cn

CHENG Wen, WANG Meng, CHEN Chen

Key Laboratory of Northwest Water Resources and Environmental Ecology of Ministry of Education, Xi’an University of Technology, Xi’an 710048, China

The flow patterns of the gas-liquid two-phase flow in a Moving-Bed Biofilm Reactor (MBBR) have a critical effect upon the mass transfer by the convection. Bubble plumes promote unsteadily fluctuating two-phase flows during the aeration. This article studies the unsteady structure of bubble plumes through experiments. The time-serial bubble plume images in various cases of the tank are analyzed. The Recursive Cross Correlation-Particle Image Velocimetry (RCC-PIV) is used to calculate the velocities in those cases, and then the time-serial vortex, the total turbulence intensity, the time-serial streamline are obtained. It is shown that the aspect ratio and the void fraction are the dominant factors influencing the unsteady structure of bubble plumes. When the aspect ratio is unity and the void fraction is high, the bubble plumes see a symmetrical vortex structure with a long residence time, which is beneficial for optimizing the aeration system and enhancing the applied range of bubble plumes.

bubble plume, Recursive Cross Correlation (RCC), Particle Image Velocimetry (PIV), flow pattern

Introduction

Bubble plumes have a great application value in projects, such as alleviating the damage of waves to building structures, preventing the invasion of the brine with air bubble curtains in the estuary, controlling the stratification structure of reservoirs and lakes to improve water quality, preventing channels and harbors from being frozen, enhancing the oxygen content for aquatic growth and so on. In view of serious pollutions caused, especially in recent years, by the oil-gas blowouts, with the large-scale exploitation of oil and gas under the sea, bubble plumes can used to control the pollution area. A great number ofstudies of simulations of bubble plumes have widened the application range[1]. The bubble plume becomes now a key issue in the current field of fluid mechanics.

The Moving-Bed Biofilm Reactor (MBBR) is a new biological wastewater treatment technology with many desirable features, such as efficient operation and low energy consumption, which takes both advantages of the traditional fluidized bed and the biological contact oxidation process with the combination of the activated sludge process and the biological film process. The suspension filler latched by the bacterial biofilm is turned into a fluidized state by the aeration and the flow promotion in the device. Therefore, the selection of a suitable aeration mode is of great importance for the stable operation of MBBR. The aeration plays its role in two aspects: one is to supply the microbiological oxidation with oxygen, another is to enhance the turbulent intensity of fluid, and they are both important for the efficiency of the wastewater treatment.

The aeration is a usual engineering measure,while the bubble plume is a complex two-phase fluid generated in the aerating process[2,3]. The optimization of aerating equipments has attracted many studies, however, some important factors as the flow pattern of the gas-liquid two-phase flow, the mass transfer between the gas phase and the liquid phase, and the mechanism of the oxygen transfer have not been paid enough attentions. The bubble plume structure is an important factor related with the aerating efficiency and the application range of the bubble plume[4]. With this in mind, experiments were carried out in this study, and the Particle Image Velocimetry (PIV), as a new measurement technique for the flow field, was employed. Some bubble plume images in various conditions were obtained through CCD cameras. The digital image is turned into a white-back graph by binaryzation. The Recursive Cross Correlation (RCC) algorithm is used to obtain the whole flow field from these images. In this article, Visual Fortran programs are developed for image preprocessing and post-processing to remove noise and reduce image distortions. The software DeltaGraph is also used to draw figures[5,6].

Fig.1 Recursive cross correlation PIV for measuring bubble velocity vectors

1. Methods to extract flow field information

The extraction of flow field information by PIV usually contains the following three steps: (1) digital image pre-processing, (2) algorithms for PIV measurements, (3) data post-processing[7].

The digital images are treated by Otsu’s method[8], which is an image segmentation method based on Density Histogram (DH) technique. In this step, the optimal threshold value can be obtained to turn the gray level in each image into 0 and 1. Then, the RCC-PIV measurement algorithm based on the gray level distribution is employed to calculate the velocity of the flow field. Combining this velocity with images of 0-1, the bubble’s flow rate in each position can be obtained. Finally, other physical parameters describing the flow field such as the time-serial vortex, the total turbulence intensity, the time-serial streamline are obtained.

The computational process of RCC-PIV is shown in Fig.1.

Fig.2 Experimental set-up

Initially, a sparse velocity distribution is obtained in a 32×32 pixel interrogation area. The result is used to direct the next search. The second process is carried out in a quarter-downsized area. Iterating this process several times, the final data are extracted at a very fine spatial resolution[9]. The cross correlation is defined as

Fig.3 The time-dependent velocity field in the same void fraction

Fig.4 Time-serial flow field with void fraction of 16% and different aspect ratios: time-averaged velocity and vorticity

wherefandgare the brightness, the indexesiandjare digitalized image coordinates, andMandNare the size of the interrogation area.

2. Experimental device and conditions

The experimental set-up is shown in Fig.2. A planer bubble plume is generated in a thin rectangular tank of 300 mm in width, 800 mm in height and 40 mm in horizontal depth. Air bubbles are injected into the bottom of the tank through three rows of capillary tubes, by which the bubble size and the number density can be precisely controlled. The purpose of the experiments is to examine the gas-liquid two-phase flow characteristics, and the whole experimental equipment is composed of two parts: the experimental device and the measurement device.

The behavior of bubble plumes is visualized and recorded for more than 50 cases by changing the gas flow rate and the water depth in experiments. The bubble plumes are recorded as digital images at a frame rate of 30 fps in total. 1 200 consecutive images are processed to calculate the time related statistic quantities of the bubble velocity fields. Preprocessing of digital images is carried out to remove noise and reduce image distortions, as mentioned before. After experiments and calculations, some results are obtained. The aspect ratio and the void fraction are fixed for cases to analyze the behavior and the structure of the bubble plume. The time-dependent bubble distribution of the plane bubble plume and the time-serial bubble plume images in different conditions are analyzed.

In this study, the running water is regarded as the liquid phase, and its kinematic viscosity coefficient is 10–6m2/s and its density is 1 000 kg/m3, while the gas phase is air and its density is 1.28 kg/m3. The experimental temperature is 13oC-15oC, andQstands for the gas flow rate,H/Wis the aspect ratio, andRis the average bubble radius in the tank measured by image processing. The initial height of the water surface without bubbles is 0.3 m.

3. Results and analyses

3.1Analysis of flow field in the same void fraction

The void fraction is defined as the percentage of total bubbles in the volume and is an important factor to influence the bubble plume movement. Based on the experimental conditions, the flow fields of bubble plumes of three different aspect ratios and the same void fraction of 16% are analyzed in this study. Table 1 shows the experimental conditions.

Table1 Experimental conditions (α=16%)

When the void fraction is increased up to 16%, the velocity of bubbles increases, and the movement in each cases is extremely tempestuously. The timedependent velocity field can be seen in Fig.3.

The bubble plume rises along the centerline with the aspect ratio of 1.0 (No. 1), and the classical “cooling tower” structure is formed, resulting in more and more bubbles entering into both sides of the bubble plume. When the bubble plume in the timedependent velocity field with the aspect ratio of 1.5 (No. 2), an unstable structure is accompanied by a meandering motion in the lower part and collapses in the upper part due to the presence of many re-circulation bubbles. With the aspect ratio of 2.0 (No. 3), accompanied with the meandering motion in the bottom, the bubble plume rises in a snake-like structure, then some eddies are formed in the free surface, affecting the motion of the bubble plume. The fluid structure is interfered with the staggered vortex, as in agreement with the observed phenomena[10,11].

From the time-dependent velocity of each case, the corresponding time-serial flow field can be obtained as shown in Fig.4. The flow field shows a great variety in the same void fraction but with different aspect ratios.

For the case of No. 1, a symmetrical and steady uniform vortex circulation appears in both sides of the bubble plume.The total turbulent intensity is shown in the side faces, and the maximum value is 0.126 m/s. More and more bubbles gather in the water, which prolongs the effective contact time between bubbles and water, and the effective area is increased at the same time. The turbulent intensity is larger for the case of No. 2, as compared to other two cases. For the main direction is vertical, and the vortex structure is unsteady, the vibration period of the bubble plume is shortened. Obvious symmetric vortex circulation structure appears on both sides of the bubble plumes of the aspect ratio of 2.0. However, some small vortex structures also appear in the upper part, which makes an impact on the motion of the bubble plume[12,13].

3.2Analysis of the flow field for the same aspect ratioThe influence of the aspect ratio on the bubble plume is also considered in this article. Table 2 shows the experimental conditions when the aspect ratio is 1.0. With the gas flow rate increasing, the void fraction and the pressure in each case are also increased. The time-serial flow field is shown in Fig.5.

Table 2 Experimental conditions (H/W=1.0)

Chahine[13]mentioned that the depth would not affect the structure of the bubble plume in their experiments. However, the void fraction in their tests was only 0.02. In order to know the influence of the aspect ratio, the time-serial flow fields in different void fractions are considered in this article[14-16].

A comparison shows that the void fraction greatly influences the movement of the bubble plume of the same aspect ratio. As the void fraction increases, the vortex structure of the bubble plume becomes more and more clear, then more and more bubbles enter into the vortex circulation. The increasing gas flow rate also enhances the bubble velocity, however, the total turbulent intensity is decreeased. The bubbles rise along the centerline of the plume in the case of No. 4, with a total maximum turbulent intensity appearing in the middle of the bubble plume, of a value of 0.138 m/s. The vortex structure is more symmetrical, but with a little curl. With the void fraction increasing, the bubble plume width increases, and the total maximum turbulent intensity gradually shifts to both sides of the plume. The angular velocity of the fluid element is increased and it begins to expand. With the further accretion of the void fraction in the case of No. 6, the total maximum turbulent intensity appears onboth sides of the plume, of a value of 0.112 m/s, which is decreased, as can be explained by a small difference between the velocity of various points in the plume and the time-averaged velocity, and the homogeneous change of velocity. A symmetric vortex structure also appears in this case, and the angles of the two directions of the vortex are increased. The increasing bubble number and the homogeneous flow field greatly extend the effective contact time and contact area between bubbles and water, which is beneficial for the aeration.

Fig.5 Time-serial flow field with aspect ratio of 1.0 but various void fractions

4. Conclusions

Based on the experiments and the analyses of the bubble plume unsteady structure, the following conclusions can be drawn:

(1) Some factors that influence the structure of bubble plumes are considered in this article. The void fraction and the aspect ratio have a great impact on the motion of the bubble plume, which are the main factors responsible for the unsteady structure of the bubble plume.

(2) Through the calculation and analysis of the time-serial flow field of the bubble plume with the same void fraction in the moving-bed bio-film reactor, it is shown that the amplitude of the fluctuation of the bubble plume is the largest for the aspect ratio of 2.0, because when the gas flow rate increases, some vortex structures would form in the free surface; when the aspect ratio is 1.0, the flow field is more homogenous.

(3) For the flow field of the bubble plume with the same aspect ratio of 1.0, the vortex structure is more stable and symmetrical when the void fraction islarger, and the turbulent intensity is smaller than those in other cases, which is beneficial for the bubbles to stay in the water, and the effective contact time and contact area between bubbles and water are greatly enhanced. The aeration systems can play an important role in practical projects.

[1] CHENG Wen, LIU Wen-hong. Experimental study on gas-liquid two-phase flows in an areation tank by using image treatment method[J].Journal of Hydrodynamics,2008, 20(5): 650-655.

[2] HUANG Ya-fei, XU Liang and XU Chun-jiang et al. Application of PIV measurement in particle separation in upward field[J].Goal Preparation Technology,2010, (2): 14-19(in Chinese).

[3] CHENG Wen. Development of new techniques for image measurement and its inverse analysis of gas liquid two phase flows[D]. Ph. D. Thesis, Fukui, Japan: Fukui University, 2004.

[4] WAN Tian, CHENG Wen and LIU Xiao-hui. PIV technology for study of gas-liquid two-phase flow in an aeration tank[J].Advances in Science and Technology of Water Resources,2007, 27(6): 99-102(in Chinese).

[5] GUO Xiao-hui, LIN Jian-zhong. Flow past two rotating circular cylinders in a side-byside arrangement[J].Journal of Hydrodynamics,2009, 21(2): 143-151.

[6] LI Na, ZHANG Rong-sheng and LI Li et al. Pre-treatment technology of micro air bubble measurement in gas-liquid flow[J].Yangtze River,2009, 40(11): 73-75(in Chinese).

[7] CHENG Wen, SONG Ce and LIU wen-hong et al. Image analysis of bubble velocity in gas-liquid twophase flow[J].Journal of Engineering Thermophysics,2009, 30(1): 83-86(in Chinese).

[8] XIAO Hao-fei, ZHOU Mei-hua. CFD simulation of gasliquid flow in aeration tank[J].Journal of Anhui Agricultural Sciences,2010, 38(4): 1955-1957(in Chinese).

[9] LIU Wen-hong, WANG Meng and ZHOU Xiao-de et al. Experimental study on oil wastewater treatment by biological aerated filter[J].Journal of Xi’an University of Technology,2010, 26(3): 260-264(in Chinese).

[10] LIU Wen-hong, WAN Tian and CHENG Wen-juan et al. Analysis on steady structure of bubble plume in the basis of image binarization[J].Journal of Hydraulic Engineering,2009, 40(11): 1369-1372(in Chinese).

[11] DEEN N. G., WILLEMS P. and SINTANNALAND Van M. On image pre-processing for PIV of single- and two-phase flows over reflecting object[J].Experiments in Fluids,2010, 49(2): 525-530.

[12] ZHANG Kai, BRANDANI Stefano. CFD simulation of particle-fluid two-phase flow in fluidized beds[J].Journal of Chemical Industry and Engineering (China),2010, 61(9): 729-733.

[13] CHAHINE Georges L., TANGUAY Michel and LORAINE Greg. Acoustic measurements bubbles in biological tiessure[J].Journal of Hydrodynamics,2009, 21(1): 47-64.

[14] YUAN Wen-qi, LIU Sui-qing. Transient numerical modeling of gas-liquid two-phase flow in water-filling pipelines[J].Journal of Tongji University (Natural Science),2010, 38(5): 709-715(in Chinese).

[15] WANG Hong-yi, DONG Feng. Calculation method for bubble volume in gas-liquid two-phase flow[J].Chinese Journal of Scientific Instrument,2009, 30(11): 2444-2449(in Chinese).

[16] GAO Zhong-xin, DENG Jie and GE Xin-feng. Simulation of bubbly two-phase turbulent flow in circular pipe bend[J].Journal of Hydraulic Engineering,2009, 40(6): 696-708(in Chinese).

December 27, 2010, Revised May 20, 2011)

* Project supported by the National Natural Science Foundation of China (Grant No. 51076130), the National Basic Research Program of China (973 Program, Grant No. 2011CB403305).

Biography: LI Shi-rong (1973-), Male, Ph. D. Candidate, Senior Engineer

CHENG Wen,

E-mail: wencheng@xaut.edu.cn

2011,23(4):510-515

10.1016/S1001-6058(10)60143-6