Scholarly article on topic 'Non-invasive imaging of shallow bubble columns using electrical capacitance tomography'

Non-invasive imaging of shallow bubble columns using electrical capacitance tomography Academic research paper on "Chemical sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Journal of Saudi Chemical Society
OECD Field of science
Keywords
{"Shallow bubble column" / "Electrical capacitance tomography" / Imaging / "Flow regime"}

Abstract of research paper on Chemical sciences, author of scientific article — Waheed A. Al-Masry, Emad M. Ali, Saleh A. Alshebeili, Fouad M. Mousa

Abstract This paper deals with application of non-invasive electrical capacitance tomography to study the hydrodynamics of shallow bed bubble columns. Two bubble columns with different height to diameter ratio were used. Air–kerosene system that represents dielectric two-phase mixture was investigated. The ECT provided good measurement of the gas holdup at different gas velocities compared to the classical pressure measurements. The ECT was able to provide the gas hold up and the bubble velocities distribution across the column diameter at different gas velocities. The study revealed that spatial gas holdup and bubble velocity distributions are sharp with parabolic shape in the small bubble column (HD /DC =5). However, in the large bubble column (HD /DC =4) the gas holdup and bubble velocity profiles were flatter indicating improvement in the mixing homogeneity and leading to well-mixed reactor. 3D graphical visualization of the flow regimes and transition points were also examined using the ECT. In the small bubble column flow regimes were heterogeneous to slugs flow especially at high flow rate, resulted in downward flow near the walls and imperfect mixing.

Academic research paper on topic "Non-invasive imaging of shallow bubble columns using electrical capacitance tomography"

Journal of Saudi Chemical Society (2010) 14, 269-280

ORIGINAL ARTICLE

Non-invasive imaging of shallow bubble columns using electrical capacitance tomography

Waheed A. Al-Masry a *, Emad M. Ali a, Saleh A. Alshebeili b, Fouad M. Mousa c

a Department of Chemical Engineering, King Saud University, Riyadh, Saudi Arabia b Department of Electrical Engineering, Riyadh, Saudi Arabia c Chemical Research, SABIC R&T, Riyadh, Saudi Arabia

Received 3 November 2009; accepted 13 January 2010 Available online 4 February 2010

KEYWORDS

Shallow bubble column; Electrical capacitance tomography; Imaging; Flow regime

Abstract This paper deals with application of non-invasive electrical capacitance tomography to study the hydrodynamics of shallow bed bubble columns. Two bubble columns with different height to diameter ratio were used. Air-kerosene system that represents dielectric two-phase mixture was investigated. The ECT provided good measurement of the gas holdup at different gas velocities compared to the classical pressure measurements. The ECT was able to provide the gas hold up and the bubble velocities distribution across the column diameter at different gas velocities. The study revealed that spatial gas holdup and bubble velocity distributions are sharp with parabolic shape in the small bubble column (Hd/Dc = 5). However, in the large bubble column (HD/ DC = 4) the gas holdup and bubble velocity profiles were flatter indicating improvement in the mixing homogeneity and leading to well-mixed reactor. 3D graphical visualization of the flow regimes and transition points were also examined using the ECT. In the small bubble column flow regimes were heterogeneous to slugs flow especially at high flow rate, resulted in downward flow near the walls and imperfect mixing.

© 2010 King Saud University. All rights reserved.

* Corresponding author.

E-mail address: walmasry@ksu.edu.sa (W.A. Al-Masry).

1. Introduction

Multiphase flow systems are encountered in many chemical and biological processes where mass transfer and chemical reactions between the phases take place. Although fundamental studies of transport phenomena in multiphase systems particularly bubble column and three-phase fluidization systems have been extensive over the past decades, the present design status is still based on 'rule of thumb' rather than first principles. This is because the local flow structure is extremely complex and the link between the micro-scale and macro-scale has not been clearly established. The successful approach toward the understanding of such complex flows require

1319-6103 © 2010 King Saud University. All rights reserved. Peerreview under responsibility of King Saud University. doi:10.1016/j.jscs.2010.02.022

Production and hosting by Elsevier

Nomenclature

DC column diameter, cm UG superficial velocity, cm/s

HD column height, cm £Q gas hold up

L distance between two planes, cm s time delay parameter, s

R column radius, cm

Ub bubble velocity, cm/s

reliable experimental data, which in turn, depends on the implementation of sophisticated measuring techniques capable of behaviour investigation as well as the ability to provide the required information over the entire flow field. For this reason, several techniques to measure the bubble size, shape and velocity were developed and tested. A review of these methods can be found elsewhere (Al-Masry et al., 2005).

Recently most of the work dealing with studying the hydrodynamics is centred on the measurement of gas holdup, pressure fluctuation and acoustic sound (Lin et al., 2001; Al-Masry et al., 2006; Kashev et al., 2005; Chilekar et al., 2005; Vial et al., 2000, 2001; Mannasseh et al., 2001). Several types of analysis tools were successfully applied to extract useful information from the previously mentioned measurements. Among these tools are statistical (Vial et al., 2000; Letzel et al., 1997; Drahos et al., 1991), spectral (Vial et al., 2000; Letzel et al., 1997; Drahos et al., 1991), fractal (Vial et al., 2000; Drahos et al., 1992), time series (Vial et al., 2000, 2001) and deterministic chaos (Vial et al., 2000; Letzel et al., 1997; Lin et al., 2001). Al-Masry et al. (2005) have employed the spectral analysis to acoustic sound measurements in air-water system to determine the bubble characteristics such as the bubble size, bubble distribution, damping factor of the bubble oscillation. Al-Masry et al. (2007) and Al-Masry and Ali (2007) have extended the work to include the addition of coalescence and non-coalescence agents to air-water system. The influence of such surfactant compounds on the overall bubble characteristics was investigated. It is found that acoustic measurements and spectral tools are helpful to unfold valuable information about the bubble properties especially the bubble size and its frequency of oscillation.

In this work, we extend our investigation by considering the application of electrical capacitance tomography (ECT) to study the hydrodynamic properties of bubble columns. The first reported ECT system was developed by the US Department of energy in Morgantown, to image the gas/solids distribution in fluidized beds (Fasching and Smith, 1991; Halow et al., 1993). Subsequently, the UMIST (UK) ECT system has been used successfully in a number of research investigations for imaging other two-phase processes, such as: gas/solid distribution in pneumatic conveyors and fluidized beds (Liu et al., 2001), combustion flames in engine cylinders (He et al., 1994), and water hammer (Yang, 1996). The performance of an ECT in flow pattern and void fraction determination in multiphase flows containing oil and water was evaluated by Pacho and Davies (2003). Most of the tomo-graphic techniques are too slow to capture the real-time image of highly fluctuating multiphase flows. Thus, the data captured are in time-averaged mode. For fully imaging, the technique is limited to bubbles with sizes larger than the spatial resolution, i.e. 5-10% of the column diameter. For smaller bubbles, the

technique only gives the information of the hold up within the pixel (Warsito and Fan, 2001). The ECT studies reveal a radial symmetry of the time-averaged solids hold up distribution for the turbulent regime.

Meanwhile, the recent progress in the development of measuring techniques such as process tomography has provided more insights into the complex multiphase flow phenomena, and gives validation of monitoring, both continuously and simultaneously, the local and global dynamic behaviour of the gas bubbles and the particles in a non-invasive manner (Williams and Beck, 1995). A remarkable progress for electrical capacitance tomography (ECT) technique has been made in the past decade, including different hardware systems, image reconstruction algorithms, and applications (Halow et al., 1990; Huang et al., 1989; Yang, 1996). The selection of a particular measurement technique has to be based on a thorough analysis of all salient and contributing factors. These include the ability of the technique to distinguish between various phases and provide quantities information about mixture contents. Among available tomography techniques, electrical tomography, either resistance or capacitance, is the most promising technique that has a relatively high temporal resolution, up to 5% of column diameter. The latest development of electrical capacitance tomography (ECT) could capture tomography data up to 1000 frames/s. Thus electrical capacitance tomography (ECT) is an emerging technique because of its simplicity, cheapness, robustness and non-intrusive property. The technique is aimed at internal visualization of industrial processes like mixing, separation and two-phase flow (Yang et al., 1995; Xie et al., 1989). Indeed, electrical tomo-graphic imaging has been applied to a broad range of chemical engineering processes, including: bubble columns, fluidized beds, pneumatic transport, liquid mixing, cyclonic separation, pressure filtration, liquid pipe-flow, polymerization, and emergency depressurization and paste extrusion.

2. Experimental setup

Two bubble columns were used in this study. The small column is made from Perspex with 15 cm diameter and 5 m in height, and fitted with an external sliding ECT sensor. The different positions of the sliding ECT sensor above the sparger are; 30, 50, 75, 80, and 105 cm. The superficial gas velocity ranged from 0.01 to 0.17 m/s. The large column was made form QVF glass sections with 45 cm diameter and 5 m in height, and fitted with a fixed external ECT sensor installed at 75 cm above the sparger. Due to the heavy weight of the ECT sensor and the geometrical variations of the QVF glass sections; it was not possible to slide the sensor on the glass column. The superficial gas velocity used in the large column ranged from 0.005 to 0.11 m/s. The sliding and fixed sensors lengths are

Pre-distributor gas supply

Direct gas supply

Liquid discharge

Figure 1 Gas distribution system.

50 and 100 cm, respectively. For the gas-liquid system, air was the sparging phase, while kerosene was the liquid phase. Both columns were fitted with 5 mm thick perforated plate with 7 mm holes. The large diameter sparger hole is important industrially, simulating exothermic petrochemical reactions in bubble columns, where the gas acts as both reacting and cooling medium. In particular, for polymeric reactions where internals contributes negatively to the behaviour of bubble columns. A pre-distributer under the sparger is used to distribute the gas over the perforated plate. Gas distribution system is shown in Fig. 1. Schematic experimental setup is shown in Fig. 2. The experimental runs were analyzed using pressure drop (PD) measurements and electrical capacitance tomography (ECT) measurements.

The ECT system consists of a capacitance sensor, data acquisition system, and computer system for image reconstruction, interpretation, and display. The data acquisition system is manufactured by Processes Tomography (UK). The standard measurements protocol involves measuring the capacitance between all combinations of single 'source' and 'detector' plates, giving E(E — 1)/2 measurements, where E is the number of electrodes located around the circumference. The rate at which images can be produced varies, depending on the data acquisition system, the measurement protocol, and the method of image reconstruction (Byars, 1999; Tapp et al., 2003). The

Figure 2 (a) Schematic diagram of bubble column. (1) Bubble column; (2) manometer; (3) mass flow controller; (4) sliding sensor; (5) needle valve; (6) three-way valve; (7) liquid line; (8) drain liquid; (9) sparger; (10) gas line; (11) PC; (12) electrode; (13) electronics; (14) wires; (15) exhaust system. (b) Schematic diagram of the electrical capacitance tomography (ECT) system. (1) Electronics; (2) PC; (3) electrode; (4) sliding sensor; (5) wires.

ECT32v2 software (Tomography Ltd., UK) is used for concentration (gas holdup) analysis. Flowan multiphase analysis software (Flowan Ltd., UK) is used for speed and flow analysis. Both installed ECT sensors consisted of 12 electrodes with dual planes for gas dynamics measurements. The ECT software converts the capacitance measurements data points into permittivity concentration image, and display it into 32 x 32 square pixel grids.

3. Effect of height to diameter on gas hold up

A typical experimental ECT data point measurement takes 30 s. In each second of the measurement, plane 1 and plane 2 were set to records 100 frames (100 fps) each. This gives the total frames recorded for one experimental data point to 3000 frames for plane 1 and 3000 frames for plane 2. Usually one complete experimental run compose of at least 10 data points. Because the number of the frames is extremely large, an averaging procedure is used. The frames values for each experimental data point will be averaged for frame 1, 1000, 2000, and 3000 only. Experimental positions of the ECT sensors are given in Table 1. For example, at HD/Dc = 5 we have one sensor positioned at 30 cm. Table 2 shows the frame values at 1, 1000, 2000, and 3000 for the high permittivity fluid, i.e. the kerosene at HD/Dc = 5.

The gas holdup determined by the ECT sensor in comparison with the gas holdup from PD measurements is shown in Fig. 3 for HD/Dc = 5. It is clear that there is a difference between the two methods of measurement. However, the trend of the curves is increasing with increasing gas input. The difference in the calculated gas hold up may be attributed to the sensitivity of each instrument and to numerical errors. Increasing HD/Dc ratio to 8 made it possible to slide the ECT sensor to two axial positions. All the measurements are shown in Fig. 4. It is clear that from the two sensor positions that the

Table 1 Dispersion height to column diameter, HD/Dc (cm/cm).

Hd/Dc (cm/cm) 75/15 (5) 120/15 (8) 180/15 (12) 240/15 (16) 180/45 (4)

ECT sensor position 1 (cm) SS30 SS30 SS30 SS25 GS50

ECT sensor position 2 (cm) - SS80 SS80 SS50 -

ECT sensor position 3 (cm) - - SS140 SS105 -

ECT sensor position 4 (cm) - - - SS160 -

Table 2 Sample permittivity concentrations for HD/Dc = 5.

Ugr (m/s) Frame value at plane 1 Frame value at plane 2

1 1000 2000 3000 1 1000 2000 3000

0.028 91.74 88.93 92.10 92.26 93.35 93.67 93.16 89.48

0.038 91.18 88.54 89.47 87.57 89.98 86.55 89.66 89.77

0.047 81.82 89.34 85.37 85.80 87.56 86.04 78.89 84.25

0.057 82.15 85.60 85.1 83.30 78.46 86.09 82.12 77.41

0.066 79.92 80.12 77.73 81.40 77.63 81.11 78.42 77.36

0.075 78.87 83.20 70.79 82.18 78.63 78.69 77.27 73.96

0.085 68.83 78.08 76.59 76.82 76.61 69.34 66.04 74.39

0.094 78.16 74.06 72.42 66.69 74.36 75.60 73.67 78.74

0.123 68.99 56.94 69.03 67.79 74.88 63.59 2.21 66.94

0.142 65.95 59.94 66.17 69.47 66.18 72.86 58.63 59.68

0.170 65.76 58.63 67.17 60.44 62.15 57.72 62.69 61.97

0 0.05 0.1 0.15 0.2

Uqr, (m/s)

Figure 3 Gas hold up for HD/Dc = 5.

Uqr, (m/s)

Figure 4 Gas hold up for Hd/Dc = 8.

PD, HD/Dc=12 ECT, SS30 ECT, SS80 ECT, SS140

Ugr, (m/s)

Figure 5 Gas hold up for Hd/Dc = 12.

area (where the sliding sensor is positioned at SS30 is taking pictures) to smaller bubbles as they travel axially where the sliding sensor is positioned, SS80. In comparison with the pressure drop measurements, the three curves are close but the ECT at SS30 is giving the lowest gas hold up. Fig. 5 shows the gas hold up at HD/DC = 12 for three different sensor positions. The general conclusion is that the ECT results are in the range of the gas holdup values obtained by the PD. There is no clear trend for the effect of the ECT sensor position which may indicate uniform gas hold up through the height of the column. However, extending the column to Hd/Dc =16 starts illustrating the effect of sensor position as shown in Fig. 6. Slight increase in the gas hold up with increasing the sensor position can be observed especially at high gas superficial velocities. This is possibly related to the smaller size of the bubbles as they travel axially through the column. One may also argue that these minor differences are within experimentation accuracy.

Ugr, (m/s)

Figure 6 Gas hold up for Hd/Dc = 16.

gas hold up measured at SS80 is higher than at SS30. Possible explanation is the break-up of large bubbles formed at sparger

4. Hydrodynamics spatial variation

In this section we utilize the ability of ECT to measure the distribution of the gas holdup across the bubble column width. ECT has the ability to estimate the bubble velocity by analyzing the gas concentration at different predefined time periods. From this analysis, a gas velocity profile can be generated at different location in column spatial direction. The spatial variations of gas holdup and bubble velocity will be calculated using the software and summarized for those with HD/ DC 6 5. It should be noted that the gas velocity is determined from the cross-correlation of the gas concentration at two consecutive planes which are L distant apart. The cross-correlation gives the time delay parameter, s from which the gas velocity is calculated as U = L/s. Fig. 7 shows the spatial variation of gas hold up as a function of superficial gas velocity. The spatial gas hold up shape for the very low two superficial gas velocities of UGR = 0.009 and 0.019 m/s is not representative of the results and represent an experimental error. This is expected as the gas flow is very low and the ECT sensors could

0 R/Dc

UGR= =0.028 m/s

UGR= =0.038 m/s

UGR= =0.047 m/s

UGR= =0.057 m/s

UGR= =0.070 m/s

UGR= =0.075 m/s

UGR= 0.085 m/s

UGR= =0.094 m/s

UGR= =0.123 m/s

UGR= =0.142 m/s

UGR= =0.170 m/s

Figure 7 Spatial variation of gas hold up at Hd/Dc = 5.

not detect any fluid dynamics in the column. For the rest of the results, the spatial profile gives low gas concentration near the column walls with maximum value at the centre. It is also evident that the spatial gas hold up profile increases with increasing input superficial gas velocity. Fig. 8 shows the spatial average gas velocity over the experimental run which is 30 s for each superficial gas velocity. The profiles are excellent and show the powerful capabilities of the new ECT techniques in capturing the fluid dynamics phenomena in bubble columns. Since the bubble velocities are averaged, some of the negative values particularly near the walls will be positive in values. However, they still clearly show that the velocity of the bubbles is low at the walls. To observe this effect, the instantaneous bubble velocity (not averaged) at t = 15 s are plotted in Fig. 8b. Now, we can see very clearly that the bubble velocities near the wall are around UBi = —0.5 m/s and at the centre

varies between UBi = 0.5 and 1.75 m/s. Negative velocity may indicate bubbles that are reversing their movement direction. Bubble close to the wall, as expected, have lower velocity. This can be attributed to wall friction and due to the fact that they get away from the air jet force. It is clear that the holdup distribution is more uniform than the bubble velocity. The latter is highly expected because of the bubble to bubble and bubble to wall collision and due to the circulated movement of the bubbles.

To examine the spatial effects on the hydrodynamics parameter in the shallow bed bubble column, we repeat the experiments for Hd/Dc = 4. Fig. 9 shows the spatial gas hold up profile. The profiles are almost flat at very low superficial gas velocities. At moderately high superficial gas velocities UG @ 0.05 m/s, the profiles start to increase at the column centre giving the famous bell-shape curve. However, the profiles

(b) instantaneous

(a) average

Figure 8 Spatial variation of bubbles velocity at HD/Dc = 5, legends are the same as in Fig. 7.

-00 R/Dc

UGR= =0.010 m/s

UGR= =0.016 m/s

UGR= =0.021 m/s

UGR= =0.026 m/s

UGR= =0.031 m/s

UGR= =0.037 m/s

UGR= =0.042 m/s

UGR= =0.047 m/s

UGR= 0.052 m/s

UGR= =0.058 m/s

UGR= 0.063 m/s

UGR= 0.068 m/s

UGR= 0.073 m/s

UGR= =0.079 m/s

UGR= 0.084 m/s

UGR= 0.089 m/s

UGR= 0.094 m/s

UGR= 0.099 m/s

UGR= 0.105 m/s

Figure 9 Spatial variation of gas hold up at HD/Dc = 4.

are less steep at the centre than those shown in Fig. 8 at HD/ DC = 5. This may give an indication that the mixing and homogeneity are improved in the shallow bed bubble column, and the wall effects are diminishing. The downward movement of the fluid near the walls can be depicted by showing the spatial bubble velocity profiles shown in Fig. 10. The spatial average bubble velocities over the plane zones are more uniform with the lowest values near the walls. Fig. 10b depicts the instantaneous bubble velocity which reflects the actual bubble velocity at t = 15 s. There are two records of negative instantaneous velocities at the walls suggesting a downward fluid movement at the walls.

5. Flow regimes

This part discusses the flow regimes using 3D plots made up of the images captured by the ECT system. These 3D images are

obtained by employing MATLAB software (Mathworks, Inc., USA). The flow regimes are identified from visual observation of the composed images shown below. Fig. 11a shows the 3D plot at UG = 0.009 m/s and Hd/Dc = 5. The image is made up of 100 frames captured at plane 2. It is evident that the general flow picture implies almost stationary liquid at this gas flow rate. The color code red = 1 means 100% liquid and the color code blue = 0.0 means 100% gas. The change in color in between gives composed picture of what is happening inside the column. This technique is important for nontransparent columns where visual or photographical techniques fail. The number from 1 to 12 at the top of the generated column is the location of the ECT sensors. Fig. 11b shows very small changes when the gas input is increased to UG = 0.028 m/s. The beginning of green color patches to appear when gas input is increased to UG = 0.038 and 0.047 m/s, as seen in Fig. 12. This indicates that the bubbly flow

R/Dc (a) average

(b) instantaneous

Figure 10 Spatial variation of bubbles velocity at Hd/Dc = 5, legends are the same as in Fig. 9.

(а) (Ъ)

Figure 12 3D plot for Hd¡Dc = S, (a) UG = 0.038 m¡s, (b) UG = 0.047 m¡s, 100 frames.

(a) (b)

Figure 13 3D plot for HD¡DC = S, (a) UG = 0.0S7 m¡s, (b) UG = 0.07 m¡s, 100 frames.

(a) (b)

Figure 15 3D plot for HD/Dc = 5, (a) UG = 0.094 m/s, (b) UG = 0.123 m/s, 100 frames.

(a) (b)

Figure 16 3D plot for HD/Dc = 5, (a) UG = 0.094 m/s, (b) UG = 0.123 m/s, 100 frames.

(а) (Ь)

Figure 18 3D plot for HD¡DC = 4, (a) UG = 0.0314 m¡s, (b) UG = 0.0419 m¡s, 100 frames.

(a) (b)

Figure 19 3D plot for HD¡DC = 4, (a) UG = 0^7б m¡s, (b) UG = 0.0б29 m¡s, 100 frames.

(a.) (b)

Figure 20 3D plot for HD¡DC = 4, (a) UG = 0.0733 m¡s, (b) UG = 0.0838 m¡s, 100 frames.

regime is still prevailing at these superficial gas velocities. However, the generated pictures also indicate chaotic movement of the gas stream left to the right. Increasing UG to 0.057 m/s shows the gas moving like plugs (Fig. 13a). The change to heterogeneous regime is clearly visible at UG = 0.07 m/s as it seen in Fig. 13b. Fig. 14 clearly illustrates the heterogeneous regime. Fig. 15a shows the beginning of flow regime transition from heterogeneous to slug flow (DC = 15 cm) at UG = 0.094 m/s. Figs. 15b and 16a show clearly slug flow regime. However, at UG = 0.17 m/s the flow becomes annular with mainly gas in the column while the liquid is at the walls as it is clearly shown in Fig. 16b.

The following set of generated 3D plots Figs. 17-22 are related to Hd/Dc = 4, i.e. DC is increased from 15 to 45 cm. Fig. 17 is at very low gas flow rate, and therefore the red color is predominant. Fig. 18 shows how the red color starts chagrining to orange color, indicating that bubbly flow regime is beginning. Fig. 19 shows clearly that the greenish color is

filling the column with the red color at the walls up to UG = 0.0629 m/s. This indicate that the bubbly flow regime is well established with the mixing homogeneity is approaching the perfect stage. As the gas velocity increases, the green color intensifies as illustrated in Figs. 20 and 21. Increasing the superficial gas velocity to UG = 0.089 m/s as seen in Fig. 21a, the perfect green color is well established in the main column except at the walls the red color is predominant. This means that the mixing is homogenous in the upward flow while thin layer of liquid is moving near the walls downwards. Fig. 22 demonstrates the effect of flow rate as high as UG = 0.105 m/s with 100 frames. Even at this high flow rate there is no clear blue color, i.e. no junk of large bubbles are moving in patches in the centreline of the column. This is in agreement with the flat gas holdup profiles discussed above. In comparison with Fig. 16, we can see clearly that at similar superficial gas velocity two different flow regimes are encountered as a function of HD/Dc. Decreasing the Hd/Dc ratio

resulted in delay of transition of flow regimes; with the bubbly to transition flow regime is predominant in the large column. The wall effects are eliminated by enlarging the column diameter.

6. Conclusions

This paper utilizes the electrical capacitance tomography to analyze the bubble characteristics inside bubble columns. Our investigation revealed that ECT is a powerful tool that can capture the spatial bubble properties with high resolution capability. Unlike the sound measurement techniques, the variation of the gas holdup and the bubble velocity with the column diameter were obtained. Comparison of these properties for two columns with different height to diameter ratio revealed that shallow bubble columns have flat gas hold up profiles with the fluid direction mainly is upward and minimal downward liquid flow. In contrast to photo-dependent techniques, the ECT can be performed in non-opaque media. The later feature expands the range of ECT utilization. Furthermore, ECT was able to provide useful information about the flow regimes and transition points as the intensity of the gas flow can be easily visualized graphically. The intensity and structure of these 3D graphical images of the gas flow can be interpreted to identify unique flow regimes and the transition points from specific regime to another. Specifically, it is found that shallow bubble columns have better homogenous flow regimes than tall bubble columns at similar superficial gas velocities.

References

Al-Masry, W., Ali, E., 2007. Identification of hydrodynamics characteristics in bubble columns through analysis of acoustic sound measurements - influence of the coalescence behavior of the liquid phase. Chemical Engineering and Processing 46, 127-138. Al-Masry, W., Ali, E., Aqeel, Y., 2005. Determination of bubble characteristics in bubble columns using statistical analysis of acoustic sound measurements. Chemical Engineering Research and Design 83, 1196-1207. Al-Masry, W., Ali, E., Aqeel, Y., 2006. Effect of surfactant solutions on bubble characteristics in bubble columns based on acoustic sound measurement. Chemical Engineering Science 61, 3610-3622.

Al-Masry, W., Ali, E., Alkalbani, M., 2007. Prediction of regime transitions in bubble columns using acoustic and differential pressure signals. Chemical Engineering Journal 133, 139-149. Byars, M., 1999. Development in electrical capacitance tomography. In: First World Congress on Industrial Tomography, Buxton, UK. pp. 1-8.

Chilekar, V., Warnier, M., Schaff, J., Kuster, B., Schouten, J., Ommen, J., 2005. Bubble size estimation in slurry bubble columns from pressure fluctuations. American Institute of Chemical Engineers Journal 51, 1924-1937. Drahos, J., Zahradnik, J., Puncochar, M., Fialova, M., Bradka, F., 1991. Effect of operating conditions on the characteristics of pressure fluctuations in a bubble columns. Chemical Engineering Processing 29, 107-115. Drahos, J., Bradka, F., Puncochar, M., 1992. Fractal behaviour of pressure fluctuations in bubble column. Chemical Engineering Science 47, 4069-4075.

Fasching, G.E., Smith, N.S., 1991. A capacitive system for 3-dimensional imaging of fluidized-beds. Review of Scientific Instruments 62, 2243-2251.

Halow, J.S., Fasching, G.E., Nicoletti, P., 1990. Preliminary capacitance imaging experiments of a fluidized bed. American Institute of Chemical Engineers Journal Symposium Series 86, 41-50.

Halow, J.S., Fasching, G.E., Nicoletti, P., Spenick, J.L., 1993. Observations of a fluidized-bed using capacitive imaging. Chemical Engineering Science 48, 643-659.

He, R., Xie, C.G., Waterfall, R.C., Beck, M.S., Beck, M.C., 1994. Engine flame imaging using electrical capacitance tomography. Electronic Letters 30, 559-560.

Huang, S.M., Plaskowski, A.B., Xie, C.G., Beck, M.S., 1989. Tomographic imaging of two-component flow using capacitance sensors. Journal of Physics E: Scientific Instruments 22 (3), 173177.

Kashev, C., Chilekar, V., Warnier, M., Schaff, J., Kuster, B., Schouten, J., Ommen, J., 2005. Detecting regime transitions in slurry bubble columns using pressure time series. American Institute of Chemical Engineers Journal 51, 1951-1965.

Letzel, H.M., Schouten, J.C., Krishna, R., van den Bleek, C.M., 1997. Characterization of regimes and regime transitions in bubble columns by chaos analysis of pressure signals. Chemical Engineering Science 52, 4447-4459.

Lin, T.J., Juang, R.C., Chen, Y.C., Chen, C.C., 2001. Prediction of flow transitions in a bubble columns by chaotic time series analysis of pressure fluctuation signals. Chemical Engineering Science 56, 1057-1065.

Liu, S., Yang, W.Q., Wang, H.G., Su, Y., 2001. Investigation of square fluidized beds using capacitance tomography: preliminary results. Measurement Science and Technology 12, 11201125.

Mannasseh, R., LaFontain, R.F., Davy, J., Shepherd, I.C., Zhu, Y., 2001. Passive acoustic bubble sizing in sparged systems. Experiments in Fluids 30, 672-682.

Pacho, D., Davies, G., 2003. Detection and monitoring of oil foams using raw capacitance data. Industrial Engineering Chemical Research 42, 636-645.

Tapp, H.S., Peyton, A.J., Kemsley, E.K., Wilson, R.H., 2003. Chemical engineering applications of electrical process tomography. Sensors and Actuators B 6702, 1-8.

Vial, C., Camarasa, E., Poncin, S., Wild, G., Midoux, N., Bouillard, J., 2000. Study of hydrodynamic behaviour of bubble columns and external loop airlift reactors through analysis of pressure fluctuations. Chemical Engineering Science 55, 2957-2973.

Vial, C., Poncin, S., Wild, G., Midoux, N., 2001. A simple method for regime identification and flow characterization in bubble columns and airlift reactors. Chemical Engineering Science 40, 135-151.

Warsito, W., Fan, L.S., 2001. Neural network based multi-criterion optimization image reconstruction technique for imaging two and three-phase flow system using electrical capacitance tomography. Measurement Science and Technology 12, 2198-2210.

Williams, R.A., Beck, M.S., 1995. Process Tomography - Principles, Techniques and Applications. Butterworth-Worth-Heinemann, Oxford, UK.

Xie, C.G., Plaskowski, A., Beck, M.S., 1989. 8-Electrode capacitance system for two-component flow identification part 1: tomographic flow imaging. Measurement and Instrumentation IEE Proceedings 136, 173-183.

Yang, W.Q., 1996. Hardware design of electrical capacitance tomography systems. Measurement Science and Technology 7, 233-246.

Yang, W.Q., Beck, M.S., Byars, M., 1995. Electrical capacitance tomography: from design to applications. Measurement Control 28, 261.