- Email: [email protected]
Study of axial mixing, holdup and slip velocity of dispersed phase in a pulsed sieve plate extraction column using radiotracer technique
ARTICLE IN PRESS Applied Radiation and Isotopes 67 (2009) 1248–1253
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Study of axial mixing, holdup and slip velocity of dispersed phase in a pulsed sieve plate extraction column using radiotracer technique Ghiyas Ud Din a,c,, Imran Rafiq Chughtai b, Mansoor Hameed Inayat b, Iqbal Hussain Khan c a
Department of Nuclear Engineering, Pakistan Institute of Engineering and Applied Sciences [PIEAS], P.O Nilore, Islamabad, Pakistan Department of Chemical and Materials Engineering, Pakistan Institute of Engineering and Applied Sciences [PIEAS], P.O Nilore, Islamabad, Pakistan c Isotope Application Division, Pakistan Institute of Nuclear Science and Technology [PINSTECH], P.O Nilore, Islamabad, Pakistan b
a r t i c l e in f o
Keywords: Liquid–liquid extraction Pulsed sieve plate column Residence time distribution (RTD) Axial mixing Holdup Slip velocity Radiotracer
a b s t r a c t Axial mixing, holdup and slip velocity of dispersed phase which are parameters of fundamental importance in the design and operation of liquid–liquid extraction pulsed sieve plate columns have been investigated. Experiments for residence time distribution (RTD) analysis have been carried out for a range of pulsation frequency and amplitude in a liquid–liquid extraction pulsed sieve plate column with water as dispersed and kerosene as continuous phase using radiotracer technique. The column was operated in emulsion region and 99mTc in the form of sodium pertechnetate eluted from a 99Mo/99mTc generator was used to trace the dispersed phase. Axial dispersed plug flow model with open–open boundary condition and two points measurement method was used to simulate the hydrodynamics of dispersed phase. It has been observed that the axial mixing and holdup of dispersed phase increases with increase in pulsation frequency and amplitude until a maximum value is achieved while slip velocity decreases with increase in pulsation frequency and amplitude until it approaches a minimum value. Short lived and low energy radiotracer 99mTc in the form of sodium pertechnetate was found to be a good water tracer to study the hydrodynamics of a liquid–liquid extraction pulsed sieve plate column operating with two immiscible liquids, water and kerosene. Axial dispersed plug flow model with open–open boundary condition was found to be a suitable model to describe the hydrodynamics of dispersed phase in the pulsed sieve plate extraction column. & 2009 Elsevier Ltd. All rights reserved.
1. Introduction Liquid–liquid extraction is a process of separation of constituents of a liquid phase by contacting it with another immiscible liquid phase. Petroleum, nuclear, chemical, metallurgical, pharmaceutical, food processing and bio-processing industries are the major beneficiaries of this technology. Column type contactors with agitation are very famous liquid–liquid extraction equipments as they offer large interfacial area, high mass transfer coefficient, high turbulence and minimum radial gradients. Counter current movement of phases in these type of columns provide high concentration gradients for efficient mass transfer but axial mixing in both phases lowers the process efficiency by lowering solute concentration gradients. It has been reported that axial mixing in extraction columns lowers the process efficiency
Corresponding author at: Department of Nuclear Engineering, Pakistan Institute of Engineering and Applied Sciences [PIEAS], P.O Nilore, Islamabad, Pakistan. Tel.: +92 51 2207381x3332; fax: +92 51 2208070. E-mail addresses: [email protected], [email protected] ( Ghiyas Ud Din).
0969-8043/$ - see front matter & 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2009.02.024
as much as 30% (Li and Ziegler, 1967). The major sources of axial mixing in the extraction columns are geometrical and operating parameters. Therefore, the presence of axial mixing in such kind of equipments is unavoidable and needs special care. A simple approach to represent the hydrodynamics of phases in these kind of columns is the axial dispersion model (ADM) (Levenspiel and Smith, 1957; Levenspiel, 1999) and the concept of residence time distribution (RTD) analysis is an important method for the estimation of axial dispersion in chemical reactors (Danckwerts, 1953). The holdup and slip velocity of dispersed phase are other parameters of fundamental importance that need to be focused in the design and operation of pulsed extraction columns. Axial mixing in the continuous phase of liquid–liquid extraction columns remained a major focus of many studies. These include studies on spray towers (Hazlebeck and Geankoplis, 1963; Henton and Cavers, 1970; Geankoplis et al., 1982), reciprocating plate extraction columns (Kim and Baird, 1976a, b; Hafez et al., 1979; Parthasarathy et al., 1984) and pulsed sieve plate extraction columns (Kumar and Hartland, 1989). Axial mixing in the dispersed phase of liquid–liquid extraction columns has either been assumed negligible or it has not been estimated. Only a few studies can be found regarding axial dispersion in the dispersed
ARTICLE IN PRESS Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
phase (Srinikethan et al., 1987) with conventional methods in which a non-radioactive tracer is injected into the system and measurements were made by a conductivity probe. The holdup and slip velocity of the dispersed phase have also been determined by arresting the flow of phases and measuring the fractional dispersed phase volume in the extraction column (Venkatanarasaiah and Varma, 1998). These experimental ap-
1249
proaches present some disadvantages including low sensitivity, requirement of phase separation before measurement and poor statistics. Most important of all is that one needs to shutdown the plant to obtain the holdup fraction at each operating parameter resulting in a high plant shutdown. Radiotracers are being used for industrial process optimization and trouble shooting from decades. They offer state of the art
Light phase outlet
Injection port D1
D2
To collection vessel
Balance leg From heavy phase feed vessel via metering pump
To collection vessel
Heavy phase inlet
Ratemeters
Data Acquisition System D3
Light phase inlet
Interface level Heavy phase outlet
From light phase feed vessel via metering pump
Pulse unit
Fig. 1. Schematic diagram of pulsed sieve plate extraction column.
ARTICLE IN PRESS 1250
Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
technology with high sensitivity, on-line measurement without disturbing the plant operation, better statistics, high benefit to cost ratio and possibility of getting information from remote locations of process equipment. A variety of industrial process investigations have been carried out using the power of radiotracer technology (Thyn et al., 2000; Thyn and Zitny, 2002; Pant and Yelgoankar, 2002; Farooq et al., 2003; Kim et al., 2005). Although numerous applications can be found in literature regarding industrial process investigations using radiotracer technology but only a few have been reported with reference to the hydrodynamics of liquid–liquid extraction columns (Arthayukti et al., 1976). The present investigations are focused to study the hydrodynamics of dispersed phase in a pulsed sieve plate extraction column using radiotracer RTD analysis. Effect of pulsation frequency and amplitude have been studied on the axial mixing, holdup and slip velocity of dispersed phase for two sets of dispersed and continuous phase superficial velocities and the system was simulated using the ADM.
2. Materials and methods The schematic diagram of pulsed sieve plate extraction column under investigation is shown in Fig. 1. The internal diameter of the column is 5 102 m and height is 2 m. Two separating chambers one at the top and the other at the bottom of the column are also part of this apparatus. The column is fitted with regularly spaced (5 102 m) sieve plates which help to increase the interfacial area between the two immiscible liquids. The column was operated counter currently with heavy phase (water) as dispersed and light phase (kerosene) as continuous phase. The kerosene is fed into the lower separating chamber with the help of a metering pump which then flows upwards through the sieve plate column to the upper separating chamber where it overflows to a collection vessel. Similarly water is fed into the top separating chamber via a metering pump from where it flows downwards through the column to the lower separating chamber and then it flows through a balance leg into a collection vessel. A pulse unit located at the base of lower separating chamber provides vertical pulses to the flowing fluids. The column was operated in the emulsion regime i.e. dispersed phase remained dispersed throughout the plate stack and no coalescence into layers occurred at the plates. A liquid–liquid interface was allowed to form at about 10 cm below the light phase inlet and this interface level was stabilized with the help of a balance leg before starting an experiment. 99m Tc in the form of sodium pertechnetate having half life of 6.02 h and gamma energy 0.14 MeV (91%) is a well known water tracer (Pant et al., 2000) but no data is available for its use in water–kerosene environment. Therefore, validation of radiotracer was carried out by thoroughly mixing about 0.1 mCi 99mTc in equal amounts of water and kerosene. Upon separation and measurements, the radiotracer was found suitable for labeling water phase in water–kerosene environment. About 0.5 mCi of 99mTc eluted from a 99Mo/99mTc generator was injected in the form of an instantaneous pulse to carry out RTD experiments for investigation of the hydrodynamics of dispersed phase (water) as per experimental plan shown in Fig. 1. The experiments were carried out for a range of pulsation frequency and amplitude as given in Table 1. The movement of radiotracer was monitored for every second with the help of lead collimated 2 in 2 in NaI(Tl) scintillation detectors mounted at D1, D2 and D3 as shown in Fig. 1. The data was acquired on-line using a multi-channel data acquisition system and stored in a computer for processing.
Table 1 Pulsed sieve plate extraction column specification and range of operating variables. Internal diameter of the column (m) Length of the column (m) Number of sieve plates Plate spacing (m) Plate diameter (m) Hole size (m) Average number of holes Average free area (%) Dispersed phase superficial velocity (m/s) Continuous phase superficial velocity (m/s) Range of pulsation frequency (s1) Range of pulsation amplitude (m)
5 102 2 38 5 102 5 102 2 103 140 per plate 25 0.34 102 and 0.44 102 0.37 102 and 0.47 102 0.7–2.0 0.8 102–1.6 102
The tracer data from detectors D2 (column inlet) and D3 (column outlet) was corrected for background, radioactive decay and normalized. The experimental mean residence time (MRT) of the system was calculated by the difference of first moments of outlet and inlet response curves. Mathematical expression for the first moment in discrete form can be written as P t C Dt (1) First moment ¼ Pi i i i i C i Dt i where C is the tracer concentration (counts/s in present case), t is the time of measurement (s), Dt is the time interval between the two measurements (s), and i ¼ 0,1,2,3,y. Overall holdup of the phase under investigation was calculated on the basis of calculated MRT using the following relationship: Hd ¼
tQ d VR
(2)
where Hd is the dispersed phase holdup, Qd is the dispersed phase flow rate, t is the MRT, and VR is the effective reactor volume. The slip velocity of the dispersed phase averaged over the whole column was estimated from the above calculated Hd using the following equation (Venkatanarasaiah and Varma, 1998): Vs ¼
Ud Uc þ Hd ð1 Hd Þ
(3)
where Vs is the slip velocity of dispersed phase, Ud is the dispersed phase superficial velocity, Uc is the continuous phase superficial velocity. The RTD is a probability distribution function that describes the amount of time a fluid element spends inside a reactor. It helps in troubleshooting of reactors and characterizes the mixing and flow within the reactors. If an impulse of tracer is injected at the inlet of a system at time t ¼ 0 and its concentration is measured as a function of time at the outlet, then E(t) representing the probability for a tracer element to have a residence time between the time interval (t, t+dt) is defined as C ðtÞ Ei ðtÞ ¼ R 1 i 0 C i ðtÞ dt such that Z 1 Ei ðtÞ dt ¼ 1
(4)
(5)
0
where i ¼ 1,2,3,y,n, Ci(t) is the tracer concentration, Ei(t) is the RTD function. RTD models have been playing a vital role for industrial process investigations from decades. They provide macroscopic lumped sum description, which is sufficient for many engineering calculations. The plug flow is an ideal condition for the flow of
ARTICLE IN PRESS Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
phases in an extraction column but some degree of axial mixing is always inevitable. The pulsation and closely spaced sieve plates inside the column continues to re-arrange the droplets of the dispersed phase randomly. Keeping in view of these considerations, the ADM was used to study the system hydrodynamics.
1251
The flow conditions are not plug type before and after the inlet (D2) and outlet (D3) boundaries, therefore, open–open boundary condition can be chosen in present situation. A uniform radial concentration in the dispersed phase is assumed due to large length to diameter ratio. The basic general differential equation of the one dimensional ADM for fluid flow in the dimensionless form is as follows:
0.03 @C 1 @2 C @C ¼ @y Pe @X 2 @X
Ud x 100 = 0.34 m/s Uc x 100 = 0.37 m/s A x 100 = 1 m f = 0.95 s-1
input (D2)
where C is the dimensionless tracer concentration ¼ c(t)/c(0), Pe is the Peclet number ¼ uL/D, X is the dimensionless axial coordinate ¼ x/L, u is the mean linear velocity, D is the axial dispersion coefficient, c(t) is the tracer concentration at time t, and c(0) is the initial tracer concentration. A detailed analysis and solution of Eq. (6) have been given in (Levenspiel and Smith, 1957; Levenspiel, 1999). A RTD analysis software package ‘‘RTD’’ developed by IAEA (2004) was used for modeling in the present investigations. Two point measurements methodology has been adopted for this model in this software package. Fig. 2 shows typical normalized RTD curves obtained at the input (D2) and output (D3) with model output response of the dispersed phase in the pulsed sieve plate extraction column in response to an instantaneous pulse injection at (D1).
E (t)
0.02 0.015 0.01
Model output
Experimental output (D3)
0.005 0 0
100
200
300
400 500 Time (s)
600
700
800
Fig. 2. Typical normalized RTD curves at the input (D2) and output (D3) with model output response of dispersed phase in the pulsed sieve plate extraction column.
200
45
100
190 180 170
Peclet Number
Experimental MRT (s)
(6)
160 150 140 130
Ud x 100 (m/s)
120
0.34 0.44
80
0.34 0.44
Uc x 100 (m/s)
A x 100 (m)
40
1.00 1.00
35
0.37 0.47
70
30
60 25 50 20
40
Uc x 100 A x 100 (m/s) (m) 0.37 0.47
90
Ud x 100 (m/s)
15
30
1.00 1.00
110
20 0.5
0.7
0.9
1.1
1.3
1.5
1.7
1.9
2.1
Peclet Number
0.025
10 0.6
0.8
1
Pulsation frequency (s-1)
1.2
1.4
1.6
1.8
2
Pulsation frequency (s-1)
2.20 0.39
Ud x 100 (m/s)
0.37 0.35 0.33 0.31 0.29
Ud x 100 (m/s)
0.27
0.34 0.44
Uc x 100 (m/s)
A x 100 (m)
0.37 0.47
1.00 1.00
Slip velocity (m/s)
Dispersed phase holdup
2.10
0.34 0.44
2.00
Uc x 100 (m/s)
A x 100 (m)
0.37 0.47
1.00 1.00
1.90 1.80 1.70 1.60 1.50 1.40
0.25 0.5
0.7
0.9
1.1
1.3
1.5
1.7 -1
Pulsation frequency (s )
1.9
2.1
0.5
0.7
0.9
1.1
1.3
1.5
1.7
1.9
2.1
Pulsation frequency (s-1)
Fig. 3. (a) Effect of pulsation frequency on the MRT of dispersed phase; (b) effect of pulsation frequency on the Peclet number of the dispersed phase; (c) effect of pulsation frequency on the dispersed phase holdup; (d) effect of pulsation frequency on the slip velocity of dispersed phase.
ARTICLE IN PRESS 1252
Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
b 60
36 215 34
Ud x 100 (m/s)
32
0.34 0.44
175
Peclet Number
Experimental MRT (s)
195
155 135 Ud x 100 (m/s)
115
Uc x 100 (m/s)
0.34 0.44
95
0.37 0.47
c
0.9
1.1 1.3 Pulsation amplitude x 100 (m)
55
1.12 1.12
50
30
45
28
40
26
35
24
30
1.12 1.12
22
25
1.5
20
20
1.7
0.7
0.9
1.1
1.3
1.5
1.7
Pulsation amplitude x 100 (m)
d
0.49
2.30
Slip velocity (m/s)
0.44 Dispersed phase holdup
0.37 0.47
f (s-1)
f (s-1)
75 0.7
Uc x 100 (m/s)
Peclet Number
a
0.39
0.34 Ud x 100 (m/s)
0.29
0.34 0.44
Uc x 100 (m/s) 0.37 0.47
f (s-1)
2.20
Ud x 100 (m/s)
2.10
0.34 0.44
2.00
Uc x 100 (m/s) 0.37 0.47
f (s-1) 1.12 1.12
1.90 1.80 1.70 1.60 1.50
1.12 1.12
1.40 1.30
0.24 0.7
0.9
1.1
1.3
1.5
0.7
1.7
0.9
1.1
1.3
1.5
1.7
Pulsation amplitude x 100 (m)
Pulsation amplitude x 100 (m)
Fig. 4. (a) Effect of pulsation amplitude on the MRT of dispersed phase; (b) effect of pulsation amplitude on the Peclet number of the dispersed phase; (c) effect of pulsation amplitude on the dispersed phase holdup; (d) effect of pulsation amplitude on the slip velocity of the dispersed phase.
3. Results and discussion
Model Mean Residence Time (s)
Fig. 3(a–d) shows the effect of pulsation frequency on the MRT, axial mixing, holdup and slip velocity of the dispersed phase for two constant sets of Uc, Ud and pulsation amplitude. It has been observed that increase in pulsation frequency increases the MRT of dispersed phase until it approaches a maximum asymptotic value. A similar trend in the axial mixing of the dispersed phase has been observed as Peclet number decreases with increase in pulsation frequency and then approaches a minimum value Fig. 3(b). The increasing trend of MRT of dispersed phase is qualitatively in agreement with those of Srinikethan et al. (1987) who have carried out this kind of investigations on a 15 cm diameter reciprocating plate column using sodium chloride solution as tracer and conductivity cell as detector. However, an asymptotic trend in MRT of the dispersed phase after achieving a maximum value has been seen in present investigations by covering a wider range of operating conditions. In fact, increase in pulsation frequency increases the droplet population density inside the column leading to an increase in dispersed phase holdup. This phenomenon continues till a maximum holdup is reached corresponding to a minimum drop size obtainable under the available operating conditions. At this stage, the column is densely packed with droplets and no further droplets can be accommodated Fig. 3(c). Similar is the reason for the behavior of MRT curve as it is directly proportional to the holdup in the
250 230 210 190 170 150 130 110 90 70 50 50
70
90 110 130 150 170 190 210 Experimental Mean Residence Time (s)
230
250
Fig. 5. Comparison of experimental and model MRTs.
present scenario as can be seen from Eq. (2). The slip velocity of dispersed phase decreases with increase in pulsation frequency until a minimum value is obtained beyond which no decrease has been observed as can be seen from Fig. 3(d). This is due to the fact that increase in droplet population density with increase in pulsation frequency leads to the situation where velocity fields around the droplets interferes with each other. Also, decrease in
ARTICLE IN PRESS Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
the size of a droplet increases the drag on this droplet and hence decreases its velocity to move counter currently through the continuous phase. The slip velocity of dispersed phase remains decreasing till a minimum drop size has been obtained under the available operating conditions. Fig. 4(a–d) shows the effect of pulsation amplitude on the MRT, axial mixing, holdup and slip velocity of the dispersed phase when Uc, Ud and pulsation frequency are not changed. Increase in pulsation amplitude also leads to an increase in droplet population density inside the column, therefore, similar trends in MRT, axial mixing, holdup and slip velocity have been observed and phenomena can be explained on the basis already described above. A comparison of experimental and model MRTs of various RTD experiments carried out during these investigations has been given in Fig. 5 which shows a good agreement between experimental and model MRTs. 4. Conclusions (a) Axial mixing and holdup of dispersed phase in a pulsed sieve plate extraction column increases with increase in pulsation frequency and amplitude until a maximum asymptotic value is achieved. (b) Slip velocity of dispersed phase in a pulsed sieve plate extraction column decreases with increase in pulsation frequency and amplitude until it approaches a minimum value. (c) The model estimated MRTs are in good agreement with experimentally measured MRTs which shows that the axial dispersion model is a suitable model to describe the hydrodynamics of dispersed phase in pulsed sieve plate extractions column. (d) Short lived and low energy radiotracers are excellent tools to study the hydrodynamics of liquid–liquid extraction pulsed columns due to their ease of handling and providing on-line information regarding the system hydrodynamics.
Acknowledgments The authors are grateful to the Higher Education Commission [HEC] for financial support in accomplishment of this study. The authors are greatly indebted to the International Atomic Energy Agency (IAEA) for providing RTD analysis software package. The cooperation and technical assistance extended by Pakistan Institute of Nuclear Science and Technology [PINSTECH] and
1253
Pakistan Institute of Engineering and Applied Sciences [PIEAS] is thankfully acknowledged. References Arthayukti, W., Muratet, G., Angelino, H., 1976. Longitudinal mixing in the dispersed phase in pulsed perforated-plate columns. Chem. Eng. Sci. 31 (12), 1193–1197. Danckwerts, P.V., 1953. Continuous flow systems, distribution of residence times. Chem. Eng. Sci. 2 (1), 1–13. Farooq, M., Khan, I.H., Ud Din, G., Gul, S., Palige, J., Dobrowolski, A., 2003. Radiotracer investigations of municipal sewage treatment stations. Nukleonika 48 (1), 57–61. Geankoplis, C.J., Sapp, J.B., Arnold, F.C., Marroquin, G., 1982. Axial dispersion coefficients of the continuous phase in liquid–liquid spray towers. Ind. Eng. Chem. Fundam. 21 (3), 306–311. Hafez, M.M., Baird, M.H.I., Nirdosh, I., 1979. Flooding and axial dispersion in reciprocating plate extraction column. Can. J. Chem. Eng. 57, 150–158. Hazlebeck, D.E., Geankoplis, C.J., 1963. Axial dispersion in a spray-type extraction tower. Ind. Eng. Chem. Fundam. 2 (4), 310–315. Henton, J.E., Cavers, S.D., 1970. Continuous-phase axial dispersion in liquid–liquid spray towers. Ind. Eng. Chem. Fundam. 9 (3), 384–392. International Atomic Energy Agency, 2004. Integration of tracing with computational fluid dynamics for industrial process investigation, Tecdoc-1412. IAEA, Vienna, Austria. Kim, H.S., Shin, M.S., Jang, D.S., Jung, S.H., Jin, J.H., 2005. Study of flow characteristics in a secondary clarifier by numerical simulation and radioisotope tracer technique. Appl. Radiat. Isot. 63 (4), 519–526. Kim, S.D., Baird, M.H.I., 1976a. Axial dispersion in a reciprocating plate extraction column. Can. J. Chem. Eng. 54, 81–89. Kim, S.D., Baird, M.H.I., 1976b. Effect of hole size on hydrodynamics of a reciprocating perforated plate extraction column. Can. J. Chem. Eng. 54, 235–237. Kumar, A., Hartland, S., 1989. Prediction of continuous-phase axial mixing coefficients in pulsed perforated-plate extraction columns. Ind. Eng. Chem. Res. 28 (10), 1507–1513. Levenspiel, O., 1999. Chemical Reaction Engineering. Wiley, New York. Levenspiel, O., Smith, W.K., 1957. Notes on the diffusion-type model for the longitudinal mixing of fluids in flow. Chem. Eng. Sci. 6 (4–5), 227–235. Li, N.N., Ziegler, E.N., 1967. Effect of axial mixing on mass transfer in extraction columns. Ind. Eng. Chem. 59 (3), 30–36. Pant, H.J., Yelgoankar, V.N., 2002. Radiotracer investigations in aniline production reactors. Appl. Radiat. Isot. 57 (3), 319–325. Pant, H.J., Saroha, A.K., Nigam, K.D.P., 2000. Measurement of liquid holdup and axial dispersion in trickle bed reactors using radiotracer technique. Nukleonika 45 (4), 235–241. Parthasarathy, P., Sriniketan, G., Srinivas, N.S., Varma, Y.B.G., 1984. Axial mixing of continuous phase in reciprocating plate columns. Chem. Eng. Sci. 39 (6), 987–995. Srinikethan, G., Prabhakar, A., Varma, Y.B.G., 1987. Axial dispersion in plate-pulsed columns. Bioprocess Eng. 2 (4), 161–168. Thyn, J., Zitny, R., 2002. Analysis and diagnostics of industrial processes by radiotracers and radioisotope sealed sources-II. Faculty of Mechanical Engineering, Department of Process Engineering, Praha, CTU. Thyn, J., Zitny, R., Kluson, J., Cechak, T., 2000. Analysis and diagnostics of industrial processes by radiotracers and radioisotope sealed sources-I. Faculty of Mechanical Engineering, Department of Process Engineering, Praha, CTU. Venkatanarasaiah, D., Varma, Y.B.G., 1998. Dispersed phase holdup and mass transfer in liquid pulsed column. Bioprocess Eng. 18 (2), 119–126.
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Study of axial mixing, holdup and slip velocity of dispersed phase in a pulsed sieve plate extraction column using radiotracer technique Ghiyas Ud Din a,c,, Imran Rafiq Chughtai b, Mansoor Hameed Inayat b, Iqbal Hussain Khan c a
Department of Nuclear Engineering, Pakistan Institute of Engineering and Applied Sciences [PIEAS], P.O Nilore, Islamabad, Pakistan Department of Chemical and Materials Engineering, Pakistan Institute of Engineering and Applied Sciences [PIEAS], P.O Nilore, Islamabad, Pakistan c Isotope Application Division, Pakistan Institute of Nuclear Science and Technology [PINSTECH], P.O Nilore, Islamabad, Pakistan b
a r t i c l e in f o
Keywords: Liquid–liquid extraction Pulsed sieve plate column Residence time distribution (RTD) Axial mixing Holdup Slip velocity Radiotracer
a b s t r a c t Axial mixing, holdup and slip velocity of dispersed phase which are parameters of fundamental importance in the design and operation of liquid–liquid extraction pulsed sieve plate columns have been investigated. Experiments for residence time distribution (RTD) analysis have been carried out for a range of pulsation frequency and amplitude in a liquid–liquid extraction pulsed sieve plate column with water as dispersed and kerosene as continuous phase using radiotracer technique. The column was operated in emulsion region and 99mTc in the form of sodium pertechnetate eluted from a 99Mo/99mTc generator was used to trace the dispersed phase. Axial dispersed plug flow model with open–open boundary condition and two points measurement method was used to simulate the hydrodynamics of dispersed phase. It has been observed that the axial mixing and holdup of dispersed phase increases with increase in pulsation frequency and amplitude until a maximum value is achieved while slip velocity decreases with increase in pulsation frequency and amplitude until it approaches a minimum value. Short lived and low energy radiotracer 99mTc in the form of sodium pertechnetate was found to be a good water tracer to study the hydrodynamics of a liquid–liquid extraction pulsed sieve plate column operating with two immiscible liquids, water and kerosene. Axial dispersed plug flow model with open–open boundary condition was found to be a suitable model to describe the hydrodynamics of dispersed phase in the pulsed sieve plate extraction column. & 2009 Elsevier Ltd. All rights reserved.
1. Introduction Liquid–liquid extraction is a process of separation of constituents of a liquid phase by contacting it with another immiscible liquid phase. Petroleum, nuclear, chemical, metallurgical, pharmaceutical, food processing and bio-processing industries are the major beneficiaries of this technology. Column type contactors with agitation are very famous liquid–liquid extraction equipments as they offer large interfacial area, high mass transfer coefficient, high turbulence and minimum radial gradients. Counter current movement of phases in these type of columns provide high concentration gradients for efficient mass transfer but axial mixing in both phases lowers the process efficiency by lowering solute concentration gradients. It has been reported that axial mixing in extraction columns lowers the process efficiency
Corresponding author at: Department of Nuclear Engineering, Pakistan Institute of Engineering and Applied Sciences [PIEAS], P.O Nilore, Islamabad, Pakistan. Tel.: +92 51 2207381x3332; fax: +92 51 2208070. E-mail addresses: [email protected], [email protected] ( Ghiyas Ud Din).
0969-8043/$ - see front matter & 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2009.02.024
as much as 30% (Li and Ziegler, 1967). The major sources of axial mixing in the extraction columns are geometrical and operating parameters. Therefore, the presence of axial mixing in such kind of equipments is unavoidable and needs special care. A simple approach to represent the hydrodynamics of phases in these kind of columns is the axial dispersion model (ADM) (Levenspiel and Smith, 1957; Levenspiel, 1999) and the concept of residence time distribution (RTD) analysis is an important method for the estimation of axial dispersion in chemical reactors (Danckwerts, 1953). The holdup and slip velocity of dispersed phase are other parameters of fundamental importance that need to be focused in the design and operation of pulsed extraction columns. Axial mixing in the continuous phase of liquid–liquid extraction columns remained a major focus of many studies. These include studies on spray towers (Hazlebeck and Geankoplis, 1963; Henton and Cavers, 1970; Geankoplis et al., 1982), reciprocating plate extraction columns (Kim and Baird, 1976a, b; Hafez et al., 1979; Parthasarathy et al., 1984) and pulsed sieve plate extraction columns (Kumar and Hartland, 1989). Axial mixing in the dispersed phase of liquid–liquid extraction columns has either been assumed negligible or it has not been estimated. Only a few studies can be found regarding axial dispersion in the dispersed
ARTICLE IN PRESS Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
phase (Srinikethan et al., 1987) with conventional methods in which a non-radioactive tracer is injected into the system and measurements were made by a conductivity probe. The holdup and slip velocity of the dispersed phase have also been determined by arresting the flow of phases and measuring the fractional dispersed phase volume in the extraction column (Venkatanarasaiah and Varma, 1998). These experimental ap-
1249
proaches present some disadvantages including low sensitivity, requirement of phase separation before measurement and poor statistics. Most important of all is that one needs to shutdown the plant to obtain the holdup fraction at each operating parameter resulting in a high plant shutdown. Radiotracers are being used for industrial process optimization and trouble shooting from decades. They offer state of the art
Light phase outlet
Injection port D1
D2
To collection vessel
Balance leg From heavy phase feed vessel via metering pump
To collection vessel
Heavy phase inlet
Ratemeters
Data Acquisition System D3
Light phase inlet
Interface level Heavy phase outlet
From light phase feed vessel via metering pump
Pulse unit
Fig. 1. Schematic diagram of pulsed sieve plate extraction column.
ARTICLE IN PRESS 1250
Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
technology with high sensitivity, on-line measurement without disturbing the plant operation, better statistics, high benefit to cost ratio and possibility of getting information from remote locations of process equipment. A variety of industrial process investigations have been carried out using the power of radiotracer technology (Thyn et al., 2000; Thyn and Zitny, 2002; Pant and Yelgoankar, 2002; Farooq et al., 2003; Kim et al., 2005). Although numerous applications can be found in literature regarding industrial process investigations using radiotracer technology but only a few have been reported with reference to the hydrodynamics of liquid–liquid extraction columns (Arthayukti et al., 1976). The present investigations are focused to study the hydrodynamics of dispersed phase in a pulsed sieve plate extraction column using radiotracer RTD analysis. Effect of pulsation frequency and amplitude have been studied on the axial mixing, holdup and slip velocity of dispersed phase for two sets of dispersed and continuous phase superficial velocities and the system was simulated using the ADM.
2. Materials and methods The schematic diagram of pulsed sieve plate extraction column under investigation is shown in Fig. 1. The internal diameter of the column is 5 102 m and height is 2 m. Two separating chambers one at the top and the other at the bottom of the column are also part of this apparatus. The column is fitted with regularly spaced (5 102 m) sieve plates which help to increase the interfacial area between the two immiscible liquids. The column was operated counter currently with heavy phase (water) as dispersed and light phase (kerosene) as continuous phase. The kerosene is fed into the lower separating chamber with the help of a metering pump which then flows upwards through the sieve plate column to the upper separating chamber where it overflows to a collection vessel. Similarly water is fed into the top separating chamber via a metering pump from where it flows downwards through the column to the lower separating chamber and then it flows through a balance leg into a collection vessel. A pulse unit located at the base of lower separating chamber provides vertical pulses to the flowing fluids. The column was operated in the emulsion regime i.e. dispersed phase remained dispersed throughout the plate stack and no coalescence into layers occurred at the plates. A liquid–liquid interface was allowed to form at about 10 cm below the light phase inlet and this interface level was stabilized with the help of a balance leg before starting an experiment. 99m Tc in the form of sodium pertechnetate having half life of 6.02 h and gamma energy 0.14 MeV (91%) is a well known water tracer (Pant et al., 2000) but no data is available for its use in water–kerosene environment. Therefore, validation of radiotracer was carried out by thoroughly mixing about 0.1 mCi 99mTc in equal amounts of water and kerosene. Upon separation and measurements, the radiotracer was found suitable for labeling water phase in water–kerosene environment. About 0.5 mCi of 99mTc eluted from a 99Mo/99mTc generator was injected in the form of an instantaneous pulse to carry out RTD experiments for investigation of the hydrodynamics of dispersed phase (water) as per experimental plan shown in Fig. 1. The experiments were carried out for a range of pulsation frequency and amplitude as given in Table 1. The movement of radiotracer was monitored for every second with the help of lead collimated 2 in 2 in NaI(Tl) scintillation detectors mounted at D1, D2 and D3 as shown in Fig. 1. The data was acquired on-line using a multi-channel data acquisition system and stored in a computer for processing.
Table 1 Pulsed sieve plate extraction column specification and range of operating variables. Internal diameter of the column (m) Length of the column (m) Number of sieve plates Plate spacing (m) Plate diameter (m) Hole size (m) Average number of holes Average free area (%) Dispersed phase superficial velocity (m/s) Continuous phase superficial velocity (m/s) Range of pulsation frequency (s1) Range of pulsation amplitude (m)
5 102 2 38 5 102 5 102 2 103 140 per plate 25 0.34 102 and 0.44 102 0.37 102 and 0.47 102 0.7–2.0 0.8 102–1.6 102
The tracer data from detectors D2 (column inlet) and D3 (column outlet) was corrected for background, radioactive decay and normalized. The experimental mean residence time (MRT) of the system was calculated by the difference of first moments of outlet and inlet response curves. Mathematical expression for the first moment in discrete form can be written as P t C Dt (1) First moment ¼ Pi i i i i C i Dt i where C is the tracer concentration (counts/s in present case), t is the time of measurement (s), Dt is the time interval between the two measurements (s), and i ¼ 0,1,2,3,y. Overall holdup of the phase under investigation was calculated on the basis of calculated MRT using the following relationship: Hd ¼
tQ d VR
(2)
where Hd is the dispersed phase holdup, Qd is the dispersed phase flow rate, t is the MRT, and VR is the effective reactor volume. The slip velocity of the dispersed phase averaged over the whole column was estimated from the above calculated Hd using the following equation (Venkatanarasaiah and Varma, 1998): Vs ¼
Ud Uc þ Hd ð1 Hd Þ
(3)
where Vs is the slip velocity of dispersed phase, Ud is the dispersed phase superficial velocity, Uc is the continuous phase superficial velocity. The RTD is a probability distribution function that describes the amount of time a fluid element spends inside a reactor. It helps in troubleshooting of reactors and characterizes the mixing and flow within the reactors. If an impulse of tracer is injected at the inlet of a system at time t ¼ 0 and its concentration is measured as a function of time at the outlet, then E(t) representing the probability for a tracer element to have a residence time between the time interval (t, t+dt) is defined as C ðtÞ Ei ðtÞ ¼ R 1 i 0 C i ðtÞ dt such that Z 1 Ei ðtÞ dt ¼ 1
(4)
(5)
0
where i ¼ 1,2,3,y,n, Ci(t) is the tracer concentration, Ei(t) is the RTD function. RTD models have been playing a vital role for industrial process investigations from decades. They provide macroscopic lumped sum description, which is sufficient for many engineering calculations. The plug flow is an ideal condition for the flow of
ARTICLE IN PRESS Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
phases in an extraction column but some degree of axial mixing is always inevitable. The pulsation and closely spaced sieve plates inside the column continues to re-arrange the droplets of the dispersed phase randomly. Keeping in view of these considerations, the ADM was used to study the system hydrodynamics.
1251
The flow conditions are not plug type before and after the inlet (D2) and outlet (D3) boundaries, therefore, open–open boundary condition can be chosen in present situation. A uniform radial concentration in the dispersed phase is assumed due to large length to diameter ratio. The basic general differential equation of the one dimensional ADM for fluid flow in the dimensionless form is as follows:
0.03 @C 1 @2 C @C ¼ @y Pe @X 2 @X
Ud x 100 = 0.34 m/s Uc x 100 = 0.37 m/s A x 100 = 1 m f = 0.95 s-1
input (D2)
where C is the dimensionless tracer concentration ¼ c(t)/c(0), Pe is the Peclet number ¼ uL/D, X is the dimensionless axial coordinate ¼ x/L, u is the mean linear velocity, D is the axial dispersion coefficient, c(t) is the tracer concentration at time t, and c(0) is the initial tracer concentration. A detailed analysis and solution of Eq. (6) have been given in (Levenspiel and Smith, 1957; Levenspiel, 1999). A RTD analysis software package ‘‘RTD’’ developed by IAEA (2004) was used for modeling in the present investigations. Two point measurements methodology has been adopted for this model in this software package. Fig. 2 shows typical normalized RTD curves obtained at the input (D2) and output (D3) with model output response of the dispersed phase in the pulsed sieve plate extraction column in response to an instantaneous pulse injection at (D1).
E (t)
0.02 0.015 0.01
Model output
Experimental output (D3)
0.005 0 0
100
200
300
400 500 Time (s)
600
700
800
Fig. 2. Typical normalized RTD curves at the input (D2) and output (D3) with model output response of dispersed phase in the pulsed sieve plate extraction column.
200
45
100
190 180 170
Peclet Number
Experimental MRT (s)
(6)
160 150 140 130
Ud x 100 (m/s)
120
0.34 0.44
80
0.34 0.44
Uc x 100 (m/s)
A x 100 (m)
40
1.00 1.00
35
0.37 0.47
70
30
60 25 50 20
40
Uc x 100 A x 100 (m/s) (m) 0.37 0.47
90
Ud x 100 (m/s)
15
30
1.00 1.00
110
20 0.5
0.7
0.9
1.1
1.3
1.5
1.7
1.9
2.1
Peclet Number
0.025
10 0.6
0.8
1
Pulsation frequency (s-1)
1.2
1.4
1.6
1.8
2
Pulsation frequency (s-1)
2.20 0.39
Ud x 100 (m/s)
0.37 0.35 0.33 0.31 0.29
Ud x 100 (m/s)
0.27
0.34 0.44
Uc x 100 (m/s)
A x 100 (m)
0.37 0.47
1.00 1.00
Slip velocity (m/s)
Dispersed phase holdup
2.10
0.34 0.44
2.00
Uc x 100 (m/s)
A x 100 (m)
0.37 0.47
1.00 1.00
1.90 1.80 1.70 1.60 1.50 1.40
0.25 0.5
0.7
0.9
1.1
1.3
1.5
1.7 -1
Pulsation frequency (s )
1.9
2.1
0.5
0.7
0.9
1.1
1.3
1.5
1.7
1.9
2.1
Pulsation frequency (s-1)
Fig. 3. (a) Effect of pulsation frequency on the MRT of dispersed phase; (b) effect of pulsation frequency on the Peclet number of the dispersed phase; (c) effect of pulsation frequency on the dispersed phase holdup; (d) effect of pulsation frequency on the slip velocity of dispersed phase.
ARTICLE IN PRESS 1252
Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
b 60
36 215 34
Ud x 100 (m/s)
32
0.34 0.44
175
Peclet Number
Experimental MRT (s)
195
155 135 Ud x 100 (m/s)
115
Uc x 100 (m/s)
0.34 0.44
95
0.37 0.47
c
0.9
1.1 1.3 Pulsation amplitude x 100 (m)
55
1.12 1.12
50
30
45
28
40
26
35
24
30
1.12 1.12
22
25
1.5
20
20
1.7
0.7
0.9
1.1
1.3
1.5
1.7
Pulsation amplitude x 100 (m)
d
0.49
2.30
Slip velocity (m/s)
0.44 Dispersed phase holdup
0.37 0.47
f (s-1)
f (s-1)
75 0.7
Uc x 100 (m/s)
Peclet Number
a
0.39
0.34 Ud x 100 (m/s)
0.29
0.34 0.44
Uc x 100 (m/s) 0.37 0.47
f (s-1)
2.20
Ud x 100 (m/s)
2.10
0.34 0.44
2.00
Uc x 100 (m/s) 0.37 0.47
f (s-1) 1.12 1.12
1.90 1.80 1.70 1.60 1.50
1.12 1.12
1.40 1.30
0.24 0.7
0.9
1.1
1.3
1.5
0.7
1.7
0.9
1.1
1.3
1.5
1.7
Pulsation amplitude x 100 (m)
Pulsation amplitude x 100 (m)
Fig. 4. (a) Effect of pulsation amplitude on the MRT of dispersed phase; (b) effect of pulsation amplitude on the Peclet number of the dispersed phase; (c) effect of pulsation amplitude on the dispersed phase holdup; (d) effect of pulsation amplitude on the slip velocity of the dispersed phase.
3. Results and discussion
Model Mean Residence Time (s)
Fig. 3(a–d) shows the effect of pulsation frequency on the MRT, axial mixing, holdup and slip velocity of the dispersed phase for two constant sets of Uc, Ud and pulsation amplitude. It has been observed that increase in pulsation frequency increases the MRT of dispersed phase until it approaches a maximum asymptotic value. A similar trend in the axial mixing of the dispersed phase has been observed as Peclet number decreases with increase in pulsation frequency and then approaches a minimum value Fig. 3(b). The increasing trend of MRT of dispersed phase is qualitatively in agreement with those of Srinikethan et al. (1987) who have carried out this kind of investigations on a 15 cm diameter reciprocating plate column using sodium chloride solution as tracer and conductivity cell as detector. However, an asymptotic trend in MRT of the dispersed phase after achieving a maximum value has been seen in present investigations by covering a wider range of operating conditions. In fact, increase in pulsation frequency increases the droplet population density inside the column leading to an increase in dispersed phase holdup. This phenomenon continues till a maximum holdup is reached corresponding to a minimum drop size obtainable under the available operating conditions. At this stage, the column is densely packed with droplets and no further droplets can be accommodated Fig. 3(c). Similar is the reason for the behavior of MRT curve as it is directly proportional to the holdup in the
250 230 210 190 170 150 130 110 90 70 50 50
70
90 110 130 150 170 190 210 Experimental Mean Residence Time (s)
230
250
Fig. 5. Comparison of experimental and model MRTs.
present scenario as can be seen from Eq. (2). The slip velocity of dispersed phase decreases with increase in pulsation frequency until a minimum value is obtained beyond which no decrease has been observed as can be seen from Fig. 3(d). This is due to the fact that increase in droplet population density with increase in pulsation frequency leads to the situation where velocity fields around the droplets interferes with each other. Also, decrease in
ARTICLE IN PRESS Ghiyas Ud Din et al. / Applied Radiation and Isotopes 67 (2009) 1248–1253
the size of a droplet increases the drag on this droplet and hence decreases its velocity to move counter currently through the continuous phase. The slip velocity of dispersed phase remains decreasing till a minimum drop size has been obtained under the available operating conditions. Fig. 4(a–d) shows the effect of pulsation amplitude on the MRT, axial mixing, holdup and slip velocity of the dispersed phase when Uc, Ud and pulsation frequency are not changed. Increase in pulsation amplitude also leads to an increase in droplet population density inside the column, therefore, similar trends in MRT, axial mixing, holdup and slip velocity have been observed and phenomena can be explained on the basis already described above. A comparison of experimental and model MRTs of various RTD experiments carried out during these investigations has been given in Fig. 5 which shows a good agreement between experimental and model MRTs. 4. Conclusions (a) Axial mixing and holdup of dispersed phase in a pulsed sieve plate extraction column increases with increase in pulsation frequency and amplitude until a maximum asymptotic value is achieved. (b) Slip velocity of dispersed phase in a pulsed sieve plate extraction column decreases with increase in pulsation frequency and amplitude until it approaches a minimum value. (c) The model estimated MRTs are in good agreement with experimentally measured MRTs which shows that the axial dispersion model is a suitable model to describe the hydrodynamics of dispersed phase in pulsed sieve plate extractions column. (d) Short lived and low energy radiotracers are excellent tools to study the hydrodynamics of liquid–liquid extraction pulsed columns due to their ease of handling and providing on-line information regarding the system hydrodynamics.
Acknowledgments The authors are grateful to the Higher Education Commission [HEC] for financial support in accomplishment of this study. The authors are greatly indebted to the International Atomic Energy Agency (IAEA) for providing RTD analysis software package. The cooperation and technical assistance extended by Pakistan Institute of Nuclear Science and Technology [PINSTECH] and
1253
Pakistan Institute of Engineering and Applied Sciences [PIEAS] is thankfully acknowledged. References Arthayukti, W., Muratet, G., Angelino, H., 1976. Longitudinal mixing in the dispersed phase in pulsed perforated-plate columns. Chem. Eng. Sci. 31 (12), 1193–1197. Danckwerts, P.V., 1953. Continuous flow systems, distribution of residence times. Chem. Eng. Sci. 2 (1), 1–13. Farooq, M., Khan, I.H., Ud Din, G., Gul, S., Palige, J., Dobrowolski, A., 2003. Radiotracer investigations of municipal sewage treatment stations. Nukleonika 48 (1), 57–61. Geankoplis, C.J., Sapp, J.B., Arnold, F.C., Marroquin, G., 1982. Axial dispersion coefficients of the continuous phase in liquid–liquid spray towers. Ind. Eng. Chem. Fundam. 21 (3), 306–311. Hafez, M.M., Baird, M.H.I., Nirdosh, I., 1979. Flooding and axial dispersion in reciprocating plate extraction column. Can. J. Chem. Eng. 57, 150–158. Hazlebeck, D.E., Geankoplis, C.J., 1963. Axial dispersion in a spray-type extraction tower. Ind. Eng. Chem. Fundam. 2 (4), 310–315. Henton, J.E., Cavers, S.D., 1970. Continuous-phase axial dispersion in liquid–liquid spray towers. Ind. Eng. Chem. Fundam. 9 (3), 384–392. International Atomic Energy Agency, 2004. Integration of tracing with computational fluid dynamics for industrial process investigation, Tecdoc-1412. IAEA, Vienna, Austria. Kim, H.S., Shin, M.S., Jang, D.S., Jung, S.H., Jin, J.H., 2005. Study of flow characteristics in a secondary clarifier by numerical simulation and radioisotope tracer technique. Appl. Radiat. Isot. 63 (4), 519–526. Kim, S.D., Baird, M.H.I., 1976a. Axial dispersion in a reciprocating plate extraction column. Can. J. Chem. Eng. 54, 81–89. Kim, S.D., Baird, M.H.I., 1976b. Effect of hole size on hydrodynamics of a reciprocating perforated plate extraction column. Can. J. Chem. Eng. 54, 235–237. Kumar, A., Hartland, S., 1989. Prediction of continuous-phase axial mixing coefficients in pulsed perforated-plate extraction columns. Ind. Eng. Chem. Res. 28 (10), 1507–1513. Levenspiel, O., 1999. Chemical Reaction Engineering. Wiley, New York. Levenspiel, O., Smith, W.K., 1957. Notes on the diffusion-type model for the longitudinal mixing of fluids in flow. Chem. Eng. Sci. 6 (4–5), 227–235. Li, N.N., Ziegler, E.N., 1967. Effect of axial mixing on mass transfer in extraction columns. Ind. Eng. Chem. 59 (3), 30–36. Pant, H.J., Yelgoankar, V.N., 2002. Radiotracer investigations in aniline production reactors. Appl. Radiat. Isot. 57 (3), 319–325. Pant, H.J., Saroha, A.K., Nigam, K.D.P., 2000. Measurement of liquid holdup and axial dispersion in trickle bed reactors using radiotracer technique. Nukleonika 45 (4), 235–241. Parthasarathy, P., Sriniketan, G., Srinivas, N.S., Varma, Y.B.G., 1984. Axial mixing of continuous phase in reciprocating plate columns. Chem. Eng. Sci. 39 (6), 987–995. Srinikethan, G., Prabhakar, A., Varma, Y.B.G., 1987. Axial dispersion in plate-pulsed columns. Bioprocess Eng. 2 (4), 161–168. Thyn, J., Zitny, R., 2002. Analysis and diagnostics of industrial processes by radiotracers and radioisotope sealed sources-II. Faculty of Mechanical Engineering, Department of Process Engineering, Praha, CTU. Thyn, J., Zitny, R., Kluson, J., Cechak, T., 2000. Analysis and diagnostics of industrial processes by radiotracers and radioisotope sealed sources-I. Faculty of Mechanical Engineering, Department of Process Engineering, Praha, CTU. Venkatanarasaiah, D., Varma, Y.B.G., 1998. Dispersed phase holdup and mass transfer in liquid pulsed column. Bioprocess Eng. 18 (2), 119–126.