Coronado, Raniella Bianca Y. Reyes, Noelle Ivonette Z. Zapata, Rosette Anne Lea T.
5ChE-D
PROBLEM B1 PRESSURE DROP AND FLOODING IN A PACKED COLUMN
I.
Abstract
Pressure drop is one of the indicators i ndicators of when flooding will occur in a packed column. It is correlated with many aspects such as liquid hold up, void fractions, and the flow rates of the gas and liquid being processed. In this experiment, the effect of liquid accumulation, the packing factor, the gas and liquid flow rates on the pressure drop along a gas absorption column was observed and discussed. This was done by measuring the height of fluid in a manometer at different gas and liquid flow rates. The experimental pressure drop is then calculated using appropriate equations discussed in section VIII of this paper. For dry packed column, the experimental results was analyzed by comparing these data to the computed values using the Ergun Equation and Robbins Equation. Meanwhile, experimental results for the pressure drop in an irrigated packed column were compared with the values calculated using the latter. Upon comparison, the results from the Ergun equation, relative to those calculated using the Robbins Equation, deviated significantly from the experimental results. On the other hand, for an irrigated column and constant liquid flow rate, a general trend wherein a sudden increase in pressure drop was observed when the gas flow rate is further increased. The packing factor and the void fraction was also c omputed. II.
Objectives
This experiment aims to determine the void fractions of the packed beds, the effects of liquid holdups on the pressure drop of the column, to familiarize with the parts of the equipment and to define the packing factor experimentally through the use of the flooding velocity calculations.
III.
References
[1] McCabe, W., Smith, J. & Harriott, P. (1993). Unit Operations of Chemical Engineering (5th ed.). Singapore: McGraw-Hill Book Co. [2] Geankoplis, C. (2003). Transport Process and Unit Operation (4th ed.). Upper Saddle River, NJ: Prentice Hall [3] McCabe, W. L., Smith, J. C., & Harriott, P. (2005). Unit Operations of Chemical (7th ed.). New York: McGraw-Hill. Engineering (7th [4] Arachchige, U. & Melaaen, M. (2012). Selection of Packing Material for Gas Absorption. European Journal of Scientific Research, 87: 87: 117-126. [5] Perfett, L. & Fisher, T. (1996). Gas Absorption Column. Retrieved from http://chem.engr.utc.edu/Webres/435F/ABS_COL/abs_col.html [6] Fahien, R. (1983). Fundamentals of Transport Phenomena Phenomena. New York: McGraw-Hill, Inc. [7] Sharma, K. (2007). Principles of Mass Transfer . New Delhi, India: Asoke K. Ghosh, Prentice Hall of India Private Limited, M-97, Connaught Circus, New Delhi-110001
[8] NPTEL (n.d.). Module 4: Absorption Lecture no. 1. Retrieved from http://nptel.ac.in/courses/103103035/module4/lec1.pdf [9] Column Diameter and Pressure Drop (n.d.) Retrieved from http://www.separationprocesses.com/Absorption/GA_Chp04a.htm#TopPage [10] Shulman, H. L., Ullrich, C. F. and Wells, N. (1955), Performance of packed columns. I. Total, static, and operating holdups. AIChE J., 1: 247 – 253. doi:10.1002/aic.690010219 [11] Zakeri, A., Einbu, A., Oi, L. & Svendsen, H. (2009). Liquid Hold-up and Pressure Retrieved from Drop in Mellapak 2X . http://ena.chemeng.ntnu.no/TrondheimJointAbsorptionSeminar/Zakeri%20_Trondheim15060 9.pdf [12] Elgin, J. C. & Weiss, F. B. (1939), Liquid holdup and flooding in packed towers. Ind. Eng. Chem, 31 (4): 435 – 445. doi:10.1021/ie50352a010 [13] Cussler, E. L. (1984). Diffusion: Mass Transfer in Fluid Systems . New York, USA: Cambridge University Press. [14] Brenner, H. (1989). Gas-Liquid-Solid Fluidization Engineering. USA: Butterworth Publishers [15] Yuan, J., Yan, L., & Hlavka, D. (2009). Flow through packed beds. Retrieved from http://www.me.rochester.edu/courses/ME241.gans/PackedBeds(11).pdf
IV.
Equipment/Materials The equipment used in this experiment was the Gas-Liquid Absorption Column.
Rachig rings
V.
Theory
Materials, in chemical engineering, are transformed or separated into useful products by means of a variety of industrial and chemical processes [1]. Gas absorption is one of the well-known unit processes and is commonly utilized in the food industry. In this process, a gas mixture consisting mainly of an inert gas and a soluble gas is made in contact with a liquid acting as a solvent to separate the gas mixture [2]. Absorption technology is widely applied in the removal of H 2S and CO2 from natural or synthesis gas by absorption in amines or alkaline salt solutions. One common instrument used in this process is the packed tower which is shown in Figure 1 below. This apparatus entails a cylindrical column or tower outfitted with a liquid inlet and gas outlet at the top and a gas inlet and liquid outlet at the bottom, and a collective mass of inert solid shapes known as tower packings [3].
Figure 1. Packed Tower
Several packing types available for gas absorption can be utilized depending on different parameters like flow rate, temperature, pressure, etc. [4]. The shape of the packing inhibits it from being to compact thus, enhancing the bed porosity [3]. Liquid will flow through the packed column and dispense uniformly over the packing surface in an ideal operation. Gas will enter the tower from below the packed section and move upward countercurrent to the flow of the liquid through the small spaces between the packing material. Efficient mass transfer can be obtained by the large amount of intimate contact between the liquid and gas streams [5]. Pressure drop can be described as the pressure loss due to the frictional resistance of the components the gas touches, velocity variations, back flows and eddy formation [2]. A linear relationship between the pressure drop and the gas velocity can be observed in a constant diameter packed column. The pressure drop increases as we increase the gas velocity at a constant liquid flow rate. In addition, the pressure drop is larger when we increase the gas velocity at a higher liquid rate [6].
For this experiment, numerous equations can be used. First is the Ergun equation formulated by a Turkish chemical engineer, Sabri Ergun. This equation is used to get the pressure drop across a certain length of packing. ∆
=
(−)
+
.7 (−)
(1)
Two equations coming from the first and second term of the Ergun equation were also formulated. The Blake-Kozeny equation derived from the first term is applicable for laminar regions and the Burke-Plummer equation which is the second term is relevant for turbulent regions. Both equations are only valid for void fractions less than 0.5 [2]. Later on, Fahien and Schriver adapted the Ergun equation and modified it for broader values of porosity and Reynolds number resulting to the following equations for laminar (2), turbulent (3) and intermediate (4) regions given below:
ϕ = ϕ =
3
(2)
(−). 2 (−).
+
.7/,
(3)
(−).
ϕ = ϕ + (1 )ϕ
(4)
Where: (),
= ϕ=
.
∆
(−)
(5)
(6)
Z- height of packed column A correlation of the pressure drop for wet packing was also made my Leva and later on, Robbins developed a pressure correlation using the similar approach by Leva:
∆ = ∆ + ∆ ∆ = 3 2 10 = 986 (0.05 ).
∆ = 0.4(0.00005 ).(∆ ) 2.
= . (0.05 ). (
)
(7)
The packing factor was estimated by Lobo et al (1945) to be:
=
(−)
(8)
VI.
Operating Conditions and Procedure
Before running the actual experiment, preliminary procedures were conducted starting with determination of the length of the packed beds, the diameter of the gas column and, the dimensions of the packing. After which, the sump tank was cleaned and was filled with water of about 75% of its capacity. It was also made sure that the parts of the equipment were properly checked such that the on-off switch knobs were turned off, the flow meter and drainage valves were closed and, the return line valve and pressure taps were fully opened. It was noted that all entrained liquid in the tubes connected to the pressure taps were drained. For the start-up, the second manometer in the equipment was filled with colored water to be used as substitute for mercury. The main switch, the compressor and the pump were turned on and gas was allowed to flow within the system at a rate of 140 L/min for 15 minutes, totally removing all the water from the column. The three-way glass cocks were also adjusted to ensure that the gas flowing out of the pressure taps were directed to the left manometer only. Once the start-up was completed, the gas rate was returned to 60 L/min. Differential pressures, in mmH2O, across the upper and lower packed beds were then measured by adjusting the three-way cocks connected to the pressure tap at the middle of the column. The adjustments were done slowly so as to minimize the effects of surface tension on the manometer fluid. Such was repeated with increasing gas flow rates with increments of 10 L/min until the rate reached 140 L/min. Additionally, pressure drop readings due to liquid holdups were acquired. The gas rate was first reset to 60 L/min and the liquid control valve was opened such that the liquid rate was set to 1 L/min. Periodic throttling was observed and controlled to lower the chances of sudden increase in pressure which could result to fluid overflowing. After obtaining the measured pressure drop, another trial was performed by setting new flow rates for both gas and liquid. The liquid rate was increased by an increment of 10 L/min until it reached 140 L/min, whereas the liquid flow rate was increased by an increment of 1 L/min until it achieved a flow rate of 7 L/min. After the measurements were read and noted, the equipment was prepared for shutdown. The pump was firstly turned off and the liquid in the flow meter was drained before fully closing the control valve. The gas rate was once again set to 140 L/min and was allowed to run for 15 minutes prior to the closing of the gas control valve. Subsequently, the compressor was turned off and so was the on-off switch of the equipment.
VII.
Data and Results
Presented in table 1 are the experimental data gathered from the experiment. Table 1 Difference between the manometer fluid heights at var ying gas and liquid flow rate Liquid Flow Rate Air Flow Rate 20 L/min
0 L/min
1
2
3
0.2
0.2
2
2.2
30
0.4
0.4
2.2
40
0.6
0.6
50
0.8
60
4
5
6
6.5
7
0.4
2.6
2
2
1.2
2
0.8
2.8
2.6
2.8
2.6
2
2.4
1.4
3.4
3
3.8
6.4
0.8
1.6
2.6
2
4
4.6
6.8
11.8
1
1
2
3
3.2
5.4
6.2
15.6
19.8 F
70
1.2
1.2
2.4
3.6
4.8
6.6
15.6
27.4 F
F
80
1.4
2
3.2
4.4
6.4
11.6
30.0 F
F
F
90
1.6
2.2
3.4
5
8.4
18.4
F
F
F
100
1.8
2.4
4.6
5.6
15
30.6 F
F
F
F
110
2
3.2
5.4
6.2
16.2
F
F
F
F
120
2.6
3.8
6
7.6
21
F
F
F
F
130
2.8
4.6
7.2
8.8
24.2 F
F
F
F
F
140
3
5.4
8
10.2
F
F
F
F
F
150
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
160
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
170
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
R (in cm)
Note: F- indicates flooding has started N/A – flow rate cannot be attained by the system
VIII.
Treatment of Results
The pressure drop across the whole column was calculated b y using the measurement of the manometer fluid level and the formula:
∆ = Where, ΔP = pressure drop across the column, Pa ρ = density of fluid, kg/m 3
R = height of manometer fluid, m
(8)
The results obtained was then plotted against the respective Reynolds number for packed beds at different gas flow rates by using the equation:
, =
(9)
(−)
Where, NRe,p = Reynolds number for packed beds D p = particle diameter,m vo =superficial gas velocity, m/s ρ = fluid density, kg/m 3 μ = fluid viscosity, Pa-s ε = void fraction of packing
350 300 250 a 200 P , P Δ - 150
100 50 0 50
100
150
200
250
300
350
400
450
500
NRe,p
Figure 2. Plot of pressure drop across the gas absorption column with dry packings vs. the
Reynolds number for packed beds The void fraction of the column was then calculated by using the modifications of Fahien and Schriver in the Ergun equation. Equations (2), (3), (4), (5), and (6) were used to perform calacutions. The void fraction of the column was calculated using three different gas flow rates and was averaged to obtain the experimental void fraction. From the graph, the flow rates used were 30, 70, and 110 L/min. Table 2 The calculated void fraction using the chosen flow rates Flow rate -ΔP Void Fraction 30 39.088 0.36 110 195.4402 0.4075 70 117.2641 0.3833 Average 0.3836
For operation involving liquid flow, the gas velocity was calculated using the continuity equation: = (10) To compare the behavior of the pressure drop across the column at differ ent gas and liquid flow rates, two graphs were plotted: log( ΔP/Z) vs. log(G) and ΔP vs Gf . 3.7 3.4 Dry
3.1
1 L/min ) 2.8 Z / P Δ 2.5 ( g o L 2.2
2 3 4 5
1.9
6
1.6
6.5
1.3
7 2.4
2.5
2.6
2.7
2.8
2.9
3
3.1
3.2
3.3
3.4
Log G
Figure 3. The logarithmic plot of pressure drop per unit of the column ( Δ P/Z) versus the
superficial gas mass velocity (G) 45 42 39 36 33 30 f 27 s p 24 , P 21 Δ18 15 12 9 6 3 0 1000
Dry 1 L/min 2 3 4 5 6 6.5 2000
3000
4000
5000
6000
7000
8000
7
Gf, lb/hr-ft2 0.5 Figure 5. The plot of the pressure drop, Δ P, versus the gas loading factor, G f =U t ρ g ;
wherein U t is the superficial gas velocity and ρ g is the gas density The packing factor was obtained by using flooding velocity calculations (Fig. 14-55 of the Handbook). The experimental pressure drop ─ ─ at a gas flow rate and a liquid flow rate of 60 L/min and 2 L/min, respectively, was used for the calculation.
Figure 6. Graph for flooding velocity
(Fig. 14-55 of the Handbook)
The packing factor, F p, was then calculated using the equation
(11) From figure 6, the obtained capacity factor, CP, was 0.46. The calculated F p is 689.35 ft -1. IX.
Analysis/Interpretation of Results In this experiment, two equations ─ namely Ergun-type equation and Robbins equation ─ were used to compare the results for the pressure drop across the dry packed column.
Table 3 Comparison of pressure drop in dry packed column Air Flow Rate -ΔP (mm H2O) L/min Experimental Ergun Equation
Robbins Equation
20
1.9544
1.6091
0.4469
30
3.9088
2.9521
1.0055
40
5.8632
4.6539
1.7875
50
7.8176
6.7146
2.7929
60
9.7720
9.1342
4.0218
70
11.7264
11.9127
5.4742
80
13.6808
15.0501
7.1499
90
15.6352
18.5464
9.0491
100
17.5896
22.4016
11.1718
110
19.5440
26.6158
13.5179
120
25.4072
31.1888
16.0874
130
27.3616
36.1207
18.8803
140
29.3160
41.4115
21.8967
As shown in Table 3, results obtained from all treatment depicts an increase in pressure drop as the air flow rate increases. However, a significant difference between the experimental values and those obtained from the Ergun Equation can be observed. This may be attributed to the assumptions and considerations made by Ergun in establishing the equation. Ergun’s equation is applicable for the range of flow rate ─ meaning from laminar flow to turbulent flow. He assumed that the viscous losses due to laminar flow and the kinetic losses due to turbulent flow are additive. Upon deriving the equation, he made an assumption wherein the friction factor during laminar flow mainly depends on the void fraction and is independent of the Reynolds number. On the other hand, it is dependent only on the Reynolds number and not on the void fraction when the flow is turbulent [15]. This assumption does not take into account the nature of the packing and the gas loading factor which are both taken into consideration in Robbins equation, explaining why the values calculated using this equation are closer to the experimental values. Figure 7 shows the comparison between the pressure drop among the three treatment method. 45 40 35 O30 2 H25 m m20 , P Δ - 15
Experimental Ergun Equation Robbins Equation
10 5 0 0.05
0.15
0.25
0.35
0.45
0.55
G, kg/hr-m2
Figure 7. The comparison between the values of pressure drop at a given flow rate obtained
by experimentation (blue), Ergun Equation (red), and Robbins Equation (green) For irrigated or wetted packed column, the experimental pressure drop at a given flow rate was compared with the pressure drop calculated using Robbins Equation since it takes into account the pressure drop contributed by the liquid loading or hold-up within the packings. Figure 3 and 8 shows the logarithmic plot of pressure drop per unit height of packing versus the superficial gas mass velocity. Theoretically, at low liquid flow rates and increasing gas flow rate, pressure drop increases in a similar behavior (slope) when no liquid flow is present in the column as seen in Figure 8. During this scenario, the pressure drop in an irrigated packed bed is higher than that of the dry column, but having similar slopes, because the liquid consumes part of the voids along the column, hence decreasing the portion where
gas can pass through. As the liquid flow rate increases, drastic and sudden increase in pressure drop can be observed at high gas flow rates. At this point, liquid is starting to load or accumulate inside the voids, hence resulting to the sudden increase in pressure drop. Generally, such trend can also be observed in the experimental results depicted in Figure 3. Some of the overlap among lower liquid flow rates and higher liquid flow rat es may be due to the effects of surface tension inside the manometer. Also, the system is not ideal, hence it may deviate from the theoretical or expected outcome. 7.8 6.8 Dry 5.8
1 L/min
4.8
2
) Z / 3.8 P Δ ( g 2.8 o L
3 4 5
1.8
6
0.8
6.5 -0.2 -1.2
2.4
2.6
2.8
3
3.2
3.4
7
Log(G)
Figure 8. The logarithmic plot of the pressure drop per unit of height packing calculated
using the Robbins Equation versus the superficial gas mass velocity Looking back at figure 5, the loading point can for the gas absorption column can be estimated. Loading point is the point wherein liquid starts to be trapped or accumulated inside the voids and spaces present, resulting to the “drowning” of the column. For a given liquid flow rate, drowning of the column will result to flooding when the gas flow rate is further increased. Loading point is signified by a sudden and drastic increase in pressure drop (see figure 9). Also, as seen in the figure, liquid flow rates below 4 L/min do not intersect the “loading line.”
À
45 42 39 36 33 30 f 27 s p 24 , P 21 Δ18 15 12 9 6 3 0 1000
Dry 1 L/min 2 3 4 5 6 6.5 2000
3000
4000
5000
6000
7000
8000
7
Gf, lb/hr-ft2
Figure 9. The black line intersects the loading point for each given liquid flow rate.
X.
Answers to Questions
What are the characteristics that a packing should have for it to be employed in mass transfer operation? Packings in mass transfer unit operations are designed to increase the interfacial area of contact between the two phases (gas and liquid) as well as enhance their flow [7]. Mass transfer operations account for the interfacial area of contact, void volume, fouling resistance, etc. of the packings. Thus, packing materials should have large interfacial area of contact for a larger pressure drop and high void of volume to maintain a low pressure drop, high fouling resistance, good mechanical strength and uniform void spaces for uniform flow of streams [8].
Explain the mechanism of gas flow through a packed bed with liquid flowing countercurrently. When there is a liquid flowing countercurrently with a gas in a packed column, irrigation of the packings occurs and the cross-sectional area available for the gas is reduced void volume in the packings is filled with liquid. During constant flow of liquid at low to moderate gas velocity, we can observe that the pressure drop resembles that of dry packings. More so, there is a systematic flow of the liquid in the column and no sign of liquid holdup. However, as we increase the gas velocity, a sudden rise in the pressure drop can be seen and liquid starts to be trapped in the packings. Increasing it further will start the build-up of the liquid which will soon result to flooding [9].
Differentiate between static and dynamic or operating holdup. How does this affect the pressure drop through a packed column? Liquid holdup ensues when the gas velocity in a packed column is further increased and is considered as a significant hydrodynamic framework for gas and liquid flow in columns. Shulman et al. (1955) reported that total liquid holdup is composed of a dynamic and static segment under gas-liquid flow conditions. Operating or dynamic holdup appears as the volume of liquid per volume of packing that gushes out of the bed right after the flow of
gas and liquid in the column is stopped. Contrarily, static liquid holdup is known to be the volume of liquid per volume of packing that is retained in the packed bed after all the flows are stopped and the bed is drained [11]. Liquid holdup gives a sudden rise of the pressure drop through a packed column due to the entrainment of the liquid by the gas [12].
Define loading and channeling. Give the relevance of these two factors in packed column operation. Loading, for mass transfer operations, is the condition when the liquid starts to accumulate in the packed column and generates pressure drop. Loading is known to be desirable for mass transfer. On the other hand, channeling is observed when the liquid or gas flow at some points are greater than other points. This is not an ideal condition for packed columns and is usually avoided by choosing the right packing material. Under normal packed column operations, both loading and channeling can occur. In order to attain a good mass transfer, high liquid flow rate can be used to generate loading and eliminate channeling [13].
How does the packing factor obtained from the flooding velocity differ from the one estimated empirically with the use of the correlation of Lobo et al? The correlation of Lobo et al for packing factor showed only a lone curve predicting packing flooding points which is based solely from bed porosity. On the other hand, the packing factor procured using the flooding velocity considers the flooding of the packed bed thereby giving a more accurate result [14]. XI.
Findings, Conclusion and Recommendation ko
Findings and Conclusion One of the objectives of this experiment is to determine the void fraction of the gas absorption column. By using equations (2), (3), (4), (5), and (6), the experimental void fraction along the column was calculated using three different gas flow rates ─ 30, 70, and 110 L/min. It was found that the void fraction is around 0.36 to 0.4075. Getting the average, the experimental void fraction of the packed column was calculated to be 0.3836. Another objective of this experiment was to determine the experimental packing factor using flooding velocity calculations. The computed packing factor, F p, using the graph for flooding velocity and equation (11) is 689.35 ft -1. Lastly, another objective of this experiment is to determine the effect of liquid hold up or loading on the pressure drop in the packed column. By comparing the experimental values for an irrigated column and the calculated values using the Robbins equation, a general trend was observed wherein the accumulation of liquid within the voids and spaces resulted into sudden and drastic increase in pressure drop. The loading point ─ time when liquid starts to hold up or accumulate inside the voids ─ on the logarithmic plot of the pressure drop per unit height of packing versus the superficial gas velocity is signified by the sudden change in slope of the curve. As the gas velocity at a given liquid flow rate is further increased, the occurrence of flooding advances which is a result of the reduction of the cross-sectional area available for the gas. The differences between the experimental results and the calculated pressure drop using the Ergun equation and Robbins equation can be attributed to the assumptions and considerations made while deriving and establishing these said equations.
Recommendation It is necessary to double check the flow of the gas from the pressure taps. Gas flow should always be directed to the left manometer which is the water manometer. In addition to this, if you only want to account for the pressure drop across the whole packed column, the three-way valve around the middle pressure tap should be closed to block the flow of the gas from the center. This way, only the gas flow from the top and bottom of the column will be accounted. Also, in order to minimize, if not prevent, the effect of surface tension on the reading of the height of fluid in the manometer, throttle the gas control valve periodically. To further understand the concept of flooding and pressure drop, the experiment can further be extended by utilizing the separate parts of the packed column. Also, to appreciate the concept of gas absorption, other experiments involving different process fluids, such as carbon dioxide and water, may be performed.
