The effect of fly ash on fluid dynamics of CO2 scrubber in coal-fired power plant

The effect of fly ash on fluid dynamics of CO2 scrubber in coal-fired power plant

chemical engineering research and design 9 0 ( 2 0 1 2 ) 328–335 Contents lists available at ScienceDirect Chemical Engineering Research and Design ...

792KB Sizes 0 Downloads 102 Views

chemical engineering research and design 9 0 ( 2 0 1 2 ) 328–335

Contents lists available at ScienceDirect

Chemical Engineering Research and Design journal homepage: www.elsevier.com/locate/cherd

The effect of fly ash on fluid dynamics of CO2 scrubber in coal-fired power plant Zhisheng Chen a,b , Derek Yates a , James K. Neathery a , Kunlei Liu a,∗ a b

University of Kentucky, Center for Applied Energy Research, 2540 Research Park Drive, Lexington, KY 40511, USA China Datang Technologies & Engineering Co., Ltd., 120 Zizhuyuan Road, Beijing 100097, PR China

a b s t r a c t Uncaptured fly ash and/or suspended solids from wet flue gas desulfurization (WFGD) scrubbing solutions are one of several factors that will influence the performance and robustness of carbon dioxide capture systems in coal-fired power plants which will be installed prior to the exhaust stack. In this study, a 100 mm ID packed column scrubber was tested with different concentrations of ash in various chemical solutions to evaluate the influence of solids on the fluid dynamics of the packing material. Data reported here are collected from three solutions including water, 30 wt% MEA (monoethanolamine), and 20 wt% potassium carbonate. The packing selected for this study was a 16 mm polypropylene pall rings. Compressed air was used to simulate flue gas at near ambient temperature and pressure. A series of three experiments was performed, and the results indicated that the flooding point of the packed column was significantly impacted by the addition of 1–3 wt% ash solids into the solution. Solutions (water and 20 wt% potassium carbonate) containing solids had a lower pressure drop at a given superficial gas velocity and early flooding start point (e.g., lower superficial gas velocity at the column flooding point) than that without ash. A higher concentration of ash in the solution correlated to a lower pressure drop at the column flooding point. However, the addition of ash to a 30% MEA solution caused the pressure drop to increase for a given superficial gas velocity. The liquid holdup in the column had a significant increase due to the presence of ash. Published by Elsevier B.V. Keywords: Fly ash; Fluid dynamics; Pressure drop; Flooding point; Liquid holdup; Carbon capture

1.

Introduction

Gas–liquid counter-current flow in a packed tower plays an important role in the modern chemical industry for gas cleanup applications. Many researchers have focused on this area, and the fluid dynamics of packed columns in various configurations have been studied extensively, including structured packing as well as both classic random packing and a group of modern new random packing media (Kouri and Sohlo, 1996; Spedding, 1986; Heymes et al., 2006; Alix and Raynal, 2008; Piche et al., 2001a,b,c). Numerous experimental data can be found in the literature, with various correlations and empirical formulas developed to describe the relationship of pressure drop, column flooding point, and liquid holdup in the column. The effect of gas and liquid loading on the pressure drop across the column packing was acknowledged in various



empirical correlations associated with dimensionless parameters of Reynolds number and Froude number. Mackowiak developed a correlation to predict pressure drop in the irrigated packed columns (Mackowiak, 1990), as well as liquid holdup from the pre-loading zone to the flooding point. In a later work he established an extended channel model for prediction of the pressure drop in single-phase flow in packed columns (Mackowiak, 2009). Billet and Schultes modeled the pressure drop (Billet and Schultes, 1991) and the liquid holdup (Billet and Schultes, 1993) by using a physical model in the two-phase counter-current packed columns in 1991 and 1993, respectively. Many researchers have studied the hydrodynamic properties of carbon dioxide capture systems, such as Pascal and Raynal (2008, 2009). They introduced parameters such as liquid distribution, holdup (Alix and Raynal, 2008), pressure drop and mass transfer (Alix and Raynal, 2009) for modern high capacity packing applications. However, none of

Corresponding author. Tel.: +1 859 257 0293. E-mail address: [email protected] (K. Liu). Received 30 September 2010; Received in revised form 2 April 2011; Accepted 27 July 2011 0263-8762/$ – see front matter Published by Elsevier B.V. doi:10.1016/j.cherd.2011.07.024

329

chemical engineering research and design 9 0 ( 2 0 1 2 ) 328–335

2.

Experimental

2.1.

Packed column and experimental conditions

Nomenclature hL liquid holdup (m3 m−3 ) ReL = VL L /L ˛ Reynolds number of liquid VG gas velocity (m/s) liquid velocity (m/s) VL G gas density (kg/m3 ) liquid density (kg/m3 ) L L dynamic viscosity (Pa s) ˛ specific surface area of packing (m2 m−3 )

the studies focused on the potential influence of fly ash or other suspended solids. It is expected that all carbon dioxide capture processes for coal-derived post-combustion CO2 enrichment will be installed and operated downstream from the particular removal device, however, due to the imperfect operation of such equipment, a very small amount of uncaptured fly ash or solid carryover from the wet-FGD system will continually accumulate within the carbon capture system. On the other hand, the CO2 capture process could also result in solids formation (particular in the bicarbonate form) in the scrubber due to chemical reaction and with the possibility of precipitation such as ammonium bicarbonate in the chilled ammonia process (Budzianowski and Koziol, 2005). Although a filter system could be installed in-line to remove the suspended solids to a large degree, the initial and ongoing cost considerations make it desirable to limit the filter capacity and mesh size to only the level required to maintain a robust operation. This study focused on the effect of fly ash on fluid dynamics of random packed columns which could be used for post-combustion carbon dioxide capture process in the near future.

The experiments were performed in a 100 mm internal diameter PVC packed column. The effective packing height was 1.75 m. Three different base aqueous solutions were tested: water, 30 wt% MEA (monoethanolamine), and 20 wt% potassium carbonate. Each base solution was tested using four sets of solids loading in the solution: no solid, 1 wt% solid, 2 wt% solid and 3 wt% solid. The content of solid in this study was selected upon our experience in the power industry and the balance between filtration and energy and capital cost associated with solid removal. Higher solvent solid content would lead to plugging/fouling and abrasion problem in piping and pump systems. In this case, a filter would be installed in the process to prevent these issues. Compressed air was used as a countercurrent flow to simulate flue gas at near ambient temperature and atmospheric pressure. The packing selected for this study was a 16 mm polypropylene pall rings; the packing specifications are given in Table 1. Fig. 1a and b shows photographs of unused new pall ring packing and water washed used packing, respectively, after extended experiments in various ash concentrations. As indicated in Fig. 1, ash particulates can be observed on the surface of the used packing. Further analysis from optical microscope showed that the ash has been coated on the packing surface. The twelve solutions studied in this work were sampled and measured for surface tension, viscosity, pH value and density. The measured data, as listed in Table 2, indicated that the addition of solid (fly ash) did not appreciably affect the solution’s physical properties for the solvents tested except for a slight increase in the pH value which may be beneficial for CO2 capture in the scrubber. The experimental setup is shown in Fig. 2. A tank with a 230 l solution storage capacity is located at the bottom of the system. A progressive cavity pump dispenses liquid to the top

Table 1 – Physical property and parameters associated with packing materials tested. Type

Size (mm)

Height (mm)

Bulk density (kg/m3 )

Surface area (m2 /m3 )

Void fraction (%)

Pall ring

16

16.0

16

330

93

Fig. 1 – Unused new and used pall ring packing.

330

chemical engineering research and design 9 0 ( 2 0 1 2 ) 328–335

Table 2 – Parameters for tested solutions. Solution

Measured parameters T

Water Water with 1% ash Water with 2% ash Water with 3% ash 30% MEA 30% MEA with 1% ash 30% MEA with 2% ash 30% MEA with 3% ash 20% K2 CO3 20% K2 CO3 with 1% ash 20% K2 CO3 with 2% ash 20% K2 CO3 with 3% ash

(◦ C)

20.8 20.4 20.6 20.6 20.8 20.3 20.3 21.2 21.1 20.9 20.8 21.1

pH 7.91 11.24 11.69 11.96 12.67 12.74 12.76 12.71 12.53 12.65 12.83 12.95

of the column, where it is distributed onto the random packing. A mass flow controller was used to meter compressed air into the column from bottom to form a counter-flow gas–liquid configuration. Two thermocouples were installed, one in the liquid inlet, and one in the gas inlet. The ambient temperature was also monitored. A liquid flow diverter was installed at the base of the column to allow the liquid to be collected in a separate vessel, to permit measurement of column liquid hold-up. The entire system was monitored and controlled through a National Instruments LabView Real-Time controller, with data output recorded for later analysis. In the literature, most of the experiments concerning fluid dynamics in packed columns were carried out using a constant liquid loading. Increasing the gas loading led to increased pressure drop and liquid holdup. In this study a constant liquid–gas (L/G) ratio was used, with changes in either liquid or gas flow as independent variables, and column pressure drop

Viscosity  (g/cc) 0.996 0.994 0.992 0.992 1.010 1.013 1.009 1.009 1.181 1.177 1.177 1.178

v (cSt)

u (cP)

1.046 1.037 1.037 1.029 2.720 2.780 2.663 2.686 1.403 1.420 1.437 1.420

1.041 1.031 1.029 1.020 2.747 2.816 2.687 2.710 1.656 1.671 1.691 1.672

y (dyns/cm) 70.02 69.87 69.73 69.73 61.25 61.44 61.19 61.19 74.88 74.63 74.63 74.69

and flooding point as dependent variables. Three different L/G ratios were studied: 5.35, 10.69, 20.05 L m−3 . In order to directly compare the impact of suspended solids on the fluid dynamics of the packed column, the pall rings were changed out each time prior a new base solvent was scheduled to be tested. This practice avoids any residual “loading” of the ash/solid on packing which occurs with new packing and makes the test results directly comparable. For each solution, 1 wt%, 2% and 3% of ash concentrations were tested in sequence. The general procedure for solution and experiment preparation is described briefly here. The system is cleaned thoroughly and the ash-coated ‘dirty’ random packing is replaced with unused clean packing. Fill water (typically 95 l) is added into the bottom solution storage tank; chemical reagents (such as MEA or K2 CO3 ) are added into the tank and thoroughly mixed to achieve the desired weight percent of chemical concentration in the tank. In a typical preparation process, the stirring will take two to three hours to guarantee a well-mixed solution. Upon the completion of all experiments for a solid-free solution, a representative fly ash mass was weighed and added into tank to achieve a 1 wt% total solid in solution, and stirred. Following the completion of the experiments requiring 1% solids, and then the same procedure was followed to achieve a 2 wt% and 3 wt% ash loading.

2.2.

Fig. 2 – Flow diagram of experiment.

Surface tension

Pressure drop measurement

Static pressure taps situated both below and above the packed portion of the column allow the pressure drop created by the fluid flow through the packing to be measured. The pressure taps were connected to a differential pressure transmitter, which was monitored by LabView. The initial pressure tap arrangement was made to be self draining, but was blown out frequently to prevent the accumulation of liquid while under operating pressure. During the facility commissioning, it was found that due to the liquid wall effect, the tip of pressure probe under the packing supporting grid would be easily covered by liquid film which would enter the pipeline under the system pressure. An extended pipe was welded to the probe tip to help eliminate this problem, see Fig. 2. In addition, a compressed air line and an appropriate valve arrangement was installed on the line connecting pressure probe and transducer to purge the pipe when abrupt changes were observed in the experiment due to liquid accumulation in the pressure lines. In this way, the most repeatable and accurate data were obtained.

chemical engineering research and design 9 0 ( 2 0 1 2 ) 328–335

331

water L/G=10.69L m-3 1%ash-water L/G=10.69L m-3 2%ash-water L/G=10.69L m-3 3%ash-water L/G=10.69L m-3

1000

Expon. (water L/G=10.69L m-3)

Pressure drop (pa m-1)

Expon. (1%ash-water L/G=10.69L m-3) Expon. (2%ash-water L/G=10.69L m-3) Expon. (3%ash-water L/G=10.69L m-3)

100

10 1

0.1 F-factor (pa0.5)

Fig. 3 – Pressure drop of different ash concentrations in aqueous solution at L/G = 10.69 L m−3 .

2.3.

Liquid holdup measurement

The liquid holdup of the packed column at varying flow rates was measured using a manual volumetric method. The system was run at the selected gas and liquid flow rates until a steady state was achieved. At that time, a ball valve V1 as listed in Fig. 2 between the column and storage tank was turned off and a timer (e.g., a stopwatch) was started. After 15 s the pump motor was powered off, liquid trapped in the column will be collected and measured by volumetric vessel through bypass valve V2. The pump feed rate was recorded. It would take several minutes before all the liquid held up in

3.

Results and discussion

3.1.

Pressure drop

No available literature was found which studies the impact of solid on fluid dynamics in packed columns for CO2 cap-

20%K2CO3 L/G=10.69L m-3 1%ash-20%K2CO3 L/G=10.69L m-3 2%ash-20% K2CO3 L/G=10.69L m-3 3%ash-20%K2CO3 L/G=10.69L m-3 Expon. (20%K2CO3 L/G=10.69L m-3) Expon. (1%ash-20%K2CO3 L/G=10.69L m-3) Expon. (2%ash-20% K2CO3 L/G=10.69L m-3) Expon. (3%ash-20%K2CO3 L/G=10.69L m-3)

1000

Pressure drop (pa m-1)

the column drained through valve V2 into the measurement vessel. The liquid holdup in column in normal operation was calculated as the difference between the captured volume and the product of pump feeding rate in 15 s. The dynamic liquid holdup was calculated as the holdup volume divided by the total volume of packing.

100

10 1

0.1

F-factor

(pa0.5)

Fig. 4 – Pressure drop of different ash concentrations in 20% K2 CO3 aqueous solution at L/G = 10.69 L m−3 .

332

chemical engineering research and design 9 0 ( 2 0 1 2 ) 328–335

MEA 1% ash-MEA

1000

2% ash-MEA 3% ash-MEA Expon. (MEA) Expon. (1% ash-MEA) Expon. (2% ash-MEA)

Pressure drop (pa m-1)

Expon. (3% ash-MEA) 100

10

1 1

0.1

F-factor (pa0.5) Fig. 5 – Pressure drop of different ash concentrations in 30% MEA aqueous solution at L/G = 10.69 L m−3 . ture processes. The experimental pressure drop data for water solution with various solid loadings at a given constant liquid–gas ratios are plotted in Fig. 3. The X-axis is the parameter called “F-factor” which is defined as a function of gas velocity and square root of gas density: F-factor = vG ·



G

The F-factor could represent gas kinetic energy. In general, the column flooding points for the water with the suspended

solids was shifted towards flooding at a lower gas flow rate as compared to that obtained from the experiment without ash. It can be observed that the more ash is present in the solution, the lower the gas flow rate is at the column flooding point. However, measurement of the gas side pressure drop shows that a lower gas resistance occurred in solutions with the presence of ash as compared to the same gas flow rate without ash for water solution. On other hand, the difference of pressure drop between water-only and solution with

800

Experimental pressure drop (Pa m-1)

700

600

500

400

300

200

100

0 0

100

200

300

400

500

Calculated pressure drop (Pa

600

700

800

m-1)

Fig. 6 – Comparison of pressure drops for dry packing calculated with experimentally result.

333

chemical engineering research and design 9 0 ( 2 0 1 2 ) 328–335

water L/G=5.35L m-3 2%ash-water L/G=5.35L m-3 water L/G=10.69L m-3 2%ash-water L/G=10.69L m-3 water L/G=20.05L m-3 2%ash-water L/G=20.05L m-3 Expon. (1%ash-water L/G=5.35L m-3) Expon. (1%ash-water L/G=10.69L m-3) Expon. (1%ash-water L/G=20.05L m-3)

1%ash-water L/G=5.35L m-3 3%ash-water L/G=5.35L m-3 1%ash-water L/G=10.69L m-3 3%ash-water L/G=10.69L m-3 1%ash-water L/G=20.05L m-3 3%ash-water L/G=20.05L m-3 Expon. (3%ash-water L/G=5.35L m-3) Expon. (3%ash-water L/G=10.69L m-3) Expon. (3%ash-water L/G=20.05L m-3)

Holdup (m3 m-3)

0.1

0.01 0.001

0.01

liquid load (m3 m-2 s-1)

Fig. 7 – Dynamic liquid holdup of different ash concentrations in water solution at different L/G ratios. solids are almost constant through the tested range of gas velocity. The same tendency for the shift of flooding points occurred in 20 wt% potassium carbonate aqueous solution illustrated in Fig. 4; however, the difference of pressure drop

between solid-free solution and solution with solids are gradually increased with the increasing in F-factor though 20 wt% K2 CO3 aqueous solution has a lower pressure drop at flooding point.

1% ash-water L/G=5.35L m-3 1% ash-water L/G10.69L m-3 1% ash-water L/G=20.05L m-3

Liquid holdup

(m3

m-3)

0.1

0.01 1

10

ReL Fig. 8 – Liquid holdup for different liquid–gas ratio in 1 wt% ash-water slurry.

100

334

chemical engineering research and design 9 0 ( 2 0 1 2 ) 328–335

1% ash-water 2% ash-water 3% ash-water Linear (1% ash-water) Linear (2% ash-water) Linear (3% ash-water)

Liquid holdup

0.1

y = 0.0036x + 0.0095 y = 0.0030x + 0.0056 y = 0.0029x + 0.0006

0.01 1

10

100

Re L Fig. 9 – Linear regression of liquid holdup to Reynolds number for different ash loading.

From the visual inspection after solutions with solids were tested, the packing materials were coated with fine ash. In addition, a weight gain method was used to evaluate the impact of solid loading on pall rings ash coating. One hundred pall rings were chosen randomly from the packing filled in the column after the completion of experiment for solid-free solution, solid-contained solution with 1 wt%, 2 wt% and 3 wt%, respectively. After air dry for overnight, these packing elements weight 48.4 g, 48.5 g, 48.8 g, and 48.7 g correspondingly. Considering the randomness of the selection, the packing elements had a slight increased weight gain when solidcontained solution was used compared to that of the ‘clean’ packing. However, the solid loading in the solution did not change the weight gain after 2 wt%. The 30% MEA aqueous solution displayed similar flooding behavior to that of the with water and potassium carbonate solution in that flooding would occur at a lower gas loading with the presence of ash, as presented in Fig. 5. However, the gas side pressure drop using the MEA solution exhibited a slightly higher pressure drop with the presence of ash as compared to the water and potassium carbonate base solutions. Billet developed a model for the prediction of pressure drop in packed columns (Billet and Schultes, 1991, 1993). A comparison of the pressure drops calculated from Billet’s model for dry packing and experimental data, reveals that the mean relative standard deviation (MRSD) is 5.8%, see Fig. 6. MRSD is defined as the following:

MRSD =

1 n standard deviation for n data. n average i=1

For a wetted packed column, the pressure drop and gas velocity at the column flooding point can be calculated from Billet’s model, but the predicted gas velocity at the flooding point was higher than that obtained from our experiments, and the predicted pressure drop was much lower compared with our experimental result. The Billet’s model also does not

correlate well the current experimental data with the presence of solids in the solution. The possible reasons for this disagreement could include to be:(1) experimental uncertainty – when the gas velocity approached the flooding point, the test conditions becomes unstable and the pressure drop associated with this condition fluctuates more and becomes more challenging to obtain representative data; (2) the pressure drop could be affected by many factors including the geometry parameter, such as the variability of the shape and size of the packing elements between manufacturer’s product, the different ratio of packing element and column diameter, and even the flow pattern in the packed column.

3.2.

Liquid holdup

In this work, the liquid holdup for an air–water system was used. According to the film theory, more liquid held in the packed column tends to lead to a smaller void fraction, which leads to higher pressure drop. As shown in Fig. 3, a lower pressure drop was found as a result of increasing ash concentration, which means lower liquid holdup should be obtained. Actually, higher liquid holdup was observed as the increasing of ash concentration in Fig. 7. The reason is still unclear and is under investigation. One possible reason may be the ash content has an interaction effect with the liquid and packing surface, the contact angle and hydrophilic properties may be changed along with the increasing ash content. Fig. 8 illustrates liquid holdup for different liquid–gas ratios in 1 wt% ash–water slurry. It was observed that the dynamic liquid holdup increased linearly with the increasing in liquid Reynolds number with the exception near the column flooding point. At a certain liquid flow rate, the gas flow rate decreased along with an increasing liquid–gas ratio, but the liquid holdup remained relatively constant. Thus, the gas flow rate has a negligible effect on liquid holdup. The liquid holdup of different ash concentration could be described as a function of

chemical engineering research and design 9 0 ( 2 0 1 2 ) 328–335

Reynolds number except in the vicinity of flooding point as follows: hL = A · ReL + B,

(1)

where, A=

2.5n2 − 6.5n + 33 , 10,000

(2)

B=

5.5n2 + 66.5n − 55 , 10,000

(3)

The symbol “n” represents ash concentration. Eq. (1) was only valid for aqueous water solutions with the presence of ash. The relevant relationship was shown in Fig. 9.

4.

Conclusions

The ash concentration in the solutions tested will influence the pressure drop, the column flooding point, and the dynamic liquid holdup in a packed column. All solutions tested in this study had a lower pressure drop across the packing at the flooding point when ash was introduced. The flooding point also had a lower gas flow rate when more ash was present in the solution. The pressure drop was slightly influenced by the ash, however, it did not maintain the same trend between high viscosity solvent (MEA) and low viscosity solvent (K2 CO3 and water). With more ash present in the solution, the aqueous MEA solution had a slightly higher pressure drop at the same gas loading, whereas the K2 CO3 solution and water solution had an inverse result. In the solid-free water solution, dynamic liquid holdup increased linearly with the increasing liquid flow rate regardless of the change of gas flow rate, at liquid–gas ratios varied from 5.35, 10.69 to 20.05 L m−3 , except the vicinity of flooding point which had a significant increase of liquid holdup. More ash present in the solution led to a higher dynamic liquid holdup at the same liquid flow rate. The liquid in the packed column caused a reduction of the free-space cavities in the packed column, resulting in a decreased void fraction of the packed column medium with the increasing liquid holdup. Normally, a lower void fraction of a packed column or a higher liquid holdup would cause a higher pressure drop, however, in this study the higher liquid holdup caused by higher ash concentration led to a lower pressure drop. One possible explanation is that the surface properties of the packing have been changed by the ash material, leading to the change of contact angle, and thereby influencing the liquid wetted area.

335

There are different kinds of CO2 capture processes, such as MEA process, potassium carbonate process and chilled ammonia process, the effect of ash in each individual process should be considered. Some other form of solid especially salts in chilled ammonia process should also be taken into account on the hydrodynamics of absorber and stripper.

Acknowledgements American Electric Power (AEP), Big Rivers Electric Corporation, Duke Energy, E-ON US, East Kentucky Power Cooperative (EKPC), Electric Power Research Institute (EPRI), Illinois Clean Coal Institute (ICCI), and the Kentucky Department for Energy Development and Independence (KY DEDI).

References Alix, P., Raynal, L., 2008. Liquid distribution and liquid hold-up in modern high capacity packings. Chem. Eng. Res. Des. 86, 585–591. Alix, P., Raynal, L., 2009. Pressure drop and mass transfer of a high capacity random packing. Application to CO2 post-combustion capture. Energy Procedia 1 (2009), 845–852. Billet, R., Schultes, M., 1991. Modelling of pressure drop in packed columns. Chem. Eng. Technol. 14 (1991), 89–95. Billet, R., Schultes, M., 1993. A physical model for the prediction of liquid hold-up in two-phase countercurrent columns. Chem. Eng. Technol. 16 (1993), 370–375. Budzianowski, W., Koziol, A., 2005. Stripping of ammonia from aqueous solutions in the presence of carbon dioxide effect of negative enhancement of mass transfer. Chem. Eng. Res. Des. 83, 196–204. Heymes, F., Demoustier, P.M., Charbit, F., Fanlo, J.L., Moulin, P., 2006. Hydrodynamics and mass transfer in a packed column: case of toluene absorption with a viscous absorbent. Chem. Eng. Sci. 61, 5094–5106. Kouri, R.J., Sohlo, J., 1996. Liquid and gas flow patterns in random packings. Chem. Eng. J. 61, 95–105. Mackowiak, J., 1990. Pressure drop in irrigated packed columns. Chem. Eng. Process. 29 (1991), 93–105. Mackowiak, J., 2009. Extended channel model for prediction of the pressure drop in single-phase flow in packed columns. Chem. Eng. Res. Des. 87, 123–134. Piche, S., Larachi, F., Grandjean, B.P., 2001a. Improved liquid hold-up correlation for randomly packed towers. Trans. Icheme 79, Part A (January). Piche, S., Larachi, F., Grandjean, B.P., 2001b. Improving the prediction of irrigated pressure drop in packed absorption towers. Can. J. Chem. Eng. 79 (August). Piche, S., Larachi, F., Grandjean, B.P., 2001c. Loading capacity in packed towers – database, correlations and analysis. Chem. Eng. Technol. 24 (4), 373–380. Spedding, P.L., 1986. Ammonia absorption into water in a packed tower I. Characterization of packing and liquid distribution. Chem. Eng. J. 32, 65–76.