AN ABSTRACT OF THE THESIS OF

DuWayne Aaron Zuehlsdorff for the degree of Master of Science

in Chemical Engineering presented on April 20, 1979

Title: Absorption of Ammonia in Water Using a Multiple Stage

Crosscurrent Packed Column

Abstract approved:
Redacted for Privacy
Dr. Charles E. Wicks

A multiple stage crosscurrent packed column suitable for gas-

liquid mass transfer was constructed and operated. The absorption

column consisted of a rectangular box constructed of plexiglass and

metal. Deflection baffles were positioned at regular intervals on

opposite sides of a central packed section to divert the gas phase

into the packing affecting a crisscross gas-liquid flow pattern.

The column was used for absorbing ammonia from air in an ammonia-air-

water system.

Pressure drop data were collected for three crosscurrent arrange-

ments and for the conventional countercurrent arrangement. The

results indicate that crosscurrent operation with a = 1 (a is a

baffle spacing variable) has a higher pressure drop per meter of

packed height than the countercurrent operation. When a is decreased

to 2 and 3 the crosscurrent column has better pressure drop charac-

teristics than the countercurrent column.

 The overall mass transfer efficienty was studied experimentally

for the three crosscurrent arrangements and for the countercurrent

arrangement. Crosscurrent operation with a = 1 is more efficient at

removing ammonia than countercurrent operation for the system and

ranges of flow rates studies. At a liquid to gas molar ration of

two (L/G = 2.0) the a = 2 crosscurrent arrangement was also more

efficient than countercurrent but this advantage is lost at lower

L/G ratios.

Overall, the multiple stage crosscurrent packed column is

comparable to and at some conditions more efficient than the conven-

tional countercurrent packed column. It is therefore a viable

alternative mass transfer device.

 Absorption of Ammonia in Water Using a
Multiple Stage Crosscurrent Packed Column

by

DuWayne Aaron Zuehlsdorff

A THESIS

submitted to

Oregon State University

in partial fulfillment of
the requirements for the
degree of

Master of Science

Completed April 1979

Commencement June 1979

APPROVED:

Redacted for Privacy

Head of Department of Chemical Engineering and Professor
in charge of major

Redacted for Privacy

Dean of Graduate School

Date thesis is presented April 20, 1979

Typed by Sharon Mosley for DuWayne Aaron Zuehlsdorff

 ACKNOWLEDGEMENTS

The author expresses his appreciation to Dr. Charles E. Wicks

and the Chemical Engineering Department of Oregon State University

for financial support in the form of a Graduate Teaching Assistant-

ship.

The assistance of Ghawamedin Bayan during construction of the

equipment is also gratefully acknowledged.

 TABLE OF CONTENTS

Page

I. INTRODUCTION 1

II. LITERATURE SURVEY 6

III. EXPERIMENTAL EQUIPMENT 9

IV. THEORY OF CALCULATIONS 18

Inlet Flow Rate of Air 18

Inlet Flow Rate of Ammonia 20
Inlet Flow Rate of Water 23
Exit Water and Air Flow Rates 23
Exit Flow Rate of Ammonia in the Liquid Stream 24
Exit Flow Rate of Ammonia in the Gas Stream 25
Number and Height of Gas Phase Transfer Units 25

V. OPERATING PROCEDURE 27

Preliminary Operating Procedure 27

Procedure for a Data Run 28
Start-up 28
Data Collection 28
Shut-down 28

VI. ANALYSIS OF ERRORS 31

VII. PRESENTATION AND DISCUSSION OF RESULTS 38

Definition of Baffle Spacing Variable 38

Basis for Comparison of Crosscurrent and 39
Countercurrent Data
Gas Phase Pressure Drop 40
Mass Transfer Efficiency 41
Absorption Efficiency 45
Height of a Transfer Unit 46
Comparison Based on Packed Area 46

VIII. CONCLUSION 58

IX. NOMENCLATURE 63

X. LITERATURE CITED 66
 Page

XI. APPENDIX A: CALIBRATION CURVES 68

XII. APPENDIX B: SAMPLE CALCULATIONS 72

XIII. APPENDIX C: DATA TABLES 79

 LIST OF FIGURES

Figure Page

1 Flow patterns in a multiple stage crosscurrent packed

column. 3

2 Multiple stage crosscurrent packed column. 10

3 Section A-A detail of multiple stage crosscurrent

packed column. 11

4 Baffle arrangements. 13

5 Equipment arrangement schematic. 14

6 Manometer system schematic. 17

7 Pressure drop vs. gas rate; liquid rate = 1.76 kg/sm2. 42

2
8 Pressure drop vs. gas rate; liquid rate = 3.51 kg/sm . 43

2
9 Pressure drop vs. gas rate; liquid rate = 5.27 kg/s'm . 44

10 Absorption efficiency vs. gas rate; baffle arrangement

#1, a = 1. 48

11 Absorption efficiency vs. gas rate; baffle arrangement

#2, a = 2. 49

12 Absorption efficiency vs. gas rate; baffle arrangement

#3, a = 3. 50

13 Absorption efficiency vs. gas rate; countercurrent

arrangement. 51

14 Absorption efficiency vs. gas rate; comparison of

arrangements at L/G = 2.0. 52

15 Absorption efficiency vs. gas rate; comparison of

arrangements at L/G = 1.0 and 4.0. 53

16 H vs gas rate; baffle arrangement #1, a = 1. 54

tOG
Figure Page

17 HtOG vs. gas rate; comparison of arrangements at

L/G = 2.0. 55

18 Pressure drop vs. gas rate; liquid rate = 3.51

kg/sm2; comparison of arrangements based on
packed area. 56

19 Absorption efficiency vs. gas rate at L/G = 2.0;

comparison of arrangements based on packed area. 57

20 Absorption efficiency vs. gas rate at L/G = 2.0;

proposed regions of gas-liquid contact. 62

21 Ammonia rotameter calibration at 283 K and 103.4 kPa. 69

22 Small liquid rotameter calibration at 283 K. 70

23 Large liquid rotameter calibration at 283 K. 71

 LIST OF TABLES

Table Page

I Maximum Probable Error of Direct Measurements 34

II Maximum Probable Error of Indirect Measurements 35

III Results of Duplicate Runs 36

IV Pressure Drop Data 80

V Mass Transfer Data - Part I 81

VI Mass Transfer Data - Part II 83

 ABSORPTION OF AMMONIA IN WATER USING A
MULTIPLE STAGE CROSSCURRENT PACKED COLUMN

I. INTRODUCTION

Gas absorption is the unit operation in which one or more soluble

components of a gas mixture are dissolved in a liquid. Objectionable

pollutants may be removed from the gas stream or valuable vapors

carried by the gas stream may be recovered for reuse. This diffusional

transfer of mass from one phase to another has always been important to

the chemical industry and is now of primary interest because of growing

environmental concerns and increasing solvent costs.

One of the most common pieces of absorption equipment is the packed

column. The packed column is a simple device consisting of a vertical

shell filled with one of numerous types of inert packing materials.

The two phases, usually one gas phase and one liquid phase, flow through

the tower and the packing. The pieces of packing cause flow disruptions

and are designed to promote contact between the two phases and thus

increase the mass transfer.

Two types of packed tower schemes are common. The first type is

the cocurrent operation in which the gas and liquid phases are intro-

duced into the tower at one end and flow in the same direction through

the packing. The second type is the countercurrent operation in which

the two phases are introduced at opposite ends of the tower, the liquid

into the top and the gas into the bottom, and flow in opposite direc-

tions. Ordinarily, countercurrent flow is preferable to cocurrent

 2

because a higher concentration driving force and thus better mass

transfer exists throughout the tower; this can be easily shown through

elementary mass transfer theory. However, there are disadvantages of

the countercurrent tower as compared to the cocurrent packed tower.

In a countercurrent tower, increasing the flow rate of either phase

tends to decrease the flow area available for the other phase. One

consequence of this is that the pressure drop required to move the gas

phase through a countercurrent tower can be significantly higher than

for a cocurrent tower. In a limiting case, the flows can be such that

the liquid phase will stop flowing downward, a condition called flooding.

Also, at low liquid rates the liquid tends to collect into small rivu-

lets and flows along localized paths through the packing leaving much

of the packing surface dry. This effect is known as channeling and is

one of the chief reasons for poor performance in large countercurrent

packed towers.

A third less well known type of flow scheme for packed towers does

exist. This is called the crosscurrent packed tower. As a single stage

unit, the liquid phase enters the top and the gas phase enters at a

side, causing the gas phase to flow perpendicularly to the liquid flow.

A device that produces a cascade of crossflow units can be constructed

as in Figure 1. While overall operation is countercurrent, phase flow

within individual stages is opposed at 900 and interstage flow patterns

are essentially crisscross in nature.

The purpose of this project was to construct a pilot-scale

multiple stage crosscurrent packed column and conduct overall mass

transfer studies in order to effect a critical comparison between

 3

GAS PHASE OUT

-01.-- LIQUID PHASE IN

'1111. GAS PHASE IN

3,
LIQUID PHASE OUT

Figure 1. Flow patterns in a multiple stage crosscurrent packed

column.
 4

crosscurrent and the more conventional countercurrent packed columns.

A detailed mass transfer study would include the following variables:

1. Liquid and gas flow rates

2. Vertical baffle spacing

3. Intercore to outer wall baffle spacing

(or open space vs. packed space ratio)

4. Types of packing material

5. Systems with different gas solubilities

The final result would be to develop a conceptual model and correla-

tions that incorporate these variables into a general design procedure

for crossflow devices. The scope of this investigation is limited to

studying the effects of the first two. Pall rings were chosen for the

packing material for all runs because previous studies reported

favorable pressure drop characteristics when used in crossflow appli-

cation [17]. The ammonia-air-water system was chosen because it is of

high commercial importance, is relatively safe and easy to work with

when proper precautions are taken, and necessary thermodynamic and

physical data are readily available.

Although published data for absorption of ammonia using a

countercurrent packed tower are available, similar data were also

taken using the same pilot-scale tower. Comparison with published

data can be misleading since small packed towers often behave unlike

large ones.

Superficially, it would appear that crosscurrent flow devices are

less desirable than countercurrent ones because the crosscurrent flow

has inherently less efficient mass transfer characteristics than the

 5

countercurrent mode. However, the vigorous contacting patterns produced

in a crossflow cascade may compensate for this loss in efficiency. Also,

other favorable characteristics such as low pressure drops, constant

redistribution of the liquid phase, and wide operating ranges may make

the multiple stage crosscurrent packed column a viable alternative mass

transfer device.

This report will first examine the literature on crosscurrent

columns. Then the equipment used for this study and all relevant calcu-

lations and operating procedures will be described. Finally, the

results of this experimental study will be presented and discussed.

 6

II. LITERATURE SURVEY

The idea of a crosscurrent packed column is relatively new.

Surprisingly little has been published about crosscurrent towers

although considerable literature dealing with countercurrent and

cocurrent packed towers is available.

The principles of mass transfer theory are well discussed by Bird,

Stewart and Lightfoot [1], Skelland [2], and Welty, Wicks and Wilson

[3]. There is also an abundance of literature on design procedures

and performance data for countercurrent packed columns and packing

characteristics. Among these arePerry's Handbook [4], Treybal [5],

McCabe and Smith [6], and Foust [7]. There are also numerous articles

published each year in chemical engineering periodicals dealing speci-

fically with the absorption of ammonia in countercurrent packed towers.

Single stage crossflow gas-liquid contactors have found extensive

use in water cooling and some use in stripping and absorption operations.

The first crossflow application was likely in a water cooling tower [8].

Crossflow cooling towers are used mainly because of low pressure drops.

Hanf [9] recognized this advantage and showed a drawing of a crosscur-

rent packed scrubber. HEW Bulletin AP-51 [10] also included a drawing

of a crossflow packed scrubber. It stated that this type of scrubber

operates with very low pressure drop and water requirements, both of

which are about 40 percent of that required for countercurrent oper-

ation.
 7

Design techniques for single stage crossflow devices have been

developed. Baker and Shryock [11] and Wnek and Snow [12] presented

design methods for cooling towers and ammonia stripping towers, and

Thibodeaux [13] discussed general design techniques for mass transfer

in single stage crossflow packed towers.

Gray, Frankenberger and Glasgow [14] first tested the feasibility

of a cascade of crossflow stages. Their tower was considered to be a

multistage packed tower with gas flow directed across the tower by use

of baffles. Fluid dynamics were studied and a limited amount of

pressure drop data were collected. Oza [15] then developed a modified

version of the Ergun equation to predict the pressure drop within the

crossflow column. An equation was also developed to measure drift

angle, the angle of deflection of the liquid from the vertical walls of

the crossflow tower, at various liquid and gas flow rates. Thibodeaux

[16] and Golshani [17] collected extensive data on the fluid dynamics

of a cascade crossflow packed column. Visual observations of the flow

patterns were made and pressure drop and some liquid hold-up data

were collected. Flow rates exceeded those possible in conventional

countercurrent operations and pressure drop was significantly less.

Thibodeaux noted that while pressure drop and gas-liquid contact

observations seemed encouraging, questions of mass transfer efficiency

still remained. Phase contact appeared vigorous but unirrigated

sections of packing immediately below each baffle were in evidence.

However, the vigorous bubbling activity at baffle sites may partially

compensate for this loss in efficiency.

 8

Pittaway [18] studied oxygen desorption from an oxygen-rich water

phase by contacting with air in a single stage packed crossflow device.

He established that the liquid phase mass transfer coefficient for his

application is intermediate to those for cocurrent and countercurrent

towers. Thibodeaux [19] considered the theoretical efficiency of the

crossflow operation by employing the number of transfer units technique

that is employed in countercurrent tower design. In general, he found

the efficiency of the single stage device to be intermediate between

countercurrent and cocurrent flow devices. However, as the number

of crossflow stages increases, the efficiency approaches that of the

countercurrent design. Thibodeaux also listed some advantages and

disadvantages of crossflow cascades.

Thibodeaux's analysis did not consider the dry zones of packing

in existence or the bubbling action on the baffles. These effects can

be assessed properly only through overall mass transfer studies on a

multistage crosscurrent packed column.

Thermodynamic and physical properties of the ammonia-air-water

system are found in many sources including Perry's Handbook [4] and in

data by Clifford and Hunter [20].

 9

III. EXPERIMENTAL EQUIPMENT

The absorption column, as illustrated in Fiaure 2, consisted of

a rectangular box constructed of plexiglass and metal. When operated

as a crossflow device, the internals of the box were arranged to

encourage gas-liquid contact in a general countercurrent fashion with

the liquid phase descending through the column and the gas phase moving

upward. Deflection baffles were positioned at regular intervals on

opposite sides of the packed section to divert the gas phase into the

packing affecting a crisscross gas-liquid flow pattern.

The box was constructed of 13-mm (1/2-in) clear plexiglass and

had internal dimensions of 2.44 m (96in) in height, 254 mm (10 in) in

width, and 127 mm (5 in) in depth. The 127 mm by 2.44 m sides were

removable to change baffle spacing. Two 127-mm square access ports,

one immediately above and the other below the packed section, facili-

tated changing of packing material. A 508-mm high section at the top

of the column remained void and served as a water droplet de-entrainer.

A similar 381-mm section at the bottom of the column was used as a

liquid accumulator.

The packed section was positioned in the center of the column and

was 127 mm by 127 mm cross-section and 1.52 m (60 in) in height. This

packed section was bounded by plexiglass on two sides and screen on the

other two sides and top and bottom. Figures 2 and 3 are detailed

drawings of the multistage crossflow packed column.

 10

--="i 183 mm

GAS OUTLET

LIQUID OE-
FRONT VIEW ENTRAINER SIDE VIEW
SPACE (VOID)

ACCESS PORT

LIQUID INLET
SHOWER HEAD

TOP VIEW

1274 mm SCREEN

H 254 ma

BOTTOM VIEW TIE-PODS

127

F.,- 254 gm

PACKED SECTION
GAS DEFLECTION 127, nn
BAFFLES
r
I

I
GAS INLET
LIQUID ACCUMULATOR
SPACE (VOID)

ACCESS PORT

LIQUID OUTLET

25 mm

Figure 2. Multiple stage crosscurrent packed column.

 QV NUT
RUBBER 0-RING .die..TIE -ROD

,.\ .-\_-\_11-,. 0e E\_1K

\--xN xv
*di
fori _
So
#690
C 1 # /z \\ - 13
--v-
13 mm
: / / \ 1 TACK WELD
pLEXIGLASS I I ) I

C I \ / t 3
VP N
V
at°W.
II rff 0 efie
(..---

0
\1 smm /d : OS '
It 6\--Am-N4-,\-v

SEALED WITH SCREEN PACKING MATERIAL BAFFLE METAL SCREW

"SILICONE SEAL"

Figure 3. Section A-A detail of multiple stage crosscurrent packed column.

 12

The screen, which formed two sides of the packed section, was

made of 0.88 mm stainless steel wire woven into a 6.35-mm (1/4-in)

square mesh. The projected open area of the screen was 74 percent.

Steel tie-rods, 178 mm long and 3 mm in diameter (7 in by 1/8-in),

were positioned at 127 mm intervals along the walls and wired to the

screen to properly position the packed section in the center of the

column. Sixteen-millimeter (5/8-in) polyethylene pall rings of

porosity 87 percent were used as packing material.

Metal baffles 127 mm by 64 mm by 3 mm (5 in x 2 1/2 in x 1/8 in)

were positioned in the open plenum on either side of the packed section.

The baffles were made of stainless steel and gasketed with rubber to

prevent air and water bypass. Two 64-mm long, 3-mm diameter rods were

tack-welded to the bottom of each baffle and inserted into 6-mm deep

holes in the plexiglass sides of the column to hold the baffles in

place. The column was operated using three different baffle arrange-

ments as shown in Figure 4. When the column was operated in a standard

countercurrent manner, the two open sides of the column were filled

with wood blocks and gasketed with rubber to prevent air bypass; this

produced a conventional countercurrent packed column that was 127 mm

by 127 mm cross-sectional area and 1.52 m in height. All four of these

arrangements were made without disturbing the packed section.

Figure 5 is a schematic diagram showing the equipment arrangement.

An air-ammonia-water system was investigated. The air and ammonia

were introduced as the gas phase and water served as the liquid phase.
 13

ARRANGEMENT NO.

CENTER SECTION
PACKED WITH PALL RINGS
SOLID LINES
DENOTE BAFFLE
POSITIONS

WOOD BLOCK

12 6 4 NUMBER OF BAFFLES

127 min 254 mm 381 mm VERTICAL DISTANCE

BETWEEN BAFFLES

Figure 4. Baffle arrangements.

 AMMONIA AND
AIR OUT

CROSSfLOW
COLUMN

/1\
r
AMMONIA 1 *.
ROTAMETER
1

---1 1

-OT 1

1
1

1
LIQUID
ROIAMETER O
1
1 1
AMMONIA
1
IN
1 *1
1 1

i
WATER
1

1
1
IN
T
ORIFICE 1

MANOMETER

LIQUID LEVEL
T SCALE

AIR AIR BYPASS

IN

MANOMETER SAMPLE LINE

AIR BLOWER
DRAIN!

Figure 5. Equipment arrangement schematic.

 15

The air was supplied by a Sutorbilt Horizontal California Series

B positive displacement blower that is designed to deliver a constant

volume of clean air against varying pressures.. Eighty-three millimeter

inside diameter (3 1/4 in ID) P.V.C. plastic pipe rated at 862 kPa was

used for the air supply and exit gas stream. Air flow rates up to
3
425 m /s were controlled with the use of an air bypass line and a

two-valve system located about 2.4 m upstream from the column. These

two values were connected with a bar to prevent accidental closing of

both valves at the same time. The blower was also protected with a

pressure safety valve set at 125 kPa located on the blower outlet line.

Air flow rates were measured with sharp-edged orifices sized for the

appropriate flow range and installed with radius pressure taps according

to A.S.M.E. standards [21]. These orifices were calibrated using a

Meriam Laminar Flow Meter, Model 50MC2-4F, and an accompanying Meriam

micromanometer. Orifice upstream pressure and pressure drops were

measured with water manometers. Upstream air temperature was measured

with a mercury therometer inserted in the air line.

The anhydrous ammonia was supplied from a 68-kg capacity liquid

storage tank and delivered through a stainless steel regulator as a

gas. The ammonia was introduced into the air stream about 1/3 meter

upstream of the column. Thirteen-millimeter (1/2-in) plastic tubing

and nylon Swagelocks were used in the ammonia system to avoid corrosion

problems. Flow rate was controlled by a needle valve and measured

with a rotameter. Pressure and temperature of the ammonia were measured

immediately upstream of the rotameter with a mercury Manometer and metal

thermometer, respectively.
 16

The liquid phase was supplied by either city water lines or a

constant-head tank, depending upon the flow rate desired. Water flow

rate was controlled by gate valves and measured with two separate

rotameters which had a combined range of 82 to 8097 kilograms of water

per second. These rotameters were calibrated using the bucket and

stopwatch method. The piping for the water was galvanized 25 mm

(1 in), schedule 40 pipe and a plastic shower head dispersed the water

over the top of the packing.

Column pressure drop was measured with a water manometer and

pressure taps were located to cover 1.27 m (50 in) of packed height.

The same combination of manometers were used for this measurement and

for orifice pressure readings. A schematic of this arrangement and

necessary value operations is included as Figure 6.

 P
P top
atm
-1

WATER

112i; VALVE
OPERATION TO
MANOMETER MEASURE
OPEN CLOSE

P1

6.35 mm (I/4-inch) 4.5 2,3 AP

orifice
STAINLESS STEEL
VALVES 2,3 4,5 AP
column
2,4 3,5 P -P
I top

6.35 mm (1/4-Inch) Pbottom

PLASTIC TUBING

CROSSFLOW
COLUMN

ORIFICE WITH
RADIUS TAPS

Z;
Figure 6. Manometer System Schematic.
 18

IV. THEORY OF CALCULATIONS

The calculations necessary to complete a material balance around

the absorption column will be discussed below.

4.1 Inlet Flow Rate of Air

The orifice placed in the pipe is a constriction which increases

the velocity of the fluid flowing through it. This increases the

kinetic energy at the point of constriction and, from an energy balance,

there must be a corresponding reduction in pressure. The rate of dis-

charge from the constriction can be calculated knowing the pressure

reduction and certain parameters about the orifice. The equation for

compressible flow through an orifice has been given as [21]:

W=YKAA(Pi-P2)pi (1)

where A2 is the cross sectional area of the orifice opening; P1 and P2

are the pressures upstream and downstream from the orifice, respect-

ively; p1 is the density of the ambient air at the upstream pressure

and temperature; K is equal to C/117where C is the orifice coeffi-

cient and a is the ratio of orifice diameter to pipe diameter; and Y

is an expansion factor. A correlation for the expansion factor for

sharp-edged orifices and radius taps is given by A.S.M.E. as [21].

Y = 1.00 - (0.41 + 0.3584) V3p1 (2)

 19

where AP is P1 -P2, the pressure drop across the orifice, and y is the

ratio of specific heats, assuming ideal gas behavior (y=1.4 for air).

Since all pressures are near atmospheric and temperatures are near

room temperature, the ideal gas approximation may be used for evaluating

the air density:

P1
p - M (3)
1 RT
1

where P and T are upstream pressure and temperature; R is the gas

1 1

constant; and g is an average molecular weight of the gas stream. The

composition of this air and water mixture can be determined using a

sling psychrometer. Measurement of the dry bulb and wet bulb temper-

atures yield z, the kilograms of water per kilograms of dry air. The

average molecular weight can be determined from the composition as

follows:

yaii.(28.97) + yH20(18.01)

but
Yair YH20

z (28.97/18.02)
and
''H2O 1.+z (28.97/18.01)

=
28.97 + 18.01(1.609 z)
therefore
1.00 + 1.609 z

where z = kg H2O / kg dry air (4)

Substitution of these relations into equation (1) and dividing by

g to get a molar flow rate yields the flow rate of air in mol/s:
 20

5.51 C A2
M 1.00-(0.41+0.35134) DP P1 AP
air (5)
1
1.4 T1
1
11:177

The constants are determined for each of the three orifices as tabu-

lated below:

A
2
d 13
2
mm (d/D) m
16.00 0.194 201.1E-06 0.57
25.40 0.308 506.7E-06 0.57
50.80 0.615 2027.0E-06 0.61

These constants can be combined into a simplified version of equation

(5) as:

m = K1 1.00 -K2 4-1.\d/P1 6'1)

air 1 2 P M T (6)
1

where the constants K and K for each orifice are given as:
1 2

d K OPERATING RANGE
2
.-rh1mm mole/s
100 1.779E-03 0.29321 0.159-0.393
25.40 4.500E-03 0.29510 0.393-1.172
50.80 20.721E-03 0.32871 0.975-4.891

28.97 + 18.01(1.609 z)
and M
1.00 + 1.609 z

z = kg H2O / kg dry air

AP = orifice pressure dropp
P = upstream pressure
1
T = orifice upstream temperature
1

4.2 Inlet Flow Rate of Ammonia

The flow rate of gaseous ammonia into the column is measured

using a Gilmont Model F1500 rotameter that has been calibrated at the
 21

factory for air and water flowing at 101.3 kPa and 294 K (1 atm, 70 °F).

The theory of correlation based on a dimensionless combination of

Stoke's law and flow coefficients was used to convert the calibration

to ammonia. Only the density and viscosity of ammonia at the conditions

of flow are required [22].

The calibration chart supplied with the rotameter is a linear plot

of the percentage change in diameter ratio, Rd, versus the percent of

scale reading of the meter. This line has the following equation:

R = 1.00 + 0.263(% of scale) (7)

d

Using this, the Stoke's number may be calculated:

3 2
St = 1.042E-06 Wf (pf-p) pRd /11 Pf (8)

where Wf is the weight of the float, 2.715E-03 kg; pf is the density

of the float, 2530 kg/m3; p is the density of the ammonia at flow

3
conditions, kg/m ; and p is the viscosity of ammonia at flow condi-

tions , Pa.s. The viscosity and density data for ammonia as a function

of pressure and temperature is available in the literature [4, 23].

The viscosity is only a function of temperature and a simple linear

interpolation of the data yields:

u = 9.2E-06 + 2.0E-08 (1.8 TNH3 - 499.67) (9)

A similar linear interpolation for density as a function of both

pressure and temperature, based upon a linearly varying multiplier of

the ideal gas law, yields:

 22

p = 1.291E-03 P /(T v)
NH3 NH3

where v = 0.6222 + 8.0E-05 (1.8 TNH3 - 499.67)

+ 6.0E-04 (15.0 - Pmu /6894.75) (10)

3
The flow coefficient, C for the rotameter is obtained as a func-
r'
tion of Stoke's number and R from a correlation chart supplied with
d
the meter. However, an analytical equation was also supplied to

calculate Cr:

C
Vb +T4ac - b
r 2a

where a = 3.08 log(Rd) - 1.25

b = 3.83 - 1.17 log (Rd)

c = log (St) - 0.111 log(Rd)

Since the value of the diameter of the float, Df, is 0.0127 m, the

volumetric rate of flow in m3/s is:

11 141 cfPf P)
q = 3.924E-02 Df Cr ' R (2 0+R /100.)(12)
d d

For ammonia at operating conditions encountered in this work, density

3
ranges from 0.6 to 0.8 kg/m . By using the approximation that

(2530-p) = 2530 and substituting in equations (8) and (12) for all

constants and converting to a molar flowrate, one obtains:

2
St = 2.829E-09 oRd /1.1 (13)

and th = 1 528E-03 Cr Rd1(2.0 + Rd/100.) (14)

NH3'in
 23

where C Rd, and p are found using equations (7), (9), (10), (11),
r'

and (13). The flow rate is in mol/s. A calibration curve at 283 K

and 103.4 kPa is given in Appendix A.

4.3 Inlet Flow Rate of Water

Two liquid rotameters were calibrated and used to measure the

flow rate of water into the column. The calibration curves are given

in Appendix A. These curves have calibration slopes of 0.1574 and

1.091 mot H 0/s(% of scale). The flow rate of water in mol/s is then
2

given by:

MH20'in = F (% of scale) (15)

where F = 0.1574 for the smaller rotameter

and F = 1.091 for the larger rotameter

4.4 Exit Water and Air Flow Rates

The water flow rate going out of the column in the exit liquid

stream is equal to the amount of water entering through the rotameters

less the amount that was absorbed in air and left with the exit gas

stream. If the humidity (z kg H2O / kg dry air) is obtained for the

exit stream, the amount absorbed by the air would be approximated by:

H2 0absorbed. 61airo.n 1.609(z - z )

out in

This amount was never more than 0.2 percent of the liquid flowrate in

all of the data taken. Therefore a good approximation is:

= m (16a)
mair,out airon
 24

and
thH 0 out thH 0 in (16b)
2 ' 2 '

4.5 Exit Flow Rate of Ammonia in the Liquid Stream

The amount of ammonia leaving in the liquid stream is determined

by a titration technique using sulfuric acid, sodium hydroxide, and

bromocresol green indicator. Approximately 0.2 normal solutions of

acid and base are used.

A sample of the exit water stream is obtained by quickly pipetting

an amount into a known volume of the sulfuric acid solution. This

"traps" the ammonia in the solution by the reaction:

H2SO4 + 2NH4OH (NH4)2SO4 2H20

Bromocresol green indicator is added and then the solution is back-

titrated with the sodium hydroxide until the color changes from yellow

through green and then to blue. The end-point is the intermediate

green color. Using V as the volume of solution, m3, and N as the

3
normality of the solution, kilogram-equivalents/m of solution, the

total equivalents of sulfuric acid at the start is equal to Nu cnA x

V The number of kilogram-equivalents of sodium hydroxide used

H SO
2 4
during the titration is equal to N V Therefore, the amount
NaOH NaOH*
of sulfuric acid that reacted with the ammonia is the difference

between the two. This is the number of kilogram-equivalents (or kmole

for NH ) of ammonia in the sample of exit water. Knowing the volume

3

of the sample taken and using the density of water as 1000 kg/m3,

one obtains in mol/s:

 25
N N V
H SO VH SO - NaOH NaOH
2 4 2 4
(18.01) (17)
thNH1,out 1000 V (thm 0 out)
sample 2 '
liquid

4.6 Exit Flow Rate of Ammonia in the Gas Stream

The amount of ammonia remaining in the gas stream is calculated

by difference:

m
NH1,out 11 NH ,in mNH1,out
3
gag liquid (18)

4.7 Number and Height of Gas-Phase Transfer Units

The concept of a transfer unit is commonly used for interpreting

and correlating mass transfer data. It is based on the idea of dividing

the packed section into a number of contact units called transfer units.

The depth of packing required by a single unit is called the height of

one transfer unit. Then the total height of the packed section, Z, is

given by:

Z = N H
tOG tOG (19)

where N is the number of overall gas transfer units and H is the

tOG tOG

height of an overall gas transfer until. The total height of packing,

Z, is 1.52 m. The number of overall gas transfer units is [5]:

Yin
r (20)
NtOG
JYout

where y* is the solute concentration in the gas corresponding to

equilibrium with the bulk liquid concentration x and (1-y)*m is the

logarithmic average of 1-y and 1-y*. In this study equation (20) can
 26

be further simplified because both the operating and the equilibrium

lines are straight (dilute solutions), Henry's law is valid (y*=Hx),

and absorption heat effects are negligible. Under these conditions

equation (20) may be evaluated explicitly as:

H
-
In [yin
(
y - H
1 - +
out xin (21)
NtOG
1 - 1

where the absorption factor is A = L/HG, L is the total liquid molar

flow rate, G is the total molar gas flow rate and H is Henry's law

coefficient. Since the concentrations of ammonia in the inlet liquid

stream is zero and is dilute in other streams, y = Y, and L/G =

mH20/mair'
A Henry's law constant of 0.95 is obtained from the

solubility data for ammonia at the conditions used in this study [4]

Equation (21) can then be simplified for dilute ammonia solutions:

lnKYin)( 1 - A
1) +
1 I
Yout
N = (22)
tOG 1
1 - -A

where A = m / 0.95 M
H 0 air
2

The height of a transfer unit can then be calculated by first finding

and then using equation (19) to evaluate HtOG.

NtOG
 27

V. OPERATING PROCEDURE

5.1 Preliminary Operating Procedures

The column and all support equipment were constructed in the

manner and dimensions previously described. A routine maintenance

check on the Sutorbilt blower was done and repeated after every one

hundred hours of operation.

The column's upper access port and top retainer screen were removed

and the central section was filled with the pall rings. This packing

material was dumped into the column in a random manner. During this

operation, the column contained no liquid and was not shaken. Once

placed in the tower, the packing material was not disturbed for the

remainder of the study. During this investigation the column was

operated as both a crossflow and countercurrent device. The necessary

internal arrangements as shown in Figure 4 were accomplished by removing

the 127 mm by 2.44 m plexiglass sides. The baffles or wood blocks were

arranged properly; the plexiglass sides were repositioned; all butt

joints were sealed with clear Silicone Seal; and, the sides were secured

by tightening the metal screws. The Silicone Seal was allowed to dry

overnight before continuing.

Solutions of sodium hydroxide and sulfuric acid of approximately

two-tenth normal were prepared and titrated against standards to deter-

mine exact concentrations.

 28

5.2 Procedure for a Data Run

After arranging the internals of the column, the following

procedure was followed to generate the required data.

Start-up. The following steps were performed:

1. Open the liquid outlet valve.

2. Open manometer system valves 1, 4, and 5.

3. Install the appropriate orifice for the desired flow

range (beveled edge on the downstream side).

4. Turn on the air blower and open air control valve

mechanism about one-third. Allow about twenty minutes

before data collection for inlet air temperature to

stabilize.

5. Open water control valve for small rotameter until

float registers about ten percent of scale.

6. Open main valve on the ammonia tank and slowly open the

regulator until 70 to 100 kPa (10 to 15 psi) registers

on the outlet pressure meter. Keep the rotameter

control valve closed and check ammonia system for

leaks (ammonia produces a white cloud in the presence of

concentrated hydrochloric acid solution).

Data Collection. An operating gas rate and liquid rate had to

be chosen. It was easiest to first set the air control valve, calculate

the air flow rate using equation (6), choose a liquid-to-gas ratio,

back-calculate the liquid rotameter setting from equation (15), and

then set that water flow rate by adjusting the appropriate liquid

rotameter control valve. Approximately ten minutes was allowed for the
 29

flow patterns and liquid hold-up within the column to fully develop and

then the liquid outlet valve was adjusted until the liquid flow out of

the column matched the inlet liquid flow. A liquid level of from 50

to 100 millimeters was maintained in the bottom of the column. The

ammonia flow rate was set (at two percent of air flow rate for most

data runs) using the calibration curve at 283 K and 103.4 kPa given in

Appendix A to set the ammonia rotameter. This calibration curve is

reasonably accurate for setting a flow rate, but the exact flow rate

must later be calculated using equation (14). About 20 to 30 minutes

was allowed for the column to reach equilibrium, depending on the

flow rates.

The following data were taken for each run and the instruments

were re-read several times during the run to check for any fluctuations.

The barometric pressure in the immediate area, P was read from a

a '

mercury manometer. The following data were recorded using the equip-

ment described in the previous section:

T T
1 NH3
P P
1 NH3
AP AP
col

T T for inlet and exit gas streams

w' d

% of scale for water rotameter

I of scale for ammonia rotameter

A sample of the exit liquid stream was obtained by quickly

pipetting 5 or 10 ml into 25 ml of 0.2 normal sulfuric acid.

Bromocresol green indicator was added and the solution was back-

titrated with 0.2 normal sodium hydroxide. After the sample was taken,
 30

the liquid in the bottom section of the column was drained and then

allowed to refill to its prior level. This sampling procedure was

repeated at least twice more at fifteen minute intervals until adequate

duplication of results was obtained. This completed one data run.

The steps outlined in this sub-section were then repeated for

new air and water flow rates.

Shut-down. After the desired data were taken, the following

procedure was followed to shut-down the column.

1. Close the valve on the ammonia tank and then close the

regulator valve and the ammonia rotameter control valve.

2. Turn off the water supply and drain all water lines.

3. Turn off the air blower and record operating time in

the maintenance log.

4. Close all valves in the manometer system to prevent

evaporation of manometer fluid.

 31

VI. ANALYSIS OF ERRORS

The significance of conclusions based upon the numerical results

of this study can be properly determined only after careful considera-

tion of the reliability of the data. Several types of errors can

contribute to the inaccuracy of the results; efforts were taken to

minimize the errors associated with poor equipment performance and

those that arise as a result of approximations and assumptions made

in the theoretical development of equations. However, as in any inves-

tigation, the results obtained were a compromise between desired

accuracy and minimization of time and expense required to obtain that

accuracy.

All measurements may be classified as either direct or indirect.

A direct measurement is made whenever the magnitude of the measured

quantity is determined by direct observation from the measuring instru-

ment. Estimates of errors associated with these measurements are based

on the reported errors given in operating manuals or on the scale

increments of the instrument; estimated error is ordinarily one-half

the scale increment that is easily read. The magnitude of an indirect

measurment is determined by calculation from the magnitudes of other

quantities directly measured using some functional relationship that

exists among the quantities. The estimate of error is more difficult

to determine for an indirect measurement than for a direct measurement

since the errors in the direct measurements may either augment or offset

each other's effect on the error of the calculated result, depending

upon their signs and the form of the functional relationship.

 32

Several methods are available for estimating the propagation of

errors in a function of several variables [24]. When the desired

quantity Q is a function of several variables ql, q2 qn

Q f(q1, (23)
q2 qn)

then a good approximation for AQ, the probable error in Q, is given as:

3f 3f 3f
(24)
AQ 7qlAql 76q2
q2
-3q-Aqn

where Aqn are the estimated errors in ql, q2 qn. The

'ql, Eq2
maximum probable error is obtained by arbitrarily assigning correct

signs to the individual error estimates so they will all act to

increase the value of iQ in equation (24). The individual errors may

offset each other and this may be considered by calculating the most

probable error based on a normal distribution of the individual errors:

2 2
42 ) (31)))6,,,2 41(7 )A,2
=
ql q2 42 aqn mn

However, Sherwood [24] recommends using the maximum probable error when

the individual accuracy measures are based, not on a series of observa-

tions, but on estimates alone. This is the most conservative practice

and will naturally indicate a larger value for the error estimates than

would be obtained by applying the probability formula. All error

estimates given below are maximum probable errors as estimated for

direct measurements and as calculated for indirect measurements using

equation [24]. For reference, the most probable error is usually about

three-fourths the maximum probable error for this data.

 33

Table I summarizes the maximum probable errors for the direct

measurements and Table III summarizes those for the indirect measure-

ments. The error is also reported as a percentage of the measured

quantity since this varies depending on the magnitude of the measure-

ment.

Examining the percent maximum error columns of Table I and II

show the accuracies of most values are reasonable and acceptable.

The largest errors arise from the calculation of M This

NH out °as'
3 '-
operation of subtraction is the source of a great loss of accuracy

because the result is the small difference between two relatively large

numbers each of which is subject to error. The inaccuracy of

SH3out,gas produces inaccuracy in Y which is then used to calculate

out
and This error could be reduced if or Y
NtOG HtOG. NH3out,gas out
could be measured directly. However, since the absorption efficiency,

% NH removed, is not subject to this inaccuracy and can also be used

3

for comparison of different columns, the investment for sophisticated

gas analysis equipment was not justifiable.

Another indicator of the accuracy of a result is its reproduci-

bility. If a statistically significant amount of duplicate runs could

be made, a reasonable estimate of the error associated with results

could be calculated. Since each run required over an hour to

complete, this approach was not pursued. However, some runs were

duplicated to verify results. These duplicate runs were made on

different days and the major results are reproduced in Table III.
 34

TABLE I. Maximum probable error of direct measurements.

Maximum
Smallest Largest Probable % Error
Measurement Value Value Error Largest Smallest

d 16.00 50.80 ± 0.05 0.3 0.1

D -- 82.55 ± 0.25 -- 0.3

Z -- 1.52 ± 0.005 -- 0.3

P a tm 99,990. 102,660. ± 68. 0.1 0.1

T 317.15 329.15 ± 0.50 0.2 0.2

1

102,500. 112,000. ± 180. 0.2 0.2

P1

AP 286. 6,146. ± 12. 4.2 0.2

T T 285.15 296.15 ± 1.0 0.4 0.3

w' d

1 279.82 289.26 ± 0.56 0.2 0.2

NH3
P 101,900. 106,500. ± 200. 0.2 0.2
NH
3
% of scale
water rotameter 7.33 35.67 ± 0.5 6.8 1.4

% of scale
NH rotameter 17.0 68.0 ± 0.5 2.9 0.7
3

AP 30. 3,470. ± 12. 40.0 0.3

col
 35

Table II. Maximum probable error of indirect measurements.

Maximum
Smallest Largest Probable % Error
Quantity Value Value Error Largest Smallest

a 0.194 0.615 ± 0.0025 1.3 0.4

C 0.57 0.61 ± 0.005 0.9 0.8

K1 1.779E-03 20.721E-03 ± 0.285E-03 16.0 1.4

K 0.29321 0.32871 ± 0.00360 1.2 1.1

2

A 201.1E-06 2027.0E-06 ± 4.0E-06 2.0 0.2

2

Ti 28.81 28.87 ± 0.02 0.1 0.1

11 9.28E-06 9.62E-06 ± 0.05E-06 0.5 0.5

P 0.730 0.798 ± 0.003 0.4 0.4

H --- 0.95 ± 0.01 --- 1.1

0.47 2.81 ± 0.04 8.5 1.4

mair
M ± 0.09 7.8 0.9
H20 1.15 10.24

L' 1.29 5.71 ± 0.07 5.4 1.2

G' 0.84 2.51 ± 0.05 6.0 2.0

m 0.0130 0.0570 ± 0.0006 4.6 1.1

NH in
3,
0.0097 0.0500 ± 0.0010 10.3 2.0
mNH3,out,liquid
0.0012 0.0181 ± 0.0016 133.3 8.8
mNH3out,gas

Y. 0.0193 0.0423 ± 0.0012 6.2 2.8

in

X 0.0041 0.0146 ± 0.0007 17.1 4.8

out
Y 0.0011 0.0077 ± 0.00T5 136.4 19.5
out
% NH Removal 62. 96. ± 6. 9.7 6.3
3

1.54 4.98 ± 0.83 53.9 16.7

N
tOG
0.31 0.99 ± 0.25 80.6 25.3
H
tOG
 36

TABLE III. Results of duplicate runs.

Date of % NH3
N H
Run # Run LI G' Removal tOG tOG

23 12-17-79 3.36 1.34 92.1 2.98 0.510

24 12-28-79 3.34 1.34 92.4 3.02 0.503

31 1-10-79 2.59 2.08 89.3 3.19 0.476

32 1-20-79 2.59 2.09 90.6 3.45 0.441

27 1-02-79 1.93 1.56 83.8 2.48 0.613

34 1-20-79 1.93 1.54 84.2 2.54 0.598

55 2-18-79 1.32 2.09 72.2 2.41 0.631

56 2-19-79 1.29 2.06 71.1 2.27 0.669

 37

The differences between the results of the duplicate runs, expressed

as a percentage are given in the table below. This duplication of

results suggests an accuracy exists that is better than indicated by

the maximum error analysis given in Table II. However, the variation

associated with N and H is still substantially larger than that

tOG tOG
for the absorption efficiency. The results of this study will be

presented in terms of absorption efficiency and HtOG with the under-

standing that these estimated errors are an essential part of their

values.

Analysis of Duplicate Runs

% Difference in Results
N H
Run #'s % NH Removal tOG tOG
3

23, 24 0.3 1.3 1.4

31, 32 1.5 8.2 7.4

27, 34 0.5 2.4 2.4

55, 56 1.5 5.8 6.0

 38

VII. PRESENTATION AND DISCUSSION OF RESULTS

The results of this experimental study are presented and discussed

below. No fluid dynamic observations are presented because the flow

characteristics of the expeimental multiple stage crosscurrent packed

column are thoroughly discussed by Thibodeaux [16] and Golshani [17].

No liquid holdup or liquid phase drift angle data were taken. The

column pressure drop and overall mass transfer data that were taken

are tabulated in Appendix C and also presented graphically in Figures

7 through 17.

7.1 Definition of Baffle Spacing Variable a

The effect of baffle spacing arrangement was the independent

variable of primary interest. A baffle spacing variable, a, was

chosen that could be related to different sized crosscurrent towers.

Thibodeaux [16] and Golshani [17] defined a as the "ratio of gas flow

area to liquid flow area". This ratio is unity for countercurrent

columns, but an extra variable found in crosscurrent columns, baffle

spacing, uncouples these flow areas. The exact areas that are

"available" for gas flow and liquid flow with each baffle spacing is

debatable. Therefore, a is redefined here as the "vertical distance

between baffles (the two successive baffles are on opposite sides of

the packing) divided by the horizontal width of packing between the

void spaces". The numerical value of this a is identical to that as

defined by Thibodeaux and Golshani. The values of a for each arrange-

ment are given in the table on the following page.

 39

Arrangement Number of Vertical Distance Width of

Number Baffles Between Baffles Packing a

1 12 127 mm 127 mm 1

2 6 254 127 2

3 4 381 127 3

7.2 Basis for Comparison of Crosscurrent

and Countercurrent Data

The cross-sectional area used for calculating L' and G' determines

the basis used for comparing the crosscurrent and countercurrent col-

umns. The area used for a countercurrent column is the total cross-

sectional area of the column or the packing. These areas are

identical for a countercurrent column but are different for a cross-

current column. Since the performance of the crosscurrent column is

being compared against the standard countercurrent column, the most

conservative procedure would be to use the cross-sectional area of the

total column in both cases. This would include the packing and void

spaces for the crosscurrent column. Therefore, comparisons would be

made on the basis of identical flow rates into columns of identical

overall dimensions.

Actually, the design of the experimental tower made it impractical

to fill the entire column with packing and operate it as a counter-

current column. The tie-rods would cause an uneven packing and create

void areas. The void spaces on the side of the packing were filled

with wood blocks creating a column of half the cross-sectional area

of the crosscurrent column. The ranges of molar flow rates of the

gas and liquid phases were correspondingly reduced by one-half.

 40

This produced mass flow rates per cross-sectional area that were

comparable to those of the crosscurrent column operation.

The data in Appendix C and Figures 7 through 17 are presented on

the basis of total column cross-sectional area. This basis is used in

sections 7.3 and 7.4 to compare the performance of the crosscurrent

and countercurrent columns. A comparison of absorption equipment

based on total size definitely has some merits, but the total cross-

sectional area of the experimental crossflow column was arbitrarily

set; the void space and thus total cross-sectional area could be

increased with little effect on performance. However, this open area

to packed area ratio should have some influence as it is decreased and,

as a limiting case, the crossflow changes to counterflow operation.

Since this spacing was arbitrarily set for this investigation, a better

basis for comparing the crosscurrent and countercurrent column might

be one that excludes this variable. A comparison based on cross-

sectional area of the packing would numerically exclude the open area

but would not necessarily be the same for different open area to packed

area arrangements. Since the packed area for this experimental

column was one-half the total column cross-sectional area, the data

for the crosscurrent arrangements can be converted to mass flow

rates per cross-sectional area of packing by multiplying L' and

G' by two. Section 7.5 uses this basis to compare the different

column arrangements.

7.3 Gas Phase Pressure Drop

The pressure drop data collected by Thibodeaux [16] and Golshani

[17] included pall rings, but were for liquid rates from 6.1 to 45.7
 41

kg/sm 2 . All data for this study was below this range; therefore,

experimental pressure drop data were obtained for liquid rates 1.76,
2
3.51 and 5.27 kg/sm . These data are tabulated in Appendix C and

presented in Figures 7, 8 and 9 as the conventional logarithmic plots

of G' vs. AP that is employed for ordinary packed beds.

col

The gas phase pressure drop for the countercurrent operation is

less than that for crossflow with a = 1, but larger than for crossflow

with a = 2 and 3. Loading occurs at lower gas rates for a.= 1 than for

the countercurrent case. This advantage for the conventional counter-

current operation seems to hold for increasing flow rates provided that

the operating range is not near the countercurrent flooding point.

However, at extremely low ranges all three crossflow arrangements

appear to have lower pressure drops than the countercurrent arrange-

ment.

A close comparison of the differences between Figures7, 8, and

9 suggests that increasing the liquid flow rate has a more pronounced

effect on increasing pressure drop for all three crossflow arrangements

than it does for the countercurrent case. There is no evidence that

indicates if this pattern will continue at higher flow rates.

7.4 Mass Transfer Efficiency

A total of 57 data runs (not including the pressure drop data

previously described) were made using the experimental column. The

absorption efficiency and height of a transfer unit were calculated

from each set of measurements. Sample calculations for Run 30 are

included in Appendix B. The first ten runs were preliminary runs and _
 42

4000

2000

OBLP
1000
800

600
=

0- 400
C-
C
Ct"
C

N 200
2
C
fa_

100
80

60

40k-At

L OBLP = Observed
Loading Point

20. I I
I. ,
1 I
I Ii; I ill!! i t 1 1 1

0.8 0.9 1.0 1.5 2.0 2.5 3.0 4.0

GAS RATE kg/s-m2

Figure 7. Pressure drop vs. gas rate; liquid rate = 1.76 kg/sm2.
 43

4000
I-
1

2000

100

800

600

400

20

100

80

60

40
OBLP = Observed
Loading Point

20
0.8 0.9 1.0 1 5 2 0 2 5 3.0 4.0
2
GAS RATE kg/sm

Figure 8. Pressure drop vs. gas rate; liquid rate = 3.51 kg/sm2.
 44

2
GAS RATE kg/sm

Figure 9. Pressure drop vs. gas rate; liquid rate = 5.27 kg/sm2.
 45

are not included in the following discussion. However, the results

of all 57 runs are tabulated in Appendix C. Runs 3 and 16 had

operational difficulties that affected their results and are not

included in the discussion.

Absorption Efficiency. The absorption efficiency was plotted as

a function of the gas rate for L/G ratios of 1.0, 2.0 and 4.0 for

the three baffle spacings and the countercurrent operation. These

results are presented in Figures 10 through 13. The data are replotted

on the same axes in Figures 14 and 15 to compare the countercurrent

and crosscurrent operations. The lines representing L/G = 1.0

are dashed lines because insufficient data were taken to justify an

exact slope. The relative position of this data set with respect to

the larger L/G ratios is significant and should be considered.

The absorption efficiency ranged between 60 and 96 percent

removal for the operating conditions used in this study. Increasing

L/G ratio significantly improves the absorption efficiency in all

cases. At larger values of a this effect is larger - the L/G lines

are farther apart. The comparison of absorption efficiencies for all

operating modes, as presented in Figures 14 and 15, shows some

interesting results. At L/G = 2.0 the crossflow column with a = 1

and 2 appears to be more efficient at removing ammonia. However at

L/G = 1.0 only the a = 1 arrangement is more efficient than counter-

current. At the highest L/G ratio, 4.0, the a = 1 arrangement again

appears superior but the other crossflow arrangements are possibly

 46

better than countercurrent and a = 1 arrangement at extremely low gas

rates (less than 1.2 kg/sm2).

Height of a Transfer Unit. The height of a gas phase transfer

unit was calculated for all runs although the error analysis indicated

a higher degree of inaccuracy for this value than for the absorption

efficiency. Both methods are indicators of mass transfer efficiency

and should give the same comparative results. Graphs of HtOG as a

function of gas rate are presented for the a = 1 arrangement, L/G =

1.0, 2.0 and 4.0 in Figure 16 and a comparison of counter current and

all crossflow arrangements at L/G = 2.0 is given in Figure 17. A

lower H value corresponds to better mass transfer efficiency, and

tOG
Figures 16 and 17 confirm the results presented for absorption effi-

ciency. Since these two measurements of mass transfer efficiency are

not indpendent of each other, this agreement merely verifies the

method used to calculate The values of HtOG are included in

HtOG.
this study for comparison to data published in the literature since

H values are a more commonly used measure of mass transfer effi-

tOG
ciency than the percent solute removed.

7.5 Comparison Based on Packed Area

The pressure drop vs. gas rate data are reproduced in Figure 18
2
for L' = 3.51 kg/sm where all mass flow rates are given per cross-

sectional area of packing. The absorption efficiency vs. gas rate for

all arrangements at L/G = 2.0 is also given in Figure 19 using the

packed area as a basis for mass flow rates. When the packed area is

used as a basis to compare crosscurrent and countercurrent data the

relative position of the crosscurrent data shifts to the right.

 47

Figure 18 illustrates that all crosscurrent arrangements have

significantly less pressure drop per meter of packed height than the

countercurrent arrangement. Also, the crosscurrent arrangements can

accommodate significantly larger gas flow rates for a specified press-

ure drop than can the countercurrent column. Loading occurs at lower

gas flow rates in the countercurrent column than in the crosscurrent

column.

When flow rates are based on the cross-sectional area of packing

all crosscurrent arrangements are less efficient than countercurrent

at removing ammonia at a given gas rate. However, Figure 19 also

illustrates that for a given absorption efficiency and the same amount

of packing material the crosscurrent arrangement can process signifi-

cantly larger flow rates.

 48

1 1 1 1 1 1 1 1 I .1
95 1

MOO

2.0
a 85

0-, I
=
z
80
>-
2=
1-1-1

Z;
G: de
LL.

75
CD
G
1 0 /

CD de
(di
Ct3
cc )13
70

65

- -,
- -
1 i I I I I 1 I I 1 I I I I t I I I_ I
60
0.5 1.0 1.5 2.0 2.5
2
GAS RATE kg/s.m

Figure 10. Absorption efficiency vs. gas rate; baffle arrangement

#1, a = 1.
 49

95

90

ae
80

75

70

65

60
0.5 1.0 1.5 2.0 2.5

GAS RATE kg/sm2

Figure 11. Absorption efficiency vs. gas rate; baffle arrangement

#2, a = 2.
 50

95

90

85

80

-
W_

75

CD
I-
CL
CD
v)
IDD

c= 70

65

60
0.5 1.0 1.5 2.0 2.5
2
GAS RATE kg/sm

Figure 12. Absorption efficiency vs. gas rate; baffle arrangement

#3, a = 3.
 51

95
I I I i I IIII
MP

4.0
r"'
90

...

L., 85
z.
c
x
w
CC
MP
M
=
Z
aQ
80

fl.

NM

75
IMO

Me
de
= 1.0 SO
de 5o Il
70
/

OW

65
=11.1

IMP MEN

60 '
I
t I I
I I I I I i 1 1 1 t 1 I 1

0.5 1.0 1.5 2.0 2.5

2
GAS RATE kg/sm

Figure 13. Absorption efficiency vs. gas rate; countercurrent

arrangement.
 52

95
I t I 1 lIt 11 1
.1 t 1

IMO

Mak .1.00

"WM 11

cc =1
2

AM!

e--

85

U.I
CC

ae 80
1

Li- 75
U./

c.

70,
14

Oa= 1

Aa= 2

65 MN.
1:::1 a= 3

countercurrent
Mae

60, '
I 1
1 1 I I , I .
t 1
1 I t III
0.5 1.0 1.5 2 0 2.5
2
GAS RATE kg/sm

Figure 14. Absorption efficiency vs. gas rate; comparison of

arrangements at L/G = 2.0.
 53

95

90

0a=1
Aa= 2
1:1 a = 3

countercurrent

65

60
0.5 1.0 1.5 2.0 2.5
2
GAS RATE kg/sm

Figure 15. Absorption efficiency vs. gas rate; comparison of

arrangements at L/G = 1.0 and 4.0.
 54

0.5 1.0 1 5
2
GAS RATE kg/sm

Figure 16. HtoG vs gas rate; baffle arrangement #1, a = 1.

 55

1.0

0.9

0.8

L. 0 7
a)

cv
+.)

= 0.6

0.5

0.4

0.3
0.5 1 0 1.5 2.0 2.5
2
GAS RATE kg/sm

Figure 17. H.I.nG vs. gas rate; comparison of arrangements at

1.7G = 2.0.
 56

4000

2000

1000
800

600

400
f r7

200

100

80

60

40

20
0.8 1.0 2.0 4.0 6.0
2
GAS RATE kg/sm

2.
Figure 18. Pressure drop vs. gas rate; liquid rate = 3.51 kg/sm
comparison of arrangements based on packed area.
 57

95

85
LAJ

80

75

70

C)
a= 1

a= 2
65 a= 3
0
countercurrent

60 I , 1
1 I I 1 I I I t 1 I 1 I t r I 1 1 1
1.0 2.0 3.0 4.0 5.0
2
GAS RATE kg/sm
Figure 19. Absorption efficiency vs. gas rate at L/G = 2.0
comparison of arrangements based on packed area.
 58

VIII. CONCLUSION

2
Pressure drop data at liquid rates 1.76, 3.51 and 5.27 kg/sm

were collected for the experimental column using air as the gas phase

and water as the liquid phase. In this range of liquid rates, the

pressure drop for the countercurrent operation is intermediate

between crosscurrent with a = 1 and crosscurrent with a = 2.

Crosscurrent operation with a= 1 has a higher pressure drop per

meter of packed height than the countercurrent operation. It is

only when a is decreased to 2 and 3 that the crosscurrent column

has an advantage in pressure drop operating characteristics. The

pressure drop in a crosscurrent operation also appears to be more

sensitive to liquid flow rates than the pressure drop in conventional

countercurrent packed column.

The overall mass transfer efficiency was calculated as percent

ammonia removed and as height of a transfer unit for the ammonia-air-

water system. Both methods indicated that crosscurrent operation

with a = 1 is more efficient than countercurrent operation for this

system and ranges of flow rates studies. At L/G = 2.0 the a= 2

crosscurrent arrangement was also more efficient than countercurrent;

however, it was slightly less efficient at L/G = 1.0. At L/G= 4.0

the relative efficiencies of the various arrangements depended upon

the flow rates involved. The data show that there is possibly a

range where the different crosscurrent arrangements could be more

 59

efficient than the countercurrent operation. This last observation

certainly requires more mass transfer studies before it can be

qualified further.

Although no formal optimization was done, the multiple stage

crosscurrent packed column with a = 2 appears to be a viable alter-

native to countercurrent towers for the ammonia-air-water system. It

operates with lower pressure drop and better mass transfer efficiency

at most conditions than does the conventional countercurrent packed

column. The a = 1 crosscurrent arrangement is more efficient on

a mass transfer basis but would have higher operating costs than

countercurrent because of its higher gas-phase pressure drop gradient.

The above conclusions are based on a comparison of absorption

equipment that have the same overall dimensions and are operating at

identical flow rates. However, the void space in the experimental

crosscurrent column was arbitrarily set and the above comparison is

valid only for crosscurrent columns with the same open area to packed

area ratio. The effect of this variable and different types of packing

material needs to be investigated before an optimum conbination can be

specified.

The multiple stage crosscurrent packed column operates with a

conbination of contact patterns - crossflow within the packing, sieve

fray type bubbling action on the baffles, and some liquid spray in the

void regions between baffles. At various gas and liquid flow rates

different contact action exists within the column. For example, at

 60

low flow rates the gas likely does not cross the packed section; it

probably penetrates the packing only enough to slip around the baffles

and then goes up the void regions. In this range of flow rates, the

liquid does not leave the packing and crossflow gas-liquid contact

does not exist. As flow rates are increased the gas begins to deflect

the liquid out of the packing and some liquid accummulates on the

baffles. A comprehensive investigation of the crosscurrent column

should examine these different flow regimes. The combination of types

of gas-liquid contact may make the crosscurrent column an efficient

mass transfer device for less soluble systems. The data taken for the

ammonia-air-water system did suggest that different contact regimes do

exist. The absorption efficiency data for the three crosscurrent

arrangements at L/G = 2.0 is again presented in Figure 20. Instead

of drawing a least-squares straight line through the data, the data are

connected with visual best-fit curves. Although the variance of the

data may not statistically allow this type of fit, it does suggest that

even for the ammonia-air-water system two regions may exist - prior to

and after liquid build-up on baffles (liquid hold-up on baffles was

visually observed for each run). Also, more extensive absorption data

should justify the curvature of the lines. Straight lines were used

in Figures 10 through 19 because this study only examined the relative

differences between the different arrangements. The straight lines

can not be extrapolated to flow rates or absorption efficiencies beyond

the range investigated.

 61

The results of this study show that the multiple stage crosscurrent

packed column is comparable to and at some conditions more efficient

than the conventional countercurrent packed column. The crosscurrent

column can process significantly higher flow rates with the same

pressure drop and absorption efficiency and with the same packed volume

as the countercurrent packed column. However, this study only consid-

ered the effect of two variables, flow rates and vertical baffle

spacing, and conclusions can not be generalized without further

research. The favorable results with the ammonia-air-water system,

pall rings, and the open area to packed area ratio of this experimental

column definitely justify further research. The effects of all possible

variables need to be studied before the performance of the multiple

stage crosscurrent packed column can be completely understood and

the design utilized.

 62

1 1 l 1 I I I I I I 1 1 1 1 1

95

90

h.*

Prior to Liquid
Build-up on Baffles

With Liquid Present

On Baffles

0 a = 1
p a = 2 -
65- a = 3

60 1 1 I I 1 1 I I I 1 I I t I 1 i I

0.5 1.0 1.5 2.0 2.5

2
GAS RATE kg/s171

Figure 20. Absorption efficiency vs. gas rate at L/G = 2.0; proposed
regions of gas-liquid contact.
 63

IX. NOMENCLATURE

Symbol Description Units

A absorption factor
m2
A cross sectional area of orifice
2

C orifice coefficient

C ammonia rotameter flow coefficient

r

diameter of orifice mm

D inside diameter of air pipe 83 mm

D diameter of float for ammonia rota- 0.0127 m

f
meter

E i.e., 2.0E-02 is equivalent to 0.02

F calibration coefficient for water mol/s(% of scale)

rotameters
2
G total molar flow rate of gas phase mol/sm

GI total mass flow rate of gas phase .kg/sm

H Henry's law constant for ammonia,

mole fraction basis: y*=Hx H=0.95

H height of an overall gas transfer unit

tOG
K C47-7-
K K constants used in orifice equation --
2

L total molar flow rate of liquid phase mol/sm2

2
L' total mass flow rate of liquid phase kg/sm

m molar flow rate of air, water, and mol/s

air,H 0 NH
2 ' 3

M average molecular weight of air stream kg/kmol

N normality of NaOH and H2SO4 solutions kgequiv/m3

 64

Symbol Description Units

N number of overall gas transfer units

tOG
P barometric pressure Pa
atm
pressure upstream from orifice Pa
P1

P pressure downstream from orifice Pa

2

AP pressure drop across orifice P1-P2 Pa

AP pressure drop per meter of packed Pa/m

col
height

P pressure of ammonia gas Pa

NH3

3
q volumetric flow rate of ammonia m /s
3
R gas constant 8.314E+03 Pam /kmolK

R percentage change in diameter ratio

d
for the ammonia rotameter

St Stoke's number

Tw,Td wet and dry bulb temperatures of

air stream

T1 orifice upstream air temperature

T ammonia gas temperature

NH
3

v linearly-varying multiplier of ideal gas --

law used to calculate ammonia density
3
volume used in titrations m

W mass flow rate through orifice kg/s

W weight of float for ammonia rotameter 2.715E-03 kg

f

Xi mole fraction of ammonia in the inlet

n,out
and exit liquid streams; solute-free
basis

expansion factor through orifice

 65

Symbol Description Units

mole fraction of ammonia in the inlet

Yin,out
and exit gas streams; solute-free basis

z humidity as kg water per kg dry air kg/kg

z total height of packing 1.52 m

a baffle spacing variable

a ratio of orifice diameter to pipe mm/mm

diameter d/D
2
p density kg/m
3
pf density of float for ammonia rotameter 2530 kg/m

ammonia viscosity Pas

 66

X. LITERATURE CITED

1. Bird, R. B., W. E. Stewart and E. N. Lightfoot, Transport Phenomena,

John Wiley and Sons Inc., New York, pp. 495-711, 1960.

2. Skelland, A. H. P., Diffusional Mass Transfer, John Wiley and Sons

Inc., New York, pp. 6-383, 1974.

3. Welty, J. R., C. E. Wicks and R. E. Wilson, Fundamentals of Momen-

tum, Heat and Mass Transfer, John Wiley and Sons Inc., New York,
pp. 442-621, 1969.

4. Perry, R. H. (ed.), Chemical Engineers' Handbook, 5th Ed., McGraw-

Hill Book Co., New York, pp. 3:96, 211, 14:1-16, 18:19-49, 1973.

5. Treybal, R. E., Mass Transfer Operations, McGraw-Hill Book Co.,

New York, pp. 153-175, 220-280, 1968.

6. McCabe, W. L. and J. C. Smith, Unit Operations of Chemical Engineer-

ing, McGraw-Hill Book Co., New York, pp. 639-677, 1967.

7. Foust, A. S., L. A. Wenzel, C. W. Clump, L. Maus and L. B. Anderson,

Principles of Unit Operations, John Wiley and Sons Inc., New York,
pp. 267-288, 1960.

8. Fordyce, H. E., U. S. Patent No. 2,776,121, Official Gazette, p. 97

(Jan. 1, 1957).

9. Hanf, E. B., "A Guide to Scrubber Selection", Envr. Science and

Tech., 4, 110 (Feb. 1970).

10. National Air Pollution Control Administration, "Control Techniques

for Particulate Air Pollutants", U. S. Dept. H. E. W. Bulletin AP-
51, Washington, D. C. (Jan. 1969).

11. Baker, D. R. and H. A. Shryock, "A Comprehensive Approach to the

Analysis of Cooling Tower Performance," J. Heat Trans., Trans. AIME,
83, 339 (August 1961).

12. Wnek, W. J. and R. H. Snow, "Design of Cross-flow Cooling Towers

and Ammonia Stripping Towers", Ind. Eng. Chem., Process Des. Dev.,
11, 343 (July 1972).

13. Thibodeaux, L. J., "Continuous Crosscurrent Mass Transfer in

Towers", Chem. Eng., 76, 165 (June 2, 1969).

14. Gray, M. L., K. Frankenberger and M. Glasgow, "Cascade Crossflow

Packed Column", University of Arkanasa, Fayetteville, AR., April
1974.
 67

15. Oza, S. B., "Fluid Dynamics of a Crossflow Device", M.S. Thesis,

University of Arkansas, Fayetteville, AR., 1974.

16. Thibodeaux, L. J., "Fluid Dynamic Observations on a Packed, Cross-

flow Cascade at High Loadings", University of Arkansas, Fayetteville,
AR., 1977.

17. Golshani, A., "Dynamics Behavior of a Cascade, Cross-Flow, Packed

Column", M.S. Thesis, Oregon State University, Corvallis, OR., 1977.

18. Pittaway, K. R., "Liquid-Phase Mass Transfer Coefficient Measure-

ments in a Crosscurrent Packed Tower", Ph.D. Thesis, University of
Arkansas, Fayetteville, AR., 1976.

19. Thibodeaux, L. J., D. R. Daner, A. Kimura, J. D. Millican and R. J.

Parikh, "Mass Transfer Units in Single and Multiple Stage Packed Bed,
Cross-Flow Devices", Ind. Eng. Chem., Process Des. Dev., 16, 325
(July 1977).

20. Clifford, I. L. and E. Hunter, "The System Ammonia-Water at Tempera-

tures up to 150°C and at Pressures up to Twenty Atmospheres", J.
Phys. Chem., 37, 101 (Jan. 1933).

21. "Fluid Meters: Their Theory and Applications", A.S.M.E. Research

Report on Fluid Meters, 6th Ed., New York, p. 208, 1971.

22. Gilmont, R. and P. W. Maurer, "A Generalized Equation for Rota-

meters With Spherical Floats", Instr. and Control Sys., 34, 2070
(Nov. 1961).

23. Van Wylen, G. J. and R. E. Sonntag, Fundamentals of Classical

Thermodynamics, 2nd Ed., John Wiley and Sons Inc., New York, p. 666,
1973.

24. Sherwood, T. K. and C. E. Reed, Applied Mathematics in Chemical

Engineering, McGraw-Hill Book Co., New York, pp. 344-386, 1939.
APPENDICES
 68

APPENDIX A

Calibration Curves
 69

80 r
I I I

70

MEM

60

M
O
* 50

IMO

40

30

Im

20

WIM

10

0 I I I

0 20 40 0

SCALE READING AT CENTER OF BALL FLOAT

Figure 21. Ammonia rotameter calibration at 283 K and 103.4 kPa.

 70

16

14

12

10

8
SLOPE = 0.1574
mol/s(% of scale)

0
0 20 40 60 80 100

SCALE READING AT TOP OF FLOAT

Figure 22. Small liquid rotameter calibration at 283 K.

 71

0 20 40 60 80 100

SCALE READING AT TOP OF FLOAT

Figure 23. Large liquid rotameter calibration at 283 K.

 72

APPENDIX B

Sample Calculations
 73

APPENDIX B

Sample Calculations

This set of calculations is for Run 30. This run was made with

the column operating in the crosscurrent manner with arrangement #2,

six baffles. The data is first presented in units as read from the

instruments and then converted to S.I. units before proceeding with

the calculations.

Measurement Observed Units S.I. Units

No. of baffles 6

Packing 5/8 in pall rings 16 mm pall rings

Orifice Diameter 2.000 in. 50.80 mm

Barometric Pressure 30.22 in. Hg 102,050 Pa

T 42°C 315.15 K
1

Pl 30.4 in. Hg 102,660 Pa

1.15 in. H2O 286 Pa

T in 62.6°F 290.15 K
d'

T , in 51.5°F 283.98 K
w
Td, out 50.0°F 283.15 K

Tw, out 48.0°F 282.04 K

Small Liquid Rotameter 15.0% of scale

T 50° F 283.15 K
NH3
P 15.00 psia 103,420 Pa
NH3
Ammonia Rotameter 30.0% of scale
 74

Measurement Observed Units S.I. Units

AP 0.35 in. H20/50 in. 68.6 Pa/m

col

N 0.212 gequiv/liter 0.212 kgequiv/m3

H SO
2 4

N 0.200 gequiv/liter 0.200 kgwequiv/m3

NaOH
3
25 ml 25.E-06 m
VH SO
2 4
3
10 ml 10.E-06 m
Vsample
3
a) 4.1 ml 4.1E-06 m
VNaOH, rqd.
3
b) 3.9 ml 3.9E-06 m
3
c) 3.9 ml 3.9E-06 m

P and P were calculated from the manometer readings as follows:

1 NH
3

(A manometer reading)/(13.6 in H20/in Hg)

P1 Patm
P1 = 30.22 + (20.9 - 18.4 in H20)/13.6 in H20/in Hg

P = 30.4 in Hg x 3376.85 Pa/in Hg = 102,660 Pa

1

manometer reading, cm Hg
P = P +
NH3 atm 2.54 cm/in

41.2.54
9 - 41.1
P = 30.22 + in Hg
NH
3

P = 30.53 in Hg x 3376.85 Pa/in Hg = 103,110 Pa

NH
3

Using psychometric charts and wet and dry bulb temperatures:

z = 0.0055 kg H 0/kg air

in 2

z = 0.0065 kg H 0/kg air

out 2
 75

Using equation (4)

28.97 + 18.01 (1.609 x 0.0055)

M = 28.87 kg/kmol
1.00 + (1.609 x 0.0055)

for d = 50.80 mm, K1 = 20.721E-03 and K2 = 0.32871, using equation (6):

286 W(102660)(286)
M = 20.721E-03 [1.00 - 0.32871
air 102660 ( 28.87)(315.15)

1.18 mol/s
mair=

Using equation (7)

Rd = 1.00 + 0.263 (30.0) = 8.89

Using equation (9)

p = 9.3E-06 + 3.0E-08 (1.8 x 283.15 - 499.67)

p = 9.40E-06 Pas

Using equation (10)

v = 0.6222 + 8.0E-05 (1.8 x 283.15 - 499.67)

+ 6.0E-04 (15.0 - 103110/6894.75)

v = 0.6230

So p = (1.292E-03 x 103110)/(283.15 x 0.6230)

NH3

2
PNH =2A171La4m
3
 76

Using equation (13)

3 2
St = 2.829E-09 x 0.755 x 8.89 /9.40E-06

St = 16980.

Using equation (11)

a = 3.08 log (8.89) - 1.25 = 1.67

b = 3.83 - 1.17 log (8.89) = 2.72

c = log (16980) - 0.111 log (8.89) = 4.12

Cr = ( 1)(2.72)2 + 4 (1.67) (4.12) - 2.72)1(2. x 1.67)

Cr = 0.955

Using equation (14)

MNH3' = (1.528E-03)(0.955)(8.89)(1575)(2.0 + 8.89/100)

in

M = 0.0236 malls
NH in
3'

Using equation (15)

M = 0.1574 x 15.0 = 2.36 mol/s

H20

Using equations (16) and (17)

= (4.1 + 3.9 + 3.9)(10-6)/3 = 3.97E-06

VNa0H,average

(0.212)(25.E-06) - (0.200)(3.97E-06)
(18.01)(2.36)
thNH1,out (1000)(10.E-06)
liluid
 77

= 0.0191 malls
mNH , ,out
11

liquid

Using equation (18)

m = 0.0236 - 0.0191 = 0.0045 molls

NH out
3'
gas

Then: Yin = M /M = 0.0236/1.18 = 0.0200

NH in air
3'

x =
out r" NH,out = 0.0191/2.36 = 0.0081
lIquid

Y = M NH1,outilhair 0.0045/1.18 = 0.0038

out
") gas

Using equation (22)

A = 2.36/(0.95 x 1.18) = 2.11

ln[(0.0200/0.0038)(1 - 1/2.11) + 1/2.11]

NtOG 1 - 1/2.11

NtOG

Using equation (19)

H = 1.52/2.24 = 0.679
tOG

The mass flow of the liquid and gas streams per total cross-sectional

area of the column are (use cross-sectional area of packing for counter-

current):
 78

mol 18 kg kmol 1
L = 2.36 x
s kmol x 1000 mol x 2
(.254 x .127) m

L' = 1.32 kg/sm2

mol 28.87 kg kmol 1

G' = 1.18 x 2
s kmol 7000 mol (.254 x .127) m

G' = 1.06 kp/sm2

The absorption efficiency is calculated as:

thNH1,out,liquid
% NH removal x 100
3
111NHvin

0 0191
% NH removal x 100 = 81%
3 0.0236
 79

APPENDIX C

Data Tables
 80

APPENDIX C

Data Tables

Table IV. Pressure drop data.

a L' G' AP a L' G' AP a L' G' AP a L' G' AP

col col col col

1 1.76 0.87 160 2 1.76 0.84 40 3 1.76 1.10 20 * 1.76 0.85 200
1.03 290 1.23 90 1.45 60 1.16 380
1.31 600 1.69 220 2.06 130 1.51 630
1.44 780 2.20 490 2.49 210 1.79 910
1.53 1570 2.68 850 2.97 320 2.05 1210
1.67 2630 2.99 1100 3.50 510 1.98 1410
1.78 3290 3.49 1940 3.75 570
3.76 1820 3.91 1310

3.51 0.90 260 3.51 0.97 80 3.51 1.16 50 3.51 0.88 240
0.98 380 1.69 430 2.02 190 1.17 430
1.03 450 2.19 780 2.67 340 1.43 650
1.14 840 2.61 1170 3.12 570 1.70 940
1.22 1490 2.97 1650 3.62 590 1.81 1110
1.27 2440 1.87 1250
1.39 3350

5.27 0.87 420 5.27 0.97 140 5.27 0.94 40 5.27 0.85 250
0.93 800 1.69 630 1.79 220 1.19 490
0.98 1170 2.21 1050 2.25 320 1.51 830
1.04 2100 2.61 1630 2.69 460 1.69 1100
1.07 3220 2.97 2200 2.96 720 1.72 1310

*
Countercurrent
 81

Table V. Mass transfer data - part I.

mNH out
Number m 3 '
AP M N20 171NN3,in
Run of col L/G air liquid
# Baffles Pa/m mol/mol mol/s mol/s mol/s mol/s

1 12 290 2.75 1.15 3.15 0.0468 0.0380

2 12 430 5.51 1.14 6.30 .0463 .0400
3 12 1530 8.34 1.13 9.45 .0456 .0386
4 12 270 2.77 1.14 3.15 .1465 .0379
5 12 240 1.39 1.13 1.57 .0462 .0280
6 12 1940 8.34 1.13 9.45 .0463 .0438
7 12 840 6.82 1.16 7.88 .0470 .0426
8 12 3470 9.11 1.12 10.24 .0476 .0442
9 12 450 5.47 1.15 6.30 .0222 .0199
10 12 290 2.74 1.15 3.15 .0222 .0188
11 12 100 2.00 0.66 1.31 .0133 .0097
12 12 160 1.99 0.91 1.81 .0176 .0133
13 12 270 1.98 1.19 2.36 .0235 .0184
14 12 410 2.00 1.47 2.95 .0292 .0248
15 12 1060 2.00 1.70 3.40 .0342 .0301
16 12 1900 2.00 1.81 3.62 .0342 .0297
17 12 2170 1.96 1.86 3.64 .0349 .0317
18 12 2390 2.00 1.86 3.72 .0370 .0356
19 12 1430 4.01 1.47 5.88 .0292 .0267
20 12 300 4.00 1.18 4.72 .0231 .0208
21 12 140 4.02 0.90 3.62 .0176 .0151

22 12 70 4.01 0.65 2.60 .0130 .0107

23 12 2170 4.00 1.50 6.02 .0299 .0275
24 12 3390 3.99 1.50 5.99 .0307 .0284

25 12 430 1.00 1.57 1.57 .0310 .0220

26 12 980 1.00 1.94 1.94 .0386 .0289

27 6 190 2.00 1.74 3.46 .0349 .0293

28 6 240 2.00 1.87 3.73 .0372 .0323
29 6 360 2.00 2.08 4.17 .0419 .0370

30 6 70 2.00 1.18 2.36 .0236 .0191

 82

Table V (continued)

thNH,out
Number
AP M thH 0 thNH in
Run of col L/G air 2 3' liquid
# Baffles Pa/m mol/mol mol/s mol/s mol/s mol/s

31 6 570 1.99 2.33 4.65 0.0473 0.0422

32 6 570 1.99 2.34 4.65 .0474 .0430
33 6 920 1.99 2.77 5.51 .0570 ".0523
34 6 200 2.01 1.72 3.46 .0340 .0286
35 6 310 4.01 1.67 6169 .0339 .0320
36 6 890 4.01 2.30 9.21 .0457 .0426
37 6 90 4.00 1.16 4.65 .0228 .0205
38 6 490 3.99 1.89 7.56 .0375 .0351
39 6 360 1.00 2.29 2.28 .0455 .0310
40 6 220 0.99 1.92 1.89 .0379 .0247
41 4 290 2.00 2.81 5.62 .0567 .0500
42 4 30 1.99 1.16 2.31 .0232 .0180
43 4 60 2.00 1.71 3.41 .0339 .0274
44 4 160 2.01 2.31 4.65 .0466 .0404
45 4 270 4.00 2.31 9.25 .0467 .0432
46 4 100 4.01 1.71 6.85 .0343 .0317
47 4 30 4.00 1.18 4.72 .0232 .0209
48 4 60 1.01 1.93 1.94 .0387 .0239
49 4 210 1.00 2.81 2.80 .0560 .0376
50 --* 390 2.00 0.65 1.30 .0133 .0103
51 400 4.00 0.65 2.60 .0133 .0119
52 760 2.01 0.90 1.81 .0179 .0150
53 840 3.99 0.90 3.59 .0179 .0166
54 1290 1.99 1.17 2.33 .0235 .0205
55 -- 1234 1.01 1.17 1.18 .0235 .0170
56 1190 1.01 1.15 1.16 .0236 .0168
57 200 4.04 0.47 1.89 .0095 .0078

*Runs 50-57 were made with the column operating as a conventional

countercurrent packed column.
 83

Table VI. Mass transfer data - part II.

2 NH
Run kg/sm 3
Y X Y N H
in out out L' G' remov. tOG tOG

1 0.0408 0.0121 0.0077 1.76 1.03 81.3 2.05 0.743

2 .0405 .0064 .0056 3.52 1.02 86.3 2.20 .692
3 .0403 .0041 .0061 5.27 1.01 - - bad run - -
4 .0409 .0120 .0076 1.76 1.02 81.5 2.06 .737
5 .0409 .0178 .0161 0.88 1.01 60.5 1.25 1.211
6 .0409 .0046 .0023 5.27 1.01 94.5 3.12 0.487
7 .0407 .0054 .0038 4.40 1.04 90.7 2.60 .585
8 .0423 .0043 .0030 5.71 1.00 93.0 2.84 .535
9 .0193 .0032 .0020 3.52 1.03 89.7 2.54 .599
10 .0193 .0060 .0029 1.76 1.03 84.8 2.37 .642
11 .0202 .0074 .0055 0.73 0.59 72.9 1.67 .910
12 .0193 .0074 .0046 1.01 0.81 76.0 1.88 .809
13 .0197 .0078 .0042 1.32 1.06 78.5 2.06 .738
14 .0198 .0084 .0031 1.65 1.31 84.6 2.56 .594
15 .0201 .0089 .0024 1.90 1.52 88.2 3.02 .504
16 2.02 1.62 - - bad run - -
17 .0188 .0087 .0017 2.03 1.66 90.8 3.54 .430
18 .020T .0096 .0008 2.08 1.66 96.1 4.98 .305
19 .0199 .0046 .0017 3.28 1.31 91.6 2.90 .523
20 .0195 .0044 .0019 2.63 1.05 90.0 2.74 .555
21 .0195 .0042 .0028 2.02 0.80 85.9 2.25 .677
22 .0201 .0041 .0036 1.45 0.58 81.9 1.97 .771

23 .0199 .0046 .0016 3.36 1.34 92.1 2.98 .510

24 .0205 .0047 .0016 3.34 1.34 92.4 3.02 .503
25 .0197 .0140 .0058 0.88 1.40 70.9 2.25 .672
26 .0199 .0150 .0050 1.08 1.73 74.9 2.78 .547
27 .0201 .0085 .0033 1.93 1.56 83.8 2.48 .613

28 .0199 .0087 .0026 2.08 1.67 86.8 2.86 .531

29 .0201 .0089 .0024 2.33 1.86 88.3 3.02 .504

30 .0199 .0081 .0038 1.32 1.06 81.1 2.24 .679
 84

Table VI (continued)

Ru n 2 %NH
Y X kg/s.m 3
# in out ''out N H
L' G' remov. tOG tOG

31 0.0203 0.0091 0.0022 2.59 2.08 89.3 3.19 0.476

32 .0203 .0093 .0019 2.59 2.09 90.6 3.45 .441

33 .0206 .0095 .0017 3.07 2.48 91.7 3.67 .414

34 .0197 .0083 .0031 1.93 1.54 84.2 2.54 .598

35 .0203 .0048 .0011 3.73 1.49 94.4 3.49 .436

36 .0199 .0046 .0014 5.14 2.06 93.2 3.15 .482

37 .0196 .0044 .0020 2.59 1.04 90.0 2.68 .567

38 .0198 .0046 .0013 4.22 1.69 93.4 3.24 .469

39 .0199 .0136 .0063 1.27 2.05 68.2 2.05 .741

40 .0198 .0131 .0069 1.05 1.72 65.2 1.80 .843

41 .0202 .0089 .0024 3.14 2.51 88.1 3.02 .503

42 .0201 .0078 .0046 1.29 1.04 77.3 1.94 .782

43 .0198 .0080 .0038 1.90 1.53 80.8 2.22 .684

44 .0201 .0087 .0027 2.59 2.06 86.8 2.81 .541

45 .0202 .0047 .0015 5.16 2.06 92.6 3.08 .493

46 .0201 .0046 .0015 3.82 1.53 92.3 3.08 .494

47 .0196 .0044 .0019 2.63 1.05 90.2 2.74 .554

48 .0201 .0123 .0077 1.08 1.72 61.8 1.54 .988

49 .0199 .0134 .0066 1.56 2.51 67.1 1.92 .792

50 .0204 .0086 .0046 1.45 1.16 77.7 1.96 .774

51 .0204 .0046 .0021 2.90 1.1.6 89.7 2.67 .570

52 .0198 .0083 .0033 2.02 1.61 83.6 2.45 .621

53 .0199 .0046 .0014 4.01 1.61 92.8 3.16 .482

54 .0201 .0088 .0026 2.60 2.09 87.0 2.89 .527

55 .0201 .0144 .0056 1.32 2.09 72.2 2.41 .631

56 .0206 .0146 .0060 1.29 2.06 71.1 2.27 .669

57 .0202 .0042 .0035 2.11 0.84 82.8 2.01 .757