Pergamon Chemical En~ineerinq S~ience, Vol. 52. No. 24. pp.

4569 4591, 1997

1997 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
PII: ~000~-2509(97)00300-X 0009 2509,'97 $1700 ÷ 0.00

Multicomponent gas absorption with

multiple reactions: modelling and
simulation of NOx absorption
in nitric acid manufacture
M. P. Pradhan, N. J. Suchak, P. R. Walse and J. B. Joshi*
Department of Chemical Technology, University of Bombay, M atunga, M umbai 400 019, India

Abstract---A mathematical model has been developed for the optimum design of plate columns
for the manufacture of nitric acid. The effects of outlet NOx composition, internal geometrical
parameters, temperature and pressure have been included in the model. The present model has
been simulated using the commercial plant data and has been compared with models published
in the literature. Sensitivity of the fixed and operating costs on the optimum values of plate
spacing, liquid submergence, excess oxygen, pressure, nitric acid concentration, etc. have been
presented. ~C 1997 Elsevier Science Ltd
Keywords: Multicomponent gas absorption; NO.~ absorption: nitric acid manufacture

INTRODUCTION Marek (1968), Carleton and Valentin (1968), Hoftizer

Absorption of NOx gas is an important step in the and Kwanten (1972), Makhotkin and Shamsutdinov
manufacture of nitric acid. Absorption of NOx gas is (1976), Holma and Sohlo (1979), Emig et al. (19791,
probably the most complex when compared with Counce and Perona (1979, 1980, 1983), Joshi et al.
other absorption operations. The reasons are: (i) The (1985), Miller (1987), Gutierrez-Canas et al. (1989~,
NO~ gas is a mixture of several components consist- Wiegand et al. (1990) have reported various aspects of
ing of NO, NO2, N204, N203, etc. and the absorp- the process design of packed columns, plate, columns,
tion of NOx gas in water results into two oxyacids, and packed bubble columns used for the manufacture
nitric acid and nitrous acid. (ii) Several reversible and of nitric acid. These attempts will be strengthened if
irreversible reactions occur in both gas and liquid the following important features are included in the
phases. (iii) Simultaneous absorption of many gases process design.
occurs followed by chemical reaction. Also, simulta-
neous desorption of many gases occurs preceded by 1. The rates of absorption of N O 2 , N 2 0 3 and
chemical reaction. For example, the absorption of N204 in nitric acid are different from those in water.
NO2~ N203 and N204 is accompanied by chemical The rates decrease with an increase in the concentra-
reaction whereas the desorption of NO, NO2 and tion of nitric acid.
HNO2 is preceded by chemical reaction. (iv) Hetero- 2. It is known that, for a given set of partial pres-
geneous equilibria prevail between gas- and liquid- sures of NO, NO2 and N 2 0 4, there exists a certain
phase components. Sherwood et al. (1975) and Joshi limiting concentration of nitric acid beyond which no
et al. (1985) have reviewed these aspects of NO~ ab- absorption of N204 and NO2 occurs (Carberry,
sorption. 1958). This heterogeneous equilibrium substantially
For the process design of NO~, absorption towers, it reduces (even 3 4 limes) the rates of absorption of
is necessary to understand the combined effects of NO2, N203 and N2Oa, and the extent of reduction
several equilibria, the rates of mass transfer and chem- increases as the nitric acid concentration approaches
ical reaction. Further, substantial heat effects are the equilibrium value.
associated with NOx absorption. Therefore, temper- 3. A substantial quantity of nitric acid is formed in
ature variations need to be taken into account in the the gas phase, particularly at high temperatures and
process design. There have been outstanding attempts high partial pressure of NO,,. Therefore, HNO3
in this direction. For instance, Koval and Peters formation needs to be included in the mathematical
(1960), Andrew and Hanson (1961), Koukolik and model.
4. A detailed energy balance also needs to be in-
corporated in the model.
5. For every mole of NO~, absorbed, 1/3 mole
* Corresponding author. is desorbed in the form of NO. Thus, the oxidation
4569
4570 M. P. Pradhan et al.
rates are linked with the absorption/desorption 1. In all the previous models, it has been assumed
rates. that the absorption of NO2, N 2 0 3 and N 2 0 , is
6. The rate of absorption of NO2 as such is negli- accompanied by a fast pseudo-first-order reaction.
gible. However, in the presence of N O , NO2 However, in some cases, the mass transfer resistance
forms N 2 0 3 and H N O 2 in the gas phase. The rates of as found to play some role in the overall rate of
N 2 0 3 and H N O 2 absorption are very high. Thus, the absorption. Therefore, a stepwise procedure has been
presence of sufficient N O enhances the rate of absorp- included in the present model wherein appropriate
tion. weightage has been given to the gas side and liquid
7. The m a x i m u m permissible concentration of side mass transfer coefficients and the chemical reac-
HNO3(I) decreases with an increase in the mole frac- tion.
tion of NO. As the absorption continues, the mole 2. Suchak and Joshi (1994) developed the model for
fraction of N O increases because of the absorption of packed columns. Therefore, they had assumed plug
N 2 0 4 , N 2 0 3 and H N O 3 (g). Therefore, the multistage flow for the liquid phase. Since the present model is
operation is needed for getting the desired concentra- for the plate columns, the liquid phase on a plate has
tion of HNO3(1). Each stage consists of an absorption been assumed to be completely backmixed. Further,
section and an oxidation section. F o r instance, in the the effects of free area and orifice diameter on plate
case of sieve plate column, the absorption occurs on have been included in the model.
the plate and the N O oxidation occurs in the empty 3. In the empty space between the plates, oxidation
space between the two plates. The plate spacing de- of N O occurs. The previous models considered the
pends upon the extent of desired oxidation. The extent oxidation under isothermal conditions. However,
of absorption on a plate is also a variable. there is usually no provision for the removal of heat in
8. The effect of temperature is multifold. The rate of the empty space. Therefore, the oxidation has been
N O oxidation decreases whereas the rate of absorp- considered under adiabatic conditions. When the gas
tion increases with an increase in temperature. The enters the liquid phase of the next plate, the gas phase
conversion of NO2 to N 2 0 4 is favourable at lower has been assumed to reach the liquid-phase temper-
temperature. Further, for a given NOx composition in ature.
the gas space, the m a x i m u m permissible HNO3(1) 4. In plate columns, the pressure drop across all the
concentration increases with a decrease in temperature. plates is likely to be in the range of 5 4 0 % of the total
9. Total pressure is a very strong parameter. The operating pressure. Since, the oxidation rate is pro-
rate of N O oxidation is proportional to the cube of portional to the cube of operating pressure, it is im-
pressure and the rates of absorption also increase with portant to include the pressure variation in the model.
an increase in total pressure. 5. Overall optimization: The variables which con-
tribute to the fixed and the operating costs are (a)
Table 1 summarizes the earlier work. T h o u g h the excess oxygen, (b) total pressure, (c) cooling water
mathematical model of Suchak and and Joshi (1994) temperature, (d) product concentration, (e) height of
considers most of these aspects, it was developed basi- weir, (f) size of orifice, (g) percentage opening of sieve
cally for packed columns. Therefore, it was thought plate, (h) tray spacing, (i) product temperature, (j)
desirable to develop a model for plate columns. The efficiency of energy recovery, (k) outgoing NOx con-
following additional points will be considered: centrations, etc.

Table 1. Summary of previous work

Model by A B C D E F G H I J K

Sobotka (1973) x x x x ,/ x x x x x x
Emig et al. (1979) x x x x ,/ ,/ x x x x x
Holma and Sohlo (1979) x x x x ,/ × x x x x x
Counce and Perona (1980) ,/ x x x x x x x x x x
Carta & Pigford (1983) × x ,/ x x x x x x x x
Miller (1987) ,/ x x x x x x x x x x
Guiterrez-Canas (1989) x x x x x ,/ x J x x x
Weiseiler et al. (1990) ,/ x ,/ x x x x x x x x
Weigand et al. (1990) ,/ x x x ,/ ,/ x x x x x
Suchak and Joshi (1994) ,/ ,/ J ,/ x ,/ x x x x x

,/: Considered x: Not considered

where A: dependence of H x / ~ on HNO3 Concentration; B: role of Heterogeneous equilibria in the rate of absorption; C:
formation of HNO3 in the gas phase; D: mass transfer of H20, HNO3 and/or HNO2 from/to gas phase to/from liquid; E:
heat transfer between the cooling system and liquid on the plate; F: gas side resistance to mass transfer; G: identification of the
regime of operation of absorption of various species followed/preceded by chemical reaction; H: backmixed in the liquid
phase of plate columns; I: heat balance during the oxidation of nitric oxide; J: pressure drop in the column; K: overall
optimization.
 Multicomponent gas absorption with multiple reactions 4571
Since, all the above parameters are strongly interac- reaction is second order with respect to NO and first
ting, detailed understanding is needed for the selection order with respect to oxygen.
of optimum design and operating parameters. The The gas phase consists of NO, NO2, N203 and
present study deals with these aspects. N204. In the presence of water vapor (which is gener-
ally the case) oxyacids (HNO2 and HNO3) are also
2. MATHEMATICAL MODEL present in the gas phase. Complex equilibria prevail in
In the commercial nitric acid plant, the mixture of the gas phase which can be described by the following
ammonia with air is passed over catalyst gauze at equations
temperature and pressure in the range of 800-850°C 2NO2(g) ,-- 21.)4{g~ 3)
and 0.3 2 MPa, respectively. The following is a major
reaction: K3
NO(g) + NOz(g) ~ N202(g) (4)
4NH3(g) + 5 0 2 ( g ) ~ 4NO(g) + 6HzO(g). (1)
K4
NO(g) + NO2(g) + H 2 0 ( g ) ~ 2 H N O 2 ( g ) (5)
This is an exothermic oxidation reaction with
a short residence time. Heat liberated is advantage- K5
ously removed in heat recovery section to produce 3NO2(g) + H20(g),-~-2HNO3(g) + NO(g). (6)
superheated stream, preheat air and in certain cases to The reaction rate constant kl and the gas-phase
preheat the tail gas. The gas stream leaving the heat equilibria constants K 2 . . . K 5 were reported by Joshi
recovery section is at a temperature above dew point et al. (1985) and are presented in Table 2(A).
(in a typical case 200'~C) and is further cooled in The total number of moles in the gas per mole of
a cooler/condenser before feeding it to an absorber. inerts is obtained by adding ratios of all gaseous
The condensate is a weak nitric acid solution which is species to one mole of inerts
added up at an appropriate location in the absorption
column. Suchak and Joshi (1994) have described the YT = YNO + YN02 + YN204 3- YN203 3- YHNO~
model for cooler/condenser. The same model has been
+ YHNO2 + YH2o + Y02 + 1 (7)
used in the present work.
where the mole fractions Y are calculated from fol-
2.1. ModeI Jbr the cooler and condenser lowing equilibrium constants:
Mainly two types of coolers are employed: (i) spray
YN204 = (K2 y2o2PT),/(YT) (8}
type, where an aqueous weak acid solution is sprayed
in the gas stream; (ii) bubble column type where the YN203 = (K3 YNOYNozPr)/(YT) (9)
gas is bubbled through a pool of liquid. In both types
YHNo2=[(K4YNoYNo2YH2oPr)/(Yr)]I'2 (10)
of coolers, the weak nitric acid solution is directly
brought in contact with the gas stream. YnNo, = [(KsYr~o2YH2oPr)/(YyoYr)] 12. (11)
Condensers employed are generally large surface
All the components of NOx can be expressed in terms of
area heat exchangers. The hot gas from the heat
recovery stream is passed over the cooler surface (a) total reactive nitrogen (Y*), mole of reactive
where part of water vapour condenses. nitrogen per mole of inerts
In cooler as well as condenser, during cooling, many
chemical reactions and chemico-physical phenomena Y* = YNO + YNO2+ 2Y'~2o, + 2YN2o3 + YHNO~+ YHNO2

occur. They are: (i) oxidation of nitric oxide; (ii) forma- (12)
tion of higher nitrogen oxides; (iii) condensation of (b) while extent of oxidation at any instance is
water and oxy acids; (iv) absorption of nitrogen oxides. expressed in the form of Y*o, moles of divalent nitro-
In the process design of nitric acid plant, the design gen oxide per mole of inerts.
of cooler/condenser is very important for obtaining
the desired concentration of the product acid from the Y*o = YNO + YN2O3 3- 0"5YHNo2 -- 0.5 YHNO3 (13)
absorption section. Therefore, a mathematical model (c) YH2o
* is moles of water in the form of oxyacids and
for the cooler/condenser is presented below. free water vapor per mole of inerts. It is expressed as

2.1.1. Gas-phase reactions and equilibria. The gas Y*~o = YH2o + 0.5YHNo~ + 0.5 Yttyo2 - (14)
entering the condenser consists of NO~ gases, nitro-
gen, oxygen and water vapor. Nitric oxide undergoes 2.1.2. Condensation and liquid-phase reactions. In
irreversible oxidation with oxygen in the gas phase. the cooler/condenser, due to reduction in temper-
The oxidation reaction is expressed as ature, water vapor from the gas phase condenses with
simultaneous absorption of nitrogen oxides thus fur-
2NO(g) + O2(g) k~ 2NO2(g). (2) ther altering the equilibria. Nitrogen oxides when
absorbed in an aqueous acid solution form nitrous
Joshi et al. (1985) have reviewed the published liter- and nitric acid. The following reactions occur in the
ature on NO oxidation. It is believed that NO oxida-
liquid phase.
tion proceeds by dimerization of NO followed by
oxidation by oxygen to form N204. The oxidation 2NO2(1) + H 2 0 ( I ) ~ HNO3(I) + HNO2(I) (15)
4572 M. P. Pradhan et al.
Table 2(A). Gas-phase reactions

Equations Equilibrium and rate constants

loglo kl = 652.1/7"-4.747 (kN/m2) -2 s- I

1Oglo K2 = 2993/T-11.232 (kN/m 2)- 2 s- 1
loglo K3 = 2072/T-9.2397 (kN/m2) - 1
loglo K4 = 2051.17/T-8.7385 (kN/m2) - 1
loglo Ks = 2003.8/T-10.763 (kN/m2) - 1

Table 2(B). Heats of reaction

Reactions Standard Heat of

Reaction (25°C)
kcal x (10- 3)

Reactions in gas phase

2NO(g) + O2(g) ~ 2NO2(g) H1 = - 13.64
per kmol NO oxidized
2NO2(g) ~- N204(g) H3 = - 13.69
per kmol N204 formed
NO(g) + NO2(g)~N203g) H4 = -- 9.55
per kmol N203 formed
3NO2(g) + H 2 0 ( g ) ~ 2HNO3(g) + NO(g) n 6 = - 4.23
per kmol HNO3 formed
NO(g) + NO2(g) + H20(g) ~---2HNO2(g) H7 = - 4.905
per kmol HNO2 formed

Reactions in the liquid phase

2NO2(g) + H20(I) --* HNO3(1) + HNO2(I) Hs = - 12.82
per kmol NO2 absorbed
N204(g) + H20(I) -* HNO3(I) + HNO2(1) H 9 = - 12.04
per kmol N204 absorbed
N2Oa(g) + H20(1) ~ 2HNO2(I) H10 = - 9.55
per kmol N203 absorbed
3HNO2(I) --* HNO3(I) + 2NO(g) + H20(1) Hll = 5.71
per kmol HNO2 absorbed
HNO3(g) ~ HNO3(1) H12 = - 9.36
per kmol HNO3 absorbed
HNO2(g) ~ HNO2(I) H13 = - 9.9
per kmol HNO2 absorbed

N203(1) + H20(1) -~ 2HNO2(1) (16) nitric acid (in the liquid phase) is described by Car-
berry (1958):
N204(1) + H20(1) ~ HNO2(1) + HNO3(1). (17)
K6 = (PNo)/(PN20,~) 3/2. (21)
H N O 2 in the bulk of liquid phase is considered to be
d e c o m p o s i n g to form nitric acid: H o l m a a n d Sohlo (1979) have given the following
correlation for the equilibrium constant:
3 H N O 2 ( g ) ~ HNO3(1) + 2NO(l) + H20(1). (18)
log KH = 30.086 -- 0.0693 T -- (0.197 -- 3.27
H N O 3 a n d H 2 0 are their s a t u r a t i o n concentra-
tions in the gas phase. V a p o r pressure of H 2 0 or
x 1 0 - 4 T ) W + 1.2271oglo [ ( 1 0 0 - W ) / W 2] (22a)
H N O 3 over the a q u e o u s nitric acid solution is a func- a n d for less t h a n 5 % by weight
tion of t e m p e r a t u r e a n d nitric acid concentration.
l o g K n = 31.96 - 0.0693 T + (3.27 x 1 0 - 4 T
Pn20 = f I-T, C o n c ( H N O 3 ) ] (19)
- 0.4193) W (22b)
PHNO3 = f [ T , C o n c ( H N O 3 ) ] . (20) where
K n = (K2) 3/2 K6(101.33) 2 (23)
2.1.3. Condensate concentration and heterogeneous
equilibria. F o r a k n o w n c o m p o s i t i o n of NO~ in the a n d W is the m a x i m u m attainable H N O 3 concentra-
gas p h a s e the m a x i m u m a t t a i n a b l e c o n c e n t r a t i o n of tion in weight percent K6 is given by eq. (21).
 Multicomponent gas absorption with multiple reactions 4573
2.1.4. Material balance over condenser/cooler. Here 2.2. Model for oxidizer
we make the following assumptions: The gases leaving the condenser are in equilibrium
with the condensate concentration. In order to pro-
(1) All reacted nitrogen oxide entering the conden-
duce acid of desired concentration, it is necessary to
ser is in the form of nitric oxide.
further oxidize the nitric oxide. This is achieved by
(2) Gasses leaving the condenser are in equilibrium
providing additional volume between the condenser
with condensate.
and the absorber. The extend of oxidation required
(3) Gases leaving the condenser are saturated with
depends on many factors such as the desired nitric
nitric acid and water vapor.
acid concentration, temperature and pressure. The
Water enters the condenser in the form of water oxidation volume is usually provided at the bottom of
vapor. Partly, it condenses and partly it reacts to form the absorption column. Here essentially an increase in
nitric and nitrous acid in both liquid and gas phases. the mole fraction of tetravalent oxides occurs. Tem-
Equating the incoming and outgoing streams of con- perature in this section is maintained in the range
denser for H 2 0 " similar to absorption temperature of the product acid.
For modeling the oxidation section, we make the
G(Yl~2o.i) = G(Y~o.~) + L~(1 + 0.5X~). (24) following assumptions: (1) gas phase flows in the plug
As stated in the assumption, NOx entering the flow manner; (2) the oxidizer is operating at the steady
condenser is in the form of nitric oxide while due to state; (3) gases follow ideal gas behavior.
oxidation, gas-phase equilibrium, condensation, and
absorption in the condenser, it is present in both gas 2.2.1 Gas-phase reactions and equilibria. The gas-
and liquid phases. The balance for reactive nitrogen phase reactions and equilibria have been described in
gives: the previous section. The set of eqs. (7}-{14) is a math-
ematical formulation for establishing composition of
G(Y~.,) = GtY*.c) + L~(X~). (25)
the gas phase. Suchak et al. (1991) have reduced these
Due to oxidation of nitric oxide to nitrogen dioxide, eight equations to four equations expressed in YNo,
oxygen concentration in the gas phase reduces. The Yr, Yn2o and YNO_,.
material balance for oxygen is given by For a known molar concentration of incrts anti
oxygen, total N O , and NO* to N* ratio, equilibrium
Gl Yo,.i) = G(Yo~.c) + L~(0.75*X~) partial pressures in the gas phase can be estimated.
+ G(Y~.~ - Y*o.c). (26)
2.2.2. Model for the oxidizer section. Mass balance
The second term on right-hand side needs some across a differential height dh at height h from the
explanation. It has been assumed earlier that all the bottom is given as:
NO~ entering the cooler/condenser is in the form of
NO. This NO is converted to HNO3 through several (a) Divalent nitrogen balance:
stages: (a) NO oxidation to NO2 according to eq. (2);
(b) conversion of NO and NO 2 to N204, N203 and dY*o S [kl(PNo)2po,e(;]. (28)
dh G
HNO2 according to eqs (3)-(5); (c) absorption of NO2,
N203 and HNO2 and desorption of NO. The desor- (b) Oxygen balance:
bed NO needs to be oxidized again. Thus, the
stoichiometric oxygen requirement is greater than dY~ _ S [kl (PNo)2po2C¢;] ' (291
that indicated by eq. (2). All the above steps may be dh G
combined to give the following equation: Equations (28) and (29) are nonlinear differential
equations. The overall oxygen and NO* balance
2NO + 1.502 + U 2 0 ~ 2HNO3. (27)
(across the oxidizer) is established on solving these
Thus, it can be seen that, for the formation of 1 tool equations.
of HNO3, 0.75 tool of oxygen is needed.
2.2.3. Heat balance. Oxidation of NO, formation
2.1.5. Method of solution. For the estimation of of N203, N204 and FINO3 in the gas phase are
gas-phase composition and quantity of condensate (at exothermic. Because of these reactions significant heat
specified temperature and concentration of conden- changes occur. The values of reaction rate constant
sate), we need to solve eqs (7)~(14) and (19)~26). Using and equilibrium constants depend upon temperature.
substitutions in eqs (7), (19), (24), (25) and (26) a system Heat liberated because of various reactions in the gas
of five nonlinear algebraic equations having five un- phase has been included in Table 2(B).
knowns namely Y'r, YNO2, Yo:, L~ and YH2O was If Qr is the total heat change per unit time in the
formulated. Under most practical conditions these differential element then the temperature change un-
five variables possess large positive value. The New- der adiabatic conditions is given by
ton Raphson method in conjunction with the
Te = (Qr + m°C~i Ti)/(m°C~,) (30)
Gauss Jordan method was employed. The objective
behind reducing these equations was to achieve single where Te and Ti are the temperature at the exist and
finite solution with minimum numerical iterations. the inlet of differential element. (a) Contribution to the
4574 M. P. Pradhan et al.
heat changes from the gas phase are due to the follow- The rate of water evaporation in the differential ele-
ing steps: ment was obtained by material balance.

(i) The rate of formation of N203 from N O and ( H E O ) v = G(Yn2o.e - - Y n ~ o , i q- ½ YHNO3,e

NO2 in the differential element is given by
-- ½ YHNOs,i + ½ YHNO2,e -- ½ YHNO2,,)
(N203) f = (N203) e - (N203) i + (N203) a (31)
+ (½RanNo3,g + ½RaHNO:,t)S dh (45B)
~--- G(YN203,e - YN203,i) (32)
where (N2Oa):, (N203)e, (N203)i and (N203)a are Q6 = (H20)vLw. (45C)
the kmoles per second N2Oa formed, leaving, entering
The total heat change in the differential volume is
and absorbed, respectively.
summation of all the heat changes:
The rate of heat generation due to N203 formation
is Qr = Q1 + Q2 + Qa + Q4 + Q5 + Q6. (46)
Q1 - ( ( N 2 0 3 ) f ) A H 4 . (33)
(ii) The rate of formation of N204 from NO2 in the 2.2.4. Method of solution. Estimation of gas-phase
differential element is given by composition: Solution procedure described by Suchak
et al. (1991) has been used.
(N2Oa)y = (NEO4)e - (N204)I (34)
Solution of ordinary differential equations: The solu-
= G(YN~o4, e -- YN2O4,1)- (35) tion of eqs (28) and (29) is a boundary value problem.
The flow rate and composition of the gas phase enter-
The rate of heat generation due to N204 formation is ing the oxidizer are known. The equations were solved
Q2 = r(N204):] AH3. (36) simultaneously by the fourth-order Runge-Kutta
method. Integration was carried out until the exit of
(iii) The rate of formation of HNO3 in the differen- the oxidizer.
tial element is given by
Solution of heat balance equations: From the above
(HNO3): = (HNOa)e - (HNO3)I (37) two steps the concentrations of all species are known
at the inlet and exit of the differential height. Assum-
= G(YHNo3,e - YHNO3,i)- (38) ing all the heat is utilized as sensible heat, temperature
The rate of heat generation due to HNO3 formation is of the gas at the exit of the differential height is
estimated. For this purpose eqs. (30)-(46) were used.
Q3 = (HNO3)fAH6. (39) The effect of step size for integration was examined
and it was selected in such a way that the oxidizer
(iv) The rate of formation of HNO2 in the differen-
volume is independent of the step size.
tial element is given by

(HNO2)f = ( H N O 2 ) e -- ( H N O 2 ) i (40) 2.3. Model for the absorption column

The absorption column employed for the manufac-
= G(YHNO2,e -- YHNO2,i)' (41)
turing nitric acid is a sieve plate column. The cool gas
The rate of heat liberated due to HNO2 formation stream (NOx) is introduced at the bottom of the
is column in the first empty section, called as pre-
Q4 = ( H N O 2 ) f A H 7 • (42) oxidizer. NO in the gas phase irreversibly oxidizes to
form NO2 affecting the equilibrium concentrations of
(v) The rate of heat generation due to the oxidation all the other constituents of NOx. Two plates are
of NO is estimated from the knowledge of the rate of separated by an empty section. The heat transfer coils
N O oxidation. It may be noted that NO is consumed are placed in the pool of liquid over the plate. A ma-
in the N 2 0 a and HNO2 formation and liberated terial balance is given by the following equations:
during HNO3 formation.
The mass balance for N O across the differential
2.3.1. Overall material balance. Over the entire col-
element is given by
umn
(NO)o = (NO)~ -- (NO)~ - (N2Oa)f (1) Reacted nitrogen balance

-- 1/2(HNO2): + 1/2(HNO3): (43) G(Y*., - Y*.I) = LpXp (47)

= G(YNo, i -- YNO, e - - YN203, e "4- YN203,1 where Y * f is mole ratio of reacted nitrogen to inerts
- - 1/2 YHNO2,e "F 1/2 YHNO2,1 at the column outlet.
(2) Water and water vapor balance
+ ½ YHNO~,i -- ½ YHNO3,i). (44)
G(Y*2o,, - Y~2o.f) = L,(1 + Xp/2) - Fw. (48)
The rate of heat generation due to NO oxidation is
given by YH20.$ is a molar ratio of saturated water vapor to
Q5 = (NO)o(AH)I. (45A) inert gas at the outlet.
 Multicomponent gas absorption with multiple reactions 4575
(3) Oxygen balance ature and nitric acid concentration
, i (60)
G(Y*~.i - Yon.r) = 0 . 7 5 L p X p + 0 . 5 G ( Y * f - Y*o.y). Pmo = [T, Conc(HNO3)]
i
(49) PHNO~ = [T, Conc(HNO3)]. (61)

2.3.4. M a s s transfer with chemical reaction in the

2.3.2. Rates of gas-phase mass transfer. The vol-
liquid film. The concentration profiles for different
umetric rates of gas-phase mass transfer are given by
species in the liquid film are given in Fig. 1. It has been
the following equations:
illustratd by Joshi et al. (1985) that the absorption of
RaNo,o = (k~a)No(P~o -- P~o) (50) NO2, N203 and N204 is accompanied by chemical
reaction within the liquid film. However, absorption
Rayo. y = (kGa)No2(P~02 -- P~O2) (51) of NO~ is also limited by nitric acid concentration in
the liquid-phase bulk and partial pressure of NO at
RaN.o4.g = (kc.a)N~o,(P~2o, -- P ~ o , ) (52) the interface.
The limiting partial pressures of NO2, N 2 0 3 and
RaN~o~.o = (kc.a)N~o~(P~o~ -- P~o3) (53)
HNO2 based on heterogeneous equilibria would be
RaHNO.~,O = (kGa)HNo3(PHNo~ - - PHNO3) (54) b ( h K 2)] 1'2
PNO2 = [ PN20~)/( (62)
RanNo2.y = (kGa)HNo2(PhNo~ -- P~NO~) (55) b b b
PN~O~ = K3PNoPNo~ (63)
Ramo.o = (kGa)H~o(P~:o -- P~i~o). (56) pbNo 2 = (K4PNoPNo:PH20)
b b b 1,2 . (64)
At this stage the following points may be noted. NO absorption takes place when PN~O4~-- pb~o.. The
net driving force is thus the difference between the
I. In the gas film, the partial pressure of NO cha- interfacial and limiting partial pressures of NO,,.,
nges with respect to distance from the bulk of inter- N 2 0 4 , N 2 0 3 and HNO2.
face. It is likely that the rate of NO oxidation in the The absorption of NO2, N203 and N204 is accom-
film is different from the rate in the bulk. Further, in panied by chemical reaction. The pertinent theory has
the gas film the transfer of NO is accompanied by been given by Doraiswamy and Sharma (1984). The
oxidation. Therefore, there is possible enhancement in overall rate of absorption is given by
the rate of mass transfer.
The volume of the gas film can be estimated on the (Ra), -- cDt(kLa)z(p[ -- p~)n~ (65)
basis of D(;, kG, H and a. It was found that the film where l is N 0 2 , N 2 0 3 , N 2 0 4 , etc q~ is the enhance-
volume constitutes about 0.005-0.01 fraction of the ment factor. We define the following parameters:
total volume. Therefore, the extent of NO oxidation in
the film was considered negligible and there was no ( ' ~ ) N O ~ = [~(kD)No~ HNo:(P~o2 -- pbo2)]':2/(kL)NO:
appreciable enhancement in the mass transfer of NO
due to chemical reaction. (66A)
2. In the gas film NO, N O 2 , N 2 0 4 , N 2 0 3 , H N O 2 ,
(x/'M)N2o3 = [(kD)N2O3] 1/2/(kL)N:O~ (66B)
HNO3 and H 2 0 were assumed to be in equilibrium at
all points. The rates of gas-phase mass transfer are (v/-M)N2O~ = (kD)N2O4)/(kL)N2O~ (66C)
likely to change. However, for simplicity the gas-
phase mass transfer rates have been written in the @, = ([ Bo](Dn2o/ D~)°5)/(Zz(p~) - p~)Ht) (66D)
form of eqs (50)-(56). Similar assumptions have been where [Bo] is the water concentration. Z is the
made by Andrew and Hanson (1961), Careton and stoichiometric coefficient in the chemical reaction.
Valentine (1968), Emig et al. (1979) and Holma et al. From eqs (15)-(17), it can be seen that the value of Z is
(1979). 0.5, 1 and 1 for NO2, N203 and N204, respectively.
The overall rates of absorption is given by equation
2.3.3. lnter}hce equilibria. At the interface NO, (65), where the enhancement factor, (I) for NO2, N203
N02, N 2 0 4 , N 2 0 3 and HNO2 are assumed to be and N204 was estimated according to the following
always in equilibrium with each other. Thus at the procedure:
interface, gas equilibrium are given by the following
equations: when ~ < 3 and v/M<<q),
K 2 = P Ni z o J ( P Ni0 2 ) 2 (57) (1) = .~/M/(tanh(x/-M ) (67a)
K3 = PNzO~/(PNo)(PNo:)
i i i (58) when . ~ > 3 and ,.//-M<<~,
K 4 = (PHNO~)
i 2 /(PNo)(PNo~)(PH~o).
i i i (59) el) = .,/ M (67b)
HNO3 and H 2 0 are at their saturation concentra- when~>3 and v/M~(I).
tions. The vapor pressure of H 2 0 or HNO3 over
aqueous nitric acid solution is a function of temper- (I) = ~ {((I). -- (l))/(q), -- 1) }1,'2 (67c)
4576 M.P. Pradhan et al.

Tai gases
)28
Tail 9oses

,~, Process

. . _ wQter

Condensate~
.-. 16

Air 25 ._.

13 1/,~ 2 6 =
i_2_
- 7 ~I, - Secondary (air -
24

Boiler feed water Nitric (acid

product

Fig. 1. Flow sheet of a typical high monopressure nitric acid process (adapted from the Chemico process).
(1) Air filter-first stage (2) Air compressor-first stage (3) Compressor intercooler (4) Air compressor-second
stage (5) Air filter-second stage (6) Ammonia vaporizer (7) Ammonia preheater (8) Air-ammonia mixing
chamber (9) Ammonia oxidation reactor (10) Tail gas preheater (11) Steam superheater (12) Catalyst-filter
(13) Tail gas preheater (14) Waste heat boiler (15) Cooler/condenser (16) Absorption column (17) Purge (18)
Stabilizer (19) Heat exchanger (20) Gas turbine (21) Low pressure steam generator (22) Steam turbine (23)
Heat exchanger (24) Air bleed stream (25) Acid condensate (26) Uncondensed gas (27) Bleaching section (28)
Stack.

when x//M >>(I)a 3. Heterogeneous equilibrium parameter K6 limit-

ing partial pressures
@ = @~. (67d)
b b b
PN2O,, PNO2, PN203, P~NO2 •
In the range of partial pressure, temperature and
kL covered in this work, eqs (67a)-(67c) were found to 4. Rates of mass transfer with chemical reaction in
hold. the liquid film:
The rate of absorption of HNO2 is given by Raio~.t, RaN2o,.t, RaN2o3.t, RaHNO~3.
(Ra)nio2 = (kL)HNO2(PHN02
i b
-- PHNo2)HHNo2. (68) 5. Rates of desorption of NO from liquid to gas-
phase RaNo 3. Two more equations are needed to
Nitrous acid in the bulk decompose to form HNO3 make the model consistent. These are provided by the
and N O in the liquid phase. Since the latter has material balance of divalent and tetravalent nitrogen
limited solubility, it desorbs. The rate of desorption of oxides across the interface.
N O is given as The NO2 balance at the interface gives

RaNod = ~3RaN20~d + 23RaN204d + 13RAN02.1 (RaNo2,g -- RaNO:d) = 2 ( R a N e o , , t - RaN2o,,o)

+ 23RaHNO2d (69) + (RaN2O3,I -- RaN2o~,g)

+ ~RaHNO2.t -- RaHno2,g ) (70)
2.3.5. Overall volumetric rates of absorption. Over- The N O balance at the interface gives
all rates of absorption are computed by solving eqs.
(21)-(23), (50)-(61) and (62)-(69). There are 23 equa- (RaNo.t -- RaNo,o) = (RaN2o3,t -- RaN2o3,g)
tions, whereas the unknowns are:
+ ½(RanNo2,t -- RanNo2,o). (71)
1. Volumetric rates of gas-phase mass transfer: Equations (60), (61) (22a) (22b) and (23) were solved
independently. Equations (21), (22a), (22b) (23),
RaNo,o RaNo2,o, RaN2oa,g, RaN2o4,o, RaHNO2,g, RaHNo3,o (50)-(59) and (62)-(71) are coupled algebraic equa-
and Ran20.o. tions in order to attain convergence under varied
2. Partial pressure of NO, NO2, N203, N204, parameter. It was found to reduce them in two equa-
HNO2, HNO3 and H 2 0 at interface: P~o, P/NO~,P~203, tions of the form P~o, and P/NO~. The procedure is
P ~ 2 o , , PnNo~,
i ~
Pm~o2, Ph2o. given in detail in Suchak et al. (1991).
 Multicomponent gas absorption with multiple reactions 4577
2.3.6. Model for empty section~space between For the stage where the condensate (from cooler/
plates. Model for empty section or the space between condenser) is introduced Ls = Lc and X s = Xc, other-
two consecutive plates is identical to that of a pre- wise Ls = O.
oxidizer described by Suchak and Joshi (1994). Mass 2. Water and water vapor balance
balance across a differential height dh at height h from
the bottom is given as G(Y*~o.N- 1 - Yn2o.u)
* = LN(1 + XN/2)

(a) Divalent nitrogen balance -- Lu+l(1 + Xu+l/2) + Ls(l + Xs/2}.

dY*o/dh (S/G)(kl P N2 o P o 2 e G ) (77)

= -- (72)
(b) Oxygen balance 2.4. Estimation of design parameters
d g ~ / d h = - (S/G)(kl p2oPo2eG). (73) For mass transfer coefficients, holdup, and gas-
liquid interracial area on a sieve plate, correlations
The adiabatic nature in the space between the suggested by Zuiderweg (1982) were used. They are:
plates is incorporated in the rate constant kl as Gas side mass transfer coefficient:
k 1 = 10 (652'1/T-4'747) (74) k~ = 0.13/pG - 0.065/p~. (78)
where the temperature (T) changes in each differential The values of ka were estimated for all gas-phase
height dh. Equations (72), (73) are non-linear differen- components such as NO, NO2, N 2 0 , , N203, HNO3,
tial equations. The heat of reaction for N O oxidation HNO2, H 2 0 , etc. For this the pertinent values of
has been given by Suchak and Joshi (1994). The rise in density were used. These values of kG were used in eqs
temperature was estimated on the basis of heat balance. (50)-(56).
Liquid side mass transfer coefficient:
2.3.7 Model for absorption stage on the plate. Ma-
kL = 2.6 x 10-5//1 °'25. (79)
terial balance over differential height
The plate column has been modelled for simulation Height of dispersion and the fractional liquid hold-up
studies. Following assumptions were made: hi = 0.6 H°wSp°ZS (Fe/b) °'25 (80)
(i) Gas-phase flows in a plug-flow manner and 25mm < H,. < 100mm
liquid phase in a backmixed manner.
(ii) Liquid hold-up is uniform throughout the 1 / V \o.s
----1 =40t~(pG/PL)O'5 ) (81)
column. ~L
(iii) Gases follow ideal gas behavior.
where
(iv) The changes in the flow rate through the froth
and through the gas space between plates are con- Fe = L ' / V ' (pG/pD °'5 (82)
sidered to be negligible while in that particular phase.
(v) Column operates at steady state. and VG is the gas velocity on the basis of bubbling
area.
Component balance in gas phase: Since gas phase Gas-liquid effective interfacial area
flow is assumed to be in plug-flow manner, the differ-
ential equations for gas phase are same as given by a = 6p~/dB (83)
Suchak et al. (1991). where
dB
Component balance in liquid phase: = l. . lt ~l VrG
,0.4/0
ao" 2 g 0 "2 . (84)

Balance for nitric acid. It is assumd that liquid Absorption section model of Suchak and Joshi
flows in backmixed behavior. The concentration of (1994) was used with design parameters of sieve plate
nitric acid throughout the plate is same and the mass obtained from the equations given above. Total pres-
balance is given by the following equation: sure drop across the plate is estimated as follows.
Dry plate pressure drop (Ray and Johnston, 1989):
G(Y*,N I -- Y ~ , N ) = L N X H N o 3 , N -- L N + I X H N o 3 , N + I
APdry = 43.432 V2(p~/pL)(I/Co) z (85)
(75)
where, Vh is velocity of gas through the holes and Co is
where XHNO3 is moles of nitric acid per unit moles of the orifice coefficient.
water. Pressure drop due to liquid head:

Overll material balance for an absorption stage APLn = H,,.PL9. (86)

Following equations establish material balance Total pressure drop across the plate is given by,
over an absorption stage. A P t = APdry + APLH. (87)
1. Reacted nitrogen balance
2.5. Method of solution
G(Y*, N- 1 -- Y*,N) = LNXN -- LN+ 1XN+ 1 - - L s X s . Computations of absorption column begins
(76) with bottom most absorption stage. The gas-phase
4578 M. P. Pradhan et al.
composition of the stream entering this absorption through the bleaching section (27) is introduced at the
stage is known from the oxidizer calculations. Con- bottom of the absorption column, bringing more
centration and flow rate of product acid withdrawn oxygen for NO oxidation and stripped NOx gases.
are also considered to be known. From the detailed The botom part of the absorber acts as NO oxidizer.
model of an absorption stage, concentrations of water The tail gases from the absorber are heated in two
vapor and reacted nitrogen in the exit stream is ob- stages (13) and (10) and expanded in the turbine (20)
tained. Concentration and the flow rate of liquid and finally leaves through the stack (28).
overflow from the subsequent absorption stage are For the overall reaction leading to the formation of
the two remaining unknowns. As mentioned earlier, 60% strength acid:
the liquid phase on a plate was assumed to be com-
pletely backmixed. The gas phase was assumed to NH3(f) + 202(g) + 1.33H20(f)
move in a plug-flow manner on the plate as well as in
-~ HNOa(f) + 2.33HzO(f)
the empty section. For this purpose eqs (76 and 77)
were solved by substituting eq. (76) in eq. (77). (60% HNOa solution). (88)
These calculations were repeated for all the absorp-
tion stages until the condition (Y~ < Y*,F) was satis- the theoretical maximum available energy from the
fied. Heat balance was also established for each above reaction is 315 kJ/mol HNO3 or 5000kJ/ton
differential volume as described by Suchak and Joshi HNO3. However, much lower energy is available in
(1994). The differential height for integration was reality. The principal reasons are (i) all the reactions
0.001 m. are carried out at temperature and pressures higher
than the standard conditions (ii) in the heat recovery
system, there is a limit to the recovery temperature
3. OPTIMUM DESIGN STRATEGY either for steam superheating or flue gas preheating
All commercial processes used for weak (50-70%) (iii) the energy required for compression is higher
nitric acid production are similar except the operating than the energy recovered by expansion. This is be-
pressure. Pressure is the most important parameter cause the moles of tail gas are lower than the moles of
both for ammonia oxidation and absorption of NOx feed gas.
gases. Atmospheric, medium (0.25-0.5MPa) and In view of the above reasons, the maximum practic-
high-pressure (0.7-1.2 MPa) processes have been de- ally available energy is about 40-50% of the theoret-
veloped for both the process steps. Monopressure ical value. This available energy is used for a variety of
systems, using the same pressure level in both the purposes. The foremost energy consumption occurs in
steps and dual pressure systems with a compression the compression system. The energy efficiencies of the
stage between ammonia oxidation and NOx absorp- compressor and the expansion turbine are low. En-
tion are equally widely used. Out of the possible ergy is also needed for overcoming the pressure drop
combinations the following three combinations are in the overall system. Further, part of the N O oxida-
commonly used (Honti, 1985): (i) medium monopres- tion, dimerisation and reactive absorption occur in
sure (ii) high monopressure and (iii) medium-high the absorber and this energy can not be recovered
dual pressure process. advantageously. Due to these reasons, very small
Figure 1 shows a typical high monopressure pro- amount of energy is recovered in the form of export
cess (Honti, 1985). Liquid ammonia is evaporated (6) steam. Honti (1985) has systematically described the
at the required pressure and preheated (7) to the various aspects of the energy balance in the nitric acid
desired temperature and mixed with the compressed plant. He has clearly brought out suggestions for
air (8). Air is filtered (1) and compressed in a two-stage future work by which the energy efficiency of the
compressor (2) and (4). The compression heat after the overall nitric acid plant can further be improved.
first stage (3) is used for preheating the boiler feed In the present paper, the optimum design of only
water and/or is delivered to absorption-type refriger- one part of the nitric acid plant will be considered, i.e.
ation system or for preheating ammonia. The com- absorption system.
pressed air after the second stage is filtered (5) and The optimum strategy to select the optimum para-
mixed with the ammonia vapor in (8). The a i ~ a m m o - meters was developed on the basis of fixed and operat-
nia mixture is passed through a converter (9). The heat ing costs. Fixed cost was considered for the
released in the converter is used for various purposes. absorption column which involves the cost of shell
The tail gas is preheated (10) and steam is superheated and internals such as plates, rings, cooling coils, etc.
(11). Steam is generated in the waste heat boiler (14), Operating cost includes the costs of compression,
superheated in (11) and expands in the steam turbine pumping and chilling. Compression cost includes the
(22). The product gas leaving the waste heat boiler (14) initial compression of gases up to the reaction pres-
enters the cooler-condenser (15). The acid condensate sure for ammonia oxidation. The re-compression of
(25) (40-50% wt) condenses here which is introduced gases is needed if the absorption column operates at
to the column at an appropriate stage (discussed a higher pressure than the ammonia oxidation. The
later). The gas passes to the absorption column at the pumping cost includes (i) cost for the pumping of the
bottom (26). The second part of the air (24), coming process water and (ii) cost for the circulation of the
from the second stage of the compressor, after flowing cooling water. The pumping cost for the cooling water
 Multicomponent gas absorption with multiple reactions 4579
circulation was estimated fom the pressure drop in the where, AC is heat transfer area of cooling coil, Fo is
cooling coils. design type factor, Fe is a pressure factor and Fu is
In order to get the optimized parameters, it was materials factor
thought desirable to study the effect of various para-
meters on total cost. The total cost per day was Fixed cost = Cshell + Cint . . . . is + Ccoi~. (97)
estimated using the fixed and operating costs. The
fixed cost consists of depreciation, interest, insurance, 3.2. Operatin9 cost
maintenance and administration costs. The additional Operating cost involves cost of compression.
expenses depend on the geographical location of the (a) Gas compression cost
plant and federal rules prevailing in a particular coun- Ccomp = Power × Cp.... /(0.5 x 0.65) (98)
try. These expenses on fixed capital can be in the
where
range of 2 5 ~ 5 % of the fixed cost. Hence, it was
Power = P1Q ln(P2/P1), kW (99)
thought desirable to study the effects on total cost at
25, 35 and 45% of the fixed capital. In order to where 0.5 is the compression efficiency and 0.65 is the
appreciate the sensitivity of fixed cost, it was thought power factor.
desirable to consider a case where the yearly expenses (b) Cost of chillin9
were considered to be 100% of the fixed capital invest- Cost of chilling was considered only when refriger-
ment. ation through ammonia vaporization is not sufficient.
The cost were estimated in U.S.S. The calculations The cost was calculated on the basis of power require-
were performed as follows. ment.
(c) Pumpin9 cost of coolin9 water and process water
3.1. Fixed cost Power required for pumping process water:
(1) Fixed cost includes costs of the shell, internals
and cooling coil. Shell cost is estimated from weight of PP = HpgQc/(3.67 x 105) (100)
the shell and correlations were used for the estimation
where H is the height of column
of cost of the internals. Cost of the cooling coil was
estimated from length of the coil. Cpumpl = PP x Cp.... /(eft). (101)
(a) Cost of shell (U.S.S.):
(d) Power required for pumping cooling water:
Cshen = Wt(Css + F~)Cy (89)
This cost was calculated against the pressure drop
where C,s is the cost of SS316, $/kg, Fss the cost of in cooling coils. The pressure drop was estimated as
fabrication, $/kg = 0.6 Cs~, and CI the contingency follows:
factor.
APe = 2 f Lc V~ pw/Ooipe (102)
Weight of the shell:
(for turbulent flow through the pipe)
Wt = rtOtRho~(H + D) (90)
where, friction factor:
Height of the shell:
f = 0.046Re o.2 (103)
H = N(H~ + Hw)l.25 (91)
where H~ is the height of empty section and Hw is the Cpump2 = APcQcCv.... /(eft) (104)
weir height. where eft = efficiency of pump.
Thickness of the shell; The operating cost was found to be due to compres-
t = P/220642.35 + to. (92) sion. The extent of energy recovery depends upon the
strategy of the energy recovery (Honti, 1985). For the
(b) Cost of internals (U.S.$): exercise of the optimum design, the energy recovery
Cint . . . . ls = (Ctray + C r i n g ) 1 . 2 was considered at different levels in the range of
60-95%.
where, and
C,~ay = 658.1 - 255.9D + 341D 2 (93) 4. RESULTS AND DISCUSSION
4.1. Simulation of a commercial plate column
C,~ng = 70.4 + 175.17D + 26.81D 2. (94)
It was thought desirable to simulate a commercial
In some countries, the cost of fabrication is high for sieve plate column for validating the present model.
trays. Therefore, as an alternative the following equa- The plant data have been reported in the published
tion was used for estimation of tray cost: literature under three different conditions of opera-
tion. These operating parameters have been sum-
Ctray = 1316.2 - 511.8D + 682D 2. (95)
marized in Table 3. The design details of the plate
(c) Cost of cooling coils (U.S.$) column have been given in Table 4.
Gutierrez-Canas et al. (1989) have also developed
C¢oil = exp(8.202 + 0.015061n(AC)
mathematical models. In one model, mass transfer
+ 0.06811 (ln(AC))z ) F o F e F u (96) terms were not considered (equilibrium model) and in
4580 M. P. Pradhan et al.
the second model mass tranasfer terms were con- data. Thus, the model developed in this work success-
sidered (transport model). The predictions of these fully predicts the operation of commercial plate col-
two models together with the plant data have been umn. Therefore, this model was further used for the
given in Fig. 2. The transport model was shown to be elucidation of parametric sensitivity and the optimum
superior to the equilibrium model. It can be seen from design.
Fig. 2 that the present model is even superior to the
transport model and favourably agrees with the plant 4.2. Effect of various parameters
In order to get an overall view regarding the num-
ber of stages needed with respect to the extent of
absorption, the parameters such as pressure, temper-
Table 3. Operating parameters of commercial plate column ature, and heights of submergence and empty sections
producing 330 tons/day of nitric acid were varied over a wide range. The capacity of ab-
sorber was 750 tons/day (100% basis) and 30% excess
Production
oxygen was used. Values of other parameters have
HNO3 as 100% (t/h) 6.25
Concentration (%) 56.0 been listed in Table 5.
Cooler-condenser The effect of total pressure, height of submergence
Condensate Temperature (°C) 34.0 and the height of empty section on the number of
Conensate Concen. (%) 36.5 plates was studied with respect to the outgoing NOx
(HNO3) concentration. The effect of these parameters was
Input gases studied up to 20 ppm level. Figures 3(A)-(C) show the
Partial presure (kPa) effect of height of empty section at weir heights of
NO 5.57 0.025, 0.05 and 0.1 m, respectively and at 1.3MPa
NO2 16.94
pressure. It can be seen that, at higher empty section
N204 20.10
heights less number of plates are required. Though,
02 16.88
N 2 243.76 the number of plates decreases, it does not necessarily
H20 4.84
Total pressure (MPa) 0.29
Output gases 60-
-A~. a COMMERCIALPLANTDATA(DETAILSIN
Total NOx (ppm) 505 "~o ~~z~ TABLES3AND41
Oxygen (%) 0.4 50- o F.QULIBRIUMMODELBYGUI'iERREZ-CANAS
+ TRANSPORTMODELBYGUTIERREZ-CANAS
Z " (1989)
~, 4o- - PRESENT MODEL
z

Table 4. Design parameters of a commercial plate column o.~

o
producing 330 tons/day of nitrc acid
o\
Tower 1 Tower 2 z 10-
o

8
Column diameter 4.8 m 4.8 m 0 . J , i . i , i , i ,

Number of plates 41 22 10 20 30 t,0 50 60 70

Plate spacing 1m 2m PLATE NUMBER FROM BOTTOM

Liquid submergence 0.07 m 0.07 m

Percent free area 4% 4% Fig. 2. Comparison between model preditions and the per-
Hole diameter 3 mm 3 mm formance of a commercial NOx absorption column (For
operating and design parameters, refer to Tables 3 and 4).

Table 5. Design parameters of a commercial plate column producing 750 tons/day nitric acid

1. Column diameter = 5.5 m

2. Percent free area = 3.87%
3. Hole diameter = 4mm
Operating parameters of a commercial plate column producing 750 tons/day nitric acid
(* Capacity is on the basis of 100% concentration)
1. Total gas flow rate before cooler-condenser = 1.69 kmol/s
2. Total NO~ flow rate before cooler-condenser = 0.1352kmol/s
3. Oxygen flow rate before cooler-condenser = 0.1356 kmol/s
4. Condensate temperature = 48°C
5. Condensate concentration = 42% (w/w)
6. Oxidizer temperature = 38°C
7. Inlet HNO3 concentration = 0%
8. Outlet HNO3 concentration = 58% (w/w)
9. Outlet NO~ = 20 ppm
 Multicomponent gas absorption with multiple reactions 4581
300 Table 6. Effect of pressure on number of stages temper-
\ PRESSURE =1.3 MPA
0./.~ WEIR HEIGHT=0.025M ature = 30°C. Outlet concentration of NOx = 700ppm
\ \ TEMPERATURE =30°(: other parameters have been listed in Table 5
).7 ~ ~ PARAMETER
=EMPTY HEIGHT(M}
~ 10(] Height of Number of stages
o..
empty Weir
section height 0.4 0.71 1.31 2.02
~
mA
IE 1 0 , ~
0.4 0.025 161 41 17 11
0.7 0.025 113 31 15 10
1.5 0.025 67 22 13 10
4 0.025 40 17 12 10
1C 10 0.025 29 15 12 9
10 100 1000 10000
(a) PPM OF NOx GASES: 0.4 0.05 154 38 15 10
0.7 0.05 105 28 13 9
300
1.5 0.05 64 20 11 9
PRESSURE =1.3 MPA 4 0.05 36 15 11 8
0 t, \ WEIR HEIGHT: 0.051,4 10 0.05 26 13 10 8
\~ TEMPERATURE =30"C
).7 ~ ~ PARAMETER-'EMPTY HEIGHT(M} 0.4 0.1 146 35 13 8
~ 100 1.5 \
~
\
OTHERDETAILS ARE IN TABLE 5
FEED PLATE=6 0.7 0.1 99 26 11 8
1.5 0.1 60 18 10 7
4 0.1 33 13 9 7
10 0.1 23 11 8 7
m
10. 0.4 0.2 144 33 11 7
0.7 0.2 97 24 10 7
1.5 0.2 57 17 8 6
4 0.2 30 12 8 6
10 100 1000 10000 10 0.2 20 10 7 6
(b) PPM OF NOxGASES

.~00 PRESSURE =1.3MPA 4 m and then to 10m. The total height substantially
0.4 ~ WEIR HEIGHT=0.1 M
\ TEMPERATURE =30°C increases at these empty heights.
t~
L~ ).7~ ~ PARAMETER =EMPTY HEIGHT(M) In addition to the effect of empty height, the effect
~ OTHERDETAILSARE IN TABLE 5 of liquid submergence was also investigated. The re-
~_ 100~
sults have been given in Tables 6 and 7 for the outlet
~-0~ NOx concentrations of 700 and 20ppm, respectively.
m
F o u r levels of liquid submergence (weir height) were
100 considered: 25, 50, 100 and 200 mm. It can be seen that
the number of plates decreases with an increase in the
height of liquid submergence. However, the extent of
1C decrease is nominal when the submergence is in-
10 100 1000 creased from 100 to 200mm. Further, it can be seen
(c) PPM OF NOx GASES
from Tables 6 and 7 that the effect of the empty
Fig. 3. Effect of the exit NOx concentration and the height section is relatively dominant over the effect of weir
of empty section on the number of plates (A) Weir height.
height=0.025m (B) Weir height=0.05m (C) Weir The effect of temperature is shown in Table 7(B).
height = 0.1 m (For operating and design parameters refer The number of plates increases with an increase in
Table 5). temperature. However, it may not be economically
advantageous to operate the absorption column at
low temperatures due to high refrigeration costs. The
liquid a m m o n i a feed stock provides for some refriger-
mean that the total height decreases. In fact, in almost ation load. The rate of heat removal due to vaporiza-
all the cases, the total height was found to increase tion of liquid ammonia can be used advantageously in
with an increase in the height of empty section. F r o m the top section where the rate of oxidation is very low.
Tables 6 and 7, it can be seen that the total height Most of the heat is usually removed by the ambient
increases with an increase in the height of empty water. Hence the operating temperature is decided by
section. The increase is nominal when the empty the ambient temperature. F r o m the foregoing dis-
height is increased from 0.4 to 0.7m. This result is cussion it can be seen that the optimum heights for
useful because the cost associated with a large number empty section and submergence section can be se-
of stages can be reduced. However, a similar result is lected directly. However, the optimum temperature
not obtained when the empty height is increased to and pressure depend upon the geographical location
4582 M. P. Pradhan et al.
Table 7(A). Effect of pressure on total column height temperature = 30°C. Outlet concentration of
NOx = 20 ppm. Other parameters have been listed in Table 5

Height of empty Height of Total column height (m)

section submergence Pressure (MPa)
HE (m) Hw (m) 0.4 0.71 1.31 2.02

0.4 0.025 # 488.25 88.00 32.50

0.7 0.025 # 502.25 99.50 40.75
1.5 0.025 # 539.75 130.00 65.75
4.0 0.025 # 668.25 225.50 145.00
10.0 0.025 # 922.50 461.25 330.75
0.4 0.05 # 465.50 83.75 28.25
0.7 0.05 # 495.50 90.75 33.75
1.5 0.05 # 530.00 110.00 51.25
4.0 0.05 # 611.50 170.00 101.25
10.0 0.05 # 834.00 321.75 221.00
0.4 0.1 # 470.50 97.00 26.50
0.7 0.1 # 505.00 101.00 28.80
1.5 0.1 # 531.25 116.75 36.80
4.0 0.1 # 586.50 176.50 74.00
10.0 0.1 # 747.50 353.50 151.50

Table 7(B) Effect of temperature on number of stages pressure = 0.4 MPa Outlet NO:, concentra-
tion = 700ppm Other parameters have been listed in Table 5

Height of empty Weir Number of stages

section height Temperature (°C)
HE (m) nw (m)
5 15 30 45

0.4 0.025 105 123 161 223

0.7 0.025 74 85 113 147
1.5 0.025 50 53 67 86
0.7 0.05 70 80 105 139
0.7 0.1 64 75 99 133
0.7 0.2 60 71 97 126

and the cost of power. The costs will be considered in N 2 0 4 and the liquid-phase concentration profile of
the optimum design strategy. H N O 3 were plotted over a very wide range of pres-
The effect of hole diameter and percentage free area sure, temperature, H N O 3 concentration and initial
on the number of plates is shown in Table 8. It can be oxidation volume has been covered as are shown in
seen that the number of plates increases with an Figs 4 and 5. The following observation can be noted:
increase in the hole diameter and decreases as the (i) The levels of N203 are much higher than N204.
percentage free area increases. The increase in number (ii) The partial pressures of N 2 0 3 and Y20,, become
of plates with an increase in hole diameter is due to very small in bottom 3 0 4 0 % of the column. In this
the decrease in the gas hold up on the plate. As hole portion the comparison of rates of absorption is more
diameter increases, the bubble size increases and important than N 2 0 4 and NO2. (iii) Because of ab-
hence the bubble rise velocity increases. This phe- sorption of N 2 0 3 , its partial pressure decreases the
nomena results into lower fractional gas hold up and the equilibrium given by eq. (4) favored. As a result,
causes lower absorption efficiencies and reflects into N O gets converted to N203. It was observed that, in
higher number of plates. If free orifice area is in- bottom portion of the column, the N O reduction
creased, the fractional gas hold-up was found to in- occurs mainly due to N 2 0 3 formation. (iv) The partial
crease. Hence, the number of plates decreases. For pressure of N O passes through a maxima. (v) In top
high free area, the possibility of weeping cannot be 50-60% of the column, the NOx gas mainly consists
ruled out. F o r the same free area it was observed that of NO. The conversion of N O occurs mainly due to
weeping occurs earlier at higher pressures. To avoid N O oxidation.
weeping, optimum values of the hole diameter and
percentage free area have to be used. 4.3. Optimum design
In order to understand the rate controlling step, the In order to understand the sensitivity of various
gas concentration profiles of N O , N O 2 , N 2 0 3 and parameters, the total cost (U.S.$/day) was estimated
 Multicomponent gas absorption with multiple reactions 4583
Table 8. Effect of hole diameter and orifice area on number at two levels of outlet N O x concentration: 700 a n d
of plates 150 ppm. F o r the two cases, cost of stainless steel was
t a k e n as 3.52 a n d 7.04 $/kg, respectively. F u r t h e r , the
Number of plates for 700 ppm outlet concentration other
cost of the plate was calculated by eqs (93) a n d (95),
parameters have been listed in Table 5
respectively. The results are given in T a b l e 9(A) a n d
Hole % Orifice area 9(B) u n d e r different c o m b i n a t i o n s of weir a n d e m p t y
diameter section heights a n d at five levels of total pressures of
(mm) 1 2 3.87 6 0.4, 0.71, 1.01, 1.31 a n d 2.02 M P a . It can be seen that
the pressure is a m a j o r p a r a m e t e r which governs the
Pressure = 0.4 MPa, absorption temperature = 30°C
2 150 150 150 148 fixed cost.
3 150 150 150 150 In order to u n d e r s t a n d the overall effect of fixed
4 150 150 150 150 a n d operating cost, the daily c o n t r i b u t i o n was esti-
5 150 150 150 WEEP m a t e d on the basis of 25 a n d 4 5 % of the fixed cost as
6 150 150 150 WEEP
the yearly (300 days) expenditure. While estimating
Pressure = 0.4 MPa, absorption temperature = 5°C
the operating cost, 9 5 % of the c o m p r e s s o r energy was
2 50* 150 92 80
3 * 150 109 93 assumed to be recoverable or 5 % of the cost of com-
4 * 150 126 105 pression was considered for the net o p e r a t i n g expen-
5 * 150 145 WEEP diture. These results were studied at a weir height of
6 * 150 150 WEEP 0.025 m with empty section height as parameter. The
Pressure = 0.7 MPa, absorption temperature = 30°C results are s h o w n in Figs 6(A) a n d (B) for 7 0 0 p p m
2 46 34 30 WEEP outlet c o n c e n t r a t i o n a n d in Figs 6(C) a n d (D) for
3 63 42 34 WEEP
150 p p m outlet c o n c e n t r a t i o n . The following observa-
4 81 51 40 WEEP
5 102 61 46 WEEP tions can be n o t e d from these figures: (i) F o r the
6 131 70 53 WEEP empty heights of 0.4 a n d 0.7 m, the total cost m o n o -
Pressure = 0.7 MPa, absorption temperature = 5°C tonically decreases with a n increase in pressure. T h e
2 38 28 23 WEEP reduction is s h a r p up to 1 M P a a n d thereafter the
3 55 37 28 WEEP reduction is nominal. (ii) F o r the e m p t y heights of 1.5,
4 74 47 34 WEEP 4 a n d 10 m, the total cost curves show minima. The
5 91 58 43 WEEP
6 110 70 50 WEEP curve for 1.5m is shallow whereas it is deep for
1 0 m a n d in between for 4 m . The o p t i m u m pressure
• Pressure reduced below 1 atm, this case is not considered. decreases with an increase in the e m p t y height.

60000 3000
O o o o SYMBOL COMPONENT
O O
50000
~8 o i NO
NO2 2500
N203
O , x N20/,
:~ 40000 2000 z
o_ }6 • [] \ O -- HN03
z z
k _O
300O0
] • NN03 \ O
/ CONC. 0\ 1500

I--
z 20000
Z 1 " A E]~ 0 0 0 1000 ~(J
0
u
10000 ,21 ,, []\ o o

X X X>
I , , ^ X A A ~ ~ l,J ~ -~_
0 2 4 6 8 - 'I0 " 12 - 14 - 'i6 - 18
(a) PLATE N U M B E R

Fig. 4. Gas-phase concentration profiles of NO, NO2, N203, N204 and the liquid phase concentration
profiles of HNO3. (i) Pressure = 0.71 MPa (ii) Temperature = 30°C (iii) Plate spacing = 0.7m (iv) Weir
height = 0.05 m

Figure Initial oxidation volume H N O 3 product

below plate 1 (m 3) Concentration (W/W)

A 0 52%
B 0 58%
C 0 64%
D 30 58%
4584 M. P. P r a d h a n et al.

60000 3500
O ' ~ . . .
~ SYMBOL COMPONENT

o NO 3000
S0000 o "~ o N02 ~8
m ^ o'~ z~ N203
o _ o O ~ "~o o o x N204 2500~
- 4OOOO
_z -O ~O O ~ O " -- NNO3 Z
t6 2000 z
~(xx)0 o Oo Q
I--
Z 0 \ o HNO3 > 24 1500
Ld A z
z~ [] X o CONC. LIJ
20000
o o\ o ° o °
,ooo
AZ~ [] ~ o o 12
10000
z~ 500
• !. Z~A^ O O
X XX X
0 ~, x x ~ , . ~. ~ = . , . ~ - = - , . ~ _ _ _ 0
0 5 10 15 20 25
(b) PLATE NUMBER

6000G z.O00
SYMBOL COMPONENT
o NO 3500
~C
'- - ° ~ 3 o-~ ~ O NO2 b~

• 4000C "°no 3000 Ig

Q.
n
z ~o ° ~ -- HN03 .z~ Z500 z
z z
o
o oooooDO~ o°°°oooo~o ° HNO3CONC. - . , 2000
3oooc
OoooOO V- O - \ 0 > r¢
p- I-
Z OO ~ ° 0 -28
1500 ~
2000C A Z~:Z~Z~Z~Z~Z~ ODD / 0°o_ z
o o
1000 ~
--~,AZ~Z~A,, " O% X u°°° "l&
10000
~00

x x x ¥ x x x x x x x x x I x x ~ v ,,, ";'@~,a ~ A A ='~l=Jmu=m--.=~"~

5 10 15 20 25 30 35 z,O
(c) PLATE NUMBER

60000 30OO
SYMBOL COMPONENT
0 ~ o OOo o o NO
50000 n b~ o D NO2 ,~ ~500
B ~ o > Z~ N203 3E
o o ~ x N20~
- 40000 ZOO0 a_
z D X o __ HNO3
o .36 z
z
o '( [] \ 0
<~ 3oooo o • 1500
t- D O n,-
z
uJ
\ \ o HN, ON . '24
bl
~:
:~ 20oo0 A O 1000 ~
Z
o O
c3
12
10000 500

x x x x x x x x ~ ~ '~ I~ o m ~ ~ _ . . . .
0 0
0 5 10 15 20 25
(d) PLATE NUMBER

Fig. 4. C o n t i n u e d .
 Multicomponent gas absorption with multiple reactions 4585

70000t ,r---,~6000
k\ ,o~?~o SYMBOL COMPONENTI I
60000[~2~,~ A:p" O o NO r I
1% ~K o o NO2 1-48 15000
I ~ \ o , ~ N203 I I
~2 5000015c15:5~"£;L \ o x N204 L I
z - % \ oo - .NO3 I..14000

, %\ °Oo I Fooo
~'30000~ ~ ~ (:~O~ HN03CONC. >t2A I i
sz %\ r._l
Q 1000
10000

~ 0
0 20 40 60 80 100
(a) PLATE NUMBER

500OO 1600
SYMBOL COMPONENT
' "~"~ o o o o NO 1400
o o :' D NO2
4O0OO
48, 1 3 ~ ~X N203 1200
0.
fl_ o o N204 0..
Q..
Z- 30000 ,i o ~ 0 -- HN03 1000 z
Z 36T HNO3CONC" --[3~ o o ZO
800
¢Y m.,
z 20000 i,-
bJ 24 ,= A A C] ~ o 600 z
Z Z
0 0
o 400
10000 12 A
200

0
X X
•
X ~"
.
X
~ %" A =, A ' ~
A
9
0 2 4 6 8 10 12 14
(b) PLATE NUMBER

Fig. 5. Gas-phase concentration profiles of NO, NO2, N203, N204 and liquid-phase concentration profile
of HNO3. (i) Temperature = 30°C (ii) Plate spacing = 0.7 m (ii) Weir height = 0.05 m (iv) HNO3 product
concentration = 58% w/w (v) Initial volume below plate 1 = 3 m 3.

Figure Pressure

A 0.4 MPa
B 1.0 M Pa

However, it must be emphasized that, the optimum timum empty heights are 1.5m, 0.7m, and 0.4m at
costs (at minima) for 4 and 10m empty heights are pressures of 0.7, 1.0 and 1.3 M P a respectively. H o w -
higher than those for 0.4, 0.7 and 1.5 m empty heights ever, if we select other cost components (shell cost
(iii) The total cost for three empty heights (0.4, 0.7 and 3.525/kg) and plate cost given by eq. (93), the results
1.5 m) is comparable in the range of 0.7 1.0 MPa. This are given in Figs 6(A) and (B). F o r this case, at pres-
is very interesting and useful result. (iv) The optimum sure of 0.4 MPa, the total cost is minimum for 1.5 m
plate spacing depends upon the fixed cost compo- empty height. Again the optimum height decreases
nents. For instance, from Figs 6(E) and (F) it can be with an increase in pressure.
seen that, at 0 . 4 M P a pressure the total cost is Figures 7(A) and (B) show the effect of pressure on
minimum for 4 m empty height. With an increase in total cost at weir height of 50 mm. The sensitivity of
pressure, the optimum height decreases. Thus the op- operating cost was considered at two levels. For these
4586 M. P. Pradhan et al.
Table 9(A). Fixed cost ($, million). Other parameters have been listed in Table 5. Concentration of outlet
NOx gas = 700ppm. Css = 3.545/kg. Tray cost = eq. (95) Fabrication cost = 60% of the material cost

Height of Height of Fixed cost ($, million)

empty Submergence of pressure (MPa)
section Hw (m)
He (m) 0.4 0.71 1.01 1.31 2.02

0.4 0.025 4.9 1.42 1.14 0.795 0.713

0.7 0.025 3.7 1.24 1.086 0.854 0.809
1.5 0.025 2.7 1.19 1.2 1.074 1.186
4.0 0.025 2.6 1.64 1.89 1.917 2.365
10.0 0.025 3.5 2.94 3.72 4.111 4.695
0.4 0.05 4.68 1.34 1.085 0.73 0.679
0.7 0.05 3.49 1.15 1.013 0.77 0.759
1.5 0.05 2.62 1.102 1.093 0.94 1.099
4.0 0.05 2.33 1.46 1.66 1.78 1.943
I0.0 0.05 3.14 2.56 3.22 3.46 4.207
0.4 0.1 4.5 1.268 0.975 0.67 0.604
0.7 0.1 3.35 1.089 0.926 0.69 0.72
1.5 0.1 2.49 1.014 0.969 0.88 0.92
4.0 0.1 2.15 1.287 1.422 1.49 1.74
10.0 0.1 3.147 2.187 2.696 2.80 3.72

Table 9(B). Other parameters have been listed in Table 5. Concentration of outlet NOx gas = 150ppm.
C~, = 7.02 $/kg. Tray cost = eq. (93) Fabrication cost = 60% of the material cost

Height of Height of Fixed cost ($, million)

empty Submergence pressure (MPa)
section H~, (m)
He (m) 0.4 0.71 1.01 1.31 2.02

0.4 0.025 # 7.05 4.22 3.87 3.57

0.7 0.025 13.26 7.28 4.96 4.98 4.86
1.5 0.025 25.46 8.95 7.03 7.93 8.37
4 0.025 32.76 14.92 13.83 17.2 19.46
10 0.025 41.8 29.75 30.11 38.82 44.93
0.4 0.05 # 7.72 3.90 3.44 2.93
0.7 0.05 # 7.21 4.29 4.0 3.86
1.5 0.05 18.76 8.15 5.68 5.68 6.2
4 0.05 46.33 12.38 9.82 9.82 13.65
10 0.05 37.53 22.75 20.12 20.12 30.37
0.4 0.1 # 12.78 4.4 3.71 2.56
0.7 0.1 # 8.2 4.65 4.38 3.14
1.5 0.1 # 8.05 5.83 6.34 4.59
4 0.1 30.47 11.2 10.3 12.63 9.3
10 0.1 35.5 19.62 21.12 27.18 21.30

two cases, 5 % a n d 4 0 % of the c o m p r e s s i o n cost was the c o m p r e s s i o n cost (respectively 95, 90, 75 a n d 6 0 %
considered in the o p e r a t i n g cost. It can be seen t h a t energy recovery was considered). In all these cases,
the t r e n d is similar to t h a t in Figures 6(A) to (F). o p t i m u m values of pressure (1 MPa), weir ( 5 0 m m )
Figure 8 shows the effect of weir height o n total cost a n d e m p t y section (0.7m) heights were considered.
with the percentage of o p e r a t i n g cost as parameter. It F r o m Figure 9 it can be seen t h a t the o p t i m u m , value
can be seen t h a t the o p t i m u m value for weir height is is at 3 0 % of excess air at all levels of the operating
0.07 m for lower levels of stainless steel cost a n d it is cost.
0.05 m at higher level of stainless steel cost.
Effect of outlet concentration of NOx gases
Effect of excess air It is possible that the values of o p t i m u m p a r a m e t e r s
Figure 9 shows the effect of excess air o n the total m a y c h a n g e at different outlet P P M of NOx gases.
cost. This effect was investigated at four levels of Hence, it was t h o u g h t desirable to study the effect of
o p e r a t i n g cost. These levels were 5, 10, 25 a n d 4 0 % of various p a r a m e t e r s o n the total cost at different outlet
 Multicomponent gas absorption with multiple reactions 4587
concentrations of NOx. It was observed that, for 1000 investigated on total cost with outlet NOx PPM as
and 300 PPM concentrations, practically similar re- parameter. It can be seen from Figures 8(A) and (B)
sults were reproduced as in the case of 700 PPM and t h a t the o p t i m a shifts to lower side as the P P M level
presented in Figs 6(A), (B), (E), (F) and 8(A) and for decreases. The range for optimum weir height can be
150 P P M s h o w n in Figs 6(C), (D), 8(B) a n d 9. F o r the considered as 0.05-0.06 m for outlet P P M in the range
case of 50 P P M outlet c o n c e n t r a t i o n , the total cost at of 50~1000.
0.4 a n d 0.7 m heights of e m p t y section was observed to
decrease c o n t i n u o u s l y with an increase in pressure. Effect of product acid concentration
However, the cost was found to be comparable at T h e effect of product acid c o n c e n t r a t i o n is s h o w n in
these two empty heights. The effect of weir height was Figure 10. Obviously, the total cost increases with an

400C I •

700C . / 1 0 0

300C

2 500C

=o
I~- 200( 4 i--- 0

°o 300C
J

o 1ooo o

1000 ~ . ~ 0.7
- 04

i i i i
01.5 1 115 2 0!5 1 1'5 2
(a) PRESSURE,MPA (b) PRESSURE, MPA

4O
50

~
10"0

...._____---4 10"0 50
ao

40

~
8 2o
30
d
~ 4.0
£ ----.o 4.0
20
10
~ 1.5 16 1-5
~ 0-7 O7
0.4 0-4

0.'5 110 115 210 2.5 0-0 o's i'o 1!s 2~ 2.5
(c) PRESSURE,, MPa (d) PRESSURE, MPA

Fig. 6. Effect of total pressure and empty height on the total cost. (i) Weir height = 0.025 m, (ii) Temper-
ature 30(?

Figure Outlet NOx C~ Tray cost Fixed cost

ppm $/kg equation component,
(%)

A 700 3.52 (931 25

B 700 3.52 (931 45
C 150 7.04 (951 25
D 150 7.04 (951 45
E 700 3.52 (95l 25
F 700 3.52 (95) 45
4588 M.P. Pradhan et al.

10.0

8000
10.0

o
J 40 tOO( 4-0

200( 1"5

o~.

I
o'.s 1~ l!s 2o
(e) PRESSURE, MP~ (0 PRESSURE,Me*

Fig. 6. Continued.

1
40O0 10.0
4000

10"0

~3000
4,0

o Q
2o00
4.0
oo 1.5
.J .J
0.4
1.5 100(3

i i i
o'5 1 l'.s ~2 01s 1 l~s
(a) PRESSURE~ MPA (b) PRESSURE, Mt~

Fig. 7. Sensitivity of the operating cost on the total cost. (i) Fixed cost component = 35%. (ii) Weir
height = 0.05 m, (iii) Temperature = 30°C, (iv) Parameter = empty height, Other details are in Table 5. (A)
recovery of compressor energy = 95% (B) recovery of compressor energy = 60%. parameter = empty
height. Outlet NOx = 700ppM, Css = 3.525/kg.Tray cost given by eq. (93).

increase in the concentration of nitric acid. However, (3) The extent of oxidation strongly influences the
in all the cases, the optimum total pressure is in the n u m b e r of stages. The extent of oxidation was varied
range of 0.7 to 1 MPa. by changing the height of empty section. The overall
equipment volume was found to increase with an
increase in the height of oxidation section. However,
5. C O N C L U S I O N S the n u m b e r of stages increases with a decrease in the
(1) A mathematical model has been developed for extent of oxidation. Hence, the extent of oxidation is
the design of plate column. The effects of NOx com- an important parameter in the prediction of optimum
position, temperature, pressure, excess air, volume of design.
interstage oxidizers and extent of absorption per stage (4) The optimum design depends nominally on the
have been included in the model. extent of absorption per stage, that is on the weir
(2) The present model has been used for the simula- height.
tion of commercial plate column. The model favour- (5) An optimum design strategy has been presented
ably agrees with the plant data. for the prediction of optimum values of various
 M u l t i c o m p o n e n t gas absorption with multiple reactions 4589

15
2000

40

1600 i
(:3 10 ~
< 40
r'~
Z
< oT 25
23
~ 120~ O3
o
10
5

_J
8
-3

O
801
5

o.o 1o'.o zo'.o 3o'.o 40'.0 so!o 6olo 70 0

EXCESS AIR,('/,)
40C
o!o2 o!o4 o!o6 olo8 o'.I 0112 0.14 Fig. 9. Effect of percentage excess air on the total cost.
(a) WEIR HEIGHT (m) Empty section height = 0.7 m, Weir height = 0.05 m, Tem-
perature = 30°C Pressure = 1.0MPa, Fixed cost c o m p o n -
ent = 35%, Css = 7.045/kg, Outlet N O x = 150ppm, Tray
cost by equation (95j, Parameter = % operating cost.
FIXED COST COMPONENT:35%
HEIGl-rf OF EMPTY SECTION : 0.7M
TEMPERATURE : 30°C
PRESSURE : 1.0 MPA
PARAMETER='/, OF OPERATING COST~o__ 3"0

El
Z
< ~ ~ 2 5 2.5
o"r

2.o
r~
1 z

I
60'.,,,

1-(

o.o2 o.'o, o.;B o. o 0-12 J

(b) WEIR HEIOHT, M

05
Fig. 8. Sensitivity of the weir height on the total cost. (i)
Fixed cost c o m p o n e n t = 35% (ii) Height of empty sec-
tion = 0.7 m (iii) Temperature = 30°C. (iv) Pressure = 1.0 MPa 0 i i i i i ,
0-2 0.4 06- 0"8 1 1.2 1-4
(v) Parameter = % Operating cost
TOTAL PRESSURE, MPa

Figure Outlet NOx C~s Tray cost Fig. 10. Effect of nitric acid concentration on total cost.
ppm $/kg (equation) Empty section height = 0.7 m, Weir height = 0.05 m, Tem-
perature = 30°C, Fixed cost component = 35%, Operating
A 700 3.52 93 cost component = 5%, C,, = 7.04, Outlet NOx = 700ppm,
B 150 7.04 95 Tray cost by eq. (95).

parameters. The o p t i m u m design was analysed for NOTATION

different levels of fixed a n d operating costs, a interfacial area, m2/m 3

(6) The effect of outlet c o n c e n t r a t i o n level of NOx AC area of cooling coil, m 2
gases on various p a r a m e t e r s was also studied. The [Bo] water c o n c e n t r a t i o n
o p t i m u m value for pressure over the p p m range of b weir length per unit b u b b l i n g area, m - 1
50 1000 was found to be in the range of 0.7 to Ccoil cost of cooling coils, $
1.01 M P a . T h e o p t i m u m value of weir height de- Ccomp gas compression cost, $
creases as the c o n c e n t r a t i o n decreases. Cy contingency factor
4590 M. P. Pradhan et al.
Cinternals cost of internals, $ Pnzo partial pressure of inert (water) in bulk of
Co orifice coefficient gas
co~ average specific heat of liquid phase, PnNO~ partial pressure of nitric acid in bulk of
kcal/(kmol K) gas
Cpower electricity cost. Rs./kw h Pr total pressure of gas, kN/m 2
Cpump power required for pumping cooling PP power required for pumping cooling
water, kw/h water
Cring cost of ring, $/kg Q1 heat changes due to formation of N203
Cshell cost of shell, $ in bulk gas, kcal/s
Css cost of SS316, $/kg Q2 heat changes due to formation of N20#
Ctray cost of tray, $/kg in bulk gas, kcal/s
D diameter of shell, m Qa heat changes due to formation of HNO3
dB bubble diameter, m in bulk gas, kcal/s
De diffusivity in gas phase, m2/s Q4 heat changes due to formation of HNO2
dh differential height, m in bulk gas, kcal/s
do orifice diameter, m Q5 heat changes due to formation of NO in
Dpipe diameter of pipe used for cooling coil, m bulk gas, kcal/s
eff efficiency of pump Qc volumetric flow rate of cooling water
f friction factor through coil, ma/s
Fo design type factor Qr total heat change per unit time in the
FM materials factor differential element
Fe pressure factor Rax, g rate of gas-phase mass transfer of com-
F~ cost of fabrication, $/kg ponent x kmol/(m3/s)
Fw feed water, kmol/s Rax,~ rate of absorption of component,
G flow rate of inerts, kmol/s kmol/(m3/s)
g gravitational acceleration, mZ/s Re Reynolds number for flow through coil
H total height of the column, m Rho~s density of stainless steel kg/m a
AH standard heat of reaction, kcal/kmol S cross-sectional area of the column, m 2
Hz Henry's low coeffÉcient, kmol/m 2 T temperature, K
(kN/m 2) t shell thickness
HE height of empty section, m V' mass flow rate of gas, kg/s
Hw weir height, m Vc velocity of the cooling water through
hi volume of liquid per unit plate area, coil, m/s
m2/m 3 Vh gas velocity through the holes, m/s
K6 heterogeneous equilibrium constant, V~ superficial gas velocity, m/s
(kN/m2)- 0.5 W weight fraction of HNO3 in aqueous
K~ heterogeneous equilibrium constant, nitric acid solution
atm -2 Wt weight of shell, kg
K2 equilibrium rate constant for reaction (3) X moles of nitric cid per mole of water
K3 equilibrium rate constant for reaction (4) Y*z moles of oxygen per mole of inerts
K4 equilibrium rate constant for reaction (5) YT total moles of gas per mole of inerts
K5 equilibrium rate constant for reaction (6) Y*, kmoles of reacted nitrogen per kmol of
kl forward reaction rate constant inerts in the gas stream from NH3 oxi-
kL liquid-side mass transfer coefficient, m/s dizer
k~ gas-side mass transfer coefficient, kmol/ YJ2, kmoles of oxygen per kmole of inerts in
[m 2 s (kN/m2)] the gas stream from NH3 oxidizer
L molar flow rate of liquid, kmol/s Y* kmol of reactive nitrogen per kmole of
L' mass flow rate of liquid, kg/s inerts
Lp flow rate of water in the product acid, Y*o kmol of divalent nitrogen per kmole of
kmole/s inerts
mo mass flow rate of liquid, kmol/s Y*~o kmol of water in form of oxyacids and
N* total moles of NOx free water in gas phase/kmol of inerts
(N203)f kmoles of (N203) formed Yx moles of gaseous component x per mole
(N203)a kmoles of (N203) absorbed of inerts
NO* total moles of divalent nitrogen oxides Z stoichiometric coefficient
Px partial pressure of component x in bulk
of gas Greek letters
p'x partial pressure of component x at gas- r/ kinematic viscosity of liquid, m2/s
liquid interface p density, kg/m 3
pO partial pressure of component x in bulk e~ fractional gas hold-up
of gas eL fractional liquid hold-up
 Multicomponent gas absorption with multiple reactions 4591
enhancement factor production of nitric acid. Comput. Chem. Engng. 3,
O~, ultimate enhancement factor 143-150.
AP, pressure drop in cooling coils, kN/m 2 Fair, J. R. (1984) In Gas Liquid Contacting Chemical
APdry dry plate pressure drop, kN/m 2 Engineer's Handbook, Eds R. H. Perry, and C. H.
APLu pressure drop against the liquid head on Chilton, McGraw-Hill, Kogakusha, Tokyo.
the plate, kN/m 2 Gutierrez-Canas, C., Arias, P. L. and Legareeta, J. A.
(1989) Industrial nitrogen oxides absorption simu-
APt Total pressure drop across a plate, lation. Comput. Chem. Engng 13, 985 1002.
kN/m 2 Hoftizer, P. J. and Kwanten, F. J. G. (1972) Absorption
of Nitrous Gases G Nonhabel, Gas Purification
Subscripts Processes for Air Pollution Control. Butterworths,
a absorption London.
C condensate H olma, H. and Sohlo, J. (1979) A mathematical model
e exit condition for an absorption tower of nitrogen oxides in nitric
F outlet acid production. Comput. Chem. Engng 3, 135 141.
Honti, G. D. (1985) Various Processesfor the man@w-
f formation
ture of Commercial Grade Nitric Acid in-Nitric Acid
G, (g) gas phase and Nitrates Fertilizer science and technology
i inlet condition series, 4, C. Keletic (ed.) Marcel Dekker, New York
L, (/) liquid phase and Basel.
l NO2, N203, N204 Joshi, J. B., Mahajani, V. V. and Juvekar, V. A. (1985)
N nth stage Absorption of NO, Gases. Chem. Eng. Commun.
P product acid 33, 1 90.
S side stream from condenser to column Koukolik, M. and Marek, J. (1968) Proceedings q[the
T total 4th European Symposium on Chemical Reaction En-
V vapour phase gineering (Suppl. Chemical Engineering Science),
347, Pergamon, Oxford.
x x in NO~
Koval, E. J. and Peters, M. S. (1960) Reaction of
Aqueous Nitrogen Dioxide. Ind. Engng Chem. 52,
REFERENCES 1011.
Andrew, S. P. S. and Hanson, D. (1961) The dynamics Makhotkin, A. F. and Shamsutdinov, A. M. (1976)
of nitrous gas absorption. Chem. Engng Sci. 14, Khim. Khim. Tekhnol. 19(9), 1411.
105 113. Miller, D. N. (1987) Mass transfer in nitric acid ab-
Carberry, J. (1958) Some remarks on chemical equilib- sorption. A.I.Ch.E.J. 33, 1351 1358.
rium and kinetics in the nitrogen dioxide-water Ray, M. S. and Johnston, D. W. (1989) Chemical
system Chem. Engng Sci. 9, 189 194. Engineering Design Project: A Case Study Ap-
Carleton, A. J. and Valentin, F. H. (1968) Proceedings proach, Topics in Chemical Engineering. 6, Gordon
q[ the 4th European Symposium Chemical Reaction and Breach, New York.
Engineering (Suppl. Chemical Engineering Science), Sobotka, V. (1973) Modelling on an absorption col-
361, Pergamon, Oxford. umn for nitric acid manufacture. Int. Chem. En~3ng
Carta, G. and Pigford, R. L. (1983) Absorption of 13, 718 727.
nitric oxide in nitric acid and water. Ind. Engng. Sherwood, T. K., Pigford, R. L. and Wilke, C. R.
Chem. Fundam. 22(3), 329 335. (1975) Mass Transfi'r. McGraw-Hill, New York.
Counce, R. M. and Perona,, J. J. (1979) Gaseous Suchak, N. J., Jethani, K. R. and Joshi, J. B. (1991)
nitrogen oxide absorption in a sieve plate column. Computer aided design of NO~ gases in water, nitric
Ind. Engng Chem. Fundam. 18, 400-406. acid and mixed solution of nitric and sulphuric
Counce, R. M. and Perona, J. J. (1980) A mathemat- acids. A.I.Ch.E.J. 37, 323.
ical model for nitrogen oxide absorption in a sieve Suchak, N. J., and Joshi, J. B. (1994) Simulation and
plate column. Ind. Engng Chem. Proc. Des. Dev. 19, optimization of NO~ absorption in nitric acid
426 431. manufacture. A.I.Ch.E.J. 40, 944-956.
Counce, R. M. and Perona, J. J. (1983) Scrubbing of Weisweiler, W., Eidam, K., Thiemann, M., Scheibler,
gaseous nitrogen oxides in packed towers. A.I.Ch.E. E. and Wieand, K. W. (1990) Absorption of
J. 29, 26- 32. NO/NO2 in nitric acid. Chem. Engng Technol. 13,
Doraiswamy, L. K. and Sharma, M. M. (1984) Hetero- 97 101.
geneous Rections: Analysis, Examples and Reactor Wiegand, K. H., Scheibler, E. and Thiemann, M.
Design. Wiley Interscience, New York. (1990) Computations for plate columns for NO~
Emig, G., Wohlfahrt, K.and Hoffmann, U. (1979) Ab- absorption by a stage to stage method. Chem.
sorption with simultaneous complex reactions in Engng Technol. 13, 289 297.
both phases, demonstrated by the modeling and Zuiderweg, F. J. (1982) Sieve trays Chem. Engng Sci.
calculation of counter-current flow columns for the 37, 1441 1464.