Documente Academic
Documente Profesional
Documente Cultură
[y
i
=(m 1)[ (3)
where, y
i
is the composition of component i in vapor phase. The
temperature of compressed vapor, T
C
(K) that enters the bottom of
the rectier is obtained as:
T
C
= T
B
_
P
C
P
B
_m1
m
(4)
where, P
C
/P
B
signies the compression ratio (CR).
3.3. Energy consumption
There are basically two approaches followed in investigating the
energetic performance improvement of a distillation column: one
is in terms of energy efciency and second one is in the aspects of
energy savings. In the former approach [1,28], the thermodynamic
efciency of both the conventional column and its heat integrated
counterpart is calculated, based on which, the energetic perfor-
mance improvement is obtained. In the later approach, the energy
savings achieved by the heat integrated scheme is estimated with
reference to the conventional stand alone column. In the present
study, the second approach is followed and it is briey presented
below (e.g., [29,30]).
For computing the heat duty of a heat integrated batch distil-
lation (HIBD), Q
HIBD
we add the reboiler duty, Q
R
with three times
the compressor duty, Q
comp
as:
Q
HIBD
= Q
R
3Q
comp
(5)
As suggested by Iwakabe et al. [30], it is logical to assume that
3 kW of thermal energy is needed to produce 1 kW of electrical
power.
We employ the above Eq. (5) for both startup period, Q
SP
and
production period, Q
PP
. Accordingly, the total energy consumption,
Q
total
is obtained as:
Q
total
= Q
SP
Q
PP
(6)
In order to calculate the energy savings achieved by the IHIBDJR
with reference to the CBD, we use:
% Energy savings =
Q
CBD
total
Q
IHIBDJR
total
Q
CBD
total
100 (7)
3.4. Economics
It is experienced that the improvement in thermodynamic ef-
ciency is usually achieved at the expense of capital investment.
Hence, once a thermodynamically efcient process scheme is
identied, the economic feasibility is tested in the next. This section
deals with the cost estimation of batch distillation columns in
terms of total annualized cost (TAC), which has the following form:
TAC($=yr) = operating cost(OC)
capital investment(CI)
payback period
(8)
Here the capital cost combines the cost of distillation column,
heat exchangers and compressor. The cost estimating formulas are
documented in Table 1. For the sake of simplicity, the operating
costs are taken to be identical to utility (heating steam, cooling
water and electricity) costs, i.e. the number resulting from the
summation of electricity (0.084 $/kW h), steam (17 $/ton) and
cooling water (CW) (0.06 $/ton) costs [13] for a year with 330
working days. Furthermore, we consider the operation of a single
batch cycle per day. For computing the TAC, we assume a payback
period of 3 years. The operating cost of the compressor is calculated
based on the bhp and a motor efciency of 0.8 [31]. It is worthy to
mention that in this cost estimation, a penalty of 10% is deliberately
added to the corresponding capital investment taking into account
the complexity involved in fabricating the heat integrated
structure.
4. A novel combination of internal and external heat
integrations: proposed VRCeIHIBDJR scheme
In the conventional batch distillation, the rectier top tem-
perature is less than the reboiler temperature. As mentioned
Table 1
Cost estimating formula [31] and parameter value [32,33].
- Column shell:
_
M&S
280
_
101:9D
1:066
c
L
0:802
c
(c
in
cmcp)
where, D
c
is the column diameter, L
c
the column height, M & S the
Marshall and Swift index (=980), and the coefcients c
in
= 2.18, c
m
= 3.67
and c
p
= 1.2.
- Column tray:
_
M&S
280
_
4:7D
1:55
c
Lc(cs ct cm)
where, the coefcients c
s
= 1, c
t
= 0 and c
m
= 1.7.
- Heat exchanger:
_
M&S
280
_
101:3A
0:65
(c
in
cm(ct cp))
where, the coefcients c
in
= 2.29, c
m
= 3.75, c
t
= 0.1 and c
p
= 1.35.
- Compressor:
_
M&S
280
_
517:5Q
0:83
comp
(2:11 F
C
)
Where, the correction factor for the compressor (F
C
) is taken as 1.0.
C
2
Distillate
Heat C
1
Condenser
Overhead
vapor
Fig. 2. The proposed VRCeIHIBDJR scheme [C = compressor].
D. Maiti, A.K. Jana / Applied Thermal Engineering 59 (2013) 405e413 407
earlier, when the temperature difference between the liquid
leaving the bottom tray and the overhead vapor is reasonably
high, the energy intensication within the framework of direct
vapor recompression mechanism is not economically attractive.
However, in case of the IHIBDJR scheme, usually the whole recti-
cation tower remains hotter than the jacketed reboiler. Therefore,
the possibility of external heat integration in the IHIBDJR is
generated, which may further strengthen its advantages over the
CBD column.
There are mainly two ways for applying the external energy
integration technique in the IHIBDJR structure. One way is to
thermally couple the overhead vapor with the bottom liquid
directly (without performing vapor compression) and the second
approach suggests the compression of overhead vapor before
coupling it with the reboiler liquid. In the later approach, attempt is
made to create an adequate temperature difference between the
coupled streams so that the latent heat of compressed overhead
vapor is optimally used in the still.
In this work, we follow the second approach that leads to the
introduction of VRC mechanism in the IHIBDJR column. This novel
combination of internal and external heat integrations is illustrated
in Fig. 2 and we call this combined structure here as the VRCe
IHIBDJR scheme. Unlike the conventional batch distillation, the
IHIBDJR column experiences a reverse trend in temperature prole
from the column top to the jacketed still (i.e., overhead vapor is
hotter than reboiler liquid). As a result, the second compressor (C
2
)
requires a reasonably low operating CR, thereby leading to an
improvement of overall thermodynamic efciency.
The conguration of VRCeIHIBDJR has already been claried
and we now attempt to develop an unsteady state heat integrated
batch rectier with following a systematic route, starting from the
thermodynamic feasibility test, through the compression ratio
tuning, to the economic analysis. In the subsequent section, the
proposed VRCeIHIBDJR mechanism is derived for a batch reactive
distillation performing several simulation experiments.
5. An illustrative example: a batch reactive distillation
5.1. Process description and mathematical model
The conventional batch reactive distillation (CBRD) operates at
atmospheric pressure and has total 16 trays, excluding the reboiler
and total condenser. The trays are counted from bottom to top;
bottommost stage is the 1st tray and topmost stage the 16th tray.
The column consists of two sections, namely rectifying section (2nd
to 16th tray) and reactive zone (reboiler and 1st tray). By con-
ducting sensitivity tests (not shown), the trays are selected for
these two sections mainly based on (i) batch time, and (ii) reaction
conversion.
5.1.1. Reaction kinetics
The representative column is a heterogeneously catalyzed pro-
cess, and it produces n-butyl acetate (BuAc) and water (H
2
O) by the
esterication of n-butanol (BuOH) with acetic acid (HAc). The
following reaction proceeds only in the liquid phase.
The rectication zone operates exactly as a nonreactive distil-
lation column, purifying product stream. n-Butyl acetate is formed
in the reaction zone and gets accumulated in the still. For the
esterication reaction, the pseudohomogeneous (PH) model has
been used. Esterication reactions are known to be reversible re-
actions of second order. Accordingly, the PH model [34] has the
following form:
r
j
= m
dx
j
dt
= d
j
M
cat
K
f
_
a
HAc
a
BuOH
1
K
a
a
BuAc
a
H2O
_
(9)
where,
a
j
= x
j
g
j
(10)
Here, r
j
denotes the reaction rate of component j (mol/s), x
j
the
liquid phase composition of jth component (mol fract.), d
j
the
stoichiometric coefcient of jth component, M
cat
the mass of the
catalyst (g), K
f
the forward reaction rate constant for esterication
(mol/g.s), K
a
the equilibrium constant of esterication, a
j
the ac-
tivity of jth component in the liquid phase, and g
j
the activity co-
efcient of jth component.
The temperature dependency of the rate constants K
j
is
expressed by the Arrhenius law:
K
j
= K
0
j
exp
_
E
j
=RT
_
(11)
The kinetic data for the PH model are reported in Table 2. The
UNIQUAC model is used to determine the activity coefcients in the
liquid phase. Values of UNIQUAC parameters and Antoine constants
are documented in Table 3.
5.1.2. Distillation model
The fundamental model [36] for the batch reactive rectier is
derived based on the following assumptions:
v Negligible tray vapor holdups
v Variable liquid holdup in each tray
v Perfect mixing and equilibrium on all trays
v No chemical reactions in the vapor phase
v Reaction occurred only in the reactive zone
v Total condensation with no subcooling
v No azeotropes formation
v Constant stage pressure drop (=0.3 kPa) and tray efciency
(vapor-phase Murphree efciency = 75%)
v Modied Raoults law (y
j
= g
j
P
0
j
x
j
=P
t
) for the vaporeliquid
equilibrium
Table 2
Kinetic parameters for the pseudohomogeneous model [34].
Reaction j K
0
j
(mol/g.s) E
j
(kJ/mol)
Esterication 1 6.1084 10
4
56.67
Hydrolysis 1 9.8420 10
4
67.66
Table 3
Thermodynamic data for the butyl acetate system [35].
Antoine constants, and the area and volume parameters for the UNIQUAC model
Component (no.) A B C r q
HAc (1) 22.1001 3654.62 45.392 2.2024 2.072
BuOH (2) 21.9783 3080.66 96.150 3.4543 3.052
BAc (3) 21.07637 3151.09 69.150 4.8274 4.196
H
2
O (4) 23.2256 3835.18 45.343 0.92 1.4
Antoine equation
ln(P
0
) = A (B/(T C)), where P
0
in Pa and T in K.
Binary interaction parameters for the UNIQUAC model (cal/mol)
U
11
= 0.0 U
21
= 68.0083 U
31
= 712.2349 U
41
= 527.9269
U
12
= 298.4344 U
22
= 0.0 U
32
= 24.6386 U
42
= 581.1471
U
13
= 131.7686 U
23
= 148.2833 U
33
= 0.0 U
43
= 461.4747
U
14
= 343.593 U
24
= 82.5336 U
34
= 685.71 U
44
= 0.0
Lattice coordination number (Z) = 10
D. Maiti, A.K. Jana / Applied Thermal Engineering 59 (2013) 405e413 408
v UNIQUAC model to compute activity coefcients (g
j
)
v Nonlinear Francis weir formula for liquid hydraulics
calculations
Note that P
0
j
denotes the vapor pressure of jth component and P
t
the total pressure. The scheme of a typical nthplate is showninFig. 3.
The plate is fed with a liquid feed mixture. Side streams are with-
drawn in state of both liquid and vapor. The model of a typical nth
plate consisting of MESH (Material balance, vaporeliquid Equilib-
rium, mole fraction Summation and Heat balance) equations is pre-
sented below.
Total mole balance
_ m
n
= L
n1
V
n1
F
n
_
L
n
S
L
n
_
_
V
n
S
V
n
_
R
f
C
j =1
d
j
r
n
(12)
Component mole balance
_ m
n
_ x
n; j
= L
n1
x
n1;j
V
n1
y
n1;j
F
n
z
n;j
_
L
n
S
L
n
_
x
n;j
_
V
n
S
V
n
_
y
n;j
R
f
d
j
r
n
(13)
Energy balance
_ m
n
_
H
L
n
= L
n1
H
L
n1
V
n1
H
V
n1
F
n
H
F
n
_
L
n
S
L
n
_
H
L
n
_
V
n
S
V
n
_
H
V
n
R
f
r
n
H
r
n
Q
n
(14)
Equilibrium
y
n;j
= k
n;j
x
n;j
(15)
Summation
C
j =1
x
n;j
= 1 (16)
C
j =1
y
n;j
= 1 (17)
In the above modeling equations, x
n,j
is the mole fraction of
component j in a liquid stream leaving nth tray, y
n,j
the mole frac-
tion of j in a vapor stream leaving nth tray, L
n
the liquid ow rate
leaving nth tray, V
n
the vapor ow rate leaving nth tray, F
n
the feed
ow rate entering nth tray, m
n
the liquid holdup on nth tray, H
F
n
the
enthalpy of feed stream entering nth tray, H
L
n
the enthalpy of a
liquid stream leaving nth tray, H
V
n
the enthalpy of a vapor stream
leaving nth tray, k
n,j
the vaporeliquid equilibrium coefcient spe-
cic to nth tray and jth component, Q
n
the heat loss from nth tray,
and H
r
n
the heat of reaction for nth tray. Here, R
f
is a multiplication
factor which takes values equal to zero for nonreactive and one for
reactive stage. The dot symbol (_) is used to represent the time de-
rivative. The time derivative of the multiplication of two variables,
say m and x, is denoted here by _ m_ x (=d(mx)/dt).
Heat of formation is considered while computing the enthalpies
of streams by means of which the heat of reaction term can be
removed from the energy balance equation. In the simulation,
algebraic formof equations [37] has been simulated to compute the
vapor and liquid enthalpies. It is worthy to mention that for a batch
distillation, values of feed ow rate (F) and bottomow rate (B) are
taken as zero. In addition, in the sample batch reactive column, no
side streams (S
L
n
= S
V
n
= 0) are involved. Negligible heat loss from
a stage to the surroundings (Q
n
= 0) is also assumed.
The internal heat integration approach will be applied on this
representative batch reactive distillation. The resulting internally
heat integrated batch reactive distillation with a jacketed reboiler
(IHIBRDJR) scheme additionally includes Eqs. (1)e(4). Like the
CBRD model, the IHIBRDJR model structure, consisting of mass
balance, energy balance and algebraic equations/correlations, is
derived in the form of differential-algebraic equations (DAE). We
have developed the computer code using the language C/C for
rigorous simulation of the DAE system.
The operating conditions and column specications are reported
in Table 4. Table 5 shows the still and reux drum compositions at
steady state and at the end of batch operation. The representative
column is treated with a distillate policy [38] based on which, the
lightest product, water is removed as distillate at the early stage of
production phase. As a consequence, the column gets progressively
richer with the main product, butyl acetate. It is obvious fromTable 5
that the nal mixture in the still consists mainly of butyl acetate and a
small amount of reactants. In order to present a meaningful compar-
isonbetweenthe CBRDandIHIBRDJR, along withthe identical startup
conditions, we attempt to keep the output specications same.
5.2. Development of internally heat integrated batch reactive
distillation with a jacketed reboiler (IHIBRDJR) scheme
5.2.1. Thermodynamic feasibility
With no heat integration between the rectifying tower and the
jacketed still, the IHIBRDJR will run with maximum condenser and
reboiler duties. To transfer any heat fromthe rectier to the still pot,
there must be a thermal driving force. After 2 h of starting the
Ln 1 + Vn
x j n , 1 +
y
j n,
S
V
n
, y
j n,
Fn
,
z j n,
S
L
n
,
x j n,
Ln Vn 1
x j n,
y
j n , 1
Tray n
(holdup =
mn
)
Fig. 3. Quantities associated with a typical nth tray.
Table 4
Operating conditions and column specications for both CBRD and IHIBRDJR.
System HAc/BuOH/BuAc/H
2
O
Feed composition (mol fract) 0.45/0.45/0.05/0.05
Total feed charge (kmol) 180 (reboiler) 16 0.9 (trays)
6 (reux drum)
Reboiler heat duty (kW) 878.43
Reboiler pressure (atm) 1.0
Distillate rate (kmol/hr) 11.44
Column diameter (m) 0.586
Column length (m) 10.12
D. Maiti, A.K. Jana / Applied Thermal Engineering 59 (2013) 405e413 409
column (startup phase), Fig. 4 displays the temperature for each
stage in rectifying section (T
N
) and that in reboiler (T
B
), before
performing any internal heat transfer. The temperature prole
along the rectier is obtained at a typical compression ratio of 2.4.
As shown in Fig. 4, one may nd that the temperature driving force
is always positive between all the rectifying stages and the jacketed
still, i.e. rectifying column is hotter. It clearly indicates the possible
heat transfer for butyl acetate system from the rectier to the still
pot, and the sample batch reactive distillation is a suitable case for
thermal integration. Because of the reason stated previously, the
feasibility region of thermal integration is selected to a height
equivalent up to 12th stage of the rectication tower at the starting
of a batch cycle.
5.2.2. Tuning of compression ratio (CR)
Now we wish to tune the CR of the IHIBRDJR by comparing its
overall performance with reference to the conventional stand alone
column. In this regard, a number of sensitivity tests have been
conducted and the results are summarized in Table 6. For a suc-
cessful thermally coupled structure, we should achieve positive
energy savings and economic benets. Here, the operating CR is
tuned on the basis of these two criteria along with the cut off value
of DT
min
of 10
C. It is reasonable to assume a driving force
(=DT =T
N
T
B
) of about 10