Chemical Engineering August 2014 - 45

Source: Dejanovi´c et al.
Chem. Eng. Process 68 (2014), in print
A
Figure 6. A fully thermally
coupled column four-product
DWC is shown here with
three partitions (a), and a
simplified, two partitions
equivalent of it (b)
B
F
C
F
C
same for typical column revamp
considerations.
If capacity increase is not a
primary goal, then it is more
probable that all stages could
be accommodated within the
available column height. For
example, a larger surface area packing
could be used. In the worst case,
the required stage count could be
distributed over two existing shells
connected in series. This is also an
option if trays need to be installed
in a DWC, and advanced tray types
could provide either higher capacity
or higher efficiency, depending on
the situation.
So a DWC is equally suitable as a
retrofit option, and this is not limited
to conventional three-product
situations only. As mentioned before,
energy savings will be equivalent
to that expected in a new design,
and this may be even above
50% if a three-column sequence
would be redesigned to accommodate
a four-product DWC. In this
case, two shells should be sufficient
to accommodate the required stage
count, but, even with such a large
reduction in boil-up ratio, it is hard
to expect that a capacity increase
could be obtained. The problem with
four-product separations is that the
pressure and temperature spreads
are rather large and that operating
pressures of individual columns
in a conventional three-column sequence
may differ to such an extent
that lower pressure columns may
become a bottleneck.
However, the situation is not
hopeless because in such a case it
can prove that a third, lowest pressure
column could be used and connected
in parallel to a second one to
accommodate increased vapor traffic.
The lower the boiling range of
components in the feed mixture, the
larger is the probability that both
a new design and a retrofit DWC
will be an attractive alternative
for the conventional sequence. ReA
B
D
D
(a)
(b)
garding
column design and rating
work required, the number of options
is large, but, as demonstrated
elsewhere [20, 21], the predictive
models introduced and used in our
simulation studies are a suitable
tool for such purposes.
If we consider the same aromatics
plant base case (Figure 5, Table
1), then in the case of a revamp,
we need only to accommodate the
internal configuration as required
in new design (Figure 7) into two
existing shells. Columns 2 and 1
from the original sequence (Figure
5) were designed to operate with
an overhead pressure of 2.5 and 1.7
bar, respectively. Both have an internal
diameter of 2 m. The former
was equipped with 59 and the latter
with 61 sieve trays. Considering
the utilized tray spacing, two shells
of columns 2 and 1 connected in series
provide a total active height (70
m) that is well above that required
in the new DWC design (64 m,
equipped with B1-350MN packing
[19]). This is a comfortable situation
that provides some flexibility. For instance,
a coarser packing to accommodate
increased vapor volumetric
flowrate could be used in the lower
pressure column. This, however, was
not needed in the present case.
Figure 10 shows the result schematically,
that is, the layout and
internal configuration of the fourproduct
DWC arranged using two
available columns connected in
series. The main feature of this revamp
is reflected in the fact that
each of the two shells contains only
Figure 7. This diagram shows the details of the
internal arrangement of the fourproduct
DWC with two partition walls
2.1
3.2
1.1 2.2
3.3
3.4
1.2
3.5
Source:
Oluji´c et al.,
Chem. Eng. Technol.
35 (2012) 1392-1404
ChemiCal engineering www.Che.Com august 2014 45
http://www.Che.Com
Chemical Engineering August 2014

Table of Contents for the Digital Edition of Chemical Engineering August 2014
