Chemical Engineering November 2013 - 30

Cover Story
was unknown since the column was
designed to operate under a much
higher feed flowrate.
The tray vendor was consulted by
communicating the symptoms, described
as follows:
i) Rapid increase in tray differential
pressure
ii) Fluctuating overhead pressure
iii) Fluctuating column sump level
iv) Reflux drum flooding
The vendor concurred that the
cause of these symptoms was flooding.
They concluded that it was downcomer
flooding.
Theories
As in a typical troubleshooting investigation,
many theories were proposed
by various members of the startup
team (including the vendor and outsiders)
to explain the premature flood.
Some could be quickly eliminated, others
took more testing.
1. Tray plugging. Tray plugging
is a common experience in refinery
sour-water strippers [4]. However, in
this tower there was little presence
of fouling compounds, such as heavy
hydrocarbons and asphaltenes. The
feed was filtered down to 2,000 micron
particles to protect the feed/
bottoms exchanger (plate-and-frame
type). If fouling was an issue, the
filters would have needed frequent
cleaning and this was not the case. In
an endeavor to address fouling, the
filter screen was changed from 2,000
micron to 250 micron. Even after
changing, the filters blocked only
once, indicating little solids in the
feed. Further, the exchanger should
have fouled before the trays and this
was not observed. Also, the problem
was first observed after a short run,
much shorter than it takes to plug the
trays of a typical sour-water stripper
even under heavy fouling conditions
(see, for instance, Ref. 4). Based on
these observations, plugging was
branded a highly unlikely root cause
for flooding in this column.
2. Troublesome reboiler. Trap-out
kettle reboilers are notoriously troublesome,
usually due to poor hydraulic
design of the lines leading to or from
the kettle, incorrect force balance on
the kettle, or liquid weep from the tray
feeding the kettle [5].
25
20
15
10
5
10
20
Early warnings of flood
(rapid rise/fall in
column pressure drop)
30
Time, min
40
LEGEND
Column feed, m3
/h
Steam rate, kg/h
Reflux drum level, %
Tray pressure drop, bar
Feed gradually increased
approximately 1 m3
every 5 minutes
/h
Steam
rate increased to
maintain constant reflux
Reflux drum level
maintained
Entrainment
of column
overhead
25
50
75
100
50
Reflux drum
level % of
inst. range
0.3
0.35
0.1
0.2
60
Tray differential
pressure bar
Figure 4. An event timing analysis of the DCS data showing the variation of reflux
drum level and stripper differential pressure as feed and steam rate were gradually
increased
In this column, an issue with the
kettle was easily ruled out by operating
the column with direct injection of
stripping steam instead of the kettle
reboiler. The flood occurred at the
same rates irrespective of whether the
reboiler or the stripping steam injection
was used.
3. Incorrect inlet/outlet nozzle designs
and velocities were closely
checked. The possibility of flashing
at the feed inlet was addressed by
reviewing operating data. The preheater
was designed to preheat the
feed from 55°C to 110°C, but during
the initial runs the feed was colder,
coming in at 40°C, and the preheat
temperature was consequently lower
at 90 -95°C, so flashing would have
been highly unlikely.
4. Incorrect tray design. One concern
was that the tray open slot area
was low, about 7.3% of the active
area, and this could result in premature
jet flood or excessive downcomer
backup. This is addressed further in
the next section.
5. Foaming. Sour water strippers are
generally considered foaming systems.
The industry's practice for system factors,
reflected by the factors listed in
Table 14-9 in Perry's Handbook (pp.
14-41 [6]), lists sour water strippers
as " Heavy Foaming " and assigns them
a system factor of 0.5 to 0.7.
30 ChemiCal engineering www.Che.Com november 2013
Further support for possible foaming
came from the observation that
there was a higher than anticipated
hydrocarbon content (about 20,000
ppm oil) in the feed to the sour-water
feed tanks (Figure 1) and an increased
quantity of particulate matter in the
feed compared to the very original
design. The filtration specification on
the sour-water-stripper feed filter was
relaxed after the vendor of the feed/
bottoms plate-and-frame exchanger
stated that a 2,000-micron filtration
level was sufficient to prevent excessive
fouling without the requirement
for the operator to clean this filter frequently.
The original specification on
the filter of 5 microns was expected to
result in frequent cleaning of this filter.
Particulate matter and heavy hydrocarbons
are well-known to promote
foaming [3, 7].
Our experience in general has been
that most sour-water strippers do not
foam. The times we experienced foaming
in sour water strippers were when
significant oil entered the system.
Solids may also catalyze foams, but
we are not aware of cause-and-effect
experiences in which solids by themselves
catalyzed foams in sour water
strippers. Among more than 60 case
studies of foaming experiences surveyed
from the open literature and
abstracted in Kister's " Distillation
Reflux
drum
floods
2,250
2,000
1,750
1,500
Feed rate reduced to
bring column out of
flood conditions
3,000
2,750
2,500
Column feed rate, m3
/h
Stripping steam, kg/h
http://www.Che.Com
Chemical Engineering November 2013

Table of Contents for the Digital Edition of Chemical Engineering November 2013
