Reliable design of sour water strippers

A simulation model aims to raise confidence in the reliability of sour water

stripper design

Nathan Hatcher and Ralph Weiland Optimized Gas Treating

O
il refining always generates sour water, and the corrosion of heat exchange surfaces,
and within a refinery there are numerous especially in areas where there are gas liquid
sources. Most refinery sour water systems interfaces and where there are sudden tempera-
contain very little CO2. The H2S content makes ture changes on heat transfer surfaces, such as
water sour, and H2S levels can become very when heat transfer is controlled by liquid level in
high. The capacity of ammonia solutions for H2S an exchanger
is a direct result of ammonias alkalinity, which Amine regenerator reflux water purges can
neutralises the hydrogen ion liberated by H2S also be a significant source of ammonia.
dissociation when it dissolves into the solution. The sour water generated in refineries is gener-
Although not the usual case, in principle, and ally classified as being either phenolic or
with enough H2S partial pressure, there can even non-phenolic. Non-phenolic water contains
be more H2S than ammonia. The potentially very almost exclusively NH3, H2S and possibly a trace
high H2S content can make sour water extremely of CO2; it is generated by refinery hydrotreating
foul and, if the H2S is not recovered, pollution (hydrodesulphurisation, or HDS) units. When
levels would be completely out of hand. Many stripped of contaminants, non-phenolic water
sour4 water sources have been noted.1 Sources can typically be recycled for reuse in the HDS as
include the following: wash water, or it can be used as make-up water
Many refining units use live steam and heat to the crude desalting process. This article
for fractionation, and live steam for velocity considers only non-phenolic water. Phenolic (or,
assist and temperature control in fired heaters. more broadly, non-HDS) water contains
Nitrogen in the presence of heat and a hydrogen compounds that can harm upstream units if used
source (such as a hydrocarbon) forms ammonia. as wash water in them. Typical contaminants
The steam is condensed and recovered in the include salts, phenols and caustic. However,
overhead circuit of the crude unit, FCC unit or make-up water to processing units must first be
coker unit treated, so maximising water reuse is desirable
In the case of refinery hydrotreating, hydrogen to minimise attendant operating costs. Other
gas and a metal catalyst are used to saturate sources of water to sour water stripping units are
olefins. Hydrotreating also converts sulphur- process drums, crude desalting units,
containing hydrocarbons to H2S, and scrubbing of hydrocarbons following caustic
nitrogen-bearing hydrocarbons to ammonia treatment for mercaptans, COS and final H2S
Although ammonia is considerably more vola- removal, TGU quench columns, and various
tile than most alkanolamines, it has a high effluent drains for removing the water used to
affinity for water. Ammonia is removed from prevent salt deposition in equipment.1
hydrocarbon products by injecting wash water It may be useful to point out that ammonia
into the gas and cooling the mixture at an and hydrogen sulphide have almost unlimited
elevated pressure to condense the water. This solubility in water when they are present
provides an irresistible invitation for ammonia together. This is a possibly interesting conse-
to enter the aqueous phase quence of the fact that the reactive component of
Wash water prevents the accumulation of salts the solvent, ammonia, is volatile and, if present

[Link]/article/1000556 PTQ Q3 2012 1

 Typical energy usage in the
Stripped SWS
water 11 8 10 AG stripping process is in the range
Cooler 7 1.0-1.5 lb of 50 psig equivalent
9 saturated steam per gallon of
Sour
1 2
sour water.
water
Feed/ When an external reboiler is
bottoms SWS
used, steam pressure can often
be higher than in an amine
regenerator to minimise the
heat exchange surface. In an
Live
3 5 steam amine regenerator, amine
4 degradation limits tempera-
Pump-1 tures. In a sour water stripper,
there is little or no ammonia in
Figure 1 Sour water stripper with live steam injection the stripped water in the
reboiler, so these concerns do
in the gas phase, it will continue to absorb as not exist. However, there is a practical limit of
long as it becomes protonated as a result of H2S 400-450F, where coking heavy hydrocarbons
co-absorption. Thus, it is conceivable that a can lead to fouling and solids deposition in the
particular sour water stream may be a lot more reboiler. Corrosion is always a concern.
concentrated in both ammonia and hydrogen Higher NH3 and H2S concentrations require
sulphide than the solubility of either component more stripping energy. Since H2S solubility relies
alone would suggest. on the presence of ammonia, the molar concen-
tration of H2S very rarely exceeds that of
Basic sour water stripping process ammonia, and then usually only in dilute
At first glance, sour water stripping is a simple systems. A typical molar ratio of H2S to ammo-
process in which either external steam, steam nia is 0.5-0.8 in the combined sour water
generated by a reboiler, or even a hot hydrocar- stripper feed water of a typical refinery.
bon stripping vapour is used to shift chemical Ammonia levels in the water are often deter-
reaction equilibria by heating the sour water. mined by upstream process conditions, and they
Stripping vapour is the gaseous solvent used to can be highly specific to the process licensor and
remove and carry the ammonia and H2S out of crude slate in operation. Obviously, higher
the system. In other words, the application of concentrations of NH3 and H2S are preferred
heat generates internal stripping steam (equiva- from a water consumption perspective. However,
lently, live steam injection or stripping gas can there is a practical limit of between one and
be used) and removes ammonia, H2S and CO2 several weight per cent ammonium bisulphide
from the water by: equivalent in the sour water feed before metal-
Heating the sour water feed to boiling point lurgy must be upgraded.
Reversing chemical reactions Trays have historically been used in sour water
Diluting the partial pressure of the gases strippers, but random packing is beginning to
stripped by furnishing excess steam. see use in units processing relatively clean water.
The process is very similar to amine regenera- Trays with directional, fixed valves have been
tion. Figure 1 shows a typical sour water stripper reported to be more resistant to fouling, because
column with heating by the injection of live the horizontal velocity imparted as the gas leaves
steam. Since a sour water stripper does not form each valve tends to sweep clean the area near the
a closed loop in the same sense that an amine valves.2
regenerator does, maintaining a water balance is Stripped sour water specifications for NH3 and
unnecessary. This means that live steam can be H2S can be highly dependent upon the locale
used as a stripping agent either alone or in where the unit is installed and the final discharge
conjunction with a conventional reboiler, and the requirements. NH3 is harder to strip than H2S,
additional water added by the condensate is and typical targets for NH3 are 30-80 ppmw in
simply added to the refinerys water inventory. the stripped water versus undetectable to less

2 PTQ Q3 2012 [Link]/article/1000556

than 0.1 ppmw for H2S. Typical recent installa- Kinetics of the carbamate-forming reaction,
tions3,4 involve 35-45 actual trays, with overall Equation 7
tray efficiencies quoted anywhere from 25 to An appreciation for mass transfer as a rate
45%. process where Henrys Law applies.
The thermodynamic framework on which the
Chemistry of ammonia with acid gases vapour-liquid calculations are based involves a
Ammonia is a relatively weak base capable of model for the activity coefficients via extended
being mono-protonated. For example, in aque- Debye-Hckel theory. Thus, the basic modelling
ous solution, it forms ammonium ion (protonated approach for ammonia is the same as for any
ammonia) to a limited extent: other amine, with one exception: ammonia is
highly volatile and this leads to different distri-
NH3 + H+ NH+4 (1) butions in absorption and especially regeneration
equipment compared with amines. Its volatility
This reaction is in perfect parallel with amine is often responsible for it getting into the sour
protonation, so in this sense ammonia can be water system in the first place (through
thought of as nothing more than just another condenser water blow-down, for example), but
reactive amine. There is a great deal of nonsense volatility also allows it to be recovered relatively
written in various books and other publications easily from the same sour water. Heat stable
concerning the reactions of H2S and CO2 with salts and alkali salt contaminants also affect the
ammonia. All reactions involve ionic species and distribution of ionic species in solution and their
all reaction products are ions. Ions do not form presence must be considered, too, and is the
ionic compounds except possibly when materials subject of a future article.
begin to precipitate. As long as all species are in
solution, they exist as individual ions and are not Simulating sour water strippers
associated with each other. All ions are freely Traditionally, sour water strippers have been
floating and the only associations are ones that modelled as a series of equilibrium stages, with
result in solution non-ideality. Thus, there is no overall stage efficiencies quoted anywhere in the
such thing as ammonium carbonate, ammonium range of 15% to 45%; that is, ranging over a
bicarbonate, ammonium bisulphide or diammo- factor of three. However, the mass transfer rate-
nium sulphide in solution. It is very hard to based approach to the simulation of amine
make sense of reaction equilibria unless one contactors and such hard-to-model distillation
discards the existence of such compounds. processes as extractive, azeotropic and reactive
The reactions that occur when H2S and CO2 distillation have been in commercial use since
dissolve in solution are the same as in any other the mid- to late 1980s. Their extension to sour
aqueous, primary or secondary amine system. water stripping is a natural progression and, in
The reactions are: December, 2011 a commercial mass transfer
rate-based sour water stripper model became
H2O H+ + OH- (2) available as part of the ProTreat simulation
H2S H+ + HS- (3) package. The remainder of this article uses the
HS- H+ + S= (4) ProTreat simulators sour water stripper model
CO2 + H2O H+ + HCO-3 (5)
to reveal some rather surprising facts and behav-
HCO-3 H+ + CO=3 ( 6)
iours of sour water strippers.
NH3 + CO2 NH2COO- + H+ (7)

Phase equilibrium
Ammonia drives all these reactions to the right The vapour pressure of ammonia and acid gases
by serving as a proton sink per reaction (1). above sour water plays a key (but by no means
Understanding sour water strippers, understand- the only) role in determining the extent to which
ing the way ammonia behaves as a solvent for a given column and process configuration will
acid gases, and understanding what happens to purify sour water and how the process operating
ammonia in amine regenerators depends on conditions affect the separation. We have devel-
knowledge of: oped an activity coefficient model for phase
Vapour-liquid equilibria behaviour for sour water that uses the

[Link]/article/1000556 PTQ Q3 2012 3

 parity plots indicating partial


 pressure ranges in Pascals, and
the goodness of fit.
C a lc ula t e d
NH pa r t ia l



Mass transfer rate model
pre ssure ,
Pa

The sour water stripper model


 in ProTreat deals with the sepa-
ration of ammonia, the acid

gases and water as governed by
the particular mass transfer rate
of each individual component.
 Ammonia is treated as reactive
  
 
 


Measured NH 3 partial pressure, Pa
towards CO2 in the same sense
that any other primary or
secondary amine is reactive, and



transfer rates are determined by
individual (vapour and liquid)
C a lc ula t e d
H2 S
pa r t ia l




phase mass transfer coefficients,
pre ssure ,
Pa


 interfacial areas and concentra-

tion difference driving forces.


The details of mass transfer rate
 modelling have been described
at length elsewhere5,6 and will
 not be repeated here. Suffice it
to say, the kinetics of the NH3-

   
 
 

 

 CO2 reaction (see Equation 7)
Measured H 2 S partial pressure, Pa have been properly accounted
for using known kinetic parame-


 ters and the usual enhancement
factor approach. The resulting
process model is a virtual plant
C a lc ula t e d
C O2 pa r t ia l




on a computer in which all the
pre ssure ,
Pa

fundamental physical and chem-



ical processes occurring are
properly accounted for. The


computer simulation is a direct
one-for-one prediction of both

column and process perform-
ance completely without
 recourse to user estimates of
  
 
 

 


Measured CO 2 partial pressure, Pa any parameters whatsoever. All
that is required is the actual
Figure 2 Parity plots of VLE data for the system NH3-H2S-CO2-H2O process flow sheet configuration,
tower internals details and
Deshmukh-Mather approach10 for activity coeffi- values for process parameters corresponding to
cients, and the Peng-Robinson equation of state such operating conditions as sour water flow and
for the vapour. The thermodynamic model has composition, tower pressures and imposed heat
been fitted to all public-domain-phase equilib- loads.
rium data for the CO2-H2S-NH3-H2O system, a The sour water stripper model itself has been
total of some 550 separate experiments, most validated against at least a half dozen sets of
involving ammonia and either one or both of the stripper performance data by two refining
acid gases H2S and CO2. Figure 2 is a set of companies and an engineering firm, with what

4 PTQ Q3 2012 [Link]/article/1000556

were reported by them as excellent results. Using when combined with the averaging that is done
a virtual plant in the form of a precise computer through overall efficiencies, and the possible use
model, the remainder of this article examines the of thermal efficiencies to force-match measured
effect of the steam-to-water ratio and stripper temperature profiles, they completely obscure
operating pressure on sour water stripper what is really going on in SWSs. The interesting
performance, as measured by such parameters behaviours that occur in sour water strippers are
as Murphree vapour efficiency and residual caused by mass transfer effects; ideal stage
ammonia and H2S in the stripped water. The approximations ignore tower internals as mass
simulation cases use live steam injection as the transfer devices and smooth out most of the
source of stripping vapour because this tends to physics of the real mass transfer device by
be the most effective way of providing thermal considering only vapour-liquid equilibrium. The
energy for stripping; however, reboiled strippers missing physics cannot be recovered by an over-
add less make-up water and can be simulated all tray efficiency for the column. Consequently,
just as easily. an ideal stage stripper model is, at best, only as
good as its guessed efficiencies. The Murphree
Exploring strippers vapour efficiency, on the other hand, provides a
Figure 1 shows the simplest possible configura- quantitative measure of how close to perfect
tion of a sour water stripper. For this exercise, equilibrium each real tray actually operates.
the stripper contained 40 one-pass valve trays Unfortunately, a mass transfer rate model is
on 2ft spacing with 2in weirs. Sour water was fed needed to predict the Murphree efficiencies, so
to Tray 6 (from the top), live 50 psig saturated one might as well use a mass transfer model in
steam entered below the bottom tray and, in all the first place and avoid efficiencies altogether.
cases, the column was sized for 70% of jet and The real value of the Murphree vapour efficiency
downcomer flood. Except for the case study is to show how far from equilibrium real trays
involving stripper pressure, where a fixed perform and how radically performance can vary
temperature approach of 20F was used, the from tray to tray. It is worth emphasising that
sour water feed to the stripper was kept at 235F. there is absolutely no relationship between
Table 1 gives the conditions of the sour water Murphree vapour and overall tray efficiencies.
used for this case study. As can be seen from Figure 3, H2S efficiency
varies enormously throughout the tower and is
Ratio of steam to sour water also a very sensitive function of the steam-to-
Four steam rates were used 0.78, 0.92, 1.20 sour-water ratio, especially when steam rates are
and 1.48 lb/USgal covering the range normally relatively high. Generalising the Murphree
used in sour water stripping. (Gantz7 shows vapour efficiency for H2S is utterly impossible
performance data for steam rates from 0.6 to 4.0 because it varies from a few per cent to 50%.
lb/USgal, although, commercially, rates at and Trays become extremely inefficient when the last
above 1 lb/USgal are more usual.) One of the vestiges of H2S are being removed from the
performance parameters of interest is the water. Obviously, the tray count and the steam
computed Murphree (1925) vapour-phase effi- required depend heavily on the treated water
ciency.8 Taylor and Kooijman9 define the specifications to be met. Thus, the designer is
Murphree vapour efficiency, Emwi,n , for component
i on tray n (from the top) as: Sour water feed conditions

MV yi,n - yi,n + 1 (8)

Ei,n = y - y Total flow, lb/h 150 000
i,n i,n + 1
Temperature, F 235
Pressure, psia 70
where y is the mole fraction Water, mol% 96.4
It should be noted that this is not an overall Hydrogen sulphide, mol% 1.5
Carbon dioxide, mol% 0.1
tray efficiency. Tray overall efficiencies are based Ammonia, mol% 2.0
on ideal stage calculations to determine the
number of theoretical (ideal) stages. ProTreat
studiously avoids such calculations because, Table 1

[Link]/article/1000556 PTQ Q3 2012 5

 ages of ammonia. The surprising
thing is the bulge, or maximum,
0
displayed at boil-up rates at the
5 Feed tray low end of the range.
10 We have observed and
reported the existence of similar
Tray from top

15
maxima in ammonia concentra-
20 tion in the context of ammonia
control in refinery amine units.11
25
0.78 Indeed, it was found that in an
30 0.92 amine regenerator only the
35 1.20 bottom few trays were effective
1.48 in stripping ammonia. In the
40 present case of a sour water
0 10 20 30 40 50 60
stripper, ammonia stripped in
H2S efficiency, %
the lower part of the column is
actually reabsorbed in the upper
Figure 3 Calculated Murphree vapour efficiencies for H2S part because the H2S is high
enough there to react with and
fix a significant part of the
0 ammonia back into the liquid.
5 Feed tray When the steam-to-sour water
ratio is high enough, however,
10
the H2S concentration is too low
Tray from top

15 everywhere to remove enough

20
ammonia to result in a maxi-
mum in the ammonia
25 concentration. We note in pass-
0.78
30 0.92
ing that there is a strictly
1.20
monotonic decrease in H2S
35
1.48 concentration as the sour water
40 moves down the column.
0 5 10 15 20 25 30 When the boil-up rate is at the
NH3 in vapour, mol% lower end of the spectrum and a
maximum in ammonia concen-
Figure 4 Ammonia concentration profiles in the SWS tration occurs away from the
ends of the stripper, the
faced with a difficult decision, because more Murphree vapour efficiency for ammonia exhib-
trays and more stripping energy lead to lower its seemingly odd behaviour. Figure 5 shows very
efficiency. These effects cannot be modelled or surprising efficiency trends from tray to tray at
accounted for with an ideal stage model, even stripping steam rates of 0.78 and 0.92 lb steam/
when a guessed overall tray efficiency is used. USgal. At the lowest steam rate, the apparent
Providing optimal, cost-effective designs using efficiency ranges from -100% to +120%, and at
theoretical stages is a rather seat-of-the-pants 0.92 lb/USgal the apparent efficiency reaches
enterprise. +200%. The reason is that the lines correspond-
Before tackling corresponding efficiencies for ing to actual and equilibrium vapour-phase mole
ammonia, it is revealing to look at the vapour- fractions cross (as they must do because a bulge
and liquid-phase concentration profiles of implies that stripping gives way to absorption at
ammonia, as well as the simulated vapour profile some point in the column) and they have a maxi-
compared directly with the equilibrium vapour mum on immediately adjacent trays rather than
composition. Figure 4 shows tray-by-tray profiles on the same tray. Figure 6 shows the actual (y)
of the liquid- and vapour-phase mole percent- and equilibrium (y*) lines for the steam rate of

6 PTQ Q3 2012 [Link]/article/1000556

0.92 lb/USgal, from which it can
be seen that both exhibit a maxi- 0
mum, the lines cross, and they
5 Feed tray
have peak values on different

Tray number from top

trays. Referring to Equation 8 10
(definition of efficiency), it is 15
easy to see that if the numerator
or denominator changes sign, or 20
the concentration difference in 25
the denominator becomes very
30
small, apparent efficiencies can
become not only large but 35
0.78 lb steam / US gal.
negative. 40
At first glance, much of this 100 50 0 50 100 150
odd efficiency behaviour may NH3 efficiency
seem a little academic; however,
when a constant efficiency is
0
applied to an equilibrium-stage
model, the maximum in ammo- 5 Feed tray
nia concentration in the interior 10
of the column, well away from
Tray from top

the ends, will not be revealed 15

and simulation results will tend 20
to be quite optimistic in terms of
25
sour water quality and stripped
gas ammonia levels. This may go 30
some way to explaining the wide 35
variations reported in tray effi- 0.92 lb steam / US gal.
ciencies (15% to 45%) in this 40
5 20 40 60 80 100
application. It is worth noting
NH3 efficiency
that when stripping steam rates
are moderate to high, the
Murphree efficiency turns out to Figure 5 Murphree vapour efficiencies for ammonia at low steam rates
be between 35% and 40%, with
variations of only one or two percentage points negative (albeit relatively small) effect on water
across the whole column. quality with respect to ammonia, but can cause
an 80-fold increase in residual H2S content of
Effect of stripper pressure the stripped water under these conditions.
From time to time, one hears it stated that higher Higher pressures do not seem to favour ammo-
stripper pressures favour sour water stripping, so nia stripping, and they certainly have a negative
we decided to put this old wives tale to the test. effect on stripped water quality vis--vis H2S.
Table 2 shows the effect of column head pressure
in residual ammonia and H2S levels in stripped Summary
water for the same process configuration and Perhaps the most important message of this arti-
sour water composition shown in Figure 1 and cle is that genuine mass transfer rate-based
Table 1, respectively. The simulated performance modelling allows the construction of a virtual
corresponds to a steam rate of 25 000 lb/h of 50 plant on a computer. There is no reliance on
psig steam (1.40 lb/USgal) with the cross what might be termed fudge factors to achieve
exchanger set to a temperature approach of 20F agreement between calculations and reality.
to minimise the effect of temperature differences A mass transfer rate-based sour water stripper
across the stripping section. The table shows model provides a virtual sour water stripper on a
that, contrary to anecdote, higher pressure has a computer. Engineers can now design sour water

[Link]/article/1000556 PTQ Q3 2012 7

 directly to sour water stripping:
5 H2S efficiency varies enor-
Feed tray
mously from tray to tray
10 Maximum at y(12) throughout a sour water strip-
Tray number from top
Maximum at y*(13) per, and the Murphree
15
efficiency is a very sensitive
20 function of the steam-to-sour-
water flow rate ratio. Any
25
attempt to generalise H2S
30 component efficiency is utterly
futile because of its huge varia-
35 y (n) tion from only a few per cent to
y* (n)
50% across a column and its
40
0 5 10 15 20 25 30 dependency on the particular
Ammonia in vapour, mole% operating conditions of each
unique situation
Figure 6 Actual and equilibrium ammonia concentrations in the vapour With respect to efficiencies
phase at low steam rates pertinent to H2S stripping, the
tray count and the steam
Effect of stripper pressure on residual NH3 and H2S levels in stripped water required depend heavily on the
treated water specifications to
Stripper head pressure, psig Treated water be met. Thus, the designer is
Ammonia, ppmw H2S, ppmw faced with really quite a difficult
10 17.2 0.00002 decision more trays and more
15 21.1 0.00013
20 26.0 0.00058 stripping energy lead to lower
25 31.5 0.00197 efficiency. What and where is
the balance? Should one use
Table 2 more or less steam, and more or
fewer trays? These effects cannot
stripper units with unprecedented accuracy and be accounted for with an ideal stage model, even
reliability. Plant operations personnel can answer when it uses efficiency. The process appears to
a wide range of what-if questions to troubleshoot be too complex to generalise into rules of thumb
operations and to optimise existing units with At stripping steam rates in the lower half of
complete confidence in the results. the normal range used in practice, it is quite
After overall tray efficiencies, Murphree vapour possible for the ammonia concentration in a sour
efficiencies are probably the most commonly water stripper to show an enormous bulge or
used type of efficiency for tray calculations. maximum in some part of the stripping section
Overall efficiencies are an invention of the 1920s, well away from the ends. At this bulge, the sour
developed to allow ideal stage distillation calcu- water stripper changes from acting as an ammo-
lations to approximate reality more closely, and nia stripper to an ammonia absorber. When (and
as 90-year-old technology they have served the where) the bulge occurs, there is a high enough
distillation business well. However, the whole H2S concentration in the liquid to attract ammo-
equilibrium-stage approach was superseded by nia into the water phase and hold it there
mass transfer rate calculation methods in the chemically. Again, ideal stage models even with
mid- to late 1980s, already more than 25 years efficiency do not show this kind of behaviour,
ago. The gas treating industry deals with far and they can lead to either gross overdesign or
more complex separations processes than distil- to a complete mis-design, depending on the rela-
lation, but, unfortunately, it has been remarkably tive H2S and ammonia concentrations in the sour
slow to catch on to the new technology. Genuine water feed
mass transfer rate-based simulation is an Murphree vapour efficiencies for ammonia in
extremely powerful tool in gas treating. sour water strippers operating at low to moder-
There are several other messages that pertain ate stripping steam rates can vary from large

8 PTQ Q3 2012 [Link]/article/1000556

negative values to positive values well over 100%. rate simulation, Laurence Reid Gas Conditioning Conference,
At high stream rates, on the other hand, Norman, OK, 2001.
Murphree vapour efficiencies for ammonia strip- 7 Gantz R G, Sour water stripper operations, API Special Report,
ping are typically 35% to 40%. This huge Hydrocarbon Processing, May 1975, 83.
variation depending on steam rate and H2S-to- 8 Murphree E V, Graphical rectifying column calculations, Ind.
Eng. Chem., 17, 1925, 747-750 and 960-964.
NH3 ratio may go a long way towards explaining
9 Taylor R, Kooijman H, ChemSep Tutorial: Efficiencies, www.
the wide range of 15% to 45% overall efficiency
[Link]/downloads/docs/ChemSepTutorial_Efficiencies.
often quoted by practitioners pdf, downloaded 4 Jan 2012.
Contrary to a commonly repeated anecdote, 10 Deshmukh R D, Mather A E, A Mathematical model for
higher stripper pressures do not favour better equilibrium solubility of hydrogen sulphide and carbon dioxide
ammonia stripping, and they certainly do not in aqueous alkanolamine solutions, Chem. Eng. Sci., 36, 1981,
favour H2S stripping where a factor of 80 loss in 355-362.
performance was predicted, even for the setup in 11 Hatcher N A, Weiland R H, Special treatment, Hydrocarbon
this study. Engineering, Jan 2012.
In conclusion, this article has pointed out
several aspects of sour water stripping that either
Ralph Weiland began working in gas treating as a graduate
do not seem to have been recognised heretofore,
student in 1965 and has been an active researcher in this field
or that offer solid scientific explanations for previ-
ever since. He received BASc, MASc and PhD degrees in chemical
ously observed and reported characteristics. engineering from the University of Toronto. He taught chemical
engineering for 30 years at universities in Canada, Australia and
ProTreat is a registered mark of Optimized Gas Treating Inc. the US, and spent 10 years doing mass transfer tray development
with Koch-Glitsch LP in Dallas, TX. He established Optimized Gas
References Treating in 1992. Email: [Link]@[Link]
1 Stevens D Mosher A, Ogg D, Fundamentals of sour water Nathan Hatcher joined Optimized Gas Treating, Buda, Texas,
stripping, Brimstone Sulfur Recovery Symposium, Vail, CO, Sept as Vice-President, Technology Development, in 2009. He has
2007. spent almost his entire career in amine treating and sulphur
2 Hauser R, Kirkey R T, Refinery tests demonstrate fixed valve recovery, and has extensive experience in the practical
trays improve performance in sour water stripper, AIChE Spring application of process simulation to operations, troubleshooting
National Meeting, 31 Mar 2003, New Orleans, LA T1-2E. and training. He holds a BS in chemical engineering from the
3 Quinlan M P, Hati A A, Processing NH3 acid gas in a sulphur University of Kansas and is currently a member of the Amine
recovery unit, Proceedings of the 2010 Laurance Reid Gas Best Practices Group. Email: [Link]@[Link]
Conditioning Conference, Norman, OK.
4 Asquith J, Moore A, sour water processing balancing needs,
Proceedings of the 2000 Brimstone Sulphur Recovery Symposium,
Links
Vail, CO.
5 Weiland R H, Sivasubramanian M S, Dingman J C, Effective More articles from: Optimized Gas Treating
amine technology: controlling selectivity, increasing slip and
reducing sulphur, Laurence Reid Gas Conditioning Conference, More articles from the following category:
Norman, OK, 2003. Water/Wastewater Management
6 Weiland R H, Dingman J C, Column design using mass transfer

[Link]/article/1000556 PTQ Q3 2012 9