Journal of Natural Gas Science and Engineering 36 (2016) 175e183

Contents lists available at ScienceDirect

Journal of Natural Gas Science and Engineering

journal homepage: www.elsevier.com/locate/jngse

Sensitivity analyses and optimization of a gas sweetening

plant for hydrogen sulde and carbon dioxide capture using
methyldiethanolamine solutions
Majeed S. Jassim*
University of Bahrain, College of Engineering, Department of Chemical Engineering, P.O. Box 32038, Bahrain

a r t i c l e i n f o a b s t r a c t

Article history: The selective removal of hydrogen sulde (H2S) in the presence of carbon dioxide using Methyl-
Received 10 August 2016 diethanolamine (MDEA) solutions is a signicant process in the petroleum and natural gas industries. An
Received in revised form industrial plant with an energy consumption of 3.36 GJ/mol CO2 was simulated and validated using
18 September 2016
rigorous kinetic and equilibrium models for H2S-CO2-MDEA-H2O system that are incorporated into the
Accepted 9 October 2016
Amine Package in Aspen HYSYS. A modied Murphee-type vapor tray efciency of 15% for CO2 and 100%
Available online 13 October 2016
for H2S was utilized to account for the non-ideal behavior of the stages. Sensitivity analyses were per-
formed on circulation rate (L/G 0.6e2 kg/kg), stripper pressure (50e220 kPa), lean amine temperature
Keywords:
Gas sweetening
(45e65  C), number of absorber stages (8e14), and concentration of MDEA (15e50 wt%). These process
Hydrogen sulde variables were evaluated with respect to percentage recovery of sour gaseous components and
Carbon dioxide normalized reboiler duty. The sensitivity analyses results showed that the circulation rate and the MDEA
Aspen HYSYS concentration are the two main factors improving process performance. The optimized scenarios
Simulation exhibited better selective H2S absorption and higher energy savings.
Absorption 2016 Elsevier B.V. All rights reserved.

1. Introduction requirements, low heat of reaction, low corrosivity, high capacity,

excellent stability, and can be utilized upto 60 wt% without
The removal of carbon dioxide and hydrogen sulde from sour appreciable evaporation losses (Kohl and Nielsen, 1997). However,
gas streams using aqueous alkanolamine reactive solutions is an the shortcomings of this solvent are its slow reaction with CO2,
essential part of natural gas sweetening industries (Kidnay et al., tendency to foam at high concentrations, and its higher cost (Bullin
2011). This absorption/stripping purication step is important et al., 1990).
because the contaminants exceedence of permitted safe levels can The advances in commercial simulators assisted researchers and
cause higher corrosion rates, hazardous health impacts, and dete- process engineers to extensively study existing amine plants or
rioration to environment. introduce innovative owsheets with emphasis on improving the
The primary, secondary and tertiary alkanolamines are the most facilities economics and performance (Freguia and Rochelle, 2003)
common chemical solvents in the gas sweetening industry (Kohl (Jassim and Rochelle, 2006) (Abu-Zahra et al., 2007) (Rodrguez
and Nielsen, 1997). Methyldiethanolamine (MDEA), as a tertiary et al., 2011) (Gao et al., 2014). The simulation challenges in reac-
alkanolamine, is developed to reduce the operating costs when the tive H2S-CO2-amine-H2O system is mainly attributed to complex-
objective of the gas sweetening plant is to remove H2S selectively to ities associated with accurate prediction of acid gas solubility,
very low concentrations in the presence of CO2. This is possible kinetics of reaction rate, and mass transfer properties. Another
because the reaction rate of MDEA with H2S is high whereas it is complexity is that every amine system in the eld of natural gas
low with CO2 (Huttenhuis et al., 2009). Furthermore, MDEA has processing represents an individual case rather than a general
favorable advantages such as low vapor pressure, low energy case because of different conditions and requirements (Lunsford
and Bullin, 1996).
Aspen HYSYS is one of the most suitable software platforms for
* Corresponding author. simulation and optimization of complex gas sweetening plants
E-mail address: drmajeed@gmail.com. because it possesses rigorous rate based capabilities (Abdulrahman

http://dx.doi.org/10.1016/j.jngse.2016.10.012
1875-5100/ 2016 Elsevier B.V. All rights reserved.
176 M.S. Jassim / Journal of Natural Gas Science and Engineering 36 (2016) 175e183

and Sebastine, 2013). simulated a prospective gas sweetening plant There are two kinetic reactions: hydration of CO2 and formation
in Kurdistan-Iraq with high concentrations of H2S and CO2 in nat- of bicarbonate (Versteeg and Van Swaaij, 1988):
ural gas feedstock using different strengths of mono-ethanolamine
(MEA), di-ethanolamine (DEA), MDEA, MEA-MDEA blend, and CO2 H2 O/H HCO
3 (9)
DEA-MDEA blend. It was observed that 40 wt% MDEA and 10 wt%
DEA blend exhibited highest H2S removal. The study showed that
CO2 OH /HCO
3 (10)
the temperature of rich amine was decreasing with higher circu-
lation rate. However, similar studies using Aspen HYSYS do not The MDEA acts as a base catalyst for the hydrolysis of CO2 to
explain the interaction between important design variables for the bicarbonate (Ko and Li, 2000):
base case conguration. In addition, researchers overlook valida-
tion of simulated base case (i, 2007) (Aliabad and Mirzaei, 2009) CO2 H2 O MDEA/MDEAH HCO
3 (11)
(Abdulrahman and Sebastine, 2013). Thus, it is difcult to justify the
accuracy of Aspen HYSYS predictions without conrming the The thermodynamic model based on vapor-liquid equilibria
reliability of the chosen thermodynamic model and property data. (VLE) and liquid phase chemical equilibria of H2S-CO2-MDEA-H2O
The aim of this parametric simulation work is to assess electrolyte system is incorporated in amine package in Aspen
comprehensively an existing MDEA industrial process plant using HYSYS. The amine package includes both the equilibrium and the
Aspen HYSYS. The non-ideal behavior complexities of the kinetics parameters that intrinsically account for the mass transfer
absorber stages inclusive of mass transfer diffusion and reaction rate in a non-equilibrium stage model. The Aspen HYSYS software
kinetics are accounted for using modied Murphee-type vapor tray allows utilization of Li-Mather thermodynamic framework to
efciency. The research performs sensitivity analyses of key oper- accurately predict VLE and heat of mixing for MDEA over a con-
ational process variables for a validated base case: circulation rate, centration range of 0-50 wt%, temperature range of 25e126.6  C,
stripper pressure, lean amine temperature, MDEA concentration, and acid gas partial pressure upto 2068 kPa (Aspen HYSYS, 2004).
and number of absorber stages. The results of the sensitivity ana- A modied Murphee-type vapor tray efciency accounts for the
lyses will assist in proposing optimized scenario to capture sour gas non-ideal behavior of stages for CO2 and H2S. The tray efciency is a
components in gas purication processes. The evaluation criteria function of mass transfer coefcient and tray geometry such as weir
are: normalized reboiler duty, CO2 recovery, H2S recovery and lean/ height, weir length and tray diameter. The simulation software
rich loadings. allows the user to specify tray efciency.

3. Flowsheet description
2. Thermodynamic model

The owsheet of the investigated MDEA amine plant is shown in

There are eight equilibrium reactions (Reactions 1e8) and three
Fig. 1. A ue gas of 3196 kg mol/h with 5.35 mol% CO2 and 0.05 mol%
kinetics reactions in the H2S-CO2-MDEA-H2O electrolyte system
H2S is fed to the bottom of absorber at 4680 kPa and 50  C. The mole
(Reactions 9e11). The liquid phase reactions of aqueous MDEA
ratio of CO2 to H2S is 107 thus selective absorption is possible using
solution with hydrogen sulde are equilibrium-based and kinetics-
aqueous MDEA solution (Kohl and Nielsen, 1997). The 23.6 wt%
based with carbon dioxide. The tertiary amine is instantaneously
MDEA lean solvent with 2846 kg mol/h and a lean loading of 0.0404
reacting with hydrogen sulde but with a slower and nite rate
molCO2/mol MDEA is fed from top of the absorber at 4620 kPa and
with carbon dioxide and it is considered pseudo-rst-order reac-
50  C. The concentration of H2S in the sweet gas is 10 ppm and the
tion. The reaction kinetics is the reason for MDEA solution to
concentration of CO2 is 2.75 mol% thus indicating selective nature
selectively remove hydrogen sulde in the presence of carbon di-
of MDEA solution in capturing sour components. The characteris-
oxide (Astarita et al., 1983).
tics of the ue gas and the lean solvent is shown in Table 1. The rich
Solubility of H2S and CO2
solution loaded with CO2 and H2S is routed from the bottom of the
H2 Sg H2 Saq (1) absorber to the third stage from top of the stripper to capture any
vaporized MDEA. The base-case reboiler duty for the regeneration
of MDEA solvent is 3.6 MW. The hot lean solution recovered from
CO2g CO2aq (2)
the bottom of the stripper is continuously recycled to the top of the
Dissociation of water absorber after heat integration with the rich solution. The approach
temperature in the cross heat exchanger is 10  C. The lean solution
2 H2 O H3 O OH (3) is further cooled to 50  C before entering to the top of the absorber.
The absorber and the stripper are modeled as tray towers and their
Dissociation of hydrogen sulde
detailed design characteristics are shown in Table 2.
H2 O H2 S H3 O HS (4)
4. Development of base-case HYSYS simulation of
Dissociation of bisulde ion conventional amine plant

H2 O HS H3 O S2 (5)

The simplied assumptions of the base-case simulation are
Protonation of amine as follows: no degradation of MDEA, no corrosion of MDEA, negli-
gible vaporization of MDEA, and the liquid-gas interface is in
MDEA H2 O OH MDEA (6) equilibrium.
The specications of the base-case simulation are: condenser
H2 S MDEA MDEA HS (7) temperature, carbon dioxide recovery in tail gas, temperature and
pressure of lean amine exit cooler, and temperature of rich amine
Dissociation of bicarbonate ion routed to stripper. The approach temperature is kept constant by
varying the temperature of the rich amine routed to the stripper
HCO 2
3 OH H2 O CO3 (8) stream.
 M.S. Jassim / Journal of Natural Gas Science and Engineering 36 (2016) 175e183 177

Fig. 1. Conguration of MDEA amine plant.

The HYSYS's programmer should suggest two active column respectively in the sweet gas stream indicating possible limitations
specications to successfully converge the stripper with minimum of Amine package thermodynamics in Aspen HYSYS. The simulated
errors in mass and energy balances. The temperature of the tail gas owrate is under-predicted due to leakage of CO2 in sweet
condenser (50  C) and H2S recovery are xed because they repre- gas stream. The energy balance deviation is well matched as the
sent existing plant acid gas specications. simulation under-predicts reboiler duty by 10.1%. The magnitude of
The mass balance of the existing plant is closed with minimum these mismatches is consistent with typical error values reported in
error by manipulating tray efciencies for sour components. The the literature using Aspen Plus which incorporates extensive
Aspen HYSYS manual conrms this approach by stressing on using correlations (Freguia and Rochelle, 2003) (Bolhar-Nordenkampf
non-unity stage efciencies to account for kinetics and mass et al., 2004) (G

asp

ar and Ana-Maria, 2011).
transfer effects (Aspen HYSYS, 2004) (Pacheco and Rochelle, 1998).
reported a value of 1.1e1.2% for CO2 and 84e90% for H2S capture 6. Results and discussions
using MDEA solution in a trayed column. The low value of CO2 ef-
ciency in their work is attributed to the utilization of Aspen This section investigates the effects of key operational param-
Ratefrac based on the Generalized Maxwell-Stefan (GMS) eters on the performance of the absorber and the stripper using
approach to multi-component mass transport (Al-Baghli et al., results of converged Aspen Hysys simulations. The following base
2001). reported a value of 15% for CO2 and a value of 60% for H2S case operational conditions are kept constant (i.e. design variables)
in a DEA solution based on a rate-based mathematical model. (i in the parametric studies to ensure comparison consistency unless
et al., 2014) reported a stage efciency value of 15% for CO2 cap- otherwise stated: MDEA concentration is 23.54 wt%, cross
ture in MEA solution using Aspen HYSYS. In this research work, exchanger approach temperature is 10  C, lean loading is 0.001 mol
the tray efciency was specied 15% for CO2 and 100% for H2S in the CO2/mol MDEA and 0 mol H2S/mol MDEA, number of stages for
absorber and the stripper to close the process mass balance. absorber/stripper is 10/20.

5. Validation of the base-case simulation 6.1. Absorber performance

The base-case simulation is validated using material and energy The tray-by-tray mass transfer performance of the simulated
balances of an existing industrial plant. Table 3 compares the mass base case absorber is monitored using resulting distribution of
balance of the simulation data with the actual industrial plant data
and Table 4 shows the energy balance of the heat utilities. The mass Table 2
Design characteristics of the mass transfer separators.
owrates of CO2 and H2S are under-predicted by 7.3% and 13.2%
Absorber Stripper

Table 1 Number of trays 10 20

Characteristics of the ue gas and the chemical solvent. Inside diameter (m) 1.75 1.4
Tray type Valve Valve
Flue gas composition (mol%) C1: 80.86, C2: 1.63, C3: 0.35, n-C4: 0.18, Tray spacing (m) 0.50 0.55
(dry basis) n-C5: 0.24, H2S: 0.05, CO2: 5.35, N2: 11.34 Weir height (mm) 53 77
Flue gas feed temperature ( C) 50 Weir length (m) 1.6 1.245
Flue gas feed pressure (kPa) 4742 Solvent feed location from top, tray 1 3
Flue gas molar ux (kgmol/hr/m2) 0.37 Bottom pressure, kPa 4680 200
Lean chemical solvent 23.5 wt% MDEA at 51  C Pressure drop, kPa 60 20
178 M.S. Jassim / Journal of Natural Gas Science and Engineering 36 (2016) 175e183

Table 3
Results of mass balance deviation between existing plant data and base-case
simulation.

Process Variable Process Plant Simulation %error

Sweet gas owrate (kg/h) 58,550a 58,845a 0.7

H2S in sweet gas (kg/h) 1.06 (10 ppm) 1.20 (11 ppm) 13.2
CO2 in sweet gas (kg/h) 3750 4024 7.3
Rich amine owrate (kg/hr) 63,539 63,396 0.2
Rich CO2 loading (mol CO2/mol MDEA) e 0.6715 e
Rich H2S loading (mol H2S/mol MDEA) e 0.0131 e
Lean amine owrate (kg/h) 60,120 60,108 0.2
Lean CO2 loading (mol CO2/mol MDEA) e 0.0028 e
Lean H2S loading (mol H2S/mol MDEA) 0 0 0
Tail gas owrate (kg/h) 4020a 3686.6a 8.9
H2S in tail gas (kg/h) 54.4 44.1 18.9
CO2 in tail gas (kg/h) 3864 3501 9.4
a
Dry basis.

temperature (T), CO2 partial pressure (PCO2), H2S partial pressure

(PH2S), and water vapor partial pressure (PH2O) as shown in Fig. 2. It
is evident that the reaction of MDEA solution with sour gas stream
containing H2S and CO2 at the bottom of the absorber lowers their Fig. 2. Composition and temperature proles of base-case absorber for (L/
G 0.9677 kg/kg).
partial pressures. The proles of H2S and CO2 are not similar
because every sour component has different nature of reaction with
the chemical solvent as explained in Section 2.
The exothermic reactions increase the temperature of solvent higher circulation rates (L/G 2) results in higher absorption at the
leading to evaporation of water. Therefore, the partial pressure of bottom of the column thus the maximum temperature is skewed to
water vapor increases at the bottom of the absorber because of the the right.
evaporation. However, it decreases at the top of the absorber The effect of circulation rate on recovery of CO2 and H2S is
because of the condensation by the colder temperature of the shown in Fig. 4. The results show that higher circulation rates en-
recycled lean amine solution. hances the recovery of sour components but with two distin-
The phenomena of reactive absorption, water evaporation and guishable behaviors. The prole of H2S indicates the instantaneous
condensation results in a creation of a temperature bulge. For the reaction nature as shown by the steep curve reaching 99%e100%
base-case, the bulge is located in the middle of the absorber (stage percentage recovery when the L/G mass ratio is less than one.
5) with a value of 77.5  C as shown in Fig. 2. According to (Kvamsdal Conversely, the prole of CO2 shows the effect of pseudo-rst-order
and Rochelle, 2008), there is a maximum temperature bulge near nite reaction kinetics nature with alkanolamine as shown by
the middle of the absorber provided by a critical solvent-to-gas lower percentage recovery. It should also be noted that both CO2
ratio. and H2S are competing simultaneously for free alkanolamine
molecules thus higher circulation rate increases availability of free
MDEA molecules for CO2 capture.
6.2. Effect of variation of circulation rate

The circulation rate is a parameter that inuences both the total

cost and the mass transfer performance of the gas sweetening
process. The total cost is inclusive of investment and operational
cost and it is proportional to circulation rate due to requirements of
larger equipment sizes and utilities cost (Rodrguez et al., 2011). The
absorber performance is affected by the circulation rate because it
provides higher effective gas-liquid interfacial area on the trays and
more free available MDEA molecules though the latter is dependent
on lean solvent loading.
The absorber temperature prole is sensitive to circulation rate
as shown in Fig. 3. The position of the bulge in the middle of the
absorber for the base-case indicates that the absorption of the heat
by the chemical solvent is uniform throughout the column
(Kvamsdal and Rochelle, 2008). The lower circulation rate (L/
G 0.6) results in higher absorption at the top of the column thus
the maximum temperature is skewed to the left. Conversely, the

Table 4
Heat duties comparison between existing plant and base-case simulation.

Utilities Process Plant Simulation %error

(GJ/ton CO2) (GJ/ton CO2)

MDEA Stripper O/H condenser 0.76 0.64 15.7

Fig. 3. Absorber temperature bulges for different L/G ratio for CMDEA 23.54 wt%, 10  C
Stripper Reboiler 3.73 3.35 10.1
approach, lean loading 0.0025 mol CO2/mol MDEA, lean loading 0 mol H2S/mol
Stripper Feed/Bottoms Exchanger 2.79 2.79 0.0
MDEA.
 M.S. Jassim / Journal of Natural Gas Science and Engineering 36 (2016) 175e183 179

Fig. 6. Effect of stripper pressure on reboiler temperature and reboiler duty.

Fig. 4. Effect of circulation rate on recovery of H2S and CO2 for CMDEA 23.54 wt%,
10  C approach, lean loading 0.0025 mol CO2/mol MDEA, lean loading 0 mol H2S/
mol MDEA. pressure on reboiler duty and reboiler temperature. It is evident
that higher stripper pressures increases reboiler temperature and
decreases reboiler duty. The stripper pressure study range
The effect of circulation rate on reboiler duty is shown in Fig. 5. (50e220 kPa) is dictated by the reboiler temperature because the
The results demonstrate that higher circulation rates increase the amine package in Aspen HYSYS allows maximum reboiler tem-
reboiler duty. The magnitude of standardized reboiler duty per unit perature to be restricted to 126.6  C corresponding to stripper
mass of captured H2S is much higher than per unit mass of captured pressure of 220 kPa. Above this temperature, corrosion of material
CO2. This might be attributed to the fact that the mass ratio of CO2 of construction and degradation of amine could occur.
to H2S is 138 in the sour gas stream thus there is higher energy Below pressure of 200 kPa, the reboiler duty increases with
requirement for removal of more diluted component. reduction of stripper pressure. This behavior could be explained
using variation of molar ratios of CO2 or H2S to water vapor across
6.3. Effect of variation of stripper pressure stripper for various stripper pressures as shown in Figs. 7 and 8. The
higher stripper pressures increase condensation of water vapor
The selective stripping of dissolved gases from chemical sol- thus the molar sour components ratios increases. The heat duty of
vents occurs at high temperature and/or low pressure because the reboiler is reduced because of condensation of water vapor as a
K-value of solutes vary exponentially with temperature and consequence of higher pressure of stripper. Above 200 kPa, the
inversely with pressure (Seader et al., 2011). The stripper pressure reboiler duty varies slightly and this behavior corresponds well
is considered an important variable that inuences reduction of with the work of (Freguia and Rochelle, 2003). Also, the results
reboiler duty (Jassim and Rochelle, 2006) (Oyenekan and Rochelle, show that the ratios are highest at the top of the stripper and lowest
2006). Fig. 6 shows the simulation results of the effect of stripper at the bottom with the CO2 ratio higher in magnitude than H2S.

Fig. 5. Effect of circulation rate on reboiler duty for CMDEA 23.54 wt%, 10  C approach,
lean loading 0.0025 mol CO2/mol MDEA, lean loading 0 mol H2S/mol MDEA. Fig. 7. Prole of CO2/H2O across stripper at different operating pressures.
180 M.S. Jassim / Journal of Natural Gas Science and Engineering 36 (2016) 175e183

Fig. 8. Prole of H2S/H2O across stripper at different operating pressures. Fig. 9. Effect of lean amine temperature on H2S and CO2 concentrations in sweet gas
for CMDEA 23.54 wt%, CO2 removal efciency average 45%, H2S removal efciency
average 91%, 10  C approach, lean loading 0.0025 mol CO2/mol MDEA, lean
loading 0 mol H2S/mol MDEA.
6.4. Effect of lean amine temperature

The temperature of lean amine inuences both the reaction gas-liquid phases (Bolhar-Nordenkampf et al., 2004). The variation
kinetics and transport properties of H2S-CO2-MDEA-H2O system of the number of trays with circulation rate and rich loading is
(Pacheco and Rochelle, 1998). showed that the resistance to mass shown in Fig. 11. The higher number of trays implies greater
transfer for this system is liquid-lm controlled thus the study of interphase interfacial area and longer residence time subsequently
this operational parameter could be helpful in enhancing recovery leading to reducing circulation rate of chemical solvent and
of sour components and saving energy. It was observed in the increasing rich loading (mol CO2/mol MDEA) at constant efciency
simulation runs with higher lean amine temperatures that water of sour gases removal. This behavior is advantageous because of the
vapor is leaking in sweet gas and tail gas streams. Therefore, the reduction of operational cost but the capital cost will increase
amount of makeup water is adjusted to close the material balance. because of higher height of absorber.
The ue gas temperature in the base case is 50  C and that is the The effect of increasing the number of stages on reboiler duty is
adiabatic saturation temperature. The variation of lean amine shown in Fig. 12. The simulation results show that as the number of
temperature requires adjusting ue gas temperature using cooling stages increases from 8 to 14 stages, the reboiler duty normalized
water (Oyenekan and Rochelle, 2006). In this work, the lean amine per ton CO2 decreases by 1.05% and the reboiler duty normalized
temperature was kept higher than ue gas temperature by upto per ton H2S increases by 6.1%. The higher number of trays reduces
2  C (Kidnay et al., 2011). the circulation rate and increases rich loading thus it requires less
The effect of lean amine temperature on concentration of H2S energy to reverse the reaction in the stripper. The behavior for H2S
and CO2 in sweet gas is shown in Fig. 9. The higher lean amine is in agreement with (Mandal et al., 2004) who reported that the
temperatures leads to increase of partial pressures of H2S and CO2.
This behavior could be explained by the fact that higher tempera-
tures decreases H2S and CO2 solubility in alkanolamine solution
and hence lowering the removal efciency (Huttenhuis et al., 2007)
(Huttenhuis et al., 2009). The magnitude of H2S concentration in
sweet gas is increased by 295% whereas the CO2 concentration is
increased by 13% with higher temperatures. This could be attrib-
uted to inuence of temperature on reaction rate constant as the
reversibility of H2S-MDEA reaction is more evident than CO2-MDEA
reaction (Mandal et al., 2004).
The effect of lean amine temperature on circulation rate and
reboiler duty is shown in Fig. 10. It is evident that the increase of
lean amine temperature by 15  C leads to 2.6% increase in reboiler
duty and 0.8% decrease in circulation rate. These results are
consistent and in agreement with (Seader et al., 2011) in which it
was highlighted that the absorber is efciently operated at lower
temperatures in order to minimize stage requirements and circu-
lation rate.

6.5. Effect of number of absorber stages

Fig. 10. Effect of lean amine temperature on reboiler duty and lean amine owrate for
The number of trays in absorber affects the degree of separation CMDEA 23.54 wt%, CO2 removal efciency average 45%, 10  C approach, lean
because it inuences the species mass and energy transfer rates in loading 0.0025 mol CO2/mol MDEA, lean loading 0 mol H2S/mol MDEA.
 M.S. Jassim / Journal of Natural Gas Science and Engineering 36 (2016) 175e183 181

Fig. 11. Effect of absorber stages on circulation rate and rich loading for Fig. 13. The prole of H2S and CO2 concentration in the sweet gas stream in the
CMDEA 23.54 wt%, CO2 removal efciency average 45%, 10  C approach, lean absorber for CMDEA 23.54 wt%, CO2 removal efciency average 45%, 10  C approach,
loading 0.0028 mol CO2/mol MDEA, lean loading 0 mol H2S/mol MDEA. lean loading 0.0028 mol CO2/mol MDEA, lean loading 0 mol H2S/mol MDEA.

rate of absorption of H2S is decreasing with increasing contact time H2S recovery is steep while the slope of CO2 recovery is gradual. The
for MDEA solutions. amine strength above 25 wt% has no signicant impact on H2S
The increase in number of trays leads to higher concentration of recovery. The mechanism of the reaction between MDEA and sour
H2S and CO2 in sweet stream thus the absorber performance is gases explains such behavior as it is instantaneous with H2S (re-
deteriorated as shown in Fig. 13. Apparently, the inuence of action 7) and fast with CO2 (reaction 11). The MDEA solution con-
reduction of solvent rate circulation outweigh the increase of centration between 30 and 40 wt% could be considered optimum
number of trays. because the percentage of H2S recovery is 99.92e100% and the CO2
recovery is 51e55% respectively. The corresponding lean loadings
6.6. Effect of MDEA concentration are 0.002e0.0013 mol CO2/mol MDEA and the rich loadings are
0.5778e0.4671 mol CO2/mol MDEA.
The range of MDEA solution strength is 20e50 wt% and the acid Fig. 15 shows the effect of MDEA strength on reboiler duty and
gas loadings upto 0.7e0.8 mole/mole are practical in carbon steel lean/rich solvent loadings. The results show that the reboiler duty
equipment without serious corrosion challenges (Polasek and and lean/rich loadings are decreasing with higher MDEA concen-
Bullin, 1984). The Aspen HYSYS software acknowledges reliable trations. The reason for such behavior is attributed to reduction of
results upto 50 wt% using incorporated kinetic models thus CO2 loading in the rich solution thus the reboiler requires less
allowing the study of mass transfer performance of absorber and thermal energy to desorb sour gases from rich solvent in the
thermal energy requirement of reboiler (Aspen HYSYS, 2004). stripper.
Fig. 14 shows that the absorber performance is increasing sub- (Freguia and Rochelle, 2003) analyses of heat sinks in the
stantially with higher MDEA concentrations. For amine strength stripper for MEA solution capturing CO2 could be extended to
range between 15 and 25 wt%, the results show that the slope of

Fig. 12. Effect of absorber stages on reboiler duty for CMDEA 23.54 wt%, CO2 removal
efciency average 45%, 10  C approach, lean loading 0.0028 mol CO2/mol MDEA, Fig. 14. Effect of amine concentration on recovery of H2S and CO2 in absorber for 10  C
lean loading 0 mol H2S/mol MDEA. approach temperature and L/G 0.97.
182 M.S. Jassim / Journal of Natural Gas Science and Engineering 36 (2016) 175e183

Table 5
Comparison of base-case with two proposed optimized scenarios.

Process Variables Base-Case Scenario 1 Scenario 2

L/G (kg/kg) 0.97 0.83 0.69

Stripper pressure (kPa) 200 200 200
Lean amine temperature ( C) 51.52 51.52 51.52
Approach temperature ( C) 10 10 10
Number of absorber stages 10 10 10
MDEA concentration (wt%) 23.6 30 40
CO2 recovery (%) 46.5 46.5 46.7
H2S recovery (%) 97.8 98.8 91.0
Lean amine loading (mol CO2/mol MDEA) 0.0028 0.0020 0.0022
Rich amine loading (mol CO2/mol MDEA) 0.6715 0.6125 0.5508
Reboiler duty (GJ/mol CO2) 3.36 3.07 2.76

The results of the sensitivity analyses presented in Section 6.6

shows that MDEA concentration is the most important process
variable in terms of energy requirement and sour gases capture.
Therefore, its effect is further investigated with 30 and 40 wt% in
scenarios 1 and 2. The latter concentration corresponds to 100%
Fig. 15. Effect of amine concentration on reboiler duty, rich and lean amine loading for capture of H2S and that is the objective of amine plant. The aim of
10  C approach temperature and L/G 0.97. the MDEA plant is to selectively remove H2S in conjunction with
minimum slippage of CO2 thus the latter recovery is kept constant.
For the suggested scenarios, the circulation rate has to be adjusted
MDEA solution capturing simultaneously CO2 and H2S. The accordingly to accommodate higher MDEA concentrations there-
condensing steam in the reboiler provides four necessary heat sinks fore the L/G ratio is lowered to equate CO2 recovery. The cross
in the stripper: heat of desorption for CO2, heat of desorption for exchanger approach temperature is kept constant to ensure com-
H2S, heat of vaporization of water and sensible heat to bring parison consistency.
chemical solvent to reboiler temperature. These terms are shown in The results of the rst scenario in comparison to the base-case
Equation (12). shows that the reboiler duty is reduced by 8.6%, the circulation
rate is lowered by 14.4%, lean amine loading is reduced by 28.6%,
Qr  nCO2 DHabs;CO2  nH2S DHabs;H2S V  nH2O DHvap;H2O and H2S capture is improved by 1.02%. The results of the second


L Cp Tbottom  Ttop scenario shows that the reboiler duty is reduced by 17.6%, the cir-
(12) culation rate is lowered by 28.8%, lean amine loading is reduced by
21.4%, and H2S capture is reduced by 6.9%. The deterioration in H2S
The number of CO2 and H2S moles desorbed in the stripper are capture is attributed to lower circulation rate and higher concen-
higher with more concentrated chemical solvent as shown in tration of the lean solvent (Jassim et al., 2007). reported that CO2 is
Fig. 14. Their heat of desorption terms are negative in Equation (12) exhibiting accelerated absorption kinetics with higher alkanol-
thus the reboiler duty is consequently reduced. The concentration amine concentrations. This behavior causes competition with H2S
of chemical solvent is inversely proportional to circulation rate. The for free amine thus reducing the absorption rate of H2S.
sensible heat term is dependent on circulation rate thus higher The temperature bulges of the three scenarios in Table 5 are
concentrated solutions requires lower reboiler duty. shown in Fig. 16. The increase in concentration leads to higher

6.7. Optimization of the base case

The one-at-a-time sensitivity analyses of the base-case

improved our understanding of the behavior of amine plant and
signicance of the investigated process variables. This section will
discuss two scenarios that would contribute towards improving
mass transfer performance and reducing energy consumption
based on selected inuential parameters.
The simulation results of the variation of the stripper pressure
showed that it has an optimum at 200 kPa with respect to reboiler
duty as illustrated in Fig. 6 thus this value will be kept constant. The
sensitivity results of the lean amine temperature showed that it is
proportionally correlated with reboiler duty and inversely propor-
tional to circulation rate as shown in Fig. 10. However, the effect of
lowering base case lean amine temperature from 51  C to 45  C
results in insignicant reduction in reboiler duty (0.7%) and circu-
lation rate (0.2%). Therefore, this parameter will be held constant
for suggested optimized scenarios.
The effect of the number of absorber stages is negligible against
the reboiler duty as shown in Fig. 12. The average percentage
reduction of reboiler duty per extra stage is only 0.17%. The number
of stages of base case will be kept constant thus providing no extra
capital cost penalty in the two suggested scenarios. Fig. 16. Absorber temperature bulges for conditions of scenarios in Table 5.
 M.S. Jassim / Journal of Natural Gas Science and Engineering 36 (2016) 175e183 183

maximum temperature in the absorber. It is also observed that the solutions of MEA and DEA. Fluid Phase Equilibr 185 (1), 31e43.
Aliabad, Z., Mirzaei, S., 2009. Removal of CO2 and H2S using aqueous alkanolamine
reduction of circulation rate moves the peak towards the top of the
solutions. World Acad. Sci. Eng. Technol. 49 (1), 194e203.
absorber. Aspen HYSYS, 2004. Release 2004, Simulation Basis. Aspen Technol., Cambridge,
MA.
7. Conclusions Astarita, G., Savage, D.W., Bisio, A., 1983. Gas Treating with Chemical Solvents. John
Wiley & Sons, New York.
Bolhar-Nordenkampf, M., Friedl, A., Koss, U., Tork, T., 2004. Modeling selective H2S
This validated simulation work shows that Aspen HYSYS is a absorption and desorption in an aqueous MDEA-solution using a rate-based
reliable software to perform sensitivity analyses of a conventional non-equilibrium approach. Chem. Eng. Process 43, 701e715.
Bullin, J.A., Polasek, J.C., Donnelly, S.T., 1990. The use of MDEA and mixtures of
MDEA gas sweetening plant that selectively captures H2S in the amines for bulk CO2 removal. Proc. Sixty-Ninth Gas. Process. Assoc. Annu. Conv.
presence of CO2 when both instantaneous and pseudo-rst-order 135e139.
chemical reactions occur in the liquid phase. Freguia, S., Rochelle, G.T., 2003. Modeling of CO2 capture by aqueous mono-
ethanolamine. AIChE J. 49 (7), 1676e1686.
The distribution of temperature bulge across the absorber in- Gao, H., Zhou, L., Liang, Z., Idem, R.O., Fu, K., Sema, T., Tontiwachwuthikul, P., 2014.
dicates the uniformity of the circulation rate. This work showed Comparative studies of heat duty and total equivalent work of a new heat pump
that the temperature peak is proportional to the concentration of distillation with split ow process, conventional split ow process, and con-
ventional baseline process for CO2 capture using monoethanolamine. Int. J.
MDEA. The sensitivity analyses showed that the most critical pro- Greenh. Gas. Control 24, 87e97.
cess variables affecting the performance of H2S-CO2-MDEA-H2O Ga
sp
ar, J., Ana-Maria, C., 2011. Dynamic modeling and validation of absorber and
gas sweetening system in terms of normalized reboiler duty and desorber columns for post-combustion CO2 capture. Comput. Chem. Eng. 35
(10), 2044e2052.
sour gas recoveries are MDEA concentration and circulation rate.
Huttenhuis, P.J.G., Agrawal, N.J., Hogendoorn, J.A., Versteeg, G.F., 2007. Gas solubility
The use of 30 wt% alkanolamine solution as an optimized sce- of H2S and CO2 in aqueous solutions of N-methyldiethanolamine. J. Pet. Sci. Eng.
nario in comparison to the baseline scenario provides a saving of 55 (1), 122e134.
8.6% in the reboiler duty, a reduction of 14.4% in circulation rate, a Huttenhuis, P.J.G., Agrawal, N.J., Versteeg, G.F., 2009. Solubility of carbon dioxide
and hydrogen sulde in aqueous N-methyldiethanolamine solutions. Ind. Eng.
reduction of 28.6% in lean amine loading, and improvement of Chem. Res. 48 (8), 4051e4059.
1.02% in H2S capture. This implies that the operational cost will be Jassim, M.S., Rochelle, G.T., 2006. Innovation absorber/stripper conguration for CO2
lower with respect to heat transfer utilities and pumping. capture by aqueous monoethanolamine. Ind. Eng. Chem. Res. 45 (8),
2465e2472.
Jassim, M.S., Rochelle, G.T., Eimer, D., Ramshaw, C., 2007. Carbon dioxide absorption
Acknowledgement and desorption in aqueous monoethanolamine solutions in a rotating packed
bed. Ind. Eng. Chem. Res. 46 (9), 2823e2833.
Kidnay, A.J., Parrish, W.R., McCartney, D.G., 2011. Fundamentals of Natural Gas
Any opinion, ndings, conclusions, or recommendations Processing. CRC Press, p. 218.
expressed herein are those of the author. This paper was prepared Kohl, A.L., Nielsen, R.B., 1997. Gas Purication, fth ed. Gulf Publishing Company,
with the nancial support of Deanship of Research in University of Houston, USA.
Kvamsdal, H.M., Rochelle, G.T., 2008. Effects of the temperature bulge in CO2 ab-
Bahrain (UOB) (Grant number: 17). The author gratefully sorption from ue gas by aqueous mono-ethanolamine. Ind. Eng. Chem. Res. 47
acknowledge the fruitful help including practical discussion with (3), 867e875.
Ashwin Murali (UOB). Ko, J., Li, M., 2000. Kinetics of absorption of carbon dioxide into solutions of N-
methyldiethanolamine plus water. Chem. Eng. Sci. 19, 4139e4147.
Lunsford, K.M., Bullin, J.A., 1996. Optimization of amine sweetening units, Bryan
Notation Research and Engineering, Inc.. Technical Papers. In: Proceedings of the 1996
AIChE Spring National Meeting. American Institute of Chemical Engineers, New
Cp Specic heat of liquid in reboiler York, NY.
Mandal, B.P., Biswas, A.K., Bandyopadhyay, S.S., 2004. Selective absorption of H2S
DHabs,CO2 Heat of absorption of CO2, kJ/mol from gas streams containing H2S and CO2 into aqueous solutions of N-meth-
DHvap,H2O Heat of vaporization of H2O, kJ/mol yldiethanolamine and 2-amino-2-methyl-1-propanol. Sep. Purif. Technol. 35
DHabs,H2S Heat of absorption of H2S, kJ/mol (3), 191e202.
i, L. E., 2007. Aspen HYSYS simulation of CO2 removal by amine absorption from a
L Liquid molar ow in reboiler, kmol/s gas based power plant. In The 48th Scandinavian Conference on Simulation and
nCO2 Moles of CO2 removed from ue gas, kmol/s Modeling (SIMS 2007), 30-31 October, 2007, Go teborg (Saro
), 73e81.
nH2O Moles of H2O condensed in stripper, kmol/s i, L.E., Brthen, T., Berg, C., Brekne, S.K., Flatin, M., Johnsen, R., Thomassen, E., 2014.
Optimization of congurations for amine based CO2 absorption using Aspen
nH2S Moles of H2S removed from ue gas, kmol/s HYSYS. Energy Procedia 51, 224e233.
Qr Reboiler duty, MJ Oyenekan, B.A., Rochelle, G.T., 2006. Energy performance of stripper congura-tions
T Temperature, K for CO2 capture by aqueous amines. Ind. Eng. Chem. Res. 45 (8), 2457e2464.
Pacheco, M.A., Rochelle, G.T., 1998. Rate-based modeling of reactive absorption of
V Vapor owrate in stripper, kmol/s CO2 and H2S into aqueous methyldiethanolamine. Ind. Eng. Chem. Res. 37 (10),
4107e4117.
References Polasek, J., Bullin, J., 1984. Selecting amines for sweetening units. Energy Prog. 4,
146e149.
Rodrguez, N., Mussati, S., Scenna, N., 2011. Optimization of post-combustion CO2
Abdulrahman, R.K., Sebastine, I.M., 2013. Natural gas sweetening process simulation
process using DEAeMDEA mixtures. Chem. Eng. Res. Des. 89 (9), 1763e1773.
and optimization: a case study of Khurmala eld in Iraqi Kurdistan region.
Seader, J.D., Henley, Ernest J., Keith, Roper D., 2011. Separation Process Principles,
J. Nat. Gas. Sci. Eng. 14, 116e120.
third ed. John Wiley & Sons, Inc., USA.
Abu-Zahra, M.R.M., Schneiders, L.H.J., Niederer, J.P.M., Feron, P.H.M., Versteeg, G.F.,
Versteeg, G.F., Van Swaaij, W.P.M., 1988. On the kinetics between CO2 and alka-
2007. CO2 capture from power plants Part I. A parametric study of the technical
nolamines both in aqueous and non-aqueous solutions-II. Tertiary amines.
performance based on monoethanolamine. Int. J. Greenh. Gas. Control 1, 37e46.
Chem. Eng. Sci. 3, 587e591.
Al-Baghli, N.A., Pruess, S.A., Yesavage, V.F., Selim, M.S., 2001. A rate-based model for
the design of gas absorbers for the removal of CO2 and H2S using aqueous