Journal of Natural Gas Science and Engineering 37 (2017) 327e340

Contents lists available at ScienceDirect

Journal of Natural Gas Science and Engineering

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

Technical and economic evaluations of the triethylene glycol

regeneration processes in natural gas dehydration plants
Mihaela Neagu*, Diana Luciana Cursaru
Petroleum Refining and Environmental Engineering Department, Faculty of Petroleum and Petrochemistry Techonology, Petroleum-Gas University of
Ploiesti, 39 Bucuresti Blvd, 100680 Ploiesti, Romania

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

Article history: A conventional natural gas dehydration plant and one based on striping gas concept, which employs
Received 12 August 2016 triethylene glycol (TEG) as the dehydrating agent, were simulated using a steady state simulator (UniSim
Received in revised form Design). The main units were included in the flowsheets, namely: absorber, flash units, heat exchangers,
11 October 2016
regenerator, and reboiler. All simulations were performed of about 25 L TEG/kg water absorbed. The
Accepted 24 November 2016
Available online 28 November 2016
equation of state (EOS) used in the simulation is the Peng Robinson (PR). The reboiler temperature of
conventional regenerator and then the hot stripping gas flowrate, have been studied for their response to
changes in the regenerated TEG concentration, dew point of sale gases, TEG losses (make-up), regen-
Keywords:
Natural gas
erator overhead vapor flowrate, and partial pressure of water vapor. Despite the increase in plant
Dehydration plant complexity, the fixed capital investment estimation proves an insignificant costs increase of the stripping
TEG gas configuration over the benchmark. It appears that stripping gas is a more effective way to improve
Stripping gas the regenerated TEG concentration and entire dehydration plant performance.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction
Among all fossil fuels, natural gas is the most environmentally
friendly and for this reason it is expected a rapidly growing in global
importance as a primary energy source. Natural gas is also used by
industry, mainly as a feedstock for downstream industry.
The natural gas must satisfy some specifications that it makes
suitable for transport in specific systems. Consequently, natural gas,
which in most cases contains contaminants (i.e. non-hydrocarbon
gases such as water vapors, carbon dioxide, hydrogen sulfide, ni-
trogen, oxygen, and helium) and heavy (liquid) hydrocarbons, must
be processed or treated, prior for delivery to the pipelines transport
(Mokhatab et al., 2006). The treatment process includes several
steps for creating pipeline-quality natural gas. The number of the
steps and the type of the techniques most often depends upon the
gas source. A dehydration step is needed to eliminate water that is
present in various amounts dependent on upstream conditions
(Carol, 2002). Water in natural gas can create serious problems
during gas transportation and processing or depreciate its quality
by decreasing the heating value. The most severe problem is the
* Corresponding author.
E-mail addresses: mpetre@upg-ploiesti.ro (M. Neagu), dianapetre@upg-ploiesti.
ro (D.L. Cursaru).
formation of gas hydrates that may block plug valves fittings or
pipelines, compression systems, process equipment and in-
struments (Gong et al., 2010; Mokhatab et al., 2007). Another
common problem in the oil and gas industry associated with nat-
ural gas and condensed water is the corrosion of materials,
particularly when CO2 and H2S are presented in the gas (ObaniJesu,
2009).
The treatment for removal of the water vapors from natural gas
consists basically of dehydration process which is accomplished by
decreasing the dew point temperature. The most common methods
of natural gas dehydration are absorption in liquid desiccant,
adsorption on solid desiccant, and refrigeration (Netusil and Ditl,
2011; Rouzbahani et al., 2014).
The liquid desiccants suitable for dehydrating of the natural gas
include the following substances: calcium chloride, lithium chlo-
ride, glycols, zinc chloride (Gandhidasan et al., 2001). Among of
these, the glycols are the main ones used for natural gas dehydra-
tion. The commonly available glycols are monoethylene glycol
(MEG), diethylene glycol (DEG), triethylene glycol (TEG) and tet-
raethylene glycol (TREG). Triethylene glycol is used for many de-
cades in vapor water absorption and by far the most suitable for
commercial application where dew point depressions of about
15e49
C are required (Gas Processors and Suppliers Association,
Engineering Data Book, 2012). When dew point depressions

http://dx.doi.org/10.1016/j.jngse.2016.11.052
1875-5100/© 2016 Elsevier B.V. All rights reserved.
328 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340
consistently higher than about 82
C are required, the adsorption-
based processes are generally specified (Kohl and Nielsen, 1997).
The most common solid desiccants used in adsorption of water
vapors are molecular sieves, silica gel or alumina. Despite their
higher costs, the molecular sieves are preferred as adsorbent when
very lower dew points are required, such as natural gas feed
streams for cryogenic hydrocarbon recovery units. The main
drawback of dehydration based on adsorption technique is the
capital cost, which is 2e3 times higher than that for absorption
(Netusil and Ditl, 2011).
Refrigeration by external vapor recompression, as a dehydration
technique, requires large pressure drop therefore; this process is
recommended when high-pressure gas is available.
Dehydration by membrane permeation or supersonic processes
are seldom used. Some of the disadvantages of membranes in the
dehydration process, such as lower selectivity, easily fouling by gas
contaminants, high costs especially for large gas flowrates appli-
cations, make the process noncompetitive (Scholes et al., 2012).
Supersonic gas processing systems were recently promoted in
this area and both condensation and separation occurs in a simple
device, i.e. supersonic nozzles. They are relatively simple to operate,
cheap, environmentally friendly and significant for offshore gas
applications (Karimi and Abedinzadegan Abdi, 2009; Wen et al.,
2011).
In the near future, it is unlikely that membrane permeation and
supersonic processes would obtain significant market share for
dehydration of natural gas applications. Many researches and de-
velopments of this dehydration processes are needed to make them
competitive with absorption in triethylene glycol or adsorption on
solid desiccant processes which are successfully used in current
practices for many years.
There are also many commercially available processes for
customized dehydration systems, especially absorption in TEG.
Several processes are available today, each employing different
strategies to enhance glycol regeneration (Rahimpour et al., 2013b).
The alternative glycol regeneration concepts available in current
practice are related to the Drizo, the Coldfinger, vacuum distillation
and the stripping gas processes (Rahimpour et al., 2013a). From all
of these concepts, the stripping gas is the simplest and most
widespread process for enhanced glycol regeneration. Stripping gas
is therefore mainly used to increase the purity between 99.1 and
99.6 wt % at the regeneration temperature and slightly over at-
mospheric pressure. Injecting the stripping gas directly into the
reboiler has the great advantage to require few additional equip-
ments (Kohl and Nielsen, 1997).
In this paper, according to the advantages noted above, the
performances of a stripping natural gas drying process which em-
ploys triethylene glycol (TEG) as the dehydrating agent are re-
examined both as technological feasibility and economics, in
comparison to those of a conventional drying process. The paper is
sequenced as follows: Section 2 describes the drying processes and
state the input data employed in our investigation. The drying
processes are presented in terms of simulation flowsheets accord-
ing to UniSim Design software facilities. In section 3, the calculation
methods for sales gas dew point, for the total equivalent work and
for economic evaluations (i.e., fixed capital investment and total
cost of production) are showed. In section 4, a systematic procedure
in order to determine the operating conditions that maximize the
regenerated TEG concentrations, is developed. The simulation re-
sults are discussed in terms of their limitations and advantages on
overall technological performance of both drying processes. The
economic evaluation is completing our investigation and its results
support the final decision. Some concluding remarks will be given
in Section 5.
2. Process description
Two process configurations are re-examined in this work: the
conventional process with a simple regenerator as benchmark and
the process with hot gas stripping injection in the reboiler of
regenerator. In help of our investigations, a commercial simulator
for chemical processes, namely UniSim Design R-380 is involved.
In the conventional drying process, the rich TEG is regenerated
by reducing the pressure and increasing the temperature, before it
is recycled into the absorption column. The most important
drawback of the atmospheric conventional process is the lean TEG
concentration of 98.8e98.9 wt% (GPSA, 2012; Piemonte et al., 2012).
Nowadays, the simplest and most used strategy employed for
enhanced glycol regeneration is the stripping gas process (GPSA,
2012; Kohl and Nielsen, 1997; Piemonte et al., 2012). In this pa-
per, a small portion of the dry gas at the regeneration pressure is
heated and introduced into the regenerator of reboiler. Stripping
gas is used in order to decrease the partial pressure of water in the
vapor phase, and thereby to obtain glycol with purities between
99.2 and up to 99.9 wt% at the regeneration temperature and near
ambient pressure. However, in the stripping configuration, the
process complexity will increase. These configurations are
described in sections 2.1 and 2.2.
It is well known that natural gas compositions can vary widely,
depending on where and from what type of reservoir the natural
gas is produced (Kotarba and Nagao, 2008; Wang and Economides,
2009). Moreover, the gas composition and properties play a major
role in gas processing plants. Available literature related to natural
gas compositions or operating conditions in the processing plants
are very limited. This is not surprising because most of technical
data are confidentially. Taking into account these aspects, a speci-
fied natural gas composition is not used in this work. More than
that, for general applicability of this work, a typical natural gas
composition was preferred. The specifications of typical sweet
natural gas are given in Table 1 (Demirbas, 2010). The pressure inlet
in gas plant dehydration is based on example 20e11 from Gas
Processors Suppliers Association (GPSA) Data Book.
2.1. Conventional drying process
Fig. 1 shows a schematic flowsheet of the conventional drying
process with TEG.
As shown in Fig. 1, a typical natural gas (see Table 1) saturated
with 0.001192 mol fraction water at equilibrium conditions enters
to the bottom of absorber [ABS] and flows upward. The dry gas
leaves the top of the absorber column. Equilibrium water content or
equilibrium dew point of the dry gas is strongly dependent by lean
Table 1
Typical sweet natural gas specification.
Properties Value
Temperature (
C) 30
Flow rate (Nm3/day) 1,000,000
Pressure (kPa) 4100
Composition (vol%)
Methane 94.9
Ethane 2.5
Propane 0.2
i-butane 0.003
n-butane 0.003
i-pentane 0.001
n-pentane 0.001
n-hexane 0.001
CO2 0.7
Oxygen 0.02
Nitrogen 1.6
 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340 329

Fig. 1. Schematic flow sheet for conventional TEG drying process.

TEG concentration. The lean TEG which coming from the regener- labeled 2 and stream labeled Lean TEG recycle.
ator unit [REG]) is recycled to the top of the absorber. The TEG flows The simulations of the processes were developed by assuming
down the column over trays and water is absorbed from the wet the input parameters of the columns listed in Table 2.
gas. All simulations were performed of about 25 L TEG/kg water
absorbed (GPSA, 2012) and lean TEG concentration in wt. %
depending to the regenerator operating temperature. The rich TEG
carrying the water leaves the bottom of the absorber and passes Table 2
through the expansion valve [VLV] to reduce its pressure from Input parameters of the columns used in the steady-state simulations.
4089 kPa to 139.5 kPa. The liquid/vapor stream enters in a vertical Parameter Value, u.m.
two phases separator [HC FLASH]. A small flow (3.37 Nm3/h) con-
General
sisting in wet hydrocarbon vapors leave the separator from top. Property package PR (Peng-Robinson)
Rich TEG at 129.5 kPa and 34.3
C passes through the lean/rich EOS solution method Cubic EOS Analytical Method
glycol heat exchanger [L/R HX] and then flows with 100
C and Absorber
119.5 kPa to the 4th stage of the regenerator. Rich TEG flows Top pressure 4080 kPa
Total pressure drop 8 kPa
downward through the regenerator [REG]. Water vapors, hydro-
Number of stages 14 a
carbons and TEG traces are removed from the rich TEG and Overall stage efficiency 0.3
condensed steam returns in the regenerator as reflux. The overhead Tray sizing utility Bubble cap
vapor flow at 122.9
C passes then from a cooler [COOLER]. The Number of flow paths 2
vapor/liquid flow enters at 30
C in a vertical two phases separator Tray spacing 0.6096 m
Max. flooding 85%
[GW]. A small wet gas stream is removed from separator feed as top Sieve tray flooding method Minimum Csb
product. Optionally, the wet gas stream can be mixed with wet Regenerator
hydrocarbon stream from [HC FLASH] and recycled together in Condenser Full reflux
absorber feed (not simulated in this research). The downstreams of Top pressure 102 kPa
Total pressure drop 18 kPa
[GW] are water and TEG traces. The best practice is to feed this
Feed temperature 100
C
stream, along with the excess water from the [GS] separator, to a Number of stages 10 b
sewage plant (Sommer and Melin, 2004). Lean TEG leaves the Reflux ratio 0.01 c
c
regenerator at operating temperature and pressure and goes Bottom temperature 175; 180; 185; 190; 195; 200; 204
Packing column utility
through the lean/rich glycol cross heat exchanger [L/R HX] and lean
Raschig rings 1/16 in. wall, CSteel, 1&1/2 in.
TEG cooler [L/DG HX] before returning with 34
C to the absorber. L/ Packing column sections Section 1: stage 1 to stage 3
DG HX is a gas/liquid heat exchanger, meaning that the dry gas at Section 2: stage 4 to stage 6
31
C (i.e., top absorber temperature) is the cooling agent for lean Section 3: stage 7 to stage 10
TEG. The final product of the conventional flowsheet is roughly HETP 0.3048 m
Section spacing 0.4 m
1,000,000 Nm3/day sale gas at 4070 kPa. The glycol lost in dry gas
Max. flooding 70%
stream, in vapor stream from [HC FLASH] and in wet gas from [GW] Packing correlation Robins
was completed by a make-up flow, in [MIX-2]. UniSim Design a
Counted from top to bottom.
software allows the user to implement RECYCLE logical operations. b
Counted from top to bottom (excluding condenser and reboiler).
In this work [RCY- 1] are shown the relationships from stream c
Column specification.
330 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340
2.2. Stripping gas drying process
As mentioned earlier, the stripping gas is a small portion that is
heated and then is fed in reboiler attached to the regenerator. Fig. 2
shows a schematic flowsheet of the stripping gas drying process
with TEG. In following, only the new units and major changes in
operating procedure over the conventional flowsheet are discussed.
As it can be seeing in Fig. 2, a new RECYCLE logical operation [RCY-
2] is implemented to show the relation between sale gas flow and
the flow fed to a splitter unit [TEE]. The sale gas is divided in two
streams. One outlet stream is the final sale gas and the other one is
stripping gas. Next, the stripping gas is routed to a new expansion
valve unit [VLV-2] in order to reduce its pressure from 4070 kPa to
130 kPa. Stripping gas must be heated before enters in reboiler. In
this work it is proposed the emplacement of a heat exchanger be-
tween lean TEG from reboiler and stripping gas in [L/G HX] the new
unit. The stripping gas temperature after [L/SG HX] is discussed in
section 4.
In simulations performed for stripping gas configuration, all
parameters from Table 2 are kept unchanged, except the temper-
ature of the reboiler witch will be chosen after conventional drying
process simulation, as optimal operating temperature.
Both conventional and stripping gas flow sheets include few
heat exchangers. The [L/R HX] and [L/DG HX] are designed as shell
and tubes exchangers (shell TEMA E, horizontal orientation, 2 tube
passes per shell). The pressure drops per shell side and per tube
side are 10 kPa, each. The [L/SG HX] is designed as double pipe heat
exchanger. The reboiler is kettle type, horizontal orientation, and
medium pressure steam is selected as heating medium. The air at
20
C is the cooling agent in [COOLER] and in the condenser of the
column, as well. No energy losses are assumed in these systems. A
75% adiabatic efficiency is assumed for pumps.
Fig. 2. Schematic flow sheet for stripping gas drying process.
3. Calculations
3.1. Dew point calculation
The outlet gas water amount specification is the final purpose of
the dehydration process design. It is common to refer to water
content of a gas in terms of water dew point. In present work, an
empirical method is selected to calculate the equilibrium dew point
of natural gas leaving the absorber column. Gandhidasan (2003)
proposes the following empirical equation:


Tdew;eq ¼ 18:228  ln 0:001685  wout Pg0:81462 (1)
where Tdew;eq is dew point in
C; wout is water content of the gas
(kilograms of water per million standard cubic meters) and Pg is
pressure (MPa).
UniSim simulation does not provide the possibility of checking
the amount of water per standard cubic meter directly. In Eq. (1),
the mass water content of the gas is provided from simulation re-
sults as well the gas flow in standard cubic meter. In order to
achieve the dew point results, a Microsoft Excel application was
accessed and here the calculation with Eq. (1) was made.
3.2. Cost calculation
In this work the fixed capital investment (FCI) and total cost of
production (TCOP) estimations are done based on data from Towler
and Sinnott (2008). The fixed capital investment is made up of:
FCI ¼ ISBL þ OSBL þ EC þ CO (2)
where ISBL is the inside battery limits investment (including direct
field costs such as: costs of major equipments, bulk items, civil
 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340
work, installation labor and supervision, and indirect field cost such
as: temporary construction costs, field expenses and services,
construction insurance, labor benefits and burdens, miscellaneous
overhead items); OSBL is the offsite cost; EC is the engineering costs
(are referred to the contractor charges, detailed design and other
engineering services); and CO is contingency charges.
In this work, the inside battery limits (ISBL) investment included
only the direct field cost. Hence, the ISBL it is reduced to a fixed
capital cost. Fixed capital cost (C) is based on an estimate of the
purchase cost of the major equipment items required for the pro-
cess, and the other costs (such as: equipments erection including
foundations and minor structural work, piping, including insu-
lation and painting, electrical power and lighting, instruments and
automatic process control systems, process buildings and struc-
tures, ancillary buildings, offices, laboratory buildings, storage for
raw materials, site preparation) are estimated as factors of each
equipment cost.
The Hand factor method propose that the C fixed capital cost of a
plant to be given as a function of the total purchased equipment
cost by the equation:
X
i¼M
C¼ ðF,Ce Þ
i¼1
where C is total capital cost of the plant; Ce is the purchased cost of
each major equipment items (i ¼ 1 … M); F are Hand's installation
factors specified for each major equipment type. The factors F are
presented in Appendix A as selection from Towler and Sinnott work.
Equation (3) can be used to make a preliminary estimation once the
flowsheet has been drawn up and the main plant equipments have
been sized.
The equipment cost Ce is estimated by the next equation:
Ce ¼ a þ b,Sn
where Ce is the purchased equipment cost on Chemical Engineer-
ing Plant Cost Index CEPCI of 478.6 available for January 2006; a and
b are cost constant; S are size parameter; n is exponent for that type
of equipment. Constants a and b, parameter S and exponent n are
shown in Appendix A as selection from Towler and Sinnott work.
The estimated cost Ce is for base equipment made from carbon
steel. To calculate the purchased equipment cost with different
materials is used a correction material factor (fm ¼ 1.3) as is rec-
ommended in the above mentioned book. The CEPCI from Chemical
Engineering Magazine is used to update the prices as of 2006, ac-
cording to Eq. (5).
Cost index in year A
Cost in year A ¼ Cost in year B,
Cost index in year B
where: cost in year A relates to year 2016; cost in year B relates to
year 2006; cost index in May 2016 is 543.5. All monetary values are
done in US dollars.
Fixed Capital Investment (FCI) is converted into a constant series
of payment for every year of the plant life. The annual capital cost
(ACC) is calculated from Eq. (6).
    concentration in the bottom of regenerator and on equilibrium dew
ACC ¼ FCI, ð1 þ iÞn  1 = ið1 þ iÞn
where i is the interest rate and n is the project life time. The
parenthesis of above equation is the annual capital charge ratio.
The annual cost of production COP is made up of:
331
COP ¼ VCOP þ FCOP (7)
where VCOP and FCOP are variable and fixed costs of production,
respectively. Variable costs of production are sum of all the variable
costs of production that are proportional to the plant output or
operation rate. The FCOP are sum of all the fixed costs of production
that are covered regardless of the plant operation. The VCOP and
FCOP are estimating as direct calculation based on operating plant
data or as factors of current and/or above calculations, as it can see
in Table 3. It is assumed that plant operate 8000 h/year.
In this work is useful to calculate a total cost of production
(TCOP) as sum of the annual cost of production and the annual
capital cost (Eq. (8)).
TCOP ¼ COP þ ACC (8)
4. Results and discussions
4.1. Accuracy
(3) As mentioned earlier, UniSim Design -R380 software marketed
by Honeywell Company has been used to model by steady-state
simulations the drying gas process. UniSim allows the user to
implement solver-methods. In this work a general solving method,
namely HYSIM Inside-out method was selected to model the ab-
sorbers and Modified HYSIM Inside-out method to model the re-
generators and implement some restrictions. The solving methods
consider the mass and energy balances including logical constraints
to provide the convergence. In all heat-exchangers, constraints, in
order to avoid the temperature crossovers between hot and cold
streams, have been introduced. It generally requires a large number
of iterations to converge on the solution; a 10,000 maximum
(4) numbers of iterations were chose for column simulations and 25
maximum numbers of iterations for heat-exchangers. The equilib-
rium error tolerance is 1.000  105 and Heat/spec. error tolerance
is 5.00  104.
4.2. Results and discussions for conventional drying process
simulation
A deep dehydration of natural gas requires high purity of the
recycled solvent. To improve this purity the pressure and temper-
ature during the regenerator can be handling. In this work, the
pressure in regenerator is kept constant and slightly above atmo-
spheric (see Table 2) in order to prevent any contamination with air
that could promote the undesired oxidation of the glycol. Therefore,
raising the temperature during the regeneration step was investi-
gated. The typical TEG dehydration unit operates to the reboiler
temperature around 190
C or higher. However, the regeneration
(5) temperature must remain below an acceptable limit to avoid the
thermal decomposition of glycol. According to Gironi et al. (2010)
and Piemonte et al. (2012), the limit of regeneration temperature
is 204
C for TEG. Some doubts about these temperatures are re-
ported in literature. For example, originates from Dow Chemical
manufacturer onset decomposition data is 240
C, but Gas
Processors Suppliers Association (GPSA) Data Book (2012) indicate
as 207
C. To develop the reboiler temperature analysis on TEG
(6) point of sale gases from absorber, a range of 175
C to 204
C (with
an increment of 5
C, except the last one) was chosen in conven-
tional drying process simulations. Fig. 3 shows that TEG concen-
trations increase with regenerator temperature, as in the dew point
of sale gases.
332 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340
Table 3
Components of COP estimation.
Variable Cost of Production (VCOP)
TEG cost (high purity)
Steam cost
Electrical energy cost
Miscellaneous (consumable materials, TEG loss, waste water treatments, waste disposal)
Fixed Cost of Production (FCOP)
Operating labor (OL)
Supervision (S)
Direct salary overhead (DSO)
Maintenance (M)
Property taxes and insurance
Rent of land (and/or buildings)
General plant overhead
Environmental charges
Fig. 3. Variation of TEG concentration and equilibrium dew point of sale gases with reboiler temperature.
Some parameters such as the overhead vapor flowrate, partial
pressure of water vapors, TEG loss in the top of regenerator and
reboiler duty depend on the regenerator temperature and was
evaluated as well. The overhead flow includes mainly the water
vapors, the glycol loss and few hydrocarbon vapors stripped from
rich glycol. The overhead flowrate and TEG loss increase because
the mass flowrate of stripping vapor generated by reboiler increase
with regenerator temperature. Fig. 4 illustrates the variation of the
overhead flowrate and of the TEG loss with regenerator
temperature.
The total reboiler heat duty has two functionalities: to provide
the sensible heat for heating the rich glycol from fixed fed tem-
perature to reboiler temperature and to provide the latent heat to
vaporize water. In fact, under fixed feed temperature and roughly
constant feed flowrate and under fixed regenerator pressure (see
Table 2), the variation of the regenerator heat duty is predomi-
nantly due to the latent heat of water vaporization. Higher reboiler
duty is achieved by increasing the reboiler temperature as it can be
seen in Fig. 5. At the same time, by increasing the reboiler tem-
perature, it was notices a decreasing in the water vapor composi-
tion in overhead flow and it is partial pressure (at constant top
2513.28 $/tonne TEG
12$/tone
0.10$/kWh
2% of VCOP
25,000$/year/each operator
25% of OL
50% of (OL þ S)
3% of ISBL
2% of ISBL
1% of ISBL þ OSBL
65% of (OL þ S þ DSO þ M)
1% of ISBL þ OSBL
pressure).
According to all previous comments, the optimal operating
condition for conventional drying process cannot be chosen easily.
However, taking into account the equilibrium dew point value of
the sale gas as main criterion, it turns out that is more favorable to
operate the regenerator at higher temperature. In our opinion, it is
adequate a temperature of 200
C in the reboiler of the regenerator,
although there are some doubts regarding the TEG degradation
temperature. Table 4 presents the input and optimum results for
conventional gas drying process with TEG at 200
C in regenerator
reboiler.
According to the simulation data from Table 4, at temperature of
200
C in reboiler of regenerator it can be obtained a 98.74 wt% TEG
in water which is recycled to the top of the absorber after cooling in
[L/R HX] and [L/DG HX] heat exchangers. The calculated TEG cir-
culation rate is 25.81 L TEG/kg water absorbed. The water content of
the outgoing gas was 89.67 mg/Nm3 (or 84.94 mg/Sm3) which
determines the sale gas dew point of about 14.59
C under the
pressure of 4070 kPa. Dew point results comparison was made
between the sales gas dew point calculated from empirical Eq. (1)
on basis of values provided by UniSim simulation and the dew
 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340 333

Fig. 4. Variation of the overhead vapor flowrate and of the TEG loss flowrate with reboiler temperature.

Fig. 5. Variation of the reboiler duty and the partial pressure of water vapors with reboiler temperature.

point result achieved with use of GPSA chart (Gas Processors
Suppliers Association (GPSA) Engineering Data Book, 2012). From
GPSA chart was found that for water content of 84.94 mg/Sm3 and
4070 kPa the equilibrium dew point is 10
C. Hence, as the GPSA
chart is well recognized as accurate for predicting the dew point of
natural gas under given conditions, it can be considered that the
values obtained with use of Eq. (1) are somewhat questionable.
Even in this case, we decided to use in next simulations the UniSim
data in Eq. (1) because these data are not affected by key error that
would affect significantly our conclusions over stripping gas drying
process strategy.
The results for conventional drying process simulation clearly
shows that the regenerator can provide only poor concentration of
TEG in recycled absorbent and in consequence, the dew point is not
accurate according to some national norms (e.g. according to Report
prepared in 2011 for the European Commission by GL Noble Denton
and Po€yry Management Consulting (2011). Nowadays, the dew
point is a characteristic of the sale gas and its value differs from
country to country (i.e. Italy - 5
C at 70 bar, Austria 8
C at 40 bar,
Greece þ5
C at 80 bar). These values can be achieved by using the
conventional process.
In the near future is desired a harmonization of the sale gas
334 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340

Table 4
Input and optimum results for conventional gas drying process with TEG at 200
C in regenerator reboiler.

Material streams

Wet gas to absorber Sales gas (after L/DG-HX) Lean TEG recycle Regenerator feed Overhead vapor

Vapor fraction 1 1 0 0 1
Temperature,
C 30.00 34.48 34.00 100.00 122.86
Pressure, kPa 4100.00 4070.00 4080.00 119.50 102.00
Mass flowrate, kg/h 31,635.30 31,595.72 1069.14 1105.61 37.28
Molar flow, kmoli/h 1862.00 1860.17 7.78 9.80 2.0244
Volume flowrate, Nm3/h or m3/h 41,708.80 41,667.81 0.948,997 0.9858 45.35
Water flowrate, kg/h 39.98 3.74 13.45 49.65 36.14
Water dew point by Eq. (1) 14.59

Component molar fraction

Total hydrocarbons 0.998,809 0.999,888 e 0.001289 0.006235

Water 0.001264 0.000112 0.096,006 0.281,302 0.991,396
TEG e 0.000000 0.903,994 0.717,409 0.002369

Component mass fraction

Total hydrocarbons 0.998,737 0.999,880 e 0.000365 0.010,794

Water 0.001264 0.000118 0.012,581 0.044,909 0.969,884
TEG e 0.000002 0.987,419 0.954,726 0.019,322

Heat flow, kJ/h Energy stream

Condenser Reboiler
1758.29 385,066.27
Pump energy, kJ/h 5575.31

quality in all EU countries, respectively a dew point of 8
C at
70 bar. Under these circumstances, it is likely that many countries
will be forced to re-tech their plants and to switch to the stripping
technology. Our paper aims to demonstrate that switching from
conventional to stripping process does not involve very high costs,
instead allows more flexibility. By changing the flow of the strip-
ping gas we can obtain the desired dew point, no matter how low
will be this value. The only drawback would be a slight decrease of
the sale gas flow, due to its utilization as stripping gas.
4.3. Results and discussions for stripping gas drying process
simulation
The stripping gas drying process simulation was developed
based on flowsheet from Fig. 2, at operating temperature in
regenerator attached to the reboiler of 200
C. The small portions
from sales gas stream are used as stripping gases and are expanded
in valve unit [VLV-2] at 130 kPa, therefore their temperatures will
decrease from 34.48
C to 15
C. Striping gases need to be heated at
120
C by heat exchange with lean TEG from reboiler, in [L/G HX]
unit.
In this study five new simulations have been performed for
different stripping gas flowrates obtained with the aid of [TEE]
splitter unit. The investigated range of stripping gas flowrate is
from 41.664 Nm3/h (or 0.1% from gas sales flowrate) to 208.320
Nm3/h (or 0.5% from gas sales flowrate). Fig. 6 shows that TEG
concentration increases with stripping gas flowrate, and with
decreasing of the dew point of sale gases.
From Fig. 6 it is obvious that stripping gas drying process in-
creases the TEG concentration in lean solvent. A comparison be-
tween Figs. 6 and 3 shows clearly that TEG concentration increases
from 98.74 wt% to 99.22e99.85 wt% depending on the stripping gas
flowrates, at 200
C constant temperature in reboiler of the
regenerator. At the same time, equilibrium dew points decrease
mainly due to the increasing of TEG concentration.
Many parameters such as the overhead vapor flowrate, partial
pressure of water vapors, TEG loss in the top of the regenerator and
reboiler duty, depends on the stripping gas flowrate and were
evaluated as well.
In this process configuration, by increasing the stripping gas
flowrate it was noticed a negative effect on the overhead vapor
flowrate and reboiler duties, as shown in Fig. 7. The negative side is
obvious and cannot be avoided because entire stripping gas stream
goes to the top of the regenerator together with the stripping vapor
generated by reboiler. This negative effect is demonstrated by linear
graphs in Fig. 7. Note that, if we analyze carefully the reboiler duty
values it can be seen that these values increase with increasing the
stripping gas flowrate, but not significantly. Furthermore, if we
compared the reboiler duties from the stripping gas drying process
simulations with the results shown in Table 4 for the conventional
drying process simulation at the same temperature in reboiler
(respectively 200
C), it can be noticed that in the stripping gas
process, the increase in reboiler duties is less than 0.6%.
In Fig. 8 is shown the evolution of the water partial pressure and
TEG loss in overhead vapors with the stripping gas flowrate. It can
be seen that by increasing the stripping gas flowrate decrease the
water vapor composition in overhead flow and its partial pressure
(at constant top pressure). By comparing these results with the
water partial pressure values from Fig. 4 at 200
C in reboiler, it is
clearly proved that the main stripping gas effect is the drastically
decrease of the water partial pressure. Another positive effect of the
stripping gas process is to decrease the overhead temperature of
the regenerator. In consequence, the TEG losses in overhead vapors
will decrease.
It is not simply to choose the optimum operating way from
entire range of stripping gas flowrate investigated in this paper
section. Taking into account only the equilibrium dew point value
of the sale gas, it turns out that it is enough 41.664 Nm3/h stripping
gas flowrate to increase the TEG concentration from 98.74 to
99.22 wt%. On the other hand, the increasing in stripping gas
flowrate is currently a loss in gas sales flowrate and it is not an
economically option.
Table 5 presents the input and optimum results for stripping gas
drying process with TEG at 200
C in regenerator reboiler and at
41.664 Nm3/h stripping gas flowrate.
According to the simulation data from Table 5, at temperature of
200
C in regenerator of the reboiler and at 41.664 Nm3/h stripping
gas flowrate heated at 120
C it can be obtained a 99.22 wt% TEG in
 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340 335

Fig. 6. Variation of TEG concentration and equilibrium dew point of sale gases with stripping gas flowrate (at 120
C stripping gas temperature inlet in reboiler).

Fig. 7. Variation of the overhead vapor flowrate and the reboiler duty with stripping gas flowrate.

water. The calculated TEG circulation rate is 24.90 L TEG/kg water
absorbed. The water content of the outgoing gas was 50.64 Nm3/h
(or 47.97 mg/Sm3) which determines the sale gas dew point of
about 24.94
C under the pressure of 4070 kPa. From GPSA chart
was found that for water content of 47.97 mg/Sm3 and 4070 kPa the
equilibrium dew point is 18
C. A comparison between the con-
ventional drying process and the one with 41.664 Nm3/h stripping
gas shown clearly a decrease with about 10
C in sale gas dew point,
in the same operating conditions. Moreover, for 208.32 Nm3/h
stripping gas the dew point will decrease with about 35
C to the
value obtained in benchmark drying process.
Finally, some remarks will be done in order to explain our
choosing regarding the stripping gas temperature. Fig. 9 shows that
increasing stripping gas temperature from 120
C to 160
C (with an
increment of 10
C) has a very low effect on the overhead vapor
flowrates and on the reboiler duties. In our opinion, a stripping gas
temperature of 120
C is a reasonable option.
4.4. Results and discussions of cost calculation
In the conventional flowsheet, the main equipments included in
the cost plant calculation are: the absorber [ABS], the regeneration
336 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340

Fig. 8. Variation of the partial pressure of water vapor and TEG loss with stripping gas flowrate.

Table 5
Input and optimum results for stripping gas drying process with TEG at 200
C in regenerator reboiler and at 41.664 Nm3/h stripping gas flowrate.

Material streams

Wet gas to absorber Sales gas (after L/DG-HX) Lean TEG recycle Regenerator feed Overhead vapor

Vapor fraction 1 1 0 0 1
Temperature,
C 30.00 34.51 34.00 100.00 95.88
Pressure, kPa 4100.00 4070.00 4080.00 119.0 102.00
Mass flowrate, kg/h 31,635.30 31,594.12 1072.35 1110.443 69.96
Molar flow, kmoli/h 1862.00 1860.08 7.55 9.66 3.97
Volume flowrate,Nm3/h or m3/h 41,708.80 41,665.79 0.951,402 0.98972 88.93
Water flowrate, kg/h 39.98 2.11 8.256 46.09 37.81
Water dew point by Eq. (1) 24.94

Component molar fraction

Total hydrocarbons 0.998,808 0.9999369 e 0.0013923 0.4707366

Water 0.001192 0.0000631 0.0607195 0.2650235 0.5288744
TEG e 0.0000000 0.9387265 0.7335842 0.0003890

Component mass fraction

Total hydrocarbons 0.998736 0.9999131 0.0000003 0.0003576 0.4561700

Water 0.001264 0.0000669 0.0076994 0.0415093 0.5405160
TEG e 0.0000020 0.9922303 0.9581331 0.0033140

Heat flow, kJ/h Energy stream

Condenser Reboiler
5208.97 387,478.87
Pump energy, kJ/h 5608.81

column [REG], two vertical separators ([HC FLASH] and [GW]), one
cooler [COOLER], two cross heat exchangers ([L/R HX] and [L/DG
HX]), one static mixer [MIX-2], one centrifugal pump including one
electrical motor ([P-100]). In the capital cost of plant is considered
that one centrifugal pump (including its electrical motor) is in
warehouse spare. For the absorber, Eq. (4) is applied separately for
column shell with ellipsoidal heads and for bubble cap trays. For
the regeneration column cost calculation are divided in following
items: cost of column shell with ellipsoidal heads, cost of non-
structured packing, cost of condenser (assumed as plate and
frame type), and cost of U-tube kettle reboiler. Please note that the
vertical separator [GS] and mixer [MIX-1] from Figs. 1e2 are not
included in cost calculation because in practice, the absorber [ABS]
is direct fed with wet natural gas from pipeline. In our simulations,
the natural gas (flow and composition from Table 1) was synthetic
mixed in [MIX-1] with a proper water flow. Only wet gas with water
at equilibrium conditions enters to the bottom of absorber [ABS]
and excess water is removed from vertical separator [GS] as bottom
product.
In the stripping gas drying process, all supplementary items
from the conventional configuration will increase the investment
costs. In stripping gas configuration, the lean TEG/stripping gas
 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340 337

Fig. 9. Variation of the overhead vapor flowrates and the reboiler duty with stripping gas temperature.

cross heat exchanger ([L/SG HX]) and gas stream splitter unit [TEE], Table 7
are the main supplementary items. The fixed capital investment (FCI) estimation for conventional.and stripping gas
drying processes.
The above main plant equipments have been sized in help with
UniSim Design-R380 facilities in order to provide the size param- Cost, US$ Conventional drying process Stripping gas drying process
eter S included in the cost calculation, according Eq. (4). A detailed ISBL 1,169,176 1,189,872
procedure for equipment cost Ce is done in Appendix A. OSBL 350,753 356,962
Table 6 shown a comparison between the fixed capital costs of EC 455,978 464,050
CO 151,993 154,683
the dying processes.
FCI (2006 prices) 2,127,990 2,165,567
According to Eq. (2), the fixed capital investment (FCI) includes FCI (2016 prices) 2,416,451 2,459,226
the ISBL cost, the offsite cost (OSBL), the engineering costs (EC), and
contingency (CO), also. As is recommended in Towler and Sinnott
(2008) work, the OSBL is assumed to be 30% of ISBL and CO is
and [COOLER] items cost is due to the stripping gas flow path.
assumed to be 10% of ISBL plus OSBL. The overall FCI is done in
Fixed Capital Investment (FCI) is converted into a constant series
Table 7. The FCI is actualized to 2016 price in help with Eq. (5).
of payment for every year of the plant life, assuming that the plant
From Table 7 it can be seen that the fixed capital investment for
life is 20 years and the interest rate is about 15%. For this interest
stripping gas drying process is only 2% higher than benchmark
rate and recovery period, the annual capital charge ratio is 0.160, so
process. This is obvious by fact that in the stripping process only
the annual capital cost (ACC) is 0.160  2,416,451$ ¼ 386,632$/year
two supplementary equipments are required, respectively the lean
for conventional drying plant, or 0.160  2,459,226$ ¼ 393,476$/
TEG/stripping gas cross heat exchanger ([L/SG HX]) and gas stream
year for stripping gas drying plant. Above calculations were done
splitter unit [TEE]. The minor changes in vertical separator [GW]
by Eq. (6).

Table 6
The fixed capital cost C (or the inside battery limits investment ISBL) estimation for conventional and stripping gas drying processes.

Conventional drying process Stripping gas drying process

Equipment type Equipment cost Fixed capital cost C or Equipment cost Fixed capital cost C or
Ce, US$ ISBL, US$ Ce, US$ ISBL, US$
Absorber [ABS] 217,651 815,465 217,651 815,465
Regeneration column [REG] 13,329 50,418 13,329 504,18
Vertical separator [HC FLASH] 8999 35,996 8999 35,996
Vertical separator [GW] 1679 6716 8999 35,996
Heat exchanger [L/R HX] 14,144 49,504 14,144 49,504
Heat exchanger [L/DG HX] 14,830 51,906 14,825 51,886
Reboiler 19,311 67,590 19,279 67,477
Condenser [COND] 2080 7280 2080 7280
Cooler [COOLER] 14,716 51,506 15,018 52,563
Mixer [MIX-2] 1042 2605 1042 2605
Centrifugal pumps including electrical motors [P-100] (one centrifugal pump 8605 30,190 8605 30,190
in warehouse spare)
TEE e e 2169 5423
Heat exchanger [L/SG HX] e e 2080 7280
338 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340

In order to complete the comparison between two drying pro- dew point calculation based on empirical method proposed by
cesses, the annual costs of production COP are estimated, in the Gandhidasan (2003) is about 14.59
C under the pressure of
following step. The variable and fixed costs of the production are 4070 kPa. Therefore, at this dew point level, the natural gas quality
done for its components from Table 3. The operating labor cost is is not properly according to certain national networks norms. In
calculated for a typical 4.8 operators per shift position and 5 shift other words, the conventional drying process can be applied only in
position operators. The steam medium pressure cost not includes few specific cases. In opposition, the simulations of stripping gas
its production (fuel fired at boilers, etc). No process water con- drying process demonstrate clearly that TEG concentration in-
sumption is listed under variable cost of production because air- creases to 99.22e99.85 wt% for stripping gas flowrate in range of
cooling medium was assumed in plants operation. But, the water 41.664 Nm3/h to 208.320 Nm3/h. At the same time, equilibrium
is consumed in process washes and its amount is difficult to esti- dew points decrease, mainly because of the increasing of the TEG
mate. It can be assumed as that all the process water needs are concentration. For example, 41.664 Nm3/h stripping gas is enough
supported by internal recycles. The electrical energy consumption to obtained a 99.22 wt% TEG in water and a dew point of 24.94
C.
in drying plants was assumed as electrical energy for all pumps (i.e. No doubt, the natural gas quality was substantially upgraded.
recycling pump, reflux pump, water pumps), electrical energy for Moreover, the economic evaluation demonstrates that the fixed
air-cooling fans, electrical energy in control devices, energy for capital investment for stripping gas drying process is only 2%
compressors, electricity for lighting, etc. It is obvious that electrical higher than for benchmark process. This result is not surprising
energy consumption is very difficult to estimate. In our opinions, because in stripping process only two supplementary equipments
1000 kW/h electrical energy is a reasonable assumption. Table 8 are required, namely the lean TEG/stripping gas cross heat
shown the annual costs of production COP. exchanger and gas stream splitter units.
Finally, the total cost of production (TCOP) as sum of the annual Finally, our technological and economic analysis demonstrates
cost of production from Table 8 and the annual capital cost previ- that stripping gas is a more effective way to improve dehydration
ously estimated, is done by Eq. (8). The TCOP for conventional system performances.
drying plant is 3,223,975 $/year and the TCOP for stripping drying
plant is 3,216,669 $/year. As a quick check, it can see that ACC is Nomenclature
roughly 12% of the total cost of production, which is typical for
chemical processes, according Towler and Sinnott (2008). [ABS] absorber
[REG] regenerator
5. Conclusions [VLV-1, VLV-2] expansion valves
[HC FLASH] vertical two phases separators for rich TEG flow
In this paper, the performances of two gas drying processes: the [GW] vertical two phases separators for overhead vapor flow
conventional process with a simple regenerator as benchmark and [GS] vertical separator for excess water removal
the process with hot gas stripping injection in the reboiler of [L/R HX] lean/rich heat-exchanger
regenerator are re-examines both as technological feasibility and [L/DG HX] lean TEG/dry gas heat exchanger
economics. The gas drying flowsheets have been modeled by using [L/SG HX] lean TEG/stripping gas heat exchanger
a process simulator (UniSim Design R-380). According to the nat- [COOLER] overhead cooler
ural gas properties (composition, pressure and temperature), the [COND] condenser
water amount in wet gas is about 909 mg/Sm3 gas. A deep dehy- [MIX-2] TEG make-up
dration of natural gas requires high purity of the recycled TEG [TEE] splitter unit
solvent. In our work, was developed a systematic procedure to [RCY-1, RCY-2] logical operations
determine the operating conditions that maximize the regenerated [P-100] centrifugal pump
TEG concentrations. Initially, the technological performance of the FCI fixed capital investment
conventional drying process was investigated. To improve TEG TCOP total cost of production
purity, the temperature from regenerator was modified in range of ISBL inside battery limits investment
175
C to 204
C in successive process simulations. The maximum OSBL offsite cost
performance of regenerator was depicted at 200
C, namely a TEG EC engineering cost
concentration in regenerated solvent of 98.74 wt%. The equilibrium CO contingency cost

Table 8
Annual costs of production COP estimation for conventional and stripping gas drying processes.

Cost, US$ Conventional drying process Stripping gas drying process

Variable Cost of Production (VCOP)

TEG cost (including TEG loss) 18,739 8726
Steam cost 18,806 18,778
Electrical energy cost 800,000 800,000
Miscellaneous (consumable materials, waste water treatments, waste disposal) 16,751 16,429
Annual VCOP 854,296 837,901
Fixed Cost of Production (FCOP)
Operating labor (OL) 600,000 600,000
Supervision (S) 150,000 150,000
Direct salary overhead (DSO) 375,000 375,000
Maintenance (M) 39,832 40,537
Property taxes and insurance 26,554 27,024
Rent of land (and/or buildings) 17,260 17,566
General plant overhead 757,141 757,599
Environmental charges 17,260 17,566
Annual FCOP 1,983,047 1,985,292
 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340
C fixed capital cost
ACC annual capital cost
COP cost of production
VCOP variable cost of production
FCOP fixed cost of production
CEPCI chemical engineering plant cost index
Appendix A. Calculation for conventional gas drying process
For conventional gas drying process, the C fixed capital cost is
calculated according to Hand factor method and is exemplifies as
follow.
A.1. Cost of the vertical vessels
In conventional gas drying process, the vessels are: absorber
[ABS], regeneration column [REG], vertical separator [HC FLASH]
and vertical separator [GW]. All vessels are cylindrical with ellip-
soidal heads.
As can be seen, the S size parameter for each type of equipment
is required in Eq. (4). The size parameter of the vessel is the shell
mass of equipment. The mass calculation of vessels requires the
following parameters: internal diameter, external diameter, height
and density of the construction material. The density of 304
stainless steel is 8000 kg/m3. The internal diameters (as well the
height) of all vessels are given from UniSim simulations. The
external vessel diameters are calculated as internal diameters plus
two times wall thickness. The wall thickness is calculated as the
minimum wall thickness plus the corrosion allowance of 3.2 mm
and the reinforcement allowance of 8.1 mm. The minimum wall
thickness is calculated in accordance with ASME Boiler and Pres-
sure Vessel Code and equation (13.41) from Towler and Sinnott
(2008). In equation (13.41), the welded-joint efficiency is 1.0 and
the maximum allowable stress for stainless steel is obtained based
on data of Table 13.2 from Towler and Sinnott (2008) and on spe-
cific operating temperature and design pressure of each vessel.
Please note that the design pressure of the vessels is 10% above the
operating pressure. In this work the operating temperature and
pressure of vertical separators [HC FLASH] and [GW] are indicated
in section 2. Tables 5 and 6 show the operating temperature and
pressure of the absorber and the regenerator columns.
The S size parameter for bubble cap trays is the internal diam-
eter of absorber and those of Raschig rings (1/16 in. wall, CSteel,
1&1/2 in) is the volume of the packing. Likewise, the constant a and
b, and exponent n are required in Eq. (4). All of these are from table
6.6, Chapter 6 of Towler and Sinnott (2008). Table A1 shows the size
parameters of the vertical vessels and the results of their cost
calculation.
A.2. Cost of the heat exchangers
In conventional gas drying process, the heat exchangers are: [L/
R HX], [L/DG HX], [COOLER], reboiler and condenser of regenerator.
Table A1
Cost of the vertical vessels (price at year 2006).
Equipment type Size parameter
S Value
Vertical separator [HC FLASH] Shell mass, kg
Vertical separator [GW] Shell mass, kg
Absorber [ABS] Shell mass, kg
Bubble cap trays Diameter, m
Regeneration column [REG] Shell mass, kg
304 Raschig rings Volume, m3
339
The first three heat exchangers are tube and shell type, the reboiler
is Kettle type and the condenser is double-pipe type. From UniSim
Design simulations are known all heat exchanger duties Qr and
logarithmic mean temperature differences (LMTD), also. From
Coulson and Richardson (1999) the overall transport coefficients U
are known according to the heat exchanger type or its hot and cold
side streams. The heat transfer areas are calculated as Qr/U∙LMTD.
Table A2 shows the heat exchanger areas results. Please note that in
Table A2, the heat exchange area marked with * represent the
minimum area as size parameter for which a, b, n constants are
available.
Finally, the heat exchanger costs are calculated based on the
transfer areas as the size parameters in Eq. (4). All heat exchanger
costs are obtained for carbon steel and then corrected to stainless
steel with the material factor (fm ¼ 1.3). Table A3 shows the size
parameters of the heat exchangers, and the results of their cost
calculation.
A.3. Cost of the pumps and miscellaneous
Two single-stage centrifugal pumps (one in warehouse spare)
including two electrical motors and one static mixers [MIX-1] are
additional items in conventional drying plant. The size parameter
of the above equipments is the flow (in liters/sec), except the
electrical motors for which the size parameter is power (kW). The
flows and power are known from simulation results. All equipment
costs are obtained for carbon steel and then corrected to stainless
steel (except the electrical motors). Table A4 shows the size pa-
rameters of the pumps and miscellaneous items, and the results of
their cost calculation.
A.4. Operational costs
From the reboiler simulation the steam consumption is 196 kg/
h. By considering the price of medium pressure steam as 12 $/tons,
the annual steam consumption involves $18,806. The cost of steam
medium pressure does not include its production. No process water
consumption is listed under operational cost because air-cooling
medium was assumed in plants operation. The electrical energy
consumption in conventional drying plants is very difficult to es-
timate. In our opinion, 1000 kW/h electrical energy is a reasonable
assumption (including: electrical energy for all pumps, for air-
cooling fans, for control devices, for compressors, for lighting,
etc). By considering the price of electricity as 10 cents/kWh, the
annual electricity consumption involves $800,000.
A.5. Hand factors for fixed capital cost
The Hand factor method proposes that the C fixed capital cost of
a plant to be given as a function of the total purchased equipment
cost. The installation F factors required in Eq. (3) are shown in
Table A5.
Constants and exponent in Eq. (4) Equipment cost Ce, US$
a b n
316.9 10000 600 0.6 8999
140.8 10000 600 0.6 1679
17,919.18 10000 600 0.6 217,651
1.52 200 240 2.0
423.4 10000 600 0.6 13,329
0.196 0 3700 1.0
340 M. Neagu, D.L. Cursaru / Journal of Natural Gas Science and Engineering 37 (2017) 327e340

Table A2
The heat exchanger areas calculations.

Equipment type Reboiler duty Qr, kW Overall transport coefficients U, W/m2·C Logarithmic mean temperature differences (LMTD),
C Area, m2

Heat exchanger [L/R HX] 50.44 100 97.33 5.2 (10*)

Heat exchanger [L/DG HX] 77.47 200 24.25 16
Cooler [COOLER] 26.53 50 35.47 15
Reboiler 107.00 300 67.17 10.3
Condenser 0.488 50 100.00 0.10 (1*)

Table A3
Cost of the heat exchangers (price at year 2006).

Equipment type Size parameter Constants and exponent in Eq. (4) Equipment cost Ce, US$

S a b n

Heat exchanger [L/R HX] Area, m2 10,000 88 1.0 14,144

Heat exchanger [L/DG HX] Area, m2 10,000 88 1.0 14,830
Cooler [COOLER] Area, m2 10,000 88 1.0 14,716
Reboiler Area, m2 14,000 83 1.0 19,311
Condenser Area, m2 500 1100 1.0 2080

Table A4
Cost of the pumps and miscellaneous items (price at year 2006).

Equipment type Size parameter Constants and exponent in Eq. Equipment cost Ce, US$
(4)

S Value a b n

Mixer [MIX-1] Flow, liter/sec 0.265 780 62 0.8 1042

Centrifugal pumps [P-100] (one in warehouse spare) Flow, liter/sec 0.265 3300 48 1.2 8605
Electrical motors for pumps (one in warehouse spare) Power, kW 1.55 920 600 0.7 3471

Table A5 pipeline flow. J. Nat. Gas. Chem. 19, 261e266.

The installation F factors for Hand factor. Karimi, A., Abedinzadegan Abdi, M., 2009. Selective dehydration of high-pressure
natural gas using supersonic nozzles. Chem. Eng. Process 48, 560e568.
Equipment type Installation factor F Kohl, L., Nielsen, R.B., 1997. Gas Purification, fifth ed. Gulf Professional Publishing,
Huston.
Absorber [ABS] and Regeneration column [REG] 4.0 Kotarba, M., Nagao, K., 2008. Composition and origin of natural gases accumulated
Pressure vessels ([GW], [HC FLASH]) 4.0 in the Polish and Ukrainian parts of the Carpathian region: gaseous hydrocar-
Heat exchangers ([L/R HX], [L/DG HX], 3.5 bons, noble gases, carbon dioxide and nitrogen. Chem. Geol. 255, 426e438.
Reboiler, Condenser, [COOLER] Mokhatab, S., Wilke, R.J., Leontaritis, K.J., 2007. A review of strategies for solving
Pumps [P-100] 4.0 gas-hydrate problems in subsea pipelines. Energ. Sources, Part A 29, 39e45.
Miscellaneous ([MIX-1]) 2.5 Mokhatab, S., Poe, W.A., Speight, J.G., 2006. Handbook of Natural Gas Transmission
and Processing. Gulf Professional Publishing, United Kingdom.
Netusil, M., Ditl, P., 2011. Comparison of three methods for natural gas dehydration.
J. Nat. Gas. Chem. 20, 471e476.
References ObaniJesu, E.O., 2009. Modeling the H2S contribution to internal corrosion rate of
natural gas pipeline. Energ. Sources, Part A 31, 348e363.
Piemonte, V., Maschietti, M., Gironi, F., 2012. A triethylene glycolewater system: a
Carol, J., 2002. Natural Gas Hydrates: a Guide for Engineers, first ed. Gulf Profes-
study of the TEG regeneration processes in natural gas dehydration plants.
sional Publishing, United Kingdom.
Energ. Sources, Part A. 34 (5), 456e464.
Coulson, J.M., Richardson, J.F., 1999. Chemical Engineering In: Fluid Flow, Heat
Rahimpour, M.R., Jokar, S.M., Feyzi, P., Asghari, R., 2013a. Investigating the perfor-
Transfer and Mass Transfer, sixth ed., vol. 1. Butterworth eHeinemann, United
mance of dehydration unit with coldfinger technology in gas processing plant.
Kingdom.
J. Nat. Gas. Sci. Eng. 12, 1e12.
Demirbas, A., 2010. Methane Gas Hydrate (Green Energy and Technology). Springer-
Rahimpour, M.R., Saidi, M., Seifi, M., 2013b. Improvement of natural gas dehydration
Verlag, London.
performance by optimization of operating conditions: a case study in Sarkhun
Dow Chemical, Triethylene glycol, available from::http://msdssearch.dow.com/
gas processing plant. J. Nat. Gas. Sci. Eng. 15, 118e126.
PublishedLiteratureDOWCOM/dh_004d/0901b8038004d042.pdf?fil
Rouzbahani, A.N., Bahmani, M., Shariati, J., Tohidian, T., Rahimpour, M.R., 2014.
epath¼ethyleneglycol/pdfs/noreg/612e00004.pdf&fromPage¼GetDoc.
Simulation, optimization, and sensitivity analysis of a natural gas dehydration
Accessed 15.10.2012.
unit. J. Nat. Gas. Sci. Eng. 2, 159e169.
Gandhidasan, P., Al-Farayedhi, A., Al-Mubarak, 2001. A. Dehydration of natural gas
Scholes, C., Stevens, G., Kentish, S., 2012. Membrane gas separation applications in
using solid desiccants. Energy 26 (9), 855e868.
natural gas processing. Fuel 96, 15e28.
Gandhidasan, P., 2003. Parametric analysis of natural gas dehydration by a tri-
Sommer, S., Melin, T., 2004. Design and optimization of hybrid separation processes
ethylene glycol solution. Energ. sources.. 25, 189e201.
for the dehydration of 2-propanol and other organics. Ind. Eng. Chem. Res. 43,
Gas Processors Suppliers Association (GPSA), 2012. Engineering Data Book, thir-
5248e5259.
teenth ed. GPSA Press, Tulsa.
Towler, G., Sinnott, R., 2008. Chemical Engineering Design: Principle, Practice and
Gironi, F., Maschietti, M., Piemonte, V., 2010. Natural gas dehydration: a triethylene
Economics of Plant and Process Design. Butterworth-Heinemann Elsevier Ltd,
glycol-water system analysis. Energ. Sources, Part A 32, 1861e1868.
€yry Management Consulting, July 2011. Study on Interop- United Kingdom.
GL Noble Denton and Po
Wang, X., Economides, M., 2009. Advanced Natural Gas Engineering. Gulf Pub-
erability, Gas Quality Harmonisation, Cost Benefit Analysis. Preliminary Report
lishing Company, Houston.
for Consultation Report Prepared for the European Commission. http://ec.
Wen, C., Cao, X., Yang, Y., 2011. Swirling flow of natural gas in supersonic separators.
europa.eu/energy/gas_electricity/consultations/doc/20110916_cost_benefit_
Chem. Eng. Process 50, 644e649.
analysis_report.pdf. Accessed 13. 01. 2013.
Gong, J., Shi1, B., Zhao, J., 2010. Natural gas hydrate shell model in gas-slurry