You are on page 1of 12

Computers and Chemical Engineering 117 (2018) 117–128

Contents lists available at ScienceDirect

Computers and Chemical Engineering


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

Simultaneous optimization and heat integration of the coal-to-SNG


process with a branched heat recovery steam cycle
Bo Huang a, Yang Li a, Rui Gao a, Yongfei Zuo b, Zhenghua Dai a,∗, Fuchen Wang a,∗
a
Key Laboratory of Coal Gasification and Energy Chemical Engineering of Ministry of Education, East China University of Science and Technology, Shanghai
200237, China
b
SEDIN Engineering CO., Ltd., Taiyuan 030032, China

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

Article history: The coal-to-SNG process is an energy-intensive process, and optimizing the heat recovery network can
Received 9 October 2017 improve the economy and energy efficiency. This study proposes a branched, triple pressure level heat
Revised 28 January 2018
recovery steam cycle (HRSC) to recover waste heat, in which one branch is responsible for recovering the
Accepted 9 February 2018
waste heat from the water gas shift (WGS) unit, and the other branch is responsible for the methanation
Available online 22 February 2018
(METH) unit. The extended Duran–Grossmann model is used to optimize two heat exchanger networks
to match the branched HRSC superstructure. The temperature/pressure/flow rates of the HRSC streams
and the operating temperature of the WGS and METH units are optimized. The optimal bypass ratio of
the WGS unit as well as the recycle ratio and split ratio of the METH unit, are 0.506, 0.681 and 0.456,
respectively. The exergy efficiency of the coal-to-SNG plant is improved by 1.28% compared with the
industrial plant, which can reach 54.17%.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction One way to improve the waste heat utilization efficiency is to


conduct heat integration and optimize the heat recovery network.
Natural gas (NG) is cleaner and more energy-efficient compared There are generally two kinds of heat integration models. One
with other fossil fuels. NG demand has increased at an annual rate model is based on temperature intervals (heat cascade/problem
of 16.4% over the last five years (Li et al., 2016) due to increased table algorithms), and it is represented by the transshipment
urbanization and industrialization in China. In 2014, the annual de- model (Papoulias and Grossmann, 1983). The original transship-
mand of NG in China was 185.5 billion m3 , and 30% of the demand ment model requires the temperature intervals to be fixed, namely,
relied on importation (Shaikh et al., 2017), which indicates that the the interactions between the chemical process and the heat re-
energy security situation is dire. Producing coal-based synthetic covery network can only be realized by using discrete operating
natural gas (SNG) is a potential way to mitigate this energy se- temperatures for chemical processes and heat recovery streams.
curity problem due to the relatively rich coal reserves in China. Elia et al. (2010) have used this transshipment model to recover
The coal-to-SNG process consists of an air separation unit heat from a hybrid coal, biomass, and natural gas to liquid (CBGTL)
(ASU), a gasification (GSF) unit, a water gas shift (WGS) unit, a sour process through steam cycles. Martelli et al. (2011) developed a
gas removal (SGR) unit, and a methanation (METH) unit, among heat integration model that is based on the problem table al-
which, the GSF unit, WGS unit, and METH unit are strong exother- gorithm (Linnhoff and Flower, 1978). This model has been used
mic processes. Take the METH unit as an example: 20% of the en- to optimize the operating conditions of the HRSC in a coal-to-
ergy in the syngas is transformed into waste heat (Topsoe, 2009). SNG plant (Martelli et al., 2011) and a coal to Fischer–Tropsch fu-
Considering the large amount of waste heat, heat recovery has an els plant (Martelli et al., 2013). Although the operating conditions
important impact on the energy efficiency of the total plant. The (pressures, temperatures, and flow rates) of the HRSC streams are
waste heat in the chemical plant is usually recovered by raising optimized, the coal-to-SNG or coal to Fischer–Tropsch fuels pro-
steam or preheating boiler feed water (BFW) through a heat re- cess cannot be optimized. To exploit the heat recovery potential,
covery steam cycle (HRSC). Navarro-Amorós et al. (2013) proposed dynamic temperature in-
tervals to take variable temperature into consideration, which has
the potential to investigate the synergy between operating condi-
tions of process streams and heat recovery steam/water streams.

Corresponding authors. Kong et al. (2017) used dynamic temperature intervals to simul-
E-mail addresses: chinadai@ecust.edu.cn (Z. Dai), wfch@ecust.edu.cn (F. Wang).

https://doi.org/10.1016/j.compchemeng.2018.02.008
0098-1354/© 2018 Elsevier Ltd. All rights reserved.
118 B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128

Nomenclature turbine turbines in heat recovery steam cycle


recycle recycle in methanation unit
ASU air separation unit
Superscripts
BFW boiler feed water
in inlet
CBGTL hybrid coal, biomass, and natural gas to liquid
iso isentropic
CCS CO2 capture and storage system
out outlet
DNS dryness of steam
Pch pinch candidate
F molar flow rate of species in the stream
FCp heat capacity flowrate
FT total molar flow rate
GAMS general algebraic modeling system
taneously realize the chemical process synthesis and heat integra-
GSF gasification
tion.
GCC grand composite curve
Another heat integration model that can account for the inter-
HENs heat exchanger networks
actions between a chemical process and a heat recovery network is
HP high pressure
the Duran–Grossmann model (Duran and Grossmann, 1986), which
HRAT heat recovery approach temperature
is based on the pinch location method (PLM). This model is widely
HRSC heat recovery steam cycle
used in the field of simultaneous process optimization and heat
HRSC–CPP heat recovery steam cycle and captive power
integration. Baliban et al. (2011) conducted several investigations
plant
on heat and power integration for CBGTL process (Baliban et al.,
h enthalpy
2011, 2012, 2013). Recently, Yu et al. (2017) also applied the Duran–
K chemical equilibrium constant
Grossmann model to an organic Rankine cycles (ORC) system re-
QH heating utility
covering low-temperature waste heat in a refinery.
LP low pressure
The aim of this paper is to exploit the synergy between op-
LVH low-value heat
erating conditions of the coal-to-SNG process and the HRSC sys-
M mass flow rates
tem, and the most relevant work was done in the abovemen-
METH methanation unit
tioned Duran–Grossmann model-based papers. However, heat inte-
METH–HEN heat exchanger network in the methanation unit
gration in these papers occurred between all the cold/hot streams
MP medium pressure
from the total site, which is not reasonable because the location
NG natural gas
of each subsystem is relatively far from one another. Dowling and
NLP nonlinear programming
Biegler (2015) extended the Duran–Grossmann model to conduct
ORC organic Rankine cycle
heat integration of different groups of process units separately.
P pressure
In this paper, we will use the extended Duran–Grossmann
PC pinch candidate set
model to simultaneously realize separated heat integration within
QA heat absorbed by cold streams in heat recovery
the WGS unit and METH unit and heat integration between the
steam cycle
WGS unit and METH unit through water/steam integration. A
QC cooling utility
branched, triple pressure level HRSC is also proposed to recover
QR heat released by hot streams in heat recovery
waste heat from a coal-to-SNG process, with one branch corre-
steam cycle
sponding to heat recovery of the WGS unit, and the other branch
QSIA heat load of cold streams above the pinch candi-
corresponding to heat recovery of the METH unit. The operating
dates
conditions of the WGS unit and METH unit, as well as HRSC sys-
QSOA heat load of hot streams above the pinch candi-
tem, are optimized simultaneously. This paper is presented in the
dates
following order: the second section describes the coal-to-SNG pro-
SGR sour gas removal
cess, the third section provides the mathematical model, the fourth
SNG synthetic natural gas
section presents the results, and the last section summarizes the
SRW set of reactors restricted by water gas shift reac-
conclusions.
tion
SRM set of reactors restricted by CO methanation re-
action
SQA set of heat absorbers in heat recovery steam cy-
cle
SQR set of coolers in heat recovery steam cycle 2. Process description and research methods
T temperature
W power The flowsheet of the coal-to-SNG process is shown in Fig. 1
WHB waste heat boiler and can be generally described as follows: The ASU separates air
WWT wastewater treatment into nitrogen and oxygen. The steam, oxygen, and coal are then
WGS water gas shift fed to a gasification unit to produce syngas (CO + H2 ). The formed
WGS-HEN heat exchanger network in the water gas shift syngas is conditioned in the WGS unit to undergo the water gas
unit shift reaction, thereby meeting the H2 /CO ratio requirement. Af-
WGS-METH water gas shift and methanation unit ter removing H2 S and CO2 in the SGR unit, the conditioned syn-
ZH heat deficit gas is finally converted to CH4 rich gas in the METH unit. The
HRSC is used to recover waste heat and provide electricity (blue
Subscripts dashed line in Fig. 1). Apart from electricity, the HRSC also pro-
g equipment vides medium pressure (MP) steam (red line in Fig. 1) for steam
pump pumps in heat recovery steam cycle users and provides preheated BFW for the gasifier jacket and the
waste heat boiler (WHB) in the gasification unit.
B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128 119

Coal
GSF
Condensate
Condensate CO2,H2S Condensate
WHB

Gasifier
Air MPS Water Sour gas
Air Methanation
separtion gas shift removal
Waste water
Quench treatment
O2
LPS
SNG

BFW S2 S3 L1 E3 S4 E4 E5
E1 S1 E2
L1
BFW
Q Label
E1 L LP steam
S1 E2
S2 S MP steam
S3
MPS HRSC power E3
E4 E Electricity
S4 E5

Fig. 1. Flowsheet of the coal-to-SNG process. (For interpretation of the references to color in the text, the reader is referred to the web version of this article.)

Coal Gas cooler


E-104 E-103 E-102

Lock
syngas
Syngas
hopper
Quench liquor
LP steam out
M-102
Condensate

Quench S-101
Gasifier Waste heat
boiler out2
steam
E-101
R-100 M-101 R-101
BFW
BFW Raw gas Steam
Main reactor
Pre-reactor

Steam&O2 Ash Lock


Fig. 3. Flowsheet of the water gas shift unit.
Gas liquor
Ash
Fig. 2. Schematic diagram of a Lurgi gasifier. steam is raised in the WHB, and this LP steam is used to treat
wastewater from the GSF unit and provide heat for the SGR unit.
The rest of LP steam is directed into the LP steam pipeline. In this
2.1. Coal-to-SNG process
paper, the surplus LP steam is superheated and used to generate
power in the HRSC.
2.1.1. Air separation unit (ASU)
Cryogenic air separation is widely used to produce high pu-
rity oxygen, which is proven to be a cost-effective technology 2.1.3. Water gas shift (WGS) unit
(Ebrahimi et al., 2015). This process separates N2 and O2 by us- The WGS unit adjusts the CO/H2 mole ratio by converting a por-
ing their different liquefaction temperatures. The ASU is driven by tion of CO to H2 through the water gas shift reaction (1) to sat-
electricity in this study, and the purity of O2 is 99.6 vol%. isfy the H2 /CO ratio requirement for the methanation reaction, in
which the favorable H2 /CO ratio is approximately 3.1. The water
2.1.2. Gasification unit gas shift reaction is an exothermic and reversible reaction; thus,
The product gas from the fixed-bed gasifier contains approx- low temperature is favored for the forward direction.
imately 8 vol% (wet basis) CH4 , so the fixed-bed gasifier is one
CO + H2 O ↔ C O2 + H2 H298
0
= −41.1 kJ mol−1 (1)
of the most suitable technologies to produce coal-based SNG. The
schematic diagram of a Lurgi gasifier is depicted in Fig. 2. Lignite The flowsheet of the WGS unit without heat recovery is shown
coal is fed from the top of the gasifier through a lock hopper, and in Fig. 3. There are two serial reactors. R-100 is used to burn out
oxygen and steam enter the gasifier from the bottom of the gasi- O2 in the raw gas, and raw gas from R-100 is split in S-101 so that
fier. The MP steam generated from the water jacket is used as the one stream is fed to the main WGS reactor (R-101), while the other
gasification agent, but this amount of steam is not enough for gasi- part is bypassed. The bypass ratio is a key parameter for the WGS
fication, which accounts for approximately 10% of the steam de- unit and its definition is shown in Eq. (2), wherein F TS−101out is the

mand for gasification (He et al., 2013). Extra MP steam is extracted total molar flow rate at splitter S-101 outlet ‘out’ (see Fig. 3). The
from the steam pipeline, and in this paper, this extra MP steam operating temperature of the main WGS reactor is 473.15–773.15 K
is provided by extraction from the high pressure (HP) turbine or (Boll et al., 2011), and a sulfur-tolerant catalyst is used due to the
MP turbine in the HRSC. The syngas exiting the gasifier enters the existence of H2 S. The units E-101 to E-104 perform as coolers or
quench cooler, in which the syngas is saturated with steam and heaters to adjust the gas temperature to meet the operating re-
enters the waste heat boiler (WHB). Saturated low pressure (LP) quirements. In E-103 and E-104, part of the water vapor in the syn-
120 B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128

Gas cooler
E-211 E-210 E-209 E-208 E-207 E-206 E-205

SNG R-205 R-204 R-203


Condensate Condensate
5th Methanator 4th Methanator 3rd Methanator E-204
P-201
E-202
Steam
out
E-200 E-201 E-203
out2
M-201 R-200 M-202 R-201 S-202
R-202
Sweet gas out2
M-203
S-201
Sulfur guard st 2nd Methanator
out Steam1 Methanator

Fig. 4. Flowsheet of the methanation unit.

out
gas is condensed and removed. Finally, the syngas is cooled to ap- F TS−202
proximately 313.15 K and fed to SGR unit to remove H2 S and CO2 . RR = in
(5)
F TS−202
out out
F TS−101 F TS−201
RB = (2) RS = in
(6)
in F TS−201
F TS−101

2.2. Superstructure of the heat recovery steam cycles (HRSC)


2.1.4. Sour gas removal (SGR) unit
The production of a large quantity of CO2 in the shifted syngas 2.2.1. Separated heat integration and water/steam integration
from the WGS unit will affect the methanation reaction; simulta- In this paper, we propose a branched HRSC to recover waste
neously, H2 S will poison the methanation catalyst. As a result, the heat from the WGS and METH units in an integral manner. Further,
SGR unit is necessary for the normal operation of the METH unit. we will conduct ‘separated heat integration’ within each subsystem
The Rectisol process is widely used in chemical plants (Yu et al., (Dowling and Biegler, 2015) and simultaneously realize heat inte-
2012), in which chilled methanol is used to absorb H2 S and CO2 . gration between different subsystems through water/steam inte-
The energy consumption of the SGR consists of the electricity con- gration. As shown in Fig. 5, the ‘separated heat integration’ means
sumption for methanol refrigeration and the heat requirement for that the cold streams exchange heat with hot streams in the same
the reboiler in the regeneration tower. unit, namely, streams in the WGS unit do not exchange heat with
streams in the METH unit, and this configuration is reasonable as
2.1.5. Methanation (METH) unit the WGS and METH units are relatively far from one another. The
The function of the METH unit is to convert syngas into the CH4 heat integration between the WGS and METH units is realized by
rich gas via CO methanation (3) and CO2 methanation reactions water/steam integration through the branched HRSC.
(4). In the METH unit, reactions (1), (3) and (4) are considered. The unique feature of the HRSC is that it has two branches. The
upper branch is responsible for recovering the waste heat from the
−1
CO + 3H2 ↔ C H4 + H2 O H298
0
= −206 kJ mol (3) WGS unit, and the lower branch is responsible for recovering the
waste heat from the METH unit. Therefore, two separate heat ex-
changer networks (HENs) can be derived, namely, the WGS-HEN
C O2 +4H2 ↔ C H4 +2H2 O H298
0
= −165 kJ mol−1 (4) and the METH–HEN. It is worth noting that to simplify the model,
the mixing of branches in the HRSC is isothermal and isobaric. It
The TREMP (Topsoe, 2009) methanation process is employed for is important that the split ratios between the upper and lower
the methanation unit in this model. The normal operating tem- branches are optimization variables and can be different at each
perature range of methanation reactors is 523.15 to 973.15 K, and splitter; this is how the water/steam integration between the WGS
the catalyst can operate below its sintering temperature (approx- and METH units is realized.
imately 1023.15 K) (Blumberg et al., 2015). The flowsheet of the
METH unit in the base plant is shown in Fig. 4. 2.2.2. Branched HRSC with a captive power plant (HRSC–CPP)
In the METH unit, six reactors in a series are used. R-200 is the In the coal-to-SNG process, the largest steam consumer is the
sulfur guard, and R-201 and R-202 are the main reactors. R-203, Lurgi gasifier, which consumes a large amount of MP steam (5 MPa,
R-204, and R-205 are gas cleaners to convert residual CO and H2 723 K). However, the steam is insufficient, even if all the waste heat
to CH4 . To control the temperature of the reactors, part of the ef- is used to generate the MP steam; therefore, a captive power plant
fluent gas from R-201 is recycled to the inlet of R-201. Eqs. (5) and (CPP) is applied to provide extra steam. In this paper, HRSC and
(6) show the recycle ratio and the split ratio, respectively, as vital CPP are integrated to provide electricity and steam. The superstruc-
operating parameters in the METH unit. The effluent of R-204 is ture of the integrated HRSC–CPP is shown in Fig. 6, the HP steam
cooled to remove excess water, thus promoting the forward metha- from the CPP boiler is fed into the HP turbine of the HRSC. We as-
nation reaction. Finally, the effluent of R-205 is cooled to 313.15 K sume that the temperature and pressure of HP steam generated by
via 3 serial coolers and knocks out condensate water. the CPP boiler are the same as those generated by the HRSC.
B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128 121

WGS Process streams

WGS-HEN

Economizer Evaporator superheater

METH-HEN METH Process streams

Condenser

Fig. 5. Schematic of heat integration.

E-301 E-302 E-303

HP
P-301
Turbine T-301
E-402 E-403
E-401
HP Economizer HP Evaporator HP Superheater
Reheater

E-305 E-306 E-307 E-304 E-404

MP steam
user
P-302

E-406 E-407
E-405
MP
MP Economizer MP Evaporator MP Superheater T-302
Turbine
LP steam from
WHB

E-308 E-309 E-310

E-311 Boiler

P-303
E-411
E-409 E-410
E-408
LP Economizer LP Evaporator LP Superheater LP
T-303
Turbine

Deaerator coal
E-312

P-304 air
Demineralized water
E-412 Condenser
BFW

E-313
To water jacket

E-413

Fig. 6. Superstructure of branched heat recovery steam cycle (HRSC).


122 B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128

In this paper, a triple pressure level branched HRSC sys- 3.2. Coal-to-SNG process model
tem is used to exploit the heat recovery potential. Water
from the deaerator is pumped by the LP/MP/HP water pumps, An operating coal-to-SNG plant in northwest China is used as
and water/steam absorbs heat in the corresponding economiz- the base plant in this paper. The operating conditions of the WGS
ers/evaporators/superheaters. Finally, superheated steams enter unit, METH unit, and HRSC system are optimized to exploit the
the corresponding turbines to generate electricity. The exhausted heat recovery potential, while the operating conditions of the ASU,
steam from the LP turbine then enters the condenser. Part of the GSF unit, and SGR unit are fixed due to their relatively narrow op-
LP steam is extracted from the LP turbine and mixes with the con- erating windows. The mass balance and energy balance model for
densed water in the deaerator, thus meeting the operating tem- the WGS unit, METH unit, and HRSC are built rigorously, while the
perature of the deaerator (377 K). In addition, the MP steam is ex- operating data used for the ASU, GSF unit, and SGR unit are re-
tracted from the HP turbine or the MP turbine for MP steam users trieved from the base plant in northwest China. In the following
(including the GSF unit, WGS unit, and METH unit). The GSF unit paper, the WGS unit and the METH unit are combined into the
produces saturated LP steam in the WHB, and part of the LP steam WGS–METH section.
is consumed by wastewater treatment and SGR units. The remain-
ing steam is superheated and injected into the MP turbine or LP
turbine. 3.2.1. The operating data of ASU, GSF unit, and SGR unit
In this paper, only the electricity consumption of the ASU
3. Research methods is considered, which is 0.6 kWh/Nm3 O2 at a pressure of 5 MPa
(Li et al., 2014).
The overall model is formulated in the General Algebraic Mod- As for the gasifier in the GSF unit, Yining coal is used as feed-
eling System (GAMS). This model is a highly nonconvex NLP prob- stock, for which the mass flow rates is 1081 kg h−1 . Tables 1 and 2
lem. There are 30 0 0+ constraints and 30 0 0+ variables, in which list the properties of coal and the consumption amounts of oxygen
20 0 0+ constraints and 20 0 0+ variables are nonlinear. The state- and steam, respectively. The conditions of syngas exiting the gasi-
of-the-art global solvers, such as GAMS/BARON, GAMS/ANTIGONE, fication battery limit, or the source gas of the WGS unit, are listed
and GAMS/Lindoglobal, failed to get to the global optimal solution in Table 3. The saturated LP steam (0.6 MPa) generated in the WHB
within a reasonable time (100 h) for this model; therefore, a local is approximately 0.494 t steam/ t coal, of which 36.5% is used in
NLP solver, GAMS/CONOPT (Drud, 1994), is used to solve it, namely the wastewater treatment unit, and 35.5% is used in the SGR unit.
the solution is a local optimal solution. In this paper, we will ob- The remaining LP steam is superheated in E-311/E-411 in Fig. 6 and
tain the initial solution with initial values provided by an Aspen then injected into the MP/LP turbines.
Plus simulation, and the initial solution will be tuned in the tun- For the SGR unit, a simple input-output relation is used to sim-
ing procedure. The recycle ratio of the METH unit (Li et al., 2014) ulate this process. The gas recovery ratio taken from the base plant
and the pressures of HP/MP/LP steam are key manipulated vari- is shown in Table 4, namely, 100% H2 S and 97.55% CO2 is removed
ables that have a relatively large influence on the efficiency of the in the SGR unit.
coal-to-SNG process. Each key manipulated variable has been dis-
cretized into equally spaced values according to its bounds. In the
tuning procedure, the model has been solved iteratively. In each it- 3.2.2. Models of the WGS-METH section
eration, only one key manipulated variable is fixed at a discretized The mass balance and energy balance model of the WGS-METH
value and other variables are free, and the best-known solution is section is built to optimize the operating conditions. In the WGS-
kept and used as the initial value for the next iteration. Finally, METH section, there are mixers, splitters, coolers, heaters, flash
the best-known solution is used as the initial value for the model tankers and reactors. The models of the mixers, splitters, coolers,
without fixed key variables, and the final solution (best known) is and heaters are straightforward and are therefore neglected here,
reported in this paper. Each iteration of the model takes about two following are models of the reactor and flash tanker.
CUP seconds (Intel Core i7-4790 3.60 GHz). In the WGS unit, the water gas shift reaction (1) is considered.
This reaction is equilibrium restricted, and the approach temper-
3.1. Thermodynamic property model ature is 48 K in the WGS unit, which is determined by regressing
the base plant operating data.
In this paper, steam is treated as a real gas, while other gases In the METH unit, reactions (1), (3), and (4) are considered, and
are treated as ideal gases, and the gas mixture is treated as an only 2 of these 3 reactions are independent. Therefore, reactions
ideal gas mixture. (1) and (3) are modeled in the METH reactors, and the gas compo-
ALAMO (Cozad et al., 2014) and Aspen Plus are used to fit the sitions at the reactor outlet are close to an equilibrium state. The
enthalpy-temperature correlation for steam in the pressure range approach temperature is specified as 0.8 K in this paper, and the
of 0.1–125 bar and in the temperature range of 298.15–10 0 0 K. In heat loss is approximately 2.2% of the reaction heat, which are also
the proposed method, ALAMO generates 10,0 0 0 or more sampling determined by regressing the base plant operating data.
points (T, P) and calls the Aspen Plus model (which uses the Equilibrium constant correlations for reactions (1) and (3) are
IAPWS-95 property package) via Python to simulate and retrieve shown in Eqs. (8) and (9) Chein et al., 2016); P denotes the total
out stands for the molar flow rates of species i at the
pressure and Fg,i
simulation results (enthalpy values), thus building an enthalpy-
temperature correlation. The R2 can reach 0.999, and the maxi- outlet of equipment g. The set of reactors that are restricted by the
mum error is less than 2%. water gas shift reaction is defined as SRW = {R-101, R-201, R-202,
The steam enthalpy-temperature correlation generated by R-203, R-204, R-205}, and the set of reactors that are restricted by
ALAMO is shown below. The units of enthalpy, T, and P are the CO methanation reaction is defined as SRM = {R-201, R-202, R-
kJ kmol−1 , K, and MPa, respectively. 203, R-204, R-205}. The composition at the outlet of the reactors is
governed by Eqs. (10) and (11).
6.44 × 107
hH2O = − 585T + 0.176(T )2 + 3.53 × 105 ln (T )  
T 4400
− 4.43 × 103 PH2O − 37.72(PH2O )2 − 6.17 × 10−3 (T )2 PH2O
KgWGS = 1.767 × 10 −2
exp ∀g ∈ SRM (8)
Tgout
+ 4.23 × 10−2 T (PH2O )2 − 2.02 × 106 (7)
B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128 123

Table 1
Properties of raw coal.

Coal proximate analysis (wt%) HHV (ad) Ultimate analysis (wt%)

Mad Aad Vad FCad MJ/kg C H O N S


13.10 10.33 28.08 48.49 23.01 60.10 3.00 12.43 0.62 0.42

Table 2 The power output of the steam turbine is calculated by Eq. (13),
Gasification agent consumption.
ηturbine and ηgenerator are the isentropic efficiencies of the turbine
Gasification agent Value Quality and generator, which are assumed to be 0.8 and 0.96, respectively
Oxygen 207.57 Nm3 O2 /t coal Purity 99.6 vol% (Ensinas et al., 2007), and the set of turbines is defined as ST = {T-
MP steama 1.482 t steam/t coal 5.0 MPa, 723 K 301, T-302, T-303}. The HP and MP turbines are backpressure tur-
MP steamb 0.282 t steam/t coal 4.1 MPa, 520 K bines, while the LP turbine is a condensing turbine, and the steam
a
This extra steam is from the HRSC. dryness at the outlet of the LP turbine is greater than 0.9.
b
This steam is from water jacket. 
Wturbine = F Tgin · ηturbine · ηgenerator · (hin
g,H2O − hg,H2O )
iso
(13)
Table 3 g∈ST
Gas composition at the gasification battery limit.
DN ST −303 ≥ 0.9 (14)
Term Value

T/K 446.6 Heat absorbed by economizers, evaporators, superheaters and


P/MPa(A) 4 reheaters is calculated by Eq. (15), and the set of heat absorbers in
Flow rates/kmol h−1 100 the HRSC is defined as SQA = {E-301,…, E-313}∪{E-410,…, E413}.
H2 O 23.61
CO 13.28 Q Ag = F Tgin (hout
g,H2O − hg,H2O )
in
∀g ∈ SQA (15)
CO2 24.46
Composition/vol% H2 S 0.18 Heat released by the condenser is calculated by Eq. (16), and
H2 30.05 the set of heat coolers in the HRSC is defined as SQR = {condenser}.
N2 0.10
CH4 7.88
O2 0.23
Q Rg = F Tgin (hin
g,H2O − hg,H2O )
out
∀g ∈ SQR (16)
C2 H 6 0.21
The pressure drop between two pressure levels is restricted by
Equation (17).
Table 4
Performance of the SGR unit. PHP − PMP ≥ 0.5 MPa and PMP − PLP ≥ 0.5 MPa (17)
Components CO H2 CO2 CH4 H2 S The temperature change in the evaporator and condenser is as-
Recovery ratio/% 99.59 99.97 2.45 100 0 sumed to be 1 K so as to avoid the infinite heat capacity of streams
under phase change.

   2 3.3.2. Boiler of the captive power plant (CPP)


−26830 F Tgout
KgMET H = 1.198 × 10 23
exp · ∀g ∈ SRM The heat absorbed by the boiler is calculated by Equation (18),
Tgout Tgout
in which ηboiler is the boiler efficiency, which is fixed at 0.85
(9) (Ensinas et al., 2007) in this paper, Mfuel is the mass flowrate of
coal fed into the boiler, and LHV represents the lower heating
out out
· Fg,H2 WGS out out
∀g ∈ SRM value of fuel coal. The amount of HP steam generated in the boiler
Fg,CO O − Kg · Fg,CO 2 · Fg,H2 = 0 (10)
is calculated by Eq. (19).

(Fg,H2 ) · Fg,CO
out 3 out
− KgMET H · Fg,CH4
out out
· Fg,H2 ∀g ∈ SRM Qboiler = ηboiler · M f uel · LHV (18)
O = 0 (11)
The effluent of the WGS reactor (R-101) and the 4th and 5th  
METH reactors (R-204 and R-205) need to be cooled to a low tem- Qboiler = F Tboiler · hout in
boiler,H2O − hboiler,H2O (19)
perature so that water vapor in the syngas is condensed. In the
flash model, we assume dry gas does not dissolve in the conden- 3.3.3. Extended Duran–Grossmann model
sate, the gas stream at the flash tanker outlet is saturated with Pinch analysis (Linnhoff, 1982) is a powerful tool to perform
water vapor, and excess water vapor is condensed. heat integration and target the heating/cooling utility. The analysis
requires fixed inlet/outlet temperature and flow rates; therefore, it
3.3. Heat integration model is difficult to use this method to simultaneously optimize the pro-
cess with variable temperature and flow rates. However, Duran and
3.3.1. HRSC thermodynamic model Grossmann (Duran and Grossmann, 1986) had proposed a mathe-
The electricity consumed by the LP/MP/HP water pump can be matical model that used floating pinch temperatures so as to op-
calculated by Eq. (12), where ηpump is isentropic efficiency, which timize the process and heat integration simultaneously. Hence, the
is assumed to be 0.8 (Ensinas et al., 2007), hing,H2O
stands for the Duran–Grossmann model has been adopted.
specific enthalpy of water/steam at the inlet of equipment g, hiso
g,H2O
As the Duran–Grossmann model is based on pinch methods, it
is the outlet enthalpy when the compression process is isentropic, obeys the rule that the pinch point always appears at the inlet of
and the set of pumps is defined as SP = {P-301, P-302, P-303, P- streams when the heat capacity flowrate (FCp) is fixed; thus, the
304}. Duran–Grossmann model treats each of the inlet temperatures of
the hot and cold streams as pinch candidates. Duran and Gross-
 F Tgin
Wpump = (hiso
g,H2O − hg,H2O )
in
(12) mann had observed that the true pinch temperature exhibits the
g∈SP
η pump largest heating and cooling utilities among all pinch candidates.
124 B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128

Table 5 WGSF , WSGR , Wcompression are values of electricity consumption of the


Specific power consumption of auxiliary devices.
pumps in the HRSC, syngas recycle compressor in the METH unit,
Devices Specific consumption References ASU, GSF unit and SNG compressor, respectively. The electricity
ASU 0.6 kWh/Nm3 O2 (Li et al., 2014) balance is constrained by Eq. (26).
GSF 0.003 kWh/kg coal (He et al., 2013)
SGR 0.024 kWh/Nm3 syngas (Yin and He, 2011) Wauxiliary = Wpump + Wrecycle + WASU + WGSF + WSGR + Wcompression
(25)

The key point of Duran–Grossmann model is therefore to calculate


Wturbine − Wauxiliary = 0 (26)
the heating utility (the same as heat deficit (ZH Pch )) of each pinch
candidate, where Q SIAPch /Q SOAPch are the total heat loads of the 3.5. Overall models and objective
cold/hot streams above each candidate. The set of pinch tempera-
ture candidates is defined as PC = {Tgin | g∈coolers or heaters in the The overall NLP model can be formulated as shown in P1:
HEN}. The maximum heating utility (QH) is determined by Eq. (21).
ESNG
The actual minimum cooling utility is determined by the heat bal- Max OBJ =
Ecoal
ance Eq. (22), where(x)is the difference between the total heat
contents of the cold and hot streams. Constraint (23) is suitable s.t.m(x ) = Wturbine − Wauxiliary = 0
for the coal-to-SNG process as it has surplus heat to recover. To (x ) = 0
use NLP algorithms, max operators in Eq. (21) should be smoothed (x ) ≥ 0
by Eq. (24), whereɛis a small number such as 10−4 . For a detailed
description, the reader can refer to the original paper (Duran and ω (x ) ≥ 0 (P1)
Grossmann, 1986). The objective is maximizing the total exergy efficiency: ESNG
represents the exergy of SNG, and Ecoal stands for the exergy of
Z H Pch (x ) = QSIAPch (x ) − QSOAPch (x ) ∀Pch ∈ PC (20)
gasification coal and the fuel coal. Fuel coal is used to fire the cap-
tive power plant boiler. The function m(x) represents the electric-
QH = max{ZH Pch (x )} (21) ity balance of this process: the electricity consumption by auxiliary
devices is provided by the steam turbines with no surplus electric-
QC = (x ) + QH (22) ity to the power grid. The function (x) represents the energy and
material balance of the WGS–METH section and the HRSC system.
The function  (x)is the design specification of this process, such as
QH = 0 (23) the molar fraction of CH4 in SNG, which is set to 97.6% (the same
as the base plant) and the molar fraction of steam at the METH
reactors inlet, which is set to be greater than 14% to avoid carbon
max{0, ς } ≈ 0.5(ς + (ς 2 + ε )0.5 ) (24)
deposition. The functionω(x) represents the constraints of the ex-
In the original Duran–Grossmann model, there is only one set tended Duran–Grossmann model Eqs. (20)–((24)).
of cold and hot streams that exchange heat with each other. In
this paper, the extended Duran–Grossmann model Dowling and 4. Results and discussion
Biegler, 2015) is used with two sets of cold and hot streams. One
set of cold and hot streams corresponds to streams in the WGS– 4.1. Model validation
HEN (see Fig. 5); the other set corresponds to streams in the
METH–HEN (see Fig. 5). So there are two sets of constraints ((20)– A set of base plant data is used to verify the thermodynamic
(24). property model and WGS-METH model. The gas conditions are the
same as the data in Table 3; the feed gas entering the WGS unit is
3.4. Electricity balance 100 kmol h−1 .
In this paper, the operating conditions of the base plant are
The electricity generated by HP/MP/LP turbines is used to meet used to conduct a simulation with an optimization model (formu-
the electricity demand of all the auxiliary devices; the power con- lated in GAMS) to verify the thermodynamic property model and
sumption of auxiliary devices is denoted as Wauxiliary , and the spe- WGS-METH model. Table 6 compares some key parameters.
cific power consumption of the ASU, GSF unit and SGR unit are The temperature and compositions of the base plant and GAMS
listed in Table 5. The total power consumption of auxiliary de- model agree very well, which indicates that the optimization
vices is calculated by Eq. (25), in which Wpump , Wrecycle , WASU , model is reliable.

Table 6
Comparison between the base plant data and GAMS model results.

R-201a inlet R-201 outlet

Base Plant Optimization Model Base Plant Optimization Model

T/K 528.2 528.5 948.2 948.9


P/MPa 3.22 3.22 3.17 3.17
H2 O 14.16 13.98 24.41 24.19
CO 10.73 10.73 2.95 2.92
Composition/vol% CO2 3.13 3.12 4.61 4.62
H2 45.11 45.26 27.58 27.76
N2 0.22 0.22 0.26 0.26
CH4 26.50 26.53 40.18 40.25
C2 H6 0.15 0.15 0.01 0.00
a
R-201 is the first main methanation reactor.
B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128 125

Table 7
Operating conditions of the WGS-METH section.

R-101 R-101 R-201 R-201 R-202 R-202 R-203 R-203 SNG


inlet Outlet inlet outlet inlet outlet inlet outlet Product

P/MPa(A) 3.92 3.87 3.22 3.17 3.14 3.09 3.06 3.01 7.00
T/K 638.7 683.9 524.9 897.7 610.3 973.1 513.2 792.8 313.2
H2 O 23.40 18.70 14.00 23.70 10.60 18.70 18.70 29.50 0.30
CO 12.60 7.90 9.00 1.70 11.50 4.40 4.40 0.30 0.00
Composition/vol% CO2 25.30 30.00 3.30 4.50 2.90 4.60 4.60 3.30 1.10
H2 S 0.20 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.00
H2 30.50 35.20 38.30 20.70 44.50 29.20 29.20 11.60 0.40
N2 0.10 0.10 0.30 0.30 0.20 0.30 0.30 0.30 0.50
CH4 7.90 7.90 35.10 49.10 30.20 42.90 42.90 55.00 97.60

0.544
1000 Grand Composite Curve of the METH unit
0.542
Grand Composite Curve of the WGS unit
900
0.540

0.538 800

Temperature /K
Exergy efficiency

0.536 Split ratio


700
Bypass ratio
0.534
Recycle ratio
600
0.532

0.530 500

0.528 400
0.526
300
0.524
0 100 200 300 400 500 600 700 800
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Ratio Enthalpy/kW

Fig. 8. Grand composite curves of the WGS unit and METH unit without stream
Fig. 7. Influence of bypass ratio, recycle ratio and split ratio on exergy efficiency.
generation.

4.2. Optimization results Fig. 8 depicts the grand composite curves (GCC) of the process
streams within the WGS and METH units. There is approximately
In the following paper, the optimal operating conditions for the 412 kW/716 kW of waste heat that could be recovered or removed
WGS-METH section and the HRSC are described. in the WGS/METH units, respectively, when the heat recovery ap-
proach temperature (HRAT) is fixed at 10 K. It is obvious that the
waste heat of the METH unit concentrates in the high-temperature
4.2.1. Optimal operating conditions for the WGS–METH section region, while it concentrates in the low-temperature region for the
Table 7 shows the optimal operating conditions of the main re- WGS unit; for this reason, we integrate the heat of these two units
actors in the WGS unit and METH unit, and apart from the tem- to realize cascaded utilization of energy.
perature, pressure, and compositions in Table 7, the bypass ratio
of the WGS unit is 0.506, namely, 50.6% of syngas bypasses the 4.2.2. Optimal operating conditions for the HRSC
WGS reactor R-101. In this case, steam in the raw gas is sufficient The electricity balance of the HRSC–CPP is shown in Fig. 9, from
for the water gas shift reaction and no extra steam is needed. For which we can see that the electricity generated by the turbines
the METH unit, the optimal recycle ratio is 0.681, which is rela- is 202.9 kW, 66.36% of which is consumed by the ASU (including
tively low compared to the ratio of 0.85 reported in the literature O2 compression). The second largest electricity consumer is SGR,
(Li et al., 2014). Such a low recycle ratio leads to a high operating which accounts for 20.76% of total electricity consumption. The
temperature in the methanation reactors. The temperature at the electricity consumption proportion is similar to that of (Li et al.,
R-201 and R-202 outlets is 897.7 K and 973.1 K, respectively, and 2014).
high temperature is advantageous for the cogeneration of electric- Apart from the electricity balance, Fig. 9 also depicts the wa-
ity (Gassner et al., 2011). The optimal split ratio is 0.456, namely, ter balance of the HRSC–CPP. It is important to note that the flow
54.4% of the syngas enters R-201, and 45.6% of the syngas enters rates of the MP steam evaporated in the HRSC is 0, although the
R-202. Fig. 7 shows the influence of the bypass ratio, recycle ra- superstructure of the HRSC has 3 pressure levels. A CPP boiler with
tio, and split ratio on the exergy efficiency. When the bypass ra- a capacity of 61.89 kmol h−1 HP steam is installed to meet the
tio of the WGS unit is between 0 and 0.6, it has a weak influence electricity demand and MP steam demand of the system. To satisfy
on the exergy efficiency; however, when it exceeds 0.6, the exergy the MP steam demand of steam users, 102.54 kmol h−1 MP steam
efficiency decreases sharply. The influence of the recycle ratio on (5.0 MPa, 723.15 K) is extracted from the MP turbine, in which
exergy efficiency is relatively straightforward, and the exergy effi- 89.00 kmol h−1 , 0.21 kmol h−1 , and 13.34 kmol h−1 steam are sent
ciency initially increases and then decreases with an increase in to the gasifier, METH unit, and wastewater treatment unit, respec-
the recycle ratio. For the split ratio of the METH unit, the influence tively. To maintain the operation of the deaerator, 3.94 kmol h−1 of
on exergy efficiency is weak when it is between 0 and 0.5; how- LP steam (0.126 MPa, 379.53 K) extracted from the LP turbine is fed
ever, the exergy efficiency decreases gradually when it exceeds 0.5. to the deaerator. 142.01 kmol h−1 of demineralized water is heated
126 B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128

61.89 kmol·h-1
86.80%
GSF
47.19 kmol·h-1 0.00%
WGS
102.54 METH
kmol·h-1 0.20%
WWT
13.00%
-4.58 kW HP Turbine T-301

MPS 66.36%
ASU
0 kmol·h-1 17.9 kW
20.76%
SGR

0.00 kW 7.13% SNG


MP Turbine T-302 131.2 kW compr Boiler
8.77 1.60%
From WHB GSF
kmol·h-1
14.43 kmol·h-1 4.15%
53.8 kW Pumps

-0.02 kW coal
LP Turbine T-303

air
3.94 kmol·h-1

142.01kmol·h-1
Deaerator
25.79 kmol·h-1
Demineralized water
-0.10 kW

16.94 kmol·h -1 Condenser


61.89 kmol·h-1
BFW
To water jacket
31.3 kmol·h-1

To WHB

Fig. 9. Water and electricity balance of the HRSC–CPP.

Table 8
Operating conditions of the HRSC.

Pressure/MPa Tin/K Tout/K Total/kmol·h-1 WGS/% METH/%

HP economizer 12.50 377.15 600.72 47.19 9.50 90.50


HP evaporator 12.50 600.72 601.72 47.19 13.00 87.00
HP superheater 12.50 601.72 833.15 47.19 0.00 10 0.0 0
Reheater 11.07 814.65 832.72 109.07 0.00 10 0.0 0
LP economizer 0.30 377.15 406.49 14.43 10 0.0 0 0.00
LP evaporator 0.30 406.49 407.49 14.43 81.40 18.60
LP superheatera 0.30 407.49 446.61 14.43 10 0.0 0 0.00
LP superheaterb 0.60 432.15 491.88 8.77 10 0.0 0 0.00
demineralized water 0.20 318.15 373.16 142.01 10 0.0 0 0.00
BFW for jacket 4.50 377.15 423.15 16.94 10 0.0 0 0.00
a
Refer to the LP superheater (E-310/E-410 in Fig. 6).
b
Refer to the superheater (E-311/E-411 in Fig. 6) that heats the LP steam from waste heat boiler.

from 318 K to the deaerator temperature so as to maintain the wa- unit, and this portion of the HP saturated steam together with
ter balance of the HRSC–CPP, and 16.94 kmol h−1 BFW is heated to other HP steam generated from the METH unit is superheated in
423.15 K and fed to the water jacket of the gasifier. the METH unit.
The pressure, temperature, total flow rates and relative amounts
of steam/water generated by the WGS/METH unit are shown in 4.2.3. Composite curves
Table 8. The data of the 6th and 7th columns in Table 8 show the To illustrate the heat recovery effect, composite curves of the
distribution of water/steam in the WGS-HEN and METH–HEN. It WGS-HEN (see Fig. 10) and METH–HEN (see Fig. 11) are shown.
can be observed that waste heat from the WGS unit is used to gen- From these two graphs, it is obvious that the cold and hot stream
erate LP steam and preheat the demineralized water as well as the composite curves are very close, indicating that waste heat is re-
BFW for the water jacket of the gasifier, while the main function covered efficiently.
of the waste heat of the METH unit is to generate HP steam. The
4.3. Comparison with the base plant
difference between this configuration and the conventional config-
uration is that a portion of the saturated HP steam (13.00%) is gen-
To show the superiority of the proposed method, the heat re-
erated in the WGS unit and not generated in a conventional WGS
covery performance of the WGS-METH section between the base
B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128 127

Table 9
Comparison of heat recovery amount between base plant and optimized plant.

Case Streams P/MPa Tin –Tout /K Flow rates/kmol h−1 Exergy/kW

Base plant MP steam 5.49 378–736 42.20 352.62


LP steam 0.80 378–444 7.00
LP BFW 1 1.00 313–411 130.00
LP BFW 2 2.10 322–395 47
Optimized plant HP steam 12.50 377–833 47.19 437.36
Reheat steam 11.07 815–833 109.07
LP steama 0.30 377–447 14.43
LP steamb 0.60 432–492 8.77
Demineralized water 0.20 318–373 142.01
BFW for jacket 4.50 377–423 16.94
a
Refer to the LP steam raised in LP economizer/evaporator/superheater in the HRSC.
b
Refer to the LP steam from WHB that is superheated in the HRSC.

Table 10
Comparison of operating conditions between base plant and optimized plant.

Case Tin/K Tout/K Effluent/kmol h−1 Bypass ratio Recycle ratio Split ratio

Base plant R-101 549.9 618.4 37.48 0.624 0.550 0.420


R-201 528.4 948.9 50.67
R-202 534.9 948.8 37.96
R-203 603.2 809.1 34.59
Optimized plant R-101 638.7 683.9 49.20 0.506 0.681 0.456
R-201 524.9 897.7 58.69
R-202 610.3 973.1 36.74
R-203 513.2 792.8 32.50

plant and the optimized plant is compared. Table 9 lists the oper-
700 Hot Composite Curve of the WGS-HEN ating conditions of the HRSC streams in the base plant and opti-
Cold Composite Curve of the WGS-HEN mized plant. Syngas entering the WGS-METH section is the same
in these two plants, and SNG exiting the WGS-METH section of
600 the base plant and optimized plant is the same in terms of ex-
Temperature /K

ergy (4392.5 kW). The base plant recovers 352.62 kW of waste heat
in terms of exergy, while the optimized plant recovers 437.36 kW.
500 The heat recovery amount (in terms of exergy) can be improved by
24.03% by optimizing the operating conditions of the WGS-METH
section and the HRSC simultaneously. The exergy efficiency of the
400 coal-to-SNG process is 52.89% for the base plant, while it’s 54.17%
for the optimized plant. The exergy efficiency of the total plant has
been improved by 1.28%. The main difference of the HRSC between
300 the base plant and the optimized plant is that HP steam is not gen-
erated in the base plant (see Table 9), and the highest temperature
0 100 200 300 400 500 600 of the steam is 736 K in the base plant, while it is 833 K in the
optimized plant.
Enthalpy/kW
Table 10 compares the operating conditions of the main reac-
Fig. 10. Composite curves of the WGS–HEN. tors in the WGS-METH section. The main reactor of the WGS unit
(R-101) has been operated at a higher temperature in the opti-
1000
mized plant than that in the base plant, which gives the WGS unit
Hot Composite Curve of the METH-HEN the ability to generate HP steam. The relatively lower bypass ratio
Cold Composite Curve of the METH-HEN of the WGS unit ensures that the effluent of R-101 is larger (the
900
temperature is also higher), thereby improving the quality of waste
800 heat from the WGS unit. The recycle ratio of the METH unit in the
Temperature /K

optimized plant is larger than that in the base plant, which elim-
700 inates the steam requirements in the optimized plant. The split
ratio of the METH unit in the optimized plant is also larger than
600 that in the base plant, allowing the outlet temperature of R-202
to reach 973.15 K and with the higher temperature, to generate HP
500 steam.

400
5. Conclusions
300
The synergy between operating conditions of the coal-to-SNG
0 100 200 300 400 500 600 700 800 900 process and the HRSC system are exploited by simultaneously opti-
Enthalpy/kW mizing the operating conditions of the HRSC streams and the coal-
to-SNG process. In this paper, a novel, branched, HRSC superstruc-
Fig. 11. Composite curves of the METH–HEN. ture for heat recovery and power generation is proposed that is de-
128 B. Huang et al. / Computers and Chemical Engineering 117 (2018) 117–128

signed to avoid the heat exchange between process streams in dif- Cozad, A., Sahinidis, N.V., Miller, D.C., 2014. Learning surrogate models for simula-
ferent subsystems and to realize heat integration between different tion-based optimization. Aiche J. 60, 2211–2227.
Dowling, A.W., Biegler, L.T., 2015. A framework for efficient large scale equation-ori-
subsystems through water/steam integration. The extended Duran– ented flowsheet optimization. Comput. Chem. Eng. 72, 3–20.
Grossmann model is used, and it covers two sets of cold and hot Drud, A.S., 1994. CONOPT—a large-scale GRG code. ORSA J. Comput. 6, 207–216.
streams to match the branched HRSC superstructure; hence, the Duran, M.A., Grossmann, I.E., 1986. Simultaneous optimization and heat integration
of chemical processes. Aiche J. 32, 123–138.
operating conditions for the two HENs are determined simultane- Ebrahimi, A., Meratizaman, M., Reyhani, H.A., Pourali, O., Amidpour, M., 2015. En-
ously. ergetic, exergetic and economic assessment of oxygen production from two
The optimal exergy efficiency of the coal-to-SNG plant is 54.17%, columns cryogenic air separation unit. Energy 90, 1298–1316.
Elia, J.A., Baliban, R.C., Floudas, C.A., 2010. Toward novel hybrid biomass, coal, and
which is improved by 1.28% compared with the base plant. The op-
natural gas processes for satisfying current transportation fuel demands, 2: Si-
timal bypass ratio of the WGS unit, the recycle ratio of the METH multaneous heat and power integration. Ind. Eng. Chem. Res. 49, 7371–7388.
unit and the split ratio of the METH unit are 0.506, 0.681 and Ensinas, A.V., Nebra, S.A., Lozano, M.A., Serra, L.M., 2007. Analysis of process steam
demand reduction and electricity generation in sugar and ethanol production
0.456, respectively. It is shown that the recycle ratio of the METH
from sugarcane. Energy Convers. Manage. 48, 2978–2987.
unit and the split ratio of the METH unit have a greater influence Gassner, M., Vogel, F., Heyen, G., Maréchal, F., 2011. Optimal process design for the
on the exergy efficiency than the bypass ratio of the WGS unit. By polygeneration of SNG, power and heat by hydrothermal gasification of waste
analyzing the exergy recovered by the HRSC, it is found that the biomass: process optimisation for selected substrates. Energy Environ. Sci. 4,
1742–1758.
heat recovery amount (in terms of exergy) is improved by 24.03% He, C., Feng, X., Chu, K.H., 2013. Process modeling and thermodynamic analysis of
compared with the base plant, and this gain makes the major con- Lurgi fixed-bed coal gasifier in an SNG plant. Appl. Energy 111, 742–757.
tribution for the improvement of exergy efficiency. Kong, L., Avadiappan, V., Huang, K., Maravelias, C.T., 2017. Simultaneous chemi-
cal process synthesis and heat integration with unclassified hot/cold process
This branched HRSC superstructure, together with the extended streams. Comput. Chem. Eng. 101, 210–225.
Duran–Grossmann model, is suitable for the determination of oper- Li, S., Gao, L., Jin, H., 2016. Life cycle energy use and GHG emission assessment of
ating conditions of the chemical plant and heat recovery network. coal-based SNG and power cogeneration technology in China. Energy Convers.
Manage. 112, 91–100.
Li, S., Ji, X., Zhang, X., Gao, L., Jin, H., 2014. Coal to SNG: technical progress, modeling
Acknowledgments and system optimization through exergy analysis. Appl. Energy 136, 98–109.
Linnhoff., 1982. A User Guide on Process Integration for the Efficient Use of Energy.
Linnhoff, B., Flower, J.R., 1978. Synthesis of heat exchanger networks: I. Systematic
Financial support from the National Natural Science Foundation
generation of energy optimal networks. Aiche J. 24, 633–642.
of China under Grant No. 21776087, the National Key R&D Plan un- Martelli, E., Amaldi, E., Consonni, S., 2011. Numerical optimization of heat recovery
der Grant No. 2017YFB0602604, and the NSFC–DFG joint project steam cycles: mathematical model, two-stage algorithm and applications. Com-
put. Chem. Eng. 35, 2799–2823.
under Grant No. 2171101208 are gratefully acknowledged.
Martelli, E., Kreutz, T.G., Gatti, M., Chiesa, P., Consonni, S., 2013. Numerical optimiza-
tion of steam cycles and steam generators designs for coal to FT plants. Chem.
References Eng. Res. Des. 91, 1467–1482.
Navarro-Amorós, M.A., Caballero, J.A., Ruiz-Femenia, R., Grossmann, I.E., 2013. An al-
Baliban, R.C., Elia, J.A., Floudas, C.A., 2011. Optimization framework for the simulta- ternative disjunctive optimization model for heat integration with variable tem-
neous process synthesis, heat and power integration of a thermochemical hy- peratures. Comput. Chem. Eng. 56, 12–26.
brid biomass, coal, and natural gas facility. Comput. Chem. Eng. 35, 1647–1690. Papoulias, S.A., Grossmann, I.E., 1983. A structural optimization approach in process
Baliban, R.C., Elia, J.A., Floudas, C.A., 2012. Simultaneous process synthesis, heat, synthesis—III: total processing systems. Comput. Chem. Eng. 7, 723–734.
power, and water integration of thermochemical hybrid biomass, coal, and nat- Shaikh, F., Ji, Q., Fan, Y., 2017. An ecological network analysis of the structure, devel-
ural gas facilities. Comput. Chem. Eng. 37, 297–327. opment and sustainability of China’s natural gas supply system security. Ecol.
Baliban, R.C., Elia, J.A., Floudas, C.A., 2013. Biomass to liquid transportation fuels Indic. 73, 235–246.
(BTL) systems: process synthesis and global optimization framework. Energy Topsoe, H., 2009. From Solid Fuels to Substitute Natural Gas (SNG) using TREMP.
Environ. Sci. 6, 267–287. Halder Topsoe Technical Report.
Blumberg, T., Sorgenfrei, M., Tsatsaronis, G., 2015. Design and assessment of an IGCC Yin, A., He, L., 2011. Comparison of two rectisol processes. Nat. Gas Chem. Ind. 36,
concept with CO2 capture for the co-generation of electricity and substitute 53–55.
natural gas. Sustainability 7, 16213–16225. Yu, C., Huang, C., Tan, C., 2012. A review of CO2 capture by absorption and adsorp-
Boll, W., Hochgesand, G., Higman, C., Supp, E., Kalteier, P., Müller, W.D., Kriebel, M., tion. Aerosol Air Qual. Res. 12, 745–769.
Schlichting, H., Tanz, H., 2011. Gas production, 3. gas treating. Ullmann’s Ency- Yu, H., Eason, J., Biegler, L.T., Feng, X., 2017. Simultaneous heat integration and tech-
clopedia of Industrial Chemistry. no-economic optimization of Organic Rankine Cycle (ORC) for multiple waste
Chein, R., Yu, C., Wang, C., 2016. Numerical simulation on the effect of operating heat stream recovery. Energy 119, 322–333.
conditions and syngas compositions for synthetic natural gas production via
methanation reaction. Fuel 185, 394–409.

You might also like