Subscriber access provided by READING UNIV
Article
Comprehensive review of the design optimization of natural gas liquefaction processes: Current status and perspectives Muhammad Abdul Qyyum, Kinza Qadeer, and Moonyong Lee Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.7b03630 • Publication Date (Web): 17 Nov 2017 Downloaded from http://pubs.acs.org on November 18, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
1
Comprehensive review of the design optimization of natural gas liquefaction processes: Current status and perspectives Muhammad Abdul Qyyuma, Kinza Qadeera, and Moonyong Leea,* a
School of Chemical Engineering, Yeungnam University, Gyeongsan 712-749, Rep. of Korea
Submitted to: Special Issue in I & ECR on “PSE Advances in Natural Gas Value Chain.” *
Correspondence concerning this article should be addressed to:
Prof. Moonyong Lee
Email:
[email protected] Telephone: +82-53-810-2512
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 73
2
Abstract
Globally, liquefied natural gas (LNG) has drawn interest as a green energy source in comparison with other fossil fuels, mainly because of its ease of transport and low carbon dioxide emissions. However, LNG production is an energy and cost intensive process because of the huge power requirements for compression and refrigeration. Therefore, a major challenge in the LNG industry is to improve the energy efficiency of the LNG processes through economic and ecological strategies. Optimizing the design and operational parameters of the natural gas liquefaction cycles has been considered as one of most effective and popular approaches to address this issue. This paper reviews recent developments in the design optimization of LNG processes. In the choice of the most suitable and competitive LNG process, the operating costs, capital costs, environmental impact, and safety concerns must be considered for the optimal design and operation of LNG processes. The challenges in comparing recent researches are also discussed, along with suggestions for future directions to improve the energy efficiency of natural gas liquefaction processes.
Keywords: Review; LNG; natural gas liquefaction; Design optimization; N2-expander; Mixed-refrigerant; SMR; DMR; C3MR; Cascade.
1. Introduction
Globally, energy demand is increasing because of the growing population and improving living standards. According to the International Energy Agency, living standards and population growth will increase by around 30% between 2016 and 2040.1 The U.S. Energy Information Administration has projected that global energy consumption will grow by 48% between 2012 and 2040.2 To fulfil this projected energy demand, natural gas (NG) has become an attractive energy source in comparison with other fossil fuels because of its
ACS Paragon Plus Environment
Page 3 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
3
lower CO2 and other air pollutant emissions. For example, for the same amount of electricity generated using NG, half the greenhouse gas emissions and less than one-tenth of the air pollutants are produced compared to using coal as a fuel.1 Therefore, compared to the demand for coal and oil, that for NG is increasing rapidly.3, 4 It is also expected that investment to find new NG reserves across the globe will increase in the next few decades.
To date, NG reserves have been found in remote areas. Thus, there is a huge geographical mismatch between NG reservoirs and customer location. This geographic mismatch is a major reason for the increase in global NG trade. In NG trading, an economically feasible transportation method is a crucial factor. Generally, NG is transported from remote areas to the customer in the form of gas (through pipelines) or liquid (liquefied natural gas). NG is converted into a liquefied natural gas (LNG) at a cryogenic temperature of –160 °C and atmospheric pressure by shrinking its volume by a factor of approximately 600. This reduction in volume makes the transportation of NG over long distances as a liquid preferable for several reasons, including economic, technical, political, and safety factors. In 2016, global LNG trade reached a record value of 258 million tonnes (MT), which is 5% higher than the 2015 LNG trade value.5 In contrast, in the previous four years, the global LNG trade growth rate was only 0.5%. Furthermore, by 2020, Shell has projected that the global LNG trade will grow by 50%.1 However, the high cost involved in the LNG value chain is the major issue associated with LNG trading. If this high cost could be reduced, the growth rate of global LNG trade (i.e., 50%) could increase dramatically.
Generally, the LNG value chain (also known as a supply chain) consists of the exploration, production, liquefaction and refrigeration, shipping, regasification, and storage. The total LNG project cost is calculated by including the individual cost of each step involved in the supply chain. The average cost breakdown of LNG project by expense category5, 6 is shown schematically in Figure 1.
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 73
4
Figure 1. Average cost breakdown of LNG projects by expense.5, 6
As shown in this figure, liquefaction and refrigeration account for around 42% of the total LNG project cost. This cost value varies from 40 to 60% depending on the LNG plant site conditions and the type of available liquefaction processes, such as N2 expander, single mixed refrigerant, or propane-precooled mixed refrigerant processes. These available liquefaction processes are used depending on specific requirements from small-scale to large-scale LNG production, and each LNG process is unique in terms of the degree of complexity, capacity, environmental impacts, and energy efficiency.
As shown in Figure 1, a major issue associated with LNG plants is the high energy consumption, as reported in the most recent survey of International Gas Union (IGU) 5. Therefore, LNG plants are considered as energy intensive facilities. Much energy is wasted because of the non-optimal design and operating variables, which contribute to process irreversibility. However, the energy efficiency of an LNG process for a
ACS Paragon Plus Environment
Page 5 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
5
given structure can be improved significantly by efficient optimization. Therefore, any small improvement in the performance of an LNG process through optimization or design modification can increase the global competitiveness of the process and enhance the economic benefits by reducing the energy consumption. Among several ways to enhance the energy efficiency of LNG plants, optimization algorithms can be used, and numerous algorithms, both deterministic and stochastic, have been used to optimize the LNG process. In addition, optimization techniques based on process knowledge (thermodynamics) have also been used to improve the energy efficiency of LNG plants.
In the past few years, many new optimization techniques with improved searching, which can be implemented to optimize the design of complex thermodynamic and nonlinear processes, have been developed and applied the processes involved in LNG processing in the academic research community. Whereas, these advanced optimization techniques have still not been widely applied and implemented to optimize the design of LNG processes in real LNG industry because of the information gap between process design practitioners and academic researchers, which leads to the need for a review regarding the design optimization of LNG processes from a practical and research viewpoint. This is the first review study that is specifically intended to give an overview of the optimal design for LNG processes that have been optimized using different optimization strategies. Furthermore, this review is organized in a way to provide literature references for design engineers regarding optimization techniques for LNG processes. Using this review, we hope that the research community can better align their research directions and goals according to the requirements of LNG process design engineers. In this study, performance enhancements based on the use of different configurations alone are not considered. However, great efforts have been exercised to analyze as much important and relevant literature as possible.
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 73
6
2. LNG production processes
Based on the refrigerants and refrigeration cycle units used (e.g., Joule–Thompson (JT) valve and expander), LNG processes are categorized based on the type of refrigeration cycle:
I. II.
N2-expander-based processes Mixed-refrigerant-based processes
The main distinguishing characteristics of these LNG processes are the energy efficiency, environmental impact of the refrigerants, capacity, and degree of complexity. These processes will be briefly explained and analyzed in the following sections.
2.1.
N2-expander-based LNG processes
The selection of key design parameters for an offshore LNG plant, in terms of the process flexibility, compactness, process safety, simplicity, and LNG and make-up refrigerant storage, is different from that of an onshore liquefaction plant. In several studies,7–9 N2-expander-based LNG processes have been reported and developed as the most promising candidates for offshore applications. In N2 expander based LNG processes, pure N2 is used as a refrigerant in a single refrigeration loop/cycle to liquefy the NG. The N2 refrigerant is maintained in the gaseous phase throughout the refrigeration loop; thus, sensible heat is transferred at a gliding temperature level. The N2 expander refrigeration cycle (also known as the reverse Brayton cycle) mainly consists of multistage compression units equipped with an inter-stage cooling system (air- or water-based). High-pressure N2 gas is used as the coolant by utilizing a turbo-expansion system. Thermodynamically, isentropic, rather than isenthalpic, expansion is involved in N2-expanderbased LNG processes. Besides the isentropic expansion, during expansion, some useful shaft work is produced as a bonus, and this can be integrated to overcome the compression power load. This bonus work is another positive point of N2 expander LNG processes. The most basic N2-expander-based LNG
ACS Paragon Plus Environment
Page 7 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
7
process is known as the “N2 single expander LNG process,” and a schematic of this process is shown in Figure 2.
Figure 2. Schematic diagram of the N2 single expander LNG process.10 This basic N2 single expander process has been used in several optimization-based research studies7, 9– 12
to improve the energy efficiency further by applying only optimization techniques or by modifying or
adding refrigeration cycle units. For example, the N2 single expander LNG process has been improved by introducing one more independent turbo-expansion system,10, 13 as shown in Figure 3. This process is known as the “N2 dual expander LNG process.” The main concept behind this process is that expansion occurs at two different pressures by splitting the refrigerant stream (N2) into low- and high-pressure streams. This splitting of N2 at two different pressures reduces the required amount of energy at a small capital cost (i.e., the extra turbo-expander). Nevertheless, the high energy efficiency, arising from the near reversible operation, compensates for the additional capital cost of the expander.
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 73
8
Figure 3. Schematic diagram of the N2 dual expander LNG process.10
The energy efficiencies of N2-expander-based LNG plants using either single expander or dual expander processes have been further improved by introducing propane/carbon dioxide precooling refrigeration cycle into the conventional N2 expansion process, as shown in Figure 4.
ACS Paragon Plus Environment
Page 9 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
9
Figure 4. Schematic diagram of the CO2/C3/R410a precooled N2 expander LNG process.7, 9, 14
Precooling cycles allow the exergy loss in the main cryogenic exchanger to be minimized by reducing the temperature gradient. Furthermore, several researchers have successfully improved the energy efficiency of N2 expander based LNG processes through optimization techniques alone or improvements to the refrigeration cycle, as discussed in Section 4.
2.2.
Mixed refrigerant based LNG processes
This type of LNG process uses a mixture of refrigerants in one or more loop refrigeration cycles. Mixed refrigerants were first used in the 1960s in large-scale natural gas liquefaction.15 The components of conventional single mixed refrigerant (MR) are given in Table 1. Accordingly, with the optimal mixture of components with different volatilities, a high specific refrigeration effect (SRE) can be achieved at a relatively low operating pressure. 15
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 73
10
Table 1. Specific refrigeration effect (SRE) of the pure refrigerants.15 Operating pressure (bar) for
Pure Component
Normal boiling point (℃)
Nitrogen
−195.806
246.0
Methane
−161.49
63.5
Ethane
−88.6
16.5
Propane
−42.04
7.9
SRE of 1 kJ/mole
The high critical temperature or a high SRE and a high dew point temperature is achieved by using high boiling components, whereas low refrigeration temperature or low bubble point is obtained from the use of low boiling components in the mixture of refrigerants. Furthermore, in comparison with pure refrigerant (e.g., nitrogen) or binary mixture of refrigerant such as nitrogen-methane, the bubble point of the refrigerant can be increased near to the subcooling temperature of feed natural gas by the addition of high boiling components (less volatile), which gives SRE as same to pure refrigerant at lower operating pressure.
Mixed-refrigerant-based LNG processes have higher efficiency in comparison with N2-expanderbased LNG processes because two types of heat transfer (sensible and latent) occur, allowing close matching of cold (mixed refrigerant) and hot (feed natural gas) composite curves inside the cryogenic heat exchangers, even in a single stage. Further potential benefits of MR in terms of high energy efficiency and large capacity LNG trains can be achieved by using a mixed-refrigerant-based LNG processes, and these have been developed with different configurations of refrigeration cycles, for example, the addition of one or more refrigeration cycles. Based on the type of mixed refrigerant and number of refrigeration cycles, MR based LNG processes can be further categorized into the following well-established processes:
•
Single mixed refrigerant (SMR) process
ACS Paragon Plus Environment
Page 11 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
11
•
Korea single mixed refrigerant (KSMR) process
•
Propane-precooling mixed refrigerant (C3MR) process
•
Dual mixed refrigerant (DMR) process
•
Cascade process
•
AP-X process
These all process are briefly explained in the following section. 2.2.1.
Single mixed refrigerant (SMR) process
In the early 1970s, Black and Veatch introduced the SMR process. Figure 5 shows a basic schematic diagram of the conventional SMR process. This process is also known as the poly-refrigerant integrated cycle operation (PRICO) process. The SMR process mainly consists of a cryogenic multi-stream exchanger (plate-fin), expansion (or JT) valves, compressors, and coolers (air or water).
Figure 5. Schematic diagram of single mixed refrigerant LNG process.16
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 73
12
In a typical SMR process, NG enters the LNG heat exchanger at elevated pressure and ambient temperature and exchanges the latent heat of vaporization with the MR. NG leaves the exchanger as a subcooled (stream-2) liquid and is then flashed to a pressure slightly higher than atmospheric pressure through the JT valve. LNG is obtained in stream-3. In a mixed refrigerant loop, the MR stream is compressed at high pressure through multiple compressor stages, each equipped with after-coolers. After lowering the pressure through the JT valve, the MR (stream-5) is vaporized inside the LNG heat exchanger and leaves as a superheated (stream-6) vapor for recompression, thus completing the cycle.
The SMR process has lower energy efficiency as compared to cascade, DMR, and C3MR processes; nevertheless, SMR is less complex and involves lower capital investment. The optimization of refrigerant composition and operating pressure can enhance the performance of the SMR process by reducing the temperature gradient between the refrigerant stream and NG stream composite curves in the cryogenic heat exchanger,17, 18 as discussed in Section 4. In terms of energy efficiency, the SMR process has also been improved by modifying the refrigeration cycle configuration, such as in the KSMR process.
2.2.2.
Korea single mixed refrigerant (KSMR) process
This is an improved version of the SMR process, which was introduced by the Korea Gas Corporation (KOGAS). The simple design of the SMR process was used as a base design to improve the energy efficiency by splitting the MR stream into more and less volatile components at two different pressures.19, 20
The process flow diagram of the KSMR process is shown in Figure 6.
ACS Paragon Plus Environment
Page 13 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
13
Figure 6. Schematic diagram of Korea single mixed refrigerant LNG process.19
Initially, two streams are generated from a single MR stream at two different pressures, and only the MR stream, which is required to liquefy NG, is compressed at high pressure. The remaining stream is compressed at low pressure, which is sufficient to achieve the desired precooling of NG. The splitting/flashing of the MR stream means that only the required portion of less volatile MR expands through the JT valve at low pressure. In contrast, the more volatile MR stream expands at the intermediate pressure level. By using the two different pressure level streams to liquefy NG into separate cryogenic heat exchangers, the KSMR process is more efficient than the DMR process.20
2.2.3.
Propane-precooled mixed refrigerant (C3MR) process
C3MR technology is licensed by Air Products and Chemicals, Inc. (APCI) and is considered one of the dominant technologies for base load LNG production. Because of its high energy efficiency, a base
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 73
14
load of about 81% LNG can be produced using the C3MR process.21 The major highlight of this process is that two refrigeration cycles are used to produced LNG: a pure propane (C3) refrigeration cycle and a mixed refrigerant (MR)-based refrigeration cycle. The propane cycle is used to precool the NG and partially liquefy the MR stream. Then, the liquefaction and subcooling of the NG stream are carried out in the MR cycle. The process is shown schematically in Figure 7.
Figure 7. Schematic diagram of propane precooled mixed refrigerant LNG process.22
The SMR and C3MR processes are similar in terms of the MR cycle, but the degree of complexity of C3MR is high compared to that of the SMR process. The same type of cryogenic heat exchanger (plate and fin brazed with an aluminum core) is used in both the SMR and C3MR process.22 Because of the high degree of complexity, the C3MR process is less attractive for offshore LNG production. The performance of the C3MR process is better than other available LNG processes because of its ability to match hot (NG condensation curve) composite and cold (MR boiling curve) composite curves.
ACS Paragon Plus Environment
Page 15 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
15
2.2.4.
Dual mixed refrigerant (DMR) process
The DMR process for LNG production was introduced in 1978. In the first designs of the DMR process, the high-temperature MR components were boiled at different pressure levels with inter-stage compression.23 Because of this boiling, multiple zone heat exchangers, vessels, and valves are used. The DMR process uses two loops of refrigeration cycles, in which mixed refrigerants are circulated to liquefy the NG, as shown in Figure 8.
Figure 8. Schematic diagram of the dual mixed refrigerant LNG process.23
Actually, the DMR process was developed to remove the propane compressor bottleneck in the C3MR process.21 In this process, a warm mixed refrigerant (WMR) composed of methane, ethane, propane, and butane is introduced to precool the NG stream and partially liquefy the cold mixed refrigerant (CMR) rather than the propane precooling cycle used in the C3MR process. This process also significantly
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 73
16
reduces the propane inventory, making this process attractive for floating LNG (FLNG) production. The dual MR loop produces LNG efficiently, even in remote and cold areas.23 Other than design optimization, recent developments in the DMR process involve the use of fewer heat exchangers, a three-stage expansion in the precooling section, and the direct utilization of the available cold energy to cool down the boil-off gas.
2.2.5.
Cascade process
Cascade LNG processes are mainly categorized into two types with respect to the refrigerants: one based on pure refrigerant, known as the pure refrigerant cascade LNG process, and the other, the mixed refrigerant (MR) cascade LNG process. The pure-refrigerant-based cascade process uses methane, ethylene, and propane to liquefy NG more efficiently and was first introduced by ConocoPhillips in 1969.21, 24 The Kenai LNG plant with 1.4 million tonnes per annum (MTPA) capacity uses the original version of this pure-refrigerant-based cascade process. The major feature of this process is that propane cycle is used to precool the NG stream and condense the ethylene refrigerant. Then, the ethylene refrigeration cycle is used to condense both feed NG stream and methane in the third refrigeration loop. This condensed methane is then employed to subcool the LNG. The process flow diagram of the pure-refrigerant-based cascade LNG process is shown in Figure 9.
ACS Paragon Plus Environment
Page 17 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
17
Figure 9. Schematic diagram of the pure-refrigerant-based cascade (ConocoPhillips optimized) LNG process.25
The operational advantages of pure-refrigerant-based cascade processes include the easy to understand process, simplified training, simple control system, and the ability to shift heat load to optimize driver power or refrigeration loops.24 Furthermore, this process can be used both plate-fin and kettle type heat exchangers with respect to train size. The largest train size utilizing this process is approximately 5.2 MTPA with a specific power requirement range of 0.29–0.35 kWh/kg LNG.
In contrast, MR-based cascade LNG processes use mixed refrigerants to provide the refrigeration load in each precooling, liquefaction, and subcooling section of the LNG production process. This process has a higher energy efficiency in comparison with the pure-refrigerant-based cascade process because of its ability to closely match the MR and NG composite curves inside the cryogenic heat exchanger. This process has been installed in northern Norway in the Snøhvit LNG plant, which has a single train capacity of 4.3 MTPA. The major disadvantage of this process over the pure-refrigerant-based cascade process is
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 73
18
that it requires the optimization of the MR composition whenever feed gas composition fluctuates. The MR-based cascade LNG process is shown schematically in Figure 10.
Figure 10. Schematic diagram of mixed-refrigerant-based cascade LNG process.26
2.2.6.
AP-X process
The AP-X processes were developed by APCI using the C3MR process as a base design. These AP-X processes are also classified into two types: AP-N™ and AP-HN™.27, 28
The AP-N™ process was developed by adding a nitrogen refrigeration loop to the C3MR process, which increases the base-load plant capacity by more than 50%.27, 28 This nitrogen refrigeration loop provides a refrigeration effect equivalent to a 1–2 MTPA capacity LNG plant. The AP-N™ process has
ACS Paragon Plus Environment
Page 19 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
19
been demonstrated by the installation of six trains in Qatar, which are currently in operation.21, 27, 28 The AP-N™ process uses a propane cycle for precooling, an MR cycle for liquefaction, and a nitrogen refrigeration cycle for subcooling. For the subcooling process, the nitrogen is compressed through compressors and introduced into an economizer, where it is cooled. The major portion of this nitrogen is withdrawn at an intermediate point; then, this nitrogen undergoes isentropic expansion through a turboexpander, generating a sufficient refrigeration effect and liquefying the natural gas. The remaining portion of the nitrogen is further cooled and used for subcooling purposes, as shown in Figure 11. Because of the non-flammability of nitrogen, this process is a promising candidate for offshore LNG production. However, the energy efficiency of this process is lower than MR precooled liquefaction processes.
Figure 11. Schematic diagram of the AP-N™ LNG process.27
The AP-HN™ process was developed to improve the overall energy efficiency of the AP-N™ process because, in the AP-N™ process, nitrogen is also used in the precooling process at higher
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 73
20
temperatures, which leads to lower energy efficiency. Therefore, a second refrigerant loop was introduced for precooling. This improves the efficiency of the process, making it equivalent to that of the SMR process.27 This process is illustrated in the process flow diagram in Figure 12.
Figure 12. Schematic diagram of the AP-HN™ LNG process.27
A hydrofluorocarbon (HFC) refrigerant is used as a second loop refrigerant. This is preferred to propane because HFCs are non-flammable. However, there are some environmental and cost concerns involved in the refrigerant make-up process during FLNG operation.
3. Criteria for LNG process selection For the installation of a successful, economic, and efficient LNG plant, the most suitable LNG process should be selected before the front end engineering design (FEED) definition stages. The LNG process should be selected from alternative processes based on the operating units, the type of equipment involved, and process utilities. Section 2 briefly summarizes available, well-established LNG processes
ACS Paragon Plus Environment
Page 21 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
21
categorized by the refrigerants used and the number of refrigeration cycles. The LNG processes differ from each other regarding the energy efficiency, degree of availability, degree of complexity, start-up and shutdown procedures, and inherent safety concerns. Furthermore, LNG processes cannot be compared based on solely energy efficiency because the energy efficiency of any LNG process depends strongly on site conditions, feed NG composition, and operating parameters. However, the energy efficiency of an LNG process has a key importance in the selection and installation of an LNG plant; nevertheless, this is not the sole criterion in the design of an LNG plant, and several other factors are also important because of their influence on the commercial aspects of an LNG project. Some of the key factors are listed below. •
Process safety and reliability
•
Environmental impact of the refrigerants and flue gases
•
Site conditions
•
Availability of operational equipment
•
Equipment lifetime
•
Availability of a cost-effective cooling medium for the inter-coolers
•
Availability of refrigerant(s) for make-up
•
Type and size of cryogenic heat exchanger
•
Type and size of compressors
•
Train flexibility
•
Required area/space
•
LNG bunkering system
•
Facility to integrate natural gas liquid (NGL) recovery
•
Ease of operational optimization under varying process conditions and feed compositions
•
Degree of complexity
•
Startup/shutdown time
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 73
22
•
Operational training
•
Capital and operating costs
•
Period of return
In case of offshore LNG production, major key factors for the process selection in terms of quality are energy efficiency, single train capacity, and complexity.29 Figure 13 shows a qualitative comparison of available processes for offshore LNG plants, adapted from a diagram by David Wood and Associates.29
Figure 13. Qualitative comparison of available processes for offshore LNG plants.29
As shown in Figure 13, there is a trade-off between the energy efficiency and the degree of complexity of the process, as well as the single train size. Thus, highly efficient but more complex LNG processes, such as CEMR, are less attractive for offshore applications compared to other expander-based and SMR processes. Furthermore, Lim et al.30 presented an available portfolio for the selection of LNG processes with respect to single-train capacity, which is shown in Figure 14.
ACS Paragon Plus Environment
Page 23 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
23
Figure 14. Portfolio for the selection of LNG processes with respect to single-train capacity.30, 31
4. Design optimization of LNG processes
This section includes a brief introduction of design optimization, optimization problem formulation, general types of optimization techniques, and design optimization analysis of LNG processes.
4.1. Design optimization
The design of any physical and profitable system involves some key variables (controlled and uncontrolled) that fulfill the system requirements or characteristics, which are defined beforehand by an end-user (customer). In the case of LNG plant design, the highest performance in terms of low energy consumption with the highest possible liquefaction rate (i.e., usually greater than 90%) is considered the major requirement and objective. The design of an LNG plant can be considered “optimal” if it achieves
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 73
24
the objective in terms of energy efficiency. The process of searching for the optimal values of the design variables for any well-defined system is known as “optimization.”
4.2. Optimization problem formulation
Any design optimization technique is chosen based on the optimization problem. This optimization problem can be categorized on the basis of many important factors, e.g., constrained, objective (single or multi), or determinacy based. Yang32 classified the optimization problems in a systematic way based on following:
I.
Objective (Single and multi-objective)
II.
Constraint (Constrained and unconstrained)
III.
Landscape (Unimodal and multimodal)
IV.
Function form ( Linear and non-linear)
V. VI.
Variables/Response (Discrete, continuous, and mixed) Determinacy (Deterministic and stochastic)
The effectiveness of an optimization technique is also strongly dependent on the problem formulation in terms of mathematical language. The major components of an optimization problem are considered to be:
•
The objective function(s) or goal(s)
•
The constraint function(s) and limit(s)
•
The design variables
•
The lower and upper bounds (search area) of each design variables
•
The design parameters
ACS Paragon Plus Environment
Page 25 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
25
The objective function(s) or goal(s) should be optimized in terms of a maximum or minimum; for example, in the optimization of an LNG process, the overall energy consumption as an objective function should be minimized and the profit should be maximized. In some cases, there is more than one objective function, and these types of optimization tasks are called “multi-objective optimizations.” These multiobjective optimization algorithms are also known as multi-criteria or even multi-attribute optimizations.32 The constraint function(s) should be satisfied by the optimum objective function(s) at the values of the optimal design variables. Generally, the design variables for simple and linear systems are selected based on experience and the designer’s intuition based on limited design variables. Nevertheless, based on experience and conventional approaches, it is difficult (or impossible) to choose or find the optimal design variables for highly nonlinear and complex systems, such as LNG processes. The design variables are varied between lower and upper bound values for each design variable to find the optimum objective function(s). During optimization problem formulation, design engineers tend to give very conservative boundary values (upper and lower) for design variables. This is mainly due to the lack of information and knowledge of the interactions between the objective and constraint functions in relation to the design variables and insufficient knowledge of the objective function behavior, especially in the early stages of design engineering optimization.33 Effective and robust optimization can only be achieved when the behavior of the objective and constraint functions on the variation of decision variables (between the lower and upper bounds) is known. Some design variables are associated with uncertainty, and knowledge of this is critical in the selection of an appropriate optimization technique; e.g., in real LNG plant optimization, there is a high possibility of uncertainty in the feed NG composition. In design engineering optimization, there are some variables that are assumed to be constant or fixed, and these variables are called “design parameters.” The identification of design variables and design parameters is made by analyzing the degrees of freedom (DOF). Equation (1) determines the total number of degrees of freedom available in an LNG process.34
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 26 of 73
26
n
DOF =
∑d
i
− (k − 2)( NC ) − 2
(1)
i =1
Here, di = DOF of the unit operation involved in the LNG process,
k = number of interconnecting streams in the refrigerating cycle, and
NC = number of components involved.
Mathematically, the optimization of the LNG process concerning overall energy consumption (defined by equation (2)) as an objective function is formulated as:
n Min f(X) = Min ∑Wi / mLNG i =1
(2)
subject to
h (X) = 0, the equality constraints function;
(3)
g (X) > 0, the inequality constraints function; and
(4)
a < X > b , the lower and upper bounds of the design variables.
(5)
Here, “X” is the vector of design variables.
Generally, the flow rates of the refrigerant component(s), operating pressures (such as condensation pressure) and evaporation pressure, and operating temperatures are selected as the key design variables. A minimum internal temperature approach (MITA) value of between 1–3 °C in the main cryogenic LNG exchanger and the zero-liquid fraction at the inlet of each compressor are chosen as the constraints during LNG process optimization.
4.3. Types of optimization algorithms
ACS Paragon Plus Environment
Page 27 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
27
Generally, optimization algorithms are divided into two types: deterministic and stochastic. These two types of optimization algorithms can be further categorized as: 32
I.
Deterministic a. Linear programming b. Non-linear programming c. Gradient based d. Gradient free
II.
Stochastic a. Heuristic b. Metaheuristic (Population and Trajectory based)
Deterministic algorithms are also known as “conventional or classic optimization algorithms.” In optimization using these algorithms, it is assumed that there is no uncertainty in the design parameters and design variables.35 These algorithms follow a rigorous procedure, and the values and path of both functions and design variables are repeatable.32 In contrast, stochastic algorithms concern the uncertainties in the design variables and parameters values. Robust optimization is a type of stochastic algorithm that provides a solution to the optimization problem that is feasible for all variations of the uncertain variables and parameters. Furthermore, stochastic algorithms always initiate the optimization from random points and values of the design variables. Nowadays, another approach, the so-called “hybrid optimization algorithm,” has been introduced to take advantage of the benefits of deterministic and stochastic approaches.36–38
All optimization algorithms belonging to the deterministic and stochastic types are achieved through numerical programming without embedding any prior knowledge of the process associated with the physical, chemical, and thermodynamic properties as well as process constraints. This can cause failure or
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 28 of 73
28
less robust result particularly when solving the issues closely related to the process limitations, technicalities and infeasible approach temperature inside the main cryogenic LNG exchanger. The efficiency and robustness of the numerical optimization algorithms depend on the interactions between the design variables, design parameters, constraints, and the objective function in the process associated with the search area and the initial values. Furthermore, during LNG process optimization, numerical based algorithms often reveal their limitations concerning highly nonlinear and complex system, such as the SMR, DMR, and C3MR processes, owing to the termination in the infeasible region (usually negative MITA values) before achieving an optimal, meaningful solution. To avoid these infeasibilities, Khan et al.22 introduced a knowledge-based optimization (KBO) algorithm, which is specific only for LNG processes. The KBO approach is simple and reliable and works based on knowledge of the entire process. The KBO is preferable when numerical optimization algorithms cannot easily handle objective functions with highly nonlinear operational constraints. Meanwhile, the efficiency of the KBO methodology is dependent on a deep understanding of the LNG process and the influential parameters, such as design variables. However, the KBO approach may have difficulties overcoming a local optimum point.22, 39 Furthermore, Khan et al.21 reported the pros and cons of deterministic, stochastic, and KBO optimization algorithms, as listed in Table 2.
ACS Paragon Plus Environment
Page 29 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
29
Table 2. Pros and cons of deterministic, stochastic, and KBO approaches.21 Property
Deterministic approach
Stochastic approach
KBO approach
Ease of implementation
Same
Same
Same
Constraint handling
Easy
Relative difficult
Easy
Connection ease with Hysys
Same
Same
Same
Calculation time
Small
Large
Intermediate
Multi-objective handling
No
Yes
Yes
Need good initial estimate
Yes
No
No
Premature
Mature
Mature
Few
Many
Few
Robustness of convergence Turning Parameters
Optimization algorithms (deterministic, stochastic, and KBO) that have been employed for the design optimization of LNG processes are further analyzed in Section 4.4.
4.4. Design optimization analysis of LNG processes Using a trial and error approach, the optimization of LNG processes can be achieved without using any special optimization algorithm. Reports24, 26, 40–43 are available concerning optimized LNG processes obtained by modifying the structure and operating parameters without applying any deterministic or stochastic algorithms. However, to avoid confusion, in this study, we have focused on deterministic, stochastic, and KBO-based design optimization studies of LNG processes. For the design optimization analysis, we have included all optimization studies of LNG processes since 1979. The rigorous optimization of refrigerant flow rates and operating pressures can improve the energy efficiency of LNG processes by reducing the gap between composite curves inside the LNG cryogenic exchanger,17,
22
which ultimately increases the global competitiveness of the process, yielding high
economic benefits related to low energy consumption. This was attempted by Ait-Ali44 in 1979 to optimize the MR system for natural gas liquefaction using the deterministic optimization algorithm. He
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 30 of 73
30
used a two-dimensional numerical search approach to find the optimal design of MR based LNG process. The method used by Ait-Ali provided a deep understanding of the MR systems. Nevertheless, the ideal solution assumption has limited the applicability of this method in real MR-based LNG processes. Design optimization analyses of LNG processes with the parameters that most influence the objective function(s), such as NG feed conditions, composition, LNG rate/temperature, isentropic/adiabatic efficiencies of the compressor/expander, and MITA values, are summarized in Table 3.
ACS Paragon Plus Environment
Page 31 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
31
1
Table 3. Design optimization analysis of LNG processes.
LNG process SMR (PRICO)
Two stage SMR
Simulator/ EOS
LNG rate (%) / Temp (oC)
Isentropic Efficiency (%) Compressor/ expander
MITA (oC)
Optimization algorithm
Objective functions
[25oC, 55 bar, N.G]
STAR,
N.G/-160
80.0/N.A
1.0-3.0
Pinch analysis and NLP
Unit power consumption
[C1,C2,C3,n-C4,i-C4,N2] =N.G [N.G oC, 48 bar, 1.0 kmol/s]
PR 94.1/ N.G
N.G/N.A
2.0
Thermodynam ic simulation optimizer
Unit power consumption
92.3/ N.G
N.G/N.A
2.0
Thermodynam ic simulation optimizer
Unit power consumption
39434.4 J/mol
Shi et al.,46 2005
90.4/N.G
N.G/N.A
2.0
Thermodynam ic simulation optimizer
Unit power consumption
20069.9 J/mol
Shi et al.,46 2005
95.1/N.G
N.G/N.A
>0.0
Hysys optimizer in original mode
Unit power consumption
122.3 kW/mol/s
Cao et al.,12 2006
90.5/N.G
N.G/N.G
>0.0
Hysys optimizer in original mode
Unit power consumption
93.2 kW/mol/s
Cao et al.,12 2006
Feed NG Conditions [T, P, F] / composition (mol %)
[32 oC, 50 bar, 4.0 kmol/h]
Thermodyn amic simulation/ Sequential modular Thermodyn amic simulation/ Sequential modular Thermodyn amic simulation, Sequential modular Hysys,
[C1,C2,C3,n-C4,i-C4,N2] =[82.0,11.2,4.0,0.9,1.2,0.7]
PR and LKP
[32 oC, 50 bar, 4.0 kmol/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,N2]
PR and
[C1,C2,C3,n-C4,i-C4,N2] =[82.0,11.2,4.0,0.9,11.2,0.9] N2-CH4 expansion
[N.G oC, 48 bar, 1.0 kmol/s] [C1,C2,C3,n-C4,i-C4,N2] =[82.0,11.2,4.0,0.9,11.2,0.9]
C3MR
[N.G oC, 48 bar, 1.0 kmol/s] [C1,C2,C3,n-C4,i-C4,N2] =[82.0,11.2,4.0,0.9,11.2,0.9]
SMR
N2-CH4 expander
ACS Paragon Plus Environment
Optimized value of objective function 1168.6 kJ/kg LNG and 1126.7 kJ/kg LNG 29642 J/mol
Ref. / Year Lee et al.,45 2002
Shi et al.,46 2005
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 32 of 73
32
Mixed fluid cascade (MFC)
CO2 precooled N2 expander process
SMR
=[82.0,11.2,4.0,0.9,1.2,0.7]
LKP
[11 oC, 61.5 bar, 1.0 kmol/s]
gPROMS, SRK
N.G / -155 at 55.1 bar
90.0/N.A
N.G
gPROMS with multiflash
Total shaft work
10.896 MW
Jensen and Skogestad 26 , 2006
HYSYS, SRK
N.G/-164
3.0
Extended Pinch Analysis and Design (ExPAnD)
Required compression power
1720 kW
Aspelund et al.,47 2007
Aspen Hysys, PR
N.G/-160
Polytropic efficiency of compressor 82.0, adiabatic efficiency of pumps 85.0, and 85.0 for both gas and liquid expanders 80.0/N.A
1.2
GA, NLP
Energy consumption
24.53 MW
Nogal et al.,48 2008
Aspen Hysys, SRK
N.G/-160
80.0/N.A
2.5
GA, NLP
Coefficient of performance (COP)
0.5488
Nogal et al.,48 2008
N.G
N.G/-150
N.G/N.G
N.G
MINLP (GAMS)
Power consumption
N.G
Hasan et al.,49 2009
Hysys, PR
N.G/-160
80.0/80.0
>2.0
NSGA-II (MPMLE Interface)
i) Total shaft work ii) Capital
i) 25.5 MW shaft work with $60
Shah et al.,50 2009
[C1,C2,C3,N2] =[88.8,5.70,2.75,2.75] [15oC, 60 bar, 8.1 kg/s] [C1,C2,C3,n-C4,i-C4,N2] =N.G
[19.85oC, 42 bar, 1.0 kmol/s] [C1,C2,C3,n-C4] =[93.0,5.0,1.5,0.5]
Cascade mixed refrigerant
AP-X
Propane precooled dual N2
[19.85oC, 42 bar, 1.0 kmol/s] [C1,C2,C3,n-C4] =[93.0,5.0,1.5,0.5] [30oC, 64 bar, N.G] [C1,C2,C3,n-C4,i-C4,N2] =N.G [30oC, 55 bar, 22.6 kg/s] [C1,C2,C3,n-C4,i-C4,N2]
ACS Paragon Plus Environment
Page 33 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
33
expander
=N.G
C3MR
[35oC, 1 bar, 1 kmol/h]
cost iii) Annualized costs (TAC), and iv) the hydrocarbon inventory (THI)
Hysys, N.G
95/N.G
N.G/N.A
3.0 and 5.0
Hysys optimizer
Compression power
FORTRAN based simulator
N.G/N.G
N.G/N.A
3.0
Integrated approach (GA and MILP) (GAMS)
Compression power
199355 kW
Nogal et al.,52 2010
Hysys, N.G
100/N.G
80.0/N.A
2.93
Tabu Search (TS) and the Nelder-Mead Downhill Simplex (NMDS)
Energy consumption
144.5 MW
Aspelund et al.,53 2010
Hysys, N.G
100/N.G
80.0/N.A
2.57
TS and NMDS
Energy consumption
122.7 MW
Aspelund et al.,53
Coal bed methane different mol% of N2 C3MR
SMR (PRICO)
[25oC, 60 bar, 5 MTPA] [C1,C2,C3,C4,C5+,CO2,N2] = [85.1,6.5,3.0,1.2,0.5,2.2,1.5] [20oC, 60 bar, 100 kg/s] [C1,C2,C3,C4,N2] =[95.89,2.96,0.72,0.06,0.37]
SMR (PRICO)
[20oC, 60 bar, 100 kg/s]
million capital cost ii) TAC from $28.7 to $32 million/year , at THI from 750 to 10000 kg. iv) 800 kg with 32 MW shaft work. 0.241 kWh/Nm3 with 0% N2 content in feed
ACS Paragon Plus Environment
Gao et al.,51 2009
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 34 of 73
34
SMR
Propane precooled N2 expander
[C1,C2,C3,C4,N2] =[88.80,5.60,3.70,1.90,0.0] [25oC, 55 bar, 165kmol/h] [C1,C2,C3,n-C4,i-C4,N2] =[82.0,11.2,4.0,0.9,1.2,0.7] [35oC, 1.0 bar, 1 kmol/h]
2010 Hysys,
N.G/-161
75.0/N.A
1.5
GA
Unit power consumption
986.55 KW (1092.4 kJ/kg LNG)
Hysys, PR
95/N.G
80.0/85.0
3.0
Hysys optimizer
Unit product liquefaction power consumption
0.75 kWh/Nm3
98.89/-160
83.0/N.A
3.0
GA
Energy consumption
100.78 MW
Alabdulka rem et al.,56 2011
92/N.G
75.0/N.A
3.0
NLP
Energy consumption
0.4244 kW
Khan et al.,18 2011
N.G/-160
80.0/N.A
2.0
SQP
Energy consumption
1900.95 kW
Wang et al.,57 2011
PR
Coal bed methane with N2 less than 70%
C3MR
[N.G oC, N.G bar, 100 kg/s]
Hysys, PR
SMR
[C1,C2,C3,n-C4,i-C4, n-C5 , i-C5, C6+,N2, CO2] =[85.99,7.5,3.5,1.0,1.0,0.2, 0.3,0.4,0.1,0.005] [32oC, 50bar, 1.0 kg/h]
PR
C3MR
[C1,C2,C3,n-C4,i-C4, n-C5 , i-C5,N2] =[91.35,5.36,2.14,0.47,0.46, 0.1,0.1,0.2] [40oC, 8bar, 1.0 kg/s] Mass% [C1,C2,C3,n-C4,i-C4, n-C5 , i-C5,N2] =[90.97,6.0,1.5,0.2,0.3,0.02, 0.01,1.0]
Hysys,
Aspen Plus, PR
ACS Paragon Plus Environment
Shirazi and Mowla54, 2010 Gao et al.,55 2010
Page 35 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
35
DMR
[27oC, 65bar, 0.748kg/h]
Hysys, PR
N.G/-160
80.0/N.A
N.G
Mass% [C1,C2,C3,n-C4,i-C4, i-C5 , ,N2] =[87.5,5.5,2.1,0.5,0.3,0.1,4] SMR
o
[30 C, 60bar, 6533 kmol/h]
C3MR
Lee et al.,36 2011
N.G/-158
N.G/N.A
>2.0
Evolutionary search algorithm
Specific power (kW/(t/d))
28.80 / 27.44 for TEALARC process
Morin et at.,58 2011
Hysys SKR
N.G/-160
80.0/N.A
1.2
NLP (gPROMS)
Energy consumption (MW)
15.17
Tak et al.,59 2011
Hysys PR
N.G/-162
75.0/N.A
3.0
Response surface methodology
Unit power consumption (kWh/kg)
Lee et at.,60 2012
[38oC, 42 bar, N.G]
Hysys,
N.G/-162
N.G/N.A
>0.0
i) UA (kJ/◦C h)
[C1,C2,C3,C4, CO2,N2] =[81.7,8.1,2.1,0.8,0.5,6.8]
PR
BOX method (Hysys optimizer)
Cycle-1 0.3204 / Cycle-2 0.3106 / Cycle-3 0.3184 i) 11,931,510
[N.G oC,8 bar,1.0 kg/s]
Aspen
[19.85oC, 42bar, 1.0kmol/s]
[35oC, 71bar, 292418 kg/h] [C1,C2,C3,n-C4,i-C4, n-C5 , i-C5, n-C6 , n-C7,N2] =[86.39,6.47,2.87,0.82,0.72, 0.31,0.41,0.31,0.15,1.54]
C3MR
854,124
Hysys, N.G
[C1,C2,C3,n-C4] =[93.0,5.0,1.5,0.5] SMR
Required power for the compressors (W)
Ez Optimizer
[C1,C2,C3,n-C4,i-C4, n-C5 , i-C5,N2] =[91.759,4.326,2.240,0.699, 0.39,0.175,0.164,0.247] SMR
Hybrid optimization method (SQP+GA) &
Hatcher et al.,61 2012
ii) Shaft work ii) Not clearly evaluate
N.G/-160
72.0/N.A
2.0
ACS Paragon Plus Environment
MINLP
Shaft work,
1475.06 kW
Wang et
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 36 of 73
36
SMR
Mass% [C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[90.97,6.0,1.5,0.2,,0.3,0.02, 0.01,1.0] [30oC, 60bar, 100 kg/s]
al.,62 2012
Plus, PR
Hysys, PR
N.G/-144
80.0/N.A
N.G
Genetic Algorithm and Nash-GA
Process cost
Consensual decisionmaking model based on game theory was presented 6615 kW
Castillo and Dorao.,63 2012
Hysys
N.G/-145
75.0/N.A
5.0
GA (MATLAB)
Shaft Power
Hysys, SRK
100/N.G
80.0/N.A
0.10
NLPQLP routine
Compressor power consumption
106.1 MW
Wahl et al.,65 2013
Hysys, SRK
100/N.G
80.0/N.A
0.10
NLPQLP routine
Compressor power consumption
90.3 MW
Wahl et al.,65 2013
Hysys, PR and SRK
N.G/-155
80.0/N.A
2.6
NLPQLP routine
Compressor power consumption
125.5 MW
Wahl et al.,65 2013
Simulator was not
N.G/ < -155
70.0/N.A
1.2 & 5
SQP
Power consumption
Approx. 352 Skaugen kW at 1.2 oC et at.,66
[C1,C2,C3,C4,N2] =[95.89,2.96,0.72,0.06,0.37]
SMR
[25oC, 55bar, 1000 kmol/h]
Yoon et al.,64 2012
[C1,C2,C3,n-C4,N2] =[96.93,2.94,0.06,0.01,0.06] SMR
[20oC, 60bar, 100 kg/s] [C1,C2,C3,n-C4,N2] =[95.89,2.96,0.72,0.06,0.37]
SMR
[20oC, 60bar, 100 kg/s] [C1,C2,C3,n-C4] =[88.8,5.60,3.70,1.90]
SMR
[25oC, 55bar, 1 kmol/s] [C1,C2,C3,C4,N2] =[95.89,2.96,0.72,0.06,0.37]
SMR
[25oC, 70bar, 0.265 kg/s]
ACS Paragon Plus Environment
Page 37 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
37
SMR
[C1,C2,C3,n-C4,N2] =[89.8,5.5,1.8,0.1,2.8]
specified, PR
[30oC, 40bar, 1.517*104 kmol/h]
Unisim,
N.G/-157
82.2/N.A
0.43
SRK [C1,C2,C3,C4,N2] =[89.7,5.5,1.8,0.1,2.9]
SMR
[30oC, 40bar, 1.517*104 kmol/h]
Unisim,
N.G/-157
81.8/N.A
1.57
[C1,C2,C3,C4,N2] =[89.7,5.5,1.8,0.1,2.9] [20oC, 60bar, 100 kg/s]
NG expansion
[32oC, 50bar, 158.5 kg/s]
Power consumption
86.6 MW with active constraints in terms of superheating temperature and distance to surge
Jacobsen and Skogestad .,67 2013
Fmincon interior-point algorithm &
Power consumption
119 MW
Jacobsen and Skogestad .,67 2013
MATLAB Hysys, SRK
N.G/-160
80.0/N.A
0.10
Simulated annealing (SA)
Specific compression power
1063 kJ/kg
Austbo et at.,68 2013
Hysys, PR & LKP
92/N.G
75.0/N.A
5.0
Aspen Hysys optimizer
Compression power
194.4MW
Yoon et al.,69 2013
Hysys,
13.55/N.G
80.0/80.0
3.0
Aspen Hysys optimizer
Energy consumption
0.03975 kWh/Nm3
He and Ju70, 2013
[C1,C2,C3,n-C4,N2] =[95.89,2.96,0.72,0.06,0.37] Cascade (C3H8N2O-N2)
2013
MATLAB
SRK
SMR
Fmincon interior-point algorithm &
MITA value , Approx. 225 kW at 5 o C MITA value
[C1,C2,C3,i-C4,n-C4,N2] =[82.0,11.2,4.0,1.2,0.9,0.7]
[15oC, 40bar,100*104Nm3/d]
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 38 of 73
38
process
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,C6,N2, CO2] =[93.72,3.53,0.48,0.09,0.08, 0.02,0.03,0.06,1.38,0.61]
PR
SMR
[45oC, 45bar, 1050 kmol/h]
Hysys,
[C1,C2,C3,C4,N2] =[92.67,3.9,1.95,0.98,0.5]
SRK
[20oC, 65bar, 42.91 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,i-C5, N2] =[87.5,5.5,2.1,0.5,0.3,0.1,4]
PR
[32oC, 50bar, 1.0 kg/h]
Unisim,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[91.35,5.36,2.14,0.47,0.46, 0.1,0.1,0.2]
PR
[32oC, 50bar, 1.0 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[91.35,5.36,2.14,0.47,0.46, 0.1,0.1,0.2]
PR and LK
[32oC, 50bar, 1.0 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5,
PR and LK
DMR
SMR
SMR
C3MR
(original mode)
93.72/N.G
2.0
BOX method
Unit power consumption
6363.2 kW (23.279 kJ/mol)
Sun et al.,71 2013
N.G/-160
82.33 and 70.7% 1st and 2nd stage polytropic efficiencies respectively /N.A 80.0/N.A
0.0
Hybrid optimization (GA and SQP)
Unit power consumption
11,675W (0.2721 kW/kg LNG)
Hwang et al.,37 2013
92/N.G
75.0/N.A
3.0
PSP
Shaft work,
0.3807 kW
Khan and Lee,17 2013
92/N.G
75.0/N.A
3.01
KBO
Shaft work,
0.4324 kW
Khan et al.,22 2013
92/N.G
75.0/N.A
3.0
KBO
Shaft work,
0.2783 kW
Khan et al.,22 2013
ACS Paragon Plus Environment
Page 39 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
39
i-C5,N2] =[91.35,5.36,2.14,0.47,0.46, 0.1,0.1,0.2] SMR
SMR
C3MR
C3MR-SP
SMR
[20oC, 60bar, 100 kg/s]
Hysys,
[C1,C2,C3,C4,N2] =[88.80,5.60,3.70,1.90,0.0]
SRK
[20oC, 60bar, 100 kg/s]
Hysys,
[C1,C2,C3,C4,N2] =[88.80,5.60,3.70,1.90,0.0]
SRK
[25oC, 50bar, 3.73*105 kg/h]
Hysys,
[C1,C2,C3,n-C4,N2] =[96.92,2.94,0.06,0.01,0.07]
PR
[25oC, 50bar, 3.73*105 kg/h]
Hysys,
[C1,C2,C3,n-C4,N2] =[96.92,2.94,0.06,0.01,0.07]
PR
[N.G oC, 50bar, 1.0 kg/s]
Aspen Plus,
100/N.G
80.0/N.A
0.1
SQP
Shaft work,
106.1 MW
Wahl et al.,65 2013
100/N.G
80.0/N.A
2.6
SQP
Shaft work,
143.0 MW
Wahl et al.,65 2013
N.G/-161.3
75.0/N.A
5.0
BOX method built-in Hysys
i) Shaft work ii) Exergy efficiency iii) Cost of cooling duty iv) Operating expenditure (OPEX)
i) 1469 Wang et MJ/tonneal.,72 2013 LNG, ii) 31.41%, iii) 5.88*108 kJ/h, iv) $125.78/ton ne-LNG
N.G/-161.3
75.0/N.A
5.36
BOX method built-in Hysys
N.G/-160
78.0/N.A
3.0
GA
i) Shaft work ii) Exergy efficiency iii) Cost of cooling duty iv) Operating expenditure (OPEX) Shaft work
i) 1464 MJ/tonneLNG, ii) 31.50%, iii) 5.92*108 kJ/h, iv)$125.77/t onne-LNG The authors studied the effect of
[C1,C2,C3,n-C4,i-C4,i-C5,
ACS Paragon Plus Environment
Wang et al.,72 2013
Xu et al.,73 2013
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 40 of 73
40
N2 expander
Propane precooled N2 expander
N2] =[87.5,5.5,2.1,0.5,0.3,0.1,4]
PR
[20oC, 30bar, 1.0 kmol/h]
Hysys,
[C1,C2,C3,i-C4,N2] =[90.0,4.0,2.0,2.0,2.0]
PR
[20oC, 30bar, 1.0 kmol/h]
Hysys,
[C1,C2,C3,i-C4,N2] =[90.0,4.0,2.0,2.0,2.0]
PR
R410a precooled N2 expander
[20oC, 30bar, 1.0 kmol/h]
Hysys,
[C1,C2,C3,i-C4,N2] =[90.0,4.0,2.0,2.0,2.0]
PR
N2 dual expander
[30oC, 40-60bar, 50*104 Nm3/d]
Hysys,
different ambient temperature on shaft work requirement 95/N.G
80.0/85.0
2.0
Aspen Hysys optimizer (original mode)
Unit power consumption
0.4669 kWh/Nm3
He and Ju,74 2014
95/N.G
80.0/85.0
2.0
Aspen Hysys optimizer (original mode)
Unit power consumption
0.3734 kWh/Nm3
He and Ju,74 2014
95/N.G
80.0/85.0
2.0
Aspen Hysys optimizer (original mode)
Energy consumption
0.3607 kWh/Nm3
He and Ju,74 2014
12.61/N.G
75.0/85.0
N.G
N.G
Exergy utilization rate (%)
19.42
He and Ju,75 2014
95.5/N.G
80.0/80.0
>3.0
GA
Energy consumption
0.5163 kWh/Nm3
He and Ju,76 2014
PR [C1,C2,C3,n-C4,i-C4,n-C5, i-C5,C6,N2] =[94.30,3.55,0.48,0.09,0.08, 0.02,0.03,0.06,1.39] Parallel N2 expander
[32oC, 5bar, 2083 Nm3/h= 1812.21 kg/h]
Hysys, PR
ACS Paragon Plus Environment
Page 41 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
41
process
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[85.82,5.62,3.47,1.35,0.72, 0.34,0.35,2.07]
C3MR
[15oC, 62.6bar, 100 TPD]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[86.89,5.10,2.13,0.45,0.44, 0.01,0.01,4.97]
PR
N2-CO2 expander (CO2 precooled N2 expander)
[32oC, 50bar, 1.0 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[91.33,5.36,2.14,0.47,0.46, 0.1,0.1,0.22]
PR and LK
C3MR
[-30oC, 44.4bar, 96.86 kg/s]
Aspen Plus,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5, C6+,N2, CO2] =[85.995,7.5,3.5,1.0,1.0,0.2, 0.3,0.4,0.1,0.005] N2 dual expander
C3MR
N.G/-160
N.G/N.A
3.0
Successive reduced quadratic programming using gPROMS
Energy consumption
344.33 kW
Lee et al.,77 2014
92/N.G
75.0/75.0
3.0
KBO
Energy consumption
0.4945 kW
Khan et al.,9 2014
N.G/-160
83.0 / Steam turbine efficiency 90.0 with exhaust 110 oC
3.0
Hybrid optimization (GA and SQP)
Fuel consumption
4.14 kg/s
Mortazavi et al.,38 2014
N.G/-155
N.G/N.A
3.0
Empirical modeling (Data-driven model)
Energy consumption
45938 kW
Song et al.,78 2014
N.G/-161.3
75.0/N.A
>0
Hysys optimizer with PengRobinson
i) Shaft work ii) Specific annual cost
i) 1547.3 MJ/tonneLNG ii)
Wang et al.,79 2014
PR
[10oC, 50bar, 1 MTPA]
Hysys,
[C1,C2,N2] =[94.1,0.2,5.7]
PR
[25oC, 50bar, 3.99*105 kg/h]
Hysys,
[C1,C2,C3,n-C4,N2] =[96.92,2.94,0.06,0.01,0.07]
PR
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 42 of 73
42
DMR
PRICO (SMR)
[25oC, 50bar, 3.99*105 kg/h]
Hysys,
[C1,C2,C3,n-C4,N2] =[96.92,2.94,0.06,0.01,0.07]
PR
[25oC, 50bar, 1.0 kg/s]
Aspen Plus,
[C1,C2,C3,n-C4,i-C4,i-C5, N2] =[87.5,5.5,2.1,0.5,0.3,0.1,4]
PR
$214.9/tonn e LNG i) 1356.5 MJ/tonneLNG ii) $213.4/tonn e LNG 1003.6 kJ/kg LNG (0.2787 kW)
N.G/-161.3
75.0/N.A
>0
Hysys optimizer with PengRobinson
i) Shaft work ii) Specific annual cost
N.G/-160
78.0/N.A
3.0
GA
Energy consumption
0.77/N.G
85.0/80.0
2.0
Hysys optimizer
Energy consumption
9.9 kW/kmol/h
Yuan et al.,14 2014
100/N.G
80.0/N.A
2.0
SQP (NLPQLP)
Energy consumption
15208 kW
Austbø81, 2015
73/-119.4
90.0/ Turbine efficiency 92.0
1.0-3.0
GA
Energy storage
0.068 kW
Fazlollahi et al.,82 2015 Fazlollahi et al.,82 2015 Fazlollahi et al.,82 2015
CO2 precooled N2 single expander
[32oC, 48bar, 4.0 kmol/h]
Hysys,
[C1,C2,C3,n-C4,i-C4] =[82.0,11.2,4.0,0.9,1.2]
PR and LK
SMR
[20oC, 60bar, 1.0 kmol/s]
Hysys,
[C1,C2,C3,C4,N2] =[95.89,2.96,0.72,0.06,0.37]
PR
[21oC, 37bar, 21200 kmol/h]
Hysys,
[C1,C2,C3]=[95.0,3.0,2.0]
PR
DMR
[21oC, 37bar, 21200 kmol/h]
Hysys,
73/-119.4
90.0/ Turbine efficiency 92.0
1.0-3.0
GA
Energy storage
0.074 kW
C3MR
[C1,C2,C3]=[95.0,3.0,2.0] [21oC, 37bar, 21200 kmol/h]
PR Hysys,
73/-119.4
90.0/ Turbine efficiency 92.0
1.0-3.0
GA
Energy storage
0.086 kW
[C1,C2,C3]=[95.0,3.0,2.0]
PR
SMR
ACS Paragon Plus Environment
Wang et al.,79 2014
Xu et al.,80 2014
Page 43 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
43
R410a precooled parallel nitrogen expansion process C3MR
C3MR
SMR
SMR
[20oC, 50bar, 1633 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5, C6~C9,N2] =[92.09,3.54,1.49,0.34,0.28, 0.05,0.06,0.03,2.1] [15oC, 62.6bar, 4902 kg/h]
PR
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[86.89,5.10,2.13,0.45,0.44, 0.01,0.01,4.97]
PR
[32oC, 50bar, 1.0 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[91.33,5.36,2.14,0.47,0.46, 0.1,0.1,0.22]
PR and LK
[32oC, 50bar, 1.0 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[91.33,5.36,2.14,0.47,0.46, 0.1,0.1,0.22]
PR and LK
[32oC, 66.5bar, 26700 kmol/h]
Hysys,
Hysys,
97.8/-158.2
80.0/85.0
3.0
GA
i) Energy consumption ii) Figure of merit (FOM)
i) 0.376 kWh/Nm3 ii) 0.5665
He and Ju,83 2015
i) 85/N.G ii) 87/N.G iii) 90/N.G iv) 95/N.G
75.0/N.A
3.0
Successive reduced quadratic programming (SRQPD) using gPROMS
Total energy consumption
Lee et al.,84 2015
92/N.G
75.0/N.A
3.0
Sequential coordinate random search (SCRS)
Total shaft work
1139.5 kW at 85% liquefaction rate, 1160.9 kW at 87%, 1214.6 kW at 90%, and 1346.1 kW at 95% 0.2629 kW
92/N.G
75.0/N.A
3.0
Sequential coordinate random search (SCRS)
Total shaft work
0.4400 kW
Khan et al.,16 2015
N.G/N.G
75.0/N.A
2.0
GA
Total energy consumption
128325 kW
Moein et al.,85 2015
PR [C1,C2,C3,n-C4,i-C4,N2] =[94.0,3.10,1.30,0.40,0.30, 0.90]
ACS Paragon Plus Environment
Khan et al.,16 2015
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 44 of 73
44
KSMR
[28oC, 90bar, 1.0 kg/h]
Hysys, PR and LK
SMR
[C1,C2,C3,n-C4,i-C4,N2] =[89.40,7.99,1.58,0.07,0.11, 0.84] [25oC, 55bar, 1.0 kg/s] [C1,C2,C3,n-C4,i-C4,N2] =[91.3,5.4,2.1,0.5,0.5,0.2]
PR
[20oC, 60bar, 1.0 kmol/s]
Hysys,
SMR
SMR
94.6/N.G
83.0/N.A
3.0
MCD
Specific power requirement
0.2208 kW/kg LNG
Park et al.,19 2015
100/N.G
80.0 /N.A
3.0
Successive reduced quadratic programming (SRQPD) using gPROMS
Specific power requirement
1037.1 kJ/kg LNG
Tak et al.,86 2015
N.G/-161
80.0/N.A
3.0
Power consumption
65.6 MJ/kmol
Austbø et al.,87 2016
[C1,C2,C3,n-C4,N2] =[95.89,2.96,0.72,0.06,0.37]
PR
sequential quadratic programming algorithm (NLPQLP)
[32oC, 80bar, 1.0 kg/s]
Aspen plus, N.G/-155 N.G before JT
78.0/N.A
3.0
GA
Hysys,
95/N.G
80.0/85.0
3.0
GA
[C1,C2,C3,n-C4,i-C4,N2] =[90.0,5.0,2.0,1.0,1.0,1.0]
PR
i) Very strong ii) 0.40 iii) 1005.76 kW 6.73 kWh/kmol LNG
Cao et al.,88 2016
[30oC, 50bar, 1.0 kmol/h]
i) Robustness ii) Exergy efficiency iii) Total shaft work Unit power consumption
95/N.G
80.0/85.0
3.0
GA
Unit power consumption
9.11 kWh/kmol LNG
Ding et al.,7 2016
95/N.G
80.0/85.0
3.0
GA
Unit power
11.09
Ding et
Hysys,
[C1,C2,C3,C4,i-C4,i-C5,N2] =[87.5,5.5,2.1,0.3,0.5,0.1,4] Propane precooled N2-CH4 expander CO2 precooled N2-CH4 expander
[30oC, 50bar, 1.0 kmol/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,N2] =[90.0,5.0,2.0,1.0,1.0,1.0]
PR
N2 single
[30oC, 50bar, 1.0 kmol/h]
Hysys,
ACS Paragon Plus Environment
Ding et al.,7 2016
Page 45 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
45
expander
C3MR
SMR
DMR
C3MR
[C1,C2,C3,n-C4,i-C4,N2] =[90.0,5.0,2.0,1.0,1.0,1.0]
PR
[27oC, 65bar, 49.179 kg/h]
Hysys,
[C1,C2,C3,C4,N2] =[87.48,5.50,2.12,0.89,4.01]
PR
[37oC, 50bar, 114155 kg/h]
gPROMS,
[C1,C2,C3,n-C4,i-C4,N2] =[91.3,5.4,2.1,0.5,0.5,0.2]
PR
[25oC, 55bar, 1 MTPA]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[87.23,6.68,3.49,0.89,0.59, 0.19,0.29,0.49]
PR
[25oC, 55bar, 1 MTPA]
Hysys,
[C1,C2,C3,n-C4,i-C4,N2] =[61.81,17.1,11.92,4.48,4.48 ,0.21]
PR
consumption
kWh/kmol LNG
al.,7 2016
92.2/N.G
80.0/N.A
3.0
GA & Pinch analysis
Specific power
0.2594 kWh/kg LNG
Ghorbani et al.,89 2016
100/N.G
75.0/N.A
3.0
successive reduced quadratic programming (SRQPD)
i) Total energy requirement ii) Total annual cost
i) 37861 kW & ii) 29.21 million USD
Lee and Moon,90 2016
96.3%
N.A
3.2
BOX
i) Specific compression power ii) Total UA
i) 0.3399 kW/kg LNG with UA= 14.6 MW/ oC
Khan et al.,23 2016
N.G/-161
N.G/N.A
3.0
ACS Paragon Plus Environment
Hybrid GA (Gradient Assisted Robust Optimization, GARO)
LNG cycle power demand
ii) 14.6 MW/ oC with 0.5256 kW/kg LNG 135.7 MW
Mortazavi et al.,91 2016
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 46 of 73
46
Three stage propane precooling Cascade
SMR
MSMR (KSMR)
C3MR
AP-X
[29oC, 75bar, 41700 kmol/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[89.74,4.96,3.43,0.73,0.79, 0.05,0.02,0.21]
PR
[40oC, 60bar, 100 kg/s]
Hysys,
[C1,C2,C3,i-C4,N2] =[95.89,2.96,0.72,0.06,0.37
N.G
[26oC, 80bar, 1.0 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,N2] =[93.18,5.05,1.09,0.05,0.08, 0.55] [25oC, 55bar, 1 MTPA]
PR & LK
[C1,C2,C3,C4,N2] =[89.7,5.5,1.8,0.1,2.9] [30oC, 50bar, 52010.62 kmol/h]
N.G/-160
N.G/ -163.75
80.0/N.A
80.0/N.A
>2.0
1.4
Energy and exergy analysis (KBO)
PSO
93.65/N.G
83.0/N.A
3.0
Multivariate Coggins + KBO
75.0/N.A
>0.5
PR
N.G/-157 , LNG rate 30.72 kg/s
Hysys,
95/N.G
80.0/85.0
3.0
Hysys optimizer (BOX and mixed method) GA
100/-107.5 at 20 bar
85.0/N.A
3.0
100/-107.5 at 20 bar
85.0/N.A
100/-107.5
85.0/80.0
Hysys,
i) COP
i) 1.65
ii) Specific power
ii) 0.205 MWh/tonne LNG
iii) Exergy efficiency i) LNG production rate ii) Specific power Specific compression energy
iii) 40.22% i) 177 kg/s
Khan et al.,92 2016
Park et al.,93 2016
ii) 1017.8 kJ/kg LNG 0.2618 kW/kg LNG
Pham et al.,20 2016
Specific energy consumption
973.93 kJ/kg LNG
Unit power consumption
4.337 kW h/kmol
Sanavandi and Ziabashar hagh,94 2016 Sun et al.,95 2016
GA
Specific power consumption
0.2088 kWh/kg LNG
Xiong et al.,96 2016
3.0
GA
GA
0.2202 kWh/kg LNG 0.3203
Xiong et al.,96 2016
3.0
Specific power consumption Specific
PR
Cascade PLNG process SMR PLNG process Single
[C1,C2,C3,n-C4,i-C4,N2] =[91.6,4.5,1.1,0.3,0.5,2.0] [35oC, 50bar, 1.0 kmol/h]
Hysys,
[C1, CO2]=[99.5,0.5]
PR
[35oC, 50bar, 1.0 kmol/h]
Hysys,
[C1, CO2]=[99.5,0.5] [35oC, 50bar, 1.0 kmol/h]
PR Hysys,
ACS Paragon Plus Environment
Xiong et
Page 47 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
47
expander PLNG process DMR
at 20 bar [C1, CO2]=[99.5,0.5]
PR
[37oC, 50bar, 114155 kg/h]
gPROMS,
[C1,C2,C3,n-C4,i-C4,N2] =[91.3,5.4,2.1,0.5,0.5,0.2]
PR
93.2 by mass/N.G
75.0/N.A
3.0
SRQPD
power consumption
kWh/kg LNG
al.,96 2016
i) Total compression energy
i) 1080.31 kJ/kg LNG
Lee and Moon,97 2017
ii) Total annual cost
ii) 10.10 million USD/year
iii) Total profit Cascade process (single stage single comp)
[5 °C above ambient seawater temperature, 60bar, 1 ton]
Cascade process (two stage single comp)
[5 °C above ambient seawater temperature, 60bar, 1 ton]
Mixed fluid cascade (MFC)
[11oC, 61.5bar, 1 ton] [C1,C2,C3,N2] =[88.8,5.70,2.75,2.75]
Carnot
[5 °C above ambient
MATLAB (numerical model), CoolProp
N.G/−155
85.0/N.A
5.0
GA
MATLAB (numerical model), CoolProp
N.G/−155
85.0/N.A
5.0
GA
HYSYS, SRK
N.G/−155
90.0/N.A
5.0
HYSYS optimizer with original method
MATLAB
N.G/−155
85.0/N.A
5.0
Developed
[C1,C2,C3,n-C4,i-C4,N2] =[87.7,5.4,2.6,0.4,0.8,3.1]
[C1,C2,C3,n-C4,i-C4,N2] =[87.7,5.4,2.6,0.4,0.8,3.1
ACS Paragon Plus Environment
iii) 84.9 million USD/year Total energy 302.1 consumption kWh/ton NG at 10 °C, and 363.0 kWh/ton NG at 30 °C Total energy 228.6 consumption kWh/ton NG at 10 °C, and 276.2 kWh/ton NG at 30 °C Total energy 196.5 consumption kWh/ton NG at 10 °C, and 243.2 kWh/ton NG at 30 °C Total energy 93.6
Jackson et al.,98 2017
Jackson et al.,98 2017
Jackson et al.,98 2017
Jackson et
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 48 of 73
48
cascade
seawater temperature, 60bar, 1 ton]
(numerical model), CoolProp
MATLAB routine
consumption
Modified DIRECT (DIviding a hyperRECTangle)
i) Total compression power
[C1,C2,C3,n-C4,i-C4,N2] =[87.7,5.4,2.6,0.4,0.8,3.1 SMR
SMR
[26.85oC, 65bar, 26.38 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,i-C5, N2] =[87.5,5.5,2.1,0.5,0.3,0.1,4]
PR
[32oC, 80bar, 1.0 kg/h]
Hysys,
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[91.33,5.36,2.14,0.47,0.46, 0.1,0.1,0.22]
PR
96.5/N.G
80.0/N.A
2.85
92/N.G
75.0/N.A
3.0
KBO
ii) Specific power requirement Total compression energy
kWh/ton al.,98 2017 NG at 10 °C, and 115.0 kWh/ton NG at 30 °C 6.71 kW Na et al.,99 2017
949 kJ/kg LNG 0.3183 kW/kg LNG
Pham et al.,39 2017
CO2 [32oC, 48bar, 1.0 kmol/h] precooling single N2 [C1,C2,C3,n-C4,i-C4,N2] expansion =[82.0,11.2,4.0,0.9,1.2,0.7]
Hysys, PR & LK
78/N.G
85.0/80.0
>2.0
GA
Unit energy consumption
8.39
Song et al.,100 2017
CO2 [32oC, 48bar, 1.0 kmol/h] precooling [C1,C2,C3,n-C4,i-C4,N2] single N2 expansion =[82.0,11.2,4.0,0.9,1.2,0.7]
Hysys, PR & LK
80.7/N.G
85.0/80.0
>2.0
Unit energy consumption
8.90
Song et al.,100 2017
Hysys & PR
95/N.G
80.0/N.A
3.0
NSGA-II (Nondominated sorting genetic algorithm) GA
Unit power consumption
4.655 kWh/kmol
Ding et al.,101 2017
MFC
[30oC, 50bar, 52010.62 kmol/h] [C1,C2,C3,n-C4,i-C4,N2] =[91.6,4.5,1.1,0.3,0.5,2.0]
ACS Paragon Plus Environment
Page 49 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Industrial & Engineering Chemistry Research
49
KSMR
[12.1oC, 63.4bar, 100 ton/d]
Hysys, N.G
N.G/ −150.7
N.G/N.A
N.G
Online steadystate optimizer based on SQP
Unit energy consumption
1051 kJ/kg LNG
Won & lee.,102 2017
Hysys, N.G
N.G/ −150.7
N.G/N.A
N.G
Real time steady state optimization (RTSSO)
Unit energy consumption
1051 kJ/kg LNG
Won & Kim,103 2017
[32oC, 50bar, 1.0 kg/h]
Hysys,
92/N.G
75.0/N.A
3.0
MCD
Unit power consumption
1344.24 kJ/kg LNG
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[91.35,5.36,2.14,0.47,0.46, 0.1,0.1,0.2]
PR
Qyyum et al.,104 2017
[C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[86.90,5.10,2.13,0.45,0.44, 0.01,0.01,4.96]
KSMR
[12.1oC, 63.4bar, 100 ton/d] [C1,C2,C3,n-C4,i-C4,n-C5, i-C5,N2] =[86.90,5.10,2.13,0.45,0.44, 0.01,0.01,4.96]
SMR
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 50 of 73
50
2
The first attempt toward the design optimization of the LNG process was made in 1979 by Ait-Ali.44
3
Table 3 and Figure 17 show that the design optimization of LNG processes has matured since 2002.
4
Furthermore, since 2013, improvements in the energy efficiency of the LNG process through optimization
5
approaches alone have remained a hot topic for process engineers and researchers. As shown in Figures
6
18 and 19, investigations have focused on the SMR process, considering it the most promising candidate
7
for small scale and offshore LNG production. In addition to the SMR process, the N2- expander-based
8
LNG processes are also considered promising candidates for small-scale and offshore LNG production
9
because, of all the LNG processes, they are green and safe processes, although the relatively low energy
10
efficiency is a major issue. Therefore, following the SMR process, N2-expander-based LNG processes
11
have been considered for optimization. As shown in Figure 18, 21% of optimization studies concern the
12
design optimization of N2-expander-based processes. Currently, APCI’s C3MR process is the dominant
13
technology and the most energy efficient. Consequently, approximately 77% of LNG plants use the
14
C3MR process.105 Nevertheless, many researchers have reported that the energy efficiency of the C3MR
15
process can be improved through optimization. Therefore, as shown in Figures 18 and 19, 20% of
16
optimization studies have concerned the C3MR process. Because of the complexity in the handling of the
17
degrees of freedom in the DMR and cascade LNG processes, few researchers have attempted to optimize
18
these processes.
ACS Paragon Plus Environment
Page 51 of 73
51
Year
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
19
20
Figure 17. Research focus on the design optimization of LNG processes by year.
SMR DMR KSMR Expander based Cascade C3MR AP-X
2%
20%
35%
11%
6% 21% 4%
21
22
Figure 18. Design optimization breakdown with respect to LNG processes.
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
52
LNG Process
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 52 of 73
23
24
Figure 19. LNG processes and design optimization studies
25
As shown in Figure 20 and Table 3, most LNG process optimization studies (more than 20) have used
26
genetic algorithm (GA)-based optimization approaches because of their simplicity and ease of
27
implementation. In addition to GA, a well-known commercial simulator, Aspen Hysys®, is often used for
28
optimization. This simulator has built-in optimization algorithms, such as the BOX method, and many
29
researchers and processes engineers prefer Hysys for optimization because of its simplicity and integrated
30
simulation–optimization framework. In particular, in optimization using Hysys, there is no need for an
31
interface/connection. More than 15 optimization studies have used the Hysys optimizer for the design
32
optimization of LNG processes, as shown in Figure 20.
ACS Paragon Plus Environment
Page 53 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
53
33
34
Figure 20. Optimization algorithm trends for LNG processes.
35
36
5. Challenges in the optimal designs comparison of LNG processes
37
As listed in Table 3, most optimization studies have focused on the minimization of operating cost in
38
terms of energy. The following design parameters have a significant impact on the objective function, as
39
well as optimization efficiency:
40
•
Feed NG conditions (pressure and temperature)
41
•
Feed NG flow rate and composition
42
•
Process simulator
43
•
Thermodynamic model (binary interactions calculation method)
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 54 of 73
54
44
•
LNG rate (LNG specifications)
45
•
Minimum driving force (MITA) value in LNG heat exchangers
46
•
Compressor and expander efficiencies
47
In considering these above design parameters, we found some difficulty in comparing the objective
48
function(s) in different design optimization studies. Each study uses different values for the LNG process,
49
resulting in different optimal conditions and objective functions. Without information concerning the
50
design parameters and only considering low energy consumption, it is impossible to select the best and
51
most suitable LNG process, even at the FEED level. Therefore, in the present study, design optimization
52
analysis was carried out by considering the design parameters that correspond to objective functions. We
53
found that, for different design parameters, the objective functions were different. Thus, it was
54
challenging to identify each design parameter during analysis, especially because some optimization
55
studies have not included all design parameters and have considered only decision (design) variables,
56
constraints, and objective functions.
57
A comparison of the operating and total annual costs was also difficult because of the differences in
58
cost correlations, cost index, and utility prices used; therefore, the consistent cost evaluation of 1 kg LNG
59
is difficult. The evaluation of the optimal design of different LNG processes becomes easy when the same
60
research group or engineers present competing optimized alternatives, such as Pham et al.20, 39 and Khan
61
et al.10, 16-18, who belong to the same research group and used consistent design parameters for the same
62
LNG processes. Similarly, Cao et al.12,
63
parameters to present their optimal LNG process designs.
88
and Lee et al.45,
60, 77, 84, 90, 97
’ also used consistent design
64
In addition, the authors recommend the following important parameters to facilitate the
65
reproducibility of the simulation–optimization framework and accurate comparison, which will,
66
ultimately, lead to practical implementation.
ACS Paragon Plus Environment
Page 55 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
55
67
•
The NG flow rate and its composition, temperature, and pressure should be declared.
68
•
The LNG process should be modeled and simulated realistically as possible.
69
•
The chosen thermodynamic method (e.g., PR or SRK.) should be stated in simulation and modeling.
70 71
•
declared for rigorous convergence of the model.
72 73
The design parameters and variables necessary to satisfy the degree of freedom should be
•
For optimization, either using an external technique or internal technique, the chosen
74
interface, optimization algorithm, constraints, decision variables, and objective function(s)
75
should be declared in the research paper.
76
•
The units of the objective function(s) should be same to allow a fair comparison with other
77
optimized objective function(s), e.g., in most studies, the total energy consumption as an
78
objective function is given as kg-LNG production, whereas, in some studies, it is given as
79
Nm3 LNG production. This inconsistency in units also creates confusion for comparative
80
studies, as well as allowing unfair comparison.
81
•
If cost analysis is considered, the capital and operating cost both should be optimized, and the
82
cost index, cost correlations, payback period, interest rate, and equipment life should be
83
declared.
84 85
•
Finally, for each LNG process alternative, the design criteria should be the same, and the processes should be optimized as rigorously as possible.
86 87
6. Future directions and conclusions
88
Potential advancements in LNG processes towards energy efficiency for ecological and economic
89
reasons can be realized by considering the following:
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 56 of 73
56
90
LNG processes have been mostly optimized using either stochastic or deterministic algorithms with a
91
set of design variables such as a pure or mixed refrigerant flow rate, mixed refrigerant evaporation
92
pressure, mixed refrigerant evaporation temperature, and mixed refrigerant condensation pressure and
93
temperature, as summarized in Table 3. Based on the analysis of Table 3, stochastic algorithms tend to be
94
more commonly used for more complex LNG processes such as C3MR. As shown in Table 3, there have
95
been few attempts to optimize the design of the DMR and cascade LNG processes in comparison with the
96
SMR and C3MR processes. Therefore, design optimization of the DMR and cascade processes is needed
97
to facilitate the selection and implementation of the most suitable LNG process.
98
Recent studies have mainly concerned the design optimization of existing LNG processes. The
99
efficiency of LNG process has been enhanced through well-known optimization approaches, mainly GA,
100
non-linear programming (NLP), and sequential quadratic programming (SQP). However, new and
101
efficient metaheuristic optimization algorithms such as vortex search106, runner root107, and many others
102
have been developed; thus, optimal designs can be further refined through the newly developed
103
optimization approaches.
104
It has been found that the energy efficiency and global competitiveness of LNG processes can be
105
improved through the improvement/replacement of the involved equipment.105,
108, 109
106
optimization of these improved processes and equipment is worthwhile. In recent studies, the optimized
107
designs of LNG processes were mostly compared by evaluating the total required compression power
108
because of this major contribution to the operating costs. However, this could result in inappropriate
109
decisions because the capital expenditure is not considered in the cost evaluation.
Therefore, the
110
The MITA value in the main LNG cryogenic exchanger is usually used as a design constraint, but this
111
might lead to incorrect decisions because it does not take into account the heat transfer area separately as
112
a main design or capacity parameter of heat exchangers. For the optimal design of LNG processes,
113
research must consider equipment costs and size rigorously, as well as the thermodynamic performance.
114
There is also a need to establish an efficient optimization framework that will allow the simultaneous
ACS Paragon Plus Environment
Page 57 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
57
115
design optimization of whole process and the detailed optimal designs of the involved equipment
116
corresponding to minimization of operating and capital costs. For example, Pattison and Baldea110 have
117
presented the optimization-oriented model of MHEX that was integrated in a SMR-LNG process. Tsay et
118
al.111 proposed an equation-oriented simulation and optimization of SMR process incorporating detailed
119
spiral-wound multi-stream heat exchanger model.
120
The following important issues must be also addressed for further improvements in LNG processes:
121
•
Process optimization including waste energy recovery and integration
122
•
Improvements in the integrated NGL-LNG processes through optimization approaches
123
•
Stochastic optimization including both capital and operating cost
124
•
Design optimization of newly presented different energy efficient LNG process configurations, for example, as recently presented by Lim et al.112
125 126 127
•
Evaluation of newly developed metaheuristic optimization algorithms for the design optimization of complex and highly nonlinear LNG processes
128 129
Concluding remarks
130
NG liquefaction processes use a tremendous amount of energy in the form of compression power,
131
which is the most energy-intensive step of all the NG value chain steps. Thus, LNG processes
132
have much scope for design and operational optimization corresponding to the minimization of
133
the operating (energy) and capital expenditures. There is also opportunity for the development of
134
new environmentally friendly refrigerants and liquefaction processes for cryogenic applications.
135
Acknowledgments
136
This work was supported by Basic Science Research Program through the National Research Foundat
137
ion of Korea (NRF) funded by the Ministry of Education (2015R1D1A3A01015621) and Priority Researc
138
h Centers Program through the National Research Foundation of Korea (NRF) funded by the Ministry of
139
Education (2014R1A6A1031189).
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 58 of 73
58
140 141
Nomenclature
142
N2
Nitrogen
143
C1
Methane
144
C2
Ethane
145
C3
Propane
146
iC5
Isopentane
147
C6
Hexane
148
C9
Nonane
149
PR
Peng-Robinson
150
SRK
Soave−Redlich−Kwong
151
LKP
Lee–Kesler–Ploecker
152
MILP
Mixed-integer linear programming
153
MINLP
Mixed-integer nonlinear programming
154
MVS
Microsoft Visio Studio
155
KBO
Knowledge-based optimization
156
GA
Genetic algorithm
157
NSGA
Non-dominated sorting genetic algorithm
158
TS
Tabu search
159
NMDS
Nelder-Mead downhill simplex
160
PSO
Particle swarm optimization
161
NLP
Non-linear programming
162
SQP
Sequential quadratic programming
163
SCRS
Sequential coordinate random search
164
MCD
Modified coordinate descent
165
N.G
Not given or not specified
166
N.A
Not applicable
ACS Paragon Plus Environment
Page 59 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
59
167
HMCD
Hybrid modified coordinate descent
168
LNG
Liquefied natural gas
169
FLNG
Floating liquefied natural gas
170
MR
Mixed refrigerant
171
NG
Natural gas
172
T
Temperature
173
P
Pressure
174
F
Flow rate
175
SMR
Single mixed refrigerant
176
KSMR
Korea single mixed refrigerant
177
MITA
Minimum internal temperature approach
178
C3MR
Propane precooled mixed refrigerant
179
DMR
Dual mixed refrigerant
180
TDCC
Temperature difference between composite curves
181
THCC
Temperature-heat flow composite curves
182
LMTD
Log mean temperature difference
183
COP
Coefficient of performance
184
FEED
Front end engineering design
185 186
References
187
(1) SHELL LNG Outlook 2017. http://www.shell.com/energy-and-innovation/natural-gas/liquefied-
188
natural-gas-lng/lng-outlook.html, (accessed 31 May, 2017).
189
(2)
190
2017).
EIA,Today in Energy. https://www.eia.gov/todayinenergy/detail.php?id=26212, (accessed 30 May,
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 60 of 73
60
191
(3) EIA Annual Energy Outlook 2017. https://www.eia.gov/outlooks/aeo/pdf/0383(2017).pdf, (accessed
192
30 May, 2017).
193
(4)
194
outlook-2017/bp-energy-outlook-2017.pdf, (accessed 30 May, 2017).
195
(5) IGU, International Gas Union World LNG Report 2017. http://www.igu.org/news/igu-releases-2017-
196
world-lng-report, (accessed 30 May, 2017).
197
(6) Songhurst, B., LNG plant cost escalation. 2014.
198
(7) Ding, H.; Sun, H.; He, M., Optimisation of expansion liquefaction processes using mixed refrigerant
199
N2–CH4. Appl. Therm. Eng. 2016, 93, 1053-1060.
200
(8) Li, Q. Y.; Ju, Y. L., Design and analysis of liquefaction process for offshore associated gas resources.
201
Appl. Therm. Eng. 2010, 30, (16), 2518-2525.
202
(9)
203
the N2–CO2 expander cycle for the economic development of stranded offshore fields. J. Nat. Gas Sci.
204
Eng. 2014, 18, 263-273.
205
(10) Khan, M. S.; Lee, S.; Getu, M.; Lee, M., Knowledge inspired investigation of selected parameters
206
on energy consumption in nitrogen single and dual expander processes of natural gas liquefaction. J. Nat.
207
Gas Sci. Eng. 2015, 23, 324-337.
208
(11) Austbø, B.; Gundersen, T., Optimization of a Single Expander LNG Process. Energy Procedia
209
2015, 64, 63-72.
BP Energy Outlook 2017. https://www.bp.com/content/dam/bp/pdf/energy-economics/energy-
Khan, M. S.; Lee, S.; Hasan, M.; Lee, M., Process knowledge based opportunistic optimization of
ACS Paragon Plus Environment
Page 61 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
61
210
(12) Cao, W.-s.; Lu, X.-s.; Lin, W.-s.; Gu, A.-z., Parameter comparison of two small-scale natural gas
211
liquefaction processes in skid-mounted packages. Appl. Therm. Eng. 2006, 26, (8–9), 898-904.
212
(13) Foglietta, J. H., LNG production using dual independent expander refrigeration cycles. In Google
213
Patents: 2002.
214
(14) Yuan, Z.; Cui, M.; Xie, Y.; Li, C., Design and analysis of a small-scale natural gas liquefaction
215
process adopting single nitrogen expansion with carbon dioxide pre-cooling. Appl. Therm. Eng. 2014, 64,
216
(1–2), 139-146.
217
(15) Venkatarathnam, G.; Timmerhaus, K. D., Cryogenic mixed refrigerant processes. In Springer New
218
York: 2008.
219
(16) Khan, M. S.; Karimi, I. A.; Bahadori, A.; Lee, M., Sequential coordinate random search for optimal
220
operation of LNG (liquefied natural gas) plant. Energy 2015, 89, 757-767.
221
(17)
222
process using the particle swarm paradigm with nonlinear constraints. Energy 2013, 49, 146-155.
223
(18) Khan, M. S.; Lee, S.; Lee, M., Optimization of single mixed refrigerant natural gas liquefaction
224
plant with nonlinear programming. Asia-Pac. J. Chem. Eng. 2012, 7, S62-S70.
225
(19) Park, J. H.; Khan, M. S.; Lee, M., Modified coordinate descent methodology for solving process
226
design optimization problems: Application to natural gas plant. J. Nat. Gas Sci. Eng. 2015, 27, Part 1, 32-
227
41.
Khan, M. S.; Lee, M., Design optimization of single mixed refrigerant natural gas liquefaction
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 62 of 73
62
228
(20)
Pham, T. N.; Khan, M. S.; Minh, L. Q.; Husmil, Y. A.; Bahadori, A.; Lee, S.; Lee, M.,
229
Optimization of modified single mixed refrigerant process of natural gas liquefaction using multivariate
230
Coggin’s algorithm combined with process knowledge. J. Nat. Gas Sci. Eng. 2016, 33, 731-741.
231
(21) Khan, M. S.; Karimi, I. A.; Wood, D. A., Retrospective and future perspective of natural gas
232
liquefaction and optimization technologies contributing to efficient LNG supply: A review. J. Nat. Gas
233
Sci. Eng. 2017, 45, 165-188.
234
(22) Khan, M. S.; Lee, S.; Rangaiah, G. P.; Lee, M., Knowledge based decision making method for the
235
selection of mixed refrigerant systems for energy efficient LNG processes. Appl. Ener. 2013, 111, 1018-
236
1031.
237
(23) Khan, M. S.; Karimi, I. A.; Lee, M., Evolution and optimization of the dual mixed refrigerant
238
process of natural gas liquefaction. Appl. Therm. Eng. 2016, 96, 320-329.
239
(24) Andress, D.; Watkins, R.; Waynert, J.; Barclay, J.; Breon, S.; Daly, E.; Demko, J.; DiPirro, M.; Hull,
240
J. R.; Kelley, P. J. In Beauty of simplicity: Phillips optimized cascade LNG liquefaction process, AIP
241
Conference Proceedings, pp 91-100, 2004.
242
(25) Qualls, W. R.; ConocoPhillips, L.; Hunter, P. In A Focus on Balance–A Novel Approach Taking the
243
Phillips Optimized Cascade LNG Process Into the Future, 2003.
244
(26) Jensen, J. B.; Skogestad, S., Optimal operation of a mixed fluid cascade LNG plant. Comput. Aided
245
Chem. Eng. 2006, 21, 1569-1574.
246
(27) Bukowski, J.; Liu, Y. N.; Boccella, M. S.; Kowalski, M. L., Innovations in natural gas liquefaction
247
technology for future LNG plants and floating LNG facilities, 2011.
ACS Paragon Plus Environment
Page 63 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
63
248
(28) Bukowski, J. D., Natural gas liquefaction technology for floating LNG facilities.
249
(29) Wood, D. Floating gas liquefaction: Competing technologies make progress.
250
http://www.epmag.com/floating-gas-liquefaction-competing-technologies-make-progress-636831,
251
(accessed 17th June, 2017).
252
(30) Lim, W.; Choi, K.; Moon, I., Current Status and Perspectives of Liquefied Natural Gas (LNG) Plant
253
Design. Ind. Eng. Chem. Res. 2013, 52, (9), 3065-3088.
254
(31) Bradley, A.; Duan, H.; Elion, W.; van Soest-Vercammen, E.; Nagelvoort, R. K. In Innovation in the
255
LNG Industry: Shell's Approach, 24th World Gas Conference, Argentina, 2009.
256
(32) Yang, X. S., Engineering Optimization: An Introduction with Metaheuristic Applications. Wiley:
257
2010.
258
(33) Wang, G. G.; Shan, S., Review of Metamodeling Techniques in Support of Engineering Design
259
Optimization. J. Mech. Design 2006, 129, (4), 370-380.
260
(34) Dimian, A. C., Integrated Design and Simulation of Chemical Processes. Elsevier Science: 2003.
261
(35) Mokhatab, S.; Mak, J. Y.; Valappil, J. V.; Wood, D. A., Handbook of Liquefied Natural Gas.
262
Elsevier Science: 2013.
263
(36) Lee, K.-Y.; Lee, J.-C.; Hwang, J.-H.; Cha, J.-H.; Roh, M.-I. In Determination of the optimal
264
operating condition of the dual mixed refrigerant cycle at the pre-FEED stage of the LNG FPSO topside
265
liquefaction process, The Twenty-first International Offshore and Polar Engineering Conference, 2011;
266
International Society of Offshore and Polar Engineers: 2011.
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 64 of 73
64
267
(37) Hwang, J.-H.; Roh, M.-I.; Lee, K.-Y., Determination of the optimal operating conditions of the dual
268
mixed refrigerant cycle for the LNG FPSO topside liquefaction process. Comp. Chem. Eng. 2013, 49, 25-
269
36.
270
(38)
271
configurations for propane pre-cooled mixed refrigerant (APCI) natural gas liquefaction cycle. Appl.
272
Energy. 2014, 117, 76-86.
273
(39) Pham, T. N.; Long, N. V. D.; Lee, S.; Lee, M., Enhancement of single mixed refrigerant natural gas
274
liquefaction process through process knowledge inspired optimization and modification. Appl. Therm.
275
Eng 2017, 110, 1230-1239.
276
(40) Kikkawa, Y.; Nakamura, M.; Sugiyama, S., Development of Liquefaction Process for Natural Gas.
277
J. Chem. Eng. Jpn 1997, 30, (4), 625-630.
278
(41) Kanoğlu, M., Exergy analysis of multistage cascade refrigeration cycle used for natural gas
279
liquefaction. Int. J. Energy Res. 2002, 26, (8), 763-774.
280
(42) Husnil, Y. A.; Lee, M., Control structure synthesis for operational optimization of mixed refrigerant
281
processes for liquefied natural gas plant. AIChE J. 2014, 60, (7), 2428-2441.
282
(43) Yin, Q.; Li, H.; Fan, Q.; Jia, L.; Weisend, J.; Barclay, J.; Breon, S.; Demko, J.; DiPirro, M.; Kelley,
283
J. P. In Economic analysis of mixed-refrigerant cycle and nitrogen expander cycle in small scale natural
284
gas liquefier, AIP conference proceedings, pp 1159-1165, 2008.
285
(44) Ait-Ali, M. A. Optimal mixed refrigerant liquefaction of natural gas. 1979.
Mortazavi, A.; Alabdulkarem, A.; Hwang, Y.; Radermacher, R., Novel combined cycle
ACS Paragon Plus Environment
Page 65 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
65
286
(45) Lee, G. C.; Smith, R.; Zhu, X. X., Optimal Synthesis of Mixed-Refrigerant Systems for Low-
287
Temperature Processes. Ind. Eng. Chem. Res. 2002, 41, (20), 5016-5028.
288
(46) Shi, Y.; Gu, A.; Wang, R.; Zhu, G., Optimization Analysis of Peakshaving Cycle to Liquefy the
289
Natural Gas-Chapter 174.
290
(47) Aspelund, A.; Berstad, D. O.; Gundersen, T., An Extended Pinch Analysis and Design procedure
291
utilizing pressure based exergy for subambient cooling. Appl. Therm. Eng. 2007, 27, (16), 2633-2649.
292
(48) Nogal, F. D.; Kim, J.-K.; Perry, S.; Smith, R., Optimal Design of Mixed Refrigerant Cycles. Ind.
293
Eng. Chem. Res. 2008, 47, (22), 8724-8740.
294
(49) Hasan, M. M. F.; Razib, M. S.; Karimi, I. A., Optimization of Compressor Networks in LNG
295
Operations. Comp. Aided Chem. Eng. 2009, 27, 1767-1772.
296
(50) Shah, N. M.; Hoadley, A. F. A.; Rangaiah, G. P., Inherent Safety Analysis of a Propane Precooled
297
Gas-Phase Liquified Natural Gas Process. Ind. Eng. Chem. Res. 2009, 48, (10), 4917-4927.
298
(51) Gao, T.; Lin, W.; Gu, A.; Gu, M., Optimization of Coalbed Methane Liquefaction Process Adopting
299
Mixed Refrigerant Cycle with Propane Pre-cooling. J. Chem. Eng Jpn 2009, 42, (12), 893-901.
300
(52) Del Nogal, F. L.; Kim, J.-K.; Perry, S.; Smith, R., Synthesis of mechanical driver and power
301
generation configurations, Part 2: LNG applications. AIChE J. 2010, 56, (9), 2377-2389.
302
(53) Aspelund, A.; Gundersen, T.; Myklebust, J.; Nowak, M. P.; Tomasgard, A., An optimization-
303
simulation model for a simple LNG process. Comp. Chem. Eng. 2010, 34, (10), 1606-1617.
304
(54) Mokarizadeh Haghighi Shirazi, M.; Mowla, D., Energy optimization for liquefaction process of
305
natural gas in peak shaving plant. Energy 2010, 35, (7), 2878-2885.
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 66 of 73
66
306
(55) Gao, T.; Lin, W.; Gu, A.; Gu, M., Coalbed methane liquefaction adopting a nitrogen expansion
307
process with propane pre-cooling. Appl. Energy 2010, 87, (7), 2142-2147.
308
(56)
309
propane pre-cooled mixed refrigerant LNG plant. Appl. Therm. Eng. 2011, 31, (6–7), 1091-1098.
310
(57) Wang, M.; Zhang, J.; Xu, Q.; Li, K., Thermodynamic-Analysis-Based Energy Consumption
311
Minimization for Natural Gas Liquefaction. Ind. Eng. Chem. Res. 2011, 50, (22), 12630-12640.
312
(58) Morin, A.; Wahl, P. E.; Mølnvik, M., Using evolutionary search to optimize the energy consumption
313
for natural gas liquefaction. Chem. Eng. Res. Design 2011, 89, (11), 2428-2441.
314
(59) Tak, K.; Lim, W.; Choi, K.; Ko, D.; Moon, I., Optimization of mixed-refrigerant system in LNG
315
liquefaction process. Comp. aided Chem. Eng. 2011, 29, 1824-1821.
316
(60) Lee, S.; Long, N. V. D.; Lee, M., Design and Optimization of Natural Gas Liquefaction and
317
Recovery Processes for Offshore Floating Liquefied Natural Gas Plants. Ind. Eng. Chem. Res. 2012, 51,
318
(30), 10021-10030.
319
(61)
320
considering operation and design objectives. Comp. Chem. Eng. 2012, 41, 123-133.
321
(62) Wang, M.; Zhang, J.; Xu, Q., Optimal design and operation of a C3MR refrigeration system for
322
natural gas liquefaction. Comp. Chem. Eng. 2012, 39, 84-95.
323
(63) Castillo, L.; Dorao, C. A., Consensual decision-making model based on game theory for LNG
324
processes. Energ. Convers. Manage. 2012, 64, 387-396.
Alabdulkarem, A.; Mortazavi, A.; Hwang, Y.; Radermacher, R.; Rogers, P., Optimization of
Hatcher, P.; Khalilpour, R.; Abbas, A., Optimisation of LNG mixed-refrigerant processes
ACS Paragon Plus Environment
Page 67 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
67
325
(64) Yoon, S.; Cho, H.; Lim, D.; Kim, J., Process design and optimization of natural gas liquefaction
326
processes. Chem. Eng. Trans. 2012, 29, 1585-1590.
327
(65) Wahl, P. E.; Løvseth, S. W.; Mølnvik, M. J., Optimization of a simple LNG process using sequential
328
quadratic programming. Comp. Chem. Eng. 2013, 56, 27-36.
329
(66) Skaugen, G.; Walnum, H.; Hammer, M.; Wahl, P.; Wilhelmsen, Ø.; Kolsaker, K. In Design and
330
optimization of compact heat exchangers in processes used for liquefaction of natural gas, Unpublished
331
paper at International Conference of Applied Energy (ICAE), 2013.
332
(67) Jacobsen, M. G.; Skogestad, S., Active constraint regions for a natural gas liquefaction process. J.
333
Nat. Gas Sci. Eng. 2013, 10, 8-13.
334
(68) Austbø, B.; Wahl, P. E.; Gundersen, T., Constraint handling in stochastic optimization algorithms for
335
natural gas liquefaction processes. In ESCAPE 23, Lappeenranta, 2013.
336
(69) Yoon, J.-I.; Choi, W.-J.; Lee, S.; Choe, K.; Shim, G.-J., Efficiency of Cascade Refrigeration Cycle
337
Using C3H8, N2O, and N2. Heat Transfer Eng. 2013, 34, (11-12), 959-965.
338
(70) He, T. B.; Ju, Y. L., Design and optimization of natural gas liquefaction process by utilizing gas
339
pipeline pressure energy. Appl. Therm. Eng. 2013, 57, (1–2), 1-6.
340
(71) Sun, H. In Simulation study of operating optimization of small-scale LNG plant with SMR cycle.
341
(72)
342
refrigerant processes. J. Nat. Gas Sci. Eng. 2013, 15, 93-105.
343
(73) Xu, X.; Liu, J.; Jiang, C.; Cao, L., The correlation between mixed refrigerant composition and
344
ambient conditions in the PRICO LNG process. Appl. Energ. 2013, 102, 1127-1136.
Wang, M.; Khalilpour, R.; Abbas, A., Operation optimization of propane precooled mixed
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 68 of 73
68
345
(74) He, T. B.; Ju, Y. L., Performance improvement of nitrogen expansion liquefaction process for small-
346
scale LNG plant. Cryogenics 2014, 61, 111-119.
347
(75) He, T. B.; Ju, Y. L., A novel process for small-scale pipeline natural gas liquefaction. Appl. Energ.
348
2014, 115, 17-24.
349
(76) He, T.; Ju, Y., A novel conceptual design of parallel nitrogen expansion liquefaction process for
350
small-scale LNG (liquefied natural gas) plant in skid-mount packages. Energy 2014, 75, 349-359.
351
(77) Lee, I.; Tak, K.; Kwon, H.; Kim, J.; Ko, D.; Moon, I., Design and Optimization of a Pure
352
Refrigerant Cycle for Natural Gas Liquefaction with Subcooling. Ind. Eng. Chem. Res. 2014, 53, (25),
353
10397-10403.
354
(78) Song, K.; Lee, S.; Shin, S.; Lee, H. J.; Han, C., Simulation-Based Optimization Methodology for
355
Offshore Natural Gas Liquefaction Process Design. Ind. Eng. Chem. Res. 2014, 53, (13), 5539-5544.
356
(79) Wang, M.; Khalilpour, R.; Abbas, A., Thermodynamic and economic optimization of LNG mixed
357
refrigerant processes. Energy Convers. Manag. 2014, 88, 947-961.
358
(80) Xu, X.; Liu, J.; Cao, L., Optimization and analysis of mixed refrigerant composition for the PRICO
359
natural gas liquefaction process. Cryogenics 2014, 59, 60-69.
360
(81) Austbø, B., Use of Optimization in Evaluation and Design of Liquefaction Processes for Natural
361
Gas. 2015.
362
(82) Fazlollahi, F.; Bown, A.; Ebrahimzadeh, E.; Baxter, L. L., Design and analysis of the natural gas
363
liquefaction optimization process- CCC-ES (energy storage of cryogenic carbon capture). Energy 2015,
364
90, Part 1, 244-257.
ACS Paragon Plus Environment
Page 69 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
69
365
(83)
He, T.; Ju, Y., Optimal synthesis of expansion liquefaction cycle for distributed-scale LNG
366
(liquefied natural gas) plant. Energy 2015, 88, 268-280.
367
(84) Lee, I.; Tak, K.; Lee, S.; Ko, D.; Moon, I., Decision Making on Liquefaction Ratio for Minimizing
368
Specific Energy in a LNG Pilot Plant. Ind. Eng. Chem. Res. 2015, 54, (51), 12920-12927.
369
(85) Moein, P.; Sarmad, M.; Ebrahimi, H.; Zare, M.; Pakseresht, S.; Vakili, S. Z., APCI- LNG single
370
mixed refrigerant process for natural gas liquefaction cycle: Analysis and optimization. J. Nat. Gas Sci.
371
Eng. 2015, 26, 470-479.
372
(86) Tak, K.; Lee, I.; Kwon, H.; Kim, J.; Ko, D.; Moon, I., Comparison of Multistage Compression
373
Configurations for Single Mixed Refrigerant Processes. Ind. Eng. Chem. Res. 2015, 54, (41), 9992-10000.
374
(87) Austbø, B.; Gundersen, T., Impact of problem formulation on LNG process optimization. AIChE J.
375
2016, 62, (10), 3598-3610.
376
(88) Cao, L.; Liu, J.; Xu, X., Robustness analysis of the mixed refrigerant composition employed in the
377
single mixed refrigerant (SMR) liquefied natural gas (LNG) process. Appl. Therm. Eng. 2016, 93, 1155-
378
1163.
379
(89) Ghorbani, B.; Hamedi, M.-H.; Shirmohammadi, R.; Hamedi, M.; Mehrpooya, M., Exergoeconomic
380
analysis and multi-objective Pareto optimization of the C3MR liquefaction process. Sus. Energ. Tech.
381
Asses. 2016, 17, 56-67.
382
(90)
383
Equipment Cost and Life Expectancy. Ind. Eng. Chem. Res. 2016, 55, (39), 10336-10343.
Lee, I.; Moon, I., Total Cost Optimization of a Single Mixed Refrigerant Process Based on
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 70 of 73
70
384
(91) Mortazavi, A.; Alabdulkarem, A.; Hwang, Y.; Radermacher, R., Development of a robust refrigerant
385
mixture for liquefaction of highly uncertain natural gas compositions. Energy 2016, 113, 1042-1050.
386
(92) Najibullah Khan, N. B.; Barifcani, A.; Tade, M.; Pareek, V., A case study: Application of energy
387
and exergy analysis for enhancing the process efficiency of a three stage propane pre-cooling cycle of the
388
cascade LNG process. J. Nat. Gas Sci. Eng. 2016, 29, 125-133.
389
(93) Park, K.; Won, W.; Shin, D., Effects of varying the ambient temperature on the performance of a
390
single mixed refrigerant liquefaction process. J. Nat. Gas Sci. Eng. 2016, 34, 958-968.
391
(94) Sanavandi, H.; Ziabasharhagh, M., Design and comprehensive optimization of C3MR liquefaction
392
natural gas cycle by considering operational constraints. J. Nat. Gas Sci. Eng. 2016, 29, 176-187.
393
(95) Sun, H.; He Ding, D.; He, M.; Shoujun Sun, S., Simulation and optimisation of AP-X process in a
394
large-scale LNG plant. J. Nat. Gas Sci. Eng. 2016, 32, 380-389.
395
(96) Xiong, X.; Lin, W.; Gu, A., Design and optimization of offshore natural gas liquefaction processes
396
adopting PLNG (pressurized liquefied natural gas) technology. J. Nat. Gas Sci. Eng. 2016, 30, 379-387.
397
(97) Lee, I.; Moon, I., Economic Optimization of Dual Mixed Refrigerant Liquefied Natural Gas Plant
398
Considering Natural Gas Extraction Rate. Ind. Eng. Chem. Res. 2017, 56, (10), 2804-2814.
399
(98) Jackson, S.; Eiksund, O.; Brodal, E., Impact of Ambient Temperature on LNG Liquefaction Process
400
Performance: Energy Efficiency and CO2 Emissions in Cold Climates. Ind. Eng. Chem. Res. 2017, 56,
401
(12), 3388-3398.
402
(99) Na, J.; Lim, Y.; Han, C., A modified DIRECT algorithm for hidden constraints in an LNG process
403
optimization. Energy 2017, 126, 488-500.
ACS Paragon Plus Environment
Page 71 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
71
404
(100) Song, R.; Cui, M.; Liu, J., Single and multiple objective optimization of a natural gas liquefaction
405
process. Energy 2017, 124, 19-28.
406
(101) Ding, H.; Sun, H.; Sun, S.; Chen, C., Analysis and optimization of a mixed fluid cascade (MFC)
407
process. Cryogenics 2017, 83, 35-49.
408
(102) Won, W.; Lee, K. S., An energy-efficient operation system for a natural gas liquefaction process:
409
Development and application to a 100 ton-per-day plant. Comp. Chem. Eng. 2017, 97, 208-219.
410
(103) Won, W.; Kim, J., Bi-level optimizing operation of natural gas liquefaction process. Comp. Chem.
411
Eng. 2017, 96, 87-102.
412
(104) Qyyum, M. A.; Minh, L. Q.; Ali, W.; Hussain, A.; Bahadori, A.; Lee, M., Feasibility study of
413
environmental relative humidity through the thermodynamic effects on the performance of natural gas
414
liquefaction process. Appl. Therm. Eng. 2018, 128, 51-63.
415
(105) Mortazavi, A.; Somers, C.; Alabdulkarem, A.; Hwang, Y.; Radermacher, R., Enhancement of
416
APCI cycle efficiency with absorption chillers. Energy 2010, 35, (9), 3877-3882.
417
(106) Doğan, B.; Ölmez, T., A new metaheuristic for numerical function optimization: Vortex Search
418
algorithm. Info. Sci. 2015, 293, 125-145.
419
(107)
420
multimodal optimization problems inspired by runners and roots of plants in nature. Appl. Soft Comp.
421
2015, 33, 292-303.
Merrikh-Bayat, F., The runner-root algorithm: A metaheuristic for solving unimodal and
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 72 of 73
72
422
(108)
423
Enhancement of LNG plant propane cycle through waste heat powered absorption cooling. Appl. Therm.
424
Eng. 2012, 48, 41-53.
425
(109)
426
Performance enhancement of propane pre-cooled mixed refrigerant LNG plant. Appl. Energy 2012, 93,
427
125-131.
428
Rodgers, P.; Mortazavi, A.; Eveloy, V.; Al-Hashimi, S.; Hwang, Y.; Radermacher, R.,
Mortazavi, A.; Somers, C.; Hwang, Y.; Radermacher, R.; Rodgers, P.; Al-Hashimi, S.,
(110) Pattison, R. C.; Baldea, M., Multistream heat exchangers: Equation-oriented modeling
429
and flowsheet optimization. AIChE Journal. 2015, 61, (6), 1856-1866.
430
(111) Tsay, C.; Pattison, R. C.; Baldea, M., Equation-oriented simulation and optimization of
431
process flowsheets incorporating detailed spiral-wound multistream heat exchanger models.
432
AIChE Journal. 2017, 63, (9), 3778-3789.
433
(112) Lim, W.; Lee, I.; Tak, K.; Cho, J. H.; Ko, D.; Moon, I., Efficient Configuration of a Natural Gas
434
Liquefaction Process for Energy Recovery. Ind. Eng. Chem. Res. 2014, 53, (5), 1973-1985.
435
436 437 438 439 440 441 442
ACS Paragon Plus Environment
Page 73 of 73
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
73
443
Table of Content
444
ACS Paragon Plus Environment