Comprehensive Review of the Design Optimization of Natural Gas

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


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


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


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.



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


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


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.



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


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


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.



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.


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


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



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


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


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



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.


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


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



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.


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


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



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.


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


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



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


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


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



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


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


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



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.


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


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



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


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


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



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


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


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



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.


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


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



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.


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


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


92.3/ N.G

N.G/N.A

2.0



39434.4 J/mol

Shi et al.,46 2005

90.4/N.G

N.G/N.A

2.0



20069.9 J/mol

Shi et al.,46 2005

95.1/N.G

N.G/N.A

>0.0

Hysys optimizer in original mode


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


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


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


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


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]


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


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


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


Gao et al.,51 2009


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


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


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


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


MINLP

Shaft work,

1475.06 kW

Wang et


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


90.3 MW

Wahl et al.,65 2013

Hysys, PR and SRK

N.G/-155

80.0/N.A

2.6

NLPQLP routine


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]


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


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]



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


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)


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


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


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,


Wang et al.,72 2013

Xu et al.,73 2013


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


Hysys,

[C1,C2,C3,i-C4,N2] =[90.0,4.0,2.0,2.0,2.0]

PR

R410a precooled N2 expander


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)


0.4669 kWh/Nm3

He and Ju,74 2014

95/N.G

80.0/85.0

2.0



0.3734 kWh/Nm3

He and Ju,74 2014

95/N.G

80.0/85.0

2.0


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


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


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



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


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


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


1.0-3.0

GA

Energy storage

0.086 kW

[C1,C2,C3]=[95.0,3.0,2.0]

PR

SMR


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


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]


Khan et al.,16 2015


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


i) Robustness ii) Exergy efficiency iii) Total shaft work Unit power consumption

95/N.G

80.0/85.0

3.0

GA


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


Hysys,

[C1,C2,C3,n-C4,i-C4,N2] =[90.0,5.0,2.0,1.0,1.0,1.0]

PR

N2 single


Hysys,


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


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


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


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


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


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


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


Hysys,

[C1, CO2]=[99.5,0.5] [35oC, 50bar, 1.0 kmol/h]

PR Hysys,


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


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


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


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


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


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]


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


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


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)


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


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



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.


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


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.



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.


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


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)



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.


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


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:



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


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


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).



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


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


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,



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


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


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



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.


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


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.



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


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


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.



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


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


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



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.


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


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



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.


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


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



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


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


73

443

Table of Content

444


Comprehensive Review of the Design Optimization of Natural Gas

Recommend Documents