Thermodynamic analysis of the apci skikda liquefaction process

Abstract

Keywords: APCI; LNG; Mixed Refrigerant; COP; Exergy Efficiency; Optimization; Genetic Algorithm

Introduction

Today, liquefied natural gas (LNG) is considered one of the most important energy carriers. It is produced by three cryogenic gas liquefaction processes: cascade cycles with pure refrigerants, mixed refrigerant cycles and nitrogen expansion cycles. Mixed refrigerant processes are mainly divided into three types: single mixed refrigerant (SMR), dual mixed refrigerant (DMR) and cascaded mixed refrigerant cycles [1]. Among these mixed refrigerant processes,

the APCI-LNG process is the most widespread in the cryogenic industry, currently belongs to the majority of liquefaction plants in the international market, accounts for about 40% of the capital of LNG complexes and has a significant impact on utilities and operating costs. Many researchers have been conducting theoretical and experimental studies and have been interested to the optimization of this liquefaction process for more than three decades. Kikkawa et al. [2] simulated a uniform refrigerant liquefaction process using a pre-cooling loop and an expander

*Corresponding author. *E-mail address: chaim.23.derb@gmail.com This paper was recommended for publication in revised form by Regional Editor Ahmet Selim Dalkılıç Published by Yıldız Technical University Press, İstanbul, Turkey Copyright 2021, Yıldız Technical University. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

with the CHEMCAD III software. Process thermal efficiency has been evaluated based on the required power consumption to produce a unit of LNG, i.e., specific power consumption. The total required power consumption of the refrigerant compressors is 81.5MW which is equivalent to a specific power of 931 kJ/kg of LNG. Cao et al. [3] designed and simulated two types of small-scale NG liquefaction processes in a skid-mounted assembly, their results show that the N2-CH4 expansion cycle precedes the mixed refrigeration cycle due to the absence of propane pre-cooling. The power consumption of the compressors is more influential to specific power consumption, so compression with intercooling should be adopted to lessen the irreversible degree of compression process. Vatani et al. [4] considered two mixed refrigeration cycles for the design of an LNG production process with integrated natural gas liquids recovery. They formulated an optimization problem with minimization of specific energy consumption. Their results showed that not only liquefaction efficiency of the process is considerable (0.414 kWh/kg LNG) but also it can recover the ethane higher than 93.3%. This process can be used for large LNG plants in the natural gas refineries. It can also be said that the overall efficiency will be higher for the leaner feed gases. Wang et al. [5] optimized two types of C3MR processes by studying different objective functions, including shaft work consumption, exergy efficiency and operating expenses. They reported that tree work consumption is the best objective function for the optimization of these processes. Khan and Le [6] optimized SMR and C3MR processes using knowledge based optimization (KBO) method. By considering boiling point difference of the MR components, they improved the energy efficiency of the system. They reduced the compression energy requirement of an SMR process by 10% with warm particles model, while the exergy efficiency of the whole process increases 5%. Kumar and Mishra [7] presented a detailed review for cryogenic systems and made energy and exergy analysis of the system for liquefaction of various gases. Mehrpooya and Ansarinasab [8] conducted an energy and exergy analysis for the Linde SMR and APCI single mixed refrigeration processes. Their results show that the exergy efficiency and exergy destruction rate of the APCI process are 45% and 72.24MW respectively. Derbal al. [9] made an exergy analysis of the mixed refrigeration APCI-LNG Skikda and determined exergy efficiency and exergy destruction of the MR compressors and gas turbine. Their results show that the inlet temperature and inlet pressure of MR compressors has an impact on the evolution of the exergy efficiency. Wang et al. [10] experimentally investigated the effect of operating pressures and refrigerant flow compositions on the cycle performance; they concluded that the openings of two throttle valves had significant impacts on operating pressures. Hajji et al. [11] performed simulation and sensitivity analysis for two objective functions, specific energy and compressor energy consumption, by the Aspen HYSYS

process simulator and this was done to improve the overall performance and heat recovery of the LNG process output streams by changing the thermodynamic variables of the C3MR mixed refrigerant process. The results obtained show a greater improvement in the objective functions compared to other studies. He et al. [12] proposed a simulation-based optimization methodology to simultaneously determine the components and their respective fractions in the MR. Then; they performed energy and exergy analysis to reveal the relationships between the MR components and the process performance. Their results indicated that energy efficiency and exergy efficiency were strongly influenced by the selection of components for the MR. Aghazadeh et al. [13] carried out a comprehensive thermodynamic analysis and optimization of two new combined refrigeration cycles using the first and second thermodynamic laws and the assumption of a constant area model for the ejector. Their results show that the maximum thermal efficiency and exergy efficiency are 77.3% and 23.7% in cycle I and 87.5% and 23.9% in cycle II, respectively, by optimizing all design parameters. Rehman et al. [14] presented a systematic approach that includes the identification of improvement potentials in the single mixed refrigeration process SMR. To this end, sources of irreversibility are identified using an advanced exergy analysis to identify further improvement potential. The results reveal that 43% of the overall energy destruction associated with the base case of the SMR process can be avoided through process optimization or redesign. Genetic algorithm (GA) is a global optimization method applied for optimizing LNG process. Alabdulkarem et al. [15] applied GA for the optimization of the propane precooled mixed refrigeration cycle and reported the energy consumption of the MR at different compression pinch temperatures using a combination of GA and sequential quadratic programming. Xu et al. [16] modeled and optimized the PRICO process using GA and Aspen Plus software; the result shows that at about the 40th generation, the search identifies the optimal solution. They obtained a relationship between energy consumption and mixed refrigerant composition. Moein et al. [17] analyzed a single mixed refrigeration cycle (SMR) to determine the optimal operating conditions and optimized it with using GA. They minimized the total required work by optimizing eleven variables including the outlet pressures of all compressors and throttling valves and the molar flow rate of the MR components. The results showed that the total required work was reduced by 14% compared to the base case. Nikkho et al. [18] carried out a simulation of two small-scale modified single refrigeration processes for NG liquefaction. They used the GA method to optimize the specific energy consumption and the total exergy destruction rate of the process. Their results show that the specific energy consumption and the total exergy destruction rate are reduced by 19.4% and 25.4%, respectively. On the other

hand, the maximum total exergy efficiency of 47.51 % is obtained. The literature review shows that the APCI cycle is the most efficient for the production of LNG. In this context, fits our contribution. The objective of our work is to determine the efficiency of the APCI Skikda (Algeria) liquefaction process, compare it to other liquefaction processes and seek its improvement. To do so, this cycle will be thermodynamically analyzed. The energy analysis will lead to obtaining the COP of the cycle while the exergy analysis will allow the detection of irreversibilities within the process equipment and the evaluation of the exergy efficiency of the cycle. The effects of important parameters (compressor inlet gas temperature, expander inlet temperature and MCHE outlet temperature) on APCI process performance will be studied and the main results will be described. The genetic algorithm will be used as a method of performance optimization. The optimization results are described and

compared to those of the literature. The details of this work are represented in the following sections. APCI Process Description The APCI process used in Skikda LNG complex includes three refrigeration cycles (mixed refrigeration, propane refrigeration and external propane refrigeration). In this descriptive section, we limit ourselves to the mixed refrigeration cycle, subject of this work. It is mainly composed of a main cryogenic heat exchanger (MCHE), three compressors (C1, C2 and C3) and chillers (c4-c5-...-c11). The process is shown in Figure. 1. The NG flows through the HP, MP, BP propane feed gas coolers (c1, c2 and c3) at (22.46°C, 63 bar). The NG coming from the c3 cooler enters the scrubber column (SC) (Stream 4) at (-35.15°C,

Thermodynamic Analysis

In order to analyze the performance of the system, the influence of the inlet temperature of the compressors, the inlet temperature of the expander, the temperature of the NG after cooling in the MCHE, on the COP and the exergy

Table 2. Main assumptions for the thermodynamic analysis of the basic cycle

Energy Analysis The steady-state energy balance equation is expressed as:

Neglecting the kinetic and potential energy variations ( and gzk), the energy balance equations for each component can be written in terms of specific enthalpy as follows [19]:

m i ,in (hi',out − hi ,in ) = Wi m i ,in (hi ,out − hi= ) ,in

i=C1, C2, C3 where h'i,out is the enthalpy at entropy constant and ηc the compressor isentropic efficiency. For the expander, the energy equation is:

W= m j ,in (h j ,in − h j ,out= ) η j m j ,in (h j ,in − h'j ,out ) (4) j

j=E1, E2 where h'i,out is the enthalpy at entropy constant and ηi the expander isentropic efficiency. For heat exchangers HX1and HX2, the energy equation is:

For the MCHE, the energy equation is obtained from Figure 1: (6)

m NG ( h1 − h4 ) + m LNG ( h9 − h44 ) + m 19 ( h19 − h23 ) + m 24 ( h24 − h28 ) Q = W ∑W − ∑W net

i=C1, C2, C3; j=E1, E2 The input energy is the net work done on the system for the production of LNG, the output energy is the heat supplied by the refrigerants in the NG cooling stages. Exergy Analysis The exergy analysis of a process complements the energy analysis and is used to assess the work potential of the input and output material and heat streams, and to determine the location and magnitude of irreversibility losses. The state-specific exergy is determined from the following equation:

In this study, variations in kinetic and potential exergy have been neglected. By assuming a constant chemical composition of the flows, the chemical exergy term is cancelled out in the exergy balance equations. Therefore, the specific exergy for each arbitrary state can be considered equivalent to the physical exergy. The physical exergy is obtained from the following equation [20]:

where (s) and (h) are the specific entropy and enthalpy at the conditions specified for the species and the subscript 0 represents the environmental conditions. The general exergy balance can be expressed as a rate as follows [21-23]:

Table 3. Formulas used to calculate the exergy destruction of the cycle components Components

where E x ,in is the exergy at the input, E x ,out is the exergy at the output and ∆E x is the exergy destruction. The exergy destruction for each process component is calculated based on the well-known equations summarized as follows in Table 3. The total exergy destruction can be found as the sum of the exergy destruction of each component [24-27]:

The exergy efficiency of the cycle is defined as follows [19]: (11) The thermodynamic properties, including temperature, pressure, mass flow rate, enthalpy, entropy, exergy and steam fraction “x” at each point in the MR cycle of the APCI process are summarized in Appendix 1.

Optimization

The genetic algorithm (GA) is a heuristic search that mimics the process of natural selection and is commonly used to generate useful solutions to optimization problems [28]. The design of the liquefaction process is a highly nonlinear problem with many local optima. To solve the optimization problem, GA is chosen as the optimization method [29-31]. The GA starts working by producing an initialization population. During the optimization, there are several constraints to ensure that the process can run safely and stably. The GA will judge whether the constraints are satisfied or not. If the constraints are satisfied, the GA will calculate the objective function; otherwise, the penalty function will replace the objective function. The best population will be found when the objective function reaches the maximum or minimum value. The ‘Stochastic uniform’ method is used as a selection method in this paper, namely it stochastically chooses a number of chromosomes from the current population and selects the one with the best ‘fitness’. Breeding is carried out

by the ‘Scattered’ crossover method. Then chain mutations occur randomly with minor probability and a new individual is finally formed. In order to save the best solution from each generation, the corresponding optimal fitness solution is compulsively copied into the next generation. The parameters used for the GA are presented in Table 4, as for the variables used for the cycle optimization are presented in Table 5.

Objective Function In this work, the actual optimization problem is the minimization of the net energy consumption of the APCI cycle. In general, the exergy efficiency of the cycle has been defined as an objective function expressed as follows:

In this work, energy and exergy analyses were performed on the APCI cycle. Basic cycle performances described under the operating conditions are presented in Table 6. It consumes a net power of about 102.35 MW, while its COP and exergy efficiency are estimated at 2.66 and 59.99% respectively.

Table 6. Results of the thermodynamic calculation of the basic APCI cycle performances

Results And Discussion

In Eq. 12, X is an adjusted variable vector including the inlet temperature of the compressors (C1, C2 and C3), the inlet temperature of the expanders (E1, E2).

Exergy analysis is applied to assess the extent of exergy destruction in each component of the cycle shown in Figure 2. It is found that compressors occupy the highest proportion of the total exergy destruction rate estimated at 30.43 %, followed by coolers and intercoolers occupying 22.17%. Other important equipments with irreversible operations are the MCHE, expanders, heat exchangers and separators. They contribute respectively with a rate of 19.71%; 8.55%; 7.2% and 7.72% of the total exergy destruction of the liquefaction process. The energy destruction of the throttle valves and pumps are the lowest. Our results are in agreement with the work of the authors [35-38]. It is clear that reducing the exergy destruction in the compressors and coolers would be the best way to reduce the total exergy destruction of this liquefaction process. The main measures to reduce compressor exergy destruction are to adopt multi-stage compression, selecting reasonable inlet parameters and compression ratio for the compressor.

Table 5. Variables limits used for the cycle optimization Parameter

Figure 2. Distribution of exergy destruction rate of the basic cycle components

Thus, the outlet temperature of the compressors is reduced and the heat transfer temperature difference in the coolers is lower. The exergy destruction of the chillers is reduced due to the reduced heat transfer temperature difference. In this work, the numerical simulation of the key process parameters was carried out using the Aspen HYSYS software. The effects of compressor inlet temperature, NG temperature after cooling in the MCHE and MR expander inlet temperature on the COP and exergy efficiency of the APCI process were analyzed. The results proposed are shown in Figure 3; Figure 4 and Figure 5.

Figure 3. Effect of NG temperature after cooling in MCHE on the COP and exergy efficiency

Figure 3 shows the effects of the NG temperature after cooling in MCHE on the COP and exergy efficiency, where the heat transfer for the NG occurs under a pressure of

Conclusions

The mixed refrigeration cycle of the LNG- APCI Skikda process was thermodynamically studied. The results of the energy analysis showed that the net power of the basic APCI cycle is about 102.35 MW and its COP is 2.66. Exergy analysis pinpointed components whose irreversibility is high. The results of the exergy analysis showed that the total exergy destruction of the cycle is approximately 41 MW and its exergy efficiency is 60%. Compressors are the main energy consumer in the APCI process; they occupy with chillers the highest rate of the total exergy destruction 30.43% and 22.17% respectively. The reduction of their exergy losses can be achieved by an adequate choice of the compression ratio as well as the inlet temperature. In order to increase the performance of the APCI cycle, some key parameters have been analyzed. The results of the numerical simulation show that the compressor and expander inlet temperature as well as the NG temperature after cooling in the MCHE affect the COP and the exergy efficiency of the APCI cycle. COP and exergy efficiency will increase with decreasing compressor and expander inlet temperature and likewise with decreasing MCHE NG outlet temperature. To improve the performance of the basic cycle, an optimization was carried. The GA was used as an optimization method considering the net energy consumption as an objective function for the cycle optimization. The optimization results indicate that the power consumed by the compressors is reduced by 2.82% and the exergy destruction of the cycle by 7.23%. The optimal COP and exergy efficiency of the cycle are 2.70 and 62.26% respectively. They are improved by 1.48% and 3.64% respectively.

Nomenclature

Specific exergy, kJ/kg Exergy, MW Enthalpy, kJ/kg Enthalpy at constant entropy, kJ/kg Flow rate, kg/s Pressure, bar Rate of heat flow, MW Entropy, kJ/kg. K Temperature, K Power, MW Vapor fraction

Exergy efficiency Expander isentropic efficiency Exergy destruction

Abbreviations

APCI Air Products and Chemicals Inc C Compressor C3MR Propane pre-cooled mixed refrigerant COP Coefficient of performance KBO Knowledge Based Optimization LNG Liquefied Natural Gas MCHE Main Cryogenic Heat Exchanger MR Mixed Refrigerant NG Natural Gas SMR Single Mixed Refrigerant

Data Availability Statement

The authors confirm that the data that supports the findings of this study are available within the article. Raw data that support the finding of this study are available from the corresponding author, upon reasonable request.

Conflict Of Interest

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethics

There are no ethical issues with the publication of this manuscript.

References

1. h Enthalpy, kJ/kg h' Enthalpy at constant entropy, kJ/kg [1] He T, Karimi IA, Ju Y. Review on the design and m. Flow rate, kg/s optimization of natural gas liquefaction processes P. Pressure, bar for onshore and offshore applications. Chem Eng Q Rate of heat flow, MW Res Des 2018;132:89−114. [CrossRef] s Entropy, kJ/kg. K [2] Kikkawa Y, Nakamura M, Sujiyama S. Development T. Temperature, K of liquefaction process for natural gas. J Chem Eng W Power, MW Japan 1997;30:625−630. [CrossRef] x Vapor fraction [3] Cao WS, Lu XS, Li WS, Gu AZ. Parameter compari- son of two small- scale natural gas liquefaction pro- Greek symbols cesses in skid-mounted packages. Appl Therm Eng ηc Compressor isentropic efficiency 2006;26:898−904. [CrossRef] 796 J Ther Eng, Vol. 9, No. 3, pp. 786−798, May, 2023
2. Vatani A, Mehrpooya M, Tirandazi B. A novel pro- cess configuration for co-production of NGL and A. Energy and exergy investigation of two modified LNG with low energy requirement. Chem Eng single mixed refrigerant processes for natural gas Process Process Intensif 2013;63:16−24. [CrossRef] liquefaction. Comput Chem Eng 2020;140:106854.
3. Wang M., Khalilpour R, Abbas A. Operation optimi- [CrossRef] zation of propane precooled mixed refrigerant pro- [19] Sayyaadi H, Babaelahi M. Exergetic optimization of cesses. J Nat Gas Sci Eng 2013;15:93−105. [CrossRef] a refrigeration cycle for re-liquefaction of LNG boil-
4. Khan MS, Lee M. Design optimization of single off gas. Int J Thermodyn 2010;13:127−133. mixed refrigerant natural gas liquefaction process [20] Kotas TJ. The exergy method of thermal plant anal- using the particle swarm paradigm with nonlinear ysis. Malabar, ABD: Krieger Publishing; 1995. constraints. Energy 2013;49:146−155. [CrossRef] [21] Kabul A, Kizilkan O, Yakut AK. Performance and
5. Kumar D, Mishra RS.Thermal analysis of cryo- exergetic analysis of vapor compression refrigera- genic systems for liquefaction of various gases. tion system with an internal heat exchanger using International J Appl Eng Res 2014;9:3755−3780. a hydrocarbon, isobutane (R600a). Int J Energy Res
6. Mehrpooya M, Ansarinasab H. Exergoeconomic 2008;32:824−836. [CrossRef] evaluation of single mixed refrigerant natural gas [22] Morosuk T, Tsatsaronis G, Schult M. Conventional liquefaction processes. Energy Convers Manag and advanced exergetic analyses: theory and appli- 2015;99:400−413. [CrossRef] cation. Arabian J Sci Eng 2013;38:395−404. [CrossRef]
7. Derbal C, Haouam A, Mzad H. Energy efficiency enhancement of mixed refrigerant process in LNG processes. 1st ed. New York: Springer; 2008. [CrossRef] plant. Energy Procedia 2018;153:154−160. [CrossRef] [24] Vatani A, Mehrpooya M, Palizdar A. Advanced exer-
8. Wang Q, Song Q, Zhang J, Liu R, Zhang S, Chen G. getic analysis of five natural gas liquefaction processes. Experimental studies on a natural gas liquefaction Energy Convers Manag 2014;78:720−737. [CrossRef] process operating with mixed refrigerants and a rec- [25] Palizdar A, Sadrameli SM. Conventional and tifying column. Cryogenics 2019;99:7−17. [CrossRef] advanced exergoeconomic analyses applied to eth-
9. Hajji A, Chahartaghi M, Kahani M. Thermodynamic ylene refrigeration system of an existing olefin plant. analysis of natural gas liquefaction process with pro- Energy Convers Manag 2017;138:474−485. [CrossRef] pane pre-cooled mixed refrigerant process (C3MR). [26] Direk M, Mert MS, Soylu E, Yüksel F. Experimental Cryogenics 2019;103:102978. [CrossRef] investigation of an automotive air conditioning sys-
10. He T, Mao N, Liu Z, Qyyum MA, Lee M, Pravez AM. tem using R444A and R152a refrigerants as alterna- Impact of mixed refrigerant selection on energy and tives of R134a. J Mech Eng 2019;65:212−218. [CrossRef] exergy performance of natural gas liquefaction pro- [27] Mert MS, Dılmac OF, Ozkan S, Karaca F, Bolat E. cesses. Energy 2020;199:117378. [CrossRef] Exergoeconomic analysis of a cogeneration plant
11. Dokandari DA, Khoshkhoo RH, Bidi M, Mafi in an iron and steel factory. Energy 2012;46:78−84. M. Thermodynamic investigation and optimiza- [CrossRef] tion of two novel combined power-refrigeration [28] Holland JH. Adaptation in natural and artificial cycles using cryogenic LNG energy. Int J Refrig systems: an introductory analysis with applica- 2021;124:167−183. [CrossRef] tions to biology, control, and artificial intelligence.
12. Rehman A, Qyyum MA, Qadeer K, Zakir F, He X, Cambridge: MIT Press; 1992. [CrossRef] Nawaz A, et al. Single mixed refrigerant LNG process: [29] Whitley D. A genetic algorithm tutorial. Stat Investigation of improvement potential, operational Comput 1994;4:65−85. [CrossRef] optimization, and real potential for further improve- [30] Mitchell M. An introduction to genetic algorithms. ments. J Clean Prod 202;284:125379. [CrossRef] Reprint ed. Cambridge: MIT Press; 1998.
13. Alabdulkarem A, Mortazavi A, Hwang Y, Radermacher R, Rogers P. Optimization of pro- refrigerant cycle integrated with NGL recovery pro- pane pre-cooled mixed refrigerant LNG plant. Appl cess for small-scale LNG plant. Ind Eng Chem Res Therm Eng 2011;31:1091−1098. [CrossRef] 2014;53:5545−5553. [CrossRef]
14. Xu X, Liu J, Cao L. Optimization and analysis of mixed refrigerant composition for the PRICO mixed refrigerant natural gas liquefaction plant natural gas liquefaction process. Cryogenics with nonlinear programming. Asia-Pac J Chem Eng 2014;59:60−69. [CrossRef] 2012;7:62−70. [CrossRef]
15. Moein P, Sarmad M, Ebrahimi H, Zare M, Pakseresh S, Vakili SZ. APCI-LNG single mixed refrigerant pro- propane-nitrogen two-phase expander refrigeration cess for natural gas liquefaction cycle: analysis and cycle for energy-efficient and low-global warm- optimization. J Nat Gas Sci Eng 2015;26:470−479. ing potential LNG production. Appl Therm Eng [CrossRef] 2018;139:157−165. [CrossRef] J Ther Eng, Vol. 9, No. 3, pp. 786−798, May, 2023 797
16. Zhu J, Zhang W, Liu S, Li Y, Liu M, Yin Q, Li Y. adopting single nitrogen expansion with carbon Experiment and dynamic simulation study on pro- dioxide pre-cooling. Appl Therm Eng 2014;64:139- pane pre-cooling double nitrogen-expander lique- 146. [CrossRef] faction process for medium-pilot LNG plant. Appl [37] He T, Ju Y. Optimal synthesis of expansion liquefac- Therm Eng 2020;170:114994. [CrossRef] tion cycle for distributed-scale LNG (liquefied natu-
17. Li QY, Ju Y. Design and analysis of liquefaction ral gas) plant. Energy 2015;88:268−280. [CrossRef] process for offshore associated gas resources. Appl [38] Nezhad SA, Shabani B, Soleimahadni M. Therm Eng 2010;30:2518−2525. [CrossRef] Thermodynamic analysis of liquefied natural gas
18. Yuan Z, Cui M, Xie Y, Li C. Design and analysis (LNG) production cycle in APCI process. J Therm of a small-scale natural gas liquefaction process Sci 2012; 21:564−571. [CrossRef]

Share and Cite