An innovative vortex-tube turbo-expander refrigeration cycle forperformance enhancement of nitrogen-based natural-gas liquefactionprocess
Muhammad Abdul Qyyuma, Feng Weia, Arif Hussaina, Adnan Aslam Noonb, Moonyong Leea,⁎
a School of Chemical Engineering, Yeungnam University, Gyeongsan 712-749, Republic of KoreabDepartment of Mechanical Engineering, FET, International Islamic University, Islamabad, 44000, Pakistan
H I G H L I G H T S
• An innovative vortex tube based N2
expansion refrigeration cycle.
• Vortex tube is hybridized with turbo-expander for LNG production.
• The overall energy requirement for theLNG process is reduced significantly.
• The proposed refrigeration cycle canbe implemented to other cryogenicprocesses.
G R A P H I C A L A B S T R A C T
Expander
LNG exchanger Condenser
Compressor
NG
LNG
Expander Vortex tube Expander
LNG exchanger Condenser
Compressor
NG
LNG
A R T I C L E I N F O
Keywords:Natural gas liquefactionLNGN2 expanderVortex tubeHybrid vortex tube and turbo expander
A B S T R A C T
Liquefied natural gas (LNG) has attracted global attention as a more ecological energy source when compared toother fossil fuels. The nitrogen (N2) expander liquefaction is the most green and safe process among the differenttypes of commercial natural gas liquefaction processes, but its relatively low energy efficiency is a major issue.To solve this issue, an energy-efficient, safe, and simple refrigeration cycle was proposed to improve the energyefficiency of the N2 based natural-gas liquefaction process. In the proposed refrigeration cycle, vortex tube as anexpansion device was integrated with turbo-expander in order to reduce the overall required energy for LNGproduction. A well-known commercial simulator Aspen Hysys® v9 was employed for modeling and analysis ofproposed LNG process. The hybrid vortex-tube turbo-expander LNG process resulted in the specific energy re-quirement of 0.5900 kWh/kg LNG. Furthermore, the energy efficiency of the proposed LNG process was alsocompared with previous N2 expander-based LNG processes. The results demonstrated that the proposed hybridconfiguration saved up to 68.5% (depending on feed composition and conditions) in terms of the overall specificenergy requirement in comparison with previous studies.
1. Introduction
Natural gas liquefaction processes are mainly divided into two typesbased on the refrigerants used in the refrigeration loop: nitrogen (N2)
expander and mixed-gas refrigerant (MR). To achieve the maximumpotential benefits of mixed refrigerants, these processes have beenimproved and modified into further categories including single mixedrefrigerant (SMR), Korea single mixed refrigerant (KSMR), dual mixed
https://doi.org/10.1016/j.applthermaleng.2018.08.023Received 21 January 2018; Received in revised form 29 July 2018; Accepted 7 August 2018
refrigerant (DMR) processes cascade, and propane precooled mixedrefrigerant (C3MR) processes. The mixed refrigerant-based processesare highly energy efficient with a high degree of complexity and capitalexpenditures as compared to the N2 expander processes. However, themixed refrigerant processes for liquefied natural gas (LNG) use highlyflammable hydrocarbon-based refrigerants that make them less attrac-tive because of their environmental hazards and safety concerns [1].Appendix A [1,2] summarizes the safety and environmental data fordifferent refrigerants including MR ingredients and nitrogen. The safetydata shows that N2 as a refrigerant has zero occupational exposure limit(OEL), and a zero lower flammability limit (LFL). Furthermore, theAmerican Society of Heating, Refrigerating, and Air-Conditioning En-gineers (ASHRAE) has standardized 34 safety groups (2010a and2010b), and categorized N2 refrigerant in the A1 category (i.e., noflame propagation), whereas, all other refrigerants are categorized intoA3 (i.e., highly flammable). The environmental data shows zero globalwarming potential (GWP) for N2 refrigeration when compared to theMR ingredients.
The selection of key design parameters for an offshore LNG plant isquite different from that of an onshore liquefaction plant in terms of theprocess flexibility, compactness, process safety, environmental con-siderations, and make-up refrigerant storage. Considering more con-scious safety and environmental concerns, the N2 expander LNG processcould be one of most promising candidates for offshore LNG production.Besides of better inherent safety, the N2 expander-based LNG processoffers a high degree of availability (minimal unplanned shut downs)with less capital investment, making them more feasible for offshoreLNG production [1,3–6].
However, the high energy requirement of LNG production is con-sidered as a major challenge associated with the N2 expander-basedLNG processes. Yin et al. [7] performed a comparative study of mixedrefrigerants and a N2-expander LNG process and reported that themixed refrigerant processes consumed only 46% of the energy requiredby the N2-expander process. This high energy consumption is a majorhurdle in the large-scale implementation of the N2 expander-based LNGprocess and makes it less attractive for offshore LNG production.However, this energy consumption varies with environmental plant siteconditions [8,9] and the types of available refrigeration cycle tech-nology for the LNG production such as the SMR, KSMR, DMR, andC3MR processes. Furthermore, the LNG processes requires a significantamount of energy, which is mostly contributed by the shaft work, thecompression stage of the liquefaction cycle. This is mainly dependenton the temperature differences in the heat exchangers [10].
The conventional N2 expander process features a single expanderthat generates the refrigeration effect by expanding N2 from highpressure to low pressure (usually from 100 to 4 bar) [6,11]. This high-pressure reduction results in a significant irreversibility in the cryogenicheat exchanger, which ultimately increases exergy losses and leads tohigh energy consumption [4]. The corresponding irreversibility alongthe length of a cryogenic exchanger is demonstrated by the large
temperature gap between the hot and cold composite curves shown inFig. 1.
One way to overcome this irreversibility is to install the expander ina series combination. By doing this, the optimal pressure drop intervalscan be designed to reduce the irreversibility. However, this approachrequires a high capital investment, high maintenance cost, and complexcontrol system for industrial operations.
Optimization of the design and operational parameters is one of themost popular approaches for improving the energy efficiency of N2
refrigeration processes [3]. Various optimization techniques have beenreported to enhance the energy efficiency of the N2 expansion basedLNG process [5,6,11–15] as well as mixed refrigerant LNG process[16–18]. Another alternative to improve the energy efficiency of theLNG plant is to enhance the refrigeration cycle units, such as com-pressors, cryogenic heat exchangers, and expansion devices [19,20].For example, Fahmy et al. [21] have enhanced the energy efficiency ofthe open cycle Phillips optimized Cascade natural gas liquefactionprocess. They improved the structure of the LNG process by replacingthe Joule-Thompson valve with expanders. Nevertheless, Vatani et al.[22] reported that structural optimization alone cannot be useful en-ough to improve the overall energy efficiency of LNG processes.
This study focused on employing a vortex-tube technique to im-prove the energy efficiency of an existing N2 refrigeration process. Thepotential applications of the vortex tube for liquefaction were first in-troduced by Georges J. Ranque (1933), followed by experimental va-lidation by German physicist Rudolf Hilsch (1947) [23]. The vortextube includes significant merits such as compactness with no movingparts, low cost, zero maintenance, and adjustable cold and hot streams
Nomenclature
N2 nitrogenKSMR Korea single mixed refrigerantSMR single mixed refrigerantDMR dual mixed refrigerantC3MR propane precooled mixed refrigerantCFD computational fluid dynamicsLNG liquefied natural gasMR mixed refrigerantHP high pressureLP low pressureMITA minimum internal temperature approach
NG natural gasTDCC temperature difference between composite curvesTHCC temperature heat-flow composite curveVT vortex tubeW shaft workΔT1 approach temperature in cryogenic heat exchanger CHE-
01ΔT2 approach temperature in cryogenic heat exchanger CHE-
02CHE cryogenic heat exchangerE coolerK compressors and turbines
Fig. 1. Composite curves of conventional N2 expander LNG process.
M.A. Qyyum et al. Applied Thermal Engineering 144 (2018) 117–125
118
[24–26]. The construction of the vortex tube consists of inlet nozzle(s),a diaphragm, a vortex generator, a chamber, a cylindrical tube, aconical valve, and a hot and cold outlet. Fig. 2 shows the schematic of avortex tube.
The working principal of a vortex tube depends on the pressuregradient that causes the energy separation (in terms of hot and cold)from compressed gas. The compressed gas is introduced tangentiallyinto the tube chamber through one or more nozzles. The swirls form atyphoon through the interchangeable vortex generator. The compressedgas leaves the tube from the cold and hot sides. A small conical controlvalve at the hot-gas side is installed to control the temperatures as wellas cold and hot fraction corresponding to the specified application ofthe vortex tube.
Detail expansion mechanism of the vortex tube still has argumentwith open issues to be addressed. However, it has been reported thatthe refrigeration cycle associated with isentropic expansion [27–29]provides a higher coefficient of performance, which is defined as theratio of the useful cooling effect provided to the required compressionenergy, in comparison with the isenthalpic expansion-based refrigera-tion cycles [20,30], as indicated in the famous Maxwell thermodynamicrelation (i.e., dH=TdS+VdP). A more exciting and potentially muchbroader use for the vortex tube is possible if it is integrated into a re-frigeration system to accomplish the required expansion process in aless-irreversible manner [31].
This study addressed the potential benefits of a vortex tube coupledwith turbo expanders to enhance the energy efficiency of the N2 re-frigerant-based liquefaction process. The structure and design para-meters of the proposed hybrid vortex-tube turbo expander were opti-mized to obtain the maximum potential benefits corresponding tominimal energy consumption.
2. Process simulation
A simulation study of the proposed LNG process was carried outusing commercial simulators Aspen Hysys® and Ansys Fluent®. Becausethe rating model of the vortex tube was not available in the AspenHysys®, the energy separation behavior of N2 as a working fluid in thevortex tube was modeled using the rigorous computational fluid dy-namic (CFD) software of Ansys Fluent®. Results from the CFD modelwere embedded into Aspen Hysys® to simulate the proposed LNG pro-cess.
2.1. Process simulation using Aspen Hysys
A Peng-Robinson fluid package with the option of Lee-Keslerequation [32] was chosen for the simulation of proposed LNG process.The thermodynamic states were determined using the Peng-Robinsonequation of state [33]. The enthalpy and entropy of process streamswere calculated using the Lee-Kesler equation of state, because, it hasbeen investigated [34,35] that the Lee-Kesler model is the most accu-rate enthalpy model for gases especially at higher pressures. The mainsimulation and modeling assumptions were:
• heat loss to the environment was negligible,
• the isentropic efficiency of each compressor, expander, and LNGcryogenic turbine were 80% [5,36], 85% [36], and 90% [37], re-spectively,
• water was used as a cooling medium in the after coolers,
• the pressure drop across each water cooler and cryogenic exchangerwas negligible,
• the LNG storage tank pressure was 2 bars, and
• the minimum internal temperature approach (MITA) was chosen as3 °C for both LNG cryogenic exchangers.
The main process simulation basis and feed condition are listed inTable 1.
2.2. CFD model of vortex tube
Energy separation in the vortex tube with N2 (refrigerant) as aworking fluid at a cryogenic temperature was investigated using AnsysFluent®. The major governing equations which are involved in thenumerical analysis of vortex tube are based on the equation of con-tinuity, momentum balance, and energy balance. These governingequations are given below as Eqs. (1)–(3). The detailed governingequations of vortex tube are provided in [26,38,39].
∇ =ρV( ) 0 (1)
−∇ −∇ −∇ + =P ρV V τ ρf( ) 0 (2)
∂∂
+ ∇ = ∇ ∇ − ∇ + ∇ρet
ρUe λ T p U τ U( )
. ( ) . ( ) . .(3)
where ρ is the density, e is the total energy, ∇T is change in tempera-ture, and τ is the viscous dissipation coefficient.
The standard k-ε turbulence model was used to simulate the flowbehavior in the vortex tube. For meshing, the ICEM-CFD applicationwas used to generate the hexahedral structural mesh of 5mm. The CFDmodel was based on the experimental investigation for the Exair™ 708slpm vortex tube used by Skye et al. [40]. Table 2 lists the geometricparameters for the CFD model analysis.
Fig. 2. Schematic of vortex tube (courtesy of AiRTX Vortex Tubes).
Table 1Process simulation basis and feed conditions [6].
Table 2Geometrical parameters of vortex tube used in study [40].
Measurement Value
Working tube length (mm) 106Working tube I.D. (mm) 11.4Nozzle height (mm) 0.97Nozzle width (mm) 1.41Nozzle total inlet area (mm2) 8.2Cold exit diameter (mm) 6.2Cold exit area (mm2) 30.3Hot exit diameter (mm) 11Hot exit area (mm2) 95
M.A. Qyyum et al. Applied Thermal Engineering 144 (2018) 117–125
119
2.2.1. Grid independence studyIn order to avoid interpolation losses, the hexahedral grid has been
generated using Ansys ICEMCFD, as shown in Fig. 3. To make sure thatthe computational results were independent of mesh size, four mesh ofdifferent sizes have been generated. A grid independence study wascarried out for the case of N2 working fluid with inlet temperature andpressure values of 139 K and 7 bar, respectively. Four generated gridswere represented by M1-M4 having sizes of 1.1, 1.8, 2.4 and 3.5 millionelements, respectively. The results of the grid independence study interms of area averaged cold-side temperature against different meshsizes are shown in Fig. 4. Grids M3 and M4 resulted a cold-side tem-perature difference of less than 1%, therefore keeping in view the ac-curacy of results and computational economy grid, M3 has been se-lected and all simulations presented in the current study have beenconducted using the same grid.
2.2.2. Operating conditionsThe selection of the operating conditions of the refrigerant, espe-
cially the optimal correspondence of temperature and pressure to low
compression power, was a critical issue. To date, most researchers[23–25,27–29,31,40–45] have used N2 as a working fluid to study theenergy separation behavior in a vortex tube at a pressure of 5–7 bar andambient temperatures (15–25 °C). Investigations relevant to energyseparation using a vortex tube at cryogenic temperature (<−120 °C)and high pressure (more than 50 bar) were not found in available openliterature by the author. Therefore, to investigate the potential appli-cation of a vortex tube to improve the energy efficiency of the LNGprocess, the authors assumed an inlet pressure of 7 bar. After fixing theinlet pressure, different low temperature values were chosen by ana-lyzing the reported temperature drops in [23–25,27–29,31,40–45]. Atthese temperature values and fixed inlet pressure, the obtained CFDresults were embedded in the Aspen Hysys® simulation model. Then,the liquefaction rate (LNG production) corresponding to the overallenergy consumption was observed as shown in Table 3.
It is clearly seen from Table 3 that at a high inlet temperature for thesame inlet pressure, energy can be saved at the expense of low LNGproduction. Because a liquefaction rate above 90% is considered op-timum, a liquefaction rate of 92% was used in the study. Fig. 5 showsseveral examples of the total temperature contours for different inletconditions. Based on these analyses, the case as shown in Fig. 6(Fig. 5d) was chosen as a vortex-tube model where the inlet tempera-ture and pressure were 139 K and 7 bar, respectively, corresponding toa 92% liquefaction rate and 0.59 kW (optimized value) overall requiredenergy. The static pressure at the cold exit boundary was fixed at 2 bar,
Fig. 3. Hexahedral grid generation for computational analysis.
122.5
123
123.5
124
124.5
125
125.5
0.5 1 1.5 2 2.5 3 3.5 4
Tem
pera
ture
(K)
No. of elements (millions)
Fig. 4. Grid independence study.
Table 3Inlet conditions for CFD modeling of vortex-tube corresponding liquefactionrate and required power.
M.A. Qyyum et al. Applied Thermal Engineering 144 (2018) 117–125
120
while the static pressure of the hot exit boundary was manipulated tovary the cold fraction. Table 4 summarizes the CFD results embedded inAspen Hysys®.
3. Process description
Fig. 7 shows a process flow diagram of the proposed vortex-tube-based liquefaction process. Different streams with the name ‘stream-x’(x= 1, 2, 3, 4, ….…) were used for the process description in Fig. 7.The nitrogen refrigerant stream-1 was compressed up to 32 bar (stream-8) through four compressor stages, each equipped with an after cooler.To avoid the large compression power and reduce the irreversibility of
Fig. 5. Total temperature contours for different inlet conditions.
Fig. 6. Total temperature contours of N2 at 139 K and 7 bars.
Table 4CFD result for N2 working fluid.
Property Value
Inlet temperature (K) 139.0Inlet pressure (bar) 7.0Cold-side temperature (K) 123.0Cold-side pressure (bar) 2.0Cold-mass fraction 0.25Hot-side temperature (K) 145.0Hot-side pressure (bar) 3.0
M.A. Qyyum et al. Applied Thermal Engineering 144 (2018) 117–125
121
Fig. 7. Process flow diagram of vortex-tube-based N2 expander liquefaction process.
M.A. Qyyum et al. Applied Thermal Engineering 144 (2018) 117–125
122
the process, the compression ratio for each compressor was chosen inthe practical range of 1:3. The pressurized stream-9 was then in-troduced into a turbo-expander booster-compression system to furtherincrease the pressure up to 74 bar (stream-10). The high-pressurestream-10 was cooled through a hybrid cooling system combined with awater cooler (E-5) and a cryogenic-plate fin exchanger (CHE-01) beforeentering the expander. After expansion, the stream-13 at 139 K and7 bar was introduced into the vortex tube. The cold-side stream of thevortex tube (stream-14) was used to cool the stream-11 through theCHE-01. The stream-16 from the hot side of the vortex tube was againintroduced into the expander K-7. Stream-17 was used to liquefy thecompressed natural gas through the cryogenic heat exchanger CHE-02.Stream-18 from the CHE-02 as a superheated vapor and stream-15 fromthe CHE-01 were mixed in the mixer Mix-1 (stream-19) and recycled tothe refrigerant compression system. The feed natural gas was com-pressed to 80 bar through a booster compressor K-NG. The LNG productwas finally obtained after the expander K-8 with 8% boil-off gas at apressure slightly higher than atmospheric, i.e., 2.0 bar.
4. Process optimization
The knowledge-based optimization method [6,12,46] was modifiedfor the optimization of the proposed LNG process. Fig. 8 illustrates themodified knowledge-based optimization algorithm used in this study.
The minimization of specific compression energy in the proposedLNG process was chosen as an objective function for optimization. Sincethe refrigerant flow rate and operating pressures have a pronouncedimpact on the overall required compression power and process irre-versibility, these were chosen as the key decision variables for opti-mization. Table 5 lists the decision variables with the lower and upperbounds.
The refrigerant pressure before and after the turbo-expander boostercompressor was designated as ‘P1’ and ‘P2’, respectively. The MITAvalue as a major constraint was chosen as 3 °C in both cryogenic ex-changers by considering the LNG cryogenic exchangers’ transfer of heatwith an MITA value as small as 1–3 °C.
Mathematically, the optimization problem can be formulated as:
∑= ⎛
⎝⎜
⎞
⎠⎟
=
f X W mMin ( ) Min. /i
n
i LNG1 (4)
subject to:
≥T XΔ ( ) 3min1( ) (5)
≥T XΔ ( ) 3min2( ) (6)
where ‘X’ is the vector of decision variables, =X P P P m( , , , )NG N1 2 2 .
5. Results and discussion
Optimization was carried out by adjusting the mean values of thelower and upper bounds of all decision variables. The pressures P1 andP2 were adjusted corresponding to the minimum compression powernecessary to achieve the vortex-tube inlet conditions. The refrigerantflow rate and feed NG boosting pressure were then optimized to achievefeasible MITA values for both cryogenic exchangers. The natural gasboosting pressure had a significant impact on the MITA value of the
main LNG cryogenic exchanger CHE-02, which ultimately affected theoverall compression power of the process. Fig. 9 shows the impact ofthe NG boosting pressure on the MITA value along the length of theexchanger.
It can be seen in Fig. 9 that by increasing the feed NG pressurewithin the lower and upper bound values, the bottom peak of thecomposite curve shifted from an unfeasible region (negative MITA va-lues) to a feasible region. The higher value of feed NG pressure, i.e.,90 bar, resulted in a MITA value greater than 3 °C, which is infeasiblefrom an economic point of view. The circled area indicates that theconstraint MITA value in the main cryogenic exchanger fully satisfiedthe optimal value of 3 °C at 80 bar pressure. Therefore, 80 bar pressurewas chosen as the boosting pressure of the feed NG corresponding to theminimum objective function at a MITA value of 3 °C.
Initially, the N2 single-expander liquefaction process was adopted[6] and modeled to set a benchmark for comparison purposes regardingthe process enhancement and optimization of the proposed hybridvortex-tube turbo-expander LNG process. Fig. 10 compares the com-posite analysis of the proposed LNG process with the conventional N2
single-expander process. Fig. 10(a) and (c) present the TDCC of aconventional N2 single expander and the proposed hybrid vortex-tubeturbo-expander LNG process, respectively. Fig. 10(b) and (d) show theTHCCs of a conventional N2 single expander and the proposed lique-faction process, respectively.
The conventional N2 expander process consists of a single expanderthat expands N2 from high to low pressure (usually 100 bar to 4 bar [6])and generates the refrigeration effect for the NG liquefaction. This high-pressure drop leads to a significant irreversibility in the cryogenic heatexchanger, which ultimately increases the exergy losses in the con-ventional N2 expander process. This irreversibility is demonstrated bythe large temperature gap between the hot and cold composite curvesalong the length of the cryogenic exchanger as shown in Fig. 10(b). Thegap between the hot and cold composite curves was closer to the idealMITA value only near the corner region of the cryogenic exchanger.This large gap away from the MITA value cannot be avoided using a N2
single-expander configuration [6]. On the contrary, the gap betweenthe hot and cold composite curves was drastically reduced in the pro-posed N2 expander process as seen in Fig. 10(d). This small gap along awide region of the cryogenic exchanger implies that the hybrid vortex-tube turbo-expander configuration effectively reduced the associatedirreversibility because of a large pressure drop that led to a largetemperature gradient inside the main LNG cryogenic exchanger.
In the conventional N2 single-expander LNG process, there are largeexergy losses because of the large temperature gradient shown as a high
Table 5Decision variables and bounds.
Decision variables Lower bound Upper bound
Boosting pressure of natural gas, PNG, bar 50 90LP of N2 (stream-8), P1, bar 25 45HP of N2 (stream-10), P2, bar 55 110Flow rate of nitrogen, mN2, kg/h 4.5 9.5
Fig. 9. Effect of feed NG boosting pressure on MITA values.
M.A. Qyyum et al. Applied Thermal Engineering 144 (2018) 117–125
123
approach temperature difference between the composite curves (TDCC)inside the main cryogenic LNG exchanger as shown in Fig. 10(a). Theseexergy losses can also be presented by analyzing the gap between thecomposite curves as shown in Fig. 10(b); there was a large gap betweenthe hot and cold composite curves compared to the composite curves ofthe proposed hybrid vortex-tube turbo-expander LNG process shown inFig. 10(d). Thermodynamically, if the space between the hot and coldTHCCs is minimum, then the liquefaction process is energy efficient.Similarly, the approach temperature difference between composite
curves along the length of a multi-stream cryogenic heat exchangershould be low, and for efficient economical heat transfer, the MITAvalue should be between 1 and 3 °C. In this study, the most-conservativeMITA value of 3 °C was used. In this context, the approach temperaturevalue inside the LNG cryogenic exchanger of a conventional N2 single-expander LNG process is satisfied only at the ends of the exchanger andwhen the MITA value approaches 50 °C as shown by the MITA peaks inthe black ellipse region of Fig. 10(a). However, the approach tem-perature peaks shown by the green circled region of Fig. 10(c) werelower than that of Fig. 10(a) because the minimization of the tem-perature gradient inside the main cryogenic exchanger ultimately re-duced the overall energy requirement for natural gas liquefaction. Theoptimization results are summarized in Table 6. As seen in Table 6, aspecific compression power of 0.5900 kWh was required to produce onekilogram of LNG at the optimal condition.
The composition (mole fraction) of LNG is provided in Table 7.Furthermore, the necessary properties including mass flow, tempera-ture, pressure, enthalpy, and entropy for each stream of optimizedvortex tube-based LNG process are listed in Table S-2 in the Supple-mentary Material lists.
Supplementary data associated with this article can be found, in theonline version, at https://doi.org/10.1016/j.applthermaleng.2018.08.023.
Table 8 compares the specific energy requirement of the proposedconfiguration with other well-established N2 expander processes. The
Fig. 10. TDCC (a) and (c) and THCC (b) and (d) of conventional N2 single expander and proposed hybrid vortex-tube turbo-expander LNG process, respectively.
Table 6Optimization results.
Parameter Value
Boosting pressure of natural gas, PNG, bar 80LP of N2 (stream-8), P1, bar 32HP of N2 (stream-10), P2, bar 74Flow rate of nitrogen, mN2, kg/hr 5.67Specific compression energy (kW/kg-LNG) 0.5900ΔT1 (°C)
unit power of the N2 expander liquefaction process with the novel hy-brid vortex-tube turbo-expander refrigeration cycle was 68.5% less thanthat of the conventional single-stage nitrogen-expansion liquefactionprocess presented by Du et al. [13], 27.3% less than that of that pro-posed by Austbø and Gunderson [11], and 20.8% less than that of theoptimized process of Khan et al. [6].
6. Conclusions
A novel hybrid vortex-tube turbo-expander configuration was pro-posed to minimize the specific compression power of the N2 expanderliquefaction process. The modified knowledge-based optimization al-gorithm was successfully applied to achieve the full potential benefit ofthe proposed process. The integration of the vortex tube with a turboexpander greatly reduced the overall energy consumption up to 68.5%for the natural gas liquefaction process, depending on the design vari-ables/parameters and feed conditions. As a result, the proposed con-figuration showed superior performance over existing N2 expanderprocesses to liquefy natural gas in an eco-friendly manner with a sig-nificantly less energy requirement. Although, the efficiency of in-tegrated vortex tube in terms of temperature drop (inlet temperature –cold side temperature) is poor, i.e., 16 K, in comparison with otherseveral published vortex tube results with nitrogen as a working fluid.The expansion efficiency of the vortex tube could be further improvedby optimizing the geometric and operational parameters (consideringcryogenic conditions) with respect to the overall liquefaction process.The application of the vortex tube is expected to provide a promisingpotential for the enhancement in the refrigeration and the energy ef-ficiency of industrial natural gas liquefaction processes.
Acknowledgements
This research was supported by the Basic Science Research ProgramFoundation of Korea (NRF) funded by the Ministry of Education(2018R1A2B6001566), the Priority Research Centers Program throughthe National Research Foundation of Korea (NRF) funded by theMinistry of Education (2014R1A6A1031189), and EngineeringDevelopment Research Center (EDRC) funded by the Ministry of Trade,Industry & Energy (MOTIE). (No. N0000990)
References
[1] M. Abdul Qyyum, K. Qadeer, M. Lee, Closed-loop self-cooling recuperative N2 ex-pander cycle for the energy efficient and ecological natural gas liquefaction process,ACS Sustain. Chem. Eng. 6 (2018) 5021–5033.
[2] J.M. Calm, G.C. Hourahan, "Physical, safety, and environmental data for currentand alternative refrigerants," Refrigeration for Sustainable Development(Proceedings of the 23rd International Congress of Refrigeration, ICR 2011, Prague,Czech Republic, 2011.08.21-26), International Institute of Refrigeration (IIR/IIF),Paris, France.
[3] M.A. Qyyum, K. Qadeer, M. Lee, Comprehensive review of the design optimizationof natural gas liquefaction processes: current status and perspectives, Ind. Eng.Chem. Res. 57 (2018) 5819–5844.
[4] M.A. Qyyum, K. Qadeer, S. Lee, M. Lee, Innovative propane-nitrogen two-phaseexpander refrigeration cycle for energy-efficient and low-global warming potentialLNG production, Appl. Therm. Eng. 139 (2018) 157–165.
[5] H. Ding, H. Sun, M. He, Optimization of expansion liquefaction processes usingmixed refrigerant N2–CH4, Appl. Therm. Eng. 93 (2016) 1053–1060.
[6] M.S. Khan, S. Lee, M. Getu, M. Lee, Knowledge inspired investigation of selectedparameters on energy consumption in nitrogen single and dual expander processesof natural gas liquefaction, J. Nat. Gas Sci. Eng. 23 (2015) 324–337.
[7] Q. Yin, H. Li, Q. Fan, L. Jia, J. Weisend, J. Barclay, S. Breon, J. Demko, M. DiPirro,J.P. Kelley, Economic analysis of mixed-refrigerant cycle and nitrogen expandercycle in small scale natural gas liquefier, in: AIP Conference Proceedings, vol. 985,AIP, 2008, pp. 1159–1165.
[8] M.A. Qyyum, L.Q. Minh, W. Ali, A. Hussain, A. Bahadori, M. Lee, Feasibility studyof environmental relative humidity through the thermodynamic effects on theperformance of natural gas liquefaction process, Appl. Therm. Eng. 128 (2018)51–63.
[9] K. Park, W. Won, D. Shin, Effects of varying the ambient temperature on the per-formance of a single mixed refrigerant liquefaction process, J. Nat. Gas Sci. Eng. 34(2016) 958–968.
[10] M. Wang, R. Khalilpour, A. Abbas, Thermodynamic and economic optimization of
LNG mixed refrigerant processes, Energy Convers. Manage. 88 (2014) 947–961.[11] B. Austbø, T. Gundersen, Optimization of a single expander LNG process, Energy
Procedia 64 (2015) 63–72.[12] M.S. Khan, S. Lee, M. Hasan, M. Lee, Process knowledge based opportunistic opti-
mization of the N2–CO2 expander cycle for the economic development of strandedoffshore fields, J. Nat. Gas Sci. Eng. 18 (2014) 263–273.
[13] H.P. Du, Y.D. Huang, H.Y. Li, Q.S. Ying, Q.H. Fan, L.X. Jia, Numerical simulationand optimization of small-scale LNG plant for skid mounted, in: 2010 Asia-PacificPower and Energy Engineering Conference, 2010, pp. 1–4.
[14] T. He, Y. Ju, Optimal synthesis of expansion liquefaction cycle for distributed-scaleLNG (liquefied natural gas) plant, Energy 88 (2015) 268–280.
[15] W. Ali, M.A. Qyyum, K. Qadeer, M. Lee, Energy optimization for single mixed re-frigerant natural gas liquefaction process using the metaheuristic vortex searchalgorithm, Appl. Therm. Eng. 129 (2018) 782–791.
[16] W. Won, J. Kim, Bi-level optimizing operation of natural gas liquefaction process,Comput. Chem. Eng. 96 (2017) 87–102.
[17] W. Won, K.S. Lee, Delta-operator-based adaptive model predictive control andonline optimization of a natural gas liquefaction process, Chem. Eng. Sci. 162(2017) 21–32.
[18] W. Won, K.S. Lee, An energy-efficient operation system for a natural gas liquefac-tion process: Development and application to a 100 ton-per-day plant, Comput.Chem. Eng. 97 (2017) 208–219.
[19] A. Mortazavi, C. Somers, A. Alabdulkarem, Y. Hwang, R. Radermacher,Enhancement of APCI cycle efficiency with absorption chillers, Energy 35 (2010)3877–3882.
[20] A. Mortazavi, C. Somers, Y. Hwang, R. Radermacher, P. Rodgers, S. Al-Hashimi,Performance enhancement of propane pre-cooled mixed refrigerant LNG plant,Appl. Energy 93 (2012) 125–131.
[21] M.F.M. Fahmy, H.I. Nabih, M. El-Nigeily, Enhancement of the efficiency of the OpenCycle Phillips Optimized Cascade LNG process, Energy Convers. Manage. 112(2016) 308–318.
[22] A. Vatani, M. Mehrpooya, A. Palizdar, Advanced exergetic analysis of five naturalgas liquefaction processes, Energy Convers. Manage. 78 (2014) 720–737.
[23] N. Bej, K.P. Sinhamahapatra, Exergy analysis of a hot cascade type Ranque-Hilschvortex tube using turbulence model, Int. J. Refrig. 45 (2014) 13–24.
[24] Y. Xue, M. Arjomandi, R. Kelso, The working principle of a vortex tube, Int. J.Refrig. 36 (2013) 1730–1740.
[25] S. Eiamsa-ard, P. Promvonge, Review of Ranque-Hilsch effects in vortex tubes,Renew. Sustain. Energy Rev. 12 (2008) 1822–1842.
[26] S.E. Rafiee, M.M. Sadeghiazad, Experimental and 3D CFD investigation on heattransfer and energy separation inside a counter flow vortex tube using differentshapes of hot control valves, Appl. Therm. Eng. 110 (2017) 648–664.
[27] O. Aydın, B. Markal, M. Avcı, A new vortex generator geometry for a counter-flowRanque-Hilsch vortex tube, Appl. Therm. Eng. 30 (2010) 2505–2511.
[28] M. Mohiuddin, S. Elbel, A fresh look at vortex tubes used as expansion device invapor compression systems, 2014.
[29] Y. Xue, M. Jafarian, A. Choudhry, M. Arjomandi, The expansion process in acounter flow vortex tube, J. Vortex Sci. Technol. 2 (2015) 1.
[30] M. Kanoğlu, Cryogenic turbine efficiencies, Exergy Int. J. 1 (2001) 202–208.[31] G. Nellis, S. Klein, The application of vortex tubes to refrigeration cycles, 2002.[32] B.I. Lee, M.G. Kesler, A generalized thermodynamic correlation based on three-
parameter corresponding states, AIChE J. 21 (1975) 510–527.[33] D.-Y. Peng, D.B. Robinson, A new two-constant equation of state, Ind. Eng. Chem.
Fundam. 15 (1976) 59–64.[34] C. Li, W. Jia, X. Wu, Application of Lee-Kesler equation of state to calculating
compressibility factors of high pressure condensate gas, Energy Procedia 14 (2012)115–120.
[35] Z. Yuan, M. Cui, R. Song, Y. Xie, Evaluation of prediction models for the physicalparameters in natural gas liquefaction processes, J. Nat. Gas Sci. Eng. 27 (Part 2)(2015) 876–886.
[36] T. Gao, W. Lin, A. Gu, M. Gu, Coalbed methane liquefaction adopting a nitrogenexpansion process with propane pre-cooling, Appl. Energy 87 (2010) 2142–2147.
[37] M.A. Qyyum, W. Ali, N.V.D. Long, M.S. Khan, M. Lee, Energy efficiency enhance-ment of a single mixed refrigerant LNG process using a novel hydraulic turbine,Energy 144 (2018) 968–976.
[38] S.E. Rafiee, M.M. Sadeghiazad, Three-dimensional and experimental investigationon the effect of cone length of throttle valve on thermal performance of a vortextube using k–ɛ turbulence model, Appl. Therm. Eng. 66 (2014) 65–74.
[39] H. Khazaei, A.R. Teymourtash, M. Malek-Jafarian, Effects of gas properties andgeometrical parameters on performance of a vortex tube, Scientia Iranica 19 (2012)454–462.
[40] H.M. Skye, G.F. Nellis, S.A. Klein, Comparison of CFD analysis to empirical data in acommercial vortex tube, Int. J. Refrig. 29 (2006) 71–80.
[41] B. Ahlborn, J.U. Keller, R. Staudt, G. Treitz, E. Rebhan, Limits of temperature se-paration in a vortex tube, J. Phys. D Appl. Phys. 27 (1994) 480.
[42] T. Dutta, K. Sinhamahapatra, S. Bandyopadhyay, CFD analysis of energy separationin Ranque-Hilsch vortex tube at cryogenic temperature, J. Fluids 2013 (2013).
[43] T. Farouk, B. Farouk, A. Gutsol, Simulation of gas species and temperature se-paration in the counter-flow Ranque-Hilsch vortex tube using the large eddy si-mulation technique, Int. J. Heat Mass Transf. 52 (2009) 3320–3333.
[45] H.R. Thakare, A.D. Parekh, Computational analysis of energy separation in coun-ter—flow vortex tube, Energy 85 (2015) 62–77.
[46] T.N. Pham, N.V.D. Long, S. Lee, M. Lee, Enhancement of single mixed refrigerantnatural gas liquefaction process through process knowledge inspired optimizationand modification, Appl. Therm. Eng. 110 (2017) 1230–1239.
M.A. Qyyum et al. Applied Thermal Engineering 144 (2018) 117–125
![Turbo Blower for 80 kW Proton Exchange Membrane Fuel Cell ...wseas.us/e-library/conferences/2005corfu/c3/papers/498-271.pdf · Fig. 1 Turbo-Expander of PEM Fuel cell system [2] 2.](https://static.documents.pub/doc/80x56/5ce7bf8588c9932d758cdddb/turbo-blower-for-80-kw-proton-exchange-membrane-fuel-cell-wseasuse-libraryconferences2005corfuc3papers498-271pdf.jpg)
