A new type of dehydration unit of natural gas and its design considerations A new type of dehydration unit of natural gas and its design considerations

A new type of dehydration unit of natural gas and its design considerations

  • 期刊名字:自然科学进展(英文版)
  • 文件大小:454kb
  • 论文作者:LIU Hengwei,LIU Zhongliang,ZHA
  • 作者单位:The Education Ministry Key Laboratory of Enhanced Heat Transfer & Energy Conversation,Shengli Engineering & Research Ins
  • 更新时间:2020-09-15
  • 下载次数:
论文简介

ROGRESS IN NATURAL SCIENCVol. 15. No. 12, December 2005A new type of dehydration unit of natural gas andits design considerationsLIU Hengwei, LIU Zhongliang", ZHANG Jian, GU Keyu and YaN Tingmi(1. The Education Ministry Key Laboratory of Enhanced Heat Transfer Energy Conversation, Beijing Municipal Key Laboratory ofHeat Transfer and Energy Utilization, College of Environmental and Energy Engineering, Beijing University of Technology, Beijing100022, China: 2. Shengli Engineering Research Institute, Shengli (I Field Lid. SINOPEC, Dongying 257026, ChinaRcccivcd March 7, 2005; revised April 15, 2005Abstract A new type of dehydration unit for natural gas is described and iis basir strueture and working principles are presentedThe key factors affecting the performance and dehydration efficiency of the unit such as nucleation rate droplet growth rate, the strengthof the swirl, and the position at which the shock wave occurs are discussed. And accordingly the design considerations of eadof the unit are provided. Experimental investigations on the working performance of the umit justified the design considerationsKeywords: supersonic, swirling now, dehydration, natural gas.Natural gas is contained at high pressure beneath 4], a so-called T M supersonic separator wasthe earth surface in large cavities or porous rock for- troduced, which overcomes all the disadvantages statmations called gas deposits. The main component of ed above, and the experimental results suggested thatthe natural gas is mcthane. Besides methane, it often the supersonic separator could obtain a maximum dewcontains water vapor and many other heavier hydro- point depression of 22-28 C, but the detailed excarbon vapors. Before the natural gas is deliveredperimental conditions were not giventhe customers, most of water and heavy hydrocarbonvapors should be removed from natural gas. TheseIn this paper, a new dehydration unit that iswater and heavy hydrocarbon vapors in natural gasbased on the theory similar to that of the twister su-not only reduce the heating value of natural gas as fuel personi separator but with a different structure is debut also result in various serious problems in gas pro-cribed. The purpose of this study is to find the factors that affect its dehydration efficiency, to give outcess facilities. Changing pressure or temperature may the comprehensive design considerations for eachcausc water and heavy hydrocarbon to condense, resulting in corrosion, water hydrate or even ice blockcomponent of the unit, and to justify the consideraage. Mercury in natural gas can also react with alloytions through the experimental investigationscomponents in downstream equipment to form amal1 Structure and fundamentalsgams. To prevent such problems gas dehydration facilities have been used to eliminate all liquid phases1 Structureand all componcnts that might condense during delinery or consurnption processesFrom the view point of thermodynamics andaerodynamics, a supersonic swirling separator is simiThe typical traditional natural gas dehydration lar to a system that is formed by a turbo-expander, atechniques include absorption, adsorption, refrigera- cyclone gas/liquid separator and a compressor 6)tion and so on. These conventional technologies have Fig. 1 shows the schematic configuration of the sumany advantages such as good dehydration perfpersoni swirling separator. The separator is commance, high reliability but they also have many dis- posed of a Laval nozzle that is used to expand the wetdvantages such as the need of huge facilities, large feed gas tnic velocity which results in a lowinvestment, energy consuming, complex mechanical temp中国煤化工 clone and separatorwork and damage to the environment 2. In Refs. [3, whCN MH Grling flow and sepaSupporledl hy the Specialized Research Fund for the Doctoral Program of Higher Education( Grant No. 20040005008)and thc Opcn Projcct of theKey laboratory of Beijing Municipality of China(2005)*Towhoncorrespondenceshouldbeaddressed.E-mail:Liuzhi(@bjut.edu.cnProgressinNaturalScienceVol.15No.122005www.tandf.co.uk/journals1149rate liquid from the gas and a diffuser that slows strength of swirling is another factor influencing thedown the flow and recovers some of the initial pres- dehydration efficiencysureIn addition, supersonic swirling separator is aval no?According to the theory of aerodynamicsthe position at which the shockwave occurs influences the performance of the unitsignificantly, so the position where the shock waveccur is also a key factor that affects the workieffiFig. 1. Schematic configuration of supersonic swirling separator1, L Aval nozzle; 2, cyclone; 3, diffuser: 4, wet gas inlct; 5, dryIn order to obtain good dehydration efficiency,gas outlet;6, liquid outle②-⑩, thermocouple;①and⑦,h-all the factors stated above should be considered formidity/ termperature transduce2. 1 Laval nozzleundanozzle, and is accelerated to the supersonicand with phasDue to the isentropic acceleration the temperalure andMacroscopically, this will result in the release of lasure of the natural gas will drop, making the nat- tent hcat to the flow and the spontaneous generationural gas mixture supersaturated. Nucleation will take of a liquid droplet cloud, whose properties stronglydepend on the coupling between the flow and the con-place and droplets will start to grow. Subsequently, a densation process itself. The non-equilibrium conden-swirl is induced in the gas nlow by means of a cyclone,which is placed in the tube behind the Laval sation can be separated into two distinct processes,nozzle. The droplets in the flow will be swirled to theamely, homogeneous nuclcation process and dropletwall of the tube. As a result, the core of the flow growth process. The homogeneous nucleation processhrough the tube becomes dry and the boundary layer phase, of stable clusters by the kinetic process of e-part of the flow ncar the tube wall contains most ofthe vapor components. The boundary layer part of droplet growth process refers to the process in whichthe flow is then drawn out of the flow and dischargedstable droplets increasc their size by gaining more andgas inDoriflow. Finally, the dry gas goes into a diffuserFor the supersonic swirling sepa-and some of the pressure is recoveredrator,nucleation rate determines the total liquidamount separated, and the droplet growth rate deter2 Design considerationsmines the droplet size at the outlet of the laval nozzle, which strongly influences the dehydration perforAlthough the principles of the unit are quite sim- mance. If the droplets are too small they will justple, when it comes to design, it is not. The fundaflow along the streamlines of the flow, and many ofmentals show that the basic condition for dehydration them will not be forced to the outside. If they are tois that the wet gas condenses into a two-phase flow in large, theill be too large for them to bethe Laval nozzle. So the nucleation rate, droplet swirled to the outside. Therefore, the successful degrowth rate and the sizes of the droplets in the outlet sign of the unit depends on a delicate balance betweenof the Laval nozzle must have a strong impact on the the flow speed, the strength of the swirl, the nucle-dehydration efficiency. Wet gas first condenses into a ation rate, the growth rate, and so on, which are alllarge amount of small droplets of various sizes, existrelating in the two-phase flow in the form of multiple dis- lendperse phases. The two-phase flow then flows through divYH中国煤化工 Increase the eftsanding of allCNMHGthe cyclone and attains certain strength of swirling2.2 Cycloneresulting in a gas-liquid separation duc to differentcentrifugal forces on the two phases. Thereforc, theCyclone is a swirl generator which is placed in1150www.tandf.co.uk/journalsProgressinNaturalScienceVol.15No.122005the tube behind the Laval nozzle. So its structure may structed of stainless steel, and was 1357 mm longstrongly influence the position at which the shock The specific dimension can be found in Fig. 1, andwave occurs. If the equivalent friction of the cyclone the details about the configuration of the cyclone canis too large the shock wave may occur inside the be found in Ref [5]. To meet the requirement of theLaval nozzle, which may result in subsonic flow and experiment, an indoor test rig was set up, and themay reduce the dehydration cfficicncy(the dehydra- overall apparatus is shown in Fig. 2. Wet air was chotion efficiency increases with gas velocity ) There- sen as the working fluid in this study. The workingfore, the best configuration of the unit is that the fluid was circulated in an open loop, which made proshock wave should be shifted to the inlet of the diffus- vision for filtering, metering, and pre-cooling. TheGA55 typc compressor provided a maximum workingpressure of 1.0 MPa, with air displacement ofIn order to shift the shock wave to the inlet of 600 m /h. The temperature of the fluid inside the su-he diffuser, the cyclone should be designed with very personic swirling separator was measured with 5 copsmall equivalent friction, and the pressure loss of the per-constantan thermocouples and their schematic lo-flow passing through it should be as small as possible. cations are shown in Fig. 1. The accuracy of the thermocouples is±0.1℃, and its estimated maximum2.3 Diffuserpossible error is 0.2 C. Humidity / temperaturea diffuser is used to slow down the flow andtransducers were mounted at both the supersoniccover some of the initial pressure, so that the pressureswirling separator inlet and the dry gas outlet. Theloss of the flow will not be too large through the unittransducer is accurate to +1. 5% in relative humidityThe configuration of the diffuser does not have immeand +0.5 C in temperature, and its estimated maxi-diate influence on the dehydration process, so il can mum error in relative humidity is 2. 4%and that intemperature is 0. 54C. The pressure gauges were located in the inlet of the supersonic swirling separator,3 Experimental apparatus and methodliquid outlet and dry gas outlet. The precision of3.1 Apparmated maximum error is 2%. The flow meters werelocated in the inlet and dry gas outlet The precisionA supersonic swirling separator was manufac- of these flow meters is +1.5%, and their estimatedtured based on the abovc considerations. It was con- maximum error is 2. 54%Fig. 2. Flow chart. 1, comprcssor; 2, humidifier; 3, supersonic swirling seperator; 4, liquid outlet; 5, dry gas outlet; 6, surge tank;7, liquid collectorIn order to study the dew point depression pressure loss through the supersonic swirling separa(ATa)at different pressure loss ratios(y), a surge tor to its inlet pressure, that is,tank was installed downstream the dry gas outletwhich could provide different back pressures for the1中国煤化工(2)supersonic swirling separator. Here the dew point de- whetCNMHG at the inlet(℃),pression(ATa)is defined as the dew point of theTd is the dew point ot the dry gas at outlet(C)let minus that of the dry gas outlet, that is,the pressure at inlet(MPa)and pe the pressure at dryΔT4-Td1-T(1)gas outlet(MPa)And the pressure loss ratio y is defined as the ratio ofProgressinNaturalScienceVol.15No.122005www.tandf.co.uk/journals11513. 2 Experimental procedureIn order to determine the performance of the su-personi swirling separator, the inlet parameterscluding pressure, temperature, humidity and flowrate were set to the prescribed values. The pressurctemperature,flow rate, relative humidity(rH)anddew points at the inlet and outlet of the supersonicseparator were recorded for a givcn back0.3040.at the outlet of dry gaTo begin an experiment, the compressor wasDew putt depression versus pressure loss ratic(Inlet: P1started, and the outlet pressure was adjusted to a giv-0.64MPa;Ta1=31.0℃;Q1≌345.0Nm3/h;RH95%)en value. After the whole system achievestate, the data needed for determiningmance of the separator were acquired by asition computer. By adjusting the dry gas outlet backoressure to another higher value, the above experi0.2mental procedure was repeated0.33Supersonic swirling separator is a Laval nozzley=0.58r=0.71based unit, so there is a critical flow rate. As weknow if the critical flow state is achieved inside thesupersonic swirling separator, then the flow rateAxial position x(mm)through the separator should be a constant and indeig. 4. Axial temperature distribution of the supersonic swirlingpendent of the back pressure according to the theoryseparator(Inlet: P1=0. 64 MPa: Q1 345 0 Nm/h; RH2thermodynanlICsssible flow[, 13)95%)Therefore, if the regulation of the back pressure ofthe dry gas outlet results in change of the volumetricThe horizontal distance is measured from the inthen the flow inside the separator is fullylet of the separator to the outlet of Laval nozzle. Itsubsonic and the experiment should be stoppedcan be seen that the generai trend of all the curves atdifferent pressure loss ratios is basically the sameResults and discussionAnd the temperature decreases suddenly inside theLaval nozzle. Later the temperature ascends gradualDew point depression implying dehydration per- ly, implying that there must be a shock wave occurformance is shown in Fig 3. It is scen from this fig. ring in the vicinity of the position, and the position isure that the dew point depression increases monotoni- shifted to the outlet side of the Laval nozzle as thecally with the increasing of the pressurc loss ratio, pressure loss ratio increasesand it can also be seen that the maximum dew pointdepression of the supersonic swirling separator isFig. 4 also illustrates the comparison of the dryabout 20C. So if the pressure of the natural gas togas outlet temperature with the inlet temperatureOne can observe that these two temperatures have litbe processed is high enough, a satisfactory dehydration performance can be obtainedtle difference, suggesting that the temperature of drygas at outlet is almost recovered completelyFig. 4 illustrates the temperature profile in the5 Conclusionssupersonic swirling separator. Though the temperatures measured by the thermocouples are not the real中国煤化工 showed that the suvalues duc to the inevitable stagnation effect inside persCNMHGatisfactory workingthe supersonic swirling separator, their primary tenpedency along the axis is valuable, through which the dew point depression of about 20 C without any needposition where the shock wave occurs can be deter- of external mechanical power and chemicals. All thishows that the design considerations presented in thiswww.tandf.oo.uk/journalsProgressinNaturalScienceVol.15No.122005paper are reliable5 Gu K.Y., Liu Z L. and Liu h. w. Swirling gas separating devicefor purification. Chinese patent: 2003101053690is, (ii) The position at which the shock wave occurs 6 Zcmansky M.W. Ilcat end Thermodynamica. New York: Me-key factor that affects the dehydration cffGraw-Hill Book Company, 19987 Zhao L J t Zeng D. L, Yuan P. et al. The thermodynamic anal-In this test, the shock wave occurs in the diyysis of the supersonic flow and shock wave of two-phase flow, Jour-part of the Laval nozzle. This may causenal of Engineering Thermophysics, 2001, 22(3): 284-286flow and may reduce the dehydration efficiency, so8 Adam S. and Scherr G. h. Instabilities and bifurexlinn ofaccording to the design considerations, further workinequilibrium two-phase flow. J. Fluid Mechanics, 199734(1):1-28should be done to design a cyclone with less equiva9 Abraham FF. Homogeneous Nucleation Theory, New York: A-lent friction10 Benson R S. Advanced Engineering Thermodynamics. New YorkReferencesPergamon Press, 197711 Shen W.D., Zheng P Z and Jiang D. A. Engineering Thermody-1 Chen J.X. Long distance delivery gas hydrate preventive measuresnics. Beijing: Higher Education Press, 1994and dehydration process, Ocean Oil, 2001, 110: 2336--233912 Kong L. Compressible Fluid Dynamics. Ji nan: Shandong Poly-2 Wang Y. D. Natural Gas Treatment and Processing Technolechnic University Press, 1991Beijing: PetroLeum Industry13 Kirllin V.A., Sychev V.V. and Sheindlin A. E. Engineering3 Okimoto F. and Brouwer J. M. Supersonic gas conditioningThermodynamics. Moscow: Mir Publishers, 1976World Oil,202,223(8):117011784 Twister B V, Kongsberg F M. C. and Subsea A.S.Dermwnstra-lion of Twister for Subsea Application, REP-0000021304, 2002中国煤化工CNMHG

论文截图
版权:如无特殊注明,文章转载自网络,侵权请联系cnmhg168#163.com删除!文件均为网友上传,仅供研究和学习使用,务必24小时内删除。