外文翻譯---基于鍍線電除塵器量綱和相似度分析的評(píng)估（英文）

1、Assessment of plate–wire electrostatic precipitators based on dimensional and similarity analysesF.J. Gutiérrez Ortiz ?, B. Navarrete, L. CañadasDepartamento de Ingeniería Química y Ambiental, Univers

2、idad de Sevilla, Camino de los Descubrimientos, s/n. 41092 Sevilla, Spaina r t i c l e i n f oArticle history:Received 2 February 2010Received in revised form 4 April 2011Accepted 6 April 2011Available online 21 April 20

3、11Keywords:Dimensional analysisPhysical similarityPilot plantModelingElectrostatic precipitatora b s t r a c tThis paper is focused towards dimensional analysis in ESP model building, showing both the reduction ineffort

4、and more effective modeling that can result. Electrostatic precipitators (ESP) are widely used inindustry today and much research has been carried out during the last decades. However, dimensionalanalysis is still an uns

5、ettled matter, in spite of it allows to reduce the number of parameters necessaryfor defining the ESP performance, provides a reliable scaling-up of the desired operating conditions fromthe pilot-scale to full-scale plan

6、t (based on the invariance of the pi-space) as well as a consistent extrap-olation within the range covered by dimensionless numbers, and gives a greater flexibility in choice ofparameters. This analysis together with th

7、e similarity analysis is presented in this work, in order toobtain a functional dependence between a target number and a set of few dimensionless numbers.The target selected has been the ratio of particle dust concentrat

8、ion at the outlet of the ESP and thatat the inlet of the ESP. Thus, after doing these analyses, several quite reduced models have been formu-lated theoretically and later tested and validated with experimental data obtai

9、ned in a pilot ESP, in orderto show an application of the study. By this way, a non-linear regression model matches well with exper-imental data from a pilot plant.? 2011 Elsevier Ltd. All rights reserved.1. Introduction

10、The electrostatic precipitators (ESPs) are mainly used for the reduction of fly ash emission from industrial boilers fired with lig- nite or bituminous coal [1]. Wire–plate electrostatic filters are commonly used in indu

11、strial applications. In wire–plate ESP charg- ing of particles occurs by using high-voltage power. The basic prin- ciple underlying the solid pollutant removal process is to charge the particulate matter by means of coro

12、na-generated ions, which then move towards the collecting plates under the effect of the ap- plied electric field. In practical ESP configurations the electric field is generated by high voltages applied to a row of emit

13、ting wires centrally placed between two parallel earthed collecting plates. The cross-section of the wires may be round, square or further complicated in order to increase the local electrical field and thus the corona d

14、ischarge efficiency. The Deutsch–Anderson equation has been used for scale-up purposes for many years. The theoretical migration velocity, w, was found by balancing the electric field (Coulomb) forces on the particle cha

15、rge against the forces of fluid drag (Stokes’s law). How- ever, many authors did not find good agreement between the ori- ginal Deutsch theory and experiments using the theoretical valuesof w. So an ‘effective’ migration

16、 velocity was used, which included any effects not explicitly recognized in the original Deutsch theory, as re-entrainment, diffusion, etc. The effective migration velocity is a function of particle size and is simply a

17、scaling parameter. For the last two decades, many mathematical models have been published and all of them were developed to evaluate particle re- moval efficiency of an electrostatic precipitator (ESP) by predicting thre

18、e interactive fields, namely electrical, gas velocity and particle velocity fields [2]. Some studies have been carried out to calculate the EHD flow field by solving the time-averaged Navier–Stokes equations with a k–e t

19、urbulence closure model [3,4], frequently using a CFD code for carrying out the computations. New modeling research use methods to overcome the simplifications followed in electrostatic precipitator modeling by previous

20、models. The use of CFD for such investigations is fast becoming a powerful and almost essential tool for engineering applications. However, the reliability of CFD analysis requires to solve large fluctuations experienced

21、 by some parameters of the turbulence model located at near wall re- gions and shear layers a particularly fine computational mesh is necessary which inexorably raises the computer run-time requirement. Other examination

22、 approach to the performance of an ESP is a dimensional-analytical and physical similarity treatment, so it is possible to predict and check the performance of a given design for an emission source. This technique is gre

23、atly useful to study the complex physical phenomena involved since it simplifies0016-2361/$ - see front matter ? 2011 Elsevier Ltd. All rights reserved.doi:10.1016/j.fuel.2011.04.006? Corresponding author. Tel.: +34 95 4

24、48 72 60/68; fax: +34 95 446 17 75.E-mail addresses: fjgo@esi.us.es (F.J. Gutiérrez Ortiz), navarrete@esi.us.es (B.Navarrete), canadas@esi.us.es (L. Cañadas).Fuel 90 (2011) 2827–2835Contents lists available at

25、ScienceDirectFueljournal homepage: www.elsevier.com/locate/fuelconsidering an infinite turbulent mixing coefficient (Deutsch mod- el) is an upper bound. According to the technique used, a relationship between the dimensi

26、onless variables is required. In this study, this relation can be described by a functional dependence between the system depuration efficiency, expressed as penetration, and the relevant independent dimensionless variab

27、les:Co Ci ¼ f1HD ; dD ; LD ; rD ;erg;erp; bUDQ g ; K1DU ; lDCiQ g ;qg Ci ;qp Ci ; wD2Q g ;DtDQg ;qvCiQ 3 gU2D5 ;qdl Ci ; dpav D!3. Physical similarityA complete similarity between the model (pilot-scale) and the tec

28、hnical performance (full-scale) implies geometric, electric and fluid dynamics, and gas and particle properties similarity. Only the latter can cause difficulties if, in model experiments, materials used are different fr

29、om those ones in the industrial plant. There- fore, there will be a higher similarity between a pilot-scale and full-scale if both of them are located in the same plant to easily achieve material similarity, just in this

30、 work. Therefore, the functional dependence can be simplified by ver- ifying a series of physical similarities, consisting of the equality of all dimensionless parameters or similarity numbers (p-groups), which are the s

31、caling invariants of a physical problem. This tech- nique can be used to predict large ESP units behaviour from the data acquired in smaller pilot units. Likewise, it can be also used to analyze the effect of varying som

32、e operation parameters on the performance of industrial equipment. Next, the physical simi- larity for pilot-scale and full-scale ESPs is identified.Geometric similarity is produced by operating not only with the same wi

33、re to plate spacing but also with same type of electrode (and equally positioned). Thus, although the following dimension- less parametersHDdD ; LD ; rDare constant, plate height and number of channels in each ESP fielda

34、re different in the two scales considered and, hence, only a partial geometric similarity is achieved. The fluid dynamics similarity is achieved when operating with the same Reynolds number, which represents the convecti

35、ve iner- tial forces to viscosity forces gas ratio:HD ? CiQ g lD ? qg Ci ¼ qgQ g lH ¼ NReBy the adequate combination of the dimensionless parameters related to electrical phenomena and material properties, some

36、 variables with physical meaning have been obtained and facilitate the explanation of the similarity from an electric point of view:K1D U represents how near to the critical stage, in which the corona dis- charge occurs,

37、 the ESP performance is, and its value is the same for both scales. Likewise, multiplying qmCiQ3 gU2D5 by dimensionless gas den- sity, it is obtained a measure of the effect of the electric cohesion on the layer of parti

38、cles collected, controlled by dust resistivity (equal for both scales) and by the non-stationary dust layer build- ing, which may be approached by a time constant built on material properties, and so the similarity is ac

39、hieved again. Multiplying wD2Qg byQg DtD, the dimensionless number wD Dt is obtained and it is a measure of the relative strengths of the drift motion and diffusive motion. This parameter must be taken into account since

40、 the turbulence levels in conventional ESP may be far from what would be required to assure uniform mixing of the particles within a short length, as Deutsch considered in his model. So it is more reasonable to model the

41、 flow in an ESP as one with finite eddy diffusivity. Thus, a high value of this dimensionless parameter leads to a lower collection efficiency (higher penetration levels). Since the eddy diffusivity is a function of the

42、turbulence, the problem to achieve the scale-up may be the prediction of the appropriate diffusivity value to use. This issue will be addressed later. Values of 1 or less for this dimen- sionless parameter are near to De

43、utsch collection. Likewise, multiplying bUD Qg by L D a new dimensionless number is obtained bU Qg=L which gives the ratio between the electric diffusivity of the ions in the gas flow and the diffusivity due to the conve

44、ctive movement of the gas flow but without considering electric contri- bution. This number is quite close to the Electro-hydrodynamics number that is the magnitude of the secondary flow interaction and describes the ele

45、ctrical wind velocity relative to the gas flow velocity [8]. Other parameters with a physical effect on the ESP performance are relative permittivity (erg;erp) and dimensionless densities of gas, particles and dust layer

46、. All of them are constant if the gas and particles characteristics do not change. Finally, the dimensionless average particle diameter (particle diameter to plate distance ratio) is equal for scales and. so it should no

47、t affect to the dependence of the efficiency on the rest of dimen- sionless variables in order to compare scales.4. Model formulationAccording to the previous analyses, the functional dependence between the efficiency an

48、d the rest of dimensionless parameters that arises by applying the p-theorem and physical similarityTable 3Variables related to gas characteristics of an electrostatic precipitator.Symbol Dimensional variable Dimensionle

49、ss variableDescription SI unitl Gas viscosity kg/ms lD CiQg qg Gas density kg/m3 qg CiTable 4Variables related to particle characteristics of an electrostatic precipitator.Symbol Dimensional variable Dimensionlessvariabl

50、e Description SIUnitCo Particle concentration at ESP outlet(variable related to unit efficiency, i.e., thetarget)kg/m3Co Ciqp Particles density kg/m3qp Ciw Migration velocity of particlesa m/s wD2Qg Dt Turbulent mixing c

51、oefficient (eddydiffusivity)m2/sDtD Qgqv Particle resistivity XmqV CiQ3 g U2D5qdl Density of the deposited dust layer kg/m3qdl Cidpav Particle mean size m dpav Da Assuming that Stokes drag law is valid, the migration vel

52、ocity of particlesemerging from a balance of drag force and electrical force is w ¼qpECc 3pldpav, where qp isthe charge that a particle has acquired, E is the local electric field strength, and Cc isthe Cunningham s