资源描述：
Solar Energy Vol.66,No.4,pp.277–289,1999 1999Elsevier Science Ltd Pergamon PIIS0038–092X9900028–6All rights reserved.Printed in Great Britain 0038-092X/99/-see front matter AEOLIAN DUST DEPOSITION ON PHOTOVOLTAIC SOLAR CELLSTHE EFFECTS OF WIND VELOCITY AND AIRBORNE DUST CONCENTRATION ON CELL PERANCE ,,† DIRK GOOSSENS***and EMMANUEL V AN KERSCHAEVER*** *Laboratory for Experimental Geomorphology,Katholieke Universiteit Leuven,Redingenstraat16bis, B-3000Leuven,Belgium **Erosion and Soil and Water Conservation Group,Wageningen Agricultural University,Nieuwe Kanaal 11,NL-6709PA Wageningen,The Netherlands ***Imec v.z.w.,Kapeldreef75,B-3001Leuven,Belgium Received13May1998;revised version accepted3March1999 Communicated by ROBERT HILL AbstractWind tunnel experiments were conducted to investigate the effect of wind velocity and airborne dust concentration on the drop of photovoltaicPVcell perance caused by dust accumulation on such cells.Perance drop was investigated at four wind velocities and four dust concentrations.I–V characteristics were determined for various intensities of cell pollution.The evolutions of the short circuit current,the open circuit voltage,the maximum power,the reduction of solar intensity received by the cells, and thefill factor variation with increasing cell pollution were examined.The depositionand accumulationof fine aeolian dust on PV cells significantly affects the perance of such cells.Wind velocity has an important impact on cell perance drop,since the drop is larger in high winds than in low winds.However,the wind also affects the sedimentological structure of the dust coating on the cell,resulting in a higher transmittanceof lightfor coatings created during high winds.The wind tunnel experiments indicate that the er effect is more important than the latter,which means that,in general,the drop in PV cell perance due to dust accumulation is larger as wind speed increases.Airborne dust concentration also affects the drop in PV cell perance,since high dust concentrations lead to a higher accumulation on the cell.Contrary to wind speed, airborne dust concentration does not seem to affect the sedimentological structure of dust coatingswith respect to light transmittanceon PV cells.1999Elsevier Science Ltd.All rights reserved. 1.INTRODUCTION ing,and contamination with airborne particulates, either of naturalsoilor industrialcarbon,soot, During the last few years,there has been an other dirtorigin.This paper focuses on the increasing interest in the natural degradation contamination with natural soil dust.According to processes that occur on solar collectors mounted Thomas et al.1985,this may be considered the outdoors.Many freshly installed collectors al- principal source of degradation of collectors ready show a reduction in their electricor mounted outdoors.Apart from diminishing the thermalperance after a few weeks of opera- reflectance of mirrors and the transmittance of cell tionsee Grassi1985for some examples.Since glazing,the presence of airborne particles may the losses continuously increase in the course of further affect the malfunctioning of solar collec-time,collector efficiency may drop to very low tors in different ways.The entry of veryfine dust values after only a few years.Many collectors are inside the electronic sun sensor window of a designed to remain operational for periods of20 collector array may,for example,cause the array years and more;hence the study of the natural to loose track of the sun,as has been reported by degradation of solar cells is of particular impor- Khoshaim et al.1983for a PV plant in Saudi tance. Arabia.Also,airborne particles in the atmosphere According to Bethea et al.1983,the primary affect the amount and properties of the radiation sources of solar collector degradation arehail, finally reaching the collectorssee Santamouris, chemical weathering processes,radiative weather- 1991and Abdelrahman et al.,1988,but this topic is outside the scope of this paper. The pollution of solar cell surfaces by airborne †Author to whom correspondence should be addressed.Tel. particles has been recognized since the early 132-16-32-64-36;fax132-16-32-64-00;e-mail Dirk.Goossensgeo.kuleuven.ac.be1960sDietz,1963.The oldest studies mainly 277 278 D.Goossens and E.Van Kerschaever deal with thermal collectors,and more specifically Probably the most complete study conducted with the effect dust accumulation rts on mirror thus far is that by El-Shobokshy and Hussein reflectance.The majority of these studies discuss1993.In the laboratory,they polluted PV sur-reflectance measurements cuted on outdoor faces with different kinds of dust and measured mirrorsRoth and Pettit,1980;Pettit and Freese,the electric output of the cells under different 1980;Roth and Anaya,1980;Bethea et al.,1981;conditions.The parameters investigated werethe Deffenbaugh et al.,1986.Several authors also short circuit current,the maximum power,the tried to simulate the deposition of dust on mirrors reduction in solar intensity received by the PV in the laboratory,i.e.under more controlled cells,and thefill factor.The effect of particle size conditions,and investigated its effect on mirror was investigated usingfive size fractions ranging reflectanceYoung,1976;Roth and Pettit,1980;from5to80m m.In addition,three kinds of dust Bethea et al.,1983.Studies dealing with thelimestone,cement and carbonwere tested.Both effect of dust deposition on the transparency of parameters significantly affected the reduction of cell glazing are more recent and mainly date from PV cell perance. the early nineties.Outdoor measurements on Although very useful,the work cuted by glazing transparency have been pered by El-Shobokshy and Hussein1993contains some Nahar and Gupta1990,El-Nashar1994and important restrictions.Probably the most impor-Bonvin1995.Laboratory simulations were re-tant of these is that all the PV surfaces they ported by Hasan and Sayigh1992and El-prepared were polluted under zero-wind condi-Shobokshy and Hussein1993.Feuermann and tions.In natural circumstances there is always Zemel1993,in a similar approach,measured some movement of the air,even in very calm, the degradation in pyranometer sensitivity due to apparently windless conditions,due to turbulence dust accumulation on the pyranometer glass.or to natural convection or advection.Since the The influence of collector design on the amount response time of small dust particulates is ex-and distribution of airborne dust on a collector tremely low,fine particle transport will occur was investigated by Goossens et al.1993and even at very low wind speeds.On the other hand, Smits and Goossens1995.Both studies illus-dust pollution of PV cells mounted in deserts trated the need for a careful construction to avoid typically occurs during high wind speeds,for large accumulation on the collector surfaces.example during dust storms,when large amounts Most studies dealing with the effect of dust of sediment are eroded from the ground and the accumulation on the electric perance of PV concentration of particles in the atmosphere is cells date from the last few years,although very high.Long-term measurements of dust depo-several outdoor measurements had been con-sition in the Negev desert have shown that the ducted15years agoKhoshaim et al.,1984;largest deposition always occurs during high wind Al-Busairi and Al-Kandari,1987.The drop in speedsGoossens and Offer,1995.In addition, short circuit current due to dust accumulation has the‘background’wind speedexcluding the un-been measured by Khoshaim et al.1984,Pande usual storm eventsis typically of the order of 21 1992,Khoshaim et al.1983and Pande and1–3m sOffer and Goossens,1990,i.e. Hill1995on outdoor cells and by Katzan and significantly different from zero.Since even low Stidham1991and El-Shobokshy and Hussein winds significantly affect the sedimentological 1993during laboratory experiments.Khoshaim structure of dust coatings onflat surfacesGoos-et al.1984and Katzan and Stidham1991also sens,1991,the zero-wind approach by El-report on the I–V characteristics of the cells.Shobokshy and Hussein1993is an oversimplifi-Measurements of open circuit voltage were per-cation of the real process. ed by Pande1992and Hasan and Sayigh Another restriction of El-Shobokshy and Hus-1992.Maximum power output was investigated sein’s work is that no natural desert dust was used by Al-Busairi and Al-Kandari1987and Hasan in their experiments.The limestone,cement and and Sayigh1992on outdoor cells and by Katzan carbon dust they used is of great significance for and Stidham1991and El-Shobokshy and Hus-cell pollution in urban or industrial areas,but in sein1993in laboratory experiments.In all these many desert plants the pollution by natural soil studies,the negative effect of dust accumulation dust is much more important.Finally,El-Shobok-on PV cell perance was remarkable.Also shy and Hussein’s work only deals with the effect Rolland et al.1990mentioned the negative of particle properties. effect of dust accumulation on PV cell per-In this study,natural soil dust is used as a ance.pollutant.We also add a meteorological com- The effects of wind velocity and airborne dust concentration on cell perance279 ponent to the topic and investigate the effect wind1000W and500W.These werefixed to a metal speed and airborne dust concentration rt on PV rail70cm above the PV cell.Special care was cell perance.Attention will also be paid to taken to avoid any shadowingdirect or indirect the sedimentological structure of the dust coating on the cell.The1000W lamp contained a self- on the PV glazing and to the effect it rts on cooling device,keeping the emission of heat to a cell perance.The aim of the study is,there-minimum.Due to the slightly different spectrum fore,twofoldwe intend to study both the of the halogen lamps compared to the standard aerodynamic and the sedimentological effect that solar simulator,lower current densities were wind speed and airborne dust concentration rt measured during the wind tunnel tests.This is not on PV cell perance.a problem provided the results are expressed in a relativenot in an absolute. The I–V characteristics of the cell were mea- 2.EXPERIMENTAL FACILITIES sured using a Hameg HM203oscilloscope.This -5 The dust experiments were conducted in the made it possible to determine the I–V characteris- aeolian dust wind tunnel of the Laboratory for tics in a very short time intervala few seconds, Experimental Geomorphology,Katholieke Uni-avoiding a warming-up of the cell surface.Tem- versiteit Leuven,Belgium.The tunnel is of the perature of the cell surface was always around closed-return type and contains two test sections.258C. All experiments were carried out in the large Light intensity of the solar simulator was section,which is7.6m long,1.2m wide and0.6measured with an Ophir Nova Laser Power– m high.Energy Monitor.The measuring range of the Dust transport in the tunnel was generated by instrument is from300nm to1100nm,which is means of an Engelhardt laboratory dust-cloud almost exactly the spectral response interval of producer which was connected to the tunnel.This the PV cell used. apparatus ensures a continuous feed to the air current of natural dust particles,and allows the 3.MATERIALS,S AND PROCEDURE operator to adjust dust discharge. A more detailed description of the wind tunnel All experiments were cuted with natural and the dust-cloud producer can be found in aeolian dust prepared from Belgian Brabantian Goossens and Offer1988.loess.The loess was dried,ground and sieved Wind velocities were measured with a standard through a63m m sieve to exclude all sand Pitot tube and a digital Furness FC016manometer particles.In the sifting,some of thefinest par- with an accuracy of0.001mm water pressure.ticles were lost in small dust clouds.The remain- Dust amount on the PV cell was determined using ing sediment consisted of95silt2–63m mand a Mettler PJ3000balance with an accuracy of5clay,2m m.It had a median diameter of 0.001g.30m m,which corresponds closely to the size of The PV cell used in the experiments was a dust particles that settle on the earth’s surface standard multicrystalline silicon cell with a rather during natural dust stormsYaalon and Ganor, low efficiency.It was covered with a titaniumox-1979.Only1of the particles was coarser than ide anti-reflective layer and was encapsulated in50m m.As can be seen in Fig.1,the dust was the following successionglass–EV A–cell–EV A–very well sorted. trilaminatetedlar–polyester–tedlar.It was Before each wind tunnel experiment,the PV chosen because of its high temporal stability and cell was carefully cleaned with a soft cloth and had a size of10cm310cm.The glass type and the I–V characteristics of the clean cell were finish are the same as for a standard production determined using the oscilloscope de- PV module.The spectral response of the cell scribed by Chenming and White1983.No ranges from340nm to1180nmapproximately,decrease was observed in the perance of the with a maximum sensitivity near800nm.clean cell in the course of the experimental Measurements carried out under standard con-programme.The cell was then put into the wind ditions showed a cell efficiency of approximately tunnel.A6m empty fetch was used to allow the 22 11.5.At1solar equivalent1000W m,windand the dustto reach equilibrium con- spectrum AM1.5the current density was29.7ditions before arriving at the PV cell.To avoid 22 mA cm.During the wind tunnel experiments,local aerodynamic disturbances,the wind tunnel no standard solar simulator was available,but anfloor surrounding the cell was carefully adjusted alternative was found in two halogen lamps of so that no roughness changes occurred near the 280 D.Goossens and E.Van Kerschaever Fig.1.Grain-size distribution of the dust used in the experiments. 21 cell’s borders.The cell was always installed in a speed was kept constant at1.86m s,but dust 2121 horizontal position.discharge was set to15kg h,10kg h and5 21 Vertical wind velocity profiles were measured kg h,corresponding to an air dust concentration 23 near the cell,at a fetch of6m.Four windat5cm above the cell surfaceof1.69g m, 2323 conditions were selected,with freestream veloci- 1.13g m and0.56g m,respectively.Thus, 2121 ties varying between0.63m s and2.59m s,four experiments are available to study the effect corresponding closely to the average background of wind velocityat constant air dust concen-speeds recorded at most desert stationsOffer and tration,and another four experiments can be used Goossens,1990.to study the effect of air dust concentrationat Seven wind
展开阅读全文
