al

1, Ok

SandSolar panel

emnglepanf throv, evm isr pl

n draas wewithgenerain deHowenels carease i[2e7]

technology have subsequently been conducted mainly as a tonersupplier in electrophotography [10e21]. Numerous other applica-tions for the electrostatic particle transport have been proposed,including control of bubbles in dielectric liquid [22], removal ofradioactive dust in a fusion reactor [23], transport of liquid droplet

use in commercialvely expensive ITOee-dimensional toer supply and in-ted and expensiveese issues we have

developed an improved system that consists of a sand-repellingglass plate with parallel wire electrodes embedded in a coverglass plate of a solar panel and a high-voltage power supply thatgenerates a single-phase rectangular voltage. The alternatingelectrostatic eld generates a standing wave that causes a ipeopmotion of the sand particles on the device, and when airborne, thesand particles are transported downward by gravity [8]. This reportdescribes a basic principle and performance of the system, and howthe performance of the systemwas further improved by improving

* Corresponding author. Tel./fax: 81 3 5286 3914.

Contents lists availab

Journal of Ele

.e l

Journal of Electrostatics 73 (2015) 65e70E-mail address: kawa@waseda.jp (H. Kawamoto).panels are not cleaned frequently.To mitigate this problem, we have developed an automatic

cleaning system that does not need scarce cleaning water butinstead utilizes an alternating electrostatic force [8]. Transportingparticles using electrostatic force was rst developed and imple-mented by Masuda et al. [9], and many investigations of this

However, this technology is not suitable formega solar systems because it requires prohibitielectrodes, the ends of the electrodes must be thrprevent the intersection of phases, and the powterconnections required are relatively complicafor large-scale commercial plants. To mitigate thpanels is not cleaned by rain over a long period of time in an aridregion, the capacity utilization of the power plant is reduced if the

to a transparent conveyer consisting of transparent indium tinoxide (ITO) electrodes printed on a glass substrate [33].Introduction

Solar power generation has growowing to increasing energy demandand economical concerns associated[1], and many mega solar powerplanned and constructed, especiallywhere the sun shines the brightest.frequently in deserts, and solar pastirred-up sand, causing a drastic decphotovoltaic power generation planthttp://dx.doi.org/10.1016/j.elstat.2014.10.0110304-3886/ 2014 Elsevier B.V. All rights reserved.stically in recent yearsll as the environmentalfossil fuel consumptiontion plants are beingserts at low altitudes,ver, sand storms occurn become covered byn the output power of a. Because sand on solar

[24], movement of blood cells in liquid [25], classication of particlesize [26], separation of seed by-products derived from agriculturalprocesses [27], and dust removal from solar panels and solarhydrogen generators [28]. Theoretical and numerical studies ofelectrostatic particle transport have been conducted to clarify themechanism and to support the development by many researchers[10,12,13,21,29]. Cleaning of lunar or Martian dust on solar panelsfor space exploration is another potential application of this tech-nology [30e33]. It has been demonstrated that more than 98% ofthe dust on a glass plate can be removed using electrostatic trav-eling waves generated by a four-phase rectangular voltage appliedElectrostatic cleaning system for remov

Hiroyuki Kawamoto*, Takuya ShibataDepartment of Applied Mechanics and Aerospace Engineering, Waseda University, 3-4-

a r t i c l e i n f o

Article history:Received 20 May 2014Received in revised form16 September 2014Accepted 27 October 2014Available online 7 November 2014

Keywords:CleanerElectrostatic forceMega solar

a b s t r a c t

An improved cleaning systsurface of solar panels. A sicover glass plate of a solarrepelled from the surface othe systemwas further impon the surface of the panelconsumption of this systeciency of mega solar powe

journal homepage: wwwof sand from solar panels

ubo, Shinjuku, Tokyo 169-8555, Japan

has been developed that uses electrostatic force to remove sand from the-phase high voltage is applied to parallel wire electrodes embedded in theel. It has been demonstrated that more than 90% of the adhering sand ise slightly inclined panel after the cleaning operation. The performance ofed by improving the electrode conguration and introducing natural winden when the deposition of sand on the panel is extremely high. The powervirtually zero. This technology is expected to increase the effective ef-ants constructed in deserts at low latitudes.

2014 Elsevier B.V. All rights reserved.

le at ScienceDirect

ctrostatics

sevier .com/locate/elstat

the electrode conguration and introducing natural wind on thesurface of the panel, evenwhen the deposition of sand on the panelis extremely high. The power consumption of this system isextremely low compared to the output power of the solar panel.This technology is expected to increase the effective efciency ofmega solar power generation plants constructed in deserts at lowlatitudes.

Fig. 1 shows a schematic diagram of the system. If the system isoperated intermittently, the sand that has adhered to the coverglass of the solar panels is repelled. On the other hand, if the systemis operated continuously, the sand that approaches the cover glassis also repelled, and thus, the system can protect solar panelsagainst the adhesion of sand.

Six types of sand, collected from desert areas around the world,were used for evaluation. Photographs of the sand particles areshown in Fig. 2, and these specications are summarized in Table 1.

Fig. 1. Schematic diagram of the electrostatic cleaning system that uses a standingwave and gravity to remove sand from a solar panel.

Table 1Specication of sand used for experiments.

Item Unit A B C D E F

Area Namib Japan Eurasia Oceania North America AfricaRelative

permittivity4.2 2.2 3.2 4.3 4.0 5.3

Elongation 0.72 0.53 0.76 0.83 0.81 0.71Angle of

reposedeg 36 38 39 31 34 35

Bulk density g/cm2 1.5 1.4 1.4 3.0 1.7 3.0

H. Kawamoto, T. Shibata / Journal of Electrostatics 73 (2015) 65e7066System conguration

Parallel wire electrodes embedded in the cover glass plate of thesolar panel was employed instead of ITO electrodes to reduce themanufacturing cost of the cleaning plate. Although the wire elec-trodes create a shadow and disturb the absorption of light, this isminimized by using a ne wire and a wide pitch conguration. Thediameter chosen for the wire electrodes was 0.3 mm, and the sizechosen for the pitch between the electrodes was 7 mm.

To mitigate the complexity of the electrode wiring, powersupply, and interconnections, we adopted a standing wave insteadof a traveling wave [34e36]. That is, a single-phase rectangularvoltage was applied to parallel wire electrodes. Because a travelingwave is not generated by the application of a single-phase voltage,particles are not transported in one direction but rather repelledfrom the plate, and when airborne the sand particles are trans-ported downward by gravity. We generated a single-phase rect-angular voltage by using a set of positive and negative ampliersswitched by semiconductor relays that were controlled by amicrocomputer. Because a high slew-rate is not required for thissystem, we employed conventional low-capacity onboard ampli-ers (HUR30-6, Matsusada Precision, Tokyo).Fig. 2. Six types of sand uSand A was commonly used in the experiments unless otherwisespecied.

Results and discussion

Effect of plate inclination

We manufactured a small device for use in the basic investiga-tion of the system. The dimensions of the substrate glass plate were100 100 3 mm. After 0.3-mm-diameter copper wires werearranged on the plate, a thin glass plate, 0.1 mm in thickness, wasadhered using transparent adhesive to make the surface smoothand to prevent insulation breakdown. Cross-sectional drawing ofthe device is shown in Fig. 3.

The device was inclined, and sand was uniformly scattered onthe cover glass. A single-phase rectangular voltage was thenapplied to the parallel electrodes. The experimentwas conducted inan air-conditioned laboratory (20e25 C, 40e60 RH). As shown inFig. 4, the sand particles on the glass plate were repelled andtransported downward, as conrmed by direct observation ofparticle motions using a high-speed microscope camera (Fastcam-max 120 K model 1, Photoron, Tokyo) [21,33,35] and numericalsed for experiments.

calculations. The numerical calculations were on the basis of athree-dimensional hard-sphere model of the distinct element

dition of the location where the solar panels are installed, the

Fig. 3. Cross-sectional drawing of cleaning device.

Fig. 5. Relationship between the inclination of the panel and the cleaning efciency(100-g/m2 surface loading, 0.86 kVp-p/mm, 1 Hz).

H. Kawamoto, T. Shibata / Journal of Electrostatics 73 (2015) 65e70 67method (DEM). Details of the numerical method are reported in theliterature [21,33,35]. The electrostatic eld that determines theCoulomb force and the dielectrophoresis force applied to the sandparticles is calculated by a two-dimensional differential elementmethod in a cyclic domain. Although the dynamic motion of theparticles cannot be conveyed by still images as shown in Fig. 4, weconrmed the calculated and observed motions are in qualitativeagreement by comparing calculated and measured movies. Asdescribed later, the calculated performance agrees well with themeasured results, not only qualitatively but also quantitatively.

Fig. 5 shows the cleaning efciency, i.e., the ratio between theweight of the sand fed onto the panel for 30 s and that after thecleaning operation of the system, versus the inclination of the plate.A cleaning experiment using a four-phase traveling wave was alsoconducted for comparison with single-phase cleaning. High per-formance was achieved even when the plate was only slightly in-clined, and the performance achieved by standing wave cleaningwas almost the same as that for traveling-wave cleaning when theinclination was greater than 20. This suggests that the systemwould even be effective at low latitudes at which solar panels areinstalled at low inclinations.

Effects of pitch, applied voltage, and frequency

Figs. 6 and 7 show the cleaning efciency versus the averagedelectrostatic eld strength determined by the applied voltagedivided by the pitch of the parallel electrodes and the frequency ofthe applied voltage, respectively. The solid curves in Figs. 6 and 7show the calculated results, which agree well with the measuredresults. We observed that producing a high eld strength achievedhigh performance; however, saturation occurred at a high value.Because the applied voltage is limited by the insulation breakdown,which is determined by the electrostatic eld, the system perfor-mance is almost independent of the electrode pitch at the thresholdFig. 4. Observed and calculated motions of dust particles during operation of thesystem (20 inclination, 100-g/m2 surface loading, 0.8 kVp-p/mm, 1 Hz).voltage. The threshold voltage was 9.8 kVp-p for the 10-mm-pitchdevice and 8.4 kVp-p for the 7-mm-pitch device.

The maximum cleaning efciency was approximately 80% andwas achieved at a low frequency (less than 20 Hz). The cleaningperformance decreased at higher frequencies because particlemotion cannot follow the high-speed change of polarity[21,33,35,36]. However, low-frequency operation is not an issuebecause high-speed cleaning is not necessary for this system.

Improved device

The cleaning performance of the system was further improvedby adopting a V-shaped conguration for the wire electrodes, asshown in Fig. 8. An angle of 0 corresponds to the horizontal(original) conguration and an angle of 90 corresponds to thevertical conguration. It is clear that a V-shaped conguration(with angles between 45 and 75) for the electrodes improved thecleaning performance of the system. Careful observation of particlemotion made using the high-speed microscope camera and nu-merical calculations suggest that when the V-shaped congurationis used, some particles on the panel are repelled not only down-ward but also toward the lateral sides of the panel, and this phe-nomenon increases the cleaning efciency.

Effect of surface loading of sand

El-Shobokshy et al. [37] reported a mean deposition rate of sandon solar panels of 0.387 g/m2/day and a cumulative dust depositionin one month of approximately 10 g/m2 in Riyadh, Saudi Arabia(latitude 24.9). A much higher level of cumulative sand deposition,more than 400 g/m2 in one month, has been recorded at Kuwaitinternational airport [38]. Because the amount of sand depositionon the panel depends on the geographic and meteorological con-Fig. 6. Measured and calculated relationships between the applied voltage (electro-static eld) and cleaning efciency (20 inclination, 100-g/m2 surface loading, 1 Hz).

Fig. 12. The ordinates of the gures represent the power con-

Fig. 7. Measured and calculated relationships between the frequency of the appliedvoltage and the cleaning efciency (20 inclination, 100-g/m2 surface loading,0.86 kVp-p/mm).

Fig. 9. Calculated and measured relationships between the surface loading of sand andthe cleaning efciency (20 inclination, 0.86 kVp-p/mm, 1 Hz). The photograph showsaggregated sand on the panel after operation. The sand bridges adjacent electrodes and

H. Kawamoto, T. Shibata / Journal of Electrostatics 73 (2015) 65e7068appropriate design criterion for sand deposition is not clear;however, it is reasonable to assume that a cleaning system must beable to remove more than 100 g/m2 of sand. Thus, we conductedexperiments and calculations to examine the performance of thecleaning system in cases of high loadings.

Fig. 9 shows the effect of the surface loading of the sand. If anamount of sand greater than 300 g/m2 accumulated on the coverglass, the performance of the cleaning system declined owing to theaggregation of sand that bridges the adjacent electrodes as shownin the photograph in Fig. 9 [24]. However, high performance wasachieved when the surface loading was less than 300 g/m2, whichcorresponds to a sand layer thickness of approximately 0.3 mm.

Although the performance of the cleaning systemwas worse forhigh surface loading conditions, it was experimentally conrmedthat the cleaning performance improved when a weak wind, withthe velocity greater than 1e2m/s, owed parallel to the plate in theinclined direction while the electrostatic cleaner was in operation.It is reported that high speed wind increases the deposition of dust[39]; however, stirring dust particles by the alternating electrostaticforce in the presence of wind enhances the cleaning.

Effect of particle diameter

To determine the sizes of the particles that can be cleaned bythis system, the sand particles were classied into ve groups ac-cording to their particle sizes, determined using sieves, and thecleaning experiment was conducted using each classied sand size.Fig. 10 shows the cleaning efciency versus the particle size. Par-ticles smaller than 25 mm in diameter and those larger than 300 mmin diameter were not cleaned efciently. The reasons for theproblems cleaning solar panels with small and large particles areFig. 8. Relationship between the angle of the V-shaped conguration of the wireelectrodes and the cleaning efciency (20 inclination, 100-g/m2 surface loading,0.86 kVp-p/mm, 1 Hz).different. For small particles, the electrostatic image force andadhesion force are relatively large compared to the Coulomb anddielectrophoretic driving forces. As a result, these particles adhereto the surface of the glass plate, which reduces the performance ofthe cleaning system [33]. For large particles, the large gravitationalforce hinders their bouncing and transport. Because ner particleshave a greater impact than coarser particles on the performance ofa solar panel [2], the cleaning system and its operational schememust be improved and optimized to enhance its performance incleaning small particles from solar panels.

Effect of sand characteristics

The six types of sand shown in Fig. 2 and summarized in Table 1were evaluated to conrm the effectiveness of the cleaning systemfor a range of sand characteristics. Fig.11 shows a comparison of thecleaning performance for the six different types of sand considered.Because many factors affect the performance of the cleaning sys-tem, it is difcult to clarify the reasons for the differences observedin the cleaning performance; however, the electrostatic cleaningsystemwas shown to be effective for a variety of sands. The systemand its operational scheme must be modied and optimized to tthe environmental conditions of the site where the mega solarplant is located.

Power consumption

The power consumption of the cleaning system is shown in

locks on the plate.sumption (the input power to the device) per unit area of the

Fig. 10. Cleaning efciencies for the classied particle sizes (100-g/m2 surface loading,20 inclination, 0.7 kVp-p/mm, 0.2 Hz).

cleaning plate assuming that the power consumption is propor-tional to the area of the plate. Because the transient current owsimmediately after the polarity change, the power consumption isproportional to the frequency. On the other hand, the power con-sumption is proportional to the square of the applied voltage ifinsulation breakdown does not occur [33]. Because the voltage limitfor insulation breakdown is 8.4 kVp-p for the 7-mm-pitch elec-trodes, and the optimal frequency is less than 10 Hz, the powerconsumption is only 0.2 W/m2 under operational conditions of7 kVp-p and 1 Hz. An important factor that inuences the energyconsumption is the operational time of the system, which depends

Fig. 11. Cleaning efciencies for six types of sand (300-g/m2 surface loading, 20

inclination, 0.86 kVp-p/mm, 0.2 Hz).

H. Kawamoto, T. Shibata / Journal ofon the operational scheme, i.e., the number of operational cyclesand the operational period. For example, if the system is operatedfor 30 min a day, the energy consumption is only 0.1 Wh/m2 a day.The energy consumption of this system is extremely low comparedto the typical energy output by the solar cell.

Demonstration

The performance of this system was demonstrated using anactual large solar panel (560 mm 320 mm). The left-hand side ofFig. 12. Power consumption of the electrostatic cleaning system.Fig. 13 shows the sand accumulated on the panel, and the right-hand side shows the panel after the cleaning operation wasapplied to the left half of the solar panel for 3 min. Fig. 13 clearlyshows that the cleaning system is effective in removing accumu-lated sand from a solar panel. The output power of the panel withthe cleaner plate (without dust) was 97% compared to that withoutthe cleaner plate and dust. The power was reduced to 60%when thedust covered the plate, and it was recovered to 90% after operation.Another experiment was conducted to demonstrate that the sandthat approaches the cover glass is repelled if the system is operatedcontinuously. The performance of the cleaning system was betterunder continuous operation than under intermittent operation.Field experiment must be conducted under desert conditions todetermine the optimal operational scheme and to demonstrate theeffectiveness of the system for the specic conditions of interest.

If the surface of the plate gets wet owing to rainfall or dewfall, orif a sandstorm and rainfall occur simultaneously, the accumulatedsand will adhere strongly to the plate owing to liquid bridgingforce. Cleaning experiments conducted under these conditionsconrmed that high performance was achieved after the platedried.

Concluding remarks

An improved cleaning system for removal of the sand that ac-cumulates on solar panels using electrostatic force has beendeveloped. This system is suitable for use in mega solar powerplants constructed in deserts at low latitudes because it is poten-tially inexpensive, requires virtually no power, and operates auto-matically without water and other consumables.

Acknowledgment

The author would like to express his gratitude to HarunaTakahashi and Shogo Shibata (Waseda University) who helped tocarry out the experiments. This work was supported by JSPSKAKENHI Grant Number 23360116.

Fig. 13. Demonstration of the effectiveness of the electrostatic cleaning system on alarge solar panel (150-g/m2 surface loading, 20 inclination, 0.7 kVp-p/mm, 0.2 Hz).

Electrostatics 73 (2015) 65e70 69References

[1] K. Komoto, E. Cunow, C. Breyer, D. Faiman, K. Megherbi, P. van der Vleuten,IEA PVPS Task8: study on very large scale photovoltaic (VLS-PV) systems, in:38th IEEE Photovoltaic Specialists Conference (PVSC), 2012, pp.001778e001782.

[2] M. Mani, R. Pillai, Impact of dust on solar photovoltaic (PV) performance:research status, challenges and recommendations, Renew. Sustain. EnergyRev. 14 (2010) 3124e3131.

[3] A.O. Mohamed, A. Hasan, Effect of dust accumulation on performance ofphotovoltaic solar modules in Sahara environment, J. Basic Appl. Sci. Res. 2(2012) 11030e11036.

[4] H.A. Kazem, T. Khatib, K. Sopian, F. Buttinger, W. Elmenreich, A.S. Albusaidi,Effect of dust deposition on the performance of multi-crystalline photovoltaicmodules based on experimental measurements, Int. J. Renew. Energy Res. 3(2013) 850e853.

[5] D. S. Rajput, K. Sudhakar, Effect of dust on the performance of solar PV panel,Int. J. Chem. Technol. Res. 5 (2013) 1083e1086.

[6] A. Ndiaye, C.M.F. Kebe, P.A. Ndiaye, A. Charki, A. Kobi, V. Sambou, Impact ofdust on the photovoltaic (PV) modules characteristics after an exposition yearin Sahelian environment: the case of Senegal, Int. J. Phys. Sci. 8 (2013)1166e1173.

[7] A. Benatiallah, A.M. Ali, F. Abidi, D. Benatiallah, A. Harrouz, I. Mansouri,Experimental study of dust effect in multi-crystal PV solar module, Int. J.Multidiscip. Sci. Eng. 3 (2012) 1e4.

[8] H. Kawamoto, T. Shibata, Electrostatic cleaning system for removing sand onsolar panels, in: 39th IEEE Photovoltaic Specialist Conference (39th PVSC),2013, p. 39.

[9] S. Masuda, K. Fujibayashi, K. Ishida, H. Inaba, Connement and transportationof charged aerosol clouds via electric curtain, Trans. Inst. Electr. Eng. Jpn 92(1972) 9e18.

[10] J.R. Melcher, E.P. Warren, R.H. Kotal, Theory for nite-phase traveling-waveboundary-guided transport of triboelectried particles, IEEE Trans. Ind. Appl.25 (1989) 949e955.

[11] J.R. Melcher, E.P. Warren, R.H. Kotal, Traveling-wave delivery of single-component developer, IEEE Trans. Ind. Appl. 25 (1989) 956e961.

[12] F.W. Schmidlin, A new nonlevitated mode of traveling wave toner transport,IEEE Trans. Ind. Appl. 27 (1991) 480e487.

[13] F.W. Schmidlin, Modes of traveling wave particle transport and their appli-cations, J. Electrostat. 34 (1995) 225e244.

[14] F.W. Schmidlin, Advances in traveling wave toner transport, in: Proc., NIP15:Int. Conf. on Digital Printing Technologies, Society for Imaging Science andTechnology, Springeld, VA, 1999, pp. 302e305.

[15] K. Taniguchi, S. Morikuni, S. Watanabe, Y. Nakano, T. Sakai, H. Yamamoto,T. Yagi, Y. Yamamoto, Improved driving characteristics for the toner trans-portation system, in: Proc., NIP16: Int. Conf. on Digital Printing Technologies,Society for Imaging Science and Technology, Springeld, VA, 2000, pp.740e742.

[16] K. Taniguchi, H. Yamamoto, Y. Nakano, T. Sakai, S. Morikuni, S. Watanabe,Y. Yamamoto, A new technique for measuring the distribution of charge-to-mass ratio for toner particles with on-line use, J. Imaging Sci. Technol. 47(2003) 224e228.

[17] M.D. Thompson, Y. Gartstein, J.T. LeStrange, Aspects of toner transport on a

[22] M. Aoyama, T. Oda, M. Ogihara, Y. Ikegami, S. Masuda, Electrodynamicalcontrol of bubbles in dielectric liquid using a non-uniform traveling eld,J. Electrostat. 30 (1993) 247e258.

[23] M. Onozuka, Y. Ueda, Y. Oda, K. Takahashi, Y. Seki, I. Aoki, S. Ueda, R. Kurihara,Development of dust removal system using static electricity for fusionexperimental reactors, J. Nucl. Sci. 34 (1997) 1031e1038.

[24] H. Kawamoto, S. Hayashi, Fundamental investigation on electrostatictraveling-wave transport of liquid drop, J. Phys. D Appl. Phys. 39 (2006)418e423.

[25] S. Masuda, M. Washizu, I. Kawabata, Movement of blood cells in liquid bynonuniform traveling eld, IEEE Trans. Ind. Appl. 24 (1988) 217e222.

[26] H. Kawamoto, Some techniques on electrostatic classication of particle sizeutilizing electrostatic travelling wave eld, J. Electrostat. 66 (2008) 220e228.

[27] L.C. Weiss, D.P. Thibodeaux, Separation of seed by-products by an AC electriceld, J. Am. Oil Chem. Soc. 61 (1984) 886e890.

[28] M. Mazumder, M.N. Horenstein, J.W. Stark, P. Girouard, R. Sumner,B. Henderson, O. Sadder, H. Ishihara, A.S. Biris, R. Sharma, Characterization ofelectrodynamic screen performance for dust removal from solar panels andsolar hydrogen generators, IEEE Trans. Ind. Appl. 49 (2013) 1793e1800.

[29] G.Q. Liu, J.S. Marshall, Effect of particle adhesion and interactions on motionby traveling waves on an electric curtain, J. Electrostat. 68 (2010) 179e189.

[30] C.I. Calle, J.L. McFall, C.R. Buhler, S.J. Snyder, E.E. Arens, A. Chen, M.L. Ritz,J.S. Clements, C.R. Fortier, S. Trigwell, Dust particle removal by electrostaticand dielectrophoretic forces with applications to NASA exploration missions,in: Proc. ESA Annual Meeting on Electrostatics, 2008. Paper O1.

[31] J.R. Robison, R. Sharma, J. Zhang, M.K. Mazumder, Computer simulation ofelectrodynamic screens for mars dust mitigation, in: Proc. ESA AnnualMeeting on Electrostatics, 2008. Paper A3.

[32] C.I. Calle, C.R. Buhler, J.L. McFall, S.J. Snyder, Particle removal by electrostaticand dielectric forces for dust control during lunar exploration missions,J. Electrostat. 67 (2009) 89e92.

[33] H. Kawamoto, M. Uchiyama, B.L. Cooper, D.S. McKay, Mitigation of lunar duston solar panels and optical elements utilizing electrostatic traveling-wave,J. Electrostat. 69 (2011) 370e379.

[34] G. Liu, J.S. Marshall, Particle transport by standing waves on an electric cur-

H. Kawamoto, T. Shibata / Journal of Electrostatics 73 (2015) 65e7070traveling wave device, in: Proc., NIP15: Int. Conf. on Digital Printing Tech-nologies, Society for Imaging Science and Technology, Springeld, VA, 1999,pp. 262e265.

[18] R. Kober, Traveling wave transport of conductive toner particles, in: Proc.,NIP16: Int. Conf. on Digital Printing Technologies, Society for Imaging Scienceand Technology, Springeld, VA, 2000, pp. 736e739.

[19] R. Kober, Simulation of traveling wave toner transport, in: Proc., NIP18: Int.Conf. on Digital Printing Technologies, Society for Imaging Science andTechnology, Springeld, VA, 2002, pp. 453e457.

[20] H. Kawamoto, N. Hasegawa, Traveling wave transport of particles and particlesize classication, J. Imaging Sci. Technol. 48 (2004) 404e411.

[21] H. Kawamoto, K. Seki, N. Kuromiya, Mechanism on traveling-wave transportof particles, J. Phys. D Appl. Phys. 39 (2006) 1249e1256.tain, J. Electrostat. 68 (2010) 289e298.[35] H. Kawamoto, T. Miwa, Mitigation of lunar dust adhered to mechanical parts

of equipment used for lunar exploration, J. Electrostat. 69 (2011) 365e369.[36] H. Kawamoto, N. Hara, Electrostatic cleaning system for removing lunar dust

adhering to spacesuits, J. Aerosp. Eng. 24 (2011) 442e444.[37] M.S. El-Shobokshy, A. Mujahid, A.K.M. Zakzouk, Effects of dust on the per-

formance of concentrator photovoltaic cells, Proc. IEEE 132 (1985) 5e8.[38] H. Qasem, T.R. Betts, H. Mullejans, H. AlBusairi, R. Gottschalg, Effect of dust

shading on photovoltaic modules, in: 26th European Photovoltaic Solar En-ergy Conference and Exhibition (26th EU PVSEC), 2011.

[39] D. Goossens, E. Van Kerschaever, Aeolian dust deposition on photovoltaicsolar cells: the effects of wind velocity and airborne dust concentration on cellperformance, Sol. Energy 66 (1999) 277e289.

Electrostatic cleaning system for removal of sand from solar panelsIntroductionSystem configurationResults and discussionEffect of plate inclinationEffects of pitch, applied voltage, and frequencyImproved deviceEffect of surface loading of sandEffect of particle diameterEffect of sand characteristicsPower consumptionDemonstration

Concluding remarksAcknowledgmentReferences