Self‐Powered Autonomous Electrostatic Dust Removal for Solar Panels by an Electret Generator

Abstract Solar panels often suffer from dust accumulation, significantly reducing their output, especially in desert regions where many of the world's largest solar plants are located. Here, an autonomous dust removal system for solar panels, powered by a wind‐driven rotary electret generator is proposed. The generator applies a high voltage between one solar panel's output electrode and an upper mesh electrode to generate a strong electrostatic field. It is discovered that dust particles on the insulative glass cover of the panel can be charged under the high electrical field, assisted by adsorbed water, even in low‐humidity environments. The charged particles are subsequently repelled from the solar panel with the significant Coulomb force. Two panels covered with sand dust are cleaned in only 6.6 min by a 15 cm diameter rotary electret generator at 1.6 m s−1 wind speed. Experimental results manifest that the system can work effectively in a wide range of environmental conditions, and doesn't impact the panel performance for long‐term operation. This autonomous system, with its high dust removal efficiency, simplicity, and low cost, holds great potential in practical applications.

Fig. S1 Charging process of a dust particle on a glass cover under an electric field.
An upward electric field is applied, and the particle accumulates positive charges.voltage is applied to the DRU using a commercial power supply with a voltage value of 4.5 kV.The trajectories of the dust particles between the electrodes are recorded with a high-speed microscopic camera.The dimensions of the DRU used in the actual experiment may be not scaled with the diagram in Fig. 2b.The gap between the mesh electrode and the lower electrode is 10 mm, and the mesh electrode pitch distance is 10 mm.without dust (Fig. S23a) and, with dust (Fig. S23b) on the glass surface.The surface potential is measured before and after the ADRS operation, which is  and  , respectively.And the ADRS is operated for 30 s, and then the measurement preparation after the ADRS operation takes about 15 s.The voltage on the DRU is about 9 kV, and each test is repeated 3 times.Based on the results, we can obtain the change in the surface potential on the glass cover before and after the ADRS operation, denoted as  =  - .Charge on the glass surface  can be calculated using the following formula: where  and  are vacuum permittivity and the relative permittivity of the glass, respectively. = 8.854 × 10 F/m and  = 3.9 .S and h is the area and thickness of the glass cover, respectively, and  = 0.01  and ℎ = 0.2 .
Note S2: Estimation of the surface charge on a particle according to the surface charge of the glass cover.
According to the dust charging mechanism, the total amount of charge carried by the particles is equal to that by the glass surface, but with opposite polarity.From Fig.
3d-i,iii, the charge carried by the glass surface, namely, the charge carried by the particles (Q) after dust removal is 192.1 nC and -239.3 nC, respectively.   of  acted on a particle is 9.9 × 10 N.
We only consider the first-order dielectrophoretic force, [S1] which can be given by: where  ⃗ is the equivalent dipole moment, ∇ ⃗ is the electric field gradient,  is the radius of the particle, and  is the real part of the complex Clausius-Mossotti (CM) factor, which is given by  = , where  and  are the relative permittivity of the particle and the medium, respectively.Substituting the particle radius  = 55 , the relative permittivity of the particle and air  = 3.9 ,  = 1 , based on the simulated electric field distribution, the z-direction value of  was obtained, as shown in Fig. S27.The approximate maximum upward z-direction value of  on the glass surface is 1.9 × 10 N.
Fig. S27 Z-direction dielectrophoretic force at the Cut Line.

Note S4: Simulation of the outputs of REG and VMC
According to our previous work [S2], the equivalent circuit of an ideal REG can be considered as a square wave current source IS parallelly connected with the generator capacitor Cg as shown in Fig. 4a.The amplitude and period T of Is equal to those of the short-circuit current Isc of REG.Isc can be expressed as follows under the ideal conditions: And the open-circuit voltage Voc can be expressed as follows under the ideal condition: where  is twice the charge density on the electret for bipolar charged electret.
Due to the fringe effect of the electric field, the waveform of Isc will be like a trapezoidal wave in the actual case, as shown in Fig. S28a.A parameter named MR can be defined to quantitatively describe the degree of the fringe effect [S3], and is

Fig. S2
Fig. S2 Schematic diagram of the experimental setup and area for high-speed microscopic observation.(a) Schematic of the experimental setup and (b) the area for microscopic observing.The DRU is placed horizontally without any inclination.High

Fig
Fig. S3 Dust removal effect under different relative humidity and surface cover materials.(a) Removal rate and Thalf versus relative humidity.(b) Removal rate versus time when relative humidity is 8%.(c) Removal rate versus time under the lower electrode (copper) and various surface materials covering the lower electrode.In Fig. S3b, the total time of dust removal was 300 s, with each data collection lasting 60 s, and there were intervals between 3 collections.

Fig
Fig. S4 Dust removal rate versus time under different particle materials.The particle materials are (a) SiO2, (b) CaO, (c) Al2O3 and (d) Fe2O3.

Fig
Fig. S5 Illustration of the z-directional forces acting on a particle resting on the glass surface.The direction of  could be upward or downward depending on the position.

Fig
Fig. S6 Schematics of the structure and working principle of the REG.(a) Structure.(b) Working principle.

Fig
Fig. S7 Photographs of (a) rotator, (b) stator and (c) the REG driven with a linger motor in the experiment system.

Fig
Fig. S8 Experiment of long-term stability of the REG.(a) Photograph of the experimental setup.(b) Long-term stability of the short-circuit current.The REG is with a diameter of 5 mm and pairs of 12.The rotation speed of the REG is 750 rpm in the test.The short-circuit current is measured once in one day.And the result depicts that the charge density of the REG maintains stability for 183 days.

Fig
Fig. S9 Photograph of the VMC.

Fig
Fig. S10 Simulated voltage on Cd with different Cd.

Fig
Fig. S11 Photograph of the experimental setup for dust removal test.

Fig
Fig. S12 SEM photographs of the particles.(a) Silica particles with an average size of 35 μm.(b) Ulanbuhe-1 Desert sand particles whose average size is about 150 μm.

Fig
Fig. S13 Results of the removal rate versus time at different L and particle sizes.

Fig
Fig. S14 Results of the removal rate versus time at different L and particle sizes.

Fig
Fig. S15 Results of the removal rate versus time at different L and particle sizes.

Fig
Fig. S16 Summarization of the dust removal effect at different L and particle sizes for g =20, 25 mm.Thalf under (a) g = 20 mm and (c) g = 25 mm.Removal rate under (b) g = 20 mm and (d) g = 25 mm.Partial time point data in Fig. S16c is missing because the dust removal rate did not reach 50% within 60 seconds.

Fig
Fig. S17 Results of the dust removal effect at different VMC stage numbers.g =

Fig
Fig. S21 Structure and output characteristic of the wind-driven REG.(a) Photograph of the wind-driven REG and (b) its Isc, (c) Voc and (e) long-term stability.

Fig
Fig. S22 Results of the dust removal performance for the particles of 15 μm diameter.(a) Removal rate versus time.When L = 10 mm, the photographs of DRU (b) before and (c) after the dust removal.When L = 20 mm, the photographs of DRU (d) before and (e) after the dust removal.

Note S1 :
Measurement of the surface potential of the glass cover and calculation of the change of the charge on the glass cover An electrostatic voltmeter (Trek 347, US) is utilized to measure the surface potential on the glass cover, then the charge is calculated according to the electric potential.The testing procedure is illustrated in Fig. S23, in which the mesh electrode connects to positive voltage and particles are negatively charged (Test 1); for opposite charging, just replace the connecting terminals (Test 2).And the photograph of the surface potential measurement setup is shown in Fig. S24.To eliminate the influence of the initial potential on the glass surface, we conducted tests under two conditions:

Fig
Fig.S23Experimental procedure of the electric potential of the glass cover before

Fig
Fig. S24 Photograph of the experimental setup for measuring the surface potential of the glass cover.

Fig
Fig. S25 (a) Schematic diagram of DRU structure in the simulation.

Fig. S26
Fig. S26 Distribution of (a) the electric potential and (b) the electric field at the Cut Plane.(c) Distribution of the value of the z-component electric field at the surface of the glass cover.(d) Value of z-component electric field at the Cut Line.

*
The values used in the simulation for Cg and Cd are the actual ones measured by a precision LCR meter (TH 2816A, CN).Note S5: Wind-driven dust removal experiments in the real environmentThe REG was driven by the natural wind with a speed between 0 ~ 4.5 m/s.Movie S4 illustrates that two DRUs for two solar panels driven by one REG and VMC can almost remove Ulanbuhe-1 dust effectively after about 30 minutes.Movie S5 shows the comparison of the dust removal performance with and without the DRU.It was found that after about 20 minutes, the dust on the solar panel with the DRU was almost removed, while the dust still accumulated on the solar panel without the DRU.In the experiments, the power supply was used to power the homemade current testing board and the light intensity sensor.Note S6: Experiment on the dust removal effect with different connection modesIn one test, the solar panel is initially left non-dust for a while after the installation of DRU.Subsequently, sand dust is uniformly spread onto the surface of the solar panel.After the dust spread is completed and a period to allow the solar panel output to stabilize, REG starts to work, and then dust is removed from the solar panel.Before the initiation of each experiment, the solar panel is thoroughly cleaned, and the initial conditions are kept consistent across the four experiments.The solar panels used in the test under different connection types and stability under high voltages have a smaller size compared to panels in the demonstration experiment.They have an area of 27 cm × 20 cm, a peak power of 5 W, and a peak voltage of 18 V.Note S7: Experiment on the influence of high voltage on solar panel output Two solar panels of the same model are used, with the front electrode and back electrode connected to the output terminal of VMC, respectively (Fig.S29).One cycle of the test process is: after REG starts working for about 3 hours, disconnect the solar panel from VMC, and measure the current output of the solar panel.After the measurement, the charge in all the capacitors on VMC and Cd is released, and then the solar electrode is connected to VMC again.Repeat the measurements for multiple cycles per day to simulate the conditions in real application scenarios.Because the wind in the real environment is intermittent, the system may also run repeatedly.

Fig. S29
Fig. S29 Photograph of the experimental setup for influence test of the high voltage on the solar panel output.

Table S1
Parameters and results of the measurement of the surface potential on

Table S2 Default
values of the parameters on the general dust removal test * The voltage on the DRU, namely, on Cd is measured by a homemade voltage division circuit.Note that the voltage on the DRU is approximately 9 kV, lower than the output voltage of the VMC, due to discharge at the tips of the upper mesh electrode.TableS3Parameters of REG, DRU and dust particles on the dust removal test

Table S4
Parameters in simulation and calculation