The effect of dust on the performance of solar photovoltaic module: Case studies in Nusa Tenggara Timur, Indonesia and Perth, Western Australia

Tanesab, Julius (2018) The effect of dust on the performance of solar photovoltaic module: Case studies in Nusa Tenggara Timur, Indonesia and Perth, Western Australia. PhD thesis, Murdoch University.

Abstract

The performance of a PV module tends to decrease as dust impinges onto its cover surface. The attached dust diminishes the illumination by absorbing and scattering sunlight received by the solar module. Degradation caused by dust is temporary, but it should not be underestimated. Many studies investigated the influence of dust accumulation on optical properties and its impact on PV module performance. However, less attention was given to the effect of dust on small scale PV systems such as residential systems. Also, most of the preceding literature were not supported with an economic analysis which can inform maintenance activity scheduling. This study aims to identify the effect of dust on the performance of solar PV modules in varying environmental conditions and cost effective maintenance schedule for both solar home system and residential grid-connected system.

To study the effect of dust on PV performance in different climate areas, research was conducted in Nusa Tenggara Timur (NTT), Indonesia and Perth, Western Australia. A series of experiments in the laboratory was performed. A solar simulator was used to measure PV modules‟ performance. A combination of a spectrophotometer, scanning electron microscope, electron dispersive spectroscope and an X-ray diffraction machine were used to examine properties of dust. In addition to the laboratory experiments, a field study was carried out to investigate the effect of dust accumulated naturally on PV performance degradation deployed in the two regions.

Characterization results revealed that dust in Perth exhibited angular shapes dominated by quartz, while porous particles with a large amount of calcium oxide were observed in dust from NTT. The grain size analysis showed that the percentage of clay and very fine silt of dust from Perth was higher than that from NTT. Therefore, at the same density, dust from Perth passed less light than that from NTT. Power output produced by PVs coated artificially with dust from Perth was lower than that from NTT although the difference was not statistically significant.

The performance degradation of PV modules deployed in the field varied with season. In Perth, power output of the modules which was maximal at the beginning of summer decreased significantly at the end of the season. The performance was then increased, approaching the initial position at the end of autumn and reached a peak at the end of winter. A similar decrease in the summer‟s performance was observed in the modules at the end of spring. In NTT, the performance which was maximal at the beginning of wet season dropped slightly at the end of the season and had significantly decreased at the end of the dry season.

PV performance variations were in agreement with dust density deposited on the examined PV modules. Seasons with less rainfall demonstrated more accumulation of dust compared to those with greater rainfall. In addition, as the tilt angle increased dust deposition decreased; as a result, the average transmittance of dust increased. For a one year period, power loss of PV modules due to dust was 4 - 6% and 16 - 18% for Perth and NTT, respectively. The greater degradation in NTT is attributed to the lower tilt angle of the PV modules, the higher relative humidity, and the longer dry season in the region.

The effect of dust on PV performance for a long time period carried out in Perth revealed that the degradation of Pmax output of PV samples deployed for almost 18 years without any cleaning procedures were 8 - 12%. These losses are higher than that measured for the one year period and indicate that natural cleaning agents such as rain and wind could not remove dust particles attached on the surface of the PV modules perfectly. In addition to the power decrease, observation results in the field showed that the modules exhibited some permanent degradation indicated by corrosion, delamination, and discoloration. This may be attributed to hot spot phenomenon caused by dust for a long time period besides the age of the examined PV modules.

Economic analysis revealed that annual cost of production losses of residential PV systems in Perth and NTT with a degradation pattern as measured in the field was higher than the maintenance cost activities. Consequently, the system in Perth needs once cleaning in a year, meanwhile twice for the system in NTT. This thesis, therefore, suggested more intense cleaning should be applied for PV modules mounted at lower latitude and deployed in a tropical climate area.

Standard dust de-rating factor (5%) stipulated by Australian/New Zealand Standard 4509.2:2010 was appropriate for modelling a grid-connected PV system in Perth, but, the system required cleaning once per year. Conversely, the standard soiling loss factor of 5% was not suitable for solar home system modelling in NTT as the estimation of the impact of dust was underestimated. Thus, this thesis recommended that the soiling de-rating factor should vary between regions and with season. This will improve the accuracy and the reliability of PV system models.

Item Type:	Thesis (PhD)
Murdoch Affiliation(s):	School of Engineering and Information Technology
Supervisor(s):	Urmee, Tania, Whale, Jonathan and Parlevliet, David
URI:	http://researchrepository.murdoch.edu.au/id/eprint/40274

Item Control Page

Downloads