Search

• Field test of cleaning robot's moving shadow on operating PV string. • Much larger power drop for landscape modules than portrait. • Larger power drop for full-cell modules than half-cut. • No cell hotspots created. • Bypass diode heated when module stopped, but not while moving. PV cleaning robots cast a shadow on the PV modules which they are cleaning, if used in daylight. The shadow affects the electrical characteristics (current and voltage) of the module string, and could potentially also create hot spots in modules' cells or bypass diodes. The severity of these effects depends on the layout of cells and bypass diodes and the module's orientation, among other factors. In this study we tested a commercial cleaning robot on strings of two kinds of PV modules with different cell formats, connected to a grid-tied inverter. It was found that the string power was greatly reduced when modules were in landscape orientation, but barely affected in portrait orientation. The robot shadow caused slightly greater power reduction for full-cell than half-cut modules. While the robot was in motion, no heating of cells or bypass diodes was observed, although diode heating did occur if the robot stopped on a module. For the worst case of full-cell landscape modules, a simplified estimate of the energy loss due to one robot pass was roughly equivalent to 0.16% of the string's daily energy production. The results suggest that if PV cleaning robots are used in daylight, the power loss from their shadow could be almost eliminated by installing modules in portrait orientation. [ABSTRACT FROM AUTHOR]

• Vibration of 6 kinds of PV module under robot cleaning were measured. • A module's size was the dominant factor in how much it vibrated. • A module's frame and glass thickness had little influence on its vibration. • Overall the amount of vibration by robot cleaning was small – less than 1 mm. • Strong wind caused more severe module vibrations than the cleaning robot. • Large-format modules and robots that run on the modules may present greater vibration risk. Automatic PV cleaning machines ("robots") are becoming widely used in desert regions. The motion of the robot and its brushing action can vibrate the PV modules being cleaned. This study measured the vibration of six different kinds of PV module, excited by a single kind of commercial cleaning robot. The robot's brushes were straight, rather than helical, which imparted a periodic excitation to the modules at approximately 7 Hz. For comparison, module vibration by strong wind was also measured. It was found that (i) among similar-sized PV modules, the amount of vibration was largely independent of the module construction (glass thickness and frame), (ii) modules tended to vibrate at their natural frequency rather than the brush (excitation) frequency, and (iii) the module deflection range by robot cleaning was approximately 0 mm upward to 1 mm downward, while in strong wind it was approximately 2 mm upward to 1 mm downward. In this study, which used modules up to 2 × 1 m, the particular cleaning robot did not cause significant vibration, however as commercial modules become larger they may also become more flexible and prone to larger deflections. [ABSTRACT FROM AUTHOR]

As PV cleaning machines become widely adopted in desert regions, there is concern that frequent dry-brushing might abrade modules' anti-reflective coating (ARC). It is common to study ARC abrasion via accelerated lab tests, which typically differ from real-world PV cleaning regarding sample format, test conditions and characterization tools. This paper reviews research on PV ARC cleaning abrasion, focusing on differences in experimental techniques between the lab and field. The main conclusions are: (1) coupon tests should be verified with full-size modules due to dependence of coating properties on sample size, (2) lab tests should be verified outdoors with non-accelerated cleaning schedules because abrasion is largely caused by the soiling layer itself when brushing, (3) reflectivity is well suited to measuring ARC abrasion because it can be performed in the field and is reasonably consistent with module optical performance, (4) different ARCs exhibit a wide range of abrasion susceptibility whereas bare glass is essentially immune to dry brushing. [ABSTRACT FROM AUTHOR]