Optimisation of PV Cleaning Practices: Comparison Between Performance Based and Periodic Based Approaches
Date
2018-07
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
The British University in Dubai (BUiD)
Abstract
Solar energy has the largest untapped reserve in energy and is one of the fastest emerging energy markets. Especially in places that experience sunny clear days throughout the year such as the Middle East region, solar energy has seen large investments in technologies such as photovoltaic panels and concentrate solar power. Photovoltaic panels convert the sun radiation directly into DC electric current through the photovoltaic effect. Although the biggest costs involved in this technology is the installation/capital costs, the subsequent running costs (although less than traditional sources of energy) such as cleaning and maintenance can increase or decrease the feasibility of these panels. Soiling, which is the deposition of materials (usually dust) that obstruct solar radiation falling on the panel and is one of the major factors of production losses in PV panels, and thus managing it will result in a positive economic and technical impact on PV plant feasibility.
Soiling on PV panels was found to have varying degrees of effects depending on various factors such as environmental, geographical, and installation factors. Higher relative humidity, drier desert climates, larger dust particles/concentration, and lower PV panel tilt angle are all examples of how soiling or its effects are increased. The losses resulting from this can lead to significant losses, especially in a place like Dubai, UAE which meet many of these factors. Soiling can also lead to long term damage to the PV panel such as degradation, delamination, and hot spots that can further influence PV feasibility. It is important to choose the correct cleaning method, frequency, and approach to ensure that the best practices are being used on the basis of both economic and technical criteria.
To optimize the cleaning practices, an algorithm was developed by the author based on the literature review and economic analysis, and was run for multiple panels and tilts at the DEWA OTF solar lab in Dubai, UAE. The PV output, soiling rates, electricity price, and cleaning costs were inputted into the algorithm and 5 different approaches were simulated:
1. Performance based manual cleaning approach where panels are cleaned only when soiling losses surpass cleaning costs.
2. Adaptive periodic based manual cleaning approach where a changing cleaning period is fixed for each season based on significant variations of seasonal averages from the performance based manual cleaning approach, taken as 1 standard deviation from the season average.
3. Optimized periodic based manual cleaning approach where panels are cleaned on a fixed period taken as the rounded average from the performance based manual cleaning approach.
4. Periodic based manual cleaning approach where panels are cleaned every 5 days as per the existing practices in the Mohammed bin Rashid 13 MW solar park.
5. Periodic based automatic cleaning approach where panels are cleaned daily with the currently installed prototype cleaning robot in the Mohammed bin Rashid 13 MW solar park.
The cases simulated were 12 different PV setups which included 6 different PV technologies on a 5 and 25 degree tilt:
1. Polycrystalline
2. Monocrystalline
3. BIPV Double Glass Polycrystalline
4. Bifacial Monocrystalline
5. CIGS Thin Film
6. CdTe Thin Film
After running the algorithm and comparing results with the existing periodic based manual cleaning approach of 5 days currently in place in the Mohammed bin Rashid 13 MW solar park, it was found that the best approaches (in order of best to worst) were:
1. Performance based manual cleaning with between 9.91% to 33.87% improvement.
2. Adaptive periodic based manual cleaning with between 8.75% to 31.75% improvement.
3. Optimized periodic based manual cleaning with between 5.78% to 28.39% improvement.
4. Periodic based automatic cleaning showed decreases in improvement of between 164.64% to 245.34%.
By adopting each of the first 3 above approaches in the current MBR 13 MW solar park, we can realize yearly savings of 110,073.6 AED, 102,429.6 AED, and 91,728 AED respectively, whereas automatic cleaning was found to be not economically feasible as currently installed. Furthermore, it was concluded that higher tilt angles resulted in lower cleaning requirements while thin film panels required less cleaning than 1st generation PV panels. It was also important to note that higher soiling did not correlate to higher cleaning requirements, as the cleaning was dependent not only on the soiling but on the PV output and cleaning costs. The algorithm was found to be an effective yet simple tool to help PV owners decide on cleaning procedures, however it does not take into account the effect of natural cleaning events. Future work can include different tilts including trackers, comparison of more cleaning methods including different robots, and different geographical locations.
Description
Keywords
PV cleaning practices, optimisation, solar energy, energy markets, solar power, Middle East region, PV technologies, United Arab Emirates (UAE)