Explore More
Solar Calculator
HomeBlogBlogPredicting photovoltaic (PV) losses in Indian solar rooftop using the Internet of Things (IoT) and data-driven estimation algorithms

Predicting photovoltaic (PV) losses in Indian solar rooftop using the Internet of Things (IoT) and data-driven estimation algorithms

Predicting photovoltaic (PV) losses in Indian solar rooftop using the Internet of Things (IoT) and data-driven estimation algorithms

Solar on the rise…with a catch!

Solar power plants are on the rise as the better and more reliable way to boost the Indian renewable energy industry. Although it isn’t the latest entry, solar power is arguably the one with major untapped potential and rapid scalability. Solar energy generation is almost as renowned for its lower regular maintenance as having their performance contingent on the surroundings.
Being an open energy system, solar generation is intermittent and highly dependent on having suitable weather. Incidentally, there are also intrinsic limitations governed by their material physics. Since any decrease in generation can result in considerable financial drawbacks, it is important to predict the possible generation. Energy estimation is only possible when we are able to effectively model the system taking into account its various parameters as well as losses. The most important parameter influencing generation is the incident irradiation. This dictates the upper bound on the energy that can be harnessed. Apart from that, photovoltaic losses play a crucial role in benchmarking the maximum generation possible. These losses depend on a plethora of factors including cloud cover, module operating temperature, ambient temperature, particulate deposition on modules and sensors, grid unavailability, etc.

Understanding photovoltaic losses to cut financial losses!

A major part of the total loss in solar rooftop generation is due to its module temperature. Incident irradiation rises the module temperature thereby decreasing its operating efficiency. This initiates a negative feedback wherein reduced efficiency further increases its temperature resulting in thermal loss. Soiling of modules due to particulate deposition is another important phenomenon contributing to a significant loss. Loss due to soiling, in contrast to thermal loss, is almost exclusively defined by the plant surroundings and its cleaning schedule. Loss due to module degradation and ageing is a more elusive one with its roots in its electrical characteristics. It is intrinsic in solar module nature to give decreasing performance as time passes even in the most favourable operating conditions. This unavoidable loss poses a slow burn of a risk in terms of revenue loss and reduced plant life.
In view of the significant trouble these losses pose to plant performance, it is important to have a proper understanding of them from design and operational standpoints. Therefore, modelling these losses is no longer a theoretical interest in energy science. It should rather be a norm in all solar industrial operations giving insights on how to have better operational efficiency.

IoT and data analytics to the rescue!

It is noteworthy at this juncture to point out the importance of having an extensive solar data acquisition system. Modern technologies like IoT are indispensable in leveraging the immense power of data. Such data is invaluable to answer and corroborate almost all kinds of decision-making exercises. Yet, this kind of data-driven approach still remains relatively unknown to many solar rooftop operations. This is the major motivation behind Amplus’ pursuit of leveraging the data to devise tools aimed at recognizing the loss patterns. As part of Amplus’ remote monitoring solution, all our plants have sensors sending plant data to our servers on a real-time basis. This is incorporated into the analytics portal and studied to understand their performance on a granular level. Various empirical algorithms are developed here to evaluate the aforementioned PV losses. Their pan portfolio level implementation gives valuable feedback to optimize plant operations. This article is a testimonial to prove the potential of data-driven performance monitoring in solar installations. It also drives home the necessary motivation towards adopting such novel techniques to better compete in the solar market.

Vijay Bhaskar

Leave a Comment

Your email address will not be published. Required fields are marked *