Mathematical model to predict rooftop PV system yield

Scientists in India developed a mathematical model to predict the output of solar cells and modules in the field. The model was developed and tested using both sun simulator and actual installed modules. The scientists state that their model can be applied to a PV installation anywhere in the world, and that by taking into account module degradation over time their forecasts can be as much as 26% more accurate than existing energy yield models.

Shoolini University in Himachal Pradesh, India. Scientists at the University developed a more accurate model for PV energy yield forecasting.

Wikimedia

Share

From pv magazine Global

Being able to accurately forecast the energy yield of a PV system is important in convincing investors of the returns solar can achieve, as well as providing a benchmark for owners and operators to better understand how a project is performing.

And as PV technology continues to diversify, with more differentiated products making their way to market, understanding how each will perform under certain conditions is an increasingly important consideration. And when a project does not meet expectations, it can cause big problems for all involved, and even harm solar’s wider reputation.

This was on the minds of a group of researchers led by Shoolini University and India’s National Institute of Technology in Hamirpur, who set out to develop a new model to predict the output of solar cells and modules over time. A key consideration for their model was accounting for the often unavoidable performance loss that PV systems experience over time. By working this into the model, the group estimates it is able to increase the accuracy of yield forecasts by as much as 26%.

Accounting for module degradation

The model is described in full in the paper A power prediction model and its validation for a roof top photovoltaic power plant considering module degradation, published in Solar Energy. The researchers measured the output and degradation of various solar cells in the lab under a sun simulator, and then validated these results with data from a PV installation on the institute’s rooftop. The paper also identifies different parameters to include in the model under different climate conditions, finding that a simpler model with only three parameters offered the most accurate forecast under low irradiance conditions, while four and five parameter models were more precise as the level of irradiation increased.

“Our model could be useful for developing PV power forecasting software capable of predicting daily, monthly, and yearly solar power generation, which is a crucial parameter for calculating the electricity supply and demand distribution, as well as for marketing PV systems,” said Shoolini University Professor Shyam Singh Chandel. “Being able to predict the output power of different PV technologies will help interested parties decide for the best option available in the market for a particular location. Based on the foreseen energy generation and the selected technology, one can estimate the total expenditure and payback period for a given project.”

This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.