Techno-economic analysis and energy forecasting study of domestic and commercial photovoltaic system installations in Estonia. (15th August 2022)

• Record Type: Journal Article
• Authors: Shabbir, Noman; Kütt, Lauri; Raja, Hadi A.; Jawad, Muhammad; Allik, Alo; Husev, Oleksandr
• Is Part Of: Energy. Volume 253(2022)
• Publication Date: 2022-08-15
• Subjects: PV Systems -- Economic analysis -- Energy forecasting -- Machine learning -- Payback time; PV RES -- RE nZEBs -- TSO NCEP -- ROI RNN -- LSTM SM -- NM ARIMA -- ANN KNN -- ANFIS CNN -- BPNN CAP, MAPE; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2022.124156
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: The Baltic countries have good potential for solar photovoltaic (PV) energy generation, as on average 15 hours of sunlight is available in summer. Another potential option is to encourage the construction of nearly zero-energy buildings (NZEBs) according to the EU framework. This study focuses on solar irradiance and energy generation potential in different regions of Estonia as a case study. Techno-economic analysis of possible solutions to use differently rated domestic and commercial PV systems' feasibility and payback periods are presented. The results illustrate that all PV systems studied in the research are self-sufficient while selling excess energy to the grid with a nominal payback period. Furthermore, for short-term energy management, we developed an efficient deep learning-based forecasting algorithm. Apart from the inherent non-linear nature of solar energy data, what makes forecasting particularly challenging is to efficiently cope with the issue of data regression and random noise. The RNN-LSTM algorithm is chosen for the prediction of solar energy. This is the first comprehensive report that can encourage potential Estonian users to invest in solar PV systems and gain economic benefits. The results presented in this study cover a broader perspective and are more useful keeping in mind the real market situation of the Baltic countries. Highlights: Compared initial investment and payback period of multiple rated PV systems. Design and development of deep learning based short-term forecasting algorithm for solar PV generation. ROI and payback period has been calculated for different rated PV systems in different regions. Economic impact on Estonian's national grid due to on-grid distributed PV systems.