| Work Detail |
Researchers in the Middle East have carried out a series of simulations to evaluate the technical and economic feasibility of building-integrated photovoltaic-thermal systems. The proposed framework could be applied to various building types and geographic locations.
Researchers at Al al-Bayt University in Jordan have created a comprehensive theoretical dynamic model to analyze the performance of building-integrated photovoltaic thermal (BIPV/T) systems.
The scientists tested their new framework through a series of simulations conducted in a residential building in the city of Mafraq, in north-central Jordan. They optimized the BIPV/T system to maximize production and efficiency while minimizing the surface area of ??the panels.
“Mafraq is one of the most suitable regions in Jordan to take advantage of the potential of solar energy and implement solar thermal technology plants,” the researchers explain. “It has a daily average of normal direct irradiation (DNI) that ranges between 6.54 and 7.29 kWh/m2 and an average annual wind speed of 4.72 m/s, it enjoys abundant solar irradiation and abundant wind resources” .
The building was simulated with MATLAB/Simulink. It had a roof area of ??200 m2, with a demand for hot water supplied by the thermal part of the panels of 10 m3 per month. The average electricity demand in winter, spring, summer and autumn was 452 kWh, 582 kWh, 443 kWh, 342 kWh and 441 kWh, respectively.
The BIPV/T panels were assumed to have a power of 320 W each and an efficiency of 16.49%. The hourly solar radiation and ambient temperature data used in the model come from an artificial neural network (ANN) developed by the group. According to the researchers, the model achieves high accuracy values ??exceeding 0.97 for the predicted solar irradiance during training, validation, testing and the global dataset.
“The comprehensive ANN-based modeling and optimization approach can help policymakers and energy professionals analyze in a reasonable time domain and with good accuracy, thus accelerating analyzes and tasks based on design and improvement of the performance,” says the team.
For the optimization task, the group chose the non-dominant sorting genetic algorithm II (NSGA-II), also integrating the order of preference by similarity with the ideal solution (TOPSIS) technique. NSGA-II optimizes multiple objectives by ranking solutions based on dominance, while TOPSIS ranks options by comparing their closeness to the best and worst possible solution.
The analysis showed that the optimal solution for winter includes the installation of 15 modules, with an electrical power of 2,606 W, a thermal power of 5,569 W and a total electricity produced of 21.95 kWh. In summer, the optimal solution is 16 modules with an electrical power of 1,780 W, a thermal power of 4,700 W and a total production of 23.63 kWh. For the spring seasons, 12 BIPV/T modules were needed, and 14 in the fall.
In addition, the scientists carried out an economic analysis. In this analysis, they assumed that BIPV/T would last 20 years, the interest rate would be 5%, and the variable cost of operating the system would be $0.1/kWh. In addition, the reference costs were set at $1,067 for the hot water tank, $180/kW for the inverter and controller, and $200/m2 for the BIPV/T panels.
Under these conditions, the levelized cost of energy (LCOE) is $0.1/kW. If the cost of equipment increases by 20%, the LCOE will be $0.12/kW, while if it increases by 40%, the price will reach $0.14/kW. If equipment prices drop by 20%, the LCOE will be $0.08/kW, and if they drop by 40%, the cost will be $0.06/kW.
“The results indicate that the suggested framework can be applied to various building types and geographic locations. Therefore, it has considerable value in advancing the utilization of solar energy with optimal energy, economic and environmental performance,” the academics conclude.
The results were presented in “ Energy and economic analysis of building integrated photovoltaic thermal system: Seasonal dynamic modeling assisted with machine learning-aided method and multi-objective genetic optimization ” seasonal assisted method assisted by machine learning and multi-objective genetic optimization), published in Alexandria Engineering Journal .
The team also included scientists from King Fahd University of Petroleum and Minerals in Saudi Arabia, University College London in the United Kingdom, Durham University and Pakistan University of Engineering and Technology. |
