Sizing, simulation and comparison of a Building Integrated and a Building Attached Photovoltaic system

This thesis project presents the design, simulation and comparison of two different configurations of a photovoltaic system installed on an office building in Santiago, Chile. One is a BAPV type consisting of mono-crystalline modules used as Sun shields on the north facade windows and also mounted on the roof facing the same direction, an installed power of 488.4 [kWp] is calculated. The second is a BIPV system with mono-crystalline semi-transparent modules on the windows and opaque panels on the walls and the same roof installation as the other case, results in 885.87 [kWp] installed. The same model of three-phase inverter is used, but 9 are required in the first case and 15 in the latter. The estimation of the demand profile is performed using the software EnergyPlus and OpenStudio, based on the Reference building model for a large office building developed by the DOE of the United States. After modifications to adjust the internal loads, an annual consumption of 4,200.8 [MWh] is obtained. This value does not present considerable variations when the solar modules are added. The simulation of the generated power is performed on an hourly basis for a 25-year period using MATLAB R2017a, based on the radiation data provided in “Explorador Solar” by the Ministry of Energy of Chile. An annual production of 1,036.1 [MWh] is obtained for Sun shield and 1,788.0 [MWh] for Curtain wall. For the Sun shield there is injection to the grid on the weekends, since during the weeks the consumption is higher, for Curtain wall there is also generation surplus during weekdays in summer. In both cases the energy self-consumed reduces the requirements to the system resulting in a peak-shaving effect due to the coincidence of the demand profile and the sunlight hours. The Sun shield configuration has a higher annual yield and reference yield, of 2,121.4 [kWh/kWp] and 2,583.2 [h], while in Curtain wall are 2,018.4 [kWh/kWp] and 2,312.0 [h]. The performance ratio is lower, 82.1 % compared to 87.3 % of the latter. An economic analysis is performed, and it is concluded that the Sun shield configuration is profitable with an NPV of 575.000 [USD] and a payback time of 5 years, while Curtain wall is not economically viable with NPV of -1,711,700 [USD]. The initial investment is 545,362.09 [USD] and 4,297,745.67 [USD] respectively but in the second case it is reduced about 30 %, to 2,990,252.70 [USD], due to savings related to the conventional components being replaced. It is noted that the periodic costs for cleaning the modules take special importance for Curtain wall because they are charged by square meter and the installed area is much bigger in this case, 5,274.67 sqm compared to 2,664.71 sqm.