As Building Integrated Photovoltaic (BIPV) systems become increasingly integral to urban energy independence, addressing their inherent safety risks has become a critical priority. Due to the direct integration of photovoltaic modules with building st...
As Building Integrated Photovoltaic (BIPV) systems become increasingly integral to urban energy independence, addressing their inherent safety risks has become a critical priority. Due to the direct integration of photovoltaic modules with building structures, electrical faults in BIPV systems can lead to immediate fire hazards and pose severe electric shock risks to first responders, thereby hindering effective emergency response. Existing safety technologies, however, have predominantly focused on mitigating individual risk factors, resulting in a lack of organic coordination required for complex disaster scenarios. To overcome these limitations, this dissertation proposes the design, implementation, and empirical verification of a high-reliability integrated safety system. This system consolidates three core safety functions into a single unified platform: Arc Fault Circuit Interruption (AFCI) to eliminate the root causes of fire, Rapid Shutdown (RSD) to ensure personnel safety during emergencies, and Pre-emptive Fault Finding to identify potential hazards such as line breakages. The proposed system architecture is comprised of the 'Sentinel System,' a hardware control unit deployed on-site, and the 'Cloud Safety Platform,' which facilitates remote monitoring and control, thereby establishing a multi-layered safety network. For fire prevention, the DC Arc Fault Detection Device (AFDD) utilizes a 30–100 kHz Band-Pass Filter (BPF) and Fast Fourier Transform (FFT) algorithms to precisely analyze arc noise signatures. For electric shock prevention, the RSD mechanism rapidly reduces the voltage of PV strings to a safe level in response to remote control signals. Additionally, a novel fault location algorithm based on the parasitic capacitance characteristics of PV strings was developed to pinpoint cable faults within an error margin of ±1 module. To validate the reliability and performance of the proposed system, an arc fault generator compliant with the UL 1699B standard was fabricated, and a 3kW-class empirical testbed was constructed. A series of 100 repetitive tests, conducted under the supervision of an accredited testing agency demonstrated the system's superior performance. The DC AFDD achieved a 100% detection accuracy with an average detection time of 1.058 seconds. The RSD system exhibited a 99.38% communication reliability, ensuring rapid voltage de-energization. In conclusion, this study presents a robust architecture for an integrated safety system capable of effectively mitigating the multi-physical hazards associated with BIPV systems. Through rigorous empirical verification, the system has proven its performance to exceed international safety standards (NEC, UL). These findings offer significant practical contributions to elevating BIPV safety standards and accelerating the deployment of safe renewable energy infrastructure in urban environments.