Abstract: The technical value of smart electricity meters data in tracing the causes of electrical fires has become increasingly prominent, yet the data characteristics and their linkage mechanisms with electrical fires still require further exploration. This study systematically analyzes the correlation between electrical characteristics and fire risks under scenarios such as load type identification, single-phase voltage fluctuations, and current anomalies through the construction of a multi-scenario fault simulation experimental platform. The platform integrates smart meter dynamic data monitoring, oscilloscope waveform capture, and infrared thermal imaging technology. Experimental results reveal substantial discrepancies in power factor and reactive power across load types, offering critical evidence for fire site load classification. In the condition of voltage deviations, overvol⁃tage elevates resistive load currents, while undervoltage triggers compensatory current surges in constant-power loads, intensifying conductor overheating. The instantaneous peak current of zero fire short circuit can reach 260 A, releasing more than 100 J of energy. The leakage point temperature of the live wire (65 mA) exceeds 300 ℃ within 1 min, and the temperature rise of the overloaded (1.5 times) wire reaches 104 ℃. The temperature of the inferior wire (0.5 mm2) after normal current flow reaches 173.6 ℃, all of which pose a direct risk of igniting combustibles. The research has verified the technical effectiveness of smart meter data in tracing the causes of fires, and proposed active prevention and control strategies such as mandatory installation of circuit breakers and standardized wire selection standards, providing theoretical support and practical reference for precise prevention, cause tracing, and safety management of electrical fires.

Key words: smart electricity meter, abnormal electricity consumption data, electrical fire, fault simulation