A Study on a Digital Twin Method for Modelling the Diffusion Process of Fault Gases Inside High-Voltage Switchgear Combining Virtual and Physical Models.

Monitoring gas generated by internal faults in high-voltage switchgear is crucial for early warning and condition assessment; however, the diffusion process between gas generation and detection is currently widely overlooked, which severely compromises the accuracy of gas-based assessments. Implementing "virtual sensing" for high-voltage switchgear using digital twin technology to supplement data from physical sensors can effectively support digitalised and intelligent operation and maintenance. To this end, taking the characteristic gases of partial discharge as the subject of study, we first constructed a 1:1 scale digital twin model of internal gas diffusion within a 10 kV high-voltage switchgear. Calculations revealed significant differences in the diffusion processes of various gases: CO diffuses rapidly and mixes strongly, making it suitable for early warning; CO2 and O3 tend to accumulate in lower regions and stagnant zones, making them sensitive to faults in the lower sections; whilst NO reflects the channelling effect within the structure. Subsequently, a test platform comprising a 10 kV high-voltage switchgear unit and a gas detector was established for experimental validation. This revealed the trends in gas concentration changes, response sequences and peak behaviour at different fault locations, thereby verifying the validity of the proposed model. The maximum error in concentration balance was within 10 percent, providing a theoretical basis and engineering reference for the optimised placement of gas sensors in high-voltage switchgear and for fault tracing and localisation.