Abstract: Research on deep well heat hazard control mainly uses two types of numerical models: three-dimensional CFD models (field models) and one-dimensional ventilation network models (network models). However, the field models require enormous computational resources and are difficult to apply for whole-mine scale simulations, while network models cannot accurately represent the three-dimensional distribution characteristics of heat sources and the local heat dissipation effects of equipment. To address these issues, taking Jiaojia Gold Mine of Shandong Gold Mining (Laizhou) Co., Ltd. as the research object, a field-network hybrid numerical simulation method for deep well heat hazard control was proposed. The mine ventilation system was composed of nodes from the network model. The extraction main ventilator drew low-temperature airflow introduced from the surface, which heated up after passing through the high-temperature mining face, then flew through the return airway and returned to the surface. The high-temperature mining face area was modeled by a three-dimensional field model. The field model obtained input boundary conditions from corresponding nodes at the inlet boundary through an area-weighted averaging method. The outlet boundary parameters of the field model were calculated via fluid control calculations and converted into node information, which was introduced into the network model as source terms, thereby realizing the field-network hybrid numerical simulation. Results showed that: ① As the length of the supply airway increased, the high-velocity area gradually concentrated toward the center axis of the airway, the peak temperature gradually decreased, and the high-temperature zone tended to shift downward overall. ② When the heat source power was too high (severe heat hazard), the convective cooling effect provided by the main ventilator was poor. At this time, increasing the ventilator airflow could further reduce the mining face temperature.