Nonrenewable fossil fuels are finite resources that will ultimately deplete in near future. Nature sheds colossal amount of renewable wind energy but humans harvest a morsel. Taking this into account a numerical study is proposed on wind energy harvesting from a speeding subway train. Subways trains generate a remarkable gust of wind that can be transferred to useful electrical energy on daily basis. To this aim, a numerical analysis is modeled by placing Savonius wind turbine in a subway tunnel to crop the wind energy produced from the speeding train. The passage of train in the tunnel generates very high velocity slipstreams along the length of the tunnel. The slipstream phenomena develop a boundary layer regime that will be absorbed by the Savonius wind turbine to self-start and generate power. In the present study, a two-dimensional numerical simulation with modified turbine blade design is carried out using open source tool OpenFOAM® with PimpleDyMFoam solver coupled with six degrees of freedom mesh motion solver sixDoFRigidBodyMotion and k–ɛ turbulence modeling, to measure the amount of torque predicted by the rotor from the gust of wind produced by the speeding train in the tunnel. Being a self-start turbine with no yaw mechanism required the turbine collects air from any direction and converts it into useful power. © IMechE 2018.