Theoretical, numerical, and experimental investigation of smoke dynamics in high-rise buildings

Abstract Smoke kills more people than the associated fire and thus predicting smoke spreading inside high-rise buildings is of paramount importance to structural and safety engineers. Here, the velocity, temperature, and concentration fields in large-scale turbulent smoke plumes were predicted using classical self-similar turbulent plume theory, which assumes a point fire source under open-air conditions. Turbulent fires of various heat release rates in a confined space were also simulated numerically using Fire Dynamics Simulator (FDS), which was verified against experimental data before being used to validate the analytical plume jet results. The agreement between analytical, numerical, and experimental results was good. This demonstrates for the first time that for realistic, wide shafts, analytical results from self-similar theory of free turbulent plumes are as accurate as the numerical simulations and appropriately describe experimental data. This allows engineers to avoid lengthy, cumbersome numerical simulations to estimate the consequences of smoke spreading in high-rise buildings using simple analytical formulae. In addition, parametric studies were conducted using plume theory for building heights up to 500 m and heat release rates up to 500 MW. Smoke velocity, temperature, and concentration fields described smoke evolution at different heights.