POD-based modeling on thermal-flow characteristics for air-cooled condenser of power plant

In order to rapidly and accurately predict the velocity and temperature distributions in air cooled condenser (ACC) of direct air cooling power plant influenced by the ambient wind, reduced order models (ROMs) for two-dimensional flow and temperature fields were studied with proper orthogonal decomposition (POD) method. The velocity vector, u, turbulent kinetic energy, k, and turbulent dissipation rate, ε, were approximated through POD procedure. The improved complementary POD approach (PODc) was adopted to construct robust velocity POD models of ACC under various ambient wind velocities, then flux matching procedure (FMP) was introduced to estimate the POD mode coefficients. And that of k, ε were obtained through cubic spline interpolation. It was found that as the case differed from the system reference point was great, the advantage of PODc became obvious. Otherwise, the mean relative errors of POD solutions were smaller than that of PODc method for the cases near system reference point. By interpolating the above three variables obtained by POD/PODc method into CFD software for energy equation, the temperature distributions could be acquired with the mean relative error of 10−3-10−4. At the meanwhile, the computing time spending on the energy equation iterating could be decreased significantly compared with that of direct CFD simulations.