Development and Validation of a Numerical Multiphysics Algorithm for Wall Condensation Heat Transfer for Steam Turbines Starting Phases

Abstract Due to the rising interest in using renewable energy for environmental impact reduction, steam turbines are facing with a radical change in design practices and operability. In particular, the unpredictable and variable production of renewable plants drives to frequent shutdowns and fast start-ups of steam turbines, which have a huge impact on structural integrity. During start-up phases, the high temperature steam could face with a below-saturation wall temperature causing condensation and then high temperature gradients on solid components. Hence, the proper thermal gradient evolution and condensing mass prediction are quite relevant in steam turbine design. Nowadays, numerical Conjugate Heat Transfer (CHT) analysis is a good industrial practice to quantify the thermal distribution in turbomachinery components. On the other hand, the usual bottleneck of these analyses is the fluid timestep and the required temporal discretization that is one or two order of magnitude smaller than the solid one: therefore a fully coupled unsteady is still unaffordable in terms of computational costs. In this context, this work presents a low dimensional multiphysics approach which models the condensation phenomenon through correlations and the solid heat transfer through discrete 1D transient heat equation. Furthermore, a 3D CHT by employing the eulerian wall film multiphase approach is presented and compared with the developed algorithm. Finally, the numerical results obtained are compared with field data on a real unit where the inlet section, as usual, is subjected to the described condensation phenomenon.