Quantitative evaluation of wetness losses in steam turbines based on three-dimensional simulations of non-equilibrium condensing flows

摘要

Wet steam flow in steam turbines leads to degraded efficiency and blade erosion in the turbine stages. The Baumann rule, which has been used to predict wetness losses, is increasingly being questioned. More recently, the non-equilibrium condensation model is being increasingly applied to analyse wet steam flow. However, most of the influences caused by wetness losses on the aerodynamics of a wet steam turbine are excluded when this approach is used. Therefore, the efficiency of a wet steam turbine calculated by the non-equilibrium approach does not match the experimentally obtained values. To improve the accuracy of evaluating a wet steam turbine as well as the wetness losses, a quantitative evaluation program of wetness losses has been developed based on the calculation results of wet steam flow with non-equilibrium condensation using the FORTRAN language. Three-dimensional (3D) simulation of the wet steam flow with non-equilibrium condensation in turbine stages is first conducted. Then the 3D results are circumferentially averaged in the meridian plane, which are subsequently used to quantitatively evaluate the wetness losses. The wetness losses are divided into five categories: thermodynamic loss, droplet drag loss, braking loss, capturing loss and centrifuge loss. The wetness losses in the low pressure (LP) cylinder of a fossil steam turbine are calculated. The results show that the thermodynamic loss is mainly generated in the nucleation stage and the last stage of the turbine where non-equilibrium condensation occurs. The droplet drag loss is small in all wet steam stages. The braking loss is the most important component of the wetness losses, except in the nucleation stage. The capturing and centrifuge losses are moderate in the wet steam stages. The total wetness losses in the LP cylinder account for 3.65% of the total output power. This is less than the 5.14% losses predicted by the Baumann rule.