In steam turbines, wetness formation usually occurs in the last stages of the low-pressure (LP) part by means of spontaneous condensation. The phase transition is a thermodynamic non-equilibrium process and is associated with additional losses compared to an equilibrium expansion. The aim of the current work is to quantify and classify the various loss mechanisms present in LP steam turbines based on computational fluid dynamics (CFD) simulations. A detailed comparison of the losses predicted with non-equilibrium steam (NES) and equilibrium steam (EQS) modeling is carried out by means of a second law analysis (SLA). The predicted entropy production rates allow for a quantification of aerodynamic losses as well as thermodynamic and kinematic relaxation losses. The aerodynamic losses are further classified into profile, wake mixing and shock losses. The method is applied to two-dimensional cascade flows and to the flow through a newly introduced generic two-stage steam turbine test case. A comparison of the loss breakdown between NES modeling and conventional EQS modeling reveals that the discrepancies are not only caused by the additional wetness losses accounted for in the NES simulations. In addition, the influence of the condensation process on the flow field leads to a different composition of the aerodynamic losses. Most affected by this is the shock loss, which is strongly dependent on the amount of latent heat released to the flow. It is demonstrated by means of the flow at 10 % of the channel height in the last stage of the turbine that an efficiency increase of 0.3 % is possible by considering non-equilibrium steam effects within the design process of the stator vane flow path.weiterlesen