Doosan Škoda Power

An analysis of the industrial steam turbine market shows that many customers require steam turbines with relatively small output and high inlet steam pressures and temperatures, causing small nominal mass flows of steam through the turbine flow parts. As a result, the blades of the first stages tend to be relatively short, which is connected with the high energy losses. This problem can be partially alleviated when the drum rotor arrangement is used. In order to increase efficiency further, the balancing slots, which are created on the rotor blades, can be used. These slots can be of different shapes, dimensions, and positions. Doosan Skoda Power has investigated this issue using measurements on an experimental turbine as well as numerical simulations. Some results of this study have already been presented. However, many details regarding the influence on efficiency have not been presented yet. This paper presents several possible balancing slots shape modifications. These modifications influence both the mass flow through the balancing slots and the torque of these slots. It can lead to an efficiency drop or an increase. A numerical model based on the experimental turbine was created to simulate several balancing slot designs. It represents the turbine flow-through part with two drum stages, including adjacent seals and investigated balancing slots. The numerical simulations were carried out using an Ansys software tools package. The computed flow fields and important integral results are shown. The influence of each design change is discussed. The resulting rules for designing the balancing slots are presented.

Low and often unstable steam parameters at the inlets and outlets of steam turbines, especially in waste-to-energy (WtE) facilities and heating plants, result in increased and often extreme demands on the flexibility and wide control range of steam turbine output. This necessitates a good prediction of machine behavior in different operating modes already at the design stage. This paper focuses on contributing to a more comprehensive computational treatment of the last stage blade erosion problem. The intensity of erosion is significantly affected by the dispersion of the coarse water phase, i.e., the size and amount of coarse water droplets formed by the breakup of water films on the blade surface. The Czech Technical University in Prague (CTU) and Doosan Skoda Power (DSPW) are currently developing a new in-house approach to predicting coarse water phase dispersion. This approach is based on empirical relationships but utilizes the Ansys commercial CFD solver in relevant steps to refine several parameters involved in these empirical relationships. By calculating the non-equilibrium condensation of steam in the turbine stages and by calculating the transport of primary droplets near the blade surface, boundary conditions for the in-house computational approach of the formation and movement of the water film on specific turbine blades were obtained. The output of the computational approach is the distribution of the coarse water phase dispersion, mainly in the trailing edge region of the stator blade of the last turbine stage along its span. Examples of the results obtained for a 34 MW DSPW turbine designed for a WtE facility are presented in the paper.

The work explores data reduction methods used in turbomachinery linear blade cascade experiment, such as area and mass averaging and Two-Dimensional Momentum Method, and their properties. Definitions of the most common methods are listed and the methods are tested on data sets acquired experimentally and also numerically. Differences in important characteristics of the blade cascade, such as kinetic energy loss coefficient, total pressure loss coefficient, or outlet angle, obtained by different methods are reported together with the change of conservative fluxes caused by reduction of the data. Information on the data reduction methods used by selected European research groups is provided.

A flow field inside steam turbine control valves can be very variable from subsonic to transonic and with or without a flow detachment. It depends on changeable operating conditions and a valve geometry. Understanding phenomena that appear in valves helps predict and reduce pressure losses, enhancing the valve design. It also ensures safe operation and a reasonable price. The primary motivation of this paper is to provide a detailed view of different flow fields at various operating conditions. For this reason, a control valve with one chamber is analyzed using numerical simulations. It is a part of the most often used valve assembly for different steam turbines with outputs from about 10 to 1000 MW. The numerical simulations were carried out using a package of ANSYS software tools. The settings and models are based on the previous comprehensive studies, where a validation using measurements on the model was performed. Eight different valve cone lifts were computed, ranging from fully closed to fully opened valves. For each valve cone lift, about twelve different pressure ratios were defined. As a result, a complete flow characteristic of the valve was created. Based on this, the most typical cases with different flow fields are shown. At first, the position of the sonic line is analyzed. It was found that it has a direct connection with the flow contraction coefficient, which is introduced and described in detail. It is also connected with the strong field non-uniformity under the valve cone. Second, the length of the backflow region under the valve cone and in the diffuser is described and analyzed. All in all, this paper complements the previous papers about valves for steam turbines, which have already been presented at ASME Turbo Expo, created by the research and development team from the Doosan Skoda Power Company and the Institute of Thermomechanics of the Czech Academy of Sciences.

Experimental test rigs such as linear blade cascades often use ambient humid air for internal aerodynamics investigation. However, experimental evidence of flow condensation can be found at higher Mach numbers, especially for lower inlet air temperatures. This paper presents a simple analytical model of the distribution of relative humidity of an air-water mixture and specific mass of condensed vapor as a function of Mach number. Moreover, based on the experimental data, a case study of the transonic flow in a linear blade cascade is performed using ANSYS CFX. The agreement between both approaches is more than satisfactory, leading to further investigations as little is known about the effect of humidity and flow condensation on the flow field.