Study on Losses in the Intermediate-Pressure Steam Turbine Inlet Chamber

. Know-how about energy and pressure losses in all steam turbine parts is crucial to guarantee enhanced operational reliability and efficiency. This paper focuses on studying pressure losses in the intermediate-pressure turbine inlet chamber. Measurements were performed on a complex model consisting of the turbine inlet chamber, a compact valve assembly situated upstream, and nozzles located downstream. These intermediate-pressure turbine parts are larger due to the greater volumetric mass flow than the high-pressure turbine parts. However, their inner parts are proportionately smaller, which causes greater pressure losses. Measurements were carried out in the Aerodynamic laboratory of the Institute of Thermomechanics of the Czech Academy of Sciences, where the model was installed in an in-draft wind tunnel. The results were complemented by numerical simulations performed in the Doosan Skoda Power company using ANSYS software tools. Pressure losses were evaluated using the total pressure loss coefficient and, as a result, can be predicted in similar turbine inlet chambers with the required accuracy.


Introduction
To some extent, each steam turbine part should be investigated in detail to enhance the know-how and predict the efficiency to create a highly competitive turbine bid. Therefore, extensive studies of losses in the most common valve assemblies have been conducted recently. There was a high-pressure valve assembly for the nozzle [1,2] and throttle [3,4] governing along with a compact valve assembly for the intermediate-pressure turbine [5,6]. A valve assembly may consist of many parts, such as a stop valve and a control valve. In these parts, pressure losses and flow fields have been investigated using measurements and numerical simulations. As a result, a complex valve flow characteristic representing the dependence between the mass flow ratio, the pressure ratio, and the relative valve cone lift was created [6]. Except for these parts, there is the turbine inlet chamber between the valve assembly and the first turbine stage. Measurement of pressure losses in such a chamber had never been performed. This paper presents results for the first time.

Experimental model
The whole model for the current measurement is in Fig. 1. It consists of the valves already measured and numerically simulated [5,6] and the turbine inlet chamber, the subject of this new study. Downstream, there are a nozzle part and outlet pipelines, which do not represent any actual turbine part but connect the model with the vacuum storage. The aerodynamic suction-type modular tunnel is in Fig. 2. The air is sucked down from ambient conditions at the inlet through the tested model to the outlet pipelines, where a control nozzle sets conditions according to the required pressure ratio. The vacuum in the chamber can reach an absolute pressure of 10 kPa. In order to study the flow field in the turbine inlet chamber, four pressures around the circumference were measured upstream. It is shown in Fig. 2 and Fig. 3 in position pA. Other pressures pB and pC were measured on the outer chamber surface. Pressure pD was measured upstream of the nozzles. Positions are shown in Fig. 4. Four pressure extractions were used for pB, see Thermometers Pt100. The relative uncertainty of pressure measurement was evaluated to be about 4 % with a confidence interval of 95 %. Because experiments were carried out with dry air at low pressure, Reynolds numbers are approximately two orders of magnitude lower than the real valve assembly with flowing steam. This difference was analysed in [7,8], There is concluded that measurements at lower Reynolds numbers still provide accurate valve characteristics. It is based on the theory of models applied to valve assemblies [9]. There is also concluded that differences are minor and show slightly greater pressure losses. It is a conservative result to be applied to real valves.      Hexahedral and tetrahedral meshes, which satisfy typical quality criteria, were created. It is shown in Fig. 7. It was also kept a recommendation for the compressible turbulence model k-ω SST to have y+<5 [10,11]. It was based on similar projects [12], where mesh sensitivity was tested. The medium was defined as dry air following the experiment where the silica-gel drier guarantees dryness, see Fig. 2. To compute the steady-state RANS solution, a highresolution scheme was used. Boundary conditions were defined according to the measured data. 10 -4 residual target values judged simulation convergence criteria and imbalances of pressure losses in the chamber were monitored to be less than 0.1 %.

Results and discussion
At first, a valve flow characteristic representing the dependence of the mass flow ratio m/mcr on the pressure ratio ε2, and the relative valve cone lift z/D, was evaluated. The pressure ratio ε2 can be calculated from outlet static p2/p1t or total pressures p2t/p1t. Index 1 means the inlet and index 2 means the outlet of the valve, respectively. m is the mass flow and mcr is the critical mass flow in the valve calculated from the inlet flow parameters and the diffuser throat diameter D. z is the valve cone lift. Fig. 8 compares the measurement results on the valve assembly alone (the dashed lines) with the new model with the turbine inlet chamber (the full lines). The resulting differences are not significant. As a result, it can be concluded that the valve characteristics, widely used for pressure loss prediction, are not dependent on the downstream geometry. It also shows that the measurement is consistent with the previous one.  The valve flow characteristic comparison was also evaluated from CFD results for z/D = 0.322, where the valve is fully open. For such a case, a pressure loss prediction is required. Using the CFD, static and total pressure ratios could be evaluated. Results are in Fig. 9. Some differences can be observed in the case of the total pressure ratios p2t/p1t < 0.96. Fortunately, this region is far away from the area of practical use, which is under the dashed line in Fig. 9.     In order to generalize the results, the pressure drop from the inlet to the outlet of the turbine inlet chamber should be evaluated not only from a few measuring points but from average values at the inlet and outlet plane. Furthermore, for practical use, the pressure drop between locations pA and pD should be better evaluated from the CFD results since the positions for measurement are limited, see Figs. 3 to 5. Results for the required case with z/D = 0.322 are shown in Fig. 13. It can be noticed that the static pressure drop is almost the same no matter whether it is calculated from the pressure at the whole outlet (before the  Fig. 4) or the measurement points (pD). Comparing experimental and numerical results, it could be seen that the differences are reasonable up to m/mcr < 0.3. For greater m/mcr, the differences are significant. The reason is that in the region of m/mcr > 0.3, strongly unsteady phenomena such as flow detachments appear, but the numerical simulations are performed as steady-state. It was described and shown in detail in [6]. Fortunately, for a practical pressure loss prediction, the valves and the turbine inlet chambers are designed to be used only for conditions where m/mcr < 0.3 and where the differences are reasonably small.
If the differences in the static pressure drop are minor (for m/mcr < 0.3), there is an assumption that differences in the total pressure drop are also small in this range. As a result, the total pressure losses should be evaluated. The standard way to evaluate and generalize the loss in the inlet turbine chamber is to use the total pressure loss coefficient Y [13]. It is defined as: The results of Y for all calculated cases are in Fig. 14 on the left. It is visible that the total pressure loss coefficient is not constant. At lower relative mass flow rates, its value oscillates. The reasons for these oscillations are going to be analyzed in further studies. However, the average value of Y, shown by the dashed line, can be used very well for practical use. Results of calculated total pressure losses using the average value Y are shown in Fig. 14 on the right by the dashed line and using the exact values of Y by the solid line, respectively. For m/mcr < 0.3, the differences in the total pressure loss between the exact and average values are lower than 0.10 %. It is a reasonable difference for a practical pressure loss estimation.

Conclusions
Measurement and numerical simulations were performed on the complex model of the intermediate-pressure turbine inlet chamber. As a result, it was found that the presented measurement is consistent with the previous one [5,6], and the valve flow characteristic is not dependent on the downstream geometry. It can be observed that differences in pressure drop in the inlet turbine chamber between experimental and numerical results are reasonable for the often-used turbine operating range, where m/mcr < 0.3. For greater mass flow ratios, the differences are significant. For practical use, total pressure loss prediction for inlet chambers of the intermediate-pressure turbine was generalized in the form of the average value of the total pressure loss coefficient.