INVESTIGATIONS ON JET BRAKE SYSTEMS IN PELTON TURBINES

Transcription

1 INVESTIGATIONS ON JET BRAKE SYSTEMS IN PELTON TURBINES G. Edinger, C. Bauer, J. Gaschl, R. Mack Abstract: Pelton turbines are often equipped with braking nozzles. There are variants with single or twin nozzle systems in use. This paper shows the braking characteristics of several variants. A coefficient of performance is defined, which allows to compare different arrangements of braking nozzles and different types of runners with different bucket shapes. Furthermore, the influence of vary the braking jet diameter was investigated. 1 Introduction Jet brakes are counter oriented brake nozzles that are used to decelerate the shaft line of Pelton turbines. While in former times also used for regular shutdown, today the jet brakes are mainly used as an emergency brake that gets active when the rise of mechanical failure occurs. The regular shutdown in modern Pelton turbine applications is done with electrical brakes supported by mechanical brake at the end of the coast down process. It is assumed that due to the extraordinary load there might be a bad influence on the durability of the runner if they are used for every shutdown. A first hint to this load and a torque path gives Lal [1], which is the only generally known experimental investigation on braking Pelton turbines with a nozzle. In [8], [9] and [10] are shown braking effects caused by discharging water of the previous bucket which are considered in the design of new runners. There are also other disadvantages occurring with braking nozzles. A leak in the valves of the nozzles hardly can be noticed, this results in a hard water hammer to the buckets. The main focus in this paper is to describe how effective the operation of a braking nozzle is and if there is any dependence on the rotational speed and to compare different arrangements regarding to their efficiency and braking effect. 2 Existing Arrangements Before set up the experimental investigations, existing arrangements had to be reviewed to get a survey witch variants are in use. In the present literature very less information can be extracted. Several arrangements were found in [2], [3] and [4]. Descriptions about mounting in the turbine casing and the position of the braking nozzles relating to the main injector can be found in [6]. Other information could be found in drawings from manufacturers of turbines. A single nozzle is most commonly used, for example at Kops I, Spullersee or Schneiderau III. There are also twin braking nozzles with two jets acting on one bucket such as at Uttendorf or Silz.

2 Also important, is the dimension of the brake jet relating to the bucket width or the main jet. Braking jet diameters are within 10% to 33% of the main jet diameter (in individual cases up to 50%). Concerning the bucket width the ratio is about 8% to 12% which is corresponding to the mentioned range above. This ratio is called β, defined in Equation 5. Another fact to be considered is the braking pitch diameter. The most common variant is to let the braking nozzles act on the main pitch diameter. But also variants with an intervention slightly inside or outside the main pitch diameter can be found in [2], [3] and [4]. 3 Experimental Setup The investigations were carried out in the Hydraulic Laboratory at the Vienna University of Technology, Institute for Energy Systems and Thermodynamics. To measure the shaft torque, a torque-measuring shaft and a hydrostatic bearing were used, which allows to consider friction losses. Furthermore the rotational speed was measured by a speed counter that is mounted on the torque-measuring shaft. The discharge of the braking nozzles was determined by the nozzle characteristics. The head was identified by measuring the static pressure considering the velocity head. Fig. 1. Experimental setup and braking nozzle To provide an easy adaption of the test setups the nozzles were mounted on an adjusting device and the water supply was provided by flexible hoses. The runners were driven by a variable speed motor-generator. This enables to brake the runner at a constant speed. That means a defined speed is preset and then the measurement is performed in a stationary condition. The range of speed variation was set to n ED =0.02 to 0.38 (see Equation 3). A speed factor of 0.38 corresponds within 70% to 80% over speed, this is in the range of a real runaway speed of Pelton turbines [7]. Zero speed could not be investigated because of the required minimal detection speed of the torque-measuring shaft. Figure 1 shows the experimental setup in the hydraulic laboratory and a braking nozzle. A detailed description of the whole setup gives [6]. The braking nozzles were equipped witch replaceable mouthpieces to vary the braking jet diameter so that different values of β could be investigated. Figure 1 shows also the braking nozzle

3 and some mouthpieces, manufactured by Voith Hydro. This paper shows the results of three different runners, Table 1 contains general information, and Figure 2 shows the bucket backsides of two runners.. A B Fig. 2. Bucket shapes runner A runner B runner C pitch diameter D (m) bucket width B (m) number of buckets (-) Table 1: General information Runner A is made of plastic. For this reason the maximum speed was limited to 1000rpm in order not to damage the buckets and to limit the impact load on the buckets. Runner B is an old design with fins on the backsides of the buckets. It is a metal runner, so the head was only limited by the operating pressure of the pipe system in the laboratory to 150m. Runner C is also a steel runner so the head could be up to 150m too. The bucket shape of runner C is similar to runner A. Figure 3 illustrate three arrangements which have been investigated Fig. 3. Arrangements of braking nozzles In Figure 3 the left arrangement (1), a single jet, is the most common found in [2], [3], [4] and in drawings from Voith Hydro. The variant in the middle (2) with two jets corresponds to the twin-braking nozzle arrangement. The jets come from outside the plane of symmetry. This is due to the bucket width and the diameter of the braking nozzle body. Because of the large outer diameter of the nozzle body this angle is required. The right arrangement (3) in Figure 3, two jets axial, does not exist at any plant but it has been investigated to show the effect, if the jet intervention is changed fundamentally.

4 4 Results To describe the efficiency of the braking effect a coefficient of performance is defined. The Coefficient of Braking Performance (CBP) is defined (Equation 1) which relates the braking power on the turbine shaft to the offered hydraulic jet power. To show the braking effect, the torque factor T ED, speed factor n ED,the specific flow rate Φ B (all corresponding [5]) and the ratio β are used (Equation 2 to Equation 5). CBP = Tω ρqe T ED = T ρd 3 E n ED = nd E (1) (2) (3) 4Q Φ B = z 0 π 2EB² (4) β = d B (5) CBP and T ED are shown depending on n ED. Since the jet diameter of a mouthpiece is constant, Φ B of an arrangement can only be changed by using a different mouthpiece or increasing the number of jets (see Table 2). The value of z 0 for two jet arrangements is also 1, because the two jets act on one bucket. Β / Φ B x10-3 β / Φ B x10-3 β / Φ B x10-3 two jets, β / Φ B x10-3 runner A 8.2% / % / 9.87 runner C 10.6% / % / % / runner B 9.3% / % / Table 2: Φ B and β of different arrangements Figure 4 shows the CBP of runner B and arrangement 1 (single braking nozzle). There are three different curves which correspond to different ratios of braking jet diameter to bucket width. The range of β is within 10.6% to 20%. The jet hits the bucket on the pitch diameter. As reference always the maximum values of a single braking nozzle acting on the pitch diameter, with β=10.6% and a runner C is used. As it can be seen in Figure 4, lower ratios of jet diameter to bucket width have an advantage in efficiency over larger ratios. At β=20% the arrangement has already a disadvantage of about 40% compared to β=10.6%. However, the torque is higher with a larger jet diameter, this fact can be seen in Figure 5. The higher flow rate does not work with the same efficiency on the buckets. This fact might be caused by increased splashing water on the braking jet and losses during the interaction with the bucket. Figure 6 illustrates an approximated braking hill-chart.

5 Fig. 4: CBP, influence of β Fig. 5: T ED, influence of β

6 Fig. 6. Braking hill-chart, runner C Fig. 7. Influence of bucket shape

7 Fig. 8. Two braking jets, influence of bucket shape Concerning the bucket shape Figure 7 shows, that runner A and C have a similar characteristic. The ratio β is 8.2% in case of runner A, 9.3% among runner B and 10.6% for runner C. With higher n ED the brake performance of runner B becomes worse than the others. This might also be caused by splashing water against the braking jet. Because of the fins on the bucket backsides of runner B, more water is thrown against the jet. In the normal operating range from n ED =0.03 to 0.20 the bucket from runner C doesn t have any disadvantage relating to the buckets of runner A. At very low speed, the bucket shape doesn t have a big influence. Noticeable is that the similar buckets of runner A and C have a different efficiency in a wide range. This might be caused of the shape of the bucket backside especially in the area of the splitter. The differences of β between the two runners might have an influence on this too. Figure 8 illustrates the characteristics of a twin-braking nozzle arrangement mentioned above. These investigations were only performed with β=8.2% in case of runner A and β=10.6% of runner B of each jet. Due to the bucket shape this variants are less efficient than the single nozzle arrangement. The braking jet is not so deflected if it hits the bucket in the backside of the splitter area. The torque factor is shown in Figure 9. Runner A achieves a better braking performance compared to a single nozzle arrangement. Due to the fins on the bucket backside of runner B the performance is worse than runner A. The water is deflected only about 90 because of the flat fins. This is similar to a jet on a plate. In case of runner A, however, the water can flow to the backside of the splitter and thus the deflection might be more than 90.

8 Fig. 9. T ED of twin braking arrangements A comparison of different arrangements is shown in Figure 10. All three variants mentioned above were compared regarding to their efficiency at β=8.2%. Arrangement 1, (single jet) and arrangement 2 (twin-braking nozzles) already have been discussed. The third arrangement (two jets axial) has a surprising good performance at higher speed. However, at very low speed, the runner is driven by the braking jets and the CBP is thus negative. In normal range of using braking nozzles, the performance is worse than other arrangements. At over speed the curve approaches the twin-braking nozzle arrangement. If the speed will be increased up to n ED >0.38 the efficiency of this axial variant might be better than the twin-braking nozzle arrangement. This fact was not to expect. Fig. 10. Runner A, comparison of different arrangements

9 Further experiments were made with a reduced braking pitch diameter. But there could not be observed a general influence. This is attributable to the shape of the bucket backsides. Sometimes a slight increase of the CBP can be measured, sometimes there isn t any significant influence. There is also a dependency on the head. At very low heads the CBP is a little bit higher at constant n ED. Above a head of approximately 40m the CBP remains constant. Arrangements with two nozzles were also tested with only one nozzle in use. Hence, a possible interaction of the two jets that has an influence on the braking performance can be identified. The experiments showed that there isn t any influence on the braking performance. CBP and T ED have at least the half value than the arrangements with two nozzles in use. 5 Conclusion In the present study several arrangements of braking nozzles with regard to their efficiency had been investigated. The experimental studies were carried out in the hydraulic laboratory of the Vienna University of Technology. To describe the braking characteristics shaft torque, flow, speed and pressure measurements were performed. In general it can be assumed that a smaller ratio β of braking jet diameter to bucket width is more efficient than bigger values of β. Twin braking nozzles are less efficient than a single nozzle, but there might be an advantage regarding the impact load. The relatively good performance of arrangement 3, two jets axial, was not expected before the tests. 6 Nomenclature CBP (-) coefficient of Φ B (-) specific flow rate braking performance T ED (-) torque factor T (N.m) braking shaft torque n ED (-) speed factor ω (s -1 ) angular speed D (m) runner pitch diameter ρ (kg.m -3 ) water density n (s -1 ) runner speed E (J.kg -1 ) specific energy Q (m 3.s -1 ) discharge CBP ref (-) reference coefficient of T ED ref (-) reference torque braking performance factor B (m) inner bucket width z 0 (-) number of nozzles β (-) ratio of jet diameter d (m) braking jet diameter related to bucket width References [1] J. Lal, Die Kennlinien einer Freistrahlturbine im Triebgebiet sowie im Bremsgebiet und die Wirkungsgrade im Triebgebiet, PhD Thesis, Federal Institute of Technology Zurich, Springer, Vienna, 1952 [2] L. Vivier, Turbines hydauliques et leur regulation, Michel, Paris, 1966 [3] G. Büchi, Le Moderne Turbine Idraulliche, Ed I Regolatori Di Velocita, Atlante, Hoepli, Milan, 1957

10 [4] G. Büchi, Le Moderne Turbine Idraulliche, Ed I Regolatori Di Velocita, Testo, Hoepli, Milan, 1957 [5] International Standard, IEC Second Edition, , 1999 [6] G. Edinger, Bremsverhalten verschiedener Bremsdüsenkonfigurationen in Pelton Turbinen, Diploma Thesis, Vienna University of Technology, Institute for Thermodynamics and Energy Conversion, 2009 [7] Zh. Zhang, Freistrahlturbinen, Springer, Berlin Heidelberg, 2009 [8] R. Mack, W. Rohne, S. Riemann, W. Knapp, R. Schilling, Using the potential of CFD for Pelton turbine development, 23 rd IAHR Symposium, Yokohama, October 2006 [9] R. Mack, W. Moser, Numerical Investigation of the Flow in a Pelton Turbine, Proceedings of the Hydraulic Machinery an Systems 21 st IAHR Symposium, Lausanne, September 2002 [10] R. Mack, T. Aschenbrenner, W. Rohne, M. Farhat, Validation of bucket flow simulation using dynamic pressure measurements, 22 nd IAHR Symposium on Hydraulic Machinery and Systems, Stockholm, July 2004 Authors DI Gernot EDINGER, Vienna University of Technology Institute for Energy Systems and Thermodynamics Getreidemarkt 9/302, A-1060 Vienna, AUSTRIA Phone: (30201), FAX , gernot.edinger@tuwien.ac.at Univ.Prof. DI Dr.Ing. Christian BAUER, Vienna University of Technology Institute for Energy Systems and Thermodynamics Getreidemarkt 9/302, A-1060 Vienna, AUSTRIA Phone: , FAX , cbauer@mail.tuwien.ac.at DI (FH) Josef GASCHL, Voith Hydro GmbH & Co. KG Linzer Straße 55, 3100 St. Pölten, AUSTRIA Phone: , FAX , josef.gaschl@voith.com DI Dr.Ing. Reiner MACK, Voith Hydro Holding GmbH & Co. KG Alexanderstraße 11, Heidenheim, GERMANY Phone: , FAX , reiner.mack@voith.com

11 Gernot Edinger graduated 2009 in Industrial Engineering from the Vienna University of Technology. He works now as a Project Assistant at the Institute for Energy Systems and Thermodynamics, Research Group Fluid Flow Machinery. Christian Bauer graduated 1991 in Mechanical Engineering from the University of Stuttgart. In 2000 he finished there his PhD in the field of Hydraulic Machines. After several years in industry Mr. Bauer is since 2008 Head of the Research Group Fluid Flow Machinery at the Vienna University of Technology, Institute for Energy Systems and Thermodynamics. Josef Gaschl passed the Engineering College HTL - St.Pölten in 1980 in mechanical Engineering. In 1981 he joined Voith Hydro as design engineer. In parallel to his carrier at Voith Hydro he graduated in Mechanical Engineering / Mechatronics from University of Applied Sciences in Mittweida / Germany. Actually Mr. Gaschl is a member of the Board of Management (CTO / COO) of Voith Hydro St.Pölten. Reiner Mack graduated from the Fachhochschule Ravensburg Weingarten in 1993 with an engineering degree in Technical Physic. He worked for different industrial research departments in Switzerland and the United States, specializing in Computational Fluid Dynamics. He joined Voith Hydro in 2001 where he is responsible for Pelton turbine development.