Shop Test Result of V64.3 Gas Turbine

Transcription

1 CS oc^) THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS 345 E. 47 St., New York, N.Y. 117 The Society shall not be responsible for statements or opinions advanced in papers or in discussion at meetings of the Society or of its Divisions or Sections, or printed in its publications. Discussion is printed only if the paper is published In an ASME Journal. Papers are available from ASME for fifteen months after the meeting. Printed in USA. Copyright 1991 by ASME 91-GT-224 Shop Test Result of V64.3 Gas Turbine M. JANSEN, T. SCHULENBERG, D. WALDINGER Siemens AG Muelheim Fed. Rep. Germany ABSTRACT The V64.3 6MW combustion turbine is the first of a new generation of high temperature gas turbines, designed for 5 and 6Hz simple cycle, combined cycle and cogeneration applications. The prototype engine was tested in 199 in the Berlin factories under the full range of operation conditions. It was equipped with various measurement systems to monitor pressures, gas and metal temperatures, clearences, strains, vibrations, and exhaust emissions. The paper describes the engine design, the test facility and instrumentation, and the engine performance. Results are given for turbine blade temperatures, compressor and turbine vibrations, exhaust gas temperature, and NO emissions for combustion of natural gas and fuel oil. INTRODUCTION As gas turbines running at 5 and 6Hz have increased more and more in power output beyond 1MW, Siemens/KWU is now introducing the V64.3, a medium size 6MW gas turbine, running at 54rpm, for application at 5Hz and 6Hz electric utilities and for large industrial applications. In particular, application of this new high temperature gas turbine is intended for - electric utilities, either simple cycle for peak load power stations or combined cycle for base load plants, with a net efficiency of more than 5%, - conventional fuels as natural gas or oil, as well as coal gas from integrated coal gasification plants, which are expected to be economical within the next years, - district heating, where medium size gas turbines could be of high interest for local applications, - cogeneration of electric energy and process steam for industries around the plant site. Fig.! V64.3 Gas Turbine Presented at the International Gas Turbine and Aeroengine Congress and Exposition Orlando, FL June 3-6, 1991 This paper has been accepted for publication in the Transactions of the ASME Discussion of it will be accepted at ASME Headquarters until September 3, 1991

2 Beginning in March 199, extensive component and load tests have been performed on the test bed in the Berlin factories.the first two units of V64.3 gas turbines have already been delivered for installation to a combined cycle power station near Helsinki. These have gone into commercial operation in autumn 199. ENGINE DESCRIPTION A cross section of the new model V64.3 combustion turbine is shown in Fig.!. It is obvious that most of the design principles of the well approved V94.2 and V84.2 gas turbines have been adopted. Again it is a single-shaft engine with interlocking disks, stressed by an axial tiebolt to ease maintenance. It has two silo combustion chambers which are now arranged horizontally, again flange mounted on the gas turbine casing, a 17 stage compressor and a 4 stage turbine. Details of the design have already been described by Becker and Ziegner (1988). Significant improvements of the thermodynamic performance were intended by increasing the pressure ratio to about 16 and by increasing the turbine inlet temperature. The first 4 compressor stages were equipped with adjustable guide vanes to improve performance at part-load. Profiles of compressor and turbine bladings were of new design. Due to the higher turbine inlet temperature the transition piece between the two combustors and the turbine inlet was heat protected by shielding plates. Fig.2 Model V64.3 Gas Turbine During Assembly With the higher pressure ratio than in current machines, an intercooler was provided for more effective blade cooling in the first turbine stage. On the other hand, the combustors are almost identical to the approved V94.2 combustors, except that only 3 burners per combustor were required for this smaller gas turbine. The same hybrid burners as in a V94.2 were used which enabled lowest NOx emissions, as described by Becker et al.(1991). Ceramic tiles were used again in the combustors to protect the flame tube. Fig.2 shows a V64.3 during assembly. In the front the mechanical drive of the 4 adjustable compressor guide vanes are visible. Moreover heat shields of the transition piece can be seen in the background. TEST FACILITY The test facility at the Berlin factories was designed to test heavy duty gas turbines under full load conditions. It is described by Deblon (1978) and by Schulenberg and Bals (1987). Instead of a generator, the gas turbine was connected with a water brake, and the released energy was submitted to 3 cooling towers. Different from previous arrangements a gear box was installed between water brake and gas turbine, which will also be required for electric utilities to transform the 9Hz rotational speed to a 5Hz or 6Hz network. The total arrangement is shown in Fig.3. Tests were run by supplying natural gas and fuel oil, both by using the same hybrid burners. The main objectives of these tests were to validate the thermodynamic performances of compressor, combustors, and turbine, to ensure proper cooling of all components which are exposed to hot gas, and to determine dynamic stresses under all load conditions. Therefore several measurements were taken in addition to the standard measurements which are required during commissioning. They are summarized in Tab.!. This list includes static pressure measurements by pressure tappings, thermocouple measurements of the gas temperature in compressor and turbine, as well as in the exhaust duct, and wall temperature measurements. The latter ones were obtained both with thermocouples and with a pyrometer which could be inserted into each turbine stage, in between the vanes, to observe the proceeding and the succeeding running blade as described by Schulenberg and Bals (1987). Blade dynamic stresses were detected with strain gauges on compressor stages 1 to 4 and 9 and on all turbine stages, each on the airfoil 2

3 H Measuring Container Gear Box Fuel Supply Starting Motor I ; Water Brake Ill. piiii IpjkI mm I^ I -& 4 4- Measuring Container L } Measuring Container Lube Oil Tank V64.3 Gas Turbine Fig.3 Arrangement in Test Bed in the Berlin Factories and on the blade root.in addition, vibrations of the fourth stage blades were observed with a laser optical system as described by Roth (198). This additional information enables to predict the vibratory stresses of the first mode of all blades of this stage, whereas strain gauges yield complete vibration spectra of only a selected number of blades. Measurements of clearances were taken in order to minimize leakages and tip losses of compressor and turbine bladings. During the tests, emissions of NO, NO2, CO, CO2 and UHC could be measured continously in the exhaust duct. All data were stored by a new data acquisition system. Thermodynamic performances were evaluated continously during operation,so that the control Type of measurement Compressor Combustor Turbine Pressure Temperature Flow temperatures Component temperatures with thermocouple with radiation pyrometer Cooling air mass flow 4-8 Blade vibration with strain gauge with laser optic system Clearances 2-16 Tab.!: Test Run V64.3; Measurement Points center could be informed about trends and behavior already during the tests. Fast signals such as blade vibrations and pyrometric temperatures were stored additionally on analogue tapes. They were digitized later on. TEST PROGRAM After installation in the test stand, commissioning of V64.3 gas turbine was started at Feb.28, 199. One week later we tested ignition and acceleration of the gas turbine to operation speed. From this test the start-up curves and the lube oil system were optimized, and the control and measuring systems were checked. On April 2, the gas turbine was fully loaded up to 65MW, using natural gas in the diffusion burner mode. From this test we obtained all material temperatures measured with thermocouples, the exhaust duct temperature pattern, all vibratory amplitudes, and the engine performance. Compressor and turbine bladings performed well. Measured data were within the design limits. Some hot metal temperatures in the combustors, however, required some modifications. A second test phase was conducted in June 199 with modified combustors. These tests included optimization of the adjustable compressor vanes, pyrometric measurements of turbine blades, and laser optical measurements 3

4 of the last stage blades. A third test phase in July 199 included also tests with natural gas in the premix burner mode and with fuel oil. In the following sections results of all these test will be discussed in more detail. ENGINE PERFORMANCE The first loading cycle in April 199 revealed already that power and efficiency of the gas turbine were higher than predicted. Differences in efficiency were mainly due to significantly higher turbine efficiency, whereas the compressor efficiency only slightly exceeded the predicted level. The measured gas turbine efficiency at the coupling, turbine inlet and outlet temperature and compressor pressure ratio are shown in Fig.4. All data have been scaled with the base load performance. The thermodynamic data at base load and peak load, corrected to ISO conditions (i.e. 15 C ambient temperature, 113mbar ambient pressure, 6% humidity) are listed in Tab.2. In these data all measuring uncertainties and tolerances of manufacturing as well as gear losses are subtracted. Therefore the power and efficiency of future machines can be expected to be higher than listed. 1W C 1 C 'v.9 m mass flow pressure ratio Power output Power output Fig.4 Performance of V64.3, Test Results Base Load Peak Load Standard fuels Gas Oil Gas Oil Low heat value (LHV) kj/kg Hu Hu Turbine output at coupling MW PK_IG PK_IS Turbine efficiency at coupling % 11 K-IG K_tS Compressor pressure ratio - IIv_iG nv_is Fuel mass flow kg/s m Br mgr Turbine exhaust mass flow kg/s m Tii mtll Turbine exhaust temperature C ltii-ig OTII-IS Tab.2: Thermodynamic Data at ISO Conditions 4

5 EMISSIONS During the first test phase the gas turbine was run with natural gas. The hybrid burner, which has successfully been applied already to all types of Siemens gas turbines, was also installed in the V64.3. This burner can be operated with natural gas, either in a diffusion burner mode or in a premix burner mode. Details are described by Becker et al. (199). Emissions of NO and CO for the diffusion burner mode are shown in Fig.5. At zero load the compressor guide vanes were closed, in order to obtain high turbine exhaust temperatures and low CO emissions. Adjustable dilution air holes in the combustors were open, by which combustion air was bypassed to minimize CO emissions. W 3C.2 pprr E 25C z 2C 15 1C 5 ^CO / max. Dilution air min. NOx 1r TY' 1 Compressor closed guide vanes op en So the NOx emissions could be minimized over the whole load range. A further, efficient reduction of NO x emissions in the diffusion burner mode will be achieved by injection of water and steam. Tests will be performed next. At base load a NOx reduction to about 1% of the dry emissions, Fig.5, is expected. A significantly better NO x reduction was obtained however with the premix burner mode. By switching over at 26% load, as shown in Fig.6, NO emissions were kept below 2ppm, for all loads up to base load. Here again, the adjustable compressor guide vanes and the dilution air openings in the combustors were controlled such that almost a constant fuel-to-air ratio was gained. Below 26% load however, the diffusion burner mode is preferable in comparison to the premix burner mode which causes higher CO emission in that range. 3 o ppm E 25 OU «x2 z 15 1C max. Dilution air closed NOx min. Compressor guide vanes open in the dry exhaust gas with 15% O 2 by Volume Power output at coupling Fig.5 Emissions of NOx and CO, Natural Gas in Diffusion Burner Mode With increasing load the compressor guide vanes were opened as soon as the maximum turbine exhaust temperature had been reached, i.e. at 5% power output as shown in Fig.4. From 26% load to 5% load the adjustable dilution air openings were closed, giving an almost constant fuel-to-air ratio in the flame within this load range. 5C /CO in the dry exhaust gas with 15% 2 by Volume Power output at coupling Fig.6 Emissions of NOx and CO, Natural Gas in Premix Burner Mode COMPRESSOR The static pressure rise in the compressor and diffuser showed good agreement with the design calculation. The anticipated compressor efficiency was in 5

6 fact exceeded during testing. Because the first four stages were fitted with adjustable vanes, the efficiency for 5% power output decreased only to 95% of the base load efficiency. The tip clearances of the unshrouded rotor blades have been minimized to achieve maximum possible compressor efficiency. Measurements of the rotor blades in rows 1, 4, 8, 13 and 16 were performed as part of the test program to check tip clearances. Two techniques were used to measure clearances. Minimum clearances were measured in each run with four abrasion pins arranged around the circumference. In addition the non-steady-state behavior of the tip clearances was investigated using a trace pin method developed by Siemens. Applying this method a pin is moved step-by-step into the radial gap until it has contact with the rotating blades. By this method it is possible to determine the local tip clearances at individual times during operation. Figure 7 shows the non-steady-state tip clearances in the first and fourth rows of rotor blades. The clearance is related to the cold gap. The associated speed curve is shown in the bottom half of Fig.7. During start-up the clearances are reduced to 48% (row 4) and 35% (row 1) of that in the cold condition as a result of centrifugal forces. The subsequent increase in tip clearances is attributable to thermal expansion. Looking at the fourth row of rotor blades the slow heat-up of the compressor disks becomes apparent only 2 minutes after rated speed is achieved. Fig.7 also shows the results obtained with abrasion pins. These results are the averages of the four measurements around the circumference. The measurements show that the Model V64.3 can be reliably operated with minimal tip clearances in favor of compressor efficiency. Vibration measurements were performed on rows 1, 2, 3, 4 and 9 to check the dynamic stresses acting on the compressor rotor blades. In each stage two blades were fitted with strain gauges. The signals were monitored with the aid of a telemetric system. Figure 8 shows the result of vibration measurements for the rotor blade in the first compressor stage. The upper half of the figure shows the Campbell diagram. It is apparent that the first natural frequency of the blade as required does not coincide with integral multiples of the rotor's rotational frequency (harmonic). The Campbell diagram also shows that, in the range from zero to rated speed, resonances between the natural frequency modes and speed hafmonics cannot be avoided. Special attc tion is paid to the effect of nozzle excitation. 4 >. f r c no 3 c LL cc2 Operating Range U E m 2 1. U 4.8 Trace pins X Abrasion pins Speed d.4 6 Max.6 6 Allowable Blade 4.3 d Q.4 I- Blade 1 U, y o..i.1 / j r"tart,/////z; ^-- Operating speed dohw n Time in min Speed ^y_. / TiSpa Fig.7 Real Time Blade Tip Clearances Fig.8 First Stage Compressor Blade Frequency and Stress

7 The alternating stressing of the blade, recorded by the strain gauges was converted for each eigen mode to the point of maximum blade stress. Here we reverted to the results of finite element calculations. The results of this analysis are shown in the lower half of Fig.8. The stresses are related to the allowable strength. Maximum stresses occur during start-up as a result of non-incident flow conditions and because of resonance excitation in the first eigen mode. The stresses encountered are, however, low by comparison with allowable stresses (fatigue strength of the structure). They are within the empirical range, of our well approved V94.2 and V84.2 gas turbines. Figure 9 shows the Campbell diagram and the cyclic stresses measured for the rotor blades in the fourth stage. 6 f Operating Range 6 I ^^ c no I 5 E 5. z C. LL st Eigenvalue 4 6 Max Allowable Speed U) U) I I N ^ I.2 I I ^ /a I.1 o tr ao ^ a o4, Speed Fig.9 Fourth Stage Compressor Blade Frequency and Stress These stresses are likewise converted to the point of maximum blade stress. The maximum stresses in this stage are also low. In Tab.3 the highest measured dynamic stresses of the first compressor blades are listed. They are related to the allowable stress of the blade. Most of these stresses occur only during start-up or shut-down. 4 3 In the operating speed range only 14% of the allowable stresses have been measured. Row Rel. Speed Alternating Stress Allowable Stress Tab.3: Maximum Dynamic Stress of Compressor Blade for Whole Speed Range TURBINE The first seven of the eight blade rows in the four-stage turbine are cooled. The moving blades in the first stage are shrouded in order to reduce tip losses. The proven combination of convection and impingment was selected as the cooling method. To verify the effectiveness of cooling, the surface temperature of the blades in the first stage was measured in addition to the gas and cooling air temperatures. Several thermocouples were installed in two guide vanes in each row. The surface temperatures of the moving blades were measured with the aid of pyrometers which were fitted between stationary vanes. They could be moved transversely across the blade height, as described by T.Schulenberg and H.Bals (1987). Figure 1 shows the surface temperature of the stationary vane in the first stage for a section at 75% of the blade height. The measurement was carried out for a load of 65MW. The temperature measured in the range of the stagnation point is higher than that calculated. Although this surface temperature is just allowable, the cooling of these vanes will be improved for future machines, to achieve a more uniform temperature distribution on the airfoil. Figure 11 shows the calculated surface temperature curve of the moving blades in stage one at 5% blade height for the same turbine inlet temperature. The data recorded with the pyrometer are within the limits of design data for this operation point. 7

8 C) E - Calculated Measured Values at 75 % Blade Height Design limit account the existing scatterband in the dynamic behavior of the blades as a result of fabrication tolerances. The dynamic stress of the blades in the fourth stage is shown in Fig.12 represented as a function of output. Figure 12 shows the measured maximum stress converted to the point of maximum blade loading. There is an adequately large margin to the permissible stress. d Mc^ CO 6d Max 6d Allowable 15 C N C) U) UE Pressure Surface LE Suction Surface T Fig.1 Temperature Profile of First Turbine Vane Power in MW a) C) a E C) I- C) Calculated Measured Values at 5 % Blade Height Design limit Fig.12 Vibration Stress of Fourth Turbine Blade In Tab.4 the highest measured dynamic stresses for all turbine blade rows are listed. The dynamic stresses are far below the allowable limit. Row d Max skj Allowable C Suction Surface LE Pressure Surface Fig.11 Temperature Profile of First Turbine Blade The dynamic stress of all turbine blades was investigated with the aid of strain gauges. In addition, vibrations of the fourth stage blades were observed with an optical method as shown in H.Roth (198). This method covers each blade in the row under observation. The results take into Tab.4: Maximum Dynamic Stress of Turbine Blades CONCLUSION - Full load tests of the gas turbine V64.3 demonstrated that the expected performances were exceeded. - Safe and reliable operation of the engine was proven. - NO and CO emissions are extremely low using the dry low NO hybrid burner. - The V64.3 is the smallest machine of a new Siemens gas turbine generation. These have been scaled in dimension and rotational speed. Therefore the perfomance of these future machines can also be estimated. 8

9 ACKNOWLEDGEMENTS The authors wish to thank Mr.Brandner for the completion of the tests. This work was supported by the Bundesministers fur Forschung and Technologie, Forderkennzeichen 3E697A. REFERENCES B.Becker and M.Ziegner, The New Siemens/KWU Model V64.3 Gas Turbine, Motortechnische Zeitschrift 49,6; , B.Becker, P.Berenbrink, T.Schulenberg, Operating Experience with Hybrid Burners, Submitted for Publication to 19th CIMAC Congress, April 15-19, B.Deblon, Full Load Tests of the 8MW Gas Turbine V93.2 Using a Water Break, ASME 78-GT-68, T.Schulenberg and H.Bals, Blade Temperature Measurements of Model V84.2, 1MW/6Hz Gas Turbine, ASME 87-GT-135, H.Roth, Vibration Measurements on Turbomachine Rotor Blades With Optical Probes, Fluids Eng. Gas Turbine Conf., New Orleans, March 1-13, 198, Publ by ASME, New York, , 198. B.Becker, F.Bonsen, G.Simon, A Simple and Reliable Combustion System, ASME 9-GT-173, 199. E