Bigger Blades Cut Costs by joshcorey


									                                              STEAM TURBINE TECHNOLOGY

Bigger blades cut costs
This year will see the commercial                                                                       turbine are from 5 to 21 kg/s/m2 (3 500 to 15
                                                                                                        000 lbm/h/ft2) end load; from 150 to 300 m/s
introduction of the world’s largest                       Figure 1. Last                                (500 to 1000 ft/s) annulus velocity; and from
steel last-stage blades for steam                         stage blade                                   4 to 10 per cent exit moisture.
                                                                                                           Based on a review of market needs, GE and
turbines.                                                                                               Toshiba chose the primary design point for
                                                                                                        the 40 in and 48 in last stage blades to be 11
                                                                                                        kg/s/m2 and 225 m/s annulus velocity (8000
Amir Mujezinovic, GE,                                                                                   lbm/h/ft2, 750 ft/s annulus velocity), and 8 per
Schenectady, NY, USA                                                                                    cent exit moisture. In addition to this prima-
                                                                                                        ry design point, two secondary design points
                                                                                                        were selected: 8 kg/s/m2 and 200 m/s annu-
                                                                                                        lus velocity (6000 lbm/h/ft2, 650 ft/s annulus

               he shape and design of a steam
               turbine blade determine how                                                              velocity); and 20 kg/s/m2 and 275 m/s annu-
               much of the energy of the steam                                                          lus velocity (15000 lbm/h/ft2, 900 ft/s annu-
               is turned into work. A longer last                                                       lus velocity). The design is optimised at the
               stage blade increases the power                                                          primary design point while the secondary de-
output capability of the steam turbine, which                                                           sign points are monitored to ensure that no
in turn leads to improved power plant effi-                                                             performance problems are introduced at
ciency and a lower electricity production cost.                                                         these points.
   In the 1980s most turbine manufacturers
had developed and put into service titanium                                                             Aerodynamic design
blades with similar annulus areas. Due to sub-                                                          The last three stages work together as a sys-
sequent improvements in both design                                                                     tem and are designed aerodynamically using
methodology and material characteristics, GE                                                            a combination of streamline curvature design
Power Systems and the Toshiba Corporation                                                               methods, two-dimensional cascade analysis,
made the decision to develop new blades                                                                 and state-of-the-art three-dimensional compu-
using steel as a blade material. The use of steel                                                       tational fluid dynamics analysis techniques.
results in a lower cost to the customer and                                                                This design employs advanced aerodynam-
avoids uncertainty in the supply, and there-                                                            ic features including meridional flowpath con-
fore price, of high quality titanium forgings.                                                          touring, axial and tangential compound lean
   GE and Toshiba recently completed the de-                                                            of the L-0 nozzle, and tailored exit profiles
velopment of new 40 in and 48 in steel last-                                                            from the L-1 stage to allow a radius ratio of
stage blades for steam turbine applications                                                             0.43 in the L-0 blade.
worldwide. In terms of annulus area, the 48                                                                To further reduce the radial pressure gradi-
in blade is the largest steel 3000 rpm last stage                                                       ent at the L-0 nozzle exit, compound tangen-
blade in the world.                                                                                     tial lean is employed in the last stage nozzle.
                                                                                                        This tangential lean has the effect of intro-
Increasing the annulus area                                                                             ducing an inward radial force on the flow –
One of the loss mechanisms in the steam tur-                                                            forcing more flow into the hub region and in-
bine is the kinetic energy of the steam as it                                                           creasing the pressure in this region.
leaves the last stage blade. The lower the ki-                                                             Combined with the flowpath contouring,
netic energy, the higher the steam turbine ef-                                                          the L-0 nozzle lean reduces the radial pressure
ficiency will be. The magnitude of loss is                                                              gradient at the nozzle exit, raises the root re-
proportional to the square of the ratio of the                                                          action, and allows for a lower hub/tip ratio.
volume flow rate of the steam through the last                                                             Axial compound lean is applied to the L-0
stage of the steam turbine and the annulus area                                                         nozzle in addition to the compound lean in
of the turbine exit. To decrease the loss, a larg-                                                      the tangential direction. The primary purpose
er turbine exit annulus area is needed.                                                                 of the axial lean is to increase the nozzle-to-
   An increase in the last stage blade annulus                                                          blade spacing at the tip while maintaining a
area can be accomplished by either using                                                                smaller axial spacing over the remainder of the
shorter blades mounted on a larger diameter
rotor (larger “hub”) or by using longer blades
mounted on a smaller diameter rotor. These                    Steam conversion valve
opposing approaches yield different radius ra-
tios, a key parameter in the aerodynamic and
mechanical design of the last stage blade. In
                                                                                                                 LP turbine
                                                        Boiler                             HP turbine
the GE–Toshiba development project, an op-
timisation of the aerodynamic and mechani-
cal considerations of the design resulted in:
● A new last stage (L-0) blade length of 1016
   mm (40 in) for 60 Hz application and 1219                               Dynamometer                                           Dynamometer
   mm (48 in) for 50 Hz application.                                            A                                                     B
● A hub diameter of 1565 mm (61.6 in) and
   1880 mm (74 in), for 60 Hz and 50 Hz L-0                                  Spray water
   blades, respectively.
   Steam turbine last stage blades typically are                                                                                  Condenser
applied to different machine configurations                                                 Flow meter
with varying outputs and operating points.
This approach requires a robust design for a
wide range of operating conditions. Typical          Figure 2. Test steam turbine system
ranges of operation for the last stage of steam

                                                                                                               February 2003 Modern Power Systems   25
                                             STEAM TURBINE TECHNOLOGY

                                                                                                         2. Analysis using a reduced frequency analy-
                                                                                                         sis calibrated to empirical data.
                                                                                                         3. Validation testing in a subscale test rig,
                                                                                                         where the last three rotating and stationary
                                                                                                         stages were tested in actual steam conditions,
                                                                                                         at a variety of operating conditions (combi-
                                                                                                         nations of the steam axial exit velocity and
                                                                                                         condenser pressure). Blade mounted dynam-
                                                                                                         ic strain gauges were used to measure blade

                                                                                                         The experimental verification of the overall
                                                                                                         stage efficiency of the new steel last stage
                                                                                                         blade was performed in an experimental low-
                                                                                                         pressure model turbine. A schematic diagram
        Figure 3. Test model turbine                                                                     of the 10 MW model steam turbine facility
                                                                                                         used is shown in Figure 2, and a photograph
                                                                                                         of the model turbine train used in this system
height. A larger axial spacing is desired at the    (aeromechanical design) of the airfoil is that       is shown in Figure 3.
tip to allow additional time for the water          the blade natural frequencies – those fre-              Model turbine tests have been performed
droplets torn from the trailing edge of the noz-    quencies at which vibration will take place in       with several different end load conditions.
zle to accelerate to flow velocity before en-       the absence of any continuing excitation at          The LP end exit axial flow velocity conditions
tering the L-0 blade. Better matching of the        running speed – must have a margin from a            varied from 94 to 212 m/s (310 to 700 ft/s).
droplet and flow velocity reduces erosion and       range of multiples of engine speeds (so-called          The results show that the newly developed
contributes to the long-term reliability of this    per-rev lines). The per-rev excitation comes         steel 48/40 in last stage blade has excellent ef-
design.                                             from different sources, most of which are as-        ficiency at the design condition and also in the
                                                    sociated with the non-uniform flow in the            partial load conditions.
Mechanical design of stages                         steam path, either upstream or downstream               Mechanical testing confirmed the mechan-
On the last stage blade the most important me-      from the blade row.                                  ical design calculations described above. The
chanical design features are the curved axial          In addition to the per-rev excitation, long       running speed to zero speed airfoil transfor-
entry dovetail (a portion of the blade that         low-pressure blades (in particular last stage        mation, static design and the dynamic design
mates with the rotor), nub and sleeve as a part     blades in some of the off-design operating           (blade natural frequencies) were validated in
span damper, and an integral blade cover (see       regimes) can experience vibration induced by         a wheel box test.
Figure 1).                                          flow (aeroelastic instability). Design solutions        The test set-up consisted of a full-size single
   The mechanical design consists of several        dealing with aeroelastic instability have been       flow rotor with the last three stages of blades
interconnected steps: transforming an aero-         primarily empirical in nature and specific to        (also full size) assembled on it (Figure 4). The
dynamic shape of the airfoil into an as-ma-         application.                                         blades are instrumented with strain gauges
chined shape; static design; dynamic design            The risk of aeroelastic instability is greatest   and the entire assembly is put into a spin cell
(or aeromechanics design); and erosion de-                                                               – basically, a bunker from which air can be
sign.                                                                                                    evacuated. The test rotors are spun to the de-
   The airfoil shape used in the aerodynamic                                                             sired speed using a drive turbine, the blades
calculations is the one that the airfoil will as-                                                        excited and blade frequencies measured at
sume at the running speed. The low-pressure                                                              various speeds.
blades untwist during acceleration from rest                                                                In addition to the blade frequency mea-
to running speed. A necessary step during the                                                            surement, a rotor end mounted torsional shak-
mechanical design is to follow a process that                                                            er was used to measure coupled rotor blade
determines the zero-speed airfoil shape that                                                             torsional frequencies. To confirm the calcu-
yields the desired aerodynamic shape at op-                                                              lated blade untwist, a strobe light was used to
erating speed.                                                                                           visualise the cover untwist. Results confirmed
   The basic premise in the static design of the                                                         the rotating speed at which adjacent covers
airfoil and its accompanying rotor wheel is                                                              engage and from that, it was inferred that the
that both the maximum average stress in all                                                              analytical transformation of the airfoil from
the wheel and blade sections – and the maxi-                                                             running airfoil to the zero speed airfoil was
mum local stress throughout the blade and the                                                            correct.
wheel – are kept under a certain level, estab-      Figure 4. New blades installed on a rotor               During wheel box testing, both static and
lished to ensure substantial margin to failure                                                           dynamic strain gauges were used for blade in-
under all operating conditions and extreme          in the region of low flow (low steam axial exit      strumentation. Static gauges were mounted
faulted conditions.                                 velocity) and high condenser pressure (a far-        both on the airfoil and in the wheel and blade
   The maximum average stress (assuming             off-design condition). In such conditions a          dovetail to measure static stress level. The re-
fixed material properties) determines the           massive flow separation from the hub of the          sults correlated very well with the analytical
overspeed margin: when compared to the              last stage blade forces the migration of the         results. The dynamic strain gauges measured
yield strength of materials (blade and wheel)       steam flow to the upper portion of the last          blade-vibrating frequency and coupled rotor
it determines the overspeed at which section        stage blade. Flow separation from the airfoil        blade torsional vibration. The measured blade
gross yielding would occur; and when com-           then can cause blade stall flutter, and flow in-     frequencies confirmed the analytical predic-
pared to the ultimate strength, it determines       stability can cause buffeting of the blades.         tions and the measured coupled blade rotor
the overspeed at which ductile failure would           GE and Toshiba took a three-fold approach         torsional vibrations confirmed the predicted
occur.                                              to guard against such instability:                   margins from two times electrical frequency.
   In contrast, the maximum local stress de-
termines the low cycle fatigue life (number of      1. Use of a last stage blade design that employs     Market introduction
cycles to a crack initiation) and also plays a      integral covers and part span dampers. This          Overall, the newly developed 40/48 in last
role in the level of stress corrosion cracking      construction increases the frequency of the          stage group is a significant contribution to
(SCC) risk. The 40 in and 48 in last stage blades   row of blades (when compared to free-stand-          steam turbine performance and reliability.
were designed to conservative stress and over-      ing blades) and is a well recognised mitigation         Beginning in 2003, GE and Toshiba expect
speed criteria in order to ensure safe and reli-    against instability. In addition, this construc-     to offer the new blades on their steam turbines
able service for the life of the machine.           tion provides additional damping at the loca-        for both combined cycle and fossil power
   The basic premise in the dynamic design          tions of contact between adjacent blades.            plant applications.                          MPS

                                                                                                                February 2003 Modern Power Systems     27

To top