Dale Ernette, Siemens KWU, Germany

With advances in combustion turbine technology and the movement of combined cycle plants to larger ratings, steam turbine technology has evolved. In the past ten years steam turbines utilized in combined cycle plants have changed from relatively small steam turbines in two pressure non-reheat cycles, to large multiple pressure reheat units with outputs in the 400 MW range.

The thermal design of the bottoming cycle is different from a conventional steam plant because there are no feedwater heaters in the cycle and the steam is typically inducted at multiple points in the cycle in order to optimize the performance of the heat recovery steam generator (HRSG). As a result, mass flow is typically 30 per cent greater at the low pressure (LP) exhaust than at the turbine throttle in a combined cycle steam plant. In a conventional steam plant the LP exhaust flow is 30 per cent less than turbine throttle flow. Combined cycle plants also differ from conventional steam plants by desiring minimum plant heights, short delivery times, and the need for significant amounts of duct firing during peak load demands.

Although the basic design concept for steam turbines remains the same whether it is for a conventional fossil fired steam plant or for a combined cycle steam plant, the use and application of steam turbine technology can vary a great deal. As a result, specific steam turbine designs have been created for use in a combined cycle plant.

Blading optimization

A short time ago, it was considered that most combined cycle plants would be standard type plants typically based upon the one-on-one (1×1) or two-on-one (2 gas turbines and 1 steam turbine) plant configuration. Due to the increasing non-regulated power industry environment, variations in the 1×1, 2×1 and even 3×1 plants are now being seen throughout the industry. The majority of the differences can be attributed to the level of duct firing or process steam requirements. These differences may require a customized high pressure (HP) and intermediate pressure (IP) blade path in order to efficiently maximize power output. However, this customization typically conflicts with short delivery times, which are desired for combined cycle steam turbines.

Titanium has a high strength/weight ratio and superior corrosion resistance in a low pressure steam turbine environment
Click here to enlarge image

The difficulty of customizing a blade path is made more complex by the use of three-dimensional blades. These blades are characterized by airfoil sections which are curved in the circumferential direction to reduce secondary loss, thereby increasing blade efficiency by up to two per cent. This complex blade geometry and the economical manufacture of the blades is made possible by the advanced methods of three-dimensional frictional flow analysis in conjunction with specially developed five-axis milling machines. However, the difficulty of designing blading to three-dimensional flow characteristics has greatly increase the complexity of the thermal and mechanical flow computer programmes.

To quickly consider and optimize all of the thermal and mechanical parameters used in three dimensional blade design, Siemens has designed an advanced information technology (IT) tool with which information for blade design, blade drawing and blade manufacture are moved seamlessly throughout design and manufacturing environments. This sophisticated computer programme allows the engineer to design the blade path entirely in a three-dimensional environment.

Since the goal of this programme is to optimize the blade path to the largest degree possible, a number of parameters are optimized in the thermal design programmes. The optimization algorithm used at Siemens determines the maximum efficiency by varying between 40 to 80 free parameters while taking into consideration between 100 and 300 boundary parameters. Some of the parameters include profile shape, blade height, aerodynamic blade loading, and inner and outer flow contours. These thermal parameters are then limited by the mechanical constraints to produces the final blade path.

This design environment has allowed a decrease of up to 80 per cent in the total design time for blading while at the same time increasing the quality by standardization of the design process. Efficiency is also maximized due to designing in a three-dimensional environment.

LP blading

Due to the thermal design of the combined cycle, the LP exhaust flow is much larger than a corresponding megawatt level in a conventional steam plant. In order to efficiently utilize the steam as it leaves the LP turbine, a larger LP exhaust annulus area is required for a steam turbine in a combined cycle plant than in a conventional steam plant of the same rating. This increased LP exhaust flow could prevent the use of the cost effective axial exhaust steam turbine or could require use of a four-flow LP turbine instead of a two-flow LP turbine. The effect of these possible changes will not only increase the cost of the steam turbine, but also increase the cost of the total power plant since there will be additional effects on the size and height of the plant foundation, and the design of the cooling water system.

Therefore, whether extending the range of an axial exhaust steam turbine or minimizing the number of LP exhaust cylinders necessary for optimal cycle performance, long last row blades are necessary to minimize the cost of the combined cycle steam plant. Siemens has developed LP last row blade designs using both steel and titanium in order to reduce the cost of the steam turbine along with the overall power plant. Of course, the longer the last row blade, the increased difficulty in designing and manufacturing this blade.

Throughout the development of steam turbines, the length of last row LP blades has continually increased over time. For 50 Hz steel blades (full-speed) the latest development has been a steel blade with a blade height of 1150 mm. Modern designs of the LP blading optimize the flow distribution over the length of the last row rotating blade. In order to accomplish this, the last stage stationary blades are of a twisted design and curved in a circumferential direction. This curvature induces an additional radial force and thus counteracts separation in the blade root area. Blade sections with transonic discharge velocities in the range of up to Mach 1.3 are designed with straight suction sides to eliminate overexpansion. Sections with higher Mach numbers are provided with converging-diverging throat passages.

High performance

The Siemens 1067 mm titanium blade (60 Hz full-speed) has a greater annulus area than steel blades. This is possible by the lighter density of the titanium. Due to the reduced damping properties of titanium, it is necessary to provide for continuous coupling of the titanium blades. This is provided by an interlocked blade design which produces a stiff, highly damped structure.

Titanium has a high strength/weight ratio and superior corrosion resistance in a low-pressure steam turbine environment. However, titanium has lower material damping than steel, thus necessitating the change of construction from free-standing to interlocked construction for increased mechanical damping.

During turbine operation the blade elastically untwists due to centrifugal forces, which causes adjacent shrouds and snubbers to make contact. At rated speed the contact forces between adjacent shrouds and snubbers allow the blading to behave as a continuously coupled structure. Continuous-coupling increases the stiffness and natural frequencies of the blade-disc assembly. The contact surfaces provide additional damping which reduces vibratory stresses.

The finished blade is fully machined from an enveloped forging of Ti-6Al-4V with bi-modal microstructure. This bi-modal microstructure has reduced variability in elastic modulus, and higher fatigue strength compared to the mill annealed condition. The yield strength of Ti-6Al-4V is the same as 17-4PH, and the weight of titanium is only 57 per cent of steel.

The blade was designed using advanced full 3-D fluid codes to provide high thermal performance at the supersonic velocities at which it operates. The upper half of the blade is designed with a converging-diverging cross section to provide fully controlled expansion of the steam.

Economic solutions

Typically for a 1×1 combined cycle plant (one gas and one steam turbine) the steam turbine is a single-flow, high-pressure turbine in tandem with a combined IP and LP turbine. The axial exhaust configuration provides the most economical solution for a combined cycle by virtue of its compact design and lower elevation requirements. Therefore, by extending the range of the axial exhaust configuration the total steam turbine cost decreases. This includes the capital cost of the steam turbine and the construction of the turbine building, this can be reduce by approximately 40 per cent over using the double flow LP turbine cylinder arrangement of similar rating.

In order to extend the range where an axial exhaust steam turbine could be applied within a combined cycle, a welded rotor was developed. Previous designs of IP/LP rotors typically used forged rotors which utilized 2.25 per cent Cr, then provide different heat treats for the HP and LP sections to achieve different rotor properties. This design worked well for inlet temperatures of 550°C, however with designs above 550°C the design became limiting.

Therefore in order to accommodate inlet temperatures above 550°C, short startup times and long last row LP blades it was decided to weld together the CrMoV rotor material used in HP and IP rotors to the 3.5NiCrMoV rotor material traditionally used for the LP turbine. These two materials are joined by a narrow-gap tungsten inert gas (TIG) welding process providing a weld with material properties equal to or better than the rotor base material.

As gas turbines exhaust temperatures increase, a rotor material with a higher chromium content can be applied to the hot to increase the application of the IP/LP rotor to temperatures above 600°C.

Duct firing

In a deregulated power environment, each power plant can have its own economic basis in order to optimize its profit structure. Traditionally, combined cycle power plants have included some measures for increasing power output during peak demand. However, in today’s deregulated environment the request for additional duct firing can increase to 80 per cent additional power as in the unfired case. In order to accommodate large amounts of duct firing, Siemens has designed the combined cycle steam turbines with the capability to pass additional flow.

For an unfired steam turbine in a combined cycle power plant, studies have shown that the economic optimum steam turbine inlet and reheat conditions are 125 bar/565°C/565°C. In order to maximize efficiency, the combined cycle is operated in a full-arc admission mode. Therefore, an increase in flow to the stream turbine will increase the inlet pressure in proportional levels to the throttle flow.

The designs used at Siemens permits continuous operation up to 170 bar/565°C/565°C. This difference between the thermal design point and the maximum mechanical design point allows an increase in duct firing capability of over 35 per cent from the rated design point. Further increases in the duct firing capability can be accommodated by lowering the turbine inlet pressure level at the design point, thereby allowing further increases up to the maximum mechanical design point.

It is important to note that designing for modest levels of duct firing does not cause an appreciable loss of efficiency in the unfired case. However designing for high levels of duct firing (greater than 40 per cent) does cause loss of efficiency in the unfired case and therefore should be considered carefully in an economic analysis.

Future design direction

Low pressure turbine steel blades using technology which is available today, the 1150 mm (50 Hz full speed) blade is already at a practical limit and no increase in this size of full-speed steel blade is foreseen in the near future. However, the blade technology available today using titanium does permit further increases in LP blade sizes beyond the currently developed titanium blades. At present, Siemens is developing the next generation of titanium blade, which will increase the LP exhaust annulus area by more than ten per cent from the present design. Further increases in annulus beyond this will likely be limited by the thermodynamic design at the tip of the inlet of the last rotating blade.

Combustion turbine firing temperatures have increased so that the 565°C steam turbine inlet temperatures are now standard, and 600°C inlet temperatures are now possible given the combustion turbine exhaust temperatures. At this point in time, given the cost of fuel and the increase in the cost of steam turbines and power plants, the efficiency improvement by going above 565°C is not economically justified. Although steam turbine materials exist today to accommodate temperatures above 565°C, research into alternate materials that have a more attractive cost/benefit ratio continues. With the inevitable rise in fuel costs coupled with development of more cost effective materials, the increase in main and reheat steam temperatures could become economically justifiable in the near future.

The next generation of turbine blading for the HP, IP and LP front stages is already under development at Siemens building upon the current generation of three-dimensional blading. This new blading will still incorporate the same features as the existing three dimensional blading, most notably the reduced secondary losses at the blade root and tip due to the blade curvature, with the added advantage of being able to define the degree of reaction individually for each turbine stage. In essence, this will be a departure from the classical principles of impulse and reaction turbine designs through the use of individual degrees of reaction for the different turbine stages. In this regard, the designer can further increase the optimization of the blade path thereby increasing overall blade path efficiency.