US utilities opt for steam turbine upgrades23 April 1999
In the deregulated environment, utilities have to get maximum benefit from their installed capacity. One method of achieving this is through upgrading existing steam turbine plant.
The trend towards privatisation of utilities, coupled with deregulation of electricity supply markets, has been a global prime mover in the drive to increase profits from electricity generation. Fast payback requirements usually dictate the capital spending programmes of privatised utilities.
Increasing pressures to reduce CO2 and other noxious emissions are being imposed by national governments seeking to satisfy targets agreed at Kyoto. One way to satisfy these pressures is through improved plant efficiency, which reduces the environmental impact of fossil plant, hence allowing less fuel to be used for the same electrical output, with corresponding reductions in gaseous emissions and heat rejection to the environment.
It is in this economic and environmentally conscious climate that many utilities have chosen to upgrade existing steam turbine plant, rather than build new generating capacity. As well as reducing emission levels per kWe, this approach can also maximise revenue by increasing the generating capacity of existing plant, often at a capital cost/additional kWe lower than that required to install new plant.
Steam turbine upgrades on fossil plant can often be justified on the grounds of performance improvement alone. The case for retrofitting is even stronger if efficiency improvements are combined with increased steam passing capability, and if the upgrade offers reduced maintenance requirements. Turbine upgrades are always more economically attractive on plant operating at high load factors.
There were a number of successful turbine retrofits in the UK in the early 1990s. Since then, several major US utilities have started to implement similar turbine upgrades, with many more poised to follow the same route.
Alstom has firm contracts with six US utilities to upgrade more than 30 turbine cylinders. One such utility is the Tennessee Valley Authority (TVA), which has ordered HP turbine upgrades for five units and LP turbine upgrades for four units from Alstom since September 1998. Design contracts are also in place for three further LP turbine upgrades.
HP and IP blading
The main performance loss mechanisms in a typical low root reaction steam turbine stage. The greatest efficiency gains on short height stages have been made by reducing blading profile and secondary losses. Alstom used advanced computation fluid dynamics (CFD) analysis, with model turbine tests, to develop a new and efficient range of blading. This new blading has since been proven by in-service performance testing.
Analysis of the velocity distribution over profile surfaces has allowed profile performance to be improved significantly. Inlet profiles are designed to tolerate potential mismatches between profile and steam inlet angles which may occur in nozzle wakes or at end wall regions.
Complex vortices and loss-producing secondary flows occur at the end wall regions of blade passages. The influence of these on stage efficiency can be reduced through the use of narrower blade chords and longer aerofoils. However, end wall losses are also influenced by the cross-channel pressure gradient. Advanced CFD analysis is used to predict this pressure field and to determine the benefits of refined passage geometry. Using this analysis supported by model turbine tests, Alstom established a number of blade geometries which are effective in controlling secondary flows. Additional benefits are obtained from advanced technology 3-D blades which reduce these effects still further.
The steam leakage flows at the rotating blade tips and rotor hubs of modern low reaction turbines are minimised by careful design using a number of standard features. Spring-backed hub seals are fitted to kinematically constrained diaphragms to permit the use of small radial clearances. Disc panel leakage ports redirect the hub leakage flow away from the main steam path, preventing inefficient flow disturbance, and also reducing the level of rotor axial thrust. The development of integral blade tip shrouding allows the use of an interleaved, labyrinth seal at the rotating blade tips, which also reduces steam leakage.
LP blading. There is a rapid increase in blade heights towards the LP turbine exhaust. This leads to a large variation in static pressure in the fixed/moving blade interspace, particularly in the last two stages, which tends to produce a negative reaction at the base of the rotating blades and increased reaction at the tips. These problems are exacerbated in older turbines, which often have irregular outer boundaries.
On older turbines, long LP blading was designed on a two-dimensional basis, with simple radial equilibrium being used to account for centrifugal effects. Axisymmetric streamline curvature techniques for throughflow calculations provide the capability to analyse flow structure in groups of stages and to recognise the problems associated with radial flows and curved boundaries. Further refinements are possible through the use of three-dimensional time-marching methods. The intensity of the interspace pressure field can be controlled and reduced by two major design options:
The effects of lean and twist are complementary, and both can be combined with smoother boundaries to control the flow structure.
Other performance factors. On plant where reheater attemperating sprays are used to prevent excessive reheat temperatures, improvements in HP turbine efficiency reduce cold reheat steam temperature, thus permitting reheat sprays to be turned down, leading to additional cycle benefits. HP turbine retrofits can also take advantage of increased steam flow capacity from over-specified fossil-fired boilers.
Reliability issues. There are various reliability issues which can influence the decision to retrofit steam turbines in fossil plant. Perhaps the most common problem in the USA is solid particle erosion (SPE), where exfoliated particles (principally magnetite) from boiler tubes are carried over into the turbine to cause erosion damage of HP turbine nozzles and, to a lesser extent, first stage IP blading. SPE can seriously compromise the mechanical integrity of the nozzles and riveted blade tip shrouding within 3-4 years of operation. HP and IP section efficiencies can fall off by up to 5 per cent over this period. The costs of regular inspection, repair or replacement of SPE damaged components can be very high.
There are other persistent reliability problems which strengthen the case for retrofitting fossil-based steam turbines. These include thermal fatigue cracking of rotors or casings, life limitation due to steady state creep damage, rotor instability problems or recurrent blading failures.
Case study – TVA
TVA leads the US generation league with an annual output in excess of 150 TWh, and operates some of the most efficient plant in the USA. TVA has embarked on a programme to optimise plant cycle efficiency of its 500-1300 MWe units. The governing principle of this programme has been to keep boiler rating levels unchanged. In this way, increases in unit MWe output are achieved entirely through improved efficiency without increasing emissions.
TVA's large units have an average age of over 30 years. The original 1960-70s steam path components have significantly lower efficiencies than those available today. A review by TVA of the available technology indicated an opportunity to enhance HP and LP turbine efficiencies by replacing existing steam path components with modern technology.
A crucial factor in TVA's assessment of a turbine upgrade is the ability to sustain high efficiency levels over longer periods so that fuel savings are assured for the long term. This concern arises from the fact that plant supplied in the USA during the 1960-70s has been particularly prone to SPE damage, which leads to rapid and continuing loss of efficiency. In the case of TVA units, this ongoing damage necessitated repair overhauls every 5-6 years.
TVA selected Alstom to carry out a series of HP turbine upgrades for the following units:
These have been followed by orders for upgrading L-1 and L-0 LP diaphragms on Paradise units 1-2 and the L-0 diaphragms on Cumberland units 1&2 (1300 MWe). Design contracts have also been obtained for LP diaphragm upgrades on a further three units.
A factor that influenced TVA's decision to select Alstom HP turbine upgrades was that Alstom had already installed the first of four similar HP turbine upgrades at a power plant in Alabama. The combination of efficiency improvement and the extension of maintenance intervals to 12 years, obtained on a unit which entered service in 1990, led to the confirmation of the order for the remaining units.
This retrofit was particularly interesting, as the original HP turbines had nozzle control down to half arc admission with an opposed flow nozzle box configuration. The double flow nozzle box had large numbers of narrow nozzles, and the early stages of the HP cylinder suffered severe SPE damage after relatively short periods following overhauls. The utility was keen to improve the baseline performance, as well as avoid the regular inspections and repairs needed to ensure plant integrity and regain lost efficiency. The change to a single, forward-flow, first-stage with full arc admission (using fewer, wider nozzles), together with integrally shrouded rotating blades, will significantly reduce the SPE problem. The Alstom retrofit solution has an optimised steam path with four more stages than the original configuration using improved aspect ratio aerofoils of advanced design.
The Paradise units 1 and 2 and Widows Creek unit 7 have similar HP turbine configurations to those developed for application on the Alabama units, and similar solutions were proposed. The larger units at Paradise 3 and Bull Run 1 already operate with full arc admission and have a single flow configuration but, once again, with large numbers of relatively narrow first stage nozzles. These larger units have also suffered badly from rapid efficiency degradation due to SPE. The retrofit solutions take advantage of one additional stage on Paradise unit 3 and two on Bull Run unit 1 applied in a similar configuration to the other units.
The Alstom HP turbine upgrades selected by TVA have cylinder percentage efficiencies (governor valve bowl to HP exhaust) approaching the mid 90s, with an expected degradation rate of approximately 1-1.5 per cent between 10 year overhauls. This is much lower than the 0.5-1 per cent per year degradation currently experienced on some TVA units. Alstom achieves these sustained high efficiencies through the application of advanced fixed and rotating blade technology in a single flow configuration which avoids SPE damage. These HP turbine upgrades will therefore provide between 8-11 per cent improvements in HP section efficiencies which correspond to 2-2.5 per cent improvements in unit heat rate and more than 30 MWe additional output on the larger units.
The LP turbines of Paradise 1 & 2 and Cumberland 1 & 2 are also being upgraded by applications of modern L-0 and, in the case of Paradise, L-1 fixed blade diaphragms. These address deficiencies in the original design to provide cost effective efficiency improvements. These LP upgrades have been developed using 3-D CFD analysis benchmarked against aerodynamic traverse data obtained from the LP exhaust hood. LP diaphragm upgrades of this type, coupled with improvements to exhaust diffusers, are expected to provide heat rate improvements of up to 1 per cent.
Having selected the Alstom turbine upgrades, TVA also intends to perform a turbine cycle optimisation study. This is an iterative process that can take several attempts to converge on the final optimisation of cycle parameters. The optimisation process will identify component limitations, such as heater capacity limitations or mismatch in BFPT steam conditions.
Retrofitting existing steam turbine plant using state-of-the-art blading technology increases the efficiency and profitability of existing generating plant. Additional generating capacity can often be obtained for less investment cost than new plant, and with no increase in operating costs. The cost of long term reliability solutions can be offset by increased revenue from performance improvements.