Gas fired combined cycle power plants mothballed after the initial dash for gas are tending to be recommissioned again now that the increase in open market electricity prices appears to be a dependable trend. New combined cycle plants entering the market will need improved flexibility and durability during load cycling operations as well as further increases in thermal efficiency to be financially bankable. Recent important advances in HRSG design may be the key to viability on this front.


Relatively few HRSG manufacturers are still competing in the power generating plant market. Those that remain are now coming up with advanced designs that will go a long way to solving some of the most urgent problems in the combined cycle power plant market.

In January this year a number of boiler manufacturers acquired the Siemens horizontal flow, once through vertical tube Benson technology which was so successfully demonstrated in the PowerGen advanced utility combined cycle power plant test bed at its Cottam power station site (Figure 1) in the UK.

The project was described in some detail in the September 1999 issue of Modern Power Systems, pp 40 to 43, and the radically new approach to HRSG design was described in the July 2000 issue of MPS, pp33 to 35, but the 500 MWe/60 per cent efficiency upgrade mentioned never materialised.

Using a horizontal flow, vertical tube Benson HRSG developed by Siemens in collaboration with Babcock Borsig Power (Figure 2), the original aim was to increase the efficiency of the uprated V94.3A gas turbine based system to over 60 % by further development of the main components, implementation of improvements in the water/steam cycle and further optimisation of the plant concept.

The gas turbine would have been replaced by the new S9 GT, a new alternator would have been fitted and an upgraded steam turbine IP/LP section installed. The further development would have increased output by some 25 % to 500 MW from the initial 400 MW, 58 % efficiency system.

Cottam’s contribution

Radical improvements in high performance combined cycle power plant steam generators were demonstrated in the Cottam Development Centre (Figure 1) combined cycle power plant installed by Siemens Power Generation in co-operation with PowerGen in the UK. Having proved with such remarkable success the advanced 400 MW combined cycle concept’s performance, why did Siemens pull out of the highly profitable joint venture late in 2003?

Well, apparently there was a bias option for PowerGen’s take over in the original agreement, but the Siemens testing agreement was completed by 2002. At the same time Siemens PG and Westinghouse in the USA were going through the merger process and prevailing market conditions made the upgrade untenable. A major refurbishment to restore the plant to “as new” condition so that it could continue to operate economically in the electricity market was carried out from October 2001 to February 2002.

Although the 500 MW system was never developed at Cottam, PowerGen, now owned by German utility E.On, has acquired a very high efficiency and remarkably profitable power generation plant which has shown great reliability and load cycling performance.

PowerGen reports that they spent some £52.5 million to acquire SPV’s 50 % of the power station in Nottinghamshire, which is little more than their initial investment in the joint venture. They estimate that the cost of acquiring the station in January 2004 equates to around £235 per kW. The plant is operated with only three people on shift and a total staff of 30.

The gas turbine was first synchronised in March 1999 and the entire plant was first on-line in summer 1999. Predicted performance was rapidly shown to be in very good agreement with simulation results and the plant has been operating at high load factors in the commercial electricity market since start-up. Output and availability over the first four years of operation have been remarkable considering the load cycling duty and gas arbitraging implications of the duty schedules. Through 2003 availability (RCA) was 98.83 % with planned outage (BD) 1.1 %.

When MPS staff visited the station in February 2004 the plant had been running at 380 MWe output for the previous 3 to 4 days. In the future it may well be more profitable for the operators, who are also gas suppliers, to arbitrage the natural gas than maintain the present high levels of electricity output. This will be driven mainly by the price of natural gas, but in terms of electricity generating plant economics it will only add to the residual economic operating life of the installation. At the same time, it is interesting to note that for the old 2000 MWe coal fired plant alongside the Cottam Development Centre a new FGD system is to be fitted reflecting increased competitivity for the more traditional fuels. Competitive edge is a moving target and clean coal is coming back into contention as gas prices continue to rise and governments begin to reconsider the nuclear option.

Following the highly successful technology demonstration at Cottam, Siemens have licensed the once-through Benson HRSG technology to Alstom in the USA to use in conjunction with their OCC (optimised for cycling and constructability) concept.

All change

Priorities for the development of the next stage of advanced gas turbine development after the V94.3A2, apparently labelled “S9” in some quarters, seem to have changed radically during the test phase of the Cottam project under two powerful influences – the changes in the electricity market and the merger of Siemens Power Generation with Westinghouse Electric in the USA.

The benefits of combining the best of the Westinghouse F, G, and H technologies, including the US DOE AT programme developments, with Siemens’ own advancements would have seemed obvious. At the same time, operating experience from the extensive fleet of Siemens combined cycle plants in the highly competitive UK power trading market was bound to lead further technological development in new directions. The emphasis has moved firmly towards improvement of HRSG reliability and performance, and the Cottam project appears to have shown the way ahead.

Gas turbine technology development requirements, on the other hand, have at least temporarily retrenched from chasing ever larger unit outputs and elusive thermal efficiency targets which require vast levels of development investment. The emphasis is now on improving dependable cycling load performance with minimum impact of frequent start-up/shut-down cycles on lifetime O&M costs, and rapid start-up capability to exploit peaking and fuel arbitrage opportunities.

Also, while the combined cycle markets in Europe have stalled, new plants are still being installed in the Americas and government funds are still being invested in advanced gas turbine development. For the time being it seems sensible to leave beating the 60 % efficiency and 500 MWe per unit psychological barriers to the US side of the Atlantic.

Breaking 60%

Exceeding the 60% combined-cycle efficiency mark is largely dependent on further development of the water/steam cycle and its key components, the heat-recovery steam generator and steam turbine, in the direction of process improvement. Figure 3 shows the areas of improvement that were expected to contribute to the 60 % efficiency target at the Cottam combined cycle power plant.

The efficiency improvement due to additional pressure stages in the heat-recovery steam generator decreases with increasing gas turbine exhaust-gas temperature, because the effectiveness of the IP and LP steam stages becomes less and less in comparison with the high-pressure stage. At very high exhaust gas temperature and limited main steam temperature, the IP and LP steam flow rates even become negligibly small, with the result that multiple-pressure processes approach the effectiveness of a single-pressure process. However, as long as it is possible for main steam temperature to follow the gas turbine exhaust gas temperature with a difference of less than 60 K, the triple-pressure reheat steam cycle is the preferred solution. A main steam pressure of 160 bar appears to yield the largest possible potential for further increases in efficiency.

For evaporator design, at the main steam pressures of 80 to 130 bar common to date either natural circulation or once-through systems are primarily implemented. With the increase in main steam pressure to 160 bar and above, the transition to a once-through evaporator becomes advantageous:

• Elimination of the thick-walled drum results in a significant improvement in the dynamic characteristics of the entire plant. This advantage is especially important for heat-recovery steam generators that have no bypass on the exhaust-gas side. The parallel startup of the gas turbine and heat-recovery steam generator thus required necessitates a thermoelastic design of the pressure parts with low wall thicknesses.

• Once-through evaporators can be made cost-effectively. For today’s typical steam pressures the weights of the high pressure components can be up to 2 % lower than for a natural circulation evaporator with its drum. At the higher pressures of the new generation of plants the weights of the pressure parts may be 3 to 5 % lower.

Once-through system design can also be used for the intermediate-pressure section. This remains a cost decision in individual cases.

Getting it together

As well as greatly improving load cycling performance, once-through HRSG design can facilitate significantly reduced start up times for combined cycle plants which can save subtantial amounts of fuel consumption.

Various papers on integrated combined cycle plant design for fast start capability at base load efficiency have been presented by people like Raymond Baumgartner in Siemens Westinghouse in Orlando, Thomas Mastronade at Alstom in Windsor, Con, and Matthias Fränkle of Siemens Power Generation in Germany. Some elements of the innovation needed have already been well demonstrated at the Kings Lynn, UK, single-shaft Siemens 1S.V94.3 combined cycle unit, not a million miles away from Cottam.

Once the gas turbine generator reaches rated speed it is synchronised and connected to the grid and loaded to minimum output. With further increase in exhaust temperature the gas turbine ramp rate become limited by the permissable temperature and pressure transients in the HRSG. These transients have to be constrained by the thermal stresses and fatigue damage limits in such areas as the HP drum, HP superheater headers and tube penetrations.

Some reductions in start-up time can be achieved by using steam sparging into the evaporator and closing the stack damper to keep the HRSG warm prior to start-up, using two-stage attemperation to control the temperature of the HP superheater outlet header, and minimising the inner diameters and wall thicknesses of the superheater and reheater outlet headers in order to maximise the allowable temperature gradients before start-up. More substantial reductions in start-up times are achieved by eliminating the HP drum completely. All of these were achieved in the once-through Benson boiler at Cottam (Figures 4, 5).

Alstom, who claim to have 30 once-through HRSG’s in service, plan to combine the once-through vertical tube Benson technology licensed from Siemens with their already commercialised OCC (Optimised for Cycling and Constructability) design concept which incorporates structural design features aimed at reducing thermal stresses during cyclic operation as well as simplifying construction. These include:

• Single rows of tubes between headers.

• Finned tubes with no bends.

• Flexible connections between pressure part sections.

• Elimination of division walls in headers.

• High creep stress materials in high temperature areas.

• Small diameter headers which can reduce thermal stress by 60 %.

• Multiple header connections to promote uniform flow and metal temperatures in superheater and reheater sections.

• Full penetration tube-to-header weld joint.

• Seamless tubes.

• Enhanced drain arrangement to prevent condensate flooding of superheat and reheat sections during gas turbine purge.

Figure 6 shows the configuration of the lower end of of a “single row” harp in an OCC superheater assembly (fins not depicted) which shows greatly improved stress analysis and temperature distribution over the earlier designs shown in Figure 7.

The OCC system will be used in the Magnolia Power 320 MWe combined cycle plant currently under construction on a site in the Burbank area of Los Angeles. The plant, which is based on a single GE 7FA gas turbine with SCR, should be on line in time to meet Burbank Water Power’s summer 2005 power demand.

Alstom is supplying one OCC HRSG for a GE 7FA gas turbine, a burner management system, a selective catalyst reduction system, an oxidation catalyst system and a boiler recirculation pump. The HRSG is to be built at Alstom’s manufacturing centre in Portugal. To reduce erection costs, risk, and schedule, it is being supplied in “C-Frame” modules that integrate the pressure parts, support steel, casing and insulation.

Alstom’s OCC design can be packaged and delivered in three alternative modularity options:

• Harp bundles (30 ton) – where transportation is restricted, large cranes are scarce and site labour is inexpensive.

• Pressure part modules (120 ton) – where transportation is less restricted and large cranes are available.

• C-Frame modules (220 ton – pressure parts integrated with support steel and casing) – where transportation is unrestricted, large cranes are available and site labour is expensive.

At the time of writing we have not heard of a first application of the new “OTCC” combined cycle HRSG’s, but we have heard rumours of the system being tendered for two different projects in Italy involving large GE turbine systems.

Size matters

The sheer size of a conventional HRSG for a 500 MWe combined cycle power plant moves the steam cycle technology into new realms of complex technology which might lead to increased outage time.

Looking at the potential markets for a 60 %-plus efficient combined cycle system with improved load cycling performance, the North American 60 Hz markets appear to have a greater potential for sufficient return on investment to justify the formidable research and development costs than Germany or any of the other EU 50 Hz states. From this point of view it no doubt seems more logical to concentrate on the government AT programme funded “G” technology rather than the Siemens V94 series for the time being.

In Germany, electric power utilities, plant manufacturers and consumers alike are clamouring with complaints that the vast programmes of power plant investment required to meet the requirements of the Kyoto protocol, and the closure of nuclear power plants, cannot possibly be carried out because of the continuing prevarication and political manoeuvring in government, resulting in the absence of a coherent energy policy. In the mean time, costs of natural gas fuel are continuing to rise rapidly and generators are beginning to look increasingly towards wind energy, renewables, and even renewed interest in nuclear energy and more coal power.

On both sides of the Atlantic, there is evident commitment to combining the best elements of Westinghouse and Siemens technology in the ultimate combined cycle products.

Once-through evaporators

For vertical HRSGs, with exhaust gas flows from bottom to top, the countercurrent configuration yields the smallest heat exchange surfaces and the lowest pressure drop. However, stability problems arise in once-through evaporators with this configuration which can only be eliminated with the installation of flow restrictors.

In contrast, the parallel-flow configuration exhibits flow stability but results in the largest heat exchange surfaces and the highest pressure drop of all configurations. A combination of the two configurations yields intermediate values for the heat exchange surface and for pressure drop, while also providing the requisite flow stability and retaining a variable evaporation endpoint.

A simple possibility would be a series configuration of evaporator tubes with alternating downward and upward flow. At low mass flux, the slip between water and steam in the tubes with upward flow forms a nearly standing water column with rising steam bubbles, while the tubes with downward flow contain a column of steam with falling water droplets. This also leads to high pressure drops in tubes with low mass flow, as the hydrostatic pressure differences in all of the series connected riser tubes are added. The result is an unstable distribution of flow in parallel tubes.

Another alternative would be a horizontal configuration of the evaporator tubes with vertical inlet and outlet headers. The vertical headers and the associated hydrostatic pressure differences result in differing inlet pressures of the individual horizontal tube banks. This results in significant differences in mass flow and outlet temperature in the parallel tube system.

This problem could be alleviated by reverting to a more complex configuration, subdividing the headers in a large number of individual sections connected to the corresponding supply lines with varying degrees of flow restriction. As well as being a complex solution this approach incurs some additional pressure drop and results in reduced plant cost-effectiveness.

These disadvantages were avoided by utilising the new configuration of heat exchange surfaces first implemented in the Cottam combined-cycle power plant illustrated in Figure 2 and the steam water circuit in Figure 3.

Flow through the parallel, vertical evaporator tubes in this horizontal heat-recovery steam generator is only upward, as in a natural-circulation boiler with horizontal gas pass. In contrast to standard once-through evaporators, revised fluid-dynamics design results in a higher mass flow rate in tubes with increased heat absorption. This leads to equilibration of enthalpy at the evaporator outlet.