Maintaining continuity of electricity production will become ever more challenging as gas supplies become more complex. Peter Marx, Martin Liebau, Belén Gasser-Pagani and Mark Stevens describe the capabilities of reheat turbines that can address this problem.

Among the fossil fuels that power stations burn, natural gas is expected to play an ever greater role in the future. Combined-cycle plants will increasingly employ this resource because it exists in substantial reserves and is a relatively clean fuel. Such facilities are also efficient and flexibile enough to compensate for the intermittency of renewables. But these plants will need to show another kind of flexibility too, one that takes into account the likelihood of variations in the composition of gaseous fuel.

Fuel supply networks are becoming more complex because of a trend towards the use of liquefied natural gas (LNG) – the exact make up and properties of which depend on its source – and synthetic fuels such as hydrogen. So how can a gas-fired plant confront the challenge of continuing to perform within its operational range if the components of the fuel change or their relative quantities vary? How can it, at the same time, adhere to emissions limits and ensure its maintenance intervals do not change?

Operational experience has shown how our GT26/GT24 sequential combustion gas turbines perform when the composition of their fuel varies or an altogether different fuel is used. Measurements on these units have also shown what their emissions profiles are over their operational ranges. The turbines’ capabilities allow the management of hydrogen co-firing and variation of the Wobbe index (WI) of fuel gases. Field experience has also demonstrated the units’ proficiency in using back-up fuel oil and switching during high loads between gas and oil, or vice versa.

Evolving global fuel supplies

There is a growing trend towards the use of LNG because it can be procured from a number of sources without the aid of long pipelines. Furthermore synthetic gases such as hydrogen may be used as storage media for over-capacity caused by renewables, in which case they are fed into natural gas supply lines. And in the case of a shortage in the supply of fuel, a gas turbine must be able to switch over to oil as a back-up without interrupting power supply. Biodiesel is becoming popular as such a reserve.

The various fuels may be mixed at a central mixing point but they may also be mixed at the gas turbine-based power station itself. In the latter, there may be rapid changes in fuel properties, which will require the fuel supply and control system to adapt quickly. Figure 1 shows the various fuel sources and the network they form in their paths towards a consumer such as a gas turbine power plant.

Figure 1
Figure 1: Various fuel sources available to a gas-fired power plant

Several typical fuel parameters affect the layout of fuel supply systems, the design of the burners and the operational concept of the gas turbine. These are: the type of fuel, whether liquid or gaseous; the reactivity of fuels such as hydrogen and high hydrocarbons such as C2+; lower heating value (LHV) according to gas composition; and the WI according to gas composition and inerts such as N2 and CO2.

Our reheat gas turbines meet the demands on their flexibility that such fuels make. This is partly due to the design of the EV and SEV burners, which are used within the first and reheat combustors, respectively, and are suited to changes in the volume of fuel flow without affecting emissions and stability. Another advantage of the GT26/GT24 gas turbine is its reheat system, which allows additional flexibility in addressing changes in fuel reactivity. Here, the fuel flow to each of the two combustors is modulated while the power output is maintained. The EV and SEV combustors even allow the burning of a different type of fuel. Table 1 shows the turbines’ capabilities while operating flexibly in the fleet.

Table 1

Figure 2 shows field experience in the burning of gaseous fuels as the WI and C2+ range changes. Gases with a WI range of 21 MJ/m3 have been used successfully in the field with the GT26/GT24 standard combustor hardware. If the WI has to extend by another 4 MJ/m3 below the lower limit of the range, the EV burner gas injection can be modified and the SEV fuel injection system can be used as is. Both of these findings have been validated in single-burner high-pressure tests.

Figure 2
Figure 2: Field experience for the GT26/GT24 gas turbine when burning gaseous fuels of various values of Wobbe index and concentrations of C2+ hydrocarbons

Combustor operation

In the fleet of reheat gas turbines that Alstom has sold, there are more than 50 GT24 and 80 GT26 units. In total, these units have accumulated more than 5 million fired hours and 83,000 starts over 16 years. The first units have reached 100,000 operating hours and 2000 starts. Several of the 49 units of the 2006 version of the GT26 that are in operation have come to their hot gas path inspection interval of 28,000 equivalent operating hours. They have shown high reliability that meets the continuously increasing demands on flexibility. The units that have reached this interval have been found to be in excellent condition and it was possible for the combustor components to be partially re-used until the next interval. A service factor of greater than 70 per cent among customers demonstrates that these units can be operated with high efficiency and to commercial benefit.

In the upgrade to the GT26 (2006), the EV and SEV burners and injectors extended the turbine’s operational flexibility and allowed higher firing temperatures at constant emissions. The EV combustor is of the annular type and has a single row of EV burners. The burners operate in a staged premixed mode in fuel gas operation, which enables low emissions over a wide operating range and stability. Figure 3 shows the EV burner flame at high loads, in the lean premix mode and at start-up to idle, in the the rich premix mode.

Figure 3
Figure 3: EV staged burner concept

At high flame temperatures the two fuel stages are optimised to give a uniform fuel distribution at the burner exit. This results in low NOx emissions at high flame temperatures. At low loads the staging ratio is adjusted to bias the fuel towards the central part of the flow, which ensures stability at low flame temperatures. Since both fuel stages are always in operation, a continuous operation is feasible without switchover points.

The SEV combustor is also of the annular type and comprises a single row of SEV burners. Figure 4 shows how each burner contains a single fuel injection lance with a single fuel gas stage and, in the dual-fuel case, a single fuel oil stage with its associated single water stage. Rapid premixing between fuel and oxidant – which is the exhaust gas from the high-pressure turbine – is achieved by injecting the fuel into vortices initiated by vortex generators that are integrated in the burner body. This combustor does not have an ignition system as the inlet temperatures allow auto-ignition. Crucial to the operation is that the auto-ignition delay time lies within a certain window.

Figure 4
Figure 4: SEV burner concept

The sequential combustion system enables the control of the SEV combustor inlet temperature by altering the fuel flow in the EV combustor. In this way the flame position in the SEV combustor can be controlled and minimises the impact on emissions with changing reactivity, for example those caused by a change in the C2+ hydrocarbon content or co-firing of hydrogen. As the SEV combustor temperature and therefore the inlet temperature to the low pressure turbine stays constant, the gas turbine power remains almost constant.

The GT26 (2006) units with the improved EV/SEV combustors have in operation produced emissions of NOx below 15 vppm (15% O2) over the entire operating range between the low load point at about 10 per cent gas turbine load and baseload at 100 per cent gas turbine load.

Hydrogen co-firing

An overcapacity of electricity at times of low demand is becoming more likely as renewables are providing an increasing share of power generation. These over-capacities can be used to power electrolysis to produce hydrogen (H2), which can either be added to the natural gas pipe system or stored in tanks for later combustion.

Hydrogen electrolysis has an efficiency of about 70 per cent, so it makes sense to burn it in a combined-cycle power plant with an efficiency of up to 60 per cent. We have investigated hydrogen co-firing with the advanced GT26/GT24 burner system.

Hydrogen is a highly reactive gas that influences the burning velocity of the flame and the ignition delay time. Tests on single EV and SEV burners have shown the combustion behaviour of H2 cofiring with mixtures of 5-18 per cent to be comparable to that of the highly-reactive C2+ gases. Tests using maximum H2 mixtures of up to 45 per cent in the case of the EV burner and 30 per cent in the case of the SEV type have also confirmed the robustness of the burner hardware and have shown its potential.

The tests also confirmed the feasibility of co-firing mixtures of C2+ gases and H2 of up to 18 per cent using the existing and proven combustor operation concept for C2+ gases. For co-firing with up to 10 per cent hydrogen, the systems for gas supply and control need only minor modifications.

With mixtures of C2+ gases and H2 the NOx emissions in the case of the SEV combustor may increase. This can be partially compensated for by reducing the SEV inlet temperature. Figure 5 shows the test results of the existing SEV burner versus the next-generation burner, which enables higher quantities of C2+/H2 while keeping nearly constant the emissions at the same firing temperature. Hydrogen at a content of higher than 10 per cent is feasible. It requires an advanced SEV burner for reduced NOx emissions and further changes in the gas supply and control system.

Figure 5
Figure 5: Test results for a high-pressure single SEV burner with C2+/H2 cofiring

Wobbe index variation

The WI indicates the interchangeability of fuel gases with respect to the thermal load of the burners. It correlates the heat input, gas supply pressure and pressure loss characteristics of a burner fuel gas system, and also reflects the gas jet momentum flux. The WI correlates mainly with the inert content but it is also temperature dependent. To compare gas compositions the WI is often calculated at ISO-conditions, termed WIiso, but for gas turbine applications it is more useful to calculate it at the fuel gas temperature, termed WInet.

The standard combustor hardware for the GT26/GT24 has run successfully in the field with gases with a Wobbe index changing by 21 MJ/m3, as shown in Figure 2.

With gas turbines an important challenge related to the WI is how to avoid an interruption in power output or modifications to the hardware as the fuel gas composition changes during operation. As fuel gas sources diversify, future power plants will find this concern to be increasingly important.

Figure 6 demonstrates how the GT26 has operated with a flexibility of up to ±15 per cent with variations in WI without changes to the hardware and while maintaining a single concept of operation. The robust combustor hardware means combustion stability and low emissions are possible as the supplied fuel gas changes.

Figure 6
Figure 6: WI flexibility of the GT26 in the field

The increasing complexity of gas supply networks means there is a growing need for gas turbines to cope with big changes in the WI of their fuels during operation. We have followed demand from the market by further optimising the concept of operation of our gas turbines so that these units can run continuously while the WI changes by 21 MJ/m3, or ±25 per cent. This is achieved by measuring the fuel gas composition and automatically adapting the control parameters for the actual measured gas composition.

A standard gas chromatograph or an alternative fast device measures the composition of the fuel gas. The control parameters are changed automatically in response. This advanced concept keeps emissions stable and allows the reliable operation of the turbine if the fuel gas or WI changes rapidly. In addition this concept of operation enables efficiency to be maximised for every type of supplied gas as it varies.

Several features ensure the reliable operation of the gas turbine as the fuel type changes over a wide range.

First, control of the fuel gas temperature depends on the measured gas composition. Pre-heating of the fuel is, of course, desirable, because it significantly enhances the efficiency of a combined-cycle power plant. However, it also reduces WInet. To avoid WI values below the lower end of the proven range for the combustor hardware, a maximum temperature can be continuously calculated, which varies for different measured gas compositions. The GT26 can maintain the fuel gas within the WI range by regulating the gas temperature during operation only when high inert content brings the WI close to its lower limit.

Second is the optimised calculation and control of heat input. Open loop mass flow control is adapted to consider the actual composition of the supplied gas and the heating value. This ensures reliability and stable combustion, especially during transient modes of operation.

Third, the fuel split between stages one and two of the EV combustor can be adapted to maintain the stability of the combustion as the WI changes.

GT26/GT24 gas turbines perform with the WI ranging widely thanks to the robust burner hardware and their flexible operation (Figure 7). Fluctuations in the index of up to ±15 per cent have been proven without the need for modifications. Even ±25 per cent is possible with only minor changes to the operating concept and without any changes to the combustor hardware. This concept ensures reliable operation and stable emissions as the WI varies, without interruption of the output.

Figure 7
Figure 7: Extended WI flexibility with improved concept of operation

But it is not just variations in fuel gas composition and type that power plants may have to contend with. For some consumers the use of oil as a back-up is an issue. What are the effects of such a change in fuel on the turbines, particularly for long-duration operation and when making the switch at high load?

Extended fuel oil operation

Dual-fuel engines can burn fuel oil as a back-up if there is a shortage of natural gas. However, gas turbines operating on fuel oil have a higher requirement when it comes to inspection and maintenance because the machines are not designed to run on such fuel for extended periods of time. The fuel oil factor, which additionally penalises the equivalent operating hours, is defined based on how the hot gas parts deteriorate in the field.

Forty six units in the GT26/GT24 fleet have accumulated a total of 32,000 fired hours and more than 2600 starts on fuel oil in commercial operation. The fleet leader for GT26 extended fuel oil operation is the Dock Sud plant in Argentina, where a limited availability of natural gas has meant two of its units burn fuel oil for 50-70 days a year. The hot gas parts in these two engines have seen up to 6000 oil-fired hours and proved to be in good condition at the 3.5 year hot gas path inspections.

Furthermore a mixture of 7 per cent biodiesel in standard diesel oil #2 – known as B7 – has been in commercial operation since 2012, with commissioning experience of a mixture of up to 20 per cent biodiesel, or B20.

The main mode of degradation related to operation using back-up fuel is hot corrosion by salts in the media streams, particularly in the fuel. Investigations into the progress of this corrosion in the case of hot gas path components have shown that the GT26/GT24 can relax the penalisation factors for extended oil operation. This finding is based on the positive experience and good condition of the engines.

Accumulated field experience demonstrated the GT26/GT24 can offer fuel oil operation without additional penalisation for typical operation periods with back-up fuel. This allows most customers to use fuel oil as back-up when necessary without any impact on availability or maintenance costs. If extended operation is required, a fuel oil factor ensures inspections are performed at reasonable periods in order to maintain the reliability and acceptable condition of parts during standard hot gas path inspections.

Successful fuel switching

When the turbine has to change its fuel source to fuel oil it must do so without interruption and with minimum variation in power output and heat energy to the HRSG. So the turbine should be flexible enough to allow an immediate switch once a fuel gas shortage is recognised.

The key operational parameters at a fuel switchover are the power and the exit temperature of the gas turbine. Both parameters are important inputs for the HRSG, and therefore for the performance of the combined-cycle power plant as a whole.

Previous versions of the GT26 required the power output to be reduced to a relative load of around 8 per cent to perform the fuel switchover in EV-only operation, in other words after switching off the second combustor, or SEV. This had the advantage of significantly reducing the requirements for fuel gas mass flow and pressure, which is important during fuel gas shortages. But it also had the commercial disadvantage of a consequential reduction in load. The EV-only switchover by push-button to fuel oil takes around 12 minutes from a relative load of 60 per cent. Now the advanced concept allows a fuel change-over in both directions on higher loads, in other words with the second, or SEV, combustor in operation.

Figure 8 shows how the high-load fuel transfer can now take place at a relative load of 60-95 per cent without any reduction in load. The low fluctuation in turbine exit temperature allows the steam turbine to stay in operation without a reduction in power. Field experience has shown a fuel transfer time of around 10 minutes from baseload fuel gas to fuel oil. In commercial units, more than 100 switchovers have been performed to date.

Figure 8
Figure 8: High-load fuel switchover at around 95% relative load in a commercial unit

The challenges in the advanced concept have been the realisation of the high-load fuel switchover with the sequential combustor design. In the previous concept the fuel transfer only took place in the EV combustor at a certain load point. The advanced version has a variable load range and includes the transfer of fuel in the sequential combustor. The transfer takes place in one combustor at a time as this avoids instabilities in operation. In the case of a switch from gas to fuel oil, first the EV combustor switches then the SEV combustor. In the case of a switch from oil to gas, the process is in reverse.

The advanced fuel switchover concept allows the customer to react according the demands of the grid or the fuel gas supply. Scenarios in which the pressure of the fuel gas drops are project specific. Whereas the EV-only fuel transfer allows a fast reduction of the mass flow and pressure of the fuel gas supply with the turbine connected to the grid, the high-load fuel switchover allows the plant a fast switchover at constant load (Figure 9). Both options deliver high flexibility independent of grid requirements or fuel gas supply.

Figure 9
Figure 9: Illustrating the extended fuel switchover concept

Field experience indicates that the fuel switchover time can be further reduced and the load range from 60-95 per cent further extended. The advanced fuel switchover concept extends the flexibility of the GT26 significantly and meets market requirements.

The challenges posed to gas-fired power plants by the changing nature of fuel supplies have been met by the GT26 and GT24 turbines. The technology platform of these units has been gradually improved to meet market demands. Upgrades to the turbines in 2011 that incorporate advanced fuel flexibility were based on the fleet experience of the 2006 upgrade of the GT26. We have demonstrated the capabilities of these new units in test rigs at our Birr facility in Switzerland and at commercial sites, capabilities that can give power plants the flexibility they now need to cope with the typical changes they will experience in fuel gas compositions.

Peter Marx is R&D Program Manager, Product Support, Gas Turbines, Martin Liebau is Project Manager, Belén Gasser-Pagani is R&D Engineer Product Support and Mark Stevens, Gas Product Manager, all at Alstom Switzerland Limited. For more information, visit

The article is based on a winner at this year’s POWER-GEN Europe Best Papers Awards.

More Power Engineering International Issue Articles
Power Engineering International Archives
View Power Generation Articles on