SGT-600 cutaway showing gas/air flow. In 2018 and 2019, the SGT-600 was tested with its 18 burners at varying load on hydrogen levels above of 60%-vol H2.
There is an increasing sense that curbing the greenhouse effect has become a decisive factor in energy policy worldwide. The Paris Agreement, negotiated at the 2015 UN Climate Change Conference, required net zero emissions to be achieved by 2045-2060, but political pressure in industrialized countries is mounting to increase the speed of transformation.
Replacing fossil fuels with hydrogen is one way to help achieve this goal. And gas turbines entirely fueled with hydrogen produced from water and surplus renewable electricity could provide a significant contribution to reducing CO2 emissions. In addition, gas turbines are very flexible, have short ramp-up times, and are very effective in balancing variable renewable energy production. Accordingly, there is a great deal of effort to reach the goal of totally renewable energy production with the help of gas turbines.
Hydrogen is seen by many as a good way of storing renewable energy. Jenny Larfeldt, Senior Expert in Combustion Technology at Siemens Industrial Turbomachinery, said: “This is why gas turbines capable of running solely on hydrogen may prove to be the missing link to establishing a green and sustainable energy sector.”
In January 2019, Siemens signed a commitment to gradually increase the hydrogen capability in gas turbines to at least 20% by 2020, and 100% by 2030. The first goal has already been achieved, with projects in place, including a 60% hydrogen project involving two SGT-600 turbines at Braskem in Brazil.
Braskem, the largest petrochemical company in Latin America, is modernizing a cogeneration power and steam plant at its Petrochemical Complex in São Paulo, Brazil. Completion of the project is scheduled for early 2021. The Braskem project provides an example of another driver towards the adoption of hydrogen-fired turbines. This modernization project involves two SGT-600 DLE (dry low emissions) gas turbines co-fired with 60% hydrogen by volume. The turbines will each generate 19 MWe, corresponding to base load operation at site ambient conditions and provide 80 tons/hour of steam.
Why use hydrogen?
Traditionally, power demand has been variable, and grid operators had to balance the network through adjustment of power supply. The growth of renewable power supply, particularly wind and solar power, has meant that increasingly power supply has become variable as well, dependent upon climate variations. Hydrogen is a good way of storing this variable renewable energy, and gas turbines using this hydrogen later can provide the necessary flexibility to balance the grid as well as minimize carbon emissions.
In addition, many industrial processes produce hydrogen as a process gas, and using this rather than shipping in natural gas can be economically attractive. Many such sites use process gas to generate heat and steam, and it is only natural (and even more efficient for process plant) to extend this to include electricity as well.
Using process gases can have issues that need to be resolved. For example, process gases may be contaminated with salts or, especially in the case of the steel industry, dust and dirt. These are easy enough to clean from the flow.
In addition, the composition of process gases can vary. At Braskem, for example, several fuel streams are available, and the process has natural variations. Consequently, the turbines need to be able to cope with the predicted range in which they are likely to operate. Anders Nilsson, Product Manager of the SGT-600 for Siemens, says that even if the process provides gas outside the planned operating range, the turbine will still be able to operate, although there may be consequences such as temporarily higher NOx emission levels and, in some cases, potential load reduction.
Addressing hydrogen issues
According to Larfeldt, one of the main issues that needs to be factored into design is that hydrogen has a very rapid flame speed. This can result in the flame being sucked back, which may damage or destroy the burner. She said: “We use dry low NOx burners (DLE), so, in order to protect the burner hardware, the speed of the fuel and air mixture upstream in the burner has to be higher than the flame speed of the hydrogen. The key to increasing the hydrogen ratio in the fuel mix lies in the burner design, and we had to adjust the fuel injection accordingly.”
The important element was the design of the burners. The introduction of additive manufacturing, or 3D printing, enabled the design process to proceed much more quickly. She said: “This technology made it possible for us to adjust the design inside the burner without changing the exterior. In turn, it has also made it easier for us to retrofit existing turbines, and to dramatically speed up the development of burner design.”
Jenny Larfeldt, Senior Combustion engineer at Siemens, used 3D-printed dry-load emissions (DLE) burners to reduce the heat load.
High flame-speed makes the flame more compact and closer to the burner tip. This results in the heat being greater at the burner tip. By using additive manufacturing, it is easy to improve burner tip cooling design to mitigate the higher heat load. Thus, additive manufacturing speeds up the design process, and allows rapid production of a prototype for testing. Printing new and modified components is a quick and simple matter.
Other modifications required include the use of flashback detection similar to that traditionally used for DLE operation with liquid fuels. Flame flashback takes place when there is upstream propagation of the flame back into the burner; the higher flame speed of hydrogen makes this more of an issue than with just natural gas as a fuel. In the unlikely event of flashback, a flashback out (FBO) system will be activated to resolve the situation and keep the gas turbine in operation.
Customers demand very high levels of availability and reliability, and this involves considerable and thorough testing. Additive manufacturing also enables changes in design to be rapidly prepared, tested and introduced.
Hydrogen development on DLE started in 2008 with high-pressure combustion testing of a standard individual SGT-600 burner in 2008, mapping it at varying engine conditions. This was followed by multiple single burner test campaigns as design was reiterated. After proving the burners in single configuration, the next challenge is to test them under real conditions. In 2017, 30 burners were tested within the SGT-800 gas turbine with up to 50%-vol levels of hydrogen. In 2018 and 2019, the SGT-600 was tested with its 18 burners at varying load on hydrogen levels above of 60%-vol H2. Since the DLE burner is the same for the SGT-600, SGT-700 and SGT-800 engines, testing creates synergies for all those products per default.
According to Nilsson, one of the limiting factors with testing is now the availability of hydrogen. Tests lasting around an hour can burn through the available stored hydrogen, which can take up to a week to replenish. It is a significant logistical challenge. Larfeldt said: “We need large quantities of hydrogen gas, but there is no large-scale production near Finspång. We sourced five containers with hydrogen in flasks at 250 bar from all over Sweden, and two in slightly larger flasks from Germany. Together, these still only give us enough hydrogen gas to do large-scale tests for less than one hour.” An inherent conflict of proving gas turbine performance on operation with the customer process gas without testing on site at a customer plant is always trying to mimic the process gas in delivery engine tests at Siemens’ Finspång manufacturing site test bed. This is not an issue at sites where hydrogen is a process gas.
How easy is it to convert?
At 30% hydrogen and below, the changes required are minor, largely in auxiliary systems. Moving up to 60% H2 might also involve redesign of gas and fire detection systems, as well as ventilation in enclosure in accordance with the safety classification, changing of gas valves and piping, in addition to changing the burners. Tuning of emissions as well as modification of the start-up sequence may be required. All these changes can be carried out during a standard outage.
CO2 reduction
Larfeldt says that reducing CO2 emissions now is the priority. While historically the NOx emissions, causing local pollution, has been in focus it is now time to think global and reduce CO2 emissions. In Europe for instance the NOx emissions from heat and power generation using gas turbines amount to about 2.2% of the total NOx emissions. The introduction of DLE combustion technology in gas turbines has contributed to this success.
The use of hydrogen as a fuel lowers CO2 emissions from the combustion process since hydrogen burns to water. Using DLE burners, the flame temperature is controlled and kept the same. However, the more compact flame, near the burner tip, tends to raise NOx emission levels slightly. This latter can most likely be improved by tuning of the burner design but also mitigated by secondary NOx emission control systems.
Current gas turbines typically generate around 500g CO2/kWh in a simple cycle configuration and around 300g CO2/kWh in a combined cycle configuration. This needs to be reduced to 250g CO2/kWh, and ideally further. Using hydrogen in co-firing applications can bring this about.
If Braskem runs the SGT-600 with 75% vol. hydrogen, it would reduce the carbon footprint by half and thereby reach this goal. It could potentially be a very significant market.
Over the next ten years, Nilsson sees the oil and gas industries moving increasingly towards green refineries and turning towards green fuels. Because of the process gas stream, the use of hydrogen to provide power is growing. Nilsson felt that big refineries and plastics producers were likely to be the next big market.
In addition, hydrogen production by variable-output renewable plants, such as wind and solar, can be used to provide energy storage to power hydrogen-fired gas turbines when demand exceeds renewable supply. Hydrogen production is the key to the process.
Currently, over 50 million tons are produced worldwide annually, mostly from steam methane reforming or autothermal reforming. Some 50% of this hydrogen is currently used to synthesize ammonia. The conventional method for producing hydrogen is the steam reforming process, in which steam reacts with natural gas to produce syngas, a mixture of hydrogen and carbon monoxide.
The carbon monoxide is shifted to CO2, which is removed. In comparison, generating hydrogen via the electrolysis of water with electricity from renewable sources is completely free of CO2 emissions from the beginning. To achieve rapid reductions in CO2 emissions, changes must be made quickly. That means that developments must consider existing infrastructure. Gas turbines provide a good complement to variable renewable energy, being able to run at variable output and with rapid ramp rates.
Braskem Project detail
Braskem, the largest petrochemical company in Latin America, is modernizing a cogeneration power and steam plant at its Petrochemical Complex in São Paulo, Brazil. Completion of the project is scheduled for early 2021.
Siemens will engineer, deploy, operate and maintain the entire plant for a period of 15 years, with a possible extension beyond that, under a long-term contract that includes performance guarantees for reliability, availability, efficiency, costs, maintenance, and emissions.
The plant includes two SGT600 DLE gas turbines, and the project will include extension of the existing high-voltage substation, three reciprocating compressors, an advanced load-shedding system, and associated software for plant control.
This project will provide steam and power energy to the complex’s cracking unit. The unit has an ethylene production system and produces raw materials for the chemical and plastic sector. The upgrade project will reduce the cracking unit’s water consumption by 11.4%, and CO2 emissions by 6.3%.
The standard power output of an SGT-600 turbine is 24 MW. For this application, each turbine will provide 19 MWe and 80 tons per hour of steam. In addition, the turbines will feature third-generation dry low emissions (DLE) technology and run on residue gas with 60% hydrogen. NOx levels from the turbines will be 25 ppm or lower.