Going all-electric?25 July 2019
A new Haldor Topsøe led project is developing a concept for the electrification of ammonia production using solid oxide electrolyser cells without the need for an air separation unit. If successful, this would be a major step forward in exploiting ammonia’s huge potential for applications in the power sector.
The potential role of ammonia as fuel, energy vector and storage medium is attracting increasing attention. It can be used as a CO2 free fuel instead of gasoline, diesel and fuel oil. Moreover, the highly energy-consuming production of ammonia for fertiliser and other purposes (estimated to account for about 1% of global energy consumption) can be based on green power instead of current energy and CO2 intensive production methods (typically employing steam reforming of natural gas). The transition to ’green ammonia’ could eliminate significant CO2 emissions and at the same time provide a way to efficiently store excess electricity from wind turbines and solar facilities.
A new research project, SOC4NH3, headed up by Haldor Topsøe, is looking at the potential benefits of a low carbon ammonia manufacturing route based on solid oxide electrolyser cells (SOEC). The project is being run in collaboration with the University of Aarhus, Technical University of Denmark, Energinet, Vestas, Equinor, Ørsted, with funding from the Danish Energy Technology Development and Demonstration Program (EUDP), which is supporting the project with DKK 15.9 million out of a total budget of DKK 26.8 million. The project leader is John Bøgild Hansen, senior principal scientist, Haldor Topsøe.
Electrolysis based schemes have been used previously for ammonia production (typically employing alkaline electrolysis) in conjunction with an air separation unit. Interest in the electrification of ammonia production is now re-emerging as concerns about CO2 emissions grow and the price of electricity from intermittent renewables decreases. The energy intensity and OPEX using classical, alkaline electrolysis are, however, high relative to natural gas based processes and the CAPEX high due to the separate air separation unit and use of state of the art electrolysers.
Solid oxide electrolysis cells offers much better energy efficiency than alkaline or PEM based electrolysers, but in a conventional scheme would still need to be coupled to an air separation unit in order to generate ammonia synthesis gas with a hydrogen/nitrogen ratio of 3 (Figure 1).
Haldor Topsøe’s new process (patent pending) for making H2/N2 ammonia synthesis gas utilises the unique capabilities of the SOEC acting as an oxygen separation membrane and its ability to make use of heat in lieu of power so that the expensive air separation unit can be eliminated (Figure 2). The nitrogen is introduced to the process by simply burning some of the hydrogen generated by the steam electrolysis.
Another way of looking at the overall scheme is to regard the SOEC as the air separation unit with the electrolyte providing highly efficient separation between the oxygen and the fuel side of the process.
Rather than using electric heaters for final preheating and generating additional steam, as proposed in previous schemes for SOEC based production of hydrogen for ammonia synthesis gas, in the new process, air is added and burned in catalytic burners instead. The new concept is called SOEC4NH3.
In these previously proposed schemes, the SOEC based hydrogen production is coupled with an air separation unit to provide the nitrogen. The key point is that the new process can be realised at a much lower investment cost due to the elimination of the air separation unit.
If the stacks are operated below thermo- neutral voltage power is saved but a temperature drop across the stack occurs.
Air can then be burned between the stacks bringing the inlet temperature to the next stack up again. This process can be repeated until a stoichiometric gas with a H2/N2 ratio of 3 is obtained suitable for the ammonia synthesis. The burning of the air generates extra water, which needs to be electrolysed, but due to the high efficiency of the SOEC – being close to 100 % – the net result is only a marginal increase in overall power consumption.
The simultaneous generation of hydrogen and nitrogen does, however, come at the cost of larger stack areas because the average current density needs to be lower when heat is added externally in order to keep the stacks overall thermoneutral.
There is thus a tradeoff between low energy consumption and investment in stack area. The optimum will depend on electricity price and stack costs. Analyses carried out by Haldor Topsøe have, however, shown that overall energy consumption for the new process is equal – within a few percent – to the energy consumption for the SOEC-hydrogen-production + air-separation-unit scheme mentioned above.
Aims of the project
In summary, the key aims of the new project (which builds on Haldor Topsøe’s world- renowned expertise in ammonia production technology, electrolysis and catalysis) are:
- demonstration of the new process for production of ammonia synthesis gas (ie, a stoichiometric mixture of H2/N2, suitable for Haber-Bosch synthesis) without the need for an air separation unit, by means of solid oxide electrolyser cells; and
- investigation of the use of ammonia as a fuel for solid oxide fuel cells (SOFC).
The envisaged ammonia synthesis gas production plant will employ a 50 kWe electrolyser. The SOFC unit test will be carried out on one stack corresponding to 1.5 kWe.
In parallel the project will include design and planning of a complete, industrial sized demonstration unit for all-electric ammonia production as well as socio and techno-economic studies on ammonia as an energy vector for storing excess electricity and using it for stationary power and heat production, shipping and other heavy transport.
The new process for ammonia synthesis gas generation is considered particularly suitable for an envisioned new generation of smaller scale ammonia plants (100 – 1000 tonne/day, corresponding to 30-300 MWe input) running entirely on renewable electricity. Such plants will most likely be decentralised and be of a size matching available renewable power sources nearby. The first electrified plants employing the new technology will not be of the very large size that conventional ammonia plants based on steam methane reforming are today, typically 2200 tonne/day, which would require 670 MWe input even with the new efficient process. Steam electrolysis by SOEC has reached a TRL level of 5-6 and catalytic burners a TRL of 9 but the combination has never been demonstrated. The new process is thus at a very low TRL level of 1-2 but would – if demonstrated at a scale of 50 kW – be moved to a TRL of 5-6. For this project, the plan is to use the existing infrastructure and expertise for SOEC testing at Aarhus University’s Foulum site for the demonstration.
The size of 50 kW has been chosen as a reasonable compromise between cost, available infrastructure at Foulum and the need for a unit of a size large enough to ensure low heat losses, etc, in order to get realistic results.
As already noted, ammonia is an excellent fuel for a solid oxide fuel cell and this was demonstrated by Haldor Topsøe and Risø-DTU amongst others around 2005 in proof of concept experiments with the first version of small Haldor Topsøe SOFC stacks.
An ammonia based SOFC plant can have a very simple lay out as there is no need for any fuel processing, steam addition or anode recycle, although the latter can be employed to obtain an overall higher fuel utilisation.
The concept is called SOFC4NH3.
The SOFC stack technology has evolved substantially since the proof of concept in 2005, with respect to lower operating temperatures, productivity and robustness.
DTU Energy Conversion has acquired testing equipment that is now capable of emulating adiabatic stack performance so that a single stack test would provide a TRL lift from 3 to 5.
Ammonia and the energy future
As touched on earlier, the two experimental parts of the project will be supplemented with analyses of the role these processes could have in moving to fossil free energy systems, both in Denmark and internationally.
Previous analyses have focused on hydrogen and renewable carbon based fuels based on renewable electricity but ammonia has only recently been investigated as an energy vector for renewable energy.
Based on the results from the experimental demonstrations, heat and mass balances will be established for:
- ammonia production via the new SOEC process;
- combined heat and power production with ammonia as SOFC fuel;
- use of gas turbines coupled with ammonia cracking for combined heat and power; and
- use of ammonia in modified diesel engines for power and heat generation.
SOC4NH3 is used to refer to the concept of ammonia employed as an energy vector based on solid oxide cell technology.
The use of diesel engines with ammonia as fuel will also be investigated for shipping and other heavy duty applications.
The project participants represent all parts of the value chain required to deploy ammonia as an energy vector based on renewable electricity, says Haldor Topsøe, noting that the companies involved are all world leaders within their respective fields and the two universities are at the leading edge of solid oxide cell R&D.
Ammonia is used as a fertiliser and is also a feedstock for a large number of basic chemicals. Less appreciated is that it is also an excellent potential fuel and storage medium.
Among available energy storage technologies, chemical storage appears to be by far the best option for large scale applications (see graph below), and the energy density of ammonia is much higher than the other carbon free energy vector, hydrogen, while the logistics of handling it are much simpler.
Ammonia is toxic and should be treated with respect, but safety analyses have shown that it is not more dangerous than conventional fuels, when handled properly.
Image: Energy capacity of different energy storage technologies vs discharge duration (source Institute for Sustainable Process Technology, Amersfoort, The Netherlands, Power to Ammonia, 2017). P2F (NH3), aka P2A (power to ammonia), looks promising