In any energy transition to a low-carbon system there will be a need for decarbonised fuels and thus a requirement for simpler, scalable hydrogen production processes. Dynamic or chemical looping processes have the potential to deliver improved routes to hydrogen production in significant part due to a breakthrough concept from my ERC Advanced Grant, SPEeD (Nature Chemistry, 11 (2019) 638). Thus chemical looping allows us to simplify a complex scheme involving multiple reactors and separation into a single unit with large capital cost savings, possibly up to one order of magnitude. Operation of such a chemical looping hydrogen production process at demonstration scale will facilitate market uptake of scalable, low-carbon hydrogen production processes and allow for refinement in the calculation of cost savings.
As a vital step in the hydrogen production process, here we consider the reversible water-gas shift (WGS) reaction,
H2O + CO⇄ H2 + CO2 Reaction 1
where H2O is reacted with CO to produce H2 and CO2. Reaction 1 is key to many H2 production processes and performed on a very large scale industrially (total current capacity of 50 million tons per annum (US DoE, www.hydrogen.energy.gov/pdfs/hpep_report_2013.pdf)). The reaction is equilibrium limited (a video describing my relevant ERC-funded work can be found at
http://nuvision.ncl.ac.uk/Play/18143(s’ouvre dans une nouvelle fenêtre)) and this leads to significant process complexity involving multiple reaction stages and separation. Thus overcoming this equilibrium limitation through the use of membranes (e.g. Gas Sep. Purif. 10 (1996) 243) and CO2 absorption (e.g. Int. J. Hydrogen Energy 38 (2013) 6065) has attracted considerable attention.
In a chemical-looping WGS reactor, H2O (high oxidising potential) is first passed over a solid phase carrier of oxygen (oxygen carrier material or OCM) that accepts oxygen resulting in the production of H2. Then in a second step CO (low oxidising potential) is passed over the solid phase, removing oxygen to produce CO2.
Step 1, H2O-feed, bed-oxidation half cycle:
H2O + [OCM] ⇄ H2 + O[OCM] Reaction 2
Step 2, CO-feed, bed-reduction half cycle:
CO + O[OCM] ⇄ CO2 + [OCM] Reaction 3
Conventional OCMs, such as metal-metal oxides, function by donating oxygen and receiving oxygen, at the fixed oxygen chemical potentials associated with their phase transitions. The practical consequence of this is that one can never have an OCM that gives a high conversion for both Reaction 2 and Reaction 3 at the same time. In fact by some measures such a reactor can perform no better than a conventional ‘mixed’ reactor (see video,
http://nuvision.ncl.ac.uk/Play/18143(s’ouvre dans une nouvelle fenêtre)).
SPEeD demonstrated that a ‘chemical memory reactor’ can be built using a single OCM capable of transferring the necessary chemical information for thermodynamically reversible operation between two chemical looping half cycles allowing unlimited conversion in both half cycles and the production of pure, separated hydrogen (Nature Chemistry, 11 (2019) 638). This ‘chemical memory reactor’ consists of a bed of a non-stoichiometric oxide OCM operated in reverse flow.
PuSH completed the design of a 0.3 Nm3(H2)/hr demonstration-scale hydrogen production unit producing ‘pure’ hydrogen from a high temperature chemical looping shift process employing such a ‘chemical memory reactor’. This included equipment sizing, a piping and instrumentation diagram (P&ID) and a full HAZOP (hazard and operability) study of the plant.
A suitable oxygen carrier material was selected and sourced at the kg scale. Additional work on the design and production of higher capacity oxygen carrier materials has been undertaken.
All components for the plant were procured and the plant was constructed. The plant is separated into five units: steam feed, flow system, reactor system including flow direction switching, downstream including gas analysis and venting and the control system. Further results from operation of the unit which will confirm development to TRL 5. There is an on-going dialogue with potential investors and technology partners. The patent protecting the intellectual property underpinning the concept has now been granted in Japan, China, USA and Russia.
