(122d) Chemical Looping Simulations Using Aspen Plus

Carbon capture from power plants is recognised as one of the primary remits of governments worldwide as part of their duty to mitigate climate change.  Chemical looping combustion (CLC) is potentially the technology best suited for the efficient, low cost and low energy capture of CO₂ by separating the CO₂ from flue gases. The process consists of cyclic reduction and oxidation of a metal oxygen carrier (in the form of particulates) which is exchanged between two reactors.  The oxidation reactor is usually a circulating fluidised bed and the combustor is often a bubbling bed reactor. The oxygen carrier transfers oxygen from the air to the fuel, hence direct contact between air and fuel is avoided. Consequently, the outlet gas from the combustor contains only CO₂ and H₂O; the latter is easily removed by condensation and the CO₂ is readily captured for storage and/or utilisation. The main attractiveness of the process is that it does not involve any penalty for extra highly intensive separation of CO₂ from the other combustion products. Indeed, separation always involves high energy consumption with the possible consequent production of additional CO₂.

A number of challenges exists for the optimal operation of the CLC set-up and those involve the characterisation of the oxidation reactor, the optimal circulation rates, the thermal output, the reactors’ residence time.

The present paper presents the process integration analysis of the CLC set-up and the implementation of a novel method to simulate fluidised beds using Aspen Plus software. The fuel reactor is modelled as PFR and CSTR reactors in series: the PFR reactor is employed to describe the bubble phase whereas the CSTR describes the emulsion phase. The air reactor is modelled via four CSTRs in series which take into account the variation of the void inside the reactor. Two different models are considered: the uniform conversion model (UCM) and the shrinking core model (SCM). In addition, various controlling steps of reaction are investigated and comparisons between the fluidised bed models proposed here and the Gibbs reactor model (usually implemented in Aspen Plus to simulate fluid beds) are carried out.

Results are shown from the two models and the net heat duty of the whole system is verified.