Driven by the search for the highest theoretical efficiency, several studies have investigated in the last years the adoption of fuel cells in the field of power production from natural gas with CO2 capture. Most of the proposed power cycles rely on high temperature fuel cells, namely Solid Oxide Fuel Cells (SOFC) and Molten Carbonate Fuel Cells (MCFC), based on the concept of hybrid fuel cell plus gas turbine cycles. Accordingly, high temperature fuel cells are integrated with a simple or modified Brayton cycle. As far as SOFC are concerned, two main plant solutions can be identified depending on the integration with the natural gas reforming/shift section: (i) systems where natural gas is — partially or totally — internally reformed in the fuel cell and (ii) systems where natural gas is reformed before the fuel cell and the cell is fed with a high hydrogen syngas. In both cases, CO2 can be separated downstream the fuel cell via a range of available technologies, e.g. chemical or physical separation processes, oxy-combustion and cryogenic methods.
Following a literature review on very promising plant configurations, this work investigates the advantages and limits of adopting an external natural gas conversion section with respect to the plant efficiency. As a reference plant we considered a power cycle proposed by Adams and Barton , whose performance is the highest found in literature for SOFC-based power cycles, with 82% LHV electrical efficiency. It is based on a pre-reforming concept where fuel is reformed ahead the SOFC which thus works with a high hydrogen content fuel. This plant was firstly reproduced considering all the ideal assumptions proposed by the original authors. As second step, the simulations were focused on revising the power cycle, implementing a complete set of assumptions about component losses and more conservative operating conditions about fuel cell voltage, heat exchangers minimum temperature differences, maximum steam temperature, turbomachinery efficiency, component pressure losses and other adjustments.
Considering the consequent modifications with respect to the original layout, the net electric efficiency changes to around 66% LHV with nearly complete (95%+) CO2 capture, a still remarkable but less attractive value, while requiring a very complex and demanding heat exchangers network. Detailed results are presented in terms of energy and material balances of the proposed cycles. All the simulations have been carried out with the proprietary code GS, developed by the GECOS group at Politecnico di Milano.