THE GRACE PROJECT: DEVELOPMENT OF OXYGEN CARRIER PARTICLES FOR CHEMICAL-LOOPING COMBUSTION. DESIGN AND OPERATION OF A 10 kW CHEMICAL-LOOPING COMBUSTOR

摘要

A comprehensive research programme was launched to develop chemical-looping combustion (CLC), a new technology of unmixed combustion with inherent capture of CO_2, using metal oxide particles for the transfer of oxygen from the combustion air to the fuel. More than 240 different oxygen-carrier particles produced by extrusion were tested in the initial screening process. Furthermore, more than 50 particles produced by freeze-granulation, and some particles produced by impregnation were also tested. The particles included active oxides of nickel, iron, copper and manganese as well as several different support materials. A limited number of particles were selected for comprehensive testing. Fluidization conditions and recirculation flows were studied in cold-flow models indicating the feasibility of both the full-scale design and a small 10 kW prototype unit. A 10 kW prototype for chemical-looping combustion was designed, built and run with nickel-based oxygen carrier particles. The prototype uses two interconnected fluidized beds, a fuel reactor where the fuel is oxidized to CO_2 and H_2O by the oxygen-carrier particles, and an air reactor where oxygen is supplied to regenerate the particles. The air reactor also serves as a riser, providing the circulation of particles between the two reactors. The gases in the two reactors are kept separate by two fluidized particle seals. As far as known this is the first unit where this process has been in continuous operation. Start-up, turn-down and operation of the process were found to be easy. A total operation time of more than 100 h was reached with the same batch of particles, i.e. without adding fresh, unused material. During night-time and during start-up of operation the system was kept at high temperature and in circulation with electrical preheating. Thus, the actual time that the particles have been circulating in the system is close to 300 h. The fuel used was natural gas, and a fuel conversion efficiency of 99.5% was accomplished, which is very close to the thermodynamic equilibrium of the NiO/Ni system. It should be pointed out that there is no such thermodynamic restraint for the other metal oxide systems studied. There was no CO_2 in the gas from the air reactor, indicating that a separation efficiency of 100% is possible. Furthermore, there was no leakage in the opposite direction, i.e. from the air to the fuel reactor, indicating that pure CO_2 can be obtained in the process, except for nickel oxide for which there is a thermodynamic limitation. Neither decrease in reactivity nor particle strength was seen during the test period. The loss of fines was small and decreased steadily during the test period. In the end of the period the loss of fines, i.e. particles smaller than 45 μm, was 0.0023% per hour. If this can be assumed to be a relevant measure of the steady-state attrition, it would correspond to a lifetime of the particles of 40 000 h. Assuming a lifetime of 4 000 h, the estimated cost for particles will be low, in the order of 1 ?/ton CO_2 captured. A technical evaluation showed that the process uses technology very similar to circulating fluidized-bed combustion. Thus, a chemical-looping boiler can be built with adaptation of well-known technology. A preliminary costing has been performed for a 200 MW_(th) chemical-looping combustion boiler for use at BP's Grangemouth refinery, indicating that CLC should feature strongly among the best options for reducing the cost of CO_2 capture.