OCR Text |
Show laboratory-scale and commercial-scale C F B C systems. The previous configuration of the C F B C system, used by de Diego et al " and W a n g et al 23"27, has been modified to accommodate a commercial gas-fired burner and a simple fuel injector. For the first time, to the knowledge of authors, a comparison of N 2 0 reduction by afterburning, between an externally accommodated commercial gas burner and direct fuel injection through a simple fuel-gas injector is made and this should provide useful information on optimising afterburning configurations in commercial-scale C F B C applications. A comparison on the effectiveness of reducing N 2 0 between three common secondary fuels, CH4, C 2H6 and C3H8 is also made after conducting afterburning tests with direct injection of each secondary fuel. Gas-phase modeling using the detailed reaction scheme, GRI-Mechanism Version 2.11, is carried out to assess the role of N 2 0 thermal decomposition in reducing N 2 0 emissions in the afterburning zone and to compare the effectiveness of C H 4 and C2H6 in reducing N 2 0 when injected as a secondary fuel. EXPERIMENTAL Figure 1 shows a schematic diagram of the C F B C system used in this study. It comprises of a riser of diameter 161 m m and length 6.2 m, primary and secondary cyclones, an external heat exchanger (EHE), an L-valve for controlling the circulation rate of solids, an afterburning fuel injector and a commercial gas-fired afterburner. The whole system was made of stainless steel, and was insulated with ceramic fibres on it's outer surface except for the secondary cyclone. The water-cooled afterburning fuel injector was installed through a horizontal port 350 m m below the exit of the riser. A nozzle mixing, gas burner (FRG M V S 1 A ) , rated at 29 k W , made by Stordy Combustion Engineering Limited, was chosen as the afterburner. A s shown in Fig. 1, the afterburner, having a 250 m m long silicon carbide flame tube, fired into the top of the riser so that the main combustion products from the riser would mix with those of the afterburner before entering the primary cyclone. The afterburner flame tube was accommodated in an insulated stainless pipe, which was at the same level as the inlet to the primary cyclone. Primary air was supplied to the base of the riser through an air distribution plate, while secondary air was injected through opposing injectors located at 1.83 m above the primary air distribution plate and tertiary air was injected through the afterburner flame tube. A n afterburning fuel could be introduced into the main combustion products in the riser, either through the fuel injector, i.e. direct fuel injection, or through the afterburner where it was ignited and partly combusted. Three fuels, namely CH4, C2H6 and C3H8, were tested with direct fuel injection and one fuel, C3H8, was tested with the burner. Silica sand, having a Sauter mean diameter (SMD) of 120 pm and a particle density of 2500 kg/m3, was used as the main circulating bed material. Either a U K bituminous coal or a Colombian bituminous coal was used as the primary fuel with Table 1 showing the proximate and ultimate analyses of the two coals. All coal particles were crushed to less than 3 m m and typically their S M D was about 0.48 m m . Coal particles were fed by a screw feeder and pneumatically transported into the riser at the height of 316 m m above the air distribution plate. Most solids entrained by combustion gases from the riser were collected by the primary cyclone and fed to the E H E via a dipleg. After exchanging heat with fluidising/combustion air of the E H E and water-cooling stainless steel tubes in the E H E , the solids were returned to the riser at a controlled rate through the L-valve. The temperature level along the riser w a s controlled by the position of water-cooling tubes inside the E H E and the average bed |