OCR Text |
Show furnace is shown in Figure 4. The furnace is 45 feet from hopper bottom to top and is 22 feet deep by 16 feet wide. It utilizes an external water spray to cool the steel furnace. Refractory is installed in the burner zone. Direct flring, multiple burner operation, and vertical orientation advantages are like those for the CFPF. It has also been characterized by testing with numerous burners. Pilot-Scale Considerations An extensive review of the literature regarding pilot-scale testing was not done for this paper, but two items that have come to our attention provide challenging comments[ll.121. This is of interest, since the demonstration site designs must be based upon pilot-scale data and modeling input. Richter121 argues that mathematical modeling must be considered as an alternative to pilot-scale testing. At the least, modeling can be used to support the scale up of pilot-scale data. He adds that it is almost impossible to achieve duplicate time/temperature histories between pilot- and full-scale furnaces. Lack of geometrical similarity and differences in local flow, mixing, heat release patterns, and radiation (which can affect ignition characteristics) can limit the extent to which pilot-scale data can be extrapolated. Similar arguments are presented in the EPRI report[ll]. Johnson and Sotter conclude from their survey of pilot-scale test facilities that among other things, NOli. and unburned combustibles results cannot be accurately scaled up. We feel that it can be argued that modeling can provide insights, but the state-of-the-art is not such that modeling alone can provide commercial design infonnation for new technologies. At B&W, numerical models are used along with engineering correlations and operating practice to scale up the pilot-scale data. We will cover some of these points (and concede some) in the remainder of this paper. Detailed discussion of comparisons to larger scale for both test programs is presented later in this paper, but some general comments are given here. The pilot- and intennediatescale pulverized coal burners (cell burner in this case) are built by geometrically scaling down the full-scale design except for the throat diameter. Throat velocity is matched to simulate near-burner flow patterns and coal grind size is duplicated (this has been successful at B& W). This, combined with proper test furnace heat transfer (heat flux to walls is simulated by using refractory), should result in nearly duplicating fleld results for combustion and emissions. Local radiation may be less compared to some opposed-wall or multi-burner row field units, but we are interested in simulating a "generic" boiler. Consequences of combustion, such as slagging, fouling, and ash characteristics (for ESP perfonnance), are harder to quantify. This is caused by the lower pilot-furnace velocity (may effect slagging/fouling) and lower S03 levels due in part to less and cooler convective surface area. Still, even these parameters were satisfactorily addressed in the test programs. The cyclone reburning technology presented a number of challenges. Building a continuously operating I8-inch cyclone was only the start! The cyclone furnace is a geometrically- scaled version of the B&W cyclone furnace, but steam is not raised in the water jacket. Rammed refractory on welded studs is used to duplicate that aspect of the field cyclone furnaces. With the larger surface-to-volume ratio and colder water jacket of the pilot-scale unit, increased heat transfer (compared to a large 8 to10-foot diameter commercial cyclone 4 |