OCR Text |
Show nonuniform, so that straight lines on Arrhenius diagrams can only be coarse approximations. Variation are particularly severe soon after the onset of devolatilization, at the beginning of tar evolution, and much later, when tar evolution ceases. Eventually the gas release rates for all coal types approach very similar values. The variations with rank segregate roughly into two categories. For ranks from lignite through hv bituminous, rank variations are moderate, especially during the later stages of devolatilzation at high temperatures. Nominal rates for these coal types vary by a factor of four at 670 K but only by 25% at 1000 K. The temperature at which devolatilzation commences also varies with rank, ranging from 550 K for the lignite to 650 K for the bv bituminous coals. (Of course, these temperatures will change for different heating rates.) Low-volatility coals comprise the second category. They begin to devolatilize at much higher temperatures and sustain significantly slower rates that the other ranks. Apparent activation energies for the two groups differ by nearly a factor of two, being only 8.6 kcaVmole for the lowrank plus bituminous group and 15.5 kcallmol for the low-volatility group. In Figure 5 b, predicted rates during unifonn heating at different rates illustrate how the nominal rates change as heating rates are varied. They show that rate variations over the entire rank spectrum are never as substantial as those for varying the heating rate by a single order of magnitude. Regardless of coal type, devolatilization rates increase in direct proportion to increases in heating rate; they increase by a factor of 6 for every order of magnitude increase in heating rate. The apparent activation energies are surprisingly unifonn, becoming only slightly larger for faster heating rates. Hence, given the sample's ultimate and proximate analyses and sufficient infonnation to assign a particle heating rate, FLASHCHAIN can be used to assign parameters in simple global rate laws for any coal type. Whereas a single fIrst order expression was analyzed here, the same approach can also be applied \vith competing 2-step or distributed activation energy rate laws. Above and beyond applications as a devolatilization submodel in detailed simulations, FLASHCHAIN can also be used as a virtual coal laboratory. In this sense, it provides the same infonnation that would nonnally be acquired in, for example, drop-tube tests, such as rapid heating volatiles yields, the partitioning between volatile- and char-nitrogen species, soot loadings, and gas compositions and heating values. In turn, these quantities can be used as regression variables in engineering correlations that relate coal properties to macroscopic boiler perfonnance characteristics. 8 |