Stochastic Reactor-Based Fuel Bed Model for Grate Furnaces

Biomass devolatilization and incineration in grate-fired plants
are characterized by heterogeneous fuel mixtures, often incompletely
mixed, dynamical processes in the fuel bed and on the particle scale,
as well as heterogeneous and homogeneous chemistry. This makes modeling
using detailed kinetics favorable but computationally expensive. Therefore,
a computationally efficient model based on zero-dimensional stochastic
reactors and reduced chemistry schemes, consisting of 83 gas-phase
species and 18 species for surface reactions, is developed. Each reactor
is enabled to account for the three phases: the solid phase, pore
gas surrounding the solid, and the bulk gas. The stochastic reactors
are connected to build a reactor network that represents the fuel
bed in grate-fired furnaces. The use of stochastic reactors allows
us to account for incompletely mixed fuel feeds, distributions of
local temperature and local equivalence ratio within each reactor
and the fuel bed. This allows us to predict the released gases and
emission precursors more accurately than if a homogeneous reactor
network approach was employed. The model approach is demonstrated
by predicting pyrolysis conditions and two fuel beds of grate-fired
plants from the literature. The developed approach can predict global
operating parameters, such as the fuel bed length, species release
to the freeboard, and species distributions within the fuel bed to
a high degree of accuracy when compared to experiments.