No abstract

A computational model of a generic commercial two-stage entrained-flow upflow coal gasifier has been used in the present work to aid the researchers of the coal partitioning project at the National Energy Technology Laboratory (NETL). Specifically, the level of variation in the predicted performance in a generic entrained-flow gasifier if both coal slurry density fractions as well as diameter fractions are properly modeled in the CFD simulation was determined. Simulations were performed using the commercial computational fluid dynamics (CFD) software Fluent for a Pittsburgh #8 Bailey coal to analyze fixed carbon conversion, exit gas volume fractions, and particle trajectory patterns. The computational simulation involved effects of flow turbulence, finite-rate chemistry, and solid-gas phase interactions. In order to represent the coal sample in a CFD code, two basic case setups were explored. The first used a Sauter-mean average diameter along with the overall mass average density of the 28 size/density fractions of the Bailey coal sample. Each surface injection was then treated with a uniform density and diameter distribution. The elemental composition of the parent coal was used in setting up the boundary conditions and volatile component distribution for the first case. The second case used 28 simultaneous injections on each surface, each of which assumed the average density of that particular fraction along with a uniform diameter distribution of the corresponding size class. The Ultimate and Proximate analysis of each fraction was determined and included in the setup of the particle compositions within the CFD model for the second case. Two different devolatilization models were explored and the results were compared. Flow field characteristics such as center-plane velocity and temperature along with exit gas compositions were useful in judging the validity of the simulations as well as to compare predictions for the various chemistry models. Particle trajectory patterns were analyzed for coal feeds with varying initial diameter, density, and elemental composition. One of the purposes of the coal partitioning project was to analyze fly ash mass flux and slag layer buildup in a commercial scale gasifier. With this in mind, different particle wall boundary conditions were explored and data was generated relating to the state of the particle when it impacts the reactor wall. Average impact velocity component normal to the wall, temperature, and char content of the impacting particle were recorded for flow fields obtained from case setups that used different injection profiles and devolatilization models. Validation studies were performed against data taken from an experimental facility at SRI International. Results for three different experimental configurations were compared with predictions from Fluent.