OCR Text |
Show Pilot-scale Numerical Modeling In conjunction with the pilot-scale testing, a series of numerical modeling cases of pilot-scale burner concepts were completed during development of the LEBS advanced 10w-NOx burner. Modeling cases were completed using B&W's in-house developed flow and combustion model, COMO [6,7]. COMO is a fundamentally based multi-dimensional model which simulates the complex interacting processes which occur in pulverized coal combustion: turbulent flow, particle transport and dispersion, gas and solid phase chemical reactions, and convective and radiative heat transfer. The model uses an Eulerian control volume based treatment for the continuous gas phase flow, and a Lagrangian particle model for the disperse solid phase flow. Beginning with the existing low-N0x. burner concept, modeling was used to evaluate concept variations that would be expensive and time consuming to build and test. These variations included changes in burner zone configuration, primary and secondary swirl, and air distribution. All configurations were compared with baseline predictions for B&W's commercial DRB-XCL ® 10w-NOx burner, and included evaluations of flame attachment and character, near and far field mixing, and predicted exit quantities such as carbon monoxide, unburned carbon in the ash, and NOx' In parallel with this modeling effort, pilot-scale testing campaigns were used to further evaluate key burner concepts identified through numerical modeling. Part of these tests included investigation of parameters that could be more effectively evaluated with experimental tests at the 5 MBtu/hr scale. Results from each of these tests were then fed back to additional model studies. This staggered series of modeling and testing provided an efficient means to evaluate concepts and further advance the burner design. Evaluations ·were completed using both two-dimensional (2-D) and three-dimensional (3-D) flow and combustion models. 2-D axisymmetric models were used to parametrically evaluate a range of burner configurations in an idealized pilot -scale burner tunnel. This approach was chosen to provide qualitative feedback on burner variations with significantly faster turn-around time than would be possible with full 3-D models. While being limited by the axisymmetric assumptions, these models have proven very effective in providing trending information on mixing, fuel penetration, internal and external recirculation zones, flame shape, and near burner temperatures, and exit conditions as the burner design was varied. Final concepts, both with and without staging, are being evaluated with 3-D predictions of the burner and the SBS facility to confirm axisymmetric trends and provide further insight into burner performance. A comparison of oxygen distributions predicted within the SBS for both staged and unstaged operation of a burner concept is shown in Figure 3. These 3-D predictions, completed for a limited number of key cases, provide an accurate representation of mixing, heat flux distribution, and residence time for the pilot-scale test facility. This is particularly important for staged burner operation, where overfire air is injected above the primary combustion zone, and accurate representation of the burner and overfire air streams is important for comparison with experimental tests. Additional work is currently ongoing to provide a comparison of predicted and measured near-flame species. Page - 6 |