Advanced Design Optimization of Combustion Equipment for BioEnergy Systems Using Sculptor with CFD Tools

Paper from the AFRC 2013 conference titled Advanced Design Optimization of Combustion Equipment for BioEnergy Systems Using Sculptor with CFD Tools by Joseph Smith.

In the past, design changes for combustion equipment including process burners, gas flares, vent incinerators, etc. was accomplished using the "build and try" method. This tried and true approach was very expensive and took a lot of time to evaluate up to three or four different designs to identify the "optimal" configuration to reduce pressure loss and/or improve mixing efficiency. Efficient optimization algorithms along with better computational fluid dynamics (CFD) tools are now available on affordable computer resources to allow a design engineer the option of examining hundreds of designs to find the "optimal" solution. Sculptor® from Optimal Solutions Software (OSS) automates engineering design shape optimization for CAE models (CFD, FEA, etc.). This is particularly useful for CFD shape optimization where chemically reacting flows and heat transfer is important in the design of combustion devices, chemical reactors, heat exchangers, etc. This paper discusses design optimization for combustion equipment used in a biomass fired furnace. Results of this work examine design concepts for co-firing biomass with coal to reduce greenhouse gas emissions while maintaining high furnace efficiency and minimizing NOx formation. This work is meant to illustrate the current state-of-the-art in design optimization using CAE tools to improve combustion equipment performance. Based on the analysis of the simulation results, implications of various design options are discussed and recommendations made regarding potential advances in basic biomass/coal burner design.

Advanced Design Optimization of Combustion Equipment for BioEnergy Systems Using Sculptor® with CFD Tools Joseph D. Smith, Ph.D. and Vivek Rao Missouri University of Science and Technology [e-mail] Mark Landon, Ph.D. Optimal Solutions Software 2825 West1700 North, Idaho Falls, ID 83402 [e-mail] ABSTRACT In the past, design changes for combustion equipment including process burners, gas flares, vent incinerators, etc. was accomplished using the "build and try" method. This tried and true approach was very expensive and took a lot of time to evaluate up to three or four different designs to identify the "optimal" configuration to reduce pressure loss and/or improve mixing efficiency. Efficient optimization algorithms along with better computational fluid dynamics (CFD) tools are now available on affordable computer resources to allow a design engineer the option of examining hundreds of designs to find the "optimal" solution. Sculptor® from Optimal Solutions Software (OSS) automates engineering design shape optimization for CAE models (CFD, FEA, etc.). This is particularly useful for CFD shape optimization where chemically reacting flows and heat transfer is important in the design of combustion devices, chemical reactors, heat exchangers, etc. This paper discusses design optimization for combustion equipment used in a biomass fired furnace. Results of this work examine design concepts for co-firing biomass with coal to reduce greenhouse gas emissions while maintaining high furnace efficiency and minimizing NOx formation. This work is meant to illustrate the current state-of-the-art in design optimization using CAE tools to improve combustion equipment performance. Based on the analysis of the simulation results, implications of various design options are discussed and recommendations made regarding potential advances in basic biomass/coal burner design. INTRODUCTION AND BACKGROUND System design optimization is an important part of meeting growing challenges related to environmental concerns related to NOx, CO and greenhouse gas emissions. Biomass combustion has received considerable attention as a way to provide electric power in regions without Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 traditional fossil energy resources. Hybrid combustion systems that combine coal with non-design fuels (i.e., biomass) have been used as a way to reduce greenhouse gas emissions. However, large differences in fuel quality for non-design biomass fuels (i.e., pine, poplar, switch grass, etc.) result in concerns about increased levels of NOx and CO. Several biomass burner designs (see Figure 1) have been developed to maintain combustion efficiency while reducing NOx and CO emissions for co-fired systems. Figure 1 - Typical Biomass/coal burners: a) Low-Emission Scroll-type Biomass/ Coal burner (Coen Sales brochure), b) BWE Bio-dust burner equipped with a gas lance and/or an oil lance (BWE Sales brochure) The present work applies a novel approach using an advanced shape optimization tool called Sculptor® from Optimal Solutions Software (OSS) to investigate engineering design optimization of a biomass burner system. Sculptor provides Arbitrary Shape Deformation (ASD) that controls and manipulates the shape change of any type of CAE geometry computational grids and/or CAD data. Using ASD, the design engineer defines control points around the shape being deformed and moves them to deform (morph) the entity into a new shape. This allows the engineer to mold (sculpt) the shape as if it was molding clay and expands the user's ability to examine any shape instead of being restricted to a single CAD model. Sculptor allows the engineer to smoothly deform the computational mesh without having to perform the extremely costly process of "remeshing" for each new shape considered. In addition, linking shape changes directly to governing physics (i.e. pressure drop, fuel/oxidizer, flow uniformity, mass flow, drag, mechanical stress, etc.) the engineer can achieve higher performance improvements not possible by the old "build-and-try" approach. a) b) 2 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 SYSTEM OPTIMIZATION: SCULPTOR BASED DESIGN FRAMEWORK Optimal Solutions Software, LLC, (OSS) has developed a revolutionary technology called Arbitrary Shape Deformation (ASD), a major advancement that helps solve the following critical problems in automated CFD and FEA engineering design: • Inadequate shape parameterization algorithms. • Inadequate algorithms for CAE (CFD and FEA) grid shape modification. Traditional CAD parameters have repeatedly failed to provide the correct parameterization that is needed in physics based shape optimization. ASD enables the user to control and manipulate the shape change of any type of geometry either CAD/CAM data or CFD/FEA computational grids. To use ASD, one defines a set of control points around the entity that is to be deformed. The control points are then moved and the underlying functions deform the entity into a new shape. This gives the user the ability to mold the entity as if he were a sculptor molding clay. This is perhaps best understood by using a physical analogy. Consider a volume of clear, flexible plastic, in which the object to be deformed has been embedded. The embedded object has the same degree of flexibility as the volume so that as the plastic volume is deformed, the embedded object is also deformed in the same manner (see Figure 2 and Figure 3). The volume is modeled as a trivariate parametric volume with its deformation controlled by a small set of control points. In creating the ASD, the user has control over the number of control points as well as where each control point is placed. The ASD technology and the accompanying graphical user interface tools to create and visualize the shape deformation process are found in OSS's commercial software package called Sculptor. ASD solves the first problem by allowing the CFD designer to freely create his own shape parameters therefore eliminating the restriction of only being able to use the parameters found in the CAD model. The second problem is solved by the fact that Sculptor performs a smooth volumetric deformation in real-time without the need to revisit the CAD model. This eliminates the extremely costly process of having to remesh the grid for every shape change. Sculptor has demonstrated the ability to perform a design change in seconds that currently takes hours or even days to complete where one goes back to the CAD model to reshape and remesh. Sculptor has also demonstrated the ability to parameterize, deform and optimize shapes into new and improved designs where the CAD model did not offer the necessary parameters or degrees of freedom. 3 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 Figure 2 - A spherical ball (mesh) is embedded in an ASD grid. Control points are moved and the ASD grid is deformed, causing the ball (the mesh) to deform Figure 3 - ASD volumes can be created to form a fit around complex geometry as shown in this Formula Race car front wing. A parametrically rectangular ASD volume is created and several design variables to change the wings' shape are selected with the initial shape and a (real-time) smoothed deformation of the middle wings' sections angle of attack which shows the smooth transition between the morphed and non-morphed locations Sculptor can be used in internal and/or external flow problems in aerospace, automotive, chemical processing, biomedical, electronics, power generation, turbo machinery, marine, basically anywhere that the shape of the geometry affects the performance of fluid dynamics, heat or mass transfer, chemical reactions, combustion, structural mechanics, Fluid Structure Interaction, etc. Here is a list of its current capabilities: (1) Import and export any CAE model (CFD and/or FEA) and/or CAD files. Works with commercial and in-house codes. Currently interfaced to Fluent, StarCD, StarCCM+, ICEM-CFD, Plot3D, Nastran, Ansys, Abaqus, Fieldview, CRUNCH, STL, IGES, STEP, and many more file formats. (2) Innovative advanced tools in a Graphical User Interface that is easy to learn and use to create the ASD volumes for any shape including irregular shapes and topologies. 4 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 (3) Create user-defined shape change parameters. Group control points together into shape change variables where each control point is defined to move in its own directions. Any shape change desired can be defined to translate, rotate, or scale, the geometry/mesh shape. (4) Fine local control of deformation for subtle shape changes. Often the best shape change for a solution that improves the performance is small and subtle, thus the need for highly accurate smooth shape deformations. (5) Large deformation capabilities for useful conceptual design exploration. (6) Ability to maintain high mesh quality for viscous surface mesh thickness and shape. (7) Ability to perform rigid body deformations of components within the flow domain mesh. (Useful for such problems as wing flap rotation, valve movement, etc.) (8) Perform deformation in real-time without the need to revisit CAD or remesh. (9) Calculate and monitor the computational mesh quality in real-time during the deformation process, and use the mesh quality measurements as constraints in the optimization process. (10) Deform multiple CFD and/or FEA grids simultaneously. (11) Optimize with a built-in Gradient-based Optimization Algorithm (GRG - Generalized Reduced Gradient) or Design of Experiments with Optimal Latin Hypercube and Response Surface Methodology. (12) Use in batch-mode execution for use with external optimization tools such as Simulia iSIGHT, ModeFrontier, Phoenix Integration, Opitmus, DAKOTA, Matlab, etc.. (13) Take the new shape back to CAD. Apply the optimal shape deformation of the mesh to the CAD model via IGES, STEP, ACIS, ParaSolids file formats. The user can also use Sculptor interactively to deform a CAD model directly as if it were clay in his hands. Sculptor is used worldwide by design engineers in aerospace, aircraft, marine, turbo machinery, biomedical, energy-oil/gas/coal, for fluid dynamics, heat transfer, chemical reacting flow, structural, electro-magnetics, etc. For the problem at hand, a biomass/coal combustion burner, we have chosen to modify the shape of the geometry to effect the change of the step height, the quarl slant, the quarl shape, and the angle of the swirl vanes. For all but the swirl vanes, a cylindrical ASD volume is created with 4 redial control points, ten control points around the circumference, and six along the length of the inlet portion of the burner (see Figure 4). The swirl vanes will be handled with different parametrically rectangular ASD volume. This model is a 72 degree section of the burner. 5 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 Figure 4 - A 4x10x6 cylindrical ASD volume with control points arranged in a fashion to change the step height, quarl slope, and quarl shape Figure 5 - A 4x10x6 cylindrical ASD volume with control points arranged in a fashion to change the step height, quarl slope, and quarl shape 6 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 Figure 6 - Close-up view of the undeformed inlet region of the burner Figure 7 - The height of the step is modified by moving the control points at the outer step edge in the radial direction. This particular deformation represents a height change of 20% (original height = 0.00783m modified height = .009396m). This also represents a height change of 20% (original height = 0.00783m modified height = .009396m). This also represents a decrease in the slope of the quarl 7 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 Figure 8 - The slope of the quarl is increased by a radial expansion of the control points located at the intersection of the quarl and the main burner geometry. This particular deformation represents a change the slope from 20° to 25° keeping the original step size Figure 9 - The shape of the quarl is changed by a radial expansion of the control points located between the step and the wall of the combustion chamber. The result of this deformation is a quarl with a concave inward shape 8 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 BIOMASS/COAL COMBUSTION Coal has been a primary choice amongst fossil fuels, which are a primary source of copious amounts of energy. Owing to combustible content and a high calorific value (HHV: 25-30 MJ/kg) [1], the widespread use of coal combustion plants has provided a reliable supply of energy to industry. With depleting coal and other fossil fuel reserves, it is imperative to switch to renewable sources of energy that may continue to provide energy of a similar magnitude by combustion processes. Biomass, being renewable matter, significantly calorific (HHV: 10-20 MJ/kg) [1], and easily available geographically, has been a prime alternative for coal as feedstock. With compositions similar to that of coal due to similar formative processes, the goal is now to design and optimize biomass firing techniques, or co-firing biomass with coal, in order to reduce the dependence of combustion processes on coal and ultimately replace coal altogether. In light of progressive feedstock, co-fired units have taken precedence as well. When appropriate proportions of biomass and coal are used as co-fire feedstock to an optimized burner design, SOx, NOx, and greenhouse gas emissions decrease [2]. With demonstrated techniques in co-firing [3], significant cost savings of power generated from combustion of biomass alone at $80,000/year to $400,000/year from co-fired units [4]. As a matter of energy security, it is important to develop sustainable energy supply processes such as combustion from renewable sources such as biomass, while utilizing advances in computational and numerical techniques. The primary goal of this paper is to show how Sculptor® from Optimal Solutions Software (OSS) is particularly useful for CFD shape optimization where chemically reacting flows and heat transfer is important in the design of combustion devices. Based on numerous coal and biomass compositions, it is seen in general, that coal has a higher C content and lower O content than biomass [1,3]. Accordingly, the contribution from volatile matter is much higher in biomass sources. The mass fraction of volatile content plays a pivotal role in the combustion mechanism: devolatilization, char oxidation and gas phase reactions. Devolatilization dominates the reaction process in the initial injection period, separating the volatile content from the char. Based on the fuel composition, the volatile content further breaks up into H2, CO, SO2 and N2. Due to the higher C and lower O2 content of coal, there is generally C(g) released in the devolatilization step. For biomass however, the higher O2 content ensures complete oxidation of C, releasing CO, and based on stoichiometry, a surplus of O2 which contributes to gas phase reactions that follow. The oxidation of chars further generates CO and H2. These two steps entail a net release of energy, which sustains devolatilization of the incoming fuel and consequently, the combustion mechanism. The reactions and kinetics are described as follows: 9 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 Two-Step Devolatilization: Biomass  (1-Yl) x Charl + Yl x Volatile (Low Temperature) (1) Biomass  (1-Yh) x Charh + Yh x Volatile (High Temperature) (2) Volatile Breakup Reaction : C0.4273H1.142O0.4634N0.003  0.57105 H2 + 0.4273 CO + 0.01805 O2 + 0.0016 N2 Char Oxidation Reactions: C +1/2 O2  CO (3) C + CO2  2 CO (4) C + H2O  CO + H2 (5) Gas Phase reactions: H2 + 1/2 O2  H2O (6) CO + 1/2 O2  CO2 (7) CO + H2O  CO2 + H2 (8) BIOMASS/COAL BASE CASE Based on the reaction scheme described in Eqs. (1) - (8), the kinetic parameters used for modeling biomass combustion are shown below. Table 1 - Devolatilization Kinetic Parameters [5] Yl 0.8 Yh 0.81038 Al (s-1) 370000 Ah (s-1) 1.5 E13 El (J kmol-1) 7.4 E7 Eh (J kmol-1) 2.5 E8 Pressure(MPa) 2 10 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 Table 2 - Kinetic Parameters of Char Oxidation and Gas Phase Reactions [6] Reaction Equation A (s-1) B Ea (J kmol-1) ΔH0(MJ kmol-1) 3 0.002 0 7.9E7 -111 4 242 0 2.75E8 172 5 426 0 3.16E8 131 6 5E10 0 1.68E8 -242 7 2.2E20 0 1.67E6 -283 8 2.6E10 0 8.4E8 -41.1 An illustration of the biomass combustion process is shown in Figure 10 below. Figure 10 - Representative schematic of the biomass combustion process Burner Geometry The Burner Engineering Research Laboratory (BERL) 300 kW burner design was selected for this optimization study. The geometry of the burner is detailed in Figure 11. The configuration is built for natural gas, and this study is unique in modifying the geometry by omitting the radial holes meant for natural gas, and allowing the air swirl from the secondary air stream to enhance the dispersion of biomass particles entrained by the carrier air. This swirling improves mixing, and consequently provides a more uniform temperature distribution near the burner outlet. Based 11 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 on a swirl number of 0.56, a circular pattern of swirl vanes was built in the 3D CAD model to imitate a realistic swirl affected to the air by the physical obstruction the swirl vanes provide. Figure 11 - BERL 300 kW burner geometry As mentioned before, optimizing on the physics of inlet conditions can provide favorable changes in the inherent characteristics of combustion i.e. location of devolatilization zones, flame height, width of flame base and the shape of recirculation zones thus formed. Accordingly, this study aims at setting up a factorial experiment with the following parameters (see Table 3) and studying the response to the ASD brought about by the Sculptor package. STAR-CCM+ v8.04.010 was used to model the combustion phenomena. Table 3 - Optimization Parameters for Initial Designed Experiment Parameter Design Change Considered Quarl Slope Increase slope from 20 to 250 Quarl Design Changed Quarl surface from straight to convex Step Height Increase by 20% in radial direction Design of Experiments and Response Surface Methodology Design of Experiments (DOE) is a strategy that can take a typical design environment consists of a number of analysis tools that are interrelated in some fashion to form a system analysis for the component being designed. The typical goal is to modify the set of design parameters so that an improved component results. Thus, knowledge of the impact of design variables on system 12 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 performance, design feasibility, and product robustness lets you make intelligent design decisions to improve the overall quality of the component. One way to use DOE is to create an approximation of the design surface with an intelligent DOE method such as an Optimal Latin Hypercube. A DOE of this fashion will populate the design space and be fit with a response surface to that the approximate surface can be explored with more efficient algorithms without having to perform an expensive analysis for each design evaluation. This is important especially when performing large expensive CAE (CFD/FEA) analyses. Experimental design, in which a prescribed set of experiments or trials (system analyses) is performed, can be used to study the effects of design parameters on the design states so that intelligent design decisions can be made. Primary considerations in experimental design are as follows: • Number of experiments that can be performed (given cost and time constraints) • Values for the parameters in each experiment • Proper interpretation of the results Although modifying one variable at a time (such as in a parameter study) sometimes leads to an improved design, it can be inefficient and ineffective when interactions among design variables induce unforeseen results. For example, you might try varying numerous inputs independently before one with significant impact is found, or you might even resort to changing many variables simultaneously in an attempt to evaluate the effect of a number of inputs at once. The former strategy cannot account for interaction effects, and the latter strategy makes separation of any effects impossible when done in an informal manner. In a more systematic fashion, a formal Design Of Experiments (DOE) method can define a design matrix specifying the values for the design parameters in each of the experiments. Sculptor provides the tools to perform the necessary formal DOE for the type of problems in this project. Optimal Latin Hypercube An Optimal Latin Hypercube is a modified Latin Hypercube, in which the combination of factor levels for each factor is optimized, rather than randomly combined. With this technique, as with random Latin Hypercubes, the design space for each factor is uniformly divided (the same number of divisions (n) for all factors). These levels are then randomly combined to generate a random Latin Hypercube as the initial DOE design matrix with n points (each level of a factor studies only once). An optimization process is then applied to this initial random Latin Hypercube design matrix. By swapping the order of two factor levels in a column of the matrix, a new matrix is generated and the new overall spacing of points is evaluated. The goal of this optimization 13 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 process is to design a matrix in which the points are spread as evenly as possible within the design space defined by the lower and upper level of each factor. Response Surface Model Response Surface Models (RSM) use polynomials of low order (from 1 to 4), kriging, or radial basis functions, to approximate response of an actual analysis code. A number of exact analyses using the simulation code(s) have to be performed initially to construct a model, or alternatively a data file with a set of analyzed design points can be used. The model then can be used in optimization and sensitivity studies with a very small computational expense, since evaluation only involves calculating the value of a polynomial (or other function) for a given set of input values. Accuracy of the model is highly dependent on the amount of data used for its construction (number of data points), the shape of the exact response function that is approximated, and the volume of the design space in which the model is constructed. In a sufficiently small volume of the design space, any smooth function can be approximated by a quadratic polynomial with good accuracy. For highly non-linear functions, polynomials of 3rd or 4th order can be used. If the model is used outside of the design space where it was constructed, its accuracy is impaired, and refining of the model is required. The problem at hand consists of 4 design variables: step height, quarl slope, quarl shape, and swirl vane angle. Using an optimal Latin Hypercube DOE and knowledge of how much computational resources are available, one decides on how many complete CFD analyses can be run. The implementation of the OLH/DOE in Sculptor will then find the ‘optimal' arrangement of the design sets within the 4-dimentional design space. Then all of the designs are executed. This is an extreme case for easy parallel computing since each design is a separate CFD case. All cases can be run simultaneously if a large enough computing resource such as a large cluster or cloud of cpus is available. Once the cases have been run, the important function values are post-processed and returned to the optimization tool where the best design can be sorted (i.e. design producing the lowest NOx, etc.). Also, a response surface model can be fit with the function data and used as a surrogate model to explore design space further. Design Variables: ASD Control Point Groups The design variables used for this preliminary analysis were defined in Table 3 and included a change in step height, the quarl slant, and the quarl curvature. The angles of the swirl vanes were also investigated, but are not shown here. Although this paper does not include a complete DOE set of design experiments, this work has been performed and will be reported in the presentation. Results completed and presented in this 14 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 paper focused on varying each design variable (step height, quarl slope, quarl curvature). The subsequent presentation includes results of the complete DOE optimization. However, results from the 3 exploratory cases illustrated the impact of changing the burner design and demonstrated how the optimization tool can be applied to combustion equipment. Design Functions: Objective Functions or Design Goals The goal of the design optimization was to find the ‘best' set of design variable values that would improve the burner operation. In this case we were looking at how the design changes affected the flare shape and and size of the Internal Recirculation Zone (IRZ). Ultimately, the optimization will examine directly the impact that design changes have on NOx reduction. DESIGN OPTIMIZATION RESULTS Biomass combustion was initially modeled with no shape changes to establish a base operating condition. The partially-premixed eddy break up combustion and half-order char oxidation models were used to simulate chemical reactions. Gas phase radiation was modeled using the Discrete Ordinates Model (DOM-S4) by specifying an optical path length calculated as the weighted sum of the constituent gray gases. Radiation was not considered for the particle phase, which was assumed to be of uniform size (100 μm). The boundary conditions for the "faces" of the three-dimensional 70° segment were assigned periodic boundary conditions with interfaces. The input conditions shown in Table 4 were used for the base case conditions. The conditions shown in Table 4 were also used to analyze biomass combustion for the three cases with modified geometries. Solutions for each of the cases were obtained on the same mesh for each case. Table 4 - Base Case Input Data Lagrangian Phase Mass flow inlet: 0.025 kg/s Particle temperature at inlet: 550 K Gas Phase Carrier air inlet velocity: 10 m/s Carrier air inlet temperature: 800 K Secondary air inlet velocity: 15 m/s Secondary air inlet temperature: 1100 K Wall Boundary Conditions Furnace wall maintained at 800 K Biomass heat Content: HHV (MJ/kg): 18.75 Biomass Elemental Composition (wt. %, dry): C: 46.78 H: 6.38 N: 0.25 O: 46.59 15 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 CFD predictions for the scalar temperature, streamlined particle temperature and velocity vectors for the input data shown in Table 4 are shown in Figure 12, Figure 13 and Figure 14. Figure 12 - Temperature profiles of biomass combustion from (a) base case (b) curved quarl (c) quarl slope increase from 200 to 250 (d) 20% increase in step height. c) d) a) b) 16 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 Figure 13 - Particle temperature streamlines in biomass combustion from (a) base case (b) curved quarl (c) quarl slope increase from 200 to 250 (d) 20% increase in step height. c) d) a) b) 17 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 Figure 14 - Vectors of velocity showing recirculation zone changes to the right of the flame in biomass combustion from (a) base case (b) curved quarl (c) quarl slope increase from 200 to 250 (d) 20% increase in step height. ANALYSIS OF OPTIMIZED DESIGNS (Vivek/Mark/Joseph) Firstly, it is observed from Figure 12, that the flame temperatures for (b)-(d) are similar compared to (a). There is a difference of approximately 400 K in the peak flame temperature between (a) and the other cases. Further, it is observed that (a) and (d) have similar and shorter flame lengths than (b) and (c), which show longer and again, similar flame lengths. It is proposed that the change in step height by a nominal 20% in (d), is seen to produce a more stable flame c) d) a) b) 18 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 anchored around the region where the first step of devolatilization occurs. A closer look at (d) reveals a ‘second flame', closer to the flow axis, originating from the region where the second step of devolatilization is suspected to occur, based on plots of mass fraction of volatile content. In good theoretical agreement, this phenomenon shows only an approximate 5% reduction in peak flame temperature compared to (b) and (c). Also, the temperature of the bulk gas in the proximity of the flame in (d) appears to be primarily below 1400 K, compared to the other cases, which show temperatures above 1600 K. Secondly, based on the streamlines of particles colored by temperature from Figure 13, it is seen in (d) that the recirculating particle paths arise primarily from backflow along the wall from top to bottom. In comparison with the unmodified geometry in (a), there is a drastic reduction in the stagnancy of particle flow caused by internal recirculation zones (IRZs). It is observed in both (b) and (c), that the IRZ from (a) to the right of the flame, gets elongated and tilts towards the wall, at the top. This may be attributed to the inference from the temperature plots (b) and (c), where the flame is anchored further away from the burner than those in (a) and (d); and this is due to relatively inefficient particle-air mixing, which delays the heating of particles and the devolatilization process. This theory is supported from the velocity vector plots in Figure 14. The base case in Figure 14(a) shows an elliptical IRZ centered to the right of the end of the flame.The geometry deformations in vector plots (b) and (c) are seen to produce an elongation of these IRZs, as mentioned before; coupled with the suggested inefficient mixing, causing the flames (in plots (b) and (c)) to be anchored higher than those in (a) and (d), due to a delayed onset in devolatilization. However, vector plot (d) shows that there is a breakdown of the velocity component tangential to the flame, contributing to more uniform mixing of regions with initially sharply varying temperatures, leading to a more homogeneous temperature distribution at steady state. This, we believe, explains the uniform bulk temperature in the immediate vicinity of the flame in (d). Based on the inferences above, it is suggested that the increased step height before ignition provides a longer draft with a smaller cone angle, allowing the swirling air from the annular inlet to mix with a narrower stream of particle laden air. The streamlines from this region are seen to anchor the flame in Figure 13 (d), producing a narrower flame than Figure 13 (b) and (c); anchored closer to the burner, with a more uniform temperature distribution within the flame vicinity and a smaller IRZ in the bulk fluid. 19 Design Optimization of Combustion Equipment AFRC 2013: Safe and Responsible Development for the 21st Century Sheraton Kauai, Hawaii - September 22 - 25, 2013 CONCLUSIONS AND RECOMMENDATIONS Combustion of biomass in the 300 kW BERL setup was modeled using the eddy break up partially-premixed model. The geometry was subjected to arbitrary shape deformations into three different cases. Of the three, it was seen that the increase in step height by 20% produced a significant reduction in the size and pattern of IRZs. The flame was also seen to be anchored closer to the burner, at the region of the first step of devolatilization of biomass particles, attributed to a probable increase in mixing efficiency with swirling air provided by the longer step size (narrower primary air-particle stream). The longer step appears to produce a smaller area for mixing particle laden air with the swirling secondary air. Although, devolatilization kinetics control the rate of volatile release which impacts local flame temperature near the burner, the increased mixing leads to earlier devolatilization and thus, anchors the flame closer to the burner. Consequently, the dissipation of energy into the bulk gas produces a more uniform temperature distribution near the flame. REFERENCES [1] Tamura, M. and Van de Kamp, W. L. Interim Research Report on the Combustion Characterization of Substitute Fuels for Co-Firing with pulvverized coals during the period 2000-2001. [2] Bradley W. Moulton, PE. Principles of Burner Design for Biomass Co-Firing. 34th International technical Conference on Coal Utilization & Fuel Systems, May 31, 2009. [3] Larry Baxter and Jaap Koppejan. Biomass-coal Co-combustion: Opportunity for Affordable Renewable Energy. [4] Federal energy Management Program (FEMP). Biomass Cofiring in Coal-Fired Boilers. [5] Chen, C., Horio, M., Kojima, T. Numerical simulation of entrained flow coal gasifiers. Part I: modeling of coal gasification in an entrained flow gasifier. Chemical engineering Science 55 (2000), 3861-3874. [6] Chen W-h et al. A comparison of gasification phenomena among raw biomass, torrefied biomass and coal in an entrained-flow reactor. Appl Energy (2013), http://dx.doi.org/10.1016/j.apenergy.2013.01.034. 20