Using CFD to Optimize Flow Distribution in Gas Flare Piping

Paper from the AFRC 2017 conference titled Using CFD to Optimize Flow Distribution in Gas Flare Piping

Design techniques for flare piping which includes the manifold headers, runners, and risers going to the tips themselves has traditionally centered around one-dimensional flow theory combined with empirical pressure-drop correlations. With increasingly complex staging and runner systems being employed in gas flaring systems such as with multi-point ground flares (MPGFs) and totally enclosed ground flares (TEGFs), CFD should play an important role in investigating and optimizing the piping system design.; This paper will provide results from a study where CFD was used to investigate the flow distribution in the original piping of an EGF and an air-assisted MPGF. Variation of flow to tips of less than 10% was desired but the CFD results indicated maximum variations that exceeded the maximum specification. For the MPGF, the flare gas headers tended to exhibit more variation than the air headers. Using only standard piping reducers and standard pipe sizes, modification to the piping system was performed to improve flow distribution uniformity for three different MPGF headers. Maximum variations for the MPGF were reduced to within the specification for the flare design. The EGF piping system was also optimized reducing maximum flow variation. An incompressible RANS version of Fluent was used for the study and was checked against using the compressible version since tip exit velocities approach sonic conditions. Results of the comparison of compressible and incompressible versions of Fluent indicated no measurable difference for this study as most of the piping system is in the incompressible flow regime.

American Flame Research Committees 2017 - INDUSTRIAL COMBUSTION SYMPOSIUM Hyatt Regency Hotel Houston, Texas - September 17 -20, 2017 Using CFD to Optimize Flow Distribution in Gas Flare Piping Joseph D. Smith1, PhD, Missouri University of Science and Technology Vikram Sreedharan2, PhD, Robert Jackson3, Zachary P. Smith4, Elevated Analytics Scot Smith, Doug Allen, Zeeco Inc. Abstract: Design techniques for flare piping which includes the manifold headers, runners, and risers going to the tips themselves has traditionally centered around one-dimensional flow theory combined with empirical pressuredrop correlations. With increasingly complex staging and runner systems being employed in gas flaring systems such as with multi-point ground flares (MPGFs) and totally enclosed ground flares (TEGFs), CFD should play an important role in investigating and optimizing the piping system design. This paper will provide results from a study where CFD was used to investigate the flow distribution in the original piping of an EGF and an air-assisted MPGF. Variation of flow to tips of less than 10% was desired but the CFD results indicated maximum variations that exceeded the maximum specification. For the MPGF, the flare gas headers tended to exhibit more variation than the air headers. Using only standard piping reducers and standard pipe sizes, modification to the piping system was performed to improve flow distribution uniformity for three different MPGF headers. Maximum variations for the MPGF were reduced to within the specification for the flare design. The EGF piping system was also optimized reducing maximum flow variation. An incompressible RANS version of Fluent was used for the study and was checked against using the compressible version since tip exit velocities approach sonic conditions. Results of the comparison of compressible and incompressible versions of Fluent indicated no measurable difference for this study as most of the piping system is in the incompressible flow regime. Introduction and Background: Optimum operation of various kinds of flare fields, especially multipoint ground flares, is dependent on the uniform distribution of flare gas to the different runners (or headers) and tips. This aspect of flare field design can be facilitated using computational fluid dynamics analyses. Presented here are the results for multiple flow distribution simulations of different multipoint flare configurations and a totally enclosed ground flare. In this study, different header size sections were implemented using the CFD tool ANSYS Fluent to ensure uniform flow distribution through the multipoint flare field tips as an iterative process. Also, this same iterative process was implemented for different tip sizes for a totally enclosed ground flare to ensure uniform flow distribution between the different flare tips. 1 Laufer Endowed Energy Chair and Professor of Chemical and Biochemical Engineering, Missouri University of Science and Technology, Rolla, MO; 573-341-4294/ [e-mail] 2 Senior CFD Engineer, Elevated Analytics, Tulsa, OK; 573-201-3321/ [e-mail] 3 Chief Engineer, Elevated Analytics, Tulsa, OK; 801- 691-4663/ [e-mail] 4 CEO, Elevated Analytics, Tulsa, OK; 918-513-2550; [e-mail] Environmental Compliance in Multi-Tip Ground Flares INDUSTRIAL COMBUSTION SYMPOSIUM September 17 -20, 2017 Houston, Texas Physical Model Description: The CFD model for flow within the runners (headers) and risers of the multi-point ground flare (MPGF) was implemented using ANSYS Fluent 16.2 with the geometry and mesh developed with ANSYS Workbench. The three different flare gas runners have different gas compositions labeled appropriately and the base design for the multi-point ground flare is shown in Figure 1. Figure 1: Multipoint ground flare - base case Two different flow conditions, that is, a high-flow and a low-flow case were analyzed for the MPGF-A and MPGF-C gas compositions, whereas only a high-flow case was analyzed for the MPGF-B gas composition. The flow distribution within the runners and risers of a totally enclosed ground flare (TEGF) were modeled for the base case again using ANSYS Fluent 16.2 for CFD modeling and ANSYS Workbench for the model mesh. The exact tip geometry is not included in the CFD model, instead a simplified tip model is placed at each tip location. The eleven stages of the TEGF are all modeled as open in the CFD model. The schematic/layout for the TEGF is as shown in Figure 2. Page 2 of 10 Environmental Compliance in Multi-Tip Ground Flares INDUSTRIAL COMBUSTION SYMPOSIUM September 17 -20, 2017 Houston, Texas Figure 2: Totally enclosed ground flare - base case Flow and Boundary Conditions: The single-phase flow model in ANSYS Fluent was used for this CFD study. This single-phase Eulerian formulation consists of conservation equations for mass, momentum and energy for the bulk phase (Versteeg and Malalasekara, 2007). The standard k-epsilon turbulence model was used to model turbulent flow (and turbulent mixing) for the model. A simplified gas chemistry with estimated fluid properties such as density, viscosity and thermal properties for a simplified gas composition. Table 1: High-flow conditions for the multi-point ground flare base case Table 2: Low-flow conditions for the MPGF base case The flow conditions for the multi-point ground flare are shown in Table 1 for the high-flow conditions and Table 2 for the low-flow conditions. As can be seen from Table 2, the MPGF-B gas header has been run Page 3 of 10 Environmental Compliance in Multi-Tip Ground Flares INDUSTRIAL COMBUSTION SYMPOSIUM September 17 -20, 2017 Houston, Texas with just a high-flow condition (no low-flow case). The "Ref. Case" mentioned in Table 1 is displayed in Table 3, and those in Table 2 are displayed in Table 4. The actual gas composition for the high-flow cases are shown in Table 3 and the simplified composition is displayed in Table 4. The actual gas composition for the high-flow cases are shown in Table 5 and the simplified composition is displayed in Table 6. Table 3: Flow conditions and gas composition for high-flow base cases for multi-point ground flares Table 4: Simplified gas composition for the multi-point ground flare high-flow base cases Page 4 of 10 Environmental Compliance in Multi-Tip Ground Flares INDUSTRIAL COMBUSTION SYMPOSIUM September 17 -20, 2017 Houston, Texas Table 5: Flow conditions and gas composition for the multi-point ground flare low-flow base cases Table 6: Simplified gas composition for the multi-point ground flare low-flow base case Page 5 of 10 Environmental Compliance in Multi-Tip Ground Flares INDUSTRIAL COMBUSTION SYMPOSIUM September 17 -20, 2017 Houston, Texas Results and Discussion: Computing resources at Elevated Analytics in Tulsa (OK) were utilized for this purpose. This consisted of two Intel Xeon CPUs (2 x 16 processors) running at 2.1GHz with 128GB RAM. Typically, each run met the convergence criteria within 1 to 2 hours of CPU time for the multi-point ground flare runs and within 8-10 hours for the TEGF runs when all 32 processors were utilized. The analyses consisted of different CFD calculations including (a) base case designs, (b) Seven high-flow cases and one low-flow case for the TO gas composition, (c) One high- and low-flow case for multi-point flare A, and (d) one high-flow case for multi-point flare B. (a) Base Case Designs As shown below, Figure 3 displays the total pressure contour in the runners and risers of the Case A highflow base case design in MPGF-A. The smaller diameter runner here is for the gas flow and the larger diameter runner is for the air flow. Tip 1 Tip 14 Figure 3 - Pressure contours for High-Flow Base Case for MPGF-A (Case A) From Figure 3, the maximum total pressure is observed at the inlet to the flare gas runner with a gradual reduction in this total pressure as the flow approaches the gas risers. Page 6 of 10 Environmental Compliance in Multi-Tip Ground Flares INDUSTRIAL COMBUSTION SYMPOSIUM September 17 -20, 2017 Houston, Texas Figure 4: Velocity contours for High-Flow Base Case for MPGF-A (Case A) Figure 4 displays the velocity contours in the runners and risers of the MPGF-A high-flow base case design (Case A). It is clearly observed that both the flare gas and air velocities are higher close to the inlet end of both risers with a gradual reduction in velocity as the fluid flows to the other end of the risers. Figure 5 displays the total pressure contour in the runners and risers of the MPGF-A low-flow base case design (Case K). Figure 5: Pressure contours for Low-Flow Base Case for MPGF-A (Case K) Page 7 of 10 Environmental Compliance in Multi-Tip Ground Flares INDUSTRIAL COMBUSTION SYMPOSIUM September 17 -20, 2017 Houston, Texas Figure 6: Velocity contours for Low-Flow Base Case for MPGF-A (Case K) Figure 6 displays the velocity contours in the runners and risers of the multi-point ground flare A low-flow base case design. Both the flare gas and air velocities are higher close to the inlet end of both risers with a gradual velocity reduction as the fluid flow approaches the other end of the risers. With the tip number ordering from Figure 3, a summary of the high-flow and low-flow case results are displayed in the Table 7 below. Table 7: Summary of the MPGF-A base case design high-flow and low-flow results Page 8 of 10 Environmental Compliance in Multi-Tip Ground Flares INDUSTRIAL COMBUSTION SYMPOSIUM September 17 -20, 2017 Houston, Texas Figure 7: Total Pressure Contours (Pa) for MPGF-B High-Flow Base Case (Case E) Figure 7 displays the total pressure contours in the runners and risers of the MPGF-B high-flow base case design (Case E). Higher total pressure is observed in the core of the gas and air flows close to their respective runner inlets. Figure 8 displays the velocity contours in the runners and risers of the same flare for the high-flow base case design (Case E). The flare gas velocities are higher close to the riser elbows for the tips in proximity to the air inlet. Figure 8: Velocity Contours (m/s) for MPGF-B High-Flow Base Case (Case E) Page 9 of 10 Environmental Compliance in Multi-Tip Ground Flares INDUSTRIAL COMBUSTION SYMPOSIUM September 17 -20, 2017 Houston, Texas Conclusions: Starting with a base design case, the fluid flow within the gas flare piping system for a MPGF and TEGF was analyzed and visualized using CFD. It was discovered that the flow distribution was non-uniform and did not satisfy the customer requirement. Elevated Analytics' engineers developed a solution, and the use of flow reducers were simulated and incorporated the for multi-point ground flare system, and different effective flare tip areas were calculated using CFD for the TEGF. This simulation process was iterative, and more than a few different configurations were simulated using the CFD model before a satisfactory design was selected. References: [1] Versteeg, H. and Malalasekara, W. (2007). An Introduction to Computational Fluid Dynamics: The Finite Volume Method (2nd Edition), Prentice Hall. Page 10 of 10