CFD analysis of liquid seal design

CFD Analysis of Liquid Seal Design Lawrence (Larry) D. Berg, PE and Joseph D. Smith, Ph.D. Systems Analyses and Solutions (SAS) Tulsa, OK Scot K. Smith, P.E. and Doug Wade Zeeco, Inc. Broken Arrow, OK Fall 2011 - AFRC Meeting Liquid Seal Analysis •  Outline -  Basic Design -  Process Conditions / Flow Conditions Fall 2011 - AFRC Meeting General Arrangement 94 Perforated Plate ends with Impingement plate. Support is provided by (4) 3x3 angle 41 Top of Teeth NLL Bottom of Teeth Impingement plate fabricated from perforated plate Perforated Plate extends to shell Fall 2011 - AFRC Meeting Slide 4 Methodology: Computational Grid Structure Overall Domain Internals Approximately 500,00 cells Supply Pipe Supply Pipe End Perforated Plates Fall 2011 - AFRC Meeting Slide 5 Process Conditions Temp   Inlet  Total  Pres   Pa  gauge   MW   (°C)   Flow  Rate¥   (kg/hr)   Design  1   24.15   59   225,667   821,660   Design  2   22.2   55   311,645   1,452,620   Design  3   22.2   40   341,637   1,666,343   Liquid Seal Max Flowv Break Flow † Min Flow Design 1 821,660 kg/hr @ 225,667 Pa gauge TDB kg/hr @ TBD Pa 6,985.85 kg/hr @ 15 Pa gauge Design 3 1,452,620 kg/hr @ 311,645 Pa gauge TDB kg/hr @ TBD Pa 6,985.85 kg/hr @ 145 Pa gauge Design 3 1,666,343 kg/hr @ 341,637 Pa gauge TDB kg/hr @ TBD Pa 6,985.85 kg/hr @ 137 Pa gauge Fall 2011 - AFRC Meeting Model Assumptions •  Standard k-ε turbulence model throughout •  Ideal gas equation to determine state of gas •  Volume of Fluid sub-model (VOF) employed to simulate liquid / gas interaction •  Outlet pressure Boundary Condition defined; CFD code calculates pressure drop through system; Outlet BC varied to achieve specified inlet pressure condition Fall 2011 - AFRC Meeting Total Pressure (psig) Max Flow Results - Typical Fall 2011 - AFRC Meeting Max Flow Pressures D C A B Location Total P A 46.6 B 45.5 1.1 C 44.2 1.3 D 43.0 1.2 Overall ΔP (A - D) Total Δ psig psig 3.6 Summary of Pressure Losses (PSIG) Fall 2011 - AFRC Meeting Max Flow Results •  Max flow pressure loss is 3.6 psig •  Specification calls for pressure loss of 2.0 psig •  Use CFD to make design changes Fall 2011 - AFRC Meeting Design Changes Perforated Plate increased open area Perforated Plate ends where Impingement plate was located. Impingement plate removed 1 Fall 2011 - AFRC Meeting Final Pressure - Max Flow Ave Total Pressure @ Outlet = 31.7 psig Total Pressure (psig) Ave Total Pressure @ Inlet = 33.7 psig Total Pressure ΔP = 2.0 psig Fall 2011 - AFRC Meeting Minimum Flows Fall 2011 - AFRC Meeting Minimum Flow Simulation achieves Steady State conditions when pressure remains constant with time Pressure Increases to Seal Depth and then levels off (slight increase above seal pressure due to gas bubbling through seal) Calculated Pressure at Seal Inlet Fall 2011 - AFRC Meeting Flare Gas Volume Fraction Minimum Flow 50% Flare Gas Volume Fraction Iso-Surface Fall 2011 - AFRC Meeting Break Flow Analysis •  Flow not determined •  Customer not certain of definition •  Proposed: -  When flow is sufficient to create continuous or "unbroken" path of gas through liquid seal •  Client accepted - type of condition to cause surging Fall 2011 - AFRC Meeting Methodology: Break Flow Estimation •  Simplify geometry to a 2D axi-symmetric model (symmetric about axis of rotation) to establish the break flow for each seal design •  Verify flow with 3D analysis - check for surging •  Approach provides accurate and efficient methodology to determine break flow and provides insight into the multiphase gas/fluid flow phenomena inside the liquid seals Fall 2011 - AFRC Meeting Methodology: Break-flow Calculation Unbroken Path Broken Path Non-continuous gas flow caused by liquid surface interrupting gas flow Fall 2011 - AFRC Meeting Typical 2D Results Fall 2011 - AFRC Meeting Results - Break Flow Determination •  Two types of flow blockage occurs: -  Water blocks flow "at the teeth" -  Water blocks flow "Downstream" of teeth Water blocks gas at pipe exit or at the teeth Water blocks gas flow downstream of teeth Volume Fraction Of Flare Gas Teeth Blockage Downstream Blockage Fall 2011 - AFRC Meeting Results - Break Flow Determination •  Downstream blockage -  Occurs very rapidly -  Induces small pressure change •  Teeth blockage -  Occurs over longer period of time -  Induces large pressure fluctuations •  Look for minimal Teeth blockage events to identify Break Flow rate 03/24/2010 2011 -Seal AFRCAnalysis Meeting Habshan Fall Liquid Results - Break Flow Determination Inlet Pressure Graphs Design 1 Simulation Inlet Static Pressure Fluctuations Mass Flow = 80 lbm/s Fall 2011 - AFRC Meeting Pressure change of about 1 psig caused by teeth blockage Results - Break Flow Determination Inlet Pressure Graphs Design 1 Simulations Mass Flow = 80 lbm/s Pressure fluctuations nearly eliminated in 150 lbm/sec flow Mass Flow = 125 lbm/s Inlet Static Pressure Fluctuations Fall 2011 - AFRC Meeting Mass Flow = 150 lbm/s •  Change Break Flow Definition to No Blockage of Teeth Region •  Use 150 lbm/sec as input to 3D model •  Ensure continuous flow path Fall 2011 - AFRC Meeting Results - Break Flow Fall 2011 - AFRC Meeting Design 1 - Break Flow % volume fraction 90% or greater gas region completely encases supply pipe until exiting past water seal Normal Liquid Level Outer Perforated Plate Iso-Surface of 90% Volume Fraction of Flare Gas Fall 2011 - AFRC Meeting % volume fraction Design 1 - Break Flow Thin film of 100% Flare Gas surrounds supply pipe up to Normal Liquid Level Normal Liquid Level Iso-Surface of 100% Volume Fraction of Flare Gas Fall 2011 - AFRC Meeting Results - Break Flow Determination Habshan 1,2 Flow Rate •  Seal water nominally outside outer perforated plate •  Design 1 and 2 have similar seal depths •  Velocity in annular space Expect Break flows to have similar velocities through annular space Supply Pipe Outer Perforated Plate Fall 2011 - AFRC Meeting Results - Break Flow Determination † Flow 2 = 227.9 Use 230 lbs/s flow for Design 2 Fall 2011 - AFRC Meeting Break Flow Analysis Fall 2011 - AFRC Meeting Design 2 % volume fraction Iso-Surfaces of Volume Fraction of Flare Gas through liquid seal 90% Gas Region Around Supply Pipe Fall 2011 - AFRC Meeting 99% Gas Region Around Supply Pipe •  Minimum flow rate analysis -  No abnormal flows observed -  Bubbling through teeth observed (as expected) •  Total pressure loss reduced to 2.0 psig max. •  Break flow calculated using 2-D axi-symmetric model to identify flow when gas has free path through seal at least 90% of time. Full 3-d simulation used to validate 2-d results -  Design 1 break flow rate = 150 lbm/s -  Design 2 break flow rate = 230 lbm/s Fall 2011 - AFRC Meeting