Purge rates: husa compared to CFD

Purge Rates: Husa Compared to CFD Larry D. Berg, Joseph D. Smith, Ph.D. and Ahti Suo-Anttila, Ph.D. Alion Science and Technology Owasso, OK Scot K. Smith, P.E. and Jay B. Modi Zeeco, Inc. Broken Arrow, OK OUTLINE • • • • • Slide 2 Review of Husa Method Review of Purge Practice Husa vs. CFD - High Purge Rate Husa vs. CFD - Range of Rates Conclusion 2006 AFRC International Symposium Review of Husa Method • Howard W. Husa - Two papers • 1964 Hydrocarbon Processing • 1977 Fall API Fire Safety Meeting • Range of rates, gasses and diameters (4" - 48") • Single Correlation for all experimental results Slide 3 2006 AFRC International Symposium Review of Husa Method n ⎤ ⎡ ⎛ ⎞ 1 20 . 9 0.65 3.46 ⎟⎟⎥ ∑ C i Q = .07068 D ⎢ ln⎜⎜ e (.065(29 MW )) ⎣ y ⎝ O2 ⎠ ⎦ 1 • • • • • • Slide 4 Q = purge rate (ft3 / hr) D = Internal diameter of flare stack (inches) y = depth down flare stack from exit plane (ft) O2 = Oxygen concentration (mole fraction) Ci = Mole Fraction of Purge Gas type MW = Molecular Weight of Purge Gas Type 2006 AFRC International Symposium Review of Husa Method • Incorrect MW Functionality As Published 21 18 15 12 Series1 9 6 3 0 0 0.25 0.5 0.75 1 M W o f Pur g e g as Slide 5 2006 AFRC International Symposium Review of Husa Method • From data published with paper, deduce actual • functional form: As Published: n ⎡ ⎤ ⎛ ⎞ 1 20 . 9 0.65 3.46 (.065 ( 29 MW )) ⎜ ⎟ Q = .07068 D ⎢ ln⎜ e ⎥ ∑ Ci ⎟ ⎣ y ⎝ O2 ⎠ ⎦ 1 • "Deduced" Equation: n ⎡ ⎤ ⎛ ⎞ 1 20 . 9 ⎟⎟⎥ ∑ C i 0.65 6.586 ⋅ e (− .065 Q = .07068 D 3.46 ⎢ ln⎜⎜ ⎣ y ⎝ O2 ⎠ ⎦ 1 Slide 6 MW ) 2006 AFRC International Symposium Review of Husa Method • "Deduced" Equation: n ⎤ ⎡ ⎛ ⎞ 1 20 . 9 0.65 3.46 ( − .065 ⎜ ⎟ Q = .07068 D ⎢ ln⎜ C 6 . 586 ⋅ e ⎥∑ i ⎟ ⎣ y ⎝ O2 ⎠ ⎦ 1 MW ) • Husa "K" factors for gas Q = .07068 D Slide 7 3.46 ⎡ 1 ⎛ 20.9 ⎞⎤ ⎟⎟⎥ K ⎢ ln⎜⎜ ⎣ y ⎝ O2 ⎠⎦ 2006 AFRC International Symposium Review of Purge Practice • Husa - Interested in Prevention of Stack Explosions • Simple "pipe" or "utility" flares Slide 8 2006 AFRC International Symposium Review of Purge Practice • Since Husa (30-40 years ago) Flares have changed Slide 9 2006 AFRC International Symposium Review of Purge Practice Multi-Point Ground Flares Enclosed Flare Slide 10 2006 AFRC International Symposium Review of Purge Practice Multi-Injection Steam Flares Not only has the equipment changed, but a secondary purpose has been added to explosion prevention: "Protection of complex combustion Equipment" Slide 11 2006 AFRC International Symposium Review of Purge Practice • Equipment did not appear overnight • Vendors have developed proprietary purge methods • CFD can provide an independent verification Slide 12 2006 AFRC International Symposium Husa vs. CFD - High Purge Rate • Customer wanted to ensure minimal air ingression • Wanted results 1'-0" from stack exit • Purge was methane, ethane, propane mix • Steam flares included "cooling" steam Slide 13 2006 AFRC International Symposium Husa vs. CFD - High Purge Rate Slide 14 Tip Husa O2 Concentration CFD (FLUENT) 58" Utility 15.2% 1.1% 34" Utility 15.3% 0.8% 66" Utility 16.2% 0.2% 56" Steam 15.9% 3.1% 46" Steam 17.5% 3.4% 2006 AFRC International Symposium Husa vs. CFD - High Purge Rate • Husa greatly over predicts O2 concentration • Husa does not include combustion • Not very interested in 1'-0" from tip exit Slide 15 2006 AFRC International Symposium Husa vs. CFD - High Purge Rate Temperature ˚C Temperature Contours Oxygen Contours O2 Mole Frac Wind = 3 m/s Wind = 8.9 m/s Wind = 18 m/s Utility Tip Slide 16 2006 AFRC International Symposium Husa vs. CFD - High Purge Rate Temperature ˚C O2 Mole Frac Temperature Contours Oxygen Contours Wind = 3 m/s Wind = 8.9 m/s Wind = 18 m/s Multi-Injection Steam Tip Slide 17 2006 AFRC International Symposium Husa vs. CFD - Range of Purge Rates • • • • Slide 18 More reasonable to compare: Range of purge rates Variation of depth No combustion 2006 AFRC International Symposium Husa vs. CFD - Range of Purge Rates • • • • • Slide 19 Range of purge rates: 0.05, 0.25, 1.00 ft/s Variation of depth: 1', 8', 15' Methane only No combustion Use steady state and transient CFD for utility tip 2006 AFRC International Symposium Husa vs. CFD - Range of Purge Rates Husa vrs. CFD - 1.0 ft/sec 34" Utility Flare Oxygen Concentration 15.00% 12.00% Husa 9.00% CFD - ISIS 6.00% CFD-FLUENT 3.00% 0.00% 0 5 10 15 Depth from Tip Exit (ft) Slide 20 2006 AFRC International Symposium Husa vs. CFD - Range of Purge Rates Husa vrs. CFD - .05 ft/sec 34" Utility Flare 21.00% 21.00% 18.00% 18.00% 15.00% Husa 12.00% CFD - ISIS 9.00% CFD -FLUENT 6.00% 3.00% 15.00% Husa 12.00% CFD - ISIS 9.00% CFD - k-e 6.00% 3.00% 0.00% 0.00% 0 5 10 Depth from Tip Exit (ft) Slide 21 Oxygen Concentration Oxygen Concentration Husa vrs. CFD - .25 ft/sec 34" Utility Flare 15 0 5 10 15 Depth from Tip Exit (ft) 2006 AFRC International Symposium Husa vs. CFD - Range of Purge Rates • At high rates (1.0, 0.25) - Husa, steady state CFD, and transient CFD reasonable agreement • At low rate, steady state analysis "OK", but transient much better Slide 22 2006 AFRC International Symposium Husa vs. CFD - Range of Purge Rates Movie of transient prediction for 50% O2 Slide 23 2006 AFRC International Symposium Husa vs. CFD - Range of Purge Rates Slide 24 2006 AFRC International Symposium Husa vs. CFD Multi-Tube Steam Flares Typical Multi-Tube (HCSX) Steam Flare Tip Slide 25 2006 AFRC International Symposium Husa vs. CFD Multi-Tube Steam Flares For Husa need: • Velocity • Diameter Typical Multi-Tube (HCSX) Steam Flare Tip Slide 26 2006 AFRC International Symposium Husa vs. CFD Multi-Tube Steam Flares D Which Diameter ? D Slide 27 2006 AFRC International Symposium Husa vs. CFD Multi-Tube Steam Flares Which Velocity ? Vtip Vriser Slide 28 2006 AFRC International Symposium Husa vs. CFD Multi-Tube Steam Flares 1.0 ft/sec & Center Steam Husa vrs. CFD - 1.00 ft/sec 46" Mult-Tube Steam Flare Riser Diamter / Riser Velocity 21.00% 21.00% 18.00% 18.00% 15.00% 12.00% Husa CFD - k-e 9.00% 6.00% 3.00% 0.00% Oxygen Concentration Oxygen Concentration Husa vrs. CFD - 1.00 ft/sec 46" Mult-Tube Steam Flare Measured Diamter / Tip Velocity 15.00% 12.00% Husa CFD - k-e 9.00% 6.00% 3.00% 0.00% 0 5 10 Depth from Tip Exit (ft) 15 0 5 10 15 Depth from Tip Exit (ft) Tip Diameter & Tip Velocity have Best Correspondence Slide 29 2006 AFRC International Symposium Husa vs. CFD Multi-Tube Steam Flares 0.05 ft/sec & Center Steam Husa vrs. CFD - .05 ft/sec 46" Mult-Tube Steam Flare Riser Diameter / Riser Velocity 21.00% 21.00% 18.00% 18.00% 15.00% 12.00% Husa CFD - k-e 9.00% 6.00% 3.00% Oxygen Concentration Oxygen Concentration Husa vrs. CFD - .05 ft/sec 46" Mult-Tube Steam Flare Tip Diameter / Tip Velocity 15.00% 12.00% Husa CFD - k-e 9.00% 6.00% 3.00% 0.00% 0.00% 0 5 10 Depth from Tip Exit (ft) 15 0 5 10 15 Depth from Tip Exit (ft) Riser Diameter & Riser Velocity have Best Correspondence Slide 30 2006 AFRC International Symposium Husa vs. CFD Multi-Tube Steam Flares Suggest: • For stack explosion analysis, use riser velocity and riser diameter • For combustion equipment analysis, use tip velocity and tip diameter Slide 31 2006 AFRC International Symposium Summary and Conclusions • Husa recommended empirical correlation to estimate • • • • Slide 32 required purge rate for "simple" gas flares Gas flare equipment changed significantly (became more complex) since Husa first analyzed purge requirements Compared Husa purge estimate to estimates based on Fluent and ISIS CFD codes Recommend approach to purge estimate based on what's needed (i.e., stack combustion analysis vs. tip analysis) Discussion 2006 AFRC International Symposium