Technical foundations to establish new criteria for efficient operation of industrial steam-assisted gas flares - presentation

TECHNICAL FOUNDATIONS TO ESTABLISH NEW CRITERIA FOR EFFICIENT OPERATION OF INDUSTRIAL STEAM-ASSISTED GAS FLARES L. Douglas Smoot, Ph.D., Combustion Resources, Inc. Joseph D. Smith, Ph.D., Idaho National Laboratory Robert E. Jackson, Combustion Resources IFRF 16th International Members' Conference Combustion and Sustainability: New Technologies, New Fuels, New Challenges 8-10 June, 2009 - Boston Massachusetts, USA Slide 2 Discussion Outline • Basic Flare Operation and Control - Operating/design issues for Pipe Flares and Ground Flares • Historical Performance Analysis - EPA and CMA Data - API Guidelines • Proposed Metrics for Flare Operation - Adiabatic Temperature vs. LHV - Combustion Efficiency vs. Steam Ratio • BAT for Flare Performance Analysis • Conclusions and Recommendations IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 3 Flares in Recent News! Decades of contamination of the water and soil from oil and gas operations mean most food must now be imported, says JNN, and the practice of "gas flaring" has put a toxic pall over many villages. WILLEMSTAD (Reuters) - A refinery on Curacao operated by Venezuela's state oil company is damaging people's health and must cut emissions or face multi-million dollar fines, a court on the Caribbean island ruled on Thursday. In 2007, a Curacao court threatened to close Isla if it cannot meet emissions standards, citing a study estimating that 18 people die prematurely every year from contaminant exposure. Van Unen also took into account excessive flaring that sometimes runs for days and weeks. The amounts to be paid were decided with the serious health consequences and the earnings of PDVSA in mind, the judge stated. Industrial flares burn off pressurized gases but also can shoot out massive amounts of noxious emissions. The Houston area has about 400 flare stacks, and they are among the largest and least- understood sources of pollution in the region, researchers said. IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 4 Operating/Design Issues for Pipe Flares Wind shortens Flame and causes soot formation on back side of flame Flame bends over in wind and licks downwind side of flare stack Air egression into stack tip can lead to internal burning and possible explosive conditions Cross winds cause air egression into flare stack P. Gogolek, CANMET Energy Technology Centre - Ottawa Natural Resources Canada IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Operating/Design Issues for Low Profile Flares High tip velocity increases air entrainment o Tip design controls air entrainment (fuel/air mixing) o High jet velocities = High Jet Noise Pressure Assist improves mixing and combustion o Slide 5 Smoke below certain tip pressure (D-stage pressure) = ambient air entrainment Tip spacing critical o Tips must cross light o Individual tip flames can merge and lengthen overall flare flame o Adjacent rows compete for ambient air (longer flames near flare center) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 6 What is a Flare or Why is there a flame on top of that stack? Flares are Safety Devices used to prevent catastrophic events in systems with highly flammable gases Flares provide safe discharge point for relief devices or in case of loss of containment API RP-521† describes proper flare design and operation but does not define them as "combustors with routine emissions" (e.g., incinerators or process heaters) Plants required to have "Flare minimization plan" as part of air permit - recovery devices operate with quasi-steady flows but can't operate over large flow ranges possible in hydrocarbon refineries or chemical plants Pollution control devices (i.e., incinerators) combust process emissions Safe and efficient flare design must handle very low hydrocarbon flows (due to fuel cost) and high hydrocarbon flows (due to safety constraints) under variable ambient conditions (i.e., wind and rain) for non-uniform gas compositions †API RP-521, 4th ed., American Petroleum Institute, Washington D.C., March (1997). IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Flare Efficiency Studies (EER) Slide 7 Flare Screening Facility and Flare Test Facility used to analyze flame stability and combustion/destruction efficiency for different fuels and different flare tips (captured/analyzed flare plume) Used 3", 6", and 12" Open Pipe Flares plus Air-assisted, Pressureassisted and Steam-assisted Flares Considered Pilot vs non-pilot operation Considered saturated/unsaturated hydrocarbon, H2S, and NH3 Correlated Flame stability vs Gas heating value and Tip exit velocity o Flame stability defined as Tip velocity > Flame velocity o Assisted flares more stable, Piloted flares more stable, unsaturated fuels more stable (but require more air) Pohl, J.H. & N.R. Soelberg, Evaluation of the Efficiency of Industrial Flares: Flare Head Design and Gas Composition, EPA-600/2-85-106, September (1985) Pohl, J.H. & N.R. Soelberg, Evaluation of the Efficiency of Industrial Flares: H2S Gas Mixtures and Pilot Assisted Flares, EPA-600/2-86-080 September (1986) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 8 Testing on Flame Stability and Heating Value Identified Min HV for Gas where Flame Speed equal Exit Velocity Characterized flows for different fuels and tip sizes Pilot and Assist media Improved Flame Stability Pohl, J.H. & N.R. Soelberg, Evaluation of the Efficiency of Industrial Flares: H2S Gas Mixtures and Pilot Assisted Flares, EPA-600/2-86-080 September (1986) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Flare Efficiency Studies by CMA/EPA/JZ (1983)Ŧ Slide 9 Tested Air-Assisted and Steam-Assisted Flares with emphasis on determination of combustion efficiency and factors that affect it Utilized commercial scale steam and air-assisted flares while burning propylene and nitrogen in varying compositions. Factors considered: o Flow rate of the relief gas, o Heating value of the relief gas, and o Steam/relief gas ratio Showed flares achieve >98% destruction efficiency if properly operated Optimal Combustion efficiency with steam to relief gas mass ratio of 0.4 - 1.5 lbm/lbm Over-steaming appeared to occur for steam/relief gas ratio greater than 4x recommended steam rates Ŧ McDaniel, M., Flare Efficiency Study, EPA-600/2-83-052, July (1983) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Factors Important in Flare Performance Slide 10 Smokeless operation¥ o o o o Black Smoke = Soot (fv< 10-5 Vol Frac, dprim< 40 nm) Smoke burns up in high temperature region (smokes if soot exits this region before being consumed) Smoke forms when C-C bonds in hydrocarbons crack and aromatic structures grow into multi-ring molecules (>3 ring = primary soot particle) Other Poly-aromatic hydrocarbons (PAH) form along reaction route to soot ¥ Moss, J.B., Stewart, C.D., and Young, K.J., "Modeling Soot Formation and Burnout in a High Temperature Laminar Diffusion Flame Burning under Oxygen-Enriched Conditions," Combustion and Flame, 101: 491-500 (1995) Gollahalli and Parthasaronthy, R.P., "Turbulent Smoke Points in a Cross-Wind," Research Testing Services Agreement No. RTSA 3-1-98, University of Oklahoma, Norman, OK, August (1999). IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 11 Factors Important in Flare Performance High Exit velocity: use "momentum ratio" to account for cross wind effect o Flare gas entrains surrounding air to enhance mixing/combustion efficiency o Assist media (steam/air) enhances entrainment o Purge conditions (low momentum ratio) results in flame deflection and poor mixing o Flare gas momentum < 10% wind momentum allows flame stabilization downwind of tip‡ ‡Pohl, J., Gogolek, P., Schwartz, R., and Seebold, J., "The effect of Waste Gas Flow & Composition Steam Assist & Waste Gas Mass Ratio Wind & Waste Gas Momentum Flux Ratio Wind Turbulence Structure on the Combustion Efficiency of Flare Flames" Gogolek, P.E.G., and A.C.S. Hayden, "Efficiency of Flare Flames in Turbulent Crosswind," Advanced Combustion Technologies, Natural Resources Canada, American Flare Research Committee, Spring Meeting, May (2002). IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Factors Important in Flare Performance Slide 12 Flare Gas Heat Content > Minimum Heating Value for Stability o Lower Heat value provides less energy to local reaction zone o Combustion at lower flame temperature o Reaction zone more susceptible to flame shearing (quenching) Pohl, J.H., R. Payne & J. Lee, Evaluation of the Efficiency of Industrial Flares: Test Results, EPA-600/2-84-095, May (1984) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Factors Important in Flare Performance Slide 13 Flare Gas Heat Content > Minimum Heating Value for Stability o Saturated HC's: stable flame at increasing exit velocities for higher gas HV (increased mixing dilutes reactants before combusted - requires more fuel to compensate) o Unsaturated HC's: stable flame at increasing exit velocities for constant/ decreasing gas HV (increased mixing with fast kinetics and reduced O2 demand for unsaturated HC) o Assist media increases stability (enhanced mixing for flare gas with pilots) o Practical operation: increasing C/H ratio (unsaturation) requires more assist media IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 14 Quantifying Flare Performance Using Flammability Ratio and Steam/Fuel Mass Ratio General Straight Chain Saturated (Aliphatic) Hydrocarbons have Heating Values ~20,000 Btu/lbm Flammability Limit represents "mixture" that has sufficient oxidant and fuel to react Other factors affecting "reaction" include: Fuel/Oxidant Mixing (turbulent versus laminar) o Ignition source (create radicals to promote reaction) o o "Stability" (radicals to propagate reaction) Example Calculation: o Propylene (fuel), Nitrogen (purge), Natural Gas (pilots and supplemental fuel), and Steam IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 15 Definitions Used in Proposed Performance Metrics:¥ m = molar, st = stoichiometric, F = Fuel, M = mixture of fuel plus air o Vent gas = Waste Fuel Gas + Supplemental Fuel Gas + Purge gas o Flare gas = Vent gas + Pilot fuel gas + Steam o (FR)m = (F/CZG) / (F/M)st o CZG = Combustion Zone Gas = Flare gas plus added Combustion Air to reach Stoichiometric Oxygen o (F/M)st = Fuel / (Fuel + Stoichiometric Air) (molar basis) o (FR)m = Flammability Ratio (molar basis) ~ % Lean Flammability Limit / % Stoichiometry ¥ Proposed Metrics by Mr. Brian Dickins, EPA Engineer, Chicago, Illinois IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 16 Proposed Performance Metrics for Steam Flares: Standard 1 - Steam/Vent Gas Mass Ratio Limit: • • • • Optimum = manufacturer's recommendation or constant Maximum ≤ 4 times manufacturer's recommended value All flows defined by operating conditions - need Purge Gas flow Purge gas flow rate can be set based on manufacturer value or estimated by Husa criteria: Purge Gas (scfh)‡ = 0.0035 D3.46K (1) where: D = flare stack diameter (inches) K = Purge gas constant for specific gases = 2.33 for methane; = 1.71 for N2 with wind = 1.07 for N2 without wind Standard 2 - Minimum Gas Heat Content Limit: • • If Flare Gas, then LHV ≥ 200 BTU/scf If Vent Gas, then LHV ≥ 300 BTU/scf (2A) (2B) ‡ Berg, L.D., Smith, J.D., Suo-Anttila, A., Price, R., Modi, J., Smith, S., "Flare Purge Rates: Comparison of CFD and Husa", AFRC Symposium, Houston, TX (2006) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 17 Application of Metrics: Required Information Input Parameters: • • • • • Flare diameter (ft) and tip design Waste fuel to be flared, including moisture Wind velocity (ft/hr) Pilot Gas flow rate (scfh) Combustion Air flow into tip zone (lb/hr) Operational Parameters (to be Specified) • Purge gas flow rate (lb/hr) • Supplemental fuel flow rate (lb/hr) • Steam flow rate to tip for air entrainment/mixing (lb/hr) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 18 Example: Calculated Adiabatic Flame Temperature for Selected Steam Flow at constant LHV Vent Steam/Flare Gas Flare Gas Gas Mass LHV LHV ratio 3 3 (lbm/hr) (BTU/ft ) (BTU/ft ) (lbm/lbm) Purge Flow (N2) Steam Flow Fuel Flow (lbm/hr) (lbm/hr) (lbm/hr) 1 2 3 4 5 6 7 8 9 28 28 28 74 74 74 148 148 148 43 208 5554 113 549 14679 227 1099 29359 14.9 55.2 1360.6 39.3 145.8 3595.9 78.6 291.5 7191.7 43 83 1389 113 220 3670 227 440 7340 552 1199 2050 552 1199 2050 552 1199 2050 200 200 200 200 200 200 200 200 200 CMA52 CMA53 2.14 1.41 201 201 0.452 0.226 2.592 1.636 251 197 112 112 Case # Vent Flow 1.00 2.50 4.00 1.00 2.50 4.00 1.00 2.50 4.00 Tip Exit Velocity FRm (ft/min) Adiabatic Flame Temp (ºF) Reynolds Froude No. No. (-) (-) (-) 6.3 10.8 156.0 16.7 28.6 412.3 33.5 57.1 824.5 2893 2856 2843 2893 2856 2843 2893 2856 2843 0.702 0.702 0.702 0.702 0.702 0.702 0.702 0.702 0.702 1005 1716 24760 2656 4536 65438 5311 9071 130876 0.016 0.028 0.397 0.043 0.073 1.049 0.085 0.145 2.098 2330 2323 0.576 0.575 Notes: • Flare Gas LHV = (Fuel / Combustion Zone Gas) / (Molar Stoichiometric Ratio Based on Air) • Steam and Fuel Flows adjusted to achieve Desired Steam/Flare Gas Mass Ratio and Adiabatic Flame Temp calculated for mixture IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 19 o Computed Adiabatic Flame Temperature, F Correlation of Steam to Hydrocarbon Flow vs Adiabatic Flame Temp (Tad) at constant LHV 3000 Propylene 2900 2800 2700 2600 2500 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Steam to Vent Gas Ratio IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 20 Example: Calculated Adiabatic Flame Temperature for Selected LHV at different Flare Operating Conditions Case # 10 11 12 13 14 15 16 17 18 19 CMA52 CMA53 Purge Flow (N2) Steam Flow Fuel Flow (lbm/hr) (lbm/hr) (lbm/hr) 74 74 74 74 74 74 74 74 74 74 2.14 1.41 81 223 397 90 280 594 151 330 814 69 201 201 7.3 15.3 25.1 16.0 38.0 74.4 76.6 145.8 332.8 63.4 0.452 0.226 Vent Gas Flare Gas LHV LHV 3 3 (lbm/hr) (BTU/ft ) (BTU/ft ) Vent Flow 81 89 99 90 112 148 151 220 407 137 2.592 1.636 130 256 390 265 539 847 862 1199 1584 768 251 197 50 50 50 100 100 100 300 300 300 400 112 112 Tip Exit Velocity (ft/min) 13.2 14.1 15.2 14.1 16.6 20.6 20.9 28.6 49.4 19.4 Steam/Flare Gas Mass Adiabatic ratio Flame Temp (ºF) (lbm/lbm) 1.00 2.50 4.00 1.00 2.50 4.00 1.00 1.50 2.00 0.50 1619 1586 1574 2283 2243 2229 3165 3150 3140 3339 2330 2323 FRm Reynolds Froude No. No. (-) (-) (-) 0.353 0.353 0.353 0.528 0.528 0.528 0.789 0.789 0.789 0.841 0.576 0.575 2091 2233 2406 2244 2633 3275 3314 4536 7838 3082 0.034 0.036 0.039 0.036 0.042 0.053 0.053 0.073 0.126 0.049 Notes: • Vent Gas adjusted to achieve desired (LHV)f level and Adiabatic Flame Temp calculated for mixture • Flare Gas LHV = (Fuel/Combustion Zone Gas) / (Molar Stoichiometric Ratio Based on Air) • (LHV)f includes Fuel + Steam + Purge • CMA data with Low LHV included for comparison purposes IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 21 Correlation of Vent Gas LHV versus Adiabatic Flame Temp (Tad ) for Selected Flare Operating Conditions o Adiabatic Flame Temperature (F) 5000 Propylene 4500 4000 3500 3000 2500 2000 1500 1000 500 0 500 1000 1500 2000 2500 Vent Gas Lower Heating Value (BTU/scf) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 22 o Adiabatic Flame Temperature (F) Correlation of Flare Gas LHV versus Adiabatic Flame Temp (Tad ) for Select Flare Operating Conditions 5000 "Propylene" 4500 Log. ("Propylene") 4000 R2 = 0.9941 3500 3000 2500 2000 1500 1000 500 0 50 100 150 200 250 300 350 400 450 Flare Gas Lower Heating Value (BTU/scf) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 23 Correlation of Flammability Ratio (~%FL/%SR) versus Adiabatic Flame Temp (Tad ) for Propylene (LHV)f = 400 (LHV)f = 300 (LHV)f = 200 3500 o Adiabatic Flame Temperature (F) 4000 3000 Inflammable (LHV) f = 100 2500 Flammable (LHV)f = 50 2000 R2 = 0.9963 1500 Propylene.4 Linear (Propylene) 1000 500 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 FRm (%Flammability/%SR) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 24 Deficiencies and Limitations Need to account for wind effects (i.e., wake stabilized mixing versus momentum mixing) o If wind effects added, need good wind data to facilitate incorporation into metrics Additional testing useful in understanding previous data that show low LHV and high combustion efficiency o CMA data indicates "stable" combustion at low heat values? o Calculation basis, definition of LHV, etc. Metrics do not account for tip geometry effects (i.e., internal burning, flame stabilizer rings, etc.) o Impossible in simplified approach to account -CFD useful for detailed analysis IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges LES Based CFD Represents "BAT" for Flare Flame Analysis Slide 25 CFD captures dynamics of flare flame o Momentum vs. Buoyancy o Wind Effect on Flame Structure o Chemical Reactions o Radiation to/from flame Approaches o Eddy Breakup (EBU) Chemistry in Reynold's Averaged Navier-Stokes (RANDS) Formulation (averaged flame structure) o Large-Eddy Simulation (LES) with Detailed Chemistry and Radiation (accurate flame structure) o LES with EBU Chemistry and Diffuse Radiation IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 26 Isis-3D Flare Model: • Provide reasonably accurate estimates of the total heat transfer to objects from large fires • Predict general characteristics of temperature distribution in surrounding objects (i.e., Fence) • Accurately assess impact of flare operations in given ambient conditions (effect of wind speed, fuel flow, fuel composition, surrounding equipment for given flare geometry) • Reasonable CPU time requirements using "standard" desktop LINUX workstations (i.e., P4 processor, 1 GByte RAM) IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 27 ISIS-3D Validation of Multi-Tip Ground Flare Analysis (no wind) ~16 m (52 ft) 12 m (39 ft) Flame height ~11 m (~36') Non-luminous region ~1 m (3') Tip Height~3 m (10') IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 28 ISIS-3D Validation: Radiation Predictions Compared to Experimental Data IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 29 Wind Effect on Radiative Flux IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges Slide 30 Conclusions and Recommendations • Flare performance greatly effected by cross winds - - - - • ratio of flare gas to wind momentum combining oxidant with fuel shearing reaction zone Cooling reaction zone Flare performance greatly effected by steam rate - steaming adds additional air to combustion zone - over steaming quenches (cools) reaction zone - over steaming dilutes reaction zone • Flare performance greatly effected by flare gas heating value - flare gas (vent + purge + pilots + steam) must be flammable when mixed with air to burn - Flammability ratio (fuel/flare gas) / (fuel/fuel+air)stio captures "combustibility" of flare gas with steam addition to combustion zone • LES based CFD represents BAT for Flare Performance analysis - Includes combustion/soot reactions plus detailed radiation - Results compare well to test data IFRF 16th International Members' Conference: Combustion and Sustainability: New Technologies, New Fuels, New Challenges