Flaring of gas mixtures containing H2S

Flaring of Gas Mixtures Containing H2S Allan Chambers, P.Eng. Alberta Research Council Inc. AFRC May 2008 Background • flaring of sour gas in Alberta - well testing - safety flares - solution gas flares (dispose of uneconomic sour gas) • environmental concerns - public safety - damage to vegetation Background cont'd • operating permits are required - dispersion models used to predict ground level SO2 concentrations - models assume >98% combustion efficiency - models assume 100% conversion of H2S to SO2 • little experimental data on sour gas flares • are the modelling assumptions correct? Past Work on Sour Flares • EPA Flare Study (Pohl and Soelberg, 1986) - 3 and 6 in. diameter pipe flares - probe sampling for H2S, SO2 and HCs • results - H2S destruction efficiency typically >98% - less than 80% accounted for as SO2 - >29% H2S in fuel resulted in grey-blue plume • propane was the fuel, not methane • flares with no cross wind Past Work: Tollefson & Strosher, 1989 • sample probe downwind of full scale flare tip • limited range of conditions • was combustion complete? Concentration (ppm) SO2 H2 S CS2 COS 2350 375 375 90 Amount of Total Sulphur 65.9% 10.5% 21.0% 2.5% Past Work: Strosher 1996 • probe sampling downwind of visible flame • 25% H2S in natural gas Concentration 3 (mg/m ) gas phase SO2 H2S CS2 COS Others volatile CS2 thiophenes, C 4H 4S non-volatile sulphur Amount of Total Sulphur 6910 126 482 64 210 74.6% 2.6% 8.8% 0.7% 453 210 8.2% 1.7% 157 3.4% Past Work: Chambers, 2003 • Differential Absorption Lidar (DIAL) measurements of a well test flare • vertical laser scans 2 to 3 flame lengths downwind of visible flame • CH4, ethylene and SO2 concentration profiles • measured fuel flow rate and composition, wind speed and direction SO2 Plume from Incinerator plume cross section 990 ft downwind DIAL Sour Gas Well Test Flare SO2 in Flare Plume R i SO2 flux of 350 kg/hr DIAL Results for Well Test Flare • only 50 to 70% of H2S accounted for as SO2 • no indication of unreacted H2S from ground monitors • hydrocarbons in plume primarily CH4 Visible Secondary Plume • forms 1 to 2 flame lengths from end of flame • possible causes - fine particulates - SO3 forming H2SO4 aerosol H2S Combustion Equilibrium Fraction of total Sulphur 1 0.8 0.6 SO2 SO3 H2SO4 0.4 0.2 0 300 500 700 900 1100 1300 1500 Temperature (C) • F*A*C*T equilibrium software Does H2S Affect HC Combustion? sweet solution gas sour solution gas sour well test 1 Tip diameter (in) 3.5 H2S content DIAL measured 0 98% 8 1.15% 74 - 92% 1 10 6.6 and 11% 91% corroded flare tip What is the impact of H2S? • evidence that significant amount of H2S is not converted to SO2 • possible impact on hydrocarbon combustion efficiency • what are the other sulfur species? • should some flaring conditions be avoided? - high cross winds - high H2S concentrations - etc. What is Needed • experimental measurements to determine H2S impacts and sulfur species formed • numerical simulations to better understand interaction of a number of flare parameters • may lead to recommendations to improve sour flares or avoid some conditions Full Scale Experiment Challenges • sampling with probes is difficult - height of sour flares (50 m) - hazards working in vicinity of sour gas flare - plume movement • site selection - prefer a sour flare with ability to vary H2S concentration, gas flow rate • cost of field experiments • how do you control the wind? Large Range of Alberta Conditions • • • • • flare tip exit velocity from 1 to 25 m/s gas heating value from 12 MJ/m3 to 40 MJ/m3 H2S content from 5% to 50% flare tip diameter from 3 inch to 12 inch pipe wind velocity from 0 to 25 m/s (90 km/h) Plume Measurements • DIAL to measure mass flux in plume of: - methane and C2+ hydrocarbons - SO2, NOx • direct sampling from the plume for other sulfur species (H2S, SO3, COS, etc.) DIAL for remote measurements Sampling for Sulfur Species? tethered He balloon denuder tube sampling Numerical Simulation of Flares • University of Utah simulation expertise • Large Eddy Simulation (LES) methods to model open flames • include prediction of: - hydrocarbon chemistry - sulfur chemistry - radiation heat transfer • large range of conditions possible Proposed Experimental Program • locate sour flare with measured fuel flow and composition • DIAL measurement of sour flare plume (SO2 & HC mass flux) • sampling from plume to measure relative amounts of sulphur compounds • mass balance on sulfur species, effect on combustion efficiency • validation data for numerical simulation Expected Outcomes of Project • better understanding of: - sulphur compounds produced by flaring - impact of H2S and other parameters on flare performance • numerical simulation results to guide experimental work • experimental data to validate numerical model • a simulation tool has the potential to improve flare design and operation Acknowledgements • funding by: - Canadian Association of Petroleum Producers (CAPP) • overseen by: - Air Issues Committee, Petroleum Technology Alliance Canada (PTAC)