Flare issues in the operating world - an owner's perspective

Flare Issues in the Operating World An Owner's Perspective Vance Varner Process Associate / Flare Specialist The Dow Chemical Company September, 2010 IFRF TOTeM 36 1 Purpose of Presentation • Flares necessary safety / environmental equipment • Need to be essentially 100 percent reliable • Very challenging in real operating world • Unacceptable to have failure be the leading indicator of emerging issues with flares CREATE / LEAD DISCUSSIONS with REAL WORLD ISSUES SEEN WITH OPERATING FLARES September, 2010 IFRF TOTeM 36 2 Background • Dow (direct or subsidiary) has a significant inventory of flares globally • Most types - - - - - Elevated / various supports Ground (enclosed / multipoint) Assisted (steam / air / pressure / fuel) Non-assisted Enriched - fuel gas / hydrogen • Multiple suppliers • Size range: - Diameter: 3" (76 mm) to over 100" (2540 mm) - Elevation: 20 feet (6 meters) to 500 feet (150 meters) • FFG ignition / HEI ignition / auto / manual / combinations September, 2010 IFRF TOTeM 36 3 Safety Issues for Flares • Loss of pilots • Oxygen in flare header / flare stack - Flash back / detonation • Flaming rain - Knock out drum over-fill • • • • • Flare stack plugging Loss of flare header purge Loss of water seal Brittle fracture of carbon steel Smoking flares (environmental issue) September, 2010 IFRF TOTeM 36 4 Other Issues • Main flame operation at very low flows - Flares must be sized for high hydraulic loads • Lack of robust automated smoke control • Pilot monitoring - Responsiveness - slow - Loss of signal • • • • • Fabrication issues Documentation timing Maintenance planning Radiant heat calculations Noise September, 2010 IFRF TOTeM 36 5 Disclaimer / Acknowledgement • Not meant to pick on any supplier or type of flare • Have had issues with all suppliers / at least one of all types / multiple locations • Some issues are operational • Some issues are design • Some issues are both • Flare systems / designs have improved - Equipment lifetime longer - Higher smokeless capacity - Improved efficiency - Larger equipment September, 2010 IFRF TOTeM 36 6 Real World Operational Changes • Lower flows during normal operation • Less continuous flows (more recovery) • Increased awareness - Environmental concerns - By neighbors (visible flames not as acceptable) - Government agencies • Larger / taller (higher ultimate capacity) - More visibility September, 2010 IFRF TOTeM 36 7 Incident Categories • Past 10 years: • 72 percent of those involve flame out Flame out - Operations 18.6% Flame out - Unknown 15.9% Flame out - Weather 15.0% Flame out - Instrumentation 6.2% Flame out - Mechanical 5.3% Flame out - Steam 5.3% Flame out - Fuel 4.4% Flame out - Pilot plugged 0.9% September, 2010 IFRF TOTeM 36 8 Incident Categories Surrounding area fire 4.5% Flare pluggage 3.6% Materials issues 3.6% Flare damage 2.7% Flashback - Operations 2.7% Flame in muffler 2.7% Instrument failure - Operations 1.8% Internal burning - coke 1.8% Noise 1.8% Flare tip crack - mechanical 0.9% Liquid out of flare 0.9% Smoke 0.9% September, 2010 IFRF TOTeM 36 9 Top Issue / Problem • Current pilot design: high wind / high rain • There are more problems with going out during thunderstorms - Lower wind speed - Larger water droplets • Pilots age range from 2 months to 10 years September, 2010 IFRF TOTeM 36 10 Loss of Pilots - Root Causes • Weather • Changing fuel composition • Wrong pressure setting - 3 psi / 0.2 bar too high or too low • Mechanical issues / age • Design September, 2010 IFRF TOTeM 36 11 Loss of Pilots - Solutions • • • • Improved location on flare tip Improved mixing Improved shielding from wind Recommendation on expected lifetime of pilot assembly • Higher heat release / improved stability?? • Improved testing for multiple conditions September, 2010 IFRF TOTeM 36 12 Other Pilot Loss Items • High nitrogen flow in flare gas - Happened on at least 3 different flares • High steam flow - Sudden lowering of flare gas (event over) Pilot operation / flame stability should be independent of the flare gas flow / composition September, 2010 IFRF TOTeM 36 13 Other Items See anything wrong with this flare tip? Steam ring not exactly in correct location September, 2010 IFRF TOTeM 36 14 Same Tip September, 2010 IFRF TOTeM 36 15 What Happened? • Tip was installed in 2007 as a replacement - Issue noticed / repaired in 2008 (new ring) - Issue reappeared in 2009 • Replacement by different supplier - Could not send out old tip drawings • Steam line shift - stayed within guides on stack • Closest support (not guide) below tip was 90 feet (27 meters) • Original tip had hard support for steam line - Missed during design / specifications • Lessons: - Look for all connections - Avoid hard piping supports on flare tip September, 2010 IFRF TOTeM 36 16 Flame Around Ring • Small flare • Relatively low flow • Flame around steam ring • More than just wind effect September, 2010 IFRF TOTeM 36 17 Flare Plugs Several examples: • Air assisted flare in Canada - ice - Winter operation - Water vapor in gas stream - New tip design contributed to issue • Air assisted flare - coke - Normally very low flow - Gas is ethylene rich September, 2010 IFRF TOTeM 36 18 Coke in Flare Tip Air assisted Coke Stalagmites Coke in arm Notice the fire all around these Fan operational - air there as well September, 2010 IFRF TOTeM 36 19 Internal Burning September, 2010 IFRF TOTeM 36 20 Details • Normally low flow / ethylene rich stream • Air assisted flare • Generally replacing tip every 4 to 5 years - same problem • Opportunity for new technology - Existing arm style allows cross mixing / burning inside - May be too small for "tube" type September, 2010 IFRF TOTeM 36 21 Flame Out - Water in Seal • • • • • • Water collection in density seal Drain line plugged with coke Sent high flow of gas to flare Pushed water out Extinguished main flame and pilots Could not re-light pilots for 4 hours - Issue with FFG - stuck check valve Lesson learned: periodic functional testing of FFG systems September, 2010 IFRF TOTeM 36 22 Fabrication Issues • This has varied - not every time • More issues with sub-contractors - Not adequate inspection - Not following drawings / specifications - With non-technology parts • Some issues with not following specifications September, 2010 IFRF TOTeM 36 23 Real World Control Issues • Control of smoke not straightforward • Design of control panels - Issue with long term operation • Protection of wiring - Burning of wires near tip - Ignition wiring become wet / shorting out • Protection of thermocouples September, 2010 IFRF TOTeM 36 24 Radiant Heat Calculations • API 521 is used as industry basis • Suppliers have adjusted / modified based on their own research - Consider these methods proprietary - Some burner styles produce less • Opposite driving influences - Safety: push designs taller - Competition / cost: push designs shorter September, 2010 IFRF TOTeM 36 25 Radiant Heat Calculations • Owners / engineering contractors cannot match supplier results • Harder to compare different proposals • Support the concept of a single method for everyone in the industry / users - Must be based on field / research data - Must be supported by suppliers / experts September, 2010 IFRF TOTeM 36 26 Enclosed Ground Flares • Issues with cracked / failed burners • Impingement on other burners • Lesson: Inspections critical • Need: More robust designs September, 2010 IFRF TOTeM 36 27 Backflash in Stack / Header • Two incidents: oxygen in header with a backflash / detonation • Both due to operational issues, not design • The pilots were the ignition source • NO INJURIES in these cases, but potential was there September, 2010 IFRF TOTeM 36 28 Backflash 1 • Procedure / operating discipline not followed • Air from purge step on equipment connected directly to flare header • Mixed with other hydrocarbons • Backflash physically moved KO drum and piping from foundations • High oxygen alarm in place September, 2010 IFRF TOTeM 36 29 Backflash 2 • • • • • Maintenance group opened wrong flange 48" flange with spacer Removed / heard "sucking" sound Vacated rather than re-insert Mixed with hydrogen in header from another source • Detonation / standing flame • No real equipment damage September, 2010 IFRF TOTeM 36 30 Backflash - Why Discuss • Lessons learned • More information on flame speed with various hydrocarbons in a pipe / stack / tip - Know or determine the minimum tip exit velocity required to prevent backflash - Laminar flame speed information known September, 2010 IFRF TOTeM 36 31 Documentation • Drawings - generally OK • Operational manual - always too late in the process - Can provide procedures early - Final drawings come after delivery - Not always specific to flare / tries to cover potential options September, 2010 IFRF TOTeM 36 32 Maintenance • • • • In general getting more attention More inspections Units working on maintenance strategy Need input from suppliers - Typical life of components - Recommended maintenance September, 2010 IFRF TOTeM 36 33 Other Subjects • Materials of Construction for long term value • General improvements in technology - Higher smokeless capacity / efficiency - Longer operation between inspection / ability to work on tip • Lower costs - New equipment - Replacement parts September, 2010 IFRF TOTeM 36 34 Conclusions • Operating flares in the real world - there are issues • Owners must pay attention to the flares - Operation - Maintenance - Inspections • Designs / technology must continue to be improved • Encourage use of conferences & forums for communications / discussions on common problems September, 2010 IFRF TOTeM 36 35