The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study...

24
The Flixborough, UK, Cyclohexane Disaster, 1 June 74 • A Vapor Cloud Explosion (VCE) Case Study • Presented to ES-317y at UWO in 1999. • Dick Hawrelak

Transcript of The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study...

Page 1: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

The Flixborough, UK, Cyclohexane Disaster, 1 June 74

• A Vapor Cloud Explosion (VCE) Case Study

• Presented to ES-317y at UWO in 1999.

• Dick Hawrelak

Page 2: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Flammability Hazards

• Bursting shock wave - TNT equiv.

• Fireball - thermal hazards

• BLEVE - flying fragments

• Flash Fire - thermal hazards

• Vapor cloud explosions - TNT equiv.

Page 3: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Two Classes Of Hydrocarbon Vapor Cloud Explosions, VCEs

• Detonation Class VCE - High level explosion.

• Deflagration Class VCE - Low Level explosion.

Page 4: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Detonation Class VCE

• Flammable vapor is ignited in a congested plant area with vertical confinement.

• High flame acceleration leads to detonation.• The sharp impulse force can be equivalent to

a TNT explosion. • Damage will be radial from the explosion

epicenter.• Overpressure range, PSO, 10 to 5 psig.

Page 5: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Deflagration Class VCE

• Flammable vapor is ignited in uncongested area with open space.

• Low flame speed leads to deflagration.• Impulse force longer duration and not as

damaging as a detonation.• Damage may be directional from the explosion

epicenter.• Overpressure range, PSO, 5 to 0.5 psig.

Page 6: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Flixborough Reactors

Page 7: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Flixborough Flowsheet

Page 8: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

The Incident

• A 20 inch diameter temporary by-pass pipe Jack-knifed and failed under thermal expansion stress.

• 40 of 120 tonnes of cyclohexane escaped into the congested reactor support structure.

• Within two minutes, the vapor cloud ignited and a Detonation Class VCE took place (35 tons TNT equiv).

Page 9: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Thermal Expansion Jack-Knife

Page 10: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

VCE Results Flammable Hazard V1.2

0.10

1.00

10.00

100 1,000 10,000

Distance From Vessel, Feet

PSO in psig

Clancey

Gugan

Flix Pts.

Edge Of Cloud

Page 11: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Flammable Hazard V.12

Type of Damage Overpressure Radial PSO, psig Dist., Ft.

Flixborough Type Damage 10.01 376Distillation Tower Overturned 7.25 468Piperack Bent - Piping Breaks 6.09 527Limit Of Major Plant Eqt Damage 5.00 603Steel Panel Building Demolished 3.48 772API Tank 50% Full Uplifts 3.00 854Non Reinf Conc Blocks Shattered 2.51 965Lower Limit Serious Struct Damage 2.20 1,053Man Knocked Over - No Ear Damage 2.00 1,125Corrugated Steel Panels Buckle 1.00 1,804Glass Damage Shattered 0.50 2,892Limit Minor Structural Damage 0.41 3,334Limit of Glass Failure 0.15 6,721

Page 12: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Basis: Know Your Insurer's Expectations, Fire Protection Design, Hydrocarbons Processing, August 1977, p-103, by Robert W. Nelson. Updated from Lihou's Table 4 - 8 Sep 96

Overpressure, psigEquipment 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10 12 14 16 18 20 22

Control Rm, Stl Rf a c d n

Control Rm, Conc Rf a e p d n o

Cooling Tower b f o

Tank: API Cone Roof d k u

Instrument Cubicle a l,m t

Furnace Heater g i t

Reactor: Chemical a i p t

Filter h f v t

Regenerator j i,p t

Tank: Float'g Roof k u d

Reactor: Cracking j i t

Page 13: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Plot Plan Showing VCE Impact

Page 14: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

The Consequences

• 28 plant people were killed.

• 53 people were wounded and required medical treatment.

• 1,800 houses were damaged in the rural area beyond the plant fence line.

• Property damage was $425MM in US funds.

Page 15: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Events Leading To The Incident.

• Two months before the incident, R-5 was found to be leaking.

• A 6 ft. long crack had developed.

• A water hose stream was directed to the crack to cool and quench the small cyclohexane leak.

Page 16: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Events Cont’d

• The cooling water contained nitrates which encourage stress corrosion of certain carbon steels.

• Thus, by trying to relieve the situation, the quenching was actually acting as a promoter of corrosion.

• Ultimately, the reactor had to be removed from service.

Page 17: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Events Cont’d

• There was no experienced works manager, WM, available on site at the time of the removal of R-5.

• The previous WM, a good maintenance engineer with 25 yrs of experience, had quit because an anticipated promotion was given to an outside person.

Page 18: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Events Cont’d

• As there was no experienced mechanical engineer on site, those remaining decided to “fast track” or “scratch pad” a solution for the intended by-pass.

• They sketched a full-scale by-pass line in chalk on the maintenance floor.

• No stress analyses calculations were performed on the by-pass connection.

Page 19: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Events Cont’d

• The by-pass line was quickly installed and the plant put into start-up mode.

• Shortly after start-up, the by-pass line failed causing 40,000 lbs of cylcohexane to leak into the confined spaces of the reactor support structure.

• Within two minutes, the vapor cloud exploded.

Page 20: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Lessons Learned

• The main root cause of this incident was the use of cooling water with nitrates to quench cyclohexane leaks on the reactors.

• Another root cause was installing a by-pass line, or any line for that mater, without stress analysis. This is a recipe for disaster.

• A third root cause was management must recognize when they are vulnerable to critical manpower changes.

Page 21: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Lessons learned Cont’d

• More control is required to conduct good engineering practices once the plant is up and running.

• Poor location and poor construction of the control room.

• Plant was too congested at the design stage.

• Must minimize hazardous inventories.

Page 22: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Lessons Learned Cont’d

• Process hazard review required at regular intervals.

• Plant must adhere to pressure vessel regulations.

• Require emergency planning with the community.

Page 23: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Deterministic Pre-planning Flammable Hazard V1.2

• Defines vapour cloud characteristics between UEL and LEL.

• BLEVE shock wave, thermal and fragmentation analysis.

• Flash Fire thermal analysis.• VCE analysis.• Space separation (ISBL, OSBL and green

belt).

Page 24: The Flixborough, UK, Cyclohexane Disaster, 1 June 74 A Vapor Cloud Explosion (VCE) Case Study Presented to ES-317y at UWO in 1999. Dick Hawrelak.

Possible Exam Questions

• How does a Detonation Class VCE differ from a Deflagration Class VCE?

• Describe the characteristics of the two type of explosions.

• What were the three root causes of this incident?