Failura Analysis Project: Case Analysis

profilePV329
ZzzzzzAaaaaaCaseStudyFAPaper1.docx

Zzzzzz Aaaaaa1

Zzzzzz Aaaaaa 7

Failure Analysis of the Flixborough Explosion

Zzzzzz Aaaaaa, University of Florida Department of Chemical Engineering

16 October 2017

______________________________________________________________

Abstract

Toxic, flammable, and corrosive chemicals are frequently used as routine reactants for chemical process plants. When chemical engineers become too comfortable working alongside these dangerous chemicals, disaster due to negligence often occurs. This paper discusses and analyzes one such disaster involving the explosion of a nylon manufacturing plant in Flixborough, England after a flammable compound leaked into the open air and ignited. Information on the processes and equipment involved in the nylon plant will be provided, as well as a discussion and analysis of the crucial decisions that led to the demise of the plant. Lastly, this failure analysis will examine the lessons learned in the Flixborough disaster to avoid future disasters in the chemical process industry.

1. Background

1.1 Context

The Nypro Limited chemical facility in Flixborough, England aimed to produce 70,000 tons of caprolactam per year [1]. Caprolactam is the main raw material used in the production of Nylon 6, and is made from oxidizing cyclohexane, a highly flammable liquid with properties similar to those of gasoline [2, 3]. The process used in the facility to convert cyclohexane to caprolactam involved six reactors set up in series configuration, with each reactor usually containing about 20 tons of cyclohexane at any given time [2].

1.2 The Explosion

Months before the explosion occurred, workers discovered a vertical crack in the stainless steel structure of the fifth reactor [1]. Workers decided to remove Reactor 5 for repairs and connect reactors 4 and 6 to continue production at a reduced yield [1]. However, the pipe used to connect the reactors was flexible, dog-legged, and different from the feed pipes connecting the other reactors [2]. Refer to Figure 1 below for the configuration of the six reactors. No hazard assessment, pressure testing, or calculations were conducted to understand the new implications of the pipe change [1, 2], and thus disaster ensued. On the afternoon of June 1, 1974, the new bypass pipe used to connect reactors 4 and 6 ruptured, allowing an estimated 30 tons of gaseous cyclohexane to volatilize and form a large vapor cloud [2]. The vapor cloud was ignited less than a minute later, resulting in a violent explosion that killed 28 people and went down in history as the largest single loss by fire or explosion in the United Kingdom [3].

1

2

3

4

5

6

Figure 1

1.3 Significance

The Flixborough explosion teaches all branches of engineering the importance of having a deep understanding of the parts and equipment involved in their facilities, and to perform extensive testing when changes to the parts/equipment are made. The Flixborough disaster set a precedent for chemical process plants in the United Kingdom, which urged other plants to increase their security, evaluate their safety precautions, and ensure that potentially dangerous substances are licensed and stored compliantly [4]. Failure in the field of engineering is a crucial lesson for all future engineering endeavors, as learning from past disasters is the best way to avoid future disasters [5].

2. Analysis

2.1 Installation

The pipes normally used to connect the reactors had a 28” diameter, but a 20” diameter pipe had to be used to connect reactors 4 and 6 because it was the largest pipe on site [6]. Since reactors 4 and 6 were at different heights, three different lengths of the 20” pipe had to be welded together to form a dog-legged shape that would accommodate the height difference between the reactors [6]. Refer to Figure 2 below for the structure of the dog-legged pipe used to connect reactors 4 and 6.

Figure 2: The dog-legged pipe

2.2 Tests and Calculations

Engineers performed calculations to confirm that the pipe diameter was large enough for the flow of cyclohexane [6], but did not take into account the pressure and hydraulic thrust on the bellows for which they were not designed [4]. All calculations performed were with respect to a straight pipe instead of dog-legged, so it was unknown whether the bellows or bypass pipe would withstand the strains [6]. Instead of drawing the pipe in accordance with the designer’s guide, the only drawing of the pipe was made in chalk on the workshop floor [4]. To relieve strain on the bellows of the bypass pipe, a scaffolding structure usually built for workmen working at elevated positions was constructed [3]. However, no pressure test was implemented on the bypass pipe nor the scaffolding structure before it was fitted [6]. After installation of the 20” bypass pipe and the scaffolding structure, a pressure test was performed on the equipment up to 9 bar (g), but not up to the safety valve pressure of 11 bar (g) [6]. The tests and calculations carried out were not in agreement with the recommendations of the United Kingdom: that any modification should be designed, constructed, tested, and maintained to the same standards as the original plant [4].

3. Conclusion

3.1 Impact

The Flixborough explosion claimed many lives, caused infrastructure devastation, and started widespread fires [6]. Twenty eight people lost their lives in the explosion, eighteen of which were located in the control room at the time of the explosion [6]. The layout of the plant was not designed to withstand any explosion or major disaster, including the control room, which suffered shattered windows and a roof collapse [1]. 433,000 gallons of flammable liquids from the plant were responsible for the fires which damaged structures up to 8 miles away [3]. The fires burned on the site for over ten days after the explosion [6], and damage extended to 1,821 nearby houses and 167 shops and factories [2]. Refer to Figure 3 below for a photograph of a destroyed section of the Flixborough reactor system, including the fragile scaffolding.

Figure 3. Destroyed reactor system [3]

3.2 Negligence

The Flixborough explosion occurred due to a combination of negligence, laziness, and lack of knowledge. The engineers at Nypro Limited in Flixborough did not take proper precautions when implementing the temporary bypass pipe by using a pipe with a smaller diameter, not taking the dog-legged pipe shape into account in their calculations, failing to properly draw the structure, testing equipment after assembly, and performing pressure tests under the safety limit [3, 4, 6]. The engineers at Flixborough were unaware of the vulnerabilities and limitations of the 20” bypass pipe bellows configuration, and did not consider the failure risk of using flammable and toxic liquids in the vicinity of temporary, minimally tested equipment [5].

3.3 Prevention

The two main causes of loss of containment are due to facility equipment failure (FEF) or personnel failure (PF) [1]. Facility equipment failure can include poor process design, a lack of understanding of equipment, corrosion/erosion mechanisms, stress/fatigue/vibration mechanisms, or aged equipment; whereas personnel failure can include noncompliance with procedures or improper installations [1]. Understanding how parts function and fail is crucial to chemical process safety and comes with the moral code that engineers accept as part of their job [5]. Only knowledgeable and experienced professionals should decide when to insist on code compliance versus allowing changes to processes in the engineering field [5]. The Flixborough explosion could have been avoided by following proper safety protocol and placing a larger emphasis on the new modification rather than on fast restart of the plant after initial discovery of the cracked reactor [1, 2].

References

[1] M. Deighton, “The Flixborough Disaster,” in Facility Integrity Management - Effective Principles and Practices for the Oil, Gas, and Petrochemical Industries, 1st ed. Cambridge, MA: Elsevier, 2016, pp. 7-8.

[2] D. Crowl and J. Louvar, “Flixborough, England,” in Chemical Process Safety: Fundamentals with Applications, 3rd ed. Upper Saddle River: Prentice Hall, 2011, pp. 23-25.

[3] R. Sanders, “The Flixborough Disaster and the Lessons We Should Never Forget,” in Chemical Process Safety - Learning from Case Histories, 4th ed. Cambridge, MA: Elsevier, 2015, pp. 196-202.

[4] “The Flixborough Disaster,” in Department of Employment. London, England: Crown, 1975.

[5] H. Bloch, “Remembering Flixborough,” in Petrochemical Machinery Insights. Cambridge, MA: Elsevier, 2016, ch. 28.5, pp. 431-433.

[6] R. Eckhoff, “The Flixborough Disaster,” in Explosion Hazards in the Process Industries, 2nd ed. Cambridge, MA: Elsevier, 2016, pp. 77-81.