Flixborough Disaster Inquiry

That any modifications should be designed, constructed, tested and maintained to the same standards as the original plant.

That all pressure systems containing hazardous materials should be subject to inspection and test by a person recognised by the appropriate authority as competent after any significant modification has been carried out and before the system is again brought into use.

That existing regulations relating to modifications of steam boilers which do not apply to pressure systems containing hazardous materials should be extended so as to apply to such systems. In framing such regulations consideration will no doubt be given to the question whether the inherent nature of the material requires greater control than applies in the case of steam boilers.

That the British Standard referring to the pressure to which pipe work should be tested (see paragraph 73 above) should be clarified. At present it is ambiguous in its reference to testing to 1 -3 x the “design pressure”. It should be made clear whether, “design pressure” means “maximum operating pressure” or “relief valve pressure” and if the former that a test to above relief valve pressure is always required.

That compliance with the British Standard requirement for hydraulic testing (paragraph 73 above) should be obligatory. The by-pass was tested pneumatically to 9 kg/cm2 for leaks without any previous test at all. Had it burst there might well have been serious if not fatal injury.

That when an important post is vacant special care should be exercised when decisions have to be taken which would normally be taken by or on the advice of the holder of the vacant post. This, in the present instance, would have involved the reference by senior management to Mr Hughes of the problems created by the defect in Reactor No. 5 and the design of the by-pass.

That the training of engineers should be more broadly based. Although it may well be that the occasion to use such knowledge will not arise in acute form until an engineer has to take executive responsibility it is impossible at the training stage to know who will achieve such a position. All engineers should therefore learn at least the elements of other branches of engineering than their own in both their academic and practical training.

It is therefore recommended that all plants whose safety relies upon nitrogen should have nitrogen supplies which are ample to cover all contingencies. This means that there must be either adequate “in-house” supply or massive reserves.

The cracked Reactor R2525 initiated the sequence of events which led to disaster. Examination of the crack by expert metallurgists showed that the crack had been caused by nitrate stress corrosion. This corrosion was created because nitrate treated cooling water had been used in the past to dilute small leakages of cyclohexane from the plant. This phenomenon had not been appreciated by the management at the time the water was used and may not even yet be appreciated by other sectors of the chemical industry. Where such cooling water is used the vessels and piping should be adequately protected. The attention of industry should be drawn to the risk.

Since many chemical plants contain zinc coated components, eg galvanised wire and walkways, it is important that the attention of industry should be drawn to these matters.

Creep cavitation fractures in stainless steel have been known for some time but previously it had not been generally known that these could, under appropriate conditions of stress and temperature be produced in a relatively short time. Such a fracture can be produced in a matter of minutes by a small fierce fire. (See Appendix II). Again the attention of industry should be drawn to this phenomenon.

It is recommended that consideration be given to installing devices or systems for recording vital plant information in a form which would survive the effects of fire or explosion. An example of such a device is the “black-box” used in aircraft.

A scheme for co-ordination between the planning authorities and the Health and Safety Executive should be devised so that the planning authorities may be advised on the safety problems involved in any proposed plant before planning permission is granted.

We have pointed out in Paragraph 194 (c) above that the present situation is unsatisfactory. We recommend review of the existing regulations.

The question of crack propagation and detection in internally clad mild steel vessels and their protection from corrosion on the outside should be investigated as a matter of urgency.