Duty holders must demonstrate that structures have sufficient robustness to ensure a low probability of catastrophic failure when subjected to accidental explosions.
Ongoing operational safety shall be demonstrated by appropriate design capacity in conjunction with an appropriate integrity management system to ensure degradation is kept within acceptable limits.
The magnitude of the explosion loads for which the installation is designed should reflect the residual risks to persons on the installation from explosion (in line with the ALARP principle). Robustness of the integrity should be measured against the resilience of the structure to resist these loads with suitable margins to cater for the uncertainties in the load estimation, the modelling schem chosen, and the material properties of the materials used.
The potential for an accidental release of a hydrocarbon mixture is ever present on hydrocarbon carrying installations and, if ignited, could lead to a serious incident involving loss of life, loss of the asset and environmental damage. Full scale tests on realistic modules and congestion confinement levels carried out on joint industry projects have shown that the severity of the threat is more severe and complex than originally envisaged. Far higher pressures have been measured in testing, capable of causing serious structural damage similar to that which led to an escalation of events on Piper Alpha.
The nature of the threat varies significantly with inventory detail, level of confinement and congestion which is almost always present on North Sea layouts. Even if confinement is reduced, high levels of congestion can actually create confinement and enhance the explosion experienced by the surrounding plant and structure. This also changes if modifications made to the structure increase inventories. Older platforms may also have higher frequency of releases as a result of ageing which will increase the likelihood of leaks due to corrosion and fatigue issues.
Prior to the Piper Alpha incident, structural design to limit the consequences of a hydrocarbon explosion was mainly confined to constructing blast walls in the well bay area. These were limited in number and design pressures used were usually low, of the order of 0.1-0.2 Bar. Post Piper, several bars of pressure were found to be a real possibility with realistic levels of congestion and confinement. Little consideration had therefore been given to explosion scenarios on many older fixed structures. As a result of this, little attention was given to design strategies dealing with explosion scenarios.
This policy should have no impact for new installations as designs should incorporate the required regulatory and technical principles. However for existing installations remedial measures may be required to demonstrate robustness of the installation and changes to inventory profile may also require some remedials. Ageing assessments should confirm that structural arrangements are still appropriate for inventory levels that exist on the installation. This technical policy is relevant to the assessment of fixed and floating installations including FPSOs, Jack-Ups and Semi-Submersible Mobile Drilling Units.
The Piper Alpha incident in 1988 and regulatory changes made since then have led to a different safety culture not only on North Sea installations but also at a global level. More emphasis on requirements for providing inherent safety is now required. Although an incident of the same magnitude as Piper Alpha has been avoided, there have been fires and explosions which could potentially have led to more catastrophic events. These could be classed as ‘near misses' and are a reminder of the need to maintain the good practice achieved and implemented.
The failure history of large ocean going vessels is a significant cause of concern in the shipping industry. This is of relevance to the offshore sector in relationship to FPSO's where potentially larger gas cloud sizes could form due to space constraints.
Prior to the Piper Alpha incident in 1988, limited design guidance for offshore structures to limit the consequences of fires and explosions existed. Some guidance from onshore petrochemical guidelines existed but this was not always appropriate for offshore applications. Generally speaking, guidance for the offshore area is still evolving. The latest industry based guidance which is currently in use is the API Recommended Practice 2FB for design of offshore facilities against fire and blast loading.
Guidance in the form of the Interim Guidance Notes published by the Steel Construction Institute were produced in 1992 as a result of a large joint industry project, ‘Blast and Fire Engineering for Topside Structures (phase 1)'. This was sponsored by 28 companies and the Health and safety Executive in order to improve understanding in the characteristics of hydrocarbon fires and explosions. A fire and blast information group was subsequently established through which a number of important technical notes have been published. Some of the interim guidance notes have also been updated through a UKOOA/HSE sponsored project. Other valuable documents include the Norsok standards and the GEE JIP project which resulted in the publication of an important handbook. Much of this data is currently being harmonised through ISO to a set of appropriate criteria.
Large scale testing in recent years has highlighted that significantly higher overpressures can occur as a result of an ignition of a hydrocarbon in a typical offshore module with realistic levels of congestion and confinement. As a result of the increasing complexity of computational fluid dynamics codes (used to model and estimate loading) and the capability for more detailed structural models, a better understanding of the load and response characteristics have been established. As a consequence there is the potential for an improvement in QRA models. In order to focus effort where it is most needed, a risk screening process is usually adopted which classifies installations and modules according to their explosion risk level. The measures for frequency and severity level are normally based on the size of the inventory and complexity of the process, the level of congestion and confinement and the staffing level of the installation.However the QRA process is quite complex and can be demanding which tends to limit the QRA by concentrating on representative scenarios. In some cases this can lead to ignoring the low probability high consequence events. Successfully dealing with extreme uncertain events has been shown in the past to be dependent on good engineering practice, as well as appropriate QRA, such as providing continuity between elements, ductile details which avoid high stress concentrations at critical points (weld details for example) so that the structure can take up the energy of the blast by stable deformations and plasticity.
Offshore Installations and Wells (Design and Construction, etc) Regulations 1996 (DCR), Regulation 4 and 5(1)(a) and (e).
Offshore Installations (Safety Case) Regulations 2005 (SCR), Regulation 12 (d), and Regulation 19;
The Offshore Installations (Prevention of Fire and Explosion, and Emergency Response) Regulations 1995, Regulations 4 and 5.
Structural elements and systems are used in provision of control and barriers for explosions and are likely to be designated safety critical elements. The performance standard required for these elements and systems should be clear to encapsulate not just the blast wall capacity, but the connections to the primary structure and the response of the primary structure, and be subject to scrutiny associated with the verification scheme for the installation.
Retrofit of barriers (usually blast walls) to existing installations plays a major part in reducing the risk and achieving ALARP criteria for accidental explosion events. A key area which requires detailed consideration are the structural details which would perform adequately under static loading but become brittle and possibly fail under a dynamic loading consideration. Some examples of this include the common practice of sniping secondary beams to minimise weld distortions when attached to webs of plate girders as part of the deck. These have been shown to act as stress raisers and potential failure sources, even at low pressures.
Some generic advice on methods that may be used to assess response are:
This method is often described as simple because of the simplicity of the idealisation and the fact you can obtain a response time history using a spreadsheet hence conducting parametric studies which are useful for screening different pressure time histories. However what is certainly not simple is the way in which the actual structure is converted into a single spring and a mass which in general requires good engineering judgement and a certain degree of experience. The method relies on a kinematic similarity between the actual and idealised spring mass, which for a beam simply means the velocity and displacement of the spring mass corresponds to a point on the beam which best describes the global behaviour. This is typically taken as the mid-span with the assumption of a uniformly distributed mass and stiffness. However large masses such as vessels commonly seen on topsides will significantly change the equivalent stiffness and mass adopted and therefore the corresponding displacements. Care is required in converting back to the actual structure to ensure the additional stresses are captured, particularly at support locations of both the beam and at equipment support points
The most complex and extensive structural modelling process adopted is the finite element method which can potentially capture all of the non-linear material and geometric behaviour together with the dynamics. Some packages can also potentially capture contact, brittle and ductile behaviour of the structure. If an installation is one that is in a high risk category which is relying heavily on ductility, this is the only technique which, if the model is extensive enough, can capture the interaction between the structural members and the true ductility of the structural system.
One of the key differences between FPSO's and fixed platforms is that the option of separation is limited, ie having the living quarters on a separate facility or a means of escape onto a separate unit is not an option. Also the option of minimising congested areas by splitting up the processing into smaller units and introducing sufficient gaps between the units is unlikely due to space constraints and economics. This can potentially lead to larger gas cloud sizes as they would not be limited by the volume of the module as on a fixed topside. Previous research has shown that as the gas cloud increases a runaway length can be reached at which point the overpressure increases significantly. Limiting a potential gas cloud size is an important issue in the design phase as well as keeping the design pressures inherently lower generally through design to minimise the effects of an explosion and fire is a key component on an FPSO facility
This policy covers the following elements: