Exploring The Differences In QRA Modeling For Offshore And Onshore Plants
In this article:
QRA is a powerful technique for assessing hazardous process plants. The technique has been applied for many years for both offshore oil and gas installations and onshore for the so-called "SEVESO" plants. However, there are some distinct differences in the purpose, method, and complexity of QRA for these two domains. The purpose of the article is to explore these differences and provide recommendations on the qra approach based on the context.
Offshore oil and gas (O&G) installations are in many ways unique installations and therefore also have a unique risk picture. The reason for this is that the installation consists of multiple facilities/operations cramped in on a small footprint typically less than a 100 m radius. The facilities found on such installations commonly consist of a hotel, process plant, harbor, and heliport. Furthermore, the installations are isolated by water and personnel cannot just remove themselves from harm's way the same way as personnel working on an onshore plant.
For an onshore process plant, it would be avoided to locate for example a hazardous process plant next to a hotel or an airport. It is typically, if not plenty of space, then at least more space available for spreading different parts of the process plant on a larger footprint introducing safety distances.
The purpose of a QRA for an offshore O&G installation is to calculate the risk to 1st and 2nd parties working on the installation. The risk to 3rd parties is normally not a concern simply because the installation is located remotely and isolated from any third parties. Normally a 500 m safety zone is implemented around the offshore installation to prevent 3rd party ship traffic in the area. To quantify the risk to 1st and 2nd parties on and offshore O&G installation it is typically necessary to consider the following hazards:
Loss of containment (LOC) of flammable, explosive, and/or toxic material;
Personnel transport risk from helicopter or vessel to/from installation;
Ship collision risk with installation;
Helicopter crash risk on installation;
Dropped and swinging load risk on installation;
Non-process related fires;
Escape, Evacuation, and rescue risks;
The purpose of QRA for onshore process plants is normally to fulfill the requirements of the SEVESO II directive. Here, the main focus is on the large loss of containment (LOC) events with the potential for affecting 3rd parties outside the border of the plant. The risk to 1st and 2nd parties is highly relevant for onshore process plants however often not quantified in the QRA. The main reasons are (1) traditions, (2) the focus from the SEVESO directive requirements, and (3) the upper limit for fatalities in a major accident hazard (MAH) is in the thousands compared to offshore O&G where fatalities rarely would reach above a hundred. There is a tendency that asset owners not often perform systematic quantitative safety/risk analyses unless regulators require this. Offshore Regulators have instructed offshore asset owners to systematically and quantitatively address risk for 1st and 2nd parties. However Onshore Regulators only require the risk to 3rd party is assessed quantitatively, and only for onshore plants considered to be SEVESO plants.
QRA LOC modelling and risk resolution
Based on the above, there is a distinct difference between offshore and onshore QRA on what hazards need to be considered. However, looking at LOC risk modelling alone for both onshore and offshore QRA, there are distinct differences between here as well. In an offshore QRA, all LOC events that can result in fatalities needs to be considered at some level meaning both small, medium, and large-size LOC events needs to be considered for both small and large inventories of hazardous materials. This can easily lead to hundreds if not thousands of different scenarios and an offshore O&G QRA becomes "big data".
For an onshore QRA typically only large size LOC from a few big inventories of particular hazardous material needs to be analyzed quantitatively to analyze the risk to 3rd parties. So typically only a handful of different LOC scenarios need to be analyzed rather than hundreds or thousands.
Hence an onshore QRA is typically much simpler than an offshore O&G QRA data-wise. The question is then if vital information is lost by this approach and if onshore QRA could gain something by increasing the risk resolution the same way as for offshore QRAs.
First of all, the onshore QRA does not quantify the risk to 1st and 2nd parties working on the process plant which in itself could be argued to be vital information even not normally a legal requirement. Normally it is argued that risk to 1st and 2nd party personnel has been managed by performing HAZOP, LOPA, and similar. However, these kinds of studies are also performed for offshore O&G installations and still, it is found relevant (and legally required) to quantify the risk to 1st and 2nd parties.
More importantly, information is lost on what can go wrong in the plant, due to the poor resolution typically found in onshore QRAs. Often LOCs that can cause exposure of 3rd parties require some sort of catastrophic failure, which will be highly unlikely unless exposed to external forces i.e. fire and blast loads from other initiating accidents. Hence domino effects often become the cause of the LOC studied in the onshore QRA. Often domino effects are only discussed qualitatively. But by increasing the resolution in the study of LOCs it is possible to determine smaller fire and explosion scenarios that can cause the domino effect quantitatively rather than relying on a pure generic failure frequency of a piece of equipment. It would therefore be beneficial to perform a fire risk assessment (FRA) and Explosion Risk Analysis (ERA) for the onshore plant quantitatively, which are the backbones of an offshore O&G QRA for LOC events.
Another problem with the poor risk resolution typically applied in onshore process plant QRAs is that the QRA will more or less only be applied for determining if the risk to 3rd parties is acceptable and that the SEVESO Directive has been fulfilled. This is of course important. But it is even more interesting if the QRA can be used actively in building and operating the plant. It will be possible to configure a specific process plant in many different ways that probably all fulfill the risk acceptance criteria for 3rd parties. But this does not mean that all the possible versions of the plant are equally safe! The design should lead to the safest version of the plant within the “As Low As Reasonably Practicable" (ALARP) principle. If used correctly and on time, the QRA methodology can be applied to provide valuable input to the design process evaluating alternative design options. Especially, the QRA could benefit the general layout of the plant and what specific safety measures should be implemented for the plant besides legal minimum requirements.
A higher risk resolution of onshore process plant QRAs can also benefit the operational phase of the plant. A detailed QRA can provide insights into what could have been the potential outcome (seriousness) of near misses etc. so the organization doesn't belittle incidents, but take appropriate actions to prevent the same type of incidents from re-occurring. Often only luck is the difference between having fatalities or not for near misses. A good consequence resolution of a QRA can also provide important information for the emergency preparedness of the plant. A high-resolution QRA will also make it possible to evaluate the effect of downgraded situations during the operational phase such as periods with impaired barriers.
Offshore O&G QRAs have typically a much better risk resolution than onshore process plant QRAs. Many onshore process plants could benefit from a better risk resolution to:
Obtain quantitative modelling of domino effects rather than qualitative;
Guide design of process plant to obtain an ALARP design;
Provide valuable input into near-miss investigations;
Provide important input for local emergency preparedness on site;
Supporting decision-making in operational effects i.e. evaluating effects of temporary impaired barriers etc.
It is important before starting a QRA project to determine if the QRA is "only" performed to meet legal requirements or if additional benefits shall be obtained from the QRA. The outcome of this decision can impact both the planning and scheduling of the QRA project and needs to be taken upfront of building the plant.
Depending on the complexity of the process plant, its surroundings, and its hazard potential the right answer could be to go for a "high" resolution QRA whereas in other cases it may be considered a waste of resources to go beyond the legal requirements. The important thing is that an enlightened decision is taken early in the project of building an onshore process plant.