Key points
Since the COMAH regulations changed in 2015 many sites have become COMAH establishments, or moved levels. This article explores the learning from working with companies to allow for efficient safety report production. The article will also explore interactions with other operators on the site and inclusion within the site reports.
History:
The Control of Major Accident Hazard Regulations (COMAH) where first implemented within the UK in 1999. This was in direct response to the update of the EU Directive update (SEVESO II). These replaced the Control of Industrial Major Accident Hazards Regulations 1984 (CIMAH).
In 2015 the COMAH regulations were updated in line with the revised SEVESO III directive from the EU. The main change relevant to the article in the COMAH regulation update was a change to the quantities of materials which can be used on a site before the regulations are applicable. The result of the changes mean that some previously non-COMAH sites moved to Lower tier and some lower tier became upper tier, without any change to their operations.
Safety studies have followed similar methods of evaluation since the inception of Hazard and Operability Studies (HAZOP) and Hazard Analysis Techniques (HAZAN) recorded in Trevor Kletz’s (ICI) Institute of Chemical Engineers (IChemE) book ‘HAZOP & HAZAN- Notes on the Identification and Assessment of Hazards’ in 1983 and have traditionally been heavily utilised for feeding into compliance demonstration for COMAH requirements on high hazard facilities.
Safety Study Records:
The first part of COMAH analysis is the review / creation of the safety analysis for the facility. The recording of these safety studies will shape the quality of the safety documents. The temptation is that the study should be completed as quickly as possible whilst satisfying the requirement for completing the study.
Some study consequence detail solely relies on the completion of risk ranking on risk matrices that on occasion can be ‘vague’ in definition (i.e. likelihoods forming opinions ‘has occurred in industry’, ‘has occurred in the company’ etc).
Satisfying the requirement to derive the reduction of risk from the causes using safeguards to an acceptable risk level, however this may not be sufficient to confirm which scenarios could lead to a major accident hazard scenario.
The correct definition of the consequences will aid the derivation of the major accident hazard scenarios, with causes, safeguards and anticipated feasible consequences. Good examples of the safety studies includes sufficient details to allow this derivation without re-visiting the whole study once more.
The scenarios in which material could be released would benefit from the detail of the scenario with some precision (or at least best estimates):
- Material released;
- Amount released;
- Expected hazard scenario (fire case, toxic unignited release);
- Ultimate consequence (effect on personnel (how many, location, population etc.), effect on environmental receptors).
Once the study team have worked through a few examples the recording becomes quicker, thus counteracting the argument that including such detail significantly increases the study time. Similarly with risk ranking on risk matrices being completed within the study workshop versus outside of the session.
Risk Matrices have their place for scenario screening but this topic itself can be a whole complete article so will not be expanded on any further within this article.
Representative Set Development:
The detailed safety study documentation allows for the ability to develop representative sets for inclusion within the Safety Report (upper tier COMAH) of Major Accident Prevention Plan (MAPP) (lower tier COMAH) in a more timely manner.
Most software available will allow for the filtering / exporting of data to allow the scenarios to be refined. By refining the data into similar scenarios allows the person determining the hazardous scenarios and subsequent representative sets can be completed without the need for additional evaluation time.
Incomplete safety study documentation will lead to rework of scenarios meaning that the list of scenarios would initially be longer to evaluate and determine whether the scenario was sufficient to become a major accident hazard in respect of the materials, quantities and consequences involved.
The list could be anywhere as much as double the required length after the first run through, if the scenarios have needed to be filtered based on risk ranking results alone.
Additional considerations for representative sets ought to be the effects of other operators in close proximity to your operations, either on your site or in close geographical locations. The relationships between organisations should be such that emergency scenarios on each side of the fence are adequately shared to allow the correct definition which could lead to major accident hazards, ensuring that appropriate controls are in place.
Learnings:
Safety studies with minimal data for the consequence definition related to causes leads to the reliance on the correct use of risk ranking. The risk ranking with the criteria being defined for loss of life or harm to the environment can be based on the judgement within the study team, generally an opinion on the day rather than a robust consistent approach. The definition of COMAH scenarios is not always clear, for example:
‘Loss of containment leading to environmental harm.’ OR ‘Loss of containment leading to harm to an operator.’
The correct definition including the details listed in the previous section will allow the person compiling the scenario list for inclusion within the COMAH sets with much more clarity and repeatability between scenarios. This can include the findings of study pre-analysis if conducted (environmental harm category in line with Chemical Downstream Oil Industries Forum (CDOIF) guidance, COMAH substance information). For example:
‘Loss of containment of 5 tonnes of Methanol, ignition and fatal injury.’ OR ‘Loss of containment of 5 tonnes of Methanol, passed to River Example and MATTE Level C.’
The pre-screening of quantity released, through dispersion modelling, can allow the safety study to screen out many scenarios within the safety study. This can be both safety and environmental aspects of evaluations.
Studies in which the definitions have been more thorough have been observed to save time in the representative set development. Therefore, it is a worthwhile exercise to complete.
Conclusions:
The correct recording of safety studies for COMAH facilities will allow the easier derivation of the representative scenarios which need to be evaluated fully to confirm the residual risk levels for the facility. The additional benefit of allowing more transparency between the different studies and the final safety report / MAPP.
When completing the safety studies it is beneficial to schedule in time to complete some up front modelling so that the consequences detailed are more informed, thus allowing some screening at these earlier stages of evaluation.
The adoption of the better definition will all in all reduce the burden for initial report production and periodic updates when changes occur.
