Domino Effects - opportunity or hindrance for COMAH sites?
The term ‘domino’ has long been a part of the Control of Major Accident Hazards (COMAH) Regulations but its interpretation and use has evolved over time. As such, it is not always clear what the requirements of COMAH and non-COMAH establishments with the potential for major accidents are, with regard to assessing domino effects.
The COMAH Regulations 2015 define a domino effect as an “increase in risk or consequences of a major accident … where the risk of consequences of a major accident may be increased because of … (a) geographical position of establishments; (b) proximity of establishments to each other; or (c) inventories of dangerous substances held by establishments” (Para 24. (1) and (2)). In short, we are talking about potential for escalation between two or more hazardous installations and, in recent years, this has taken to include both COMAH and non-COMAH establishments (Schedule 3, Para 3 (c)).
Surprisingly, the Regulations do not specifically talk about ‘escalation’ hazards but do infer that these are important within the site boundary as well; for example, through the requirements for the identification of all areas on the site that can cause a major accident and of all initiators for major accident hazards. Therefore, it could be interpreted that domino effects could exist not only between hazardous installations, but within the boundaries of an establishment itself.
In the following simplified example, A and B represent two hazardous installations sharing a boundary. Along this boundary, both establishments have each placed a single pressurised storage vessel with highly flammable contents. Within the hazard identification studies of each establishment, all major causes for loss of containment are identified, but both studies dismiss the potential for escalation, since each facility consists of a single storage vessel. In this instance, the potential for flame impingement and escalation from one storage vessel to another could result in disastrous consequences and yet a simple, low cost risk reduction measure may not be identified through dismissal of the initiating cause.
In a second example, a hazardous installation has two processes within close proximity. The hazards associated with process 1 are significantly greater than 2, but the likelihood of occurrence is demonstrated to be much lower. In this example, where hazards from process 2 could affect process 1, failure to identify domino effects within the site boundary could omit dominating initiating events and present an inaccurate risk profile.
Furthermore, the examples above could be interchangeable. If, for example, site B does not have a suitable hazard identification and management system, the dominating risk to harm on site A could be from external initiators. Whilst B could be a non-COMAH site, thus not subject to these regulations, the interpretation of ‘duty of care’ under the much broader Health and Safety at Work (etc.) Act 1974 implies that any site, regardless of COMAH status should manage safety risks proportionately. Therefore, it is vital that domino groups are established and effective communication takes place.
For ongoing demonstration of ALARP principles in industry, it is essential that all hazardous installations understand the risks that can be posed to people and the environment, both within and external to the establishment boundary. For small sites, this can be managed through the development of a hazard identification register, safety management system and emergency response protocols. To achieve this, COMAH establishments – some of which have engaged in these practices now for more than 2 decades – could play an important role in helping non-COMAH establishments lower the risk profiles of both sites.
The domino group implementation under COMAH should thus be looked at as an opportunity, not a hindrance.