Good Safety practices equal Good business
The “Good Safety practices equals Good business” saying is probably in every safety engineer’s mind when promoting new initiatives within a company. However, sometimes it’s hard to speak a common language and explain the benefits, especially to senior management. In safety studies and QRA you quantify the losses from postulated accidents but when looking beyond safety and compliance it’s hard to justify investment for losses you haven’t had but may have. When performing RAM studies, the benefits are easily quantifiable in more common language – money. Lifecycle cost studies are perfect to quantify the return on investment (ROI) when changing design configurations or performing maintenance resource optimisation.
History has shown us that safety incidents have a direct impact on production. According to the AGCS Global Claims Review 2015, “Fire & Explosion” has been the top cause of Business Interruption (BI) loss by number of claims and value in the energy sector (followed by machinery breakdown and power interruption).
So what are the business interruption costs that relate to safety hazards?
In the energy sector, the average claim is almost €4m for in BI. According to Allianz Global Corporate & Specialty, “Fire & Explosion” accounted for 59% of their 1,807 BI claims globally, in a five-year period. An interesting aspect of these data is that “Fire & Explosions” incidents costs €1.7m in business interruption losses on average.
Moreover, according to AGCS insurance claims analysis, the vast majority of insured BI losses are not caused by natural catastrophes. Non-natural hazard events account for 88% of BI losses, according to value. The main causes of BI for onshore energy in recent years have been fire and explosions, often linked to a leak of hydrocarbons related to maintenance problems or faulty equipment.
Most of these incidents could be avoided if better understanding of the impact of fire and explosions were developed and preventive actions were put in place. Quantifying the impact of an event can lead to consideration of more preventive measures or greater mitigation to minimise the impact and thereby losses. Assessing the likelihood along with severity of an event helps prioritisation of efforts!
This is where accurate consequence and quantitative risk analysis (QRA) delivers significant benefits.
One example these methods play an essential role is when deciding on the location and construction of occupied buildings in the vicinity of hazardous installations. A number of factors must be considered during the design and operational phases – temporary buildings, for example.
Traditional QRA calculates the vulnerability for people indoors as a function of the probability of death when a certain level of hazardous effect is exceeded. The hazardous effects range from explosion overpressure, radiation from fires, flame impingement or toxicity. The main problem with the traditional approach is accounting for the impact to people without considering the building type they reside within. This is obviously a significant limitation to using the results of traditional QRA in selecting appropriate building types. It also limits the ability to define the safest place to locate buildings from the standpoint of risk to the occupants.
So, it is clear that the building location and construction must be taken into account in order to develop a full picture of risk to building occupants. The 3D explosion modelling capability in both the Phast and Safeti programs can be used to describe the impact of building vulnerability and occupied building analysis.
Let’s take a look at an example – two control rooms and an office building have been defined in Figure 1. The two building types will assist us in demonstrating the influence of different building vulnerabilities. This example accounts for a range of pressurised methane gas release scenarios which are located at the centre of each obstructed region (indicated by the pink circles in Figure 1). These scenarios are defined for a range of hole sizes, each with a typical frequency of occurrence assigned to it. In addition to this, information related to the weather, population, ignition points and wind rose data are used to fully describe the plant.
In addition, some obstructed regions are defined to account for confinement of the vapour cloud – a detailed explanation on how 3D explosion works can be found here.
Before looking at the results, it is important to note that building types will present a different probability of death for a specific explosion strength. Extensive guidance has been provided in this topic by a number of sources including, CCPS (1994), CIA (1998), API (2003) and BEVI (2009).
So, for our example, two methods are accounted for: typical building vulnerability data for the discrete over-pressure method (similar to the purple book approach) and, for the interpolated vulnerability method, similar to the CIA approach.
By using a range of building types with different vulnerability properties for the control rooms and office in the model, we can assess the overall societal risk against some suitable acceptance criteria. Figure 3 shows F-N curves for 4 different building configurations.
The Blue F-N curve is for a simple discrete overpressure vulnerability model and we can see that the criteria shown are not satisfied.
Moving to an interpolated overpressure vulnerability model where different buildings types are taken into account – Orange, Light-blue and Green FN Curves – we can that the overall societal risk has been reduced.
The scenario where brick buildings are used to define the building types (Orange F-N curve), the risk criteria is not met. It still exceeds the criteria in the 4 to 10 fatality range.
By replacing the control rooms with stronger concrete structures (Light-blue FN Curve), we can see that the overall societal risk reduces further and is now within the ALARP region! The final curve illustrates how further risk reduction can be achieved by replacing the control rooms with blast resistant structures and upgrading the office block to a concrete construction. Further cost-benefit analysis can then be carried out to assess whether the risk reduction which can be achieved warrants the extra construction costs!
If you interested in looking into the full paper, contribution from Dr. Nic Cavanagh, please download it from here.
Many times quantifying the benefits of consequence and QRA studies is extremely difficult – most times the technical terminology gets in the way and the message gets lost. Cost data gives a clear message and can be used to easily highlight the benefits of performing risk analysis.