There 5 main components of QRA. These are Hazard identification, consequence modelling, frequency modelling, risk estimation and risk management and mitigation.
The first step of QRA is hazard identification in which potential hazards for a proposed installation are identified and analysed. The procedure involves critical examinations of all situations in which potential hazards may exist, followed by disciplined analyses of the combination or sequences of events that could transform these hazards into an accident. These are typically conducted through a series of brainstorming sessions and critical reviews of the process. List of potential hazard associated to every installation and operations are then produced. Hazards that are judged to present a significant risk are further scrutinised in more details. Possible causes are identified to facilitate further stages of QRA.
2. Consequence Modelling
Consequence analysis determines the impact of each hazard scenario identified earlier using various source and dispersion models. These models consider various factors pertinent to each failure case including release conditions, ambient weather and local terrain. Based on the extent of damage, hazard zones are determined.
Hazard zone or range depends on the severity of the effect. For example, for vapour cloud explosion, the hazard range depends on the coverage of the area under the cloud. Effect of toxic release depends on the distance the release travels as well as the concentration of chemicals involved. The impact of fire and explosion is dependent of the dispersion of heat in the surrounding of area. In this case, the consequence is dominated by energy released through radiation from the heat sources.
3. Frequency Analysis
In Frequency Analysis, the likelihood each possible incidences are computed using statistical data and design of the proposed system. Two methods are typically used, these are, Fault-Tree Analysis and Event Tree Analysis. In both cases, data on failure frequency for each component that constitute the equipment installed in the plant are obtained from the literature, or based on the manufacturer’s documentation. Note that the data are based on statistics and tend to improve over time due to more reported cases and improvement of technology. It is important to ensure that the source of data must be authoritative.
4. Risk Estimation
Risk to people can be expressed in two complementary forms, i.e., individual risk and societal risk. The former is the risk experienced by an individual person whilst the latter is a risk experienced by the whole group of people exposed to the hazard. All quantitative risk estimates are subject to some degree of uncertainty. The uncertainties arise from a number of sources including:
- Errors, in the formal sense of scientific measurement.
- Uncertainties in the modelling process and in biological effects such as vulnerability.
- Uncertainty in whether or not there is indeed an effect to be incorporated within an estimate.
- Omission of possible causes of risk due to incomplete analysis of the mechanical or engineering sources of plant failure, non-quantification of human error as well as omission of extreme natural causes.
5. Risk Mitigation
Risk tolerability and acceptability varies among jurisdictions. The limits may vary from one country to another but the general principle can be explained by the ALARP framework. Both individual and societal risks are used in deciding the tolerability. In Malaysia, as long as the location specific individual risks (LSIR) is less than one fatality in a million population per year (1 x 10-6) for residential population, the proposed installation can be considered tolerable (See guideline from Department of Environment).
In addition, various mitigating measures should also be proposed in order to reduce the various risks. These include engineering and administrative controls that are to be observed during construction, installation and operation of the proposed facility.