Marvin Rausand - Risk Assessment

Figure 5.1The ALARP principle.

The ALARP principle is relevant primarily for risk to humans. The vertical axis in Figure 5.1(not shown) is thus a measure of individual risk, such as AIR. The UK HSE has defined the upper and lower limits for AIR in Table 5.1.

Table 5.1ALARP limits in the United Kingdom.

Limit	Probability per year	For whom
Upper	1 in 1000		The workforce
	1 in 10 000		The public (for existing industrial plants)
	1 in 100 000		The public (for new industrial plants)
Lower	1 in 1 000 000		The public

The risk is usually in the ALARP region, which is the region where the risk must be reduced to an ALARP level. The principle as such does not define what is “reasonably practicable.” Additional guidance is therefore required. An important first principle in ALARP is that risk reduction measures should be implemented, unless there are good arguments for not doing so. This means that if no evidence is available to show that a risk reduction measure is not practicable, it should be implemented. Some other aspects that are important are the following:

1 Good practice should be adhered to when deciding on what risk reduction measures to implement. The argument that it is not practicable (due to cost or other reasons) to implement a measure cannot be used if it is common practice to do it. Good practice can be found in standards, industry guidelines, regulations, and other types of documents.

2 Systematic identification of possible risk reduction measures should be undertaken. This needs to be based on a good understanding of the accident scenarios and how they can be prevented or mitigated.

3 The severity of the hazardous event in question. The more serious the event is, the more is expected to be done to reduce risk.

4 The state of knowledge about the hazardous event, and the availability and suitability of ways of preventing or mitigating its effects. If it is practically difficult to implement measures to reduce risk, this can be taken into account.

5 The cost of preventing the hazardous events or mitigating its effects.

In practical application of the ALARP principle, there tends to be focus on the final item in this list and how cost–benefit analysis supports decisions about whether to reduce risk or not. The other points clearly show that many other aspects need to be considered and that cost–benefit assessment by no means is the only input into the decisions.

ALARP implies that a cost–benefit approach can be adopted to decide what constitutes a “practicable” level of risk. In the ALARP region, the notion grossly disproportionate is central. It requires that a risk reduction measure be implemented if the cost of the measure is not grossly disproportionate to the benefit gained. A disproportion factor may be calculated as

(5.2)

The cost of the risk reduction measure is an estimate of the total cost, which covers not only purchase, installation, and training but also cost implications related to the operation of the system, such as reduced productivity. If reduced productivity has a strong influence on a decision not to implement, the company should show that phasing or scheduling the work to coincide with planned downtimes (e.g. for maintenance) would not change the balance. In the cost estimation, it is common to take a societal view on the costs. For a company, cost associated with accidents can often (at least partially) be covered by insurance. There may also be other effects that reduce the net cost. For society at large, this is not relevant, because insurance is just a redistribution of the costs on many instead of just one.

The benefit of implementing a risk reduction measure is an estimate of the “cost” reduction implied by fewer injuries and fatalities (expressed through VSL) and also of possibly reduced input resources and/or improved system productivity.

The evaluation of disproportionality is carried out by first defining a disproportionality limit . If the actual factor as calculated by ( 5.2) is less than , the risk reduction measure should be implemented, and if , it should not be implemented. A disproportionality limit of, for example, means that for a measure to be rejected, the costs should be more than three times larger than the benefits. There are no strict authoritative requirements on what limit to employ, but it is reasonable to use a higher value of for high risk (i.e. close to the upper limit) than for lower risk (HSE 2001).

A challenge with the cost–benefit approach is that it raises the problem of expressing not only the costs but also the risk reduction benefits in monetary terms. This is a particularly sensitive issue when it comes to putting a value on human life as was discussed earlier. To guide decision‐making, some companies use internal criteria for the value of human life. An alternative to such explicit valuation of human life is simply to calculate the cost–benefit ratio for any risk reduction measure and to look out for any clearly unreasonable situations. If the value of life is not quantified, then resources, which also have value, cannot be allocated rationally to develop and implement countermeasures to protect life. For other types of consequences, such as environmental damage, costs also need to be calculated. Clean‐up costs are sometimes used in connection with spills, but these costs do not reflect irreversible damage to species affected by spills.

A particular problem when a comparison of cost and benefit is done in this way is that the costs are deterministic, whereas the benefits are probabilistic. If we decide to implement a risk reduction measure, we know that there will be a cost associated with it. The benefit we gain is a reduction in the probability of an accident occurring or a reduction in the consequences, should an accident occur. Often, the probability of an accident is very low – regardless of whether individual risk measures are introduced or not – such that the accident will not occur even if the risk measure is not introduced. The benefit is thus purely probabilistic.