6 Risk assessment and prioritisation

Note: items in italics throughout the report are those for which further definition and / or information can be found in the glossary section.

Estimated or calculated values of risk (as outlined in section 5) are the fundamental prerequisites for prioritising natural hazard management efforts. However, quantitative or qualitative measures of risk are insufficient in themselves to justify management decisions.

Risk assessment is the process of determining the importance and relevance (significance) of calculated or estimated risk,with reference to the social and physical context within which it occurs. This process determines whether risk is intolerable, tolerable or acceptable. Risk assessment may involve considerations of risk perception, risk / benefit analysis, risk acceptability and risk comparison, with the aim of prioritising natural hazards and developing some appropriate level or form of response. Implicitly or explicitly, this process involves balancing risk with the benefits associated with exposure to that risk.

6.1 Risk perception

There are factors other than an objective calculation of risk that dictate how individuals or communities perceive the risks from natural hazards and, ultimately, whether they view them as intolerable, tolerable or acceptable. The perception of risk, rather than its absolute value, determines whether local government authorities judge that a given risk requires management, and influences the management approach adopted. Furthermore, the perception of risk by the public determines the extent to which they will accept the proposed level of management. Techniques for ranking and evaluating risk are given in following sections. Factors influencing public risk perception are given in Table 9.

Table 9. Factors influencing public risk perception (Smith, 2004:47)

 

Factors increasing risk perception	Factors decreasing risk perception
Involuntary hazard (radioactive fallout)	Voluntary hazard (mountaineering)
Immediate impact (wildfire)	Delayed impact (drought)
Direct impact (earthquake)	Indirect impact (drought)
Dreaded hazard (cancer)	Common hazard (road accident)
Many fatalities per event (air crash)	Few fatalities per event (car crash)
Deaths grouped in space/ time (avalanche)	Deaths random in space/time (drought)
Identifiable victims (chemical plant workers)	Statistical victims (cigarette smokers)
Processes not well understood (nuclear accident)	Processes well understood (snowstorm)
Uncontrollable hazard (tropical cyclone)	Controllable hazard (ice on highways)
Unfamiliar hazard (tsunami)	Familiar hazard (river flood)
Lack of belief in credibility of, and messages from, authority (private industrialist)	Belief in credibility of, and messages from, authority (university scientist)
Much media attention (nuclear power plant)	Little media attention (chemical plant)

Risk communication and public awareness programmes should aim to close the gap between objectively evaluated levels of risk and perceived levels of risk.

6.2 Risk / benefit analysis

Risk / benefit analysis involves assessing the benefits of inhabiting areas at risk, of carrying out various activities in them and of putting oneself at risk, against the costs of damage and losses in disaster (including estimated future costs of the event) and of mitigation works.

The perception of risk, and ultimately its acceptability, is influenced implicitly or explicitly by the risk / benefit ratio. This means that the level of risk an individual or community is prepared to accept or tolerate depends to a large extent on the benefits associated with exposure to that risk. A good example of this is one of the highest risk activities an individual is likely to expose themselves to i.e. travelling in vehicles. The risk associated with this activity has continued to be tolerated because of the benefits it accrues.

An increasingly accepted standard to judge the acceptability of societal risk is implicit in the ALARP (As Low As Reasonably Practicable) principle outlined below (Figure 2). The principle recognises that above a certain level (threshold) risk will not be accepted by individuals or the community, despite any benefits associated with being exposed to that risk. There is also a low level of risk which will generally be accepted without resort to specific mitigating measures - this is represented by the acceptance threshold. Between these two thresholds is a region where risks may be tolerated as long they are kept As Low As Reasonably Practicable. Risks within this region are tolerated because of their risk / benefit ratio or because the costs of further mitigation are not acceptable. A quantitative approach to determining the ALARP thresholds is given in section 6.3.

Figure 2. ALARP principle

View full sized figure (including text description)

Source: AS/NZS (2004b)

 

6.3 Risk acceptability

Before risk reduction options are adopted, risks need to be evaluated in terms of their acceptability. Risks can be classified as intolerable, tolerable, or acceptable.

For a few hazards, standards are available for establishing the acceptability of risk. The most common approach to deriving acceptability standards is represented by the ALARP principle outlined above (refer to section 6.2).

The ALARP (As Low As Reasonably Practicable) approach presents a standard for determining the acceptability of risk. It is based on the assumption that not only the magnitude of impact, but also its frequency, will determine the acceptability of risk. In its original form (referred to as a "FN diagram", see Figure 3), "magnitude" is represented by the number of deaths in a given event and "frequency" by its probability of occurrence in years. It should be noted that the most commonly used version of this approach was developed for risk from accident in the nuclear power industry (please note that the example given in Figure 3 refers to landslide risk). The acceptability thresholds represented will vary for different sources of hazard and different cultural and social conditions. The approach identifies an upper threshold above which risks are generally unacceptable, and a lower threshold below which risks are generally acceptable and require no action. Between these two thresholds is a region where risks are tolerated only on the basis that they are kept As Low As Reasonably Practicable. The principle also takes into account that tolerability will be influenced by the risk / benefit ratio associated with exposure to that risk and the cost benefit / ratio of mitigation measures.

Figure 3. FN Diagram for landslide risk

Source: Australian Geomechanics Society (2000)

6.4 Risk as a guide to management approaches and treatment options

Conceptually, the risk benefit / ratio can be used as a guide to management approaches / treatment options.

Figure 4 illustrates a conceptual relationship between the risk / benefit ratio (see section 6.2) and the appropriate (or acceptable) level of treatment. The main concept is that up to a certain limit it is not the actual level of risk that dictates the acceptable treatment option, but rather the degree to which that risk is offset by the benefits accrued. Thus the type and severity of the treatment option from 'do nothing' to 'prohibition' is a function of the risk / benefit ratio. The flattening of the line on the figure at highest risk values suggests that there might be levels of risk which are so great that offsetting benefits become irrelevant in terms of action taken. While there are no quantitative examples available that have put these concepts into practice, the principle underpins many treatment decisions.

Figure 4. The conceptual relationship between the risk / benefit ratio and treatment options

Source: Crozier (1993)

There are many other factors that are less quantifiable that dictate the nature of treatment options adopted, such as those known to affect risk perception (section 6.1,Table 9). Probably the first decision to be made with respect to the extent and type of treatment is whether there is risk is to human life, as opposed to solely economic values. Clearly cost / benefit ratios of proposed measures as well as available resources are also important factors.

Table 10 provides an example of how risk level helps to determine appropriate management approaches and treatment options, using the example of landslides.

Table 10. Qualitative risk level implications for landslides (Australian Geomechanics Society, 2000)

Risk level	General guide to management implications
Very high risk	Extensive detailed investigation and research planning and implementation of treatment options essential to reduce risk to acceptable levels; may be too expensive and not practicable.
High risk	Detailed investigation, planning and implementation of treatment options required to reduce risk to acceptable levels.
Moderate risk	Tolerable provided implementation plan is implemented to maintain or reduce risks. May be accepted. May require investigation and planning of treatment options.
Low risk	Usually accepted. Treatment requirements and responsibility to be defined to maintain or reduce risk.
Very low risk	Acceptable. Manage by normal slope maintenance procedures.

6.5 Risk comparison

Risk comparison can be used to evaluate and communicate risk, as well as guide management priorities.

Risk comparison can be achieved using either qualitative or quantitative measures of risk. This process aids the understanding of risk by allowing the comparison of a poorly understood risk with a better understood risk. Table 11 shows an example where the risk of death from natural hazards is compared to other forms of hazard.

Table 11. Comparison of individual risk of death from hazards in New Zealand (population 3.5 million), annual average between 1840 and 1990 (Ministry of Civil Defence, 1994)

Hazard	Deaths per year	Probability of death per person per year
Smoking	4,000	1.1 x 10-3
Road accident	600	1.7 x 10-4
Suicide	380	1.1 x 10-4
Falls	300	8.6 x 10-5
Drowning	120	3.5 x 10-5
Homicide	50	1.4 x 10-5
Fire	32	9.0 x 10-6
Natural hazards	6	1.6 x 10-6

6.5.1 Risk comparison methods

FEMA ranking and the SMUG system are two useful methods to compare risks, and are summarised in this section.

It should be noted that risk comparisons made using these methods reflect only the parameters / criteria used, therefore other factors such as risk/benefit analysis (section 6.2) should also be considered when prioritising risk management actions.

FEMA ranking

This system has been described in section 5.3.3 of this report.

SMUG system

The SMUG system is used to compare and prioritise hazards based on the Seriousness, Manageability, Urgency and Growth (SMUG) characteristics of each hazard. (Note: SMUG can also be used as a method of qualitative risk calculation - refer to section 5.3.3).

SMUG uses the following criteria:

Seriousness the relative impact in terms of people and / or dollars

Manageability the relative ability to reduce the risk (through managing the hazard or the community or both)

Urgency the measure of how imperative or critical it is to address the risk (associated with the probability / likelihood of the risk from the hazard, including return period considerations)

Growth the rate at which the risk will increase (through an increase in the probability of the extreme event occurring, an increase in the exposure of the community, or a combination of the two).

Each of the criteria is applied to each risk. Generally, one risk is chosen as a benchmark on the basis that it is likely to represent the highest score and the other risks are compared to this. Once all of the risks have been rated, a numeric score is assigned to the results:

Seriousness	High = 3	Medium = 2	Low = 1
Manageability	High = 1	Medium = 2	Low = 3
Urgency	High = 3	Medium = 2	Low = 1
Growth	High = 3	Medium = 2	Low = 1

If this level of comparison shows risks falling within a very small range then more refinement may be desired. The risk evaluation can be made more specific by breaking down the analysis of each SMUG component into detailed rankings from say 1-5, or considering part scores as opposed to whole numbers. Consideration must be given to whether this level of complexity and detail is required.

The Ministry of Civil Defence and Emergency Management (2002a) provides a good overview of the SMUG system, and identifies detailed ranking criteria, largely using AS/NZS (2004a) descriptors. These recommended criteria are shown in Table 12.

The SMUG system advocated by the Ministry in 2002 has been subsequently modified by many CDEM groups. Many groups removed the 'urgency' component (as it may be adequately covered under 'seriousness') and expanded the 'manageability' table to include the subcomponents of 'difficulty' (how difficult the hazard is to manage) and 'effort' (how much effort is currently being put into managing the hazard). Manageability ratings were then given to each subcomponent for each of the 4Rs (reduction, readiness, response and recovery), thereby a manageability rating derived from eight manageability values.

An example of the application of the SMUG detailed ranking criteria by Taranaki Regional Council is given in Figure 5.

Table 12. SMUG detailed ranking criteria (Ministry of Civil Defence and Emergency Management, 2002a)

 

Seriousness

Where each impact area (human, social, economic, infrastructure, geographic) is given a rating between 1-5 using the AS/NZS (2004a) descriptors then the average taken.

	Descriptor	H	S	E	I	G	Av
1	Insignificant
2	Minor
3	Moderate
4	Major
5	Catastrophic

Manageability

Management difficulty	Current effort	Rating
Low	High	1
Low	Med	2
Med	High
Med	Med	3
High	High
Low	Low	4
Med	Low
High	Med
High	Low	5

Urgency

Level rank	Descriptor
A (5)	Almost certain
B (4)	Likely
C (3)	Possible
D (2)	Unlikely
E (1)	Rare

 

Growth

Event occurrence probability rise	Changing community exposure	Rating
Low	Low	1
Low	Med	2
Med	Low
Med	Med	3
Low	High
Med	High	4
High	Low
High	Med
High	High	5

Figure 5. Example of the application of SMUG detailed rating criteria

View full sized version of figure (including text description)

Source: Taranaki Regional Council (2004b)