Principles of risk assessment
5.5 The risk assessment process
5.5.4 Evaluating residual risk
The fourth stage in the risk assessment process is to evaluate the level of residual risk, i.e. taking into account any current control measures (workplace precautions and risk control systems) that are already in place. Some risks may also be adequately controlled leaving only a low residual risk; these will require no further action other than a review of the validity of the assessment on a periodic basis (see below).
Hazard presents a risk which causes an accident/incident which in turn causes injury, ill health, damage, or loss.
As an example, oil can be considered as a hazard. When it is in the container it has potential but no likelihood of achieving its potential, therefore there is no risk. When it is spilt on a table there is a small possibility of causing minor harm (ill health perhaps), there-fore the risk exists but is low. When the oil is spilt on the fl oor there is a better chance of it causing greater harm (slip), so the risk could be termed medium. When the oil is spilt on a heater unit or close to other potential ignition sources there is an even greater chance of it causing serious harm or even death (start-ing a fi re), so the risk could be termed high or very high. It can be seen then that a hazard is an inherent quality of something, whereas risk is dependent entirely upon the circumstances surrounding the hazard.
An example is a contractor who has been engaged to install IT cabling within an exist-ing workplace. The work location will include the main access corridor to the main stair-case within the building. Those that could be put at risk from the operation would be the contractor, the occupier of the premises, vis-itors to the premises, other occupiers (if the building is occupied by more than one) and other contractors.
Risks could include the more obvious fall-ing from height, fallfall-ing materials, slips and trips. In addition the less obvious, but nev-ertheless safety critical, is the potential obstruction of escape routes and breaches of fi re compartment walls. Many people could be affected during the works and potentially many more over the life of the building if fi re compartmentation is breached by running cables or trunking through fi re breaks.
Principles of risk assessment
As no numerical scales have been introduced with the above simple estimator it is diffi cult for a precise indication of the level of risk to be determined. Neither does this approach provide an easy mechanism to confi rm if the risk has been reduced to the lowest level reasonably practicable.
However, by introducing the above subjective estimations of the two elements of risk into a simple risk matrix, a qualitative assessment of risk can be made.
Figure 5.10 is an example of a simple risk matrix.
When undertaking risk assessments for a number of work activities the application of a consistent method of qualitative evaluation of risk will allow for actions to be prioritised.
For example, a systematic qualitative evaluation of an organisation’s risk will result in some risks that are evaluated as ‘high’, some that are ‘low’ and some that are
‘insignifi cant’, management decisions can then be made of a basis of risk vs cost. This qualitative method provides a basic evaluation of risk and will allow an organisation to consider what may or may not be considered a reason-ably practical level of safety.
There are a number of methods for evaluating risk.
The method applied for any particular risk will depend on a number of factors, such as the complexity of the activ -ities carried out and the type and nature of the workplace.
For many of the day-to-day risks that people in the work-place are exposed to, including fi re, a simple qualitative assessment will suffi ce, for more complex risks a quanti-tative or semi-quantiquanti-tative assessment may be needed.
Qualitative analysis
Qualitative analysis describes a quality of the risk.
Typical of the qualities most often described is that of quantum, i.e. size or magnitude. For example, when assessing the means of escape in the event of a fi re, fi re risk has historically been rated as high, normal or low.
Qualitative analysis of risk is a subjective measure, based upon the risk assessor’s judgement. As with all methods of risk evaluation, a qualitative assessment will need to allow consideration of the two aspects of risk, i.e. the likelihood of a particular occurrence and the severity of the consequences.
The HSE, in its guidance document HSG65, have suggested a basic estimator as appropriate for a simple evaluation of risk. The HSE estimator, shown in Table 5.1, uses purely subjective measures of likelihood and uses RIDDOR events to ascertain the qualitative descriptors of the seriousness of the outcome.
Likelihood Severity
High Where it is certain or near certain Major Death or major injury (as
that harm will occur defi ned by RIDDOR) or
illness causing long-term
disability
Medium Where harm will often occur Serious Injuries or ill health causing
short-term disability
Low Where harm will seldom occur Slight All other injuries or illness
Table 5.1 The HSE estimator
Likelihood of an accident/incident occurring Outcome of the potential accident/incident
Major injuries may
occur
Serious injuries may
occur
Slight injuries may
occur High
chance of an event
High risk Medium risk
Medium risk Medium risk
Low risk
Low risk
Low risk
Low risk
Medium chance of an
event Low chance
of an event
Insignificant risk
Figure 5.10 Example of a simple risk matrix using the two aspects of risk to determine the magnitude of risk, expressed in qualitative terms
Qualitative analysis – describes the quality of risk using words.
Quantitative analysis – quantifi es the risk with numerical data.
Semi-quantitative analysis – uses numbers to quantify qualitative data.
Quantitative analysis
In this method risk is evaluated, not by subjective judge-ment, but by numerical data. Quantitative evaluation of the risk is therefore more demanding than a qualitative approach but provides a more rigorous evaluation.
A quantitative approach, when used to recalculate the effectiveness of controls (after implementation of risk control measures) also provides clear evidence that confi rms that risks have in fact been reduced.
It should be noted that specifi c raw data is required when using this technique to analyse the magnitude of risk. Data that will be required to allow a quantitative evaluation of risk will be found in the local and national records of:
➤ Hazard reports
➤ Injury accidents and incidents
➤ Ill health and sickness
➤ Health monitoring systems
➤ Environmental monitoring systems
➤ Fire-related incidents.
Details relating to the types of safety events that occur together with the likelihood of occurrence gathered from such records are an essential tool when evaluating risk as they provide statistics that confi rm the severity (or potential severity if a near miss) and how frequently the event occurs.
For most organisations gathering statistics in relation to fi re to provide a quantitative evaluation of fi re risk can be diffi cult to achieve. However, data from the
insurance industry via the statistics produced by the Fire Protection Association (FPA) and from the fi re services via DCLG is available. Generally the statistics produced by the insurance industry focus upon estimated fi nancial losses, whereas those produced by DCLG relate to fi re deaths and injuries.
Both sets of statistics provide information on the causes of fi re and the numbers of fi res occurring in given areas or sectors. As the severity rating or potential outcome of a fi re is death or multiple deaths, reducing the likelihood of a fi re occurring and managing the secondary hazards associated with fi re (not being able to escape, etc.) must be seen as being a high priority.
Gathering meaningful data that gives a clear indica-tion of frequency and severity is very often diffi cult (unless the industry, sector, or organisation is large enough to have suffi cient statistics available).
While personal injury accident data and ill-health data are generally readily available due to the frequency of events, the statistics relating to the numbers of fi res and false alarms refl ect the relatively low numbers of incidents; however, the outcomes are quite often more severe, so pure quantitative analysis can be very often diffi cult to achieve.
Pure quantitative analysis is generally only needed in a small select group of high risk industries, such as nuclear and offshore. The fi nal and probably most widely used evaluation of risk is a combination of the qualitative and quantitative approaches and is referred to as ‘semi-quantitative’ risk.
Figure 5.11 Graph showing the initiators of fi res that resulted in more than £250 000 damage
0 10 20 30 40 50 60 70
number of fires
Deliberate ignition Electrical
Friction, heat and sparks Smoking materials
Naked lightCooking
Spontaneous combustion Oil
Fireworks
LPG cutting equipment Spread from secondary fire
Under investigation Unknown
source
How the £250 000 plus fires in 2004–2005 started
Principles of risk assessment
Semi-quantitative analysis
The use of semi-quantitative techniques for risk assess-ment is widespread and it is often referred to as a quan-titative method; however, it is easily seen that risk is expressed as a numerical value, the estimation of the magnitude of the risk is in fact subjective and therefore qualitative. A semi-quantitative evaluation of risk allows numerical values to be assigned to both severity and likelihood in the absence of data.
Like the qualitative approach, the semi-quantitative technique is essentially subjective and is based upon the risk assessor’s personal interpretation of the level of risk, but with the added assistance of a risk matrix so that a risk factor can be calculated numerically.
For example, the likelihood of a particular event occurring is assigned numerical values as shown in Table 5.2.
While it is easy to differentiate between the two opposing defi nitions at the end of each scale, there is often confusion with the words used to separate out the mid range.
There is no laid down criteria for either the words used to describe the value or the numerical values within the scale. A 5 by 5 scale is arbitrarily used here, though there are a wide variety of scales used from 3 through to 8 and sometimes even 10.
Just as the likelihood rating is assigned numerical values, so too is the severity rating. Again using a scale of 1 to 5 and adapting the HSE’s simple qualitative estima-tions of severity, an example of how numerical values may be assigned is shown in Table 5.3.
In addition to the severity of the outcome of an event for an individual, the total numbers of people that may be affected must also be considered, as in the case of calculating the likelihood of the rise in fi res started by discarded smokers’ materials in places of public entertainment (if a smoking policy exists). The resultant effect upon persons being able to escape safely may also revolve around the numbers of persons present.
Having assigned numeric values to each element of risk, a risk matrix can then be completed to provide a calculation of an overall ‘risk factor’.
The resultant risk matrix can be used to provide and develop an action plan which may also be assigned numbers so that priorities can be identifi ed. Risk matrices are often colour coded to provide a visual concept of whether or not the residual risk is tolerable or acceptable.
An example of a semi-quantitative risk matrix, which incorporates the numerical scales discussed above, is shown in Figure 5.12.
The qualitative description of the severity Numerical value that
of a particular event may be assigned
First aid injury or illness requiring minimal attention 1 (plaster, etc.)
Minor injury or illness. Includes those where a 2 person could spend up to 3 days away from work
‘3 day’ injury or illness. Those where the person 3 would be off work more than 3 days (as per RIDDOR)
‘Major’ injury or illness (as defi ned per RIDDOR) 4 Fatality or disabling injury or illness preventing return
to work 5
Table 5.3 An example of how numerical values may be assigned
The qualitative description Numerical value that of the likelihood of a may be assigned particular event occurring
No evidence of occurrence 1
Foreseeable but remote 2
Has occurred but only infrequently 3 Has occurred fairly frequently 4 Has occurred regularly and will 5 occur again
Table 5.2 Example of numeric values given to subjective, qualitative descriptions of risk, of a typical semi-quantitative assessment
Note: When assigning numerical scales to severity the assessor must take into account chronic or long-term effects of any particular event and/or the effects of the long-long-term exposure to a perceived ‘low’
hazard.
Once individual risks have been evaluated using a semi-quantitative approach, it is possible to establish formalised action plans based upon the risk grading – a sample is shown in Table 5.4.
The type of risk analysis or evaluation method used will vary from organisation to organisation; whether a numerical matrix or high, medium, low matrix is used will very much be dependent upon the risk levels of an organisation’s activities and workplaces.
5.5.5 Applying additional risk control measures