Introduction to Fire Safety Management

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.

---

Introduction to Fire Safety Management
92
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.