Risk assessment may have a fundamental part to play in the development of fire engineering designs that deliver adequate fire safety in terms of life safety, business continuity and/or asset protection.
Regulatory regimes generally require that designers and managers of non-domestic premises assess the risk posed by fire in those premises and that they take suitable measures to reduce fire risk to an acceptable level. This can be achieved by:
incorporating fire protection in the design of the premises
implementing effective and appropriate management controls.
Insurers may require fire risk reduction measures as a condition of their insurance, in order to control their exposure to financial loss. Clients may also require designers to incorporate measures to protect their assets in case of fire as a means of ensuring business continuity.
The findings from a risk assessment can be used to inform decisions as to whether fire precautions and fire safety management procedures are sufficient to control fire risks to a satisfactory level, or whether additional risk reduction measures are required.
Risk assessment can also be used to perform a systematic comparison of different fire risk control and/or reduction options so that a decision can be made as to the optimal design or management solution. It is not, however, appropriate to carry out a risk assessment in an attempt to justify a decision that has already been made. Risk assessment is input to the decision-making process, not output from that process.
The techniques used to assess risk vary from very simple qualitative analyses to sophisticated quantitative risk analysis (QRA) techniques of the type commonly found in the nuclear, transport and chemical processing industries. No single approach is correct for all applications. For example, QRA may be inappropriate for assessments where straightforward adherence to good industry practice is reasonable. On the other hand, in some situations, simple checklists (e.g. .tick box. list-based techniques) are likely to be completely inappropriate for assessing complex process fire risks. While risk assessment methodologies vary, they are likely to include the following steps:
(1) Identify the hazards.
(2) Identify the possible consequences and estimate their likelihood.
(3) Evaluate the risk.
(4) Take action to reduce risk to as low as reasonably practicable.
(5) Record the findings.
(6) Monitor and review as appropriate.
Before embarking upon a risk assessment, it is important to determine the scope and purpose of that assessment and, if appropriate, to agree that scope with those who will refer to it . this may include clients, managers, premises owners and regulators.
Almost all risk assessment includes an element of judgment, in either the identification of the hazards, the analysis of possible consequences or the estimation of their likelihood. For this reason, it is important that the risk assessment is undertaken by persons with skills and experience appropriate to the fire risks being assessed. In cases where the assessment involves a straightforward and unvarying application of good industry practice (e.g. government guidance), the assessor might need no detailed knowledge of fire behaviour. However, where the assessment uses techniques that might result in solutions that depart significantly from guidance, it will be necessary for the assessors to have the relevant competence in fire safety engineering and/or fire safety management to appreciate the consequences of that departure on fire risk in the premises. This will require understanding of the fire hazards or fire risks that the guidance addresses and the reasons why the guidance recommends a particular controlling measure. It is only with this knowledge that the assessor can make informed decisions as to the significance of variation from that measure.
It may not be possible for a fire engineer to assess the fire risks associated with certain hazards (e.g. in determining whether certain equipment is .critical. if assessing fire risk to business or process continuity). In this case, it will be both necessary and appropriate to draw on the skills and experience of others in order to undertake an adequate assessment of fire risk.
Hazard, risk and risk assessment
A number of definitions are available for the above, and these concepts are fundamental to any risk assessment process. It is therefore essential to define what is meant by these terms:
A hazard is that which has the potential to cause harm.
Risk is a function of both the likelihood of a specific hazard being realised and the consequence of that realisation.
Risk assessment is the process by which reasonably foreseeable hazards are identified, the likelihood of occurrence of specific undesirable events (the realisation of the identified hazards) is estimated, and the severity of the harm caused is assessed. This may be coupled with a judgment concerning the significance of the results.
In other words, it is through the risk assessment that the risk will be evaluated (using either a qualitative or a quantitative approach).
It is essential that all reasonably foreseeable fire risks are identified and considered in the risk assessment process. It is not always necessary to carry out a detailed risk assessment (i.e. to consider every fire risk or all possible fire risk scenarios). Nevertheless, it should be recorded that those fire risks have been considered and have been found to be .acceptable.. It is important to note that the fact that a specific fire risk was not foreseen does not automatically mean that is was not .reasonably foreseeable. Those undertaking the hazard analysis should have the competence and knowledge to identify the fire risks that need to be assessed, whether they had been foreseen up to that point in time or not.
It is also relevant to point out that the hazard may always exist, but that it is possible to reduce substantially the risk of this hazard being realised (i.e. to take place within a particular environment). It is true that to eliminate the risk to zero is almost impossible. However, to reduce it to an acceptable level is possible, and, in fact, this should be the core aim of a risk assessment.
Defining the scope of the risk assessment
In some cases, it may be straightforward to define the purpose and scope of the risk assessment, e.g. where the assessment is aimed only at satisfying fire safety legislation in simple premises. In other cases, the scope may need more careful definition, especially:
when the purpose of the risk assessment is to support the acceptability of the design of complex or novel buildings
where fire risks to assets or business continuity are being assessed.
In simple terms, the questions that need to be asked may fall into the following types:
risks of what. (e.g. fatality, major injury, loss of assets, loss of business, reputation, etc)
risks to what/whom. (e.g. employees, visitors, members of the public, firefighters, assets or premises)
risks .from what. (e.g. accidental ignition, nature and distribution of potential fire load, arson, etc).
Once the scope and purpose of the risk assessment is defined, this will inform the decision about which technique (or combinations of techniques) is (are) most appropriate for undertaking the assessment.
General acceptability criteria
When undertaking these risk assessments, it is important to understand what can be regarded as an acceptable level of risk from fire.
As mentioned previously, it is practically impossible to achieve zero risk. In reality, society neither expects nor requires such an extraordinary standard of safety. The fire risk has to be acceptable to those who have interest in controlling it. In the case of life safety, it will be the regulations defined by society that define acceptable risk. In the case of insurance requirements, it will be the insurers or their representatives. In the case of risk to business continuity, it will be the management of the relevant organisation. In most cases, the objective will be to reduce risks to be .as low as is reasonably practicable. (ALARP). This involves assessing the fire risk against the effort, time and money needed in order to control it. Thus, ALARP describes the level to which it is expected that fire risks are controlled. If the fire risk reduction benefit is proportionate to the time, effort and money necessary to implement the relevant fire risk reduction measure(s), then that fire risk reduction measure must usually be implemented.
In fire safety, the practical definition of the level of fire risk that can be regarded as ALARP tends to be set by national guidance . adherence to such guidance (where relevant and appropriate) will tend to demonstrate that fire risks from fire are acceptably controlled. Where Duty Holders wish to vary from that guidance, then the normal expectation is that they use alternative risk control measures to achieve the same level of safety by other means.
Assessment techniques
Application of good industry practice
In many cases, it is possible to assess fire risk quite straightforwardly by reference to relevant good industry practice. Indeed, it should be the case that, before any risk assessment is carried out, the assessor should review whether relevant good industry practice exists and, if so, whether it can be straightforwardly applied. It is normally accepted that, if good practice can be applied, it is adhered to(2). The following would be possible exceptions:
If it were to be applied to existing premises, the cost of compliance with the guidance would be grossly disproportionate to the fire risk reduction achieved.
The situation under consideration has inherently and significantly lower or greater fire risk than that for which the good practice was developed.
The operations or works include alternative means of controlling the risks to a comparable or better level.
Good practice encompasses industry and regulatory codes, .approved codes of practice. (ACOPs) and regulatory guides, as well as practices adopted successfully by similar organisations.
Where life safety is concerned, relevant good practice is likely to reflect the minimum expectations of the relevant society (government, the public, regulators, etc). It is therefore of use both to those who will use it directly to assess risk and to those who will assess risk in other ways (be they quantitative or qualitative). When one is able to demonstrate a level of risk equivalent to that represented by the application of good practice in premises typical of the type being examined, then that should be acceptable.
If it is found that a design or management solution results in a situation where fire risk is higher than would be delivered by the application of good practice, then it is questionable whether that can be regarded as acceptable. In practice, if relevant good practice exists and is adopted for all reasonably foreseeable hazards, further detailed evaluation of risk need not usually be made. The risk assessment duty is discharged by the appropriate adoption of that good practice. It is therefore very important to ensure that the good practice is:
appropriate to the activities being considered
up to date
both relevant to and covers all significant fire risks from the circumstances being considered.
Qualitative risk assessment
Qualitative risk assessment (or analysis) can be defined as the application of methods for ranking the identified risks according to their potential consequences. This type of risk assessment relies upon the training and experience of the assessor(s) to:
identify the relevant hazards
make a judgment as to the likelihood of that hazard resulting in harm
assess whether the resultant risk is acceptable.
This will normally be used in conjunction with relevant good industry practice, which the assessor will apply where it is reasonable to do so. Where this is either impracticable or alternative solutions offer the same or a better level of safety at lower cost or in a manner more suited to the operational requirements for the premises, the assessor should use their judgment to determine the acceptability of those variations from good practice.
Good industry practice sometimes gives guidance on how to assess risk and how to apply that risk assessment in order to influence the design in a qualitative but structured manner. For example, BS 7974: 2001(3) as well as the more recent BS 9999: 2008(4) introduce the concept of .risk profiling. as a tool to inform the design of such aspects as means of escape and structural fire resistance. They introduce the concept of .occupancy characteristics., considering whether the occupants are likely to be awake and aware of their environment, and whether they will be familiar with it or not. The concept also considers the probable fire growth rate in the premises (this will necessarily be a matter of judgment) and combines the two to produce a ranking of risk. That ranking is used to indicate recommended design criteria (such as maximum means of escape distance, structural fire resistance, etc).
Variation of the risk profile is possible by application of certain risk reduction measures (such as automatic sprinkler systems or fire detection and alarm systems), which allows more flexibility in the design of other risk reduction measures. In some cases, this allows risk to be controlled by less costly and/or less intrusive engineering measures than would be demanded by a .code-compliant prescriptive solution.
For more details of this approach, reference should be made to the current versions of those standards.
Another example of a qualitative risk assessment is the .risk matrix. technique . commonly called a qualitative risk assessment (or analysis) matrix. The risk matrix is a comparative table in which the likelihood and the consequence(s) are related to each other according to a qualitative raking. This comparison will provide a qualitative estimation of the level of risk. Table 5.1 shows an example of a risk matrix.
In summary, the level of risk will be shown in each cell of the matrix. The level of risk is represented on a scale:
E extreme risk
H high risk
M moderate risk
L low risk.
Though well used as a technique in the assessment of risks for fire safety management purposes, the application of risk matrices to design risk assessment tends to be less useful because of the difficulty of agreeing the acceptability criteria. (For example: Are .medium. risks acceptable? Are .low. risks always acceptable?) It is more useful as a technique for comparing risks than determining absolute acceptability.
In addition to standard and well-known risk matrices, alternative methodologies are starting to be used for the same purpose, such as multi-criteria decision-making models.
Whichever technique is used, these analyses should be documented in a manner that records how the assessment has been undertaken and includes the rationale for concluding that risks are acceptable. Where the risk assessment forms part of the design solution for a building, it should be included in the Fire Safety Strategy document.
Quantitative risk assessment (QRA) and cost.benefit analysis (CBA)
Quantitative risk assessment (QRA) is a technique whereby risks are evaluated by assigning numerical values to hazard, to the probability that the hazard will be realised and cause harm, and to the resultant fire risk. This is carried out in order to be able either to compare risk reduction measures on a .like-for-like. basis or to ascertain whether risks are tolerable in absolute terms. The QRA is commonly used in industries such as nuclear power generation and transportation to assess all safety risks in a structured and rigorous way. It is often used to determine if it is reasonably practicable to make safety improvements in existing or altered conditions, or to define safety objectives for new works.
There are several QRA techniques, such as: hazard and operability study (HAZOPS); standard logical trees, fault tree analysis (FTA), event tree analysis (ETA); and new logical trees, such as the continuum net-value work diagram. The Health and Safety Executive (HSE)(6), the Occupational Safety and Health Administration (OSHA)(7) and the American Institute of Chemical Engineers (AIChE)(8) provide good guidance documents for using such techniques. Figure 5.1 shows an example of an event tree used for describing possible fire scenarios if a fire occurs. Figure 5.2 shows an example of a continuum net-value work diagram, which describes the complexity of human behaviour within fire emergency situations.
The two examples shown in Figures 5.1 and 5.2 illustrate graphically how QRA techniques can be used. For each event represented in each diagram, there would be an associated probability. Therefore, as mentioned before, the assessor will need to use his/her prior knowledge and/ or use some historical data as a basis for estimating the probabilities. (For more complex scenarios, such as large spaces, high-population-density environments, etc, the assessor might use numerical optimisation techniques. It is also relevant to mention that, when using the QRA techniques, the assessor can also perform cost.benefit analysis (CBA) if necessary and/or requested.
Both QRA and CBA need not be restricted to safety-related decision making. They may be usefully applied to decisions concerning property and asset protection as well. For example, using knowledge regarding the probability of a significant fire during a relevant period of time, its consequences and the potential financial loss (in terms of both assets damaged or destroyed and lost revenue), informed judgment can be made as to the practicability and desirability of fire protection as a loss control measure. It may commonly be found that the case for inclusion of such engineering is far stronger as protection to assets than life.
QRA may use statistical or historical data to inform judgment on probability, or expert judgment may be used to assign probabilities to hazards that may result in harm. The resultant risk can be expressed as the likelihood that an unwanted and harmful event occurs in a particular period of time; for example, the probability of a fatality per year of operation might be 1 × 10−7.
In some industries (e.g. transport), guidance exists on the value that society is willing to pay to prevent a fatality as a result of the operations of that industry(10). It is important to note that this does not constitute the .value. of a life . that is unquantifiable . but it does give an indication of the cost that society is willing to pay against an assessed reduction in risk to life, with regard to that industry. This is called the .value of preventing a fatality. (VPF). It can be used in a CBA whereby the cost of the risk reduction measure is assessed against the risk reduction it achieves.
Events that could result in multiple fatalities (e.g. death in fire) typically have value multipliers assigned to the VPF, or have a higher VPF in order to recognise that society has less tolerance of multiple fatalities than it does of single events. Society also expects lower levels of risk exposure for members of the public than for employees. A further value multiplier is often applied to the cost part of the analysis to define the level at which the risk reduction measures are deemed .grossly disproportionate.. Typically, costs (for risk reduction measures) less than three timesthe value of risk reduction achieved are regarded as indicating that it is reasonably practicable to implement that risk reduction measure. However, this does not mean that, if cost is more than three times the value of risk reduction that would be achieved, then it is ALARP not to implement that measure . other criteria (such as societal concern or comparison to relevant good practice) might apply.
Guidance on the application of probabilistic risk assessment is given in PD 7974-7(11). This document advises that it is most straightforward to apply QRA and CBA where comparisons are being made of alternative risk reduction measures (e.g. a fire engineered solution compared to a .code-compliant. one), but that establishing .absolute. quantified values for acceptability is far less straightforward. If contemplating the use of such an analysis, it is therefore important that the techniques to be used and the input data (including the VPF and application of all relevant .value multipliers.) are agreed with all those with an interest in controlling fire risk (including the relevant regulators) before embarking upon the analysis.
Where using CBA to assess whether it is reasonably practicable to implement measures to reduce risk to business, assets or property, it may be more straightforward to quantify the negative benefit of the loss of that property or functionality. However, it is no less important to agree with all stakeholders the input data to be used and the .success criteria. for what residual risk is regarded as being tolerable before embarking upon the analysis.
A CBA on its own:
does not constitute an ALARP case
cannot be used to argue against statutory duties
cannot justify risks that are intolerable
cannot justify what is evidently poor engineering design.
If carrying out a CBA, it is crucial that the same level of discipline is used in estimating costs as is used in assessing the risk. Only costs directly related to safety can be used in the analysis . costs associated with non-safety requirements (e.g. aesthetic appearance, security or revenue protection requirements) cannot be considered in a safety-related CBA. It is, however, acceptable to include installation, training and any additional maintenance costs, and any business losses that would follow from assets being taken out of service solely for the purpose of putting the measure into place. The corollary of this is that any cost savings that result from the implementation of the risk reduction measure should also be considered these might include improved availability of assets, for example. These should be offset against the cost of the risk reduction measure(s) in the CBA.
The cost used must be that for the minimum safe engineering or management solution. The non-safety requirements may be entirely legitimate, but they are subject to a different cost.benefit case, unrelated to safety, andcannot influence the decision as to whether the measure is ALARP or not. Similarly, only costs that fall on the Duty Holder should be used . costs to third parties (e.g. members of the public) should not be used.
Whilst QRA is a useful and respected tool, there are known pitfalls to its use:
QRA is not always appropriate. It should not be used where established good industry practice exists, is relevant and is straightforwardly applicable.
It should be used with caution when considering low-frequency and/or serious consequence events (such as fire, in many premises).
Historical. data should be used with caution and statistics based on limited sample periods should be used with care. History shows that, even where many years have passed without significant incident, this does not necessarily indicate that risk is acceptable. Indeed, fire safety legislation is often driven by public reaction to infrequent events that would not necessarily have been predicted beforehand using probabilistic assessment techniques.
QRA should not be used to justify removal of risk reduction measures on the basis of cost saving alone, unless it can be demonstrated that fire risk is maintained at equivalent or lower levels by other risk reduction measures.
Numerical levels of probability might mistakenly be regarded as predictive .fact. and be given undue prominence in the judgment of acceptable risk. This will be especially relevant if it is viewed that their precision automatically means that they are accurate, whereas in most cases there will be significant uncertainty in the probabilities generated during the assessment process.
The quantified .success criteria. for determining whether fire risk is tolerable or not may be difficult to establish.
The last point is particularly relevant where fire risk is being assessed. While general levels of ‘tolerability’ for risk to individuals are reasonably well defined numerically in guidance and standards, where multiple fatalities in fire are concerned, society tends to be much less tolerant of risk. There is a greater-than-normal expectation of safety from that particular hazard. This is generally known as .societal concern. and is not straightforward to quantify.
There is no widely agreed and quantified maximum level of risk that satisfies societal concern the benchmark level can be regarded as being equivalent to that set by means of the recommendations in national and/or governmental guidance. Therefore, QRA should normally only be used to demonstrate acceptable fire safety by comparison with accepted levels of risk against established relevant good practice.
Since the above can be viewed as an .absolute. maximum level of risk, it is not acceptable to .offset. fire risk againstother safety risks if that results in fire risk higher than the accepted level. For example, it might be claimed that there is more risk to building occupants due to them being assaulted with fire extinguishers than the fire risk reduction that is realised by their presence, so justifying the removal of the extinguishers. However, good practice as regards control of fire risk would be for their provision. The correct approach in these circumstances would be to provide them, thus achieving an acceptable level of fire risk, but to introduce physical and management controls to reduce the risk of assault to an acceptable level (e.g. tamper-resistant cabinets or tethers; siting them in management-controlled areas; provision of detection that alerts a fire team armed with extinguishers). This would then significantly reduce the risk of the extinguishers being used as weapons.
Societal concern
Where a significant number of persons could be affected by the consequences of a particular fire hazard in the premises, or where those persons might be regarded as particularly .vulnerable. in case of fire, consideration should be given to possible societal concern about the risk or the measures proposed to reduce the risk. The factors to be considered within this determination should include those where:
the extent to which the risk arises from a failure that could result in a major accident, which society would be unaware of, or would assume was already well controlled there might be public aversion to the scale of the injuries should the risk be realised
public disquiet and loss of confidence that would arise from a key failure occurring within the accident sequence, even if not leading to serious consequence (e.g. a near miss)
the risk is inequitably shared, particularly where a vulnerable group (such as children or persons with a disability) may be involved
the decision may lead to loss of public trust in the Duty Holder.s ability to learn from serious incidents and/or adopt good practice
the adoption of the risk reduction measure would have a significant adverse effect on the Duty Holder’s operations, which the public may perceive as being disproportionate to the safety risks.
The above will be particularly relevant for public bodies (e.g. health authorities, transport infrastructure providers, education authorities) or those offering access to large numbers of members of the public (managers of sporting and entertainment venues, Duty Holders in shopping malls, etc).
Risk to firefighters
It is expected that firefighters are likely to be exposed to risk (when carrying out their fire and rescue duties) that would be comparatively intolerable for members of the public.
Fire and rescue operations are normally undertaken on the basis of a dynamic risk assessment upon arrival at the incident (based upon the type of premises, the severity of the fire and whether it is believed that there are persons at risk from the fire), coupled with appropriate personal protective equipment (PPE) (e.g. heat-resistant clothing and/or breathing apparatus). It is not, therefore, either practicable or necessary to control risk to firefighters during their operations to levels equivalent to those for other occupants.
Having said the above, the risk to firefighters undertaking their duties during a fire should be considered when designing a building. It is not acceptable to ignore the fact that their duties under law are likely to include doing all that is reasonable to protect both life and property in case of fire. The fire protection provided should be such that these duties can be undertaken without exposing the firefighters to intolerable risks. In practice, this will usually mean ensuring that works either comply with good industry practice and/or offer an equivalent level of fire risk (to the firefighters) by other means.
Within this perspective, some technological tools can be useful when assessing the fire risks to firefighters within buildings. For instance, the use of people movement (PM) models (commonly called evacuation models) as well asfire models can assist an assessor when undertaking a risk assessment.
Risk assessment pitfalls
General
It may be that a view could be taken that risk assessment is an opportunity to dispense completely with prescriptive standards and to reduce costs by assessing out established risk reduction measures. If properly applied, risk assessment does allow targeted risk reduction, perhaps resulting in lower risk than the .prescriptive. solution or the same level of risk at lower cost. However, care should be taken when using risk assessment techniques to depart from established prescriptive codes, and some previously published examples(2) of poor practice in risk assessment are included in the following sections.
Considering only the probability of fire
It is unlikely to be legitimate to conclude that fire hazard is so low that the probability of having a fire that can cause harm is negligible. It is expected that, where a lowfrequency but serious consequence event such as a large fire is concerned, it should be assumed that a fire could occur and the risk should be assessed on that basis. The management controls that would be required to reduce to negligible the probability of a significant fire starting are so demanding that, in most industries, it is not sensible to rely upon them being applied throughout the life of premises.
Reverse ALARP
It might be attempted to justify the removal of existing fire protection measures on the basis that the cost of ongoing maintenance or renewal is grossly disproportionate to the fire risk reduction benefit that they achieve. This is not acceptable because there is a responsibility to maintain existing fire protection measures (this is usually enshrined in law) and those existing measures reduce risk to what must have been regarded (when they were implemented) as an acceptable level. By providing those measures, the Duty Holder has demonstrated that it is reasonably practicable to do so, and by so doing it is reasonable to achieve the resulting level of risk. Increasing that level of risk can therefore not be ALARP. This unacceptable form of argument is commonly known as reverse ALARP’.
This does not mean that fire protection can never be removed; if one can reasonably argue that fire risk has not been increased at all by that removal, then it may be acceptable to do so. This might be by applying one or more of the following criteria:
The risk reduction measure to be removed or modified addressed a hazard that is no longer present.
Alternative risk reduction measures no less effective than the measure being removed will be applied and maintained, so resulting in risk not being increased.
In all cases the removal of the risk reduction measure does not increase risk beyond that whichwould be achieved by the application of relevant and current good practice.
Using the cost of remedial works in a CBA
It might be the case that works have been designed and implemented in an unacceptable manner. For example, it might be discovered that they do not comply with good industry practice nor do they offer an equivalent level of safety. In this case it has been known for CBA to be used (in either a qualitative or quantitative risk assessment) to justify why it is acceptable for those variations from acceptable risk to remain. It is often the case that those making the argument use the .trouble. (i.e. cost, disruption and impact upon programme) of correcting the issue as the measure against which the risk reduction benefits are judged.
This is not good practice and should be avoided. When using CBA either qualitatively or quantitatively, the judgment should be made against the cost of the relevant works when they have been competently and correctly designed, supplied and installed, not against the cost of correcting works designed, supplied or installed incorrectly.
Incorrect reference to good practice
Some attempts to justify departure from relevant good practice refer to inappropriate guidance (e.g. standards written to address fire risk in premises with less significant fire hazards than those in question). It may be the case that a Duty Holder in a hotel refers to guidance on offices, instead of guidance that addresses the risks commonly encountered in hotels (e.g. offices . occupants usually awake, familiar and aware of their surroundings; hotels occupants may be asleep, sensory impaired and/ or unfamiliar with their surroundings). It is important that those assessing risk are mindful that the guidance they use, either directly or as a .benchmark., is appropriate to the environment that they are considering.
Another example might be to cite reference to design solutions used elsewhere, but where the context is different in crucial ways. An example might be a railway rolling stock manufacturer who wishes to offer vehicles to the operatorof an established underground rail system. That rail system has been in operation for many years and the infrastructure is built to standards long-superseded. While the vehicles might be entirely satisfactory when used on modern infrastructure, compliant with current standards, it may be necessary to compensate for the higher risk inherent in operating on much older infrastructure by reducing the fire risk associated with the rolling stock. In this case, comparison of the risk posed by a part of the system, rather than the whole system itself, is also of questionable validity.
Not considering risk to particularly vulnerable occupants
When assessing fire risk, those undertaking the analysis should be fully aware of the occupancy profile of the premises. They should ensure that the assessment considers whether there are any occupants who are likely to be present whose response to a fire emergency in that premises might be delayed or their ability to make good theirescape might be hindered by a sensory or physical impairment (whether permanent or temporary). Examples might include:
young children
the elderly
persons with a sight and/or hearing impairment
persons with restricted mobility (e.g. wheelchair users)
persons who are bed-ridden.
In all cases, it will be important for the risk assessment to consider the risk to each individual type of occupant and to conclude whether the existing or proposed risk reduction measures are adequate to control risk to an acceptable level.
It is normally very important to consider whether relevant good practice exists and it would be appropriate to apply that good practice wherever it is reasonable to do so. Where varying from that guidance, it is strongly recommended that those undertaking the risk assessment are able to construct a robust case for the proposed risk reduction measures being equivalent to that good practice.
It is recommended that the assessors do not base the assessment only upon the current occupants of the premises. One should also consider whether it is foreseeable that vulnerable occupants might be in the premises even if they are not currently present. For example, if a building has step-free access to all or part of it, then it should be considered that wheelchair users might be found in all accessible parts of those premises, even if it is not evident that they are, or if there is no particular reason for them to be in that part of the premises.
The risk assessment should therefore take into account the possible presence of wheelchair users, and appropriate procedures and/or physical protection measures should be provided to ensure that they can be safely evacuated.
It is unlikely to be acceptable to argue that, because few vulnerable people are likely to be in their premises, the probability of both having a fire of significant size and having a vulnerable person in the premises at the same time is so small as to render the cost of any fire risk reduction measure aimed solely at that group grossly disproportionate to the risk reduction achieved. This is not viewed as good practice, because it may place a vulnerable group at a significantly higher risk than other building occupants, and it fails to maintain risk at levels equal to or better than relevant good practice. Whether it can be claimed that individual risk is low or not, this approach is unlikely to satisfy the test of societal concern, which makes its acceptability highly questionable.
By making their premises accessible to those vulnerable groups, it is expected that the Duty Holder takes steps to reduce their risk from fire to a level comparable to that for the rest of the occupants of that premises.