Introduction
When assessing against SOLAS chapter II-2, regulation
17, an analysis shall show that an equivalent level of safety is achieved
by the alternative design and arrangements with regard to introduced fire hazards.
Guidelines for such analysis are found inMSC.1/Circ.1002, as amended by MSC.1/Circ.1552. However, when considering FRP composite structures, it
may also be relevant to consider MSC.1/Circ.1455 which contains guidelines that have been developed to
provide a consistent process for the coordination, review and approval of
alternative design and arrangements in general, i.e. not only fire safety. This may
be particularly appropriate when the use of FRP composite affects other aspects of
safety than those related to fire. Further assistance may be found in guidance notes
for MSC.1/Circ.1002, as amended by MSC.1/Circ.1552 and in guidelines on fire safety engineering applied to
buildings. Below follow discussions on the required method for analysis, evaluation
and approval of FRP composite structures, with regard to uncertainty treatment,
sophistication and the practical process. Reference is made to the guidelines
referenced in SOLAS, MSC.1/Circ.1002, as amended by MSC.1/Circ.1552, and also to MSC.1/Circ.1455. It is particularly pointed out that the assessment must
stand in relation to the current scope of the proposed design and arrangements; a
simple and well-protected structure in FRP composite should not require a
complicated or time-consuming assessment.
C.1 Uncertainty treatment
1 Even the most detailed risk assessment contains limitations; uncertainties are
involved throughout the whole process. The uncertainties that arise when determining
the frequencies and probabilities of events are often perceived as the dominating
sources of error. Data is insufficient or not fully relevant for the particular
events. Common reasons are that statistics have simply not been recorded or that the
data is aged and does not comprise updates in legislation and novel technology.
However, even if statistical information is often considered to be "the truth", it
should be handled with care since the figures are always changing and may have great
errors. Furthermore, statistics can give an image of something that has happened in
the past but evaluations of novel ship designs need to be carried out before the
ship is put into practice, which implies that statistical data will not be available
for such parts of the ship. A general statistical representation may be available
for the prescriptive design but the fire risk of the alternative design and
arrangement needs to be calculated from knowledge.
2 Attempting to compare a calculated risk of alternative design and arrangements with
a statistical representation of a prescriptive design, or an absolute risk
criterion, may become extremely uncertain since the different approaches contribute
with fundamentally different uncertainties. It could therefore be recommendable to
carry out a relative risk assessment, as described in MSC.1/Circ.1002, as amended by MSC.1/Circ.1552, even when carrying out a SOLAS regulation
II-2/17 assessment at a more sophisticated level. Thereby, uncertainties
can be minimized by founding the risk estimations of the ship designs on similar
assumptions (e.g. in models, expert judgement, statistical data, etc.). In order to
expose the differences in fire safety it is also recommendable that the assessment
concerns only the alternative design and arrangements and thereby relevant parts of
the ship (a risk measure for the ship as a whole may give a wrong representation of
the safety).
3 When determining consequences of events, uncertainties depend on how systematic and
detailed the approach is. Models used when estimating the consequences and the
experience in the expert group are also sources of uncertainties. In the hazard
identification, uncertainties are also often linked to the method used, how detailed
it is and the competence of the expert group examining the systems. Lack of
routines, knowledge and experience are drawbacks which need to be considered when
designing a ship with novel technology. The uncertainties can result in missing or
wrong scenarios when identifying hazardous events, which can have great effects on
the proceeding analysis. A common feature of all the steps of the risk assessment is
that many simplifications are made in order to model complicated systems. Much
because of the complex matter of assessing the impact of human behaviour when
modelling, they tend to be focused on machines and technical components. Leaving the
effects of organizational aspects, safety management systems and operator actions
outside the scope of the risk assessment will, however, not reduce
uncertainties.
C.2 Required method
1 Many different methods for risk assessment, of varying sophistication, can be used
to evaluate uncertainties in a ship design, which is the focus when adopting a
risk-based approach. All ship designs contain uncertainties and all risk assessments
contain uncertainties. As a result, all decisions will be made under some measure of
uncertainty. If a risk assessment would result in an absolute certain probability
density function of the possible consequences, a decision would be truly
"risk-based". However, since uncertainties cannot be eliminated, it is important to
analyse them and to appraise the effects of uncertainties on the result and the
total effect when these uncertainties are considered. Methods for risk assessment
are often classified based on the inclusion of quantitative measures
(qualitative-quantitative) or on the consideration of the likelihood of outcomes
(deterministic-probabilistic). A more suitable classification includes the previous
features but depends on how uncertainties are treated with varying thoroughness.
2 The guidelines in MSC.1/Circ.1002, as amended by MSC.1/Circ.1552, outline a plausible worst-case approach for analysis
and evaluation which can be described as a deterministic risk assessment. This kind
of consequence analysis, commonly referred to as "engineering analysis", is
described in several engineering guides to performance-based analysis of fire
protection in buildings, which have formed the basis for the guidelines. MSC.1/Circ.1002, as amended by MSC.1/Circ.1552, makes clear that the scope of the analysis depends on
the extent of deviations from prescriptive requirements and on the extent of the
alternative design and arrangements. However, increased uncertainties do not only
increase the scope of the analysis but also affect the required accuracy and
sophistication of the method for verification of safety. A more sophisticated
approach will further increase the engineering efforts but may be necessary if
safety margins are to be kept reasonable and risks are to be properly managed when
for example deviations are numerous, significant or concern many areas or when the
design and arrangements are large, complex, novel or outside the scope of
prescriptive requirements. Therefore, the approach outlined in MSC.1/Circ.1002, as amended by MSC.1/Circ.1552, may or may not be sufficient to adequately assess fire
safety. Furthermore, if the case is simple, a less complicated kind of risk
assessment should be sufficient. Therefore, MSC.1/Circ.1002, as amended byMSC.1/Circ.1552, "only" presents guidelines; the required sophistication
of the method used to assess safety depends on whether it is sufficient to describe
the current design and arrangements in terms of fire safety. The adaptability of the
method used to verify fire safety and its dependence on the current scope is clearer
in MSC.1/Circ.1455 (paragraph 4.13.2). Since the term "engineering
analysis" refers to a certain kind of risk assessment, the more general term "SOLAS
regulation
II-2/17 assessment" is used in these guidelines.
3 Moving to SOLAS regulation II-2/17, the stated ultimate requirement for alternative
design and arrangements is sufficient safety; an alternative design and arrangements
shall be at least as safe as if prescriptive requirements were complied with (regulation
II-2/17.3.4.2). If the scope of the deviations posed by the alternative
design and arrangements is great it may be relevant to carry out an assessment at a
higher level and determine an index of safety for the whole (or a considered part of
the) ship. However, if effects on safety can be managed within the areas of one or a
few regulations separately, this will allow for an assessment at a lower level (e.g.
limited to evaluations of fire growth potential or containment of fire). This is
also why it was decided to have regulation
II-2/17.2.1 read: "provided that the design and arrangements meet the
fire safety objectives and the functional requirements", without mentioning whether
it is the functional requirements in SOLAS regulation
II-2/2 or in any other regulation. "Minor" alternative design and
arrangements should be possible to analyse and compare to single affected functional
requirements of deviated regulations. As long as those functional requirements are
met it may not be necessary to evaluate safety at a higher level through the overall
fire safety objectives and functional requirements. However, this requires that risk
control measures are found which target potential deficiencies in the areas of the
individual deviated regulations.
4 It should be noted when considering FRP composite structures that a sole ASET-RSET
evaluation, common in fire safety engineering, may provide an insufficient
assessment. Effects on safety from use of FRP composite may go beyond what is
captured by such assessment, e.g. effects appearing after escape from the fire or
disproportionate damage. In any case, it should be proven that the ship can survive
a set of relevant design fires and be its own lifeboat. The design fire scenarios
must be specified to represent all the affected safety barriers, i.e. not only those
presented as functional requirements in SOLAS,
as further elaborated below.
C.3 Establishment of approval basis
1 Modern ships (in particular passenger ships) are built with several fire safety
functions or barriers. This will provide an integrated and redundant system that
takes into account that some safety systems do not work as intended. A ship (partly)
built of FRP composite structures should provide similar robustness and the design
process should document that safety system can fail without loss of important safety
functions or disproportionate consequences. However, all safety barriers are not
clearly stated in the regulations and may be hard to identify.
2 According to SOLAS regulation II-2/17, alternative design and arrangements for
fire safety should provide a degree of safety at least equivalent to that achieved
by compliance with the prescriptive requirements. To form an approval basis, it is
stated that the SOLAS regulation II-2/17 assessment should include an
identification of the prescriptive requirement(s) with which the alternative design
and arrangements will not comply (regulation
II-2/17.3.2). This is also a foundational part in MSC.1/Circ.1002, as amended by MSC.1/Circ.1552, where it is stated that the regulations affecting the
proposed alternative design and arrangements, along with their functional
requirements, should be clearly understood and documented (paragraph 5.1.2). This is
further stressed in paragraph 4.3.4, where it is stated that the preliminary
analysis should include a clear definition of the regulations which affect the
design and a clear understanding of the objectives and functional requirements of
the regulations (i.e. the purpose statement in figure 5). The objectives and
functional requirements of the deviated prescriptive requirements can thereafter be
used (along with the fire safety objectives) to define performance criteria, as
described in paragraphs 4.4 and 6.3.2 of the Guidelines on alternative design and
arrangements for fire safety (MSC.1/Circ.1002, as amended by MSC.1/Circ.1552) and in regulation II-2/17.3.4.
3 When FRP composite is used, the fundamental deviations concern requirements on
non-combustibility. However, due to limitations in current regulations, an
identification of deviated prescriptive requirements and their associated purpose
statements may not form a sufficient basis to evaluate the safety of FRP composite
ship designs. The regulations are based on assumptions regarding the design and
arrangements and therefore not all safety requirements are apparent. In particular,
many requirements are made up around steel designs, leaving many implicit
requirements unwritten. Therefore, use of FRP composite will affect fire safety in
many ways, some of which are not covered by the fire safety regulations. An approval
basis for equivalent safety may therefore not be sufficiently defined based only on
deviations from prescriptive requirements, which is clearer in MSC.1/Circ.1455 (paragraph 4.7.1) than in MSC.1/Circ.1002, as amended by MSC.1/Circ.1552 (paragraph 5.1.2).
4 Depending on the scope of the proposed alternative design and arrangements,
additional investigations may be called for to consider how the implicit level of
fire safety represented in the Convention is affected. This may be relevant for an
assessment of any design and arrangements which are truly novel (not simple
extensions of the corresponding prescriptive requirements) since all hazards are not
addressed by the Convention. A simple comparison with existing prescriptive
requirements may not be sufficient and the assessment may therefore require special
attention.
5 Investigations of effects on the implicit level of fire safety, or identification
of missing requirements, can also be claimed necessary regardless of the novelty of
the proposed alternative design and arrangements. To further complicate the
comparison of safety levels, many prescriptive requirements have unclear connections
with the purpose statements of their regulations and with the fire safety objectives
of the fire safety chapter, which are supposed to define "fire safety". Some
functional requirements could, for example, be claimed missing based on the
prescriptive requirements and for some functional requirements listed at the
beginning of regulations there are no associated prescriptive requirements.
Deviation from one prescriptive requirement may affect the achievement of a
functional requirement of a different regulation, etc.
6 A SOLAS regulation II-2/17 assessment involving FRP composite structures, as any
SOLAS regulation
II-2/17 assessment, must be sufficient to describe the introduced novelty
in terms of fire safety. Determining the approval basis only based on deviated
prescriptive requirements may not be sufficient and additional investigations of
effects on the implicit level of fire safety may be necessary. These guidelines
attempt to clarify such potential explicit and implicit effects on fire safety when
using FRP composite compared to what is implied by the prescriptive requirements
from a broad perspective in section 3 (Important factors to consider when evaluating
FRP composite structures with starting point in the regulations of SOLAS chapter II-2). However, it could also be the case that further
investigations are needed regarding how the proposed design and arrangements affect
the fire safety implied by prescriptive requirements. Investigations could, for
example, be carried out to clarify effects on the fire safety objectives and
functional requirements of the fire safety chapter, effects on the structure of the
fire safety (effects on the source, exposure or effect on part of the fire
protection), effects on properties of the fire protection (e.g. effects on the
flexibility, sensitivity, complexity, vulnerability, reliability or human
intervention) or effects on fire development (effects on a fire in the incipient,
growth, fully developed or decay phase). There are also many established methods for
hazard identification which may be used.
7 In order to manage all the identified pros and cons of the alternative design and
arrangements with regard to fire safety, it is also suggested that they are managed
in a better way than the way in which it is described in MSC.1/Circ.1002, as amended by MSC.1/Circ.1552 (paragraphs 5.2.1.2 and 5.2.1.3), e.g. by collection and
rating in a risk-based presentation, such as a ProCon List or Risk Matrix. This will
be of significant value when forming fire scenarios. In general, when novel design
and arrangements are managed, it is recommendable to have a larger focus on the
initial stages of the SOLAS regulation II-2/17 assessment, particularly on the
identification, collection, rating and selection of fire hazards.
C.4 Approval process
It should be stressed that the sophistication of the risk assessment may vary
depending on the scope of the proposed design and arrangements, so may the practical
process of the assessment. MSC.1/Circ.1002, as amended by MSC.1/Circ.1552, describes an approach where the assessment is reviewed
at two stages by formal approval of reports. The guidelines in MSC.1/Circ.1455 include the Administration more in the process by
putting a larger focus on monitoring and having review and approval of the
assessment in several more but smaller stages. Regardless of which guidelines are
referred to, it should be emphasized that the actual process may include more steps
than in the guidelines but it may also be significantly simplified. For example,
proposing the use of FRP composite for interior structures, a limited part of the
ship or structures which are ubiquitously thermally insulated, may not require a
lengthy, detailed or very time-consuming assessment.