6.3 General considerations

6.3.1 In the following, general issues and procedural considerations are dealt with. Prevalently, the responsibility for ensuring documentation quality rests with the Submitter. A party wishing to enter into the approval process may therefore preferably have safety and quality management systems in place for their own processes as well as for subcontracted processes, as this substantially supports a controllable documentation. Basic formal issues have to be followed, ensuring document control and facilitating the adherence to any existing or future management systems. Such formal issues include stating on the submitted papers:

project title and document title
responsible person(s), including company and contact information
scope statement/project description
project identification number
distribution list

authorization signatures
date
revision number/letter
other controlling documents/processes, if necessary

Table 2 – THE APPROVAL MATRIX
Project Category	Known application of proven technology (conventional process)	Known application of a technology with a limited field history/New application of proven technology	New application of a technology with a limited field history/Known application of a new or unproven technology	New application of novel or unproven technology	Activity performed by:
Requirements	(1)	(2)	(3)	(4)
*A) Basic risk assessment*	Not required	Required (unless rule challenge deemed insignificant or of negligible impact on safety and environment)	Required	Required	Submitter (yard, supplier)
*B) Further analysis requirements*	Not required	Depending on basic risk assessment outcome. Hazards medium or high, if any, may be examined further, at least by semi-quantified analysis	Semi-quantified assessment. All hazards medium and high may be examined by means of quantified analysis	Quantified risk assessment to all risk contributions (due to the novelty of the design, it may not be possible to rank such hazards credibly. Hence, all may be examined in depth)	Submitter in cooperation with the Administration
*C) Qualifications of analysts*	N/A	Operational experience General knowledge of risk assessment techniques	Operational experience. In-depth experience with risk assessment. Some knowledge of analysis techniques	Operational experience Risk assessment and analysis experts	N/A
*D) Applied rules and guidance*	Existing prescriptive rules (SOLAS, MARPOL, relevant codes, national, regional and international legislation, prescriptive class rules)	Existing prescriptive rules where no rule challenge prevails (SOLAS, MARPOL, relevant codes, national, regional and international legislation, prescriptive class rules) applicable standards if available from other industrial sectors, class guidance on risk-based approval as applicable	IMO circulars on alternative arrangements, class guidance on risk-based approval, other relevant industry standards	IMO circulars on alternative arrangements, class guidance on risk-based approval	N/A
*E) Potential additional tests, surveys and compliance control (after commissioning)*	As per Safety Management System (SMS) and existing regulation	Internal surveying. Additional review at safety related events subject to recording and corrective action	Internal/external surveying, recording and additional intermediate surveys of risk-based features, if deemed necessary	Continuous monitoring and review subject to reporting to the Administration until a sufficient level of experience is gained	Submitter (operator)
*F) Review by third party*	Considered	Considered	Considered	Recommended	N/A

6.3.2 Drawings and layouts

6.3.2.1 Any design detail deviating from conventional best practice should be described comprehensively to facilitate a full understanding of the extent of the novel type equipment or detail with intended design basis including environmental conditions and operational scheme.

6.3.2.2 Safety critical details have to be documented.

6.3.3 Design parameters

All parameters applied in the design process should be explicit. Processes may be presented as diagrams or described in prose. The method (how) and stage (when) of application should be clear from the description.

6.3.4 Requirements for risk control measures

6.3.4.1 For alternative designs the acceptable risk level will typically be set by a regulation compliant reference design (see sections 4 and 5).

6.3.4.2 In the absence of a regulation compliant reference design to compare with, the evaluation criteria may be specified by means of risk analysis. Identified risk levels will usually be categorized to belong to three categories: intolerable risks which should be reduced, negligible risks which do not require any action, and risks in the ALARP area which should be reduced to as low as reasonably practicable.

6.3.4.3 If several possible options can efficiently reduce the same risk, the passive options, which are usually more verifiable and reliable, should be chosen. As a majority of incidents are strongly influenced by human error and operational faults, the team may seek solutions that minimize potential human error, if at all viable and efficient.

6.3.4.4 If risk control measures are operational, then their implementation into management systems should be documented, to ensure that the crew is fully informed and familiar of such special measures.

6.3.5 Reporting format

6.3.5.1 The documentation of the project prerequisites, any assumptions and exclusions made, the HazId, the risk assessment, (if any), recommendations and conclusions as per the approval process may take several forms, all with their own advantages and drawbacks, such as:

.1 PC based work sheets and documents;
.2 technical reports; and
.3 programs tailored to the purpose.

6.3.5.2 Regardless of the format, the content should be verifiable. Also, as the HazId and ensuing conclusions will be reviewed during follow-up meetings, means of document control should be applied to ensure that only controlled versions of the information yielded are distributed.

6.3.6 Calculation/analysis requirements

6.3.6.1 When making decisions based on analysis techniques, care should be taken to evaluate their adequacy. This requires expertise on various types of risk assessment methods to ensure that the most suitable will be selected for the application in question. As is deductible from prior elaborations, the level of novelty/regulation challenge is variable and, depending hereon, the most suitable methodologies may also vary.

6.3.6.2 Below are nine reminders when performing in-depth risk analyses:

.1 apply best industry practice and be consistent with IMO FSA Guidelines when selecting risk assessment techniques;
.2 perform a high-level assessment of the design type for which approval is sought;
.3 ensure that the specific risk assessment (based on the generic high-level assessment) meets the requirements for methodology and depth level acceptable by the Approval Authority;
.4 ensure that the applied model reflects the as built and operated ship or system as accurately as possible. If necessary, the process should be conducted iteratively, as the design process progresses, to ensure all safety critical aspects are covered;
.5 apply assumptions on a sound basis and perform frequency and consequence analyses based on relevant and consistent data;
.6 check for consistency between the level of detail in the assessment and the assumed risk control measures and the system safety testing and -management program (programmed maintenance, safety management systems), especially if such assumed control- or mitigating measures are of an operational character;
.7 include internal and external events in the analysis;
.8 include normal operational modes as well as states of emergency in the analysis; and
.9 include sensitivity analysis, uncertainty analysis and importance measures.

6.3.7 Errors and uncertainties

6.3.7.1 To be able to compute or determine the different parameters which are to be applied in the risk analysis or which are found in any form of design equation, it is necessary to have access to various types of data.

6.3.7.2 Uncertainty reflects either lack of knowledge about the actual value of a variable (epistemic uncertainty) or variability intrinsic to the parameter (aleatory uncertainty). In standard risk assessment methods, however, uncertainties involved can be accounted for within the method (e.g. conservative assumptions).

6.3.7.3 Although uncertainty on variables may be considerable, expected values can be estimated. Therefore, it may be sensible to choose a reasonable estimate reference. Choosing values implying the worst conceivable case could result in an exaggerated picture of the risks involved.

6.3.7.4 To examine the impact of specific variables on the final results, a sensitivity analysis should be carried out, where the effect of for example doubling the value of a variable may be examined to decide whether the originally selected values are conservative enough or if they deserve a more precise/detailed analysis or not.

6.3.7.5 Hence the following issues which may require investigation at any given stage require consideration:

.1 variation in the input data;
.2 the impact of simplifications/assumptions of problems;
.3 the effects of various characteristics of the scenarios; and
.4 the reliability of systems.

6.3.7.6 Variables which are found to have a major impact in the sensitivity analysis may justify a more conservative or precise approach than variables of lesser importance. The sensitivity analysis may indicate the variables of major impact and how uncertainties of such variables are handled.