Like any classification society, LR is always evolving its classification rules. Today is no different. In fact the speed of new rule development is currently very high. There is a different look and feel to some of these new rules however – for they are no longer completely prescriptive. Some might even say they are no longer rules, since they don’t tell you what to do. So what is wrong with traditional prescriptive rules? The answer is very simple – nothing! In fact prescriptive rules are excellent providing there is a high degree of confidence that the hazards of a technology are understood.

Prescriptive classification rules, typically, are one solution to managing the unwanted hazards of a known technology. Hence classification rules are generally pragmatic and cost effective and have become the accepted route to managing a known technological hazard.

One of the many challenges the maritime industry faces today is the rate of deployment of new, complex and integrated technology to meet new regulation and maintain a competitive edge. Prescriptive rules, on their own, are not a good way of managing the hazards of new, complex and integrated technology, since the hazards are not fully understood. Prescriptive rules also offer only one solution to a known problem. Some might say they do not facilitate innovation.

Hatch covers

To give an example. Hatch covers on bulk carriers are standard and are the subject of classification rules. In this case the safety hazards being managed are:

1) Cargo liquefaction due to the cargo becoming wet from moisture (from rain or sea spray entering through the cargo hold hatch)

2) Loss of excess buoyancy or ship stability by green seas coming into the holds through the hatch. A ‘lid’ on the hold does this simply and neatly, and is an obvious solution.

A hatch cover will also help keep the cargo in spec for the end user. Clearly hatch covers are not the only way these safety hazards could be managed. Frequently the rule requirement asks for a specific hardware measure, since there is considerable confidence in the measure, providing it is adequately maintained.

Here we touch on an important element which any classification society requires – confidence in the measure. There is a high level of confidence in hatch covers, given deep sea trade and varying weather conditions. This is not to say that other solutions to managing the hazard are not acceptable. For instance, in benign sheltered waters, management controls might form part of an alternative approach, instead of using hatch covers. Gaining the same level of confidence in management controls however involves considerably more effort.

So why do an increasing number of LR’s new rules additionally require risk based techniques to be used? Using prescriptive rules (for instance requiring a specific hardware measure) is adequate when small incremental steps are made in evolving or developing new rules involving a known technology. In such cases prescriptive rules can be produced, with a high degree of confidence, that ensure the technological hazard is adequately identified, understood and managed.

New technologies

What about new technologies or increasingly integrated and complex systems where the hazards are not fully understood? Using or modifying existing prescriptive rules gives very little confidence that the potential new hazards are understood and adequately managed. A simple example is LNG as a fuel. If prescriptive class rules developed for HFO/MDO were used as they are or were modified to manage LNG as a bunker fuel, it’s entirely probable incidents would result. The process used to evolve the existing rules would not be sufficiently robust and systematic.

The fact that LNG is cryogenic and that it is used in the gas state will most probably lead to catastrophic incidents. When technologies involve inherently large hazards, are expensive or have high expectation of the stakeholders, developing rules by overt reliance on learning through incidents is not an acceptable way of developing rules.

I hear you saying that LNG and methane are well understood, as indeed they are in certain industries and applications. LNG’s use on board ship as a fuel, outside of boil of gas in LNG tankers, is very new, and standard designs and their safety implications on ships have yet to be fully understood. In such cases prescriptive rules alone do not give the adequate level of confidence that class requires to approve a design. Additional methods need to be used, of sufficient rigour, to give class the confidence that there has been a robust and systematic approach to the identification and management of new technological hazards.

Prescriptive rules

There is good news however! As experience of a technology develops, and standard designs are evolved, prescriptive rules can be developed. It is quite possible that within two to five years of a new technology being introduced prescriptive rules can be produced. The industry is embracing, at an ever increasing rate, new and complex technologies. Increasingly these technologies have to be effectively integrated to ensure that unwanted hazards are managed, and that the benefits of the expenditure on new technology are realised. As an ex-seagoing engineer I fondly remember a rotating mechanical engine governor, a wonderful piece of kit. When did you last see one of these?

Somewhere in my car there is a reciprocating engine, buried beneath mounds of electronics. Solid state logic circuits are the control systems of today and the future, and they are everywhere – microprocessors can be tiny things. Ensuring we have a robust and systematic understanding and management of the hazards of new technologies, complexity and integration is the only thing that will provide adequate confidence for class approval to be given.

If you want to be at the cutting edge of performance by gaining the benefits of new technology, complexity and system integration, robust and systematic risk assessment will be a large part of the design development. This is the reality of today. If you are prepared to forgo a competitive performance edge, then wait a number of years until a particular new technology, complexity and integration is better understood, and when wholly prescriptive rules are likely to have been developed.

Source: Lloyd's Register