In July 1988, graphic television footage showed the horror of the
fire on the North Sea offshore platform, Piper Alpha. The death
toll from that disaster was high: 167 people lost their lives. As
in the case of a plane crash, the incident was fully analysed, and
regulatory authorities tried to learn from the tragedy. The result
was a complete re-thinking about the safety of offshore structures.
Bodies such as the United Kingdom Health and Safety Executive embarked
on a process which, it is hoped, will prevent any recurrence of
Piper Alpha. They recognised the need to design in greater safety
measures at the construction stage. New software and methodologies
were required to reduce fire risk and optimise hazard assessment.
Analysing the risks....
This driving force led to the instigation and implementation of
a successful pan-European project, supported by the European Commission,
under the BRITE-EURAM programme. The catalyst was the introduction,
in 1990, of new UK requirements for safety on offshore structures.
Included in this legislation was the necessity of carrying out risk
analysis of the various types of incident that could occur. Certificates
- the equivalent of those given to ships for seaworthiness - would
only be granted if the operator of an offshore structure could demonstrate
that any risks associated with the design and operation were kept
to the minimum reasonable level.
As Piper Alpha demonstrated, one major risk is that of fire, and
a key objective of the new safety requirements was to minimise injuries
and deaths if it occurred. This project set out to tackle exactly
The project was co-ordinated by the Italian Classification Society
(RINA), and included partners from Denmark, Germany, Italy and the
United Kingdom. The partners' roles were well defined and tailored
to their particular fields of expertise: skills in safety regulations
were provided by RINA and Germanischer Lloyd, both being certification
and classification organisations. Tecnomare possessed offshore installation
design know-how, Snamprogretti had considerable experience in risk
analysis of petrochemical plants, and WS Atkins and the UK Atomic
Energy Authority were active in two areas: structural engineering
design and risk analysis. Completing the line-up of partners were
the Technical Institute of the University of Lisbon, and the Institute
for Computational Safety and Reliability, which both brought computing
skills to the table.
....developing a concept....
According to the project co-ordinator, Mario Dogliani, the main
focus of the project was structural engineering. "One of the
important requirements was to develop the concept of a safe refuge,"
he explains. "If there is a major fire or explosion, there
must be a shelter on the structure where people can gather until
the rescue services arrive. One goal of our risk analysis was to
determine the best position for such a refuge and to optimise its
Project tasks were divided into three main themes: collecting data
(such as the type of requirements demanded by relevant legislation
and the capability of existing methodologies and software), devising
new analytical computing tools, and using these tools in the development
of new structural designs.
"The computing tools were important," says Dogliani. "They
perform two major functions: first we needed to develop the capability
of modelling an incident, such as a pipe rupture, then we had to
analyse the potential consequences of such an incident." In
the real world, this is a more complex concept than you might imagine.
Major occurrences, according to Dogliani, are rarely the result
of one large incident. They are usually caused by two or more small,
unlikely, unrelated problems happening simultaneously and having
a knock-on effect.
....and implementing the results
"The trick then," continues Dogliani, "is to take
this theoretical knowledge and apply it to real cases. We did this
in two ways. Firstly, we used the risk analyses to produce a new
set of rules for designers, and we wrote software to help them achieve
the new specifications. Secondly, we were able to show that the
use of information-based optimisation data was a much more powerful
design tool than had previously been realised."
The results have not gone unnoticed. "We believe that some
major oil companies are using them in their rig design procedure,"
says Dogliani, "but it's not easy to know exactly how, or to
what extent because they are very cagey. We also know that there
is a good deal of interest from naval architects, who are aiming
to design safer ships." Technology borne of a disaster could
make sea travel just that bit safer.