Design and operation of large infrastructure projects - A risk-based approach

Soren Randrup-Thomsen, M.Sc., Ph.D., Head of Department, Risk & Safety, Ramboll Denmark

Designing large infrastructure projects – bridges, tunnels, ports etc. – includes fulfillment of various requirements from codes and standards to ensure that the structural safety is not compromised and further to ensure that the structure is not disrupted in longer periods. It is however noted that these requirements can be fulfilled in various ways when also cost issues must be considered. Operational Risk Analysis (ORA) is a way to document that safety, availability and financial requirements are all fulfilled when the final design has been chosen, the construction project is finalized, and the bridge is in daily operation.

The risk-based approach has been used for design and construction of the Norwegian Public Roads Administration (NPRA) floating bridge ‘Bjørnafjorden Floating Bridge’ crossing the 500 m deep Bjørnafjord between Stavanger and Bergen–a 5 km long bridge to be built on 38 floating pontoons and a land-based pylon. Floating bridges are quite unusual in bridge design and the structural behavior of 

the bridge deviates significant from more traditional bridge designs and especially risks related to ship traffic accidents challenges the design. For this reason, the risk-based approach has been used to ensure that selected design alternatives do not pose unacceptable high risks for user safety or for disruption of the bridge.

In general, risk analysis covers several activities to be carried out, all aiming at somehow quantifying the considered risk – in this case, safety for users of the bridge and availability of the bridge in operation. The risk activities run through the following procedure.

As seen from the sketched procedure, the operational risk analysis includes identification of hazards – events that under various circumstances will lead to accidents in terms of human injuries/fatalities and/or to a closure of the bridge in a period. Examples are ordinary traffic accidents, truck fires and ship collisions with pontoons or girders. The hazard identification activity must hence cover all imagined types of accidents, why the identification process often includes workshops with experts within various fields.

When hazards are identified the risk analysis part is initiated. This includes analysis of the identified hazards to determine how often the hazard will occur (often based on statistics and presented as occurrences per year) and also modelling of the consequences, if the hazard appears. Consequences will be expressed as number of fatalities or number of down time hours of the bridge given the accident. The consequence modelling part may also be based on statistics but often are more detailed models used (e.g., computational models to estimate fires and toxic releases, structural models to estimate damages due to ship collisions). By combining the frequency and the consequence for a given hazard, a risk contribution in terms of number of fatalities per year or number of disruption hours per year is estimated.

The final risk is then calculated by summing up all risk contributions and compared to an established risk acceptance criterion.

 This criterion may be given by national authorities setting the limit for annual number of user fatalities for the considered structure. However, often such criteria are not available and may be set when starting up the project. A commonly used approach is to allow the same number of annual fatalities as seen on comparable projects – or to strengthen requirements just a little to increase safety. When looking at disruption periods this may be more locally determined and depends very much on the traffic patterns close to the bridge and the consequence for the users not being able to use the bridge and hence must take other delaying routes