FPSOs and other floating offshore facilities typically follow "prescriptive based" classification rules for design of the hull, mooring and marine systems. In some cases the process facilities are also classified. An alternative approach is to use a more open framework, "risk based" design approach that allows variation from prescriptive rules provided system risks are maintained at acceptable levels. The various classification societies currently allow such risk-based alternatives [2, 3]. Although numerous "component" risk studies for FPSOs have been conducted and published, this is one of the first that accounts for the integration and linking of risks and risk tradeoffs among the hull, mooring system, marine systems, topside process plant and the utility, power and control systems that support them. The model and basis are first described, followed by application to a prototype deepwater, turret moored FPSO with gas handling. Example cases are shown to demonstrate use of the model to make design decisions to the various FPSO components.


Classification rules are established based on engineering principles, experience, testing & expert judgement. They are intended to ensure probabilities of accidents are low, but this is not explicit. Changes to developing and implementing ABS rules are being explored through risk based approaches. An alternative, risk-based approach to classification of Floating Production, Storage and Offloading systems is being investigated by ABS as part of a major internal technology development project. The project is comprised of model, database and methodology development and training at multiple levels throughout the organization. The prototype model was completed early in 2001 and is the focus of this paper. The model represents systems failures more comprehensively than any other offshore risk assessment known to the authors. This level of detail was sought in order to achieve specific goals:

  1. Develop explicit risk measures of class rules to facilitate prioritization and optimization, thus allowing one to focus resources on the greatest risk contributors,

  2. Develop a consistent means for performing risk tradeoffs (i.e. demonstration of equivalent level of safety), and

  3. Provide a vehicle for expansion of Class or Group services to risk significant systems, components, structures, or human actions not currently included in Class scope.

Model Development

The overall model is comprised of a collection of initiating events, facility response model and consequence calculations. These are categorized into discrete damage states (see Figure 1). Also illustrated in Figure 1, the facility response model is comprised of support event trees, frontline event trees and end states. The support event trees represent the failed/operational state of support systems (e.g. utilities, instrumentation and control, emergency) required for successful operation of the frontline (main system) event trees. The end states are simply a discrete categorization of the various failed configurations of the facility. Frontline systems were segregated into three categories for convenience: process, marine and structural systems. Each of these model partitions are described in summary in Table 1, and the development activities are described in the following sections.

This content is only available via PDF.
You can access this article if you purchase or spend a download.