Keywords: safety, intelligence, enabling causes, triggering events, safety procedures or devices, fault tree, fuzzy logic, fuzzy sets.

Throughout the lifecycle of a constructed facility, safety issues constantly change as the facility is designed, constructed, put into service, and years later, as it finally is taken out of service by being deconstructed or demolished. Attention given to safety at its very conception is crucial to the safety performance throughout the lifecycle of the facility. Project stakeholders have the largest influence on safety in the conceptualization phase where their values influence the resources devoted to safety. Furthermore, the influence of stake holders decreases throughout the facility's lifecycle. Once the commitment to safety is made during the conceptualization phase, it is essential to "design in" safety.

The clients of safety and safety considerations change throughout a facility's lifecycle. In the construction phase, the facility is not structurally complete; presenting many opportunities for structural instability, caused by the ever changing forces that structural elements exert upon each other as the form of the building progresses. As the structure nears completion, demobilization commences, occupants begin to move in, and the need for safety transitions as the building transitions. This phase presents its own unique safety challenges, for example zone overlap as workers pass through areas of the structure that are in use and occupants are exposed to construction hazards. During the service life of the building, safety of both the structure and the occupants is tied to effective maintenance. The level of safety for occupants is a synthesis of design, maintenance, and managerial effort. At the end of the facility's lifecycle, deconstruction or demolition pose another set of safety challenges. In the case of deconstruction, and even to some extent, in the case of demolition, safety at the end of the building's life is linked to decisions made during when the building is conceived and later designed.

Although safety concerns change during the lifecycle of a constructed facility, the basic relationships among factors affecting safety remain constant. Figure 1 shows that safety can be modeled as the interaction of three elements; enabling causes, triggering events and safety Procedures Or Devices (POD)s.

Enabling causes are often referred to as safety hazards that include sites without signs or fences to prevent public access, structurally unsound temporary structures such as scaffolding and equipment parked on sloping surfaces. By definition, enabling events are inside the realm of project control or internal to the structure. The enabling cause illustrated in Figure 1 is that of a backhoe parked on a sloping surface.

Triggering events set a failure chain leading to injury into action. Such an event is illustrated by the rainstorm in Figure 1 that causes the sloping surface where the backhoe is parked to become slippery. Once the soil fails or the coefficient of static friction decreases to the point where the frictional forces can not longer counteract the gravitational force of the backhoe along the slope, the backhoe begins to slide down the surface. By definition, triggering events such as severe weather are outside the realm of project control.

Safety PODs come into play in two ways. The first type of safety POD can be used prevent accidents from occurring. A safety checklist that requires parking and storage of equipment on level surfaces is an example of a POD that prevents accidents from happening. A second type of safety POD mitigates damages as do evacuation plans. Some safety PODs function to both keep workers out of harm's way and to mitigate damages.

This paper describes a fault tree model for safety based on enabling causes, triggering events and safety PODs. It demonstrates how enabling causes, triggering events and safety PODs can be modeled using an example of crane failure. The fault tree presented in the paper is based on binary logic. A key feature of this model is the introduction of a new gate to represent the action of safety PODs. The model can be used to calculate outcome probabilities in much the same way as probability of the top undesired event are computed in conventional fault trees that do not model safety PODs.

The model can be used to achieve the desired level of safety by evaluating the possible outcomes of various configurations of safety elements. "What if?" scenarios can be generated to determine the effect of reducing or eliminating different enabling causes. They also can be used to evaluate the action of PODs on enabling causes and triggering events and to identify PODs that could be effective in increasing the level of site safety.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £95 +P&P)