News

Risk Analysis – FTA / ETA

Written by GM International | Feb 03, 2021

FTA (Fault Tree Analysis) and ETA (Event Tree Analysis) are deductive analytic techniques to analyze faults and accidents, determining their causes using Tree Structures.

Fault Tree Analysis or fault tree is one of the most popular and well-known methods of risk analysis. FTA should be counted among the deductive analysis methods because, starting from a general and overall analysis of the type of failure or unwanted event on the system, it identifies components’ failures. FTA and other analytical methods originated around the 1960s at Bell Telephone laboratories supported by reliability theory and Boolean algebra. Since the 1960s, FTA methods have found increasing application in the manufacturing and service sectors and are now one of the most simple and effective methods for analyzing the reliability and safety of systems.

 

In a similar form, analysis using event trees (ETAs) is an inductive method developed forward in time. Basically, it is assumed that the safety systems are either available or faulty; in other words, they either fully meet expectations or do not function at all when they are put into operation. In this way, a tree structure with binary branches is created, resulting in a number equal to the possible accidental sequences of 2n, where n is the number of safety systems considered. The purpose of ETA techniques is therefore complementary to that of FTA techniques. Basically, an FTA explains how an undesired event can be the result of a series of failures in the subsystems that make up the analyzed system, while the ETA examines all the possible consequences of such undesired event.

 

FTA (Fault Tree Analysis)

The FTA allows, in a graphical and logical way, the linking of faults of the components in a system. The main purpose is not so much to identify the causes of failures but, starting from a failure on the system (undesired event), to put it in functional relation with the failures on the components (basic events). It is a deductive technique that focuses on a major incident (Top Event) and it provides a method for identifying its possible causes. The fault tree is a graphical model that explicitates the logical connections between concurrent fault and human error events that can occur in a system.

The strength of FTA as a qualitative tool lies in its ability to identify combinations of events that may lead to the relevant undesirable event. By evaluating the probability of failure of the individual components it is possible to calculate the probability of occurrence of the Top Event and by reducing the tree, it is possible to identify the so-called MCS (Minimal Cut Set), concepts derived from the graph theory that generally correspond to the combinations sufficient to cause the Top Event with the maximum probability of occurrence. This analysis is strongly recommended for systems with high redundancy. For single MCS systems it is convenient to use techniques oriented to the evaluation of single failure events such as FMEA and HAZOP, while the FTA technique can be used to investigate in detail some subsystems.

 

ETA (Event Tree Analysis)

The Event Tree Analysis is a technique, which is often used for the definition of accidental hypotheses with FTA to identify potential events when boundary conditions vary, which allows to determine the potential consequences of an initiating event, in terms of accidental scenarios. This technique is among the most difficult to apply. In fact, meaningful results can only be obtained if the undesired (or even desired) events leading to the branching of the tree (the boundary conditions) are totally known at prior. It is clear that the application of this method requires solid practical experience, to anticipate all possible events in the system and to explore all possible consequences of such events. The sequence of events is generally defined by the barriers that may or may not be successful.

ETA is therefore an excellent methodology for risk analysis, being used to identify possible accident scenarios. As part of an investigation of an event, it is usual to highlight the actual path on the event tree that led to the manifestation of the scenario. Adopting a quantitative approach based on numerical valuation, the frequency of occurrence of each accident scenario is determined starting from the frequency of the initiating event. Those data are combined with the probabilities of occurrence of the intermediate events put in correspondence of the nodes of the tree, from where, then, the ramifications will be produced.