Leveson (2004) argued that event based accident models, such as FTA, RBD and FMECA, have limitations due to their emphasis on linear causality and inability to deal with non-linear relationships such as feedback, and may give only a superficial explanation about why a disaster may have occurred.