Establishing an onshore pipeline incident database to support operational risk management in Brazil - part 1: Defining architecture

Abstract

This work aims to propose an interdisciplinary research for establishing a national onshore pipeline incident database. The creation of a failure database has the potential to allow effective demonstration to control authorities and the society of safety improvements implemented by operators over time, support the development of dynamic risk assessment and management models and comparison with international benchmarks metrics, among other benefits. Therefore, in this part 1 a robust approach is proposed in order to previously define an appropriate data collection architecture that reflects the current context of this activity in Brazil. Defining architecture properly is a key issue to ensure that future data collection would be able to generate relevant information (and them knowledge) to support and strengthen operational risk management. Through a systematic literature review, the state of art regarding several international benchmarks was established for mapping the parameters that typically compose pipeline incident databases taxonomies. Categories regarding events of failure, pipeline network and database management were identified. In sequence, an anonymous survey was carried out considering this information, involving 180 pipeline specialists of different profiles/backgrounds. Statistical tests were proposed for assessing consistency of its results and selecting the parameters that would finally compose Brazilian pipeline database architecture according to expert elicitation. Based on it, in part 2 a bowtie diagram is elaborated, depicting causes, consequences, prevention and mitigation controls associated with pipeline failure events. In addition, the calculation of failure statistics from historical data obtained from a Brazilian pipeline operator demonstrates that the architecture proposed in part 1 has practical applicability and can be used for the purposes of this work.