Using Advanced Analytics to Identify the Most Probable Locations of Corrosion Under Insulation
Autor: | Nivedita K. Kumar, Kjersti H. Løken, Bruce Mackenzie |
---|---|
Rok vydání: | 2019 |
Předmět: |
Computer science
business.industry 020209 energy Bayesian network 02 engineering and technology 021001 nanoscience & nanotechnology computer.software_genre Corrosion under insulation Analytics 0202 electrical engineering electronic engineering information engineering Data analysis Data mining 0210 nano-technology business computer |
Zdroj: | Day 2 Wed, September 04, 2019. |
Popis: | Over 20 percent of major oil and gas (O&G) incidents reported within the European Union (EU) since 1984 have been associated with corrosion under insulation (CUI) [1]. Challenges are particularly acute when the source of risk is hidden, as in the case of CUI. With data being continuously generated, significant effort is required to manage data and mitigate risk. Using bayesian networks (BNs) Oceaneering has developed a decision support system for effective CUI risk management. The Bayesian model can be incorporated into existing risk-based assessment (RBA) systems. A key feature of the model is the ability to predict corrosion hotspots while quantifying uncertainties. The model uses probabilities based on objective data as well as subject matter expertise, which makes analytical techniques in business accessible to a wide range of users. With a case study we illustrate how BNs can be used to assess the risk of a fuel gas line on a live asset in the North sea. The most likely estimated remaining life (ERL) is forecasted in the range of 13 to 24 years, with a worst case of 6.7 years and best case of 40 years. By comparison, the customer CUI tracker reported an ERL of 9.7 years. BNs increase flexibility for scheduling inspection intervals, enabling more targeted inspection planning. This is a significant advancement from current RBA methodologies. |
Databáze: | OpenAIRE |
Externí odkaz: |