| Autor: |
Iraj Ershaghi, Robert House, Ke-Thia Yao, S. Raghavendra, Mona Sharifi, Jacques Blouin |
| Rok vydání: |
2015 |
| Předmět: |
|
| Zdroj: |
All Days. |
| DOI: |
10.2118/173995-ms |
| Popis: |
Offshore oil and gas platforms facilities include extensive piping components, pressure vessels, and other equipment. These are monitored regularly to minimize non-conformances for safety and integrity of platforms. This paper focuses on using field measurements of thickness in pipes to alert engineers to take appropriate actions when thickness falls below the minimum thickness. Data from the field at various thickness measurement locations (TMLs) are used to analyze for finding anomalies. Inspectors periodically measure at TMLs on pipes and record the thickness values. Based on these measurements they determine the remaining life time at a location before thickness falls below the minimum thickness. They do this by estimating corrosion rates and then calculate the next measurement date. When the measured thickness value at a TML is below the minimum they will issue action reports. With multiple measurements of thickness values at TMLs, we have developed a machine learning framework and algorithms to generate action alerts before corrosion lead to non-conformances. This approach is based on machine learning techniques to learn from examples when certain TMLs will have below the minimum thickness. Such alert prediction models are from a purely data driven approach and can help in prioritizing inspection schedule. The results show that our approach using machine learning is effective in capturing potential nonconformances in TMLs. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
