Abnormal Event Detection in Hydrocracking Units
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The Hydro Desulfurization and Cracking (HDC) unit in petroleum refineries faces challenges in timely detection of abnormal operations due to its complex and dynamic nature, leading to potential safety and economic issues, as current monitoring technologies often result in late notifications and high false alarm rates.
Innovation Solution
A method using multivariate statistical models and engineering models, combined with fuzzy logic and Principal Component Analysis (PCA), to compare process unit operations against a model of normal behavior, providing early alerts and hierarchical displays to operators for diagnosing abnormal conditions, thus enabling timely corrective actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If conventional DCS alarm systems are used to monitor critical process measurements, then the system provides automated detection assistance, but the notification is delivered too late to enable sufficient time for operator action
Solution Approach 1:
The system performs preliminary analysis by continuously comparing current sensor readings against historical data and process models to detect abnormal patterns before they trigger conventional alarms. This early detection mechanism provides operators with advance warning, enabling timely corrective actions while maintaining high reliability through pattern recognition rather than simple threshold violations.
Solution Approach 2:
The system transitions from static threshold-based alarming to dynamic pattern recognition that adapts to changing process conditions. By using historical data and process models that evolve with operating conditions, the system maintains detection sensitivity while reducing false alarms, thereby improving both response time and reliability simultaneously.
2Device complexity
If static operating ranges are used for alarm triggers, then the system provides simple monitoring, but the system produces high false alarm rates in complex processes
Solution Approach 1:
The system changes from fixed static thresholds to dynamic parameter ranges that adapt based on historical operating data and current process conditions. By continuously updating expected parameter ranges using statistical methods and process models, the system maintains simplicity in implementation while dramatically improving alarm accuracy and reducing false positives in complex HDC processes.
3Loss of information
If operators survey all critical sensors in tabular and trend format, then complete monitoring coverage is achieved, but the onset of abnormality can easily be overlooked
Solution Approach 1:
The system extracts and highlights only the critical abnormal patterns from the vast array of sensor data by comparing current readings against historical norms and process models. This extraction mechanism presents operators with a focused subset of relevant information rather than requiring survey of all sensors, maintaining complete monitoring coverage while dramatically reducing operator workload and preventing oversight of abnormal conditions.
Data Source
AI summary
The present invention is a method for detecting an abnormal event for process units of a hydrocracking unit. The method compares the operation of the process units to a model developed by principle components analysis of normal operation for these units. If the difference between the operation of a process unit and the normal operation indicates an abnormal condition, then the cause of the abnormal condition is determined and corrected.