Abnormal Event Detection in Delayed Coking Units
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Delayed Coking Units (DCUs) face challenges in detecting abnormal operations due to the complexity and high-severity nature of the process, leading to premature coking, foam carryover, and reliability issues with the coker fractionator, which can result in significant economic losses, safety hazards, and equipment damage.
Innovation Solution
A method and system for abnormal event detection (AED) in DCUs that compares online measurements to models of normal operation, automatically detects deviations, and provides the operator with timely information to take corrective actions, using fuzzy logic and Principal Component Analysis (PCA) to aggregate evidence and indicate the probability of problems, thereby reducing the likelihood of missed events and enabling early intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional alarm systems are used to detect abnormal operations in DCUs, then the system structure remains simple, but the detection precision is insufficient leading to late notification and missed events
Solution Approach 1:
The detection system is segmented into multiple specialized modules including PCA module for pattern recognition, fuzzy logic module for uncertainty handling, and multiple detection algorithms针对不同 types of abnormalities. Each module processes specific aspects of the complex DCU operation data, improving detection precision without overwhelming the overall system management.
Solution Approach 2:
The system transitions from traditional single-dimensional threshold-based alarm detection to multi-dimensional analysis by incorporating PCA transformation that projects high-dimensional process data into principal component space. This dimensional transformation enables detection of subtle abnormal patterns that single-parameter monitoring cannot capture.
2Loss of time
If traditional monitoring methods are used, then the device complexity is low, but the response time is delayed resulting in economic losses and safety hazards
Solution Approach 1:
The system performs preliminary detection and analysis continuously in the background using PCA and fuzzy logic algorithms, preparing abnormality assessments before actual alarm conditions occur. This preliminary action enables the system to notify operators of developing abnormalities before they escalate into critical failures, reducing response time and preventing economic losses.
Solution Approach 2:
The system implements continuous feedback loops where detection results feed back into process control and operator notification systems. The fuzzy logic module continuously adjusts abnormality assessments based on incoming process data and historical patterns, providing real-time feedback that enables timely corrective actions before abnormalities worsen.
Data Source
AI summary
The present invention is a method for detecting an abnormal event for process units of a Delayed Coking Unit. The method compares the operation of the process units to statistical and engineering models. The statistical models are developed by principal components analysis of the normal operation for these units. The engineering models are based statistical and correlation analysis between variables. If the difference between the operation of a process unit and the normal model result indicates an abnormal condition, then the cause of the abnormal condition is determined and corrected.