Pipe Vibration Spectrogram Detection of Passing Valves
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Oil and gas plants face significant challenges due to passing valves, which allow fluid to pass through closed valves, leading to costly leakages, environmental hazards, and business losses. Current methods lack effective detection and corrective measures for these issues.
Innovation Solution
A system and method utilizing a piezoelectric sensor to collect vibration data from pipes near valves, generating a spectrogram, and classifying it using a trained convolutional neural network (CNN) to detect passing valves. The system performs corrective actions, such as generating alerts or automatically closing upstream valves.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional valve monitoring methods are used, then the system is simple and easy to operate, but passing valves cannot be detected early leading to costly leakages and environmental hazards
Solution Approach 1:
The patent replaces traditional mechanical valve monitoring systems with acoustic vibration sensing and machine learning-based detection. Piezoelectric sensors capture vibration signals from the pipe, which are then processed through a convolutional neural network to detect passing valves. This substitution enables early detection of valve closure issues without requiring complex mechanical inspection systems.
Solution Approach 2:
The patent introduces acoustic vibration signals as an intermediary to detect valve closure status. Instead of directly monitoring valve position or flow, the system uses vibration signals from the pipe as a mediator that carries information about valve closure completeness. The convolutional neural network processes these intermediary signals to identify passing valve conditions.
2Object-affected harmful factors
If no detection system is implemented, then the system is simple and low-cost, but leakages cause environmental hazards and business losses
Solution Approach 1:
The patent implements a self-service detection system where the infrastructure itself (the pipe) generates the detection signals through its natural vibration response to valve closure. The piezoelectric sensor attached to the pipe captures these self-generated vibration signals, eliminating the need for external active testing or manual inspection. The system uses the infrastructure's own characteristics for self-diagnosis.
Solution Approach 2:
The patent establishes a feedback loop where vibration signals from the pipe are continuously monitored and processed by the convolutional neural network. When a passing valve is detected, the system can trigger alerts or automated responses, creating a closed-loop feedback system that enables timely corrective actions to prevent environmental hazards and business losses.
3Loss of time
If manual valve inspection is performed, then the system is simple and low-cost, but detection is delayed and requires human intervention
Solution Approach 1:
The patent implements preliminary action by continuously monitoring vibration signals in real-time, enabling detection of passing valves before they cause significant problems. The system proactively identifies valve closure issues as they develop, allowing for early intervention. The convolutional neural network is pre-trained to recognize patterns indicating passing valves, enabling immediate automated detection without waiting for manual inspection.
Solution Approach 2:
The patent replaces manual visual inspection or physical valve checking with automated acoustic sensing and machine learning analysis. The piezoelectric sensor and convolutional neural network system automatically detect passing valves by analyzing vibration patterns, eliminating the need for human operators to manually inspect each valve. This automation dramatically reduces detection time and enables continuous monitoring.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution enables early and automatic detection of passing valves, allowing for timely corrective actions that reduce environmental impact, operational costs, and business losses, while also improving safety and resource management.
Implementation Method 1
A small sensing device (e.g., piezoelectric sensor) can be placed on or near the valve to collect high frequency vibration (e.g., acoustic) data.
Data Source
AI summary
This disclosure describes systems and methods for detecting passing valves. A method includes measuring vibrational data with a sensor coupled to a pipe adjacent to a valve; generating a spectrogram representing a time variation of frequencies of the vibrational data; detecting that the valve is a passing valve by classifying the spectrogram as including frequencies representative of a passing valve, the classifying being based on a trained convolutional neural network; and in response to detecting the passing valve, performing a corrective action to resolve the passing valve.