Flow Detector Leak Detection Using Statistical Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing leak detection systems in liquid and gas transportation networks are ineffective in detecting small leaks, determining flow direction, and distinguishing between leaks and normal flow variations, often requiring complex and expensive hardware and taking too long to detect small changes accurately.
Innovation Solution
A system comprising a flow detector and controller that measures and analyzes the rate and direction of flow in a transportation network, using statistical analysis and algorithms to identify leaks, determine their severity, and potentially take corrective action, with the ability to learn the network's flow patterns and differentiate between static and dynamic states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If acoustic or electromechanical detectors are used to detect leaks, then major leaks can be detected, but small flow changes (1% or less) cannot be detected
Solution Approach 1:
The patent replaces complex acoustic or electromechanical detectors with a flow measurement system that uses a flow sensor and statistical analysis algorithm. Instead of relying on complex physical detection mechanisms, the system measures flow rate and compares it against expected flow patterns to detect leaks, thereby reducing device complexity while improving sensitivity to small leaks.
Solution Approach 2:
The system changes the detection parameter from acoustic signals or mechanical responses to flow rate measurements. By monitoring flow rate and comparing it against statistically determined expected values, the system can detect small leaks that would be imperceptible to acoustic or electromechanical detectors, thus improving measurement precision.
2Measurement precision
If multiple sensors are placed at different points in the transporting network to measure flow changes, then leak detection accuracy improves, but system complexity and cost increase
Solution Approach 1:
The patent extracts the leak detection function from a multi-sensor system and implements it using a single flow sensor combined with statistical analysis. Instead of distributing multiple sensors throughout the network, the system uses one sensor to measure flow and employs an algorithm to determine whether the measured flow deviates from expected patterns, thereby achieving accurate leak detection with minimal hardware.
Solution Approach 2:
The patent introduces statistical analysis as an intermediary between the flow sensor and leak detection decision. Rather than directly comparing flows at multiple points, the system uses statistical models to represent expected flow behavior and compares actual measurements against these models, enabling accurate leak detection with a single sensor.
3Measurement precision
If statistics tools are used to measure flow rate over long periods to detect small changes, then detection accuracy improves, but detection time increases and accuracy decreases
Solution Approach 1:
The patent applies partial action by using statistical analysis only when necessary - specifically, when the measured flow deviates from expected patterns. The system continuously monitors flow and applies statistical comparison in real-time, triggering detailed analysis only when anomalies are detected, thereby reducing overall detection time while maintaining accuracy for small changes.
Solution Approach 2:
The system uses feedback by continuously comparing measured flow against expected flow patterns and adjusting its detection behavior accordingly. When flow measurements align with expected patterns, the system operates in normal monitoring mode. When deviations occur, the system triggers leak detection algorithms, creating a responsive feedback loop that reduces detection time while maintaining accuracy.
4Loss of information
If prior art detectors are used to detect leaks, then major leaks can be identified, but the direction of flow and location of leaks cannot be determined
Solution Approach 1:
The patent adds the dimension of time to flow measurement by continuously monitoring flow rate and comparing it against time-dependent expected patterns. By analyzing how flow changes over time relative to scheduled operations and historical patterns, the system can infer leak location and direction without requiring multiple spatial sensors, thus gaining information in the time domain.
Data Source
AI summary
Described herein are systems, devices, and methods for detecting leaks of liquids or gases in a transporting network. In one embodiment, a flow detector for detecting volume and direction of flow is attached to a system transporting such liquid or gas, changes are sensed over time, and data is sent to and from a controller. In alternative embodiments, static and dynamic states are identified and differentiated to identify relatively small leaks. In some embodiments, the transporting network is emptied in whole or in part, thereby allowing the measurement of relatively small leaks. In some embodiments, aggregate measurement data is processed to identify usage and performance features particular to the transporting network, which allows a continuous improvement in the measurement of leaks and flow direction. In some embodiments, corrective action is taken automatically, while in other embodiments human operators order corrective action.