Model-Based Leak Detection in Pipe Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current leak detection methods in water distribution systems are inaccurate and labor-intensive, failing to effectively locate leaks due to simplistic demand modeling and lack of real-time data, leading to inefficiencies and increased non-revenue water losses.
Innovation Solution
A Leak Detection System utilizing pressure sensors and advanced metering infrastructure, combined with real-time hydraulic modeling and proprietary algorithms, to continuously monitor pressure discrepancies and calculate potential leak locations, reducing manual effort and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional leak detection methods are used, then labor-intensive manual inspection is required, but detection accuracy remains low and time consumption increases
Solution Approach 1:
The patent replaces manual mechanical inspection methods with an automated computer-based system that uses hydraulic models, pressure sensor data, and consumption meter data to detect leaks. The computer automatically performs calculations and comparisons to identify potential leak locations, eliminating the need for labor-intensive field inspections while significantly improving detection accuracy and reducing time consumption.
Solution Approach 2:
The patent introduces a computer-based hydraulic model as an intermediary between raw sensor data and leak detection conclusions. The model acts as a mediator that processes pressure and consumption data, compares actual measurements with modeled expectations, and identifies discrepancies indicating leaks, thereby automating the detection process while maintaining high accuracy.
2Measurement precision
If simplistic demand modeling is used, then model complexity is reduced, but leak detection accuracy deteriorates due to inability to accurately attribute consumption to service nodes
Solution Approach 1:
The patent changes the parameters used in demand modeling from simple historical averages to dynamic parameters that incorporate real-time consumption meter data, pressure sensor readings, and service node characteristics. By adjusting these parameters to reflect actual system conditions and using algorithms to attribute consumption to specific service nodes based on multiple factors, the model achieves higher accuracy without becoming prohibitively complex.
Solution Approach 2:
The patent implements feedback mechanisms where the hydraulic model continuously compares expected consumption and pressure values with actual measurements from meters and sensors. This feedback loop allows the model to refine its attributions and improve accuracy over time, justifying the increased complexity through measurable performance improvements in leak detection.
3Measurement precision
If historical average data is used for demand estimation, then data collection is simple, but the resolution and accuracy for particular times or days are insufficient
Solution Approach 1:
The patent transitions from static historical average data to dynamic real-time data collection and processing. The system continuously updates consumption estimates using current meter readings, pressure sensor data, and time-varying demand patterns, allowing accurate estimation for particular times or days rather than relying on smoothed historical averages, thereby increasing both accuracy and data resolution.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method, apparatus, and system for leak detection in a distribution network having consumption meters. The distribution network is divided into zones having an upstream location and a downstream location. An upstream pressure sensor detects the upstream pressure at the upstream location and the downstream pressure sensor detects the detected downstream pressure at the downstream location. A downstream pressure lookup table is used to determine an expected pressure at each downstream location based on a range of hypothetical upstream pressures at the corresponding upstream location and consumption data from the consumption meters. The expected pressure and the detected downstream pressure at each downstream location are compared to determine if the calculated discrepancy exceeds a discrepancy threshold. If a discrepancy exceeds a discrepancy threshold, a leak location lookup table containing a set of potential leak locations based on a range of hypothetical discrepancies is used to determine a set of probable leak locations.