Water ingress detection in wastewater networks

EP4659174A1Pending Publication Date: 2025-12-10STORMHARVESTER IPR LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
EP2024702996
Authority / Receiving Office
EP · EP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-03-23
Filing Date
2024-01-31
Publication Date
2025-12-10

AI Technical Summary

Technical Problem

Conventional wastewater networks face challenges in accurately and efficiently detecting water ingress, which leads to increased maintenance costs and damage due to the difficulty in pinpointing infiltration sites, affecting treatment processes and infrastructure integrity.

Method used

A computer-implemented method using machine learning models to predict wastewater flow by analyzing time series ground water level and rainfall data, identifying water ingress through wastewater flow delta calculations, and pinpointing affected regions within the network.

Benefits of technology

This approach enables precise detection and targeting of water ingress areas, reducing maintenance costs and extending the lifespan of wastewater infrastructure by facilitating timely and focused maintenance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure EP2024052411_08082024_PF_FP
    Figure EP2024052411_08082024_PF_FP
Patent Text Reader

Abstract

Methods, apparatus and / or systems are provided for detecting water ingress at a wastewater asset of a wastewater network For example, a method, apparatus and / or system configured to include: receiving a first set of time segments associated with a positive rate of change of a time series ground water level dataset associated with the wastewater asset; receiving a second set of time segments associated with negative rate of change of the time series ground water level dataset associated with the wastewater asset; applying a first set of time series rainfall data for the wastewater asset that is associated with the first set of time segments to a trained machine learning (ML) model configured for predicting a set of wastewater flow data associated with wastewater flow through the wastewater asset, where the trained ML model is trained based on a training dataset comprising historical wastewater measurement data for the wastewater asset that is associated with the second set of time segments and a second set of time series rainfall data for the wastewater asset that is associated with the second set of time segments; calculating a set of wastewater flow delta data for the wastewater asset based on the difference between a set of measured wastewater flow data for the wastewater asset that is associated with the first set of time segments and the predicted set of wastewater flow data; and detecting water ingress for the wastewater asset based on identifying those portions of the set of wastewater flow delta data associated with water ingress.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Cited By

  • Gully malfunction detection

    EP4678834A1