Smart Water Network Management for Urban Utilities

A regional water utility was losing significant water volume through undetected leaks and inefficient pressure management across a distributed pipe network serving 400,000 residents. Sequoia built a sensor-driven monitoring platform that reduced network losses by 35% and cut leak response times by 60%.

The Challenge

Water loss in distribution networks is a major operational and financial problem for utilities. Undetected leaks waste treated water, damage infrastructure, and increase energy costs through unnecessary pumping. Manual inspection cycles were insufficient for a network of this scale — leaks could persist for weeks before detection.

The utility had limited visibility into real-time pressure and flow conditions across the network. Pressure anomalies that typically indicate pipe stress or burst events were identified only after customer complaints or visible surface damage.

“We knew we were losing water — we just couldn't pinpoint where. By the time a customer reported a burst pipe, it had often been leaking for days. We needed to find these events before they became incidents.”

— Network Operations Manager

Our Approach

Sequoia designed a layered monitoring and analytics platform built around three capabilities:

• Distributed sensor network — 180+ pressure and flow sensors installed at key network nodes across the distribution system, transmitting telemetry at 15-minute intervals to a centralised cloud platform.
• Anomaly detection algorithms — Statistical pressure pattern analysis combined with network topology modelling to identify deviations consistent with leak events, with spatial triangulation to narrow source locations to within 200 metres.
• Automated alert and dispatch system — Prioritised alert workflow integrating with the utility's existing field crew management system — generating workorders with GPS coordinates and estimated leak severity for field response.

The Outcome

Field crews now respond to accurately located, prioritised leak alerts rather than broad reported areas. The reduction in search time has compressed repair cycles significantly and reduced the volume of water lost between detection and repair.

The utility is now extending the platform to support long-term infrastructure planning, using network stress data to model asset replacement schedules and defer capital expenditure on sections of pipe that remain in acceptable operating condition.

Published by the Sequoia Team