11.6.1 — Utility Infrastructure — maturity: live

Water Network Leak Indicators

Using satellite-derived soil moisture, surface deformation and thermal anomalies to flag probable leak zones in buried water distribution networks before catastrophic failure.

Multispectral and thermal satellite imagery can flag anomalous soil moisture and surface temperature patterns above buried water mains before a single litre reaches the surface — giving utilities days or weeks of advance warning.

Water utilities in most countries lose between 20% and 40% of treated water to leakage before it reaches a tap. Ground crews with acoustic loggers can only survey so many kilometres per year, and the backlog of undetected leaks grows faster than inspection capacity. Satellite observation changes the economics: persistent wide-area coverage can rank every pipe corridor by leak probability and direct ground crews only where the evidence is strongest.

Three complementary payloads do the work. Synthetic aperture radar interferometry (InSAR) detects millimetre-scale ground subsidence that accumulates above slow, chronic leaks. Thermal infrared imagery spots the cooler surface signatures of water migrating upward through soil. Multispectral bands catch the anomalous vegetation greenness that thrives above persistent moisture. Combined, these form a probabilistic leak-likelihood layer updated on each satellite pass.

The operational outcome is a ranked work-order queue delivered to field maintenance teams each morning. Utilities that have piloted similar approaches report a 30–50% reduction in physical survey kilometres while actually increasing the detection rate of significant leaks. For a sovereign operator, the network map and loss statistics never leave national infrastructure — a material consideration when water security is a strategic priority.

What matters

InSAR phase coherence degrades over vegetation; bare-soil or paved corridors give the clearest subsidence signal, and urban networks are precisely where leaks are most expensive.
A single undetected trunk-main failure can void months of water treatment investment and trigger a public health emergency within 48 hours.
Commercial InSAR services are export-controlled or subscription-gated, meaning a foreign vendor can suspend access during a dispute or crisis at exactly the moment national infrastructure is most stressed.
Revisit frequency drives detection latency: a 6-day repeat misses fast-developing failures, so a national constellation targeted at domestic pipe corridors can be tasked daily.

Quick facts

Economic cost of non-revenue water globally: $39 billion per year (2023) — World Bank Water Global Practice — NRW Overview
Thermal infrared resolution achievable in LEO microsatellite constellations: 3–5 m GSD (2024) — ESA — Earth Observation for Water Management
Average share of distributed water lost to leakage in OECD utilities: 23% (2022) — OECD — Water Governance in OECD Countries
Planet Labs daily imagery archive growth rate: ~1.5 million km² per day (2024) — Planet Labs — Planet Basemaps and Monitoring

Sovereignty score: 8/10 — A nation's water distribution network is critical national infrastructure whose vulnerability map must never reside on a foreign vendor's servers.

Leak-location data is, in effect, a detailed map of pipe condition and pressure weak-points; handing it to a commercial foreign provider creates an intelligence exposure that adversaries could exploit to plan sabotage or coercive pressure during a dispute.
Export-control regimes and vendor contractual clauses routinely allow suspension of InSAR data services, meaning a utility could lose its primary leak-detection feed at the precise moment of a national emergency or political tension.
Water loss rates and infrastructure condition are politically sensitive; a sovereign system lets the government control when, how and whether that data is disclosed to regulators, media or foreign partners rather than having it inferred from a vendor's archive.

Reference architecture

Payload: C-band SAR (5.405 GHz) for InSAR subsidence mapping at 5m resolution, 80km swath; secondary longwave thermal infrared (8–12 µm) channel at 30m resolution for surface moisture anomaly detection; optional 4-band multispectral (B, G, R, NIR) at 10m for vegetation index cross-check
Bus class: ESPA-class microsat, 120–160 kg, 600 W payload power; SAR antenna deployable to 2m × 0.5m planar array; thermal channel integrated as a compact pushbroom sensor on the same bus
Orbit: Sun-synchronous LEO at 520–560 km altitude; 6-satellite walker constellation providing 1–2 day revisit over national territory; dawn-dusk orbit preferred for stable solar power and consistent thermal acquisition geometry
Ground segment: 2-station national X-band downlink network (primary at capital, secondary at geographic antipode for resilience); S-band TT&C; on-site Level-0 ingest servers air-gapped from public internet; SatNOGS UHF beacon for housekeeping backup
Data pipeline: On-board L0 compression and store-forward → ground L1 radiometric calibration → InSAR processing stack (SNAP or ISCE open-source, hosted on sovereign GPU cluster) → ML anomaly classifier trained on historical leak events → probabilistic leak-likelihood raster at 10m grid posted to national GIS within 4 hours of downlink
End-user delivery: Web GIS dashboard for national water utility asset managers showing ranked pipe-corridor anomaly scores overlaid on network schematics; daily digest email with top-20 priority zones; REST API for integration with utility SCADA and work-order management systems; quarterly aggregate loss-index report to water regulator
Time to launch: First SAR demonstrator satellite in 22 months from contract award, enabling InSAR baseline accumulation; full 6-satellite constellation operational in 42 months; thermal channel validated against ground-truth acoustic logger data during commissioning phase
Caveats: C-band SAR modules from US primes (JPL heritage) may face ITAR restrictions; use European (Airbus, OHB) or Indian (ISRO-affiliated) SAR subsystem suppliers. InSAR coherence is poor over dense forest canopy; rural forested network segments will rely more heavily on thermal and multispectral channels. A GEO platform is not appropriate here — the resolution and coherence requirements demand LEO.

Frequently asked

What satellite bands are actually useful for detecting water leaks?

Thermal infrared (TIR) bands — typically 8–14 µm — are the primary workhorse because leaked water cools or warms the surface relative to dry surroundings. Short-wave infrared (SWIR, ~1.4–2.5 µm) and multispectral visible bands complement TIR by detecting soil moisture anomalies and vegetation stress caused by chronic seepage. Synthetic aperture radar (SAR) in C- or L-band can add coherent-change detection for subsurface moisture even through cloud cover.

How often does a satellite need to revisit the same pipe corridor to be operationally useful?

Most utility network managers consider a 24-hour or better revisit cycle the minimum for operational alerting; slower cadences are still valuable for monthly condition-scoring of the asset base. A 12-satellite LEO constellation at 500–550 km altitude can achieve sub-24-hour revisit at latitudes up to about 60°. Single-satellite or dual-satellite configurations are better suited to monthly strategic surveys than real-time leak hunting.

Can satellite data replace acoustic leak-detection crews entirely?

Not yet, and probably not at the sub-metre pipe-diameter level within this decade. Satellite imagery is most powerful as a prioritisation layer — flagging the 5–10% of the network most likely to have active leakage so that ground crews concentrate their acoustic or correlator surveys where they will be most productive. This can cut survey vehicle-kilometres by 30–50% while improving detection rates.

Why should a government own its own water-leak satellite capability rather than buying imagery from Planet or ICEYE?

Commercial providers can withdraw, reprice, deprioritise tasking or share data with third parties under terms outside the purchasing nation's control. Water supply is a life-critical infrastructure; during a drought emergency or geopolitical tension, a sovereign constellation guarantees uninterrupted access and full data sovereignty. Ownership also allows the government to set orbit, revisit and encryption parameters to national security standards — something no commercial service level agreement can replicate.

What is the typical capital cost of a 12-satellite LEO microsatellite constellation for this purpose?

A 12-unit constellation of 50–150 kg class microsatellites with thermal and multispectral payloads, ground segment and five years of operations currently falls in the $80–200 million range depending on procurement strategy, launch vehicle choice and ground-station infrastructure. Shared-government bus programmes and rideshare launches can compress this substantially. The World Bank estimates the global cost of non-revenue water at $39 billion per year, making even a $150 million constellation a compelling return-on-investment proposition for a mid-sized utility nation.

How is the satellite data integrated with a utility's existing GIS and SCADA systems?

Most modern delivery pipelines publish change-detection rasters and vector alert layers via OGC-compliant Web Coverage Service or Web Feature Service endpoints (OGC 06-121r9), which map directly into ESRI, QGIS or Bentley infrastructure platforms. SCADA integration is typically indirect: analysts convert satellite alerts into work-order tickets in the utility's asset management system (e.g. IBM Maximo or SAP PM) rather than feeding raw rasters to operational control rooms.

How does weather affect the reliability of satellite leak indicators?

Persistent cloud cover is the single biggest operational constraint for optical and TIR sensors — equatorial and maritime climates can see usable clear-sky windows on fewer than 50% of days. SAR-equipped satellites (such as ICEYE or Capella Space) provide cloud-independent surface-moisture change detection and should be part of any all-weather sovereign architecture. WMO's Guide to Remote Sensing for Hydrology (WMO No. 1160) provides regional cloud-climatology data useful for constellation sizing.

What accuracy rate can utilities realistically expect from satellite-derived leak alerts?

Published pilot studies report leak-indicator precision (true positives as a share of total alerts) in the range of 55–75%, with recall (share of actual leaks flagged) dependent heavily on pipe depth, soil type and season. These numbers improve significantly when satellite alerts are fused with historical break records, pipe-age data and pressure-zone maps held by the utility. Treating satellite outputs as probabilistic risk scores rather than binary leak confirmations is the operationally mature approach.

Glossary

GSD (Ground Sampling Distance): The distance between the centres of adjacent pixels on the ground, used to express the spatial resolution of a satellite image — a smaller GSD means finer detail.
TIR (Thermal Infrared): The 8–14 µm portion of the electromagnetic spectrum where objects emit heat radiation detectable by satellite sensors, used to identify surface temperature anomalies caused by subsurface water.
SWIR (Short-Wave Infrared): Wavelengths roughly between 1.4 and 2.5 µm that are particularly sensitive to liquid water content in soils and vegetation, complementing thermal sensors in leak-detection workflows.
SAR (Synthetic Aperture Radar): An active microwave sensor that transmits and receives its own radar pulses, enabling cloud-penetrating surface and subsurface moisture mapping independent of sunlight or cloud conditions.
Non-Revenue Water (NRW): Water that is produced and distributed by a utility but lost before reaching a paying customer — through physical leaks, meter error or unauthorised connections — typically expressed as a percentage of system input volume.
Coherent Change Detection (CCD): A SAR analysis technique that compares the phase and amplitude of two or more radar passes over the same area to detect subtle surface or near-surface changes, including soil moisture variation caused by leaking pipes.
Revisit Time: The elapsed time between successive satellite observations of the same ground location; shorter revisit times enable near-real-time monitoring of dynamic infrastructure conditions.
Data Sovereignty: A nation's legal and practical authority to control where its satellite-derived data is stored, who can access it, and under what conditions — critical when the infrastructure being monitored is classified as nationally critical.
Tasking: The process of commanding a satellite to point its sensor at a specific target location on a specific pass, as opposed to passive collection of whatever falls within a standard imaging strip.
WCS (Web Coverage Service): An OGC standard interface that allows clients to request and receive gridded raster data — such as satellite-derived temperature or moisture maps — directly into GIS platforms over the internet.

References

Guide to Remote Sensing for Hydrology — WMO No. 1160 — WMO's authoritative guide covers the application of satellite-derived soil moisture, thermal infrared and SAR datasets to water resources management, including urban distribution system monitoring and anomaly detection in buried infrastructure.
OECD Water Governance in OECD Countries: A Multi-Level Approach — The OECD benchmark survey found that average physical water losses across OECD member utilities stand at 23% of system input volume, with significant variation between well-funded and under-resourced networks, underlining the financial case for improved monitoring technology.
ESA Earth Observation for Water Management — ESA's Copernicus programme and dedicated EO research activities have demonstrated that fused optical, thermal and SAR data can identify soil moisture anomalies indicative of subsurface leaks at scales relevant to municipal distribution networks, with ongoing pilot programmes in southern Europe and Africa.
Planet Labs — Monitoring Product Documentation — Planet's daily imaging archive — growing at approximately 1.5 million km² per day — provides the temporal density needed for change-detection workflows targeting chronic low-level infrastructure leakage, though spatial resolution and thermal band availability remain constraints at the commercial tier.

Related applications

11.6.3 — District Heating Performance (Utility Infrastructure)
11.6.4 — Gas Distribution Surveillance (Utility Infrastructure)
11.6.5 — Telecom Tower Inventory (Utility Infrastructure)
11.1.1 — Upstream Asset Monitoring (Oil & Gas Intelligence)
11.1.2 — Inventory & Storage Tracking (Oil & Gas Intelligence)
11.1.3 — Refinery Throughput Estimation (Oil & Gas Intelligence)
11.1.4 — Flaring Detection (Oil & Gas Intelligence)
11.1.5 — Pipeline Integrity Surveillance (Oil & Gas Intelligence)