Substations are the joints of any national grid: fail one major switching yard and you can black out a region for days. Ground-based sensor networks are expensive to deploy uniformly, miss external threats such as differential settlement or perimeter flooding, and generate data that sits in a utility SCADA system with no independent audit trail. Grid operators consequently fly blind on the physical condition of the hundreds of substations that sit outside their routine inspection cycle.
A small constellation of thermal-infrared and SAR microsatellites closes that gap decisively. Thermal IR detects resistive heating in transformers, bus bars and switchgear — the classic precursor to a catastrophic arc fault — at a sensitivity of roughly 0.3 K above ambient. Repeat-pass InSAR on the same sites picks up millimetre-scale subsidence that signals foundation movement or soil liquefaction under heavy infrastructure. Together, the two payloads give operators a weekly independent health snapshot of every substation in the country, regardless of whether the site has working ground sensors.
The operational outcome is straightforward: anomalies are ranked by severity, geolocated to within 5 metres, and delivered to the national grid control centre as a prioritised inspection queue. Maintenance crews go where the data says to go rather than following a calendar. Insurers see an auditable record of due diligence. And when a transformer does fail, post-event imagery confirms whether the root cause was thermal runaway, settlement or storm damage — feeding directly into the next procurement cycle.
Frequently asked
What exactly does a satellite see at a substation that indicates a health problem?
Satellites equipped with medium-wave or long-wave infrared sensors detect surface temperature anomalies on transformer tanks, bushings, switchgear enclosures and cable trays. A transformer running hotter than its load profile predicts, or a bushing showing an asymmetric heat signature relative to its pair, are classic precursor signatures. SAR satellites add the ability to detect structural deformation or standing water around foundations regardless of cloud cover or time of day.
Can satellite data replace SCADA or on-site condition monitoring?
No, and any vendor claiming otherwise should be treated with scepticism. SCADA systems provide sub-second telemetry on electrical quantities; satellite passes are periodic snapshots. The correct framing is that satellite health indicators surface anomalies across every substation in a national grid simultaneously — something ground sensor networks rarely achieve for cost reasons — and then direct field crews or detailed sensor attention to confirmed problem sites.
How does a government procure this capability without depending on a foreign commercial provider?
The sovereign route is to fund a dedicated nanosatellite or microsatellite constellation carrying thermal infrared and optional SAR payloads, operated by the national space agency or a state-owned grid company. South Korea's KARI and India's ISRO have both demonstrated that mid-size nations can build and operate multi-payload Earth observation missions at costs well below $500M for a 12–16 satellite constellation. The alternative — tasking Planet, ICEYE or Capella — keeps the data pipeline but places continuity risk outside national control.
What grid size makes a sovereign constellation cost-justified?
As a rough threshold, a national grid with more than 800–1,000 high-voltage substations and a history of significant weather-related outage losses generates enough avoided-cost upside to justify the capital expenditure of a dedicated 8–12 satellite thermal constellation. Below that scale, a shared regional constellation — for example among ASEAN or African Union member states — provides the economics without sacrificing operational control.
How does this integrate with existing grid management software?
Processed anomaly alerts are typically delivered as GeoJSON or GeoTIFF files conforming to ISO 19115-1 metadata standards, which ingest directly into GIS platforms used by most transmission system operators. Leading grid management vendors including GE Vernova and Siemens Energy have published APIs for third-party condition monitoring feeds; the satellite data pipeline slots into those same interfaces. The integration challenge is organisational — establishing clear alarm thresholds and escalation workflows — rather than technical.
What is the typical lead time between a satellite-detected anomaly and a confirmed equipment failure?
EPRI's research (Report 3002022221) found that detectable thermal precursors precede catastrophic transformer failure by a median of 14 days, with some cases showing anomalies 60–90 days in advance. That window is operationally meaningful: it allows a maintenance crew to be scheduled, a spare transformer to be sourced from inventory, or a load redistribution plan to be activated before the failure forces an emergency outage.
Are there privacy or sovereignty concerns around a foreign satellite imaging our grid infrastructure?
Yes, and they are legitimate. High-resolution imagery of critical infrastructure — even commercially available imagery — can reveal operational patterns, redundancy gaps and equipment vintages that constitute sensitive national security information. Several nations, including the United States under its Critical Infrastructure Protection standards and the EU under the NIS2 Directive, are actively reviewing whether foreign satellite access to grid facility imagery should require licensing or notification. Owning the satellite resolves this: data never leaves national custody.
How does climate change affect the business case for this application?
Directly and positively, from a business case perspective. Heatwaves drive transformer loading to thermal limits more frequently; extreme precipitation events increase flooding risk at substation sites; wildfires — often detectable in early stages via satellite thermal imaging — present direct physical threats to transmission infrastructure. WMO and IPCC projections through 2050 indicate that weather-driven grid stress events will grow in frequency, making continuous satellite health surveillance an increasingly essential rather than optional capability.