Mining subsidence monitoring for underground cable duct integrity assessment

Persistent-scatterer and small-baseline InSAR from Sentinel-1 and COSMO-SkyMed Second Generation can detect millimetre-scale differential settlement along cable duct routes above legacy mine workings, giving network operators early warning of conduit joint failure risk before a service outage forces the diagnosis.

Sensors

Sentinel-1 (ESA, C-band, 5.6 cm wavelength): Interferometric Wide Swath mode delivers 5 x 20 m ground range resolution across a 250 km swath. Repeat pass over the UK is 6 days with a single satellite, 6 days effective with the A/B pair before Sentinel-1B failure reduced the constellation. Archive extends to 2014. Free and open; ideal for long time-series PS-InSAR and SBAS processing over legacy mine districts.
COSMO-SkyMed Second Generation (ASI, X-band, 3.1 cm wavelength): Stripmap mode achieves 3 x 3 m resolution; Spotlight modes reach approximately 1 m. Four-satellite constellation gives revisit of 12 hours to a few days depending on latitude and tasking. Shorter wavelength improves sensitivity to small displacements but increases phase decorrelation over vegetation. Commercial tasking required.
TerraSAR-X / TanDEM-X (DLR/Airbus, X-band): Stripmap at 3 m, High-Resolution Spotlight at approximately 1 m. Eleven-day repeat; shorter revisit achievable through right- and left-look combinations. Well-validated PS-InSAR archive over European industrial areas. Useful for cross-validating Sentinel-1 displacement time-series at higher spatial resolution along critical duct segments.
ICEYE X-band SAR constellation: Sub-metre Spotlight products with revisit configurable to daily over a target area through commercial tasking. Useful for rapid-cadence monitoring of an active subsidence event once a risk zone is identified by the longer archive analysis. No open archive; cost scales with tasking frequency.

Why legacy mine workings are a cable engineer's hidden liability

Britain alone has more than 100,000 recorded mine entries and an unknown number of uncharted shallow workings, many of them beneath urban areas where telecom ducts were laid decades after the mines closed. Germany's Ruhr, the Belgian coalfields, and parts of the Appalachian basin in the United States present comparable problems. The issue is not dramatic collapse. It is the slow, differential settlement of the overburden as voids consolidate or shallow pillar systems yield incrementally over years.

A buried duct can tolerate modest uniform settlement. What breaks conduit joints is differential movement: one section of duct descending faster than its neighbour, imposing shear and angular deflection at each coupling. A subsidence bowl 200 m across with a maximum settlement of 40 mm and a tilt gradient of 1 in 500 at its edge can generate enough angular displacement to open a push-fit joint by several millimetres, admitting groundwater and eventually causing cable failure. The problem is invisible until either a service outage or an expensive CCTV survey reveals it.

What InSAR actually measures, and what it cannot tell you

Synthetic aperture radar interferometry compares the phase of radar returns from the same patch of ground across two or more satellite passes. Where the ground has moved toward or away from the satellite along the line of sight between acquisitions, the phase difference encodes that displacement. Persistent-scatterer InSAR (PS-InSAR) identifies individual point targets, typically buildings, road furniture and hard infrastructure, that maintain stable radar reflectivity across a long image stack. Small-baseline subset (SBAS) methods use distributed scatterers and shorter temporal baselines to improve spatial density in less urbanised areas. Both approaches can resolve displacement rates down to approximately 1 mm per year over multi-year time-series under good coherence conditions.

The honest limits matter here. InSAR measures displacement in the satellite line-of-sight direction only. Decomposing that into vertical and horizontal components requires at least ascending and descending geometry acquisitions, and even then east-west horizontal displacement is better constrained than north-south. Atmospheric phase screen artefacts, particularly tropospheric water vapour gradients, can mimic or mask real deformation signals at the millimetre scale; time-series methods reduce but do not eliminate this ambiguity. Dense vegetation and loose soil cause decorrelation, reducing PS density in rural or disturbed ground. Critically, InSAR tells you nothing directly about void geometry or depth. A 5 mm per year settlement signal above a worked seam at 30 m depth and one at 150 m depth look identical at the surface. Geophysical follow-up, microgravity survey or seismic refraction, is required to characterise the subsurface risk.

Separating seasonal breathing from irreversible loss

Clay-rich overburden above many coal measures shrinks in dry summers and swells in wet winters, producing annual displacement cycles of 5 to 20 mm that are entirely elastic and reversible. A duct engineer who sees 12 mm of apparent settlement in a six-month Sentinel-1 interferogram and panics is probably looking at shrink-swell clay, not mine consolidation. Distinguishing the two requires a time-series long enough to observe at least two full annual cycles, typically three to five years of data.

The diagnostic signature of irreversible mine settlement is a monotonic downward trend that persists through seasonal cycles, often with a spatial pattern that correlates with known panel layouts or pillar grids in historical mine plans. Sentinel-1's archive from 2014 onward is long enough to resolve this separation reliably over most of the UK and Western Europe. Where the archive predates the duct installation, it can establish a pre-disturbance baseline against which future acquisitions are compared. COSMO-SkyMed Second Generation or TerraSAR-X data, at higher spatial resolution, can then be used to map the displacement gradient precisely at joint locations once the risk zone is defined.

From displacement map to joint failure probability

The output of a PS-InSAR time-series is a set of displacement velocities and time-series at each scatterer point. Converting that into something a cable engineer can act on requires two further steps. First, the point cloud must be interpolated onto the duct alignment to estimate differential settlement between successive joint positions, typically every 3 to 6 m for standard duct systems. The relevant metric is angular deflection at each joint, derived from the gradient of the settlement profile along the route.

Second, that angular deflection must be compared against the manufacturer's published joint tolerance, which for push-fit polyethylene duct systems is commonly in the range of 1 to 3 degrees before leakage risk becomes significant. Joints already near tolerance from original installation geometry have less margin. The analysis produces a ranked list of joint locations by residual angular margin, which is a direct input to a CCTV inspection prioritisation schedule. It does not replace inspection; it tells the maintenance team where to look first. Satellize applies this workflow, combining open Sentinel-1 time-series with commercial X-band tasking for high-gradient zones, and has developed the displacement-to-joint-risk translation layer as part of its analytics library alongside work such as the Kingdom of Tonga crop-estimation programme.

Practical constraints on data acquisition and processing

PS-InSAR requires a minimum of roughly 20 to 30 coherent acquisitions over the target area to produce reliable velocity estimates; with Sentinel-1 at 6-day repeat, that means a minimum of four to six months of data, though two to three years is preferred for the seasonal separation described above. Processing time for a 50 km duct corridor using a full SBAS stack runs to several days of compute on standard cloud infrastructure. Delivery of a displacement velocity map and joint-risk ranking for a corridor of that length is typically achievable within two to four weeks of data order, assuming the Sentinel-1 archive covers the area.

Urban areas present a practical advantage: building facades and road infrastructure provide abundant persistent scatterers, sometimes exceeding 500 PS per square kilometre, which is more than sufficient to interpolate confidently onto a duct alignment. Rural sections of a route, particularly where the duct passes through fields or woodland, may have PS densities below 10 per square kilometre, requiring SBAS methods or supplementary X-band data to maintain spatial continuity. Any gap in PS coverage longer than two or three joint spacings should be flagged as an unresolved segment requiring ground survey.

Typical figures

Spatial resolution (Sentinel-1 IW mode)	5 x 20 m ground range; PS point spacing in urban areas typically 5-30 m
Spatial resolution (COSMO-SkyMed SG Stripmap)	3 x 3 m; Spotlight modes to ~1 m
Minimum detectable displacement rate (PS-InSAR, multi-year stack)	~1 mm/year under good coherence; ~3-5 mm/year in lower-coherence rural settings
Revisit period	6 days (Sentinel-1 single satellite); 12 hours to several days (COSMO-SkyMed SG, tasked)
Sentinel-1 archive depth	2014 to present over most of Europe and UK
Radar frequency / wavelength	C-band 5.405 GHz / 5.6 cm (Sentinel-1); X-band ~9.6 GHz / 3.1 cm (COSMO-SkyMed SG, TerraSAR-X, ICEYE)
Displacement measurement direction	Line-of-sight only; vertical/horizontal decomposition requires ascending + descending pair
Atmospheric phase screen uncertainty	Typically 2-10 mm per acquisition; reduced to <1 mm/year in long time-series by temporal averaging
Delivery format	GeoTIFF displacement velocity raster, CSV/shapefile PS point cloud, duct-alignment joint-risk ranking (CSV or GIS layer)
Processing latency (new corridor, Sentinel-1 archive)	2-4 weeks from data order to joint-risk deliverable for a 50 km corridor

Analytics Satellize can run

Displacement velocity map along duct corridor	PS-InSAR time-series (StaMPS or similar) over Sentinel-1 C-band stack	GeoTIFF raster and PS point shapefile with mm/year velocity and uncertainty per point
Seasonal vs. irreversible settlement decomposition	Harmonic regression on displacement time-series to isolate annual cycle and residual trend	Per-point time-series CSV distinguishing elastic seasonal component from monotonic settlement trend
Subsidence bowl geometry characterisation	Spatial interpolation of PS velocities; bowl extent, maximum rate and tilt gradient computed along duct alignment	Annotated GIS layer showing bowl perimeter, maximum settlement rate, and tilt gradient contours
Joint angular deflection risk ranking	Settlement gradient interpolation onto duct joint positions; comparison against published joint angular tolerance	Prioritised joint inspection schedule as CSV ranked by residual angular margin, flagging joints below threshold
High-resolution differential settlement at critical joints	SBAS or PS-InSAR on commercial X-band (COSMO-SkyMed SG or TerraSAR-X) for flagged high-risk segments	Sub-3 m resolution displacement map for targeted segments; updated joint-risk ranking
Ongoing settlement monitoring alert feed	Incremental PS-InSAR update on each new Sentinel-1 acquisition; threshold exceedance detection	Automated alert (email or API) when displacement rate or gradient at a monitored joint exceeds defined threshold
Geophysical survey prioritisation report	Overlay of InSAR risk zones against historical mine plan digitisation and known seam depths	PDF report identifying segments where InSAR signal cannot resolve subsurface ambiguity and ground survey is required

Sources

Who does the work

We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.