How GPR Surveys Prevented a Construction Disaster Along the Nairobi Expressway

In 2023, Geopin Consult was commissioned by the Kenya National Highways Authority (KeNHA) to conduct a ground penetrating radar (GPR) utility mapping survey along 32 kilometres of the Nairobi Expressway corridor — the elevated highway running from Mlolongo through Syokimau, South B, and Lusaka Road to Westlands. The assignment brief was to locate, characterise, and map all buried utilities within the construction zone prior to piling operations and service diversion works. What the GPR found, at three locations along the corridor, was buried infrastructure that neither KeNHA's utility records nor the utility companies' own as-built drawings had flagged. This is the story of those finds — and what they would have cost to hit.

Why GPR Before You Dig — The Stakes in Nairobi

Nairobi's subsurface utility environment is, in the polite language of the industry, complex. Decades of uncoordinated infrastructure development by multiple utility authorities — Nairobi City Water and Sewerage Company, Kenya Power and Lighting Company (KPLC), Nairobi City County, Safaricom, Telkom Kenya, and a succession of other telecoms operators — have left a subsurface environment where utilities regularly occupy the same corridor without shared records, where as-built drawings reflect design intent rather than as-installed reality, and where decades of road resurfacing have buried infrastructure at depths and positions that no drawing accurately shows.

The consequences of striking buried utilities during construction are not theoretical. A 600 mm water main supplying a significant portion of Nairobi's south and east carries millions of litres per day; a strike during piling operations would cause immediate supply disruption affecting tens of thousands of households, take days to weeks to repair, and generate claims against the contractor running into tens of millions of shillings. A high-voltage underground power cable strike is worse: it is immediately life-threatening to site workers, causes widespread power outages, and can destroy construction equipment. A fibre optic duct bank strike interrupts the communications backbone for thousands of businesses and can trigger claims from multiple operators simultaneously.

The Survey Method: How GPR Works Underground

Ground penetrating radar works by transmitting short pulses of electromagnetic energy into the ground from an antenna dragged along the surface, then recording the time taken for reflected signals to return. When the transmitted pulse encounters a buried object or boundary between materials with different electromagnetic properties — a water-filled pipe, an air-filled duct, a concrete structure, a cable — part of the energy is reflected back to a receiving antenna at the surface. The depth of the object is calculated from the two-way travel time of the reflection and the velocity of the electromagnetic wave through the soil, which depends on the soil's dielectric constant.

The result, displayed in real time on the GPR control unit and processed in post-survey software, is a radargram: a vertical cross-section of the subsurface showing the amplitude of reflections against depth and survey position. Buried utilities appear as distinctive hyperbolic reflections — the characteristic arch shape produced by a point reflector as the antenna passes over it. An experienced GPR operator reads these hyperbolas to identify utility type (the shape and amplitude of the reflection vary with pipe material, size, and fill), depth, and orientation.

The Three Critical Finds

Of the 47 utility features mapped along the 32 km corridor, the majority were consistent with records held by the utility authorities — confirming their location and depth, but not producing surprises. Three finds were different. Each represented buried infrastructure absent from all utility records provided to the project and invisible to the surface inspection that had preceded the GPR survey.

Survey Output: What KeNHA Received

The deliverables from the 32 km GPR assignment were structured to meet the requirements of both the KeNHA design team and the construction contractor responsible for service diversions. Each utility feature detected was assigned a unique identifier, located by GPS coordinates at the surface, characterised by depth and estimated diameter or cross-section, attributed to the most probable utility owner based on depth, material signature, and spatial context, and classified by confidence level — High (confirmed by trial pit), Medium (strong hyperbola, characteristic of known utility type), or Low (anomalous reflection, source uncertain, investigation recommended).

The Lesson: What Utility Records Don't Tell You

The three critical finds on the Nairobi Expressway assignment share a common characteristic: they were all absent from the utility records that the project team had assembled before GPR was commissioned. This is not unusual — it is, in Nairobi's infrastructure environment, the norm rather than the exception. Utility records in Kenya suffer from several endemic deficiencies that make pre-construction utility searches necessary rather than optional on any project involving excavation.

As-built drawings are frequently not updated after installation changes or network upgrades. Utilities installed by contractors are often recorded only at the design stage, with deviations during installation never fed back into the asset register. Ownership of utilities changes when authorities are restructured or privatised, and records are not always transferred with the assets. Utilities abandoned in place rather than removed disappear entirely from active records. And in rapidly developing urban areas like Nairobi, the rate of new utility installation frequently outpaces record-keeping capacity.

GPR does not replace utility records — it supplements them. The starting point for any utility mapping assignment is always a thorough desk study: collecting all available records from all utility authorities, reviewing historical aerial imagery for evidence of past works, and conducting a surface inspection for visible markers (valve boxes, manhole covers, marker posts). GPR then surveys the ground to verify what the records show and — crucially — to find what the records do not show. On the Nairobi Expressway assignment, the desk study and surface inspection identified 44 of the 47 features ultimately mapped. The three critical finds were all in the GPR-only category.