GPR Scanning Before Drilling: How Confidence Zoning Changes Site Decisions

GPR scanning before drilling is standard practice on most structural projects in Australia. What is less standard is how teams interpret the output. A scan report is not a simple pass or fail. It is a spatial map of what the operator could resolve, what remained ambiguous, and where the data quality dropped below the threshold needed to make a safe call. The decisions that follow depend on reading that map correctly.

What GPR Actually Produces

Ground penetrating radar emits pulses of electromagnetic energy into a substrate and records the time and amplitude of reflections. Reinforcing bars, post-tensioning tendons, conduits, and voids all produce characteristic reflection signatures. The operator processes these hyperbolic returns to estimate depth, spacing, and orientation.

The output is typically a plan-view map overlaid on the scanned area, sometimes accompanied by B-scan radargrams showing the raw signal cross-sections. Both matter. The plan view tells you where targets are located. The radargrams tell you how clearly they were resolved and what the signal quality looked like at depth.

Neither output is infallible. GPR has physical limits that every project team should understand before making penetration decisions.

The Variables That Affect Confidence

Concrete Condition and Mix

High-chloride concrete, carbonated concrete, and concrete with elevated moisture content all attenuate the radar signal more aggressively than sound, low-permeability concrete. In older structures, particularly those with a history of water ingress or marine exposure, signal penetration depth can drop from 300 mm to less than 150 mm depending on frequency and mix. The scan may look complete, but the data below a certain depth is unreliable.

Reinforcement Density

Dense reinforcement creates a reflective barrier. When top-mat bars are closely spaced, the signal reflected from that layer can mask everything below it. Bottom-mat reinforcement, post-tensioning tendons, and embedded conduits may not be visible at all if the top layer is congested. This is not a failure of the equipment. It is a physical consequence of the geometry, and it needs to be stated clearly in the report.

Antenna Frequency

Higher frequency antennas (1.6 GHz and above) give better resolution for shallow targets but reduced penetration. Lower frequency antennas (400 MHz to 900 MHz) penetrate deeper but resolve closely spaced targets less precisely. The choice of antenna affects what can and cannot be detected, and scan reports should state which antenna was used and what depth range it was suited to.

Access Constraints

Not every surface can be scanned in two orthogonal directions. Walls with fixed plant, slabs with equipment in place, and ceilings with limited clearance may only allow scanning in one direction. Single-direction scans can locate bars running perpendicular to the scan lines but will miss bars running parallel to them. When access restricts scan coverage, the confidence in target identification drops accordingly.

Confidence Zoning: What It Means in Practice

Confidence zoning is the process of dividing a scanned area into regions based on the reliability of the data obtained. It is not a formal standard in the way that AS 1012.14 governs UPV testing, but it is a professional practice that responsible GPR operators apply when preparing reports for engineering use.

A three-tier framework is common:

• High confidence: The signal was clear, targets were resolved in two orthogonal directions, depth estimates are consistent, and no anomalies were detected in the proposed penetration zone. Drilling can proceed subject to normal hold-point procedures.
• Moderate confidence: The signal showed some attenuation or the scan was completed in one direction only. Targets were identified but spacing or depth estimates carry greater uncertainty. Verification by a second method, or adjustment of the penetration location, is recommended before proceeding.
• Low confidence or indeterminate: Signal quality was poor, dense reinforcement masked the zone of interest, or access prevented adequate coverage. The data is insufficient to clear the location for penetration. Engineering review is required before any work proceeds.

These zones should be marked on the plan-view output and referenced explicitly in the report recommendations.

How Project Teams Should Use the Output

Deciding Where to Core

High-confidence zones with clear separation between identified targets are the starting point for core location selection. The proposed penetration point should sit within the clear zone, not at its edge. A 50 mm margin from a resolved rebar is not the same as a 50 mm margin from the edge of a high-confidence zone. The latter requires the operator to confirm that no unresolved targets exist within that margin.

For concrete core extraction, the core diameter also matters. A 75 mm core through a zone where bars are spaced at 150 mm centres leaves less tolerance than the numbers suggest, particularly if depth estimates carry a plus or minus 10 mm uncertainty. Selecting a penetration point with the widest possible clearance from resolved targets is always preferable to threading a core through a tight gap.

Deciding Where to Verify

Moderate-confidence zones should not be dismissed or cleared on the basis of GPR alone. The appropriate response is to apply a second method. Ferroscan or cover meter surveys can confirm reinforcement position and depth in the upper 60 to 80 mm of a slab or wall with high accuracy, independent of the GPR signal quality. Where the GPR uncertainty relates to depth rather than position, UPV testing or direct probing may add useful information.

The combination of methods does not need to be expensive or time-consuming. A targeted Ferroscan pass over a moderate-confidence zone takes minutes and can convert an ambiguous result into a workable clearance. The cost of that additional pass is negligible against the cost of a severed tendon or a structural repair.

Deciding Where to Stop

Low-confidence zones require a different response entirely. The site team should not attempt to work around the data gap by moving the penetration point slightly or proceeding on the assumption that the target is probably not there. If the data cannot confirm it, the penetration should be placed on hold.

The hold-point should trigger an engineering review. The engineer of record needs to assess whether the penetration can be relocated to a zone where adequate data exists, whether additional investigation methods can resolve the uncertainty, or whether the penetration is structurally unnecessary and can be eliminated from the scope. That decision belongs to the engineer, not the drilling contractor.

Depth Limits and What Lies Below Them

Every GPR scan has an effective depth limit. That limit varies by substrate, antenna, and signal conditions, but it is always finite. A report that shows a clear slab cross-section to 200 mm depth does not confirm that nothing exists below 200 mm. It confirms that the signal did not resolve anything below that depth, which may mean the slab ends there, or it may mean the signal attenuated before reaching deeper targets.

For slabs on ground, this is usually not a concern. For suspended slabs with services, for walls with embedded conduits at varying depths, or for transfer structures with multiple reinforcement layers, the depth limit is a material constraint. The report should state the estimated effective depth of reliable data, and project teams should not assume clearance for targets that fall outside that range.

When depth limits prevent adequate coverage of the full cross-section, the appropriate action is to note the limitation, adjust the scope if possible (lower frequency antenna, additional passes, alternative access), or escalate to engineering review.

What the Report Should Contain

A GPR report prepared for penetration clearance should include the scan date and operator credentials, the antenna frequency and equipment used, the scan direction or directions, a plan-view output with identified targets and confidence zones marked, radargrams for locations of interest, a clear statement of depth limits and any signal quality issues, and specific recommendations for each proposed penetration point.

Reports that consist only of a plan-view image with coloured lines and no written interpretation are not adequate for engineering decision-making. The operator's professional judgement about what the data shows and what it does not show is as important as the image itself.

Site Workflow Around Scan Outputs

On projects where GPR is integrated into the programme from the start, the scan output feeds directly into the penetration permit process. High-confidence clearances are issued with the report. Moderate-confidence locations are flagged for verification before the permit is signed. Low-confidence or indeterminate locations are escalated to the structural engineer before any permit is issued.

This workflow prevents the common failure mode where a scan is completed, the report is filed, and the drilling crew proceeds on the assumption that silence means clearance. The report needs to be read, interpreted, and actioned. That requires someone on the project team with the authority and the knowledge to do it.

For facility managers and project managers who are not structural engineers, the practical takeaway is straightforward: treat the confidence zones on the scan output as binding instructions, not suggestions. High confidence means proceed with normal care. Moderate confidence means verify before proceeding. Low confidence means stop and get an engineer involved.

A Note on Combined Investigation Programmes

GPR is the most widely used pre-penetration method in Australia, but it works best as part of a broader investigation programme rather than as a standalone tool. On post-tensioned decks, combining GPR with Ferroscan provides independent confirmation of tendon and rebar positions. On older slabs where signal quality is uncertain, UPV testing can characterise concrete condition and inform the interpretation of GPR results. On structures where corrosion is a concern, half-cell potential mapping adds a layer of information that GPR cannot provide.

The right combination depends on the structure, the substrate, and the scope of work. SiteOps designs investigation programmes around those variables, with NATA-accredited testing and reporting that supports engineering decisions at every stage. For more information on GPR scanning, confidence zoning, and combined NDT programmes, visit siteops.au.