The article What is GPR concrete scanning? already explains the principles, antenna frequencies and achievable scan depths. This article picks up from there: once a rebar scan report lands on their desk, how should building owners, contractors or site supervisors read the radar images, the summary table and the limitations at the end of the report? It uses the most common case: counting reinforcing bars in concrete columns of an existing building.
What Does a GPR Rebar Scan Report Contain?
A complete, readable GPR rebar scan report has 5 main parts. If any one is missing, the results cannot be traced and checked.
- Scan location plan — a building plan marking point codes, e.g. G1–G20, with the floor and member type (column, beam, slab, wall).
- Radar images for each point — horizontal and vertical scans for each point, with arrows marking the arcs interpreted as bars.
- Sketch cross-section derived from the scan — showing the bar positions found, and distinguishing which bars were "actually detected" and which are "assumed".
- Summary table or text for each point — number of bars found per side, number of sides scanned, stirrup spacing and the height range scanned.
- Limitations and engineer's signature — conditions that may cause errors in the results, plus the signature of the inspecting engineer with their license registration number.
How to Read a Radargram in Practice? One Arc = One Bar
A radargram is a cross-section image that the GPR unit builds from reflected waves as it is rolled along the concrete surface. When the unit moves perpendicular to a bar, the wave reflects off the bar both before the unit reaches it and after it has passed, so the image forms an inverted arc. The apex of the arc is where the unit sits directly above the bar. Every radar image is read from 3 elements.
1. Horizontal axis = distance along the scan line
The horizontal axis is the distance the unit travels, in meters. The distance between adjacent arc apexes is the bar spacing; for example, apexes about 0.10 m apart mean the bars are spaced about 10 cm apart.
2. Vertical axis = depth from the surface
The GPR unit measures the two-way travel time of the wave and converts it to depth using the wave velocity in concrete. This velocity depends on the dielectric constant of the concrete, which changes with moisture content, so the depth axis is only as accurate as the velocity setting. The concrete cover section below explains how to verify this value.
3. Arc apex = bar position and depth to the top of the bar
The arc apex gives 2 values at once: the position along the scan line and the depth to the top surface of the bar, which is taken as the approximate concrete cover. Most display software shows the image as a color map, where arc apexes appear as rows of bright spots. Color and grayscale images are read using the same principles.
Can Column Bars Be Fully Counted from a One-Side Scan?
No. A one-side scan can only count the main bars adjacent to that face, which includes 2 corner bars. The total number of bars in the column is therefore a value calculated by assuming the bars are arranged symmetrically on all sides, not a count of every individual bar. For a rectangular column with corner bars at all 4 corners, the following geometric relationship applies.
Rectangular column, n₁ bars on the short side and n₂ bars on the long side → total bars = 2 × (n₁ + n₂) − 4
This formula only applies when the bars really are arranged symmetrically. If a column sits against a wall, is tiled, or is blocked by an electrical cabinet so that only one side is accessible, the report should state clearly "1 side scanned" and that the bars on the remaining sides are assumed values. In the report's sketch cross-section, the word Assume is commonly used to label bars that were not scanned directly.
← Scroll to see the full table →
| Sides scanned | What can be confirmed | What remains assumed |
|---|---|---|
| 1 side | Bar count on that side, bar spacing, stirrup spacing | Bar count on the other 3 sides and the column total |
| 2 perpendicular sides | Bar count on the short and long sides | Opposite sides (assumed equal) |
| 3–4 sides | Nearly the full bar count | Only bar size, which GPR cannot measure directly |
How Is Stirrup Spacing Read from a Vertical Scan?
Stirrup spacing is read from scans made by moving the unit up and down the column height, because this direction is perpendicular to the stirrups. Each arc apex is one stirrup, and the distance between apexes is the stirrup spacing in that zone.
The key thing to read is the height range over which the stirrup spacing changes. Many construction drawings call for closer stirrup spacing near the column ends, so a good report should not give a single stirrup spacing for the whole column but should state it by zone, e.g. "stirrups at 10 cm from 0–0.80 m and at 20 cm from 0.80 m up to the scan height limit", which can then be compared with the building's construction drawings. Judging whether the spacing found is adequate is the job of the structural engineer carrying out the assessment.
The scanning technician can usually only reach arm's height or ceiling level, so conclusions about stirrup spacing apply only to the height range actually scanned. The upper end of the column below the beam that was not scanned has no data.
How Accurate Is Concrete Cover from GPR?
Concrete cover from GPR is an approximate value. Its accuracy depends on how well the wave velocity setting matches the actual concrete. The way to verify it is to chip open the concrete at a calibration point, measure the actual distance to the bar with a tape measure or vernier caliper, and then adjust the wave velocity so the depth in the image matches the measured value. Factors that cause errors are as follows.
- Moisture in the concrete — moist concrete slows the wave and weakens the signal. If the velocity for dry concrete is used, the depth read will differ from the true depth.
- Plaster or applied finish layers — the depth read is measured from the surface the unit sits on. If the scan is made over plaster or tiles, the figure includes the thickness of the finish layer, so the report must state the condition of the scanned surface.
- Rough or cracked surfaces — the unit cannot sit flush on the surface, causing signal noise and blurred arcs.
If precise cover at specific points is needed, for example when accepting newly cast concrete, a Cover Meter is commonly used. It is a magnetic-field-based instrument that works well alongside GPR.
How to Read the Scan Summary Table? A Worked Example
The table below is an illustrative example showing how to translate the statements in a scan report for 4 columns into usable conclusions. The figures are assumed values to explain the reading method, not results from an actual project.
← Scroll to see the full table →
| Point | Sides scanned | Main bars found | Stirrup spacing found | What it means |
|---|---|---|---|---|
| C1 | 2 sides | 3 per side | @10 cm (0–0.80 m) · @20 cm (above 0.80 m) | Consistent with 8 bars arranged symmetrically; closer stirrups in the lower zone |
| C2 | 1 side | 3 bars | @20 cm throughout the scanned range | 8 bars is an assumed value; no closely spaced stirrup zone found within the scanned range; compare with drawings |
| C3 | 1 side | 4 bars | @10 cm (0–1.20 m) | More than the drawings; may be a lap splice zone; rescan at another height |
| C4 | 1 side | Arcs unclear | Unreadable | Signal interference, e.g. a damp surface or surface covering; open the surface or move the scan location |
Point C3 is a common case. More bars than shown on the drawings within a certain height range usually indicates a lap splice, where two bars overlap. Another cause is metal conduits or embedded steel plates, which produce arcs similar to bars. Rescanning at a different height helps tell them apart, because a lap splice disappears beyond the lap length.
Scan Results Don't Match the Drawings: What Next?
Scan results that do not match the drawings are not yet a conclusion that the structure is deficient. The steps to take, in order, are to confirm the results first and then pass them to a structural engineer for assessment.
- Rescan on another side or at another height — rule out lap splices and other embedded objects first.
- Perform an area scan at critical points — scan as a grid in both X and Y, e.g. over a 0.60×0.60 m area, to see a map of the bars across the whole panel.
- Chip open the surface to confirm — the only way to confirm bar size. Open at least 2–3 points per job, and use them as cover calibration points at the same time.
- Supplement with other tests — such as a Core Test when the actual concrete compressive strength is needed, or Half-Cell Potential when reinforcement corrosion is suspected.
- Pass to a structural engineer for assessment — use the confirmed bar counts to check load capacity under actual use, which must be done project by project. For an overview, see Structural assessment of old buildings
What Limitations Should a GPR Report State?
A report should state at least 6 limitations. The limitations at the end of a report are not a disclaimer to protect the inspector; they are conditions that tell the reader how far the figures can be relied on. The American Concrete Institute guide ACI PRC-228.2-13 likewise states that GPR is well suited to locating reinforcement but is not a method for measuring bar size. The limitations to include are:
- Closely spaced or overlapping bars, such as lap splices or bundled bars at column corners, produce overlapping reflections that make the true count hard to determine.
- Outer bars shadowing inner bars in structures with multiple layers of reinforcement or thick sections, so deeper bars are subject to larger errors than outer bars.
- Moisture and chlorides weaken the signal, especially in structures that are regularly wet.
- Other metal objects, such as electrical conduits, embedded steel plates or anchors, may be read as reinforcement.
- Narrow scan areas or one-side access mean some results must rely on the assumption that bars are arranged symmetrically.
- Results apply only to the locations scanned and cannot stand in for the whole member or the whole building.
7-Point Checklist Before Accepting a GPR Rebar Scan Report
- Point codes match in every part — the location plan, radar images, sketch cross-sections and summary table all use the same point codes.
- Radar images in both directions — horizontal for counting main bars, vertical for stirrup spacing, with arrows marking the arcs used in the conclusions.
- Number of sides scanned stated for each point — clearly separating detected values from assumed values.
- Stirrup spacing given by height range — with the maximum height scanned stated.
- Depth calibration method stated — with chipped-open points or the velocity value used. If neither is given, treat the cover as a rough estimate.
- Scanned surface condition recorded — bare concrete, plaster, tiles or a damp surface.
- Limitations and engineer's signature included — signed by an engineer holding a professional engineering license (COE), with the registration number.
Summary: How to Read GPR Rebar Scans?
Reading GPR rebar scan results starts with the radar image, then covers 5 points in order.
- One arc — is one bar.
- Horizontal scan — used to count main bars and their spacing.
- Vertical scan — used to read stirrup spacing by height range.
- Apex depth — is the approximate cover; calibrate it against chipped-open points.
- Total bar count from a one-side scan — is an assumed value based on symmetry, and bar size can only be confirmed by opening the surface.
If you are planning a rebar inspection in an existing building, see our service details at Non-destructive structural testing (NDE), or see the factors that affect pricing in NDT structural testing prices 2026. GPR scanning fees depend on the number of points, the member types and site access. Contact us for a free quotation within 24 hours. Prices do not include travel and equipment transport, which are charged additionally based on the distance to site.
The radar images in this article are conceptual simulations, and the example table uses assumed values to explain the reading method; they are not results from any project. The bar-count formula is a geometric relationship that applies when bars are arranged symmetrically. Whether the reinforcement found is adequate for its use must be assessed by the responsible structural engineer for each project.