Vol. 13, No. 1

TNT · April 2015 · 1

The American Society forNondestructive Testing

www.asnt.org

FOCUS

Introduction and BackgroundHaving regularly conducted investigationsof concrete structures, I’ve found thatmany challenging projects stretch thecapability of concrete imaging technology.In this paper, I highlight the use of groundpenetrating radar (GPR) imaging at anuclear plant.I was retained to complete a GPR

concrete imaging investigation at theClinton Power Station, south of Chicago,Illinois. The objective of the investigationwas to map the spatial location and depthof all embedded reinforcement bars andconduits that existed within a 1.8 3 m (6 10 ft) area prior to the installation of a new crane hoist. It was important tobe able to use an investigative tool thatwould allow the intelligent placement ofanchor bolts for the crane hoist. The data acquisition and interpretation of the results had to be delivered on site.The site engineer required a

nondestructive approach to be used for this project, since the structuralintegrity of the building could not becompromised in any way. X-rays wereruled out for health and safety reasons,since the building operated constantly and

had an open concept. Other concerns withX-rays involved the speed and simplicity of data acquisition. The exposure times for conventional radiography would havebeen too long given the size of the workarea and thickness of the concrete floor.GPR scanning was selected as the methodof choice since it provided rapid scanningof large areas with immediate onsite results. High-frequency radar scans were used to

image a 350 mm (14 in.) thick suspendedconcrete slab. Complications included thethicker than average slab, which was due to the presence of two beams, and havingto operate in an extremely noisy and highsecurity area. High-resolution data ofexcellent quality enabled definition ofembedded structural elements. Themaintenance engineers were able to use the results to effectively plan the placementof anchor points for the hoist withoutdamaging the integrity of the structure.

Ground Penetrating RadarImagingMeasurements were made with a readilyavailable, high frequency (1 GHz)commercial GPR concrete scanning

Ground Penetrating Radar Imaging ofConcrete at a Nuclear Power Plantby Peter Giamou

From NDT Technician, Vol. 14, No. 2, pp: 1–4.Copyright © 2015 The American Society for Nondestructive Testing, Inc.

system. The system, shown in Figure 1,consisted of an antenna transducer, handle,wheel odometer, battery, and digital videologger. The video logger provided a harddisk for data storage and a display screenfor viewing data during acquisition. Theunit was self-contained in a portable,wheeled carrying case that doubled as adisplay stand on site. GPR uses echosounding principles. The

radar system produces a short durationpulse of radio wave energy that istransmitted into the concrete. Changes inmaterial composition (which can changethe electrical character) cause some of theenergy to be reflected back. The reflectedsignals are detected and amplified at thereceiving antenna and stored on the datalogger. Objects (reinforcement bars,post-tensioned cables, metallic/non-metallicpipes, and conduits) and other features(voids, honeycombing, delamination,cracks) embedded in the concrete can bedetected because these objects will havemarkedly different electrical properties thanthe host concrete.

In the last few years, GPR imaging hasbecome widely used. Data, carefullyacquired over a grid area, are processed toproduce depth slice maps of the subsurface.The resulting images are similar to X-rayimages, but with lower resolution.

Survey ProceduresThe work site was inside a five story, cast inplace, concrete reinforced structureadjacent to one of the nuclear reactors.This building houses generators, supportequipment, and supplies for the operationof the plant. The area of interest was on thefifth (top) level, near the edge of amezzanine floor overlooking the centralportion of the building. The floor consistedof a 350 mm (14 in.) thick suspendedconcrete slab with #4 and #5 reinforcingbars placed in a bidirectional pattern at300 mm (12 in.) centers on upper andlower mats. The floor was covered with apolyurethane coating. The work area wasopen, very smooth, flat, well lit, and clean,but noisy due to all the operating machinery.The area of interest was intentionally

placed over intersecting beams so that thebulk of the load could be directlytransferred to the beams. A total of six scangrids were collected, each 1.2 1.2 m

(4 4 ft), covering a total area of 2.4 3.6 m (8 12 ft; Figure 2). Theinitial equipment setup and calibration,and the establishment of the grid layout,took approximately 2 h to complete.The data were acquired in a bidirectional

pattern at 100 mm (4 in.) line spacingintervals along a total of 39 lines in onedirection and 26 lines in the orthogonaldirection. Several velocity calibrations werecompleted at various locations on the slabto provide an accurate estimate of the radarwave velocity within the slab—a necessarystep to properly process and generate adepth scale for the data. The radar wavevelocity in the concrete at this location ofthe building was 115 mm/ns (4.5 in./ns),which is considered moderately fast formost concrete types. Data acquisition forall six grids took approximately 1 h tocomplete.

Data Processing andInterpretationThe six grids of data were processed toproduce depth slices and were interpretedindividually on site. Small tick marks weremade on the floor on all four sides of eachgrid, based on measurements read from theGPR screen view. The tick marks weremade with pencil, since it was requested by the site engineer that no permanentmarkings be left. Site contacts were warnedto protect the work area from any water,cleaning, or traffic, in order to preserve themarkings until construction was complete.The ticks were connected using duct tapeto create a final interpreted plan view of theembedded features within the concretefloor. This process took approximately 3 hto complete.A line-by-line review of all the individual

radar profiles was carried out to confirmthe interpretation. This should be astandard quality assurance procedure withany concrete scanning project. There aresome situations where the eye of anexperienced user can pick out an unusualfeature that the processed plan map depth

FOCUS | GPR Imaging of Concrete

2 · Vol. 14, No. 2

Figure 1. Concrete imaging system inoperation.

Figure 2. Plan map of survey grid.

3 6

2 5

1 4

1.2 m

1.2 m

slices do not easily reveal. Reading radarprofiles is somewhat different thaninterpreting depth slices. It requires a greaterunderstanding of how GPR systems work,how electromagnetic signals propagate in amedium, and the characteristic signatures ofreflections from various features.Nonetheless, any technician can be taughtthese tricks, enabling a one-person crew tocomplete radar scanning surveys on mostsites.On site, it was possible to see the slab

bottom reflection and beam boundaries inthe GPR profiles. The operator sees eachradar section as it is acquired, providinggood quality control. Figure 3 shows asample radar cross-section from grid 3 (asshown on the map in Figure 2). The sectionshows two upper reinforcement bar mats,one bottom reinforcement bar mat (on theright side only), the slab bottom (on theright side only), the beam edge, and thesignal response from a stirrup parallel to theprofile line direction.The presence of the beams made the

interpretation more challenging because theamount of reinforcement bar loading in the

beams was significantly greater than in the other locations. The thicker concrete,tighter bar spacing, and additional mats of reinforcement limited signals frompenetrating deeper than the first layer ofreinforcement bar in the beam locations.Areas not underlain by a beam exhibitedboth a top and bottom set of reinforcingbars at fairly consistent spacing. Theprocessed radar data for the six grids areshown, pieced together, in Figure 4a, for adepth slice from 75 to 100 mm (3 to 4 in.)below the floor surface. The matching fromgrid to grid indicates accurate gridregistration and careful data acquisition.This composite image of all six gridstogether made interpreting the location of the intersecting beams quite easy. Theadditional reinforcement bars (inferred to be stirrup cages) stand out in the image. A photo of a typical cage is shown in Figure 4b for reference.More detailed views of the top-right grid

are shown in Figure 5. The GPR imageshown in Figure 5a shows a beam on the left half of the image. The stirrups can becounted and the unevenness of their

TNT · April 2015 · 3

Figure 3. Sample cross-section showing typical ground penetrating radar responses.

4 · Vol. 14, No. 2

placement can be measured. The deepersection (Figure 5b) indicates a diagonalfeature within the slab that carried current(as measured by a power line signaldetection device) and was interpreted to bean electrical cable conduit.

Summary and ConclusionGPR was successfully used in this projectwith minimal intrusion and disruption tothe operation of the nuclear plant. Theentire job was completed in one day, withthe results produced on site. Given the sizeof the area that needed to be scanned andthe thickness of the concrete, GPR provedto be the best technology available to meetthe stringent requirements of working at anuclear plant.The current case history demonstrates

the utility of GPR for rapid noninvasive

testing of a concrete structure. The wholeproject was completed on site and satisfiedthe project engineer’s needs. Some keybenefits of using GPR at this site were:� it presented no safety hazards that wouldrequire the work area to be cleared ofpersonnel;

� the ability to provide rapid measurementswith immediate results;

� the ease of adapting to site conditions;� that access needed to be only from oneside of the floor;

� it gave an accurate estimate of structuredepth;

� it provided results that were readily under-standable to customers;

� it recorded digital results for futurereference;

� it caused no disruption or damage to thestructure.GPR is regularly used on many

engineering and construction projects (onfloors, walls, suspended and slab-on-gradestructures, bridge decks, columns, and soon). With growing industrial experienceand understanding of constructionpractices, GPR has become an essentialcomponent of building forensics services.Recent developments in GPR technology

since the initial publication of this articleinclude advances in data processing speedand 3D file generation. Hardwareimprovements include a reduction in thesize and weight of the equipment allowingfor tighter access and use in more difficultsettings (walls, soffits) as well as remoteoperation and automatic triggering forhigh-precision positioning. h

AUTHORPeter Giamou: Golder Associates, Ltd., 6925Century Ave., Mississauga, ON L5N 7K2,Canada; (905) 567-6100; fax (905) 567-6561;e-mail pgiamou@golder.com.

ACKNOWLEDGMENTThis is an updated version of a Back to Basicspaper that originally appeared in MaterialsEvaluation, Vol. 64, No. 12, 2006, pp. 1143–1145.

FOCUS | GPR Imaging of Concrete

Figure 4. Radar and visible images: (a) composite plan map formed from the six grids representing a depth slicebetween 75 and 100 mm (3 and 4 in.), with the beam cages crossing through the center of the area; (b) example of aprefabricated stirrup cage used in concretebeam construction.

(a)

(b)

Figure 5. Grid 6: (a) display presenting the details of the beam structure at 75 to100 mm (3 to 4 in.); (b) detailed viewshowing a weak diagonally-trendingfeature between 300 and 325 mm (12 and 13 in.) depth, later determined to be an embedded electrical conduit.

(a)

(b)