Use of Parallel-Seismic and Induction-Logging Tests for Foundation Depth Evaluation Under Difficult Conditions , a Root-Pile Foundation Embedded in Rock

A case study on the application of the parallel-seismic (PS) and induction-logging (IL) tests combined to evaluate the depth of a 25-cm diameter root-pile belonging to the foundations of a telecommunication tower located in Santana de Parnaíba, São Paulo, Brazil, is presented. The foundation element is embedded in a friable and altered meta-sandstone rock. The relatively small diameter of the pile and the presence of rock as surrounding material tend to limit the interpretation of the PS test. Nevertheless, the technique applied in this study showed satisfactory results. On the other hand, the existence of electromagnetic interference from the antenna itself limited the interpretation of the IL test in this case. The estimated depth of the foundation element was evaluated from the PS test as ranging from 6.5 to 7.5 m, which was later verified to be in agreement with the pile as-built documentation. When the possibility to drill boreholes at a site under investigation exists, the methods presented in this paper may become good options for the determination of the unknown depth of foundation elements in altered rock.


Introduction
The need for determining the unknown depth of existing foundation elements arises whenever a project requires the use of existing foundations under new loadings and design documentation is scarce or unavailable.The most common situation occurs when the retrofit and new use of a structure causes changes in the loads acting on the foundation, so that verifications need to be carried out.The stratigraphic profile, foundation type, dimensions and constructive method, as well as the depth of each of the foundation elements, are required as input information for bearingcapacity and settlements analyses traditional in foundation engineering practice.
With respect to foundations of telecommunication towers, such a situation occurs often as, due to cost optimization, additional or heavier and more robust antenna elements are frequently placed and installed on existing towers.Also, in this type of structures, some of the foundation elements are commonly subjected not to compression but to tension, as a result of the horizontal forces arising from wind action on the tower structure.
When foundations are shallow, direct methods for unveiling the foundations geometry and depth are likely viable.However, in the case of deep foundations, the use of an indirect method is required.To achieve this goal, various destructive and nondestructive tests have been developed.
Nondestructive tests based on geophysics methods have a high application potential, are damage-free, costeffective and time-saving.Thus, they can be extended to a large population of tested elements reducing uncertainty at a given site.Most of the tests can be classified as either reflection or direct-transmission methods.Reflection methods are in general faster and more cost-effective, whereas direct-transmission methods allow better results for deeper elements.
An example of a non-destructive reflection method, the low deformation integrity testing methodology, commonly known as Pile Integrity Testing, or PIT (also known as sonic echo / impulse response test), has recently been used in Brazil for evaluating the depth of existing piles and caissons belonging to the foundations of telecommunication towers (Souza et al., 2015).In the low deformation integrity testing, an impulse hammer generates a compressional wave (P-wave) which travels down the element until a change in acoustic impedance is encountered causing the wave to be reflected back and detected by a receiver placed next to the impact point, where the signal is recorded.The reflection of the wave occurs at the bottom of the tested element or at a depth corresponding to a discontinuity, such as a crack along the shaft.Based on the returning wave signal, the element depth, as well as the presence and location of a crack, may be determined since the velocity of an elastic wave travelling along a pile can be estimated.However, in the case of deep foundations partly embedded in rock, the low deformation integrity test has been reported to become erratic and difficult to interpret, restricting the applicability of this method (Cunha & Costa, 1998;Cunha et al., 2002;Foá et al., 2000).
The non-destructive direct-transmission geophysical methods rely on the use of one or more external boreholes drilled in the vicinity of the foundation element to be tested, and are, in general, capable of providing superior results in terms of the determination of the depth of longer foundation elements.Examples of tests include the boreholebased ground penetration radar, cross-hole sonic logging, the parallel seismic test and the induction logging test.The parallel seismic test has been used for determining the unknown depth of foundation elements for bridges and other structures in the USA for many years (Davis, 1995).On the other hand, the induction logging test has mostly been applied for subsurface mapping at environmentally-contaminated sites.Recently, an application of the PS test in the metropolitan area of São Paulo, Brazil, for evaluating the depth of a 230-cm diameter caisson without an enlarged base in a clayey-sandy silt provided good results when the obtained depth was compared to the depth depicted in the caissons documentation (Gandolfo et al., 2015).
When the pile has a relatively small diameter (such as a root-pile) and the surrounding material is stiffer than soil, such as altered rock, the applicability of the PS test, which depends on the contrast in stiffness between the pile and surrounding material, may be limited.Additionally, when significant electromagnetic interference occurs at the site, the data obtained in the IL test may be of difficult interpretation.
This paper presents the results of an application of the parallel seismic and induction logging tests to determine the depth of a root-pile that is embedded in friable metasandstone.When the foundation element is partly embedded in rock, the methods chosen for this study become more appropriate than the low deformation integrity test, even though these methods are generally more expensive than the non-destructive reflection method.

Theoretical Background
As opposed to surface-based geophysics methods, borehole and logging tests are based on the use of drilled boreholes, in which probes (or sondes) and sensors are lowered to obtain data on geologic materials in depth and identify interfaces of different geophysical signatures (Ellis & Singer, 2008).In general, geophysical profiling is carried out in open wells, where different types of probes can be used, such as electrical, acoustic, or caliper (for measuring well diameter).When the borehole is cased, such as a polyvinyl chloride (PVC) casing, only probes based on electromagnetic induction and on the physical property of natural gamma radiation, or the parallel seismic test can be used.

Parallel seismic testing
The parallel seismic (PS) test is a geophysical technique developed in France several decades ago for the determination of the unknown depth of foundation elements, which has a methodology similar to that of the downhole test.Also, studies have been conducted to develop the PS test for evaluation of pile integrity (Liao et al., 2006;Huang & Chen, 2007;de Groot, 2014).Practical advantages of the PS test are the applicability of the test for different foundation materials (e.g., concrete, steel, wood and masonry) and the possibility of testing even when the pile head is not accessible.
The principle of the PS test is that a pulse is generated by the impact of a small hammer, equipped with a trigger switch, hitting against an exposed part of the structure connected to the foundation, or on the exposed top of the foundation element, if accessible.Elastic waves, predominantly of the compressional type (P-waves), are produced and propagate through the vertical element.If desired, shear waves (S-waves) may also be produced by laterally impacting the opposite sides of the block connected to the foundation.
Due to the great contrast between the elastic modulus of the foundation element (generally consisting of a reinforced-concrete element) and the surrounding soil materials, the P-waves are refracted and are detected by receivers such as three-component geophones or hydrophones placed within an encased borehole near the foundation element.The waves arriving at the geophones are recorded on a seismograph at regular intervals.The working principle of the PS test is illustrated in Fig. 1.
The preparation of the borehole to be used in the PS test should follow the guidance for the crosshole and downhole testings (ASTM D4428 and ASTM D7400).The vertical borehole must be drilled less than 1.5 m from one edge, and be encased with a PVC casing capped at the bottom.The annular space between the casing and the borehole wall must be grouted with cement to ensure good contact with the surrounding soil.In terms of depth, the borehole must be drilled to a depth that exceeds the expected foundation depth by at least 3 to 5 m.When using geophones, the casing must be kept dry, whereas when using hydrophones, the casing must be filled with water prior to testing.Also, another important consideration is to ensure the borehole verticality.
The PS measurements are usually recorded using a sequence of geophones vertically spaced along depth at 0.5-to 1.0-m regular intervals.A seismograph is used to record the signal, or seismic trace, from each geophone position.The time of first arrival of the P-wave at a geophone is obtained as the first break observed in the seismic trace (Fig. 2).Thus, the travel time of the wave from the source (i.e., the point of impact of the hammer) to the corresponding geophone can be calculated.The set of all seismic traces results in a seismogram.Connecting the first-arrival points for all geophones allows the fitting of two straight lines; the point of intersection of the two adjusted lines (the inflection point) corresponds to the interpreted depth of the foundation element.The slope of the top line provides an indication of the wave propagation velocity within the foundation element, whereas the slope of the bottom line indicates the velocity within the surrounding soil.Usually, the top line is steeper than the bottom line, due to the greater wave velocity in the more rigid material that constitutes the foundation element than in the soil.However, when the pile has a relatively small diameter and the surrounding material has a high modulus, the distinct pattern in Fig. 2 is not observed.In this case, the depth of the foundation element may still be determined by observing the depth at which there is a significant drop in signal amplitude (energy) of the traces, and diffraction of the wave energy (both occurring below the bottom of the foundation), as previously reported by Sack & Olson (2010).coil inside a probe is energized with an alternating electric current at the audio frequency, originating a primary variable magnetic field which induces alternate-current flow in the vicinity, i.e., in the surrounding conductive geologic materials, that, in turn, induce an electric potential (and secondary magnetic field) in one or more receptor coils placed also in the probe (Fig. 3).

Induction logging testing
Tests with Electromagnetic Induction measures are widely used to map groundwater contamination plumes, for groundwater exploration, and for general geological mapping.Surface inductive surveys for measuring layer conductivities are used to detect conductive features such as buried metal objects, ore bodies, and fluid-filled fractures, and to map conductive plumes of inorganic chemicals, such as landfill leachate, or saltwater intrusion (McNeill, 1980;Mondelli, 2008).
Induction Logging for measuring conductivity of the formation around the borehole can be used to identify the placement of screening in ground-water monitoring wells, monitor contamination levels outside of cased wells, and detect or monitor contamination plumes in the vadose zone.The use of two or more receiver coils allows the investigation at different radii from the well center.
The IL sonde operates at low values of induction number, as defined by McNeill (1980).Thus, the ratio of the secondary to the primary magnetic fields, or the intensity of the electric potential measured in the receiver coil, is assumed to be linearly proportional to the bulk apparent terrain electrical conductivity, such as (Doll, 1949;Moran & Kunz, 1962): where s a is the bulk apparent subsurface electrical conductivity (which has usually units of mS m -1 ), w is equal to 2pf, where f is the frequency, m 0 is the magnetic permeability in the vacuum, s is the intercoil spacing, and H s and H p are, respectively, the secondary and primary magnetic fields measured at the receiver coil.Equation 1 represents the conductivity in the isotropic, homogeneous semi-space.
The apparent conductivity measured in a well can be found in McNeill (1990).
For some probes, the linear relationship is not valid when the media present high electric resistivity, i.e., greater than 100 ohm-m (Scott et al., 1986), and a non-linearity is observed between the electric potential and the conductivity of the medium.
Besides the electric conductivity, the induction probe is equipped with sensors that can record the logging of natural gamma ray in the surrounding formation, i.e., the concentration in counts per second (CPS) of natural gamma ray emitting radioisotopes from the uranium (U) and thorium (Th) decay series, and potassium (K)-40.These radioisotopes tend to be more abundant in clays as a result of potassium-rich feldspar and mica decomposition, and of uranium and thorium concentrations in the clay due to adsorption and ion exchange.Gamma-emitting radioisotopes of anthropogenic origin cannot be differentiated from naturally occurring isotopes in natural-gamma ray logging.Variations in the gamma log are used to indicate lithologic changes in the formation surrounding a borehole, i.e., the presence of clayey materials (Keys, 1989, Luthi, 2001).
More innovative than the aforementioned applications, is the use of the IL test to determine the depth of a steel or continuously-reinforced concrete foundation element based on the contrast between the magnetic field strength recorded along the element, and that occurring below the tip of the foundation element, i.e., representative of the geologic material.

Site description
The PS and IL tests were performed at a telecommunication-tower site located in Santana de Parnaíba, São Paulo, Brazil (Fig. 4). Figure 4a shows a lateral view of the three-legged tower structure, and Fig. 4b depicts a planview schematic of the location of the tower in reference to the site.
A vertical borehole to be used for the PS and IL tests was drilled at the site at the position marked in Fig. 4b,  1.1 m apart from the foundation block.The borehole was drilled to a depth of 11.0 m, exceeding by 3 to 5 m the expected foundation depth.The borehole was encased with an 85-mm internal-diameter PVC tube, closed at the bottom, and the annular space was totally filled with cement grout.
The stratigraphic profile at the location of the borehole is depicted in Table 1, and includes a superficial 2.0-m   thick fill layer consisting of a brown sandy silt followed by a layer (2.0 to 11.0 m) of an altered meta-sandstone belonging to the São Roque Group, light brown in color, friable (C4), very altered (A4) and fragmented (F5).The local water table was not encountered above the depth of 11.0 m.
Each tower leg was supported by a block with four piles.The piles consisted of so-called root-piles, or micropiles, which were circular in cross-section, with 25-cm nominal diameter, and molded in-situ (Fig. 5).The constructive method involved drilling a cylindrical hole through the fill and meta-sandstone layers using high pressure hydraulics or pneumatics, inserting a reinforcing steel member, and injecting cement grout from the bottom up.Micropiles have mainly been used as foundation elements, as well as for the reinforcement of slopes and existing foundations.The foundation element tested belonged to foundation block "B1", so that the distance between the borehole and the element tested was 1.1 m (Fig. 4b).

Parallel seismic testing
The following equipment was utilized for performing the PS test at the site (Fig. 6): a 12-channel seismograph (SmartSeis, Geometrics), a 1.8-kg sledgehammer with a trigger switch, and a set of 21 triaxial 8-Hz borehole geophones with a pneumatic clamping mechanism.A schematic of a geophone and the clamping mechanism is provided in Fig. 7.
The exposed top of foundation block "B1" was impacted vertically by the hammer, to generate a P-wave (Fig. 8).Also, in order to acquire S-waves, the structure was hit horizontally with the hammer, against two opposite sides of the block, a typical procedure to generate waves with opposite polarities.The seismic traces were recorded at regular 0.5-m intervals, from near the bottom of the borehole, at a depth of 10.5 m, to near the surface level, at a depth of 0.5 m.

Induction logging testing
The equipment utilized for performing the IL test consisted of a dual spaced induction sonde, model DUN 10290, made by Robertson Geologging (UK), connected to a register Micrologger II that was managed by the Winlogger 1.5 software, a mini winch provided with a 150-m long cable, and a tripod with encoder velocity transducer, as illustrated in Fig. 9.
The IL sonde utilized had a length of 225 cm, and contained two pairs of transmitting and receiving coils, the transmitting coil located 47 cm from the tip of the probe, and the receiving coil located 80 cm from the tip.The gamma-radiation sensor was located at a distance of 35 cm measured from the top of the sonde and contained a scintillation meter made of sodium iodide crystal.When the crystal is exposed to gamma rays, photons are emitted, which are amplified and converted into electric pulses for register in the form of pulse counts per second.
The IL measurements could be taken with the sonde moving downwards or upwards along the borehole.Repeatability between measurements taken using both procedures indicates good quality data.In down hole measures, the surface was used as depth reference, whereas in up hole measures the reference was the base of borehole.Both conditions can be accommodated by the Micrologger.However, the alternative modes entail different coverage of the sensors.For the gamma-radiation sensor located 35 cm below the top of the probe, the logs started at an initial depth of 45 cm, since normally the reference level was established when the top of the probe was at ground level.Different coverage also occurs when using the different types of probes (ILM and ILD).Due to the positioning of the sensors, gamma radiation data was collected when the sonde was moving downwards, whereas IL data when moving upwards.This way, each data type was collected under conditions of best possible sensor coverage.

Paralell seismic testing
The parallel seismic test results are presented in Fig. 10 showing the seismogram obtained from the record of one of the horizontal components of triaxial geophones.The criterion generally employed to interpret the seismogram and determine the depth of the foundation element, i.e., the intercept point of the two fitting straight lines (Fig. 2), could not be applied in this case, since the first arrivals beyond the 7.5-m depth were not noticeable.In Fig. 10, the depth corresponding to the end of the foundation element (pile tip) was inferred based on the observation of the signal attenuation, i.e., sharp decline, or drop, in signal amplitudes occurring at the depth of 7.5 m.When there is small stiffness contrast between the foundation element and the rock, this may become the significant criterion for pile depth estimation based on PS testing.On the other hand, Gandolfo et al. (2015) obtained a seismograph similar to that in Fig. 2 for a PS test performed on a 230-cm caisson embedded in a clayey-sandy silt.
Figure 11 presents the seismogram obtained from the record of the S-waves applied to block "B1".The traces depicted show the inversion in signal polarity, corresponding to the seismic signals recorded for the two impacts applied.The seismogram for the S-waves revealed a drop in signal amplitudes occurring at the depth of 6.5 m.
Therefore, the data analyses based on the results obtained from the PS tests performed, using both P-and S-wave seismograms, allowed to estimate the depth of the tip of the foundation element as ranging from 6.5 to 7.5 m.

Induction logging testing
The results obtained with the IL and gamma-radiation sonde are shown in Fig. 12.These results include the stratigraphic description of the layers crossed by the probe (Fig. 12a), the profile of gamma-radiation, in counts per second (Fig. 12b), and the profiles of electric conductivity obtained by induction logging, in mS m -1 (Fig. 12c), which included short and long induction logging results (solid and dashed lines) and the amplitude-envelope profile (thicker solid line).The gamma-radiation value is an indicator of the content of clay minerals rich in potassium feldspars present in the soil or rock.As shown in Fig. 12b, an analysis of the gamma-radiation profile indicated that above the "Z1" level the count numbers were higher than below this level; the "Z1" level indicates the contact between the fill layer and the meta-sandstone.The acronyms in Fig. 12a correspond to "AT" for anthropogenic material (fill), and "MA" for meta-sandstone.The gradual variations in gama-radiation below the "Z1" level suggest texture or clay-content variations in the layer.The variations are interpreted as being inherent to the material and not affected by the foundation element, with no evidence that the longitudinal reinforcement of the foundation element might have affected the gamma-radiation counts, which, in general, are more affected by the materials (soils and rocks) surrounding the borehole.The two peaks in gamma radiation observed between levels "Z1" and "Z2" can be interpreted as indications of the presence of remnant clay deposited during the sedimentation.
As shown in Fig. 12c, the short and long induction-logging profiles displayed an oscillatory pattern.The IL amplitude envelope displayed a decaying pattern following approximately an exponential decay.With depth, the envelope amplitude declined from a maximum value of ~35 mS m -1 to an approximately constant value of 10 mS m -1 below the depth "Z2".The fact that beyond the depth "Z2" the amplitude envelope remained nearly constant and at a lower value may correspond to the background value of electrical conductivity of the geological medium.In this case, "Z2" may correspond to the position of the end of the pile.However, additional data, such as measurements of electrical conductivity of the altered rock, would be required to confirm this interpretation and the length of the pile in the IL test could not be defined with certainty.
Additionally, the exponential-decay pattern with depth observed in the amplitude envelope above the "Z2" level was not expected and is not compatible with the expected electromagnetic induction response in the vicinity of a metallic structure (Telford et al., 1990).Preliminary tests conducted at a different site helped explain the reason for the amplitude envelope to decline exponentially with depth.These tests allowed to identify the interference of electromagnetic waves (plane waves) generated at the surface -probably by the antennas on the telecommunication tower itself -on the results.Given that the electromagnetic field generated by the sonde is not expected to decay with depth, the results in Fig. 12c indicate that the field generated by the sonde may be considered negligible in comparison to the surface field generated by the antennas.If confirmed, this hypothesis indicates that the sonde would have stopped working as a magnetic-field inducer to act only as a receptor.
In order to confirm these hypotheses, it would be necessary to conduct additional investigation, such as carrying out measurements along time with the probe in the air, and then at every meter in depth, and analyzing these measurements for differences in the obtained spectral contents.The only limitation to these measurements is the sampling rate of the sonde, which will not allow identification of high frequencies (MHz), but only frequencies within the kHz range.Another possible test is to utilize a portable spectral measuring device to measure the spectral content (kHz to MHz) from the surface down to the bottom of the borehole.The part of the signal that attenuates with depth should show a shift to lower frequencies, whereas the part that is irradiated by the foundation element should maintain the frequency peak, attenuating only below the base of the foundation element.

Conclusions
The parallel-seismic (PS) and induction-logging (IL) tests were performed on a 25-cm diameter root-pile embedded in altered meta-sandstone rock.The foundation element was part of the foundations of a telecommunication tower located in Santana de Parnaíba, São Paulo, Brazil.At this site, both tests were performed under challenging conditions.The presence of altered rock surrounding the rela-tively slender root-pile resulted in low contrast in stiffness, and a seismograph that did not exhibit the two-adjustable line pattern.Nevertheless, the seismograph interpretation method of verifying the depth at which significant signal attenuation occurs allowed to estimate the root-pile depth as ranging from 6.5 to 7.5 m, which was later found to be in agreement with the root-pile as-built documentation.
For the induction logging (IL) test, the amplitude envelope profile exhibited a pattern that was not expected, rendering the test inconclusive in terms of an estimate for the pile depth.A possible interpretation in this case was that the electromagnetic noise generated by the antenna interfered with the measured electromagnetic signals in this test.For future work, factors that interfere with good quality sig- nal acquisition will be further assessed for each test, procedures to enhance the consistency of the analyses will be studied and, finally, the consistency of the results under different geological settings and foundation typologies will be evaluated.
Broadly, induction logging (IL) testings are based on Faradays laws of Electromagnetism whereby a transmitter Soils and Rocks, São Paulo, 39(3): 261-272, September-December, 2016.263 Use of Parallel-Seismic and Induction-Logging Tests for Foundation Depth Evaluation Under Difficult Conditions...

Figure 1 -
Figure 1 -Illustration of the working principle of the Parallel Seismic test (Niederleithinger, 2012).

Figure 2 -
Figure 2 -Example of a seismogram for an array of various geophones installed at 0.5-m intervals in a PS test (Gandolfo et al. 2015).

Figure 3 -
Figure 3 -Illustration of the working principle of the Induction Logging (McNeill, 1990).

Figure 4 -
Figure 4 -Santana de Parnaíba site: (a) lateral view of the telecommunication tower at the testing site, and (b) plan-view location schematic.

Figure 5 -
Figure 5 -Foundation block detail: (a) plan view of a block, and (b) elevation of the block depicting two root-piles.

Figure 6 -
Figure 6 -Equipment used for the PS test: (a) seismograph, (b) small sledgehammer with a trigger, and (c) borehole geophone.

Figure 7 -
Figure 7 -Detail schematic showing the assemblage of a geophone and air bladder clamping mechanism.

Figure 8 -
Figure 8 -Proceeding with the PS test at block "B1": (a) placement of the geophones inside the borehole, and (b) generation of a compressive wave.

Figure 9 -
Figure 9 -(a) General layout of the IL test equipment, including the notebook and sonde, (b) pulley and tripod, (c) mini winch and batteries, (d) dual spaced induction sonde, and (e) micrologger.

Figure 10 -
Figure 10 -Parallel seismic test results showing the seismogram for one horizontal component after the application of a P-wave, and visualization of a drop in trace amplitudes occuring at 7.5 m.

Figure 11 -
Figure 11 -Parallel seismic test results showing the seismogram for one horizontal component after the application of two S-waves, and visualization of a drop in trace amplitudes occuring at 6.5 m.

Figure 12 -
Figure 12 -Induction-logging and gamma-natural sonde results: (a) lithological description, (b) gamma-natural profile in counts per second, and (c) induction-logging profiles in mS m -1 .

Table 1 -
Stratigraphic profile at the location of the borehole drilled at the site (SM-02).