i i

main 2015/5/11 11:50 page 673 #1

i i

Revista Brasileira de Geofsica (2014) 32(4): 673-679

2014 Sociedade Brasileira de Geofsica
ISSN 0102-261X
www.scielo.br/rbg

COMPARISON BETWEEN RESULTS OF SEISMIC REFRACTION

AND STANDARD PENETRATION TEST (SPT) TO STUDY SHALLOW GEOLOGICAL
SUBSURFACE IN AN URBAN AREA OF BRASILIA, BRAZIL

Pedro Vencovsky Nogueira, Marcelo Peres Rocha, Welitom Rodrigues Borges,

Eduardo Xavier Seimetz and Marcio Maciel Cavalcanti

ABSTRACT. The most common procedure for an engineering project/construction is the use of direct survey, borehole and Standard Penetration Test (SPT). This
provides punctual information of the geology at the site, and many boreholes are necessary along the construction site, representing a significant amount of the budget
for the construction and to help develop a better geological understand/map of the site. The use of geophysical methods allows to study the subsurface by indirect
means, with low cost, and enable to cover large areas if compared to direct surveys. Geophysical methods are increasingly being used in engineering works, however, in
Brazil the use in engineering projects is still scarce. In this work was used shallow seismic refraction method to study the shallow subsurface in an area along the future
track of the subway system of Braslia, Brazil. The refraction results (P-wave) were compared with previous existing data from Standard Penetration Test (SPT), and soil
profile description. The seismic was used to study the subsurface geology, and SPT data were used to compare the seismic results. We observed a good correlation
for the depths obtained through each method, mostly in the north portion of the line, when the SPT was near the line, indicating that its results are influenced by the
same mechanical parameters, related to soil strength. Our results motivate the use of seismic refraction as a tool to optimize the direct investigation methods for better
geotechnical characterization of the medium.

Keywords: shallow seismic refraction, standard penetration test (SPT), geotechnical study.

RESUMO. O procedimento inicial mais comum em um projeto de engenharia e o uso de pesquisa direta, por meio de sondagens e Indice de Resistencia a` Pene-
tracao (SPT, em ingles). Estas ferramentas fornecem informaco es pontuais acerca da geologia local, sendo necessarias diversas sondagens para desenvolver um bom
entendimento geologico/geotecnico da regiao, fazendo com que as sondagens representem uma quantidade significativa do orcamento da obra de engenharia. O uso
de metodos geofsicos permite estudar a subsuperfcie por meio indireto, com baixo custo, e possibilita cobrir grandes areas, quando comparado ao uso exclusivo de
sondagens diretas. Metodos geofsicos estao sendo cada vez mais utilizados em obras de engenharia, no entanto, o seu uso em projetos de engenharia no Brasil ainda
e escasso. Neste trabalho foi utilizado o metodo de ssmica de refrac ao rasa para estudar a subsuperfcie em uma area ao longo do futuro trecho do sistema de metro de
Braslia, Brasil. Os resultados de refrac ao (onda P) foram comparados com os dados pre-existentes de SPT e descric a o do solo. A ssmica foi empregada para estudar
a geologia da subsuperfcie, os dados SPT foram utilizados para comparar com os resultados ssmicos. Observou-se uma boa correlacao para as profundezas obtidas
atraves de cada metodo, principalmente na porcao norte da linha, regiao em que o SPT esta mais proximo da linha, indicando que os seus resultados sao influenciados
pelos mesmos parametros mecanicos, relacionados com a resistencia do solo. Nossos resultados motivam o uso de refrac ao ssmica como uma ferramenta para
aperfeicoar os metodos de investigacao direta, com objetivo de gerar uma melhor caracterizac ao geotecnica do meio.

Palavras-chave: ssmica de refracao rasa, ndice de resistencia a` penetracao, estudo geotecnico.

Universidade de Braslia, Campus Universitario Darcy Ribeiro, Institute of Geosciences ICC, 70910-900 Braslia, DF, Brazil.
Phones: +55(61) 8123-0671; +55(61) 9952-6668; +55(61) 9646-9090; +55(61) 9609-8492; +55(61) 8177-7417
E-mails: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]

i i

i i
 i i

main 2015/5/11 11:50 page 674 #2

i i

674 COMPARISON BETWEEN RESULTS OF SEISMIC REFRACTION AND STANDARD PENETRATION TEST (SPT) IN AN URBAN AREA OF BRASILIA, BRAZIL

INTRODUCTION shallow geological structure in the study area and also compare
the geological model, seismic results, with the Standard Penetra-
In the beginning of engineering works the knowledge and the
tion Test (SPT) data.
conditions of the subsurface is crucial because the geological
structures can be a complex factor. A typical approach is to study
METODOLOGY
the subsurface in a direct way, using boreholes and Standard
Penetration Test (SPT) (Hettiarachchi & Brown, 2009). Gener- For the data acquisition of shallow seismic refraction, were used
ally, direct survey provides high quality information, however, 48 receivers spaced 2 meters apart (Fig. 2). The seismic source
represent punctual information. Several surveys are necessary used was a hammer with 8 kg, struck 15 times against a metal
to have a greater understand of the geology, increasing the cost plate placed on the ground. A total of six seismic sections were
of the work. Direct surveys are invasive techniques, where the acquired, all with 94 meters in length (Fig. 3), composing a seis-
study area can be permanently affected, and at some cases, limit mic line with 564 meters (Figs. 1 and 3). The positions of the
the application at urban areas (McDowell et al., 2002). Several source for each line were at: 2, 47 and 94 meters, from the
methods, such as resistivity, ground penetrating radar (GPR) and first geophone. The data was acquired by using the Geode (Geo-
seismic can be used combined with direct geotechnical surveys metrics) seismograph and for the data processing was used the
to reduce its quantity and cost (McDowell et al., 2002; Consenza program package SEISIMAGER 2D. The technique used to pro-
et al., 2006). cess the data was the time-term inversion.
The main characteristic of geophysical methods is to provide The SPT is a classic geotechnical method used to measure
information by indirect means, covering both large and small mechanical properties of the medium. A sample tube is driven
areas, and in different scales. The cost of geophysical surveys into the ground using a standard weight, which is dropped freely
is considerably smaller when compared to the exclusive use of from a standard height. According to international standards
direct survey, usually spending less time to be done. (ASTM, 2008), the number of blows required for this tube to
The SPT is a common method used in engineering projects penetrate 15 cm into the soil is recorded and resulting a graph
to define soil strength. The number of blows needed per depth for of blows quantity with respect to depth. This graph can be re-
the STP can be related to vertical resistance to ground penetra- lated to variations in the strength of the material in the subsurface,
tion (Yagiz, 2008; Alves, 2009). Several authors use both seismic since the blow counts are related to density of the ground (Brown
refraction and SPT to define soil conditions such as, type of soil, & Hettiarachchi, 2008; Murley & Hettiarachchi, 2011). For the
rippability and liquefaction (Basarir & Karpuz, 2004; Basarir et region of Braslia, when the SPT blow count is higher than 50,
al., 2008). Generally, SPT data is compared with shear wave (Vs ) the material is considered impenetrable. Meaning that, from this
measurements, providing better results, since shear wave prop- depth on, the soil provides good ground stability for engineer-
agation is not affected by the water content of the medium ing purposes (Alves, 2009). This was the definition used in this
(Hasancebi & Ulusay, 2007; Anbazhagan & Sitharam, 2008). work to underline the depth of the impenetrable. We selected the
However, in this work, due to equipment restrictions, the SPT data eight SPT probing most close to the seismic line. These eight SPT
was compared with primary wave (P-wave). data where separated in three groups, according to soil descrip-
The seismic refraction is one of the main geophysical tech- tion and depth of the impenetrable level. Group 1 includes SPTs
niques used in geotechnical problems (e.g. Khalil & Hanafy, numbers 935, 938 and 940 (yellow circles in Fig. 1), Group 2
2008). Among the most common applications of seismic refrac- contains SPT numbers 952 and 953 (orange circle in Fig. 1).
tion in a geotechnical context is to measure the depth of the Group 3 includes SPTs 985, 986 and 993 (Red circles in Fig. 1).
bedrock, for the construction of large and tall buildings, dams The SPT data used in this work were acquired by engineering
and highways (Telford et al., 1990). There are some examples of companies with the supervision of the Department of Civil En-
the use of seismic refraction in Brazil (Prado, 2000; Martnez & gineering (ENC) of the Universidade de Braslia (UnB).
Mendoza, 2011). For more information about application of geo-
physical methods in Braslia, see Seimetz (2012). RESULTS
In this work, we used shallow seismic refraction to study the Figure 2 shows an example of a seismogram obtained by the
subsurface for a future subway station in Braslia, Brazil (Fig. 1). method of shallow seismic refraction. In the y-axis is the time,
The main objective was to generate a geological model based and the x-axis represents the geophone offsets. It is observed
on shallow seismic refraction results, in order to understand the that for the geophones more distant from the shot point (where

Revista Brasileira de Geofsica, Vol. 32(4), 2014

i i

i i
 i i

main 2015/5/11 11:50 page 675 #3

i i

NOGUEIRA PV, ROCHA MP, BORGES WR, SEIMETZ EX & CAVALCANTI MM 675

Figure 1 Study area on the Asa Norte region of Braslia, between the blocks 112 and 113. The red line is the seismic profile. The SPT data (color circles) were
separated by groups (different colors) related to their distance from the seismic line, depth of the impenetrable layer (from SPT results) and soil description.

Figure 2 Example of seismograms acquired in this study. In this case, the source position was 2 meters related to the first geophone. The purple line represents
the marking of first breaks.

Brazilian Journal of Geophysics, Vol. 32(4), 2014

i i

i i
 i i

main 2015/5/11 11:50 page 676 #4

i i

676 COMPARISON BETWEEN RESULTS OF SEISMIC REFRACTION AND STANDARD PENETRATION TEST (SPT) IN AN URBAN AREA OF BRASILIA, BRAZIL

Table 1 Summary of information obtained with the methods of seismicrefraction and SPT.

SPT Position on Distance from Thickness Thickness

number seismic line (m) seismic line (m) seismic (m) SPT (m)
935 17 105 3,5 4
938 47 106 4 6
940 66 112 5 6
952 155 106 7,5 9
953 179 106 7,5 8
986 541 267 11 22
985 543 253 11 23
993 557 229 10 21

the arrival of the wave occurs later) the signal/noise ratio is lower. interface and the velocity of the second layer. The velocity of the
This is probably due to the decrease of the seismic signal en- first layer was not considered as a factor for this division because
ergy far from the source, which makes the record more suscepti- can be highly affected by engineering works, as adding embank-
ble to the effects of pedestrian circulation nearby the geophones, ment or removing original soil.
cars passing on roads nearby and the influence of trees when the In all of the three regions, we assume that the depth of the
weather started to get strong winds at the study site. However, it impenetrable SPT level is related to the interface between the
was possible to identify the first arrivals at all geophones. two layers in the seismic model, since both methods rely on soil
The comparison between the results obtained from seismic strength (Basarir & Karpuz, 2004; Basarir et al., 2008). The com-
refraction and SPT are shown in Figure 3. A summary of informa- parison between the seismic section and SPT data from group 3
tion obtained from these results is presented in Table 1. is more susceptible to mistakes because of the distance be-
Six velocity models (one for each sub-section) were generated tween them (260 m).
for the study area (Fig. 3B), where two layers were observed with In the North region, the soil has a first layer of about 2,5 to
different velocities, according to the time-distance curves gener- 3 m of embankment, followed by a layer of clays until the depth of
ated from the picking of the first arrivals (Fig. 3C). The velocities about 10 m. There is no presence of siltstone. The SPT impene-
obtained for the first layer varies from 402 to 540 m/s, and the trable level depth is shallow, about 4 to 6 m, suggesting that the
velocity for the second layer varies from 1519 to 1791 m/s. clay is responsible for the impenetrable level. Section 1 of seis-
SPT data and soil description were compared to parts of the mic, showed that the velocity for the first layer (433 m/s) is related
seismic line (Fig. 3A). The SPT graph is represented in yellow to embankment, and the velocity of the second layer (1519 m/s)
lines, where the abscissa represents the number of blows (from is related to clay. The second layer has the lowest velocity in all
0 to 60 blows) and the ordinate axis represents the depth. The of the six seismic sections. The depth of the seismic interface
soil description is represented by the square blocks. Their com- showed good correlation to the depth of the impenetrable level
position varies from embankment, originated from construction of SPT data.
works, clay and siltstone. The soil description of the Center region showed no layer of
embankment, the first layer is associated with clays, with a thick-
DISCUSSION ness of 6 m, followed by a layer of siltstone until a depth of 13 m.
The seismic line was separated in three different regions: North, The SPT impenetrable level was found at a depth of 8 m, within
Center and South regions, according to soil description data, N- the layer of siltstone. In this case, it was not the change of mate-
SPT values, seismic velocities data and the depth of the seismic rial from clay to siltstone that resulted in the impenetrable layer,
refraction interface. The North, Center and South regions are re- it was the strength of the siltstone layer that increased in depth.
lated to SPT groups 1, 2 and 3, respectively (Fig. 1). Each region The velocity of the first layer (402 to 524 m/s, in sections 1 to
was correlated with a different seismic section; the North region 5) is related to clay and siltstone, the velocity of the second layer
includes only section 1. The Center region of the line includes (1620 m/s to 1791 m/s) is related to siltstone. The seismic inter-
sections 2, 3, 4 and 5, and the South region is related with sec- face was found at depth of about 7,5 m, similar to the depth found
tion 6. This division was mostly based on the depth of the seismic with the SPT method.

Revista Brasileira de Geofsica, Vol. 32(4), 2014

i i

i i
 i
i

i
i

Brazilian Journal of Geophysics, Vol. 32(4), 2014

main 2015/5/11 11:50 page 677 #5

Figure 3 Comparison between results of: A) SPT, soil description and seismic refraction. B) The six seismic sections together, with the velocity of each layer from each section in m/sec.
C) Travel-Time curves, obtained through the marking of the first arrivals on seismograms.
NOGUEIRA PV, ROCHA MP, BORGES WR, SEIMETZ EX & CAVALCANTI MM
677

i
i

i
i
 i i

main 2015/5/11 11:50 page 678 #6

i i

678 COMPARISON BETWEEN RESULTS OF SEISMIC REFRACTION AND STANDARD PENETRATION TEST (SPT) IN AN URBAN AREA OF BRASILIA, BRAZIL

The South portion of the line is where the correlations be- REFERENCES
tween seismic and SPT does not match, because of the distance
ALVES PC. 2009. Cartografia Geotecnica para Obras Subterraneas:
from the seismic line (260 m), and different geotechnical context.
Condico es de Construco es de Garagens Subterraneas e Metro no Plano
The soil description shows a layer of embankment of about 5 m, Piloto de Braslia. Master dissertation on Geotechnics, Publicaca o GDM
followed by clay and siltstone. The interface between the clay and 178/09, Departamento de Engenharia Civil e Ambiental, Universidade de
siltstone varies from 14 to 19 m deep. The SPT impenetrable level Braslia, Braslia, DF, Brazil, 168 pp.
was found at a depth of about 22 m, in the siltstone layer. The first
ANBAZHAGAN P & SITHARAM TG. 2008. Mapping of average shear
seismic layer (velocity of 540 m/s) is related to the embankment
wave velocity for Bangalore region: A case study. Journal of Environ-
and clay layers, and the second seismic layer (1580 m/s) is as-
mental and Engineering Geophysics, 13(2): 6984.
sociated to siltstone. The seismic interface seems to relate with
change in soil strength of the siltstone layer. Although there is ASTM. 2008. Standard test method for standard penetration test (SPT)
and split-barrel sampling of soils. Annual book of standards, D 1586-
great difference between the depths found with the SPT (22 m)
08, American Society of Testing and Materials, West Conshohocken, Pa.
and seismic (11 m), both method show increase in the depth
p. 132141.
from north to south.
The analysis of the SPT data and the seismic results has BASARIR H & KARPUZ C. 2004. A rippability classification system for
shown that neither of the methods could be used exclusively to marls in lignite mines. Engineering Geology, 74: 303318.
define layers in terms of their material composition. On most of BASARIR H, KARPUZ C & TUTLUOGLU L. 2008. Specific energy based
the SPT data, the impenetrable layer was not related to direct rippability classification system for coal measure rock. Journal of Ter-
change in composition, but change in soil strength, defined by ramechanics, 45: 5162.
factors such as density, compaction and porosity.
BROWN T & HETTIARACHCHI H. 2008. Estimating shear strength
properties of soils using SPT blow counts: An energy balance approach.
ASCE Geotechnical Special Publication No. 179.
CONCLUSIONS
CONSENZA P, MARMET E, REJIBA F, CUI YJ, TABBAGH A & CHARLERY
For the study area there was a good correlation between the re-
Y. 2006. Correlations between geotechnical and electrical data: A case
sults of seismic refraction and SPT, which showed an increase in
study at Garchy in France. Journal of Applied Geophysics, 60: 165178.
thickness of the shallow layer toward the South, although STP
from group 3 could not be directly correlated. FONSECA AVD, CARVALHO J, FERREIRA C, SANTOS JA, ALMEIDA
F, PEREIRA E, FELICIANO J, GRADE J & OLIVEIRA A. 2006. Charac-
The seismic model showed that the depth of the interface be-
terization of a profile of residual soil from granite combining geologi-
tween the two layers is about 4 m in the northern part of the line,
cal, geophysical and mechanical testing techniques. Geotechnical and
increasing to 11 m in the south. The depth of the seismic model
Geological Engineering, 24(5): 13071348.
is compatible with the SPT data in the center and north portion of
the line. HASANCEBI N & ULUSAY R. 2007. Empirical correlations between
For the study area, the correlation between seismic and SPT shear wave velocity and penetration resistance for ground shaking as-
sessments. Bulletin of Engineering Geology and the Environment, 66(2):
showed that for seismic velocities greater than 1500 m/s, the SPT
203213.
blow count is higher than 50, outlining the impenetrable strata.
As expected, the variations in the velocity of seismic waves HETTIARACHCHI H & BROWN T. 2009. Use of SPT blow counts to esti-
should be related to variations in material resistance as observed mate shear strength properties of soils: Energy balance approach. Jour-
with the data from SPT, since seismic refraction is an efficient nal of Geotechnical and Geoenvironmental Engineering, 135(6): 830
834.
geophysical method to determine soil compactness (Sturaro et al.,
2012). Furthermore, seismic results and SPT do not relate exclu- KHALIL MH & HANAFY SM. 2008. Engineering applications of seismic
sively on lithology, but other factors such as density, compaction refraction method: A field example at Wadi Wardan, Northeast Gulf of
and porosity. Suez, Sinai, Egypt. Journal of Applied Geophysics, 65(34): 132141.
These results provide arguments to increase the joint em- MARTINEZ K & MENDOZA JA. 2011. Urban seismic site investigations
ployment of geophysical and geotechnical methods, following for a new metro in central Copenhagen: Near surface imaging using
the examples of Fonseca et al. (2006) and Sudha et al. (2009), reflection, refraction and VSP methods. Physics and Chemistry of the
specially for bigger engineering projects. Earth, 36(16): 12281236.

Revista Brasileira de Geofsica, Vol. 32(4), 2014

i i

i i
 i i

main 2015/5/11 11:50 page 679 #7

i i

NOGUEIRA PV, ROCHA MP, BORGES WR, SEIMETZ EX & CAVALCANTI MM 679

McDOWELL PW, BARKER RD, BUTCHER AP, CULSHAW MG, JACK- Universidade de Braslia, Braslia, DF, Brazil, 97 pp.
SON PD, McCANN DM, SKIPP BO, MATTHEWS SL & ARTHUR JCR.
STURARO JR, LANDIM PMB, MALAGUTTI FILHO W & DOURADO JC.
2002. Geophysics in Engineering Investigations. Construction Industry
2012. Analysis of soil compactness of the urban area of Bauru/Sao
Research and Information Association Report, C-592. 252 pp.
Paulo state using standard penetration tests and seismic refraction.
MURLEY SC & HETTIARACHCHI H. 2011. Predicting shear strength Geociencias, 31(3): 331338.
properties for low-sensitivity granular-cohesive soils from SPT results.
SUDHA K, ISRAIL M, MITTAL S & RAI J. 2009. Soil characterization
Geotechnical Special Publication, 43234332.
using electrical resistivity tomography and geotechnical investigations.
PRADO RL. 2000. A ssmica de reflexao e o radar de penetraca o no solo Journal of Applied Geophysics, 67(1): 7479.
na investigaca o geologico-geotecnica em ambientes urbanos: Um es-
TELFORD WM, GELDART LP & SHERIFF RE. 1990. Applied Geophysics.
tudo na cidade de Sao Paulo SP, Brasil. Doctorate thesis, Instituto
de Geociencias e Ciencias Exatas, UNESP, Campus de Rio Claro SP, Cambridge University Press, 792 pp.
Brazil, 174 pp. YAGIZ S. 2008. Utilizing rock mass properties for predicting TBM
SEIMETZ EX. 2012. Estudos geofsicos ao longo do futuro trecho norte performance in hard rock condition. Tunnelling and Underground Space
do metro de Braslia. Master dissertation, Instituto de Geociencias, Technology, 23(3): 326339.

Recebido em 22 setembro, 2013 / Aceito em 26 setembro, 2014

Received on September 22, 2013 / Accepted on September 26, 2014

NOTES ABOUT THE AUTHORS

Pedro Vencovsky Nogueira. Geologist graduated from the Universidade de Braslia and Master in Applied Geosciences, area of concentration in Applied Geo-
physics from the same university. Has worked on projects regarding with shallow geology/geophysics, focusing on seismic refraction, resistivity, induced polarization
and ground penetrating radar.

Marcelo Peres Rocha. Graduated in Full Degree in Physics from the Universidade Federal de Mato Grosso (2000), Master of Science (Geophysics) from the
Universidade de Sao Paulo (2003) and Doctor of Science (Geophysics) from the Universidade de Sao Paulo (2008) with 6 months stage during the year 2006 at the
Institute of Earth Sciences Jaume Almera in Barcelona, Spain. Has experience in Geosciences with an emphasis on Geophysics, working mainly on Seismology and
Earthquake. Currently, is an Adjunct Professor at the Universidade de Braslia and teaches classes in undergraduate courses of Geophysics and Geology, as well as
post-graduation at the Institute of Geosciences.

Welitom Rodrigues Borges. Graduated in Geology from the Universidade Federal do Mato Grosso (2000), completed masters degree (2002) and doctorate
(2007) in geophysics from the Universidade de Sao Paulo. Consultant in processing GPR data in agreement between the company and Unicamp Brain Technology
(2004-2006). Geophysical consultant in the company Geopesquisa Geological Investigations (2006-2007). Consultant at GPR company SIGEO Integrated Solutions
in Geotechnology (2005-2007). Professor of Geophysics in the Undergraduate Geology (CCA/UFES 01/2009 the 05/2009). Currently, Professor of Geophysics at
the Institute of Geosciences, the Universidade de Braslia (IG/UnB). Coordinated the team Geology/Geophysics Tocantins Working Group, in search of traces of the
missing event known as the Araguaia guerrilla movement in the years 2009 and 2010. Since 2011 operates in the Brazilian Geophysical Society (SBGf) as Secretary of
the Midwest Regional. Tutor Phygeo (Junior Company of Geophysics) since its inception in 2010.

Eduardo Xavier Seimetz. Graduated in Physics from the Universidade Catolica de Braslia (2006). Has experience in Physics, with emphasis on General Physics.
Completed the graduate program at the Institute of Geosciences in applied geosciences at the Universidade de Braslia UnB. Experience with geophysical methods:
shallow seismic refraction and electrical resistivity.

Marcio Maciel Cavalcanti. Doctorate in Applied Geosciences (UnB/2013), Master in Applied Geosciences (UnB/2013), BS in Environmental Engineering (Universi-
dade Catolica de Braslia/2008). Practice in Applied Geophysics, methods and Electrical Resistivity Ground Penetrating Radar (GPR) and electromagnetic.

Brazilian Journal of Geophysics, Vol. 32(4), 2014

i i

i i