LIQUEFACTION HAZARD MAPS OF THE HARBOUR AREA OF GIOIA TAURO (ITALY) BY GEO-STATISTICAL METHODS
Johann Facciorusso and Giovanni Vannucchi Department of Civil Engineering, University of Florence, Via di S. Marta, 3 - 50139 Florence (Italy)
ABSTRACT
This paper reports the results of a case study of liquefaction assessment and mapping carried out in the harbour area of Gioia Tauro, in the southern part of Italy. The area, lying on a flat plain prevalently constituted in the first 20 meters below the ground level by cohesionless loose soils, struck by several historical earthquakes with intensity superior to VIII MCS, was the object of extensive seismological, geological and geotechnical surveys. The cyclic resistance ratio and the liquefaction potential index were estimated from numerous but not equispatially distributed in the area CPT and SPT profiles. Therefore geo-statistical method were applied to draw up reliable liquefaction hazard maps including uncertainties of estimated risk.
1. INTRODUCTION
Liquefaction hazard zonation is a complex procedure based on the results of in situ and laboratory tests.
The final product consists of liquefaction hazard maps drawn with reference to a liquefaction cumulative parameter, liquefaction potential index, which is calculated for each explored vertical profile. The reliability of these maps, generally finalised to urban planning and preservation of existent buildings, depends on many factors: the procedure followed to evaluate the liquefaction potential index from tests results, the density and the quality of experimental data and finally the methods applied to estimate the liquefaction parameter all over the investigated area, where no tests are performed. Interpolation of spatial data is a very well known problem and it can be treated by means of different kinds of approaches:
deterministic, statistical (Fenton and Vanmarcke, 1998) or with more advanced procedures like artificial neural network (Juang GIOIA
Fourth International Conference of Earthquake Engineering and Seismology 12-14 May 2003 Tehran, Islamic Republic of Iran
Figure 1. Geographical position of Gioia Tauro
et al., 2001). Liquefaction risk was recognised as one of the most significant seismic geotechnical hazards for the area of Gioia Tauro (Figure 1) because of the high seismicity, the soil properties, the economic and industrial relevance of this place. For this area the numerical results and the liquefaction risk maps resulting from the deterministic and statistical approaches are reviewed and compared.
2. SEISMOLOGICAL SETTING
Gioia Tauro and the surrounding area were struck in the past by several seismic events with intensity superior to VIII MCS, and several phenomena attributable to liquefaction occurrences were observed, especially in the event of 1783 that completely destroyed the town (Figure 2).
As can be seen from the map of maximum observed macroseismic intensity in Figure 3a (Molin et al., 1996), Gioia Tauro falls within the area of maximum observed value (Imax > X MCS). The map of predicted values (with a return period of 475 years) of Peak Ground Acceleration, PGA, obtained for Italy from extensive seismic hazard analyses (Albarello et al., 1999), indicates for this area a maximum expected PGA of 0.35 ÷ 0.40g (Figure 3b).
Figure 2. Historical liquefaction evidences during the earthquake of 1783 at Oppido (a) and Rosarno (b) (Kozak and Ebel, 1996)
Figure 3. a) Map of maximum observed macro-seismic intensity, Imax (Molin et al., 1996); b) map of
predicted values (with a return period of 475 years) of Peak Ground Acceleration, PGA (Albarello et al., 1999).
On the basis of more thorough study of seismic hazard analyses carried out specifically for this area, a maximum expected event (with a return period of 475 years) of a magnitude of 7.3 and peak ground acceleration (at the ground surface) of 0.45g was estimated.
a) b
a) b)
3. GEOLOGICAL AND GEOTECHNICAL SETTING
From a geological point of view, Gioia Tauro lies on a flat plain the origin of which was a depression, spreading along its length in a N-S direction from the Valley of Mèsima to the Massif of the Aspromonte, and filled by continental sediments of Quaternary Age. This plain is prevalently constituted by granular saturated soils in the surficial layers (up to a depth ranging from 50m to 70m from the ground level) overlying a layer of compacted clays and silty clays of great thickness (500m or more) followed by the bedrock, located at a depth of 500 ÷ 600m.
The harbour area, which is a limited part of the municipality area and of greater interest from an economical and industrial point of view, has been considered in this research (Figure 4).
Extensive geological and geotechnical surveys were performed in the past (CPT and SPT tests, stratigraphic and geotechnical boreholes, laboratory tests, etc.) to characterize this area, which should have hosted at first a thermoelectric plant, never built, then the existent harbour infrastructures.
Soil stratigraphy has been drawn from boreholes results along several sections. Stratigraphy relative to one of these sections, depicted in Figure 7 with CPT Figure 4. Aerial photo of the harbour area of Gioia Tauro
and SPT tests and boreholes locations, is shown in Figure 5.
As shown in Figure 5, the cohesionless deposit is constituted by a thick layer of debris followed in order of superimposition from the higher to the lower by:
− Soil A: coarse to medium loose aeolian sands, having a thickness of about 3 ÷ 5m;
− Soil B: coarse, coarse to medium sands with polygenic gravels or sandy polygenic alluvial gravels, having a thickness of about 10m;
− Soil C: medium, medium to fine dense sands, having a thickness ranging from 30m to 70m, constituted by a sequence of lens and thin layers of sands, gravelly sands and fine silty sands; the top of this layer is at a depth ranging from 7 to 19m from the ground level;
− Soil D: medium to fine, fine dense silty sands and sandy silts of a thickness decreasing from the interior part of the area of interest to the coast, where it is not present at all. The top of this layer is located at a depth of about 30 ÷ 40m from the ground level with a maximum thickness of 70m.
S 218 S 119 S 203 S 209 S 213 S 212 S 244
LEGENDA
Riporto Formazione A Formazione B Formazione C Formazione D Formazione E 6 m
0 m s.l.m.
10 m 20 m 30 m 40 m 50 m 60 m 70 m 80 m 90 m
200 m 0 m
Debris soil Soil A Soil B Soil C Soil D Soil E
S 218 S 119 S 203 S 209 S 213 S 212 S 244
LEGENDA
Riporto Formazione A Formazione B Formazione C Formazione D Formazione E 6 m
0 m s.l.m.
10 m 20 m 30 m 40 m 50 m 60 m 70 m 80 m 90 m
200 m 0 m
Debris soil Soil A Soil B Soil C Soil D Soil E
The water table level is estimated at 2.3m (above sea level), that is the highest one recorded in this area (with daily fluctuations from the minimum and the maximum level of 35 ÷ 40cm).
In the first 20 m of ground below the water table level, where the liquefaction phenomena are possible, just the first three layers (A, B and C) are present.
Comparing the results of stratigraphic bore- holes and CPT tests, it was possible to classify the soil according to the scheme proposed by Robertson(1990) and identify the layers
Figure 5. Stratigraphic section 3
susceptible to liquefaction from a lithological point of view. On the basis of two adimensional factors F and Q (function of tip resistance, qc, and friction ratio, fs) and of the classification index, Ic, soils are grouped in 9 classes.
Liquefaction can only take place for soils falling into classes 4 (silt mixtures; clayey silt to silty clay), 5 (sand mixtures; silty sand to sandy silt), 6 (sands; clean sand to silty sand) and 7 (gravelly sand to sand).
Debris soils fall into class 6 and 7; soil A falls into class 6 and soil C falls into classes 5, 6 and 7. CPT tests gave no results for soil B, because it was not possible to carry out soil penetration, given the high resistance of soil and its lithological composition (prevalently gravel): this kind of soil was thus considered not liquefiable.
4. DATA SET ANALYSIS
Finally, summarising geological, geotechnical, hydrological and seismological features of the site, it is possible to state that the seismic potential of the harbour area of Gioia Tauro, one of the most important trade port junction of Southern Europe, is very high, and then, the liquefaction risk is considered very high and a liquefaction hazard zonation is opportune.
The results of 54 geotechnical boreholes (with maximum investigated depth between 16.95m and 91.30m), of the laboratory tests on the samples extracted from the boreholes, of 115 profiles of mechanical CPT tests and of 121 profiles of SPT tests, were utilized to carry out the liquefaction hazard zonation.
5. EVALUATION OF SAFETY FACTOR AGAINST LIQUEFACTION (FSL) AND OF LIQUEFACTION POTENTIAL INDEX (PL) FROM IN SITU TESTS
Risk of liquefaction is expressed in terms of safety factor, that is the ratio between the liquefaction resistance of soil (Cyclic Resistance Ratio, CRR) and the earthquake-induced loading (Cyclic Stress Ratio, CSR): FSL = CRR/CSR.
0
GWL
CSR is evaluated by the formula of Seed and Idriss (1971):
' d 0 v
0 v max '0
v
cyc r
g 65 a . 0
CSR= = ⋅ ⋅
σ σ σ
τ
where amax = 0.45g, is the maximum expected acceleration at the ground surface, σv0 is the vertical total stress at the considered depth and rd is the reduction factor to consider the maximum shear stress decrease (Liao and Whitman, 1986).
The cyclic resistance ratio CRR with reference to earthquakes of 7.5 magnitude (CRR7.5) was calculated from CPT tests (Robertson and Wride, 1998) and from SPT tests (Seed et al., 1985).
The Robertson and Wride method also allows exclusion from the liquefaction risk analysis all the layers not susceptible to liquefaction from a lithological point of view (with a classification index greater than 2.6).
The magnitude scaling factor, MSF, to consider the resistance of soil corresponding to a magnitude M, different from 7.5, was determined by empirical relationships of Idriss (1990) and Andrus and Stokoe (1997).
For every CPT and SPT profile and at each depth below the water level and susceptible to liquefaction from a lithological point of view, where the in situ parameter are measured, the CSR values corresponding to the maximum expected event of magnitude M = 7.3 (with a return period of 475 years), the cyclic resistance ratio, CRR, and the safety factor against liquefaction were so calculated.
The value of FSL indicates whether or not liquefaction is expected to occur (greater or less than one) and the “intensity” of the phenomenon (much lesser than one or close to one) at each layer investigated, but since the effects of liquefaction phenomena at a certain site are the resultant of the contributes of all the underlying layers, it is necessary to define for each explored profile a final cumulative parameter of liquefaction.
The liquefaction potential index, PL (Iwasaki et al., 1978), is the synthetic cumulative parameter adopted to estimate the liquefaction risk.
An example of the evaluation of the safety factor against liquefaction and of the liquefaction potential index from a CPT test profile is shown in Figure 6.
0
5
1 0
1 5
2 0
0 1 0 2 0 3 0 4 0 5 0
qc ( M Pa )
z (m)
0
5
1 0
1 5
2 0
0 0 . 2 5 0 .5 0 .7 5 1 fs ( M Pa )
0
5
1 0
1 5
2 0
0 0 . 1 0 .2 0 . 3 0 . 4 0 .5 0
5
1 0
1 5
2 0
0 0 . 5 1 1 . 5 2
F S L
0
5
1 0
1 5
2 0
0 5 1 0 1 5 2 0
∫ ⋅ ⋅
−
=
z
L F z w z dz
P
20
) ( ) (
Low liquefaction risk High liquefaction risk Very high liquefaction risk
CSR CRR
Figure 6. Evaluation of the safety factor against liquefaction and of the liquefaction potential index from a CPT test profile
6. LIQUEFACTION RISK MAPS: DETERMINISTIC AND GEOSTATISTICAL INTERPOLATION
The liquefaction hazard maps for the harbour area of Gioia Tauro is expressed in terms of liquefaction potential index, represented by means of equipotential lines. In particular the equipotential lines delimiting the areas in which the liquefaction risk is very low (PL < 1), low (PL < 5), high (PL < 15) and very high (PL > 15), are pointed out.
The equipotential lines are obtained from the quoted plan in points for which the PL value is calculated by the aforesaid procedure, namely in points corresponding to the CPT and SPT profiles. Nevertheless it is necessary to consider that:
1. the calculated values of PL are affected by different sources of errors (methods for estimating CSR and CRR, accuracy of the measures and of the tests position, etc..), somewhere the values vary abruptly , and that
2. the locations of the in situ tests are not equispatially distributed in the area, and the reability of the interpolation depends on the density, in addition to the quality of data.
Therefore a comparison of results obtained from different procedures is useful.
The first procedure was the inverse distance deterministic method, by which the liquefaction hazard map shown in Figure 7 was obtained. In this figure the zones corresponding to different risk level and the localization of the in situ tests profiles are represented.
Figure 7. Liquefaction hazard map from deterministic method
In order to improve the resolution of the hazard maps, to estimate rationally and systematically liquefaction potential index at points where it is not calculated, considering the whole data set and to have an estimate of the errors and consequently of the reliability of PL value, the geo-statistical kriging
Legenda
Prove CPT Prove SPT
Linee equipotenziali
Rischio di liquefazione molto basso (0 < P <1)L
Rischio di liquefazione basso (1 < P <5)
Rischio di liquefazione alt o (5 < P <15)
Rischio di liquefazione molto alto (P >15 )
L
L
0 500 m L
M A R E T IR R E N O
PO R T O C A N A L E
Rischio di liquefazione molto basso (0 < P <1)L
Rischio di liquefazione basso (1 < P <5)
Rischio di liquefazione alt o (5 < P <15)
Rischio di liquefazione molto alto (P >15 )
L
L
L
Very low liquefaction risk (0 < PL < 1)
Very low liquefaction risk (1 < PL < 5)
High liquefaction risk (5 < PL < 15)
Very high liquefaction risk (PL> 15)
CPT TESTS SPT TESTS
Section 3 (Figure 5)
Legenda
Prove CPT Prove SPT
Linee equipotenziali
Rischio di liquefazione molto basso (0 < P <1)L
Rischio di liquefazione basso (1 < P <5)
Rischio di liquefazione alt o (5 < P <15)
L
L
Rischio di liquefazione molto basso (0 < P <1)L
Rischio di liquefazione basso (1 < P <5)
Rischio di liquefazione alt o (5 < P <15)
Rischio di liquefazione molto alto (P >15 )
L
L
L
Very low liquefaction risk (0 < PL< 1)
Low liquefaction risk (1 < PL< 5) High liquefaction risk (5 < PL< 15)
Very high liquefaction risk (PL> 15)
CPT TESTS SPT TESTS
method (Chiasson et al., 1995) was also used. The hazard maps of the predicted values of liquefaction potential and of the uncertainties are shown in Figures 9 and 10.
From the results obtained it is possible to state that:
− kriging interpolation method allows a more conservative map to be obtained, where high and very high risk zones are wider than those obtained by inverse distance weighted method;
− the greater part of the area of interest is characterized by a very low or low liquefaction risk;
− the zone of highest liquefaction risk is located where the port canal flows into the sea and in the northeastern part of the harbour area;
− the higher the density of in situ tests (and thus of the calculated values of PL), the higher the reliability of the model and the lower the uncertainties of the estimated values.
Figure 8. Liquefaction hazard map from kriging method
Legenda
Prove CPT Prove SPT
Linee equipotenziali
Rischio di liquefazione mo lto basso (0 < P <1)L
Rischio di liquefazione basso (1 < P <5)
Rischio di liquefazione alto (5 < P <15)
Rischio di liquefazione mo lto alto (P >15)
L
L
0 500 m L
M A R E T
P O R T O C A N A L E M A R E T IR R E N O
PO R T O C A N A L E
Rischio di liquefazione molto basso (0 < P <1)L
Rischio di liquefazione basso (1 < P <5)
Rischio di liquefazione alt o (5 < P <15)
Rischio di liquefazione molto alto (P >15 )
L
L
L
Very low liquefaction risk (0 < PL < 1)
Very low liquefaction risk (1 < PL < 5)
High liquefaction risk (5 < PL < 15)
Very high liquefaction risk (PL> 15)
CPT TESTS SPT TESTS
Legenda
Prove CPT Prove SPT
Linee equipotenziali
Ri schio di liquefazione molto basso (0 < P <1)L
Ri schio di liquefazione basso (1 < P <5)
Ri schio di liquefazione alt o (5 < P <15)
L
L
Rischio di liquefazione molto basso (0 < P <1)L
Rischio di liquefazione basso (1 < P <5)
Rischio di liquefazione alt o (5 < P <15)
Rischio di liquefazione molto alto (P >15 )
L
L
L
Very low liquefaction risk (0 < PL< 1)
Low liquefaction risk (1 < PL< 5) High liquefaction risk (5 < PL < 15)
Very high liquefaction risk (PL> 15)
CPT TESTS SPT TESTS
Legenda
Prove CPT Prove SPT 3.3 < E *< 3.6std
3.6 < E < 3.8
3.8 < E < 4
4.0 < E < 4.2
std
std
std
MARE TIRRENO
PORTO CANALE
CPT TESTS SPT TESTS 0.73 < ESTD <
0.79 < ESTD <
0.84 < ESTD <
ESTD > 0.88
Legenda
Prove CPT Prove SPT 3.3 < E *< 3.6std
3.6 < E < 3.8
3.8 < E < 4
4.0 < E < 4.2
std
std
std
CPT TESTS SPT TESTS 0.73 < ESTD < 0.79 0.79 < ESTD < 0.84
0.84 < ESTD < 0.88 ESTD > 0.88
Very high estimate reliability
High estimate reliability
Low estimate reliability Very low estimate reliability
Figure 9. Map of uncertainties expressed in terms of mean standardized errors, obtained by multiplying the mean predicted error by the estimated kriging standard errors
7. ACKNOWLEDGMENTS
This research has been carried out within the framework of a project financed by the MURST (Ministry of University and Scientific and Technologic Research) with the aim of performing geotechnical and seismological analyses in the Messina Strait area.
8. REFERENCES
1. Albarello, D., Bosi, V., Bramerini, F., Lucantoni, A., Naso, G., Peruzza, L., Rebez, A., Sabetta, F., Slejko, D. (1999) - New seismic hazard maps of the italian territory. Servizio Sismico Nazionale – Gruppo Nazionale per la Difesa dai Terremoti.
http://www.serviziosismico.it/PROG/2000/carte_pericolosita/index.html
2. Andrus, R.D. Stokoe II, K.H. (1997) - Liquefaction resistance based on shear wave velocity. Proc. of NCEER Workshop on Evaluation of Liquefaction Resistance of Soils, Nat. Ctr. for Earthquake Engineering Res., State University of New York at Buffalo, USA, pp. 89-128.
3. Chiasson, P., Lafleur, J., Soulié, M., Law, K. T. (1995) - Characterising spatial variability of a clay by geostatistics. Canadian Geotechnical Journal, Ottawa, Canada, 32, pp. 1-10.
4. Fenton, G.A., Vanmarcke, E.H. (1998) - Spatial variation in liquefaction risk. Geotechnique, 48(6), pp. 819-831.
5. Idriss, I.M. (1990) - Response of soft soil sites during earthquakes. Proc. of H. Bolton Seed Memorial Symposium, Vol. 2 , BiTech Publishers, Ltd., Vancouver, Canada, pp. 273-290.
6. Juang, C.H., Jiang, T., Cristopher, R.A. (2001) - Three dimensional site characterisation: neural network approach. Geotechnique, 51(9), pp. 799-809
7. Kozak, J., Ebel, J.E. (1996) - Macroseismic Information from Historic Pictorial Sources. Pure and Applied Geophysics, vol. 146, no.1, 1996. http://nisee.berkeley.edu/kozak/
8. Liao, S.S.C., Whitman, R.V. (1986) - Catalogue of liquefaction and non-liquefaction occurrences during earthquakes. Res. Rep. Department of Civil Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA
9. D. Molin, M. Stucchi, G. Valensise (1996) - Massime intensità macrosismiche osservate nei comuni italiani. Dipartimento della Protezione Civile.
http://emidius.mi.ingv.it/GNDT/IMAX/max_int_oss.html
10. Robertson, P.K. (1990) - Soil classification using CPT. Canadian Geotechnical Journal, Ottawa, Canada, 27(1), pp. 151-158.
11. Robertson, P.K., Wride, C.E. (1998) - Evaluating cyclic liquefaction potential using the cone penetration test. Canadian Geotechnical Journal, Ottawa, Canada, 35, pp. 442-459.
12. Seed, H.B., Idriss, I.M. (1971) - Simplified procedure for evaluating soil liquefaction potential.
Journal of the Soil Mechanics and Foundation Division, ASCE, Vol. 107, No SM9, pp. 1249-1274.
13. Seed, H.B., Tokimatsu, K., Harder, L. F., Chung, R. M. (1985) - The influence of SPT procedures in soil liquefaction resistance evaluations. Journal of Geotechnical Engineering, ASCE, 111(12), pp.
1425-1445.
14. Iwasaki, T., Tatsuoka, F., Tokida, K., Yasuda, S. (1978) – A practical method for assessing soil liquefaction potential based on case studies at various sites in Japan. Proc 2nd Int. Conf. On Microzonation for Safe Construction. Research and Application, san Francisco, California, vol. 2, 885- 896.