Design scenarios for piles in ground

subject to liquefaction
A.R. Rolfe, S.J. Palmer & E.L. Chin
Tonkin & Taylor Ltd, Wellington.

ABSTRACT
Seismic shaking and liquefaction with its associated loss of soil strength and ground
movements/deformations can be damaging to buildings and their foundations. Understanding
ground behaviours and soil-structure interaction during various stages of strong earthquake shaking
is critical in developing a resilient foundation and building design. Ground behaviour in strong
earthquake shaking cannot be reliably predicted, but we can develop scenarios of possible ground
behaviour and test our foundation design against these. Four scenarios are proposed for
consideration in design; 1) pre-liquefaction representing the start of an earthquake; 2) liquefaction
without lateral ground movement; 3) liquefaction-induced cyclic displacements; 4) liquefaction-
induced lateral spread, representing towards the end of an earthquake or post- earthquake. All or
some of these scenarios would be considered in design depending on the particular site conditions.
The percentage of design building inertial loads (base shear) to be considered in conjunction with
each of these design scenarios depends on the natural periods of shaking of the building and of the
ground, plus other factors. Various percentages of base shear are proposed in the literature and
discussed in this paper.

1 INTRODUCTION
A piled building located on land prone to liquefaction can be subject to inertia loads (base shear), loss of soil
strength, cyclic (transient) ground displacement (kinematic loading) and possibly permanent (cumulative)
ground displacement (lateral spread) as a consequence of strong earthquake shaking. These actions
individually or in combination, may damage foundation piles and compromise their performance.
(Tokimatsu and Asaka 1998) identified and investigated these effects as a consequence of the 1995
Hyogoken-Nambu earthquake. Design to allow for these potential effects requires geotechnical and structural
engineering collaborative input. This needs to consider: 1) how the ground will behave; 2) how the building
and foundations will behave; 3) the interaction between the ground, foundations and building. It is proposed
that possible scenarios of ground behaviour as a consequence of strong earthquake shaking be identified and
the building and foundation design be developed to accommodate these. This paper discusses identification

Paper 46
NZSEE 2020 Annual Conference
of these possible scenarios and their application to design of the building and foundation. Pseudo-static
analysis such as that proposed by (Cubrinovski et al. 2012) is discussed.

2 GROUND BEHAVIOUR
Prior to developing a pile design, it is critical to define the site ground conditions and assess the potential for
and consequences of liquefaction. This includes developing an understanding of the site’s geology via a
desktop study and then planning and implementing ground investigations and analysis. A ground model is
created to represent the site’s conditions. This ground model is normally presented on a series of cross
sections, or possibly a 3D model. The ground model is developed in stages as more information becomes
available including:
• Geology;
• Site development history and evidence of behaviour in historic earthquakes;
• Historic and new; borehole, penetrometer, laboratory and other investigation data;
• Groundwater level monitoring data and an understanding of factors that could affect groundwater level
during the life of the structure;
• Topographical data as inputs to slope stability and lateral spread assessment;
• The results of liquefaction susceptibility analysis;
• The results of assessments of cyclic displacement and lateral spread potential;
• Assessing and nominating soil parameters with and without liquefaction effects to be applied in
subsequent soil structure interaction analysis.
Ground conditions can be expected to vary across the site and the tools available to us for the assessment of
liquefaction, cyclic displacement and lateral spread potential provide an indication of a range of possible
behaviours rather than absolute numbers. Ground behaviour cannot be reliably predicted. The designer must
therefore identify scenarios to represent possible ground behaviour which are to be considered in the design.
Moderately conservative design scenarios should be selected in line with guidance in Module 4
(MBIE/NZGS 2016).
An example and common outcome of this development of a ground model and design scenarios for a site
prone to liquefaction is as follows:
• A soil profile comprising a liquefied layer sandwiched between a non-liquefied crust layer at the ground
surface and non-liquefiable base. An intensity of shaking to trigger that liquefaction and an
understanding if that liquefaction is continuous/widespread across the site and beyond, or discontinuous.
• If liquefaction is assessed to be widespread, cyclic displacement is modelled. A profile of cyclic
displacement with depth is identified. That displacement profile would normally be zero at the base of
the liquefied layer and increase approximately linearly to a peak at the top of the layer. and
• If liquefaction is assessed to be widespread and the site is sloping or adjoins a free edge (e.g. a river or
coast) lateral spread could be expected and is modelled. A profile of lateral spread displacement with
depth is modelled. A model of how the magnitude of this lateral spread could vary across the site is
defined (e.g. reducing with distance from a free edge).
Where a potential for lateral spread is identified, design should separately consider cyclic displacement and
lateral spread scenarios because predictions of which mechanism will occur may not be reliable and it is
possible that cyclic displacement could develop early in the earthquake and lateral spread later or after the
shaking.
In selection of soil strength and stiffness parameters to be applied in soil-structure interaction analysis lower
estimate values should be selected when soil is providing resistance to piles (beneficial to the design) and

Paper 46 – Design scenarios for piles in ground subject to liquefaction

NZSEE 2020 Annual Conference
upper estimate values should be applied where soil displacement is applying kinematic loads to the piles
(adverse to the design).

3 BUILDING BEHAVIOUR
The selection of the concept design for the building’s foundations and structure needs to consider the
assessed ground behaviour and the performance brief for the design. The performance brief will include
building code minimum requirements plus specific expectations of the client. These specific client
expectations could include a level of life safety greater than code and/or low damage. This concept design
should be undertaken collaboratively by the structural and geotechnical engineers in liaison with the client.
Important aspects of this concept design which influence the soil structure interaction will include:
• Pile type, diameter, flexural stiffness and capacity. Degree of pile head fixity.
• Embedment depth of any basement or substructure.
• The ability of the structure to tie all pile heads together.
Preliminary and subsequent design including consideration of soil-structure interaction could dictate
modification of some of these aspects of the concept.

4 SOIL-STRUCTURE INTERACTION
The interaction between the soil and piles and buried substructure will be dictated by the relative stiffness of
these two components; i.e. will the piles bend with ground displacement (a flexible pile) or will the soil
displace past the piles (a stiff pile)? An example of a flexible pile could be a screw pile, and an example of a
stiff pile could be a large diameter bored pile. Further insight on this aspect is provided by (Cubrinovski et al.
2006).
Another important aspect of this interaction is: will the shaking of the building (base shear loading) be in-
phase or out-of-phase with the shaking of the ground (kinematic loading due to cyclic displacement); or
somewhere between these two extremes. Further insight on this aspect and its relation to the natural periods
of shaking of the building and the ground, and pile loading is provided by (Tokimatsu et al. 2005) and
(Tamura and Tokimatsu 2005).
These factors are considered with reference to a number of design scenarios in the following sub-sections. As
discussed in Section 2, it is not possible to reliably predict ground behaviour as a consequence of
liquefaction and therefore the designer needs to develop possible scenarios of ground behaviour on which to
base the design. In the following sub-sections, four generic scenarios are presented and discussed. For a
specific design some or all of these scenarios can be expected to be relevant, depending on the site
conditions.

4.1 Design Scenario 1: Pre-liquefaction

The first scenario represents early in the earthquake before liquefaction develops, or tests the possibility that
liquefaction is not triggered by the shaking, as indicated by Figure 1. Base shear loads from the structure are
resisted by pile bending and passive resistance of the soil against the piles and buried substructure.

Paper 46 – Design scenarios for piles in ground subject to liquefaction

NZSEE 2020 Annual Conference
Figure 1: No liquefaction & 100% BS

BS: Base shear from building inertia.

FP: Passive soil resistance against buried elements including substructure and basement.
Analysis model
The analysis can be undertaken within the structural engineering computer model or separately as a pseudo-
static analysis in a geotechnical computer model such as LPile. Both analyses are recommended to provide a
check on the assessed pile actions (bending moments, shear forces and displacements). Structural properties
of the piles and their head fixity details are modelled. The soil resistance against the piles can be modelled as
non-linear springs. (Cubrinovski et al. 2006) provides some guidance on selection of these parameters. The
base shear (BS) load from the building inertia can be modelled as a point load.
Soil strength/stiffness parameters
The passive resistance of the soil against the buried substructure or basement (Fp) can be modelled as a non-
linear spring or as a point load. If a point load is assumed an assessment will be necessary to check that
sufficient displacement occurs to develop the assumed point load resistance.
Soil parameters should be lower estimate values because they are being relied on for resistance.
Base shear (BS)
The foundation system is to be designed to resist the full design inertia load from the building (100% BS).

4.2 Design Scenario 2: Liquefaction without lateral ground displacement

In this scenario the earthquake shaking has reached an intensity and duration which triggers soil liquefaction,
but not sufficient to develop widespread liquefaction and cyclic displacement. The consequence is reduced
lateral resistance to the piles but no kinematic loading as indicated by Figure 2. The designer should consider
modelling a portion of the piles in Scenario 2 conditions and the remainder in Scenario 1 to represent a
portion of the site in liquefied condition. This could represent variable conditions across the site. Assuming
pile heads are tied together (displace equally) the consequence of this analysis would be for piles in non-

Paper 46 – Design scenarios for piles in ground subject to liquefaction

NZSEE 2020 Annual Conference
liquefied soil (Scenario 1) attracting high shear loads which may be critical to the design. The designer needs
to assess what portion of the piles are to be modelled in liquefied soils (Scenario 2) and what portion in non-
liquefied (Scenario 1). In assessing this consider the variability of site ground conditions and the % ULS
shaking triggering liquefaction. If the % ULS shaking triggering liquefaction is relatively low, a greater
proportion of the site being liquefied should be allowed for. In a recent project the authors nominated 70% of
the piles to be in liquefied ground and 30% in non-liquefied ground.

Figure 2: Liquefaction & 100% BS.

BS: Base shear from building inertia.

FP: Passive soil resistance against buried elements including substructure and basement.
Analysis model
As for Scenario 1
Soil strength/stiffness parameters
The passive resistance of the soil against the buried substructure or basement (Fp) can be modelled as a non-
linear spring or as a point load. If a point load is assumed an assessment will be necessary to check that
sufficient displacement occurs to develop the assumed point load resistance.
An additional consideration is that the amount of passive pressure transferred from the crust to the sub-
structure could be influenced by the presence of the liquefied ‘weaker’ layer underneath. (Ashford et al.
2011) provides further discussion and guidance on this aspect.
Soil parameters should be lower estimate values because they are being relied on for resistance. In instances
where ground displacements are imposing kinematic loads on the piles and substructure upper estimate
values should be used.
Module 3 (MBIE/NZGS 2016) provides several references for the estimation of liquefied soil strength. For
guidance on the reduction of soil stiffness due to liquefaction and other associated factors, refer to
(Cubrinovski et al. 2006), (Tokimatsu and Asaka 1998), (Yasuda and Berrill 2000), and (O’Rourke et al.
1994). There is typically great uncertainty in predicting these parameters, and the designer should therefore

Paper 46 – Design scenarios for piles in ground subject to liquefaction

NZSEE 2020 Annual Conference
perform appropriate sensitivity analysis. The sensitivity of the pile response to soil strength/stiffness
parameters is influenced by the soil-structure interaction, this is investigated by (Cubrinovski et al. 2012).
Base shear (BS)
If the % ULS shaking triggering liquefaction is low (say 50%) liquefaction could be expected to develop
relatively early in the earthquake shaking making it more likely that design base shear loading could be felt
in conjunction with liquefaction; i.e. it may be appropriate to consider 100% BS in conjunction with this
scenario. If the trigger for liquefaction is approaching 100% ULS shaking a lower % BS could be considered;
i.e. it is less likely that 100% BS would be felt in conjunction with liquefaction.

4.3 Design Scenario 3: Liquefaction-induced Cyclic Displacement

With widespread liquefaction, cyclic displacement could be expected to develop as indicated by Figure 3. As
a result, the piles are subject to passive pile loading from the displacing crust and liquefied layer. The non-
liquefiable base layer provides passive resistance to the piles.

Figure 3: Liquefaction & +/- 80% BS & Cyclic Displacement & +/- FP.

BS: Base shear from building inertia.

FP: Passive soil resistance against buried elements including substructure and basement.
Analysis model
As for Scenario 1
Soil strength/stiffness parameters
Fp, the passive force on the substructure/basement, could be a load if the piles are relatively stiff and the sub-
structure displaces less than the ground, or it could be a resistance if the piles are relatively flexible and the
sub-structure displaces more than the ground. The designer needs to assess this and model as a load or
resistance as appropriate.

Paper 46 – Design scenarios for piles in ground subject to liquefaction

NZSEE 2020 Annual Conference
The magnitude of the passive resistance or load (FP) will differ depending on whether it is resisting (the
structure pushing into the ground) or if it is loading (the ground pushing into the structure), (Cubrinovski et
al. 2006) and (Ashford et al. 2011) provide further information on this.
Refer Scenario 2 for further comments on liquefied soil strength and stiffness.
Base shear (BS)
The designer needs to consider what portion of the design base shear is reasonable to assume to occur
cumulatively with kinematic loading from cyclic displacement. It is reasonable to assume less than 100%
because kinematic and base shear loading are not likely to be 100% in phase, and because cyclic
displacement could be expected to develop later in the shaking and thus full design base shear loading is less
likely to occur concurrently. 80% base shear is commonly assumed but may not apply in all instances.
Further discussion and guidance on the % of base shear to be allowed for in combination with kinematic
loading is provided by the Bridge Manual (NZTA 2018), (Cubrinovski et al. 2012), (Tokimatsu et al. 2005),
(Ashford et al. 2011), (Tamura and Tokimatsu 2005), (Liyanapathirana and Poulos 2005) and (Haskell
2009).

4.4 Design Scenario 4: Liquefaction-induced Lateral Spreading

The fourth scenario represents lateral spread (cumulative permanent displacement in one direction) as
indicated by Figure 4. This is likely to develop toward the end of an earthquake or post-earthquake. In this
Scenario the piles are subject to passive lateral loading within the crust and liquefied layer, and resisting
passive pressures within the non-liquefiable base layer.

Figure 4: Liquefaction & +/- 25% BS & Lateral Spread & FP.

BS: Base shear from building inertia.

FP: Passive soil resistance against buried elements including substructure and basement.
Analysis model
As for Scenario 1
Soil strength/stiffness parameters

Paper 46 – Design scenarios for piles in ground subject to liquefaction

NZSEE 2020 Annual Conference
With lateral spread, ground could be pushing against the structure developing full passive pressure against
the buried substructure as a load. The designer needs to assess if this is a reasonable assumption for the
particular site conditions.
Refer Scenario 2 for further comments on liquefied soil strength and stiffness.
Base Shear (BS)
Because lateral spread is likely to develop late in the earthquake shaking, or possibly post the shaking, it is
less likely that a high proportion of the design base shear loading will be felt in conjunction with lateral
spread. As discussed under Scenario 2 if the % ULS shaking triggering liquefaction and lateral spread is
relatively low, a higher % of the design base shear load could be expected to be felt in conjunction with
lateral spread, compared to the case of % ULS shaking liquefaction trigger approaching 100%. Furthermore,
the intensity of shaking of possible aftershocks is also a consideration. The designer should consider these
and other factors in selecting the % BS to be allowed for in conjunction with Scenario 4. 25% Base shear is
commonly assumed (Bridge Manual, NZTA 2018) but may not apply in all instances, refer to Scenario 3 for
further comments.

5 CONCLUSIONS
We cannot reliably predict the behaviour of ground during strong earthquake shaking and liquefaction. This
behaviour can be damaging to a building and its foundation. We can approach this uncertainty by;
developing an understanding of the site conditions, and undertaking analysis to provide an indication of
liquefaction, cyclic displacement and lateral spread potential. We can then apply this information and
engineering judgement to identify scenarios of possible ground behaviour for which the building design is to
be tested.
A companion paper by (Smith et al. 2020) presented at this conference reports a case study in which the
procedures proposed in this paper were applied.

REFERENCES
Ashford, A.S., Boulanger, R.W. & Brandenberg, S.J. 2011. Recommended Design Practice for Pile Foundations in
Laterally Spreading Ground. Pacific Earthquake Engineering Research Centre. PEER Report 2011/04. University of
California, Berkeley.
Cubrinovski, M., Kokusho, T. & Ishihara, K. 2006. Interpretation from large-scale shake table tests on piles undergoing
lateral spreading in liquefied soils. Soil Dynamics and Earthquake Engineering, Vol 26(2-4): 275-286.
Cubrinovski, M., Haskell, J.J.M. & Bradley, B.A. 2012. Analysis of Piles in Liquefying Soils by the Pseudo-Static
Approach. In: Sakr, M. & Ansal , A. (ed.), Special Topics in Earthquake Geotechnical Engineering. Geotechnical,
Geological and Earthquake Engineering, Vol 16: 147-174. Dordrecht: Springer.
Haskell, J.J.M. 2009. Pseudo-static modelling of the response of piles in liquefying soil. Research report, University of
Canterbury.
Liyanapathirana, D.S. and Poulos, H.G. (2005). Pseudostatic approach for seismic analysis of piles in liquefying soil.
ASCE Journal of Geotechnical and Geoenvironmental Engineering, Vol. 131 (12): 1480-1487.
NZ Transport Agency. 2018. SP/M/022 Bridge Manual. Third edition, Amendment 3.
NZGS & MBIE. 2016. Earthquake geotechnical engineering practice - Module 3: Identification, assessment and
mitigation of liquefaction hazards.
NZGS & MBIE. 2016. Earthquake geotechnical engineering practice - Module 4: Earthquake resistant foundation design.
O’Rourke, T.D., Meyersohn, M.D., Shiba, Y. & Chaudhuri, D. 1994. Evaluation of pile response to liquefaction-induced
lateral spread. In: Proc., 5th U.S.-Japan Workshop on Earthquake Resistant Design of Lifeline Facilities and
Countermeasures against Soil Liquefaction: 457-479. National Center for Earthquake Engineering Research.

Paper 46 – Design scenarios for piles in ground subject to liquefaction

NZSEE 2020 Annual Conference
Smith, V.A., Chin, E.L., & Palmer, S.J. 2020. Soil-structure interaction of piled buildings on ground subject to
liquefaction and lateral displacements. Wellington: NZSEE 2020 Conference Proceedings.
Tamura, S. and Tokimatsu, K. 2005. Seismic earth pressure acting on embedded footing based on large-scale shaking
table tests. ASCE Geotechnical Special Publication, Vol 145: 83-96.
Tokimatsu, K. & Asaka, Y. 1998. Effects of liquefaction-induced ground displacements on pile performance in the 1995
Hyogoken-Nambu earthquake. Soils and Foundations, Vol 38 (Special): 163-177.
Tokimatsu, K., Suzuki, H. & Sato, M. 2005. Effects of inertial and kinematic interaction on seismic behaviour of pile
with embedded foundation. Soil Dynamics and Earthquake Engineering, Vol 25: 753-762.
Yasuda, S. & Berrill, J.B. 2000. Observation of the earthquake response of foundations in soil profiles containing saturated
sands. In: ISRM International Symposium. International Society for Rock Mechanics and Rock Engineering.

Paper 46 – Design scenarios for piles in ground subject to liquefaction

NZSEE 2020 Annual Conference