Ground Improvement Proceedings of the Institution of Civil Engineers

Volume 167 Issue GI2 Ground Improvement 167 May 2014 Issue GI2
Pages 108–121 http://dx.doi.org/10.1680/grim.12.00012
Reliability of the Priebe method for Paper 1200012
estimating settlements Received 20/02/2012 Accepted 12/07/2012
Published online 29/11/2012
Douglas and Schaefer Keywords: embankments/foundations/risk & probability analysis

ICE Publishing: All rights reserved

Reliability of the Priebe method

for estimating settlements
j
1 S. Caleb Douglas MSc, PhD, PE j
2 Vernon R. Schaefer MSc, PhD, PE
Manager, Special Projects – Geotechnical, Union Pacific Railroad, Professor of Civil Engineering, Iowa State University, Ames, Iowa,
Omaha, Nebraska, USA USA

j
1 j
2

Many estimating methods have been developed for predicting the settlement of stone column reinforced ground.
The Priebe method is the most common method used in practice. Even though the Priebe method does not capture
all the parameters that affect the performance of stone column reinforced ground, the method is preferred due to its
simplicity. An extensive literature search provided data to evaluate the Priebe method. The concept of reliability was
incorporated to provide the framework for analysing the prediction method. The Priebe method was found to have
an approximately 89% probability that the measured settlement will be smaller than the estimated settlement. The
Priebe method was found not always to be conservative, with some possibility of settlement exceeding the
estimated amount.

1. Introduction working loads and measured/estimated soil properties. No allow-

Vibratory ground improvement methods originated in Germany in ances for variability in the loads or resistances are considered in
the 1930s. Initially, the vibratory methods were used to densify settlement analyses.
clean, granular materials at depth, which was termed vibro-
compaction. To improve weak, cohesive soils, the vibro-replace- Since the 1990s, the literature is clear on the inadequacies of
ment technique to construct stone columns was developed using present settlement prediction methods for stone column rein-
identical equipment to that for vibro-compaction. Stone columns forced ground. For example, Allen et al. (1991) stated that
have also been used in granular soils and fill materials. Stone ‘improvements in the semi-empirical settlement prediction meth-
column development, including both the vibro-replacement and ods involving stone columns are needed’. Clemente and Davie
vibro-displacement techniques, has been well described by Barks- (2000) found that ‘the results from full-scale tests show more
dale and Bachus (1983), Charles and Watts (2002), Elias et al. improvement than predicted by theoretical procedures, although a
(2006), and Kirsch and Kirsch (2010). McCabe et al. (2009) large scatter was observed’. Abdrabbo and Mahmoud (2002)
provided a recent description of the construction methods for stated that ‘there is no reliable procedure for settlement calcula-
stone columns. tion of improved geomaterial by stone columns’. Raman (2006)
found measured settlements in stone column treated areas to be
The advantages of stone columns include increasing bearing 42% of the predicted settlements. The lack of a validated design
capacity, increasing global stability, and decreasing settlements. procedure for estimating settlements is best illustrated by the
Design of stone columns is typically done by iteratively determin- results of a settlement prediction exercise for an embankment
ing the most economical layout pattern of columns (triangular or built on weak fine-grained soils summarised by Mestat et al.
rectangular), spacing, and depth of stone columns to meet project (2006). Seventeen participants submitted settlement predictions
requirements. A common purpose of utilising stone columns is to utilising the finite-element method, the discrete-element method,
satisfy project requirements with regard to settlement, and the the Priebe method, and other methods. Mestat et al. (2006)
settlement-specific analysis typically governs the final stone concluded that the exercise showed that ‘the calculation of
column configuration. Bearing capacity and global stability settlement of improved soil by stone columns is complicated and
analyses require ultimate limit state design approaches with remains a problem for practical applications’. McCabe et al.
partial factors based on Eurocode 7 or a factor-of-safety approach (2009) found that confident predictions of settlement performance
in the USA. Whereas ultimate limit state analyses include factor- for stone column treated ground were problematic.
of-safety or reliability considerations, settlements are explicitly
computed using a serviceability limit state design approach with Over 250 literature records were collected that specifically address

108
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

stone columns. Existing methods of estimating settlements for 1. Bouassida et al. (2003) and Normes Françaises (2005) are two
stone columns were identified, and case histories with measured additional methods, but these could not be located in English and
settlements were reviewed. Although many methods of estimating are not shown in Table 1.
settlements exist, the Priebe method is the focus of this paper.
McCabe et al. (2009) stated that the Priebe method of estimating The estimation of settlements of reinforced ground using the
settlements is at present the most favoured design approach of Priebe method can generally be broken down into two steps. The
leading stone column designers. Greenwood and Kirsch (1984) first step requires completion of an untreated settlement estimate
concluded that the simplicity of the Priebe method applying an considering the influence of the load. The second step then
improvement factor to conventional calculations is attractive to determines the settlement of the reinforced ground based on an
engineers, which results in the method being widely used. The improvement factor. Each step has the potential for contributing
improvement factor is defined as the untreated settlement divided to the wide range of outcomes for reinforced ground observed in
by the treated settlement. The case histories identified provide the the case histories.
data to allow a comparison of estimated and measured settlement
for both untreated ground and treated ground. The estimated and 2.1 The Priebe method
measured settlements utilised in this evaluation were obtained from Priebe initially published his design procedure in 1976 in Ger-
published case histories and are considered representative of man. Following the initial work (Priebe, 1976), Priebe adapted,
common geotechnical practice. This is a key consideration to extended, and supplemented the design procedure, as reported in
reduce bias in the evaluation because the writers did not have to Priebe (1991), and the process culminated in the procedure set
complete any calculations to develop the data points. forth in Priebe (1995). Priebe (1995) provided design procedures
and design charts for various aspects of stone column design,
Previous evaluations of projects that measured settlements in including settlement reduction, bearing capacity, shear values of
stone column treated areas compared the estimated improvement improved ground, settlement of footings, and liquefaction. Priebe
factor with the field-measured improvement factor (Balaam and (1995) contrasted vibro-replacement with vibro-compaction and
Poulos, 1983; Besançon et al., 1984; Charles, 2002; Charles and concluded that only considerable efforts such as large-scale load
Watts, 2002; Clemente and Davie, 2000; Ellouze et al., 2010; tests can prove the benefit of stone columns. Priebe (1995) stated
Greenwood and Kirsch, 1984; McCabe and Egan, 2010; McCabe that ‘the design method refers to the improving effect of stone
et al., 2009; Meyerhof, 1984). These improvement factor evalua- column in a soil which is otherwise unaltered in comparison to
tions resulted in two classes of data. The first class of data the initial state’. If the installation changes the engineering
resulted from projects in which both untreated and treated areas properties of the soil between the columns, the soil must be
were loaded, which allowed a direct computation of the settle- evaluated before the design of vibro-replacement can be accom-
ment ratio. The second class of data resulted from projects in plished. The assumptions and procedures associated with analys-
which the untreated settlement was estimated, and the settlement ing the reduction in settlements were well documented in Priebe
ratio was determined using the field-measured settlement of the (1995), and the reader is referred to that paper for a description
treated area. The complete process of estimating settlements was of the estimating procedure. Ellouze et al. (2010) provide a
not considered in the evaluations of the improvement factor. critique of the Priebe method. The Priebe method results in an
improvement factor based on the area replacement ratio and
2. Methods of estimating stone column strength of the column material. The estimated treated settlement
settlements is calculated as the estimated untreated settlement divided by the
Greenwood (1991) concluded that, under widespread vertical improvement factor.
loads, ground strengthened by arrays of columns behaves in
complex ways. Early methods of estimating settlements of stone 3. Difficulties of estimating settlements
column reinforced ground were strictly empirical and semi- As Osterberg (1986) stated, ‘[T]he realities of foundation en-
empirical. Theoretical models of the relationship between the gineering are that we never find actual conditions the same as we
stone columns and the in situ soil were presented in the 1970s. anticipated.’ Three areas that can potentially contribute to the
Since the 1970s, 17 or more design methods have been developed wide range of estimate outcomes are (a) design parameter
to estimate the settlement of stone column reinforced soil. The selection, (b) installation effects, and (c) stress distribution.
design methods developed have been based on elastic theory,
limited field data, a combination of theory and field data, 3.1 Design parameter selection
laboratory experiments, and/or numerical modelling studies. Set- Poulos (2000) said in relation to untreated sites that ‘settlement
tlement analyses of stone columns remain semi-empirical for predictions are far more sensitive to the geotechnical parameters
day-to-day designs. and site characterisation than to the method of analysis’. A
quality site investigation is required to properly identify the
The Priebe method is cited most often for estimating settlements geotechnical parameters and variability of those parameters
of stone columns. Methods of estimating settlements that do not across the site. The site investigation should leave no areas of
require advanced computing to complete are summarised in Table serious doubt concerning soil conditions, engineering properties,

109
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

Method/reference Method details Comments

Unit cell Equal strain Method theory Untreated

idealisation assumption settlement
(Yes/No) (Yes/No) required
(Yes/No)

Equilibrium method Yes Yes Elastic/empirical No Uses the stress concentration ratio to
Barksdale and Bachus determine stress reduction in soil to
(1983) estimate settlements
Finite-element method Yes Yes Elastic-plastic No Incorporates load-dependent behaviour of
design charts overall system
Barksdale and Bachus
(1983)
Priebe (1995) Yes Yes Elastic Yes Considers infinitely wide reinforced area
originally, modified for footings in 1995
Greenwood (1970) No No Empirical Yes Empirical correlation with spacing of
columns and strength of clay soil
Hughes and Withers Yes Yes Plastic Yes Early design method for widespread
(1974) loading
Balaam and Booker Yes Yes Elastic Yes Results similar to Priebe method
(1981, 1985) Considered rigid foundation
Incremental method Yes Yes Elastic-plastic Yes Considers load intensity in elastic-plastic
Goughnour and Bayuk behaviour
(1979a);
Goughnour (1983)
Balaam and Poulos Yes Yes Elastic-plastic Yes Results similar to Priebe method. Both
(1983) rigid and flexible loading
Van Impe and De Beer No Yes Elastic Yes Uses plane strain simplification. Design
(1983) charts to estimate settlements
Chow (1996) Yes Yes Elastic Yes Simple method developed for sand
compaction piles. Similar results to Aboshi
et al. (1979) and Balaam and Booker
(1981)
Alamgir et al. (1996) Yes No Elastic-plastic Yes Allowed surrounding soil to settle more
than stone column
Poorooshasb and Yes Yes Elastic-plastic Yes Priebe method is special case of general
Meyerhof (1997) equation derived for study
Pulko and Majes (2005) Yes Yes Elastic-plastic Yes Considers rigid footings
Ambily and Gandhi Yes Yes Elastic-plastic Yes Similar results to the Priebe method
(2007)
Borges et al. (2009) Yes Yes Elastic-plastic Yes Results in the range of Priebe method,
and Balaam and Booker

Table 1. Summary of methods for estimating settlements

chemical properties, and groundwater conditions (Slocombe, 3.2 Installation effects

2001). Numerical and analytical models are of limited value for Vibrated stone column installation methods have significant
settlement prediction due to the difficulty of obtaining accurate variation in performance as a result of the construction technique
soil and stone properties (Ashmawy et al., 2000). In the absence (Bell, 2004; McCabe et al., 2009). Case histories confirm the lack
of a site investigation, it is difficult to assign engineering proper- of a consistent response of the in situ soils due to stone column
ties for fill and natural soils as well as for demolition debris and installation (Castro and Karstunen, 2010; Egan et al., 2009;
refuse. Elshazly et al., 2008; Guetif et al., 2007; Kirsch, 2006, 2009;

110
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

Kirsch and Kirsch, 2010; Watts et al., 2000; White et al., 2002). ture review confirms the conclusion by Barksdale and Bachus
Specific equipment operating on a specific site using a specific (1983) nearly 30 years ago, and more recently by McCabe et al.
installation method will result in a unique effect on the in situ (2009), that there is a lack of field studies that appropriately
soil properties post installation. No clear accepted means of capture all the information required to develop a complete
anticipating installation effects has been identified, but what is understanding of the behaviour of stone column reinforced
clear is that the installation effects influence the performance of ground.
the stone columns (Egan et al., 2009). An advantage of the Priebe
method in this respect is that the method quantifies the improve- The case histories varied greatly with regard to the information
ment that results from the inclusion of the stone column without provided for site conditions, soil parameters, design considera-
any quantification of the densification of the soil between stone tions, construction process, and settlement monitoring. The initial
columns. However, the problematic installation effects of stone goal was to identify case histories that provided detailed site and
columns in sensitive soils have been well documented (Gue and design information, which would allow completion of the Priebe
Tan, 2003; McKenna et al., 1975; Oh et al., 2007a, 2007b; design method and comparison with measured settlements in the
Wijeyakulasuriya et al., 1999). field. Even with the numerous case histories found, very few case
histories provided sufficient details to allow completion of the
3.3 Stress distribution Priebe method. Fortunately, case histories were identified that
Settlement estimates typically include both the treated zone and contained settlements estimated with the Priebe method and
the underlying untreated zone. An initial consideration in evaluat- measured settlements.
ing the stress distribution is whether the foundation is rigid or
flexible. Balaam and Poulos (1983) found the reduction in 4.1 Evaluation of estimating settlements in untreated
settlement of a flexible foundation supported by stone columns to areas
be slightly less than that of a rigid foundation. The behaviour of An evaluation of settlements in untreated areas was completed
stone columns differs quite significantly, from an isolated stone based on reported estimated and measured settlements found in
column supporting a footing to a group of stone columns the literature, as detailed in Table 2 and illustrated in Figure 1.
supporting a rigid footing, to a large array of stone columns This comparison showed that six of the 12 measured settlements
supporting an embankment (Wehr, 2004, 2006). Although not were more than estimated, or unconservative. A similar compari-
explicitly discussed in Priebe (1995), matching of the stresses in son of 124 footing settlements on sands developed by Duncan
the example calculations provided in Priebe (1995) indicate that a (2000) resulted in a similar data trend and spread. The 12 data
Boussinesq-type analysis was used to estimate the stresses in the points do trend along the estimated-equals-measured line, which
untreated soils. Approximations of stress distribution were pre-
sented for similar aggregate column systems by Aboshi et al.
(1979), Bowles (1982), Lawton et al. (1994), Fox and Cowell 120
Greenwood (1970)
(1998) and Sehn and Blackburn (2008). Litwinowicz and Smith (1988)
Clemente and Davie (2000)
One of the details not identified in the literature search is to what 100 Clemente and Parks (2005)
depth settlements should be determined below an embankment or
at ent
Estimated settlement: cm

ed
tim m

structure constructed on stone column reinforced soils. In the

e
lin
es ttle

80
d
an se

design of footings, Eurocode 7 allows the analysis to consider

re
th ss

su
Le

ea

only the zone where the increase in effective stress is greater than
m
ls-

20% of the in situ effective stress (Bond and Harris, 2008). 60

a
qu

Common US practice is to consider the zone to where the

-e

ed t
at en
ed

tim m

increase in effective stress is greater than 10%. Numerical

at

es ttle
im

40
t

modelling does offer the benefit of providing information regard-

an se
Es

th ore

ing distribution of stresses and strains (Ashmawy et al., 2000;

M

Barksdale and Bachus, 1983).

20

4. Evaluation of settlement estimates

Over 100 stone column case histories were identified during this 0
study. Case histories where stone columns performed satisfacto- 0 20 40 60 80 100 120
Measured settlement: cm
rily are summarised in the supplementary data and sorted by the
following conditions: predominantly fine-grained soil case his- Figure 1. Comparison of estimated and measured settlements in
tories, predominantly coarse-grained soil case histories to miti- untreated areas. This figure illustrates that current methods of
gate static settlements, predominantly increasing resistance to estimating untreated settlements appear satisfactory based on the
liquefaction case histories, and predominantly improvement of limited data set
fill/demolition debris/refuse case histories. This extensive litera-

111
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

Reference Project Type of project Soil type(s) Estimating Estimated Measured

method settlement: cm settlement: cm

Greenwood (1970) Bremerhaven test Embankment Soft peat and clay Not stated 6.7 7.7
over fine sands
Litwinowicz and Schulz Canal/ Embankment Soft clays over One-dimensional 110 97
Smith (1988) Pound Creek, mudstone consolidation
Brisbane, Australia
50 a 39 a
30 25
Nundah Creek, Embankment Soft clays over One-dimensional 19 a 23 a
Western mudstone consolidation
Approach,
Brisbane, Australia
16 a 25 a
22 a 21 a
Clemente and Test Site 1, 2.5 m square Sands underlain Elastic 4.4 9.5
Davie (2000) Area C footing by soft silts and
clays
Test Site 2 4 m square Silts and clays Elastic 4.5 5.9
footing underlain by loose
silts and sands
Test Site 3 3.6 m square Heterogeneous fill Elastic 2.0 0.7
footing underlain by sands
2.6 1.2
Clemente and Parks Power plant, UK Storage tank Heterogeneous fill Elastic 3.5 4.7a
(2005) underlain by sands
a
Settlement shown taken as the average from range of the values reported.

Table 2. Summary of estimated and measured settlements in

untreated areas

represents the state where estimated settlements equal measured two case histories where more settlement was measured than
settlements. predicted by the Priebe method.

4.2 Evaluation of the Priebe method in treated areas A correlation coefficient of variation (COV) of 63% was
The evaluation of the Priebe method involved comparing esti- determined for the estimated-equals-measured line in Figure 2.
mated and measured settlements, as detailed in Table 3 and The correlation COV determination considered the expression
illustrated in Figure 2. From the review of the case history ‘measured settlement equal to estimated settlement’, based on the
information, the Priebe method was used most commonly. No method for computing COVs of empirical correlations, as
distinction was made in this study of which Priebe method (1976, described by Duncan et al. (1999) from the work of Ang and
1991, 1995) was utilised in the case history, as each revision Tang (1975). This determination is similar to an evaluation of 54
extended the previous procedure. The Priebe method under- footings on unreinforced sands with measured settlements greater
predicted the settlements for six of the 38 data points, which than 1.3 cm, which resulted in a correlation COV of 67%
resulted in field settlements of 110% to 143% of the estimated (Duncan, 2000).
settlements. For the 32 of the 38 data points that overpredicted
the settlements, the settlements were overestimated up to about In order to evaluate the Priebe method, the 18 data points with
300%. Although most of the data points in Figure 2 are above the measured settlements greater than 1 cm and less than 8 cm were
estimated-equals-measured line, which indicates a conservative selected. This range represents settlements that would be typical
estimation, the writers acknowledge that projects with more of a serviceability limit state analysis where limiting settlements
settlement than estimated can be considered unsatisfactory and is a project requirement. A typical settlement limit is 5 cm
commonly result in the case histories not being published. For according to Bond and Harris (2008). In US practice, structural
example, Raju (1997) mentioned, but did not explicitly detail, and embankment settlements are typically limited to 2.5 cm and

112
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

Reference Project Soil type(s) Settlement Estimated Measured

measurement settlement: cm settlement: cm
location

Greenwood and Kirsch Silo, Germany Clays and silts over Foundation 5.3 6.5
(1984) marl
Embankment fill, Very soft clay and Centre of loaded area 22.2 30.0
Hampton, VA, USA silt over sand
Meade and Allen (1985) US 42, Very soft silts and Embankment 20.3 25.4
KY, USA clays
Kirsch et al. (1986) Tank A Loose/soft very Tank shell 3.0 0.8
sandy silts
Tank B Tank shell 3.2 1.0
Tank C Tank shell 3.1 1.3
Tank D Tank shell 3.1 0.8
Tank E Tank shell 2.9 0.5
Tank F Tank shell 2.9 0.8
Tank G Tank shell 2.9 1.1
Tank H Tank shell 2.9 0.8
Tank I Tank shell 2.9 1.9
Tank J Tank shell 2.4 0.8
Tank K Tank shell 2.4 0.5
Water tank Silty sand Tank shell 6.2 3.2
Clemente and Davie (2000) Test site 4 Loose sands and Footing test 1.6 0.2
silts
Edge of tank 9.6 2.9a
Renton-Rose et al. (2000) Plant, Bahrain Sea-dredged sand Footing test 1 0.7 0.4
and gravel underlain
by marine sands
Footing test 2 1.7 0.6
Footing test 3 0.7 0.2
Footing test 4 1.3 0.6
Maduro et al. (2004) Multiple buildings, Sand fill, weak silt, Villa area 12.5 9.1
Puerto Rico sand and peat
S/E building 38.0 19.0
Clemente and Parks (2005) Power plant, UK Heterogeneous fill Combination turbine 4.2b 0.9b
underlain by sands
Generator 4.4b 1.1b
Steam turbine 3.7b 1.9b
Raman (2006) Railroad, Malaysia Soft clays and Embankment 8.3 4.8
loose sands
Embankment 6.4 3.2
Embankment 8.8 2.9
Embankment 6.8 2.4
Embankment 5.6 2.0
Mestat et al. (2006); Wehr Class A embankment Fill, compressible Centreline 6.4 5.0
and Herle (2006) settlement prediction fine-grained soils
exercise
10.0 12.0
Top of slope (at 6.4 4.5b
shoulder)
10.0 11.0b
Table 3. Summary of Priebe method estimated settlements and
measured settlements (continued on next page)

113
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

Reference Project Soil type(s) Settlement Estimated Measured

measurement settlement: cm settlement: cm
location

Bouassida et al. (2009a); Oil tank, Tunisia Loose silty sands Edge of tank 2.1 3.0
Ellouze et al. (2010) underlain by marl
stone
Bouassida et al. (2009b) 2 m high Soft alluvial clay Embankment 3.0 1.8
embankment
Mohamedzein and Embankment Soft clay underlain Centre of 27.0 24.3
Al-Shibani (2011) by sands embankment
a
Reported average settlement.
b
Settlement shown taken as the average from range of the values reported.

Table 3. Continued

40
at ent
ed
tim m
es ttle
an se
th ss
Le

30
ed t
at en
tim m
es ttle
Estimated settlement: cm

an se
e

th ore
lin

M
d
re
su
ea
m

20
s-
al
qu
-e
ed
at
tim

Greenwood and Kirsch (1984)

Es

Meade and Allen (1985)

Kirsch et al. (1986)
Clemente and Davie (2000)
10 Renton-Rose et al. (2000)
Maduro et al. (2004)
Clemente and Parks (2005)
Raman (2006)
Wehr and Herle (2006)
Bouassida et al. (2009a)
Bouassida et al. (2009b)
Mohamedzein and Al-Shibani (2011)
0
0 10 20 30 40
Measured settlement: cm

Figure 2. Comparison of estimated and measured settlements in

treated areas using the Priebe method. This figure illustrates all of
the data found for the Priebe method and the wide range of
measured settlements reported, which range from less than 1 cm
to about 30 cm

114
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

5 cm, respectively. Where settlements exceed 8 cm, stone col- previously found the Priebe method to provide conservative
umns are likely providing stability to the structure or embank- results.
ment, and settlement determinations do not always control the
final configuration. Figure 3 illustrates the data points in this Upon initial study of Figure 4, a simple conclusion would be that
range. A correlation COV of 61% for the estimated-equals- the Priebe method could be improved. One could hypothesise that
measured line was determined for the data shown in Figure 3. an acceptable, or even ideal, method would have a regression line
along the estimated-equals-measured settlement line. However,
A linear regression for the selected data was completed and is this hypothetical method would result in approximately 50% of
shown in Figure 4. With an r 2 value of 0.27, the correlation is the data points in the unconservative range. Would a prudent
poor. To enhance the usefulness of the regression, 95% con- engineer use that hypothetical method to estimate settlement on a
fidence intervals were added to the plot. The 95% confidence project with strict settlement requirements? To address this
intervals shown do not contain 95% of the data, but illustrate the question, the concept of reliability is introduced to further
bounds of 95% of the possible regression lines. Based on the evaluate the Priebe method.
location and trend of the regression line and confidence intervals,
the Priebe method is shown to provide a conservative design up 5. Reliability of settlement estimates
to settlements of 5–6 cm. The regression line trends somewhat The prediction of settlements is still difficult. Kirsch and
parallel to the estimated-equals-measured line and indicates that Sondermann (2003) estimated reinforced ground settlements
the Priebe method typically overestimates settlements by 150– below an embankment using the Priebe method to be 192 cm
200%. Elias et al. (2006) and McCabe et al. (2009) have compared with 240 cm estimated with a finite-element model and

10
at ent
ed
tim m
es ttle
an se
th ss
Le

8
ed t
at en
tim m
es ttle
e
lin

an se
Estimated settlement: cm

th ore
re

6
su

M
ea
m
s-
al
qu
-e
ed
at
tim
Es

Greenwood and Kirsch (1984)

Kirsch et al. (1986)
2 Clemente and Davie (2000)
Clemente and Parks (2005)
Raman (2006)
Wehr and Herle (2006)
Bouassida et al. (2009a)
Bouassida et al. (2009b)
0
0 2 4 6 8 10
Measured settlement: cm

Figure 3. Estimated settlements using the Priebe method for

measured settlements greater than 1 cm and less than 8 cm. This
figure illustrates the data found for the Priebe method indicative
of a range of settlements where minimising settlements is a
project requirement. The data in this figure were used to draw
the conclusions regarding the Priebe method

115
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

10
(2000) approach elicited much discussion (see Discussion to

e
Duncan, 2000), Christian and Baecher (2001) described Duncan’s

lin
95% confidence

d
re
intervals approach in the discussion as ‘a straightforward exposition of

su
ea
8 reliability methods without mystification’. Prior to discussing the

m
ls-
Priebe method in terms of reliability, observed variations from

ua
Estimated settlement: cm

q
-e
ed reported projects can provide a reference framework for compar-
at

ing the expected variability inherent to the estimating process to

tim

6
Es

r 2 ⫽ 0·24
the variability of site conditions.

5.1 Field variability from case histories

4
The results of monitoring an untreated water storage tank
described by Clemente and Parks (2005) yielded settlements
ranging from 2.5 to 6.8 cm across the 17.5 m diameter tank.
2 Settlements were estimated to be 3.5 cm. The measured settle-
ments were 71–193% of the estimated settlement for the
untreated water tank. This example is provided to illustrate (a)
0 that much larger settlements than predicted are possible across an
0 2 4 6 8 10 individual project site, and (b) that the probability of settlements
Measured settlement: cm
in excess of 175% of the estimated settlements is real as
Figure 4. Linear regression and 95% confidence intervals for modelled using reliability by Duncan (2000).
possible linear regressions using the Priebe method for measured
settlements greater than 1 cm and less than 8 cm. This figure The field test for stone column reinforced ground in Hampton,
illustrates the traditional statistical analysis of the data set. The Virginia, USA included a loaded area 6.1 m by 6.1 m, as
95% confidence intervals do not relate to future data points, but presented by Goughnour and Bayuk (1979b). Settlements were
indicate the range for probable linear regressions based on future monitored below the centre and at the four corners of the loaded
data. This figure shows the Priebe method to be in the area. The settlements at the four corners after 130 days were
conservative range 8.1 cm, 9.7 cm, 12.5 cm, and 13.2 cm. From these four readings,
the site COV for settlement was 22%. This value compares well
with an untreated case history by Wu et al. (2011), which found a
concluded that the match ‘appears satisfactory’. Example calcula- site COV for settlement of 21%. The site COV is due to the
tions presented by Priebe (1995) estimated a final settlement of change in soil parameters and profiles across the site. Note that if
38 cm where Greenwood (1991) measured settlements of the rate of consolidation is included in the evaluation, a much
40–41 cm. Thus, even the example calculation shown by Priebe larger site COV will result due to the variable drainage conditions
(1995) underpredicted the known settlement by 5–8%. (Alonso and Jimenez, 2011).

The Priebe method data shown in Figures 2 and 3 resulted in 5.2 Reliability of the Priebe method
correlation COVs of 63% and 61%, respectively. Considering the With a correlation COV of 60% for the Priebe method consider-
reliability approach presented by Duncan (2000) with a 60% ing the estimated-equals-measured line, Figure 5 illustrates how
COV for the Priebe method, there is a 10% probability that the probabilities of exceedance of 1%, 5% and 10% compare with
settlement may be larger than 175% of the estimated settlement. the selected data from the case histories. These probabilities
With this 10% possibility of much larger settlements than result in estimated settlements 175–300% of the values along the
estimated, a prudent designer would be compelled to consider the estimated-equals-measured line. Figure 5 represents the highest
following questions. likely COV from the Priebe method.

(a) How much variation is inherent in the estimating method? Inherent site variability results in a site COV of 20–25%,
(b) How much variation is inherent to in situ properties? regardless of the analysis method. This site COV represents the
(c) Should estimated settlements be presented as a single number lowest variation that could be reasonably assumed in a settlement
or a range? evaluation. Considering a COV of 25%, Figure 6 graphically
(d ) What are the consequences if the settlements are under illustrates how probabilities of exceedance of 1%, 5% and 10%
predicted? compare with the selected data from the case histories. Con-
(e) What percentage probability for exceeding a settlement sideration of these probabilities results in estimated settlements
threshold is the designer or owner willing to accept? 135–175% of the values along the estimated-equals-measured
line.
The concept of reliability can be used to provide the framework
to assist in answering these questions. Although the Duncan Through iteration of lines similar to the probability lines of

116
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

10 10

%
0%

0%
5

.5
.1

.1
ed.

e
ed

lin
e
ed

d
ce
lin
xce

d
ee

re
. 1%

xce

ex
d

xc

su
re
of e

of

ea
e

su
8 8

of

m
of

ea

b.
d

%
b.

s-
xcee

b.
Pro
b.

al
.1
ls-
Pro

Pro

qu
Estimated settlement: cm

Estimated settlement: cm
Pro

ua

ed

-e
of e

ed
xce
-e
ed

at
e
6 6
.

tim
at
Prob

of
tim

Es
b.
Es

Pro
4 4

2 2

0 0
0 2 4 6 8 10 0 2 4 6 8 10
Measured settlement: cm Measured settlement: cm

Figure 5. Reliability of the Priebe method estimates considering a Figure 6. Reliability of the Priebe method estimates considering a
COV of 60% with regard to the variation of the data from the COV of 25% with regard to site soil conditions. This figure
estimated-equals-measured line. This figure illustrates the illustrates the probability of exceedance using the Priebe method
probability of exceedance using the Priebe method and the and the lowest likely site COV, which is 25%. The realistic COV
highest likely correlation COV, which is 60% value to use lies somewhere between the 25% and 60% range

Figures 5 and 6 and determination of the correlation COV of each

10
possible line using the Ang and Tang (1975) approach, the best- at ent
ed
tim m

fit line shown in Figure 7 was produced and resulted in a

es ttle
an se

correlation COV of 48%. In order to answer the question posed at

th ss
Le

the close of Section 4, the concept of reliability provides an 8

assessment of the conservativeness of the Priebe method. Consid- ed t
e

at en
lin
Estimated settlement: cm

tim m
d

ering a correlation COV of 50% for the best-fit line, there is an

es ttle
re
%

su

an se
48

89% probability that settlements will be smaller than those

ea
n C ne

6
th ore
m
OV
Es latio -fit li

M
s-

estimated with the Priebe method. Stated differently, the Priebe

al
qu
rre est

method tends to overestimate measured settlements by approxi-

-e
B

ed

mately 160%.
at

4
tim
Co

The evaluation included case histories where stone columns were

utilised to reinforce clay, silt/sand, and sand soils. Figure 7 also Clay soil sites
illustrates the performance of stone columns in the different soil 2
Silty sand soil sites
types. Clear conclusions regarding the Priebe method and the Sand soil sites
applicability to different soil types could not be developed from
Figure 7. 0
0 2 4 6 8 10
Measured settlement: cm
6. Conclusions
The stone column case histories identified varied greatly with Figure 7. Priebe method best-fit line and data sorted by site soil
regard to the information provided for site conditions, soil conditions. This figure illustrates the best-fit line using the
parameters, design considerations, construction process, and reliability approach. The best-fit line corresponds to a reliability of
settlement monitoring. This study confirms the work of previous the Priebe method resulting in overpredicting settlements 89% of
authors in that there is a lack of detailed, research-oriented case the time. The data points have also been sorted by soil type to
histories that fully document the design, construction, and per- visually indicate that no conclusions could be developed for the
formance of stone column reinforced ground (Barksdale and Priebe method based on the soil type
Bachus, 1983; McCabe et al., 2009). Settlement predictions of

117
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

both untreated and treated ground are highly dependent upon the Alonso JA and Jimenez R (2011) Reliability analysis of stone
quality of the site investigation data upon which they are based. columns for ground improvement. In GeoRisk 2011: Risk
Assessment and Management. Geotechnical Special
The Priebe method is used most commonly for estimating Publication 224. ASCE, Reston, VA, USA, pp. 493–500.
settlements of stone column reinforced ground. Even though the Ambily AP and Gandhi SR (2007) Behavior of stone columns
Priebe method does not capture all the parameters that affect the based on experimental and FEM analysis. Journal of
performance of stone column reinforced ground, the method is Geotechnical and Geoenvironmental Engineering, ASCE
preferred due to its simplicity. Since no safety factors or margin 133(4): 405–415.
of error are currently considered in settlement analyses, reliability Ang AH-S and Tang WH (1975) Probability Concepts in
provided a framework in which to evaluate the Priebe method. Engineering Planning and Design: Volume 1 – Basic
Considering data with a maximum measured settlement of 8 cm, Principles. Wiley, New York, NY, USA.
there is an 89% probability that settlements estimated with the Ashmawy AK, Rokicki R and Plaskett ME (2000) Predicted and
Priebe method will exceed measured settlements. As shown in actual settlement of soils improved by vibro-replacement.
the case histories and the reliability study, the Priebe method is In Advances in Grouting and Ground Modification.
not always conservative, and settlements may exceed those Geotechnical Special Publication No. 104. ASCE, Reston,
estimated. As more complicated models and methods continue to VA, USA, pp. 281–295.
be developed, the geotechnical community should consider Balaam NP and Booker JR (1981) Analysis of rigid rafts
whether less conservative methods than the Priebe method should supported by granular piles. International Journal for
be used to estimate settlements of stone column reinforced Numerical and Analytical Methods in Geomechanics 5(4):
ground. 379–403.
Balaam NP and Booker JR (1985) Effect of stone column yield on
In closing, a philosophical statement from Terzaghi (1936) over settlement of rigid foundations in stabilized clay.
75 years ago may be the best reminder to practising engineers International Journal for Numerical and Analytical Methods
when contemplating settlements: ‘Whoever expects from soil in Geomechanics 9(4): 331–351.
mechanics a set of simple, hard and fast rules for settlement Balaam NP and Poulos HG (1983) Behaviour of foundations
computation will be deeply disappointed. . .. The nature of the supported by clay stabilised by stone columns. Proceedings of
problem strictly precludes such rules.’ the 8th European Conference on Soil Mechanics and
Foundations Engineering. Balkema, Rotterdam, the
Acknowledgments Netherlands, 1, 199–204.
This study was funded by the Strategic Highway Research Barksdale RD and Bachus RC (1983) Design and Construction of
Program 2 of the National Academies, with Dr James Bryant as Stone Columns, Vol. I, and Vol. II, FHWA/RD-83/026.
programme manager. The opinions, findings and conclusions Federal Highway Administration, Washington, DC, USA.
presented herein are those of the authors and do not necessarily Bell AL (2004) The development and importance of construction
reflect those of the research sponsor. technique in deep vibratory ground improvement. In Ground
and Soil Improvement (Raison CA (ed.)). ICE Publishing,
REFERENCES London, UK, pp. 103–111.
Abdrabbo FM and Mahmoud ZI (2002) Control of ground Besançon G, Iorio JP and Soyez B (1984) Detailed analysis of the
settlement by stone columns – case study. Proceedings of the parameters governing the design of stone columns.
4th International Conference on Ground Improvement Proceedings of In Situ Soil and Rock Reinforcement
Techniques, Malaysia, 179–186. (Renforcement en place des sols et des roches). Presses de
Aboshi H, Ichimoto E, Enoki M and Harada K (1979) The l’ecole de nationale des Ponts et Chaussées, Paris, France, pp.
‘Compozer’ – a method to improve characteristics of soft 119–126 (in French).
clays by inclusion of large diameter sand columns. Bond A and Harris A (2008) Decoding Eurocode 7. Taylor and
Proceedings of an International Conference on Soil Francis, New York, NY, USA.
Reinforcement, Reinforced Earth and other Techniques, Paris, Borges JL, Domingues TS and Cardoso AS (2009) Embankments
pp. 211–216. on soft soil reinforced with stone columns: numerical analysis
Alamgir M, Miura N, Poorooshasb HB and Madhav MR (1996) and proposal of a new design method. Geotechnical and
Deformation analysis of soft ground reinforced by columnar Geological Engineering 27(6): 667–679.
inclusions. Computers and Geotechnics 18(4): 267–290. Bouassida M, Guetif Z, De Buhan P and Dormieux L (2003)
Allen TM, Harrison TL, Strada JR and Kilian AP (1991) Use of Estimation par une approche variationnelle du tassement
stone columns to support 1–90 cut and cover tunnel. In Deep d’une fondation sur sol renforcé par colonnes ballastées.
Foundation Improvements: Design, Construction, and Testing. Revue Française de Géotechnique 102: 21–22
ASTM STP 1089. American Society for Testing and (in French).
Materials (ASTM), West Conshohocken, PA, USA, Bouassida M, Ellouze S and Hazzar L (2009a) Investigating
pp. 101–115. Priebe’s method for settlement estimation of foundation

118
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

resting on soil reinforced by stone columns. Geotechnics of Manual. Geopier Foundation Company, Mooresville, NC,
Soft Soils – Focus on Ground Improvement. Taylor and USA.
Francis, New York, NY, USA, pp. 321–325. Goughnour RR (1983) Settlement of vertically loaded stone
Bouassida M, Hazzar L and de Buhan P (2009b) A software columns in soft ground. In Proceedings of the Eighth
programme for designing columnar reinforced soils. European Conference on Soil Mechanics and Foundations
Geotechnics of Soft Soils – Focus on Ground Improvement. Engineering. Balkema, Rotterdam, the Netherlands, 1,
Taylor and Francis, New York, NY, USA, pp. 327–332. 235–240.
Bowles JE (1982) Foundation Analyses and Design, 3rd edn. Goughnour RR and Bayuk AA (1979a) Analysis of stone column–
McGraw-Hill, New York, NY, USA. soil matrix interaction under vertical load. Proceedings of the
Castro J and Karstunen M (2010) Numerical simulations of stone International Conference on Soil Reinforcement: Reinforced
column installation. Canadian Geotechnical Journal 47(10): Earth and Other Technologies, Paris, vol. I, pp. 271–277.
1127–1138. Goughnour RR and Bayuk AA (1979b) A field study of long term
Charles JA (2002) Ground improvement: the interaction of settlements of loads supported by stone columns in soft
engineering science and experience-based technology. ground. Proceedings of the International Conference on Soil
Géotechnique 52(7): 527–532. Reinforcement: Reinforced Earth and Other Technologies,
Charles JA and Watts KS (2002) Treated Ground – Engineering Paris, vol. I, pp. 279–285.
Properties and Performance. CIRIA C572. Construction Greenwood DA (1970) Mechanical improvement of soils below
Industry Research and Information Association, London, UK. ground surface. In Ground Engineering, Proceedings of the
Chow YK (1996) Settlement analysis of sand compaction pile. Conference on Ground Engineering. Institution of Civil
Soils and Foundations, Japanese Geotechnical Society 36(1): Engineers, London, UK, pp. 11–22.
111–113. Greenwood DA (1991) Load tests on stone columns. Deep
Christian JT and Baecher GB (2001) Discussion of factor of safety Foundation Improvements: Design, Construction, and Testing.
and reliability in geotechnical engineering. Geotechnical and ASTM STP 1089. ASTM, West Conshohocken, PA, USA, pp.
Geoenvironmental Engineering, ASCE 127(8): 700–703. 148–171.
Clemente JL and Davie JR (2000) Stone columns for settlement Greenwood DA and Kirsch K (1984) Specialist ground treatment
reduction. Proceedings of GeoEng 2000, Melbourne, by vibratory and dynamic methods. State of the Art Report:
Australia, CD-ROM. Piling and Ground Treatment. Thomas Telford, London, UK,
Clemente JLM and Parks C (2005) Stone columns for control of pp. 17–45.
power station foundation settlements. In Innovations in Gue SS and Tan YC (2003) Prevention of failure of bridge
Grouting and Soil Improvement. Geotechnical Special foundation and approach embankment on soft ground.
Publication 136. ASCE, Reston, VA, USA, CD-ROM. Proceedings of the 2nd International Conference on Soft Soil
Duncan JM (2000) Factors of safety and reliability in Engineering and Technology, Selangor. See
geotechnical engineering. Journal of Geotechnical and www.gnpgeo.com.my (accessed 26/02/2011).
Geoenvironmental Engineering, ASCE 126(4): 307–316. Guetif Z, Bouassida M and Debats JM (2007) Improved soft clay
Duncan JM, Navin M and Patterson K (1999) Manual for characteristics due to stone column installation. Computers
Geotechnical Engineering Reliability Calculations. CGPR and Geotechnics 34(2): 104–111.
No. 17. Center for Geotechnical Practice and Research, Hughes JMO and Withers NJ (1974) Reinforcing of soft cohesive
Virginia Polytechnic Institute and State University, soils with stone columns. Ground Engineering 5(7): 42–49.
Blacksburg, VA, USA. Kirsch F (2006) Vibro stone column installation and its effect on
Egan D, Scott W and McCabe B (2009) Installation effects of ground improvement. Numerical Modelling of Construction
vibro-replacement stone columns in soft clay. Geotechnics of Processes in Geotechnical Engineering for Urban
Soft Soils – Focus on Ground Improvement. Taylor and Environment. Taylor and Francis, New York, NY, USA,
Francis, New York, NY, USA, pp. 23–29. pp. 115–124.
Elias V, Welsh J, Warren J et al. (2006) Ground Improvement Kirsch F (2009) Evaluation of ground improvement by groups of
Methods, vol. 1. FHWA Publication No. NHI-06–020. vibro stone columns using field measurements and numerical
Federal Highway Administration, Washington, DC, USA. analysis. Geotechnics of Soft Soils – Focus on Ground
Ellouze S, Bouassida M, Hazzar L and Mroueh H (2010) On Improvement. Taylor and Francis, New York, NY, USA, pp.
settlement of stone column foundation by Priebe’s method. 241–247.
Proceedings of the Institution of Civil Engineers – Ground Kirsch K and Kirsch F (2010) Ground Improvement by Deep
Improvement 163(2): 101–107. Vibratory Methods. Spon, New York, NY, USA.
Elshazly H, Elkasabgy M and Elleboudy A (2008) Effect of inter- Kirsch F and Sondermann W (2003) Field measurements and
column spacing on soil stresses due to vibro-installed stone numerical analysis of the stress distribution below stone
columns: interesting findings. Journal of Geotechnical and column supported embankments and their stability.
Geological Engineering 26(22): 225–236. Proceedings of the International Workshop on Geotechnics of
Fox NS and Cowell MJ (1998) Geopier Soil Reinforcement Soft Soils – Theory and Practice. VGE, Essen, Germany.

119
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

Kirsch K, Heere D and Schuhmacher N (1986) Settlement Southeast Asian Geotechnical Conference, Kuala Lumpur. See
performance of storage tanks and similar large structures on http://www98.griffith.edu.au/dspace/bitstream/handle/10072/
soft alluvial soils improved by stone columns. Building on 19556/44221_1.pdf?sequence¼1 (accessed 26/02/2011).
Marginal and Derelict Land. Proceedings of a conference Oh EYN, Balasubramaniam AS, Surarak C et al. (2007b)
organized by the Institution of Civil Engineers Glasgow, UK. Interpreting field behaviors of embankment on estuarine clay.
Thomas Telford, London, UK, pp. 651–663. In Proceedings of Seventeenth International Offshore and
Lawton EC, Fox NS and Handy RL (1994) Control of settlement Polar Engineering Conference, Lisbon. See http://
and uplift of structures using short aggregate piers. In www98.griffith.edu.au/dspace/bitstream/handle/10072/19562/
Proceedings of IN-SITU Deep Soil Improvements. ASCE, 46736_2.pdf?sequence¼1 (accessed 9/05/2011).
Reston, VA, USA, pp. 121–132. Osterberg JO (1986) The realities of foundation engineering.
Litwinowicz A and Smith IK (1988) A brief review of Proceedings of the International Conference on Deep
geotechnical aspects and monitoring of Gateway Arterial Foundations. China Building Industry Press, Hong Kong, PR
Roadworks north of the Brisbane River. In Fifth Australia- China, pp. 1.203–1.208.
New Zealand Conference on Geomechanics: Prediction Poorooshasb HB and Meyerhof GG (1997) Analysis of behavior
Versus Performance. Institution of Engineers, Barton, ACT, of stone columns and lime columns. Computers and
Australia, pp. 298–304. Geotechnics 20(1): 47–70.
Maduro IJ, Molina CR, Castillo LV, Renoud-Lias B and Salvi GJ Poulos HG (2000) Foundation settlement analysis – practice
(2004) The use of stone columns on settlement and versus research. The Eighth Spencer J. Buchanan Lecture, 03/
liquefaction susceptible soils. Proceedings of the 5th 09/2000, delivered at College Station Hilton, College Station,
International Conference on Case Histories in Geotechnical TX, USA. See https://ceprofs.civil.tamu.edu/briaud/
Engineering, New York, USA, Paper No. 1.08. buchanan%20web/Lectures/Eighth%20Buchanan%20
McCabe BA and Egan D (2010) A review of the settlement of Lecture.pdf (accessed 29/07/2012).
stone columns in compressible soils. In Ground Improvement Priebe HJ (1976) Abschatzung des Setzungsverhaltens eins durch
and Geosynthetics. Geotechnical Special Publication 207. Stopfuerdichtung verbessertan Baugrundes. Die Bautechnik
ASCE, Reston, VA, USA, pp. 197–204. 53(H.5): 160–162.
McCabe BA, Nimmons GJ and Egan D (2009) A review of field Priebe HJ (1991) Vibro-replacement – design criteria and quality
performance of stone columns in soft soils. Proceedings of control. In Deep Foundation Improvements: Design,
the Institution of Civil Engineers – Geotechnical Engineering Construction, and Testing. ASTM STP 1089. ASTM, West
162(6): 323–334. Conshohocken, PA, USA, pp. 62–72.
McKenna JM, Eyre WA and Wolstenholme DR (1975) Priebe HJ (1995) The design of vibro-replacement. Ground
Performance of an embankment supported by stone columns Engineering 28(10): 31–37.
in soft ground. Géotechnique 25(1): 51–59. Pulko B and Majes B (2005) Simple and accurate prediction of
Meade BW and Allen DL (1985) Evaluation of Stone Column settlements of stone column reinforced soil. Proceedings of
Stabilized Embankment Foundation. UKTRP-85–27. the 16th International Conference on Soil Mechanics and
Transportation Research Program, University of Kentucky, Geotechnical Engineering, Osaka, Japan, 1401–1404.
Lexington, KY, USA. Raju VR (1997) The behavior of very soft cohesive soils improved
Mestat Ph, Magnan JP and Dhouib A (2006) Results of the by vibro-replacement. Ground Improvement Geosystems –
settlement prediction exercise of an embankment founded on Densification and Reinforcement. Thomas Telford, London,
soil improved by stone columns. In Numerical Methods in UK, pp. 253–259.
Geotechnical Engineering (Schweiger H (ed.)). Taylor and Raman S (2006) Comparison of Predicted Settlement Behaviour
Francis, London, UK, pp. 471–476. to the Field Measurement of Stone Column Improved Ground.
Meyerhof GG (1984) Closing address. In Piling and Ground M. Eng. Report, University of Technology, Malaysia.
Treatment. Thomas Telford, London, UK, pp. 293–297. Renton-Rose DG, Bunce GC and Finlay DW (2000) Vibro-
Mohamedzein YEA and Al-Shibani IH (2011) Performance of an replacement for industrial plant on reclaimed land, Bahrain.
embankment supported on soft soil reinforced by stone Géotechnique 50(6): 727–737.
columns. Proceedings of the Institution of Civil Engineers – Sehn AL and Blackburn JT (2008) Predicting performance of
Ground Improvement 164(4): 213–224. aggregate piers. Proceedings of the 23rd Central
Normes Françaises (2005) Recommandations sur la conception, le Pennsylvania Geotechnical Conference, Hershey, PA, USA.
calcul, l’exécution et le contrôle des colonnes ballastées sous Slocombe BC (2001) Deep compaction of problematic soils. In
bâtiments et ouvrages sensible au tassement. NFP 11–212. Problematic Soils. Thomas Telford, London, UK, pp. 163–
Revue Française de Géotechnique, 2nd trimester (111), 3–16 181.
(in French). Terzaghi K (1936) Discussion of settlement of structures.
Oh EYN, Balasubramaniam AS, Surarak C et al. (2007a) Proceedings of the 1st International Conference on Soil
Behaviour of a highway embankment on stone columns Mechanics and Foundation Engineering, Cambridge, UK, 3,
improved estuarine clay. In Proceedings of the Sixteenth 79–87.

120
Ground Improvement Reliability of the Priebe method for
Volume 167 Issue GI2 estimating settlements
Douglas and Schaefer

Van Impe W and De Beer E (1983) Improvement of settlement columns – Priebe method and FEM. In Numerical Methods
behavior of soft layers by means of stone columns. in Geotechnical Engineering (Schweiger H (ed.)). Taylor and
Proceedings of 8th European Conference on Soil Mechanics Francis, London, UK, pp. 773–776.
and Foundations Engineering. Balkema, Rotterdam, the White D, Wissmann K, Barnes A and Gaul A (2002) Embankment
Netherlands, 1, 309–312. support: a comparison of stone column and rammed
Watts KS, Johnson D, Wood LA and Saadi A (2000) An aggregate pier soil reinforcement. In Proceedings of the 55th
instrumented trial of vibro ground treatment supporting strip Canadian Geotechnical and 3rd Joint IAH-CNC and CGS
foundations in a variable fill. Géotechnique 50(6): 699–708. Groundwater Specialty Conferences, Niagara Falls, Ontario,
Wehr J (2004) Stone columns – single columns and group Canada, CD Rom.
behavior. Proceedings of the 5th International Conference on Wijeyakulasuriya V, Hobbs G and Brandon A (1999) Some
Ground Improvement Techniques, Kuala Lumpur, Malaysia, experiences with performance monitoring of embankments on
329–340. soft clays. Australia–New Zealand Conference on the
Wehr J (2006) Stone columns – group behaviour and influence of Geomechanics, Hobart, Tasmania, Australia, pp. 783–788.
footing flexibility. In Numerical Methods in Geotechnical Wu TH, Gale SM, Zhou SZ and Geiger EC (2011) Reliability of
Engineering (Schweiger H (ed.)). Taylor and Francis, settlement prediction – case history. Journal of
London, UK, pp. 767–772. Geotechnical and Geoenvironmental Engineering, ASCE
Wehr J and Herle I (2006) Exercise on calculation of stone 137(4): 312–322.

WHAT DO YOU THINK?

To discuss this paper, please email up to 500 words to the
editor at [email protected]. Your contribution will be
forwarded to the author(s) for a reply and, if considered
appropriate by the editorial panel, will be published as a
discussion in a future issue of the journal.
Proceedings journals rely entirely on contributions sent in
by civil engineering professionals, academics and students.
Papers should be 2000–5000 words long (briefing papers
should be 1000–2000 words long), with adequate illustra-
tions and references. You can submit your paper online via
www.icevirtuallibrary.com/content/journals, where you
will also find detailed author guidelines.

121
Copyright of Ground Improvement: Proceedings of the Institution of Civil Engineers is the
property of Thomas Telford Ltd and its content may not be copied or emailed to multiple sites
or posted to a listserv without the copyright holder's express written permission. However,
users may print, download, or email articles for individual use.