GROUND IMPROVEMENT USING STONE COLUMNS:

PROBLEMS ENCOUNTERED

A.Verghese Chummar1

ABSTRACT

The stone column technique of ground improvement is extensively used to improve the strength of weak
soil layers. Stone columns essentially increase the bearing capacity of cohesive soils. In cohesive soils, the
drainage path provided by stone columns accelerates the rate of residual settlement due to consolidation.
The paper gives the design and construction methods of stone columns and highlights the errors that can
affect stone columns. Two case studies of stone column failure are presented in detail. These studies reveal
the errors that occurred in the design and construction methods. The case studies also highlight the probable
field conditions that could create excessive settlements. Based on these observations, conclusions are
derived indicating the specific care that should be taken in the design and execution of stone columns.

INTRODUCTION

Stone column technique for ground improvement is being extensively used to undertake constructions in
weak soils. The stone columns essentially increase the bearing capacity of loose cohesionless soils. In
cohesive soils, along with the increase of bearing capacity, the consolidation settlement of the ground under
loading is also considerably reduced. In addition, in cohesive soils stone columns act as drainage paths to
accelerate the rate of consolidation of the residual settlement. Even though stone columns are very useful for
these purposes, designs made without proper concept of the behaviour of the stone column and execution of
work not understanding the stone column and behaviour pattern of the non-treated ground leads to
complications. Failures have occurred where stone columns have been used for ground improvement. This
paper attempts to highlight these factors. A case study where the foundation failure of a structure
constructed on a soil improved by stone columns highlights the various aspects discussed.

CONCEPT OF DESIGN

Stone columns are essentially designed to take load when the columns bulge
under loading and the surrounding soil offers passive resistance. In a cohesive
soil, when stone columns are constructed, a considerable portion of improved
ground can be reduced to as much as 30% of the natural ground. In addition,
the stone columns act as a drainage path, which would accelerate further
settlement under loading.
In a cohesionless soil like loose silty sand, when stone columns are
constructed by vibro-floatation, the ground is densified by the reduction in
voids and the intrusion of gravel and stones. Thus, in totality, the bearing
capacity increases. The drainage and residual settlements are not important in
such soil conditions.
The general behaviour of the stone column is indicated in Figure 1.

Figure 1 : Single Granular

Pre-Bulging Failure Mode

1
Director, M/S. F.S.Engineers Pvt. Ltd., 109 Velachery Road, Guindy, Chennai - 600 032, INDIA
LIMIT CONDITIONS IN THE DESIGN OF STONE COLUMNS

Even though the individual stone column capacity is

computed based on the resistance offered by the surrounding
soil, the upper limit to which the bearing capacity of a cohesive
soil can be improved in relation to its natural strength is limited.
As indicated in Figure 2, this would show that however close the
stone columns are placed, the upper limit to which the strength
could be improved is nearly 25 times its original value of
undrained shear strength. Given a safety factor of 3, the safe
bearing capacity could be improved to a maximum of 8 times the
Cu value.
In the same manner the settlement of the ground can be
reduced only to a limiting value as indicated in Figure 3.
Spacing of stone columns closer than the limit value will not
have any additional effect in reducing the settlements. Similarly
the spacing of the column beyond a certain limiting value would
not have the required effect on reduction of settlements due to
consolidation.

Figure 2 : Relation between

Undrained Sheer Strength of
Cohesive Deposit and Allowable
Vertical Stress on Granular Pile
(After Thorburn, 1975)

Figure 3 : Spacing of Granular Pile and Settlement

of Treated Ground in Uniform Soft Clay

The above factors clearly indicate the limit conditions that must be considered when designing stone
columns, even though the design is based on the individual capacities of the columns.
SUITABILITY OF THE GROUND FOR STONE COLUMNS

Two common methods of stone column constructions are:

• Vibro-floatation
• Bored rammed system
In the floatation technique (Figure 4), a vibrating needle working with the water jet reduces the friction of
the surrounding cohesionless soil, filling the voids and thus achieving compaction. This creates a cavity into
which borrowed granular material is filled and compacted. Thus,
in general, the density of the ground increases with the increase
in friction angle. As a result, bearing capacity increases. This
technique is therefore possible only in cohesionless soils.
The bored rammed stone columns are used in cohesive soils.
In this technique (Figure 5), a casing pipe is used to remove the
cohesive soil protecting the sides of the bore, thus minimising
disturbance to the surrounding soil. The stones are laid into the
bore and rammed
to a larger
diameter as the
casing pipe is
withdrawn.
These columns
achieve their
strength by the Figure 4 : Vibro-Compaction Process
lateral restraint
offered by the surrounding soil. It is therefore very essential
that the shear strength of the surrounding soil not be reduced
by the construction of the stone column. Hence, the stone
column technique could be adopted in clays of low
sensitivity. These columns also act as drainage paths to
Figure 5 : Cased Rammed Stone Column
accelerate settlements under loading.

ERRORS IN CONSTRUCTION

As indicated in the vibro-floatation technique above, compaction of surrounding soil creates a cavity.
This cavity is filled by borrowed granular material, which has to be cohesionless. If the soil has cohesion,
the disturbance reduces the strength of the surrounding soil, resulting in no densification.
Even with the use of cased rammed stone column technique when ramming, the surrounding soil gets
disturbed to some extent. Sensitive clays do not adequately regain shear strength. Due to this, ground
improvement by stone column cannot be achieved in clays with sensitivity greater than 3. This aspect is
often neglected in adopting stone columns for ground improvement.
In cohesive soil as explained, apart from increasing the bearing capacity, the stone columns act as a
drainage path to accelerate the settlement due to consolidation under loading. It is therefore necessary to
have a free drainage path through the stone columns and through a blanket of clean cohesionless soil on top
of the stone column to allow pore pressure dissipation. This aspect is often neglected. The top surface of the
stone column is not properly cleaned and proper care is not taken to provide a 15 cm thick , blanket drainage
layer of sand on top. This results in breakage of drainage path and prevention of consolidation settlement
under loading.
Often, due to the problems in handling heavy casing pipes, bentonite circulation is adopted to protect the
sides of the bore. This definitely blocks the entire drainage path and spoils the essential functioning of the
stone column.
CASE STUDY OF A FOUNDATION FAILURE

For the installation of spherical liquid petroleum gas storage tanks, raft foundation system was suggested
in very soft, sensitive clay. The clay layer initially had a value of cohesion of 0.1 kg/cm2 and sensitivity of 5.
The stone column technique was adopted to improve the layer of 11 to 12 M thickness.
Vibro-floatation technique of stone column construction was adopted. Each stone column of dia 1 M was
designed to take the load of 25 T and spacing was adjusted such that the bearing capacity of the soil was
improved to 2.5 kg/cm2.
Plate load test conducted on individual stone column confirmed the design value of load carrying
capacity.
After the stone column was created, the structure was constructed on top. Total collapse and failure of the
entire foundation occurred when hydro-testing for the storage tank was at just 1/3 its capacity.
Analysis of the cause of failure indicated the following errors in the design and construction of the stone
columns:
1. The original soil bearing capacity of 0.1 kg/cm2 had a limit to which it could be increased to 0.8
kg/cm2. Considering only individual capacity of the stone column the design was made to improve
the ground capacity to 2.5 kg/cm2 of loading by reducting the spacing of the column.
2. The soil in which the ground was to be improved had a sensitivity of 5. The vibro-floatation
technique thoroughly disturbed the surrounding soil. The residual strength of the disturbed sensitive
clay came down to nearly 0.02 to 0.03 kg/cm2. The only possible construction technique that could
be adopted for the type of clay layer is cased rammed stone column with controlled ramming
ramming giving minimum disturbance to the sensitive clay.
3. The failure occurred at a load intensity of 0.6 kg/cm2 indicating the value of cohesion that remained
after disturbance as low as 0.02 kg/cm2.
4. The top surface of the stone column was not cleaned before laying the sand blanket resulting in
blockage of pore pressure dissipation.
5. The storage tanks were filled at a fast rate not giving any time for possible pore pressure dissipation
even through the partially blocked drainage path. This would have gradually increased the shear
strength of the surrounding soil with the consolidation settlement.
6. The load test of an individual stone column using a small plate did not reveal these defects.

ERRORS IN TESTING STONE COLUMN

As indicated, individual stone columns achieve the capacity by the resistance offered by the surrounding
soil. If the testing of 1 M dia stone column is done, using a 30 cm plate in the middle, an abnormally high
value is obtained for obvious reasons. It is therefore necessary to do testing on a group of stone columns.
Even then, the testing of the stone columns particularly in cohesive soil will not indicate the residual time-
bound settlement that could take place.

THE NECESSITY OF PRE-LOADING

Stone columns in cohesive soils reduce the settlement due to consolidation by a maximum of 70%. The
balance 30% of the settlement has to be taken out by pre-loading. In the construction of storage tanks,
hydro-testing of the tank with gradual loading could take care of this settlement. If the structure constructed
does not have the facility for pre-loading distress in the structure could take place with the residual
settlements which can be more than the permissible limits. Stone columns in cohesive soils is therefore
suitable only if pre-loading facility is available, as in the case of storage-tank construction.
CONCLUSIONS

From the analysis it is concluded that to avoid errors in design and execution of the stone column the
following precautions have to be taken.
1. Understanding the limits to which the stone columns can improve the ground capacity and reduce the
settlement irrespective of the behaviour of the individual stone column.
2. Choice of the correct technique of the stone column construction depending upon the soil properties.
3. Assessment of the properties of the soil layers that can be improved by stone column.
4. Proper execution of the stone column to take sure that the functions of the stone columns is not
blocked by the technique of construction adopted.
5. The necessity for pre-loading to remove the residual settlement in cohesive soils.

REFERENCES

Greenwood, D.A. (1970). “Mechanical Improvement of Soils below Ground Surface”, Proceedings
Ground Engineering Conference, Inst of Civil Engineering, London.
Hughes, J.M.O Withers, N.J. (1974), “Reinforcing of Soft Cohesive Soils with Stone Columns”,
Ground Engineering, London, Vol.17, No.3, pp. 42-49.
Thornburn, S. (1975), “Building Structures Supported by Stabilized Ground”, Geotechnique, London,
Vol. 25, No.1, pp. 83-94.
Chummar, Verghese A., “Sub-soil Exploration Reports of F.S.Engineers (P) Ltd.”