580
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
occurred, which caused tank shell deformations. Such
deformations affected normal operation of the floating roof and
consequently made the use of the tank impossible. The case
described above has not been fully documented, but this is the
reason why the investor imposed strict requirements for tank
behaviour.
Based on the documented cases of soil yielding during tank
foundation works (Bell & Iwakiri, 1980) criteria for design,
construction and use of tanks have been established.
Criteria for maximum total and differential settlement of
tanks were determined according to Marr, et al. (1982), API-650
and API-653. Allowable differential settlements are the
maximum allowable design limits for deformation of the tank
after allowance has been made for construction tolerances.
These comprise combinations of: (a) tilt of the tank; (b) tank
floor settlement along a radial line from the perimeter to the
tank centre; and (c) settlement around the perimeter of the tank.
Foundation and foundation soil are subjected to the highest
load during hydro tests when a tank is filled with water having a
density of 1 t/m
3
. Later, during tank use, loads on foundations
and foundation soil are lower by about 15 % because the tanks
are filled with crude oil having a density of 0,85 t/m
3
.
3
OVERVIEW OF GROUND INVESTIGATION WORKS
At the site, 45 geotechnical boreholes were drilled of which
three were 70 m deep. In addition to the boreholes, 18 CPTU
tests were also carried out. From the boreholes, undisturbed soil
samples were continually taken or SPTs performed. A
piezometer was installed in one borehole and a level of ground
water monitored over a number of years. Soil classification tests
as well as strength, stiffness and water permeability tests were
carried out in a laboratory. The investigations showed that the
soil is horizontally stratified.
4
DESCRIPTION OF FOUNDATION
The foundation soil was improved with hundreds of stone
columns. After the soil had been prepared in this way, the tank
shell and bund wall were installed on rigid reinforced concrete
ring while the tank bottom was placed directly on the bedding
prepared.
4.1
Soil improvement
As the foundation soil is horizontally stratified, the soil under
all the tanks has almost the same properties. Because of its
small stiffness and low water permeability, the foundation soil
for each tank was improved with about 660 stone columns.
The depth of the improved soil was approximately 18 m. The
spacing between stone columns varied depending on their
location on the layout plan. Considering that tank structure is
susceptible to planar tilt settlement and non-planar settlement,
stone columns were spaced more closely on the perimeter below
the foundation ring and centre to achieve stronger effect of
improvement.
The quality of improvement was checked by CPTU and
SASW tests and geodetic surveys carried out in control fields
before and after soil improvement. In addition, the data relating
to the installation of stone columns were analyzed. Among other
things, the volume of the gravel pressed into foundation soil
was determined. For each tank, it was found to be about 3% of
the volume of the foundation soil improved. As geodetic
surveys showed negligible soil upheave (a few millimetres), it
can be considered that all the stone pressed into the soil
increased directly its density, i.e. soil compaction.
4.2
Concrete ring foundation
The shells of both the tanks and bund walls were mounted
directly on a rigid reinforced-concrete foundation ring of
rectangular cross-section b/h=350/(260-370) cm, with a central
drainage gutter having a width b=60 cm and variable height
h=40–80 cm. Outside and inside ring diameters are
D
out
=79,00 m and D
in
=72,00 m respectively.
4.3
Bedding of steel tank bottom
The bottoms of the steel tanks were mounted directly on the
multi-layer bedding prepared as described below.
The foundation soil was levelled and a layer of gravel of
grain size 0-64mm was placed. The layer had 70 cm in
thickness. To prevent soil pollution in case of tank leak, a
HDPE geomembrane was installed in the bedding. The
geomembrane was placed between a geosynthetic clay liner and
clean sand to protect it from damage. Above the geomembrane,
cathodic protection was installed. Additional reinforcement of
the soil below the tanks was achieved by placing sand in
geocells of 20 cm in height.
5
GEOTECHNICAL DESIGN ANALYSIS
On the basis of in situ and laboratory tests, an axisymmetric
numerical model was created in Plaxis 2D-V8. The material
behaviour is represented by the Hardening Soil model.
In a numerical analysis, soil materials of five types were
used and their description and some properties are shown in
Table 1. The analysis included tank installation stages, hydro
tests and tank exploitation.
Table 1. A description of stratified foundation soil
Layer
depth [m] description
(1)
approx
0 –6
Surface layer of stiff clay
k=0,0002[m/day]; E
oed
ref
=9,4[MPa];
E
ur
ref
=30[MPa]; p
ref
=100[kPa];
m=0,409
(2)
approx
6 –13
Layer of soft clay
k=0,0002[m/day]; E
oed
ref
=7,8[MPa];
E
ur
ref
=23,4[MPa]; p
ref
=100[kPa];
m=0,376
(3)
approx
13 –20
Sand with silt and clay
k=0,02[m/day]; E
oed
ref
=16[MPa];
E
ur
ref
=50[MPa]; p
ref
=100[kPa]; m=0,5
Soil below the tanks was improved
with stone columns as designed
k=1[m/day]; E
oed
ref
=35[MPa];
E
ur
ref
=90[MPa]; p
ref
=100[kPa]; m=0
(4)
approx
20 –70
Alternating layers of clay and sand
with silt
k=0,0002[m/day]; E
oed
ref
=10[MPa];
E
ur
ref
=40[MPa]; p
ref
=100[kPa];
m=0,376
k - permeability; E
oed
ref
reference edometric modulus at
reference stress p
ref
; E
oed
=E
oed
ref
(
/p
ref
)
m
edometric modulus;
E
ur
ref
= unload/reload modulus
6
HYDRO TEST
6.1
Introduction
In the case of the Sisak tanks, hydro tests were conducted as
part of technical monitoring to determine a set of documented
and interconnected activities which would provide proof of
proper behaviour of all elements of a tank structure.
In case of malfunction or critical deviations from the
expected behaviour, such activities would ensure that these
deviations are removed or corrected on time by taking necessary
measures approved by experts.
Under a procedure for a hydro test, the phases of tank and
bund filling and emptying were defined; after each phase had
