Missouri University of Science and Technology

Scholars' Mine

International Conference on Case Histories in (1984) - First International Conference on Case

Geotechnical Engineering Histories in Geotechnical Engineering

07 May 1984, 11:30 am - 6:00 pm

Evaluation of a Differentially Settled Tank

S. M. Gazioglu
D'Appolonia Consulting Engineers, Inc.

J. L. Withiam
D'Appolonia Consulting Engineers, Inc.

Follow this and additional works at: https://scholarsmine.mst.edu/icchge

Part of the Geotechnical Engineering Commons

Recommended Citation
Gazioglu, S. M. and Withiam, J. L., "Evaluation of a Differentially Settled Tank" (1984). International
Conference on Case Histories in Geotechnical Engineering. 16.
https://scholarsmine.mst.edu/icchge/1icchge/1icchge-theme1/16

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been
accepted for inclusion in International Conference on Case Histories in Geotechnical Engineering by an authorized
administrator of Scholars' Mine. This work is protected by U. S. Copyright Law. Unauthorized use including
reproduction for redistribution requires the permission of the copyright holder. For more information, please
contact scholarsmine@mst.edu.
 Evaluation of a Differentially Settled Tank
S. M. Gazloglu and J. L W"dhlam
Senior Project Engineer and Staff Consultant, D'Appolonla Consulting Englneera, Inc.

SYNOPSIS The paper discusses studies undertaken to identify the cause(s) for differential settle-
ments experienced by a large floating roof tank. The studies included an evaluation of existing
subsurface and tank performance data, additional subsurface exploration and laboratory test pro-
grams, a monitoring program during the restricted use of the tank and recommended remedial measures
to allow full use of the tank. It is concluded - that the affected portion of the tank was sited over
a thicker and more compressible soil layer than the remaining portions and that releveling by mud-
jacking would allow unrestricted future use of the tank.

INTRODUCTION Prior to tank construction, a site investigation

In 1978, the Department of Energy (DOE) began
the construction of the St. James Terminal fa-
cility which included six tanks having a total
capacity of two million barrels. These tanks
were to serve as surge tanks for transfer of
crude oil to and from the Strategic Petroleum
Reserve (SPR) sites at Weeks Island and Bayou
Choctaw where crude oil is stored in underground
caverns in salt domes.
ASSUMPTION
PARISH
was performed to develop recommendations for
design of the tank foundations and foundation
preparation. Because of a concern for poten-
tially large tank settlements, these recommenda-
tions included either (1) a monitored preload of
the site with a soil surcharge of 2,000 pounds
per square foot (psf) prior to tank construc-
tion, or (2) a controlled and monitored stage
loading of the tanks with water prior to their
placement into service. Due to the urgency of
putting the tanks quickly into service, the
second alternative was selected and the tanks
were constructed. Tank Nos. 1 through 5 were
satisfactorily preloaded with water and placed
into service. However, during the preloading of
Tank No. 6 from June through October 1979, the
northeast quadrant of the tank experienced
greater settlements which occurred at a faster
rate than the remaining portions of the tank.
Tank No. 6 was subsequently emptied and a field
investigation program was implemented to deter-
mine the cause(s) for the observed differential
settlements.
This paper describes the studies that were un-
dertaken to analyze the Tank No. 6 settlements,
including a comparison of observed settlements
and tank distortion with various tank perform-
ance criteria and the efforts that were made to
put the tank into service on a restricted basis
by monitoring the comprehensive soil and tank
instrumentation.

TANK NO. 6 CONSTRUCTION AND LOADING

Figure 1. Site Location
As shown in Figure 1, St. James Terminal is
situated within approximately 2,000 feet of the
Mississippi River. Because the site is located
within the Mississippi River flood plain, the
soils underlying the tank foundations consist of
a highly variable and compressible stratigraphy
to a depth of 50 to 60 feet.
Tank No. 6 is a floating roof 300 feet in ·dia-
meter tank with a usable product storage capac-
ity of 33 feet and a storage volume of 400,000
barrels. As shown in Figure 2, Tank No. 6 is
located at the easternmost portion of the site
within 2,000 feet of the Mississippi River.
Foundation designed for Tank No. 6 consists of
an 18-inch wide by 3.5 feet deep concrete ring-
wall to support the tank superstructure and a

133
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
 I I I I
~ ~
--- _ ........ •
~ ==-~- ::ft'± -
-.x.. '-..:r___,-:;;-;:._-..:_
•
t--·---.... •. . . .
--·'"':".---+ ;;:;
~;--=----!..~"""""
<--.
j_ __
.......
I

I -
. . ................
..............
---·--- -------.......
..,.,llil

. ...
_
.., ............ __ _

------------··....:::::::::::::::::.::-::: ::::·::::-:~= ~:·~- _) \

-- l
··--------~----~--:. _____ ~~-:;·\ ,I
..
--------+·-----------------k--·- \+ (
r-..::-::=--+-=---·1+- -------------· •I
---------~ ~~/
PROPER~- LINEJ
Figure 2. Details of Tank No. 6 Subsurface Investigation

-- -- -- -- -- --

I
I
!I I I ! I !

Figure 3. Tank No. 6 Preloading and Settlement Data

three to six foot sand pad within the ringwall Tank settlements were surveyed using 32 sett
foundation to support the tank floor and product ment survey points established on the ring•
load. The sand pad was crowned approximately of the tank. Ringwall settlements measured
three feet at the center of the tank during four survey points are plotted in Figure 3.
construction to accommodate future tank settle- indicated in the figure, Tank No. 6 shell 1
ments. The tank shell consists of four courses tlement was approximately 8. 5 inches at Set I
of butt welded steel panels varying from 1.041 ment Survey Point (SSP) 6 -and approximat
inches at the bottom to 0.375 inches at the three inches at SSP 21. Furthermore, settl~
top. The bottom plate of the tank includes surveys indicated that the tank shell was c
5/16-inch thick lap-welded steel panels. A ring ferentially distorted downward for a maximw
stiffener at the top and a floating roof com- 5.5 inches over a relatively short perim1
prise the remainder of the tank structure. length of 220 feet in the northeast quadr.
The remainder of shell settlements were t •
As shown in Figure 3, Tank No. 6 was filled with tively uniform. Figure 4 shows plots of ·
wat.e r in stages to consolidate the underlying pressures from four piezometers located at
soU.s prior to placing the tank into service depths below the center of the tank and i ·
following completion of construction activities cates that the pore pressures induced by
i;11 May, 1979. During staged loading, a field water loading of the tank were dissipated r
pro~ra_m w~s instituted to monitor the behavior
tively quickly. Inclinometer data shown
of tbe tank and foundation soils by means of Figure 5 indicates that maximum lateral
etb'l,ement surveys, pore pressure measurements placements in the foundation soils were on
aod ~lope indicator measurements.
order of one inch in the northeast quadrant
'-
134
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
 -I-.... :::::.,.. --
t
....,,.
' ' ' ' -- ' '
..-...
' ' ' ' --- -- -- 1::.
' ' ' '~ ' '
....n•- ' -.

-
I -...:
1...
- 1-
- - ~ ...
. .
0

... •
..
..
~

.. .
,.. .
(INITIAL ATtC POftE P'RESIIItf. -1&5 MSLI

! .

II ...
u
.. sr.cn iot: .... • .~ ....
...
H

.... ,,.,. SUIC

"""" •
....
_ , ..
...
. ~ -·-WLI'Ja
"* ·..- - - .

Figure 4. Tank No. 6 Preloading and Pore Pressure Data

contours in Figure 6.
The original ground surface near Tank No. 6 was
approximately at Elevation +13 feet mean sea
level (El +13 MSL) and was raised to El +15 MSL
Q.S
OE.tUC:ttON, INCHES
1,0 I.S
during construction of the perimeter road .
Three borings (Boring Nos. JD-1, JD-2 and JD-3)
were drilled and four test pits (Test Pit Nos . 1
10 through 4) were excavated at the locations shown
in Figure 2 to suplement pre-construction data
-s lO
obtained by others to more clearly define the
subsurface soil conditions around the perimeter
t; of the tank. The subsurface conditions encoun-
~ •15
tered from the borings and test pits are shown
§ in Figure 7 . The upper ten feet of soil under-
~-ZSL---~::;:.....- lying the tank consists of brown to gray silty
clays which are overconsolidated probably due to
10 desiccation. Below this layer, alternating
• .M.T U,lt7t sequences of soft to medium stiff clays, loose
• 4$
e
A
JULY tt, ttH
AUGUST M,ll?l
to medium dense silts and fine silty sands are
10
WCUIIOMC.Tllll O.TA
encountered to El -50 MSL underneath the north-
noM Tl-11
ClAST - fJ/1 TANK MO. •1
east quadrant and to El -30 MSL underneath the
-55 TO remaining portions of the tank. These alterna-
ting layers of clays, silts and fine sands con-
Figure 5 . Tank No. 6 Inclinometer Data sist of recent alluvial deposits which were
underlain by stiff silty clay and dense fine
sand deposits of Pleistocene Age.
Due to the proximity of the Mississippi River to
compared to negligible displacements in the the tank, the ground water table is strongly
remaining portions of the tank. influenced by the river stage. Based on piezo-
meter readings, the ground water table at Tank
Starting in late August 1979, Tank No. 6 was un- No. 6 was conservatively established near El +12
loaded in stages and the tank was completely MSL for the analyses .
empty by the middle of January 1980. An inves-
tigation program was then begun to determine the Upon completion of the field investigation pro-
cause(s) for differential settlements exper- gram, laboratory testing was performed to char-
ienced in the northeast quadrant of the tank . acterize the strength and compressibility of
various soil layers. Laboratory testing in-
cluded water content determinations, Atterberg
Limits, grain size analyses, unit weight deter-
FIELD INVESTIGATION PROGRAM minations, consolidation tests, consolidated
undrained triaxial tests with pore pressure
A survey of Tank No . 6 ringwall and floor eleva- measurements and torvane tests.
tions was made in late January 1980 to define
the distortions of the tank shell and floor. EVALUATION OF TANK NO. 6
Survey indicated a localized depression near
SSP's 6 and 7 as shown by the floor settlement The integrity of Tank No . 6 was analyzed using

135
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
 .J.UUI.

..
KT1UfriUT .-v!Y ,_.,. 011 M
~R......._,__.TIOI

MTTLIMllll &liNn "*fT '70 •

IIIOIITO·O~ITMII~

tarsnta 100' IUII'I"'OtT f'CISTI

COMTOUitS 011'-Ua - · IUIM-
"ANWo..,, .-10 1'1 IT, Mil-l
li.Of'l ""CATDIII TON ....'f'OMC
DUfUNI 1'4NII ""P..OotoDUII
ptUO..ITIItl TO 8 1 MOII'fMI
DU,.MI TAINll'ltl~

ITitlii·ITII.llM ..wet .,...,._.

OUitiMG lAIIII HIIU,OADIMt

TANK N0.6

!!WI:
t.e&.aYATIONI Ul iAitO OII M . . . . •
!liil!!iil!!e!IJ!!!"~~:-iiiili rm
,1!1
.
IIIAIIICMIICIIII.... °C0 ,Q.aW.TIGM
14.Ni r&IT, IQL.
l.flafleiMIIM.fUIIMILOCATIOII. M
D'n.IWIO• fiO.U.

Figure 6. Tank No. 6 Floor Settlement Contours

SECTION A- A
, _ .. . IIIDimll

.t rr ..... L -:-;
Figure 7. Subsurface Profile Beneath Tank No. 6

136
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
 :rANK N0.6
RINGWAU.

·~~~~~---~--~-~~-~~~
....,. , ,.,, ...... I'Ut

SECTION A-A

. . . .
I&Tn.oe.IIIIICt ..... - - - : ' MTW t1Ln731"!!-III&J\
..
~ ~....._ - ............ 1--. v .I
I·'· -··"--
Jl!!ll PLAN
=IITTWCJff
e .£._
·I
I. .'THI. ......... IIDI'ATICII. ,_ IDT· 'CT

a.
TIU' PUiC ..... .......a
\:
~
IEI"t\.DDIWW'UMYfiOIWfiONltC.....U.MC
..,_ltf-...miLCIIDIP,_OTMIIIIII* Sl
-=-=-~

. . . . . .
---
RINGWALL SETTLEMENT AROUND TANK PERIMETER

Figure 8. Evaluation of Tank No. 6 Ringwall Settlements

differential ringwall and floor settlements. three adjacent settlement points on the shell,
The cause(s) for the observed differential set- computed in relation to the best-fit rigid-set-
tlements were evaluated with respect to bearing tlement tilt plane through all of the settlement
capacity failure and consolidation settlements. survey points. The results of this analysis
indicated that the maximum differential settle-
ment (referenced to the best-fit rigid-settle-
ment tilt plane) was 0.13 percent in January
Differential Settlements 1980 and 0.16 percent in September 1979. These
maximum out-of-plane settlements were within the
limiting criterion of 0.22 percent as suggested
Survey data obtained during January 1980 and on by Belloni et al. (1974) as a "working hypothe-
September 20, 1979, when the tank experienced sis" for safe operation of large tanks with
the maximum differential settlements, were anal- floating roofs. Sullivan and Nowicki (1974)
yzed to assess the existing and most distressed found that large tanks with diameters up to 360
condition of the tank. The results of these feet could experience as much as 1. 2 inches of
analyses were then compared to various tank differential settlements (with respect to best-
performance criteria that were available in the fit rigid-settlement plane) without problems,
literature at that time. but that out-of-roundness problems occurred for
differential settlements greater than 1.8 inches
regardless of the settlement distribution or
Tilting: The best-fit rigid-settlement tilt tank diameter.
plane for Tank No. 6, as shown in Figure 8, was
determined using procedures suggested by Bell The angular distortion, which is defined as dif-
and Iwakiri (1980) and Greenwood (1974). This ferential settlement between two points divided
evaluation was made to determine the signifi- by the distance between those points along the
cance of differential settlements with respect tank perimeter, was computed to be 0.39 percent
to bending or distortion of the tank shell or between SSP's 3-4 and 8-9. This differential
tank floor. The angle of the best-fit rigid- settlement is in excess of the allowable differ-
settlement tilt plane was determined to be 0.07 ential shell settlement criterion of 0.35 per-
percent, which is well below with the limiting cent for floating roof tanks (Belloni et al.
criterion of 0.5 percent. Greenwood (1974) sug- 1974). DeBeer (1969) suggested a similar cri-
gested that an average tilt of possibly more terion.
than 0.5 percent of the best-fit rigid-settle-
ment tilt plane could be experienced before the
distortion (i.e., out-of-roundness) at the top
ring girder of a floating roof tank would cause Bearing Capacity
binding between the roof and the shell.
Differential Tank Shell Settlement: Figure 8 The measured settlement configuration of Tank
also presents the results of an analysis of No. 6, as shown in Figure 3, suggests no indica-
distortional shell settlements following the tion of foundation soil failure since the under-
method described by Belloni et al. (1974). They lying soils were still consolidating and gaining
define the maximum out-of-plane perimeter set- strength at the time of maximum loading. An
tlements as the maximum change in slope between analysis was performed to assess the factor of

137
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
 40
to
10
t; 0
it
- >0
-40
_,.
-10
Figure 9.
safety against a soil bearing failure below the
tank.
Local Bearing Capacity: The factor of safety
against a local bearing capacity failure at the
no~tbeast quadrant of Tank No. 6 was computed to
be on the order of 1.6 with a full water preload
of 2,000 psf. The computed factor of safety 1.6
is somewhat greater than the conventional mini-
mum factor of safety against a local bearing
faHure ·Of 1.5. A two dimensional, limit equil-
ibrium stability program (Siegel, 1978) was used
t:Q compute the factors of safety against lo-
calized bearing failure. This computer program
is based om. a modified Bishop procedure (Bishop,
195:5). which utilizes circular failure sur-
faces. Various searching techniques were em-
ployed to define the most critical failure sur-
faces. The undrained shear strength used in the
analysis and the most critical bearing failure
su1:faces. are summarized in Figure 9.
Overall Bearing Capacity: Because of the flexi-
bili~y of large tank structures, the possibility
of a general bearing capacity failure is remote
and is usually controlled by localized subsur-
face soi 1 conditions. Nonetheless, an over all
factor of safety against a general bearing fail-
ure should be examined. A simplified procedure
for evaluating this aspect of tank behavior
consists of averaging the shear strength of the
s ·o il to a depth of approximately two thirds of
the foundation's width (Skempton, 1951). Using
this approach, a factor of safet y against an
overall undrained bearing failure of at least
·4.5 was obtained for the design load of 2,000
psf. Standarg design practice for overall
bearing capacity requires a safety factor of at
l-east 3. 0 for rig,id foundations. Because the
soils beneath the tank were only partially con-
solidated ym.de:r wat er test loads, the factor of
lif\f,et:y fQr both l.e~;al.i zed and overall bearing
capa9ity1, fa i\,.u£e wou ld impr ove with t i me.
,I p.
~tas R.l~ :!iow.; When the, induced sbe·a r stresses
~e ~c :, the' sheaJ1i-pg s t rength of tJhe .s oil, a con-
d!i~on, · ~nQWDJ a s '" plast-i ·c flow" develops in
!i9 1\~
··~ .
$o U.s i.n the plastic s t a t e deform under
.
;r~rJ, ~ :'
First International Conference on Case Histories in Geotechnical Engineering
W,!•
Missouri University of Science and Technology
~;i~llf~Y,.1 ,<.·
http://ICCHGE1984-2013.mst.edu
.,.~·.:1~., ~~
40
Evaluation of Tank No. 6 Bearing Capacity
a constant shear stress (Jorgensen, 1934). Tbc
distortions produced by plastic flow a l ter .t \
structure of the soil skeleton causing a loss c
shear strength, a further increase in st~:a i1
and subsequent redistribution of stresset
Assuming homogeneous, isotropic a_nd elastic s o :
conditions, the maximum shear stress induced ·
the soil under a circular footing applying
uniform bearing pressure, P, is approximate:
P/ n , or about one third of the foundat i•
bearing pressure. The maximum shear stre1
develops in the soil along a bowl-shaped sur f a c
with a maximum depth of approximately 0.4 tim•
the diameter of the loaded area as shown :
Figure 8. The shear stresses a r e lower abo,
and below this surface.
The average undrained shear strength of t l
upper 60 feet of soils underlying the northea<
quadrant of Tank No. 6 is approximately 675 p
based on labor a tory test data. Using the a ·
proach outlined above, bearing pressure of 2, 0
psf resulted in a maximum shear stress o f 6
psf in the soils below the tank. Since t .
average undrained shear strength of 6 75 ps f
greater than the maximum induced shear stress
635 psf, plastic flow probably did not signi f
cantly contribute to the overall settlements
the tank although localized overstressing m
have occurred. Figure 5, which shows a plot
inclinometer data near SSP's 6 and 7 , i nd icat
that plastic flow may have resulted on the ord
of one inch settlement below the northeast qua
rant of the tank.
Settlement Analyses
Using one-dimensional consolidation theor
ana lyses were performed to determine the tot
primary and secondary settlements below t
northeast quadrant and other locations of t
tank. Localized soil profiles (See Figure
were developed at e.ach location considered us i
data acquired during the subsurface explorati
programs. The soil properties used in t
analyses are provided in Table I.
138
 Table I. Soil Parameters for Settlement Analyses

Northeast Quadrant of the Tank Remaining Portions of the Tank

Layer Layer
Layer Thickness Cc Ccr OCR Layer Thickness Cc Ccr OCR
(ft) (ft)
1 3.0 0.03 0.003 5.0 1 2.0 0.03 0.003 5.0
2 8.0 0.24 0.042 4.5 2 10.0 0.24 0.042 3.0
3 15.0 0.19 0.015 1.0 3 18.0 0.16 0.011 2.0
4 8.5 0.20 0.020 1.0 4 10.0 0.20 0.020 1.7
5 7.0 0.21 0.019 1.0 5 6.0 0.31 0.031 1.6
6 9.0 0.30 0.030 1.2 6 19.5 0.05 0.005 1.5
7 13.0 0.41 0.040 1.5 7 11.5 0.15 0.015 1.5
8 3.0 0.05 0.005 1.5 8 8.5 0.20 0.020 1.5
9 10.0 0.15 0.~5 1.5 9 25.5 0.10 0.010 1.5
10 8.5 0.20 0.020 1.5 10 12.0 0.15 0.015 1.5
11 25.0 0.10 0.010 1.5
12 12.0 0.15 0.015 1.5
Note: Consolidation parameters, Cc, Ccr• and OCR, are estimated based on laboratory
tests from pre- and post-construction investigations.

The soil profile and soil parameters used for
settlement analysis at the center of the tank
were compiled by averaging the data obtained
from investigations conducted before and after
tank construction.
PrimaTy Settlements: Using the laboratory con-
solidation test results, the max i mum total pri-
mary settlements with a design load of 2,000 psf
were estimated to be approximately 20 inches at
the northeast quadrant and at the center and
four inches for the remaining portions of the
tank. Based on the settlement survey data, the
northeast quadrant of the tank would have exper-
ienced 12 inches of additional settlement s as
TIME IN DAYS
10
0
~-J_J_~ ]. ._
compared to only one inch for the remaining
portions of the tank. These additional settle-
ments would have resulted in unacceptable tank
distortions and could potentially have resulted
in structural failure of the tank.
Settlement data were also analyzed to determine
the expected max i mum settlements underneath the
northeast quadrant and other locations below
Tank No. 6 using a procedure suggested by Su
(Transportation Research Board, 1976). Using
the settlement survey data and this procedure,
the total primary settlements at SSP's 6 and 21
were estimated to be 12.8 and 4.4 inches, re-
spectively as shown in Figure 10.
Settlement Rate: Rates of primary settlement
wer e computed by uti l izing pore pressure
measurements obtai ned during the June-October
zo 50 100 200 500 1979 preloading shown in Figure 4, the ringwall
I . settlement data shown in Figure 2 and coeffic-
15 CYCLES

--- • • INITI.__ 111.1..1NG

'I
ients of consolidation from laboratory tests.
These analyses indicated that approximately 140
~4TTL£MEHT =i to 280 days would be required to achieve 90 to
• "'- ~~
SURVEV

j - IJ· -- .~
2 POINT 21
! ill
100 percent of anticipated primary settlement
-~~t
I , J l I
under a design loading of 2,000 psf. Since the
tank had been preloaded approxi mately 150 days
4
~~ETT~EME~T 8 •4 4 ::::::J since June 1979, the est i mated degr ee of consol-
~ ...
:uovtY INCHES
POINT 6 I
i i
I
idation was approximately 50 to 60 percent below
the northeast quadrant of the tank and 90 to 100
i
; percent elsewhere.
: I I
...
ffi ' ~~.~ I Secondary Settlements: Secondary settl ements
8 were computed to be approximately four i nches
::1
......~
I
! ,,
I
i
I i. ! and two inches, respectively at the northeast
~
I
i
I! ! ' I i =I quadr ant and remaining portions of the tank over
t he anticipated 20 year life. Ther ef ore, if the
! ;' Ii .:" ~
:I:
. .\ ~r~
10
I
i !
effect of localized differential settlement
I II .I :I i i assoc i ated with the consolidation of underlying
12
I 1
' ~+-+ soils could be corrected, it was concluded that
II ~INCHEs ~ I !
;
i secondary settlements would not adversel y affect
! , !a1• 1z
the structural integrity of the tank due to
14
I'I I
: 1I. _j I
I
their relative uniformity.
i ' iI
I
! : ! I
I I i I' I
I
I
16
REMEDIAL MEASURES
Figure 10. Analyses of Tank No. 6 Settlement The analyses pe't'formed to assess the integrity
Data of the tank indicated that Tank No. 6 was struc-
139
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
 turally sound in its present unloaded condi-
tion. However, because the soils below the
northeast quadrant were only 50 to 60 percent
consolidated under the design load of 2,000 psf,
recommendations were made to relevel the tank
prior to its further preloading with water to
minimize the possibility of structural dis-
tress. Remedial measures included mudjacking
below the northeast quadrant to lift this area
to an equal or higher level with respect to the
remaining portions of the tank and preload and
monitor the behavior of the tank prior to
placing it into service.
PERFORMANCE MONITORING
Tank No. 6 remained empty from January 1980
until March 1981 without any attempt to relevel
the tank by mudjacking.. In March 1981, an ele-
vation survey of the ringwall foundation
performed which indicated two inches of unif
rebound had occurred since emptying the tank
January 1980. At this time, the DOE indiest
that it would be desirable to utilize Tank O• 6
to accommodate upcoming oil transfer operatio s
to the Weeks Island and Bayou Choctaw sites.
The analyses of March 1981 survey data abo ed
that Tank No. 6 could be loaded intermittently
provided that its behavior was carefully oni-
tored. Consequently, strain gages mounted in a
rosette pattern were installed at several loca-
tions on the tank shell to measure changes in
strain level as product was loaded into tbe
tank. In addition, piezometers installed during
the post-construction investigation program we~e
prepared for monitoring and arrangements were
made to survey the ringwall elevations on a
daily bas is.
Initially, Tank No. 6 was filled to height of 20

r
....
-~
.~
- --
·--- i/ll!tl
..... .......
,-t
1- ~ ,_ .. 4--41--1-4-l-4-+-l-+-1-l-4--1--41-+-l-+-1-:--tl-l

.......... ~
•1--1--""'
1=1:' I :.. .,.,

Figure 11. Tank No. 6 Oil Loading and Settlement Data

: : 1 - < o_..
!·-+
..,...- _ _ _+-it-l-.,-t""'-+.. ~~ -+,,.-++-H""'-I-""-HI-II--+-t..,.-11H•~·~:= i
..... H- I

I
II
-- ..ux.aa
IJ -~-

Figu-ce 12. Tank No. 6 Oil Loading and Pore Pressure Data

140

First International Conference on Case Histories in Geotechnical Engineering

Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
 -- -- -- ... .....
r,- Ill
I
1- t
1- I

-
f.-if- f.-
I ~+ f.-
I
f.-f.- f.-
bda.
M! !!,
~ I"
1-
i ~ !··. I
I 1- I
r-. .. I*'
... ~
1-

Figure 13. Tank No. 6 Oil Loading and Stress Data

....._ __
-· ...... .... , ••••• w •••••••••

kY.~lJ:U-
~ •••••
-
- b. ~::-~i="-1~
~ ·~
~r- k!:: -I"' •I'"'
- -...
.,.r-

" - p.~

--· -- ' ~ -
_.~:--

1\ VI :r:L•:,:.--::;:
I
1-- I' 1-r...... ~ f-'..::,=*~=·
'"'~"'

I- "~ ~~
I' I' v~ Or-- 1--V'
t-1-
I' I'
I' II

r--l\ I"~ v ~=--

.. :0:.~':.~'" 1- ...
---·-----·-·
.....,....,,..,..,...,_._

__..... ..
•••• , . _ , . . , . .· -IDNII'

Ltnft.DDIT _.,,_,.--.&. ...

................ ...........
,., ,..
.-..
..
..
"I\
1\f' I- VII
1/
I/

-
••
f'v v
-·
.....,...._IIYUIIDtl
- ...
Figure 14. Evaluation of Tank No. 6 Ringwall Settlements

feet before the oil was transferred a few days As shown in Figure 13, maximum and minimum prin-
later. Subsequently, Tank No. 6 was loaded to cipal stresses in the tank shell ranged up to
25 and 30 feet of oil with careful monitoring of approximately 14,000 pounds per square inch
the soil and tank instrumentation. Figures 11, (psi) and 9,000 psi, respectively. These shell
12, and 13 show the oil loading of Tank No. 6 stresses were well below the allowable stress
and the observed behavior of ringwall settle- level of 21,000 psi for the steel tank although
ments, pore pressure and stress measurements. the effect of residual stresses remaining in the
tank shell before installation of the strain
Figure 14 provides a comparative plot of differ- gages was unknown.
ential shell settlements observed during this
portion of the monitoring program. The maximum
out-of-plane differential settlement during oil
loading of Tank No. 6 was 0.14 percent which CONCLUSIONS
compared favorably with previous maximum out-of-
plane differential settlement of 0.16 percent The behavior of Tank No. 6 and its foundation
and limiting criterion· of 0.22 percent. Maximum soils were analyzed to determine the cause(s)
differential shell distortion of 0.41 percent, for differential settlements observed in the
wh icb occurred on July 7, 1981, did not exceed northeast quadrant of the tank. The probable
the previously experienced maximum angular dis- cause for the differential settlements was the
tortion of 0.42 percent, but did exceed the presence of thicker, normally consolidated com-
suggested criterion of 0.35 percent. pressible clay and silt layer below the north-

141
First International Conference on Case Histories in Geotechnical Engineering
Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu
east quadrant of the tank as compared to the Geotechnique, Vol. 5, March.
soil below remaining portions of the tank. This
differ·e nce in soi 1 layer thickness and soil DeBeer, E.E. (1969), "Foundation Problems of
behavior may have been caused by siting a por- Petroleum Tanks", Annales de L'Institut Belge
tion of the tank over a filled river meander. du Petrole, No. 6, pp. 25-40.
The studies confirmed that the settlement pat-
terns and rate of settlements were generally Greenwood, D.A. (1974), "Differential Settlement
consistent with consolidation theory and that Tolerance of Cylindrical Steel Tanks for Bulk
the tank foundation soils did not experience a Liquid Storage", Proceeding of the Conference
bearing failure. on Settlement of Structures, British
Geotechnical Society, Cambridge, England,
Tank No. 6 was allowed to be used to store pro- April.
duct for short time periods with its behavior
monitored carefully. Previously experienced Jorgensen, L. (1934), "The Application of
levels of differential settlements were used as the Theories of Elasticity and Plasticity of
upper limit guideline to allow restricted use of Foundation Problems", Journal of the Boston
the tank. The unrestricted use of Tank No. 6 as Society of Civil Engineers, July.
a surge tank will require its releveling by mud-
jacking and preloading until the primary set- Seigel, R.A. (1978), "Computer Analysis of
tlements in the northeast quadrant are com- General Slope Stability Problems", Joint
pleted. Highway Research Project No. JHR.P-75-8,
Purdue University and Indiana State Highway
During preloading and performance monitoring of Commission.
Tank No. 6, differential ringwall settlements
were evaluated with respect to the best-fit Skempton, A.W. (1951), "The Bearing Capacity of
rigid-settlement tilt plane and compared with Clays", Building Research Congress, London,
published guidelines for allowable differential England.
movements. The following conclusions are drawn
from tbese observations. Su, H.L. (1976), "Manual of Practice, Estimation
of Consolidation Settlement", Transportation
(1) The suggested allowable out-of-plane Research Board, Special Report 163,
shell settlement of 0.22 percent ap- Washington, D.C., 25 pp.
pears to be a reasonable guideline for
maximum allowable out-of-plane move- Sullivan, R . A. and J.G. Nowicki (1974),
ments for floating roof tank shells. "Differential Settlements of Cylindrical Oil
The maximum out-of-plane shell Tanks", Proceedings of the Conference of
settlement for Tank No. 6 was 0.16 Settlement of Structures, British
percent. Geotechnical Society, Cambridge, England,
April, pp. 420-424.
(2) Tbe suggested allowable angular dis-
tortions of 0.35 percent appears to be
too conservative. Tank No. 6 has
experienced maximum angular distortion
of 0. 42 percent and has not shown any
signs of structural distress.

ACKNOWLEDGEMENTS
The autbora are thankful to the Department of
Energy Strategic Petroleum Reeerve personnel for
their support during the investigation and per-
formance monitoring of Tank No. 6 and their per-
mission for the publication of this paper.

REFERENCES
Bell, T.A. and J. lwakiri (1980), "Settlement
Comparison Used in Tank-Failure Study",
Journal o£ the Geotechnical Division,
American Society of Civil Engineers, Vol.
106., No. G'f2, Februuy, pp. 153-169.
Belloni, L • ., A. Garassino and M. Jamailkowski
(1974), "Differentia~, Settlements of
Pe,.roleuJD Steel (Tanks) , Proceedings of the
C.~mfel'enete_ on Settlement of ·S tructures,
·S .rtto:hh t;le'Oteccb,ntcal Sodety, .C.a ml)ridge,
&ng)l.a ud 1 A~i ~ .• ··
aisbnp, A.W. (19'55), "T.be\Jse of tbe Slip Circle
in the St:ability · ~ody'Bie of Slopes", .

142

First International Conference on Case Histories in Geotechnical Engineering

Missouri University of Science and Technology
http://ICCHGE1984-2013.mst.edu