1

FOUNDATIONS FOR THE NEW INTERNATIONAL AIRPORT IN BANGKOK, THAILAND

ASCE Seattle SectionGeotechnical Group

May 16, 2009

THAILANDBengt H. Fellenius

FOUNDATIONS FOR THE NEW INTERNATIONAL AIRPORT IN BANGKOK, THAILAND

Bengt H. Fellenius

A presentation to the ASCE Seattle Section 2009 Spring Seminar on

Recent Developments and Case Histories in Deep Foundations.

AbstractThe Suvarnabhhumi airport, the new international airport in Bangkok, Thailand, covering an area of 8 Km by 4 Km (8,000 acres), is located in a former swamp, a flat marine delta about 30 Km outside Bangkok. The soil profile consists of a thin weathered crust on typical soft to stiff, compressible Bangkok clay deposited on a sand layer at a depth of about 25 m extending to about 47 m. Below the sand lies an about 10 m thick layer of hard silty clay followed by very dense sand to large depth. Most of the area is devoted to runways, roadways, and parking, which required extensive ground improvement to minimize settlement. The structures consist of several units sharing footprints: terminal building, conourse, trellis structure, parking garage, and elevated roadways. The structures are founded on three types of piles installed to the sand below the clay layer, 1,000 mm bored pile, 600 mm diameter bored piles, and 600 mm driven cylinder piles. The stress-bulbs from the various foundations overlap resulting in a complicated settlement analysis. A total of 25,000+ piles were installed. The design of the airport started in 1995 and construction was completed in 2005 at a total cost of close to us$30 billion. The lecture will present aspects of the soil improvement work, analysis of results from pile tests, and the design of the piled foundations for capacity, settlement, and downdrag. The main part of the presentation consists of information quoted from papers published in Geotechnical Engineering Special Issue, Vol.37, No. 3, December 2006.

2

The New International Airport, Bangkok Thailand

Foundation Design by TAMS/Earth Tech, NY

with Dr. Bengt H. Fellenius as outside consultant

The presentation is primarily based on papers published in the Special Issue of the Journal of South-East Asian Geotechnical Society, "Geotechnical Engineering", December, 2006, as listed in the next slide.

Buttling, S. 2006. Bored piles and bi-directional load tests. Special Issue of the Journal of South-East Asian Geotechnical Society, December 2006, 37(3) 207-215.

Cortlever, N.G., Visser, G.T, and deZwart, T.P., 2006. Geotechnical History of the development of the Suvarnabhumi International Airport. Special Issue of the Journal of South-East Asian Geotechnical Society, December 2006, 37(3) 189-194.

Moh, Z.C. and Lin, P.C., 2006. Geotechnical History of the development of the Suvarnabhumi International Airport. Special Issue of the Journal of South-East Asian Geotechnical Society, December 2006, 37(3) 143-170.

Seah, T.H., 2006. Design and construction of ground improvement works at Suvarnabhumi International Airport. Special Issue of the Journal of South-East Asian Geotechnical Society, December 2006, 37(3) 171-188.

AND

Fox, I., Du, M. and Buttling, S, 2004. Deep Foundations For New International Airport Passenger Terminal Complex in Bangkok. Proceedings of the Fifth International Conference on Case Histories in Geotechnical Engineering, New York, April 13-14, Paper 1.22, 11 p.

33

44

55

Suvarnabhumi

International AirportAn 8 Km x 4 Km

area close to the Gulf of Thailand

66

Data from:

Fox, Du, and Buttling, (2004), Buttling

(2006), Moh

and Lin (2006), Seah

(2006), Cortlever, Visser, and deZwart

(2006)

Basic Soil Parameters

0

5

10

15

20

1,000 1,500 2,000

Density, ρt (%)

Dep

th (

m)

0

5

10

15

20

0 50 100 150

Water Content, wn (%)

Dep

th (

m)

0

5

10

15

20

0 25 50 75 100

Vane Strength, τu (KPa)

Dep

th (

m)

τu

Firm

Stiff to Very

Stiff

Very softto

Soft

wP

≈

30wL

≈

80 -

100

FILL

7

0

5

10

15

20

0 5 10 15 20

Virgin Modulus Number, m

Dep

th (

m)

0

5

10

15

20

0 50 100 150

Reloading Modulus Number, mr

Dep

th (

m)

Compressibility

Red Lines show distribution for designCircles are data points

CR = Cc

/(1+ e0

) m = ln10 (1 + e0

)/Cc

0

5

10

15

20

0.0 0.2 0.4 0.6 0.8Compression Ratio, CR

Dep

th (

m)

88

Comparison between the Cc

/e0

approach and the Janbu Modulus Number method

Data from a 20 m thick sedimentary deposit

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40 0.60 0.80 1.00 1.20

VOID RATIO, e0

CO

MPR

ESSI

ON

IND

EX, C

c

Do these values indicate a compressible soil, a medium compressible soil, or a non-

compressible soil?0

5

10

15

20

25

30

35

0.400.600.801.001.20

VOID RATIO, e0VI

RG

IN M

OD

ULU

S N

UM

BER

, m

The Cc

-e0

approach (based on Cc

) implies that the the

compressibility varies by 30±

%.

However, the Janbu methods shows it to vary only by 10±

%. The modulus number, m, ranges from 18 through 22; It would be unusual to find a clay with less variation.

99

0

100

200

300

400

500

600

700

1975 1980 1985 1990 1995 2000 2005

YEAR

SETT

LEM

ENT

(mm

)

Settlement of the Ground Surface Observed at Airport SiteRegional settlement occurs at and around the airport area due to

mining of ground water

Start of design work

10

Current and Future (long-term)

Pore Pressure Distribution

0

5

1015

20

25

30

35

40

45

50

0 100 200 300 400 500

Pore Pressure (KPa)

Dep

th (

m)

Long-Term

Short-Term (Current)

0

10

20

30

40

50

60

70

1975 1980 1985 1990 1995 2000 2005 2010

YEAR

Dep

th to

Gra

ounw

ater

Tab

le (

m)

DesignPhase

ConstructionPhase

Nearby Observations of Groundwater Table

Pumping (mining) of groundwater has reduced the pore pressures in the Bangkok delta resulting in significant regional settlement. In 1996, coinciding the beginning of the design process, pumping in the area was stopped. Pore pressure measurements indicate that the desired effect is being reached; the pore pressures are rising and the distribution may become hydrostatic

in the future.

1111

0

20

40

60

80

100

120

140

YEAR

WA

TER

DEP

TH (

m)

132a- 14m217 - 26m216a- 39m115 -153m209 -159m111 -161m501a-180m912 -206m114a-261m

618 -267m606 -301m501b-365m132b-442m114b-480m

1925 1935 1945 1955 1965 1975 1985 1995 2005 2015

SHALLOW WELLS

DEEP WELLS

The lowering of the groundwater table due to mining of water in the Bangkok delta is not unique. Below is a compilation of depth to

the water table measured in the San Jacinto-Houston-Pasadena area in Texas.

12

-83 KPa

105 KPa

34 KPa

85 KPa

105 + 34 + 85 = 224 - 83 141 KPa

110 m

38 m74 m

Briaud et al. 2007; Fellenius and Ochoa 2008

0

50

100

150

200

250

300

3501936 1946 1956 1966 1976 1986 1996 2006

SETT

LEM

ENT

(mm

)

YEAR

SETTLEMENT OF MONUMENT RELATIVE TO BENCHMARK

0

50

100

150

200

250

300

350

400

1 10 100

YEAR

SETT

LEM

ENT

(mm

)

1936 1937 1940 1945 1950 1960 1975 2000

LINEAR PLOTLOWER SCALE

LOGARITHMIC PLOTUPPER SCALE

1936 1946 1956 1966 1976 1986 1996 2006

13

0

50

100

150

200

250

300

350

400

1 10 100

YEARSE

TTLE

ME

NT (

mm

)

1936 1937 1940 1945 1950 1960 1975 2000

DEPTH TO WATER TABLE

SETTLEMENT

0

25

50

75

100

125

DEP

TH T

O W

ATE

R T

AB

LE (

m)

Monument only

Measured depths to water table and measured settlement of the Monument plus estimated settlement of the Monument had there been no drawdown of the water table.

1414

And in the San Joaquin Valley in California:

1977

1955

1925

Subsidence at San Joaqu in Valley, California

Devin Galloway and Francis S. Riley, U.S. Geological Survey

Approximate location of maximum subsidence in United States identified by research efforts of Joseph Poland (pictured).Signs

on pole show approximate altitude of land surface in 1925, 1955,

and 1977. The pole is near benchmark S661 in the San Joaquin Valley southwest of Mendota, California,

1515

Circles are preconsolidation data pointsDashed green line shows distribution for design“Short-term”

: Effective stress distribution at time of design“Long-term”

: Effective stress distribution after groundwater table is raised

0

5

10

15

20

0 50 100 150 200

Stress (KPa)

Dep

th (

m)

Long-term

σ'0

σ'c

Short-term

σ'0

Stress ProfileBut back to Bangkok:

1616

MainTerminal

Acess Roads

Concourses

17

-400

-300

-200

-100

0

100

200

300

400

500

6000 40 80 120 160 200 240

DAYS

SETT

LEM

ENT

(mm

)

Construction Period

Full Height = 3.0 m

Calculated Immediate Settlement

Calculated Total Settlement

Measured Total Settlement

HE

IGH

T (m

)

3.0

2.0

1.0

0

Calculated Consolidation Settlement

40 m 10 m

First a few words on the soil improvement work

Measured and Calculated Settlement at center line of for a 3.0 m

Embankment during 200 days. No drains, i.e., incomplete consolidation.

1818

Settlement at center line of a 3.6m Embankment on Wick Drains

AVERAGE MEASURED SETTLEMENT

DESIGN CURVE FOR THISSURCHARGE (75 KPa)

1.0 m

FINAL HEIGHTOF FILL

SETT

LEM

ENT

(mm

)

Wick Drains Installed

≈200 days

For reference, the curve of the 3.0m "undrained" embankment

19

Settlement and Horizontal Movement for the 3.6 m Embankment

Time from start to end of surcharge placement = 9 monthsObservation time after end of surcharge placement = 11 months

WICK DRAINS TO 10 m DEPTH

1.0 m

2.0 m

WICK DRAINS TO 10 m DEPTH

WICK DRAIN

Settlement was monitored in center and at embankment sides and horizontal movement was monitored near sides of embankment

2020

Horizontal Movement versus Settlement at Different Test LocationsH

OR

IZO

NTA

L M

OVE

MEN

T (c

m)

SETTLEMENT (cm)

2121

Lateral spreading

Settlement with risk for downdrag

The Problem

Piles

2222

2323

These photos of the bridge foundations illustrate a common problem affecting maintenance ($$$!), as well as, on occasions, being one compromising safety.

2424

The problem is not limited to bridge foundations

2525

Consolidation Coefficient, cv (m2/s)

Consolidation Coefficient, cv (m2/s)

Settlement after 4 months

as a function of the Consolidation Coefficient

and Drain Spacing

Degree of Consolidation after 4 months as a

function of the Consolidation Coefficient

and Drain Spacing

vv c

HTt2

=

SETTLEMENT (m)

CONSOLIDATION (%)

2626

The clay is soft and normally consolidated with a modulus number

smaller than 10.

All foundations —

the trellis roof, terminal buildings, concourse, walkways, etc.

—

are placed on piles. The stress-bulbs from the various foundations will overlap each other’s areas resulting in a complicated settlement analysis.

1,000 mm Bored Piles

600 mm Bored Piles 600 mm Driven Closed-toe Cylinder Piles 1,000 mm Bored Piles

600 mm Bored Piles

2727

Horizontal soil movement ("lateral spreading") toward piles can be critical

TRELLIS ROOF PILE CAP

FOOTPRINTS OF PILE CAPS FOR TERMINAL BUILDING NEAR TRELLIS ROOF

TYPICAL FOOTPRINT LAYOUT OF TRELLIS AND BUILDING FOUNDATIONS

Additional features, such as Embankments, Aprons, Concourse foundations, Area Fills, etc. adversely affect the piles and the piled foundations.

The stress interference between the foundations is significant and must be

considered in the design analyses

2828

To minimize lateral spreading toward adjacent foundations, some embankments were "supported" on soil-cement columns. For others, vacuum surcharge was employed together with fill surcharge.

Vacuum surcharge will cause the perimeter soil to move inward. Combining vacuum and fill surcharge can minimize the horizontal soil movement.

Vacuum surcharge can theoretically reach a stress of 100 KPa, but in practice, the maximum stress is about 60 KPa, equivalent to a fill height of about 3 m.

The final design employed a vacuum surcharge (considered tobe effective at 60 KPa) combined with an about 3 m surcharge fill (= 56 KPa) and a 0.9 m c/c

triangular drain spacing. The target time for 60 % consolidation was 4 months at which time the extra surcharge was removed to bring the degree of consolidation to about 85 %.

2929

3030

0

1,000

2,000

3,000

4,000

5,000

6,000

0 5 10 15 20 25 30

MOVEMENT (mm)

LOA

D (

KN

)

STATIC LOADING TEST600 mm, 35 m Long Bored Pile

0

5

10

15

20

25

30

35

40

0 1,000 2,000 3,000 4,000 5,000 6,000LOAD (KN)

DEP

TH (

m)

3131

Stage 1Lower Cell activatedUpper cell closed

Stage 2Lower Cell openUpper Cell activated

Stage 2Lower Cell closedUpper Cell activated

Data fromFox, I., Du, M. and Buttling,S. (2004)Buttling, S. (2006)

Bi-directional Static Loading Test, "O-cell" test —

1,000 mm Pile

3232

O-Cell tests

Downward movements during test phases 1, 2, and 3

Concern was expressed that the toe resistance (Phase 1) was ≈3,000 KN and the shaft resistance for the lower segment was ≈5,000 KN (Phase 2), while in Phase 3 the combined shaft and toe resistances were only ≈6,000 KN. Should not the Phase 3 resistance be ≈8,000 KN rather than ≈6,000 KN (i.e., the sum of the values ≈5,000 KN and ≈3,000)?

0

25

50

75

100

125

150

175

0 2,000 4,000 6,000 8,000 10,000

LOAD (KN)M

OVE

MEN

T (m

m)

Active CellInactive, Open CellInactive, Closed Cell

P1 P2P3

1 2 3

3333

Downward toe movements

0

25

50

75

100

125

150

175

0 2,000 4,000 6,000 8,000 10,000

LOAD (KN)

MO

VEM

ENT

(mm

)

Active CellInactive, Open CellInactive, Closed Cell

P1 P2P3

1 2 30

25

50

75

100

125

150

175

0 2,000 4,000 6,000 8,000 10,000

LOAD (KN)

DO

WN

WA

RD

MO

VEM

ENT

(mm

)

Active CellInactive, Open CellInactive, Closed Cell

P2

P1 and P2 data combined

P3

1 2 3

are best plotted per sequence of testing. Particularly when considering toe resistance, one must evaluate the load-movement response in comparing Phase 1 + Phase 2 to Phase 3 (i.e., P2 shaft below cell plus P1

toe).

3434

0

5

10

15

20

25

30

35

40

45

0 2,000 4,000 6,000

Load (KN)

Dep

th (m

)

Stage 1

101 mm toeMovement

0

5

10

15

20

25

30

35

40

45

0 2,000 4,000 6,000

Load (KN)

Dep

th (m

)

Stage 2

62 mm DownwardMovement

0

5

10

15

20

25

30

35

40

45

0 2,000 4,000 6,000

Load (KN)

Dep

th (m

)

Stage 2

Stage 3Max Load

Load Distributions

3535

0

5

10

15

20

25

30

35

40

45

0 2,000 4,000 6,000

Load (KN)

Dep

th (m

)

Stage 3

48 mm DownwardMovement

0

5

10

15

20

25

30

35

40

45

0 2,000 4,000 6,000 8,000 10,000 12,000

Load (KN)

Dep

th (m

)

Stage 2 Stage 3Stage 1

??

3636

The test data —

settlement data as well as pile test data —

were applied to

the design of the piled foundations employing the Unified Design Method.

3737

The Unified Design Method is a three-step approach

1. The dead plus live load

must be smaller than the pile capacity

divided by an appropriate factor of safety. The drag load is not included when designing against the bearing capacity. [The capacity of the pile toe should be defined from a movement criterion].

2. The dead load plus the drag load

must be smaller than the structural strength

divided with a appropriate factor of safety. [The live load must not be included because live load and drag load cannot coexist].

3. The settlement

of the pile (pile group) must be smaller than a limiting value. The live load and drag load are not included in this analysis. [The value(s) of acceptable settlement often governs a piled foundation design].

3838

Construing the Neutral Plane and Determining the Allowable Load

393939

A repeat: Distribution of unit shaft shear and of load and resistance

SHAFT SHEAR

DEP

TH (–)

(+)

LOAD

DEP

TH

Settling Soil

Non-Settling

Soil

1

0

5

10

15

20

25

30

35

40

45

0 500 1,000 1,500 2,000 2,500 3,000

LOAD and RESISTANCE (KN)

DEP

TH (

m)

0

5

10

15

20

25

30

35

40

45

0 10 20 30 40 50 60 70 80

SETTLEMENT (mm)

DEP

TH (

m)

NEUTRAL PLANE

TOE MOVEMENT THAT MOBILIZES THE TOE

RESISTANCE

SETTLEMENT OF PILE HEAD

TOE RESISTANCE

PILE "CAPACITY"

DEAD LOAD

DRAGLOAD

*) Portion of the toe resistance will have developed from the driving

*)

The Unified Method (typical example)

41

Force and settlement (downdrag) interactive design. The unified pile design for capacity, drag load, settlement, and

downdrag

0

5

10

15

20

25

30

0 2,000 4,000 6,000LOAD (KN)D

EPTH

(m)

O-cell

SiltSand

Clay

Till

Qd

Pile toe load in the load distribution diagram must match the toe load induced by the toe movement (penetration), which match is achieved by a trial-and-error procedure.

0

5

10

15

20

25

30

0 50 100 150 200

SETTLEMENT (mm)

DEP

TH (m

)

0

5

10

15

20

25

30

0 2,000 4,000 6,000LOAD (KN)D

EPTH

(m)

O-cell

SiltSand

Clay

Till

0

1,000

2,000

3,000

4,0000 50 100

TOE

LOA

D (K

N)

Qd

Pile toe load in the load distribution diagram must match the toe load induced by the toe movement (penetration), which match is achieved by a trial-and-error procedure. PILE TOE PENETRATION (mm)

Pile Cap Settlement

Soil Settlement

q-z relation

4242

Applied to the Bangkok Airport case

Several static loading tests on instrumented piles were performed to establish the load-transfer conditions at the site at the time of the testing, i.e., short-term conditions. Effective stress analysis of the test results for the current pore pressures established the coefficients applicable to the long-term conditions after water tables had stabilized.

A total of 25,400+

piles were installed.

4343

The extensive testing and the conservative assumption on future pore pressures allowed an Fs

of 2.0. The structural strength of the pile is more than adequate for the load at the neutral plane: Qd

+ Qn

≈

1,500 KN.

Example of calculated resistance distribution for 600 mm diameter bored pile installed to a 30 m embedment depth.

0

10

20

300 1,000 2,000 3,000

LOAD (KN)

DEP

TH (

m)

Qn = 770 KN

Qd = 1,040 KN RULT = 2,870 KNFs = 2.0

Short-Term

Fs = 2.0 on long-term capacity

0

10

20

300 1,000 2,000 3,000

LOAD (KN)

DEP

TH (

m)

Long-Term

Qn = 500 KN

Qd = 1,040 KN RULT = 2,080 KNFs = 2.0

Clay

Sand

4444

The settlements for the piled foundations were calculated to:

Construction

Long-term

TotalTrellis Roof Pylons

20 mm

90 mm

110 mm

Terminal Building

30

15

45

Concourse

35

20

55

* * *

4545

Contour lines of settlement of ground surface and pile caps near

a trellis roof pile cap

4646