Geotechnical Aspects of Underground Construction in Soft Ground, Mair & Ta y/or (eds)

Settlement effects of bored tunnels

R.J. MairGeotechnical Consulting Group, London, UK

i

© 1996 Balkema, Rotterdam. ISBN 90 5410 856 8

ABSTRACT: This Report reviews 35 papers submitted to the Symposium which concern settlement effects of boredtunnels. The papersare reviewed in three groups: open face tunnelling; closed face tunnelling; and effects on structures,services and tunnels.

l. INTRODUCTION

Ground mov.ements and ,settlements are inevitablycaused by bored tunnel construction in soft ground. Inurban areas. the potential effect of settlement is animportant consideration in the design and construction oftunnels, and is the subject of Session 5 of the Symposium.In this Report, the papers in Session 5 are divided _into threegroups:

l. Open Face Tunnelling

2. Closed Face Tunnelling- Slurry Shields- Earth Pressure Balance (EPB) Shields

3. Effects on Structures, Services and Tunnels

Relevant papers concerning settlement effects ofbored tunnels from other Sessions of the Symposium arealso reviewed.

The volume loss, sometimes referred to as groundloss, is a useful and simple means- of quantifying themagnitude of ground movements. It is defined as the totalvolume ofthe surface settlement trough per metre length oftunnel (i.e the cross-sectional area) divided by the cross­sectional area of the tunnel. This parameter is quoted inmany of the case-history papers in the Symposium.

2. OPEN FACE TUNNELLING

2.1 General

Open face tunnelling describes the types of tunnelconstruction where the face is effectively unsupportedduring the excavation process. This means that the face isgenerally stable without support, at least in the short term,and therefore most of the papers placed in this category,listed in Table 1, concem mainly stiff clays or even more

43

competent ground. Many of the papers describe casehistories of tunnelling in London Clay.

Table l lists sin( papers from Session 5 falling into thecategory of open face tunnelling, and there are another five

Table 1: Papers on Open Face Tunnelling

Construction Ground VolumeMethod Conditions Loss (%)

Session 5Bowers et al SCL (NATM) London Clay 1.1-1.5

Kimmance et al SCL (NATM) London Clay

Marshall et al Shield + Pipejack London Clay

Oteo and SCL (NATM) Stiff clays, marlsSagaseta, Shield + Segments

Umney and Shield + Segments London Clay 1.5-1.8Heath

Standing et al Shield + Segments London Clay 2.9-3.3

Other Sessions

Egger SCL (NATM) Gypsum-bearingclaystones

Farias and Assis SCL (NATM) Porous collapsibleresidual soil

Grose and Eddie SCL (NATM)Shield + Segments 0.5

London Clay 1.0

SCL (NATM) Weak rocks 0.2Kavvadas et al

TBM + Segments ("Athenian Schists")

Negro et al SCL (NATM) Stiff/hard clays,Shield + Segments Medium dense/

dense sands

SCL (NATM) = Sprayed Concrete Linings (New Austrian Tunnelling Method)

papers from other sessions. These other five also containinteresting data on settlement measurements. Theconstruction method described 'in each of the papers islisted in Table 1, together with the ground conditions. Theuse of sprayed concrete linings (SCL), often referred to asthe New Austrian Tunnelling Method (NATM), forms thesubject of the majority of the papers. Where the volumeloss is quoted by authors the values are shown in Table l.

2.2 London Clay

More 'than half of the papers concem tunnelconstruction in London Clay. Standing, Nyren, Longworthand Burland describe the high quality instrumentationinstalled on two sites for the construction of the JubileeLine Extension' project in London. One site involves EPBtunnelling in dense sands. The other site involvestunnelling with a mechanized shield and segments inLondon Clay. The paper gives a detailed account of theinstrumentation used and some ground movement data forthe London Clay_ site. A _particular feature of theinstrumentation is the use of electro-level inclinometers.Unusually high volume losses of 3.3% for the first tunneland 2.9% for the second tunnel were obtained. The reasonfor this is not yet fully understood. The authors also presentvaluable subsurface settlement data.

Subsurface settlement data are also the subject of thepaper by Umney and Heath. The data were obtained duringconstruction of a 5.4 m diameter tunnel in London Clay.The tunnel was excavated using a mechanized shield andlined with segments. Although it must be recognized asapproximate, the volume loss deduced by the authors fromthe subsurface data is l.5%. At another location describedby the authors, where subsurface data were also obtained,the deduced volume loss is stated to be 1.8% for a 7.8mdiameter station tunnel.

Subsurface ground movement measurements were alsopresented by Marshall, Milligan and Mair in their paper ona 1.8m diameter pipe jack in London Clay. Theconstruction method was hand excavation within a shield.The pattems ofrsurface and subsurface ground movementdata were found to be very similar to those obtained forlarger tunnels in ‘London Clay. Changes in ground stressesdue to construction of the tunnel are also presented, andthese are compared with simple elastic solutions.

Grose and Eddie describe tunnel construction inLondon Clay at Heathrow. The paper quotes a maximumvolume loss of 1% for a 4.5m diameter tunnel constructedwith a mechanised shield and expanded linings. Some ofthe length of tunnels were constructed using sprayedconcrete linings (NATM) and the quoted volume loss in thepaper for these tumrels is 0.5%. This is significantly lowerthan obtained for the Heathrow Express Trial Tunnels(Bowers et al) using sprayed concrete linings.

Bowers. Hiller and New describe the HeathrowExpress Trial Tunnels in London Clay and focus on thepost-construction settlement measurements. Three differentexcavation sequences were used, all employing sprayed

44

concrete. The equivalent diameter of the tunnel was 8.7mand its depth to axis was 2lm. The volume losses obtainedat the time of construction were between 1.1 and l.5%. Thetechnique of sprayed concrete linings, which was its firstuse in London Clay, clearly demonstrated that good controlof ground movements was possible with the method.Figure 1 shows the surface settlement measurements aboveone of the sections of trial tunnel taken immediately afterconstruction and 3 years later just before installation of the_secondary concrete lining. The line through the longer termdata shown as Equation 4 corresponds to an equationproposed by the authors to characterize post-constructiontime-dependent settlements. It can be seen that there is arelatively uniform settlement increase across the wholesettlement profile in the 3 year period. In particular itshould be noted that there are only very small increases indistortion or deflection ratio. This has importantimplications for long term settlement effects on buildings.Figure 2 shows the short and longer term horizontal 'strains at the ground surface above one of the' tunnels. Bothsets of measurements are in reasonably good agreementwith the derived horizontal strain predicted from theclassical Gaussian distribution based on the short-termsettlement data. As in the case of distortion or deflection

0lf--._4;___..;p-~ n;-5 _ .--1- ,_ O"A I+/ ,,CE .10 _ ,vE *. ,fr fjj -15-- 1* Q*5 +`~l- -5-"ii /E -20 _ ~‘\ \"- "I !.m \ -`._ _FBI ,

E -25 - 5* 7/@ -30 ‘ \ /D \`. /,’-35 - "g,_40 4 |- ` | r | | | | - r-15 -10 -5 0 5 10 15 20 25 30 35

Horizontal offset from centreline (m)o Long term data - - Equation 4+ Short temr data ------ - Fitted Gaussian

Fig. 1. Immediate and post-construction surfacesettlements above tunnel in London Clay (Bowers et gl)

o.oo1o f _ _ ` _/ 7 ’ ` ` `o.ooos - » ` \ ‘A /// X _ g ¥ \ ` ` 52 o_oooo - o/ '* /C /

-0.(X1}5 " Q / QC //Q -0.0010 - >,gC X 1.~- /g -0.0015 - /U7 /

-0.0020 E , /-0.0025 ' » » v ~o 5 10 15 20 25 30Horizontal offset from centreline (m)

0 Short term data >< long term data- - Derived strain (see text)

Fig. 2. Immediate and post-construction horizontal strainsabove tunnel in London Clay (Bowers et al)

ratio for the settlement profiles, there is very little changein horizontal strain in the 3 year period. This too hasimportant implications for buildings.

2.3 Other stifclays, marls and weak rocks

, Oteo and Sagaseta present data of horizontal groundmovements at the surface, as well as settlements. The casehistories given are for tunnels in Madrid, in which theground conditions are primarily stiff clays and marls. Onecase history involves the use of sprayed concrete linings,but the remaining case histories are for shield tunnels. Theauthors compare the data with results of elastic finiteelement analyses and with solutions for incompressiblefluid flow.

Kavvadas, Hewison, Laskaratos, Seferoglou andMichalis describe tumiel construction for the Athens metro

in weathered rocks, known as "Athenian schists", usingsprayed concrete linings and also TBM construction withsegments. The "Athenian schists" are a complex sequenceQf thinly bedded clayey .and calcareous sandstones,altemating with siltstones, slates, marls and limestones.Sprayed concrete lining techniques were used to _constructthe 16.5m wide station tunnels; the maximum volume lossfor these is stated to be 0.2%. On one of the TBM drivesface instabilities occurred, Withlarge settlements occurring.The authors highlight the difficulties of sampling andcharacterizing the behaviour of complex weathered rocksin what in reality is a soft ground tunnelling project.

Negro. Sozio and Ferreira present comprehensivedata on 25 case histories of tunnels in Sao Paulo, Brazil.Most of these were constructed with sprayed concretelinings although some were constructed with shields. Themajority of the tunnels are in over consolidated tertiarysoils, which 'are principally stiff/hard clays or mediumdense/dense sands. The volume losses generally obtainedwere between 0.5 and 2%.

2.4 Unusual Soils|\

Another tunnel iri‘.Brazil, this time in Brasilia, isdescribed by Farias and Assis. This tunnel, 9.6m indiameter and constructed with sprayed concrete linings(NATM), is in a porous collapsible residual soil.Instrumentation was installed to measure subsurface andsurface movements. Figure 3 shows the very unusualpattem of settlements obtained above the centre line of thetunnel. In most cases, less settlement is obtained at theground surface than at the tunnel crown, but in this case thereverse was measured. A ground settlement of 70mm wasdeduced at the tunnel crown but at the ground surface amuch larger settlement of 170mm was measured. This is afeature of the particular porous collapsible clay.

Unusual ground movement behaviour is alsoreported by Egger in another difficult type of ground, thistime in Stuttgart, where the tunnel was constructed ingypsum-bearing claystones ("Gipskeuper"), which wereleached near the ground surface. A 16m diameter three lane

0

,_4­

@6­i _8 _

10'12'14 0 50 1(1) 150 2(1)

Settfenent (rrm)

Fig. 3. Surface and subsurface settlement above tunnel inporous collapsible residual soil (Farias and Assis)

E 6E 4 E 5 O15010050 o .. _. 501;. 1 ‘_ | mm

§1[mmi E E 6 eu5 - | 5 II I gI t :3: l : 4 a 'l i' 3 lI ‘ A\ T - - E I

\ __ -I -:L-:-gg: '-:;:;.;.¢.;.: -.-.- {

it 7 i'l 30.3.89 i - --- 7-12-87E -9.00 E -30.3.39if i '

il

Fig. 4. Settlements above tunnel in leachedgypsum-bearing claystones (Egger)

road tunnel was constructed using sprayed concrete linings(NATM). Figure 4 shows that the ground movement dataobtained by the instrumentation indicated almost constantground movement with depth, and indeed considerableground settlements to the side of the tunnel and beneath it.The explanation given by the author for these unusualmeasurements is the consolidation of the leached grounddue to drawdown of groundwater caused by the tunnellingoperation.

2.5 Surymary

The following key points can be concluded from thepapers on open face tunnelling:

l. In London Clay the volume losses obtained are

generally between 1 and 2%. An exception to thiswas the significantly higher ground losses of around3% obtained at the St James's Park measurementsite for the Jubilee Line Extension project.

2. Construction with sprayed concrete linings (NATM)is effective in controlling ground movements in stiff.clays, including London Clay. This may not be ofgreat surprise to those who have long experience ofsprayed concrete linings, but their use in London isrelatively recent and there was initially somescepticism as to how effective the technique wouldbe in controlling ground movement.

3. Significant post construction settlements may occurabove tunnels in clay, but in the case of LondonClay measurements indicate that there are only verysmall increases in deflection ratio and horizontalstrain.

4. Large settlements can occur above tunnels inunusual ground such as gypsum-bearing orcollapsible soils.

3. CLOSED FACE TUNNELLING

3.1 General

.Closed facetunnelling methods are used in unstableground conditions where the face requires support at alltimes. This principally applies to permeable ground belowthe watertable (i.e. mainly sands or mixtures of sands andclays) or soft clays. '

Table 2 lists 8 papers from Session 5 on closed facetunnelling, and there are another 6 papers from othersessions. The case histories of slurry shield constructioninvolved tunnelling mainly in sands (with the exception ofthe paper by Hou, Liao and Zhao dealing with soft clays inShanghai). Earth pressure balance machines are used moreuniversally for all types of unstable ground conditions, ascan be seen from Table 2. Where the volume loss is quotedby the authors (or where there are sufficient data tocalculate it) the values are shown in Table 2.

3.2 Slurry Shields

Kastner, Ollier and Guibert describe comprehensiveinstrumentation for three experimental sections for twin6.3m diameter slurry shield tunnel construction for theLyons metro. The tunnels are in alluvial deposits, whichmainly comprise fine silty' sands and soft clays, as shownin Figure 5. The development of ground settlementsimmediately above one ofthe tunnels (VI) is shown inFigure 6. The authors identify four components ofmovement:

1. Approach of the shield2. Passing of the shield (there was a 30 mm bead)3. Passing of the tail ofthe shield and grouting of the

linings (heave is detectable)4. ConsolidationIt is relevant to note the slow progress rate for this

Table 2: Papers On Closed Face Tunnelling

Construction Ground Volume LossMethod Conditions (%)

Session 5

Atahan et al SS Sands/gravels `1.5Dyer et al EPB Loose sands 20Hashimoto et al EPB Soft claysHou et al SS Soft clays <2.0Hwang et al EPB Silty sands/

soft clays

Kastner et al SS Sands/clays 0.8Moh et al EPB Silty sands/ 1.3

soft clays

Simic and Gittoes EPB Sands/ 0.8-1.2soft clays

Other Sessions

Forbes and Finch EPB Firm claysKuzuno et al SS Sands/claysLinney and EPB Dense sands/ 1.0Friedman stiff clays

Phienwej EPB Soft clayShirlaw et al EPB Dense sands/ <l.0

silts

Steiner SS Dense silts/ 0.4sands/gravels

SS= Slurry Shield, EPB= Earth Pressure Balance

§Il'l

lfll'IRI'|51

I I5RetainingWall '

ll

wg

'EXl7&ll7

;oEi”

OO‘ | 4 -158 z° £30 <> I*IEEE! l

IEEE-5911-Us C' um $3 1

lili' imma! 32% f '64°o 00° G cS°o °oo o oo :Q QQ °°o oo O O ooo' oO Q Q Oo o 0 0,_Q

0 I,L un _ a '5'; m ‘i.'ll .. . r - ‘ (\L_83it -7 .; *r_'T ; 7_7 _; .4_4 9_6m_ _g _T L.; - .L..I. - _L ' T7 ' : : I .l....I_Ié" g _ g j§§_ ;1-=~= l _ "‘-ji"1-2 = - minL-;::i§€§§§€§;§:1~ r -' -°.' ~.-.__ _ -.-.-.-._.-.-.- 1| _.­__l1 #gm .W 1 3/ , y , _y

.=;Q.-Q._'Q .,.;Q_‘__,?-51__.{Q§_.1_ .1'Q'r' .drg 355% `.5."°, g" 6:;,§_w'1__ _.’:__ J':__ J‘1__ _,'1__ _,':__ J °:__ ;°: _ ;'1___ , °1__, Pm: `egg §;@'.;'-°~ 525;-°-%5£§f§E£g§ Q15 ¢5s5§QI- ’ .-?-697-és' Q:-6 Qt-5' Q:-§ 97- = ” QFD' _ 1- ' Q:-Q”

Fig. 5. Instrumented section for slurry shield construction(Kastner et al)

particular tunnel. This was near the beginning of the wholetunnel drive, and 78 hours lapsed during the passage of theshield past the measuring section. Even including thesubsequent further consolidation settlements, the volumelosses (calculated in terms of the observed surfacesettlements) were found to be 0.8% at the start ofthe tunneldrive, decreasing to values well below 0.5% later on in thedrive. This illustrates how well ground movements can becontrolled by a slurry shield.

Extensometer EX11 (P1-Sl)

r

I 5/U7/93 I 7/07/93 l 9/07/F3 2| /07/9] 13/'07/930 1 e ' 1 A1 ' ' 1 ' 1i f` -2.E IE " TE I 3

`_ u -61 S­5 - 55 _g 4. on ' +|lI'|'|U Gm o_,Z r all 'STime (days)

|- depth 9.6 m - depth 12,3 m 'v-Surface]

Fig. 6. Development of surface and subsurfacesettlements above slurry shield (Kastner et al)

Atahan. Leca and Guilloux describe aninstrumentation project for construction ofa 3.4m diametertunnel constructed with a slurry shield, at shallow depth inalluvial sands and gravels below the water table,.as shownin Figure 7. The segmental linings were also instrumented.The maximum volume loss expressed in terms of theobserved surface settlements can be calculated asapproximately 1.5%. The authors also give data onhorizontal ground movements at depth, and theydemonstrate a relationship between ground movements andface pressures.

A much larger slurry shield is described by Steiner.An 1l.6m diameter tun_nel was constructed in dense silts,sands and gravels below the water table for the Grauholztunnel in Switzerland. The tunnel was at very shallowdepth (approximately 12m to axis level) at the section forwhich a full surface settlement profile is reported. Thevolume loss obtained was 0.4% or even less. To achievegood control of ground movements the author emphasisesthe need to carefully control face support and the groutingof the tail void. _ '

Kuzuno. Takasaki. Tanaka and Tamai describe aremarkable method o_f construction of a 17m wide stationtunnel complex close beneath a building in clays and sandsbelow the water table, as shown in Figure 8. A triplecircular face slurry shield machine was constructed for thispurpose, so that the entire cross section was simultaneouslyconstructed. The ground above the tunnel and the buildingwere well instrumented. The maximum settlement of thebuilding amounted to only 7 mm, which is an excellentachievement.

Hou Liao and Zhao describe a slurry shield of l 1.2mdiameter in the soft clays of Shanghai, China. Details oftheground movement measurements are given and the volumeloss obtained was less than 2%. The paper also presents acomparison of the slurry shield with a previous tunnel ofsimilar size -in the same soft clay, using a face grid tomaintain stability. In this latter case the volume loss wasabout 15%. The authors identify the same four componentsof ground movement as identified by Kastner et al, listedearlier in this Report.

rl):

E1-I1 E2-I2 E3-|3E4-I4 ' .0 _North , _ _ _ South 3.3 Earth Pressure Balance Machznes7 S _ = ‘ r! _Mannqpade }~_215m l 2 15m Moh, Ju and Hwang describe a detailed'-- Q ___~_-;f-31 _ _. it instrumentation project for a 6m diameter tunnel partly rn' " _.: ' -!, l ` . » - . I

2.5 -- _ _ " 7 " " soft clay but prrncrpallyrn srlty sands below the water table~ _ (and at a depth of 18.5m to tunnel axis level) for the Taipei/ S" ' _ ~ Rapid Transit System. The instrumentation was designed

5 to measure surface and subsurface settlements. The authors__1_ _ . _"V Sam' 1Ei§§§5E5E5E§35§3E3§3E§3§§5E§E§§3E§E55525 §§5E§E§§§E§25§5I»“ ‘ \ f"“‘“w>‘==sw"‘en'»

réééié ig; ‘§ 1 i 3? 7 ‘@§§§¥§ rt; <§;§ 5 ¥¥§§§r;§;r§§%;_ U U ~ Lisis; 1 § §.§‘§ '§§>_<§ 5 7 ,, ,,- G- ° C," ` ° 5 - 5555534E ' §§§i§;£; i i § l ei ‘F i igiiiiiéi IMP Bldg (Bl level S~_, 7.5'-- hs ‘d Sl = 'ii § ,;;;;§§§§ it 3 Ll LJ 'f'= SENSE. ia we a ll Q g .§§i@e§;§» ° ' c l% §§‘;=§‘»§f§§§‘§§;§ °. >§§ yi §§r§;;§;§§`§ E C°“°°`"S° ,faQ ggiglié -f ps 5 f , _. @§§ “ ;§§§i;§§;;§ v '5' m, `>:§ _i= ' 1 >> ?=3fz§= ~ ' » ii » 25353 gg ­§;.-\ 5 ,; sg 5.5.3 § §s,,,,;§, ~<r.>faf§=§=§ §§ 3. ! \ gag iriliiiigg _ .§§;;¢;=§<;§ o _ g»1||\;2| '­10_|_ §§¥‘§ _. r is§§ _____ r ____ W .s _. as=uz._¢» _ _ xrsrrerhi _ -5 I0 Dnwal' eeeeeeeaeraaeaaaeaf Qeeeieeieggge Q iii ch ' ~-4* '-'-'2'f'¢'f'1*i5§:¥ ,` ?-:=~k¢*’ 1%-.-. --= _ _,, .,¢j§¢','¢,.&._q;<<.5.__ _§,; _.__.______ __,4,_,,.,_ ,_ t __ WL --1| ~ -zo 57532 Qmw' as12 5L " (0 _5) fe-ef; F' . ~ i ift5:3:5:5Z225:5i7cZ:25§L¢'25'£¢'.?.5.'5L’ ‘Z$:5§:25i Z%“*» »;&1 §5:1 3 3V Semi Oven -25 _. . O- - 1 2,.§.,. -.._ ¢,_,,¢,,_¢,~;.w,.».Qp»..,,..,_.,Wap. ..~._A.._, ,,..a.1%._2_ F-=1... I)||uv|~|\ L0.1 limestone ‘ ­day \ A A _ "

I _ .1 :::;zz=:¢.~<¢;»»,;.;a.:;-z;:»¢;f.: -és#-;,~:.;.<; ~" f11“=:'.@'l":~‘~':;¢., -35- l7. 000 ( ml- mm)i (01-5)Fig. 7. Instrumentation for slurry Shield 11101161 F ig. 8. Use of triple circular face slurry shield machine(Atahan et al) 1Kuzuno et al)

identify similar components of ground movement asdescribed by Kastner et al for slurry shield construction.This is illustrated in Figure 9, from which it can be seenthat some settlementidevelops as the shield advances pastthe measuring point, and then significantly more settlementoccurs as the tail passes, followed by a consolidationcomponent. In this case the volume loss associated withtunnel construction (ignoring consolidation) was found tobe 1.3%. It was also found that there isa substantiallywider trough width at depth (measured in terms of thedistance to_the point of inflexion) than would be obtainedby applying a constant trough width parameter K to givei = K (zo -z), as .shown on Figure 10. This conclusion issimilar to the results of recent analysis of measurements ontunnels in clays (Mair et al, 1993), which shows theparameter K to increase with depth below the groundsurface.

Also for the Taipei Rapid Transit System, Hwang,Moh and Chen present interesting and detailed porepressure measurements for two tunnels, one in silty sands,the other in silty clays. The authors demonstrate that thepredominant cause of excess pore pressures induced in theground is the grouting for filling the tail void. This hasimplications for the control- of subsequent consolidationsettlements for tunnels in clays.

Hashimoto. Hayakawa, Kurihara, Nomoto, Ohtsukaand Yamazaki review Japanese experience with earthpressure balance machines in soft -alluvial groundconditions and compare. these with slurry shields. Astatistical study appears to show that greater surfacesettlement tends to be associated with earth pressurebalance machines than with slurry shields, but this may bebecause the former tend to be more often employed inalluvial clays. Their paper, which is a sequel to the paperby Nomoto et al (1995), deals with various grounddisplacement mechanisms, and compares fieldmeasurements with data from centrifuge model tests.

Phienwe|` presents a review of experience ofconstruction of small diameter tunnels (between 2 and3.5m) using earth pressure balance shields and othertechniques in the soft clay of Bangkok, as well as in thestiff clay and sand strata. Ground movement measurementsare presented.

Days afier the Passing of the Head

TailPassing0.1 I io |000| =' ' 5 ' li '~ |‘lI,|= wVl2'~! . i n.51 'F-|||.=i ,§g:l,i“"i! 41 |- .- .. _ 'fl-if' "E '° l ~- 11 e=:|@|D !E "5 _ I. 'li '||| liiiE -zo I I i_ .i .I I§ _ZS | | Depth§ .Jo "" 1 -__i i i’0mV’ 1 ?E~ !=f1|;|| `°T°“°`%b-ees5m-as L fii "iz: A i 95m» : .i iii I, | - 1 'i l4`5m5 Sh` ld Adv ing I Tail Void ‘ Consolida `

Fig. 9. Development of surface and subsurface settlementabove EPB shield in silty sands (Moh et al)

lb

20.00 y

er, m

M K T15.00 - 1' Z/

/I

I

I// Z0/

/I

/500 /

I/

Cent

T, .§ 10.00 - .

ce toTallcal Distrti

-1 Clough- -~ - O'Reilly*'This_ Studygn |||||!_-4 468g 10 12

-5.00 ­Transverse Distance to Point of Infiection, i, m

Fig. 10. Variation of trough width above EPB tunnelconstruction in silty sands (Moh et al)

Linney and Friedman describe earth pressure balanceshield construction in the very stiff clays and dense sandsof the Woolwich and Reading Beds in London. Thetunnels, 4.8m in diameter, were part of the Jubilee LineExtension project. The volume loss was estimated to beabout 1%. The authors highlight the need for good controlof face pressures and steerage of the shield in order toproperly control ground movements.

Simic and Gittoes describe construction of a 9.7mdiameter tunnel for the Lisbon metro in Miocene sandsbelow the water table. The volume losses obtained werebetween 0.8 and 1.2%. The authors give details of theirapproach to settlement prediction when_ tunnelling in softalluvial clay, and the method adopted for assessment of riskto buildings. `

Dyer, Hutchinson and" Evans present subsurfaceground movement measurements made during constructionof a 1.4m diameter tunnel in loose sands below the watertable. An earth pressure balance shield combined withpipejacking was used. As shown on Figure 11 they foundan increase in the trough width parameter K with depth, asfound by Moh et al for the tunnels in Taipei sands, whichis very similar in principle to the behaviour of tunnels inclays (Mair et al, 1993). The authors highlight the largevolume losses obtained as a result of the jacking of the"pipe, causing the loose sands to compact. .Tacking of the'pipe resulted in an additional volume loss of 14%, leadingto a total volume loss of 20%. These large volume lossesare partly a result of the very small tunnel diameter beingused in the calculation.

K=i/(Zo-Z) 3.4 Summary0 0.2 0.4 0.6 0.8 1 1.20 _ | | | '| | | The following major points emerge from the papers

on closed face tunnelling:| 02 ' 1. There is potential for a high degree of settlement

O 0.4NR’ 0.6 - 2.

0_8 ­

I..

Fig. 11. Variation of trough width settlement parameterwith depth for tunnel in loose sands 1Dyer et al)

Shirlaw. Pennington and Yi describe construction ofa 3m diameter tunnel in dense sands and silts below thewater table in Toronto, Canada using an earth balancemachine. Figure 12 clearly demonstrates the "leamingcurve" experiencedduring the early stage of construction.After about 100m of the' tunnel u had been driven, verysmall surface settlements were_obtained, corresponding tovolume losses of between 0 and 1%, but at the beginning ofthe tunnel. drive very high settlements were obtained,corresponding to volume losses in excess of 10%. Thishighlights the importance of getting the leaming curveunder control rapidly with earth pressure balance machinesbefore tunnelling beneath zones where settlement controlis critical. , p

Difficulties with the learning curve are also' describedby Forbes and Finch in their paper on the St Clair RiverTunnel. A 9.2m earth pressure balance shield was used infirm clays. It was found that the settlements were verysensitive to the face pressures. In one instance 25mm ofheave was obtained, whereas 85mm of settlement occurredat another location, but in general good settlement controlwas achieved. The authors refer to the initial difficulties incontrolling the face pressures.

` ` _ . as -5-46 48 _ 50 53 67-700 I 36 "" . 4- ' N24 28 37 ` ‘7 1° 54 59 5°

-25 -_.___._i6 ...... _-sz .......... _ ...... ___._- .__ ...... __ ._ .... _ ............... _.-§- ............... _ .... _.-.___. ; i _.50 __.-____ ..... _._...l.. ._.........._.. 1_ _ | I¢'\ 1 3E i . _5 _75 ___.12 _ .__. _.____._. .._.._ _ ._ _......._.._....._ _Q ...._..._._..._..._. -.-.-._.‘E _ V _ QE -100 .___ ___... _._ ._.__. LEGEND: _-_2 .3 ,125 _ ._'..._. _ _.._._ ° Da|a point .... _

.150 - ......_ as rdenrinearion nurnber3 of setllemenl poinl.115 ...._..._..... ‘ __..___.___.._... ______.___..._.._. ._._5 I 3 . _I s _ e2000 100 200

Chainage (m)

300 400 500

Fig. 12. Surface settlements above EPB tunnel in verydense sands and silts (Shirlaw et al)

control, and volume losses are often as low as0.5%.

Provided that there is a high degree of control onface pressure, the principal causes of groundmovement are (a) the shield passage (which maylead to settlement through closure of the soil ontothe shield due to the overcut, and also shearing ofthe soil may lead to settlement), and (b) the tailvoid. An additional cause can be grouting of thetail void causing excess pore pressures in theground.

3. A key element in control of ground movementswith slurry shields or earth pressure balanceshields is the provision of the appropriate facepressure. It is clear that there is often a significantlearning curve in achieving this, particularly in thecase of earth pressure balance shields.

4. EFFECTS_ ON STRUCTURES, SERVICES ANDTUNNELS

4.1 General

There are seven papers to Session 5 which relate tosettlement effects of bored tunnels on structures, servicesand other tunnels, and four papers from other sessions.Excluded from this group of papers are those concemingthe use of ground treatment, such as compensationgrouting, to protect structures from the effects oftunnelling. These are reviewed in the report for Session 3.

4.2 Structures

Forth and Thorley present a case history of the effectsof tunnel construction in Hong Kong on a piled building.Two 7.9m diameter tunnels were driven adjacent to thepiles using open face tunnelling shields and compressedair. The ground was completely weathered granite, belowthe water table. The maximum recorded settlement of thebuilding at the side nearest to the tunnels was 12 mm,presumably largely due to negative skin friction effectscaused by the tunnelling. There are generally very few casehistories in the literature of the effects of tunnels on piledbuildings, and there is a need for more measurements.

An unusual example of a structure affected by a tunnelis given by Eggg. A 10m diameter tunnel was constructedin weathered claystones using sprayed concrete beneath anice skating rink in Stuttgart. The maximum settlement wasl7mm, and the resulting settlement profile was equivalentto a volume loss of 0.5%. Unlike many structures, the icerink behaved in a fully flexible manner. Nevertheless it wasparticularly sensitive to differential settlements, and verycareful control was exercised in construction of the tunnel.

A methodology for the assessment of risk of damageto masomy buildings from bored tumrelling is presented byMair, Taylor and Burland. If a building is fully flexible, itwill confonn to the' "green field site" settlement trough asdepicted in Figure 13. The deflection ratios (Ah/Lh orA S/LS, corresponding to the hogging or sagging part of thestructure), together with the horizontal strain induced in thebuilding will determine the potential degree of cracking inthe masonry. Having estimated the deflection ratios andthe horizontal strains likely to be induced in the building,the potential damage can then be assessed, based on theclassification of damage to masonry developed at theBuilding Research Establishment in the UK.

The important assumption that is frequently made inassessments of potential damage is that the buildingfollows the profile of the ground as if it were a green fieldsite. This is in many cases conservative, because buildingsfrequently do not behave as perfectly flexible structures.The actual settlement profile of a building can beconsiderably wider and shallower than the corresponding"green field site" Gaussian settlement trough, withcorrespondingly much lower deflection ratios anddistortions.

The difference . "between the actual buildingperformance and the' predicted performance based onperfectly flexible behaviour is of major importance and isthe subject of a collaborative research programme beingundertaken during construction of -the Jubilee LineExtension project, described by Burland, Mair, Linney,Jardine and Standing in Session 6 of this' Symposium.

The influence of building stiffness on settlementprofile has been studied parametrically by Potts andAddenbrooke using finite element analyses. About l_00analyses were undertaken of different configurations oftunnel and building dimensions. Two importantdimensionless parameters are introduced by the authorsi

Hogglng SagglngZone Zone` |14 Building :

2 A~f“'~~j I ` ` ` ` ` }` ~ _ _ 4),-15 Ln 2 A~ "N111` : L,i i

Fig. 13. Deformation of flexible building due totunnelling §Mair et al)

MDRsag:1 _- e/B=0- .2_ 0.4 '_ 0.60io'“i I 26 I "4 I i '2 -0 if0; | | I 1 p,`| » VJ $2/B<0.2_ 0.4 _

D 1'M Rhogi

Fig. 14. Modification factors for deflection ratio(Potts and Addenbrooke]

the relative bending stiffness (p*), which expresses therelative stiffness between the building and the underlyingground, and the corresponding relative axial stiffness. InFigure l4,\.modification factors to the deflection ratios thatwould be obtained from the "green field site" settlementprofiles are presented as different curves for ratios of theeccentricity of the tunnel (e) from the centre line of thebuilding (e/B), where B is the width of the building. Thesevary with the relative bending stiffness, p*. Many buildingshave p* values exceeding 10'2. Figure 14 indicates that inthese cases the modification factor will be in the range of0.1 - 0.2, i.e. the deflection ratio that would be predictedassuming the building performs perfectly flexibly isreduced to only 10-20% of that value by the stiffness of thebuilding. This has major significance on our methods ofprediction and understanding of how buildings actuallybehave in response to tunnelling.

There is a pressing need to monitor_buildings andrecord their detailed behaviour during tunnelling. _Wardle and de Rossi describe the use of electrolevels tomonitor buildings and structures during tunnelconstruction. Electrolevels are comparatively new devicesand have been used with great success_, although there area number of potential difficulties which can be overcomewith care. One problem is the influence of temperaturechange on electrolevels, and this is discussed by theauthors.

4. 3. Services

Settlement effects on services are often neglected untila rather late stage in a proposed tunnelling project.Bracegirdle. Mair, Nyren and Taylor set out a simplemethodology for evaluating potential damage to buried castiron pipes from tunnelling operations.

Another paper relating to services is that by Chapman,which is a review paper on the ground movementsassociated with pipe jacking operations. In many cases ofsmall diameter pipe jacking the effect of groundmovements is of most relevance to adjacent services. The

author illustrates the heavy dependence on workmanship,and highlights the occurrence of heave which is oftenassociated with pipe jacking operations.

4.4 Tunnels

Kimmance, Lawrence, Hassan, Purchase and Tollingergive useful measurements made on the Jubilee LineExtension project when new tunnels passed adjacent to orclose beneath existing tunnels. The measurementspresented are of value to tunnelling engineerscontemplating similar construction in the future. The cleardistance between the invert of existing tunnels and thecrown of a new 5.4m diameter tunnel constructed inLondon Clay using sprayed concrete linings range from 2mto as little as 0.4m. Even in the case where the clearancewas only 0.4m, the settlement of the invert of the existingtunnel was only 15 mm and no visible distress of the castiron linings was observed; the corresponding increase invertical diameter was only 2 mm. The authors also presentvaluable data on the effect of tunnelling alongside existingtunnels. They show that the increase inhorizontal diameterof the existing tunnels amounted to less than 0.1%.

A parametric study of the influence of twin tunnelconstruction has been undertaken by Addenbrooke and1295. They undertook finite element analyses of a numberof geometric configurations, modelling construction of arunning tunnel of 4m in diameter in London Clay eitherabove or alongside an existing tunnel. The authorsexpressed the results of the response of the existing tunnellining to the passage of adjacent tunnels in terms of changein diameter. Figure 15 shows the predicted effect of"piggyback" tunnels (constructed above an existing tunnel)as well as "side by side" tunnels. Given that I a 0.05%diameter change is usually very small, and in many caseshardly measurable, this parametric study indicates that theinfluence of one side by side tunnel in London Clay on theother tunnel becomes negligible when the clear spacingexceeds about two tunnel diameters.

The data from ¢_Kimmance et al (expressed aspercentage diameter changes) are shown on Figure 15.Taken with the measurements by Ward (1969), which arealso shown on Figure 15, these are in 'reasonable agreementwith the predictions by Addenbrooke and Potts.

Kim. Burd and Milligan describe an ingenious set ofmodel tests in the laboratory investigating the influence ofconstruction of a tunnel in soft clay on an adjacent tunnel.Their results are broadly consistent with the conclusionsreached in the theoretical studies by Addenbrooke andPotts, in that negligible effects are implied if the clearspacing between the tunnels exceeds about two diameters.

Bracegirdle, Jefferis, Tedd, Crammond, Chudleigh andBurgess describe an unusual case history where a pair oftunnels were subject to acid attack which led to excessivelyhigh deformations ofthe tunnel linings. The tunnels are ofcast iron construction built in the early part of this century.Over part of their length considerable inward horizontalmovements of the linings took place with severe cracking.

change in diameter

(% °fi‘1iti='1 Value) . side-by-side;horizionta1 diameter0_3_ 0 » side-by-side; vertical diameter '

lengthening n piggy-back; horizontal diamater

0»2__ , piggy-baek;ver1ica1'diameter ,, side~by-side data

_ (Ward, 1969)0.l- ¢0 ` 0 Kimmarrce et al.' 0` »

0 P , I: 2 ae-+ .9 9 e ‘ ­if-0.1- Q

-0.2-0_ shortening

-0.3-| s | I | I | | | A |0 2 4 8 10pillar dimension (multiple of diameters)

Fig 15. Predicted response -of tunnel lining to passage ofadjacent tunnel (after Addenbrooke and Potts)

\.

Both tunnels lay partly in the sands and partly in the claysof the Woolwich and Reading Beds in London. The reasonfor the acid attack is complex: generation of the acidrequired the presence of both oxygen and water, so thatoxidation of pyrites present in the soil could occur. Theonly part of the tunnels that were affected was in a sand­filled valley in the clays, where there was a relatively smallamount of perched water. Elsewhere along the tunnellength, the sands were relatively dry and no acid attackoccurred.

4. 5 Summary

Theoretical parametric studies demonstrate theimportance of soil-structure interaction and demonstratehow the building stiffness plays an important role inmodifying "green field" settlement troughs. Little is knownabout the extent to which services might modify"greenfield" settlement troughs and, more importantly, thehorizontal ground movements. `

Measurements show that tunnel construction close toexisting tunnels in stiff soils such as London Clay lead torelatively small changes in diameter. These observationsare consistent with theoretical and model studies.

It is often the effects on structures, services and othertunnels which are of paramount importance in any urbansoft ground tunnelling scheme. Despite this, it is noticeablethat of the 116 papers submitted to this Symposium, thereare very few case records of the detailed performance ofstructures, services and tunnels subjected to movementsinduced by tunnelling. As is so often stated, there is apressing need for more field measurements, and many ofthese are now being obtained on buildings affected by theJubilee Line Extension project (Burland et al). With suchfield measurements and data on other tunnelling projects it

is to be hoped that engineers will soon be better equippedto make more realistic predictions of the effects of boredtunnels on structures, services and tunnels.

REFERENCES

Mair, R.J., Taylor, R.N. and Bracegirdle, A. (1993).Subsurface settlement profiles above tunnels in clay.Geotechnique 43; 2; 315-320.

Nomoto, T., Mori, H. and Matsumoto, M. (1995).Overview of ground movement during shieldtunnelling: a survey on Japanese shield tunnelling.Proc. Int. Symposium on Underground Constructionin Soft Ground, New Delhi, 1994, 345-351,Balkema.

Ward, W.H. (1969). Deep excavations and tunnelling insoft ground - discussion of paper by Peck, R.B. Proc.7th ICSMFE, Mexico, Vol. 3, 320-5.

Symposium papers reviewed in this Report:

Session 5

Atahan, C., Leca, E. and Guilloux, A. Performance of ashield driven sewer tunnel in the Val-de-Mame,France.

Bowers, K. -H., Hiller, D.M. and New, B.M. Groundmovement over three years at the _Heathrow ExpressTrial Tunnel.

Bracegirdle, A., Jefferis, S. A., Tedd, P., Crammond, N. J .,Chudleigh, I. and Burgess, N. The investigation of acidgeneration within the Woolwich and 'Reading Beds atOld Street and its effect on tunnel linings.

Bracegirdle, A., Mair, R. J ., Nyren, R. J. and Taylor, R. N.A simple methodology for evaluating the potentialdamage to buried cast iron pipes from groundmovements. I

Chapman, D. N. Ground movements associated withpipejacking operations.

Dyer, M. R., Hutchinson, M. T. and Evans, N. Sudden‘ Valley Sewer: a case history.

Forth, R. A. and Thorley, C. B. B. Hong Kong Island Line- prediction and performance.

Hashimoto, H., Hayakawa, K., Kurihara, K., Nomoto, T.,Ohtsuka, M. and Yamazaki, T. Some aspects ofground movement during shield tunnelling in Japan.

Hou, X. Y., Liao, S. and Zhao, Y. Field measurementsfrom two tunnels in Shanghai.

Hwang, R. N., Moh, Z-C. and Chen, M. Pore pressureinduced in soft ground due to tunnelling.

Kastner, R., Ollier, C. and Guibert, G. In situ monitoring ofthe Lyons Metro D Line extension.

Kimmance, J .P., Lawrence, S., Hassan, G., Purchase, N.J.and Tollinger, G. Observations of deformationscreated in existing tunnels by adjacent and crosscutting excavations.

Mair, R. J ., Taylor, R. N. and Burland, J _ B. Prediction of

ground movements and assessment of risk of buildingdamage due to bored tunnelling.

Marshall, M. A., Milligan, G.W.E. and Mair, R.JMovements and stress changes in London Clay due tothe construction of a pipe jack. _ I I

Moh, Z. C., Ju, D. H. and Hwang, R. N. Groundmovements around tunnels in soft ground.

Oteo, C. S. and Sageseta, C. Some Spanish experiences onmeasurement and evaluation of ground displacementsaround urban tunnels.

Price, G., Wardle, I. F. and de Rossi, N. Monitoring oftunnels, surrounding ground and adjacent structures.

Simic, D. and Gittoes, G. Ground behaviour and potentialdamage to buildings caused by the construction of alarge diameter tunnel for the Lisbon Metro.

Standing, J _ R., Nyren, R. J ., Longworth, T. I. and Burland,J _ B.- The measurement of ground movements due totunnelling at two control sites along the Jubilee LineExtension. `

Umney, A. R. and Heath, G. R. Recorded settlements fromthe DLR tunnels to Bank.

Other Sessions

Addenbrooke, T. I. and Potts, D. M. Twin tunnelconstruction - _ground movements and liningbehaviour.

Burland, J. B., Mair, R. J., Linney, L. F., Jardine, F. M.and Standing, J. R. A collaborative researchprogramme on subsidence damage to buildings:prediction, protection and repair.

Egger, P. Tunnel construction in Stuttgart: problems ofsettlements and swelling rock.

Farias, M. M. and Assis, A. P. Numerical simulation of atunnel excavated in a porous collapsible soil.

Forbes, J. and Finch, A. P. Application of compensationgrouting to the St Clair River Tunnel Project, NorthAmerica.

Grose, W. J _ and Eddie, C. M. Geotechnical aspects of theconstruction of the Heathrow Transfer BaggageSystem tunnel.

Kavvadas, M., Hewison, L. R., Laskaratos, P. G.,Seferoglou, O. and Michalis, I. Experiences from theconstruction of the Athens Metro. _

Kim, S. H., Burd, H. J _ and Milligan, G. W. E. Interactionbetween closely spaced tunnels in clay.

Kuzuno, T., Takasaki, H., Tanaka, M. and Tamai, T.Driving control and ground behaviour of triple circularface shield machine.

Linney, L. and Friedman, M. Protection of buildings fromtunnelling induced settlement using perrneationgrouting.

Negro Jr, A., Sozio, L. E. and Ferreira, A. A. TunnellingA in Sao Paulo, Brazil.Phienwej, N. Geotechnical experiences from previous

tunnel projects in Bangkok soils.Potts, D. M. and Addenbrooke, T. I. The influence of an

existing surface structure on the ground movementsdue to tunnelling. _

Shirlaw, J. N., Pennington, B. N. and Yi, X. Monitoringduring the construction of the Allen Sewer Tunnel,Toronto, Canada.

Steiner, W. Slurry penetration into coarse grained soils andsettlements from a large slurry shield tunnel.