ags_2006-2_article_4.pdf

8/10/2019 AGS_2006-2_article_4.pdf

1/11

ACTA GEOTECHNICA SLOVENICA, 2006/2 45.

METHODS FOR CONTROL OF SEEPAGE IN RCCDAMS WITH WATERTIGHT AND DRAINAGE

MEASURESYUEMING ZHU, STEPHAN SEMPRICH, ERICH BAUER, CUIPING YUAN and

DONGMEI SUN

About the authors

Yueming ZhuInstitute o Hydrostructure Engineering,Hohai UniversityNanjing, P.R. ChinaE-mail: [email protected]

Stephan SemprichInstitute o Soil Mechanics and Foundation Engineering,Graz University o echnologyGraz, AustriaE-mail: [email protected]

Erich BauerInstitute o Applied Mechanics,Graz University o echnologyGraz, AustriaE-mail: [email protected]

Cuiping YuanHohai UniversityNanjing, P.R. China

Dongmei SunHohai UniversityNanjing, P.R. ChinaE-mail: [email protected]

Abstract

Te technologies for construction of roller-compactedconcrete (RCC) dams have been considerably developedduring recent years in China. At the time being, they have

been successfully applied to the constructions of evenextreme-high gravity dams and medium to high archdams. Tere are a few of hundreds of RCC dams (RCCD)under design and/or construction in China. One of themain concerned technical problems according to theconstruction is about the understanding of the property ofseepage in RCCDs and the relevant theory and methods

for the control of the seepage. In order to overcome theproblem, the senior author has been engaged in a wide

study on the property and methods for control of seepagein RCCDs for more than 10 years. Te property of seep-age, measures for watertightness and drainage, optimaldesign and construction schemes for control of seepagein the dams have been essentially understood either intheory and practice. Te results have been applied for the

construction and the backanalysis of several dams. Tepaper describes the research ndings in detail with respectto the theoretical fundament and their application for ahigh RCC gravity dam.

Keywords

gravity dam, Roller-compacted concrete, RCC, seepage,anisotropic permeability, drainage

1 INTRODUCTION

Te Horizontal Layer Method used in roller-compactedconcrete (RCC) construction has been replaced by theSlope Layer Method, which importantly reduces thearea o exposed young RCC and increases RCC placingrates. Te intrinsic permeability o RCC is very low (itequals 1.010-9to 1.010-12cm/s), whilst the perme-ability o the lif surace including the joint lif suraceis proportional to the cubic width o a hydraulic joint.Teory and engineering practice have proved that the lifsurace is the main pathway o seepage, in which leakagegenerally occurs. According to the present technology

or constructing o RCC dams, the body o a RCC damis considered to be a strong anisotropic medium withmore than 2 to 4 or even 6 to 7 orders o magnitude (orexample Willew Creek RCC gravity dam, USA) o ratioo the tangential major coefficient o permeability to thenormal major coefficient o permeability o the layer andlif joint suraces. Tis causes the seepage properties oRCC dams to be totally different rom those o conven-tional concrete dams.

8/10/2019 AGS_2006-2_article_4.pdf

2/11

ACTA GEOTECHNICA SLOVENICA, 2006/246.

Since the early development stages o RCC damconstruction, there have been a lot o problems withseepage due to the lack o knowledge about the propertyo seepage in RCC dams. One problem, or exampleis an extremely high position o the exit point on the

downstream ace, so that exit-owing discharge may behigher than expected. Tis is caused by the anisotropichydraulic resistance according to the dam structureconstructed by layers. So, a lot o dams were required toincrease their waterproo capability by grouting duringthe beginning period o operation. Such was the casewith Willew Creek RCC dam and Chinese ShimantanRCC gravity dam o the height o 40.5 m.

Tis paper analyses in detail optimal design schemes orthe control o seepage. Te analysis is based on a wideand in-depth study on the property and methods or thecontrol o seepage in RCCD and on the senior authorsexperiences in studying characteristics o permeability

and control techniques to handle problems o seepageor RCC dams.

2 PROPERTIES OF SEEPAGE OFRCC AND RCCD

Te system o RCC or RCCD consists o body layer oconcrete with homogeneous and isotropy property, alayer and a lif interace o anisotropy property whichbecomes an inhomogeneous multi-laminated medium.Actual properties o seepage o RCC and RCCD are

inuenced by the properties o the body o concrete,which can be described by Darcys coefficient o perme-ability k

RCC, and the properties o the layer and the lif

interace which is related to the width o a hydraulicjoint, roughness o the joint, connectivity o the joint,stress and strain behaviour o the layer surace, loadhistory, etc. Te engineering practice has already provedthat the layer and the lif interace are the main pathwayo seepage and represent weak suraces according totension and sliding. In general, RCC and RCCD can beconsidered as strongly anisotropic medium.

An experiment recorded in literature [1, 2] presentsthe model o an in-situ concrete block connected inparallel and in series within the Longtan RCC project,which is shown in Fig. 1. Te experiment shows thatthe ratio o average coefficient o permeability in theparallel model to that in the series model is about 1 to 4orders o magnitude. And the seepage ow mainly exitsthrough the layer and the surrounding region o greatskeletal material, which proves that the layer and the lifinterace are main pathways o seepage [3].

Figure 1. Parallel and series models o a rectangular RCCblock in the permeability test.

Gong et al. [3] show the results o in-situ water pressuretest, which includes the tests on Willew Creek RCC damin the USA, Chinese Guanyinge and ongzijie projects.Te tangential major coefficient o permeability o

the layer in Willew Creek RCC dam in the USA evensurpasses 310-3cm/s, and the average value o thetangential major coefficient o the layer permeability isin the range o 110-3to 110-8cm/s. So, heavy leakageoccurs through layers and the location o the exit-ow-ing line is very high. Compared to the tangential majorcoefficient o layer permeability, the normal major coe-cient o layer permeability is the same as the coefficiento permeability o the body o RCC. Te propertieso seepage o RCC and RCCD mainly depend on thetangential property o seepage. Te system o RCC andRCCD should be considered as strongly anisotropic, andthe ratio o anisotropy may reach 1 to 4 or even moreorders o magnitude.

Teoretically, the properties o ow through layer andlif joint suraces agree with the properties o ssureow. Te properties o seepage ow through layer andlif joint suraces in RCC and RCCD can be usuallydescribed by the cubic law o an advective ssure ow,which has been proved by home and oreign engineeringexperts and which is in accordance test with results in

various countries [3, 4].

Te property o RCC considered as inhomogeneousmulti-laminated medium can be described by tangentialand normal major coefficients o the permeabilityo the layer which depend on the body coefficient o

permeability, cubic law and the thickness o the layer.Eqs. (1) and (2) show the tangential and normal majorcoefficients o the permeability o the layer.

kB

B d kg

dt f RCC f = +1

123[( ) ]

(1)

kBk

B dn

RCC

f

=

(2)

Y. ZHU ET AL.: METHODS FOR CONTROL OF SEEPAGE IN RCC DAMS WITH WATERTIGHT AND DRAINAGE MEASURES

8/10/2019 AGS_2006-2_article_4.pdf

3/11


where kt and kn are average tangential and normalmajor coefficients o permeability o the layer, respec-tively, B denotes the thickness o the layer includingthe body thickness o the layer and the width o thehydraulic joint df , kRCC denotes body coefficient o

permeability o RCC, denotes the kinematic viscosityo water with 0.013 cm/s at 10 C.

Because df is much smaller than B , the normal majorcoefficient o the permeability o the layer mainlydepends on the body coefficient o permeability o RCCand also on the coefficient o bedding cushion, whenthere is one, whilst the tangential major coefficient opermeability mainly depends on the width o a hydraulic

joint.

Figure 2. Cross section o seepage control measures or RCCD.

Due to a strong anisotropic property o RCC, the seep-age control measures or RCCD become more difficult

than conventional concrete. Fig. 2 shows the generallayout o measures or watertightness and materialzones. It will be a reliable anti seepage structure o adistorted concrete layer with a thickness o 0.3 to 1.0 mon the upstream ace, which plays an important role owater-resistance in the ront o the dam. Other orms oanti seepage structure on the upstream ace o RCCDsare available, such as a RCC layer, a reinorced concreteslab, an asphalt concrete layer, and a combination o

different anti-seepage structures. Te majority o allprojects under construction or under design adopt adistorted concrete layer as the anti seepage structure.Some inltrating ow through the anti-seepage structurewill be drained by drainage holes behind a distorted

concrete layer, in which the uplif pressures in the layerand lif surace are almost zero. In order to prevent theseepage ow rom exiting on the downstream ace, athin distorted concrete layer is also required on thedownstream ace. When the upstream water level is veryhigh, a curtain o drainage holes hanging upward thedatum plane is needed. Sometimes horizontal drainageholes are alternatively arranged in the layer and the lifsurace additionally helps drainage and pressure relie.Te principles o seepage control or RCCDs oundationmostly agree with conventional concrete dams whichmainly depend on a curtain o watertightness and drain-age holes.

3 NUMERICAL ANALYSIS ON THERCCD SEEPAGE FIELD WITH FEM

3.1 NUMERICAL MODEL OF SEEPAGEFIELD

Due to special inhomogeneous and strongly anisotropicmulti-laminated structures in RCCDs, the system oRCC should be considered as ractured rock masses

medium according to the property o the layer seepage,which can be simulated by an equivalent continuummodel, a discrete ractured network model and a dualporosity model [4-7].

Equivalent continuum models based on the balanceequation o mass consider layer and lif suraces asequivalent continuum porous media, which are analyzedby a mature theory o continuum porous media. Butthe equivalent seepage discharge and seepage pres-sure cannot be obtained with this method. A discreteractured network model assumes that the body o RCCis impervious, so that the properties o RCC mainlydepend on the widths o hydraulic joints, location andconnectivity o the layer, and the coefficient tensoro permeability is determined by the cubic law andDarcys law. Tis method is able to actually simulate thebehaviour o seepage ow between layers, but due tothe uncertainty o the joint and lif surace, it is moredifficult to simulate ractures. Te objects within adual porosity model involve seepage behaviour both inporous media and ractured media.


8/10/2019 AGS_2006-2_article_4.pdf

4/11


Combining the merits and aults o three modelsmentioned above, the authors proposed inhomogeneousmulti-laminated media elements model, which uniesthe equivalent continuum model and the non-contin-uum mixed model, and which has the advantages o an

equivalent continuum model involving ewer elementsneeded in the calculation [8].

3.2 INHOMOGENEOUS MULTI-LAMINATEDMEDIA ELEMENT

An inhomogeneous multi-laminated media elementmodel assumes that the dam body consists o a bodylayer o roller-compacted concrete, a lif interace, a lif

joint interace o the course o one continuous construc-tion o lifs, a bedding plane and normal concrete mate-rials, which become inhomogeneous multi-laminatedmedia elements. Te element conductivity matrix is

expressed with Eq. (3).

k B KB J d d d B K B J d d ij i

j i

k jkk

k

= = 1

1

1

1

1

1

1

1

1

1

1

( )=k

N

d1

(3)

Here we introduce the integral variable transormationshown in Eq. (4).

=

+ k k k k1 1

2 2 (4)

So we have Eq. (5).

k B K B J d d d ijk k

k

N

i

k j= ( )

=

11

1

11

1

1

1

2 (5)

Here Nis the number o layers o inhomogeneousmulti-laminated media, Kk is the coefficient tensor opermeability o the k th layer, Eqs. (9), (10) and (11), Bi ,Bj and J are unctions o the local coordinates o ,and .

3.3 PLANAR ELEMENT WITHOUT THICK-

NESS FOR FISSURESBearing in mind numerous lif interaces and lif jointinteraces and occurrence o horizontal cracks and cleav-age cracks in case o emergency o a dam, the authorsproposed the planar element without the thicknessmodel ([9] and [10]), which can be used to simulate theexact behaviour o the seepage ow in RCC and whichhas been proved to be a successul model up to now.

Because the normal major coefficient o the permeabilityo a RCC body and the width o the joint are so smallthat normal head loss in the joint lif suraces is almostzero, the seepage ow through joint lif suraces can beconsidered as a two dimensional ow model (index ).

We then have Eq. (6).

= =

xk

h

xi j

if ij

f

jf

0 1 2 , , (6)

where xif is the corresponding local coordinate, kij

f is atwo dimensional coefficient tensor o permeability in theplanar element without the thickness model, and krs

fe isthe element conductivity matrix which can be calculatedrom the below Eq. (7):

k kN

x

N

xk

N

x

N

xk

N

x

N

rs

fe f r

f

s

f

f r

f

s

f

f r

f=

+

+

11

1 112

1 222

2

2 ssfs x

dsf

2

r s m= 1( , )

where sf is a subdomain o a joint lif surace element,Nr and Ns are interpolating unctions o the joint lifsurace element, m is the number o nodal points in the

joint lif surace element.

According to balance equation o mass [11], we obtainthe below FEM governing equation or a seepage eld

( ) , , ,Q Q i niRCC

if

e

+ = = 0 1 2 (8)

where n is the total number o nodal points, QRCC andQF are nodal uxes o the point i which are contributedby a 3-D equivalent continuum media element o RCCand a 2-D planar element without the thickness element.

3.4 COEFFICIENT TENSOR OF PERME-ABILITY IN THE RCCD SEEPAGEFIELD

In the study o seepage properties in RCCDs, i the layero RCC is horizontal, the coefficients o permeability

can be described by major coefficients o permeabilityshown in Eqs. (1) and (2). In a sloped layer the directiono major coefficients o permeability are not the same asa coordinate axis and the properties o seepage shouldbe expressed by the coefficient tensor o permeabilityshown as Eqs. (9), (10) and (11). Eq. (9) is suitable orthe layer incline to the upstream, whilst Eqs. (10) and(11) are suitable or the layers incline to the lef bankand the right bank, respectively.

(7)


8/10/2019 AGS_2006-2_article_4.pdf

5/11


K k

r r

r

r r

n[ ]=

+ ( )

( ) +

cos sin sin

sin sin cos

2 2

2 2

01

21 2

0 0

1

21 2 0

(9)

K k

r

r r

r r

n[ ]= + ( )

( ) +

0 0

01

21 2

01

21 2

2 2

2 2

cos sin sin

sin sin cos

(10)

K k

r

r r

r r

n

[ ]= + ( )

( ) +

0 0

01

21 2

01

21 2

2 2

2 2

cos sin sin

sin sin cos

(11)

where is the slope angle, kt and kn are tangential and

normal major coefficients o permeability, respectively, rk

kt

n=

is the ratio o tangential major coefficient o permeabilityto normal major coefficient o permeability, K[ ] is thecoefficient tensor o permeability.

4 THE DESIGN METHODS FORWATERTIGHT AND DRAINAGEMEASURES

4.1 WATERTIGHT STRUCTURES ON THEUPSTREAM FACE

Lif suraces on the upstream ace are probably apathway o leakage, so it is important to take watertightmeasures in the upstream. Te distorted RCC and2-grade RCC are mostly used as watertight measureson the upstream ace in present RCCDs. 2-grade RCC

represents a dened RCC quality, which is higher thanthe 3-grade RCC. Based on the authors experiencesabout methods o seepage control in RCCDs, the ollow-ing schemes or the control o seepage are proposed.First, a thin layer o distorted RCC and additional2-grade RCC are arranged on the upstream ace, andseepage control treatment is done on the lif suraceswithin the width o 2m. Second, a thin layer o distortedRCC and additional 3-grade RCC are arranged on the

upstream ace, and seepage control treatment is done onthe lif surace within the width o 2 m (i.e. to spread alayer o mortar o the thickness o 1.0 ~ 1.5 cm in time).Different projects will take different schemes o controlo seepage. emperature control and crack prevention

are very important on the upstream ace to prevent theoccurrence o penetrating cracks. Quality o structure oseepage control is more important than its width.

Uplif pressure tolerances in the lif surace, the joint lifsurace and the datum plane are shown in Figs. 3 and 4,where Hu and Hd are depths o upstream and down-stream respectively and unit weight o water.

Figure 3. Cross section o design uplif pressure in the dam body.

Figure 4. Cross section o design uplif pressure on the datumplane.

4.2 WATERTIGHT STRUCTURES ON THEDOWNSTREAM FACE

Due to the strong permeability o the lif surace on thedownstream ace o RCCDs, the location o the exit-owing line may be very high and the leakage o seepagemay be so great that it will endanger stability saety othe dam. A layer o distorted RCC with a thickness o0.30~0.50 m will be arranged on the downstream ace.Inside the distorted RCC layer, the seepage control treat-ment will be perormed to the depth o 1 to 2 m o the


8/10/2019 AGS_2006-2_article_4.pdf

6/11


lif suraces below the water level o downstream; alter-natively; 2-grade RCC can ollow inside the distortedRCC layer.

4.3 DRAINAGE HOLES CURTAINS INTHE DAM BODY

Drainage holes are required both in a conventionaldam and a RCC dam to discharge the seepage ow. Tecurtains o exit-owing drainage holes will be arrangedin the zone o 3-grade RCC ollowing 2-grade RCC in theupstream region, in which the intervals o drainage holesare about 4 to 5 m. In the downstream region o a 2-gradeRCC zone or within the RCC, treated by seepage control,a curtain o exit-owing drainage holes can be alsoarranged, in which the interval between drainage holesis recommended to be 6 m. For a high dam with a widedatum plane, the curtains o overowing drainage holes

are needed on the datum plane near the upstream region.

4.4 BEDDING PLANE OF CONCRETE ONTHE DATUM PLANE

Due to cracks generated requently, difficulties inconstruction and more cement used in conventionalconcrete, the bedding plane built o conventionalconcrete has been replaced by bedding plane built with2-grade RCC o the thickness o 0.5 m to play the role oproviding at and resisting water. Te bedding plane andwatertight structures together orm a watertight structure.

4.5 WATERTIGHT STRUCTURES ANDDRAINAGE MEASURES ON THEDATUM PLANE

Watertight structures and drainage measures are alsorequired in the zones o the dam heel and the dam toeon the datum plane. Engineering practices have provedthat curtains relative watertightness just needs to be 3times lower than o the surrounding rock masses. Tegrout curtain must be complete and the curtain drainageholes should be unblocked. Te interval o drainageholes on the datum plane is 4 to 5 m, the same as in thedam body.

4.6 TRANSVERSE JOINT SEAL ANDWATERTIGHT STRUCTURES ONSIDES OF JOINTS

Compared to conventional concrete dams, the RCCDsneed better transverse joint seal because there are lifsuraces in RCCDs. In order to prevent water rom inl-trating into the dam through the sides o the transverse

joint, watertight structures on the sides o the joint arestrongly required. Tis can be realized by a distortedconcrete layer o a thickness between 0.3 and 0.5 m andshould be connected very well with the bedding plane.

4.7 DESIGN PRINCIPLES FOR THEWATERTIGHT STRUCTURE ANDDRAINAGE MEASURE

Briey, the design principle o a watertight structure andthe drainage measure is to resist in ront and to drainat the back in the upstream region o the dam as wellas in the zone o the dam toe in dam oundation, whilstin the downstream region o the dam the principle isto drain in ront and to block at the back as well as inthe zone o the dam heel in dam oundation. Here, toresist in ront in the upstream region means to arrangea layer o distorted RCC and 2-grade RCC or the zone

o lif surace and lif joint surace treated by seepagecontrol method o the width o 2 to 3 m; and to drain atthe back means to arrange the main curtain o drainageholes. o drain in ront in the downstream region meansto arrange the main curtain o drainage holes near theupstream, and to drain at the back means to arrange alayer o distorted RCC.

In the zone o the dam toe in dam oundation, to resistin ront means to arrange the main grouting and thewatertight curtain, and to drain at the back means toarrange the main curtain drainage holes behind thegrouting and watertight curtain. In the zone o thedam heel in dam oundation, to drain in ront means

to arrange the curtain o drainage holes behind thewatertight curtain near downstream, and to resist inthe back means to arrange the watertight curtain neardownstream.

In act, the unction o water-resisting o the watertightcurtain and the unction o drainage and pressure relieo the curtain o drainage holes are not independent.I there is only a watertight curtain and no curtaino drainage holes in dam oundation, the uplif pres-sure will be very high unless the permeability o thewatertight curtain is very small, or example 2 orderso magnitude smaller than that o surrounding rock

masses. Tis is difficult to perorm in engineeringpractices. I there is only a curtain o drainage holes andno watertight curtain in dam oundation, though theuplif pressure can be greatly reduced, the rock masses inoundation may be destroyed by a great hydraulic gradi-ent surrounding the drainage holes, which is absolutelyirrational or the long-term saety o dam.


8/10/2019 AGS_2006-2_article_4.pdf

7/11


5 APPLICATION

5.1 STUDY ON THE ENTIRE SEEPAGEFIELD

We have taken the 6th segment o Guangzhao RCCDas a case study, which is located in the middle reacheso the Longtan River in the Guizhou province. On theace o upstream, there is a layer o distorted concreteo different thickness along the height, 0.8m above theelevation o 615.0 m and 1.0 m below the elevationo 615.0 m. Te watertight structure in the upstreamregion is composed o a layer o distorted concreteand additional 2-grade RCC and 3-grade RCC, exceptor the zone above the elevation o 710.0 m, wherethere is a layer o distorted concrete and an additional3-grade RCC. Te depth o 2-grade RCC is in the range

o 3 to 13m according to the water head on the aceo upstream. Tere is a drainage curtain in the dambody, and the watertight curtain and the main drainagecurtain are arranged in the zone o the dam toe becausethe water level in downstream is lower than the elevationo the datum plane. Te normal water levels are 745.0 mand 583.5 m in the upstream and downstream, respec-tively. Empirical coefficients o permeability o variouswatertight materials in the dam, the watertight curtainin dam oundation and the grout curtain are shown inable 1.

According to the seepage control methods proposedin this paper, Fig. 5 and Fig. 6 show the distributions

o water head contour lines and uplif pressure headcontour lines, respectively, in a short orm.

Te entire seepage eld presents exact seepage propertyand obvious regularity, which indicates that the seepagecontrol method proposed in this paper plays the role o

Figure 5. Water head contour lines.

Figure 6. Uplif pressure head contour lines.

Table 1. Coefficients o permeability o body material and various watertight materials.

ype o concrete Body coefficiento permeability

(cm/s)

Normal coefficient opermeability o lif surace

(cm/s)

angential coefficient opermeability o lif surace

(cm/s)

notes

Conventional concrete 110-10 anisotropic

Distorted concrete 110-9 anisotropic

2 grade RCC 110-9 110-9 110-7

3 grade RCC 110-9 110-9 510-7

Watertight curtain 110-5 isotropic

Consolidated grouting 110-4 isotropic


8/10/2019 AGS_2006-2_article_4.pdf

8/11


resisting in ront and draining at the back. In the dambody, the potential o high water head is mostly lostthrough the isotropic layer o distorted concrete. Toughthe coefficient o permeability o distorted concrete isthe same as the normal major coefficient o permeability

o 2-grade RCC and 3-grade RCC behind the layero distorted concrete, it is watertight compared to thetangential major coefficient o permeability o 2-gradeRCC, because the coefficient o permeability o layero distorted concrete is 2 orders o magnitude smallerthan the tangential major coefficient o permeabilityo 2-grade RCC. Contour lines in the layer o distortedconcrete are probably parallel to the ace o upstream.In the layer o distorted concrete above the datum planeabout 1 m the potential o water head is reduced aboutover tens meters, in which the cracks always occur andquality o the layer o distorted concrete should be paidmore attention. In the adjacent zone o 2-grade RCCand 3-grade RCC the contour lines have a distinctlyhorizontal kink towards downstream, which is causedby the act that hydraulic gradients in 2-grade RCC and3-grade RCC are much smaller than those in the layer odistorted concrete. Due to the easiness o seepage owthrough 3-grade RCC, the quality o a watertight struc-ture ahead 3-grade RCC should be paid more attentionto ensure the saety o RCCDs.

Some seepage ow moves around the drainage holes,and it will exit rom the ace o downstream i there isno watertight structure arranged. Fig. 5 shows a layer odistorted concrete on the ace o downstream shown inenlargement. Due to no resistance on this seepage ow

in the lif surace and lif joint surace, the value o uplifpressure will be near zero. A drainage hole is anotheruseul seepage control structure, by which almost all othe seepage ow is discharged. So, in order to ensurethat the drainage hole perorms successully, the qualityo the drainage hole should be paid more attention tothe construction phase and the drainage hole should bechecked regularly during the operation phase.

Te uplif pressure contour lines at the elevation o 685.0m, 655.0 m and 625.0 m are shown in Fig. 6. In thesetypical elevations, the uplif pressure is low, and thedrainage curtain in the dam body, the grout curtain andthe main drainage curtain in dam oundation play an

important role or pressure relie.

Briey, the seepage control methods proposed in thispaper work very well. Different watertight structuresplay different roles and also depend on each other. Fromthe results mentioned above we can see that the layer odistorted concrete and drainage holes play an importantrole o water-resisting and pressure relie. A big dealo discharge will be drained by the drainage holes. Te

uplif pressure will be very high i there are no drainageholes. From the distribution o uplif pressure on thedatum plane, we can conclude that the grout curtainand the drainage curtain in dam oundation are veryimportant or water-resisting.

5.2 SENSITIVITY ANALYSIS ONSEEPAGE CONTROL METHODS

5.2.1 Water-resisting Function of2-grade RCC

Fig. 7 shows the computed result when the 2-gradeRCC on the ace o upstream is removed on the basis oseepage control structures mentioned above, rom whichwe can conclude that there are no obvious changes inthe entire seepage eld. We can so come to the sameconclusions as those obtained by engineering practice.

First, the water-resisting unction o 2-grade RCC canbe ignored. Second, the layer o distorted concreteon the ace o upstream only works in the unction owater resisting. Tird, the arrangement o 2-grade RCCwill make the construction more difficult, reduce theconstruction rate and make the advantages o RCCDade away. Fig. 8 shows the distributions o uplif pres-sure heads, rom which the same conclusions can beobtained.

Figure 7. Te contribution o hydraulic head contour lines inthe dam body when 2-grade RCC is removed.


8/10/2019 AGS_2006-2_article_4.pdf

9/11


Figure 8. Te distribution o uplif pressure head contour linesin different elevations when 2-grade RCC is removed.

5.2.2 The Study of Cracks in the DamBody

Cracks in the dam body are o the key importance orthe saety in hydraulic engineering. Figs. 9 and 10 showthe computed results when there is a vertical cleavagecrack in the layer o distorted concrete on the ace oupstream. Te width o this cleavage crack is 0.2 mmand it is located in the middle plane between two drain-age holes. Figs. 11 and 12 show the computed resultswhen there is a horizontal penetrating crack at intervalso 20 m only in the layer o distorted concrete on theace o upstream. Te width o this horizontal penetrat-ing crack is 0.1 mm.

When there is a vertical cleavage crack in the layer odistorted concrete on the ace o upstream, the unctiono water-resisting o the layer o distorted concrete disap-pears and the main watertight structures are 2-gradeRCC and drainage holes behind the layer o distortedconcrete. Te hydraulic head contour lines in 2-gradeRCC are parallel to the ace o upstream, and the uplifpressure head in the dam body is still small. But due tothe anisotropic property o 2-grade RCC, the uplif pres-sure in 3-grade RCC is a somewhat higher compared tothe distribution o uplif pressure shown in Fig. 6.

Fig. 11 shows the distribution o hydraulic head contour

lines when there are several horizontal cracks on theace o upstream. Te values o uplif pressure meet thedemands desired. Except or the zones near the cracks,the inuence o the cracks on the seepage eld can beignored. Under the watertight unction o 2-grade RCCand drainage holes, the uplif pressure head contourlines are horizontal and small enough. Because there is ahorizontal crack in the elevation o 685.0 m, the water-tight unction o the layer o distorted concrete vanishes,

so the uplif pressure contour line is horizontal in thelayer o distorted concrete.

Figure 9. Distribution o hydraulic head contour lines whenthere is a vertical cleavage crack on the ace o upstream.

Figure 10. Distribution o uplif pressure head contour lines indifferent elevations when there is a vertical cleavage crack onthe ace o upstream.

Fig. 9 to 12 show the computed results which take into

account o various adverse conditions. We can concludethat under the protection o other watertight structuresthere are no atal uplif pressures in the seepage eld,and the dam is in a temporary sae state. But we are allaware that the tendency o cleavage cracks and horizon-tal cracks is difficult to predict, so we have to pay moreattention to the temperature control and cracks preven-tion measures, especially to the quality o the layer odistorted concrete.


8/10/2019 AGS_2006-2_article_4.pdf

10/11


Figure 11. Distribution o hydraulic head contour lines whenthere is a horizontal penetrating crack at intervals o 20m inthe layer o distorted concrete on the ace o upstream.

Figure 12. Distribution o uplif pressure head contour linesin different elevations when there is a horizontal penetratingcrack at intervals o 20m in the layer o distorted concrete onthe ace o upstream.

6 CONCLUSIONS AND SUGGESTIONS

Tis paper generalizes the properties o seepage in RCCand RCCDs. Constructive suggestions o seepage controlmethods in various zones o the dam are proposed, andthey have been applied to engineering practice. Becausethe RCC can be considered as inhomogeneous multi-laminated medium with strong anisotropy, the property

o seepage o RCC are mainly depends on the propertyo the lif surace, in which the coefficient o permeabil-ity is 2 to 3 orders o magnitude higher than that o theRCC body. In construction stages the quality o the lifsurace should be paid more attention.

Due to the anisotropy o RCC, the property o seepage oRCC is more difficult than that o conventional concrete.Te design principles o watertight structures in RCCDsare to resist in ront and drain at the back. On the aceo upstream the arrangement o a watertight structureis strongly required, such as reinorced concrete slab, aconventional concrete slab and other orms. In presentengineering practice the layer o distorted concreteand additional 2-grade RCC are recommended: thelayer o distorted concrete plays the primary role inthe watertight unction, whilst 2-grade RCC plays thesecondary role in the watertight unction. In order todrain a small amount o seepage ow through watertight

structures, drainage holes should be arranged behindthe watertight structure in the upstream. In order toensure that drainage hole play a permanent role inpressure relie, the diameter o a drainage hole shouldbe greater and the distance between two drainage holescan reach 4 to 6m. In order to prevent the seepage owrom exiting too high on the ace o downstream, a layero distorted concrete is strongly required. Te design owatertight control in the dam oundation is the sameas that in conventional dam oundation, in which thegrout curtain and main drainage curtain are absolutelyneeded. Based on the study o RCC or over ten years,the design o seepage control methods and the numeri-

cal model in RCCDs have been completely solved. Teproblem o temperature control and crack prevention isstill to be researched in uture.

REFERENCES

[1] Su, B. et al. (2002). Statistical eatures o perme-ability coefficients o Jiangya RCCD core samples.

Journal of Hohai University, Vol. 30(2), 1-5.[2] Su, B. et al. (2000). Research o permeability o

roller compacted concrete o Jiangya Dam.Journalof Hohai University, Vol. 28(2), 7-11.

[3] Gong, D., Zhu, Y. (2002) Research on seepage oRCC Dams. International Journal HydroelectricEnergy, Vol. 20(2), 23-25.

[4] Wittke, W. (1990). Rock Mechanics Teory andApplications with Case Histories. Springer-VerlagBerlin Heidelberg.

[5] Schrader, E.K., Namikas, D. (1988). Perormance oroller compacted concrete dams. Proceedings of theSixteenth Congress on Large Dams, Vol. 2, 339-364.


8/10/2019 AGS_2006-2_article_4.pdf

11/11


[6] Iwai, K. (1976). Fundamental studies o uid owthrough a single racture. Ph.D. Tesis. Univ. oCaliornia, Berkeley, USA.

[7] Sheng, J. (2000). Studies on 3-D numerical simula-tion o ractured rock masses taking into account

the coupling o seepage and stress and application.Ph.D. Tesis. Hohai University, Nanjing, P.R.China.

[8] Zhu, Y. et al. (2002). Analysis o seepage controlcharacteristics o the optional scheme o TreeGorges or RCC gravity dam on the Yangtze River.

Journal of Hohai University. Vol.28(2), 1-6.[9] Zhu, Y., Gong, D., Zhang H. et al. (2003). Plane

element simulation o racture seepage or analyz-ing seepage in RCCD.Journal of Hydraulic Engi-neering, Vol 3, 63-68.

[10] Zhu Y., Liu C., Li J. (2005). Study on seepagecontrol methods and watertight and drainagemeasures in RCCDs. Guizhou Water Power, Vol.19(3), 5-11.

[11] Zhu Y. (1997). Te computing method o equiva-lent nodal virtual discharge in Darcys equation.

Journal of Hohai University. Vol. 25(4), 105-108.[12] Zhu Y., Zhang, L., Pang Z. (1999). Study on perme-

ability properties o RCC Dams and Longtan grav-ity RCCD (1). Hongshui River. Vol. 18(1), 2-8.

[13] Zhu Y., Zhang L., Pang Z. (1999). Study on perme-ability properties o RCC Dams and Longtan grav-ity RCCD (2). Hongshui River. Vol. 18(2), 8-11.


ags_2006-2_article_4.pdf

Documents