LITERATURE CITED I. 2.

L. M. Garkun, "Control of the thermal stress state of massive concrete in the outer zone of a dam," Gidrotekh. Stroit., No. 6 (1979). S. N. Starshinov, V. B. Idel'son, and L. I. Markin, "Thermal stress state of blocks with zonal separation of the concrete," l~nerg. Stroit., No. 8 (1986).

D E S I G N AND W A T E R T I G H T E F F E C T I V E N E S S O F THE UNDERGROUND CONTOUR OF THE SAYANO-SHUSHENSKOE DAM L. I. Mayshev and V. G. Skokov

UDC 624.152.612.3:627.825

The foundation of the Sayano-Shushenskoe dam is composed of strong para- and ortho-schists. The rock mass of the foundation is dissected by six steeply and gently sloping joint systems accompanied by zones (up to 5 m) of increased jointing and feather joints. Zones of reactively wide, gently sloping joints are noted in the channel part of the dam foundation at a depth up to 30-40 m. The joints are generally filled with mineral quartz- carbonate materials and iron oxide, dense clayey and grus filling is present along large tectonic zones. A decrease of water permeability with depth is characteristic for the dam foundation. Maximum permeability is noted in the channel part of the dam foundation. Here a specific water absorption of 1-10 liters/(min.m2) is found in the 30-m nearcontact zone of the foundation, and it exceeds 10 liters/(min.m2) in zones of intersection of tectonic disturbances with gently sloping joints. The upper decompressed zone with a depth of 20-30 m has average specific water absorptions of 0.2-0.5 liter/(min-m2). The lower horizon of the discharge zone at a depth from 30 to 70 m is characterized by specific water absorption of 0. I-0.2 liter/(min.m2). Below the discharge zone, to a depth of 100-120 m, are found individual permeable zones, which were found more often in the central and right-bank parts of the channel than in the left-bank. The permeability of rocks in the bank abutments of the dam is less than in the channel part. However, here too are permeable zones with a specific water absorption greater than 0.1 liter/(min.m2) at a depth to 100 m, especially on the left bank. The boundary of the practically impervious rocks having a permeability coefficient not greater than 0.01-0.02 m/day is located at a depth of 90-100 m, it drops to 120-130 m in the central part of the channel within sections 25-32 of the dam. The deformation characteristics of the rocks have a scatter of values, depend on the place of determination and state of the rocks, and their relation to the gently sloping joint system and tectonic zones is noted. The calculated values of the modulus of deformation of the rocks were taken in the design equal to: outside the tectonic zones, from 1 x 104 in the upper to 1.6 x 104 MPa in the lower discharge horizons, 1.8 • 104 MPa in the preserved rocks, and from 0.7-0.9 to (1.2-1.4) x 104 MPa in the zone of influence of the tectonic disturbances. The minimum velocities of longitudinal waves confined to zones of tectonic disturbances and the gently sloping joints (before grouting) were 2.5 km/sec for the channel stretch and up to 5-6 km/sec in the preserved rocks. When determining the parameters of the watertight and strengthening measures (Figs. 1-4) in the dam foundation, the following were taken as the basis: The engineering-geological conditions of the foundation - substantial inhomogeneity of the seepage and deformation characteristics, presence of large joints and zones of tectonic disturbances, depth of the discharge zone and presence of a pronounced aquiclude at a depth of less than 0.6H (H is the head acting on the dam); Results of calculating the stress-strain state (SSS) of the foundation at the stage of completing the design, which indicated that during and after constructing the dam and filling the reservoir, the SSS changes, the rocks under the upstream face of the dam are decompressed and under the downstream face they are compressed; Translated from Gidrotekhnicheskoe Stroitel'stvo, No. I, pp. 17-24, January, 1992.

0018-8220/92/2601-0023512.50

9

Plenum Publishing Corporation

23

"~20

320. EIO-

"3';9

",YO0

,Y00-

.x. . . . .

~-295 ~1~0-170 o.,

\

"

7 90 N_-,sa..-sa-.._ -~,o, 1s v~ss-9o) ,;~o_,'~;Y s

/

290-

-290 '

[;,~,-;;~

, \"

r

280-

-280

2 70-

- 270 -2G8

260250 2~'0'

"g50

I pU3-zqs

I

"2~'0

~FI

230

220 2

,PL7-2~5

U-

I I

-~

- 230

i

-2zo

PLS-Z20 glO

210-

200-

t .._3

200

1Jo

190

Fig. 1. Seepage and deformation characteristics, ultrasonic and temperature anomalies in the dam foundation of section 18 (velocity sections along dam section 25): 1) permeability coefficient (m/day) at elevation of upper pool level (UPL) below 500 m; 2) same at elevation of UPL 540 m; 3) ultrasonic (Vp) and temperature (T) anomalies and year of recording them; 4) velocity of ultrasonic waves (krn/sec) and modulus of deformation (102 MPa) measured in October 1990 (UPL 540 m); 5) same in July 1985 (UPL 515 m); 6) intake of piezometer PL7-245. The results of calculating and modeling steady seepage in the dam foundation on schematized hydrogeological models not taking into account the changes and influence of the SSS, the prescribed calculated diagram of uplift on the base of the foundation; Domestic and foreign experience in constructing and operating grout curtains and galleries in foundations of high-head dams. This experience shows that in the foundations of high-head dams, as a consequence of the occurrence of tensile stresses and decompression of the rocks, discontinuities of the grout curtains and intensification of seepage in the foundation are possible (e.g., the Santa Maria dam in Switzerland, Schleigas and Kolnbrain in Austria, Bratsk and Ust'-Ilimsk in the USSR, etc. [1, 2, 3]). The underground complex at the Sayano-Shushenskoe hydrostation includes: a grout curtain the base and bank abutments of the dam, blanket grouting of the dam foundation, drainage curtain, blanket grouting and well drainage of the stilling basin, and powerhouse. The grout curtain (Fig. 5) was designed and made vertical, its axis is located at a distance of 15 m from the upstream face in the zone of small stresses. The depth of the curtain is about 100 m, it joins the rocks of the aquiclude and thus provides coverage of the front of the active region of seepage. To a depth of 60 m the grout curtain is 2-row (rows D and E) with a distance of 2 m between rows, and lower it is 1-row (row E). In the upper part, in the zone of contact with the concrete dam, the curtain is strengthened by three rows of contact grout holes (rows A, B, C). The spacing of the holes in the rows of the grout curtain and contact grouting is 3 m. Grouting was carried out by the method of gradual convergence of the hole in three phases. The grout curtain in the channel stretch was made from a gallery and in the banks from adits and galleries of the dam.

24

cs/sec

0 0 P,

,5~-0

517

;

5"2Z

I

~

!

e4es"D

'

,),s

t~ 0

4-1

I>

/ /

I

I

_,ol I I;..

.......

i
V/ k,

:At.

kl",

jl 1.

7986

i

7-987

, 198B

\ j ,,,, 1989

o

1990

Fig. 2. Seepage discharge in dam foundation: 1) total in foundation; 2) total drainage; 3) unorganized drainage in gallery No. 1; 4) drain wells in gallery No. 1; 5) drainage of left bank; 6) drainage of right bank; 7) elevation of upper pool level; 8) opening of contact joint in section 18 according to strain gauge 118-12.

With consideration of the expected decompression of the rock foundation under the upstream face, the watertight elements in the foundation of the channel part of the dam were broadened toward the upper pool by a 15-m-long concrete blanket and by a grout curtain made in the form of a cutoff under the blanket on the upstream side. Blanket grouting is provided for equalizing the modulus of deformation of the foundation and for increasing it on the weakened fractured stretches and was done on 75% of the area of the dam foundation with a depth of 30 m under the downstream third and fourth columns and 15 m in the middle part of the profile under the second column, in the foundation of the bank sections of the dam the depth is 10-25 m. The spacing of the holes of the blanket grouting is 4 m with convergence to 2 m in fractured zones. The drainage curtain is the main engineering measure reducing heads in the foundation and uplift on the base of the dam. It is located along the entire upstream face and is made from grout galleries and adits in the form of inclined holes (with a spacing of 3-6 m) from the grout curtain toward the lower pool with a depth of 45 m in the foundation of the channel dam and up to 80 m in its bank abutments. Hole drainage is carried out also under the draft tubes of the powerhouse and stilling basin at the abutment to the overflow dam. Grouting operations in the dam foundation were carried out as the field of operations was prepared, mainly on stretches of blanket grouting under the II-IV columns of the dam and contact grouting along rows A and B. Grouting along row B of contact grouting and D of the second row of the grout curtain was carried out after completing works on adjacent rows of holes. The upstream grout curtain was made as the blanket was prepared. The grouting pressure was taken so as not to hydraulically fracture the rocks and was from 0.5-0.7 MPa in the upper zones of the blanket grouting holes to 3-4 MPa in zones of grouting the curtain holes deep in the rock mass. Grouting of the rocks with a sufficiently high pressure and its increase with depth and in phases as the holes converge helped to achieve a maximum possible grouting density. Thus, after completing blanket grouting in the foundation of the II-IV columns of sections 14-50 of the dam, the average specific water absorption in control tests of 506 zones of holes was 0.017 liter/(min.m2), 76% of which had values of not more than 0.02 liter/(min.m2), 17% within 0.021-0.05 liter/(min'm:Z), and only 7% more than 0.05 liter/(min'm2). 25

~2051o -

i ;,~ ---"-"Y

t MO

v

2,90260 -

~290 I 280

Z50 -

r 250

,,o- t,../.,,

)

20D 1.qo780. 170150

r "~ ~"

[190

,// //,,7 ,

F

t - 160

"

~2 0 "

170

--

~

-

-

r320

l

-,YID

510 300

'

~ .~lJO

"/

"

.o_ 270-

575 ~

-270

$5 /

z6o25"0 -

5

....

zzo-

I280

/

.//i

t

-7

/

~

__

/

~.1

,'

I

. . . . . . .

~

!III ~,

i ii7

'%

~ , ,, ~

IU[~"

l

I

~

I |.,,,o ]

I " : ;<0

L

]GO

b

Fig. 3. Lines of equal heads in dam foundation: a) according to on-site data in section 18 at different reservoir levels; b) according to data of model - dashed lines for a permeability coefficient of the curtain of 0.01 m/day and of blanket grouting of 0.02 m/day; solid lines - the coefficients are shown on the model. On the basis of the results of blanketing grouting of the foundation, section 30 and stretches adjacent to it (sections 26-36) were singled out in the foundation of the II-IV column, in which passes a zone of tectonic crushing. Here in the foundation of the II-III columns the absorption of cement was 476 kg/m, which is 3.6 times greater than the average for the channel stretch of the dam (sections 14-50). Absorption of cement under the IV columns of sections 26-36 is also greater than the average values. Additional grouting was carried out within these sections 26-36 under the III-IV columns in 1987-1990. The specific water absorption in control holes in the base of the bank abutments of the dam is less than 0.02 liter/(min.m 2) in 70-75% of the cases and more than 0.05 liter/(min'm 2) in 5%. The grouting density in the body of the grout curtain (along rows D and E) is quite high, the specific water absorption in control tests on 90-95% of the area corresponds to a value of not more than 0.01 liter/(min.m2) and not greater than 0.03 liter/(min-m2) on the remaining part of the area (sections 29-36, 1, 67). The average absorption of cement for the channel curtain was 105, on the left bank 100, and on the right bank 70 kg/m. Thus, as a result of conducting grouting operations on blanket (area) and deep (grout curtain) grouting in the dam foundation and its bank abutments, the upper, most permeable (decompressed) zone was consolidated and a deep grout curtain was created with a permeability characterized by a permeability coefficient of not more than 0.01-0.03 m/day. The initial permeability of the rocks was thus reduced by 1-1.5 orders of magnitude as a result of grouting (also from the dam's weight) in the upper zone of the foundation and in the body of the grout curtain (Fig. 1). However, these hydrogeological characteristics of the grouted rock masses are only the initial ones; they correspond to the stage of constructing the dam and filling the reservoir, during which the stress-strain state in the foundation insignificantly changes the permeability of the rock mass. 26

h.33

\

X>

\

\

7"-I. 22 8

Fig. 4. Isopiestic-water-table contour lines in foundation and bank abutments of dams, discharge of drainage in gallery No. 1: 1) drainage; 2) isopiestic (water-table contour) line on September 27, 1990, UP 540.15 m; 3) same on August 27, 1984, UP 498.1 m: 4) same on August 2, 1982, UP 465 m; 5) same on May 13, 1981, UP 408 m; 6) diagrams of inflow of water into gallery No. 1 at elevation UP 539.92 m; solid lines: into drain wells; dashed lines: unorganized inflow. The reliability of the underground contour, the effectiveness of the watertight measures, the parameters of the seepage flow for various levels and fluctuations of the reservoir, and the effect of the SSS on permeability of the rock mass and on the seepage regime in the foundation and bank abutments of the dam have been being monitored since 1977 continuously by hydrogeological observations and model and geophysical investigations. A permanent network of piezometric and geophysical boreholes located at control sections every 50-100 m along the dam front was created for on-site observations and investigations. The piezometers at the sections were arranged so that (Fig. 1) they monitor the heads in the rock-concrete contact zone at 5-7 points of the cross section, front and behind the grout curtain at various depths, on the downstream side under the blanket grouting, and in zones under the up- and downstream faces of the dam. The geophysical boreholes are also arranged as control sections, in each of which are 5-7 boreholes with a depth of 45 m; together with the drain wells they enable conducting regular investigations in the foundation of the dam, including seismological investigations, ultrasonic logging, temperature logging, and flow measurements in boreholes. The seepage discharges are measured in individual drain wells, springs, totally in the channel part of the dam (sections 14-50), along the left and right banks, including along their levels. The works conducted by the All-Union Trust for Special Hydraulic Engineering Works (Gidrospetsstroi) are an integral part of the set of investigations and on-site observations; their results supplement the existing data on the state of the foundation and underground contour of the structures of the Sayano-Shushenskoe (SSh) hydrostation [4, 5]. It is known [1, etc.] that the SSS of a foundation constantly changes during construction and filling the reservoir. But of interest is not only the very fact of a change in the SSS but also the dependence on the SSS of the deformation and seepage characteristics the rock mass of the foundation, characteristics of the seepage flow, and effectiveness of the underground contour

27

~00q il ~_[._'l---h].._t 320

~En-

,

/

-. . . . .

J.

I I I "~

.~1/

;:~0-

420

~805-00-~ 52D

Fig. 5. Diagrams of the seepage head on the base of the dam: a) in section 18 at various levels of the upper pool (UP): I) August 27, 1984, UP 498.1 m; 2) August 9, 1988, UP 533.94 m; 3) September 13, 1990, UP 539.91 m; 4) according to the building code; b) at an elevation of the upper pool level of 539.51 m: 1) section 18; 2) section 25; 3) section 33; 4) section 39; 5) section 45; b) according to the building code.

and watertight elements. The seepage characteristics are most sensitive to changes in the SSS and can serve as an "indicator" of the state of the underground contour of the dam. The period of construction and filling the reservoir can be divided into two stages according to the character and rate of change in the deformation characteristics and permeability of the rock mass and seepage in the foundation: the first up to an elevation of 480-510 m and the second up to the elevation of the normal pool level (540 m). The first stage (1978-1984) is characterized by the fact that mainly compressive stresses occurred in the foundation during the entire period, fluctuations of the reservoir level did not lead to substantial changes in the permeability of the foundation (Table 1), seepage discharge (Fig. 2), or distribution of heads in the underground contour (Fig. 3) - practically the entire head died out in the stretch from the upstream face to the drainage. At elevations of the reservoir level up to 480-510 m, only a slight increase of permeability and heads were recorded under the upstream face of the dam and they again decreased when the reservoir level was drawn down. The average permeability coefficient of the grout curtain according to the on-site data of this period was equal to: K e = Q/hcBI = 0.0055 m/day, where Q is the drainage discharge gallery No. 1 in the powerhouse stretch with length B = 300 m; h e = 100 m is the depth of the grout curtain; I is the average gradient of the head in the curtain. The density of the curtain, characterized by a permeability coefficient of 0.006-0.010 m/day, remained stable along the front and depth of the curtain during the first stage. The state of the foundation can be characterized as stable on the basis of the seepage parameters.

28

TABLE 1. Permeability Coefficient of Soils in Zone of Intakes of Piezometers Permeability coefficient, m/day, at UPL, m Piezometer

18PU1 18PU2 18PL4 18PL5 18PL8 18PLll 18PL13 25PU1 25PU2 25PU3 25PL6 30PL8 33PU 1 33PU2 33PU3 39PL7 45PUI 45PU2 45PU3 PC45 45PL6 45PL9 PC46 PC47 50PU 1 50PU2 50PL4

Elevation of intake, m

305 288 295 303 220 305 305 305 288 245 345 300 305 288 245 220 305 288 245 305 275 3O5 305 305 385 292 330

29.04.82 403.6

20.04.83 382.7

0.011 0.11 -

0.009 0.125 0.027

12.10.90 539.0

12.04.89 498

30.05.85 482.3 0.016 0.054 0.0136

0.06 0.21 0.032 0.086 0.11

0.001

0.011 0.14 0.031 0.022 0.026 0.085

0.080 0.190 0.030 0.095 0.301 0.18 0.115

0.24 0.037 0.042 0.203 0.23 0.078 0.009 0.053 0.038 0.002 0.002 0.083 0.108

1.402

0.119 0.013 0.06 O.O07

0.018 0.018 0.006

0.096 0.093 0.007 0.031

0.,92 0.0088 0.069

0.009 0.007

0.288 0.02 0.007 0.113

0.0019

0.0026

0.137

0.91 1.05 0.046

0.058 0.049 0.0018 0.0018 0.1 0.143 0.04 0.07

OO0533

0.039 0.0079

0.024

0.034

0.042

0.050 0.230 0.315 0.082 0.004 0.053 0.045 0.0023 0.0018 0.070 0.114 0.146 0.275 0.218 1.08 0,124 0.120

0.003 0.0325 0.05 0.054 0.009

TABLE 2. Seepage Discharges in the Powerhouse Stretch Measurement date

UPL, m

Average acting head, total, liters/see

in

24.06.86 23.12.86 08A0.87 24.04.90 26.06.90 13.08.90 18.09.90

498 499.03 530 499.61 520.14 534.02 539.92

Seepage discharge

188 189 220 189.6 210 224 230

15.3 14.6 26.3 16.6 29.60 79.60 107.60

reduced, liter/(sed-min)

averge per well, llters/see

0.0813 0.0772 0.1195 0.0878 0.141 0.3554 0.04678

0.7285 0.695 1.25 0.790 1.409 3.791 5.124

The second stage (1985-1990) is characterized by the occurrence of tensile stresses under the upstream face, decompression of the rocks in the foundation of the first column of the dam and other places of the cross section at various depths, increase of seepage in the drainage and from the abandoned grout holes, disproportional increase of the reservoir level, and increase and fluctuation of the heads in the foundation of the first column of the dam. The rock mass in this period was subjected to deformations to the entire depth of the geophysical investigations (to 45 m). Characteristics regions (Fig. 1) are distinguished in the contact zone with a depth of 3-7 m inside and near the lower boundary of blanket grouting in the directions of large joints and tectonic disturbances. A change in the properties of the rocks is recorded by all types of geophysical monitoring and encompasses a considerable part of the foundation. Maximum decompression is noted under the upstream face, its depth increases with increase of head. An increase of the permeability coefficient in various degrees (by 2-5 times and more) is recorded by direct measurements (Table 1) in the contact zone of the 29

TABLE 3. Gradient of the Head in Cross Sections of the Grout Curtain Gradient ofhead in sections Level, m 18

[

25

33

39

45

11.1 15.8 11.4 5.0

12.27 13.93 10.57

12.52 13.9 11.24

12.7 21.16 11.88 7.87

13.18

11.16

11.88

13.69 12.98

13.78

13.82

11.27 10.44

14.01 13.45

13.45

13.82

10.95 8.45

14.76 13.58

25.12.1986 UP 498.36

LP 322.20

8.15 7.15 6.32 3.27

I 1.27 11.96 11.22

09.03.1988 UP 533.94 LP 325.45

10.2 11.2 8.0 3.42

m

13.08.1990 UP 533.72 LP 322.64

12.2 12.0 11.0 4.02

13.4 26.16 11.93 7.08 13.09.1990

UP 539.51 LP 323.64

11.19 11.97 10.8 4.08

12.9 25.16 11.47 5.6

foundation under the first column of the dam in zones of the blanket grouting and curtain, at a depth to elevations of 245-220 m (sections 18, 25, etc.) on both sides of the grout curtain. The total seepage discharge in the dam foundation in 1988 at an elevation of the UPL of 534 m was (Fig. 2) 180 liters/see, in the channel part of the dam from the drain wells 80 liters/sec and from the unorganized inflow up to 84 liters/see (the noncoincidence of the peaks (Fig. 2) of the discharge of the drainage and unorganized seepage is explained by the placement of plugs in the outpouring wells in this period), in the left bank 42 liters/see, and in the right bank 20 liters/see. Grouting of places of water shows (wellheads) in 1988-1989 practically eliminated the unorganized inflow and reduced the seepage discharge, which at an elevation of the UPL of 526 m was 128 liters/see in 1989 versus 160 liters/see in 1988. An increase of the reservoir level to 540 m in the final stage of its filling (1990) caused further decompression of rocks and an increase of their permeability, especially in the contact zone of the foundation under the first and partially second columns of the dam. The permeability coefficient of the grouted rock mass that earlier had a value of 0.01 m/day increased in the contact zone to 0.10-0.30 m/day and more (Table 1, Fig. 1). The zone of decompression is developed especially in the stretch of contact grouting and at high UPLs intersects the grout curtain and here and there the drainage line in gallery No. 1 (it has a length up to 20-25 m). The permeability over the depth of the rock mass was determined by measuring the flow rates of the piezometers (Fig. 1) installed on both sides of the grout curtain at a depth to 100 m. Opening of the contact joint, cracking, and decompression of the rocks in the upper part of the foundation led to an increase of the drainage discharge in the channel stretch of the dam (Table 2). Let us examine in what zones in the dam foundation the seepage discharges increase. The picture of reduction of the head of the seepage flow and its gradients in sections of the grout curtain at the control sections (Fig. 3a, Table 3) changes with increase of elevation of the UPL from 498 to 540 m, the gradient of the head in the curtain increases in the upper and lower cross sections on average by 1.5-2, amounting at 540 m to 11-14 in the upper cross sections of the curtain and to 4-8.45 in sections 18, 25, and 33 and to 13.6 in section 45 in the lower cross sections. In the drainage zone I = 26 in individual cross sections of the curtain.

30

The average permeability coefficient of the curtain at an UPL of 534-540 m was within 0.024-0.041 m/day. However, its value is not constant over the depth (Fig. 1). A change (increase) in the permeability coefficient of the rock mass occurs due to its decompression and opening of cracks and change in the structure (density) of the mass in zones of tectonic disturbances and jointing under the effect of the SSS. Deformations have a distributed character, as is shown on the geophysical section (Fig. 1). The permeability coefficient of the grout curtain in the 30-100 m interval of depths (below the blanket grouting) is determined from the condition that within these depths the rise of the foundation in the vicinity of the curtain was about 2 mm (according to the data of the subsurface marlcs) and the width of individual joints is 0.2 turn; then [6] /(= .~!6::,/12vh (l- I .4/8) ~.0,0072

where b is the width of the joints; ~ is the coefficient of kinematic viscosity; hj is the distance between joints, taken equal to 10 m; A is the roughness parameter. With consideration of the initial permeability we obtain K = 0.0132 m/day, and we consider it possible to take 0.015 m/day as the calculated permeability coefficient of the curtain at depth below the blanket grouting at an elevation of the UPL of 540 m. Analogous calculations for the upper zones near the base of the dam give higher values of the permeability coefficient, especially for the central sections. Maximum decompression and increase of permeability and seepage discharge are confined to the contact zone of the foundation, where tensile stresses act under the first column of the dam and opening of the contact joint in the monitored sections reaches several millimeters, for example, more than 4 mm in section 18 (Fig. 2). A rise (decompression) of the rocks in the second column of sections 25 and 30 on the downstream side of the drainage of up to 2 mm is recorded by deformometers with a measurement base of 2000 mm in the contact zone and up to 0.5 mm at a depth of 10 m. Real decompression in the contact zone is represented in the form of opening of the existing grouting-sealed joints, formation of new cracks in the rock mass, and opening of the concrete-rock contact joint. The water-conducting system has a different origin, nonuniform distribution, and variable parameters of the joints in the decompressed rock mass. If it is assumed that one of the largest water-conducting systems conveys the seepage discharge in the contact zone or the system of channels is reduced to one joint, then for the seepage discharges recorded on October 11, 1990, we can obtain the following widths of joints at the outlet of the decompression zone into the drainage: 0.51 nun in section 18 with a discharge of 4.87 liters/see, 0.58 mm in section 20 with a discharge of 7.98 liters/see; and 0.73 mm in section 34 with a discharge of 18.25 liters/see. If we consider that jointing is uniform and distributed in a decompressed contact zone with a thickness of 3-5 m, then permeability coefficients of the zones of 0.83-0.5, 1.23-0.77, and 2.95-1.77 m/day correspond to the indicated discharges. In reality the widths of the joints and permeability coefficients on the greater part of the foundation will be somewhat less, since up to 1/3 of the discharge passes through lower parts of the curtain, but in some place they can be greater, if the nonuniform distribution of the discharge between holes is considered. The seepage discharge along the drainage front and over the depth of the drain wells is nonuniform, the discharge in the sections (Fig. 4) varies from 0.6 (section 41) to 18.40 (sections 28, 34) with an average of 6.7 liters/see. The maximum discharges in individual wells reach 5-6 liters/see (sections 28, 34, 35), whereas at an elevation of the UPL of 498 m they were not more than 1 liter/see. We note for comparison that at the Chirkey hydrostation at an elevation of the UPL corresponding to the NPL, the maximum discharge of the drain wells in the right-bank stretch of the dam foundation is at the level of 4-5 liters/see. At the Inguri Hydrostation, the discharge of many drain wells is greater than the maximum discharges of the drain wells at the SSh hydrostation. However, the nature of the increased discharges of the drain wells at these hydrostations is different: at the SSh hydrostation they are related to decompression of rocks and opening of the contact joint at high heads, at the Chirkey and Inguri hydrostations they are related to the presence of large joints and tectonic disturbances and bypass seepage flow along them. The results of the effect of a change in the SSS of the foundation during filling of the reservoir above an elevation of 500 m can be traced from the changes in the temperatures and ultrasonic velocities and change in the location of their isohypses and values of the moduli of deformation (Fig. 1). These data indicate that deformations occur at the entire depth of investigations in the form of the formation of individual zones scattered in space and confined to the contact zone, lower boundary of the blanket grouting, tectonic disturbances, and to joints. Deformation zones were recorded at low elevations of 31

the UPL and then they intensified or weakened during fluctuations of the level and filling of the reservoir, but generally they kept their location. The ultrasonic velocities and moduli of deformation decreased by 10-20% with rise of the UPL to the design elevation in comparison with the initial. The character of the change indicates the presence of decompression under the first column to a greater extent than at other places. Despite the presence in the foundation of individual occurrences related to a change in the properties and characteristics of the rocks under the effect of the SSS, we can note that the velocities of the ultrasonic waves do not drop below 4.2 km/sec and the moduli of deformation not below (1.0-2.0) x 104 MPa, with average static values of (1.5-1.7) x 104 MPa, which is higher than the calculated. The presence in the section of zones (lines) of deformations oriented toward the lower pool shows that the process of deformation (mainly decompression) occurs discretely in the mass, in the form of opening of individual joints or local jointing zones. The changes in character of permeability of the rock mass during filling of the reservoir led to insignificant changes in the distribution of heads in the underground contour of the foundations both in section (Fig. 2a) and in plan (Fig. 4). Grouting operations, pumping out the stilling basin, nonuniform operation of the turbine-generator units, and other factors unconditionally had a definite effect on the character of distribution of the heads. Nevertheless, with increase of head, the permeability increased and the seepage resistances of the rock mass under the first column more intensely than in other places, the heads under the curtain increased, and the heads and distribution of the seepage pressure remained practically the same on the downstream side of the drainage. Only in the last stage of filling the reservoir did the heads in the bank abutments of the dam (sections 18, 45, 50, etc.) on the downstream side of the curtain increase by 5-7 m (Fig. 5a). On the whole the distribution of the seepage pressure on the base of the dam in all sections remained lower than according to the building code (Fig. 5b), but a rise of the head is noted in section 18 behind the drainage line (Fig. 3a), which is related to opening of the contact joint, effect of the zone of decompression of the bank abutment, and bypass seepage. A decrease of the head is observed in sections of the powerhouse part of the dam at the outlet of the flow into the lower pool owing to drainage of the draft tubes, which operate here in a vacuum (suction) regime. The picture of the distribution of heads in cross sections of the foundation was thoroughly studied for various permeability of the foundation and grout curtain by solving seepage problems on a computer by the finite-element method. Figure 3b gives lines of equal heads in the cross section of the powerhouse part of the dam with a permeability of the foundation most closely corresponding to real conditions at an elevation of the UPL of 540 m (solid lines) and for the initial permeability (dashed lines) under the assumption that it does not change with increase of elevation of the UPL. Comparing the model and prototype lines of equal heads (Fig. 3a and b), we can note their close correspondence, i.e., conditions under which the character of distribution of heads in the outlet stretch of the flow is the same are found. The model investigations show the following. The drain wells in gallery No. 1 play the main role in reducing the heads in the underground contour of the channel part of the dam; with the drainage disconnected, reduction of the head on the grout curtain does not exceed 30-35%, in the case of combined operation the curtain protects the drainage from high inflows, especially at elevations of the UPL close to the NPL, and drainage provides reduction of the heads. The gradient of the head in the curtain decreases with depth, and it is 10-15 in the upper part and 6-8 in the lower; the linear seepage discharge in the channel cross section of the dam for the model permeability of the foundation (Fig. 3b) and elevation of the UPL of 540 m is equal to 40.65 m2/day and the total within sections 16-48 is 235 liters/sec. For an initial permeability coefficient of the grout curtain of 0.01 m/day, the total discharge in the drainage of the channel part of the dam would be equal to 69 liters/sec. CONCLUSIONS Seepage in the foundation of the SSh hydrostation has a nonuniform but regular character and substantially depends on the acting head and SSS. At an elevation of the UPL of less than 520 m, a change in the permeability of the rock mass is observed mainly on the upstream side of the grout curtain. The intensity of decompression of the rocks increases with increase of elevation of the UPL, especially above 534 m. At 540 m the zone of decompression of the contact part of the foundation in some places intersects the grout curtain and in individual cases intercepts the drainage; the discharge of the drain wells increases by 5-7 times in comparison with the discharge at an elevation of the UPL of 500 m. The width of the joints near the drainage (according to calculation) averages from 0.3-0.5 to 0.73 mm; here and there it can reach 1.5-2 mm. The average gradient of the head in the grout curtain is within 10-15; at individual places near the drainage it reaches 26, which is at the level of the allowable for a rock mass.

32

A method of controlling seepage is mainly used in such cases - repeated grouting at low elevations of the UPL, including with the use of plastic grouts, which in individual cases accelerate setting of the additives. As measures for improving the seepage conditions and quality of monitoring, we can recommend plugging places of water shows through the grouting holes at an elevation of the UPL of 500-510 m, intensifying drainage in stretches of increased inflows and passages of the head beyond the drainage line, and broadening of geophysical and hydrogeological investigations for a detailed study of the character of change in permeability and seepage during fluctuations of the reservoir, especially at high elevations. During operation of the dam it is necessary not to exceed the attained rates and level of filling the reservoir (elevation 540.25 m), since at levels close to the NPL, the seepage discharge increases intensely. The underground contour of the Sayano-Shushenskoe dam is of high quality and operates effectively and reliably under conditions of a varying stress-strain state of the foundation during fluctuations of the reservoir level. LITERATURE CITED .

2. 3. 4. 5. 6.

V. N. Durcheva, On-Site Investigations of the Solidity of High Concrete Dams [in Russian], MI~AI (1988). V. Demmer and H. Ludescher, Measures to Reduce Uplift and Seepage on the Kolnbrein Dam [Russian translation], Fifteenth Congress, 58, 81 (1985). V. Tischler and I. Schlosser, Sealing the Foundation of the Schlegais Arch Dam [Russian translation], Fifteenth Cognress, 58, 62 (1985). 1~. K. Aleksandrovskaya, "Static work of the Sayano-Shushenskoe dam in the last stages of filling the reservoir," Gidrotekh. Stroit., No. 10 (1988). V. A. Ulyashinskii, S. N. Starshinov, and V. V. Tetel'min, "Mechanism of opening of the concrete-rock joint at the Sayano-Shushenskoe dam," Gidrotekh. Stroit., No. 12 (1989). V. N. Zhilenkov, Guide to Methods of Determining the Seepage and Piping Properties of Rock Foundations of Hydraulic Structures [in Russian], I~nergiya, Leningrad (1975).

33