Journal of Rock Mechanics and Geotechnical Engineering.

2011, 3 (4): 329–342

Key issues in rock mechanics of the Three Gorges Project in China

Qixiang Fan*, Hongbing Zhu, Xuchun Chen
China Three Gorges Corporation, Yichang, 443002, China
Received 29 April 2011; received in revised form 4 July 2011; accepted 26 July 2011

Abstract: The Three Gorges Project is one of the essential key projects for flood controlling and water resources regulation in
the Yangtze River. The project includes a river-crossing dam, underground powerhouses, and navigation structures. Because of
the huge size and complicated construction technologies, the project faced a series of challenging engineering issues. In terms of
rock mechanics, there are many key technical issues, including the sliding resistance and stability of the dam section along the
foundations of powerhouses No.1–5, the slope stability of the double-line five-stage shiplock, excavation of large-scale
underground powerhouses, and curtain grouting under the dam. With decades of scientific research and 16 years of practical
construction experiences and reservoir operations, these key technical issues in construction of the Three Gorges Project are
successfully resolved, which will attribute to the development of hydropower technology. On the basis of the monitoring data during
construction and normal operation periods of the Three Gorges Project, this paper presents a systematic analysis of these key rock
mechanical issues in terms of behaviors, solutions, dynamic controlling, monitoring arrangement and integrated assessment.
Key words: Three Gorges Project; rock mechanics; dam sliding resistance and stability; high shiplock slope; underground
powerhouses; curtain grouting

1 Introduction


1.1 Description of the Three Gorges Project

The Three Gorges Project is located in Yichang City,
Hubei Province, China. The project has a normal pool
level of 175 m, a storage capacity of 3.93×1010 m3, and
a flood storage capacity of 2.215×1010 m3. It is able to
effectively regulate floods from the upper reaches of
the Yangtze River and protect the plains in the lower
reaches against floods. The Three Gorges dam, a
concrete gravity dam, has a crest height of 181 m and a
total axial length of 2 309.5 m. The hydropower
station has an installed capacity of 22.50 GW, with an
average electricity generation of 1.024×1011 kW·h in a
year, making it the world’s largest hydropower plant.
The double-line five-stage continuous shiplock
enables 10 000 t (ton) freighters to sail through in one
time and the one-stage vertical ship lift can accommodate
3 000 t ships in the tank. Figure 1 gives a panoramic
view of the Three Gorges Project after completion.
Doi: 10.3724/SP.J.1235.2011.00329
*
Corresponding author. Tel: +86-717-6276666;
E-mail: fan_qixiang@ctgpc.com.cn
Fig.1 A panoramic view of the Three Gorges Project after
completion.
The Three Gorges Project was built in three phases.
Official construction started on December 14, 1994,
and river closure was completed on November 8, 1997,
and closure for the diversion channels was completed
on November 6, 2002. In June 2003, the third phase
stated as the roller compacted concrete cofferdam was
put into service to block the river, and the water level
in the reservoir rose to 135 m. After that, the
double-line five-stage shiplock began trial operations.
On July 18, 2003, the first group of generating units
was connected to the power grid and began to operate.
On May 20, 2006, the dam reached its designed
elevation of 185 m. In October, water level in the
reservoir rose to 156 m, and the hydropower complex
began its initial operative phase. On June 11, 2007, the
first group of generating units in the right bank
powerhouse started their commercial operation. In late
2008, all of the twenty six 700 MW hydro-turbine
330
generating units in the powerhouses on the left and
right banks of the river behind the dam were put into
operation. Thus, the hydropower complex was
completed (excluding the construction-postponed ship
lift), and the reservoir was ready to impound to the
normal level of 175 m. The Three Gorges Reservoir
started a trial water storage at the end of rainy season
in 2008, and in December 2010, the Three Gorges dam
had experienced three years of trial water
impoundment. The water level in the reservoir reached
172 and 175 m after the end of rainy season of 2009
and 2010, respectively. All monitoring data and
analytical results show that the operation of the Three
Gorges Project is normal. Since its water
impoundment from 2003, the Three Gorges Project
has shown comprehensive benefits in flood control,
power generation, navigation, water supply, and
environmental protection, greatly contributing to the
sustainable development and steady growth of the
Chinese economy.
1.2 Key rock mechanical issues
Key rock mechanical issues of the Three Gorges
Project include the sliding resistance and stability of
the deep layers of the dam sections of powerhouses
No.1–5, the slope stability in the double-line five-
stage continuous shiplock, the excavation of large-
scale underground powerhouses, and the curtain
grouting under the dam. To cope with these technical
issues, numerous field geotechnical investigations and
scientific experiments were conducted in the 1950s,
allowing the engineers to identify the characteristics
and fundamental parameters of the rock masses in the
dam area. In addition, a number of research institutes,
universities, design institutes and several foreign
Fig.2 Typical foundation profile at the dam section of left-bank powerhouses No.1–5 and dam foundation reinforcement (unit: m).
Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342
experts were engaged to the scientific research,
computational analyses and optimization design of the
project. Moreover, the required productive
experiments were conducted, and related feedback
information was achieved. With these efforts,
systematic and scientific technical requirements for
engineering design and construction of the project
were formulated [1–6].
2 Sliding resistance and stability of
left-bank powerhouses No.1–5
2.1 Solutions
The dam sections of powerhouses No.1–5, on the
left bank of the Three Gorges dam site, are located in
the riverfront slopes of mountainous region. The host
rock masses in the dam foundation are primarily
composed of fresh flash-cloud plagioclase granite,
containing crevices with long and large low-angle dips
with an occurrence of NE-NNE, inclined SE (inclined
towards the left bank of the lower reach of the river),
as well as a small number of crevices with
intermediate inclination towards the lower reach of the
river. The powerhouses are arranged behind the dam.
The dam section of the powerhouses has a foundation
elevation of 90 m. The powerhouses have an elevation
of 22.2 m, resulting in a steep slope behind the dam,
which has a gradient of 54°, a temporary height of
67.8 m and a permanent height of 39.0 m, as shown in
Fig.2. These factors make it possible for the deep
layers at the dam section of the left-bank powerhouses
No.1–5 to slide.
Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342 331

To meet the required stability of deep layers in the the dam foundation ranged from –0.7 to 0.6 mm. After
dam, the following engineering measures were reservoir impoundment, the variation slightly
introduced: (1) slightly reducing the foundation increased, ranging from –0.8 to 3 mm. After the
elevation, and setting a cutoff trench at the dam heel; experimental water impoundment in the Three Gorges
(2) increasing the width of the dam bottom towards Reservoir, the water level reached 175 m in 2010, and
the upstream, and moving the curtain water drainage the variation in horizontal displacements of the dam
forward; (3) setting up a contact grouting system foundation ranged from 1.89 to 2.69 mm, with an
between the powerhouses and the rock slope of the annual change of 0.44–0.88 mm. The horizontal
dam to ensure the integrity of concrete and the rock displacements of deep foundation rock masses for the
slope; (4) setting keyways at the cross joints on the left-bank powerhouses No.1–5, which have an
dam section of the left-bank powerhouses No.1–5, and elevation below 95 m, ranged from –1.27 to 3.3 mm,
filling these joints with grouting to strengthen the and the increment in the horizontal displacements at
overall functions of the dam sections; (5) carrying out the foundation drainage tunnels before and after the
solid grouting in certain locations of the tail water experimental impoundments, in 2008 and 2010,
channels of the hydropower plant to ensure the ranged from –0.09 to 0.17 mm. The annual increments
functions of the downstream resistance body; (6) in displacement are basically consistent. The minimum
adding prestressed anchorage cables on the crevices increment occurred in 2010. Foundation deformation
with long and large low-angle dips from the on the dam section of the left-bank powerhouses
downstream slope surface; (7) considering a down- No.1–5 is converged.
stream drainage gallery with an elevation of 26 m 2.2.2 Monitoring of settlements at dam foundation
below the foundation of the dam section of the Figure 4 shows the measured settlements of the
left-bank powerhouse No.3; (8) adding 3 000 kN-grade upstream and downstream grouting galleries of the
prestressed anchorage cables on the structural plane of left-bank powerhouses No.1, 3 and 5.
crevices with long and large low-angle dips at the LD01ZC013 (No.1)
left-bank powerhouses No.1–3; and (9) carrying out 0
Displacement (mm)

backfill concrete treatment to the unfavorable rock 5 LD17ZC053 (No.5)

structural plane that protrudes above the surface of the 10
LD11ZC033 (No.3)
15
dam foundation. With these structural design and
20
foundation treatment measures, the dam and its rock
25
foundation have an anti-sliding factor of safety above
1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010
4.0, which can meet the design requirements and
Date
provide a considerable safety margin. (a) Upstream.
2.2 Monitoring results of the dam sections of the LD02ZC013 (No.1)
0
Displacement (mm)

left-bank powerhouses No.1–5 5

2.2.1 Monitoring of horizontal displacements at the 10 LD02ZC053 (No.5) LD01ZC033 (No.3)
dam foundation 15
Figure 3 shows the measured horizontal 20

displacements of the foundation for dam sections of 25

1997
1998

1999

2000

2001

2002

2003
2004

2005

2006

2007

2008
2009
2010

left-bank powerhouses No.1, 3 and 5.

Date
5 EX03ZC033 (No.3) (b) Downstream.
Displacement (mm)

3 Fig.4 Time-dependent displacement curves of the upstream and

1 downstream foundation galleries of left-bank powerhouses
1 IP02ZC05 (No.5)
No.1, 3 and 5.
IP02ZC01 (No.1)
3
5 It can be observed from Fig.4 that the upstream
2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

foundation grouting gallery had a settlement of

Date 18.88–23.48 mm in December 2010, with an annual
Fig.3 Time-dependent curves of horizontal displacements of the change of 0.81–1.09 mm, while the downstream
dam foundation from 2001 to 2010.
foundation gallery had a settlement of 20.83–22.94
It can be observed from Fig.3 that before the mm in December 2010, with an annual change of
commencement of water impoundment in the Three 1.03–1.32 mm. It is clear that the two adjacent dam
Gorges Reservoir in 2003, horizontal displacements of sections have approximately the same settlement,
 332 Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342

typically about 0.5 mm, indicating that there was no been basically stabilized since 2001.
uneven settlement in the dam foundation. Statistical analyses of the downstream horizontal
2.2.3 Monitoring of deformations inside rock displacements of the left-bank powerhouses No.1 and 5
foundation indicate that the downstream horizontal displacements
It can be observed from Fig.5 that horizontal in the foundation can be divided into two parts, one is
displacements inside the deep rock masses (with a approximately 1.4 mm induced by water pressure and
depth of 45–55 m) measured by a clinograph were another one is less than 2.5 mm induced by temporal
small. Measuring holes in direction A (upstream and effect. The temporal effect component varied mildly,
downstream) had a maximum cumulative downstream but not completely constrained. It is consistent with
displacement of 15.6 mm, and the rest had a the pattern of changes in the temporal effect
displacement of –5.8–9.3 mm. The measured values component at the foundation galleries of right-bank
fluctuated around the median value, while powerhouses No.24–26 and left-bank powerhouse
measurement had a measuring error of ±6 mm. No.14. The incomplete displacement constraint of the
Instrument IN1CF5 did not detect any dislocation of temporal effect is largely attributed to the relatively
the rock strata in a hole at depth of 50 m, indicating short duration of impoundment in the reservoir.
that the rock masses remained stable. Therefore, further monitoring is needed. Generally, the
rock masses at the foundation of the left-bank
20 powerhouses No.1–5 are stable.
Displacement (mm)

15 2.2.4 Monitoring of deformation at dam crest

10 During the experimental impoundment, under the
5 effects of temperature drop and water pressure, the
0 crests of the left-bank powerhouses No.1–5 had a
5 maximum downstream horizontal displacement,
1996

1997

1998

1999
2000
2001

2002
2003
2004

2005
2006
2007

2008
2009
2010

ranging from –3.07 to 10.32 mm, and the

Date displacement increment in dam crest before and after
Fig.5 Time-dependent curves of cumulative horizontal impoundment ranged from 6.66 to 7.3 mm, which was
displacements of holes inside the rock foundation of the dam
section of left-bank powerhouse No.4.
smaller than the deformation at the crest on the
riverbed, but was consistent with the changing patterns.
Displacement of the none-steel-pipe segment of the The downstream horizontal displacement of the dam
left-bank powerhouse No.5 was measured using crest was largely attributed to temperature changes.
multi-point borehole extensometers. The typical 2.2.5 Monitoring of stresses of anchorage cables
cumulative displacement curves are shown in Fig.6. The slopes behind the left-bank powerhouses
No.1–5 were reinforced with prestressed anchorage
1.2 cables and rods. Figure 7 shows the results of the
measurements of anchorage cables in pipe trench No.5
Displacement (mm)

0.9
0.6 of the left-bank powerhouse using a dynamometer.
0.3 SF1CF05 No.5 (88.0 m) Tonnage: 3 030.5 kN
#1 (5.0 m) #2 (10.0 m)
0.0 31 Loss rate 2 950
#3 (25.0 m) #4 (35.0 m)
25 2 790
Loss rate (%)

Force (kN)

0.3
2 630
1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

19
13 Anchoring force 2 470
Date
7 2 310
Fig.6 Time-dependent curves of cumulative displacement at 1 2 150
each measuring point measured by multi-point borehole
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010

extensometer.
Date
Fig.7 Time-history curves of force changes in anchorage cable
It can be observed from Fig.6 that horizontal measured by a dynamometer.
displacement of rock masses was small at the depth of
35.0 m, and had a springback of less than 1.1 mm in It can be observed from Fig.7 that over 1.5 years
the downstream direction. The other eight boreholes after installation, the prestressed anchorage cables
also had a small magnitude of deformation, ranging experienced a considerable loss of prestress.
from 0.6 to 2.3 mm with an annual change of –0.03– Afterwards, the loss rate decreased, and the prestress
0.06 mm. The displacement curves indicate that the change had a close relation with temperature, i.e.
deformation of the rock masses behind the dam has rising when temperature went up and falling when
Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342 333

temperature went down. After 2006, the prestress
remained stable, indicating that the bedrocks behind
the dam were stable.
2.2.6 Monitoring of drainage and seepage pressures at
foundation
In a global sense, the dam sections of the left-bank
powerhouses No.1–5 experienced insignificant changes
in seepage pressure before and after experimental
impoundment in each year. The seepage increased
slightly after experimental impoundment and
exhibited an upward tendency. The uphole of drainage
tunnel No.1 had a flow of less than 2.0 L/min. It was
generally free water, indicating that the dam
foundation behind the main drainage curtain at an
elevation of 74 m (and upward) was dry. The drainage
tunnel had a total seepage rate of less than 135 L/min.
The total seepage declined slightly after 2003, and the
drainage tunnel had a measured seepage rate of
approximately 85.1 L/min on November 30, 2008,
77.5 L/min in January 2010, and 66.85 L/min on
October 30, 2010. Preliminary analysis suggests that
the steady decline in total seepage at the dam
foundation is basically attributed to the following
factors: (1) silting was increased in front of the dam;
and (2) seepage crevices inside the bedrocks became
compacted due to the water pressure from the
reservoir basin filled up by silt. The steady decline of
total seepage at the foundation of the Three Gorges
dam was basically consistent with the variation
patterns of seepage in other projects after
impoundment. Water levels in the piezometer tube at
geological conditions. A water pressure distribution
chart was prepared on the basis of actual
measurements after impoundment at 175 m, as shown
in Fig.8. It can be observed from Fig.8 that the
measured groundwater level is lower than the slip
plane; therefore, no impact will be induced on the
stability of the dam body. It also shows that the
measured uplift pressure is smaller than the designed
value. From 2008 to 2010 after experimental
impoundment at 175 m, the measured uplift pressures
at the foundation of the left-bank powerhouse No.3
were 48.81%, 49.38% and 49.6% of the designed
uplift pressures, respectively. The measured uplift
pressure was smaller than the designed uplift pressure.
Analysis of data obtained from piezometer tubes on
the left-bank powerhouses No.1–5 shows that the
uplift pressure coefficients behind the upstream and
downstream curtains at the dam foundation are smaller
than corresponding designed values of 0.25 and 0.30.
The water level between the upstream and
downstream drainage tunnels is far below the
deep-layer slip plane. The water levels between
upstream and downstream drainage tunnels are
basically the same and do not change with the rise and
fall of the upstream water level. The measured uplift
pressures on the two typical deep-layer slip planes are
40%–56% of the designed value.
▽185
▽175 Sliding path
▽171.26 Designed pressure
Measured water level
Measured pressure
▽153.06
20+000.000

the left-bank powerhouses No.1–5 before and after

20+23.000

▽136.89
20+35.000

1 Unit: m
experimental impoundment in each year did not show
▽108
any significant change. The seepage pressure at the ▽106.6
2
20+118.000

3
dam foundation behind the main drainage curtain on ▽90 ▽91.10
A ▽81.7
the upstream left-bank powerhouses No.1–5 was ▽85
▽ 74 B ▽75
4
▽65 ▽65.4
below the elevation of 52 m. Actually measured
▽ 51.89 ▽ 53.44 C
▽ 50
seepage pressure in the deep holes of the piezometer F
▽42
5
175 m ▽36.47
tube in the middle of the foundation of the left-bank I
20+015.600
20+007.200

▽ 26.34 ▽ 25
▽23.5 ▽23.6
powerhouse No.3 was 49–52 m (41–38 m below the
169.750

I
110.000

foundation surface). No water was discharged into the

20+

20+

▽0

drainage holes of the drainage tunnel No.1, indicating

that the downstream dam foundations of the main
drainage curtain and the upstream of drainage tunnel
No.2 were basically dry, and that the water level at the
dam foundation was underneath the structural plane at
a low-angle dip.
Inside the dam foundation, there were crevices with
low-angle dips towards the downstream, thus the
designers assumed that a potential slip plane at depth
may exist inside the dam foundation in terms of
Fig.8 Distribution of uplift pressure at the foundation of the
left-bank powerhouse No.3.
The uplift pressure coefficient at the drainage
curtain of the upstream grouting gallery on the
left-bank powerhouses No.1–5 ranges from 0 to 0.11,
smaller than the designed value of 0.25. The
maximum seepage rate at the dam foundation is 67.96
L/min on October 31, 2010, which is smaller than the
334
designed value.
2.3 Integrated estimation
An integrated analysis, from Figs.2–8, suggests that
after the Three Gorges Project was completed and
began to impound in December 2010, the foundation
of the dam section of left-bank powerhouse No.5 had a
cumulative maximum displacement of 2.68 mm
towards downstream and –0.57 mm towards the left
bank. The foundation grouting gallery had a maximum
displacement of 23.48 mm. Surveys using clinometers
and multi-point borehole extensometers display that
rock strata that are 45–55 m below the foundation
surface remain approximately stable. The uplift
pressure coefficient at the discharge curtain of the
upstream foundation grouting gallery and the uplift
pressure at the dam foundation are smaller than the
corresponding designed values. The groundwater level
is lower than the elevation of the slip plane, indicating
that all reinforcement measures are effective. Analyses
of measured uplift pressure, using the definitive
sliding model at a water level of 175 m, and the
anti-sliding stability of a typical deep section of the
dam show that the deep section of the dam has an
anti-sliding stability coefficient slightly higher than
originally designed value. The anti-sliding factors of
safety under the two most unfavorable definitive
sliding models at elevations of 85 m (ABE) and 106.6 m
(ABCFI) on the left-bank powerhouse No.3 are 3.37
and 4.20, respectively. Therefore, the anti-sliding
stability in the deep layer of the left-bank powerhouses
No.1–5 can meet the design requirements [5]. This
indicates that, with the engineering measures for the
left-bank powerhouses, the anti-sliding stability
coefficients of the dam and its rock foundation can
satisfy the design requirements and have a
considerable safety margin [3].
3 Stability analysis of the high slopes
in the double-line five-stage shiplock
3.1 General description
The double-line five-stage shiplock of the Three
Gorges Project is hosted in a hill body on the left bank
of the Yangtze River. From its top, the hill was cut
along its ridge, forming two chute-shaped slopes. The
chutes contain the shiplock head and the sidewalls of
the chambers. Each chamber has the dimensions of
280 m×34 m×5 m (length × width × minimum water
depth), and spacing of the central axes of the chambers
is 94 m. The length of main structure of the shiplock is
1 621 m, with total excavation of 55×106 m3. The
highest slope reaches 170 m. The vertical walls of the
Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342
upper chamber range from 45 to 68 m in height, and a
57 m wide central divider stands between the two lines
of chambers.
The high slopes of the shiplock are resultant from a
deep cut of the hill, featured with large heights,
complex forms, a broad extent and full release of
stress, exhibiting heterogeneous characteristics. The
stability and small deformation of the slopes are
necessary for the normal operation of the miter gate.
The complicated geological conditions, including
faults and unstable monoliths, made the shiplock
construction more difficult. The shiplock construction
also encountered the interference of other works and
had to be completed with a tight schedule. Excavation
of deep and steep chambers with narrow floors had to
be carried out in accordance with excavation of
underground diversion tunnels and the valve shafts of
the shiplock. Moreover, the rock masses of the central
divider must keep intact. Both construction and
blasting operations must ensure minimal damage to
the rock mass and absolute safety [7].
3.2 Integrated solution to high slope stability in the
shiplock
After extensive research, an integrated solution was
adopted for the high slopes in the shiplock. The
instrument layout of cross-section 17-17 (stake
15+675) of the shiplock is shown in Fig.9 as an
example.
(1) Based on an integrated analysis of the
engineering geology, hydrological geology, field rock
mass dewatering tests, and 2D and 3D underground
seepage field tests of the high slopes, a scheme was
formulated for the underground drainage design of the
high slopes, which played a vital role in ensuring the
stability of the high slopes.
(2) According to the characteristics and actual
construction conditions of the high slopes, dynamic
design approach and methodology were introduced to
resolve various technical problems encountered during
construction.
(3) Blasting operations were strictly controlled, and
bolt bracing plus shotcrete was timely provided.
Prestressed anchorage cables were also promptly
applied, and systematic high-performance anchorage
rod and other construction reinforcement measures
were adopted.
(4) To ensure the stability of the rock masses in the
vertical slopes, various anchoring schemes were
considered. Eventually, 1 000 kN-grade double-
protection unbonded anchorage cables (with
corrugated tubes) and 3 000 kN-grade double-
protection terminal unbonded anchorage cables were
used [2].
Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342 335

Fig.9 Instrument layout of cross-section 17-17 (stake 15+675) of the shiplock (unit: m).

3.3 Monitoring slopes had maximum displacements of 71.57 and

Generally speaking, experimental impoundment in 53.90 mm, respectively, both towards the central axes
the reservoir has not exerted a significant impact on of the chambers. The southern and northern sides of
the deformation or seepage of the shiplock. The the central divider had maximum cumulative
surface deformation of the slopes in the shiplock, the displacements of 22.01 and 31.30 mm, respectively,
prestress of the anchorage cables, and the stress of the while those of the chambers were 6.08 and 4.71 mm,
high-performance anchorage rods all meet the respectively. The shiplock head had a maximum
requirements of design, and the measured data display displacement of 5.22 mm towards the chambers,
a stable state of the slope. indicating that the shiplock head was approximately
3.3.1 Displacement monitoring of high slopes and stable.
shiplock head During the process of chamber filling, the shiplock
In December 2010, the northern slope of the typical head had a displacement not more than 0.5 mm. Thus,
cross-section 17-17 (stake 15+675) in the shiplock had
it is believed that the normal operations of the miter
a maximum displacement of 50.18 mm and a
gates can be ensured, satisfying the requirements of
minimum of 36.61 mm, while the southern slope had a
design.
maximum of 21.15 mm and a minimum of 21.15 mm
It can be observed from Fig.10 that deformation of
(Fig.10). The rock masses in the southern and northern
the rock masses in the slopes primarily occurred
Excavation elevation (m)

TP29GP02 (245 m) TP09GP01 (170 m) during excavation and increased with increasing depth
295
Displacement (mm)

60 TP27GP02 (200 m) TP11GP01 (200 m)

50 260
40
Excavation outline
225
and further excavation. The deformation rate dropped
30
20
190 gradually after excavation was completed. The current
10 155
0 120 deformation rate is within ±2.5 mm per year,
10 85
indicating that the rock masses in the slopes are stable.
199401

199601

199712

199912

200312

200512

200712

200912
200112

Figure 11 shows the time-history curves of the

Date maximum displacement of southern and northern
(a) Northern slope.
slopes in the shiplock. It can be observed that the
TP29GP02 (245 m) TP28GP02 (230 m)
deformations of the high slopes were largely triggered
Excavation elevation (m)

TP98GP02 (160 m) TP28GP02 (170 m)

60 295
Displacement (mm)

Excavation outline
50 260 TP/BM39GP02, stake number: 15+850,
40
Excavation elevation (m)

225 80 200
30 elevation: 215 m, (No.3, south)
70 185
displacement (mm)

20 190
170
The maximum

10 155 60
50 155
0 120
10 40 140
85 TP/BM15GP01, stake number: 125
30
199401

199601

199712

199912

200312

200512

200712

200912

15+851
200112

20 110
Excavation elevation 95
10
0 80
Date
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

(b) Southern slope.

Fig.10 Time-history deformation curves of cross-section 17-17 Date
(stake 15+675) towards the chambers in northern and southern Fig.11 Time-history curves of the maximum displacement of
slopes of the shiplock. southern and northern slopes in the shiplock.
336 Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342

Excavation elevation (m)

by excavation unloading and increased with depth and 140 160
120 150
further excavation. After excavation was completed in

Stress (MPa)
100 140
80 Stress 130
September 1999, the deformation rate slowed down 60 120
and was gradually stabilized. In December 2010, the 40 110
20 Excavation elevation 100
northern and southern slopes of the shiplock had 0 90
20 80
maximum displacements of 53.9 and 71.57 mm,

199701

199901

200101

200301

200501

200701

200901

201301
201101
respectively. After the experimental impoundment to
elevation of 175 m in 2010, slope displacement varied Date
Fig.13 Time-history curve of stress changes in the anchorage
within an observational error of ±1.5 mm, indicating rods for cross-section 17-17 (stake 15+675) of the southern
that impoundment had no drastic effect on slope slope above the shiplock head.
deformation.
3.3.2 Monitoring of stresses of anchorage cables for excavation. The stresses of the anchorage rods,
shiplock slopes ranging from –116.38 to 136.65 MPa, had a negative
Figure 12 shows the time-history curve of stress relation with temperatures, and 52% of the total
changes in the anchorage cables at the cross-section anchorage rods had a stress less than 50 MPa. The
17-17 (stake 15+675) of the northern slope above the stresses of most anchorage rods fluctuated with
shiplock head. temperature, with an annual change of 5.07–57.81
MPa and accumulative stress change of 15 MPa. This
No: SF12GP01, stake number: 15+675, elevation: 180 m indicates that the anchorage rods and cables are
Anchoring force (kN)

3 000
2 840 effective in reinforcing the high slopes of the shiplock.
2 680
The displacement has been controlled since the
2 520
1 2 3 anchorage rods were installed, and the slopes were
2 360
2 200 stable overall [7, 8].
199701

199901

200101

200301

200501

200701

200901

201301
201101

3.3.4 Monitoring of drainage system of the shiplock

Measures for water interception, prevention and
Date
drainage have been considered for the high slopes in
Fig.12 Time-history curve of stress changes in the anchorage
cables for cross-section 17-17 (stake 15+675) of the northern the shiplock. Figure 9 shows the distribution of
slope above the shiplock head. groundwater levels in the piezometer tubes in the
drainage tunnels in various layers. Results indicate
It can be observed from Fig.12 that the stress that almost all of the water levels in the drainage
changes can be divided into three stages. The first tunnels (from the 2nd to 7th layers) were below the
stage, a stage of rapid stress loss, lasted for about six elevation of bottom plates, and that the variation in
months, during which the anchorage cables had an accumulative water level was less than 2 m. The
average prestress loss of 6.87%. The second stage, a groundwater level is lower than that proposed in the
stage of stress fluctuation, lasted for about 2.5 years, design for slope stability computations. For example,
during which the anchorage cables had an average the osmometers imbedded at the depth of 90–110 m of
prestress loss of 3.71%. The third stage was a stage of the drainage tunnel in the 7th layer indicated that the
periodic stress changes. The prestress increased with actual water pressure inside the rock masses was
increasing temperatures, and vice versa. After two merely 40% of the designed value. The rock masses
years, the prestress was basically stable. From the between the groundwater levels and slope excavation
measured results, the prestress of the anchorage cables faces on both sides of the slopes remain dry, and it is
on the high slopes had a loss rate of 2.9%–16.3%. In favorable for the stability of the slopes. The
total, the prestress ranged from 881 to 3 342 kN, with groundwater levels also indicated that drainage tunnels
an average loss of 11.3%. with higher elevations had higher groundwater levels
3.3.3 Monitoring of the stresses of anchorage rods for and vice versa, which meant that the micropores inside
shiplock slopes rock masses had poor connectivity. Water levels at the
Figure 13 shows the time-history curve of stress main drainage curtain in the foundation gallery of the
changes in the anchorage rods for the cross-section shiplock head did not change much, indicating that the
17-17 (stake 15+675) of the southern slope above the water levels were not significantly impacted by the
shiplock head. water storage in the reservoir. The granites exhibit the
It can be observed from Fig.13 that the stresses of characteristics of unsaturated pore flows. The
the anchorage rods declined with high-elevation groundwater levels also demonstrate that the drainage
excavation, and became stable after the completion of measures, such as drainage tunnels, drainage holes and
Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342
drainage trenches, are effective for the project.
Before the impoundment in September 2010, total
seepages in the drainage tunnels on the 1st to 7th
layers in the southern and northern slopes were 981.07
and 920.99 L/min, respectively. After the
impoundment in October 2010, the seepage flow
dropped to 60.08 L/min, indicating that impoundment
at elevation of 175 m had no impact on seepage in the
high slopes. Seepage in the drainage tunnels was
primarily affected by rainfall, and the duration of
rainfall reduced agreed well with the period of
impoundment. Generally, the seepage of shiplock
drainage system exhibited a downward tendency
before and after impoundment.
As the shiplock head and the chambers were
supported by concrete lining, a crisscross drainage
system was arranged at the high slopes and vertical
walls behind the lined walls. Monitoring data
indicated that the shiplock and the back of the
chamber walls had a small seepage pressure, typically
with water head less than 1.0 m. The shiplock head
and the chambers basically remained dry. Only in a
few locations, the measured water head reached 2.4 m,
but still much lower than the designed value of 8.0 m.
The water head on the bottom was typically below 4.0 m,
with the maximum of 6.9 m (the northern line of
Chamber 1), indicating that the drainage system was
effective. At present, seepage flows in the foundation
drainage galleries on the southern and northern lines
were 2 537.12 and 2 215.64 L/min, respectively,
highest in winter and lowest in summer. The change of
seepage flow is related to temperatures, water-proof
and impoundment. The back of the shiplock walls
typically has a water head less than 1.0 m, and the
highest one, as measured in certain location, is 0.023
MPa. The highest seepage pressure on the bottom of
the chambers is 0.068 7 MPa.
Actual monitoring data show that the deformation
of the rock masses in the slopes primarily occurred
during excavation and increased with further
excavation. The deformation rate dropped gradually
after excavation was completed. The slopes can be
seen as stable as all monitoring data are within the
designed ranges.
4 Key excavation technologies for
underground powerhouses
4.1 Excavation technologies for the Three Gorges
underground powerhouses
337
The main underground powerhouse of the Three
Gorges Project is equipped with six 700 MW
hydro-turbine generating units. The cross-section of
the chamber of the main powerhouse was roughly
rectangular with an arch-curved roof. The crest of the
chamber is set at the elevation of 105.3 m.
Powerhouses below the rock anchorage beams have a
span of 31.00 m, and 32.60 m for those above the rock
anchorage beams. The dimensions of powerhouse are
87.3 m high and 311.30 m long. The cross-section of
underground powerhouse No.4 and distribution of
deformation of surrounding rock masses are shown in
Fig.14.
The main underground powerhouse is located at the
right bank of the Three Gorges. The surrounding rock
masses of the powerhouses are primarily fresh
plagiogranite and diorite rocks of the Presinian system.
Fractures were developed in the rock masses, and the
major faults include F20 and F22 with a NNW
orientation, and F84 (a fault zone) and F10 with a
NE-NEE orientation. The surrounding rock masses of
the powerhouses are slightly permeable strata due to
faults and strong weathering. Groundwater mainly
comes from the upstream reservoir, right hill,
atmospheric precipitation, and construction-induced
water.
To ensure that the excavation quality of the main
powerhouse could meet the design requirements, a
series of indoor and field experiments were carried out.
Significant results were achieved: (1) The excavation
precision of the main powerhouse was ensured by
adopting a series of advanced construction technologies.
A significant improvement on excavation precision
was observed. The average over-excavation at the top
arch was controlled at 8.5 cm. More than half of
over-excavation was controlled within 20 cm. (2) To
further improve the effect of tensile anchorage rods,
experiments on grout injection equipment and
techniques for anchorage rods, as well as water-
cement ratio, were conducted and optimized. As a
result, the compactness of cement around the
anchorage rods is more than 90%. (3) Some
significant breakthroughs were made in environmental
protection and occupational health by introducing new
concepts and measures. (4) Adding access road,
smooth blasting and presplitting blasting was adopted
to ensure normal construction of the main powerhouse.
These efforts ensured the top arch deformation to be
controlled within 2 mm. Thus, the stability of the
surrounding rock masses of the large chambers during
construction is ensured [10, 11].
338 Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342

Fig.14 Cross-section of underground powerhouse No.4 and distribution of deformation of surrounding rock masses (unit: mm).

4.2 Monitoring of underground powerhouse blocks, six multi-point borehole extensometers were
4.2.1 Monitoring of deformation of surrounding rock set on top of the arch of the powerhouses. The
masses measured displacement was relatively small, ranging
Figure 15 shows the time-history curves of from 1.24 to 1.86 mm.
displacement and excavation changes in M13DC04. 4.2.2 Monitoring of anchorage rod stresses
#1 (34.1 m) Figure 16 shows the time-history curves of stress
Excavation elevation (m)

Excavation elevation 80
Displacement (mm)

30
25 70 changes in the anchorage rods on the upstream and
20 60 downstream sidewalls and arch crown of generating
15 #2 (22.9 m) #3 (13.3 m) 50
10 #4 (3.9 m) 40 unit No.4.
5 30
0 20
R23DC04 #4 upstream arch abutment
5 10
200601

200701

200801

200901

201001

Excavation elevation (m)

201101

25 R20DC04 #4 arch crown 12

20 Excavation elevation 10
Stress (MPa)

Date
15 7
Fig.15 Time-history curves of displacement and excavation 10 R13DC04 #4 downstream sidewall 5
changes in M13DC04. 5 2
0 0
It can be observed from Fig.15 that the deformation
200501

200601

200701

200801

200901

201001

201201
201101

of surrounding rock masses of the main powerhouse

was primarily induced by blasting operations during Date
construction, and it increased with the depth of Fig.16 Time-history curves of stress changes in the anchorage
rods on the upstream and downstream sidewalls and arch crown
excavation. After the excavation was completed and of generating unit No.4.
surrounding rock masses were reinforced, deformation
was totally controlled. At present, deformation of the After excavation and supporting work were
surrounding rock masses was 5.72–26.16 mm. The completed in December 2007, anchorage rod stresses
deformation of the top arch was 0.45–2.14 mm, within were stabilized and typically below the allowed value
the allowed value of ±10.0 mm in design. The of 175 MPa in design. At some locations, however, the
deformation of the upstream arch abutment ranged stress of anchorage rod exceeded the limit. For
from 0.68 to 7.79 mm, while that of the downstream example, the stress at elevation of 99.1 m on the
arch abutment was 0.20–5.42 mm, mainly of tensile upstream arch abutment of generating unit No.30
strain. The horizontal deformations of upstream reached 228.86 MPa, and the tensile stress was about
sidewalls at elevations of 86.0 and 76.74 m reached 50 MPa, with a tensile stress increase of 178.86 MPa.
6.16 and 15.64 mm, respectively. Deformation was Fortunately, the stress was stable thereafter. 47
considerably large, ranging from 9.67 to 16.67 mm, in anchorage rod stress gauges were set in 14 key blocks
the vicinity of downstream sidewalls at elevations of on the arch crown of the main powerhouse, with
85.0–87.0 m. To monitor the deformation of unstable stresses ranging from 1.94 to 105.98 MPa.
Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342
4.2.3 Monitoring of anchorage cable stresses
Monitoring was carried out from 2006 to 2011 on
the section of generating unit No.1 (the 4th layer) of
the upstream sidewalls, key block No.7 of the main
powerhouse, the section of generating unit No.4 (the
2nd layer) of the downstream sidewalls. The measured
data of downstream keyway of generating unit No.4
were used to analyze the process of changes in
anchoring force of the anchorage cables, as shown in
Fig.17.
SF22DCXYQ (No.4, keyway)
3 250
Anchoring force (kN)
3 000 SF12DCXYQ (No.4, the 2nd layer)
2 750
2 500
2 250 SF02DCSYQ (No.1, the 4th layer)
2 000
1 750
SF05DCDG (#7 of the main powerhouse)
200601
200701
200801
200901
201001
201201
201101
Date
Fig.17 Time-history curves of changes in the anchoring force of
prestressed anchorage cables.
From Fig.17, we know that the prestress loss of the
anchorage cables primarily took place in the first year
of construction and decreased thereafter, displaying a
negative correlation with temperature. Measured data
from 34 dynamometers in the main powerhouse
showed that the anchorage cables had a prestress of
2 008.0–2 921.2 kN. The total prestress loss was
2.02%–25.47% with an average of 10.96%, and a
post-locking average loss was 4.39%, which were all
within the design requirement (not larger than 15%).
The forces of the anchorage cables were basically
below 2 610 kN. Measured data of 10 dynameters
installed on the structural blocks of arch crown of the
main powerhouse indicated that the forces of the
anchorage cables were 1 958.7–2 444.2 kN, with a
total loss of 10.22%–27.6%, an average loss of
19.07%, and a post-locking total loss of 7.4%. The
forces have been stable since 2006.
4.2.4 Monitoring of rock anchorage beams
Measured data from 31 dynamometers mounted on
the upstream and downstream rock anchorage beams
for the generators and installation section II in the
main powerhouse showed that the anchorage beams
had a stress of 225.37–350.62 MPa, and a tensile
stress not more than 30 kN. Two dynamometers and
18 joint meters were mounted on the upstream and
downstream rock anchorage beams of the main
powerhouse. Both of the two dynamometers were in a
compression state. At present, the maximum stress
was 0.22 MPa. With regard to the opening in the
concrete of the rock anchorage beams and the
339
sidewalls, the opening width of generating unit No.4
was 5.59 mm, while 0.0–0.58 mm at other locations
(0.58 mm at arch crown of access tunnel). Currently,
the rock anchorage beams on K0+30.0 to K0+471.0,
upstream segment of installation section II of the main
powerhouse, have an internal temperature of 6.0 °C–
11.5 °C, while those on K0+51.0 to K0+61.0 have an
internal temperature of 8.3 °C–13.0 °C. Measured data
on the rock anchorage beams showed that the stress
ranged from 9.7 to 33.99 MPa, within the allowable
limits.
4.2.5 Monitoring of seepage
A total of 56 piezometer tubes were set inside the
drainage tunnels around the powerhouse. Water levels
in all these tubes were lower than the elevation of the
openings, and groundwater levels ranged from 58.99
to 129.05 m. Drainage tunnels at higher elevations had
higher groundwater levels, for example, drainage
tunnel A had an elevation of approximately 129.0 m
and a groundwater level of 114.71–129.05 m, and
drainage hole C2 had an elevation of 60.31–67.78 m
and a groundwater level of 58.99–67.54 m. At present,
groundwater levels and seepages in surrounding rock
masses around the underground powerhouse remain
stable.
An integrated analysis shows that, in December
2010, the surrounding rock masses of the main
structure of the underground powerhouse were stable
and deformations of the arch crown and upstream
sidewalls were controlled.
5 Grouting technology for the dam
foundation curtain
5.1 Profile description
The impervious curtain of the Three Gorges dam
consists of upstream main curtain and downstream
enclosed curtain. During curtain construction, the
depth of holes for the main curtain was generally
60–80 m. The depth of holes on the deep-channel dam
sections was 125 m, and the rocks at this point were
fractured. Other locations were with strong water
impermeability, and the deepest hole was 140 m. The
depth of holes for the enclosed curtain typically
ranged from 40 to 60 m, but the depth of holes on the
deep-channel sections reached 84 m. Layout of the
impervious curtain and drainage arrangements on the
foundation of a typical dam section are shown in
Fig.18.
5.2 Grouting technology for the dam curtain
To ensure a high-quality grouting for the dam
curtain, pre-grouting tests were carried out prior to
340 Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342

185.00 185.00 Dam axis

Drainage gallery
120.00 (250 cm  300 cm)
Drainage gallery
108.00
98.00 (300 cm  350 cm)
85.00 90.00
Dam axis 75.00 82.00
Drainage gallery Grouting hole
(300 cm  350 cm) 45.00 (hole distance 200 cm, 50.00
Drainage gallery
Drainage gallery row distance 80 cm) (250 cm  300 cm)
(250 cm  300 cm) 20.00 23.00 25.00
15.00 Drainage gallery
Drainage hole
Grouting hole 0.00 2.00 Grouting hole
(hole distance 300 cm)
(hole distance 250 cm) Drainage hole 14.00
Drainage hole 27.00 (hole distance 200 cm)
(hole distance 200 cm) 66.00
80.00
Left powerhouse section profile No.3

120.00 Discharging section profile No.2

Fig.18 Layout of the impervious curtain and drainage arrangements on the foundation of a typical dam section (unit: m).

construction, and high-pressure grouting tests were
conducted during construction. During dam curtain
grouting, a variety of measures were adopted,
including pressure-increasing grouting, intense
grouting, postponed treatment and re-grouting,
acrylate grouting, and high-performance chemical
(elastic polyurethane) grouting, etc. During grouting,
the complex geological conditions were carefully
considered, such as water inrush in the drilling holes,
grouting intensity due to water loss, poor adhesion of
grouting, strongly weathered strata, pervious deep
channels, and weak fault belts. In general, curtain
grouting was conducted in three phases, with a total
grouting length of approximately 2.0×105 m. The
main curtain of the dam had a total grouting length of
1.307×105 m, and the holes in phases I, II and III had
an average cement injection of 18.42, 6.90 and 4.49
kg/m, respectively. The enclosed curtain of the dam
had a total drilling routing length of 6.81×104 m. In
addition, the holes in phases I, II and III had an
average cement injection of 28.32, 9.96 and 6.23 kg/m,
respectively.
After grouting was completed, water was filled up
for checking. Results showed that the main curtain had
a seepage rate of 0.01–0.11 Lu (L/min), and 99.56% of
the sections had a seepage rate no more than 1 Lu. For
the enclosed curtain, 99.06% of the sections had a
seepage rate no more than 1 Lu, which was below the
allowable limit of 1 Lu. It indicated that the
micro-crevices had been effectively filled up by
grouting. Single-hole and cross-hole sound wave tests
were conducted. A pre-grouting average velocity of 5
100–5 400 m/s and a post-grouting average velocity of
5 300–5 600 m/s were observed, increased by
2.45%–8.40%. Measured results of large-diameter
drilling holes showed that the rocks after cement
grouting were compacted, integrated and well-bonded
[12, 13].
Table 1 presents the results of an analysis after the
dam curtain grouting.

Table 1 Results of analysis of hole permeability of the dam curtain with filled water after grouting.
Seepage rate ≤1 Lu Seepage rate >1 Lu Permeability (Lu)
Curtain Location Dam section Hole No. Total section Frequency Frequency
Sections Sections Maximum Average
(%) (%)
Left non-overflow
35 421 416 98.81 5 1.19 3.70 0.09
section
Left
Powerhouse section 53 773 769 99.48 4 0.52 4.40 0.11
bank
Flood discharge
56 866 862 99.54 4 0.46 1.95 0.31
Main section
curtain Right powerhouse
56 831 831 100 0 0 0.74 0.02
Right section
bank Right non-overflow
3 39 39 100 0 0 0.11 0.01
section
Total 203 2 930 2 917 99.56 13 0.44 4.40 —
Left powerhouse
21 268 267 99.63 1 0.37 1.47 0.32
section
Left
Flood discharge
bank 45 550 529 96.18 21 3.82 4.30 0.63
section
Enclosed Left powerhouse 26 239 239 100 0 0 0.20 0.02
curtain Right Powerhouse
Right 59 867 867 100 57 0.94 0.00 —
section
bank
Right powerhouse 18 160 160 100 0 0 0.40 0.01
Total 464 6 071 6 014 99.06 44 0.72 0.40 —
 Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342 341

5.3 Monitoring during dam curtain operations

156.0 m (post storage, 20061030)
5.3.1 Seepage monitoring 0.25
175 m (post storage, 20101026)
Before and after the impoundment of the reservoir to 0.20
level 175 m, the seepage rates on the left-bank dam

Coefficient 1
0.15 172.8 m (post storage, 20081109)
foundation (including that of the powerhouse) were 0.10
261.75 and 275.46 L/min, respectively, with an 0.05
increase of 13.71 L/min. The seepage rates on the
0.00
right-bank dam foundation (including that of the
0.05
powerhouse) were 310.58 and 351.33 L/min,

L-N6
L-N8
L-N10
L-N15
L-P2
EIII
L-P10
L-D1
D5
D10
D15
D20
R-V2
R-P18
R-P22
R-N1
respectively, with an increase of 40.75 L/min. Figure 19
Measuring point
shows the time-history curves of seepage rates on the
Fig.20 Distribution of the uplift pressure coefficients at the
left- and right-bank dam foundations. drainage curtain of the upstream grouting gallery of the Three
Gorges dam.
1 400 210
Seepage rate (L/min)

1 200 Water level 180

Water level (m)

1 000 150 on the left-bank non-overflow dam section, the
800 120 left-bank dam section, and the flood discharge dam
600 90
400
Seepage rate of left bank 60 section were 0.18, 0.25 and 0.11, respectively. The
200 30
0 0 maximum uplift pressure at the drainage curtain on
200201

200301

200401

200501

200601

200701

200801

200901

201001

201201
201101

section of the dam between the right-bank powerhouse

and right-bank dam section No.3 was 0.14. Figure 21
Date
(a) Left-bank dam.
shows the distribution of the uplift pressure
Water level coefficients at the downstream drainage curtain.
Seepage rate of right bank
Seepage rate (L/min)

500 200
175 m (post storage, 20101026)
Water level (m)

400 160 0.40

20091120
Coefficient 2

300 120 0.30

Dam of powerhouse 0.20
200 80
and drainage hole Right powerhouse 20081109
100 40 0.10
0 0 0.00
D11
L-D1
D1
D2
D3
D4
D5
D6
D7
D8
D9
D10
D12
D13
D14
D15
D16
D17
D18
D19
D20
D21
D22
D23
R-V1
R-V2
R-V3
200601

200701

200801

200901

201001

201201
201101

Measuring point
Date Fig.21 Time-history curves of distribution of uplift pressure
(b) Right-bank dam. coefficients at the downstream drainage curtain on the left
Fig.19 Time-history curves of seepage rate on the left- and diversion-right horizontal dam section.
right-bank dam foundations.
In Fig.21, all uplift pressure coefficients at the
5.3.2 Seepage pressure upstream drainage curtain of the downstream enclosed
Water levels in the piezometer tubes in front of the gallery were smaller than 0.50. The maximum uplift
curtain rose with increasing reservoir water level. After pressure coefficients on the left-bank powerhouse dam
water level reached 175 m, the curtain on the left-bank section, the flood discharge dam section, the left-bank
dam experienced the maximum water rising of 16.13 m powerhouse, and the right-bank powerhouse dam
to hit 171.58 m, 3.42 m lower than the reservoir water section were within the ranges of 0.00–0.37, 0.04–0.34,
level. The water level behind the curtain fluctuated 0.06–0.30, 0.00–0.37, and 0.08–050, respectively.
within a narrow range of 0.27–5.0 m. Water level in Figure 22 shows the distribution of uplift pressure
front of the curtain on the right-bank dam experienced on discharge dam section No.2.
a maximum increase of 13.26 m to hit 160.87 m. Water
Downstream surface

0
level behind the curtain fluctuated within the range of 30
60
Depth (m)

0.0–1.48 m. 90
20081109
20091120
Figure 20 shows the distribution of the uplift 120 Uplift pressure (20101026)
150 20101026
pressure coefficients at the drainage curtain of the
180
upstream grouting gallery of the Three Gorges dam. 0 20 40 60 80 100 120 140
Distance to upstream (m)
Monitoring results of the impervious curtain of the Fig.22 Distribution of uplift pressure on discharge dam section
dam show that all uplift pressure coefficients at the No.2.
upstream drainage curtain were below the designed
limit of 0.25. The maximum uplift pressure coefficients On October 26, 2010, when water level reached 175 m,
342 Qixiang Fan et al. / J Rock Mech Geotech Eng. 2011, 3 (4): 329–342

the actual uplift pressure on discharge dam section (3) Various grouting methods were considered to
No.2 was 35 555.2 kN/m, 63.31% of the designed ensure the quality of dam curtains.
value. The actual uplift pressure was smaller than the The monitoring data from 16 years of construction
designed level, helpful for the anti-sliding stability of experiences and related tests show that the optimized
the dam site. Figure 23 shows the time-history curves schemes are very successful. It will also contribute to
of seepage in left- and right-bank dam foundations the advancement and development of hydropower
during the impoundment periods from 2002 to 2010. technology in the future.

1 400 200 References

Seepage rate (L/min)

1 200 Water level 175

Water level (m)

1 000 150
800 125 [1] Zhang Chaoran, Dai Huichao. Technical breakthroughs in the
600 Total seepage rate 100
Seepage rate of construction of TGP. Engineering Science, 2003, 5 (2): 8–14 (in
400 left bank Seepage rate of 75
200 50 Chinese).
right bank
0 25 [2] Fan Qixiang. Key technical issues of TGP permanent shiplock.
200201

200301

200401

200412

200512

200612

200712

200812

200912

201012
Engineering Science, 2004, 6 (1): 48–52 (in Chinese).
[3] Ge Xiurun, Ren Jianxi, Li Chunguang, Zheng Hong. 3D-FEM
Date
analysis of deep sliding stability of #3 dam foundation of left
Fig.23 Time-history curves of seepage rate at dam foundation.
powerhouse of the Three Gorges Project. Chinese Journal of

It can be observed from Fig.23 that the seepage Geotechnical Engineering, 2003 25 (4): 389–394 (in Chinese).
[4] Feng X T, Zhang Z Q, Sheng Q. Estimating mechanical rock mass
exhibited a downward tendency after water level
parameters relating to the Three Gorges Project permanent shiplock
reached 135 m, possibly due to the effects of silting on
using an intelligent displacement back analysis method. International
the reservoir bottom near the dam. The forefront Journal of Rock Mechanics and Mining Sciences, 2000, 37 (7): 1 039–
blocking line of the dam currently had a seepage rate 1 054.
of 614.53 L/min, and those in the left- and right-bank [5] Changjiang Institute of Survey, Planning, Design and Research.
dam foundations were 276.22 and 338.31 L/min, Report on design quality management in 2010 on the Three Gorges
respectively. In total, all uplift pressure coefficients at Project. Wuhan: Changjiang Institute of Survey, Planning, Design and

the drainage curtain in the upstream and downstream Research, 2011 (in Chinese).
[6] China Three Gorges Corporation. Report on the quality and operations
enclosed galleries were within the permitted ranges in
of the Three Gorges hydropower complex in 2010. Yichang: China
the design. Actual uplift pressures were smaller than
Three Gorges Corporation, 2011 (in Chinese).
the values allowed, and seepage declined gradually, [7] Zhang Chaoran. Practice and experimental verification of TGP’s
indicating that the impervious curtain at the foundation permanent shiplock slope. Engineering Science, 2001, 3 (5): 22–27 (in
of the Three Gorges dam was effective [13, 14]. Chinese).
[8] Fan Qixiang, Gu Wenhong. Field testing study and quality control of
6 Conclusions high strength structure bolt for the Three Gorges permanent shiplocks.
Chinese Journal of Rock Mechanics and Engineering, 2001, 20 (5):
There are many rock mechanical issues in the Three 657–660 (in Chinese).
Gorges Project, but the sliding resistance and stability [9] Zhou Yu, Qian Xingxi, Fan Qixiang. Development and application of
full-variable diameter slipforms of inclined shafts. Engineering
of the dam section of the foundation for powerhouses
Science, 2002, 4 (9): 75–80 (in Chinese).
No.1–5, the slope stability of the double-line five-stage
[10] Li Yongquan, Wu Liang. Study on the law of blasting vibration
shiplock, excavation for large-scale underground
attenuation regarding the excavation of Three Gorges underground
powerhouses, and curtain grouting for the dam are
power plant. Engineering Blasting, 2009, 15 (2): 7–10 (in Chinese).
more complicated. [11] Wu Aiqiang, Xu Ping, Xu Chunmin, Yu Yong. Researches on stability
(1) Some special measures were adopted at for surrounding rock masses of underground power house in the Three
powerhouses No.1–5. New technologies such as Gorges Project. Chinese Journal of Rock Mechanics and Engineering,
dynamic design approach and methodology, the 1 000 2001, 20 (5): 690–695 (in Chinese).
kN-grade double-protection unbonded anchorage [12] Dai Huichao, Su Huaizhi. Stability against sliding in intake dam
cables (with corrugated tubes) and 3 000 kN-grade section of Yangtze River Three Gorges Project. Rock and Soil
double-protection unbonded anchorage cables, were Mechanics, 2006, 27 (4): 643–648 (in Chinese).
used to ensure the stability of the vertical sections of [13] Zhou Hougui, Li Yan. New construction techniques for dam

the slopes of the shiplock. foundation grouting and anti-seepage cofferdam of Three Gorges

(2) With new slip-casting method, significant Project. Water Resources and Power, 2009, 27 (1): 140–144 (in Chinese).

breakthroughs in environmental concepts were applied [14] Yang Xiying, Yang Zhenfeng. Curtain grouting under special
geological condition. China Three Gorges Construction, 2004, 11 (1):
to underground powerhouses to ensure stability of the
25–27 (in Chinese).
surrounding rock masses.