Escolar Documentos
Profissional Documentos
Cultura Documentos
Contents
1 INTRODUCTION 2
9 CONCLUSION 14
2
1 INTRODUCTION
Key words
Unlined drill & blast tunnel, TBM tunnel, friction losses, singular losses, statistical
distribution, overbreak, measured head loss, roughness, lined invert, shotcrete,
Cross sections and shapes of an unlined drill & blast tunnel vary randomly from one
section to another. The flow in it lacks longitudinal and lateral symmetry. Most
convenient measurable dimensions of a tunnel at a construction site are the cross
sections at 5 to 10 m intervals. The surface of the excavated rock has irregular
projections due to blasting that offer high resistance to flow. This is commonly termed
as micro-roughness. Variation in areas from one location to another causes continuous
expansion, contraction and change of direction, called macro-roughness. Based on the
statistical distribution of the measurements from site, methods to evaluate friction losses
in an unlined tunnel, taking both these roughness together into account, are available in
literature. Analytically, it is difficult to separate both the roughness. Addition of micro
and macro roughness shall give equivalent hydraulic roughness of the unlined drill &
blast tunnel.
2x72 MW Chimay hydropower project in central Peru has a pressurised water conductor
system which consists of an unlined drill & blast (D&B) tunnel (D-shape, nominal base
width 7.0 m, height 6.15 m) of 5130 m long; a 5.7 m diameter TBM excavated tunnel
of 4021 m long; a 4.5 m diameter concrete lined vertical pressure shaft of 107 m
height; a 3.8 m diameter steel lined high pressure horizontal tunnel of 220 m long; and
two penstocks, 2.65 m diameter, of 63 m long each. Head losses in the water conductor
system have been measured between the entry to the tunnel and at the end of the
penstocks. The paper deals with the design discharge of 82 m 3/s.
v2 L
hf = f (2.1)
2g D
0.25
f = 2
(2.2)
k 5.74
log 3.7 D + 0.9
Re
b = m (2.3)
m = x y z (2.4)
Where
R
y = F 0 , ratio of radius of curvature of the bend and the equivalent hydraulic
D
diameter of the tunnel
h
z = F , ratio of height and width of the cross section of the tunnel
w
Where
2
e= Coefficient of correction = + sin 2 for /6
5
= for /6/2
4 2
= shape factor of the conduit = 1 for circular, 0.75 for open channels
A1, and A2 are the areas upstream and downstream of the expanded flow.
2
1
c = 1 (2.6)
Where
=
( )(
1 (1 a ) 1.032 b + 1.38 a 1.48 b 0.7 1.495 b 0.49 )
1.03 0.03 b
A2
a = ratio of downstream area to upstream area=
A1
b=
180
Other local losses like entry, combination, division, obstructions are referred from
standard literatures.
A99% A1%
= 100 % (3.1)
A1%
The relative equivalent roughness in the rock tunnel can be expressed through
15
0.5 = 0.105 log (3.2)
k/R
f = 2.75 10 3 (3.3)
roughness k of the surface. The value of the friction factor f is calculated from equation
(3.4).
t m R 1.5
f = 0.55 (3.4)
( R + t m ) 2.5
4
k = Dm Dn = ( Am An ) (3.5)
A50% A1%
tn = (3.10)
0.5 ( P50% + P1% )
tn 1
=2 100 % (3.11)
R 50% tn
(1 ) 2
2 R 50%
After knowing the values of the relative overbreak , friction coefficients are read out
from the graphs provided by the author for an exposed drill & blast rock surface, or with
a concrete lined invert of a drill & blast unlined tunnel.
Ai
ki = + (3.12)
Ai
For m measured cross sections along a drill & blast tunnel stretch, the absolute
roughness is calculated according to equation (3.13).
1 m Ai
k = + (3.13)
m 1 Ai
Where, and are experimentally determined constants having the values 0.15 m and
0.37, respectively.
Ai Ai 1
ki = (3.14)
Pi
1 m
k avg = ki
m i =1
(3.15)
Total roughness of the tunnel, k = wall roughness (kw) + area roughness (kavg)
0.667
k
f = 0.49 (3.16)
4 R avg
(x
i =1
i x) 2
rms j = (3.17)
r
For m numbers of such lines (at least 3 numbers of lines on the blasted rock surface
parallel to a reference line are required for which distance measurements are taken at
0.25 m to 0.50 m intervals), the wall roughness is expressed in equation (3.18).
(rms
j =1
j )2
rms wall = (3.18)
m
The area roughness for any stretch (measured at 0.5 m to 1.0 m intervals) is calculated
as
(A A 0.5 )
0.5
j
j =1
rms Ai = 0.53 (3.19)
r
(rms Ai )2
rms A = i =1
(3.20)
m
Average friction coefficient with unlined invert and shotcrete on wall and roof
f = 0.0519
Average friction coefficient with lined invert and shotcreted on wall and roof
f = 0.0411
covering around 92% length of the drill & blast tunnel. The statistical distribution of the
surveyed areas and the calculated perimeter are shown in figure 4.1 and 4.2,
respectively. Out of 903 measured cross sections, 888 number of cross sections have
been chosen (204 in km 1, 190 in km 2, 126 in km 3, 199 in km 4, and 169 in km 5).
Rest of the cross sections are disregarded due to incompatibility.
110
100 km 1
90
km 2
80
Number of samples
70 km 3
60
km 4
50
40 km 5
30
20 all
10
0
-5 0 5 10 15 20 25 30 35 40 45
variation of area over the nominal area %
Figure 4.1: Chimay, Statistical distribution of area with respect to nominal area of
37.792 m 2.
110
km
100 1
90 km
2
80 km
Number of samples
3
70
km
60 4
50 km
5
40 all
30
20
10
0
-2 2 6 10 14 18 22
variation of perimeter %
Absolute roughness k and friction coefficient f have been calculated directly for each km
by applying the suggested methods in section 3. The values are presented graphically in
figures 4.3 and 4.4, respectively. In the figure, the values are joined by representative
smooth curves, but it is not to be understood that k values and f values keep on changing
from one point to another. They remain same for a particular km. The k and f values are
for the exposed rock in a drill & blast tunnel. These values shall be corrected when
either the invert is lined or the surface is shotcreted or both.
0.8
Rahm
0.6
Colebrook
k (m)
0.4 Huval
Wright
0.2
Johansen
0 Solvik
0 1 2 3 4 5 6 IBA
km
Rahm
0.12
0.1 Colebrook
0.08 Huval
f [-]
0.06 Priha
0.04 Reinius
0.02 Wright
0 Johansen
0 1 2 3 4 5 6
Solvik
km
IBA
Area variation in the TBM tunnel is not erratic, but it remains constant over a stretch.
Each stretch shall have its flow velocity and effective hydraulic radius, therefore, a
different friction coefficient. The roughness of the TBM exposed rock and the shotcrete
layer is of the same magnitude. The roughness of the shotcrete and the rock is varied
from 4 mm to 10 mm. Average friction head losses in the TBM is of the order of 8.23
m. The friction coefficient f of the TBM tunnel varies from 0.0181 to 0.0233.
10
26
24
22
20
18
16
2
Area, m
14
12
10
8
6
4
2
0
5100 5600 6100 6600 7100 7600 8100 8600 9100
chainage,m
Area variation in a TBM bored tunnel is not abrupt. The difference in cross section area
is gradual by applying shotcrete between two adjacent stretches. The transition angles
vary from 20 to 45. Average head losses due to expansion and contraction due to area
variation in the TBM tunnel are of the order of 0.28 m and 0.19 m, respectively.
On average, without considering the effect of the expansion and contraction, the TBM
tunnel has a friction coefficient of f = 0.02, k = 6.03 mm, and K=59.24 m1/3s-1. Taking
into account the effect of local expansion and contractions, TBM tunnel has an average
f = 0.0211, k = 7.48 mm, and K=57.67 m1/3s-1.
The cross section area of the rock trap goes on varying. Therefore, taking an average
roughness of the wall surfaces, friction head losses of 0.14 m is calculated. Similarly in
the transition between the D&B tunnel and the TBM tunnel a friction head loss of 0.02
m is expected.
Losses due to entry at the tunnel, expansion, contraction and transition in the defined
geometry of the tunnel are estimated at 0.32 m.
Similarly head losses due to flow division, combination and bifurcation in the system
works out to another 0.32 m.
Head losses due to obstruction of flow in the rock trap caused by the beams and the
concrete wall is of the order of 0.1 m.
Absolute roughness of the exposed rock surface is calculated as per the existing
methods briefed in section 3. Where shotcrte is applied, the surface roughness of the
blasted rock is, therefore, reduced by 11 cm.
12
Pr f r + Pc f c + Ps f s
f = , and (7.1)
P
P = Pr + Pc + Ps (7.2)
After taking into account the influence of invert lining and shotcrete, as applicable,
theoretically, the expected friction head losses for every km and for the total tunnel is
furnished in the table 7.2.
Table 7.2: Chimay; friction head losses in the D&B tunnel calculated by various
methods (*only recommended friction factors considered).
RAHM
HUVAL
PRIHA
REINIUS
WRIGHT
IBA
CZARNOTA
MEASURED
COLEBROOK
JOHANSEN
SOLVIK
Head losses for the entire water conductor system, amounting to 21.8 m, has been
measured at the design discharge of 82 m3/s. Deducting the losses already calculated in
sections 5 and 6, balance losses are accounted towards the measured head losses in the
unlined D&B tunnel. Theoretically calculated head losses and the measured one are
compared graphically in figure 7.1.
It is seen that Rahms method overestimates the losses by 10.8%, Colebrooks method
underestimates it by 30%, Huvals method overestimates by 4.6%, Prihass method
overestimates by 20%, Reiniuss method underestimates by 5.2%, Wrights method
overestimates by 4%, Johansens method underestimates by 20%, Solviks method
underestimates by 39%, and IBA method overestimates by 0.6%, than the measured
values. The methods that estimate the friction losses of the D&B tunnel within an
acceptable range are retained and the others are discarded to arrive at an average
calculated value of the friction head losses. Therefore Rahm, Huval, Reinius, Wright
and IBA methods are retained.
If the rock surface of the D&B tunnel were completely exposed, friction head losses of
around 10.9 m would have been expected. Since 80% of the D&B tunnel length is invert
lined and 45% of surface is covered by shotcrete, reduction of head loss has reduced by
2.86 m (26% of the losses). The invert lining contributes to 83% of this reduction, and
balance 17% is from the shotcrete layer.
Equivalent average friction coefficient of the completely exposed rock surface would
have been f = 0.0813, K = 28.41 m1/3s-1 and k = 454 mm. After 80% of the tunnel invert
has been lined and 45% of the surface is shotcreted, the new parameters have been f =
0.0600, K = 33.06 m 1/3s-1 and k= 273 mm.
The graph of nominal area versus mean excavated area is shown at figure 8.1.
14
Nominal area versus excavated area of a drill and blast unlined tunnel
240
230
220
210
200
190
180
170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Nominal area, m2
Figure 8.1: Nominal area vrs. Average excavated area of a D&B tunnel
9 CONCLUSION
Head losses in an unlined D&B tunnel can be estimated as per some of the existing
methods (Rahm, Huval, Reinius, Wright and IBA). Correction for invert lining and for
shotcrete shall have to be applied in real site condition. Enough and systematic record of
cross section data are required for better statistical distribution. Cross sections formed
due to geological overbreak or due to intentional enlargement shall be excluded from
the standard distribution.
If the invert of a D&B tunnel of Chimay size is lined, around 27% of the head losses,
with respect to complete exposed D&B rock, can be reduced. If shotcrete is applied on
the wall and the roof, around 10% of the head losses can be reduced.
Cross section of a TBM excavated tunnel may vary due to the placement of supports
within the excavated diameter. In addition to the friction losses, there shall be losses due
to expansion and contraction. The surface roughness of such a tunnel can be between 4
mm to 10 mm.
K value for an unlined D&B tunnel is around 28 m1/3s-1. This corresponds to a friction
coefficient f = 0.0813.
Separation of macro and micro roughness is difficult in a D&B tunnel. Both are to be
treated integrally as one roughness.
15
Subscripts
1%, 50%, 99% cumulative frequency of 1%, 50%, 99% for statistical
distribution.
avg average
m mean, average
n nominal, natural
t theoretical
16
Abbreviations
ch. Chainage
D&B Drill and Blast
log Logarithm function to base 10
m asl Metre Above Mean Sea Level
mean Arithmetic mean
mwc Metre of water column
rms Root mean square
TBM Tunnel Boring Machine
Bibliography
Borch E. Unlined high pressure tunnels in areas of complex topography, Water Power & Dam
Construction, November 1984.
Carstens T. and Hansen S.E. Rehabilitation in the unlined rock tunnels of Nedre Rossaga after 40
years, Hydropower97.
Cuesta L. Unlined hydroelectric tunnels, Rock Mechanics and Power Plants, Pages 289-292, ISBN
90 61918 278.
Dann H.E.- Unlined tunnels of the Snowy Mountains Hydro-electric Authority, Australia, Journal of
the Power Division, Proceedings of the American Society of Civil Engineers, October 1964.
Hydraulic considerations (Section 6) describing method [Solvik, 1984] for unlined pressure tunnels,
received from Kukule Ganga Hydropower Project, Sri Lanka.
I.E. Idelcik- coefficients de pertes de charge singulirs et de pertes de charge par frottement, traduit
par Mme. Meury, 1979.
Lecocq R. et Marin G. Evaluation des pertes de charge des galleries damnee deau forees au
tunnelier et non revetus, 1986 (collected from Prof. Schleiss, EPFL, Laussane).
17
Metcalf J.R. and Jordaan J.M.- Hydraulic roughness change in the Orange-Fish Tunnel: 1975-1990,
The Civil Engineer in South Africa, August 1991.
Nord G.- Drilling Accuracy in Underground Construction, World Tunnelling, December 2000.
Pennington M.S- Hydraulic roughness of bored tunnel. Paper on internet, IPENZ Transactions,
Vol.25, No. 1/CE,1998.
Petrofsky A.M.- Contractors view on unlined tunnels, Journal of the Power Division, Proceedings of
the American Society of Civil Engineers, October 1964.
Reinius E. Head losses in unlined rock tunnels; Water Power July / August 1970.
Ronn and Skog New method for estimation of head loss in unlined water tunnels. Hydropower
1997.
Solvik O. and Tesaker E.- Floor paving in unlined hydropower tunnels, Hydropower 1997.
Spencer R.W.- Unlined tunnels of the Southern California Edison Company, Journals of the Power
Division, Proceedings of the American Society of Civil Engineers, October 1964.