Embed Size (px)
DESCRIPTION
Intakes and outlets in dams
Citation preview
2. Intakes and outlets
Hydraulics of Dams and River Structures – Yazdandoost & Attari (eds)
© 2004 Taylor & Francis Group, London, ISBN 90 5809 632 7
Air entrainment at Guri Dam intake operating at low heads
G. Montilla, A. Marcano & C. CastroC.V. G. EDELCA, Hydraulics Department, Basic Engineering Division, Macagua, Edo. Bolívar, Venezuela
ABSTRACT: Experimental investigations of air entrainment inside the intakes of Guri DamSecond Powerhouse operating below critical submergence reservoir level, were carried out ona non-distorted 1 : 30 scale Froudian Physical Model. Scale effects were considered taking intoconsideration the – state of art – experience on physical modeling practice of submerged intakes.Prototype observations were done and compared with model results to achieve similarity betweenthe two physical systems. Visual observations and measurements were taken inside the intakeand, in the reservoir approach region to asses the behavior of the flow phenomenon under study.Turbine flow operation conditions where vortex formation occur were identified and curves for-free air entrainment flow-inside the intakes, were developed below elevation 240, design criticalsubmergence operation of the project which adds in operating safely the 630 MW units.
1 INTRODUCTION
Guri (10000 MW) Project is located on the Caroni river Basin at southeastern Venezuela and ispresently generating 52000 GWH/year firm energy, which accounts for the 50% of the total nationalelectricity demand. Guri Dam houses 20 Francis generators, including ten 725 MW units, whichcan discharge between 300 and 600 m3/s, for a net design head of 142 m, at normal pool elevation of270. Design of the intakes, carried out in the 60ths (Fig. 1.a) contemplates two 9.6 width 23 m heightstreamlined rectangular water passages, provided with 3 gate slots connected to the atmosphereto allow placing of maintenance stop-logs and, service and emergency intake gates. The intakestructure is connected to a 10.5 m diameter penstock by means of a convergent hydrodynamic curveof 29 m radius. Intake roof and invert intake elevations are set to El 236,59 and 217 m, respectively.During the last 3 years, very low inflow to Guri reservoir combined with required over-explotationof the dam planned firm energy led to unusual pool levels, and there is some possibility than in 2004dry season, units may operate bellow critical design pool elevation, El 240. This situation createdsome warning on EDELCA operators due to the potential occurrence of air being taken by theintakes with undesirable performance on the turbine operation. Air entrainment prediction insideintakes is complex due to many factors involved, due to its unstable nature and, to its relation withflow parameters, intake geometry and, approach conditions of a particular project. To predict theGuri intakes operation at very low heads and particularly to evaluate the air entrainment potential,a non-distorted 1 : 30 Scale Physical Model was built (Figs. 1.a and 1.b) in transparent plexiglass,consisting on one full geometry intake, provided with the trashrack (Fig. 1.c) and, the 3 slots toplace the stop-log, service and emergency gates. Guri reservoir was reproduced in the model by aconstant elevation tank sufficiently large to allow symmetric laboratory approach conditions. Theinvestigations were divided into two parts: the first part aimed to describing the phenomenon of airbubbles and air dragged mechanism inside the intake and, the second part documents the tendencyof vortex formation in the reservoir.
2 MODEL SIMILARY
2.1 Dimensional analysis
For engineering purposes, vortex formation, and air entrainment and drag into the intake depends onfluid properties, flow characteristics, approach and, intake geometry. To allow for the phenomenon
53
a) Intake model geometry, key elevations
b) General view of the Guri physical model scale 1:30
c) Trashrack arrangement-prototype and model
4.4m
4.8
m2.8
m
179,50msnmReducing cone Platform
Regulation
P L A N T
S E C T I O N
Intake
Tank
Tank
Pipe
Gate
Gate
Platform
2.6
m
Reducing cone Platform
Regulation
Valve
P L A N T
S E C T I O N
Tank
Pipe
Gate
Gate
Platform
2.6
m
Prototype TrashrackA free
A obs1.20
A free
A obs1.17
Model Trashrack 1:30
PierFlow
P L A N T
0,4
7m
0,4
0m
0,3
8m
0,508m
0,896m
S E C T I O N
Gate slots
Reducing cone
Intake
Trashrack
D
h
Trashrack
217,0
222,2 P L A N T
0,4
7m
0,4
0m
0,3
8m
0,508m
0,896m
S E C T I O N
D
h
240,0
217,0
222,2
V
Tank
Figure 1. Physical Model of Guri intake Scale 1 : 30.
54
investigation and report of results of any flow system to be independent of the unit system, it is con-venient to use classical dimensional analysis, in terms of the important non-dimensional parameters.
The functional expression (1) showing the non-dimensional parameters describing the phe-nomenon under study is (Fig. 1.a):
f
(
h
D,V · D
v, V ·
√
ρ · D
σ,
V√
g · D,
Ŵ
D · V
)
= f (S, Re, We, Fr, Nτ) = 0 (1)
where S = Submergence; Re = Reynolds Number; We =Weber Number; Fr = Froude Number;Nτ = Circulation Number; D = Penstock Diameter; V = Flow Velocity; Ŵ = Flow circulation;σ = Flow Surface Tension; v = Kinematic Fluid Viscosity; g =Acceleration Due to Gravity.
Functional equation (1) suggested that being the two systems similar geometrically wise andwith similar approach flow patterns, results from the model system will depend on gravitational,viscous, circulation and, surface tension forces.
2.2 Geometry comparison
In the Scale 1 : 30 physical model, every geometric detail of the prototype was reproduced, in orderto keep similarity of the solid conveyance boundaries to guarantee adequate visual observationsand its extrapolation to prototype performance. However, trashrack prototype dimensions of theelements thickness were not practical to be reproduced in the model and a criteria of the obstructedarea of the prototype trashrack was adopted resulting in a distortion factor (1 = 2), or in a factorof free area for the intake flow of 55% and, 54% for prototype and model, respectively which wasconsidered satisfactory for model reproduction in that respect (Fig. 1.c).
The expression (2) shows the distortion relationship used to maintain reciprocity of the trashracksflow areas between model and prototype:
Am
Ap= Xr · Lr = 1 · L2
r (2)
where Am = Model Area; Ap = Prototype Area; Xr = Horizontal Scale; Lr =Vertical Scale;1 = Distortion.
2.3 Viscous and surface tension effects
Physical modeling of vortex formation and air dragged into intakes has been controversial throughdecades and up to present there is not a standard methodology to approach this phenomenon thatinclude consideration of viscous, gravity, surface tension and flow turbulence, as the most importantones. To reproduce all these forces simultaneously in the model as they are present in the prototypewill result in satisfying equation (1) for both physical systems which is conflictive (Ettema, 2000).
In laboratory practice, criteria for similarity of centrifugal forces are used and the remainingforces acting on the phenomenon are accounted for by approximated methods that may not reflectrigorously the flow behavior. As a result of this approximation “scale effects” – term that normallyjustifies deviations from model to prototype performance – are brought about, and practical exper-tise suggest reducing them as it is possible either by building a model as large as economicallyfeasible in a given laboratory installations and/or, by operating the model with flow conditionsresembling more like prototype behavior. In this particular case the phenomenon is directly linkedto the gravitational force, this criterion suggests using Froude similarity. However, viscous, surfacetension, and turbulence level of the flow are considered as scale effects. Model scale is then selectedso working conditions of the model flow are acceptable, and model flow conditions are controlledto reduce remaining scale effects. Vortex originates by fluid rotation and whether they appear andtheir intensity will be related to the rotational streamlines patterns that occurred in the intake neigh-borhood. For this reason many investigations on model vortex formation have demonstrated thatscale effects are negligible when Reynolds (Re) and Weber Numbers (We) are sufficiently high.Daggett & Keulegan (1974), demonstrated that viscous effects are negligible when Re > 3.2.104
being in Guri Physical Model Scale 1 : 30, Re = 4.4.105 and Re = 2.2.105 for flows of 600 and300 m3/s, respectively, the latter suggest that viscous effects are suppressed if the model is operatedby using the Froude law.
55
With regard to surface tension, Jain (1978), who used fluids of different surface tension demon-strated that vortex and air entrainment in model studies are not affected forWe > 11.0, this conditionis satisfied by the Guri 1 : 30 Physical Model which was operated at 43.0 <We < 86.0.
2.4 Exaggeration of model discharge
A technique used by some authors to account for scale effects is to increase the operating dischargeduring the model tests. Model discharges are increased and so is flow velocity, then model operationin terms of hydraulic total roughness are plotted against Re until the first becomes independent ofRe (Semenkov, 2003). However, a difficulty arises when applying this technique since model flowpatterns and the Circulation Number change as a result of the increasing discharge. For this reasondifferent authors based on previous investigations, Denny & Young (1957), consider this methodconservative and should be used with reserve. In Guri 1 : 30 Scale Model discharge was increasedto exaggerate the flow patterns thus enhancing flow conditions for the vortex to be formed in thereservoir, up to 2.3Q, being Q the project discharge.
3 MODEL TEST CONDITIONS
Tests were executed in two stages: first group of tests were done inside the intake and, a secondstage tests were done in the reservoir region. First group of tests included examination of air bubbleformation and vortex development mechanisms inside the intake. Second group of tests includereservoir vortex formation in the free reservoir elevation and, their interaction with the trashrack.Project conditions of the tests were as follows: (1) Guri reservoir levels 240.0, 237.0, 235.0 and232.0 m, (2) Model flows between 300 and 1400 m3/s which include normal and exaggerated Q,(3) With trashracks, and without stop-logs or gates placed on the slots.
4 TESTS RESULTS
4.1 Velocity distributions
Figure 2 shows a sample of the model flow velocity distribution along the left intake bay, as measuredupstream of the intake for reservoir El = 240.0 m.This distribution is rather uniform when the intakeis completely submerged, (El > 236.7 m). However, when the intake is not submerged, a series ofstationary waves on the free surface appear as the flow upper streamlines hit the intake upperboundary, these waves may contribute to inhibit vortex formation on the free surface. When theintake is submerged (El = 240.0 m. and, with exaggerated discharge Q > 1000 m3/s) local velocity
5m10m
20m
Q=
1400m
3/s
10008
00
600
450
300
Q=
1400m
3/s
1000
8006
004
503
00
Q=
1400m
3/s
300
800
W a
t e
r E
l
e v a
t i
o n
(
m )
5m10m
20m5m
10m20m
P r o t o t y p e v e l o c i t y U p s t r e a m ( m / s )
Figure 2. Velocity distribution upstream of intake.
56
pulsations with deviations-up to the 10%-at the point of the maximum velocities – (Fig. 2) wererecorded, this behavior suggests a trend for vortex formation due to higher velocity concentrationsnear the intakes.
4.2 Vortex formation and air dragged
Figure 3 shows, for El = 240.0 m, that the average type of vortex (Knauss, 1987) for Q < 600 m3/s(Fr < 0.7) is less than 2. Moreover, the maximum frequency of occurrence of vortex formation is32% (5 minutes observation time), which is estimated to be a low frequency of vortex presence and,it may not represent a hazard to the turbine. However, when Q is exaggerated, Q = 1.7Q, vortex ofthe Types 3, 4 and 5 start to show on the free surface, Fr > 1.1. In the prototype (April 2003, forGuri reservoir elevation of 244.56 m.), it was observed vortex formation, Types 1 and 2 (h/D = 2.1,Figs. 4–8). This limited 2003 and 1985 prototype data and its comparison with similar model tests
14%
25%28%
32%
15%
0
1
2
3
4
5
6
7
0.3 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.3 1.4 1.5 1.6
Froude Number
Av
erage
Vort
ex T
yp
es
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
Prototype Discharge (m3/s)
100%
300 400 500 600 700 800 900 1000 1100 1200 1300 1400
Aver
age
Fre
qu
ency
of
Vort
ex T
yp
es13%
35%37%
40%
48%50%
53%
61%
70%
77%
82%
87%
16%
52%51%49%
Average Vortex Types
Average Frequency of Vortex Types
Prototype
Operation Range
Figure 3. Occurrence and frequency of types of vortex, El 240.0.
a) b)
Figure 4. Vortex formation, Type 2 in the prototype, April 2003.
57
0.7
0.8
0.9
1.0
Cri
tica
l S
ub
mer
gen
ce (
h /
D)c
r
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
0 100 200 300
Prototype Discharge (m3/s)
400 500 600 700
5
4
3
2
1
1
2
3
4
5
6
Vor
tex
Type
5
4
3
2
1
1
2
3
4
Vor
tex
Typ
e
5
4
3
2
1
5
6
5
4
3
2
1
Prototype Operation Range
Froude Number (V/(g D) ^ 0.5)
0.0 0.1 0.2 0.3 0.5 0.6 0.7 0.8
Wa
ter
Ele
vati
on
(m
)
243.3
242.2
241.2
240.1
239.1
238.0
237.0
235.9
234.9
233.8
232.8
231.7
230.7
229.6
Vortex Types (IAHR, Hydra-
ulic structures design manual)
Figure 5. Vortex formation in the slots.
0.0
0
I
II
III
IV
Prototype
Operation Range
Neither vortex formation nor
air dragged into the turbine
Vortex Types 4 and 5,
air dragged into turbines
Vortex Type 5, severe air
dragged into the turbine
Froude Number (V/(g D) ^ 0.5)
0.1 0.2 0.3 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.3 1.4 1.5 1.6
Wate
r E
levati
on
(m
)
Cri
tica
l S
ub
mer
gen
ce (
h/D
)cr
243.3
242.2
241.2
240.1
239.1
238.0
237.0
235.9
234.9
233.8
232.8
231.7
230.7
229.6
100 200 300 400 500 600
Prototype Discharge (m3/s)
700 800 900 1000 1100 1200 1300 14002.0
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
Vor
tex
Typ
es 2
and 3
, ai
r
bubble
s ge
ner
ated
by
tras
hra
ck,
bubble
s ri
se thro
ugh
slo
ts
Figure 6. Occurrence of vortex at slots, air bubble formation and drag to the turbine.
58
0
225.0
226.0
227.0
228.0
229.0
230.0
231.0
232.0
233.0
234.0
235.0
236.0
237.0
238.0
239.0
240.0Gate Slot 1
Gate Slot 1
Gate Slot 1
Gate Slot 1 & 2Gate Slot 3
Gate Slot 3
Gate Slot 3
Gate Slot 3
Gate Slot 2
Gate Slot 2
Gate Slot 2
Gate Slot 1
Gate Slot 1
Gate Slot 1
Gate Slot 1 & 2Gate Slot 3
Gate Slot 3
Gate Slot 3
Gate Slot 3
Gate Slot 2
Gate Slot 2
Gate Slot 2
Gate Slot 1
Gate Slot 1
Gate Slot 1
Gate Slot 1 & 2Gate Slot 3
Gate Slot 3
Gate Slot 3
Gate Slot 3
te Slot 2
Gate Slot 2
Gate Slot 2
Gate Slot 1 NSE=240Gate Slot 2 NSE=240Gate Slot 3 NSE=240Gate Slot 1 NSE=237Gate Slot 2 NSE=237Gate Slot 3 NSE=237Gate Slot 1 NSE=235Gate Slot 2 NSE=235Gate Slot 3 NSE=235Gate Slot 3 NSE=232Gate Slot 1 y 2 NSE=232
Gate Slot 1
Gate Slot 1
Gate Slot 1
Gate Slot 2
Gate Slot 2
Gate Slot 2
Gate Slot 3
Gate Slot 3
Gate Slot 3
Gate Slot 1 & 2
Gate Slot 3
Prototype Discharge (m3/s)
100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.5
0.4
0.3
Su
bm
erg
ence
(h
/ D
)
Froude Number (V/(g D) ^ 0.5
Ga
te S
lot
Wa
ter
Ele
va
tio
n (
m)
0.00 0.1 0.2 0.3 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.3 1.4 1.5 1.6
Figure 7. Operation range of the turbines, water levels at the slots and reservoir elevations.
h/D = Fr ( with vortex suppression )
h/D = 1+Fr ( no vortex suppression )
Reddy & Pickford ( 1972 )
h/D = 1.7Fr ( symetrical a
pproach )
h/D =
1.7Fr ( la
teral appro
ach)
Gordon ( 1970 )
h/D = 1+2.3Fr (
suggeste
d by Gord
on )
h/D = 0.5+2Fr Pennino &
Hecker ( 1979 )
Pater
son
& N
oble(
p
Prototype
Operation Range
( 2003 )
( 1985 )
h/D = Fr ( with vortex suppression )
h/D = 1+Fr ( no vortex suppression )
Reddy & Pickford ( 1972 )
h/D = 1.7Fr ( symetrical a
pproach )
h/D =
1.7Fr ( la
teral appro
ach)
Gordon ( 1970 )
h/D = 1+2.3Fr (
suggeste
d by Gord
on )
h/D = 0.5+2Fr Pennino &
Hecker ( 1979 )
Pater
son
& N
oble(
Prototype
Operation Range
( 2003 )
( 1985 )
h/D = Fr ( with vortex suppression )
h/D = 1+Fr ( no vortex suppression )
Reddy & Pickford ( 1972 )
h/D = 1.7Fr ( symetrical a
pproach )
h/D =
1.7Fr ( la
teral appro
ach)
Gordon ( 1970 )
h/D = 1+2.3Fr (
suggeste
d by Gord
on )
h/D = 0.5+2Fr Pennino &
Hecker ( 1979 )
Pater
son
& N
oble(
Prototype
Operation Range
( 2003 )
( 1985 )
h/D = Fr ( with vortex suppression)
h/D = 1+Fr ( no vortex suppression)
Reddy & Pickford( 1972 )
h/D = 1.7Fr (sy
mmetrical approach)
h/D =
1.7Fr ( la
teral a
pproach
)
Gordon ( 1970 )
h/D = 1+
2.3Fr ( suggest
ed by G
ordon)
h/D = 0.5+
2Fr Pennino & H
ecker ( 1979 )
Model vortex upstream < type2 - reservoir
Model vortex upstream < type4 - slot 1
Model vortex upstream < type6 - slot 1
Prototype vortex upstream < type2 - reservoir
Prototype
Operation Range
( 2003 )
( 1985 )
Crit
ica
l S
ub
merg
en
ce (
h /
D)c
r
Froude Number (V/(g D ) ^ 0.5
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Paters
on & N
oble (
1982 )
Figure 8. Occurrence of vortex formation – Guri model and prototype.
59
(Fig. 8) permits verifying that the model is capable of reproducing the prototype behavior evenwithout need to exaggerate the model Q, meaning that the model fulfills flow pattern similarity inthis respect. Inside the intake, tests indicate that vortex are not formed by lack of submergence butfor flow separation from the slots for h/D < 1.4 (Fig. 5).The vortex intensity becomes important forh/D < 1.25 when vortex Type 4 and 5 appear. These vortices are capable of generating air bubblesthat are dragged downstream by the flow (Fig. 6).
The dashed region (Fig. 7) shows the operation range of the turbine to avoid air bubble formationbelow critical submergence elevation (El 240 m). Some points of the model and prototype observa-tions from Guri Intakes reported herein are plotted together with curves advanced by other authors(Fig. 8). This data include conditions of normal Q (Vortex Type 1 and 2) and exaggerated Q, whichinclude vortex Types 3, 4, and 5 (Knauss, 1987).
5 CONCLUSIONS
In this investigation viscous and surface tension effects were considered, the test results togetherwith the prototype limited data permits concluding that the Model Scale 1 : 30 is sufficiently largeto allow for scale effects due to the relatively low Re and We Numbers to be higher than minimumvalues reported in the literature, thus these effects can be reported as negligible. Boundary geometryalong with slots and sheared flows resulting from flow interaction with these features are wellrepresented and its correct model reproduction resulted of prime importance in vortex generation,air bubbles formation and air drag from the slots to lower reaches of the penstock and eventuallyto the unit. Free area for the flow passing the trashrack was respected by the model construction,with a criteria that looks acceptable for correct full scale reproduction.
The technique of increasing the flow discharge proved useful in enhancing the vortex formationpotential of the flow and its interaction with the boundaries, vortex intensity and frequency, airbubbles formation and eventual drag into the penstock are promoted as Q is increased.
Velocities profiles were developed in the model and identification of flow local velocities devi-ations up to 10% were recorded. These pulsations were unsteady and its occurrence are closedassociated to vortex formation.
Based on this investigation, an operation range for the turbines is proposed. However, the amountof air dragged at the slots as seen in the model (Vortex Types 4 and 5), bubble size and air volumeassociated, has to be further investigated since marginal volumes of air entrained at atmosphericpressure, may not be necessarily detrimental for turbine operation.
REFERENCES
Ettema, R. 2000, “Hydraulic Modeling: Concepts and Practice”, Sponsored by Environmental and WaterResources Institute of the American Society of Civil Engineers.
Semenkov, V. 2003, “Report About Hydraulic Department”.Gordon, J.L. 1970, “Vortices at Intakes”, Water Power.Knauss, J. 1987, “Swirling Flow Problems at Intakes”, Hydraulic Structures Design Manual, IAHR 4.Dagett, L.L. & Keulegan, G.H. 1974, “Similitude in free-surface vortex formations”, ASCE, Journal of
Hydraulic Engineering, 100, HY11.Jain, A.K., Ranga Raju, K.G. & Garde, R.J. 1978, “Vortex Formation at Vertical pipe Intakes”, ASCE, Journal
of Hydraulic Engineering, 104, HY10.Denny, D.F. & Young, G.H.J. 1957, “The Prevention of Vortices and Swirling at Intakes”, IAHR Congress
Lissabon, Paper C1.
60
