Scale Modeling of Tarzout Dam

Danica Starinac1, Predrag Vojt1, Dimitrije Mladenović1

 

1 Jaroslav Černi Institute for the Development of Water Resources, Jaroslava Černog 80, 11226 Pinosava-Belgrade, Serbia; E-mail: This e-mail address is being protected from spambots. You need JavaScript enabled to view it

 

Abstract

The design of the Tarzout Dam (Biskra Province, Algeria) has been verified by scale model analyses. The length scale of the physical model was 1:30 and the hydraulic parameters were scaled applying Froude similarity's law. The tests focused on a spillway with a Piano Key Weir (PKW), a stepped chute and a stilling basin. The initial design of the chute was optimized on the basis of model testing.

Keywords: scale model, Piano Key Weir, PK weir, PKW, spillway, Tarzout.

Introduction

 

Spillways play a major role in ensuring the flood safety of dams. Insufficient spillway capacity has been the cause of one-third of all dam failures. The discharge capacity of free crest or ungated overfall spillways is directly proportional to the length of the crest or the weir for a given upstream head. Its length can be increased by using curved, undulated or corrugated weirs instead of straight linear weirs. Consequently, the discharge for a given head also increases. With the motivation of maximizing the crest length, the following crest geometries of weirs have so far been developed: duckbill spillway or bathtub spillway, fan spillway, type Y spillway, daisy-shape spillway, and morning glory spillway. In a further step, the development of labyrinth spillways began in the 1930's. After 2000, the Piano Key Weir (PKW) was introduced as an evolution of the traditional labyrinth weir (Schleiss, 2011).

The PKW spillway is a variation of the traditional labyrinth weir. The planform has a rectangular shape (Fig. 1). Contrary to a labyrinth weir, the apex is not vertical but inclined by turns in both the upstream and downstream direction. This arrangement explains the name Piano Key Weir. According to the chosen slopes of the inlet and outlet keys, they have a certain upstream and downstream overhang. This results in a smaller footprint of the structure compared to a rectangular labyrinth weir with vertical walls. Therefore, besides improved hydraulic performance, the PKW has the advantage that it may easily be installed even where the foundation space very limited, for example on gravity dam crests. This is also the reason why PKW spillways are an efficient and economical solution for increasing the flood discharge capacity of existing gravity dams.

 

fig01
Figure 1: 3D sketch of PKW with main geometric parameters (Machiels et al., 2011).

 

The first PKW was installed in 2006 at Golours Dam in France. Since then, PKWs have been used on numerous dams (France, Vietnam, India, China, etc.).

A PKW is characterized by complex three-dimensional flow, so the discharge over such a spillway depends on many parameters. Therefore, a large number of laboratory experiments are required to establish general applicable design rules. Preliminary design criteria were presented by Lempérière & Ouamane (2003), which were based on experiments performed at Biskra University in Algeria and at Roorkee University in India. Since then, more systematic laboratory experiments have been performed at EDF National Hydraulic Laboratory (EDF-LNHE Chatou) in France, HACH-Hydraulic Laboratory of the University of Liège, Belgium and the Laboratory of Hydraulic Constructions (LCHEPFL) in Lausanne, Switzerland (Schleiss, 2011; Blancher et al. (2011); Cicero et al. (2011); Leite Ribeiro et al. (2012); Machiels et al. (2011)).

Although PKW research has intensified, the amount of experimental data is still limited and empirical formulas proposed in the literature need to be applied with due caution. Each new PKW is a unique structure, such that scale models are recommended to assess the design.

This paper presents the results of model tests of Tarzout Dam, conducted at the Hydraulics Laboratory of the Jaroslav Černi Institute (Vojt, 2017).

 

Tarzout Dam Overview

Tarzout Dam is being built across the Tarzout River (Biskra Province, Algeria). The main functions of the dam are to regulate the discharge of the Tarzout River and provide irrigation water supply. Additionally, the dam and its appurtenances need to ensure a safe discharge of Q = 965 m³/s, which is equivalent to a 10,000-year flood.

The dam will create a reservoir with a full reservoir level (FRL) at 80.60 m above sea level. The design calls for an embankment dam, with a clay core (Fig. 2). The dam crest has been set at an elevation of 86.00 m.a.s.l. The width of the crest is 11.00 m. The slope of the upstream face is 1V:2.1H, and of the downstream face 1V:1.65H. The maximum structural height of the dam is 69 m.

The spillway structures are located at the left dam abutment. They comprise a Piano Key Weir (PKW), stepped chute and stilling basin.

The cross-sectional width of the spillway is 37.8 m (width between sidewalls), while the total length of the weir crest is 189 m. The PKW is positioned at the stepped downstream face of the spillway, whose slope is 1V:2H. There the steps are 1.0 m long and 0.50 m high.

Downstream of the PKW, there is a stepped chute whose function is to partially dissipate the energy of the rushing water. The chute width is 37.8 m, while the steps are 1.80 m long and 1.00 m high. Considering the longitudinal slope, the chute can be divided into two sections – the upstream section (from elevation 70.20 m.a.s.l. to 50.0 m.a.s.l.), whose slope is 1V:2H, and the downstream section (from elevation 50.00 m.a.s.l. to 17.0 m.a.s.l), with a 1V:3.3H slope.

The chute extends into a concrete stilling basin, 36.5 m long, 37.8 m wide, and elevation 17.0 m.a.s.l. The stilling basin had been designed as an USBR II type structure, with a 1.65 m high dentated sill at the end.

 

fig02
Figure 2: General layout of Tarzout Dam.

 

Given the complex hydraulic conditions associated with a PKW on one hand and a two-sectional stepped chute on the other, as well as the fact that the Tarzout project is highly specific, scale model testing was conducted for the purposes of the final design.

 

Methodology

Physical Model and Effect of Scale

The physical model of Tarzout Dam (whose length scale was 1:30, Fig. 3), was built in February 2017 at the Hydraulics Laboratory of the Jaroslav Černi Institute for the Development of Water Resources, Belgrade, Serbia.

 

fig03
Figure 3: Scale model of Tarzout Dam at the Hydraulics Laboratory of the Jaroslav Černi Institute for the Development of Water Resources.

 

The model reproduced a 50 m long section of the reservoir, main flood mitigation components (PKW, stepped chute and stilling basin), and a 200 m long downstream river channel.

In the central part of the model, the stepped chute was made from fine concrete, while the walls, PKW and stilling basin were fabricated using transparent Plexiglas plates. All the components of the model reflected the design dimensions and the roughness of the materials met all similarity requirements to prototype conditions.

The selected scale of 1:30 allowed for effective modeling of the spillway capacity, the generation of standing waves in the stepped chute, and the wave propagation in the stilling basin. Given that the Froude similarity requirements were fulfilled, a model of this scale could not effectively reproduce the flow of the air-water mixture (Boes, 2000), such that the chute depths were smaller and the residual energy at the end of the chute higher. This meant that the model could not reliably assess the height of the sidewalls of the chute, while the dimensions of the stilling basin were on the safe side (Kapor et al., 2013).

 

Model Calibration

The objective of calibrating the model was to ensure similarity between prototype flow and the flow on the scale model. Hence, appropriate boundary conditions needed to be established: the flow rate at the upstream boundary of the model and the water level at the downstream boundary.

The boundary condition for the flow rate was archived by the model's water supply system, via weir boxes from which a pre-defined amount of water was supplied to the model.

The boundary condition for the water level is generally achieved by adjusting the water level at the most downstream point, based on rating curves (from a numerical model or field measurements). In this case, the rating curve obtained from a numerical model was used.

 

Measurements on the Model

Different types of measurements were undertaken as part of the model tests. The methods applied for the various parameters are described below.

 

Water Level and Depth

Staff-level gauges (with nonius scale) were used to measure water levels in the reservoir, along the chute, in the stilling basin and in the downstream section. The measurement accuracy was ±0.1 mm.

The water levels in the reservoir and at the downstream boundary cross-section were also measured continuously, by a magnetostrictive water level sensor whose accuracy was ±0.25 mm. These measurements were performed in a stilling well to reduce water level fluctuations. Continuous measurement was supported by a suitable data acquisition unit. Due to the high longitudinal gradient, the depths of the water and the air-water mixture in the stepped chute were measured perpendicular to the streamlines, from the pseudo-bottom determined by the external edges of the steps.

 

Air Concentration in Water

The air concentration in the stepped chute was measured applying a method based on electrical conductivity (Da Silva, 2008). This method relies on the circumstance that the electrical conductivity of water is about a thousand times greater than that of the air. An original electrical conductivity probe was used. The methodology was explained in detail in a previous paper (Starinac et al., 2014a).

 

Velocity

Water velocities along the section downstream from the stilling basin were measured by means of a laboratory-type current meter. The air concentration cross-correlation method via electrical conductivity was used to determine the velocities of the air-water mixture in the stepped chute (Starinac et al., 2014a). The flow velocity was measured by the same probe that was used to measure air concentrations.

 

Flow Rate

Flow rates were measured by means of Bazin type weir boxes, where the head was measured by staff gauges (accuracy ±0.1 mm), resulting in a flow measurement accuracy of ±0.5%. The flow rate was also measured continuously by an ultrasonic flow meter, whose accuracy was ±1.5%. Continuous recording was supported by a suitable data acquisition unit.

 

Pressure and Hydrodynamic Load

Pressure was measured at 46 test points on the bottom and right sidewall of the stilling basin. Pressure sensors with an accuracy of ±0.1% were used. Appropriate data acquisition units were available for synchronous data acquisition.

A standard analysis, previously validated in several other scale-model tests, was applied to assess the hydrodynamic loads in the stilling basin. The main points of the mathematical analysis were described in a previous paper (Starinac et al., 2014a).

Pressure variation time-series for up to 20 points per plate were available for the hydrodynamic load analysis in the present case. The analyzed time series contained 100,000 data points, collected at time intervals of Δt = 0.005 s.

 

Scale Model Results

PKW Investigations

The scale model tests of Tarzout Dam began with preliminary testing of the design concept at three characteristic flow rates: Q0.01%=965 m³/s, Q0.1%=740 m³/s and Q1%=478 m³/s. The hydraulic conditions in the spillway, chute, stilling basin and downstream river channel were analyzed and the most important parameters were measured, to assess dam performance.

Based on recorded reservoir water levels, it was concluded that the spillway – design flow rate of Q0.01%=965 m³/s was achieved with the design length of the weir crest and a head of 3.18 m (design head was 3.4 m). Therefore, the capacity of the spillway was 4.7% higher than designed, so the maximum capacity at the design head of 3.4 m (84 m.a.s.l.) was 1010.8 m³/s. Since the flow conditions at the weir were satisfactory in all the tested cases (Fig. 4), the shapes of the PKW and the piers were retained.

 

fig04
Figure 4:  PKW flow conditions at a flow rate of Q0.01%= 965 m³/s.

 

A spillway rating curve (Fig. 5) was determined for the span from the FRL (80.60 m.a.s.l.) to the MRL (84.00 m.a.s.l.). At the design flow rate of Q0.01%=965 m³/s, the reservoir reached a water level of 83.78 m.a.s.l.

 

fig05
Figure 5:  Spillway rating curve, obtained from the scale model.

 

Stepped Chute Investigations

Flow conditions along a stepped chute are highly specific due to its geometry, characterized by a changing longitudinal slope. The breakpoint between the upstream and downstream sections represents a disturbance in the flow conditions generated in the upstream section. This disturbance is brought downstream by the flow and is progressive, meaning that a small disturbance upstream can result in a large disturbance downstream and cause many potential problems. Therefore, the geometry of the breakpoint (transitional zone) must be designed in such a way that the flow is as smooth as possible. One of the main tasks of model investigations was to define the most favorable geometry of that transitional zone (section). Two alternatives (Fig. 6 and Fig. 7) were investigated at three characteristic flow rates: Q0.01%=965 m³/s, Q0.1%=740 m³/s and Q1%=478 m³/s.

 

fig06
Figure 6:  Alternative 1, one large step.

 

fig07
Figure 7:  Alternative 2, additional small step.

 

The two tested alternatives exhibited similar flow conditions at Q0.01%=965 m³/s (Fig. 8). Photos taken at Q0.1%=740 m³/s (Fig. 9) show a larger depression of the water level in the transition zone for Alt. 1 than for Alt. 2. Tests at Q1%=478 m³/s (Fig. 10) confirmed the above trend.

 

fig08
Figure 8:  Flow conditions in the chute at Q0.01%=965 m³/s; Alt. 1 (left) and Alt. 2 (right).

 

fig09
Figure 9:  Flow conditions in the chute at Q0.1%=740 m³/s; Alt. 1 (left) and Alt. 2 (right).

 

fig10
Figure 10:  Flow conditions in the chute at Q1%=478 m³/s; Alt. 1 (left) and Alt. 2 (right).

 

The water depths along the chute at characteristic flow rates were also assessed for both alternatives. Due to the large gradient of the chute, the depths were measured perpendicular to the bottom of the chute, in six longitudinal directions: center line of the chute and in the vicinity of the right sidewall, at distances of 1.29 m, 4.59 m, 9.15 m and 14.1 m from the right sidewall. This test method was selected in order to register the depth distribution along the cross-section of the chute, as well as any standing waves in the chute. Since the chute is symmetrical, measurements were performed at the right half of the chute, but the results are presented for the full chute width (Fig.11-13; horizontal axis values - distances from the right wall, vertical axis values – distances from the breakpoint, colour – depth according to colour bar at the top of the graph).

The maximum water levels for both alternatives were observed at Q0.01%=965 m³/s in the most upstream cross-section of the stepped chute, as a consequence of PKW influence (Fig. 11). Flow conditions just downstream of the chute breakpoint are formed as a combination of spillway and breakpoint influences. For Alt. 1, in the zone 10-30 m downstream from the breakpoint, the stream lines were complex and the depths were characterized by local depressions and prominences. Water levels for Alt. 1 showed a significant disturbance at the breakpoint, as a consequence of the stream that hit the bottom of the transitional section, causing a high local depth increase at the cross-section 20 m downstream of the break. These phenomena were also observed for Alt. 2, but the transitions were much slighter. Therefore, Alternative 2 was proposed as final.

At all the tested discharges, the water depths down the chute were smaller than the height of the chute sidewalls. However, it should be kept in mind that because of the length scale (1:30) and its effect on the generation of the air-water mixture in the model, the measured depths along the chute need to be increased applying empirical formulas for flow aeration, to reliably determine the appropriate height of the chute sidewalls. Such calculations are especially recommended for the downstream section of the chute, marked in Fig. 14, where the sidewall height is minimal.

 

fig11
Figure 11: Water depths in the chute at Q0.01%=965 m³/s; Alt. 1 (left) and Alt. 2 (right).

 

fig12
Figure 12: Water depths in the chute at Q0.1%=740 m³/s; Alt. 1 (left) and Alt. 2 (right).

 

fig13
Figure 13: Water depths in the chute at Q1%=478 m³/s; Alt. 1 (left) and Alt. 2 (right).

 

fig14
Figure 14: Water levels along the chute at Q0.01%=965 m³/s.

 

Stilling Basin Investigations

The flow in the stilling basin was satisfactory in terms of energy dissipation and hydraulic jump position. At Q0.1%=740 m³/s (Fig. 15), which was used as the design flow rate, as well as at lower flow rates, the hydraulic jump was kept within the stilling basin. No significant difference was observed between the two tested alternatives.

 

fig15
Figure 15:  Flow conditions in the stilling basin at Q0.1%=740 m³/s; Alt. 1 (left) and Alt. 2 (right).

 

The water levels along the center line of the stilling basin, at all test discharges, are shown in Figure 16.


fig16
Figure 16: Water levels in the stilling basin; Alt. 2.

 

It is apparent in the plot that the hydraulic jump position was correct at the design flow rate of Q0.1%=740 m³/s. At Q0.01%=965 m³/s, the hydraulic jump was pushed a bit downstream, but that was acceptable due to design criteria.

Based on the tests described above, it was concluded that the initial design of the stilling basin complied with structural safety requirements, so it was retained.

 

Testing of the Modified Design Concept

Comprehensive measurements of various hydraulic quantities relevant to subsequent design stages were undertaken for the proposed layout of the entire project (Vojt, 2017).

Hydrodynamic loads on the bottom and right sidewall of the stilling basin were assessed at three characteristic flow rates: Q0.01%=965 m³/s, Q0.1%=740 m³/s and Q1%=478 m³/s. Pressure fluctuations were generally proportional to the average pressures. The average pressures at the bottom of the stilling basin were considerably higher than the pressures on the sidewalls, while the fluctuations were of the same order of magnitude.

The results – average values and pressure fluctuations at the design flow rate of Q0.1%=740 m³/s – are shown in Figures 17 and 18 for test points on the bottom of the stilling basin, and in Figures 19 and 20 for the right sidewall of the stilling basin. The highest average pressures were recorded at the downstream end of the stilling basin, while the highest pressure fluctuations were registered at the upstream end of the stilling basin, immediately after the inrush of water leaving the chute. No threat of negative pressures in the entire stilling basin at flow rates up to Q0.1%=740 m³/s was revealed.

 

fig17
Figure 17: Distribution of average pressures at the bottom of the stilling basin, Q0.1%=740 m³/s.

 

fig18
Figure 18: Distribution of pressure fluctuations at the bottom of the stilling basin, Q0.1%=740 m³/s.

 

fig19
Figure 19: Distribution of average pressures on the right sidewall of the stilling basin at Q0.1%=740 m³/s.

 

fig20
Figure 20: Distribution of pressure fluctuations on the right sidewall of the stilling basin at Q0.1%=740 m³/s.

 

The results obtained at the maximum flow rate of Q0.01%=965 m³/s (Fig. 21 – Fig. 24) showed a less favorable pressure distribution in the stilling basin, as a consequence of the hydraulic jump position (which is not submerged in the stilling basin, but pushed downstream). The distribution of average pressures at the bottom of the stilling basin (Fig. 21) was similar to that recorded at the design flow rate (Fig. 17), with the highest values recorded at the downstream end of the stilling basin. On the other hand, the highest pressure fluctuations were recorded in the downstream half of the stilling basin (Fig. 22), due to the turbulence produced by the hydraulic jump. The situation on the right sidewall (Fig. 23 and Fig. 24) was analogous. However, this flow rate was tested as the control rate, so a considerable risk of negative pressures was acceptable.

 

fig21
Figure 21: Distribution of average pressures at the bottom of the stilling basin, Q0.01%=965 m³/s.

 

fig22
Figure 22: Distribution of pressure fluctuations at the bottom of the stilling basin, Q0.01%=965 m³/s.

 

fig23
Figure 23: Distribution of average pressures on the right sidewall of the stilling basin at Q0.01%=965 m³/s.

 

fig24
Figure 24: Distribution of pressure fluctuations on the right sidewall of the stilling basin at Q0.01%=965 m³/s.

 

Hydrodynamic load assessments included analyses of the total hydrodynamic forces that acted on the surface of each plate. As a logical consequence of the pressure distributions, the average forces in the stilling basin were considerably higher at the bottom than on the sidewalls, while the fluctuations were of the same order of magnitude. Given that the fluctuations were considerably smaller than the average force, no part of the stilling basin was found to be threatened by negative forces, even at the highest test discharge.

Detailed water level and flow velocity measurements were made at characteristic flow rates in the section downstream from the stilling basin. Figure 25 shows the results for the maximum flow rate of Q0.01%=965 m³/s. All the velocities along each vertical are shown as vectors, whose magnitude is indicated by arrow length. The raster pattern is related to the velocities at the bottom of each vertical, which are of key importance for determining any erosion threat. The highest bottom velocities were recorded in cross-section No. 5 (100 m downstream of the end of the stilling basin), reaching about 5.5 m/s. That cross-section represented a natural terrain narrowing – constriction, and could be considered as the critical – control section, which ensured the submerging of the hydraulic jump in the stilling basin. In view of its importance, the recommendation is to ensure its stability and thus protect the riverbed against potential scouring due to high velocities.

 

fig25
Figure 25: Water velocities downstream from the stilling basin at Q0.01%=965 m³/s.

 

Conclusions

Scale model testing of a specific engineering project on a physical model helped detect and solve problems that could not be duly examined in the design.

The original design of the stepped chute required certain modifications. Model tests revealed that superior flow conditions were achievable with Alternative 2, which was adopted as final.

Scale model testing of the adopted design concept showed that it ensured safe evacuation of water at all the considered discharges.

 

Acknowledgments

The authors express their gratitude to the Ministry of Education, Science and Technological Development of the Republic of Serbia for providing financial support for Technology Development Projects TR 37010 and 37014.

 

References

Anderson, R.M. and B. Tullis (2011). Influence of Piano Key Weir geometry on discharge. Proc. Int. Conf. Labyrinth and Piano Key Weirs Liège B, pp. 75–80, CRC Press, Boca Raton, FL.

Anderson, R.M. and B. Tullis (2012). Comparison of Piano Key and rectangular LabyrinthWeir hydraulics. J. Hydraulic Eng. 138(4), pp. 358–361.

Boes, R., M. (2000). Scale effects in modelling two-phase stepped spillway flow. Proceedings of the International Workshop on Hydraulics of Stepped Spillways, A.A. Balkema, Rotterdam, Netherlands, ISBN 90 5809 135 X, pp 53 – 60, 2000.

Blancher, B., Montarros, F. and F. Laugier (2011). Hydraulic comparison

between Piano Key Weirs and labyrinth spillways. Proc. Int. Conf. Labyrinth and Piano Key Weirs, Liège B, pp. 141–150, CRC Press, Boca Raton, FL.

Cicero, G.M., Menon, J.M., Luck, M. and T. Pinchard (2011). Experimental study of side and scale effects on hydraulic performances of a Piano Key Weir. Proc. Int. Conf. Labyrinth and Piano Key Weirs, Liège B, pp. 167–172, CRC Press, Boca Raton, FL.

Da Silva, M., J. (2008). Impedance Sensors for Fast Multiphase Flow Measurement and Imaging, Doctoral thesis, Electrical and Computer Engineering, Technische Universität Dresden, Dresden, (2008).

Kаpоr, R., Sаvić, Lj., Zindоvić, B., Žugić, D., Stаrinаc, D. and P. Vојt (2013). Hidrаuličkа аnаlizа struјаnjа nа stеpеnаstоm brzоtоku [Hydraulic analyses of stepped spillway flow]. Grаđеvinski kаlеndаr, Vol. 46, December 2013, p. 70-127.

Leite Ribeiro, M., Pfister, M., Schleiss, A.J. and J-L. Boillat (2012). Hydraulic design of A-type Piano Key Weirs. Journal of Hydraulic Research, Vol. 50, No. 4, 2012, pp. 400–408.

Lempérière, F. and A. Ouamane (2003). The piano keys weir: A new cost-effective solution for spillways. Hydropower & Dams 10(5), pp. 144–149.

Machiels, O., Erpicum, S., Dewals, B.J., Archambeau, P. and M. Pirotton (2011). Experimental observation of flow characteristics over a Piano Key Weir. Journal of Hydraulic Research Vol. 49, No. 3 (2011), pp. 359–366.

Schleiss, A.J., (2011). From labyrinth to piano key weirs: A historical review. Proc. Int. Conf. Labyrinth and PianoKey Weirs, Liège B, pp. 3–15, CRC Press, Boca Raton, FL.

Starinac, D., Vojt, P., Damnjanović, M., Žugić, D., Savić, Lj., Kapor, R., Zindović , B. and R. Glišić (2014). Scale Modeling of the Bouzina Dam Flood Mitigation Structures, Journal of Serbian Water Pollution Control Society „Water Research and Management", ISSN 2217-5237, Vol. 4, No. 1, pp. 31-42, 2014.

Starinac, D., Kapor, R., Savić, Lj., Vojt, P., Žugić, D., Damnjanović, M., Zindović, B., Đajić, P., (2014). Air-Water Flow on a Labyrinth Spillway, Journal of Serbian Water Pollution Control Society „Water Research and Management", ISSN 2217-5237, Vol. 4, No. 3, pp. 11-20, 2014.

Tullis, J. P., Amanian, N. and D. Waldron (1995). Design of Labyrinth Spillways. Journal of Hydraulic Engineering. Vol. 121, No. 3, March 1995, pp. 247-255.

Vischer, D. L. and W.H. Hager (1999). Dam Hydraulics, John Wiley & Sons, NewYork, USA.

Vojt, P. (2017). Hidrаuličkа mоdеlskа ispitivаnjа brаnе Tarzut - Kоnаčаn izvеštај [Scale model analyses of the Tarzout dam – Final Report]. Jaroslav Cerni Institute for the Development of Water Resources (JCI), Belgrade, July 2017.