Ciro Galo Menéndez Alcázar
Centro de Investigaciones y Estudios en Recursos Hídricos Cierhi
Departamento de Ciencias del Agua
Escuela Politécnica Nacional. Quito -Ecuador
P.O.Box 17-01-2759. Fax 593-2-647848. E-mail: cierhi@mail.epn.edu.ec
Abstract: The work gives the experimental results, carried out in a physical model scale of 1:40, of the flow over a lateral channel spillway and on the other discharge of flow rate works of the Chongón dam, located on the same name river in the Guayas Province of Ecuador. The most relevant hydraulic parameter and characteristics of flow patterns during the verifications tests on original design of side spillway, outlet discharge , side channel, rectangular chute, chute transition, and on the stilling basin for energy dissipater, are described and analysed. In the same way, the test results of the optimal alternative propose are indicated. Finally, the conclusions express the arrived results as a consequence of the discharge hydraulic system optimisation, with several test for a range of flow rate corresponding to different return periods, with the maxim probable flow rate included.
Keywords: chongón, side channel spillway, trough, deflector bucket, tunnel outlet
A side channel spillway consist of an ogee crest and is commonly used to release flow from a reservoirs in sites where the side is steep and rises to a considerable height above the dam. By this way, the water falls over the spillway crest into a channel, where the flow is parallel to the crest in order to carry the water directly to the discharge channel. Because of both, turbulence and vibrations originated in the side channel flow, a side channel design is ordinarily no considered. However, it must be done where a good foundation such as rock exists. A trapezoidal cross section is the most often employed for the side channel throat. Besides, a control section is usually constructed downstream from the side channel throat in order to produce a point of critical flow.
The overflow on a side channel spillway is considered as a spatially varied flow. A substantially corrected form of its fundamental background of the differential equation with increasing discharge was probably first established by Hinds [1926] for the design of lateral spillway channel. Then, a more complete equation was developed by Favre [1933] including a friction term and a component of inflow velocity. The method developed by Hinds and Favre requires a step computational procedure with successive approximations. Camps [1940] and then Li [1955] integrated the differential equation of the flow for prismatic rectangular and trapezoidal channel with uniform inflow throughout the channel length. Theoretical and experimental studies of the flow have been also carried out by De Marchi [1941], Citrini [1941], Forchheimer, Schoklitsh, and their results have been applied to design side channel spillways of many dam that have been built around the world.
The theoretical concepts indicated above, were the base of the lateral spillway design of the Chongón dam which is considered as the main Infrastructure work of the Interbasin Project from the Daule river to the Peninsula of Santa Elena in Ecuador. The main goal of the dam is to store both flow from the river and flow coming from the Interbasin. This storage will be used for the irrigation of 44.000 ha., and for water supply purposes. The dam is a 45 m height earth dam with a capacity of 100 Hm³. The flow in excess discharge through the side channel spillway and eventually by the tunnel outlet as well. The present paper is a summary of the hydraulic model experiments carried out for design verification purposes and hydraulic performance improvements of the lateral spillway, the discharge channel, the chute, the stilling basin and the tunnel outlet
In order to observe and calibrate the approaching flow patterns from the reservoir water body close to the side channel spillway, a model was reproduced 300 m upstream from dam axe, 300 m to the right and 100 m to the left of the discharge channel axe, taking into account the topographical concerns, ground elevation and boundary conditions. The discharge tunnel from inlet to outlet was simulated as well. Fig. No1
It is a fact that the flow discharge through the spillway and the open channel is free, where the gravity forces are predominant. Therefore, the Froude Similitude Law with a ratio of 1: 40 for the physical scale model was adopted. Regarding the reservoir basin configuration and the river bed downstream from the stilling basing, both of them were built using a soil - cement mixture. The side channel spillway and the channel walls were built in Plexiglas and for the bottom, wood was used. For discharge measurements, a 90–degree V-Notch weir was used and to measure the longitudinal and cross water surface profiles, some points and hook gages were used. Finally, the flow velocity was measured using a propeller type current meter and total head pitot tubes.
In order to measure positive water pressures at the spillway profile, the side channel, the chute, the transition, and the stilling basin bottom, 94 piezometer taps were installed and connected each one by flexible tubing to a single leg manometer, and a some point, where the pressure fluctuation was considered as critical, a transducer was used. The water recovery bed channel downstream was simulated 200 m downstream of the stilling basin. For the river channel, a mobile weir was installed and the end.
In order to know the hydraulic performance for both the spillway and the lateral channel, several tests were carried out in the scale model. These tests were performed for a set of discharges of 186, 250, 481 m3/s. The different test considered not only free flow discharge but also submerged flow discharge. In the approaching channel to the spillway the flow pattern observed and measured into the left side was 7% more accelerated than that one on the right side. This was because of the higher friction resistance on the right bank. This concern does affect neither flow distribution nor discharge coefficient on the spillway for any discharge condition.
The stage in the side channel trough is characterised for its high turbulence and it is subcritical, due to the control section located downstream from the side channel. For discharges higher than 400 m³/s., the discharge is submerged in the side channel trough and strong oscillations of the surface water are observed. Fig. No3. The side channel work is an energy dissipater basin where a swirl horizontal flow is generated as a consequence of flow falling from the ogee and from the opposite, confined, side channel.. Furthermore, the swirl flow is developed downstream into the chute.
Regarding the state flow into the chute, this is characterised for the presence of cross waves at the initial reach. Besides, this situation becomes critical for discharges higher than 400 m3/s. when an overflow takes place on the chute´s wall. However, as the flow is accelerated downstream, the water depth is lower and the overflow disappears. When the flow reaches the transition or the cross wave diffuser channel, such waves generated upstream tend to disappear. Then, the cross section water profile becomes uniform. Photo No1. Regarding the pressure distribution along the chute, negative piezometric pressure was not registered. Besides, the maximum mean flow velocity reached a value of 25 m/s., just at the toe of the chute for the highest simulated discharge.
Due to the flow discharge coming from the chute, just in front of the tunnel outlet two effect are observed. The first one is an a deficient aeration of the tunnel outlet and the second one is an overflow through the top walls of the central channel located in the stilling basin. In the second case, low piezometric heads are recorded on the inner and outer side wall due to the flow fluctuation caused by the water that is falling.
The stilling basin worked efficiently in such manner that the energy dissipated was about the 70 % for all of the discharges tested. The mean velocities recorded at the downstream end were 1.25, 1.40 and 3.10 m/s. for the three discharges tested. The lowest piezometric head recorded was 1.7 m on the stilling basin. On the other hand, the simultaneous discharges of water by the side spillway and the tunnel outlet permitted us to observe that for partial opening of the bottom outlet gates, the flow discharge is not free anymore at the end of the tunnel. This is because of the fact that a hydraulic jump is formed in that place and then, this standing wave travels upstream in such a way that the tunnel discharge is not aerated any longer.
In order to improve the flow pattern through different flood structures, several alternatives were tested on different places as the following: the approaching channel to the spillway and the lateral configuration of the bottom transition between the channel though and the chute. Flow acceleration, a better cross profile at the initial reach of chute, and less overflow on the walls were obtained. On the other hand, in order to avoid the interference of flows coming from the spillway and from the tunnel outlet, some deflector buckets with different dimensions and profiles were tested. The experiments were carried out in such a way that the high-velocity flow that is leaving the deflector lip is directed upward to the stilling basin pool. In this way, a free discharge of the tunnel outlet and a high entrained of air by the water yet were obtained. Those effects help to damp the dynamic pressure fluctuations in the stilling basin.
After several tests, depending on the hydraulic performance observed at the model and taking into account some technical criteria, the final alternative was defined. The changes to the preliminary design are basically the following: in the transition between the lateral channel and the control section on the side upper part, the link alignment was modified from one straight reach to one curve with a radius of 40 m. Fig. No2. The height of the chute walls at the initial reach was increased in 1.0 m in order to avoid lateral overflow.
A 5.6m wide final deflector flow bucket was built on the transition chute, 2.63m upstream from the tunnel outlet exit. The bucket has a curvature radius of 4.8 m. Its lateral walls are 4.0 m long and its height is variable from 1.6 m at the downstream end to 0.0 upstream at the intersection with the bottom of the transition chute. Fig. No 4. Regarding the other structures as the spillway ogee, the transition chute, the stilling basin and the tunnel outlet, they kept their preliminary hydraulic architecture. Velocities, flow profiles and pressure distributions were measured with these discharges.
The oscillation effects, upstream from the chute, determine the existence in depth between the water depth registered at the central axis of the channel and that registered at the walls. This is a dumping effect caused by the interference of waves generated upstream. This difference decreases up to 0.1 or 0.2 meters along the cross section in the diffusing channel. The minimum water pressure registered is of 1.38m located in the site where the slope changes.
The discharges tested at the deflector site were of 100, 186, 269 and 328 m3/s. with a reach between 29 and 36 m. Moreover, the water impact is observed directly on the pool of the stilling basing. It allows a free discharge (without interference) from the tunnel outlet and no hydraulic jump is formed. Photo No2
Finally, regarding the high turbulence at the stilling basin , there are wave fluctuations of about 2.0 m. This value was registered when the maximum discharge was tested. There is also a damping effect for these fluctuations downstream at the stilling basing due to the big mass of water present at that site. The flow velocities, after the dissipation of energy, varied between 2.16 and 1.5 m/s.
There is a bi-dimensional condition in the observed flow due to the high turbulence originated inside the lateral channel, the generation of swirl horizontal flow and its downstream propagation through the discharge channel. Therefore, those factors were not considered for the one-dimension flow model and the experimental results are necessarily different from those calculated.
The hydraulic behaviour observed for both the spillway and the modified transition is acceptable for the whole set of discharges that were tested. However, for security reasons, it was necessary to recommend to increase the height of the chute walls on its initial part
The best choice to avoid interference of both the flow that is coming from the spillway and that one coming from the bottom outlet was the incorporation of the deflector buckets. The effect due to the impact of the water jet, downstream at the stilling basin , was insignificant. Moreover, a bigger damping effect of the strong mass of water pulsation at the stilling basin was achieved because of the fact that more air was entering to the system.
References
A water Resources Technical Publication. Design of Gravity Dams. Denver 1976, USA.
Bureau of Reclamation. Model Studies of Spillway, Boulder Canyon Proyect, Final Report. Denver 1938.
Menéndez C. Experimental studies in scale
physical model of the outlet works of the Chongón Dam. Final Report . Escuela Politécnica
Nacional . Julio 1989.
Ven Te Chow . Open Channel Hydraulics. 1959 McGraw – Hill Book Company.
Fig.1
Fig.2