Hydraulic modelling and validation for the study of smolt migration
Conference contribution, 2004
In order to assess the prerequisites for downstream migration of smolt the flow field upstream two hydropower stations in two large Swedish rivers have been made, and the simulations have been validated against stream velocities measured on site. The validated models have been used to compare the flow paths with the measured swimming paths of smolts. The simulations were done by solving the three dimensional Reynolds-averaged Navier-Stokes equations using the Fluent computational fluid dynamics (CFD) package.
One of the studied rivers, Ume älv, whose main course is completely regulated, has a large unregulated tributary river, Vindelälven, which holds important spawning grounds for salmon (Salmo salar) and trout (Salmo trutta), and downstream of the confluence of the two rivers there is only one hydropower plant, the Stornorrfors power plant. This constitutes a major hindrance for fish migrating between the Bothnian Sea and the spawning grounds of the river Vindelälven. The juvenile fish, smolt, leaving the Vindelälven downstream towards the ocean can migrate either through the turbines and deep rock tunnels of the power plant or through the surface weirs at the regulating dam. Because of the flow regime during the migration period of smolts most smolts are believed to follow the main stream through the turbines leading to a mortality of approximately 25 %.
The other studied river, Pite älv, also contain some spawning grounds in smaller tributaries upstream its first power plant. Although this plant is more like a-run-of-the-river plant most smolt passes the turbines instead of going over the spillways.
The objective of the study is to determine the migration paths of salmon and trout smolts and relate their behaviour to hydraulic factors. The results from the study could then be used to improve smolt survival by taking appropriate measures to divert the smolts from the turbines, e.g. by changing spill rules or installing diversion systems.
In Ume älv in Spring 2002, 22 salmon and 22 trout smolts were tagged with internal, individually coded radio transmitters (ATS). After the release approximately five kilometres upstream of the dam at Stornorrfors power plant their passages were registered at three stationary recorders and their movements tracked manually in the horizontal plane along their swimming paths. Simultaneously, velocity profiles were measured in selected cross-sections using an Acoustic Doppler Profiler (ADP). In spring 2003 16 smolt were tagged, released and tracked in the same way in Pite älv. In 2004 also the swimming depth was recorded.
Results from the tracking, flow measurements and flow simulation indicate that the smolts move along the main stream of the river. In Ume älv the tracking data only yielded information on the smolt positions in the horizontal plane but it was found that the water velocities in the upper part of the water column along the smolts migration paths were close to the mean transport velocity of the smolt. In Pite älv Spring 2004 the swimming depth was measured to between 0.5 to 2 m below the surface. It can thus be concluded that the smolt migration is passive i.e. the smolt drift passively with the water.
This study is part of a more extensive project that comprises studies of both upstream and downstream migration of fish. The objective of the project is to study the correlation between flow hydraulics and fish migratory behaviour. The ultimate goal is to create numerical fish models that are programmed to act as real fish in the CFD model. To get realistic models it is then first necessary to map real fish migratory behaviour to get to know the rules of the fish behaviour.