3D RANS MODEL TO PREDICT STRATIFICATION EFFECTS RELATED TO FISH PASSAGE AT HYDROPOWER DAMS

The objective of the present study is to construct and validate a 3D CFD model capable of simulating the complex hydrodynamic and thermal conditions in the forebay and turbine intakes of McNary Dam on Columbia River [Figure: McNary1.jpg] and to demonstrate the potential of the CFD model to analyze the impact of structural and operational changes under consideration by USACE on the temperature distribution within the fish passages and gatewells. 

Sub-yearling fall chinook salmon emigrate past McNary Dam [Figure: McNary4.jpg] in the summer when large thermal gradients are observed in the forebay and within the gatewells of McNary Dam [Figure: McNary2.jpg].  In particular, this effect is very pronounced in the shallower part of the forebay at the southern end near the dam where surface waters are warmer.  These factors contribute to the drawing of warmer surface water into the gatewells through which the diverted fish is collected and transported via fish orifices and the collection channel.  This thermal shock negatively affects the chances of the fish, especially of the juvenile salmonids, to survive. 

The model contains all the relevant geometrical details of the hydraulic structures and forebay bathymetry including all the 14 powerhouse units [Figure: McNary5.jpg], the 22 spillway bays, and a reach of the forebay extending about 13,000 ft upstream the dam.  Over this distance, the forebay bathymetry is highly variable, especially near the river bank areas which are very shallow.  The model also includes the vertical barrier screens (VBS) in the gatewells of all 14 units, the extended bar screens (ESBS) and the navigation lock [Figure: McNary2.jpg].  Inclusion of all these structures was necessary due to their influence on the transport and mixing of water temperature within the forebay and dam.  The mesh of the Full Forebay model contains close to 6 million cells. 

The Boussinesq flow module in FLUENT is utilized to obtain the flow hydrodynamics (the k-e model is used) and temperature distributions.  The model is first validated using temperature field data collected during summer 2004 at 46 stations within the forebay and at the gatewells of the 14 powerhouse units by the US Army Corps of Engineers (USACE).  It is found that the model is able to accurately predict the salient features of the thermally stratified flow in the forebay and at the gatewells for two test cases (with no spill and with spill) corresponding to strongly stratified conditions.

Figure: Sketch showing the position of the stations (6 transects, 46 stations) where temperatures were measured inside the forebay [McNary3.jpg]

Figure: Comparison between measured and predicated temperature profiles at stations T1P1 to T4P6 (Validation Case_2: with spill, June 30, 2004).  Continuous blue line: simulation; circle symbols: mean field data measurements, thin horizontal lines: one standard deviation of experimental data about its mean (4 hours average) [Graphic2.tif]

Figure: Comparison between measured (2h mean +/- one standard deviation) and simulated gatewell temperatures in various gate slots of the 14 powerhouse units (main bay of each unit is shown) (Case_2: with spill, August 16, 2004) [gatewell.JPG]

Figure: Comparison between measured (4h mean +/- one standard deviation) and simulated gatewell temperatures in various gate slots of the 14 powerhouse units (main bay of each unit is shown) (Case_1: with no spill, August 16, 2004) [gatewll2.JPG]

Several solutions to alleviate fish passage problems are considered.  They include introduction of a barrier curtain upstream the intake units situated near the southern shore that would act like as a selective withdrawal barrier curtain to prevent movement of warm surface waters into the turbine intake and modification of the roof geometry of the intake units.  Once developed and tested, the model will be useful for future studies related to fish passage, correlation of fish behavior with local hydrodynamics, etc.