Computational analysis of fish survival at the John Day Powerhouse Jim Kiel Laurie Ebner
Introduction Design a CFD model that can be used to predict the pressures in the turbine runner environment. What is the probability that fish passing through a turbine runner will encounter a pressure drop that negatively affects mortality? Determine the nadir pressure through the turbine runner environment
CFD Turbine Model Development Utilized industry experience and chose a turbine designer to perform the modeling VATech currently Andritz Utilized software typically used by turbine designers ANSYS CFX CFD model was designed for more inclusive sections of the prototype. Forebay, trash racks, VBSs, STSs, Tailrace
CFD Model
Domains Two Domains: Runner and Draft Tube Intake
John Day CFD Pressure Predictions Maximum uncertainty in predicting the absolute mean pressure in the turbine runner of approximately ± 2 PSI in very small zones on the pressure side of the runner with better accuracy throughout the rest of the domain.
Operating Conditions The following tables demonstrate results for all operating points Comparison of flow rates were performed Upper 1% Peak and Lower 1% Kaplan Blade (at one point +/-1% off cam)
Upper 1% Variable OP3 (Upper 1%) OP5-C (OP3 - corrected Q) GV Angle (Absolute) 40.8 40.8 Blade Angle (Absolute) 31.6 31.6 Discharge (cfs) given 19900 20895 Recomputed Discharge (cfs) 20895 20895 % in Discharge 4.76% 0.00% Heat (ft) given 102.9 102.9 Output (MW) HDC 154.97 154.97 Head (ft) computed 83.0 100.1 Runner Output (MW) computed 125.65 159.16 Efficiency (%) computed 88.65 90.13 Difference in Head 19.33% 2.69% Difference in Output 18.92% -2.70% Volume of Zone Below Threshold 7.25 PSIA 0.000316 0.00066 10 PSIA 0.000565 0.0028 14 PSIA 0.00727 0.0286
Validation/Results Good Match Between Prototype Flow Rate and Model Flow Rate for: Peak Upper 1% Kaplan Blade (on cam, + 1%)
Probability Analysis CFD post-processing, Using streamlines (massless, neutrally buoyant particles) to simulate the passage environment and compute statistics. 1. Streamlines can be generated from a section plane upstream of the runner zone using off-the-shelf software (TecPlot, FIELDVIEW). 2. The value of the lowest pressure (nadir) along each streamline is determined and associated with the release point. 3. The pressure nadirs are plotted and contoured at the release section.
Streamline Method section plane The CFD model provided by VA TECH has radially periodic boundaries Generating streamlines more difficult than with a complete 360-degree model As streamlines exit a periodic boundary, a new streamline is generated At corresponding opposite boundary, it continues the particle s passage through the model.
Streamline Method When radial duplication is turned on (i.e., the model is replicated the appropriate number of times about the axis of symmetry), the streamline appears in the expected manner.
Streamline Method To characterize as many passage options as possible, we densely seed the section plane. The above figure shows the approximately 65,000 seeds used in the analysis.
Streamline Method Only a very small percentage of paths have nadirs below a critical value of 7.25 psia. These nadirs occur mainly on the outside of the runner blade.
Streamline Method Uniform Fish Distribution Total area of section = 6.75 m 2 Area of critical nadir = 0.035 m 2 Probability of encountering critical nadir = 0.035 / 6.75 = 0.52% If we assume that fish crossing this section plane are uniformly distributed, then the area enclosed by the critical nadir value (7.25 psia) divided by the total area of the section is the probability of fish encountering this critical nadir.
Comparison of Volumetric Fluid (Iso-Surfaces) to Streamlines Operating Point Comparison Volume at 7.25 PSIA Streamlines at 7.25 PSIA OP1 (Peak) 0.000124 0.00512 OP2 (Lower 1%) 0.0000106 0.00125 OP3 (Upper 1%) 0.000316 0.00693 OP4 (Between Peak and Upper 1%) 0.000308 0.00612 OP5-A (OP4 + 1%) 0.000406 0.00474 OP5-B (OP4-1%) 0.000309 0.00981 OP5-C (OP3 - corrected Q) 0.00066 0.00937 Volume 0.0007 0.0006 0.0005 0.0004 0.0003 0.0002 0.0001 0 at 7.25 psia y = 9.469401x 2-0.019852x + 0.000017 R² = 0.978381 0 0.002 0.004 0.006 0.008 0.01 Streamlines
Streamline Method Comments The streamline method for determining probabilities of passage through zones of interest in the turbine runner is viable and is automated to a large degree. The method can be performed using commonly available commercial CFD software (Tecplot and Excel). The design of the model (number and types of blocks, rotating coordinates, transient solutions, etc.) will affect the application of this technique to other situations. The ability to use known fish distributions to weight the results of the nadir contouring should significantly improve the prediction accuracy of this technique.
Sensor Fish to CFD Comparison Sensor Fish can track acceleration and pressure through a runner environment Sensor can be released from specific locations- mid blade, hub released, etc The following slides demonstrate a comparison of this data in terms of nadir pressures in the turbine runner.
Sensor Fish to CFD Comparison Peak 1 0.9 0.8 0.7 Probability 0.6 0.5 0.4 0.3 0.2 0.1 0 0 10 20 30 40 50 60 Nadir PSIA Sensor Fish All CFD Sensor Fish Tip Sensor Fish Mid
Results Demonstrate the Following: CFD results tend to be to the right of sensor fish data Use the bead data, below are cross sections taken from the observational model at ERDC: JOHN DAY TURBINE MODEL Q = 20,510 CFS SENSOR FISH - MID BLADE TO HUB RELEASE JOHN DAY TURBINE MODEL Q = 20,510 CFS SENSOR FISH - TIP RELEASE UPSTREAM UPSTREAM DISTANCE FROM CENTER O RUNNER, FT 15 10 5 0-5 -10-15 DISTANCE FROM CENTER O RUNNER, FT 15 10 5 0-5 -10-15 -15-10 -5 0 5 10 15-15 -10-5 0 5 10 15 DISTANCE FROM CENTER OF RUNNER, FT DISTANCE FROM CENTER OF RUNNER, FT DS RING HUB Series3 DS RING HUB Series7
100% Sensor Fish and CFD Data for John Day OP5-C - Upper 1% approximate blade angle of 31.6 Degrees 90% 80% Probability of Occurrence (%) 70% 60% 50% 40% 30% 20% 10% 0% 0 5 10 15 20 25 30 Nadir Pressure (PSIA) Sensor Fish Upper 1% CFD - Upper 1% with Radial Offset CFD - Upper 1% without Radial Offset
100% Sensor Fish and CFD Data for John Day OP1 - Peak approximate blade angle of 26.20 Degrees OP5-C - Upper 1% approximate blade angle of 31.6 Degrees OP2 - Lower 1% approximate blade angle of 19.6 Degrees 90% 80% Probability of Occurrence (%) 70% 60% 50% 40% 30% 20% 10% 0% 0 5 10 15 20 25 30 35 40 45 Nadir Pressure (PSIA) Sensor Fish Peak Sensor Fish Upper 1% Sensor Fish Lower 1% CFD - Peak with Radial Offset CFD - Upper 1% with Radial Offset CFD - Lower 1% with Radial Offset
Analysis and Limitations of Tools CFD Very poor performance near blade tips due to mesh resolution, boundary layer and turbulence model issues, etc. Sensor Fish Could be in relative motion with fluid in blade tip region due to presence of tip vortices. Rotation of sensor fish will create a local decrease of static pressure on its surface Surface contacts (or impacts?) will affect pressure readings Beads Trajectories in the physical model may not be representative of prototype May not represent sensor fish trajectories Streamline method No inertial and turbulence effects No collision estimate can be done in future
Recommendations and Conclusion This analysis and investigation demonstrates the viability of utilizing CFD to describe the pressure conditions within the turbine runner environment. Use the tools within the context of their capabilities The findings here were applied, as already seen, to our Ice Harbor fish friendly runner replacement