USACE_Detrot_Forebay_CFD_Modeling_Report.docx

Draft Report

Detroit Dam Long Term Temperature Control Alternatives Study Computational Fluid Dynamics Modeling

CFD Image

April 2012 i

ABBREVIATIONS AND ACRONYMS

Detroit Dam Fish Passage Alternatives Study CFD Modeling Draft Report

September 2011 Draft CFD Modeling Report

Detroit Dam Fish Passage Alternatives Study Computational Fluid Dynamics Modeling

Table of Contents

Abbreviations and Acronyms Executive Summary

LIST OF TABLES

Table 1: Detroit CFD Modeling Plan 19

LIST OF FIGURES

Figure 1: Detroit CFD model domain.

Figure 3: Polyhedral grid at the WTC tower.

Figure 4: Polyhedral grid in the WTC tower cul-de-sac, and region boundaries. 7 Figure 5: Polyhedral grid upstream of the WTC tower cul-de-sac.

Figure 6: Upstream boundary locations used in all alternative analysis runs. 11 Figure 7: Downstream boundaries used as outlets in the CFD baseline and alternatives runs.

Detroit Dam Long Term Facility Alternatives Study CFD Modeling Draft Report

April 2012 Introduction Background National Marine Fisheries Service (NMFS) 2008 Willamette Biological Opinion describes several Reasonable and Prudent Alternatives (RPAs) for downstream fish passage that need to be evaluated and implemented within the next 15 years. NMFS considers these measures essential to recovery of Endangered /species ACT (ESA) listed fish in the Willamette basin.

RPA 5.2 requires investigation of the feasibility of improving downstream temperatures and reducing Total Dissolved Gas (TDG) exceedances in the North Santiam River for ESA – listed fish species. Interim steps must begin no later than 2010, which may include feasibility studies, a design report, authorization and appropriation, and plans and specification if appropriate. As part of the effort, evaluation of alternatives to achieve both temperature control and fish passage will be required. A Major Milestone date near the end of 2011 will be established to determine a “go/no go” on the feasibility of temperature control facilities at Detroit Dam. The Goal is to complete construction of any structural temperature control facilities by December 2018, with permanent temperature control operations beginning by March 2019.

The objective of the overall project is to prepare an Alternatives Report (AR) that addresses the feasibility of and project-specific alternatives for a long term temperature control facility that reduces the number of TDG exceedances in the North Santiam River. The design will also take into account the potential for inclusion of downstream fish passage facilities. Alternative evaluation level of design of those facilities will be included in the AR. The AR will develop a broad range of temperature control structures and TDG reduction alternatives, evaluate, reduce, and refine the alternatives, and present relative costs.

A CFD model of the Detroit Reservoir was recommended to provide information on forebay hydraulics and temperatures with and without proposed alternatives. To analyze the existing conditions, impacts of alternatives, and investigate operations for the DDR. The CFD model will ideally be capable of simulating hydraulic zones of influence and temperature mixing and stratification. The CFD model will provide hydraulic and temperature information that will be used to refine alternatives and evaluate alternatives for further development in the DDR.

The structural and operational alternatives for Long Term Facility at Detroit Dam will be updated and described in a 60% Alternatives Report based on the above analyses. The preferred Long Term Facility alternative will be carried to the 90% and Final Alternatives study level by the Detroit Product Development Team (PDT) in FY12 prior to moving to Design Documentation Report (DDR) and Plans and Specifications (P&S).

A hydraulic model of the Detroit Reservoir was recommended to provide information on forebay hydraulics and temperatures with and without proposed alternatives. To analyze the existing conditions, impacts of alternatives, and investigate operations for RM&E, the model will ideally be capable of simulating hydraulic zones of influence and temperature mixing and stratification. The model will provide hydraulic and temperature information that will be used to refine alternatives, evaluate alternatives, develop RM&E test plans, and interpret the results of RM&E tests.

Objectives The Portland District Hydraulic Design Section conducted a hydraulic modeling study of Detroit Reservoir to meet the following objectives:

· Develop and document a hydraulic model of the entire Detroit Reservoir with adequate resolution to provide hydraulic flow patterns in the just upstream of Detroit Dam and around potential fish passage structures that will be identified in the 90% Alternatives Report.

· The model should have the flexibility to conduct analyses (steady state) at a range of water surface elevations from Minimum Power Pool (1425 ft) to Maximum Conservation Pool (1563.5 ft).

· Information from the hydraulic model should provide information on forebay hydrodynamic conditions to support:

· Further development of long term facilities alternatives;

· Evaluation of long term facilities alternatives;

· Development of alternatives to be selected for further study in the Design Documentation Report (DDR);

· Interpretation of results for the DDR.

CFD Model Development Model Selection Based on the objectives described in Section 1.2, a Computational Fluid Dynamics (CFD) Model was selected for the hydraulic model due to its advantages for this application over physical modeling or hydrodynamic models with a hydrostatic pressure assumption. The Detroit Reservoir CFD Model was developed using the Star CCM+ software, CFD software developed by CD Adapco. The advantages and disadvantages of the selected CFD model are summarized below:

Advantages:

· The entire Detroit Reservoir can be included in the model domain in a CFD model while still maintaining a reasonable model size (number of grid cells and file sizes for results). The model can be divided into sub-regions for computation of hydraulic conditions in all or just a portion of the reservoir.

· Several water surface elevations can be modeled using vertically divided sub-regions to represent different water levels so that the model grid does not have to be rebuilt for runs at different water surface elevations.

· Relevant geometric features of Detroit Dam including the spillway, conduits, regulating outlets, penstock intake structure, and structural alternatives can be included in the CFD model grid.

· Model results can be queried at any location in the model domain for velocity, pressure, turbulence, or other hydraulic variables. Particles seeded into the model results can be tracked to provide comparison for flow paths for alternatives or to RM&E results.

· Alternatives (operational or structural changes) can be included in the CFD model relatively efficiently.

· The Star CCM+ software suite was chosen in an assessment of available CFD programs as the preferred CFD software for application by the Portland District because of engineer experience, software capability, software availability, and software advancements.

· CFD models can be maintained on a computer system in backup files and, if compatible with future software versions, can be used for many years as a modeling tool with little maintenance.

Limitations

· CFD model runs will require approximately 24 hours for each scenario to run in steady state configuration. It is likely that transient (or time dependent) scenarios will not be feasible for the majority of scenarios due to computational time constraints. The model boundary conditions will need to be chosen carefully for each steady state run to adequately represent conditions. The influence of transient conditions in the prototype needs to be considered when interpreting model results.

· Velocities in the Detroit Reservoir are generally well less than 0.1 ft/s, and collection of calibration/validation velocity data at these low velocities is difficult due to limitations of instrument accuracy, flow field variability, and influences of meteorological conditions. At this time, the model has been built with a model grid representing the reservoir bathymetry and structures to a reasonable level of detail and the geometric features are considered to drive the flow field. Model results are considered adequate for comparing overall flow patterns for baseline and alternative conditions. Field velocity data will be required for validation if the model is to be used for detailed design of alternatives (near-field velocities, pressures, or flow patterns).

· Meteorological conditions, such as wind/waves, are not included in the CFD model. Surface flow patterns will be interpreted relative to baseline flow patterns to determine significant changes due to operational or structural alternatives. The model will not be used to predict small-scale specifics of the surface flow field that may be influenced by meteorological conditions not modeled.

· Temperature is not included in the model at this time due to the additional complexity and computational time required to simulate temperature. If further investigation of temperature impacts/influence is required, calibration data will be needed to calibrate the temperature computation or the model results can be combined with results from a water quality model, such as CE-QUAL-W2.

· Real time viewing of results in a CFD model is limited to available computing resources.

Model Domain The Detroit reservoir CFD model domain is the entire inundated area (up to el. 1563.5 ft.) upstream of Detroit Dam. The model domain extends approximately 33,000 ft (6.2 miles) from the head of reservoir at full pool downstream to the dam.

The upstream extents of the model domain can be defined at multiple locations, as shown in. There are 4 upstream boundaries located on the Kinney, Blowout Creek, Breitenbush, and North Santiam tributaries. There is one boundary off the Cumley tributary near field to the Detroit Dam. These boundaries will represent inflow and flow splits from the tributaries.

Figure 1: Detroit CFD model domain.

The model domain extends from the dam (including spillway, test conduit intake, regulating outlets (RO) intakes, and penstock intakes) and terminates at 5 tributary locations upstream of the dam The tributaries include Cumley Creek, Kinney Creek, Blowout Creek, the N. Santiam River, and Breitenbush River.

Detroit Dam

Figure 2: Detroit CFD Forebay model extents.

Geometry As the first step in CFD model development, a solid model of the Detroit reservoir bathymetry and structures was created. All major Detroit Dam features were included in the model. Features that would require many computational cells to define, but not significantly enhance the forebay model were omitted to reduce computational cells and model run time; with the exception of the penstock intake trash structure, there were no trash screens, gate slots or guides, or other small features included in the model. The gates themselves are not represented as geometric features. The locations downstream of the dam structure where flow from the forebay exit the model domain represents outlets.

The Detroit solid model was constructed by creating then combining three separate parts (Figure 3 and 4). (reservoir bathymetry, dam structure, and reservoir lid). The three parts were then exported into STL file formats and imported into StarCCM+ for grid generation. The development of each solid piece is explained below.

· The Detroit reservoir bathymetry file was created from the USGS data obtained from their website (insert website). GIS Section converted the information to cloud point data and imported into a Microstation (.dgn) seed file. The seed file was formatted with the geographic reference system Oregon North NAD83 in units of feet. The Microstation cloud point file (DE1600Elev_30grid2.dgn) developed by GIS contained elevation geographical data points ranging from 1183’ to 1600’ on a 30’x30’ grid. Using the cloud points, a surface mesh was created. The Mesh Modeling feature and Mesh from Points subcategory in Microstation was used on the entire domain to create the bathymetric surface. Once the bathymetric mesh was created, the dam solids were imported into the Microstation file and the referenced location was translated and rotated into place matching approximately the geo-referenced coordinates. Once the dam was roughly in the right location a satellite view of Google Earth was used to verify the location of the structures relative to the North Santiam River. The solid was slightly manipulated to match the satellite photo and the archival photos DET0808.jpg and DET0833.jpg taken August 27, 1952 and found in the directory below. Once imported into StarCCM+, further manipulation was done to refine the bathymetry near the dam structure.

Z:\\nwp\ETDS\Willamette\ESA_Consultation\BiOp_Implementation\Detroit Longterm Temp Control\Reference_Material.

· The basic solid model of the Detroit dam structure was developed in Microstation then exported to SolidWorks for further modification using the STL file format. The drawings used in developing the solids include DE-12-4, DE-110-2, DE-122-1, DE-122-2, DE-122-3, DE-122-4, DE-122-5, DE-122-15, DE-122-34, DE-122-36, DE-122-37, DE-122-53, and DE-122-102. Once imported into SolidWorks the solid was modified to include the trashrack structure over the penstock intake using drawings DE-122-25, DE-122-26 and DE-122-54. The dam structure was also modified to a reference elevation and to include the geo-referencing location found in the drawing number DE-110-2. The drawing contained two coordinates lying on the dam axis at deck elevation 1579’ expressed in State Plane, NAD27 3601 Oregon North in U.S. feet. The coordinates were converted into State Plane, NAD 83 in U.S. feet to match the current geographic coordinates system for bathymetric data obtained from the USGS website. The coordinates were located on the dam axis at Station 1+10 (NAD27: Northing/Y – 388746.98, Easting/X – 1545210.76) and Station 16+94.25 (NAD27: Northing/Y – 390250.37, Easting/X – 1545708.55). The conversion to NAD83 resulted in the following coordinates: Station 1+10 (Northing/Y – 388700.55, Easting/X – 7746997.326) and Station 16+94.25 (Northing/Y – 390203.935, Easting/X – 7747495.132).

· A lid for the CFD model was created in order to “close” the fluid domain for the StarCCM+ Surface Wrapper. A surface was created in Microstation on a flat plane roughly encompassing the outline of the reservoir. The surface was thickened to an arbitrary thickness and the bottom moved to elevation of 1595 feet.

Figure 3: Dam and reservoir solid model.

Figure 4: Detroit reservoir bathymetry and lid.

Grid Development The Detroit CFD model grid was created in Star-CCM+ version 6.02 (CCM+). The development of the model grid was an iterative process that involved testing and adjusting grid development strategies; it was not a direct linear process, but essentially followed the workflow discussed below.

First, the outer surface of the model domain was split into different surfaces (bathymetry, dam, tributaries etc.) according to the boundary conditions and/or mesh conditions desired at or near each surface. Using the Surface Wrapper, Surface Mesher, Volume Mesher, and Prism Layer functions in CCM+ a polyhedral grid was developed for the Full Forebay model with the desired boundary conditions (Figure 5). More information on boundary conditions is provided in section 2.5. Test model simulations were run on the full forebay grid to ensure a fully functioning grid before moving on to the next part of the grid development process. Table 1 shows the test parameters and a solution run with good results. The results shows that total flow in into the solution domain is approximately equal to total flow out indicating that the solution grid is free of leaks or cracks in the grid. A visual QC of velocity contours of various slices in the forebay model showed that the solution grid is free of “hot spots” which if present would show high velocities in unexpected locations indicating a problem grid.

Figure 5: Full Forebay polyhedral grid at Detroit Dam before splitting into region layers.

Table 1: Full forebay test run and solution results.

The next step in the grid development was to construct to allow for eight different reservoir water surface elevations: 1563.5 feet, 1552 feet, 1541 feet, 1510 feet, 1480 feet, 1450 feet, 1425 and 1370 feet. The model will be set up to allow different parts elevations of the model domain to be excluded from the solution domain effectively altering the forebay water surface elevation. Then, to increase the flexibility of the model, the single model region (the entire model domain) will be split into multiple regions (Figure 6 - 9). These regions, like eight eight-layer cakes placed side by side, allow the model to be set to run with water surfaces at any of eight elevations (1563.5 feet, 1552 feet, 1510 feet, 1480 feet, 1450 feet, 1425 feet and 1370 feet), and also allow for the upstream boundaries to be set at the tributary locations. The tributary sections not included in the model domain at this time can be included back into the model domain if needed in the future. The multiple regions can also be simplified to create alternative configuration runs; one region (one layer of one layer cake) can be altered or replaced without affecting the rest of the model domain.

The model grid will contain roughly 4.2 million cells consisting of a core of polyhedral cells with a layer of prism cells along solid surfaces at the dam structure. Prism layer cells will be created on the dam structure hard surfaces and also on structures used in alternative runs. Prism cells on the dam structure outlet faces will not be created (allowing them to be set as either outlets or walls), on the bathymetry, or at region interfaces.

Figure 6: Region boundaries (layers).

Region boundaries

Figure 7: Region boundaries (layers).

Figure 8: Region boundaries (Layers) at N. Santiam inlet.

Figure 9: Bottom of bathymetry and region boundary layer contours.

Boundary Conditions The details of the Detroit CFD model upstream, downstream, and other boundaries are described in the sections below.

Upstream Boundaries The upstream boundaries can be set at the different locations shown in by turning off the physics in regions upstream of the desired upstream boundary, deleting interfaces at the upstream boundary and setting appropriate boundary conditions. The top of Breitenbush upstream boundary will be set as a pressure boundary (to not over constrain the model) for all runs; all other upstream boundaries will be set as velocity inlets. Note that all regions (tributaries) upstream of the upstream boundaries will have physics set to “none” to exclude them from the solution domain, and will not have interfaces connecting them to the solution domain. An example of this concept is shown in Figure 11 which shows the confluence of Blowout Creek and the N. Santiam River. The upstream boundary is highlighted in yellow. The section of the tributary upstream of the boundary will not be included in the solution domain.

Upstream boundary locations

Figure10: Upstream boundary locations.

Figure 11: Upstream boundary at Blowout Creek.

Downstream Boundaries At Detroit Dam all RO intakes, the penstock intake, test flume intake, and spillway gates can all be set as downstream outlets. When downstream boundaries are not active outlets, they are set as walls.

Figure 3: Downstream boundaries used as outlets in the CFD baseline and alternatives runs.RO Intakes Test Conduit Intake Intakes Spillway Gates Penstock Intakes

Figure 8: Various layers for Detroit Dam.

Other Boundaries Boundaries at the desired water surface are set as symmetry boundaries. Note that all regions above the desired water surface boundaries have physics set to “none”, to exclude them from the solution domain, and do not have interfaces connecting them to the solution domain. All other region boundaries are set as default wall boundaries, including the bathymetry, Detroit Dam, and inactive flow outlets.

CFD Model Runs Baseline Conditions CFD Model Runs Baseline conditions CFD model simulations will be conducted to evaluate hydraulic conditions in the forebay upstream of Detroit Dam for existing flows and geometry for three different pool elevations:

· 1563.5 (Maximum Conservation Pool),

· 1552 (Spillway Operation), and

· 1450 (Minimum Conservation Pool).

The boundary conditions for these baseline runs can be obtained from the baseline run of a HEC-ResSim model that simulates years of project outflow using current dam operations. The project outflow for each water surface elevation simulated can be estimated from HEC-ResSim by extracting the Detroit outflow for the given water surface elevation, plus or minus 1 ft. The 50% exceedance outflow for the given water surface elevation range was used as the project outflow as shown in Table 1.

For each CFD model run chosen, the simulation run about 3000 iterations and results will be post-processed using Tecplot. Figures of run results will be in the end of the CFD report. The CFD report will be included as an appendix to this report. These figures will show colored velocity magnitude counts for 5 feet below each modeled water surface elevation. Stream traces will show paths particles take from evenly spaced location in the solution regions in front of Detroit Dam to the model outlets. Additional contour plots will show velocities on two vertical surfaces roughly centered on the outlets.

These figures will show the WTC tower cul-de-sac west of the region 1/ region 2 boundary, looking down from the southeast. Surfaces showing colored velocity magnitude contours are set 5 feet below each modeled water surface elevation. Blue stream traces show paths particles take from evenly-spaced locations on the region 1/ region 2 boundary, through the WTC tower cul-de-sac, to model outlets. For all 1571 feet pool elevation runs, there are also velocity contour plots showing velocities on two vertical surfaces roughly centered in the right and middle WTC tower weir gates perpendicular to the face of the tower.

At pool elevation 1690 feet (Figure 7), velocities in the WTC tower cul-de-sac range from less than 0.02 ft/s to 4.6 ft/s, with velocities greater than 0.1 ft/s concentrated around the tower weir gate outlets. Flow enters the tower cul-de-sac uniformly then becomes complicated. Flow that enters along the right bank goes more or less directly to the tower. Most of the flow that enters at lower elevations and along the left bank crosses in front of the tower across the tower cul-de-sac to the bank left of the tower where it flows up and spirals to the tower weir gates, some circles into the south end of tower cul-de-sac and takes a very indirect route to the tower.

At pool elevation 1571 feet, the velocities for the 400 cfs outflow run (Figure 8, Figure 9, and Figure 10) range from less than 0.02 ft/s to 3.7 ft/s, with velocities greater than 0.1 ft/s concentrated around the WTC tower weir gates. Approximately 0.05 ft/s flow enters the tower cul-de-sac and makes a smooth arc through the center of the cul-de-sac to the tower; there are slow-moving eddies on both sides of the arc. The vertical velocity contour plots show a swatch of approximately 0.05 ft/s from the tower cul-de-sac floor up to the WTC tower weir gates. The velocities for the 1180 cfs outflow run (Figure 11, Figure 12, and Figure 13) are higher than those in the 400 cfs outflow run, ranging from less than 0.02 ft/s to 4.9 ft/s, but the overall results look very similar. Flow enters the tower cul-de-sac and makes a smooth arc to the tower where higher velocity flows are located, with eddies to either side of the arc.

At pool elevation 1532 feet (Figure 14, and Figure 15), velocities range from less than 0.02 ft/s to 0.8 ft/s at the WTC tower RO and Penstock bypass gates. A stream of 0.25 ft/s flow enters the cul-de-sac and slows to about 0.15 ft/s as it makes a 90 degree arc to the tower. There is a more pronounced eddy circling back along the right bank than was seen in the runs at pool elevation 1571 feet.

Alternative Conditions CFD Model Runs CFD model simulations were run for two different long term facilities alternatives identified in the 60% Detroit Long term facilities Alternatives Study Report (USACE, 2010): floating screen structure, and floating surface collector. The model grid changes, boundary conditions, and results for the alternative runs are described in the following sections. All alternative runs were run for 3000 iterations and results were post-processed in FieldView.

Floating Screen Structure Alternative CFD Model Runs Floating screen structure (FSS) alternative conditions CFD model runs were run for the same pool elevations as the existing conditions model runs: 1690 feet, 1571 feet, and 1532 feet. The baseline CFD model was modified to include an FSS solid (approximately 160 feet long, 50 feet wide, 16 feet deep structure, with a 30-foot wide, 15-foot tall entrance, and a 1000 cfs capacity) in front of the WTC tower and a 52.5-foot long, 30-foot wide, 4-foot deep holding barge beside it (Figure 4). Boundary conditions (Table 1) were set based on baseline HEC-ResSim model runs, and the computational grid. FSS run results are described below.

At pool elevation 1690 feet (Figure 16), velocities in the WTC tower cul-de-sac range from less than 0.02 ft/s to 2.2 ft/s, with velocities greater than 0.1 ft/s concentrated in the middle of the tower cul-de-sac around the FSS entrance; the concentration looks very similar to what was seen in the baseline runs, only in the center of the cul-de-sac and not at the tower. Some flow entering the tower cul-de-sac near the water surface goes directly to the FSS inlet, most flow either enters a spiral below the intake that spirals up toward the FSS or crosses below the FSS where it enters spirals in the south end of the tower cul-de-sac.

At pool elevation 1571, the velocities for the 400 cfs run (Figure 17, Figure 18, and Figure 19) range from less than 0.02 ft/s to 0.9 ft/s, with velocities greater than 0.1 ft/s concentrated at the FSS entrance. Nearly all flow in the WTC tower cul-de-sac is affected by a large, clockwise-spinning eddy (velocities range from about 0.03 to 0.1 ft/s) at the southeast corner of the FSS before entering the FSS. The vertical velocity contour plots show slightly higher velocities than the baseline runs centered in the middle of the cul-de-sac, not at the tower. The velocities for the 1180 cfs outflow run (Figure 20, Figure 21, and Figure 22) range from less than 0.06 ft/s in much of the tower cul-de-sac to 1.1 ft/s at the penstock bypass gate and 2.2 ft/s at the FSS entrance, with velocities greater than 0.3 ft/s concentrated at the entrances to the FSS and penstock bypass gate. The flow pattern is similar to what was seen at 400 cfs outflow, with generally higher velocities, and higher velocities at the active penstock bypass.

At pool elevation 1532 feet (Figure 23, and Figure 24), velocities range from less than 0.02 ft/s to 1.2 ft/s, with velocities greater than 0.03 ft/s concentrated at the FSS entrance. Flow enters a large (0.15 ft/s) eddy as it enters the WTC tower cul-de-sac; the approximately 150-foot diameter eddy dominates the flow pattern in the tower cul-de-sac.

Figure 4: WTC tower cul-de-sac with FSS and Holding Barge structures (green, red inlet) at El. 1690 ft.

The results of the initial FSS alternative runs showed that the zone of influence of the FSS is centered in the WTC tower cul-de-sac, away from the shorelines and the WTC tower. The effectiveness of a structure in attracting fish at this location is unknown, Therefore, a smaller capacity FSS that could provide a zone of influence closer to the shorelines and WTC tower was modeled to provide a range of flows, structure sizes, and location for the FSS alternative. Additional runs using a smaller, 200 cfs capacity, FSS structure (Figure 5) were set up and run at the same pool elevations (1690 feet, 1571 feet, and 1532 feet). This smaller structure was modeled by placing the solid that was created for the Floating Surface Collector (see section 3.2.2) in front of the WTC tower and setting the structure outlets as walls. Boundary conditions for these runs (Table 1) were based on baseline HEC-ResSim model runs. The holding barge was not used in these runs. Results from the 200 cfs capacity FSS runs are described below.

At pool elevation 1690 feet (Figure 25), velocities in the WTC tower cul-de-sac range from less than 0.02 ft/s to 3.3 ft/s at the FSS entrance and 5.2 ft/s at the WTC tower middle weir gate, with velocities greater than 0.1 ft/s concentrated around the FSS entrance and the tower. Flow enters the tower cul-de-sac uniformly then the majority of the flow rises up and passes underneath the FSS structure on a fairly direct path to the tower weir gates, some flow crosses in front of the tower rises as it approaches the west bank of the cul-de-sac and spirals up to the tower weir gates. The flow that enters the FSS comes mostly from the south and west parts of the cul-de-sac, not directly from the cul-de-sac entrance.

At pool elevation 1571 feet (Figure 26, Figure 27, and Figure 28), velocities for the 400 cfs outflow run range from less than 0.02 ft/s to 3.3 ft/s at the FSS entrance and 4.4 ft/s at the WTC tower middle gate. The results for this run look very much like the baseline run results with slightly higher (~0.1 ft/s) velocities in the eddy to the right of the smooth 0.05 ft/s arc. No 1180 cfs outflow at pool elevation 1571 feet was modeled for this alternative.

At pool elevation 1532 feet (Figure 29, and Figure 30), cul-de-sac velocities range from less than 0.02 ft/s to 3.3 ft/s at the FSS entrance and 0.7 ft/s at the two RO bypass gates, with velocities greater than 0.3 ft/s concentrated at the FSS entrance and RO bypass gates. The flow pattern in the WTC tower cul-de-sac is nearly identical to the baseline run with exception of the area within about 50 ft of the FSS and tower, where flow is influenced by the FSS inlet.

Figure 5: WTC tower cul-de-sac with smaller (200 cfs) FSS structure (green with red inlet) at El. 1690 ft.

Floating Surface Collector Alternative CFD Model Runs Floating surface collector (FSC) alternative conditions CFD model runs were run for the same pool elevations as the existing conditions model runs: 1690 feet, 1571 feet, and 1532 feet. These runs were set up by putting an FSC solid (90 feet long, 40 feet wide, 14 feet deep, with a 10-foot wide, 6-foot tall rectangular entrance, and four 2-foot diameter and two 1-foot diameter structure outlets) and a surface representing nets in the WTC tower cul-de-sac (Figure 6), setting boundary conditions (Table 1) based on baseline HEC-ResSim model runs, and remeshing the computational grid. Results from the FSC alternative runs are described below.

At pool elevation 1690 feet (Figure 31, and Figure 40), velocities in the WTC tower cul-de-sac range from less than 0.06 ft/s to 3.7 ft/s at the FSC inlet, 15.8 ft/s at the FSC outlet portals, and 4.6 ft/s at the tower weir gates, with velocities greater than 2.0 ft/s concentrated around the WTC tower weir gates and FSC inlet and extending out from the FSC outlet portals about 100 feet. Net velocities are less than 0.1 ft/s except at the FSC entrance (~0.2 ft/s) and where the net edge meets the dam (above elevation 1610 feet). Flow in and around the WTC tower cul-de-sac is dominated by the outflow from the FSC structure, but with a full depth net, fish would be excluded from this area. The outflow from the right side of the FSC continues near the surface, hits the bank east of the tower (at about 1.0 ft/s) where some splits to the west and snakes around into the tower and some splits to the east and continues upstream through the net before circling down and flowing back through the net (moving downstream). The outflow from the left side of the FSC continues south until it hits the south end of the tower cul-de-sac (at approximately 0.8 ft/s) follows the bank to the west and around to the north. Upstream of the net, the attraction flow to the FSC flows to the FSC inlet, with the remaining flow passing through the net relatively uniformly. There is a large, slow moving eddy in the south end of the cul-de-sac and another behind the FSC.

At pool elevation 1571 feet, the velocities for the 400 cfs outflow run (Figure 32, Figure 33, Figure 34, and Figure 41) range from less than 0.06 ft/s to 3.7 ft/s at the FSC inlet, 15.8 ft/s at the FSC outlet portals, and 3.7 ft/s at the tower weir gates, with velocities greater than 2.0 ft/s concentrated at the tower weir gates and FSC inlet and extending out from the FSC outlet portals about 100 feet. Upstream of the net, the attraction flow to the FSC flows to the FSC inlet, with the remaining flow passing through the net relatively uniformly. Net velocities are less than 0.15 ft/s except at the FSC entrance and where the upper edges of the net meets the shoreline. Flow in and near the tower cul-de-sac is dominated by the outflow from the FSC structure, but with a full-depth net fish would be excluded from this area. Most flow passes through the net under the FSC, rises to the FSC outlet flow and is pushed to the north or south bank and circles around at the bank. Some flow moves upstream through the net along the north shore (right bank). The velocities for the 1180 cfs outflow run (Figure 35, Figure 36, Figure 37, and Figure 42) range from less than 0.06 ft/s to 3.7 ft/s at the FSC inlet, 15.8 ft/s at the FSC outlet portals, and 4.9 ft/s at the tower weir gates, with a flow pattern similar to what was seen at 400 cfs outflow and slightly higher velocities at the tower and through the middle of the net.

At pool elevation 1532 feet, the velocities for the 530 cfs outflow run (Figure 38, and Figure 43) range from less than 0.06 ft/s to 3.7 ft/s at the FSC inlet, 15.8 ft/s at the FSC outlet portals, and 0.8 ft/s at the WTC tower RO and penstock bypass gates, with velocities greater than 2.0 ft/s concentrated around the FSC entrance and extending out from the FSC outlet portals about 100 feet. Net velocities are less than 0.25 ft/s except at the FSC entrance and where the upper edges of the net meets the shoreline (from flow moving upstream through the net). Flow conditions look very similar to what was seen at pool elevation 1571 feet. The velocities for the 2200 cfs outflow run (Figure 39, and Figure 48) range from less than 0.06 ft/s to 3.7 ft/s at the FSC inlet, 15.8 ft/s at the FSC outlet portals, and 3.3 ft/s at the tower bypass gates, with velocities greater than 2.0 ft/s concentrated around the tower bypass gates and the FSC inlet, and extending out from the FSC outlet portals about 100feet. Net velocities are higher than in the 530 cfs outflow run, approximately 0.2 ft/s in the middle of the net with velocities greater than 0.3 ft/s at the FSC inlet and along the bottom of the net. The cul-de-sac flow pattern is similar to what was seen in the 530 cfs outflow run with exception of the clockwise eddy upstream of the net on the right bank and flow along the west cul-de-sac bank going more directly to the RO bypass gates.

Outlet Portals (on both sides)

Figure 6: WTC tower cul-de-sac with FSC structure (green with red inlet) at El. 1690 ft and nets (blue).

CFD Model Runs for RM&E Support Two runs were conducted to support RM&E efforts and analyze weir operations that may influence the weir box alternative from the 60% Alternatives Report, as shown in Table 1. The runs were conducted with pool elevation 1650 and baseline outflow of 600 cfs, but with two different WTC weir configurations to investigate the changes in the zone of influence near the WTC tower. Figures 49 and 51 show the plan view forebay velocity magnitude near the surface for each run, and indicate that the change in the zone of influence with the change in weir operation is limited to the area very near the WTC weirs. The changes in velocity near the WTC weirs are further demonstrated in the isosurface plots in Figures 50 and 52. These figures show regions with velocities of a common value for comparison. The velocity at the WTC weir is prescribed by the boundary condition, and is higher (5.6 ft/s) with a 4 foot weir submergence than with a 10 foot submergence (2.2 ft/s) for the same outflow. Because the flow approaching the weirs converges from a large forebay volume, including the region below the weirs, the velocity near the piers is approximately the same in either case. The flow accelerates more quickly to the overflow section in the case with 4 ft submergence. RM&E tests may indicate whether changes to the weirs can be used to optimize alternatives like the weir box for collection inside the tower or passage through operational alternatives. On the other hand, operations that deter fish from entering the tower may be used to encourage fish to use a surface collector upstream of the WTC tower, while passing flow exceeding a collector’s capacity through the WTC tower with minimal numbers of fish.

Table 1: Detroit CFD Modeling Plan

Figure 7: Baseline Run, Pool El. 1690 feet, Total Outflow 1000 cfs

Figure 8: Baseline Run, Pool El. 1571 feet, Total Outflow 400 cfs

Figure 9: Baseline Run, Pool El. 1571 feet, Total Outflow 400 cfs, Vertical Slice Through Right Weir Gate and Penstock Bypass

Figure 10: Baseline Run, Pool El. 1571 feet, Total Outflow 400 cfs, Vertical Slice Through Middle Weir Gate and RO Bypass

Figure 11: Baseline Run, Pool El. 1571 feet, Total Outflow 1180 cfs

Figure 12: Baseline Run, Pool El. 1571 feet, Total Outflow 1180 cfs, Vertical Slice Through Right Weir Gate and Penstock Bypass

Figure 13: Baseline Run, Pool El. 1571 feet, Total Outflow 1180 cfs, Vertical Slice Through Middle Weir Gate and RO Bypass

Figure 14: Baseline Run, Pool El. 1532 feet, Total Outflow 530 cfs

Figure 15: Baseline Run, Pool El. 1532 feet, Total Outflow 530 cfs (note different scale)

Figure 16: FSS Alternative Run, Pool El. 1690 feet, Total Outflow 1000 cfs

Figure 17: FSS Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs

Figure 18: FSS Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs, Vertical Slice Through Right Weir Gate and Penstock Bypass

Figure 19: FSS Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs, Vertical Slice Through Middle Weir Gate and RO Bypass

Figure 20: FSS Alternative Run, Pool El. 1571 feet, Total Outflow 1180 cfs

Figure 21: FSS Alternative Run, Pool El. 1571 feet, Total Outflow 1180 cfs, Vertical Slice Through Right Weir Gate and Penstock Bypass

Figure 22: FSS Alternative Run, Pool El. 1571 feet, Total Outflow 1180 cfs, Vertical Slice Through Middle Weir Gate and RO Bypass

Figure 23: FSS Alternative Run, Pool El. 1532 feet, Total Outflow 530 cfs

Figure 24: FSS Alternative Run, Pool El. 1532 feet, Total Outflow 530 cfs (note different scale)

Figure 25: FSS (200 cfs) Alternative Run, Pool El. 1690 feet, Total Outflow 1000 cfs

Figure 26: FSS (200cfs) Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs

Figure 27: FSS (200 cfs) Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs, Vertical Slice Through Right Weir Gate and Penstock Bypass

Figure 28: FSS (200 cfs) Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs, Vertical Slice Through Middle Weir Gate and RO Bypass

Figure 29: FSS (200 cfs) Alternative Run, Pool El. 1532 feet, Total Outflow 530 cfs

Figure 30: FSS (200 cfs) Alternative Run, Pool El. 1532 feet, Total Outflow 530 cfs (note different scale)

Figure 31: FSC Alternative Run, Pool El. 1690 feet, Total Outflow 1000 cfs (note scale)

Figure 32: FSC Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs (note scale)

Figure 33: FSC Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs, Vertical Slice Through Right Weir Gate and Penstock Bypass

Figure 34: FSC Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs, Vertical Slice Through Middle Weir Gate and RO Bypass

Figure 35: FSC Alternative Run, Pool El. 1571 feet, Total Outflow 1180 cfs (note scale)

Figure 36: FSC Alternative Run, Pool El. 1571 feet, Total Outflow 1180 cfs, Vertical Slice Through Right Weir Gate and Penstock Bypass

Figure 37: FSC Alternative Run, Pool El. 1571 feet, Total Outflow 1180 cfs, Vertical Slice Through Middle Weir Gate and RO Bypass

Figure 38: FSC Alternative Run, Pool El. 1532 feet, Total Outflow 530 cfs (note scale)

Figure 39: FSC Alternative Run, Pool El. 1532 feet, Total Outflow 2200 cfs (note scale)

Figure 40: FSC Alternative Run, Pool El. 1690 feet, Total Outflow 1000 cfs (note scale)

Figure 41: FSC Alternative Run, Pool El. 1571 feet, Total Outflow 400 cfs (note scale)

Figure 42: FSC Alternative Run, Pool El. 1571 feet, Total Outflow 1180 cfs (note scale)

Figure 43: FSC Alternative Run, Pool El. 1532 feet, Total Outflow 530 cfs (note scale)

Figure 44: FSC Alternative Run, Pool El. 1532 feet, Total Outflow 2200 cfs (note scale)

Figure 45: FSC Alternative Run, Pool El. 1532 feet, Total Outflow 2200 cfs (note scale) Figure 46: FSC Alternative Run, Pool El. 1532 feet, Total Outflow 2200 cfs (note scale)

Figure 47: FSC Alternative Run, Pool El. 1532 feet, Total Outflow 2200 cfs (note scale)

Figure 48: FSC Alternative Run, Pool El. 1532 feet, Total Outflow 2200 cfs (note scale) image41.jpeg image42.jpeg image43.jpeg image44.jpeg image45.jpeg image46.jpeg image47.jpeg image48.jpeg image49.jpeg image50.jpeg image51.jpeg image52.jpeg image53.jpeg image54.jpeg image55.jpeg image56.jpeg image57.jpeg image58.jpeg image59.jpeg image60.jpeg image61.jpeg image62.jpeg image2.png image3.png image4.jpeg image5.png image6.jpeg image7.jpeg image8.png image9.png image10.png image11.png image12.png image13.emf image14.png image15.png image16.jpeg image17.jpeg image18.jpeg image19.jpeg image20.emf image1.png image21.jpeg image22.jpeg image23.jpeg image24.jpeg image25.jpeg image26.jpeg image27.jpeg image28.jpeg image29.jpeg image30.jpeg image31.jpeg image32.jpeg image33.jpeg image34.jpeg image35.jpeg image36.jpeg image37.jpeg image38.jpeg image39.jpeg image40.jpeg