Hells Canyon Complex Total Dissolved Gas Study
Ralph Myers Project Limnologist
Sharon E. Parkinson Principal Engineer
Technical Report Appendix E.2.2-4 March 2002 Revised July 2003 Hells Canyon Complex FERC No. 1971 Copyright © 2003 by Idaho Power Company
Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
TABLE OF CONTENTS
Table of Contents ...... i
List of Tables...... ii
List of Figures ...... ii
List of Appendices ...... iii
Abstract ...... 1
1. Introduction ...... 2
2. Study Area...... 3
3. Plant Operations ...... 4
4. Methods...... 5
5. Results and Discussion...... 6
5.1. TDG and Project Operations...... 6
5.1.1. Measured TDG Data ...... 6
5.1.2. Predictive Numerical Model Development...... 8
5.2. Spill Manipulation Test...... 10
5.2.1. Brownlee Dam...... 10
5.2.2. Hells Canyon Dam ...... 11
5.3. Downstream Dissipation of TDG...... 11
5.4. Physical Modeling of Hells Canyon Dam...... 12
6. Summary and Conclusions...... 13
7. Acknowledgments...... 14
8. Literature Cited ...... 14
Hells Canyon Complex Page i Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
LIST OF TABLES
Table 1. Physical characteristics of Brownlee, Oxbow, and Hells Canyon powerhouses and dams...... 17
Table 2. Equations used to describe spill vs. total dissolved gas relationships at the projects of the Hells Canyon Complex...... 17
LIST OF FIGURES
Figure 1. Total dissolved gas study area...... 19
Figure 2. Brownlee Dam study area showing reservoir, turbine discharge, and spillway sampling sites...... 21
Figure 3. Oxbow Dam study area showing reservoir, turbine discharge, and spillway sampling sites...... 22
Figure 4. Hells Canyon Dam study area showing reservoir and tailwater sampling sites...... 23
Figure 5. Mean daily Snake River flow (cfs) measured downstream of Hells Canyon Dam from October 1996 through August 1999...... 24
Figure 6. Total dissolved gas levels in water at six locations: (1) Brownlee Reservoir, (2) Brownlee Project discharge (spill and turbine), (3) Oxbow Reservoir, (4) Oxbow Project discharge (spill and turbine), (5) Hells Canyon Reservoir, and (6) Hells Canyon Project discharge on 8 dates in 1997...... 25
Figure 7. Total dissolved gas levels in water at six locations: (1) Brownlee Reservoir, (2) Brownlee Project discharge (spill and turbine), (3) Oxbow Reservoir, (4) Oxbow Project discharge (spill and turbine), (5) Hells Canyon Reservoir, and (6) Hells Canyon Project discharge on 8 dates in 1998...... 26
Figure 8. The relationship of spill and total dissolved gas measured during spill at Brownlee Dam, 1997−1998...... 27
Figure 9. The relationship of spill and total dissolved gas measured downstream of Oxbow Dam during spill, 1997–1998...... 27
Page ii Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
Figure 10. Change in percentage of total dissolved gas saturation between Oxbow Reservoir and the Oxbow spillway...... 28
Figure 11. Percentage of TDG saturation measured at the Hells Canyon boat ramp downstream of Hells Canyon Dam, 1997–1988...... 29
Figure 12. The relationship of spill and total dissolved gas measured during spill at Hells Canyon Dam, 1997–1999...... 30
Figure 13. Total dissolved gas levels measured while releasing 39,000 cfs of spill from Brownlee Dam on June 4, 1998, and 28,000 cfs of spill at Hells Canyon Dam on June 3, 1998...... 31
Figure 14. Downstream dissipation of total dissolved gas within Hells Canyon relative to the 110% saturation standard...... 32
Figure 15. General sectional view of the Hells Canyon Dam spillway model constructed by researchers at the Iowa Institute of Hydraulic Research (IIHR)...... 33
Figure 16. Location of crest gates, nappe deflectors, and sluice gates on the Hells Canyon Dam spillway...... 34
Figure 17. General view of the modified deflector configuration for the Hells Canyon Dam spillway developed using the two-dimensional model...... 35
Figure 18. Deflector performance from the evaluation of the IIHR two-dimensional physical model...... 36
LIST OF APPENDICES
Appendix 1. Hells Canyon Flow Deflector Design Report...... 37
Hells Canyon Complex Page iii Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
This page left blank intentionally.
Page iv Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
ABSTRACT
As is the case with most hydroelectric projects, spilling water at projects in the Hells Canyon Complex causes total dissolved gas (TDG) levels to exceed the 110% saturation limit for protecting aquatic biota. Levels in the tailwater of Hells Canyon Dam have been measured as high as 135% saturation, with resulting in-river levels in excess of 110% saturation for up to 60 miles downstream. Measured levels in the tailwater of Brownlee Dam approach 125% saturation, with little dissipation downstream through Oxbow and Hells Canyon reservoirs. A predictive numerical modeling tool was developed as a part of this study to provide TDG boundary condition information to water quality and habitat models for resource impact evaluation. Based on the results of spill tests, spilling water from the upper spill gates at Brownlee Dam helps minimize TDG levels in the tailwater. Also, physical modeling of Hells Canyon Dam shows that flow deflectors installed on the dam’s spillway may reduce TDG levels at total spill flows of less than 30,000 cubic feet per second. Although flow deflectors show promise for reducing elevated TDG levels at a qualitative modeling level, it is difficult to predict what levels of TDG will occur below the project under all spill conditions with the deflectors in place.
Hells Canyon Complex Page 1 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
1. INTRODUCTION
Total dissolved gas (TDG) is a measure of the sum of partial pressures of all dissolved gases, including water vapor. Typically, in most natural waters, TDG is a measure of how much nitrogen, oxygen, argon, carbon dioxide, and water vapor is dissolved in a given amount of water. A TDG level of 100% means that the water is saturated relative to atmospheric conditions and can hold no more dissolved gas. Levels exceeding 100% saturation (called supersaturation) can be detrimental, or even lethal, to aquatic biota. The states of Idaho, Oregon, and Washington have set a limit of 110% TDG saturation to protect aquatic resources, particularly juvenile salmonids. The study described in this report addressed supersaturation within and downstream of the Hells Canyon Complex (HCC).
The effects of hydroelectric operations on TDG levels are well documented (Weitkamp and Katz 1980). Although slightly elevated levels of TDG can occur naturally in rivers, levels in excess of the 110% supersaturation standard commonly occur below large-scale hydroelectric facilities along the Columbia and Snake rivers. Gas supersaturation downstream of a dam typically occurs when air becomes entrained in water that is released over a spillway and plunges deep into a stilling basin. The hydrostatic pressure at depth causes entrained atmospheric gases to be absorbed into solution. This process creates supersaturation of gases relative to surface or atmospheric pressures. Rivers with extensive hydroelectric development can become supersaturated during periods of high runoff when inflowing water exceeds the hydraulic capacity of these projects and must be released over the spillways.
The solubility of atmospheric gases in water is primarily affected by temperature and pressure. Increases in pressure on a liquid enhance its capacity to hold dissolved gases. Therefore, the pressure at depth, caused by greater hydrostatic head, allows deeper water to hold more dissolved gas than shallow water (Weitkamp and Katz 1980). Consequently, the effects of supersaturation on aquatic organisms depend on the depth distribution of the organisms. Each meter of depth increases the solubility of the dissolved gases to compensate for approximately 10% of the supersaturation. For example, a surface reading of 120% corresponds to a compensated TDG of 110%, just 1 m below the surface. Therefore, in large rivers with elevated TDG levels, most of the water volume is unlikely to be supersaturated (Weitkamp 1974). Excessive TDG levels relative to the surface represent a greater threat to organisms in shallow water than in deeper water.
Fish experience gas bubble trauma when the sum of dissolved gas pressures in their body fluids exceeds the compensating pressures of hydrostatic head, blood, tissue, and water surface tension. Gas bubbles form in the blood and tissues of fish, creating physiological dysfunction (Bouck 1980). Common external signs of gas bubble trauma include bubbles or blisters under the skin of fin rays, the head, the lining of the mouth, and along the lateral line. Bubbles and blisters often cause lesions and hemorrhaging. Exophthalmia, a condition commonly known as popeye, may occur. However, this problem is relatively rare and usually requires long-term exposure to uncompensated gas pressure. If a fish is exposed long enough to uncompensated gas pressure, gas bubble trauma can lead to its death. In general, death is caused by anoxia resulting from stasis of the blood or by secondary infection of lesions. Invertebrates are also susceptible to gas
Page 2 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study bubble trauma. In fact, long-term exposure to uncompensated gas pressure can be as lethal to invertebrates as it is to fish (Nebeker et al. 1976).
The severity of the effects of gas bubble trauma on fish varies among life stages. In early life stages, tolerance to supersaturation decreases from high tolerance in the egg stage to low tolerance in the juvenile stage (Weitkamp and Katz 1980). Tolerance then increases with life stages following the juvenile stage. Adulthood appears to be the most tolerant free-swimming life stage.
Tolerance to supersaturation might be increased by intermittent exposure to elevated TDG levels. Beyer et al. (1976) found that fish can withstand high levels of supersaturation for short periods of time, as long as they are able to equilibrate later. However, tissues generally require more time to desaturate than they do to saturate. It is generally believed that fish cannot detect, and therefore actively avoid, water that is supersaturated (Weitkamp and Katz 1980). However, there is some evidence to suggest that chinook salmon (Oncorhynchus mykiss) respond both laterally and vertically to elevated TDG levels (Blahm et al. 1976, Dawley et al. 1976, Stevens et al. 1980).
In the 1970s, Parametrix (1974) identified elevated TDG levels associated with spill at Hells Canyon Dam. While TDG levels increased during spill episodes, Parametrix (1974) concluded that there were no major differences in TDG levels regardless of which spill gates were used. However, the effect of spill TDG levels was identified as an issue that needed to be studied during efforts to renew the Federal Energy Regulatory Commission (FERC) license for the complex (FERC No. 1971). The Aquatic Resources Work Group set a goal of maintaining TDG levels below 110% saturation. With input from the Aquatic Resources Work Group, Idaho Power Company (IPC) developed a TDG study with the following objectives:
1. Define the relationship between TDG levels and HCC project operations.
2. Develop measures to predict TDG levels under a full range of operational scenarios.
3. Define the dissipation of TDGs downstream of Hells Canyon Dam.
With the study completed, IPC can use the results for two purposes. The first purpose is to determine the need for project mitigation and enhancement measures to minimize the occurrence of elevated TDG levels. The second purpose is to identify and evaluate operational measures that might minimize supersaturation.
2. STUDY AREA
The HCC is located on the Snake River, approximately 100 miles (mi) upstream of Lewiston, Idaho. Our study area included the Snake River from Brownlee Dam (river mile [RM] 284.5) downstream to RM 142, approximately 3 mi upstream of Lewiston (Figure 1). Most of our data collection effort focused on the tailwater areas of the three dams (Figures 2, 3, and 4)
Hells Canyon Complex Page 3 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company and approximately 60 mi of unimpounded Snake River from a point downstream of Hells Canyon Dam to the confluence with the Salmon River.
3. PLANT OPERATIONS
Hells Canyon, on the Oregon–Idaho border, is the deepest canyon in North America and home to IPC’s largest hydroelectric generating complex, the HCC. The HCC includes the Brownlee, Oxbow, and Hells Canyon dams, reservoirs, and power plants. Operations of the three projects of the complex are closely coordinated to generate electricity and serve many other public purposes.
Currently, over 400,000 customers rely on IPC’s hydro and thermal generation system for power. The HCC is an integral part of IPC’s generation system. Its winter and summer operations are particularly important because energy needs are highest during those seasons. In wintertime, customers need extra electricity for lighting and heating. During the summer, they need extra electricity for air conditioning and irrigation pumping.
IPC operates the complex to comply with the FERC license, as well as to accommodate other concerns, such as recreational use, environmental conditions, and voluntary arrangements. Among these arrangements are the 1980 Hells Canyon Settlement Agreement, the Fall Chinook Recovery Plan adopted in 1991, and, between 1995 and 2001, the cooperative arrangement that IPC had with federal interests in implementing portions of the Federal Columbia River Power System (FCRPS) biological opinion flow augmentation, which is intended to avoid jeopardy of the FCRPS operations below the HCC.
Brownlee Reservoir is the only one of the three HCC facilities—and IPC’s only project—with significant storage. It has 101 vertical feet of active storage capacity, which equals approximately 1 million acre-feet of water. On the other hand, Oxbow and Hells Canyon reservoirs have significantly smaller active storage capacities—approximately 0.5 and 1.0% of Brownlee Reservoir’s active volume, respectively.
Brownlee Dam’s hydraulic capacity is also the largest of the three projects (Table 1). Its powerhouse capacity is approximately 35,000 cubic feet per second (cfs), while the Oxbow and Hells Canyon powerhouses have hydraulic capacities of 28,000 and 30,500 cfs, respectively.
Target elevations for Brownlee Reservoir define the flow of water through the HCC. However, when flows exceed powerhouse capacity for any of the projects, water is released over the spillways at those projects. When flows through the HCC are below hydraulic capacity, all three projects operate closely together to re-regulate flows through the Oxbow and Hells Canyon projects so that they remain within the 1-foot per hour ramp rate requirement (measured at Johnson Bar below Hells Canyon Dam) and meet the daily peak load demands.
In addition to maintaining the ramp rate, IPC maintains minimum flow rates in the Snake River downstream of Hells Canyon Dam. These minimum flow rates are for navigation purposes and IPC’s compliance with Article 43 of the existing license. Neither the Brownlee Project nor the Oxbow Project has a minimum flow requirement below its powerhouse. However, because of the
Page 4 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
Oxbow Project’s unique configuration, a flow of 100 cfs is maintained through the bypassed reach of the Snake River below the dam (a segment called the Oxbow Bypass).
As mentioned above, when the hydraulic capacity of any of the projects within the HCC is exceeded, water is released over the spillway. Due to Oxbow’s unique configuration, spill over Oxbow Dam enters the Oxbow bypass. This release typically occurs between December and July when inflow to the projects is high. Other unusual situations, including emergencies or unexpected unit outages, can induce a spill episode at any of the projects.
4. METHODS
We monitored TDG levels in the HCC from March through June 1997, from April through June 1998, and from March through July 1999. We measured TDG pressure, barometric pressure, and water temperatures by using either a Common Sensing TB-F gas meter or a Hydrolab multiprobe sonde equipped with a TDG sensor. The sensor was immersed to a depth of 1 m and allowed to stabilize for a minimum of 15 minutes. At low-velocity sites, the probe was periodically agitated to reduce the potential for bubble formation on the sensor. Gas supersaturation was reported as a percentage calculated by dividing the total gas pressure by the barometric pressure.
In 1997, we monitored dissolved gas levels only when water was spilled at Brownlee Dam. In 1998, TDG levels were monitored during spill periods at each project, regardless of spill conditions at Brownlee Dam. In 1999, we monitored only downstream of Hells Canyon Dam. We installed a Hydrolab Surveyor 4 system at RM 246, approximately 1.5 mi downstream of Hells Canyon Dam, and then we configured the device to record hourly TDG readings. Continuous monitoring took place between March 3 and July 20, 1999.
We selected the sampling sites for 1997 and 1998 based on their locations relative to turbine discharge and spill at each of the three projects. We routinely monitored eight sites (Figures 2, 3, and 4). Samples were collected upstream of the spill gates at all three projects, from spilled water below Brownlee and Oxbow dams, and from the turbine discharge of Brownlee Unit #5 and of the Oxbow powerhouse. Because of the configuration of Hells Canyon Dam, we did not separately monitor spill and turbine discharge at that project.
Specific locations for sampling sites between Hells Canyon Dam and Lewiston were not determined prior to the actual fieldwork. Instead, we set a target distance of 5 to 10 mi between sampling sites, with discretion to sample more frequently depending on river conditions. Samples collected over a two-day period were treated as a synoptic survey.
We used two techniques to develop predictive equations to relate operations to TDG levels. These two techniques were data mining and empirical equation fitting. The amount of available data determined which technique would be used for each project. Our objective was to develop a predictive model or tool with a relatively high degree of confidence in its predictive ability to interface with water quality or habitat models of reservoirs and free-flowing reaches by providing model boundary conditions.
Hells Canyon Complex Page 5 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
As a part of this study, we also assessed structural measures that could be constructed on the Hells Canyon Dam spillway to reduce TDG levels during spill events. Researchers at the Iowa Institute of Hydraulic Research (IIHR) constructed a 1:48 scale model of the Hells Canyon Dam spillway for use in developing a flow deflector configuration that could reduce the gas entrainment and resultant TDG levels downstream of the dam (see Appendix 1 for details).
5. RESULTS AND DISCUSSION
5.1. TDG and Project Operations
5.1.1. Measured TDG Data
The Snake River experienced periods of extremely high runoff during the 1997 water year. Daily average flows measured at the Hells Canyon Dam gauge ranged from a low of 9,507 cfs in October 1996 to a high of 98,958 cfs in January 1997 (Figure 5). With the exception of two days in January, river flows exceeded the Brownlee Dam plant capacity of 35,000 cfs from December 30, 1996, to May 11, 1997, and again from June 5 to July 3, 1997. The turbine capacities of Oxbow and Hells Canyon dams were exceeded from December 29, 1996, to July 5, 1997.
In 1998, Snake River flows in Hells Canyon were again above normal, but they were not as high as those in 1997. Periods of high runoff occurred later in 1998 than they had in 1997. Flows ranged from a low of 8,617 cfs on August 16 to a high of 92,403 cfs on May 28 and 29 (Figure 5). River flows exceeded the plant capacity of Brownlee Dam between May 12 and June 17, 1998. Flows exceeded the plant capacity of Oxbow Dam between October 14 and October 18, 1997, and again between March 20 and June 29, 1998. The plant capacity of Hells Canyon Dam was exceeded during the following periods: October 14 to October 18, 1997; March 31 to April 22, 1998; and May 1 to June 22, 1998.
During 1999, flows from Hells Canyon Dam ranged between 8,285 cfs on August 8 and 63,621 cfs on March 27 (Figure 5). Brownlee Dam’s plant capacity was continuously exceeded between February 26 and May 7, and it was periodically exceeded between May 26 and June 22. Flows continuously exceeded Oxbow Dam’s plant capacity between February 20 and May 26, and they periodically exceeded plant capacity between January 19 and February 20 and between May 30 and June 30. At Hells Canyon Dam, flows continuously exceeded plant capacity between February 25 and May 13, and they periodically exceeded plant capacity between May 17 and June 28.
Total dissolved gas levels were measured in the HCC on 12 dates between March 5 and June 25, 1997, and on 21 dates between April 1 and June 9, 1998. On 8 of the dates in both 1997 and 1998, we measured TDG levels in the reservoirs immediately upstream of the project intakes and also in the project discharges (Figures 6 and 7). On the dates sampled in 1997, the rate of spill over the three dams fell within the following ranges:
Page 6 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
Dam Range of Spill Rate (cfs) Brownlee 3,000–49,000 Oxbow 2,000–34,000 Hells Canyon 1,500–34,200
On the dates sampled in 1998, the following ranges of spill rates were observed:
Dam Range of Spill Rate (cfs) Brownlee 0–59,563 Oxbow 4,000–62,000 Hells Canyon 2,400–66,340
In 1999, TDG levels were measured continuously below Hells Canyon Dam from March 3 to July 20. The rate of spill over Hells Canyon Dam during this monitoring period ranged between 800 and 37,000 cfs.
The TDG levels in Brownlee Reservoir averaged 100.1% saturation in 1997 to 102.4% saturation in 1998. No recorded levels exceeded the accepted Idaho and Oregon state standards of 110% saturation. The maximum TDG level in 1997 (107.9%) occurred on June 25. The maximum TDG level recorded in 1998 (108.1%) occurred on April 21. The cause of these elevated TDG levels in Brownlee Reservoir is unknown. Concurrent dissolved oxygen measurements at the site did not show supersaturated oxygen conditions.
Turbine operation of Unit #5 at Brownlee Dam seemed to have little effect on TDG levels. The average TDG levels in turbine outflow were 99.3% in 1997 and 99.2% in 1998. When we compared TDG levels in the reservoir with levels in the turbine discharge measured on the same date, we found that the turbine discharge level was lower. Although the decrease in TDG from the reservoir to the turbine discharge was significant (P < 0.005), the effects on the system as a whole are most likely negligible.
However, TDG levels downstream of the Brownlee Dam spillway were significantly higher than reservoir levels (P < 0.005). At spill rates over 3,000 cfs, gas levels in spilled water always exceeded the Idaho and Oregon state standards of 110% saturation (Figure 8). During spill episodes, TDG levels ranged up to 128.0% saturation (spill rate = 49,000 cfs).
In both 1997 and 1998, TDG levels in Oxbow Reservoir exceeded the 110% standard when water was spilled at Brownlee Dam (Figures 6 and 7). The excessive TDG levels in Oxbow Reservoir ranged up to 125.3% saturation.
Turbine operation at Oxbow Dam caused no significant changes in TDG levels (P = 0.09). However, on specific sampling days, TDG levels increased by up to 7.2% and decreased by up to 11.3% as water passed through the turbines. These fluctuations could not be attributed to varying turbine flows because those flows remained constant (between 26,250 and 27,000 cfs) for all
Hells Canyon Complex Page 7 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company dates monitored in 1997. Sampling methods may have contributed to the fluctuations. The time that elapsed between collecting reservoir samples and collecting turbine discharge samples may have allowed for the sampling of different water masses. Also, the depth of the Oxbow Dam turbine intakes (73 ft) may have contributed to the observed fluctuations. The levels of TDG in the turbine discharge were compared with surface TDG levels in the reservoir, not with TDG levels at the depth of the turbine intakes. Parametrix (1974) found that dissolved nitrogen levels varied considerably throughout the profile of Oxbow Reservoir.
While the effects of spill at Oxbow Dam were similar to those observed at Brownlee Dam (Figure 9), TDG levels for Oxbow Reservoir varied greatly depending on the rate of spill at Brownlee Dam. In an effort to isolate the effects of spill at Oxbow Dam, we evaluated the changes in percentage of saturation between the reservoir and the spillway. When TDG levels below the Oxbow Dam spillway were compared with levels within the reservoir on specific dates, both increases and decreases in TDG were observed (Figure 10). The largest increase in saturation (20.5%) occurred during a spill rate of 12,000 cfs. The largest decrease (13.2%) occurred during a spill rate of 2,000 cfs. At Oxbow Dam, spill rates under 2,000 cfs and over 24,000 cfs lowered TDG values in the spilled water, while spill rates between 5,000 and 24,000 cfs increased TDG levels in the spilled water. Seattle Marine Laboratories (1972) found that dissolved nitrogen levels decreased on all days sampled as a result of spill at Oxbow Dam, but they did not address rates of spill.
Because of the configuration of Hells Canyon Dam, we did not separately measure TDG levels in water from turbine discharge and from spill. We measured TDG levels approximately 0.8 mi downstream of Hells Canyon Dam at the Hells Canyon boat ramp (owned and managed by the U.S. Forest Service) in 1997 and 1998. In 1999, we recorded hourly measurements approximately 1.5 mi downstream of Hells Canyon Dam. Both locations were presumed to provide a mixture of water from turbine discharge and from spill. In 1997 and 1998, TDG levels in the Hells Canyon Dam tailwater were significantly higher than reservoir levels during periods of spill (P < 0.005), ranging up to 133% saturation (Figure 11). In 1997 and 1998, all rates of spill over 2,500 cfs caused TDG levels to exceed the 110% standard. Hourly measurements taken in 1999 ranged up to 136.3% saturation, showing a clear relationship between spill rates and TDG levels despite considerable variability in TDG at similar spill rates (Figure 12). Nearly all rates of spill produced TDG levels exceeding 110% saturation.
5.1.2. Predictive Numerical Model Development
It is very difficult to accurately predict water supersaturation downstream of dams. A common method for predicting TDG levels downstream of a hydroelectric project involves the use of empirically derived equations used by the CRiSP 1.6 (Columbia River Salmon Passage) model (Shaw 2001). These equations use spill rate in thousands of cubic ft per second (kcfs) as the only parameter to predict TDG for each of the eight federal dams included in the ongoing U.S. Army Corps of Engineers gas abatement study (COE 1996). These equations are expressed in the following form:
c×Qs %TDG = a + b×exp
Page 8 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study where Q is the amount of spill in kcfs and a, b, and c are constants obtained by fitting empirical data at each site to the above equation form.
We used the measured data from the Brownlee and Oxbow projects shown in Figures 8 and 9 to develop the coefficients for equations shown in Table 2. These particular equations become invalid when spill rates approach zero. Therefore, they are not very robust when creating time series boundary conditions for use in other models. However, this technique works well for situations where data are limited, as is the case for both the Brownlee and Oxbow projects.
We used data mining tools and techniques to identify the relationships between TDG levels downstream of Hells Canyon Dam and other measurable parameters. The set of measurable parameters expected to affect downstream TDG saturation levels used in the ANN exercise (see below) include 1) upstream TDG saturation, 2) upstream temperature, 3) gross head available (forebay elevation minus tailwater elevation), 4) spill rate, 5) total flow, and 6) downstream temperature.
The data mining techniques we used in this analysis are those of artificial neural networks (ANNs) and genetic programming (GP). ANNs are capable of taking large data sets and developing a relationship between the input parameters and the desired output parameters through a learning algorithm that establishes the best values for the internal weight distribution. After learning, the weight distribution is fixed and the “trained” ANN can then be used to make predictions from subsequent data sets. Although predictions by ANNs may be extremely accurate, a major disadvantage of the ANN approach is that ANNs do not provide a resulting equation that can easily be transferred to text and used elsewhere. With GP, on the other hand, the evolutionary force is directed toward the creation of models that take a symbolic form (Koza 1992). In this evolutionary paradigm, evolving entities are presented with a collection of data, and the evolutionary process is expected to result in closed-form symbolic expressions that describe the data.
To perform these analyses, large data sets are required to train the models. Total dissolved gas data from Hells Canyon Dam were the largest data set. This data set also had a significant amount of time series information used in these analyses for training and for subsequently verifying the predictive model techniques and equations. Data sets for Brownlee and Oxbow dams had relatively few points and no time series information. Therefore, we did not apply either of the data mining techniques to data from those sites and instead fit empirical equations as stated earlier.
To estimate the effect of dam layout and operational strategies on TDG levels downstream of Hells Canyon Dam, it is necessary to relate quantifiable system parameters to the downstream TDG saturation level. We ultimately developed an ANN model and trained it using only the two most significant input parameters: upstream TDG saturation and spill rate. This 2-input ANN model performed with a very similar accuracy to the original 6-input ANN model. Moreover, because the 2-input ANN model has a much simpler architecture, it is easier and faster to train.
In an attempt to derive usable equations for the prediction of TDG levels from system parameters, we also applied GP techniques to the same data sets described earlier. Of the
Hells Canyon Complex Page 9 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
GP expressions evolved from the same data used to train the ANNs, the following is the best performing expression:
= − × −4 TDG% 79.51 3.582 10 Qs + × −1 × 4.738 10 TDGu − × −4 × 2 7.647 10 TDGu where Qs is the flow rate over the spillway in kcfs and TDGu (expressed as a percentage) is the upstream saturation term. However, a better performing equation resulting from GP techniques also included temperature from the same data sets:
− − − = × 4 − + × 1 + × 1 − × TDG% ((1.12 10 Qs 4.57 3.14 10 T 8.44 10 TDGu 1.2 × −1 − 2 × × −4 − − × −5 − × −1 + (7.05 10 T 7.06) (1.76 10 Qs 2.14) (3.4 10 Qs 4.13 10 × −3 − × −3 × −3 − × −2 × −1 − (1.44 10 TDGu 6.96 10 )((7.45 10 TDGu 3.6 10 )(8.71 10 T 7.06) + × − 6 − × −1 + × −3 × −4 − 5.46 10 Qs 1.02 10 7.45 10 TDGu ))(1.76 10 Qs 2.14) + × −5 − × −1 × −1 − + × × × −1 − (6.18 10 Qs 7.51 10 )(8.71 10 T 7.06) 22.2) 5.5) 7.65 10 6.48 where Qs is the flow rate over the spillway in kcfs, T is temperature in degrees C and TDGu (expressed as a percentage) is the upstream saturation term. We later used this expression to develop TDG boundary condition information below Hells Canyon Dam for use in other models.
5.2. Spill Manipulation Test
5.2.1. Brownlee Dam
We conducted the Brownlee Dam spill test on June 4, 1998, during a spill rate of 39,000 cfs. We began the test by spilling all 39,000 cfs through the upper two spill gates at Brownlee Dam. Spill was maintained through these gates at that rate for approximately one hour and then moved to the lower two spill gates. When the transition from upper to lower spill gates was complete, the spill was held through the lower gates for an additional hour. We used the Hydrolab sensor to monitor TDG levels downstream of the spillway along the Oregon shore of the river. Readings were recorded at five-minute intervals throughout the entire procedure.
The TDG levels downstream of Brownlee Dam averaged 114% while spilling 39,000 cfs through the upper gates (Figure 13). Levels increased during the transition period and averaged 127.7% during spill through the lower gates, an increase of nearly 14%. Therefore, at a spill rate of 39,000 cfs, spilling water from the upper gates at Brownlee Dam appears to benefit the HCC system. Because spill at Brownlee Dam is the most significant factor affecting TDG levels within Oxbow and Hells Canyon reservoirs, decreasing TDG in spill at Brownlee Dam could reduce the elevation of TDG levels in the downstream reservoirs.
Page 10 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
5.2.2. Hells Canyon Dam
We conducted the Hells Canyon Dam spill test on June 3, 1998, at a spill rate of 28,000 cfs. We began the test with spill through the lower spill gates. Spill was maintained at this level for approximately one hour and then moved to the upper spill gates. When the transition from lower to upper spill gates was complete, we maintained the spill through the upper gates for an additional hour. Sampling was conducted at the Hells Canyon boat ramp (owned by the U.S. Forest Service) located approximately 0.8 mi downstream of the dam. Total dissolved gas levels were monitored using the Hydrolab sensor, and readings were recorded at five-minute intervals throughout the entire procedure.
Spill of 28,000 cfs from the upper gates produced higher TDG levels than spill from the lower gates (Figure 13). The levels of TDG downstream of Hells Canyon Dam averaged 135% saturation during spill from the lower gates. Gas levels increased during the transition period and averaged 139% during spill through the upper gates. Therefore, to minimize the effects of spill, we recommend that the lower gates be used to spill water at Hells Canyon Dam whenever possible. The effect of spill gate elevation on TDG was notably larger (average 14% difference) at Brownlee Dam than at Hells Canyon Dam (average 4% difference).
Our results are consistent with those by Parametrix (1974), who conducted their study at Hells Canyon Dam in the early 1970s. In an effort to determine the effects of project operations on TDG supersaturation downstream of Hells Canyon Dam, Parametrix (1974) monitored nitrogen levels in water spilled through combinations of the four spillways. They found no differences in dissolved nitrogen levels downstream of the dam as a result of changing the locations of spill. Our study found only a very small difference (4%).
5.3. Downstream Dissipation of TDG
We monitored the dissipation of TDG downstream of Hells Canyon Dam at locations between RM 142 and 247 during seven different spill episodes (Figure 14). We found a declining trend in TDG levels as the distance from the dam increased. Generally, TDG decreased approximately 0.3% saturation per river mile when discharge TDG levels exceeded 120% saturation. We found a direct relationship between rate of spill and distance from the dam at which TDG levels exceeded the 110% standard. All measured spill episodes over 19,000 cfs caused TDG levels to exceed 110% at all sites upstream of RM 180 (67 mi downstream of the dam), while spill episodes of 9,000 and 13,400 cfs caused levels to exceed 110% downstream to RM 200 (47 mi downstream of the dam). A spill rate of 2,400 cfs caused TDG levels to exceed 110% upstream of RM 230 (17 mi downstream of the dam). Between May 19 and 23, 1997, no water was spilled at Hells Canyon Dam, but water was spilled at Brownlee Dam at a rate of 16,500 cfs. During this period, the 110% standard was not exceeded downstream of Hells Canyon Dam. Our data are consistent with the findings of Seattle Marine Laboratories (1972). They found dissolved nitrogen levels of 125% immediately downstream of Hells Canyon Dam. The levels decreased to 107% nearly 60 mi downstream near the mouth of the Salmon River. However, Seattle Marine Laboratories (1972) did not report rates of spill.
Hells Canyon Complex Page 11 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
In addition to the simplistic plotting of empirical data to evaluate downstream TDG levels, IPC used data collected during this study to develop and calibrate MIKE 11, a one-dimensional hydrodynamic model that can estimate TDG levels under various operating scenarios (Parkinson 2001).
5.4. Physical Modeling of Hells Canyon Dam
The Hells Canyon Dam spillway is configured with three high-level outlets, or crest gates, and two low-level outlets, or sluice gates, for releasing up to 300,000 cfs of water past the dam. Figure 15 shows a general section through the spillway of Hells Canyon Dam. Water spilled through the upper crest gates is discharged over nappe deflectors, which deflect water off the concrete surface, creating a nearly free-falling vertical jet of water into the stilling basin (Figure 16). However, water released from the lower sluice gates plunges into the stilling basin along the concrete face of the spillway.
As mentioned earlier, IIHR researchers constructed a 1:48 scale model of Hells Canyon Dam both to understand the gas entrainment process at the site and to design a flow deflector configuration that might reduce TDG levels downstream of the dam.
A flow deflector design that accommodated spill from the upper crest gates failed to create acceptable flow characteristics for minimizing gas entrainment in the stilling basin. Therefore, the focus of the design process turned to the lower sluice gates. These gates also provide operational flexibility of the spillway when flow conditions are higher or when energy dissipation is critical. However, the high head at Hells Canyon Dam required a deflector design different from designs for the Columbia River projects. Ultimately, the deflector designed to maintain the best flow regimes under the greatest range of flows was 16 ft long, located at an elevation of 1,468 ft on the spillway face, and featured a 5° upward lip angle (Figure 17). This design maintains the most desirable performance up to 15,000-cfs discharge per sluice gate (Figure 18) or a total spill rate of 30,000 cfs. This total flow rate of 60,000 cfs past the dam, 30,000 cfs through the turbines and 30,000 cfs through the spillway, equals approximately 98% of all recorded flow occurrences. For more detailed information on the physical model evaluations, see Appendix 1 to this report.
Preliminary velocity measurements and observations revealed the need to evaluate the deflectors’ potential effects on bank erosion and bed scour because of the change in flow pattern they would produce. A three-dimensional physical model was constructed to further evaluate the flow and velocity patterns resulting from the proposed deflector configuration along the bank line, riverbed, and other nearby structures. The model would also compare these effects to effects produced by current spill procedures using the upper crest gates. The model included the Hells Canyon Dam powerhouse and approximately 3,300 ft of river downstream of the dam. Tests with the model illustrated how, with the deflectors in place, bed-material movements differed under current spill operations. Although the model tests illustrated a representation of material movement, the deflectors did not appear to produce detrimental effects to nearby structures. However, at flows approaching 30,000 cfs total spill, more material was pulled back onto the spillway apron. These materials could erode the concrete apron as they swirl around in the modified flow pattern produced by the deflectors. Therefore, if the deflector design is
Page 12 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study implemented, proper operational protocol must be followed and the crest gates used occasionally for spill release to flush the accumulated material from the apron.
6. SUMMARY AND CONCLUSIONS
Spilling water at any of the three projects within the HCC can increase TDG to supersaturation levels that exceed the 110% protective standard. Measured levels in the Brownlee and Oxbow tailraces generally range from 120% to 125% saturation during spill episodes at those two projects. In the Hells Canyon Dam tailwater, measured levels peak around 135% saturation. Supersaturation declines in the Snake River as water flows downstream of Hells Canyon Dam. However, levels in excess of 110% saturation can persist downstream to the confluence with the Salmon River.
This study did not include biological sampling to evaluate the site-specific effects of the elevated TDG levels on biota downstream of the dams. However, juvenile fall chinook surveys conducted within a time frame similar to the one used in this study failed to identify symptoms related to elevated TDG (W. Connor, U.S. Fish and Wildlife Service, pers. comm.). Adult steelhead and chinook captured at Hells Canyon Dam have exhibited symptoms of gas bubble disease (Burton 1988a,b; Bertellatti and Young 1990; Snider 1993). However, it could not be determined whether the symptoms resulted from exposure to elevated TDG levels immediately downstream of Hells Canyon Dam or exposure to elevated TDG levels produced by the federal lower Snake River projects located further downstream. The potential for detrimental biological effects downstream of Brownlee and Oxbow dams (in Oxbow and Hells Canyon reservoirs) is logically lower than the potential downstream of Hells Canyon Dam. Not only are TDG levels generally lower downstream of Brownlee and Oxbow dams than those downstream of Hells Canyon Dam, but deeper areas within the impoundments provide more potential refuge areas. Organisms in areas of depth greater than 2 m would not be exposed to the effects of supersaturation, even under the worst-case spill scenario.
Spilling water from the upper spill gates at Brownlee Dam rather than the lower spill gates appears to minimize TDG levels in the Brownlee Dam tailwater. Although levels still exceed 110%, it might be possible to keep them below 120% saturation by releasing water only from the upper spill gates. Selective spill gate operation may have a small effect (approximately 4%) on supersaturation below Hells Canyon Dam. However, installing the flow deflectors designed by IIHR might reduce supersaturation below Hells Canyon Dam. We are not able to predict actual TDG levels or the amount of reduction if the deflectors are installed. However, the physical model qualitatively indicates that improvement over current conditions might be possible by preventing the entrained air bubbles from plunging to depth and supersaturating the water. Also, the physical model shows that the flow deflectors on Hells Canyon Dam would be ineffective at spills greater than 30,000 cfs. At that flow rate, the deflectors could not keep the entrained air from plunging to depth in the stilling basin. However, the IIHR design appears to be effective for approximately 98% of recorded flow occurrences at the project. We were unable to identify any measures that would guarantee complete compliance with the 110% saturation standard.
Hells Canyon Complex Page 13 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
7. ACKNOWLEDGMENTS
We want to thank Stan Pierce, Mark Stute, Jim Trainer, and Karen Parks for their assistance in data collection and reporting. In addition, Dr. Larry Weber of the University of Iowa and Duncan Hay, an independent engineering consultant, provided invaluable expertise for the successful development of the Hells Canyon Dam spillway model and deflector configuration. We also thank Chris Randolph, Scott Larrondo, and Dave Meyers of IPC for their support throughout the study.
8. LITERATURE CITED
Beyer, D., B. G. D’Aoust, and L. Smith. 1976. Responses of coho salmon (Oncorhynchus kisutch) to supersaturation at one atmosphere. Pages 47–50 in D. H. Fickeisen and M. J. Schneider, editors. Gas bubble disease. CONF-741033, Technical Information Center, Energy Research and Development Administration, Oak Ridge, TN.
Bertellatti G. F., and D. S. Young. 1990. Oxbow Hatchery 1989 Annual Report, steelhead and spring chinook. Idaho Department of Fish and Game, Boise, ID.
Blahm, T. H., B. McConnel, and G. R. Snyder. 1976. Gas supersaturation research National Marine Fisheries Service Prescott facility, 1971 to 1974. Pages 11–19 in D. H. Fickeisen and M. J. Schneider, editors. Gas bubble disease. CONF-741033, Technical Information Center, Energy Research and Development Administration, Oak Ridge, TN.
Bouck, G. R. 1980. Etiology of gas bubble disease. Transactions of the American Fisheries Society 109:703–707.
Burton G. F. 1988a. Oxbow Hatchery Annual Report, 1986 Steelhead brood year, 1986 spring Chinook trapping. Idaho Department of Fish and Game, Boise, ID.
Burton G. F. 1988b. Oxbow Hatchery Annual Report, 1987 Steelhead brood year, 1987 spring Chinook trapping. Idaho Department of Fish and Game, Boise, ID.
COE (U.S. Army Corps of Engineers). 1996. Dissolved gas abatement. Phase I, Technical Report. U.S. Army Corps of Engineers, Portland District, Walla Walla District, Walla Walla, WA.
Dawley, E. M., M. Schiewe, and B. Monk. 1976. Effects of long-term exposure to supersaturation of dissolved atmospheric gases on juvenile chinook salmon and steelhead trout in deep and shallow test tanks. Pages 1–10 in D. H. Fickeisen and M. J. Schneider, editors. Gas bubble disease. CONF-741033, Technical Information Center, Energy Research and Development Administration, Oak Ridge, TN.
Page 14 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
Koza, J. 1992. Genetic programming: on the programming of computers by means of natural selection. MIT Press, Cambridge, MA.
Nebeker, A. V., D. G. Stevens, and J. R. Brett. 1976. Effects of gas supersaturated water on freshwater aquatic invertebrates. Pages 51–65 in D. H. Fickeisen and M. J. Schneider, editors. Gas bubble disease. CONF-741033, Technical Information Center, Energy Research and Development Administration, Oak Ridge, TN.
Parametrix, Inc. 1974. Snake River 1973 dissolved gas studies. Report to Idaho Power, Boise, ID. 58 p.
Parkinson, S. K., editor. 2001. Project hydrology and hydraulic models applied to the Hells Canyon reach of the Snake River. In Technical appendices for new license application: Hells Canyon Hydroelectric Project. Idaho Power, Boise, ID. Technical Report E.1-4.
Seattle Marine Laboratories. 1972. Final report: 1972 nitrogen monitoring studies. Report to Idaho Power, Boise, ID. 74 p.
Shaw, P. 2001. Gas generation equations for CRiSP 1.6. Available at
Snider, B. R. 1993. Oxbow Fish Hatchery and Hells Canyon fish trap 1988 Annual Report. Idaho Department of Fish and Game, Boise, ID.
Stevens, D. G., A.V. Nebeker, and R. J. Baker. 1980. Avoidance responses of salmon and trout to air-supersaturated water. Transactions of the American Fisheries Society 109:751–754.
Weitkamp, D. E. 1974. Dissolved gas supersaturation in the Columbia River system: salmonid bioassay and depth distribution studies, 1973 and 1974. Report of Parametrix, Inc. to utility cooperative, in care of Idaho Power Company, Boise, ID.
Weitkamp, D. E., and M. Katz. 1980. A review of dissolved gas supersaturation literature. Transactions of the American Fisheries Society 109:659–702.
Hells Canyon Complex Page 15 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
This page left blank intentionally.
Page 16 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
Table 1. Physical characteristics of Brownlee, Oxbow, and Hells Canyon powerhouses and dams.
Number of Hydraulic Plunge Pool Maximum Head Project Generating Units Capacity (cfs) Depth (ft) (ft) Brownlee 5 35,000 35 272 Oxbow 4 28,000 NA 117 Hells Canyon 3 30,500 67 220
Table 2. Equations used to describe spill vs. total dissolved gas relationships at the projects of the Hells Canyon Complex.
Dam Equation Coefficients
Brownlee %TDG = a + b exp(cQs) a = 122.1; b = -48.3; c = -0.3931
Oxbow %TDG = a + b exp(cQs) a = 121.2; b = -16.0; c = -0.1918 − − − Hells Canyon TDG % = ((1.12 × 10 4Q − 4.57 + 3.14 × 10 1T + 8.44 × 10 1TDG − 1.2 × s u × −1 − 2 × × −4 − − × −5 − × −1 + (7.05 10 T 7.06) (1.76 10 Qs 2.14) (3.4 10 Qs 4.13 10 × −3 − × −3 × −3 − × −2 × −1 − (1.44 10 TDGu 6.96 10 )((7.45 10 TDGu 3.6 10 )(8.71 10 T 7.06) + × − 6 − × −1 + × −3 × −4 − 5.46 10 Qs 1.02 10 7.45 10 TDGu ))(1.76 10 Qs 2.14) + × −5 − × −1 × −1 − + × × × −1 − (6.18 10 Qs 7.51 10 )(8.71 10 T 7.06) 22.2) 5.5) 7.65 10 6.48
Hells Canyon Complex Page 17 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
This page left blank intentionally.
Page 18 Hells Canyon Complex Lewiston River Mile 139
OREGON
ver Ri Grande Ronde
er iv r n o
m
l
a
S IDAHO
Rm 188.2
r r r r
e e e e
v v v v
i i i i
R R R R
a a a a
h a n m I
r e v i R
e k a n S
r
e
v
i
R
a
h
a
n
m
I
Hells Canyon Dam River Mile 247.6
Figure 1. Total dissolved gas study area.
. s Legend e R n o y n Perennial River or Stream a C s River or Reservoir ll e H Dam
Oxbow Dam River Mile 272.8 An IDACORP Company
k ee . Cr s e ne i R P 3 1.5 0 3 6 9 12
w
o Miles b x r e O iv R Brownlee Dam se River Mile 284.5 or d h Wil Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
This page left blank intentionally.
Page 20 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
Bridge
Spillway Site
Turbine Discharge Site
Brownlee Spillway
Brownlee Turbines
Reservoir Site
Brownlee Reservoir N
W E
S
Figure 2. Brownlee Dam study area showing reservoir, turbine discharge, and spillway sampling sites.
Hells Canyon Complex Page 21 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
Turbine Discharge Site
Oxbow Spillway
Pine Creek Oxbow Turbines
Spillway Site
Oxbow Reservoir Reservoir Site
N
W E
S
Figure 3. Oxbow Dam study area showing reservoir, turbine discharge, and spillway sampling sites.
Page 22 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
Tailwater Site
Hells Canyon Boat Launch
Hells Canyon Dam Turbine Discharge Hells Canyon Dam Spillway
Reservoir Site
Hells Canyon Reservoir
N
W E
S
Figure 4. Hells Canyon Dam study area showing reservoir and tailwater sampling sites.
Hells Canyon Complex Page 23 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
Snake River Flow Brownlee Project Hydraulic Capacity 100,000 Oxbow Project Hydraulic Capacity Hells Canyon Project Hydraulic Capacity
80,000
60,000 Flow (cfs) 40,000
20,000
0 10/1/1996 5/29/1997 1/24/1998 9/21/1998 5/19/1999 1/29/1997 9/26/1997 5/24/1998 1/19/1999 9/16/1999
Figure 5. Mean daily Snake River flow (cfs) measured downstream of Hells Canyon Dam from October 1996 through August 1999.
Page 24 Hells Canyon Complex Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
March 5, 1997 March 12, 1997 Flow = 57,258 Flow = 52,355 cfs 130 130
120 120
110 110 Turbine Turbine 100 100 Spill Spill
90 90 123456 123456
March 18, 1997 March 26, 1997 Flow = 54,830 cfs Flow = 50,301 cfs 130 130
120 120
110 110 Turbine Turbine 100 100 Spill Spill
90 90 123456 123456
April 16, 1997 April 22, 1997 Flow = 41,795 cfs Flow = 59,372 cfs 130 130
120 120
Percent Saturation Percent 110 110 Turbine Turbine 100 100 Spill Spill
90 90 123456 123456
April 30 , 1997 May 6, 1997 Flow = 64,296 cfs Flow = 40,279 cfs 130 130 125 120 120 115 110 110 Turbine Turbine 105 100 Spill Spill 100 95 90 123456 123456 Location
Figure 6. Total dissolved gas levels in water at six locations: (1) Brownlee Reservoir, (2) Brownlee Project discharge (spill and turbine), (3) Oxbow Reservoir, (4) Oxbow Project discharge (spill and turbine), (5) Hells Canyon Reservoir, and (6) Hells Canyon Project discharge on 8 dates in 1997.
Hells Canyon Complex Page 25 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
April 1, 1998 April 2, 1998 Flow = 31,786 cfs Flow = 31,755 cfs
130 Turbine 130 Turbine
120 Spill 120 Spill
110 110
100 100
90 90 123456 123456
April 15, 1998 April 21, 1998 Flow = 31,817 cfs Flow = 31,612
130 Turbine 130 Turbine
120 Spill 120 Spill
110 110
100 100
90 90 123456 123456
May 13, 1998 May 19, 1998 Flow = 40,160 Flow = 67,810
130 130 Percent Saturation Percent 120 120
110 110 Turbine
100 Turbine 100 Spill Spill 90 90 123456 123456
May 27, 1998 May 28, 1998 Flow = 93,940 cfs Flow = 93,100 cfs
130 130
120 120
110 110 Turbine Turbine 100 100 Spill Spill
90 90 123456123456
Location
Figure 7. Total dissolved gas levels in water at six locations: (1) Brownlee Reservoir, (2) Brownlee Project discharge (spill and turbine), (3) Oxbow Reservoir, (4) Oxbow Project discharge (spill and turbine), (5) Hells Canyon Reservoir, and (6) Hells Canyon Project discharge on 8 dates in 1998.
Page 26 Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
130
125
120
115
110 Percent Saturation Percent
105 1997 1998 100 0 10,000 20,000 30,000 40,000 50,000 60,000 5,000 15,000 25,000 35,000 45,000 55,000 65,000 Spill (cfs)
Figure 8. The relationship of spill and total dissolved gas measured during spill at Brownlee Dam, 1997−1998.
130
125
120
115
110 Percent Saturation Percent
105 1997 1998 100 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 5,000 15,000 25,000 35,000 45,000 55,000 65,000 Spill (cfs)
Figure 9. The relationship of spill and total dissolved gas measured downstream of Oxbow Dam during spill, 1997–1998.
Hells Canyon Complex Page 27 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
30 1997 1998 20
10
0
(10)
(20) Change in Percent Saturation Percent Change in
(30) 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 5,000 15,000 25,000 35,000 45,000 55,000 65,000 Spill (cfs)
Figure 10. Change in percentage of total dissolved gas saturation between Oxbow Reservoir and the Oxbow spillway.
Page 28 Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
135
130
125
120
115
110 Percent Saturation
105 1997 1998 100 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 5,000 15,000 25,000 35,000 45,000 55,000 65,000 Spill (cfs)
Figure 11. Percentage of TDG saturation measured at the Hells Canyon boat ramp downstream of Hells Canyon Dam, 1997–1988.
Hells Canyon Complex Page 29 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
R2 = 0.88
Figure 12. The relationship of spill and total dissolved gas measured during spill at Hells Canyon Dam, 1997–1999.
Page 30 Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
Brownlee Dam Spill Test
140
130
Upper Gates Transition 120 Lower Gates % Saturation
110
100 0800 0854 0914 0934 0954 1014 1034 1052 Time
Hells Canyon Dam Spill Test
140
130 Upper Transition Gates Lower Gates
120 % Saturation
110
100 1100 1207 1227 1247 1307 1327 1347 Time
Figure 13. Total dissolved gas levels measured while releasing 39,000 cfs of spill from Brownlee Dam on June 4, 1998, and 28,000 cfs of spill at Hells Canyon Dam on June 3, 1998.
Hells Canyon Complex Page 31 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
132 130 April 9 -10, 1997, spilling 21,000 cfs April 16-17, 1998, spilling 2,400 cfs 128 April 23-24, 1998, spilling 2,400 cfs 126 May 14-15, 1998, spilling 13,400 cfs May 20-21, 1998, Spilling 29,000 cfs 124 April 19-20, 1999, spilling 19,000 cfs 122 May 4, 1999, spilling 9,000 cfs 120 118 116 114 112 110 108 106 104 102 Total Dissolved Gas (Percent Saturation) 100 98 250 240 230 220 210 200 190 180 170 160 150 140 River Mile
Figure 14. Downstream dissipation of total dissolved gas within Hells Canyon relative to the 110% saturation standard.
Page 32 Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
AXIS OF DAM MAX. W.S. EL.1693.00 FULL W.S. RADIAL GATE EL.1688.00 MODEL SCALE = 1:48 CREST 0 20 40 60 80 100 EL.1638.00 PROTOTYPE SCALE (feet) SPILLWAY 012 TRAINING WALL MODEL SCALE (feet)
CREST EL.1549.00 RADIAL EL.1540.00 GATE MAX. T.W. EL.1515.00 EL.1520.00
MIN. T.W. EL.1467.00 DEFLECTOR EL.1426.00
EL.1400.00
LAB FLOOR E.1350.00
Figure 15. General sectional view of the Hells Canyon Dam spillway model constructed by researchers at the Iowa Institute of Hydraulic Research (IIHR).
Hells Canyon Complex Page 33 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
Figure 16. Location of crest gates, nappe deflectors, and sluice gates on the Hells Canyon Dam spillway.
Page 34 Idaho Power Company Hells Canyon Complex Total Dissolved Gas Study
MAX.T.W. EL.1515.00
16’-0” MIN.T.W. R15’-0”4’-0” 4 3/16” EL.1468.00 EL.1467.00
5.0O
13’-6”
68’-0”
EL.1400.00
Figure 17. General view of the modified deflector configuration for the Hells Canyon Dam spillway developed using the two-dimensional model.
Hells Canyon Complex Page 35 Hells Canyon Complex Total Dissolved Gas Study Idaho Power Company
Hells Canyon Project Performance Curve for 16.0' Deflector at 1468.0' with 5O lip
No Powerhouse 1 P.H. Unit 2 P.H. Units 3 P.H. Units 1500.0
1498.0
1496.0
1494.0 1492.0 SURFACE JUMP 1490.0 1488.0
1486.0 1484.0
.) 1482.0 1480.0
1478.0 SURFACE JET Tailwater (Elev 1476.0
1474.0 1472.0 1470.0 VENTED 1468.0 SURFACE JET PLUNGING FLOW 1466.0 1464.0
1462.0 1460.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 Discharge Per Sluiceway Bay (kcfs)
Figure 18. Deflector performance from the evaluation of the IIHR two-dimensional physical model.
Page 36 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
October 13, 2000
Sharon Parkinson Hydro Production Department Idaho Power 1221 Idaho Street Boise, Idaho 83702
RE: Hells Canyon Dam Flow Deflector Design
Dear Sharon:
The details related to the design of sluiceway flow deflectors for Hells Canyon Dam are contained in this letter report. The procedure for analyzing a specific flow deflector in the 1:48 scale model is first described. The flow deflector designs tested and flow regime performances of these deflectors are then presented. Finally, general velocity profiles obtained from the model are presented and discussed.
Introduction
Background
The distinctive geometry of Hells Canyon Dam presented some interesting challenges in developing deflectors for mitigation of total dissolved gas. The upper nappe deflectors, higher head than previous projects with spillway deflectors, the deep and short stilling basin, inclusion of the lower level sluiceways, and the higher unit discharge were all important considerations in development of an effective deflector design for Hells Canyon Dam. After initial operation of the model it became quite apparent that the unique flow developed by the upper nappe deflectors would create problems in designing a deflector for the upper spillways. Specifically, for all flows originating from the upper spillway gates the flow is deflected away from the concrete spillway surface and becomes a nearly unattached, free-falling jet. This flow phenomenon, along with the relatively large head, result in very large deflectors for the upper spillway gates. Such large deflectors when tested in the model had a very negative characteristic, which resulted in flows above the total dissolved gas design discharge to impact the riverbed downstream of the stilling basin. Hence, causing great concern for dam safety during passage of large spillway flows.
Based on these early tests it was determined that if an acceptable deflector design could be developed for the lower level sluiceways then the operational flexibility of the dam would remain. Specifically, the lower level sluiceway could then be operated for flows in which total dissolved gas levels are important and the upper spillway gates can be
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 1 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
operated for high spill discharges when energy dissipation and dam safety become critically important. Additionally, because of the flatter surface and sidewall containment for the sluiceways deflector design performance was improved.
Baseline Conditions
A balance must be maintained between energy dissipation and bubble entrainment during development of flow deflectors. The photo shown in Figure 1 illustrates the current conditions for 10.0 kcfs discharged through the upper spillway with no flow deflectors at Hells Canyon Dam. The stilling basin is the primary means of energy dissipation for the dam. It is crucial that enough energy dissipation is achieved to maintain the structural integrity of the dam with flow deflectors in place.
Figure 1. Baseline Condition with No Deflectors
The purpose of flow deflectors is also demonstrated in the photo. Large amounts of air entrainment and bubbles brought to depth are occurring in the present conditions. Flow deflectors change the flow regime so that fewer bubbles are brought to depth and the potential for total dissolved gas (TDG) is reduced.
Procedure
Previous research performed by the U.S. Army Corps of Engineers (USACE) and the Iowa Institute of Hydraulic Research (IIHR) on deflector implementations was used as a guideline for this model study. Although the procedures documented by the USACE in their Dissolved Gas Abatement Study - Phase II Report (USACE, 1999) and by IIHR (Nielsen et al., 2000) set forth basic methods, the high head of Hells Canyon Dam provided a unique dimension to the project.
Testing of each deflector in the 1:48 scale model followed the same general method. A prototype discharge of 2.5 kcfs was set in both sluiceways, and the forebay was allowed
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 2 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
to stabilize at elevation 1686.0’ for 15 minutes. The tailwater elevation approximately 920 prototype feet downstream from the end sill was then set to elevation 1500.0’. This was accomplished by using a point gage referenced to the spillway crest of the model.
The corresponding flow regime for this tailwater was observed and recorded. The tailwater was then gradually lowered until a change in flow regime was detected. This procedure was repeated until a tailwater corresponding to the plunging flow regime was observed. The same process was implemented for prototype flows of 5.0, 7.5, 10.0, 12.5, and 15.0 kcfs per sluiceway bay. Four distinct flow regime classifications were observed during this procedure. A description of these regimes is as follows:
a. Surface Jump. High submergence, flow rolls back onto jet within the sluiceway. Jet begins to ramp above water surface (see Figure 2). b. Surface Jet. Jet is swept out, flow is deflected 5-10 degrees, entire jet is surface oriented and relatively flat (see Figure 3). c. Vented Surface Jet. Begins as submergence is lowered to the point where the nappe begins to intermittently aerate. Jet is still surface oriented and appears to deflect off of the downstream water surface (see Figure 4).
d. Plunging Flow. Jet is consistently aerated and entrains bubbles to depth (see Figure 5).
The figures shown below illustrate these four flow regimes for a spill discharge of 5.0 kcfs per sluiceway bay using a flow deflector 16.0’ long at elevation 1468.0’ with a 5o lip angle.
Figure 2. Surface Jump
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 3 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Figure 3. Surface Jet
Figure 4. Vented Surface Jet
Figure 5. Plunging Flow
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 4 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
A performance curve was then generated from this data by the following method. The tailwater elevations corresponding to a change in flow regime classification were plotted against their respective discharge per sluiceway bay. Distinct boundaries between the four flow classifications were apparent and denoted on the plot. Also included on these plots, were tailwater curves for sluiceway discharges with 0, 1, 2, and 3 powerhouse units operating. A prototype discharge of 10 kcfs per powerhouse unit was used to obtain these curves. The combination the performance curve and tailwater curves on a single plot creates a powerful analysis tool for flow deflector design. These plots form a basis for comparing different deflector designs, in each case trying to maximize the overlap of normal operating conditions with the surface jet flow regime.
Deflector Designs
Background
The primary design components for a flow deflector include the elevation, length, transition radius, and lip angle. These features are interrelated, which does not allow for a simple solution. Variations in head, gate geometry, and tailwater conditions prevent designing a deflector based solely on past projects. Close examination of the design goals does help determine a logical range for these values. By reducing some of these design factors, the iteration procedure involved in the design is kept to a minimum. IIHR’s past experience with flow deflectors at Wanapum Dam and Rock Island Dam also provided a source of reliable information for this project.
The preliminary elevation of the flow deflectors was determined by analyzing the tailwater curves for total river discharges at or below 60.0 kcfs. This discharge value seems quite logical since it includes over 98% of flows recorded from 1965 to 1999 (USGS gauge no. 13290450 Snake River at Hells Canyon Dam ID-OR State Line). Therefore, with three powerhouse units operating at their hydraulic capacity (10 kcfs) each, the maximum spill discharge would be 30.0 kcfs. The flow deflectors were designed with the intent of passing as close to 30.0 kcfs of total spill flow as possible. The unique geometry of the dam favors implementation of flow deflectors in the sluiceways rather than the upper spillways. By analyzing the tailwater curves up to discharges of 15.0 kcfs per sluiceway bay, an initial elevation of a deflector was determined. The deflector elevation must remain below the tailwater elevation to prevent vented surface and plunging flows from occurring. These flow regimes result in large amounts of air being carried to depth, increasing the potential for TDG. Furthermore, the deflector elevation should be high enough to keep the performance within the surface jet flow regime for high tailwaters, since the submerged surface jump regime may result in higher levels of TDG. Based on this analysis, an initial deflector elevation of 1468.0’ was proposed.
Another critical design component of the flow deflectors is length. The length of the deflectors must be long enough to successfully deflect the flow, but kept short to
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 5 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
minimize construction cost. Cavitation is another crucial element affecting the length of the flow deflectors. Flow deflectors for the sluiceway bays at Hells Canyon Dam could very easily cause cavitation problems if they protrude past the spillway face of the dam. If this protrusion was severe enough, cavitation damage could occur over time to the deflectors from impact of the upper spillway nappe. From the model study, it appears that the flow impact from the upper spillway nappe on the recommended deflectors is minimal due to the upper nappe deflectors. These factors helped determine the preliminary design length of 14.3’, which extended 6” beyond the face of the spillway.
The transition radius refers to the radius of an arc connecting the sluiceway face to the horizontal face of the deflector. Various deflector designs implemented and tested by the IIHR, USACE, and other agencies have shown good results with a 15.0’ transition radius. This transition radius was initially used to minimize fish injury. However, these studies also showed this radius had a positive impact on performance, so this base value was used for all of the deflectors tested in this study.
The final important element in flow deflector design is lip angle. Several previous deflectors have been designed with the upper face angled slightly upward. The length of this angled portion and the magnitude of the angle are essential to the performance of the Hells Canyon Dam deflector. Initially, all deflector designs analyzed by IIHR used a 0o lip angle. Several different angles were also examined during the course of this model study to obtain better performance at Hells Canyon Dam up to the 30 kcfs design spill level.
Deflector Designs Tested
A preliminary deflector design length of 14.3’ with an elevation 1468.0’ was tested. This deflector had a 0o lip angle and a 15.0’ transition radius. A profile sketch of the general deflector location is shown in Figure 6 below.
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 6 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Figure 6. Sketch of General Deflector Location
After studying the preliminary 14.3’ deflector performance at elevation 1468.0’, the same size and shape of deflector was analyzed at elevation 1464.0’. Based on the results from the first two deflector tests, some brief tests were implemented for a 17.3’ deflector at elevation 1468.0’ with various lip angles up to 45o. It was then decided to perform a full analysis of an 18.3’ deflector at elevation 1468.0’ with no lip angle. Lip angles of 5o, 10o, and 15o were also investigated for this deflector. A similar evaluation was done for the 14.3’ deflector at 1468.0’ with a 5o and 10o lip angle. The performance of these deflectors is discussed in the following section.
The results obtained from the previously mentioned deflectors led to the design of a 16.0’ deflector at 1468.0’. A 0o, 5o, and 10o lip angle was investigated for this deflector. The recommended 16.0’ deflector at elevation 1468.0’ with a 5o lip angle is illustrated here in Figure 7.
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 7 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Figure 7. Sketch of 16.0’ Deflector at Elevation 1468.0’
Deflector Performance Results
The primary method of analysis for flow deflector design used by IIHR is development of performance curves. As described earlier, these curves are generated using four basic flow regime classifications. Table 1 summarizes the basic design parameters and general performance of each deflector tested.
Length Elevation Lip Angle Performance Comments (feet) (feet) (degrees) Operations fall into surface jump and 14.3 1468.0 0 plunging flow regime
Surface jump flow regime exhibited 14.3 1464.0 0 for small discharges
Plunging flow at low tailwaters, 17.3 1468.0 Varying up to 45 drastic changes in flow at high angles
18.3 1468.0 5, 10, and 15 Too drastic of change in flow regimes
Good for larger spills, surface jump 18.3 1468.0 0 flows for small spills
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 8 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Cleaner and smoother flows within 14.3 1468.0 5 and 10 surface jet regime; bordering plunging flow at high spills
High angle borders on surface jump 16.0 1468.0 15 regime for small spills
Good overall performance, angle 16.0 1468.0 10 slightly too drastic at low and high spills Cleanest and smoothest flow, minimal amount of aeration, remained in 16.0 1468.0 5 surface jet regime for normal operations with 1, 2, and 3 powerhouse units
Table 1. Summary of Deflectors Tested
After observing the performance of the 14.3’ deflector at elevation 1468.0’ with a 0o lip angle, a performance curve was generated (Figure A-1, Appendix A). It was clear from this analysis that small spill discharges occurring with three powerhouse units at full capacity fall into the surface jump flow regime. Larger spill flows occurring with only one powerhouse unit at full capacity enter the plunging flow classification. These results demonstrated the need for further study because the typical range of operations and conditions was difficult to meet within the surface jet flow regime.
The same 14.3’ deflector was tested at elevation 1464.0’. This elevation was chosen to provide a lower boundary for the deflector design. A full performance curve was generated for at this elevation with 0o lip angle (Figure A-2, Appendix A). Similar to the previous elevation of 1468.0’, the 14.3’ deflector at elevation 1464.0’ exhibited surface jump flow characteristics at small spill discharges under normal operating conditions. Improvement was seen for higher spill discharges under operating conditions for one or no powerhouse units at full capacity. These conditions no longer produced plunging flow for the 14.3’ deflector at elevation 1464.0’. The evidence of surface jump flow for small spill discharges was again cause for further deflector development to expand the range of performance.
Analysis of these two preliminary performance curves revealed that the flow regimes could be adjusted relative to the tailwater elevations by altering the deflector elevation. This principle formed the basis for looking at deflector designs with a lip angle. If the flow regimes could be shifted at a slight angle relative to the tailwater curves, the normal operating conditions would fall directly in the surface jet category. A few brief experiments were performed with a 17.3’ deflector at elevation 1468.0’ with various lip
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 9 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
angles up to 45o. A longer deflector was chosen for these tests so that plunging flow would not occur at lower tailwater elevations.
The concise investigation of the 17.3’ deflector with various lip angles led to the development of an 18.3’ deflector at elevation 1468.0’ with a 0o, 5o, 10o, 15o lip angle. Each of these deflector designs was explored; however, the 5o, 10o, and 15o lip angles changed the flow too drastically. Not enough energy was dissipated to prevent the jet from reaching or surpassing the end sill of the stilling basin. The 18.3’ deflector at elevation 1468.0’ with a 0o lip angle was analyzed for a full performance curve (Figure A-3, Appendix A). This deflector exhibited improved performance for larger spill discharges when compared to the 14.3’ deflector at elevation 1468.0’, while nearly staying within the surface jet flow regime for small spill discharges.
The similarity in performance between the 14.3’ and 18.3’ deflectors at elevation 1468.0’ provided reasoning for lip angles to be tested for the 14.3’ deflector. Although the lip angles brought about extreme changes in the flow regimes for the 18.3’ deflector, it was postulated that the same lip angles would produce more reasonable variations in the flow regimes using the 14.3’ deflector. Lip angles of 5o and 10o were examined for the 14.3’ deflector at elevation 1468.0’. The advantage of these deflectors over the 14.3’ deflector at elevation 1468.0’ with no lip angle is not identifiable through performance curves. Although the desired flows remained in the surface jet flow regime, the performance of the deflectors for operations within the surface jet regime was much cleaner and smoother with the inclusion of the lip angles. Less aeration occurred deep in the basin, and the jet remained closer to the surface than with no lip angle. However, the performance was still bordering plunging flow for higher spill discharges at tailwaters corresponding to one and two powerhouse units at full capacity.
Examination of these results led to extending the deflectors to a length between 14.3’ and 18.3’ to remedy the problem of plunging flow for higher spill discharges. Good results were obtained for these flows with the 18.3’ deflector, but the lip angles performed poorly. A deflector length of 16.0’ was proposed at elevation 1468.0’ with lip angles of 0o, 5o, and 10o. Basic testing of these three designs revealed the 5o lip to provide the best performance. A full performance curve was developed for this deflector as shown in Figure 8.
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 10 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
SURFACE JUMP
SURFACE JET
VENTED SURFACE JET PLUNGING FLOW
Figure 8. Performance Curve for 16.0’ Deflector with 5o Lip at Elevation 1468.0’
The normal tailwater conditions for 1, 2, and 3 powerhouse units at full capacity with spill discharges ranging from 2.5 kcfs to 15 kcfs per sluiceway bay all remain within the desired surface jet flow regime. The lip angle of 5o also provides a very smooth, clean flow within the surface jet flow regime. Although tailwater conditions under zero powerhouse flow fall outside of the desired flow regimes, this type of operating condition is unlikely to occur in the field. A sketch of the 16.0’ deflector at elevation 1468.0’ with a 5o lip and 15’ transition radius was shown previously in Figure 7. Figures showing the performance of this deflector at tailwater conditions for three powerhouse units and no powerhouse units for spill discharges of 2.5, 5.0, 7.5, 10.0, 12.5, and 15.0 kcfs per sluiceway bay are included in Appendix B.
Deflector Design Summary
This model study demonstrated that the 16.0’ deflector at elevation 1468.0’ with a 5o lip angle had the highest potential for minimizing TDG. The tailwater curves for normal conditions with 1, 2, and 3 powerhouse units in operation remain in the surface jet flow regime for total discharges of 30 kcfs through the sluiceways. This performance satisfied
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 11 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
the design criteria of passing 98% of the flow conditions within the surface jet flow regime. The potential for cavitation damage was minimized with a deflector length of 16.0’. The sluiceway deflector design also provides flexibility in operating the upper spillway gates for total spill discharges exceeding 30 kcfs.
Velocity Profile Measurements
Procedure and Setup
Flow deflectors dramatically change the velocity profile of the river downstream from the dam because of the loss of energy dissipation in the stilling basin. The current model does not include 3-D effects of the deflectors on velocity. However, the present 2-D model is still useful in distinguishing relative differences in velocity profiles with and without deflectors.
Velocity measurements in the 1:48 scale model were made using a SonTek Acoustic Doppler Velocimeter (ADV) with a side-looking probe. Three cross sections were positioned at 436’, 686’, and 936’ downstream from the end sill of the dam. At each of these three transects, the velocities were measured at five profiles. The left-most and right-most profiles were each 21.6’ prototype from the left bank and right bank (when looking downstream), respectively. The left bank wall of the 1:48 scale model flume is built parallel to the left guide wall of the spillway. The right bank wall is 216’ prototype from the left bank wall in the model. The three remaining profiles at each cross section were equally spaced between the left-most and right-most profiles resulting in a spacing of 43.2’ prototype. Velocity measurements were taken at 0.2, 0.4, 0.6, and 0.8 of the flow depth at all five profiles of each cross section. This procedure was performed for spill discharges of 2.5, 5.0, 7.5, 10.0, 12.5, and 15.0 kcfs per sluiceway bay with the final deflector design of 16.0’ with a 5o lip angle. Baseline velocity measurements were also taken without the deflectors for spill discharges of 5.0, 10.0, and 15.0 kcfs per sluiceway bay. All tailwater elevations were set for the three-powerhouse unit operation of 30 kcfs.
Results
The velocity profiles for the downstream (x-direction) velocity were plotted as shown in Appendix C. The lower baseline condition with no deflectors in place results in a fairly uniform velocity profile for a spill discharge of 5.0 kcfs per sluiceway bay. Spill discharges of 10.0 and 15.0 kcfs per bay exhibited a small amount of recirculation (-1 to –2 ft/s) along the right bank wall with most of the higher positive flows (+8 to +9 ft/s) occurring on the left bank side primarily toward the surface.
Velocity profiles for conditions with the recommended 16.0’ deflector at elevation 1468.0’ with a 5o lip angle were noticeably different from the baseline results. The main discrepancy between the two conditions is the magnitude of the velocities and the amount of recirculation. Positive velocities as high as 19 ft/s and negative velocities up to –10
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 12 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
ft/s were observed with the deflectors in place. The condition for a spill discharge of 2.5 kcfs per bay did not seem to exhibit the expected recirculation along the right bank. The spill discharge of 5.0 kcfs per bay displayed a rather uniform flow profile near the surface with a maximum velocity near +6.0 ft/s. Signs of return flow near the bed of the channel on the order of –1.5 ft/s were also observed at this discharge. A noticeable recirculation began to occur along the right bank of the channel for spill discharges of 7.5 kcfs per bay and higher. This return flow varied between –8 and –10 ft/s for these discharges in the cross section closest to the end sill. Similarly, the left portion of the first cross section exhibited a steady increase in positive velocity from +13 to +19 ft/s for discharges of 7.5 to 15.0 kcfs per bay.
Most of the variance in velocity profiles occurs in the cross section nearest the end sill. The recirculation begins to diminish as the flow moves further downstream. Small amounts of return flow can still be detected along the right bank of the second and third cross sections. Negative velocities up to –4 and –5 ft/s were recorded in these areas for spill discharges ranging from 10.0 to 15.0 kcfs per bay. Meanwhile the left portion of the channel displays a rather uniform velocity profile in the second and third cross sections.
These initial velocity measurements provide a basic understanding of how the recommended deflectors change the flow patterns below Hells Canyon Dam. Some of the general patterns were present in the tailrace with and without deflectors. However, the magnitude of the velocities with deflectors proved to be significantly higher than without deflectors, especially near the water surface. The flow regimes produced by the deflectors shifts the location of energy dissipation from deep in the stilling basin toward the water surface and farther downstream from the dam. This displacement of energy dissipation, coupled with pockets of recirculation, is the primary concern of the changes in velocity profiles produced by the recommended deflector. These modifications of flow characteristics suggest the need for a more extensive investigation into the impact of the recommended deflectors on bankline erosion and bed scour.
Closure
The 1:48 scale model study performed for Hells Canyon Dam has produced a recommended deflector configuration that meets the specified design goals. The 16.0’ deflector at 1468.0’ with a 5o lip angle has the highest potential for minimizing total dissolved gas for sluiceway discharges up to 15.0 kcfs per bay, the 98% flow exceedence level. This deflector maintains desirable flow performance under typical operating conditions with 1, 2, and 3 powerhouse units. Additionally, the recommended sluiceway deflectors retain the operational flexibility to pass high spill flows through the upper spill gates while dissipating enough energy to maintain the structural integrity of the dam. However, preliminary velocity measurements emphasize the need to further explore the effects these flow deflectors might have on general flow patterns, bed scouring, and bankline erosion. Therefore, it will be necessary to investigate the recommended deflector’s performance in a 3-D model.
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 13 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Please feel free to contact us at your convenience with any questions or comments.
Sincerely,
Larry Weber, Ph.D., P.E. Paul B. Dierking Research Engineer Graduate Research Assistant
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 14 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
References
Nielsen, Kevin D., Larry J. Weber, and Pete Haug. “Hydraulic Model Study for Fish Diversion at Wanapum / Priest Rapids Development. Part XVI: 1:32.5 Scale Sectional Model of Wanapum Dam Spillway Deflectors.” IIHR Limited Distribution Report No. 284. University of Iowa, March 2000.
United States Army Corps of Engineers. Dissolved Gas Abatement Study – Phase II 60% Draft Report. CD-ROM. March 1999.
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 15 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Appendix A: Performance Curves
SURFACE JUMP
SURFACE JET
PLUNGING FLOW
VENTED SURFACE JET
Figure A-1. Performance Curve for 14.3’ Deflector with 0o Lip at Elevation 1468.0’
1496.0 1494.0
1492.0
1490.0
1488.0 SURFACE JUMP 1486.0
1484.0
1482.0
1480.0 ) . v e
l 1478.0 E ( r
e 1476.0 t
a SURFACE JE T w
l 1474.0 i a T 1472.0
1470.0
1468.0
1466.0 VENTED PLUNGING FLOW 1464.0 SURFACE JET 1462.0
1460.0
1458.0
1456.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 Discharge Per Sluiceway Bay (kcfs)
Figure A-2. Performance Curve for 14.3’ Deflector with 0o Lip at Elevation 1464.0’
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 16 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
1500.0 1498.0 1496.0 1494.0
1492.0 SURFACE JUMP 1490.0 1488.0 1486.0 1484.0 ) . SURFACE JET v e
l 1482.0 E (
r 1480.0 e t a
w 1478.0 l i a
T 1476.0
1474.0 1472.0 1470.0 VENTED 1468.0 SURFACE JET PLUNGING FLOW 1466.0 1464.0
1462.0
1460.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 Discharge Per Sluiceway Bay (kcfs)
Figure A-3. Performance Curve for 18.3’ Deflector with 0o Lip at Elevation 1468.0’
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 17 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Appendix B: Photos of 16.0’ Deflector with 5o Lip Angle
Figure B-1. Q = 2.5 kcfs Per Sluiceway Bay with No Powerhouse Units in Operation
Figure B-2. Q = 2.5 kcfs Per Sluiceway Bay with Three Powerhouse Units in Operation
Figure B-3. Q = 5.0 kcfs Per Sluiceway Bay with No Powerhouse Units in Operation
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 18 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Figure B-4. Q = 5.0 kcfs Per Sluiceway Bay with Three Powerhouse Units in Operation
Figure B-5. Q = 7.5 kcfs Per Sluiceway Bay with No Powerhouse Units in Operation
Figure B-6. Q = 7.5 kcfs Per Sluiceway Bay with Three Powerhouse Units in Operation
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 19 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Figure B-7. Q = 10.0 kcfs Per Sluiceway Bay with No Powerhouse Units in Operation
Figure B-8. Q = 10.0 kcfs Per Sluiceway Bay with Three Powerhouse Units in Operation
Figure B-9. Q = 12.5 kcfs Per Sluiceway Bay with No Powerhouse Units in Operation
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 20 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Figure B-10. Q = 12.5 kcfs Per Sluiceway Bay w/ Three Powerhouse Units in Operation
Figure B-11. Q = 15.0 kcfs Per Sluiceway Bay with No Powerhouse Units in Operation
Figure B-12. Q = 15.0 kcfs Per Sluiceway Bay w/ Three Powerhouse Units in Operation
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 21 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Appendix C: Velocity Profiles
Isometric View of X-direction Velocity Profiles for Hells Canyon Dam Vel-x 16.0' Deflector at Elevation 1468.0' with 5 Degree Lip 3.5 Q=2.5 kcfs per Sluice Bay with 3 P.H. Units 3 T.W.=1478.98 2.5 2 1.5 1 0.5 0 -0.5 -1 -1.5 -2
950 900 1490 850 1480 800 ) E t 1470 l (f l l e 750 Si v 1460 nd a 700 E
t 1450 m i ro o f 650 m n 1440 rea 200 600 st Z D wn is 150 Do tan 550 e ce nc fr 100 500 ta Y X om Dis R ig 50 450 ht Ba 0 400 nk (f t) Figure C-1. Velocity Profiles for Q = 2.5 kcfs Per Sluiceway Bay with Deflectors
Isometric View of X-direction Velocity Profiles for Hells Canyon Dam Vel-x 16.0' Deflector at Elevation 1468.0' with 5 Degree Lip 6 Q=5.0 kcfs per Sluice Bay with 3 P.H. Units 5.25 T.W.=1480.14 4.5 3.75 3 2.25 1.5 0.75 0 -0.75 -1.5
950 900 1490 850 1480 800 ) E t 1470 l ( f l l e 750 Si v 1460 nd a 700 E
t 1450 m i ro o f 650 m n 1440 rea 200 600 st Z D wn is 150 Do tan 550 e ce nc fr 100 500 ta Y X om Dis R ig 50 450 ht Ba 0 400 nk (ft)
Figure C-2. Velocity Profiles for Q = 5.0 kcfs Per Sluiceway Bay with Deflectors
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 22 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Isometric View of X-direction Velocity Profiles for Hells Canyon Dam Vel-x 16.0' Deflector at Elevation 1468.0' with 5 Degree Lip 13 Q=7.5 kcfs per Sluice Bay with 3 P.H. Units 11 T.W.=1481.23 9 7 5 3 1 -1 -3 -5 -7 -9
950 900 1490 850 1480 800 ) E t 1470 l ( f l l e 750 Si v 1460 nd a 700 E
t 1450 m i ro o f 650 m n 1440 rea 200 600 st Z D wn is 150 Do tan 550 e ce nc fr 100 500 ta Y X om Dis R ig 50 450 ht Ba 0 400 nk ( ft)
Figure C-3. Velocity Profiles for Q = 7.5 kcfs Per Sluiceway Bay with Deflectors
Isometric View of X-direction Velocity Profiles for Hells Canyon Dam Vel-x 16.0' Deflector at Elevation 1468.0' with 5 Degree Lip 14 Q=10.0 kcfs per Sluice Bay with 3 P.H. Units 12 T.W.=1482.24 10 8 6 4 2 0 -2 -4 -6 -8 -10
950 900 1490 850 1480 800 ) E t 1470 l ( f l l e 750 Si v 1460 nd a 700 E
t 1450 m i ro o f 650 m n 1440 rea 200 600 st Z D wn is 150 Do tan 550 e ce nc fr 100 500 ta Y X om Dis R ig 50 450 ht Ba 0 400 nk ( ft)
Figure C-4. Velocity Profiles for Q = 10.0 kcfs Per Sluiceway Bay with Deflectors
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 23 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Isometric View of X-direction Velocity Profiles for Hells Canyon Dam Vel-x 16.0' Deflector at Elevation 1468.0' with 5 Degree Lip 16 Q=12.5 kcfs per Sluice Bay with 3 P.H. Units 14 T.W.=1483.20 12 10 8 6 4 2 0 -2 -4 -6 -8
950 900 1490 850 1480 800 ) E t 1470 l ( f l l e 750 Si v 1460 nd a 700 E
t 1450 m i ro o f 650 m n 1440 rea 200 600 st Z D wn is 150 Do tan 550 e ce nc fr 100 500 ta Y X om Dis R ig 50 450 ht Ba 0 400 nk ( ft)
Figure C-5. Velocity Profiles for Q = 12.5 kcfs Per Sluiceway Bay with Deflectors
Isometric View of X-direction Velocity Profiles for Hells Canyon Dam Vel-x 16.0' Deflector at Elevation 1468.0' with 5 Degree Lip 19 Q=15.0 kcfs per Sluice Bay with 3 P.H. Units 16 T.W.=1484.09 13 10 7 4 1 -2 -5 -8
950 900 1490 850 1480 800 ) E t 1470 l ( f l l e 750 Si v 1460 nd a 700 E
t 1450 m i ro o f 650 m n 1440 rea 200 600 st Z D wn is 150 Do tan 550 e ce nc fr 100 500 ta Y X om Dis R ig 50 450 ht Ba 0 400 nk ( ft)
Figure C-6. Velocity Profiles for Q = 15.0 kcfs Per Sluiceway Bay with Deflectors
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 24 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Isometric View of X-direction Velocity Profiles for Hells Canyon Dam Vel-x Baseline Data with No Deflectors 1.6 Q=5.0 kcfs per Sluice Bay with 3 P.H. Units 1.4 T.W.=1480.14 1.2 1 0.8 0.6 0.4 0.2 0 -0.2
950 900 1490 850 1480 800 ) E t 1470 l ( f l l e 750 Si v 1460 nd a 700 E
t 1450 m i ro o f 650 m n 1440 rea 200 600 st Z D wn is 150 Do tan 550 e ce nc fr 100 500 ta Y X om Dis R ig 50 450 ht Ba 0 400 nk ( ft)
Figure C-7. Velocity Profiles for Q = 5.0 kcfs Per Sluiceway Bay with No Deflectors
Isometric View of X-direction Velocity Profiles for Hells Canyon Dam Vel-x Baseline Data with No Deflectors 8 Q=10.0 kcfs per Sluice Bay with 3 P.H. Units 7 T.W.=1482.24 6 5 4 3 2 1 0 -1 -2
950 900 1490 850 1480 800 ) E t 1470 l ( f l l e 750 Si v 1460 nd a 700 E
t 1450 m i ro o f 650 m n 1440 rea 200 600 st Z D wn is 150 Do tan 550 e ce nc fr 100 500 ta Y X om Dis R ig 50 450 ht Ba 0 400 nk ( ft) Figure C-8. Velocity Profiles for Q = 10.0 kcfs Per Sluiceway Bay with No Deflectors
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 25 THE UNIVERSITY OF IOWA Iowa Institute of Hydraulic Research
Isometric View of X-direction Velocity Profiles for Hells Canyon Dam Vel-x Baseline Data with No Deflectors 9 Q=15.0 kcfs per Sluice Bay with 3 P.H. Units 8 T.W.=1484.09 7 6 5 4 3 2 1 0 -1
950 900 1490 850 1480 800 ) E t 1470 l (f l l e 750 Si v 1460 nd a 700 E
t 1450 m i ro o f 650 m n 1440 rea 200 600 st Z D wn is 150 Do tan 550 e ce nc fr 100 500 ta Y X om Dis R ig 50 450 ht Ba 0 400 nk (ft) Figure C-9. Velocity Profiles for Q = 15.0 kcfs Per Sluiceway Bay with No Deflectors
Iowa Institute of Hydraulic Research Hells Canyon Dam Deflector Design 404 Hydraulics Lab October 13, 2000 Iowa City, IA 52242-1584 Page 26