Design Methods for In-Stream Flow Control Structures (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

58 What follows is a discussion on design considerations for in- stream structures, starting with current practice and expand- ing to include new design insights based on the results from the ISL, OSL, and VSL3D experiments. These results address three major components of structure installation: angle of orientation, location and spacing, and footer depth. The range of all stream characteristics (both physical and virtual) used in this project is detailed in Table 7-1. The recommendations in this report and the associated design guidelines are limited to sand and gravel streams and rivers (up to ~30 m in width) with low (2 to 3.3) Rc/B ratios or straight channels, although many of these guidelines can be applied with an appropri- ate evaluation of the baseline (pre-structure) hydraulics at the project site. 7.1 Rock Vanes/J-Hooks Rock vanes direct the faster portion of the flow toward the center of the channel and create quiescent flow conditions near the bank (Johnson et al., 2002a). A series of vanes in the stream-wise direction is required to create a secondary flow cell, or secondary circulation, which creates scour at the mid- dle of the channel while simultaneously backfilling the bank, effectively relocating erosive flow patterns induced by channel curvature (Johnson et al., 2001). Other structures are slightly modified RVs, including JHs, CVs, and WWs. A JH consists of a RV with additional boul- ders placed at the tip of the vane in a hooking pattern with gaps between them. This layout creates scour by forcing the flow in the center of the channel to converge between the gaps. While this modification is primarily to create habitat, it also provides energy dissipation (Harman et al., 2001; Rosgen, 2001). The hook portion of the structure provides a longer, deeper, and wider scour pool than that created by a rock vane only (Rosgen, 2001). RVs can reduce or eliminate the need for armor since high flows are redirected away from banks; additionally, they compound the effectiveness of vegetative restoration tech- niques (McCullah and Gray, 2005). Vanes have previously been studied and used in other forms such as submerged (Iowa) vanes and BWs. Odgaard and Kennedy (1983), Odgaard and Spoljaric (1986), Odgaard and Mosconi (1987), and Odgaard and Wang (1991) provide many examples of laboratory work and some case studies associated with sub- merged vanes. Odgaard and Kennedy (1983) and Odgaard and Wang (1991) suggest that similar hydraulic principles are present in RVs as in submerged vanes and that for a large river with deeper pools they would be an attractive solution. However, Odgaard and Wang (1991) stress that their design guidelines are not valid for RVs. When the results from the ISL, OSL, and VSL3D experi- ments are combined, a number of trends emerge. First, as the angle of orientation increases, the ratio of protected bank to channel width (B) increases (Figure 7-1). Second, as the angle of orientation increases, the ratio of maximum scour depth to channel depth (H) also tends to increase in the straight and VSL-S channel (Figure 7-2). This was not the case, however, in the VSL-G channel. It should also be noted that for higher angles of orientation, the location of the maximum scour hole is at the structure tip in the vicinity of high bed fluctuations due to passing bed forms, indicating a high potential for structure failure if the footer rocks are not deep enough (see the Zrms in Figures 3-7 and 3-8). For the 20° single RV case, there was no scour at the structure tip for VSL-S or VSL-G. Straight Channel Recommendations Velocity data from the ISL experiments (Appendix D) show that altered flow patterns and turbulence from the structures propagate far downstream of the structure location. A larger angle (30° instead of 20°) shifted the higher shear stresses farther away from the bank, thereby creating a longer region of low bank shear stress downstream of the structure along C H A P T E R 7 Compilation of Experimental and Numerical Results to Develop Design Guidelines

59 0 1 2 3 4 5 6 OSL 30 VSL G 20 VSL S 20 VSL G 30 VSL S 30 L Ba nk Pr ot ec o n/ B Lu/B Ld/B Figure 7-1. Angle of orientation of rock vanes compared to the ratio of downstream protected bank, Ld, and upstream protected bank, Lu, for OSL, VSL-G, and VSL-S with a single rock vane at the meander apex. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 OSL 30 ISL 20 ISL 30 VSL G 20 VSL G 30 VSL S 20 VSL S 30 Sc M AX /H Figure 7-2. Ratio of maximum scour, ScMAX, to average channel depth, H, for a single rock vane installed in OSL, ISL, VSL-G, and VSL-S channels at 20 or 30. Depth Width Transport Mode Aspect Ratio Median Grain Size Discharge Bed Slope Sinuosity Wavelength Radius of Curvature Rc/B H B B/H D50 Q So p Rc m m mm L/s m m ISL-fixed 0.17 0.91 Fixed 5 6 36 0.0007 1 – – – ISL-mobile 0.17 0.91 Clear water 5 2 55 0.0003 1 – – – OSL 0.3 2.7 Live bed 9 0.7 280 0.0007 1.3 25 5.7 2.11 VSL-S 1.35 27 Live bed 20 0.5 48400 0.0007 1.5 266.7 57.6 2.13 VSL-G 0.9 27 Live bed 30 30 36000 0.003 1.15 328.1 89.1 3.30 Table 7-1. Summary of channel parameters for indoor (ISL), outdoor (OSL), and numerical (VSL3D) experiments.

60 the bank to which it is attached; however, larger angles (30° instead of 20°) produced greater scour depths, which could also create greater risk of structure undermining and failure. All RV angles provided net erosion within the zone one chan- nel width downstream. These structures tended to result in net deposition in zones that were two or more channel widths downstream of their location. In a straight channel, JHs produced 25% less scour than the RV alone. The J-hook produced a single deep scour hole just downstream of the hook portion of the vane that extended approximately two channel widths downstream of the struc- ture. This scour hole was wider and longer than the scour hole produced for the 30° RV. The 30° JH moved the high shear zone further from the bank toward the center of the channel; however, the bank protection provided by each structure was similar. Based on the ISL straight channel experiments, the following recommendations are made for RV and JH instal- lation in straight channels. • Monitoring should be conducted well downstream of the structure (>5 channel widths) because the flow structure could initiate scour and erosion of the opposite bank. • To protect the bank where structures are to be attached, a larger angle (30°) is recommended as it provides a greater length of bank protection. • Structures should be installed two or more channel widths upstream of the infrastructure they are designed to protect. • Larger footer rocks should be installed with greater angle of orientation. • Footer rocks should be sized appropriately at positions of large scour depth and large hydrodynamic forces (e.g., at the structure tip). • JHs are appropriate in projects where a larger area of scour hole is deemed beneficial to in-stream habitat. • In low aspect ratio streams, the top slope of RV and JH structures may need to be adjusted to ensure that the struc- ture slopes from the top of the bank to the channel bed. Meandering Channel Recommendations Observations from the OSL (Appendix D) and VSL3D (Chapter 3) meandering channel experiments indicated that RVs were successful at protecting a portion of the outer bank from erosive effects due to the 3-D helical flow in meander bends. A single RV structure was not able to protect the entire bank due to the extension of the shear layer from the tip of the RV. With multiple RV structures, the scour hole at the tip of the downstream RV can be significant. RVs downstream of the apex experienced significant scour around the tip of the structure, while RVs installed upstream of the apex expe- rienced very little scour but were subject to deposition. Con- versely, the most upstream RV structure was the most likely to experience excessive deposition and consequential flanking. RVs were successful at moving the high-velocity core away from the outer bank; however, in the OSL experiments, sig- nificant local scour was experienced at the downstream RV. RVs placed at the apex of the meander and downstream effec- tively moved the region of high shear stress away from the outer bank. In meandering channels, the installation of JHs resulted in both deeper and larger scour holes. JHs were successful at moving the high-velocity core from the outer banks, but in some structure scenarios, threatened the stability of the inner bank. JHs protected a longer zone of the outer bank, but this was not consistent since all VSL3D cases exhibited scour adjacent to the outer bank just downstream of the structure and before the protected area of bank. Based on the OSL and VSL3D experiments, the following recommendations are made for JH and RV structure installation in meandering channels. • When used for meander bend protection, RVs and JHs should be installed in series of at least two structures to protect the entire region of the outer bank subject to scour. • Footer rocks should be sized appropriately at positions of large scour depth and large hydrodynamic forces (e.g., at the structure tip). • For RVs, the structure array should begin at the mean- der apex in less sinuous channels, but for channels with low radius of curvature to channel width, RV and J-hook arrays should be shifted upstream so that the first struc- ture is located approximately one channel width upstream of the apex. For JHs, structure arrays should be shifted upstream by one channel width. • In highly sinuous channels, RV spacing should be decreased. A vector analysis with an appropriate offset angle is sug- gested for determining structure spacing. • JHs are appropriate in projects where a larger area of scour hole is deemed beneficial to in-stream habitat. • RVs and JHs pose a high risk of flanking due to deposition upstream of the structure. Bank keys are required to miti- gate the risk of flanking, and monitoring should evaluate the sediment deposition upstream of each structure. 7.2 Bendway Weirs/Stream Barbs Stream barbs and BWs interrupt the helical flow patterns of secondary currents typically associated with channel mean- ders. The presence of these in-stream structures relocates the erosive flow patterns from the vulnerable outer bank toward the center of the channel (Derrick, 1998). Stream barbs spe- cifically protect the bank by disrupting velocity gradients in near-bed regions, deflecting currents away from the bank

61 by forcing flow perpendicularly over the weir and shifting the channel thalweg to the stream-wise end of the barbs (Matsuura and Townsend, 2004). An effective series of barbs will induce a subcritical zone of backwater, which should reach the next upstream barb. This upstream progression of subcritical reaches controls erosion and eventually leads to sediment deposition in the near-bank region (NRCS, 2013). Bendway weirs are primarily used along meanders in larger rivers and tend to work best in high-flow, high-energy conditions but have been observed to function effectively in low-flow events (Derrick, 1998; Abad et al., 2008). When the results from the ISL and OSL (Appendix D) and VSL3D (Chapter 3) experiments are combined, BWs, which are submerged at most flows, behaved fundamen- tally differently than RVs. First, as the angle of orientation increases, the ratio of protected bank to channel width (B) decreases. For BWs in VSL-G and VSL-S, smaller angles resulted in greater bank protection, with one exception for the 60° structure in VSL-S. The scour depth was inde- pendent of the angle of orientation; however, scour depth in the vicinity of the structure decreased as aspect ratio increased (Figure 7-3). Straight Channel Recommendations Velocity data from the ISL experiments show that altered flow patterns and turbulence from the structures propagate far downstream of the structure location. As the angle of ori- entation increased, the scour depth and scour hole dimensions decreased, with no visible difference in bank protection. In all cases, scour around the structure extended to the upstream side of the structure tip, eventually exposing the footer rocks. Based on the ISL straight channel experiments, the following recommendations are made for BW structure installation in straight channels. • Monitoring well downstream (>5 channel widths) of the structure is important because the flow structure could still initiate scour and erosion of the opposite bank. • In straight channels, a BW with an angle of orientation of 80° will provide bank protection for less material cost because of the shorter structure length. • Bendway weirs are not recommended in low aspect ratio streams. If they are installed in low aspect ratio (<6) streams, footer rock depths need to be increased since significant scour is expected near the structure. Meandering Channel Recommendations Use of a single BW structure at the apex of an outer bend can protect a portion of the outer bank from erosive effect, although, to fully protect an eroding meander, multiple struc- tures will be required. With multiple BW structures, the scour hole at the tip of the downstream structure can be significant. Similar to RVs, BWs downstream of the apex experienced significant scour around the tip of the structure, while BWs installed upstream experienced very little scour but were sub- jected to deposition. In the OSL experiments, less local scour was experienced at the lower BW than at the lower RV. Scour depth at the BW tip was found to be correlated to channel aspect ratio. Larger aspect ratios had less scour. The practitio- ner survey indicated that the single reported unsuccessful BW project was in a low aspect ratio channel, although this was true of all rock structures considered (Radspinner et al., 2010). Based on the OSL and VSL3D experiments, the following Sc/H = 0.0298(B/H) + 1.0247 R² = 0.9329 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 5 10 15 20 25 30 35 Sc ou rD ep th /H Aspect Rao, B/H Figure 7-3. Scour depth in the vicinity of BWs normalized by channel depth (H) as a function of aspect ratio (B/H). All results from ISL, OSL, and VSL3D experiments have been superimposed.

62 recommendations are made for BW structure installation in meandering channels. • To provide bank protection for a full meander bend, at least three BWs should be used. • Footer rock depth should be similar to RVs. • Bendway weirs are less susceptible to flanking from over- bank flows, but structure layout should be such that the bank between and above structures should be protected at high flows. • Bendway weirs are not recommended in low aspect ratio streams. If they are installed in low aspect ratio (<6) streams, footer rock depths need to be increased since significant scour is expected near the structure. 7.3 Sill Structures (Cross Vanes/W-Weirs) CVs and WWs are variations of the RV but have a primary function as grade-control structures while still providing a reduction in near-bank shear stress. The structure raises or maintains the bed elevation, so it is normally installed within a section of little or no turbulence for larger streams or at the head of a riffle for smaller streams (Doll et al., 2003). CVs are constructed by connecting the tips of two RVs from oppo- site banks with rocks arranged perpendicular to the flow. The CV is expected to create a variety of habitat in the channel by increasing flow diversity and substrate complexity (McCullah and Gray, 2005). The vane sections raise the water level in the near-bank area, providing more cover, the scour pool allows a holding area in high- and low-flow periods, flow separation zones become feeding lines, and the downstream end of the scour pool is ideal for spawning beds. Like CVs, the WW is primarily a grade-control structure that facilitates scour pool habitat. WWs are normally only used in larger rivers since they are relatively large structures. The layout of this structure is a “W” shape looking downstream. This arrangement creates dual thalwegs and, therefore, can lead to more flow diversity. Cross vanes were tested in the ISL and OSL (Appendix D) and VSL (Chapter 3). Because they were installed in straight or crossover reaches in each channel, the recommendations are not divided between meandering and straight channels. Two configurations of cross vanes, a standard U-shaped CV, and an A-shaped CVA, were evaluated. The deepest scour was within the structure for the U-shaped CV, and immediately downstream of the cross piece of the CVA structure. WWs were tested in the ISL and VSL. WWs were not installed in the OSL because the channel was too small. Because WWs were installed in straight or crossover reaches in each chan- nel, the recommendations are not divided between meander- ing and straight channels. The deepest scour was within the structure and immediately downstream. At the early stages of scour, WWs provide some protection from scour down- stream of the middle point of the W; however, as the system progresses, this region also scours to a single large scour hole downstream. The results of the physical and numerical experiments led to the following recommendations for CVs and WWs. • For bank protection, an angle of orientation of 30° should be used instead of 20°. • A stepped CVA should be used to provide additional pro- tection for the upstream rocks; however, an A-shaped structure poses a greater risk to the stability of banks imme- diately downstream and should not be used if bank protec- tion is the project goal. • Monitoring should be focused within and immediately downstream of a CV or CVA (within one channel width) for structure and bank stability. All sill structures showed a risk to the banks in the vicinity of the scour hole in this region. • Sill structures should be installed a minimum of two chan- nel widths upstream of bridge piers or other infrastructure to be protected. For CVAs, this should be extended to 2.5 to 3 channel widths. An alternative would be to install struc- tures in series to minimize the dimensions of the scour hole immediately downstream of a CV. • WWs will provide short-term scour protection mid- channel but should not be used for long-term scour pro- tection for bridge piers and other structures installed in the middle of the river channel. • Sill structures were successful at creating large scour pools downstream of each structure; however, if the project goal includes fish habitat, sill structures need to be evaluated to ensure that the drop height created by the structure does not create a fish passage barrier. • Sill-type structures created a moderate risk of flanking due to deposition upstream of the structure. This area needs to be monitored to minimize flanking risk. • Undermining of the structure is expected to be the pri- mary failure mechanism due to the growth of the large scour pool; footer structures should be 2 to 3 times deeper than the maximum scour present in the structure-free channel. 7.4 Structure Selection Following the detailed evaluation of experimental and numerical results described previously for five structure types (with an additional CV modification, CVA), Table 7-2 com- piles these recommendations on structure selection based on site characteristics and project goals. This is a modification of the summary tables presented in the MWCG (Maryland Department of the Environment, 2000).

63 Three major categories of structure failure mechanism were identified: flanking (circumvention combined with aggradation), local scour (which can lead to rock displacement), and rock dis- placement (based on hydrodynamic forces, not local scour). The practitioner survey also reported winnowing (or scour between rocks) as a significant failure mode. This failure mechanism was not evaluated in NCHRP Project 24-33. Table 7-3 compiles the evaluation of the susceptibility to these failure mechanisms for each structure type. Recommendations based on the systematic use of VSL3D to evaluate a range of structure types in typical sand and gravel channels were used to develop updated design guidelines, which are presented in Appendix F. These recom- mendations are specific to non-cohesive sand or gravel channels with subcritical flow and relatively immobile banks compared to the bed sediment. Structure performance was evaluated at bankfull flow conditions for channels in the ranges of aspect ratio 5 to 30, sinuosity 1 to 1.5, slope 10-4 to 10-3, and grain size 0.5 to 30 mm. Each project should be evaluated based on site-specific characteristics. Table 7-2. Recommendations for in-stream flow control structure selection based on site characteristics and project goals. Site Characteristics Project Goals Steep Streams Low Aspect Ratio (Low B/H) Tight Meanders (Low Rc/B) Fine Bed Material (Sand) Protect Bank Toe Redirect Flow (Central Thalweg) Scour Pool Habitat Minimize Lateral Channel Adjustments Grade Control RV o o o o o o – o – JH o o o o o o + o o BW – – – o + + – + – CV + o o o – o + – + CVA + – o – – o + – + WW + – o – – o + – + – Not appropriate o Moderately appropriate with design modifications + Well suited Table 7-3. Evaluation of susceptibility to failure for in-stream flow control structures. Susceptibility to Failure Mechanism Local Scour Flanking Rock Displacement RV o + o JH + o + BW + – – CV + o o CVA + o + WW + o + – Low risk o Moderate risk + High risk See Section 5.2 for discussion on local scour. See Section 5.3 for discussion of flanking. See Section 5.4 for discussion of force distribution.