مؤشرات

ABSTRACT
PUCKETT, PAIGE ROLLINS. The Rock Cross Vane: A Comprehensive Study of an

In-Stream Structure. (Under the direction of Gregory D. Jennings.)
The rock cross vane, an in-stream boulder structure, consists of a U-shaped weir
with the apex upstream at bed elevation and upward sloping arms that tie into
downstream banks. The structure provides grade control, bank protection and scour
pool development via a protected drop and arms that turn flow away from the banks.
A physical model was used to measure the velocity distribution changes caused by a
range of geometric configurations of the structure. Results showed linear and
quadratic effects of drop ratio, cross product effects of drop ratio and arm angle,
drop ratio and arm slope, and arm angle and arm slope on the velocity ratio (the
average center velocity to the average outer velocity) of a cross section located two
bankfull widths downstream of the drop. The Rock Cross Vane Rapid Assessment
Tool (RCV-RAT) was developed to rate rock cross vane failures and assess the
causes of structure weakness on NC stream restoration projects. Failure ratings
were analyzed for significant correlations to project parameters, class effects of
project, and correlations between dependent variables (alpha =0.05). Of the 120
rock cross vanes observed, 109 had at least one incidence of failure ranging from
slight to extreme. From this data, an FMEA was developed to compare the risks of
the various modes of failure for the rock cross vane and to adapt the FMEA for use
in North Carolina stream projects. It was concluded that sufficient drop is necessary
for scour pool development, while steep, narrow arms leading to a constricted flow
area minimally contribute to scour pool development and may risk bank instability at
the structure and at downstream banks. Rock cross vanes that can tolerate rapid
lateral migration of the stream post-construction should be developed and tested
due to the observed problems of structure side cutting. The final chapter reviews
literature for the main effects of stream restoration on benthic macroinvertebrate
communities and the expected return of benthic macroinvertebrate communities
post-restoration to address what impacts the rock cross vane might have on benthic
macroinvertebrates and to recommend practices that might reduce these impacts.
THE ROCK CROSS VANE:
A COMPREHENSIVE STUDY OF AN IN-STREAM STRUCTURE
by
PAIGE ROLLINS PUCKETT
A dissertation submitted to the Graduate Faculty of

North Carolina State University
in partial fulfillment of the
requirements for the Degree of
Doctor of Philosophy
BIOLOGICAL AND AGRICULTURAL ENGINEERING
Raleigh, North Carolina

2007
APPROVED BY:
Dr. Gregory D. Jennings

Chair of Advisory Committee
Dr. Michael D. Boyette

Dr. Garry L. Grabow
Dr. James D. Gregory
Dedication
For Joe.
What has been will be again, what has been done will be done again;
there is nothing new under the sun.
Ecclesiastes 1:9
ii
Biography
Paige Rollins Puckett was born in Chattanooga, Tennessee on December 4, 1980.

Around the age of 8, her family moved to the country where most summer days were
spent playing in the creek in the back field. She intended on going to medical school,
but in the fall of 1999 she began her undergraduate degree in Biological Engineering
at North Carolina State University as a Park Scholar. It was then she discovered she
could continue to play in creeks as long as she desired and began to develop an
interest in stream restoration. In May of 2003 she graduated and stayed in the
department to begin her graduate work under the direction of Dr. Greg Jennings.
She will finish the program with her doctoral degree in May 2007.
iii
Acknowledgements
This research was funded by the National Science Foundation Graduate Research
Fellowship program and Department of Biological and Agricultural Engineering at
North Carolina State University.
I would also like to recognize:
Joe Puckett, Greg Jennings, Mike Boyette, Garry Grabow, Jim Gregory, John
Parsons, Rich McLaughlin and the team that conducts research at the Lake Wheeler
Land Application and Demonstration Center, Angela Gardner, Will Clayton, James
Puckett, Sarah Payonk, Mike Shaffer, David Godly, Robert Langlager, Jess Roberts,
Desiree Tullos, Dan Clinton, Barbara Doll, Dave Bidelspach, and the Rollins and
Puckett families.
iv
Table of Contents
List of Tables.......................................................................................................... vii
List of Figures .......................................................................................................... x
Chapter 1:
History of the use of hydraulic structures related to the rock cross vane ......... 1
Literature Cited ...................................................................................................... 21
Chapter 2:
Effects of Rock Cross Vane Geometry on Velocity Distribution ....................... 24
Literature Cited ...................................................................................................... 64
Chapter 3:
Rock Cross Vane Rapid Assessment Tool and Failure Guidebook .................. 65
Literature Cited .................................................................................................... 107
Chapter 4:
Occurrence, Probability and Risk of Rock Cross Vane Failures in North
Carolina Stream Restoration Projects ............................................................... 108
Chapter 5:
Restoration and Cross Vane Impacts on Benthos............................................ 170
Chapter 6:
Summary Conclusions and Future Research.................................................... 192
v
Appendix A:
Tables and Figures .............................................................................................. 204
Appendix B
Rock Cross Vane Qualitative Assessment Data ............................................... 212
vi
List of Figures
Chapter 1
Figure 1.1 Profile View of Cross Vane Arm............................................................................................ 2
Figure 1.2 Plan View of Rock Cross Vane ............................................................................................. 3
Figure 1.3 Salmon weir at Quamichan Village on the Cowichan River, Vancouver Island.................... 5
Figure 1.4 Stone Fish Weir on Menai Strait ........................................................................................... 6
Figure 1.5 Chester Weir on the Dee River ........................................................................................... 10
Figure 1.6 Plan View of Rock Vortex Weir ........................................................................................... 12
Figure 1.7 Plan View of Rootwad Revetment....................................................................................... 13
Figure 1.8 Bendway Weir Theory ......................................................................................................... 14
Figure 1.9 Installation of Concrete Iowa Vanes at Low Flow ............................................................... 16
Figure 1.10 Plan View of Rock Vane and J-Hook ................................................................................ 17
Figure 1.11 Plan Views of Three Types of Long Crested Weirs .......................................................... 17
Figure 1.12 Formation of a Waterfall .................................................................................................... 19
Chapter 2
Figure 2.1 Plan View of the Velocity Distributions at Rock Cross Vane............................................... 28
Figure 2.2 Diagram of a Step in a Channel .......................................................................................... 29
Figure 2.3 Hydraulic Jump.................................................................................................................... 31
Figure 2.4 Length of Jump for a Rectangular Channel ........................................................................ 32
Figure 2.5 Site Layout .......................................................................................................................... 34
Figure 2.6 Observed Stage vs. Time of Storage Pond Drawdown ...................................................... 36
Figure 2.7 Plan View of Test Flume ..................................................................................................... 38
Figure 2.8 Profile View of Rock Cross Vane Model ............................................................................. 40
Figure 2.9 Profile View of Flume Stilling Basin and Baffle Outlet – Baffle Combination ...................... 41
Figure 2.10 Baffle Oulet – Baffle Combination ..................................................................................... 41
Figure 2.11 Test Run with the Pygmy Meter and AquaCount Calculator............................................. 42
Figure 2.12 Flume test of drop ratio = 1, angle = 20˚, slope = 3%....................................................... 43
Figure 2.13 Predicted and Actual Velocity Ratio for Drop, Angle, and Slope for Cross Section C...... 50
Figure 2.14 Predicted Velocity Ratio versus Arm Angle and Drop Ratio for Cross Section C............. 51
vii
Figure 2.15 Predicted Velocity Ratio versus Drop Ratio and Arm Slope for Cross Section C............. 52
Figure 2.16 Residuals at Cross Section A............................................................................................ 53
Figure 2.17 Residuals at Cross Section B............................................................................................ 54
Figure 2.18 Residuals at Cross Section C ........................................................................................... 55
Chapter 3
Figure 3.1 Arm Washout....................................................................................................................... 76
Figure 3.2 Cross vane with a portion of the sill shifted or missing ....................................................... 77
Figure 3.3 Head cut at rock cross vane looking downstream .............................................................. 78
Figure 3.4 Bank erosion at cross vane leading to side cutting and headcut........................................ 80
Figure 3.5 Flow Expansion with Downstream Erosion ......................................................................... 81
Figure 3.6 Shallow Pool with Vegetation.............................................................................................. 82
Figure 3.7 Diagram of Drag and Lift ..................................................................................................... 84
Figure 3.8 Constricted Rock Cross Vane ............................................................................................. 86
Figure 3.9 Piping caused by boulder spacing and poor backfill ........................................................... 87
Figure 3.10 Cross vane with improper alignment................................................................................. 91
Figure 3.11 Cross vane backed into a pool.......................................................................................... 91
Figure 3.12 Undersized Cross Vane in Bend ....................................................................................... 92
Figure 3.13 Vane on Bedrock............................................................................................................... 93
Figure 3.14 Downstream scour from short steep arms ........................................................................ 94
Figure 3.15 Steep Arm on Right Bank.................................................................................................. 95
Figure 3.16 Cross vane sill installed too high....................................................................................... 96
Figure 3.17 Sill and backfill shifted downstream to pool ...................................................................... 97
Figure 3.18 Poor Boulder Spacing ....................................................................................................... 97
Figure 3.19 Boulders placed in pool to protect drop............................................................................. 98
Figure 3.20 Undersized Boulders ......................................................................................................... 99
Figure 3.21 Lack of vegetative bank protection.................................................................................. 100
Figure 3.22 Evidence of piping caused by insufficient backfill ........................................................... 101
Figure 3.23 Undercut Banks at Arms from insufficient footers........................................................... 101
Figure 3.24 Oversized rock cross vane.............................................................................................. 102
Figure 3.25 Bank erosion and head cut caused by an undersized vane ........................................... 103
Chapter 4
Figure 4.1 Boulder of a Cross Sill in Assumed Conditions................................................................. 115
Figure 4.2 Free Body Diagram of Assumed Cross Sill Conditions..................................................... 115
Figure 4.3 Cross Sill with Downstream Pool ...................................................................................... 116
viii
Figure 4.4 Cross Sill with Head cut .................................................................................................... 117
Figure 4.5 Free Body Diagram of Cross Sill with Head cut................................................................ 117
Figure 4.6 Cross Still Prime Conditions for Tipping............................................................................ 118
Figure 4.7 Free Body Diagram for Tipping Conditions ....................................................................... 118
Figure 4.8 Projects by County ............................................................................................................ 125
Figure 4.9 Example Fault Tree for Sill Washout................................................................................. 132
Appendix A
Figure A-1 Profile View of Cross Vane Arm ....................................................................................... 206
Figure A-2 Plan View of Rock Cross Vane......................................................................................... 206
ix
List of Tables
Chapter 1
Table 1.1 Citations of Fish Weirs and Related Structures in North Carolina ......................................... 7
Chapter 2
Table 2.1 Pipe Elevation, Length, and Slope ....................................................................................... 35
Table 2.2 Time Discharge Calculation ................................................................................................. 35
Table 2.3 Predicted Normal Depth and Observed Depth..................................................................... 39
Table 2.4 Velocity Ratio for Three Test Runs at each Cross Section .................................................. 44
Table 2.5 Factor Levels and Values ..................................................................................................... 45
Table 2.6 Velocity Ratio for Cross Section by Test Run ...................................................................... 48
Table 7 LSMEANS for Class Effects of Cross Section......................................................................... 49
Table 2.8 Analysis of Variance for Response Surface Regression Model at Cross Section A............ 53
Table 2.9 Analysis of Variance for Response Surface Regression Model at Cross Section B............ 54
Table 2.10 Analysis of Variance for Response Surface Regression Model at Cross Section C ......... 55
Table 2.11 SAS output for Parameters of Response surface Regression Model at
Cross Section C....................................................................................................................... 56
Table 2.12 Flow Regime Calculations for Cross Section ..................................................................... 58
Table 2.13 Area Ratio at Critical Depth and Brink Depth for 7.2 cfs .................................................... 60
Chapter 3
Table 3.1 Rock Cross Vane Rapid Assessment Tool .......................................................................... 74
Table 3.2 Failures and Major Indicators ............................................................................................... 75
Table 3.3 Types of Side cutting ............................................................................................................ 88
Table 3.4 Potential Failure Paths ....................................................................................................... 103
Chapter 4
Table 4.1 Failure and Major Indicators............................................................................................... 122
Table 4.2 Projects and Site Data........................................................................................................ 124
Table 4.3 Occurrences of failure and failure paths............................................................................. 134
x
Table 4.4 Median rating of failure indicators by project...................................................................... 135
Table 4.5 Frequency of Failure Indicators.......................................................................................... 136
Table 4.6 Frequency of Primary Causes of Failure............................................................................ 138
Table 4.7 Frequency of Secondary Causes of Failure ....................................................................... 139
Table 4.8 Occurrences of Failure Paths............................................................................................. 141
Table 4.9 Correlation of Project Variable ........................................................................................... 143
Table 4.10 Effects of Project Variables on F1: Arm Washout ............................................................ 145
Table 4.11 Effects of Project Variables on F2: Sill Washout.............................................................. 146
Table 4.12 Effects of Project Variables on F3: Head Cut................................................................... 147
Table 4.13 Effects of Project Variables on F4: Bank Erosion at the Structure................................... 148
Table 4.14 Effects of Project Variables on F5: Downstream Bank Erosion ....................................... 149
Table 4.15 Effects of Project Variables on F6: Insufficient Scour Pool Development........................ 150
Table 4.16 Class Effects of Project on F1: Arm Washout .................................................................. 151
Table 4.17 Class Effects of Project on F2: Sill Washout .................................................................... 152
Table 4.18 Class Effects of Project on F3: Head Cut......................................................................... 153
Table 4.19 Class Effects of Project on F4: Bank Erosion at the Structure......................................... 154
Table 4.20 Class Effects of Project on F5: Bank Erosion Downstream of Structure ......................... 155
Table 4.21 Class Effects of Project on F6: Lack of Scour Pool Development ................................... 156
Table 4.22 Partial Correlation Matrix for Failure Indicators ................................................................ 157
Table 4.23 Failure Modes................................................................................................................... 160
Table 4.24 Consequence Categories ................................................................................................. 162
Table 4.25 Detection Rating ............................................................................................................... 163
Table 4.26 FMEA of a Rock Cross Vane, General Stream Restoration Usage................................. 164
Chapter 5
Table 5.1 Summary of Effects on Benthic Macroinvertebrates .......................................................... 176
Table 5.2 Restoration Priority Levels.................................................................................................. 185
Appendix A
Table A-1 Fractional Factorial Design Test Points ............................................................................. 205
Table A-2 Velocity (ft/s) at Given Point for Test Run.......................................................................... 206
Table A-3 Velocity Ratio for Cross Section by Test Run.................................................................... 210
Table A-4 Mean Depth (in) at Cross Section for Test Run................................................................ 211
xi
Chapter 1:
History of the use of hydraulic structures related to the

rock cross vane
1
Introduction
The rock cross vane is a structure used to maintain stream grade and to turn flow
away from banks. It also provides a drop structure that creates a downstream pool
for habitat and allows for a drop in bed elevation thereby allowing energy dissipation
and lower bed slopes upstream and downstream where sinuosity is restricted. It is
unique in that it acts as a weir, vane, and drop structure all in one. The upward
sloping arms of the rock vane are similar to those of single arm vanes used to
protect the outer stream bank of a curve where velocities increase. Figures 1.1 and
1.2 are profile and plan views of the rock cross vane.
Figure 1.1 Profile View of Cross Vane Arm
Source: Rosgen 2001
2
Figure 1.2 Plan View of Rock Cross Vane
Source: Rosgen 2001
If the current solution in stream restoration stability is the rock cross vane we must
know what the initial question was? What was or is the problem that warranted the
development of this structure? What are its roots and what was it intended for? Are
there examples in history that might shed light on its strengths and short comings or
provide an alternate practice? As researchers develop current stream restoration
practices, it is vital that science is coupled with modeling and with historical
reflection. Science can tell us how, and often, history can tell us why. Answering the
why and how often can help us determine a future projection, “to what extent?”
For centuries and even millennia humans have been mesmerized and inspired by
stream and river systems. Some have sought to capture them with art and poetry
and other have sought to harness them for the purposes of fishing, irrigation,
hydroelectric power and maximization of land use. Other groups of people have
inadvertently damaged them by extensive land development resulting in changes to
3
the watersheds. The most recent trend in surface water system modification is
stream restoration. Others may call it stream reconstruction or stream rehabilitation,
but the common purpose is to attempt to fix what humans have harmed in their
disregard and misuse of nature’s water systems. Urbanization and agriculture have
often left streams straightened, eroded and cut off from their natural flood planes.
Physical functions have been altered, and the quality of water is unable to support
the level of richness and diversity in aquatic flora and fauna that once could be found
in pristine streams. Past damage and the concern of continued degradation of
streams has led many scientists and engineers to search for solutions within the
watershed and within the stream. This chapter traces some of the earliest forms of
stream and river modification to the current practice of stream restoration and
specifically the rock cross vane.
Early Structures
While stream restoration is often deemed the newest and latest practice in stream
and river research, there is nothing new under the sun. Technologies used in stream
restoration quite possibly date back to ancient ancestors’ fishing practices. The rock
cross vane is used for grade control and to direct flow away from the banks. It
borrows objectives and practices from several structures in history. One of the
earliest known of such structures is the Native American fish weir.
The fish weir
The stone fish weir, a distant ancestor of the cross vane, has a function similar to the
cross vane in that flow is directed towards the center of the stream when the apex is
upstream. Two main types of fish weirs were used by Native Americans, thought to
be the originators. The first, more common in coastal areas, was made of stakes
and nets. The vertex pointed out towards the ocean so when the tide receded, fish
would be trapped in the nets (Lutins 1992). The second type of fish weir made of
stones is common in streams and rivers. The vertex is pointed either upstream or
4
downstream. In Nevada, streams are small and snow fed, and thus, the stake and
net method was employed by the Early Necada Paiute Indians. The fisherman
would sit and wait at the center of the weir with a net. As the fish came to the weir,
they would travel upstream towards the apex to find a way through. There, the
fisherman could catch multiple fish with a net (Gilmore 1953). Salmon weir at
Quamichan Village on the Cowichan River, Vancouver Island, circa 1866 is shown in
Figure 1.3.
Figure 1.3 Salmon weir at Quamichan Village on the Cowichan River, Vancouver Island
Source: “Fishing Weir” 2007
Stone weirs with the vertex pointed downstream, such as the Fair Lawn/Paterson
Fish Weir, provided a catchment for fish during the spawning season (Lutins 1999).
The remains of an ancient stone fish weir in the tidal Menai Strait in Wales are
shown in Figure 1.4.
5
Figure 1.4 Stone Fish Weir on Menai Strait
Source: “Fishing Weir” 2007”
While those seen in North Carolina tend to have upstream apexes, the apexes of
many riverine fish weirs found in the literature are downstream (Lutins 1992). Weirs
with the apex downstream would be better suited for catching fish migrating
downstream that were carried downstream by the current (such as eel as cited by
Lutins 1992) or herded downstream into the gap at the apex. Fish weirs with the
apex downstream would also direct flows outwards thus create a current that would
carry the fish towards the banks. Weirs with the upstream apex would be better
suited to catch fish migrating upstream. In these cases, fish could be trapped in the
center with a net or basket, otherwise, they would be caught along the banks. It
would be easier to spear them from the banks than to take out canoes to the center
to catch the game.
Cultural stories can be a reflection of the deep-rooted conflicts about human

interference with natural systems. The fish dam is mentioned in Native American
6
folklore. “How Coyote Destroyed the Fish Dam at the Cascades: Distributing
Salmon in the Rivers” was recorded in October 1916. This particular story told how
a coyote destroyed a fish trap at Celilo Falls, Oregon so that people upstream would
also be able to fish (Trafzer 2005). Another story told by the Snoqualime Indians of
Washington State described how the moon turned a fish trap into a waterfall to
establish order among the people saying, “’Game of every kind shall be found by the
people for their sustenance’” (Tollefson 1993).
There have been very few fish weirs researched and mapped in North Carolina
(Lutins 1992). However, there are several documentations of such structures on
historic sites as well as paddlers’ guides and archeological surveys. These are
listed in Table 1.1.
Table 1.1 Citations of Fish Weirs and Related Structures in North Carolina
Source Structure Location Notes
National Register Poe Fish Weir Jefferson located in Ashe Early Woodland and
of Historic Places County Mississippian peoples as well
(1978) as European pioneers, dates
as far back as 8999 B
Kirchen (1998) “V” shaped Lower Yadkin River used to trap fish just upstream
stone weir of a proposed dam location
Simpkins (1981) Fish Weirs Eno, Haw, and Dan River appears to be a natural
Drainages formation
NC Northstar fish traps Dan River where the Mayo Remains of fish traps
(2005) River feeds into the Dan
River
NC Northstar Fish Weir Mayo River near Anglin visible from the bridge
(2005) Mill Bridge
Google Maps W shaped fish Yadkin River near Winston visible by aerial photography
weir Salem
Lutins (1992) Fish Weir Town Creek Indian Mound None

in Montgomery County
and in the Yadkin and Pee
Dee Rivers
Culture & fish weir Catawba River northwest none

Heritage complex of the intersection with
Museums (2005) State Highway 77
7
While many stone fish weirs have been washed out or destroyed, many are still
intact. It has been nearly 200 years since the early 1800’s Trail of Tears and the
movement of Native Americans to reservations which leaves nearly two centuries of
nature’s forces tearing at the structures since the last of the Native Americans would
have been able to make repairs and adjustments to the structure. Not all of the
remaining fish weirs are fully intact. It is possible to make out the general shape of a
weir without having all the stones present. Lacking historical documentation, it is be
difficult to determine the original height of the weirs. That aside, there must be a
reason why they are not all completely washed out. One explanation is these fish
weirs were constructed in fairly shallow waters, with the stones visible to barely
visible during base flow. Base flow does not likely have the power to move these
stones. The stress acting on the bed is related to the water depth and water surface
slope. Because the stream is shallow, during high floods, the flow spills over onto
the flood plain. This prevents very deep fast flows that might easily move large
stones.
While the stone fish weir is functional either with the apex upstream or downstream,
the rock cross vane would lose a major function by being inverted. The arms are
able to protect the banks at low flow by forcing flow towards the center as the
overflow surface of the arms angle the water inwards, away from the banks. At
higher flows, the rock protects the bank as a barricade and by directing flow over the
vane towards the center as with lower flows. The fish weir, from observation of
photographs, does not appear to have arms that increase in elevation as they
extend downstream. The fish weir does have a slight opening at the apex, but
otherwise is partially above the water surface for the rest of the vane with water
flowing through the cracks between stones. Instead of constricting flow and directing
flow away from the banks causing a concentrated flow in the center, the vane
creates relatively calm, still water upstream behind the vane. The cross vane is only
partially inundated at base flow, but at bankfull flow, the cross vane typically just
becomes fully covered. At bankfull flow, a fish weir would be inundated and hidden
8
and have less of a hydraulic impact than the cross vane. However, there is striking
similarity in shape and the use of vanes to turn water at the higher flows. While the
Native American tribes that built fish weirs may not have been focused on bank
protection in the development of their structures, they understood the concept of
harnessing flow to serve a human purpose.
Dujiangyan Irrigation Scheme of ancient China
Another culture with a deep history with flow manipulation is the Chinese culture.
The oldest know irrigation system is the Dujiangyan Irrigation Scheme of ancient
China in the Sichuan Province. It was constructed in 256 BC by Li Bing, the
governor of Shu. The three major parts of the Dujiangyan Irrigation Scheme
headworks are: Dujiang Yuzui (fish mouth), Feishayan, and Baopingkou, consist of a
1) diversion dam, 2) a spillway and 3) an intake which work to, respectively, 1) divide
the river in two, 2) control silt and sediment during floods and 3) provide a steady
flow (Li 2006). According to Li and Xu (2006), “It is a model of harmonious
coexistence between mankind and nature, and is also included as a World Heritage
Site by the United Nations Educational, Scientific and Cultural Organization
(UNESCO) as the first water project in the world”.
Horseshoe Weirs
Horseshoe weirs are curved weirs that provide a protected drop in the river and
direct flow towards the center. In essence, they are weirs that also share functions of
rock cross vanes. Wales also has a long history with weirs. In the 11th century, a
horseshoe weir was constructed on the River Dee in Wales to guide flow towards
the last two arches of the Old Dee Bridge in order to power the mills. Chester Weir in
Figure 1.5 was built by a Norman, Hugh d’Avranches.
9
Figure 1.5 Chester Weir on the Dee River
Source: “River Dee, Wales” 2007
Chester Weir was rebuilt in 1281 after damage from a flood. Horseshoe Falls also on
the Dee River of Wales is a horseshoe weir constructed in 1808 by civil engineer
Thomas Telford (Cragg 1997).
Evolution of the Rock Cross Vane
In the United States, river work and maintenance began in the late 1800’s to early
1900’s. In 1887, Congress established the Mississippi River Commission to address
the needs for improvement as a result of highly damaging floods in 1849 and 1850
(USACE 2007). In the Western United States, around 1925 stream maintenance
effort continued with the development with levee systems and dams on the Colorado
River in Yuma, Arizona (LCRMSCP 2004). The term “stream and river restoration”
is not in reference to manipulation of waters for public use, but rather to water quality
and habitat improvements. White (1996) attributes the first American stream
restoration efforts to the work that began in Michigan by public works groups in 1927
to improve fish habitats in trout streams. By 1970, the main force of the practice
10
moved to the west where wood and boulder structures were implemented to reverse
the effects of logging, soil cultivation, and grazing (White 1996). It is hard to
determine exactly when this work began in the United States, especially as it relates
to in-stream structures; however, Duff and Banks (1988) conducted a
comprehensive literature review as a means of documenting the history of stream
restoration as it relates to improving fish habitat. They classify in-channel structures
as habitat management, which were second to bank stabilization practices. In-
channel structures were mentioned in 68% of articles since 1970 and 46% of articles
since 1980 (Duff and Banks 1988).
Design guidelines for rock cross vanes were first documented by Dave Rosgen in
1996 and arose out of a need for softer forms of bank protection (Rosgen 2001).
Soft typically refers to banks that have not been hardened by the application of rip-
rap or similar practices. Efforts from the Army Corps of Engineers and researchers
and practitioners such as Richard Hey and Dave Rosgen have explored a variety of
measures which have included hardening of the banks with boulder structures and
flow redirection as well as combinations of the two.
1980’s Rock Vortex Weirs and Rootwads
Vortex rock weirs and rootwads have been used since the 1980’s to improve fish
habitat. From the Stormwater Manager’s Resource Center, the definition of rock
vortex weir is, “a structure designed to serve as grade control and create a diversity
of flow velocities, while still maintaining the bed load sediment transport regime of
the stream” (2007). Figure 1.6 is a plan view of the structure.
11
Figure 1.6 Plan View of Rock Vortex Weir
Source: Stormwater Manager’s Resource Center 2007
Rootwads are the trunk of a tree and fan of roots in which the trunk is recessed into
the banks and the root fan protects the banks. Figure 1.7 is a diagram of a rootwad
revetment.
12
Figure 1.7 Plan View of Rootwad Revetment
Rosgen (1996) states that after fifteen years of monitoring, rootwads actually lead to
worse bank erosion during high flows due to eddies. While eddies provided nice
areas of scour for fish, the associated scour became problematic when it lead to
bank instabilities. Root wads are still commonly used and are often installed in
combination with other structures as a way to increase woody debris in the stream
and protect banks.
1988 The Bendway Weir
The bendway weir consists of a low profile rock sill angled upstream at 60 to 90
degrees from the bank. The bendway weir is similar to a flattened arm of the rock
cross vane or a flattened single arm vane but not sloped upwards in the downstream
direction. The angle to the bank is also larger than the arms of the rock cross vane.
13
The United States Army Corps of Engineers (USACE) warns against the angle
approaching parallel to flow lest it is behaves more as a flow divider. This is a
problem commonly seen at the arms of rock cross vanes placed in a bend, where
the flow attempts to bypass the arm and scours out the adjacent bank. Figure 1.8 is
a diagram that demonstrates the concept of the bendway weir.
Figure 1.8 Bendway Weir Theory
Source: Adapted from USACE 2006
Mr. Thomas J. Pokrefke Jr. initially conceived the bendway weir for use in a model of
the Mississippi River. According to a publication produced by the USACE Costal
Hydraulics Laboratory, the model addressed the following problems via subsequent
solutions:
1. Excessive maintenance dredging costs via deposition at the toe of the

revetment on the outside bend. The thalweg is shifted away from the outer
bank towards the ends of the weirs and resulting in lowered costs to retrofit
weirs.
14
2. Constricted navigation channel via a widened channel through the bend and
downstream of the bend
3. High velocities via controlled flow and increased weir effectiveness at high
flow and high energy conditions
4. Damaging high-flow current patterns via increased uniformity of surface

velocities
5. Insufficient navigation channel widths in the crossing downstream of the bend

via the widening mentioned earlier and a flow path moved towards the inside
bend
6. The loss of Least Tern nesting areas via improvements to aquatic habitat and
bank stabilization (USACE 2006).
1987- 1989 Iowa Vanes and Submerged Concrete Vanes
Although, the concept of Iowa or submerged vanes was first introduced by Russian
engineers in 1947, it wasn’t until 1983 (Odgaard and Kennedy) and 1986 (Odgaard
and Spoljaric) that the first attempts to develop a theoretical design were made
(Odgaard 2006). Iowa vanes are short, vertical vanes along the outer bend of the
channel that reduce secondary currents and high velocities on the outer banks
(Odgaard 1983). Figure 1.9 is of Iowa Vanes under construction.
15
Figure 1.9 Installation of Concrete Iowa Vanes at Low Flow
Source: “How to Control Streambank Erosion” 2006
Several years after the initial theoretical designs were made, Paice and Hey (1989)
introduced the use of submerged concrete vanes which would not only control the
secondary circulations to protect the banks but provide much needed fish habitat
(Rosgen 2001). The supplemental function of fish habitat was one of the driving
factors that would later appear in the goals of the rock cross vane.
1990’s Rock Cross Vane
The rock cross vane is best applied to stream restorations where flood plane
restrictions limit increasing stream sinuosity to decrease bed slope. The drop
structure provides energy dissipation and scour pool development and flow diversity
for fish habitat and spawning. Field observations show that rock cross vanes perform
much better when they are not preceded or followed closely by a bend where the
incoming flow might approach parallel to the arms. J-hooks and single arm vanes,
shown in Figure 1.10, are more suited for the bends, although observations show
that these too can suffer from bank erosion when built too steeply.
16
Figure 1.10 Plan View of Rock Vane and J-Hook
The arms protect banks by turning water towards the center of the stream which
results in scour pool development. Ironically, the rock cross vane was observed to
carry a high risk of bank erosion both at the arms of the structure and downstream of
the structure in Chapter 4 of this study. Bank erosion was primarily attributed to side
cutting as a result of alignment, placement, and undersized rock cross vanes.
The rock cross vane is both a weir and a vane. It is somewhat similar to a duckbill
weir and a labyrinth weir, which are both long-crested weirs. Long crested weirs are
those that have a longer overfall length than the actual width of the channel. Figure
1.10 is of three types of long crested weirs.
Figure 1.11 Plan Views of Three Types of Long Crested Weirs
Source: Williams et all 1993
17
The labyrinth weir is the only weir in Figure 1.11 that is not contracted. A difference
between the three long crested weirs above and the rock cross vane is that the weirs
have a constant crest elevation. Rock cross vane arms increase in elevation in the
downstream direction. The flow area over the rock cross vane is shaped most like a
Cipolletti weir. This is deceptive because the water is not only flowing in one
direction and the overflow surface is much longer than in the front view of the vane.
These inclined arms are what give the rock cross vane the “vane” part of its name.
Present-day Challenges
The focus on restoring streams and rivers to a “natural” state has resulting in the
development of small structures that can blend in with the environment. Structures
are built with materials that could be found in proximity to the stream such as logs
and boulders (as opposed to steel and concrete). Boulder vanes (cross vanes, j-
hooks, single arm vanes, rock vortex weirs, etc) provide a measure of stability
dependent on the design of the structure, the placement and construction, and
largely on the experience of the designers and construction contractors as well as
chance – not the most preferential tool to work with. Streams and rivers migrate
slowly over time. Sometimes they migrate rapidly in cases of severe environmental
events and changes.
The evolution of structures used for stream stability has not necessarily led to more
durable or effective practices. Many structures are altered on a trial and error basis
and lack a methodical approach is not used due to a poor of understanding of
stream functions. Rosgen (2001) states,” Structures in river engineering are
designed to help stabilize channel boundaries. However, monitoring their
effectiveness have indicated that many structures, contrary to the intended design,
caused river instability. Structures are often selected and installed without an
understanding of sediment transport and violate the dimension, pattern and profile of
the stable river.”
18
Rock cross vanes are designed according to a standard plan and are sized
according to bankfull width. The standard range of slopes and arm angles are used
to provide flexibility in tying arms to bankfull when possible. Monitoring has shown
that these structures can cause a variety of problems around the vane which may
influence stream and river instability (Rosgen 2001). Highly eroding banks around
structures disturb the sediment balance downstream and change the geometry of
the channel at the structure.
Rock cross vanes are designed to mimic natural rock outcroppings. The formation of
a waterfall occurs in a similar fashion to the upstream migration head cut past a
failed rock cross vane. Figure 1.12 is a diagram of the formation of a waterfall.
Figure 1.12 Formation of a Waterfall
Source: Mann 2006
Undercutting at the base of the drop leads to a collapse in the overhang causing the
plunge pool to develop again as the waterfall moves upstream of the initial site. And
thus, the cycle continues. Waterfall evolution differs from sill failure because its long
19
layer of hard rock results in slower upstream migration of the headcut. Because the
sill is usually only one boulder thick along the bed, once the sill is bypassed, headcut
migration occurs more rapidly. Rock structures “harden” banks and actually
accelerate bank erosion as they begin to fail (Rosgen 2001).
In nature, drop structures change over time. In pristine systems where there is little
human presence, watersheds are mostly stable, and change occurs slowly. Stream
restoration is conducted in regions of minor to severe human intrusion where the
watershed can be rapidly changing. Researchers should expect and design for, at
the least, the changes that occur in nature. There should be efforts to develop in-
stream structures that tolerate change over time. Streams are dynamic systems and
should include structures that are dynamic as well. It is unreasonable to expect hard
fixed structures to remain intact and functional in an unstable environment.
The following chapters explored the nature of the rock cross vane – how it
redistributes the velocity profile, its most common paths of failures, and its design
inadequacies – in an effort to improve upon and advance the technology. The
underlying problem is that researchers are searching for a fixed structure to stabilize
banks and stream bed in a dynamic system. Rosgen (2001) advocated for “softer”
structures and presented descriptions of Cross-Vanes, W-Weirs and J-Hook Vanes.
However, this research concluded that the rock cross vane in its original design may
itself be still too hard of a structure and recommends how implement the functional
of the rock cross vane with a design that would allow for the minor shifts and
adjustments that are likely to occur post-restoration.
20
Literature Cited
“Bendway Weirs.” Coastal and Hydraulics Laboratory - Engineer Research and

Development Center. USACE. Available at:
http://chl.erdc.usace.army.mil/chl.aspx?p=s&a=ARTICLES;109. Accessed 28
November 2006.
Cragg, Roger. 1997. Civil engineering heritage: Wales and west central England.
London: Thomas Telford Ltd for Institution of Civil Engineers.
Gilmore, H. W. 1953 . “Hunting Habits of the Early Nevada Pauites.” American

Anthropologist, New Series, 55(1): 148-153.
“Fishing Weir.” Available: http://en.wikipedia.org/wiki/Fish-weir. Accessed 19 March

2007.
Hatcher, R. D., Howell, D. E., and Talwani, P. 1977. “Eastern Piedmont fault
system: Speculations on its extent.” Geology. 5: 636-640
“How to Control Streambank Erosion.” Iowa Department of Natural Resources.

2006. Available at:
http://www.iowadnr.com/water/stormwater/forms/streambank_man.pdf.
Accessed 19 March 2007.
Kirchen, R. W. 1998. “Archeological Survey and Testing of the Yadkin Intake Dam
Project Area, Prepared for the City/County Utilities Commission, Forsyth
County, North Carolina.” Wake Forest Archeology Laboratories.
Li, K. and Xu, Z. 2005. “Overview of Dujiangyan Irrigation Scheme of ancient China
with current theory.” Irrigation and Drainage. 19th ICID International
Congress, Beijing, 55(3):291-298.
Lutins, A. H. 1992. “Prehistoric Fish Weirs in Eastern North America.” Unpublished

Masters thesis, State University of New York, Binghamton.
Lutins, A. H., DeCondo, A. P. 1999. “The Fair Lawn/Paterson Fish Weir” Bulletin of
the Archaeological Society of New Jersey, 54.
National Register of Historic Places (NRHP) 2005. Available:

www.nationalregisterofhistoricplaces.com. Accessed 20 August 2005.
“Natural & Cultural Features of the McColl Property” Culture & Heritage Museums.
Available at: http://www.chmuseums.org/mlesitefeatures.htm. Accessed 20
August 2005.
21
Odgaard, A. J. “Iowa Vanes – An Inexpensive Sediment Management Strategy.” The
University of Iowa. Available at:
http://www.iihr.uiowa.edu/projects/IowaVanes/index.html. Accessed 28
November 2006.
Odgaard, A. J., and Kennedy, J.F. 1983. “River-bend bank protection by submerged
vanes.” J. Hydr. Engrg., ASCE, 109 (8): 1161-1173.
Odgaard, A. J., and Spoljaric, A. 1986. “Sediment control by submerged vanes.” J.

Hydr. Engrg., ASCE, 112 (12): 1164-1181.
“River Dee, Wales.” Available at: http://en.wikipedia.org/wiki/River_Dee%2C_Wales.

“Rockingham County River Country” NC Northstar. Available at:

http://www.ncnorthstar.com/rc_rwebsections.shtml. Accessed 18 July 2005.
Rosgen, D. L. 2001. “The Cross-Vane, W-Weir and J-Hook Vane Structures...Their

Description, Design and Application for Stream Stabilization and River
Restoration.” Wildland Hydrology, Inc. 22 pp.
Simpkins, D. L. and Petherick , G. L. 1986. “Second Phase Investigations of Late

Aboriginal Settlement Systems in the Eno, Haw, and Dan River Drainages,
North Carolina” Research Report No. 6. Research Laboratories of
Anthropology, The University of North Carolina at Chapel Hill.
Stormwater Manager’s Resource Center. Available at:

http://www.stormwatercenter.net. Accessed 19 March 2007.
“The Mississippi River and Tributaries Project.” USACE. Available at:

http://www.mvn.usace.army.mil/pao/bro/misstrib.htm. Accessed 19 March
2007.
Tollefson, K. D. and Abbott, M. L. 1993. “From fish weir to waterfall. (Snoqualmie

Falls' significance to Snoqualmie Tribe)” The American Indian Quarterly.
Trafzer, C. E., ed. 1998. “Grandmother, Grandfather, and Old Wolf: tamánwit ku
súkat and traditional Native American narratives from the Columbia Plateau.”
Michigan State University Press.
Mann, J. C. “Formation of a Waterfall.” Wikipedia. Available at:

http://en.wikipedia.org/wiki/Waterfall. Accessed 28 November 2006.
Williams, M. L., Reddy, J. M., and Hasfurther, V. “Calibration of Long Crested Weir
Design Coefficient.” Water Source Data System Library: Water Resources
Center Publications. Department of Civil & Architectural Engineering
22
University of Wyoming, May 1993. Available at:
http://library.wrds.uwyo.edu/wrp/93-13/93-13.htm. Accessed 16 June 2005.
23
Chapter 2:
Effects of Rock Cross Vane Geometry on Velocity

Distribution
Abstract
This flume study assessed the velocity ratio (average center velocity to average
outer velocity) influenced by a range of rock cross vane geometric variables
implemented in practice. A stream segment was modeled with a rectangular
channel. A fractional factorial design of three geometric variables the cross vane
was tested in a flume study at North Carolina State University. Results showed
linear and quadratic effects of drop ratio (drop depth to bankfull depth), and
interaction effects of arm angle and drop ratio, arm slope and drop ratio, and arm
slope and arm angle on the velocity ratio. The area of flow at the overfall of the
structure was calculated and compared to the area of the drop without the structure.
This area ratio (flow at structure to flow of simple drop) was used to predict whether
constriction or overtopping of bankfull conditions was occurring. Calculations of flow
area at the overfall predicted flow constriction over the structure at a high arm slope
within the prescribed ranges of arm slope (0.03 – 0.07) and arm angle (20 – 30
degrees). Results of the flume study and area calculations imply that sufficient drop
is essential to creating the higher velocity ratios necessary for scour pool
development, while high arm slopes and arm angles may result in bank instability.
Keywords. Rock cross vane, flume study, stream restoration, velocity distribution,
hydraulics, velocity ratio, flow constriction.
24
Introduction
Functions of the Rock Cross Vane
The main function of vanes in rivers and streams is to change the direction of the
flow of water. The rock cross vane combines 1) flow redirection, 2) grade control,
and 3) scour pool development. The vertical drop and vane-like arms direct the
water towards the center of the channel causing flow to be slowed near the banks
and a pool to be scoured just downstream of the rock cross vane.
Geometric Variables of the Rock Cross Vane
There are four main geometric components of the rock cross vane: arm angle, arm
slope, drop, and width ratio of arms to center of the vane. The first three are
accompanied by accepted ranges. The fourth is fixed. Referring to Appendix A and
Figures A-1 and A-2, the following definitions and recommendations of arm angle
and arm slope are given:
1. Arm angle is measured where the arm projects from the downstream bank.
The range prescribed by most designers is a 20 to 30 degree angle.
2. Arm slope is measured as the rise from the horizontal over run of the top
surface of the arm. The upstream end of the arms of the vane is at a lower
elevation than the downstream end. The arm slope range prescribed by most
designers is 3 to 7%. Ideally, the arms of the vane tie in at bankfull; however,
a low slope vane will tie in at a lower level than a high slope vane for a fixed
arm angle.
3. Drop is measured as the distance from the top of the sill of the structure
located at the stream thalweg to the bed at downstream edge of the sill (the
center of the structure installed at bed elevation and used to control grade).
For fish passage, drops less than 1 foot are recommended. Drop ratio is
defined in this study as the ratio of the drop to the bankfull depth. Effective
25
drop is a term used to describe the drop and the pool depth. Effective drop is
not used in this study.
Rock cross vanes are designed such that both arm and sill comprise one-third of the
bankfull width of the channel. This study addresses the first three components of
cross vane geometry which are varied in practice.
The Current Problem with Rock Cross Vane Design Recommendations
The effects of the given ranges of the geometric variables of the rock cross vane on
velocity distributions of flow have not been modeled. Dave Rosgen, the first to
publish design recommendations and diagrams for rock cross vanes and other
structures, has updated some of his rock cross vane designs (Rosgen 2004). His
diagrams of the cross vane design in his 2001 publication are commonly used in
North Carolina streams and are the focus of this study (Figure A-1 and Figure A-2 of
Appendix A). Recommendations for rock cross vanes are constantly changing, and
the design of the rock cross vane will continue to evolve. Currently, design
judgments in practice are made based on a combination of required durability and
function. Changes to the design are made when observations show that the rock
cross vane is not performing as expected. Rather than develop the structure based
on trial and error, it would be better to determine the optimum design for the rock
cross vane through careful testing. This study focused on one aspect of rock cross
vane performance: the effects of the design variables on the velocity distributions
upstream and downstream of the structure.
Hypothesis and Study Goals
It was observed in this study and by other practitioners that higher arm angles
tended to be coupled with more sediment deposition at the end of the arms.
Because higher arm angles do not turn water inward as much as lower arm angles
of the same slope, it is expected that greater sedimentation at the banks is caused
26
by increased eddies at higher arm angles. Rock cross vanes with higher arm slopes
have been observed to form deeper pools, leading to the inference that they too
cause greater flow concentration in the center of the stream. It has been observed
that higher drops form deeper pools, and it was hypothesized that the effects of drop
height on the velocity ratio could be due to a stronger impingement jet and a
changed flow distribution. It was hypothesized that the design variables of the rock
cross vane have significant effects on the upstream and downstream velocity
distributions. It was also hypothesized that there are potential values of these
variables that would have adverse effects on rock cross vane performance due to
reduction of flow area over the structure. These values were estimated, but negative
effects such as bank erosion were better observed by visual inspection of actual
structures in streams.
There were two main goals of this study:
1. To explore and quantify the effects of the design variables arm angle, arm
slope and drop ratio on the flow distributions up and downstream of the rock
cross vane with a physical model.
2. To apply the knowledge gained about the effects of geometry with

recommended improvements to the design of the rock cross vane and
recommend future research.
Effects of Geometric Change on Velocity Distribution
Bank protection occurs downstream because the flow near the banks is slowed
down. Scour pool development occurs because of the jet impingement and
increases velocities in the center of the velocity distribution. While the approach flow
has a nearly uniform velocity distribution, the downstream velocity profile is more
curved. See Figure 2.1.
27
Figure 2.1 Plan View of the Velocity Distributions at Rock Cross Vane
One way to quantify the resulting shape of the velocity distribution is to compare the
center velocity to the outer velocity. The term velocity ratio is used to quantify the
comparison of the two velocity regions. Velocity ratio is the ratio of the average
center velocity to the average outer velocity. In Figure 2.1, this translates to the
average of the velocity at points 2, 3, and 4 to the average of points 1 and 5. The
averages account for a lack of symmetry in the velocity distribution.
By knowing the relationship of the rock cross vane’s arm angle, arm slope, and drop
ratio, a designer may use the predicted mean velocity to calculate the velocities near
the banks and in the center of the stream. This would allow the designer to predict
potential bank erosion and scour pool development. This study sought to find the
optimum arm slope, arm angle, and drop ratio combinations to maximize the velocity
ratio.
Potential Adverse Effects of the Rock Cross Vane on Flow
Because the arms are inclined rather than flat, the arms of the structure act as a
broad crested weir and a drop structure, and there are circumstances where the
area of flow might actually be reduced. Reducing arm angles and increasing arm
28
slopes reduces the flow area over the structure for a fixed depth. For this study, area
ratio was defined as flow area at the overfall of the rock cross vane to the flow area
at the overfall of a simple drop without the structure. When this ratio is less than one,
there are two possible conditions: (A) the flow is constricted, or (B) the water surface
elevation has increased and flow has overtopped the structure, depending on the
flow and channel conditions. A simple impediment of a step in the channel can be
used to simplify what is occurring with the rock cross vane. Figure 2.2 is a diagram
of a step in a channel.
Figure 2.2 Diagram of a Step in a Channel
If the energy of the approach flow (potential and kinetic energy) at 1 minus the
height of the step is greater than the minimum energy (found at critical depth) of the
approach flow, then the flow will become shallower and the velocity will become
greater as the flow goes over the step. This is referred to as flow constriction. When
the energy head of the approach flow minus the height of the step is less than the
minimum energy for the given unit flow rate, flow will become deeper and slower to
increase head in order to pass over the step. In this study, this is referred to as
overtopping. At bankfull depth, water would flow over the banks during overtopping
assuming bankfull was also top of bank. When the area ratio is less than 1, the rock
cross vane acts as an impediment or a step. Were this step quantifiable, designers
could predict whether flow is expected to constrict or overtop. The danger of flow
29
constriction over the arms of the rock cross vane is increased velocities against the
adjacent banks and the bed of the boulders causing scour and erosion. There were
cases in the flume study where constriction was likely occurring. The risk of the
structure overtopping, which is expected for larger storm events that the bankfull (1.5
year return interval), is that upon re-entry scour occurs on the banks at the ends of
the arms of the structure. If this is frequently occurring, it can eventually lead to side
cutting and a decrease in structure durability.
Hydraulic Jump at the Rock Cross Vane
Hydraulic jumps have been observed downstream of rock cross vanes. The jump
helps dissipate energy and create the scour pool.
Chow (1959) developed methods for developing a drop number DN and depths
surrounding the hydraulic jumps based on the following equations:
q2
DN = (1)
gh3
Where,
q = Unit discharge, cfs/ft
h = Effective height of drop, ft.
From the drop number, drop length, pool depth under nappe, depth of upstream of
the hydraulic jump and downstream depth can be determined.
LD
= 4.30 DN
0.27
(2)
h
yD
= 1.00 DN
0.22
(3)
h
y1
= 0.54 DN
0.425
(4)
h
30
y2
= 1.66 DN
0.27
(5)
h
Where,
LD = Jump length, ft
yD = Pool depth under nappe, ft
y1 = Depth upstream of the hydraulic jump, ft
y2 = Subcritical subsequent depth, ft.
Figure 2.3 is a schematic of a hydraulic jump. In the physical model, there was no
pool under the nappe, so yD is not depicted.
Figure 2.3 Hydraulic Jump
The length of the drop can be calculated from Equation 2 or from the Froude number
and Figure 2.2
v1
Fr = (6)
gy1
Where,
v1 = velocity at y1, cfs/ft
31
Figure 2.2 is used to estimate the length of the hydraulic jump.
Figure 2.4 Length of Jump for a Rectangular Channel
Source: FHWA 2007
The drop over the rock cross vane is not rectangular as it is assumed to be in
Chow’s equations. Flow dropping from the length of the arms is directed inward
towards the center of the channel. It is possible that the jump may be experienced
further downstream than that which is calculated for a rectangular drop as flow may
drop off the full length of the arm. This complicates the calculations of the expected
depths and length of the hydraulic jump.
Related Studies
There is limited completed research on rock cross vanes or other boulder structures
used in Natural Channel Design. Until recently, study on the effects of cross vane
geometry has been lacking. One researcher, Russ Lawrence, in Oregon is exploring
32
the relationship between the vane geometry and the resulting downstream pool
length, depth and location. This study is observational (Lawrence 2006). A graduate
student from New York, Margaret M. Soulman, recently studied 12 cross vanes
constructed on the Batavia Kill in the Catskills. Soulman studied the cross vane
geometries and the greater stream context to better understand the variability in the
pools depths below the cross vanes. She found connections between the arm
slopes and the pools’ width/depth ratios and an even stronger connection between
the arm width – center width ratios and the width/depth ratios (Soulman Email
correspondence 2 June 2005).
A structure similar in function to the rock cross vane, the bendway weir, is used on
the outside of bends to reduce outer bank velocity and push flow back towards the
centerline. This results in deepened channels and stabilized bank erosion. Thornton
et al. conducted a flume study to test how different designs of the weir affected the
outer bank velocity to center velocity ratio. They modeled two bends on the Middle
Rio Grand reach with Froude similitude at a 1:12 scale. A Sontek Acoustic Doppler
velocimeter was used to measure the point velocities (Thornton et. al. 2005). Their
measure of velocity ratio was used in this study as a good way of quantifying the
shape of the velocity distribution. They used three ratios: 1) the maximum outer bank
velocity to the maximum center baseline velocity, 2) the maximum inner bank
velocity to the maximum center baseline velocity, and 3) the maximum center
velocity to the maximum center baseline velocity.
The velocity ratio in this study allowed for a simple quantification of measurable
changes to the flow distributions, and the averages, rather than maximums, of the
flow regions accounted for the circumstances where there was a lack of symmetry in
the distribution due to channel/flume irregularities.
33
Materials and Methods
Site Layout and Flow Source
The test site was located at the NCSU Lake Wheeler Road Field Laboratory in
Raleigh, NC. The flume received flow from a storage pond located uphill of the
flume. The flume testing site is shown in Figure 2.5.
Figure 2.5 Site Layout
34
The inflow pipe for the flume had a one foot diameter. Table 2.2.1 gives the slope
and length of the inflow pipe and the initial water elevation in the pond for tests.
Table 2.1 Pipe Elevation, Length, and Slope

96.13 Pipe elevation in pond, ft 83.65 Upper pipe length, ft
94.95 Pipe elevation at juncture, ft 83.74 Lower pipe length, ft
92.99 Pipe outlet elevation , top, ft 0.0141 Upper pipe slope, ft/ft
3.14 Total Drop, ft 0.0234 Lower pipe slope. ft/ft
167.39 Length of pipe, ft 99.78 Initial water elevation in pond, ft
The storage pond was surveyed and Malcom’s (1989) methods were used to
calculate the stage-storage of the pond and to model the expected discharge coming
out of the pipe using Malcom’s Chainsaw Method (1989) in Table 2.2. The
coefficient of discharge used for the pipe was 0.6 and the weir coefficient, Cw, was
3.33.
Table 2.2 Time Discharge Calculation

T, min Qin, cfs S, cu.ft. Z, ft. Qout, cfs
0 0 6776 3.65 6.70
1 0 6374 3.51 6.55
2 0 5981 3.36 6.39
3 0 5598 3.22 6.23
4 0 5225 3.08 6.06
5 0 4861 2.93 5.89
6 0 4508 2.79 5.72
7 0 4165 2.65 5.54
8 0 3833 2.51 5.35
9 0 3511 2.37 5.16
10 0 3202 2.23 4.96
11 0 2904 2.09 4.76
12 0 2618 1.95 4.55
13 0 2345 1.82 4.33
14 0 2085 1.68 4.10
15 0 1839 1.55 3.86
The observed and calculated stage versus time relationships are given in Figure 2.6.
35
4
3.5
3
stage (ft)
2.5
1.5
0.0 5.0 10.0 15.0
time (min)
observed calculated
Figure 2.6 Observed Stage vs. Time of Storage Pond Drawdown
The slightly flattened portion of the observed stage-time relationship at time zero to
ten seconds was caused by the opening of the electro-mechanical actuated valve
(refer to Figure 2.4). Drawdown was slower than predicted which could have been
caused by algal blockages at the outlet or friction losses through the valve.
Flume Design
The flume design and testing methods provided a physical model that was a
simplified model of a stream of equal size with a length ratio of 1:1. All the variables
(arm angle, arm slope, and drop ratio) were dimensionless. The response variable,
velocity ratio, was also dimensionless. This eliminated scale effects and made the
model applicable to streams of a similar width/depth ratio and flow regime. A
rectangular flume was chosen for construction purposes.
The flume was constructed with two segments: the upper flume and the lower flume.
The upper segment remained fixed, while the lower segment was raised and
36
lowered by jacks to change the height and therefore drop ratio. The upper segment
was 20 ft long, and the lower segment was 24 ft long. Bed slopes were constant
throughout the study, while the drop between them varied.
The chosen bankfull width/depth ratio was 14:1, which fell into the C channel
category. C channels are often accompanied by rock cross vanes during restoration
because the high width to depth ratio is ideal for maintaining less forceful shallow
flows over the structures increasing structure durability. The stream segment was 44
ft long, 6.5 ft wide with a bankfull depth of 0.46 ft which was the observed depth in
the upstream segment of the flume before the drop structure was incorporated. The
bed of the upper flume was flat and the lower flume had a bed slope of 0.0025. It
should be noted that this is only meant to represent a small segment of a stream
with a sinuosity of 1. Figure 2.7 shows the plan view of the flume. This is not the full
view, but rather the middle segment where the testing and drop occurred.
37
Figure 2.7 Plan View of Test Flume
Planed, treated wood was used for construction. The use of a smooth even surface
resulted in minimized head losses and flow directional changes induced by stream
and rock cross vane irregularities. The flume was covered in low-cost marine carpet
to provide consistent roughness. Before adding the marine carpet, velocities were
higher at the walls than the center due to plastic and caulking at the joints of the
38
walls and bed. The installation of marine carpet resulted in a relatively low Manning’s
n value that fell between the planed wood Manning’s n of .012 and the short grass
value in a dredged channel of 0.022 (Chow 1959). In order to check the chosen
Manning’s n, the normal depth ranges in the upper and lower flume were calculated
by using the Manning’s n range (.012 - .022). Normal depth was calculated for the
given slopes and channel dimensions under the assumption that the flume segments
were hydraulically long enough to achieve normal depth conditions. The upper
segment average depth was taken at Cross Section A during base conditions. The
lower segment average depth was taken at Cross Sections B and C during base
conditions. Base conditions were the running of the model with the drop without
arms. This is shown in Table 2.3.
Table 2.3 Predicted Normal Depth and Observed Depth

Predicted Depth
Flume segment n=.012 n=.022 Observed average Depth
upper 0.20 ft 0.29 ft 0.43 ft
lower 0.22 ft 0.32 ft 0.30 ft
In the upper flume, flow was deeper than the predicted values of depth for Manning’s
n. This can be attributed to sagging in the upper flume. Although not planned, the
sagging was appropriate for the physical model because oftentimes, rock cross
vanes are installed at the ends of pools. In the lower flume, the average observed
depth fell within the predicted depth range given the Manning’s n between 0.012 and
0.022. The best guess for Manning’s n in the flume is 0.02, interpolated from the
lower flume average depth and the predicted depths for the range of Manning’s n.
The expected dead and live weight of the flume were calculated to determine the
number of joists to support the system. The flume was set up on the ground just
below the outlet of the 1 ft diameter pipe. The rest of the flume downstream rested
on screw-jacks and block supports to create both stability and mobility for changes in
drop and fine-tuning of slope. Arms were constructed of plywood and attached with
two hinges each to the end of the upper flume. Arms were supported by blocks
39
positioned on the bed of the downstream segment. Arm slope was fine-tuned by
adding shims and blocks. Figure 2.8 is the profile view of the location of the arms
between the upper and lower flumes.
Figure 2.8 Profile View of Rock Cross Vane Model
Another component of the flume was added at the head of the flume:
1) A baffle outlet to help reduce the effects of head changes in the basin
due to flow rate changes.
2) Triangular flow baffles or dissipaters, which helped create a uniform

velocity distribution at the head of the flume.
At the pipe outlet, flow initially splashed 3 feet high. The baffle outlet was installed
3.5 feet downstream from the pipe outlet to reduce splashing. Triangular flow
dissipaters were installed just upstream and downstream of the baffle outlet. The
combination resulted in a stilling basin at the head of the flume which helped
dissipate the energy generated from the steep inlet pipe. To increase the center
velocity to help establish a more uniform velocity distribution (also a function of the
marine carpet), the downstream triangular dissipater was trimmed to allow the water
to flow more freely through the center of the dissipater than near the walls. Figure
2.9 is the profile view of the baffle outlet setup at the head of the flume. Figure 2.10
shows this set-up during test flow conditions.
40
Figure 2.9 Profile View of Flume Stilling Basin and Baffle Outlet – Baffle Combination
Figure 2.10 Baffle Oulet – Baffle Combination
41
Velocity Test Points Locations and Methods
Velocity was measured with a USGS Pygmy Meter Model 6205 and an AquaCount
Calculator shown in Figure 2.11. Figure 2.6, indicates where test points were taken.
Figure 2.11 Test Run with the Pygmy Meter and AquaCount Calculator
Velocity was measured over 20 second intervals at each of the five points across
each of the three cross sections. The cross sections were located as follows: (A)
one bankfull width (6.5 ft) upstream of the beginning of the vane, (B) one bankfull
width downstream of the beginning of the vane, and (C) two bankfull widths
downstream of the beginning of the vane. The points were located so that each was
in the center of five equally spaced segments across the width of the flume: 0.66 ft,
1.98 ft, 3.30 ft, 4.62 ft, and 5.64 ft.
After finding the velocity distribution at each cross section, the velocity ratio was
calculated as the average velocity of the center three points (2, 3, and 4) to the
average velocity of the exterior points (1 and 5). A flow depth was measured at the
side of the flume at each cross section and the average depth was calculated from
42
this measurement the elevation survey taken at each point across the cross sections
to account for minor elevation changes across each cross section. Flow rate was
calculated from the average velocity and depth. There were hydraulic jumps coming
off the sill and both arms which collided in the middle of the structure or expected
pool region. The shape of and distance to the hydraulic jump varied with the drops
and arms widths and slopes and were not measured.
Figure 2.12 shows the cross vane with water flowing from the bottom of the picture
to the top.
Figure 2.12 Flume test of drop ratio = 1, angle = 20˚, slope = 3%
Testing for Effects of Time Variations during Velocity Measurements
The ideal flume test would have lasted under six minutes with initial and final
discharges of 6.7 cfs and 5.7 cfs respectively, as estimated from Table 2.2.
Occasionally due to errors in the velocity meter, the velocity measurements were re-
run lengthening the test time. There was concern that if each measurement of flow
43
distribution of a cross section was not done at the same time point during the pond
draw down, flow distributions would not be represented accurately because the flow
coming from the inlet pipe would be less, and there would be a lower energy head.
The velocity distribution at each cross section was measured three times over a
pond drawdown. This was replicated three times for a total of nine test runs. These
tests showed that velocity ratios at the cross sections were unaffected by the stage-
time effects during a run. See Table 2.4 below.
Table 2.4 Velocity Ratio for Three Test Runs at each Cross Section
For each test run at the cross sections, velocity ratio was measured three times. Velocity
Ratio was measured as the average center velocity – points 2, 3, and 4 - divided by the
average outer velocity – points 1 and 5.
Cross Section – Velocity Ratio Velocity Ratio Velocity Ratio

Test Run Measurement 1 Measurement 2 Measurement 3
A – Test Run 1 1.21 1.25 1.17
A – Test Run 2 1.29 1.29 1.34
A – Test Run 3 1.26 1.26 1.34
B – Test Run 4 1.20 1.07 1.14
B – Test Run 5 0.98 1.02 0.97
B – Test Run 6 1.02 0.99 1.00
C – Test Run 7 1.32 1.30 1.44
C – Test Run 8 1.53 1.42 1.48
C – Test Run 9 1.40 1.33 1.39
From these test runs, it was inferred that the drawdown of the pond did not affect the
velocity ratio. Test runs to observe head losses showed that even though the pond
water level was dropping, the flow rate in the fume remained constant for the length
of the study because the baffle outlet reduced the effects of head loss. The flow rate
was not constant at each cross section but was not changing much over time due to
leakages. Larger leakages occurred between the upper and lower flume which
accounts for the larger observed difference in flow rate between cross section A and
44
B. Minor leakages occurred during the length of the lower flume which accounts for
the minor difference in flow rate between cross section B and C.
Experimental Design
The experimental design was a factorial analysis where each variable was tested at
three values. Table 2.5 lists the values for each of the three levels of each variable in
the study.
Table 2.5 Factor Levels and Values

Factor Arm Angle Arm Slope Drop Ratio
Level (deg) (ft/ft) (ft/ft)
1 15 0.01 0.25
2 25 0.05 0.75
3 35 0.09 1.25
This resulted in a total of 27 combinations of variables tested in the flume with three
test runs each which acted as replications for statistical purposes. Also tested were
the velocity distributions across the three cross sections at each drop without the
arms attached. This was referred to as base condition. This was included so that the
effects of the inclusion of arms on the depth of flow could be observed. All test points
are included in Table A-1 of Appendix A.
Calculation of Flow Area of over the Rock Cross Vane
As mentioned in the introduction, there are concerns of the potential of the rock
cross vane to constrict flow or lead to overtopping of bankfull depth when the flow
area is reduced at the overfall. Area Ratio was calculated as the ratio of the
predicted flow area at the overfall surface of the rock cross vane to the predicted
flow area of the rectangular channel at the drop when arms are not installed.
45
The area over the arms is calculated as twice the average depth over one arm
multiplied by the water surface length over one arm. The average depth of water on
the arms is the average of the depth over the end of the arm and the depth of flow
where the arm connects at the sill. The depth at the end of the arm is the difference
between the depth of flow and the elevation of the arm at that point which can be
calculated by the given arm slope, angle and width, shown in the first part of
Equation 7. The length of the overflow surface is the horizontal projection of the
length along the diagonal of the arm, which is the water surface length over the arm.
This is calculated from the given arm slope and arm angle as shown in the second
part of Equation 7, which assumes arms are fully inundated.
2
⎡ W sin( s ) ⎤ ⎛ W cos( s) ⎞
= ⎢2harm − arm Warm + ⎜⎜ arm ⎟⎟
2
Aarms ⎥ (7)
⎣ tan(α ) ⎦ ⎝ tan(α ) ⎠
Where,
harm = depth of flow over the arms, ft
α = angle of the arm from the bank, degrees
s = slope of the arm from the horizontal, degrees
Warm = width of arm laterally across the stream, ft
Equation 8 gives the adjusted lateral arm width for conditions when the arm is not
fully inundated.
hsill tan(α )
Wnew = (8)
sin( s )
Where,
hsill = depth of flow over the sill
Flow area over the sill is calculated as the depth of flow multiplied by the width of the
sill.
Asill = hsill Wsill (9)
46
Where,
Wsill = width of sill, ft
Total area at the overflow surface of the rock cross vane is:
Atotal = Aarms + Asill (10)
The depth of flow at the overfall is actually shallower than the upstream depth. As
water flows over the rock cross vane, two different depths can occur. When there is
a free overfall, water flows over a brink at a depth known as brink depth. When the
drop is low enough that the tail water depth prevents a free-overfall, the structure’s
arms behave as broad crested weirs and flow is at critical depth.
Critical depth, yc is calculated from Henderson (1966) as:
q2
yc = 3 (11)
g
Brink depth is estimated as:
y brink = 0.7 y c (12)
Plugging this estimated for depth over the arms and sill gives an estimate of the flow
area over the rock cross vane. Expected flow area without the structure is simply the
multiplication of the brink depth and channel width. These equations were used to
predict potential constriction through the rock cross vane at a drop ratio of 0.75.
Results and Discussion

Velocity Ratios Observed
Table 2.6 lists the velocity ratios calculated from the velocity measurements in Table
A-2 of Appendix A.
47
Table 2.6 Velocity Ratio for Cross Section by Test Run
VRA is the velocity ratio at cross section A. VRB is the velocity ratio at cross section B. VRC
is the velocity distribution at cross section C. D = drop ratio. A = arm angle. S = arm slope.
1,2,3, indicate factor levels (see Table 2.5). VRA = velocity ratio at cross section A. VRB =
velocity ratio at cross section B. VRC = velocity ratio at cross section C.
Test Run VRA VRB VRC Test Run VRA VRB VRC
D1A1S1 1.09 2.96 1.18 D2A2S3 1.25 2.99 2.22
D1A1S1 0.98 3.08 1.17 D2A2S3 1.25 2.87 2.13
D1A1S1 1.01 2.47 1.15 D2A2S3 1.23 3.10 2.25
D1A1S2 1.01 2.14 1.64 D2A3S1 1.24 3.45 2.25
D1A1S2 1.04 2.57 1.40 D2A3S1 1.29 3.11 2.06
D1A1S2 1.04 1.88 1.36 D2A3S1 1.29 3.44 2.24
D1A1S3 0.96 1.96 1.48 D2A3S2 1.25 2.77 2.32
D1A1S3 1.00 1.75 1.45 D2A3S2 1.30 3.04 2.36
D1A1S3 1.03 1.81 1.42 D2A3S2 1.29 3.05 2.18
D1A2S1 0.97 1.37 1.33 D2A3S3 1.24 2.94 2.33
D1A2S1 0.98 1.50 1.28 D2A3S3 1.20 2.84 2.33
D1A2S1 1.04 1.47 1.29 D2A3S3 1.23 2.98 2.25
D1A2S2 1.00 1.88 1.37 D2BASE 1.21 1.07 1.43
D1A2S2 0.99 1.57 1.27 D2BASE 1.29 1.02 1.48
D1A2S2 1.04 1.77 1.26 D2BASE 1.26 0.99 1.39
D1A2S3 1.00 2.65 1.38 D3A1S1 1.29 2.07 2.00
D1A2S3 1.00 2.67 1.47 D3A1S1 1.29 1.84 1.84
D1A2S3 1.02 2.62 1.42 D3A1S1 1.32 1.76 1.80
D1A3S1 1.28 1.53 1.27 D3A1S2 1.32 2.73 1.62
D1A3S1 1.07 1.29 1.09 D3A1S2 1.30 2.45 1.67
D1A3S1 1.07 1.37 1.25 D3A1S2 1.32 2.06 1.67
D1A3S2 1.08 1.60 1.46 D3A1S3 1.25 2.38 1.61
D1A3S2 1.11 1.78 1.35 D3A1S3 1.29 2.12 1.70
D1A3S2 1.14 1.69 1.38 D3A1S3 1.29 2.99 1.66
D1A3S3 1.18 2.24 1.52 D3A2S1 1.28 3.57 1.95
D1A3S3 1.14 2.31 1.49 D3A2S1 1.29 3.56 1.89
D1A3S3 1.00 2.19 1.46 D3A2S1 1.26 3.43 1.92
D1BASE 1.01 1.13 1.08 D3A2S2 1.20 2.67 1.84
D1BASE 0.99 1.14 1.19 D3A2S2 1.28 2.85 1.78
D1BASE 1.01 1.09 1.13 D3A2S2 1.25 2.67 1.92
D2A1S1 1.26 1.62 1.96 D3A2S3 1.16 3.59 2.00
D2A1S1 1.27 1.54 1.78 D3A2S3 1.14 3.50 1.97
D2A1S1 1.33 1.53 1.76 D3A2S3 1.21 3.63 1.99
D2A1S2 1.33 3.18 1.71 D3A3S1 1.28 3.71 2.28
D2A1S2 1.24 3.55 1.71 D3A3S1 1.28 4.33 2.16
D2A1S2 1.30 3.50 1.80 D3A3S1 1.30 4.37 2.14
D2A1S3 1.20 2.70 1.67 D3A3S2 1.28 3.80 2.34
D2A1S3 1.26 3.35 1.61 D3A3S2 1.27 3.44 2.30
D2A1S3 1.23 3.06 1.69 D3A3S2 1.24 3.16 2.32
D2A2S1 1.21 3.01 1.94 D3A3S3 1.25 3.78 2.18
D2A2S1 1.21 3.28 2.07 D3A3S3 1.30 3.42 2.23
D2A2S1 1.20 2.80 2.00 D3A3S3 1.29 3.46 2.16
D2A2S2 1.23 2.57 1.99 D3BASE 1.32 1.55 1.72
D2A2S2 1.22 2.69 2.03 D3BASE 1.30 1.54 1.75
D2A2S2 1.30 2.59 2.15 D3BASE 1.35 1.62 1.67
48
Statistical Modeling
PROC GLM LSMEANS was run to determine whether the cross sections were
statistically different shown in Table 7. Pdiff with Tukey test showed that the cross
sections had significantly different means (p<0.0001).
Table 7 LSMEANS for Class Effects of Cross Section
Cross Section LSMEANS

1 – Cross Section B 2.66575
2 – Cross Section C 1.77396
0 – Cross Section A 1.18826
PROC RSREG was run to model the linear, quadratic and interactive effects of the
variables on the velocity ratio. Equation 13 is the response surface regression model
of velocity ratio.
Velocity Ratio = β 0 + β1 Drop + β 2 Angle + β 3Slope + β 4 Drop × Drop + β 5 Angle × Drop +

β 6 Angle × Angle + β 7 Slope × Drop + β 8Slope × Angle + β 9 Slope × Slope
(13)
Where,
Velocity Ratio = as defined in the introduction, ft/s / ft/s
drop = drop ratio as defined in the introduction, ft/ft
angle = arm angle as defined in the introduction, degrees
slope = arm slope as defined in the introduction, ft/ft
β0 – β9 = parameters to be estimated in SAS given in Table 2.6 below.
This model was constructed and run in PROC RSREG. Figure 2.13 shows the
predicted values and observed values of velocity ratio. On the x-axis, drop ratio is
held constant for the first group of nine treatment combinations, and angle is held
constant for 3 treatment combinations as the arm slope is varied. Values of factors
increase from left to right.
49
Figure 2.13 Predicted and Actual Velocity Ratio for Drop, Angle, and Slope for Cross Section C
50
Figures 14 and 15 present the predicted velocity ratios by paired combinations of the
three factors, drop, angle and slope. Drop effects were maximized just over 100%
drop and begin to diminish.
pr edi ct ed
2. 36
2. 08
1. 80
125
1. 52
100
75
1. 24 Dr op
35
30 50
25
Angl e 20
25
15
Figure 2.14 Predicted Velocity Ratio versus Arm Angle and Drop Ratio for Cross Section C
Predicted velocity ratio was more sensitive to arm angle as drop ratio increased as
shown in Figure 2.14. This likely occurred because as the drop increased, less area
of the vane arms were submerged by the flow below the drop. The predicted velocity
ratio became less sensitive to arm slope as drop increased as shown in Figure 2.15.
At higher drops, less of the arms was inundated by the tail water. As water flowed
over the arms, hydraulic jumps occurred closer to the arms and sill which indicated a
weaker jump and drowning of the jump.
51
pr edi ct ed
2. 36
2. 08
1. 80
9
1. 52
5
1. 24 Sl ope
125
100 3
75
Dr op 50
1
25
Figure 2.15 Predicted Velocity Ratio versus Drop Ratio and Arm Slope for Cross Section C
Fit of Model
The model was run for all data by cross section. Based on the residuals plots, cross
section C was the only model that was deemed to have a strong fit. The models for
cross sections A and B did show significant effects of several variables, but the
model cannot be used to accurately predict velocity ratio for these cross sections.
While the model for cross section A seemed strong (R2 = 0.8749), the plot of
residuals showed potential poor fit, likely caused by inundation of the structure at the
lowest drop. The velocity ratio was close to 1.25 for most tests at drop ratios of 0.75
and 1.25 but for the drop ratio of 0.25, the velocity distribution was close to uniform
(Velocity Ratio = 1.0). This is supported by the plot of the residuals for cross section
A , shown in Figure 2.16 which shows lumping of the data.
52
Figure 2.16 Residuals at Cross Section A
Table 2.8 presents the Analysis of Variance for the Response Surface Equation
Model at Cross Section A.
Table 2.8 Analysis of Variance for Response Surface Regression Model at Cross Section A
Type I Sum
Regression DF of Squares R-Square F Value Pr > F
Linear 3 0.676991 0.6473 122.42 <.0001
Quadratic 3 0.203805 0.1949 36.85 <.0001
Crossproduct 3 0.034247 0.0327 6.19 0.0008
Total Model 9 0.915043 0.8749 55.16 <.0001
Cross section B initially appeared strong (R2 = 0.6537) but was determined to have a
poor fit due to high residuals (-1.0 – 1.25). The high residuals were likely caused by
the close proximity to the ends of the longer arms, where it was likely that water was
circulating. Figure 2.17 shows the residuals for cross section B.
53
Vr at i o = 1. 8068 +0. 0101 Dr op - 0. 0241 Angl e +0. 0623 Sl ope - 0. 0001 dr op2 - 0. 0001 angl e2
+0. 0084 sl ope2 - 0. 0004 sd - 0. 0037 sa +0. 001 da
1. 25 N
81
1. 00 Rsq
0. 6537
Adj Rsq
0. 75 0. 6098
RM SE
0. 4904
0. 50
0. 25
0. 00
- 0. 25
- 0. 50
- 0. 75
- 1. 00
1. 50 1. 75 2. 00 2. 25 2. 50 2. 75 3. 00 3. 25 3. 50 3. 75 4. 00
Pr edi ct ed Val ue
Figure 2.17 Residuals at Cross Section B
Model at Cross Section B.
Table 2.9 Analysis of Variance for Response Surface Regression Model at Cross Section B
Type I Sum
Linear 3 19.86432 0.4029 27.54 <.0001
Quadratic 3 2.694427 0.0547 3.74 0.0149
Crossproduct 3 9.66647 0.1961 13.4 <.0001
Total Model 9 32.22521 0.6537 14.89 <.0001
Cross section C (R2 = 0.9030) showed a strong model fit based on the random
scatter of the residuals in Figure 2.18, The residuals range from -0.3 to 0.4, which is
much less than the residuals of cross section B.
54
Vr at i o = 0. 919 +0. 0223 Dr op - 0. 0189 Angl e - 0. 0031 Sl ope - 0. 0001 dr op2 +0. 0002 angl e2
+0. 0012 sl ope2 - 0. 0004 sd +0. 0011 sa +0. 0003 da
0. 4 N
81
Rsq
0. 3 0. 9139
Adj Rsq
0. 9030
RM SE
0. 2
0. 1127
0. 1
0. 0
- 0. 1
- 0. 2
- 0. 3
1. 2 1. 3 1. 4 1. 5 1. 6 1. 7 1. 8 1. 9 2. 0 2. 1 2. 2 2. 3
Pr edi ct ed Val ue
Figure 2.18 Residuals at Cross Section C
Model at Cross Section C.
Table 2.10 Analysis of Variance for Response Surface Regression Model at Cross Section C
Type 1 Sum
Linear 3 6.598780 0.6295 173.05 <.0001
Quadratic 3 2.168167 0.2068 56.86 <.0001
Cross product 3 0.813994 0.0776 21.35 <.0001
Total Model 9 9.580941 0.9139 83.75 <.0001
Model and Parameter Estimates for Cross Section C
From this point on, the results discussed will be for cross section C only because of
the strength of fit of the model. Table 2.11 presents the SAS output for the
regression model. Significant effects (alpha=.05) are in bold.
55
Table 2.11 SAS output for Parameters of Response surface Regression Model at Cross
Section C
Standard
Parameter DF Estimate Error t Value Pr > |t|
Intercept 1 0.918961 0.191127 4.81 <.0001
Drop 1 0.022325 0.001934 11.54 <.0001
Angle 1 -0.018873 0.013869 -1.36 0.1779
Slope 1 -0.003078 0.021866 -0.14 0.8885
Drop*Drop 1 -0.000138 0.000010629 -13.01 <.0001
Angle*Drop 1 0.000250 0.000037580 6.66 <.0001
Angle*Angle 1 0.000234 0.000266 0.88 0.3806
Slope*Drop 1 -0.000356 0.000093951 -3.79 0.0003
Slope*Angle 1 .001076 0.000470 2.29 0.0249
Slope*Slope 1 0.001219 0.001661 0.73 0.4655
Significant Effects on Velocity Ratio
Significant effects on velocity distribution were:
1. Positively correlated linear effects of drop ratio
2. Negatively correlated quadratic effects of drop ratio
3. Positively correlated cross product effects of arm angle and drop ratio
4. Negatively correlated cross product effects of arm slope and drop ratio
5. Positively correlated cross product effects of slope and angle.
The positively correlated linear effects of drop ratio are stronger than the negatively
correlated quadratic effects of drop ratio. As drop ratio increases, velocity ratio
reached a maximum potential. Increasing drop ratio lessened the potential for the
tailwater to drown the hydraulic jump at the structure. It is likely that the hydraulic
jump enabled by the drop resulted in increased velocities in the center of the velocity
distribution. Lower velocities at the outer parts of the distribution were reflected by
eddying. As drop ratio increased, the effects of arm angle increased. This effect is
likely caused by the reduction of tailwater interference. The arms were able to turn
56
the water more as the tailwater depth decreased. Increasing the arm angle resulted
in more defined eddy regions which subsequently lead to higher velocity ratios.
Increasing arm angle also prevented a decrease in flow area from the rectangular
upstream portion to the overflow surface of the arms and sill for a. given flow rate.
As drop ratio decreased, arm slope effects became greater. As the tailwater depth
began to flood the structure, increasing the slope of the arms served as resistance to
flow of the sides forcing more water through the center. At higher drop ratios, these
effects became negligible as the effects of drop ratio and arm angle were dominant.
The positively correlated cross product effect of arm slope and arm reflected in
interactive effects of drop ratio and arm slope and arm angle.
Sources of Error
The model for cross section C had a close fit to the observed data (R2 = 0.9030), but
there were potential sources of error found in the physical model. The first was
human error in the construction of the flume and inadequate control of boundary
conditions. Blocks were used to support the arms of the vane. When possible,
blocks were laid parallel with the overflow surface underneath the arms along the
entire length. At the lowest drop, concrete blocks could not fit underneath the arms
so small wood blocks were used. Rebar was used to keep the arms from floating up.
As the arm angle decreased, the sharp angle with the bank did not allow room for
the blocks to fit without protruding out from under the arms. The blocks did not
prevent water from circulating underneath the arms. These may have caused slight
changes to flow pattern.
Environmental conditions were another possible source of error. The flume was
exposed to the elements and subject to extreme conditions. Shortly after
construction of the flume, it was lifted by a strong windstorm and fell off its supports.
The flume was subsequently tethered to the ground. Extremely windy days caused a
set of faulty data which had to be dropped from the study. Because the testing was
conducted in the winter, the flume was subject to freezing when water was left
57
standing over night. The icy conditions may have lead to variations in roughness of
the flume.
A third source of error was variation in the flume setup at each new drop level. The
storage pond had an algae bloom during the months of testing which began to clog
the outlet pipe that led to the flume. This may have affected flow rates. The plywood
was subject to warping which may have caused slight changes in flow direction in
the flume. Finally, leakages in the flume may have led to decreased flow rates
downstream.
Flow Regimes Observed in the Flume
The Froude number was calculated from Equation 6 for each channel segment
(upper section and lower section). According to Table 2.12, during base conditions
(no arms), flow was transitional coming out of the drop as calculated from the
average depths observed.
Table 2.12 Flow Regime Calculations for Cross Section

Cross Section mean depth, in mean Q, cfs Fr Flow Regime
A, base 5.11 7.29 0.71 Subcritical
B, base 3.67 6.20 0.99 Transitional
C, base 3.43 5.80 1.03 Transitional
A, with arms 5.28 6.74 0.63 Subcritical
B, with arms 4.02 5.25 0.73 Subcritical
C, with arms 3.47 5.22 0.91 Subcritical
Transitional flow is difficult to maintain, and the visual observation of rapid

undulations is often a sign of critical flow or a weak hydraulic jump. The undulations
disappeared as the arms were added and the resulting stronger jumps forced the
flow more quickly into subcritical flow. When the arms were added, the hydraulic
jump was observed within the length of the arms before cross section B, 6.5 ft
downstream of the drop. This observation of the hydraulic jump being located within
58
the arms was supported by a calculation of the length of the hydraulic jump by
equations 1 - 6.
Using a drop ratio of 0.75 and a bankfull depth of 0.46 ft, the effective drop is 0.34 ft.
There was no pool in the channel model to add to the drop height. Using a flow rate
of 5 cfs and channel width of 6.5 ft, the unit discharge is 0.770 cfs/ft. The resulting
drop number is 0.47 and y1 (the depth of the subcritical flow coming off the drop) is
0.13 ft. Using Equation 2, the length of the hydraulic jump is calculated as 1.2 ft.
According to these calculations, the jump was complete before Cross Section B
which is located 6.5 ft downstream of the drop. As mentioned previously, because
the overflow surface of the rock cross vane is three dimensional, this was not an
accurate calculation of the length of the jump but gives an estimation of the jump
coming over the sill.
Estimated Flow Area over the Rock Cross Vane during Tests
Area ratio predicts potential overtopping or constriction of the rock cross vane. When
the area ratio is less than 1, flow area over the structure is less than the expected
flow area at a drop without the structure. This would result in overtopping of the
structure at bankfull flow or flow constriction. As an example, the average base flow
rate and critical depth at a drop ratio of 0.75 were used to predict area ratios at the
different levels of geometric variables. A free overfall was observed at this drop;
however, calculations were made for both a free overfall and where tail water depth
prevented a free overfall. Average flow rate was 7.2 cfs. Critical depth is calculated
as 0.20 ft from Equation 11. Brink depth was calculated as 0.14 ft from Equation 12.
Calibrations of the model would be required to determine an actual critical depth

over a rock cross vane. Table 2.13 presents calculated flow areas for base depth
and estimated critical depth. Conditions of potential constriction are in bold.
59
Table 2.13 Area Ratio at Critical Depth and Brink Depth for 7.2 cfs
Area Ratio at Area Ratio at

Arm Angle Arm Slope Critical Depth Brink Depth
0.01 2.60 2.47
0.02 2.30 2.03
0.03 1.99 1.60
0.04 1.68 1.28
15 0.05 1.42 1.09
0.06 1.24 0.97
0.07 1.11 0.87
0.08 1.01 0.81
0.09 0.94 0.75
0.01 1.80 1.76

0.02 1.70 1.60
0.03 1.59 1.45
0.04 1.48 1.29
25 0.05 1.37 1.14
0.06 1.26 1.01
0.07 1.16 0.91
0.08 1.06 0.84
0.09 0.98 0.78
0.01 1.44 1.42

0.02 1.39 1.34
0.03 1.34 1.27
0.04 1.28 1.19
35 0.05 1.23 1.12
0.06 1.18 1.04
0.07 1.13 0.97
0.08 1.07 0.89
0.09 1.02 0.83
The calculations indicated that flow area was not expected to be constricted at arm
slopes of 0.3. At critical depth, flow area over the rock cross vane is also expected to
be constricted at an arm slope of 0.09 and arm angles of 15 and 25 degrees. When
there is a free overfall, constriction is also expected to occur at arm slopes 0.7 and
greater and the combination of an arm slope of 0.6 and an arm angle of 15 degrees.
This shows that constriction could occur within the given design range of arm slope
60
and degree arm angles. It also shows that increasing arm slope and arm angle
increases the expected occurrence of constriction.
Conclusions
Summary of Results
The statistical model was cumbersome for design use, so independent and
dependent variables were lumped in ranges to simplify the effects on the velocity
distribution of flow coming out of the rock cross vane at cross section C.
1. For drops less than or equal to 25 % bankfull depth, the velocity ratio ranged
from 1.25 – 1.5. Increasing slope increased the velocity ratio.
2. For drops greater than 25% bankfull depth, the velocity ratio values ranged
from 1.75 – 2.25. Increasing arm angle increased the velocity ratio.
Simplifying this by using only highs and lows for each parameter and high, low and
medium for the velocity ratio gave the following relationships:
1. Low Drop = Low velocity ratio
2. High Drop + Low Angle = Medium Velocity ratio
3. High Drop + High Angle = High Velocity ratio
Where,
low velocity ratio is less than 1.5
medium velocity ratio is 1.5-2.0
high velocity ratio is greater than 2.0
low drop is 25% bankfull depth or less
high drop is greater than 25% of bankfull depth
low angle is 25 degrees or less
high angle is greater than 25 degrees
Slope is not shown because varying slope never increased velocity ratio from a low
to a high or a high to a low velocity ratio.
61
The example calculations for area ratio at the estimated critical depth at the structure
for drop ratio of 0.75 predict conditions of overtopping constriction within the
prescribed range of arm slope (.03 – .07 ) and arm angle (20 – 30 degrees). This
implies that area calculations should be incorporated into rock cross vane design.
Increasing flow area (area ratio>1) helps slow down velocities across the structure.
As the area ratio reduces to less than one, velocities will increase over the structure
increasing the risk to bank failure at the structure, or the water surface level will
increase and result in an increased risk of bank erosion at the point of flow re-entry
into the stream just downstream of the arms.
Recommendations for Changes to Rock Cross Vane Design
There are many components to the success of a rock cross vane. It is

recommended that arm angle and arm slope be set to achieve an area ratio that
ensures protection of the physical structure at all drops and be adjusted to maximize
the velocity ratio at low drops where the velocity ratio is expected to be lower. At the
lowest drop ratio (0.25), arms were observed to be nearly fully inundated, and tail
water depth caused backing up over the vane causing arm slope to be an important
factor for increasing center velocity. High arm slopes could increase the velocity ratio
at low drops, but designers must ensure that the cross vane is wide enough and the
arm angles are great enough to prevent constriction of the flow over the vane. While
scour pool formation is important, sedimentation from eroded banks due to poor
construction might negate habitat benefits a pool provides. The qualitative
assessment in Chapters 3 and 4 led to the conclusion that higher arm slopes tend to
be coupled with problems with piping and side cutting. This supports the
aforementioned example of flow area constriction or overtopping of the bankfull
depth in the test flume at the estimated flow depth for the base discharge rate.
Constriction would lead to higher velocities on the arms and potentially scour around
them. Constriction might lead to downstream bank erosion from expansion after the
constriction through the rock cross vane.
62
Downstream velocities are expected to be higher at greater drops, and precautions
should be made to protect the banks from high near-bank velocities. At high drops,
less tailwater was observed, and the velocity ratio was influenced more strongly by
the angle of the arms than the arm slope. Increasing arm angle was resulted in
higher velocity ratios with swift undular flow after the drop with nearly still water at
the walls. This observed relationship is consistent with field observations of
deposition at the foot of the arms and scour in the thalweg between the arms. At
medium to high drop ratios (>0.25), arm slope could be lowered to increase arm and
bank stability with a negligible decrease in velocity ratio. At high arm slopes, higher
arm angles are essential to prevent constriction and overtopping as higher angles
increase cross-sectional area for a fixed arm slope.
Future Research
While a mathematic model incorporating three dimensional effects of geometry on

velocity distribution was not generated, a model could be developed pending
calibration of coefficients of discharge and further model studies of rock cross vanes
and eddies. Some of this could be accomplished by flume studies. The difficulty of
trying to achieve constant conditions in the flume with variability from sources such
as warping plywood, changing flow, and unpredictable elements of the weather,
could have been prevented. This would require:
1. A concrete or pvc channel to increase stiffness and uniformity
2. A water source with a pump to recycle flow to maintain a constant head
3. An indoor facility to shield from influences of wind and extreme temperatures
The rock cross vane acts as a long-crested weir because the overflow surface is
greater than the width of the channel. It would be beneficial to calibrate a weir
equation which would be used for better modeling of flow depths and discharge
around the structure.
63
Literature Cited
Chow, V. T. 1959. Open-Channel Hydraulics. McGraw-Hill, Inc.
Henderson, F. M. 1966. Open Channel Flow. Prentice-Hall, Inc.
“Hydraulic Design of Energy Dissipators for Culverts and Channels

Hydraulic Engineering Circular Number 14, Third Edition.” U.S. Department of
Federal Highway Administration. Available at:
http://www.fhwa.dot.gov/engineering/hydraulics/pubs/06086/hec14ch06.cfm.
Lawrence, R. A. “The Effects of Vanes on Pool Development.” PowerPoint

presentation. Available at:
http://www.rrnw.org/Skamania2005/presentations/session%208%20-
%20Lawrence.pdf. Accessed 20 October 2006.
Malcom, H. R. 1989. “Elements of Stormwater Design.” North Carolina State

University.
Rosgen, D. L. 2001. “The Cross-Vane, W-Weir and J-Hook Vane Structures...Their

Restoration.” Wildland Hydrology, Inc, 22 pp.
Rosgen, D. L. 2004. Unpublished Material.
Thornton, C. I., Heintz, M. L., Abt, S. R., Baird, D. C., and Padilla, R. S. 2005.
"Effects of Bendway Weir Characteristics on Resulting Flow Conditions."
ASCE Conference Proceedings.
64
Chapter 3:
Rock Cross Vane Rapid Assessment Tool
Abstract
While hydraulic modeling of rock cross vanes is beneficial to better understand the
effects of the rock cross vane on flow, a qualitative assessment is needed to
evaluate how well the structures are performing in practice. Due to the lack of a
standard assessment tool for the rock cross vane, several stream restoration sites
were observed to build a Rock Cross Vane Rapid Assessment Tool (RCV-RAT)
which includes a field assessment form and a failure guidebook. In conjunction with
rating levels and occurrences of various failure types, the tool helps the assessor
identify potential primary and secondary causes of the failure, via construction or
design insufficiencies. The failure guidebook was developed to supplement the field
assessment form and provide rating criteria for the given set of failures.
Keywords: rock cross vane, failure paths, Rock Cross Vane Rapid Assessment
Tool, filed assessment form, failure guidebook
65
Introduction
Dilemmas in Discussion over the Rock Cross Vane
The rock cross vane is a widely used stream restoration structure in North Carolina
stream restoration projects. Because of their high cost and frequency of use,
designers should factor in a level of risk of structure failure into stream design. The
current dilemma in documenting these failures and calculating risk is the lack of
standards for failure rating and identification. There is no standard assessment form
to provide a common language to be used in the discussion. Once this method is
established, studies can be conducted, compared and used to quantify the
associated risk. As the rock cross vane evolves and new varieties are implemented,
knowledge of failure risk can be used to determine the best practice for the given
circumstance.
The definition of failure is often a source of confusion. Rock cross vanes would
rarely be considered a failure if the assessor were solely observing durability. Rock
cross vanes are designed to withstand 100 year storms at the least, and therefore,
typically only washout during cataclysmic events such as hurricanes and the intense
flooding that ensues. There have been several studies conducted on structure failure
(Brown 2000; Frissel and Nawa 1992; Roper et al. 1998), though none specifically
on the rock cross vane. When defining failure in terms of performance of a given set
of functions, standards for assessing performance are lacking. This problem of how
failure is defined becomes a hindrance to comparing and coordinating studies.
Failure Studies
Roper et al. (1998) studied 4,000 rock and log structures in the Pacific Northwest for
durability, defined by structure shift and structure absence. They used USGS stream
gages to estimate flow events. They found that as the return period of the flood
increased, the durability decreased. Also, as stream order increased, the durability
decreased. Higher flow rates and depths are expected in higher return period
storms. Similarly, higher order streams and rivers are likely to experience higher
66
peak flows than first and second order streams in the same watershed. There was
also interaction between the two factors (Roper et al. 1998). According to this Pacific
Northwest study, the structures studied had fairly high durability. Roper commented
that the watersheds they observed had lower peak flows than in other studies, which
may have contributed to the high durability they observed. The researchers found
comparing their study to other studies to be problematic because there was not
standard measurement for success (Roper et al. 1998).
Brown (2000) conducted a qualitative study on structure performance in urban

streams based on function rather than durability, though durability was also
observed. Frissel and Nawa looked at both functional failure and outright failure or
damage of in-stream structures, however, no rock cross vanes were observed
(1992). Brown ranked cross vanes as excellent in all five of his study categories that
define failure: percent remaining intact, achieved design objective, unintended
erosion/scour, unintended sedimentation/deposition, and achieved habitat
enhancement. Of the 15 rock cross vanes observed, 93% had excellent ratings in all
but the habitat category where 100% were deemed excellent. One problem with the
assessment was that the rock cross vane was only observed for one function: grade
control when there were clearly several other important functions such as bank
protection and pool development. Related structures such as the rock vortex weir,
rock vane, and double wing deflector were listed in the functional category of
deflection/concentration, which is essentially bank protection and pool development
(Brown 2000).
Current Rapid Assessment of Stream Restoration and Structures

Currently, the Clean Water Management Trust Fund along with Dave Penrose of
North Carolina State University’s Water Quality Group is working on development of
a rapid assessment tool for monitoring and assessing streams (Penrose, Personal
correspondence 2006). The Center for Watershed Protection offers Rapid Stream
Assessment Technique (RSAT), a tool used to assess stream conditions based on
qualitative and quantitative observations (Galli 2006). Barbara Doll of Sea Grant and
67
North Carolina State University is developing a rapid assessment tool for completed
stream restoration projects as a monitoring tool. Her tool incorporates stream
functions, habitat and structures (Doll, Barbara Email correspondence. January 23,
2007). With a growing focus on improving design and function of stream restoration
and holding engineers accountable to solid work, it is vital that stream restorations
be assessed for the condition of the whole system and well as the conditions of
individual components.
Rock Cross Vane Design, Construction, and Catastrophes

Causes of rock cross vane failure can be broken down into three main categories:
design flaw, poor construction, and catastrophic events. While the first two are within
the engineer’s control, the third is unpredictable. Several mountain stream
restoration projects were damaged during the floods of tropical storm Frances and
hurricane Ivan in the fall of 2004. Though rock cross vanes are designed to
withstand 100 year storms, compound events such as these create extreme
conditions.
Structures should be designed to handle daily and monthly flow patterns as well as
the expected large scale events. Base flows and small storms are mainly
responsible for the bulk of the sediment load of the stream. The bed and filler around
the structure should be designed to withstand normal flows, neither aggrading nor
degrading. Mid-size events are more likely to cause trouble soon after installation
before the stream has had a chance to settle, stabilize and regain vegetative cover
on banks. According to Bhuiyan and Hey (2001), J-vanes go through a period of
time after installation where sediment is moving through a transforming thalweg
before it beings moving over the point bar. They also may be responsible for bank
erosion around structures. Stream channels are designed for a bankfull flow of a 1.5
- 2 year average recurrence interval. Floods over 2 year average recurrence interval
expected to enter the floodplain. However, the stream will adjust to transport the
supplied sediment (Wilcock 2004). Mean monthly flows can be used to estimate the
shear stress on the banks. While these normal flows may not cause mass
68
movement of structures, they are responsible for maintaining stability adjacent to the
structure. If the bank around a cross vane arm has been continually eroding, the arm
has a greater risk of being pushed over during a catastrophic event.
Catastrophic events should also be designed for as they are the ones that do the
greatest damage in a single event. According to a risk table published by Stream
Systems Technology Center (1998), if a cross-vane has a design life of 25 years
with a 10% chance of failure, it should be designed to withstand a 238 year storm.
While monitoring records may not contain a storm of that size, peak flow estimates
could be determined from rainfall probability curves.
A Qualitative Assessment for the Rock Cross Vane

The remainder of this chapter provides the tool for assessing the rock cross vane.
The Rock Cross Vane Rapid Assessment Tool (RCV-RAT) includes a field
assessment form and a failure guidebook. These help rate failure according to four
functions and two categories of durability and identify causes of failure.
Terminology Used
The following terms are used in the RVC-RAT:
1. Failure indicator describes the visual clues that a function is not being
performed or that there is a lack of durability.
2. Primary causes of failure are the actions of the stream that directly cause
failure. Primary causes may interact with each other to cause failures, or they
may be independent.
3. Secondary causes of failure concern design or installation flaws that cause

primary causes of failure. They in themselves do not always imply rock cross
vane failure, but may lead to failure via the primary causes. There are 21
identified secondary causes of failure.
69
4. Mechanisms of failure are the combination of a primary and secondary cause
of failure that lead to a failure indicator.
5. Path of failure refers to the combination of one failure indicator, one primary
cause and one secondary cause.
Step by Step Process for Field Use

Only failure indicators are rated. These are what the success or failure of the
structure is based on. The primary and secondary causes are not rated. They are
identified as possible contributes to the given failure indicator. The following
procedure explains how to assess a rock cross vane using the field assessment
form and failure guidebook.
1. The structure should be assessed for the presence of failure indicators: arms
washed out, sill washed out, head cut, bank erosion at the structure,
downstream bank erosion, and lack of scour pool development. These are
discussed in more detail in the failure guidebook.
2. Assign a rating of 0, 1, 3, or 5 to each failure indicator based on the guidance

of the failure guidebook.
3. Identify the primary causes of the failure indicator based on clues outlined in
the failure guidebook.
4. Identify the secondary causes of the primary causes based on clues outlined
in the failure guidebook.
5. At the end of an assessment, the structure should have a rating of each

failure indicator with primary and secondary causes identified. These ratings
are not combined to give an overall score for the structure. Combining the
ratings might mask an extreme occurrence of failure of one function when no
other indicators were present.
70
Use of the Failure Indicator Ratings
There are several potential benefits for gathering rating information. The ratings
could provide a standard measurement of the structure failure and funding agencies
could specify a threshold rating for each function used to determine whether or not
repairs need to be made. The following are suggested actions for ratings:
1. Rock cross vanes that have a function rating 1 should be monitored for further
changes and if the condition does not worsen can be considered stable.
2. Rock cross vanes with a function rating of 3 should be more closely

monitored and possibly repaired. For bank erosion, efforts should be made to
armor the banks with more vegetation. For problems with head cut formation,
grade should be restored upstream of the sills and footers.
3. Rock cross vanes with functions rating 5 should be repaired by replacement

or other method depending on the functional impairment.
Having accepted standards of function ratings would improve communication about

structure failure and aid in the development of standards for assessing stream
restoration projects. The second use of the ratings is to achieve both estimates of
occurrence of a given type of failure and severity to pinpoint weakness in rock cross
vane design and develop more robust designs. Finally, the occurrence of the failures
can be used to help determine the risk of the structure. This is useful when
calculated Risk Priority Numbers to help determine the best structure for a specific
design application (Johnson 2004).
Purpose of Failure Path Identification

The path of failure is the combination of a failure indicator, primary and secondary
cause. Identifying paths of failure enables researchers to improve their practices by
understanding which elements of the design are key components of failure and
make recommendations for future installations and designs. Identification of failure
paths might also help an assessor determine the best way to repair the structure.
71
For instance, if the secondary cause of a given failure was identified as the rock
cross vane being undersized, the assessor could recommend that the structure be
widened upon repairs.
Tool Development
Five sites were initially evaluated in the development of the Rock Cross Vane Rapid
Assessment Tool (RCV-RAT). Yates Mill, Shepherd, and Little Brasstown and Little
Brasstown Tributary were each visited two times. Beaver Creek and Rocky Branch
were visited once each.
Development of the field assessment form began with a list of seven failure
indicators: durability, head cut, bank erosion at the vane, downstream bank erosion,
insufficient pool development, and downstream sediment deposits. After the initial
visits, durability was subdivided into sill washout and arm washout as the two have
very different implications of functional failure associated with the lack of durability. A
sill washout could lead to problems with grade control. An arm washout could lead to
problems with bank erosion at the vane and immediately downstream of a cross
vane. “Sill missing” was a failure indicator that when identified shifted to a secondary
cause of failure during data analysis because it resulted in undercutting, a primary
cause of failure. A missing sill does not inherently indicate a failure as the vane
could have been improperly installed without one, and if placed on bedrock have no
incidence of a head cut. Downstream sediment was removed as a failure indicator
as linking stream sediment deposits to one point source can be problematic in
streams with high sediment movement or many spots of active scour.
An extensive list of possible causes of each failure indicator was initially drafted,
reorganized and trimmed down as redundancies were identified. The causes of
failure were divided into two categories: primary causes of failure and secondary
causes of failure. Secondary causes (such as steep arms) are not deemed as
inherent failures unless they lead to a failure indicator. Therefore, the data collected
can be used to assess the frequency of the failure indicators and which primary and
72
secondary causes initiated them, but it cannot be used to assess how a unique
combination of primary and secondary causes initiated a failure indicator. As the
field assessment form was revised and initial stream visits were made, it was
imperative to develop a guidebook to document how levels of a failure indicator were
ranked and how to identify the indicators’, primary and secondary causes. These
ratings were adjusted as the scope of failures was observed in the field.
The field assessment form and failure guidebook should not be used in place of a
full stream monitoring and assessment plan, but rather could be incorporated into
the plan as a way to assess the rock cross vane. Its methods could also be used to
develop similar tools for other in-stream structures.
Field Assessment Form

Table 3.1 is the field assessment form of the RCV-RAT. This form was developed
through an iterative process of visiting streams and identifying all potential failures
and then grouping these in more general groups.
73
Table 3.1 Filed Assessment Form for the Rock Cross Vane Rapid Assessment Tool
Cross vane # indicator primary secondary
Date/Time S.05 arms not tied in
P.01 direct
Notes S.09 arms washed out
contact of flow
with banks S.12 spacing of boulders
S.17 exposed banks
P.02 flow S.01 improper alignment
score each indicator from 0-5 based on guidebook directed at banks S.03 placed in a bend
indicator primary secondary S.12 spacing of boulders
P.05 piping
S.01 improper alignment S.18 insufficient backfill
S.03 placed in a bend S.01 improper alignment
P.03 drag
S.05 arms not tied in S.03 placed in a bend
and lift or
tipping S.12 spacing of boulders F4. bank S.05 arms not tied in
S.14 undersized boulders erosion at S.06 arms too short
vane P.07 side cutting
S.18 insufficient backfill S.09 arms washed out
F1. arm
S.01 improper alignment S.10 sill too high
washout
S.03 placed in a bend S.12 spacing of boulders
S.08 arms too steep S.21 undersized rcv
P.08
S.16 drop too high S.01 improper alignment
undercutting
S.18 insufficient backfill S.03 placed in a bend
S.19 no footers S.08 arms too steep
P.08
S.21 undersized rcv S.16 drop too high
undercutting
S.10 sill too high S.18 insufficient backfill
P.03 drag S.12 spacing of boulders S.19 no footers
and lift or S.14 undersized boulders S.21 undersized rcv
tipping S.18 insufficient backfill S.05 arms not tied in
P.01 direct
S.21 undersized rcv S.06 arms too short
contact of flow
F2. sill S.02 backed into a pool with banks S.09 arms washed out
washout S.08 arms too steep S.17 exposed banks
S.10 sill too high P.02 flow S.01 improper alignment
P.08 directed at banks
S.16 drop too high S.03 placed in a bend
undercutting
S.18 insufficient backfill P.04 flow S.08 arms too steep
S.19 no footers expansion out of S.13 boulders in pool
S.21 undersized rcv F5. vane S.21 undersized rcv
S.12 spacing of boulders downstream S.01 improper alignment
P.05 piping bank erosion
S.18 insufficient backfill S.03 placed in a bend
S.01 improper alignment S.05 arms not tied in
S.03 placed in a bend P.07 side cutting S.06 arms too short
S.05 arms not tied in S.09 arms washed out
P.07 side S.06 arms too short S.10 sill too high
cutting S.09 arms washed out S.12 spacing of boulders
S.10 sill too high S.09 arms washed out
P.09 weak jet /
F3. head S.12 spacing of boulders S.15 drop too short
low velocity ratio
cut S.21 undersized rcv S.20 oversized rcv
S.02 backed into a pool S.09 arms washed into pool
P.06 protected
S.03 placed in a bend S.11 sill washed into pool
from scour
S.08 arms too steep S.13 boulders put in pool
F6.
P.08 S.10 sill too high S.04 placed on bedrock
insufficient
undercutting S.11 sill washed out S.07 arms too flat
scour pool
S.16 drop too high P.09 weak jet S.09 arms washed out
S.18 insufficient backfill S.15 drop too short
S.19 no footers S.20 oversized rcv
74
Failure Guidebook
This guidebook is to be used in association with the rapid assessment tool to assess
the rock cross vanes. It includes four sections: indicators of failure, primary causes
of failure, and secondary causes of failure, and failure paths.
Indicators of Failure
The rock cross vane has three main functions: 1) maintaining grade of the channel
by fixing the bed elevation at a rock checkpoint, 2) protecting banks by redirecting
flow away from banks, and 3) establishing a scour pool at the vane by a drop and
increasing center velocities relative to near bank velocities (in Chapter 2 this is
referred to as increasing the velocity ratio). When any of these functions are not
occurring, the vane has failed unless it was not designed to perform the function.
Therefore the three main failures of cross vanes are: 1) Lack of Grade Control, 2)
Lack of Bank Protection, and 3) Lack of Pool Development. A final expectation of the
cross vane is that it be durable. When the boulders washout from the structure, the
cross vane lacks durability. Table 3.2 below lists the failures and their main
indicators.
Table 3.2 Failures and Major Indicators

Failure Indicator
Arms washed out
Lack of Durability
Sill washed out
Lack of Grade Control Headcut
Bank Erosion at Vane
Lack of Bank Protection
Downstream Bank Erosion
Lack of Pool/Pattern
Insufficient Scour Pool
Development
75
F1. Arms washed out
An arm washout is the movement of one to many of the arm boulders due to a heavy
force or tipping. The arm boulders may actually move downstream, shift out of place,
or simply fall into the pool. This mainly occurs when the boulders are too small for
the stream size, but it can also occur when the scour pool expands so large that
undercutting of the arms causes the boulders to tip into the pool. Also, strong side
cutting may be related to this indicator. Arm washouts lead to bank erosion and can
also lead to a head cut if the main flow begins to travel through the washout at a
lower grade than the sill. This particular indicator, once developed, can become a
secondary cause of failure as it allows for side cutting which can lead to multiple
other failures. In Figure 3.1, a boulder on the left arm is missing. Most of the base
flow is side cutting through that segment which has led to a head cut as seen by the
sill stone(s) being fully exposed when it ought to be level with the stream bed.
Several things might have caused the arm to wash out.
Assessment Scale: If less than 10% of the arms have washed out (in small streams,
this translates to one or two boulders missing) or if observable signs of boulder
shifting are present, the score for this indicator is 1. If 11% to 30% of the boulders
are missing (Figure 3.1), the score for this indicator is 3. If greater than 30% of the
boulders are missing, the score indicator is 5. If this indicator is not present, the
score is 0.
76
Figure 3.1 Arm Washout
F2. Sill washed out
A Sill washout is the movement of one or multiple boulders of the sill due to heavy
force or tipping. The sill boulders may actually move downstream, shift out of place
or simply fall into the pool, but there is no longer a boulder in place to maintain
grade. This mainly occurs when the boulders are too small for the stream size, but it
can also occur when the scour pool expands so large that undercutting of the sill
causes the sill to tip into the pool. A sill washout can lead to a head cut migration
upstream. This particular indicator, once developed, cycles back to being a
secondary cause of failure.
Assessment Scale: If observable signs of boulder shifting are present (Figure 3.2) or
if there is moderate to extreme side cutting and/or undercutting of the sill stones, the
score for this indicator is 1. If any boulders are missing, the score for this indicator is
3. If there are no boulders, the score indicator is 5. If this indicator is not present, the
score is 0.
77
Figure 3.2 Cross vane with a portion of the sill shifted or missing
F3. Headcut
A head cut is defined by the NC Division of water quality as “an abrupt vertical drop
in the bed of a stream channel that is an active erosion feature” (2005). It can also
be referred to as an upstream migration of bed degradation. A head is observed by a
lower bed elevation just upstream of the sill stones of the rock cross vane than the
sill stones. This could be an indication that the upstream bed is degrading. Another
way to observe a head cut is to feel under the sill so see if the bed is eroding out
from under the sill.
Assessment Scale: If observable signs of boulder shifting are present or if there is

slight side cutting and/or undercutting of the sill, the score for this indicator is 1. If the
undercutting travels all the way under the sill or if the bed around the sill has begun
down-cutting (as opposed the sill simply being installed above bed elevation), the
score for this indicator is 3. If a visible head cut has occurred and/or is migrating
upstream (Figure 3.3), the score indicator is 5. If this indicator is not present, the
score is 0.
78
Figure 3.3 Head cut at rock cross vane looking downstream
F4. Bank Erosion at Vane
Erosion at the vane may be observed by scoured portions of the banks, gaps
between the boulders of the arm and bank with exposed soil, and a vertical toe of
the banks where they connect with the arms. Slight erosion at the vane is visible by
slight scour behind the arms and a lack of toe at the base of the arms. Moderate
erosion is visible by decreased vegetative cover on banks, larger portions of scour
behind the arms with flow tunnels through and around the arms, and scour at the toe
of the banks. Extreme erosion at the vane is visible by scoured banks with large
gaps between arm boulders and the banks and clear flow paths under or around the
arms.
Assessment Scale: If the vegetation has been washed off at the toe of the banks
and there are sparse patches of erosion on the banks, less than 25% of the area,
the score is 1. If the patchy erosion covers 25 to 50% of the banks, bank slumping
has occurred but is still covered by vegetation, vegetation has been washed off the
banks at half-bankfull, or there are small bank scour spots, the score is 3. If most of
the vegetation up to bankfull has been washed off, if the banks have slumped or
79
sheared off, of if there are large areas of deep scour (Figure 3.4), the score is 5. If
there is no indication of bank erosion, the score is 0.
Figure 3.4 Bank erosion at cross vane leading to side cutting and headcut
F5. Downstream Bank Erosion
Downstream bank erosion usually occurs immediately downstream of the ends of

the rock cross vane’s arms to about two bankfull widths downstream of the sill. Slight
erosion is visible by the vegetation mostly covering the banks with small spots of
exposed soil and scour. Moderate erosion is visible by the vegetation only partially
covering the banks and larger portions of bank scour. Extreme bank erosion is
visible by no vegetative cover and large portions of bank scour and sloughing.
Downstream bank erosion typically occurs when there is a sharp bend coming out of
the vane causing high shear stress on the banks or when flow is constricted through
the vane and then expands upon exiting. The arms of the vane redirect flow to the
center of the vane, and therefore, when the arms are malfunctioning, the near bank
velocities may remain high causing bank erosion. In cases where serious side
cutting behind the arms is occurring, downstream bank erosion may also be a
problem.
80
Assessment Scale: If the vegetation has been washed off at the toe of the banks
and there are sparse patches of erosion on the banks (less than 25% of the area)
the score is 1. If the patchy erosion covers 25 to 50% of the banks, bank slumping
has occurred but is still covered by vegetation, vegetation has been washed off the
banks at half-bankfull, or there are small bank scour spots, the score is 3. If most of
the vegetation up to bankfull has been washed off, if the banks have slumped or
sheared off, of if there are large areas of deep scour (Figure 3.5), the score is 5. If
there is no indication of bank erosion, the score is 0.
Figure 3.5 Flow Expansion with Downstream Erosion
F6. Insufficient Scour Pool
An insufficient sour pool for this study is defined as a scour pool that is not stable
and errs on the side of aggrading rather than continued scour. Characteristics of this
are fines in the pool indicating flow slow enough for particles to settle out. Another
characteristic of poor scour pool development is boulders and cobbles in the pool
from the members of the cross vane. A slightly insufficient scour pool would have
some fines or a couple cobbles but still have some depth. A moderately insufficient
81
pool would be much shallower with more apparent accretion or filling in. An
extremely insufficient pool would be the lack of a pool at all.
Assessment Scale: If the pool is shallow as compared to the designer’s intensions

(compare pool to other pools in the pattern sequence or ask designer), the score is
1. If the pool is shallow and there is accretion of larger material in the pool or
washed out boulders from the vane, the score is 3. If the pool is non-existent or if the
pool is shallow and has accretion of fines (Figure 3.6), the score is 5.
Figure 3.6 Shallow Pool with Vegetation
Primary Cause of Failure

Primary causes of failure are the actions of the stream that directly cause failure.
Primary causes may interact with each other to cause failures, or they may be
independent.
P01. Direct contact of flow with banks
Exposed banks are those which are not protected by vegetation or armor. This
leaves the banks vulnerable to erosion and scour because they are in direct contact
with the flow. This is compounded or may be initiated when the flow is directed at the
82
banks or there is flow constriction through the vane. Direct contact of flow with the
banks is identifiable by raw banks with active or inactive sour or erosion. Because
this is the same evidence of bank erosion, a general lack of protection on the banks
would help show the stream was designed or constructed with a lack of bank
protection. This primary cause can be cyclical as exposed banks leads to bank
erosion, which leads to further exposure.
P02. Flow directed at banks
Flow that is directed at the banks usually occurs from incorrect placement of the rock
cross vane or improper alignment. Flow should be entering the vane parallel and on
center to the vane. When the banks are exposed, this can lead to severe bank
erosion and scour.
P03. Drag and Lift or Tipping Forces on Boulders
When the drag force is great enough, boulders may actually be lifted into the bulk of
the flow and carried downstream. Also, boulders may be tipped off their footers and
roll downstream, especially when the boulders have been undercut or are
overhanging the footers. Movement of the boulders into the downstream pool can
lead to filling in of the scour pool or downstream sediment bars forming. It also
weakens the durability of the vane. Excessive drag force on boulders, when the
boulders become separated from the banks of the bed can cause the banks to erode
and scour as well as a head cut to form as the cross vane is no longer holding
grade.
Drag force is in the x direction:
D = ∫ dFx = ∫ p cosθdA + ∫ τ w sin θdA (1)
Where,
D = Drag Force
p = Pressure on surface
83
A = Area of surface
τw = Shear stress on area
θ = Angle of flow to the incremental surface area as a function of location
along the body (Munson et al. 1990).
Lift force is in the y – direction (Munson et al. 1990):
L = ∫ dFy = ∫ p sin θdA + ∫ τ w cosθdA (2)
Figure 3.7 is a diagram of drag and lift forces.
Figure 3.7 Diagram of Drag and Lift
When forces of drag and lift exceed the forces of the water weight and opposing
friction forces, the boulder will move.
There are several components of design and installation which might lead to
excessive drag force on the boulders. First, increasing the amount of exposed
surface area perpendicular to the flow decreases the amount of force needed to
move the boulder. When sills are placed above the bed elevation, less force is
required to move them. Increased velocity causes a higher drag force on a
stationary body and relative velocity is simply the velocity of the flow. When the arms
84
are in a bend, the arms on the outside of the bend experience higher velocities and
are therefore more at risk of failure than those on the inside of a bend. Boulders with
rough angular surfaces rather than stream-lined surfaces also experience a higher
drag force. This is problematic in stream construction as most boulders used are
quarried stone.
P04. Flow area constriction through rock cross vane
Flow expansion coming out of the vane occurs when there is constriction of flow
through the cross vane. Constriction occurs when the rock cross vane is undersized,
when the sill is elevated above bed elevation, or when the pool is crowded by
boulders, meaning the cross sectional area is small compared to the representative
cross sectional area of the stream. Flow expansion is harmful to the banks just
downstream of the cross vane especially when the banks are not armored or
protected with vegetation. In the case where the flow backs up over the vane and
enters the floodplain, re-entry into the stream over the downstream banks may also
cause bank erosion. Clues to this are heavy bank erosion at the downstream ends of
the arms from re-entry of flow or a large scour pool just downstream of the arms of
the cross vane. Evidences of constriction through the rock cross vane in Figure 3.8
are the side cutting of the arms and the downstream scour. Causes are potentially
an undersized structure and crowding of the pool by boulders.
85
Figure 3.8 Constricted Rock Cross Vane
P05. Piping
Piping is the occurrence of flow between the boulders of the cross vane. It becomes
a problem when the piping becomes so large at the sill that a head cut forms or
when the durability of the vane is threatened by exposing the side surfaces of the sill
stones (Figure 3.9). Piping will not cause a washout, but it weakens the connections
to the bed and other boulders and decreases the amount of drag force needed to
move the boulders of the sill or arms, which can eventually lead to a washout. A
small amount of piping is expected to occur but large amounts can be prevented by
careful boulder placement, where the faces of the boulders that touch do not have
large gaps between them. Also, fabric matting and good backfill prevent piping.
86
Figure 3.9 Piping caused by boulder spacing and poor backfill
P06. Boulders in the Pool Region Hinder Scour
While this primary cause deviates from the definition of primary causes, effects of
the flow, it is classified here as a primary cause because the flow is incapable of
scouring out the pool as intended. However, the problem lies not in an insufficient jet
based on the vane design, but in the fact that boulders have washed out or been
placed into the area where a pool should be scoured.
P07. Side cutting
Side-cutting refers to the main flow of the stream flowing to either side of the sill
stones and eroding or cutting out a new thalweg. In the case of a Type 1, 2 or 3
side-cut, head-cuts may occur as the cross vane is no longer maintaining grade.
Type 2 and 3 side-cuts may eventually lead to bank failure as the arms are not
shielding the banks and the flow is not traveling in the center of the stream. Table
3.3 includes descriptions of each type.
87
Table 3.3 Types of Side cutting
Type 1
Water bypasses the sill and flows to one side or another of the sill, still within the
center portion of the vane (not behind the arms). This happens mainly when the
sill is too high above the bed elevation and is no longer part of the thalweg.
Type 2
Water flows over a portion of either arm that has either washed out or by fault of
construction is lower than the sill. Because water is flowing towards the arm, flow
is no longer in the center and the banks are at risk.
Type 3
Water flows behind the arm(s) of the cross vane. Not only have the banks
experienced damage, but the boulders of the arm are no longer stable and risk
tipping into the scour pool.
P08. Undercutting
Undercutting is used to describe the scour that occurs at the toe of the banks or
under the boulders of the rock cross vane. Undercutting is not detrimental until is
threatens the durability of the structure or until a large amount of flow cuts through
an undercut region. This can also lead to bank erosion and scour. A head cut can be
caused by an undercut sill where the scour pool is eating out the bed under the sill
from downstream of the sill or where the stream is cutting down under the sill from
upstream (can usually be prevented by good backfill and fabric matting coupled with
appropriate placement of the vane). Undercutting maybe be caused by too high of a
drop, the vane being backed into a pool, the sill installed about the stream bed,
improper alignment, insufficient backfill, no footers, or because the vane was placed
in a bend.
Around boulders, there are two main forms of undercutting: from the upstream side
and from the downstream side. From the upstream side, undercutting can be
initiated by problems with piping or a lack of fill behind the boulders. Another
common problem observed is the placement of cross vanes at the end of pools in
the riffle-pool sequence. When the vane is backed into the pool, which actively
scours and deposits sediment during larger storms, the bed just upstream of the
88
boulders may be scoured out weakening the durability of the boulder, increasing its
susceptibility to tipping or rolling.
From the downstream side of boulders, undercutting can be initiated by high drops
causing an expanding scour to work its way to the base of boulders and constricted
flow scouring beneath structures as energy attempts to dissipate. An excessive jet
impingement is caused by overflow from the step falling too far a distance for the
designed pool depth. An excessive jet impingement is one that sours out a larger
pool than designed for and one that is leading to an expanding scour pool or a pool
that is continuing to grow. As the pool grows, more and more of the bed of the pool
is scoured which undercuts the arms and sill making them vulnerable to tipping into
the pool. High shear is a problem when it leads to degradation of the stream,
meaning the sediment transport capacity is too high for the stream. Through the rock
cross vane, high shear can lead to bank erosion at the vane and downstream of the
vane. It can also lead to an enlarged scour pool that may threaten the durability of
the arms and sill boulders if it is coupled with severe undercutting.
P09. Weak Jet
Insufficient shear is expected to occur near the banks through and after a rock cross
vane as a measure of protection via slowing of near bank velocities and increasing
center velocities. A weak impingement jet is a problem when it prevents proper
souring of a pool and deposition through the pool. This can be caused by a vane that
is oversized (the cross sectional area is too large) or by an inadequate drop. Also,
bed material may simply be too large to be scoured out. A weak jet impingement is
one that is not strong enough to scour out a sufficient (as determined by the
designer) scour pool. This jet is primarily determined by the height of the drop and
steepness of the arms.
The rock cross vane is designed to increase the center velocities of the velocity
distribution to the outer velocities of the distribution. This protects the banks and
scours the pool and thalweg appropriately. If this fails to occur, the outer velocities
89
remain high, and banks are not protected. Also, the scour pool may end up too
shallow. Insufficient increase in center velocities leads to insufficient shear in the
areas where the shear is designed to be higher. This is not differentiated from
insufficient shear as it is very difficult to identify. An insufficient increase in center
velocities was shown to be mainly a function of drop in Chapter 2. Insufficient drop
would lead to both an overall low shear and low velocity ratio.
Secondary Causes of Failure

Secondary causes of failure concern design or installation flaws that cause primary
causes of failure. They in themselves do not always imply rock cross vane failure,
but may lead to failure via the primary causes. There are 21 identified secondary
causes of failure.
S01. Improper alignment
Flow should approach the vane perpendicular to the vane. If the flow approaches at
an angle (Figure 3.10), it has improper alignment. Improper alignment does not allow
the vane to function as it was designed and can leave the banks vulnerable.
Sometimes improper alignment is simply an installation error and oftentimes it
occurs do to shifting of the thalweg caused by rapid lateral migration of the stream.
The RCV-RAT does not differentiate between the two as global stream design errors
are likely to carry over to cross vane design errors.
90
Figure 3.10 Cross vane with improper alignment
S02. Rock cross vane placed in a pool
If the rock cross vane sill begins in a section of a pool (Figure 3.11), backfill may
easily scour out, and piping, undercutting and a head cut may develop.
Figure 3.11 Cross vane backed into a pool
91
S03. Rock cross vane placed in a bend
Flow should enter and exit perpendicular to the vane. Placing the vane within a bend
makes the flow enter at an angle other than perpendicular to the structure leaving
the banks at the vane vulnerable (Figure 3.12). If there is a tight bend out of the
vane, the flow will be directed at downstream banks. Because there are pools in the
bends, this category may at times overlap with the previous one, backed into a pool.
The distinction comes with the bend leading to problems with bank vulnerability and
the pool leading to problems with sill and head cut vulnerabilities.
Figure 3.12 Undersized Cross Vane in Bend
S04. Placed on bedrock
Placing the vane on bedrock prevents a scour pool from developing. On bedrock,
even high shear would be too low to scour out a pool. Bedrock serves as a natural
checkpoint so the grade control component of the rock cross vane is unneeded. The
only benefit of a rock cross vane on bedrock is bank protection, which could be
achieved by single arm vanes. Occasionally designers will place a rock cross vane
on bedrock with bank protection as the main goal of the structure. If bedrock
92
placement is the cause of a lack of scour pool development, it should be questioned
whether or not this was an intended function of the structure (Figure 3.13).
Figure 3.13 Vane on Bedrock
S05. Arms do no tie into banks
The ends of the arms are to be keyed into the banks. This does not mean they
should be tied in to the top of bank, but rather be buried into the banks act the ends.
The keying of the arms adds extra protection in the case of bank sloughing or
erosion as it can provide a barricade for backfill and bank material washing around
the sides of the cross vane. This is a difficult secondary cause to notice and may
only be noticeable in the cases of severe bank erosion; in these cases, the key
stones may have been installed but washed downstream.
S06. Arms too short
Arms that are too short allow large volumes of flow over the ends. These and
especially ends that do not get buried into the banks allow for bank scour at the end
of the arms when coupled with flow constriction through the vane. Because the
seam between boulder and bank is highly vulnerable, especially where there is low
93
vegetation, it is essential to slow down the flow at the ends of the arms. This can be
achieved by ensuring that the area of flow over the arms does not constrict flow.
Therefore, short flat arms or longer arms would have a lower vulnerability to this
downstream scouring and short steep arms (Figure 3.14).
Figure 3.14 Downstream scour from short steep arms
S07. Arms too flat
Flat arms do not necessarily threaten durability or stability, but flat arms do not
create as high of a velocity ratio as steeper arms. This is the only secondary cause
that was not observed in the subsequent evaluation of North Carolina rock cross
vanes. Instead, observations showed that flatter arms allowed for increased bank
stability.
S08. Arms too steep
Steep arms can cause flow constriction or overtopping when the rock cross vane is
not widened to accommodate the steep arms and the resulting flow area over the
rock cross vane is less than the approach flow area. Steep arms are also vulnerable
to scour upstream of the boulders between the boulders and the banks. The abrupt
94
change in bed slope causes the water to slam up against the upstream side of the
arms. Steep arms are identifiable by visual inspection (Figure 3.15). Typically they
tie in to bankfull even at higher arm angles. Also, arms that are too steep and lead to
constriction will often be eroded at the ends where water that has overtopped the
vane re-enters the channel.
Figure 3.15 Steep Arm on Right Bank
S09. Arms washed out
If arms have washed into the pool they might prevent scour pool development. They
might also lead to bank erosion because they leave banks unprotected, allow side
cutting, and no longer serve to redirect flow to the center.
S10. Sill too high
The sill of the cross vane should be placed at bed elevation. If the sill is higher, then
the flow will find a way around it by side cutting or by beginning to scour out around
the sill and cause piping and potentially cause undercutting. Often times it can be
difficult to determine whether the sill has been installed too high or if scour has
already occurred around the sill stones of the sill (Figure 3.16).
95
Figure 3.16 Cross vane sill installed too high
S11. Sill washed out or missing
A lack of a sill could be due to a sill washout or that a head was simply not installed.
The vane is not maintaining grade if there is no sill. This does not imply that there is
a headcut, as cross vanes are occasionally installed in areas of the stream where
there is natural grade control. If the sill has washed into the pool, it may prevent
proper pool scouring and lead to a head cut (Figure 3.17). If a sill was not installed, a
headcut could occur.
96
Figure 3.17 Sill and backfill shifted downstream to pool
S12. Poor spacing of boulders
Boulders should be placed tightly together. The faces of boulders should have
similar surface shapes there are no gaps between to prevent piping (Figure18).
Figure 3.18 Poor Boulder Spacing
97
S13. Boulders placed in pool
It is not uncommon to see boulders placed in the expected pool region of the rock
cross vane (Figure 3.19). This can be a corrective measure for vanes with too steep
of a drop aimed at preventing excessive erosion.
Figure 3.19 Boulders placed in pool to protect drop
S14. Undersized Boulders
Boulder size is specified by the designer and is dependent on the shear stress and
sediment transport of the stream. Undersized boulders lack the weight needed to not
shift during large flows and may easily move. When boulders washout (Figure 3.20),
a head cut may occur or banks may be left unprotected.
98
Figure 3.20 Undersized Boulders
S15. Drop too short
The drop is the main mechanism of scour pool development. The drop provides the
head differential that produces the impinging jet. If the drop is short, then the
impinging jet will not be powerful enough to scour out the pool. It only becomes a
problem when pool habitat is prevented.
S16. Drop too high
A drop that is too high is one that creates an impingement jet that is scouring out the
pool such that an expanding scour pool develops and causes undercutting at the
base of the boulders. A drop less than one foot is recommended for fish passage.
S17. Exposed Banks
Bank armor at the vane consists of vegetation, the boulders of the arms and large
backfill material. Downstream, bank armor can be applied at the toe to prevent scour
from expansion. Bank armor is not a common practice with cross vanes since the
purpose of increasing the velocity ratio is to protect the banks by slowing down the
near-bank velocity. In circumstances where the vane is too narrow and flow
99
expansion is scouring the banks, bank armor might be a good measure for
prevention of continued scour. Vegetation helps prevent bank erosion and scour by
stabilizing the banks with root networks and leaf coverage. When these are missing,
the banks are left exposed and can easily scour even at low velocities (Figure 3.21).
Figure 3.21 Lack of vegetative bank protection
S18. Insufficient backfill and or Fabric Material
Boulders should be supported by diversely sized materials from cobbles to fines just
upstream to prevent water from piping. The smaller materials fill in the gaps and the
larger ones prevent scour. Backfill also adds stability by providing a smooth
transition from the bed upstream to the top of the boulders. If backfill material is too
small (Figure 3.22), it will easily wash out. If it is large homogeneous material, piping
may still occur. Fabric matting is also essential to the prevention of piping and scour
around the boulders. It can be very difficult to identify whether or not fabric matting
has been used unless significant scouring reveals it is non-existent.
100
Figure 3.22 Evidence of piping caused by insufficient backfill
S19. No footers
Sill boulders are placed on top of footers stones recessed into the stream bed. If
footers are missing or poorly spaced, sills may be undercut (Figure 3.23).
Figure 3.23 Undercut Banks at Arms from insufficient footers
101
S20. Oversized cross vane
If the rock cross vane is too wide (Figure 3.24), the flow area is mush greater than
the representative cross sectional area of the stream and flow will spread out rather
than form a strong jet for pool development.
Figure 3.24 Oversized rock cross vane
S21. Undersized Cross Vane
A rock cross vane that is too narrow is one that has a smaller flow area than that the
approach flow area. In general, steep arms can cause a decrease in cross sectional
area. A narrow vane constricts flow through the structure causing flow expansion
upon exiting the vane. Undersized rock cross vanes inevitably will separate from one
bank or another, potentially both (Figure 3.25). This also increases the vulnerability
to the durability of the boulders because water forces can apply forces to the
boulders in new directions.
102
Figure 3.25 Bank erosion and head cut caused by an undersized vane
Failure Paths
A failure path consists of a failure indicator, a primary cause and a secondary cause.
The Rock Cross Vane Rapid Assessment Tool in Table 3.1 which includes a total 93
failure paths. Table 3.4 lists all failure paths.
Table 3.4 Potential Failure Paths

Path Failure indicator Primary causes Secondary causes
1 F1. arm washout P.03 drag and lift or tipping S.01 improper alignment
2 F1. arm washout P.03 drag and lift or tipping S.03 placed in a bend
3 F1. arm washout P.03 drag and lift or tipping S.05 arms not tied in
4 F1. arm washout P.03 drag and lift or tipping S.12 spacing of boulders
5 F1. arm washout P.03 drag and lift or tipping S.14 undersized boulders
6 F1. arm washout P.03 drag and lift or tipping S.18 insufficient backfill
7 F1. arm washout P.08 undercutting S.01 improper alignment
8 F1. arm washout P.08 undercutting S.03 placed in a bend
9 F1. arm washout P.08 undercutting S.08 arms too steep
10 F1. arm washout P.08 undercutting S.16 drop too high
11 F1. arm washout P.08 undercutting S.18 insufficient backfill
12 F1. arm washout P.08 undercutting S.19 no footers
103
Table 3.4 (Cont’)

13 F1. arm washout P.08 undercutting S.21 undersized rcv
14 F2. sill washout P.03 drag and lift or tipping S.10 sill too high
15 F2. sill washout P.03 drag and lift or tipping S.12 spacing of boulders
16 F2. sill washout P.03 drag and lift or tipping S.14 undersized boulders
17 F2. sill washout P.03 drag and lift or tipping S.18 insufficient backfill
18 F2. sill washout P.03 drag and lift or tipping S.21 undersized rcv
19 F2. sill washout P.08 undercutting S.02 backed into a pool
20 F2. sill washout P.08 undercutting S.08 arms too steep
21 F2. sill washout P.08 undercutting S.10 sill too high
22 F2. sill washout P.08 undercutting S.16 drop too high
23 F2. sill washout P.08 undercutting S.18 insufficient backfill
24 F2. sill washout P.08 undercutting S.19 no footers
25 F2. sill washout P.08 undercutting S.21 undersized rcv
26 F3. head cut P.05 piping S.12 spacing of boulders
27 F3. head cut P.05 piping S.18 insufficient backfill
28 F3. head cut P.07 side cutting S.01 improper alignment
29 F3. head cut P.07 side cutting S.03 placed in a bend
30 F3. head cut P.07 side cutting S.05 arms not tied in
31 F3. head cut P.07 side cutting S.06 arms too short
32 F3. head cut P.07 side cutting S.09 arms washed out
33 F3. head cut P.07 side cutting S.10 sill too high
34 F3. head cut P.07 side cutting S.12 spacing of boulders
35 F3. head cut P.07 side cutting S.21 undersized rcv
36 F3. head cut P.08 undercutting S.02 backed into a pool
37 F3. head cut P.08 undercutting S.03 placed in a bend
38 F3. head cut P.08 undercutting S.08 arms too steep
39 F3. head cut P.08 undercutting S.10 sill too high
40 F3. head cut P.08 undercutting S.11 sill washed out
41 F3. head cut P.08 undercutting S.16 drop too high
42 F3. head cut P.08 undercutting S.18 insufficient backfill
43 F3. head cut P.08 undercutting S.19 no footers
44 F4. bank erosion at vane P.01 direct contact S.05 arms not tied in
45 F4. bank erosion at vane P.01 direct contact S.09 arms washed out
46 F4. bank erosion at vane P.01 direct contact S.12 spacing of boulders
47 F4. bank erosion at vane P.01 direct contact S.17 exposed banks
48 F4. bank erosion at vane P.02 flow directed S.01 improper alignment
49 F4. bank erosion at vane P.02 flow directed S.03 placed in a bend
104
Table 3.4 (Cont’)

50 F4. bank erosion at vane P.05 piping S.12 spacing of boulders
51 F4. bank erosion at vane P.05 piping S.18 insufficient backfill
52 F4. bank erosion at vane P.07 side cutting S.01 improper alignment
53 F4. bank erosion at vane P.07 side cutting S.03 placed in a bend
54 F4. bank erosion at vane P.07 side cutting S.05 arms not tied in
55 F4. bank erosion at vane P.07 side cutting S.06 arms too short
56 F4. bank erosion at vane P.07 side cutting S.09 arms washed out
57 F4. bank erosion at vane P.07 side cutting S.10 sill too high
58 F4. bank erosion at vane P.07 side cutting S.12 spacing of boulders
59 F4. bank erosion at vane P.07 side cutting S.21 undersized rcv
60 F4. bank erosion at vane P.08 undercutting S.01 improper alignment
61 F4. bank erosion at vane P.08 undercutting S.03 placed in a bend
62 F4. bank erosion at vane P.08 undercutting S.08 arms too steep
63 F4. bank erosion at vane P.08 undercutting S.16 drop too high
64 F4. bank erosion at vane P.08 undercutting S.18 insufficient backfill
65 F4. bank erosion at vane P.08 undercutting S.19 no footers
66 F4. bank erosion at vane P.08 undercutting S.21 undersized rcv
67 F5. downstream bank erosion P.01 direct contact S.05 arms not tied in
68 F5. downstream bank erosion P.01 direct contact S.06 arms too short
69 F5. downstream bank erosion P.01 direct contact S.09 arms washed out
70 F5. downstream bank erosion P.01 direct contact S.17 exposed banks
71 F5. downstream bank erosion P.02 flow directed S.01 improper alignment
72 F5. downstream bank erosion P.02 flow directed S.03 placed in a bend
73 F5. downstream bank erosion P.04 flow expansion S.08 arms too steep
74 F5. downstream bank erosion P.04 flow expansion S.13 boulders in pool
75 F5. downstream bank erosion P.04 flow expansion S.21 undersized rcv
76 F5. downstream bank erosion P.07 side cutting S.01 improper alignment
77 F5. downstream bank erosion P.07 side cutting S.03 placed in a bend
78 F5. downstream bank erosion P.07 side cutting S.05 arms not tied in
79 F5. downstream bank erosion P.07 side cutting S.06 arms too short
80 F5. downstream bank erosion P.07 side cutting S.09 arms washed out
81 F5. downstream bank erosion P.07 side cutting S.10 sill too high
82 F5. downstream bank erosion P.07 side cutting S.12 spacing of boulders
83 F5. downstream bank erosion P.09 weak jet S.09 arms washed out
84 F5. downstream bank erosion P.09 weak jet S.15 drop too short
85 F5. downstream bank erosion P.09 weak jet S.20 oversized rcv
86 F6. insufficient scour pool P.06 protected pool S.09 arms washed in pool
105
Table 3.4 (Cont’)

87 F6. insufficient scour pool P.06 protected pool S.11 sill washed in pool
88 F6. insufficient scour pool P.06 protected pool S.13 boulders put in pool
89 F6. insufficient scour pool P.09 weak jet S.04 placed on bedrock
90 F6. insufficient scour pool P.09 weak jet S.07 arms too flat
91 F6. insufficient scour pool P.09 weak jet S.09 arms washed out
92 F6. insufficient scour pool P.09 weak jet S.15 drop too short
93 F6. insufficient scour pool P.09 weak jet S.20 oversized rcv
Conclusions
These tools were used to assess rock cross vanes in North Carolina stream
projects. Chapter 4 presents the results of the assessment.
106
Literature Cited
Bhuiyan, A. B. M. F., Hey, R. D. 2001. “Instream J-Vane for Bank Protection and
River Restoration.” Proceedings of the Congress-International Association for
Hydraulic Research, 161-166.
Frissel, C. A., and Nawa, R. K. 1992. "Incidence and causes of physical failure of
artificial habitat structures in streams of western Oregon and Washington."
NAJFM, 12(1): 182-197.
Galli, J. "Rapid Stream Assessment Technique." Available at:

http://www.stormwatercenter.net/monitoring%20and%20assessment/rsat/smr
c%20rsat.pdf. Accessed 25 October 2006.
Munson, B. R., Young, D. F., Okiishi, T. H. 1990. Fundamentals of Fluid Mechanics,

Third Ed. Update. New York: John Wiley and Sons.
NC Division of Water Quality. 2005. “Identification Methods for the Origins of

Intermittent and Perennial streams, Version 3.1.” North Carolina Department
of Environment and Natural Resources, Division of Water Quality. Raleigh,
NC.
Roper, B. B., Konnoff, D., Heller, D., and Wieman, K. 1998. "Durability of Pacific
Northwest Instream Structures Following Floods." North American Journal of
Fisheries Management, 18(3): 686–693.
107
Chapter 4:
Occurrence, Probability and Risk of Rock Cross Vane

Failures on North Carolina Stream Restoration Projects
Abstract
The Rock Cross Vane Rapid Assessment Tool (RCV-RAT) was used to assess rock
cross vanes on North Carolina stream restoration projects. Failure indicators were
identified and ranked. Potential primary and secondary causes of failure were
assessed. The data were used to determine the most common failure modes and
points of weakness in rock cross vane design. Failure ratings were also regressed
versus stream data such as watershed area, age of project, D50, and project length,
slope and sinuosity. Out of the 120 rock cross vanes observed, 109 had at least one
incidence of failure. Eighty-one rock cross vanes had at least one incidence of slight
failure, 50 had at least one occurrence of moderate failure, and 47 had at least one
occurrence of extreme failure. Other forms of uncertainty were addressed through
literature review. From this, a Failure Modes and Effects Analysis (FMEA) was
developed to compare the risk of the various modes of failure for the rock cross
vane.
Keywords: rock cross vane, paths of failure, rock cross vane rapid assessment tool,
probability of failure, Failure Modes and Effects Analysis
108
Introduction
North Carolina Stream Restoration
Bernhardt et al (2005) conducted a survey of all stream restoration projects

nationwide. Cost data was available for 58% of the projects. In North Carolina, there
were 8.25 projects per 1000 km of streams and rivers costing a total of $11,359,000
per 1000 mi. Of these, about 35% have undergone some level of monitoring. In
November 2006, Jeff Jurek of the N.C. Ecological Enhancement Program reported
the NCEEP as being responsible for 61% of constructed and planned stream
restoration projects out of a total of 1,449,985 linear ft of projects to date. According
to Jurek, stream project costs have averaged $205/ft in rural areas and $332/ft in
urban areas (Jurek, J. Email correspondence, November 7, 2006).
The Need for Structure Evaluation
With so many feet of stream restoration, there are bound to be many stream
structures. Many projects are designed with natural channel design and follow the
concepts of Dave Rosgen including incorporation of boulder structures. Rock cross
vanes are expensive. Doll and Jennings (2007) indicate that the installed cost is
between $3000 and $8000 for a rock cross vane depending on size. At Yates Mill, a
project within this study constructed in 2002, rock cross vanes were on the low end
at $2000 - $3000 per structure depending on size. On NC State University’s main
campus, Rocky Branch is undergoing restoration. Rock cross vanes constructed in
2006, cost $5900 each, likely the upper end of the cost range. This included nearly
95 tons of rock at $63/ton for boulders and $43/ton for class 1 rock (Doll and
Jennings 2007). Because of the high cost and frequency of implementation, it is
essential to have an assessment of how rock cross vanes are maintaining their
designed functions and durability in order to make appropriate design changes
where functions are lacking.
109
There has been a recent effort to evaluate structures on North Carolina’s streams. In
research unpublished to date, Jerry Miller from Western Carolina University has
evaluated nearly 40 sites and concludes that 70% of in-stream structures fail to meet
their given functions (Miller, J. Email correspondence, July 10, 2006.). Dave Penrose
from the NCSU Water Quality Group has also evaluated stream projects for the
Clean Water Management Trust Fund. This study conducts a qualitative assessment
on the structure’s condition, ability to perform its designed functions and habitat
provision. According to Penrose, structure failure was due to excess scour and poor
construction such as poor placement of rocks and logs resulting in piping.
Successful structures were well constructed, sealed to prevent piping and had small
elevation drops as opposed to large unstable ones (Penrose Personal
correspondence 2006).
Study Goals
The need for assessment goes further than evaluation of rock cross vanes’
effectiveness. Study goals are:
1. To determine frequency of rock cross vane failures in NC stream projects

through a snapshot of the state of the practice
2. To identify the causes of ineffectiveness and the paths of failure of the rock
cross vane
3. To use understanding of structure weaknesses to bring improvement to the

rock cross vane design.
This study also adopts the Failure Mode Effects and Analysis methodology
demonstrated by Johnson and Niezgoda (2004) in order:
4. To outline detection methods and the ease of determination,
5. To list to corrective measures necessary for the various failure modes, and
110
6. To determine a risk priority number for each failure mode for future
comparison to other structures.
This assessment is meant to be more that a simple mirror that reflects the state of
the practice. Rather it acts as a magnifying glass to bring clarity to the strengths and
weaknesses of the design and hopefully lead to better future decision making and
more efficient stream designs. It does so by finding failure occurrence, failure
severity, and failure paths. Because this study assesses one practice - the rock
cross vane - the FMEA is not to be used as a method of determining the best
practice for a given design problem, but rather to assess risk associated with the
functional failures within the given practice.
This project is one step towards more robust and predictable stream designs as they
relate to structure performance. A risk comparison of the rock cross vane and other
practices or combinations of practices which might be substitutive for the rock cross
vane would be a very beneficial follow-up study.
Defining Rock Cross Vane Performance
This study assessed the probability of durability and functional failure and calculated
failure risk from observations of structure performance. This and the following two
section entails a discussion of how uncertainty would need to be better explored to
be able to make these predictions through hydraulic models. Before addressing
uncertainty in vane performance, performance must be defined. Performance could
be one of three things:
1. The description of how much and in what manner water flows over the rock
cross vane.
2. The durability of the structure.
3. A measure of how well the structure performs its designed functions.
111
Uncertainty of Performance as Defined by Hydraulic Models
First, for a designer to apply the hydraulic relationships in Chapter 2 to a project for
estimation of the downstream velocity distribution, bank scour and pool scour, she
would first have to obtain the mean approach velocity for a given design storm such
as the 1.5 to 2 year storm used to predict bankfull area. Included in these
calculations are the bankfull area of the stream, the associated flow rate e as
recorded by the gage or as predicted by historical data and watershed information,
Manning’s n of the stream, bed slope and friction slope, etc. There is uncertainty for
each of these values and measurements. There is also uncertainty of future changes
to the watershed. Watersheds that are not fully developed may gradually have
increasing peak flows changing the standard return periods. There is uncertainty in
structure design as well. The size and density of the rock can vary. Exact placement
of the structures is difficult due to their large size and irregularities in the channel.
The geometric variables are assigned tolerance levels or acceptable ranges which
adds more uncertainty to predicting the downstream velocity distribution.
Sources of uncertainty in generating probability distribution for rock cross vane

performance through hydraulic models are: variety of design, disconnect between
design and construction, limited knowledge on all the system processes, and
immense variability of both streams and the designed structure and the interactions
between the two. Even with the flume study which provided relationships of vane
geometry to resulting flow distribution, the rock cross vane functions remain
somewhat of a black box, considering the expected increased variation in field
application. Effects that were found to be significant might be masked by the
presence of other system parameters not incorporated into the flume study.
Johnson (1996b) discusses the need for quantifying hydraulic parameter

uncertainties in order to obtain accurate probability and risk assessments. She
presents a table of coefficients of variation for Manning’s n, channel slope, particle
size, friction slope, sediment specific weight, and flow velocity. The values were
112
obtained from literature, experiments, and field observations (Johnson 1996b).
Uncertainty analysis for rock cross vanes in a given stream could be conducted
based on a combination of 1) the hydraulic parameter uncertainties, 2) watershed
assessment uncertainties from research like Johnson’s, and 3) installation
uncertainty analysis based on a comparison of design plans and as-built drawings. A
problem with the third component is that installation plans for structures are general
sketches used for all structures of the specific type as opposed to individual plans for
each structure.
According to Johnson (1996b), the sources of general model uncertainty are

ambiguity and cognitive uncertainty. Ambiguity comes from: 1) Physical
Randomness, 2) Statistical Uncertainty from Sampling, 3) Lack of Knowledge of the
System Processes, 4) Modeling Uncertainty from Oversimplification and
Assumptions of System Processes. Cognitive uncertainty comes from: 1) Definition
of System Parameters, 2) Other Human Factors, and 3) Definition of the
Interrelationships of Parameters. In another study Johnson approaches the
problems of the laboratory’s inaccuracies by using bias factors and fuzzy linear
regression (Johnson 1996a). This allows data to be applied to models in ranges
such as slight, moderate and extreme rather than discrete numbers. This enables
the model to take the focus off the infinite details that could go into the model design
while generalizing relationships and effects of parameters.
Uncertainty of Durability
Instead of defining performance as the behavior of the flow over the rock cross vane,
some researchers define performance of structures as durability – what forces the
structure can withstand. Here a simple model of durability emphasizes the
complexity of the calculations. A sill boulder is in the center of the rock cross vane
that serves to maintain channel grade upstream of the drop. The following
assumptions were made:
113
1. There are no gaps between the boulder and surrounding boulders/appropriate
fill material.
2. Suction force at the vertical face of the boulder on the downstream end is
negligible.
3. The base of the boulder is flush with the footer and appropriate fill material
and fabric matting.
4. There is a footer (some designers have not included these).
5. There is no side-cutting, bed degradation or undercutting of the sill.
6. Flow is perpendicular to the boulder.
7. The boulder is within the acceptable tolerance of the weight and size as
specified by the designer.
8. Tail water conditions do not have an effect on boulder movement. Once these
assumptions are in place, sketches give insight to the forces taking place on
the boulders.
Figure 4.1 is the assumed configuration of a boulder of the cross sill. In this diagram,
there is no vertical face against the flow as the top of the boulder is level with the
upstream bed.
114
Figure 4.1 Boulder of a Cross Sill in Assumed Conditions
Figure 4.2 is the free body diagram of Figure 4.1.
Figure 4.2 Free Body Diagram of Assumed Cross Sill Conditions
For a boulder only exposed on the downstream vertical face with no undercutting,
entrainment and tipping is very difficult. There is friction force on 3 sides of the
boulder if it is level with the bed: the bottom and the 2 vertical sides. The bottom
friction force is dependent on the boulder and water weight and roughness of the
boulder and the adjacent material. The side friction is dependent on the roughness
of the boulder and/or the adjacent material as well as the pressure applied by the
adjacent material. The top is exposed to the flow, and the right side is also exposed
to the flow or entrained air depending on tail-water depth. In order to move the
boulder, the resultant force of drag and lift drag must exceed the sum of the
horizontal friction forces and the weight of the boulder and vertical friction forces that
115
occur during lifting and hydrostatic force. Finally, the expected forces according to
flow event would be calculated to determine if the boulder of the given weight would
move. In reality, a scour pool is initially dug out during installation (Figure 4.3).
Figure 4.3 Cross Sill with Downstream Pool
This pool may change depth and size and may eventually migrate closer to the sill or
farther from the sill. The changes of flow patterns and swirling that may occur in the
pool are not accounted for by the assumed conditions in Figure 4.1. Assumption of a
flat bed downstream of the sill does not affect the major forces that would cause sill
movement. Also, pool shapes and sizes vary dependent on the conditions of the
vane. Assuming a general pool would potentially be as erroneous as assuming a flat
bed. Regardless, it is a simplification that adds a component of uncertainty.
Figure 4.4 represents a head cut formation or scouring of the bed during high flow
before re-deposition. The term head cut is defined by the NC Division of Water
Quality as “an abrupt vertical drop in the bed of a stream channel that is an active
erosion feature” (2005). The vertical face is open to drag or momentum force from
the flow. With a head cut, wall friction forces decrease since the likelihood of gaps
and lack of fill material. The friction between the sill boulder and the footer is
expected to decrease as the fill material is eroded. Therefore, it takes less drag and
lift force to move the cross sill boulder.
116
Figure 4.4 Cross Sill with Head cut
Figure 4.5 is a free body diagram of head cut conditions. As the boulder becomes
more exposed to the flow, entrainment requires less force. The force from flow
perpendicular to the upstream face of the exposed boulder is uncertain and likely
dependent on the length, depth of head cut, recirculation and flow patterns.
Figure 4.5 Free Body Diagram of Cross Sill with Head cut
Figure 4.6 shows the combination of a head cut and undercutting of the sill via the
overlap of its footer and scour at the base of the header. This situation has the
greatest potential for tipping. While the overhang is good for fish habitat, it is
structurally less stable than a rock cross vane without an overhang. The force to
move the boulder has decreased from the assumed condition Figure 4.1.
117
Figure 4.6 Cross Still Prime Conditions for Tipping
In Figure 4.7 the free body diagram of Figure 4.6, a moment is created about the
point of overhang. The greater the overhang, the less force needed to tip the boulder
into the downstream scour pool.
Figure 4.7 Free Body Diagram for Tipping Conditions
Predicting Functional Performance through Observation
Finally, performance of the rock cross vane could be defined as how well the rock
cross vane performs its four other functions: arm durability, sill durability, grade
control, bank protection at the vane, bank protection downstream of the vane, and
scour pool development. The purpose of quantifying each cross vane and stream
parameter’s uncertainty is to be able to construct a mathematical model for a given
function that can predict the range of performance. There is no existing
118
mathematical model to predict grade control for the rock cross vane. As shown, a
simplified model of durability is fairly easily determined mathematically using force
diagrams with assumed conditions; however, attempting to model functions such as
bank protection, grade control and pool development are much more complex as
they are dependent on the aforementioned parameters, the bank soil structure and
vegetative protection, assumptions of changes over time, installation of the rock
cross vane, and variable interactions. These require further research and model
development. An available option for determining function based performance and
durability is field observations of actual vanes. Assessing cross vanes at a variety of
stream restoration sites interrogates all uncertainties into the observed probability of
rock cross vane failure.
In this study, site information was collected to approximate range of hydraulic

variables. Therefore, data is presented with a range of ages, watershed sizes,
restored length, D50, mean bed slope, sinuosity and associated probabilities of rock
cross vane failure modes. The failure probabilities were used to calculate the risk of
failure using an FMEA.
FMEA and Risk Quantification
According to Stream Notes to Aid in Securing Favorable Conditions of Water Flows,

there is a difference between uncertainty and risk Uncertainty is “when the potential
outcome cannot be estimated on historical events.” Risk is “the calculated likelihood
of an unacceptable event occurring” (1998). Risk is also calculated as the cost of
failure times the probability of failure, which estimates risk as a dollar amount. The
higher the cost or probability of failure is, the higher the risk. Failure must be defined
before risk of failure can be calculated. A problem with using historical data of
structure performance is that there is not one consistent measure of failure risk. If
there were, the risk of structure performance could be assessed by an evaluation of
similar structures in similar watersheds and their success and failure rates. However,
it is difficult to calculate the risk of failure when there is uncertainty in the parameters
119
used to calculate the risk such as insufficient historical data on flow for a given
stream and on structure performance.
Johnson and Niezgoda (2004) demonstrate how to use a Failure Modes and Effects
Analysis to select the appropriate bridge scour counter measure for a design. There
are several steps to set up an FMEA:
1. List the possible countermeasures for the design problem
2. For each countermeasure, identify possible failure modes, possible effects on

other components, and possible effects on the whole system.
3. For each failure mode, identify the methods of detection and the
compensating provisions and assign a rating for ease of detection.
4. Predict and categorize the outcomes of failure for each failure mode (namely
loss of life and economic impact) and assign a consequence rating for the
outcomes of failure.
5. Determine and categorize the various likelihoods of occurrence and rate each
failure mode based on its observed category.
6. Multiply these three ratings – consequence, occurrence, and detection – to

calculate a risk priority number.
Johnson and Niezgoda emphasize the simplicity of their method and provide three
stream and bridge scour examples to demonstrate the FMEA process used to
determine the best scour countermeasure.
120
Materials and Methods
Rock Cross Vane Failures Assessed
A total of six functions were assessed for this study. Because there are eight
potential functions on Rosgen’s list alone (2001) the list was narrowed down to three
root functions based on their importance and potential to be rapidly observed and
assessed:
1. Grade control by fixing the bed elevation upstream and downstream of a drop
with a rock knick point
2. Bank protection by redirecting flow away from banks.
3. Scour pool development at the vane by a step pool and flow concentration
(increased velocity ratio).
Bank protection may be divided into bank protection at the structure (around the
arms and sill) and bank protection downstream of the structure (at the ends of the
arms). Structure durability may be divided into durability of the sill and durability of
the arms. This leads to a total of six possible failure indicators. Indicators are the
observable clues from which failure might be inferred. Table 4.1 lists the failures and
their main indicators.
121
Table 4.1 Failure and Major Indicators
Failure Indicator
Lack of Grade Control Head cut
Bank Erosion at Vane
Lack of Bank Protection
Downstream Bank Erosion
Lack of Pool/Pattern
Insufficient Scour Pool
Development
Arms washed out
Lack of Durability
Header washed out
For each failure, there are indicators of failure and mechanisms of failure. The
mechanism of failure is comprised of a primary cause of failure (what the water does
to cause the indicator) and a secondary cause (the design or construction flaws that
lead to the primary cause and indicator. A failure path is a mechanism of failure and
the failure indicator:
Secondary Cause Æ Primary Cause Æ Failure Indicator.
Descriptions of primary and secondary causes of failure are provided in Chapter 3.
There are cases where the rock cross vane may not be intended to provide all
functions and therefore cannot be classified as failed if lacking a given function.
However, designers’ intentions may not always be known by the assessor. The
function that is most likely to be nonessential to the design is scour pool
development. Structures are occasionally placed on bed material that is unlikely to
scour. It might be more beneficial for the designer to employ single arm vanes in
such cases, and the poor choice could be attributed to inexperience, lack of
knowledge about the bed conditions in the given location, or a simple affinity for the
rock cross vane. Nonetheless, the wrong structure is chosen for the location; and
failure, though attributed to the structure as failure, should be attributed to the
designer’s choice of structures.
122
Rock cross vane aesthetics deserve attention. While evaluation of aesthetics is
subjective, there are several components which potentially can be objectively
evaluated:
1. How well the rock blends in with the natural environment as related to
coloring, shape, and visibility
2. How visible active signs of failure are
3. Exposure of unnatural components such as fabric matting.
This study does not assess rock cross vanes for aesthetics; however, it does
present several examples of aesthetically pleasing vanes and prescribes how in the
future vanes might be designed to better blend in with their natural environment.
Sites
Nearly twenty stream restoration projects were observed, and 16 projects on 11

streams were used for data collection. These projects are included in Table 4.2.
123
Table 4.2 Projects and Site Data
Year Length DA D50 Avg
Project/Stream Class Sinuosity
Const ft. sqmi mm Slope
Avon Creek 2002 B5C 985 0.75 0.04 0.0046 1.03
Avon Tributary 2003 B5C 630 0.07 0.04 0.0243 1.07
Beaver Creek 2002 E5 4210 5.9 0.52 0.0055 1.3
Hyatt/Brasstown 2002 C4 2700 62 38.5 0.002 1.26
Oland/Brasstown 2001 C4 600 -- -- -- --
Greasy Creek/
2004 E4/C4 1472 1.5 3 0.015 1.24
Brasstown
South Muddy/ Hoppers
2004 E4/C4 572.1 9.1 1.1 0.0017 1.4
Creek
Campbell/Little
2001 C4 3900 9.5 0.06 0.002 1.18
Brasstown
Long Branch/
2001 C5b 475 0.07 0.09 0.028 1.03
Brasstown
Mason-Stalcup/Little
2002 E4 1900 3.8 13.7 0.008 1.12
Brasstown
Sheppard/Little
2003 C4 4000 5.6 9.65 0.004 1.48
Brasstown
Rocky Branch I 2002 B4C/C4 3300 0.7 5.7 0.01 1.15
Sharpe Creek 2001 C4/1 1040 3.46 6.5 0.01 1.3
Upper Laurel 2003 B4 1500 2.2 34 0.025 1.25
Yates Mill I 2002 E 3000 0.2 12.5 0.011 1.5
Muddy/Young’s Fork
1999 B4c/C5 1440 9.15 1.9 0.0022 1.8
Creek
Sites were located in watersheds with a variety of land uses, ranging from Rocky
Branch on a college campus in a fully developed watershed to Beaver Creek in a
mainly agricultural and forested watershed. These were predominantly piedmont
streams though some were located in mountainous regions. Figure 4.3 is a map of
North Carolina counties. Counties with sites used in this study are highlighted.
124
Figure 4.8 Projects by County
Seven of the projects, those in Clay and Cherokee Counties were all part of the
Hiawassee Watershed. These projects’ rock cross vanes were each unique in their
design, construction, and success in providing the expected functions.
Avon Creek
Avon Creek is an urban project located in a community park with high levels of
human access and virtually no vegetated buffer. There was a large amount of
sediment moving through the system with high sand content. This proved
problematic to the rock cross vanes because those that were not in the process of
being buried had noticeable piping problems due the contrast between the size of
the gaps between the boulders and the size of the bed material and bed load. There
was next to no pattern in this project and the rock cross vanes did very little to help
establish and maintain pools. In some locations boulders were oversized and
unneeded.
125
Avon Creek Tributary
Avon Creek Tributary was a small stream in which the rock cross vanes were used
to provide a step-pool system for a large drop. The pools were very shallow given
the height of the drop. The rock cross vanes seemed oversized as there were fines
in the pools.
Beaver Creek
Because the bankfull width was undersized during the planning and construction
phases, many of the rock cross vanes were totally bypassed due to large bank
failures. Though they were bypassed, most were still fully intact and experienced
very little movement. It was difficult to determine whether two of the structures were
rock cross vanes or j-hooks. One looked more like a j-hook with a protected toe on
the opposite bank and the other was in a location of severe lateral migration where
one arm might have been buried. They were both counted as rock cross vanes as
they were for all intents and purposes functioning as rock cross vanes.
Hyatt, Little Brasstown Creek

Hyatt had two remarkable rock cross vanes whose arms both tied in below bankfull
and had a very gentle slope. Water inundated a higher portion of the arms than
typically seen. The pools were well established with depositions of fines at the base
of the arms. The fines were likely due to a high sediment load and the exaggerated
width of the vane allowed some deposition at the toe of the arms and banks.
Aesthetically, these structures were the most pleasing to look at. Because the arms
were mostly inundated, the structure did not stand out to the eye. Also, the entire
structure was more parabolic in shape than the typical angular formation in other
streams. This gave the structure a more organic feel.
Oland, Little Brasstown Creek

Oland was a shorter stretch of Brasstown Creek with bank stabilization and one rock
cross vane. The pool was well established, and like Hyatt, had fines deposited at the
126
base of the arms. The arms themselves were steep and seemed to have minor
problems with piping and vegetative establishment behind them. The rock cross
vane was located on a straight-away of the creek, downstream of a bridge.
Campbell, Little Brasstown Creek

Campbell had more than the one rock cross vane assessed, but these were very
difficult to access. In general they all seemed to be having the same problems of
boulders being shifted and/or washed out. The channel was very high in clay content
and seemed to be experiencing flashy flows. The durability problems appeared to be
due to undersized widths of the structure and undersized boulders.
Sheppard (Rick’s Branch) Little Brasstown Creek

Eleven rock cross vanes were assessed, but several cross vanes could not be
located. There were large areas of bank failure around the rock cross vanes. This
was frequently due to the vanes’ placement in bends, improper alignment, and
constriction through the vanes. The site was visited two times, once in development
of the RCV-RAT and once for the qualitative assessment. Many of the structures
were unrecognizable due to increased vegetation and overall changes in the
channel. This leads to the suspicion that channel changes might be partially
responsible for some of the failures seen in conjunction with inadequacies of
structure design and construction.
Mason/Stalcup, Little Brasstown Creek

Mason/Stalcup was structurally very stable with very little evidence of failures.
Aesthetically, the project suffered slightly because the rock cross vanes’ quarried
stone and fabric matting were plainly visible and vegetation had yet to fully develop
on the banks.
127
Long Branch, Brasstown Creek
Long Branch rock cross vanes were difficult to locate as the sediment in the system
had buried most cross vanes. This lead to poor pool development and the
observation that the cross vanes were unnecessary to the system. Because
covering by sediment is not considered a functional failure in this study, lack of scour
pool development was the only recorded evidence of this problem.
South Muddy Creek/Hopper’s Creek

There was one located rock cross vane on this project. It had three cross sill stones.
The right hand boulder was much lower than the others causing flow to side cut the
thalweg and scour the right bank at the beginning of the cross vane. This cross vane
was at the head of the project and it appeared that a head cut might be migrating
upstream; however, a well established riffle just upstream of the project prevented
the head cut from traveling upstream. Also, repairs had been made to the cross
vane prior to the assessment. The extent and type of repairs is unknown.
Sharpe Creek
The success of Sharpe Creek’s rock cross vanes was largely attributed to the dense
vegetation protecting the banks and the gentle sloping of the arms. Typically the
placement near bends would have led to bank erosion, but vegetation armored the
banks. The success of the vegetation was potentially fostered by the gentles slopes
of the arms. The vanes were fairly stable and the pools were well developed. One
concern is the low canopy coverage. Assuming the grass on the banks is playing a
large role in the stability of the banks around the rock cross vanes, this stability
might be lost as the canopy grows up and chokes out some of the sunlight causing
the grasses to die back.
Rocky Branch
Rocky Branch’s quality of cross vanes was divided up into stream segments.
Upstream, cross vanes were stable and necessary due to the grade of the channel.
128
At the bottom segment of the stream, cross vanes experienced high levels of erosion
due to the vanes being installed too high and potentially being undersized for the
given flow. They also did little to establish pattern due to a very mild slope across the
segment. Some rock cross vanes on the lower portion of the stream had been
repaired in the last year due to high erosion of the banks at and downstream of the
structures. Some of the pools had also been filled with boulders to prevent over-
scouring. This raised concerns about how to incorporate interactive effects of human
interference on the success of the structures.
Upper Laurel
Upper Laurel was a difficult stream to assess for rock cross vanes because several
hurricanes had caused most of the structures to completely wash out and some to
be buried. Repairs had been made in places and some rock cross vanes with an
arm intact had been converted to j-hooks or single arm vanes. It was difficult to
determine which had been repaired and which were still washed out. All the
locations of the rock cross vanes were determined by a plan view drawing and either
a new structure or remnants were observed. In general, the rock cross vanes were
not sufficiently creating pools. The entire project seemed to consist of one large run.
This project brings up the concern of how to assess the effects of catastrophic
events on the on the success of the structure. This assessment attributes failure to
design and constructions problems, when there are clearly other interactions that
may be occurring. The hope of having a variety of project locations and ages is that
some of this uncertainty is captured within the data.
Yates Mill Phase II

The first portion of construction did not include fabric matting and piping and head
cutting was evident. The cross vanes were often placed too close to a bend leading
to bank erosion at the arms when backing into a bend and downstream erosion
when a bend followed too closely. Boulders appeared oversized for the system and
the cross vanes appeared to be oversized for the channel.
129
Muddy Creek/Young’s Fork, Marion County
The one cross vane present was placed in a bend causing flow to be directed right
of the designed thalweg, which created side cutting and piping. This was
exacerbated by poor boulder spacing and a very large drop. The vane was placed
on bedrock so the expected downstream pool in the vane was not able to develop.
Data Collection
Data collection methods are described in the Rock Cross Vane Rapid Assessment
Tool of Chapter 3. 120 rock cross vanes were assessed. Methods are summarized
below.
1. Only failure indicators were rated. These were what the success or failure of
the structure is based on. The primary and secondary causes were not rated.
They were identified as possible attributors to the given failure indicator.
2. The structure was assessed for the presence of failure indicators (arms
washed out, sill washed out, head cut, bank erosion at the structure,
downstream bank erosion, and lack of scour pool development).
3. Each of the six failure indicators was assigned a rating of 0, 1, 3, or 5

according to the instructions in the failure guidebook.
4. The primary causes of the failure indicator were identified based on clues
outlined in the failure guidebook.
5. The secondary causes of the primary causes were identified based on clues
outlined in the failure guidebook.
6. Photographs of the structure were taken.
7. At the end of an assessment, the structure had a rating of each failure

indicator with primary and secondary causes identified. These ratings were
130
not combined to give an overall score for the structure. Combining the ratings
might mask an extreme occurrence of failure of one function when all the
others were nonexistent.
8. The occurrence of failure indicators (whether or not they were present

regardless of severity) was used to assign an occurrence rating, which was
used in the calculation of the Risk Priority Number in the FMEA.
Data Analysis
Initially, a fault tree analysis was the intended method of determining the probability
of failure. Johnson’s definition of fault tree analysis is “a systematic method of
analyzing events that lead to an undesirable event” (1999). This method can be
qualitative or quantitative. In order to determine the probability of an undesired
event, the probabilities of the contributing circumstances must also be known. This is
problematic in that many of the circumstances are directly connected with vane
design and may go undetected until an actual failure occurs. Therefore, any
calculated failure probabilities would be correct, but the probabilities of the
circumstances would be less than might actually occur. Figure 4.9 is an example of
a fault tree for a sill washout in which probabilities would be assigned at each level
for the fault tree analysis.
131
Secondary Causes Primary Causes Failure Indicator
Header too High --------------|

Insufficient Backfill* ----------|
Poor Boulder Spacing -------| Æ Movement by
Undersized Boulders --------| drag and lift Æ --------------|
Undersized Cross Vane* ---| |
| Sill
| Washout
Arms too Steep ---------------| |
Backed into a Pool -----------| |
Drop too High ------------------| Æ Movement by Æ --------------|
Sill installed above Bed -----| Undercutting
Insufficient Backfill* ----------| and Tipping
No Footers ---------------------|
Undersized Cross Vane* ---|
Figure 4.9 Example Fault Tree for Sill Washout
Each item would be assigned a probability of occurrence for a fault tree analysis. Because the
items in the first two columns are often not noticed without the occurrence of the item in the
third column, the observed probabilities would not reflect the actual probabilities of
occurrence. *These are listed twice to prevent crossing over the diagram, but would have one
probability of occurrence.
Instead of using a failure tree analysis, observations were made to establish primary
and secondary causes of failure based on the occurrence of the failure indicator.
Therefore, the probability of a failure indicator occurrence, and the probabilities of
the various primary and secondary causes can be determined from the data.
Incorporation of Global Stream Design Effects on Failure
Johnson (1999) emphasizes the relation of bridge scour problems to overall stream
design problems (widening, lateral migration and degradation) along with local
problems. There were several streams where lateral stream movement was
obviously occurring. The rock cross vanes were still assessed. Some of the vanes
remained intact with one or two of the functions failing. A variety of restorations were
132
studied to represent the current range of stream restoration conditions. Causes of
failure were never attributed to sources outside the rock cross vane design and
construction because problems such as underestimated bankfull width are expected
to infiltrate all areas of design, included structures. Therefore, an assessor could
assume that failures of the rock cross vane are due to improper rock cross vane
design and not simply blame stream movement, though it may compound the
effects.
While the main intent of the study was to generate data on the occurrence and
severity of rock cross vane failures for better understanding of the strengths and
weaknesses of the design, SAS was also incorporated to locate potential effects on
cross vane performance. Data of failure occurrence and severity were regressed
over stream parameters in a general linear model to detect significant effects of site
parameters. Correlations between site parameters and site parameters, site
parameters and failures, failures and failures were observed.
FMEA Development for the Rock Cross Vane
The occurrences of failure indicators were used to cater the FMEA for North
Carolina stream projects. The FMEA was developed using Johnson and Niezgoda’s
(2004) steps outlined in the introduction and by implementing their rating criteria.
Data and Results
Failure, Cause and Path Occurrences
Out of the 120 rock cross vanes observed, 109 had at least one incidence of failure.
Eighty-two rock cross vanes had at least one incidence of slight failure, 51 had at
least one occurrence of moderate failure, and 47 had at least one occurrence of
extreme failure. Assessments of each structure are included in Appendix B.
133
Table 4.3 gives the number of occurrences for each level in the failure path. This
Table 4.is presented to show that of the 109 rock cross vanes with some level of
failure, there were 252 failure indicators, 386 occurrences of a failure indicator and a
primary cause, and 520 possible paths of failure (one indicator, one primary cause,
and one secondary cause). In reality, a mix of failure paths as defined by this study
may cause a given failure. However, this study preserves them as separate entities
for simplification.
Table 4.3 Occurrences of failure and failure paths

Path Level Occurrences
Failure Indicator 255
Indicator + Primary 387
Indicator + Primary + Secondary 522
Table 4.4 shows the median rating of failure indicators by project.
134
Table 4.4 Median rating of failure indicators by project
F1 is arm washout. F2 is sill washout. F3 is head cut migration also known as upstream bed
degradation. F4 is bank erosion at the rock cross vane. F5 is bank erosion downstream of the
rock cross vane. F6 is insufficient scour pool development.
Stream F1 F2 F3 F4 F5 F6
Avon Creek 0 0 1 0 0 5
Avon Creek Tributary 0 0 0 0 0 1
Beaver Creek 0 0 0 3 0.5 1
Greasy Creek 0 0 1 1 0 0
Hoppers Creek 0 0 3 3 0 0
Hyatt 0.5 0 0 0 0.5 0.5
Campbell 0 0 0 3 1 0
Long Branch 0 0 0 0 0 0.5
Mason Stalcup 0 0 0 1 0 0
Rocky Branch 0 0 0.5 1 0.5 0
Sheppard 0 0 5 1 1 0
Sharpe Creek 0 0 0 0.5 0 0
Upper Laurel 5 5 0 0.5 0 4
Young’s Fork 0 0 0 1 0 3
Yates Mill 0 0 1 1 0 1
According to this table, failure was not evenly spread across all projects. This
question is addressed further by correlations of vane performance over site
parameters. The median value was chosen as opposed to mean because the
number of structures on each project greatly varied with some projects having one
structure.
Table 4.5 includes the occurrence of the failure indicators that is used as probability
of occurrence in the FMEA.
135
Table 4.5 Frequency of Failure Indicators
Column 1 lists the failure indicators. Occurrences are the number of rock cross vanes with at
least one incidence of a failure indicator.
Occurrences by % Occurrence by
Rating Rating total % of Path
Failure indicator
Vanes Occurrences
1 3 5 1 3 5
Arm Washout 13 4 3 10.8 3.3 2.5 16.7 31

Sill Washout 0 0 4 0.0 0.0 3.3 3.3 4
Head cut 31 16 15 25.8 13.3 12.5 51.7 161
Bank erosion at
37 24 17 30.8 20.0 14.2 65 192
structure
Downstream
16 11 12 13.3 9.2 10.0 32.5 65
bank erosion
Lack of scour
22 16 14 18.3 13.3 11.7 43.3 69
pool development
Bank erosion at the vane (F4) had the highest frequency of occurrence with 65% of
all the vanes experiencing slight, moderate or extreme bank erosion. Side cutting
from improper alignment was the number one mechanism of failure (primary cause
and secondary cause together). This could be attributed to rapid lateral migration of
the streams, but this was not determined by the study. Tied for the number two path
to bank erosion was undercutting of the arms due to insufficient backfill. The number
three path to bank erosion at the cross vane was direct contact of flow with the
banks due to loose boulder spacing.
Bank erosion was followed by 52% of vanes experiencing some level of a head cut
(F3). The number one path to head cutting was the sill being installed too high, not
at bed elevation, leading to side cutting. Other leading paths to head cutting were 1)
undercutting caused by insufficient backfill and being backed into a pool and 2) side
cutting due to loose boulder spacing.
Forty-three percent of cross vanes experienced insufficient scour pool development

(F6). The main failure paths for this were 1) an insufficient jet caused by either a too-
short drop of an oversized rock cross vane and 2) the pool region being protected by
boulders placed in the pool. Boulders are typically placed in the pool region in cases
136
where the pool has scoured out too large, and boulders have been added to prevent
continued scour which might lead to problems with the durability of the rock cross
vane.
Thirty-two and a half percent of the rock cross vanes experienced some downstream
bank erosion (F5). This was primarily caused by 1) expansion out the vane due to
either the vane being undersized or the pool being crowded with boulders and 2)
direct contact of the flow coming out of the vane because the banks were either
missing vegetative protection or armor protection.
Durability was much less a problem than the other functions. Twenty of the vanes
experienced boulders from the arms washing out (F1), which was mainly due to
undersized boulders, improper alignment or being placed in a. Four of the rock cross
vanes experienced sill washouts (F2), which was attributed to small boulders and
improper alignment.
The top five paths of failure and their percentage of occurrence were:
1) Improper Alignment Æ Side cutting Æ Bank Erosion at the Structure (23%)
2) Sill Installed too High Æ Side cutting Æ Head Cut (22%)
3) Insufficient Backfill or matting Æ Undercutting Æ Bank Erosion at the

Structure (16%)
Insufficient Backfill or matting Æ Undercutting Æ Head Cut (16%)
Placed in a Bend Æ Side Cutting Æ Bank Erosion at the Structure (16%)
In Table 4.5, the path occurrences column shows how many potential paths of
failure were identified for the vanes with the given failure indicator. The column
"Vanes" indicates how many rock cross vanes experienced some level of the failure
indicator. “Path occurrences” is a mark of how many potential paths caused the
given indicator. For example, for the 78 occurrences of bank erosion at the structure,
there were 192 paths they may have let to them. This means there were 2.5 failure
137
paths for each occurrence of bank erosion. In actuality, each occurrence of failure
has one overarching cause of failure comprised of one to several failure paths.
Table 4.6 gives the percentage of rock cross vanes having a failure indicator that
was caused by the given primary causes. The path occurrences column shows how
many potential paths of failure were identified for the vanes with the given primary
cause of failure. Table 4.18 reveals that the top contributors to failure were side
cutting and undercutting. Side cutting can be a problem with installation, but it can
also be contributed to stream-wide design problems such as an under-estimation of
radii of curvature, bankfull flow, and shear stress on banks. Undercutting is primarily
a problem with installation: placement of boulders, backfill and fabric matting, and
sizing of the entire structure. Undercutting can occur from the upstream or
downstream direction.
Table 4.6 Frequency of Primary Causes of Failure

Column 1 lists the primary causes of failure. Column 2 gives the number of rock cross vanes
with at least one occurrence of the primary cause leading to a failure indicator. Column 3 is
the percentage of total rock cross vanes with at least one occurrence of the primary cause of
failure. Column 4 is the number of failure paths that include the primary cause of failure.
Path
Primary Vanes % of Vanes Occurrences
P01. Contact 38 31.7 49
P02. Directed 27 22.5 34
P03. Drag + Lift 12 10.0 24
P04. Expansion 23 19.2 30
P05. Piping through RCV 23 19.2 28
P06. Protected Pool Region 27 22.5 29
P07. Side Cutting Thalweg 75 62.5 151
P08. Undercutting 61 50.8 136
P09. Weak Jet 32 26.7 41
Table 4.7 gives the percentage of rock cross vanes having a failure indicator that
was caused by the given secondary causes. The path occurrences column shows
how many potential paths of failure were identified for the vanes with the given
secondary cause of failure.
138
Table 4.7 Frequency of Secondary Causes of Failure
Column 1 lists the secondary causes of failure. Column 2 is the number of rock cross vanes
with at least one occurrence of the secondary cause of failure leading to a failure indicator.
Column 3 is the percentage of total rock cross vanes with at least one occurrence of the
secondary cause of failure. Column 4 is the number of failure paths that include the secondary
cause of failure.
Secondary Vanes % of Vanes Path Occurrences

S01. Alignment 48 40.0 75
S02. Backed into Pool 18 15.0 18
S03. Placed in Bend 34 28.3 56
S04. Placed on Bedrock 18 15.0 6
S05. Arms not Tied in 2 1.7 3
S06. Arms Short 5 4.2 5
S07. Arms Flat 0 0.0 0
S08. Steep Arms 10 8.3 10
S09. Arms Washed Out 14 11.7 19
S10. Sill High 34 28.3 50
S11. Sill Washed Out 6 5.0 8
S12. Boulder Spacing 41 34.2 54
S13. Boulders Placed in Pool 12 10.0 26
S14. Small Boulders 8 6.7 11
S15. Drop Short 16 13.3 16
S16. High Drop 9 7.5 12
S17. Exposed Banks 17 14.2 19
S18. Backfill/Fabric Matting 40 33.3 58
S19. No Footers 9 7.5 11
S20. Oversized Vane 17 14.2 17
S21. Undersized Vane 31 25.8 48
The highest secondary contributors were improper alignment, placed in a bend, high
sill, backfill/fabric matting, boulder spacing, and undersized vane. The most
surprising result was the lack of occurrence of flat arms (arms with a low arm slope)
as a cause of failure. This was expected to lead to downstream bank erosion and
139
insufficient scour pool development due to inadequate flow contraction. However,
there were no cases of these two failure indicators that could be contributed to the
arms being too flat. On the contrary, it was observed that flat arms typically had
more stable banks. Flat arms prevent under-sizing the cross sectional area of flow
over the rock cross vane and allow the banks to remain soft and protected by
vegetation. Steep arms cover a greater percentage of the banks along their length
with boulder protection which hardens the banks.
Table 4.8 is gives the Occurrence of failure paths (the combination of one failure
indicator, one primary cause and one secondary cause).
140
Table 4.8 Occurrences of Failure Paths
Values on this Table 4.indicates the number of occurrences of the groupings of the primary and secondary causes of the failure
indicator. Values of 0 indicate where occurrences were predicted but not observed.
INDICATOR Æ F4 F5 F4 F5 F2 F1 F5 F3 F4 F6 F3 F4 F5 F2 F1 F3 F4 F5 F6
P04. EXPAN./ RE-
UNDERCUTTING
P02. DIRECTED
P09. WEAK JET

P01. CONTACT
SIDECUTTING
PROTECTED
P03. DRAG +
P05. PIPING
ENTRY
LIFT
P06.
P07.
P08.
PRIMARY
CAUSES
SECONDARY CAUSES:
S01. ALIGNMENT 14 9 5 15 27 0 1 0 4
S02. BACKED INTO POOL 0 18
S03. PLACED IN BEND 6 5 5 14 19 0 3 4
S04. BEDROCK 6
S05. ARMS NOT TIED IN 1 1 1 0 0 0
S06. ARMS SHORT 5 0 0 0
S07. ARMS FLAT 0
S08. STEEP ARMS 0 4 0 0 6
S09. ARMS WASHED OUT 6 0 9 1 0 1 1 1
S10. SILL HIGH 0 26 9 0 15
S11. SILL WASHED/MISSING 6 2
S12. BOULDER SPACING 17 0 0 8 3 18 8 0
S13. BOULDERS PLACED 12 14
S14. SMALL BOULDERS 3 8
S15. DROP SHORT 0 16
S16. HIGH DROP 0 2 6 4
S17. EXPOSED BANKS 6 13
S18. BACKFILL 0 2 11 6 0 1 19 19
S19. NO FOOTERS 0 1 2 8
S20. OVERSIZED VANE 0 17
S21. UNDERSIZED VANE 14 3 10 0 3 3 15
141
Data Analysis
The following data analysis explores whether there were other measurable
influences on the structures’ performance other than the primary and secondary
causes such as parameters of the projects and watersheds and failures themselves.
Three more types of information were gathered from the data: 1) correlations
between site parameters and, 2) effects of site parameters on failure indicators, 3)
and the effects of failure indicators and class on failure indicators. Project was
assigned as a class variable. Each project had a given set of variables (age, length,
DA, D50, slope, sinuosity) that are consistent across all structures within the class.
However, within each class, there were a different number of structures to assess.
Three different SAS designs were used to determine the above relationships.
1. A correlation matrix was used to find correlation between project variables.

This data set included the 15 projects with historical site data (Oland was
excluded) and their six project variables.
2. A general linear model was used to estimate effects of project variables on

failure indicators. F1-F6 represent the six different failure indicators as listed
in Table 4.5
F1 = β 0 + β1Age + β 2 Length + β 3DA + β 4 D50 + β 5Slope + β 6Sinuosity (1)
142
3. A general linear model was used to estimate effects of failure indicators (F1-
F6), class effects of project on failure indicators, and a partial correlation
matrix of the dependent variables.
F1 F2 F3 F4 F5 F6 = project (7)
Correlation of Project Variable
The Oland site data was dismissed for the effects of project variables on failure
because data of these during construction is missing. Table 4.9 gives the
correlations between project variables.
Table 4.9 Correlation of Project Variable

Pearson Correlation Coefficients, N = 15
Prob > |r| under H0: Rho=0
Age Length DA D50 Slope Sinuosity
Age 1.00000 0.06694 0.06874 -0.14205 -0.18595 0.29813

0.8126 0.8077 0.6136 0.5070 0.2805
Length 0.06694 1.00000 0.18700 0.12737 -0.46370 0.20637

0.8126 0.5045 0.6510 0.0817 0.4606
DA 0.06874 0.18700 1.00000 0.63157 -0.39881 0.10414

0.8077 0.5045 0.0116 0.1409 0.7119
D50 -0.14205 0.12737 0.63157 1.00000 0.05967 0.05140

0.6136 0.6510 0.0116 0.8327 0.8557
Slope -0.18595 -0.46370 -0.39881 0.05967 1.00000 -0.44804

0.5070 0.0817 0.1409 0.8327 0.0940
Sin 0.29813 0.20637 0.10414 0.05140 -0.44804 1.00000

0.2805 0.4606 0.7119 0.8557 0.0940
There was one significant correlation (alpha = 0.05): drainage area and D50 (the
median bed sediment size). The larger the watershed was, the larger the D50 tended
143
to be. While there were no other strong correlations, relationships appeared as
expected simply in how the parameters were correlated (positive/negative). Overall,
the weak correlations show the high level of variability within watersheds and stream
systems. The first parameter, age, had weak correlations to the other parameters,
and there was no logical reason it would be strongly correlated to the other site
parameters. There is a five year span in age and it seems unreasonable to be able
to glean much about the effects of age from the limited age range. The strongest
correlation of age was with sinuosity (p=0.28). One instance in which age might have
a strong correlation to the others is it the practice was shifting in trend towards one
end of the site parameters over time, such as most streams being designed with
increasing sinuosity. However, there is not sufficient data to support this. It could
also be expected that the older projects would have undergone a larger variety of
storms and incurred more damage than new projects. It could also be expected that
structure design and installation improves over time as experience levels increase
and more is known about the practice. Because length of the project is not related to
the total length of the stream, there is no reason why there should be strong
correlations to the other parameters. Project length depends more on the availability
of funds and accessibility to adjacent properties. If the valley lengths of all projects
were equal, then the stronger positive correlation between length and slope and
length and sinuosity would be expected. There was also a strong but not significant
correlation (p=0.09) between slope and sinuosity. A very sinuous stream is likely to
have lower slope.
Project Effects on Failure Indicator
Second, the effects of the site parameters on failure rating were determined. Instead
of using class which lumps all the site parameters into a class effect, class was
ignored.
Tables 4.10 – 4.15 give the effects of project variables on the failure indicator
ratings. Significant effects at a 95% confidence level are in bold.
144
Table 4.10 Effects of Project Variables on F1: Arm Washout
Dependent Variable: F1
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 6 24.1506074 4.0251012 5.22 <.0001
Error 112 86.4040144 0.7714644
Corrected Total 118 110.5546218
R-Square Coeff Var Root MSE F1 Mean

0.218450 261.3033 0.878330 0.336134
Source DF Type III SS Mean Square F Value Pr > F

Age 1 0.06270422 0.06270422 0.08 0.7761
Length 1 3.99905003 3.99905003 5.18 0.0247
DA 1 0.07303410 0.07303410 0.09 0.7589
D50 1 9.37007588 9.37007588 12.15 0.0007
Slope 1 5.14758935 5.14758935 6.67 0.0111
Sinuosity 1 0.69033976 0.69033976 0.89 0.3462
Table 4.10 shows that D50 (p=.0007) and slope (p=0.0111) had significant effects on
the rating of F1 (α>.05). The higher each of these parameters was, the higher the
failure rating. This is a logical relationship because arm washouts are a typically a
function of boulders being undersized for the system. Higher slopes also lead to
higher velocities. A high D50 indicates that the system carries a larger bedload.
Length also had a significant effect on F1(arm washout), however, length (p=0.0247)
is somewhat arbitrary depending on funds and land access.
Table 4.11 gives effects of project variables on sill washouts.
145
Table 4.11 Effects of Project Variables on F2: Sill Washout
Sum of
Model 6 32.28319651 5.38053275 9.36 <.0001
Error 112 64.35545895 0.57460231

0.334061 451.0250 0.758025 0.168067

Age 1 1.36979738 1.36979738 2.38 0.1254
Length 1 0.28489625 0.28489625 0.50 0.4828
DA 1 2.38336104 2.38336104 4.15 0.0440
D50 1 17.44619360 17.44619360 30.36 <.0001
Slope 1 1.16805259 1.16805259 2.03 0.1567
Sinuosity 1 0.44498523 0.44498523 0.77 0.3807
Table 4.11 shows that D50 (p=0.0440) and drainage area (p <0.0001) had significant
effects on F2 (sill durability). A higher drainage area typically means higher flows,
so it is unknown why there would be a negative effect of drainage area. The D50
relationship is likely the same as that with arm durability. It is unknown why slope
was not significant for sill durability as well.
Table 4.12 gives effects of project variables on head cut formation.
146
Table 4.12 Effects of Project Variables on F3: Head Cut
Sum of
Model 6 86.5220732 14.4203455 6.11 <.0001
Error 112 264.1838091 2.3587840

0.246708 118.6780 1.535833 1.294118

Age 1 19.24631075 19.24631075 8.16 0.0051
Length 1 2.61116256 2.61116256 1.11 0.2950
DA 1 3.74046385 3.74046385 1.59 0.2106
D50 1 0.38983520 0.38983520 0.17 0.6851
Slope 1 0.04138577 0.04138577 0.02 0.8949
Sinuosity 1 20.97169117 20.97169117 8.89 0.0035
F3 (head cutting) was unique in its relationship to age and length with it worsening
over time and being worse on projects with higher sinuosity. The age correlation
could be due to the head cut developing over time. Yet, as mentioned previously,
age was a weak variable due to the limited range. Higher sinuosity might have been
related to increased improper alignment and placement in bends which were a major
cause of head cutting.
Table 4.13 gives effects of project variables on bank erosion at the structure.
147
Table 4.13 Effects of Project Variables on F4: Bank Erosion at the Structure
Sum of
Model 6 47.5831810 7.9305302 2.82 0.0135
Error 112 314.8537938 2.8111946

0.131287 104.4622 1.676662 1.605042

Age 1 0.32499088 0.32499088 0.12 0.7345
Length 1 20.28018595 20.28018595 7.21 0.0083
DA 1 0.43179211 0.43179211 0.15 0.6959
D50 1 8.09677192 8.09677192 2.88 0.0925
Slope 1 0.51601103 0.51601103 0.18 0.6692
Sinuosity 1 0.47544873 0.47544873 0.17 0.6817
The single significant effect on F4 (bank erosion at the structure) was length
(p=0.0083). As mentioned earlier, length was somewhat arbitrary and there is no
logical reason why an increased length of project would cause increased bank
erosion.
Table 4.14 gives effects of project variables on bank erosion downstream of the rock
cross vane.
148
Table 4.14 Effects of Project Variables on F5: Downstream Bank Erosion
Sum of
Model 6 59.0656407 9.8442735 4.31 0.0006
Error 112 256.0940232 2.2865538

0.187415 165.0864 1.512136 0.915966

Age 1 1.17674990 1.17674990 0.51 0.4746
Length 1 29.49202934 29.49202934 12.90 0.0005
DA 1 0.18359788 0.18359788 0.08 0.7774
D50 1 1.53550163 1.53550163 0.67 0.4143
Slope 1 0.00004914 0.00004914 0.00 0.9963
Sinuosity 1 0.36536691 0.36536691 0.16 0.6901
The single significant effect on F5 (bank erosion at the structure) was length
(p=0.0005). As mentioned earlier, length was somewhat arbitrary and there is no
logical reason why an increased length of project would cause increased bank
erosion.
Table 4.15 gives the effects of project variables on poor scour pool development.
149
Table 4.15 Effects of Project Variables on F6: Insufficient Scour Pool Development
Sum of
Model 6 15.5478920 2.5913153 0.86 0.5235
Error 112 335.7462257 2.9977342

0.044259 147.1687 1.731397 1.176471

Age 1 0.75272377 0.75272377 0.25 0.6173
Length 1 3.86232659 3.86232659 1.29 0.2588
DA 1 1.97488010 1.97488010 0.66 0.4187
D50 1 1.65225049 1.65225049 0.55 0.4594
Slope 1 0.17648072 0.17648072 0.06 0.8087
Sinuosity 1 1.30789772 1.30789772 0.44 0.5103
There were no significant effects of the project variables on the rating of lack of
scour pool.
Class Effects on Failure Indicators
Finally, a general linear model using all the data was run in SAS to determine the
effects of project and the significant correlations between the failure indicators.
Table 4.16 gives the class effects of project on arm washout.
150
Table 4.16 Class Effects of Project on F1: Arm Washout
Sum of
Model 15 55.2524138 3.6834943 6.91 <.0001
Error 104 55.4142529 0.5328294

0.499269 218.9855 0.729952 0.333333
Source DF Type I SS Mean Square F Value Pr > F

project 15 55.25241381 3.68349425 6.91 <.0001

project 15 55.25241381 3.68349425 6.91 <.0001
According to Table 4.16, the effect of project on arm washout was significant
(p<0.0001). The rating of arm washout showed trends across projects, rather than
being consistently random across all projects. This could imply that the design of the
rock cross vanes was designed similarly within a project, meaning potential under
sizing of boulders was similar across the structures in the project. It could also imply
that structures within a project incurred similar levels of damage due to the
experiences storm events.
Table 4.17 gives the class effects of project on sill washout.
151
Table 4.17 Class Effects of Project on F2: Sill Washout
Sum of
Model 15 55.18939394 3.67929293 9.23 <.0001
Error 104 41.47727273 0.39881993

0.570925 378.9131 0.631522 0.166667

project 15 55.18939394 3.67929293 9.23 <.0001

project 15 55.18939394 3.67929293 9.23 <.0001
According to Table 4.17, the effect of project on sill washout was significant
(p<0.0001). The rating of sill washout showed trends across projects, rather than
being consistently random across all projects. This could imply that the design of the
rock cross vanes was designed similarly within a project, meaning potential under
sizing of boulders was similar across the structures in the project. It could also imply
that structures within a project incurred similar levels of damage due to the
experiences storm events.
Table 4.18 gives the class effects of project on head cut formation
152
Table 4.18 Class Effects of Project on F3: Head Cut
Sum of
Model 15 118.1248620 7.8749908 3.50 <.0001
Error 104 234.2418047 2.2523250

0.335233 116.9435 1.500775 1.283333

project 15 118.1248620 7.8749908 3.50 <.0001

project 15 118.1248620 7.8749908 3.50 <.0001
According to Table 4.18, effect of project on head cutting was significant (p<0.0001).
The rating of head cut showed trends across projects, rather than being consistently
random across all projects. This could imply that the design of the rock cross vanes
was designed similarly within a project, meaning potential improper backfill/fabric
matting, boulder placement, or alignment was similar across the structures in the
project. It could also imply that structures within a project incurred similar levels of
damage due to the experiences storm events.
Table 4.19 gives the class effects of project on bank erosion at the rock cross vane.
153
Table 4.19 Class Effects of Project on F4: Bank Erosion at the Structure
Sum of
Model 15 79.6302877 5.3086858 1.95 0.0262
Error 104 283.1697123 2.7227857

0.219488 103.1304 1.650087 1.600000

project 15 79.63028769 5.30868585 1.95 0.0262

Project 15 79.63028769 5.30868585 1.95 0.0262
According Table 4.19,, the effect of project on bank erosion at the structure was
significant (p=0.0262). The rating of bank erosion at the structure showed trends
across projects, rather than being consistently random across all projects. This could
imply that the design of the rock cross vanes was designed similarly within a project,
meaning potential lack of backfill/fabric matting, placement in a bend, or alignment
was similar across the structures in the project. It could also imply that structures
within a project incurred similar levels of damage due to the experiences storm
events, under-estimation of bankfull events, or a general lack of bank protection
through vegetation.
Table 4.20 gives the class effects of project on downstream bank erosion.
154
Table 4.20 Class Effects of Project on F5: Bank Erosion Downstream of Structure
Sum of
Model 15 76.6866611 5.1124441 2.22 0.0098
Error 104 239.3050055 2.3010097

0.242686 166.9990 1.516908 0.908333

project 15 76.68666115 5.11244408 2.22 0.0098

Project 15 76.68666115 5.11244408 2.22 0.0098
According to Table 4.20, the effect of project on downstream bank erosion was
significant (p=0.0098). The rating of bank erosion downstream of the structure
showed differences across projects, rather than being consistently random across all
projects. This could imply that the design of the rock cross vanes was designed
similarly within a project, meaning potential improper placement or alignment was
similar across the structures in the project or all cross vanes were too narrow. It
could also imply that structures within a project incurred similar levels of damage
due to the experiences storm events, under-estimation of bankfull events, or a
general lack of bank protection through vegetation.
Table 4.21 gives the class effects of project on insufficient scour pool development.
155
Table 4.21 Class Effects of Project on F6: Lack of Scour Pool Development
Sum of
Model 15 124.9178874 8.3278592 3.80 <.0001
Error 104 227.7487793 2.1898921

0.354210 126.8424 1.479828 1.166667

Project 15 124.9178874 8.3278592 3.80 <.0001

project 15 124.9178874 8.3278592 3.80 <.0001
According to Table 4.21, the effect of project on a lack of scour pool development
was significant (p < 0.0001). The rating of a lack of scour pool development showed
differences across projects, rather than being consistently random across all
projects. This could imply that the design of the rock cross vanes was designed
similarly within a project, meaning potential over sizing of the width of structures or
inadequate drop was similar for all structures in a project. It could also imply that
structures within a project incurred similar levels of damage due to the experiences
storm events or lack of digging out a large enough pool during construction.
The partial correlation matrix of the failure indicators (Table 4.22) shows that
indicators may have influenced each other.
156
Table 4.22 Partial Correlation Matrix for Failure Indicators
Partial Correlation Coefficients from the Error SSCP Matrix / Prob > |r|
DF = 104
F1 F2 F3 F4 F5 F6
F1 1.000000 0.381617 0.176517 0.334877 -0.007702 0.053169

<.0001 0.0717 0.0005 0.9378 0.5901
F2 0.381617 1.000000 0.127968 -0.006291 0.127748 0.093535

<.0001 0.1933 0.9492 0.1941 0.3426
F3 0.176517 0.127968 1.000000 0.448091 -0.204628 0.045982

0.0717 0.1933 <.0001 0.0363 0.6414
F4 0.334877 -0.006291 0.448091 1.000000 -0.034597 0.037071

0.0005 0.9492 <.0001 0.7261 0.7073
F5 -0.007702 0.127748 -0.204628 -0.034597 1.000000 0.017645

0.9378 0.1941 0.0363 0.7261 0.8582
F6 0.053169 0.093535 0.045982 0.037071 0.017645 1.000000

0.5901 0.3426 0.6414 0.7073 0.8582
According to the partial correlation matrix, F1 (arm washout) and F4 (bank erosion at
the structure) were significantly positively correlated (p=0.0005). Because bank
protection at the arms requires the arms to be intact, it is logical that these two are
correlated. F1 and F2 (sill washout) were also significantly positively correlated (p <
0.0001). This is likely due to the choice in boulder size that is typically the same for
boulders of the sill and arms.
F3 (head cut) and F5 (bank erosion downstream of the structure) are significantly
negatively correlated (p=0.0363). This can be explained by the effects of flow
constriction through the structure. If the rock cross vane is undersized, it is likely that
the downstream banks will be eroded due to flow expansion. If the sill washes out
and a head cut forms, the constriction is lessened and the banks will incur less
damage. Finally, F3 and F4 were significantly positively correlated (p < 0.0001).
157
Because both were strongly connected to side cutting, this relationship is logical.
Side cutting can also explain the strong positive correlation between F1 and F3.
Because the several different failure modes were significantly partially correlated,
the true failure paths become infinitely more complex – more so than just a
secondary and primary cause and an indicator. The true failure path could not easily
be determined by a one time use of the Rapid Assessment Tool. Rather,
determination of the true failure path requires photo documentation and knowledge
of the streams parameters, flow events, and history of disturbance post-restoration.
Correlation did not imply that failure indicators are one in the same because it is
expected that there is a cause an effect that may happen in one direction and not
the other as in the positive correlation between head cut and downstream bank
erosion. The head cut lead to downstream bank erosion, not vice versa.
FMEA Development
Johnson and Niezgoda’s paper (2004) served as a roadmap for the development of
the Failure Modes and Effects Analysis of the Rock Cross Vane via implementation
of the data from the qualitative analysis of NC rock cross vanes.
Possible Countermeasures
Typically a problem is defined and various solutions are considered. For this study,
the FMEA for just one structure is being developed, and other structures are not
compared to the rock cross vane.
Possible Failure Modes
Including every possible mode of failure in the FMEA would be impractical. There
are a total of 91 expected paths of failure and 64 unique paths observed. A path is
comprised of a failure indicator, a primary cause of failure and a secondary cause of
failure. In reality, multiple combinations of causes lead to an even greater
158
occurrence of distinct paths. There were a total of 252 failure indicators observed in
120 rock cross vanes, each with a unique or recurring path of failure. The
occurrence of primary and secondary causes is a non-exhaustive assessment
because the only time failure causes were evident was in the presence of a failure
indicator. Due to the large number of potential failure paths, for this study, “mode of
failure” is defined as type of failure as marked by failure indicator, shown in Table
4.1.
This is a partial FMEA of a rock cross vane; partial because a true FMEA contains
multiple countermeasures. Alone, it cannot help a designer determine which
structure to use. The designer would need to compare the rock cross vane to other
potential structures to determine the best application. Instead, the FMEA of the rock
cross vane can help pinpoint the main sources of risk in design so that the
technology can be improved by keeping the strengths and diminishing the
weaknesses. It can also help prioritize which aspects of the design should be made
more robust.
Table 4.23 lists the potential consequences on other components and on the whole
system. Overall, though many possibilities are listed, little system-wide effects are
expected. There are several possibilities for compensating provisions. This decision
depends on the choice of the designers and property owners in cases where bank
erosion is damaging private property. Choice of compensating provisions depends
on the severity of the failure and failure progression as determined by photo
documentation.
159
Table 4.23 Failure Modes
Failure Indicators are used as failure modes as they signify the functional failure of the rock cross vane: arm durability, sill
durability, grade control, bank protection at the cross vane, bank protection downstream of the cross vane, and scour pool
development.
Failure Effects on Other Components of Effects on the Whole Methods of Compensating Provisions
Mode RCV System Detection
Arm Washout - May cause bank erosion as banks - Possible - Visual Inspection - Re-install missing boulders.
are left bare where boulders once sedimentation - Photo - Repair banks from incurred
were attached. downstream. Documentation scour or erosion with plantings
- May lead to a Head cut formation - May diminish bed a matting.
as water bypasses the sill. and pattern
- May lead to downstream bank structure.
erosion as contraction is reduced. - May reduce
- May hinder scour pool potential for habitat.
development as contraction is - Decrease in stream
reduced and boulders fall into aesthetics.
pool
Sill Washout - May lead to a head cut formation - Head cut may - Visual Inspection - If grade control is a concern, re-
as water bypasses the sill. travel upstream - Photo install cross sill boulders using
- May hinder scour pool and take out other Documentation appropriate size and installation
development as boulders fall into structures. techniques.
pool - Possible - If drop is small and grade
sedimentation control is not a concern, adjust
downstream. cross vane to be a double-wing
deflector.
160
Table 4.23 (cont’)
Failure Effects on Other Components of Effects on the Whole Methods of Compensating Provisions
Mode RCV System Detection
Head cut - Drop is not sustained and may - Head cut may - Visual Inspection - Repair cross sill with
decrease potential for Scour Pool travel upstream - Photo appropriate footers and backfill
Development. and take out other Documentation material
- Flow may lose ability to access structures. - Install a new knick-point
arms and be constricted leading - Possible structure at the current location
to Downstream Bank Erosion. sedimentation of the head cut.
downstream.
Bank Erosion - Deposition may occur in scour - Decrease in stream - Visual Inspection - Repair banks from incurred
at vane pool preventing scour pool aesthetics. - Photo scour or erosion with plantings
development. - Possible Documentation and matting.
sedimentation - Increase bank armor.
downstream. - If cross vane is too narrow, rock
cross vane might need to be
widened and re-tied in to the
banks
Bank Erosion - None. - Decrease in stream - Visual Inspection - Repair banks from incurred
Downstream aesthetics. - Photo scour or erosion with plantings
of Vane - Possible Documentation a matting.
sedimentation - Increase bank armor.
downstream.
- Decrease in habitat
potential.
Lack of - None. - Decrease in pattern - Visual Inspection - Dig out a scour pool.
Scour Pool and habitat - Photo - Leave as is.
Development potential. Documentation
161
Failure Outcomes
Johnson and Niezgoda (2004) assign weighted ratings for the consequences of
failure based on economic impact. Table 4.24 shows four categories of
consequence, 1 being the least economic consequence and 10 being the highest.
Table 4.24 Consequence Categories
Source: Johnson and Niezgoda (2004)
The consequence rating for the various modes of failure for the rock cross vane are
low to high, depending on the compensating provisions. In the cases where projects
are used to protect bridges, the consequence rating would be critical. However, the
projects observed for this study were mainly rural and where urban, causing minimal
threat to infrastructure or utilities. For bank erosion, especially in cases of large bank
blow-outs from rapid lateral migration, there is an economic impact due to damage
to adjacent properties. The resulting consequence of bank erosion at and
downstream of the vane was rated high is high (7). Lack of scour pool development
has minimal economic impacts and was rated low (1). Head cutting can cause
problems to other structures but it could also stabilize on its own. It was assigned a
rating of marginal (4). Finally, both cases of durability are not economically impacting
unless the other functions suffer as a result and it was deemed that the rock cross
vane should be repaired. These were assigned a rating of marginal (4).
162
Likelihood of occurrence
Johnson and Niezgoda rated the likelihood occurrence based on how well tested the
practice was and how likely it was that the failure mode will occur. The highest rating
in their study was “Reasonably probable; has previously occurred frequently; never
been tested in the field; effectiveness unknown; high maintenance.” The probabilities
of failure indicator assessed by this study were used in the ratings of likelihood of
occurrence. This deviated from Johnson and Niezgoda’s intentions but made the
study more relevant to North Carolina stream projects. A rating of 2 signified less
than 1% moderate to extreme occurrence, a rating of 4 signified a 1-5% moderate to
extreme occurrence, a rating of 6 signified a 5-10% moderate to extreme
occurrence, a rating of 8 signified a 10-30% moderate to extreme occurrence and a
rating of 10 signified a greater than 30% moderate to extreme occurrence.
Ease of detection method
Table 4.25 gives Johnson and Niezgoda’s rating system for detection methods.
Table 4.25 Detection Rating
Source: Johnson and Niezgoda (2004)
All of the modes of failure were easily detectible as they are found through visual
inspection. However, for the field observations to be more dependable than a
snapshot in time, multiple visits with photo points or bank pins are required. All were
given a detection rating of 4.
163
The FMEA for the Rock Cross Vane
Multiplying these three ratings – consequence, occurrence, and decision – gave a

risk priority number. The higher the priority risk number, the less desirable a practice
was. For this FMEA of one practice, a higher risk priority number can help pin-point
the functions that have the highest vulnerability and cost of failure. Table 4.26 shows
that the greatest risk of the rock cross vane was bank erosion at the vane and bank
erosion downstream of the vane due to the high consequence and occurrence
ratings. This was followed by 1) head cutting and arm washout with lower
consequence ratings, 2) sill durability, and 3) lack of scour pool development, which
has minimal economic impact.
Table 4.26 FMEA of a Rock Cross Vane, General Stream Restoration Usage
Consequence Occurrence Detection Risk
Failure Mode Rating Rating Rating Priority
Number
Arm Washout 4 8 4 128
Sill Washout 4 4 4 64
Head cut 4 8 4 128
Bank Erosion at Vane 7 10 4 280
Bank Erosion
7 8 4 224
Downstream of Vane
Lack of Scour Pool
1 8 4 32
Development
Johnson and Niezgoda (2004) emphasized the simplicity of their method and
provided three stream and bridge scour examples to demonstrate the FMEA process
used to determine the best scour countermeasure. By looking at the various priority
risk numbers for counter measures, they were able to identify the best solution for
the give problems. Here the data might be used for comparison to other
technologies looking to accomplish the same functions as data on other structures is
collected. It should be noted that bank erosion ranked higher in this study than in
Johnson and Niezgoda’s (2004) evaluation of rock cross vanes for bridge scour
164
counter measures. This was due to the aforementioned difference in occurrence
rating.
Conclusions
Usefulness of the Qualitative Assessment
The qualitative assessment was successful in providing a rapid assessment of the

state of the practice of installing rock cross vanes in North Carolina stream
restoration projects over the past five years. From this assessment, it was evident
that cross vanes have high durability. Although the occurrence of failures were high
with 92% of the structures incurring some level of failure, the severity of the various
failures across the data set was low to moderate on average. The frequency of two
of the failures, bank erosion at the structure and head cutting, showed that more
attention should be paid to constructing the vane so boulders are properly spaced,
cross vanes are sealed with backfill and fabric matting, and the structures are
correctly located and aligned perpendicular to the flow and not too close to upstream
or downstream bends.
While the rapid assessment is beneficial, in order to understand more fully the
progression of failure, vanes should be documented from installation onward. Photo
documentation could reveal the speed of the process and onset of stabilization and
potentially reveal the interactions of failures with each other. For this study, photo
documentation would not have added pertinent information towards the current state
of the practice, but could have provided a better linear regression of structure failure
over time. The range of age of projects was limited and it was difficult to determine if
age truly had a significant effect on the failures indicators ratings. Also, the one-time
assessment gives no indications on potential effects of catastrophic events and
human and animal interactions with the structures. Watershed parameters cannot
fully explain the success of the various functions including durability, and it must be
assumed that other components affect the success and failure rates. This study
165
attempts to identify the causes by what is seen, the obvious inadequacies in design
and construction.
Because there are class effects of project, it might also have been beneficial to
observe more stream projects so the data could be more all encompassing of
representative streams in North Carolina. It can be difficult to locate and visit enough
projects to conduct the study on as not all projects incorporated the rock cross vane
or even this particular version of rock cross vane design.
Use of the FMEA for North Carolina Streams
In this study, the FMEA shows that for the rock cross vane, bank erosion at and
downstream of the structure maintains the highest risk. While head cutting has a
high occurrence in the qualitative assessment, the economic cost of failure and
repairs is low in comparison to bank erosion and subsequently had a lesser risk. The
greatest benefit to habitat, scour pool development, has the lowest risk of failure
because there is no adequate way to economically quantify the cost of damaged
habitat.
By gathering data on rock cross vane performance in North Carolina Stream

projects, the FMEA was demonstrated to be a relevant tool for use in North Carolina.
Further development of the FMEA through data collection of failure occurrences for
other structures would provide a useful tool for designers. The tool would provide a
risk calculation for commonly used structures from which a designer could choose
the best solution for their design problem and be fully aware of the associated risk.
Standard thresholds for risk could be enforced on stream projects. Likewise
standard thresholds of failure ratings should also be adopted for quality control of
projects.
One stipulation is that designers update the FMEA for changes in the technology
and the site parameters of the structures studied so that the FMEA is applicable to
their given project design. It is also important to keep in mind that the qualitative
166
assessment for failure occurrence is a snapshot approach at capturing the state of
the structure. Future human interactions with the stream and watershed and
catastrophic storm events can bring about unpredicted changes that a one-time
assessment can not capture. Even observing the structure over a period of time
might not be able to give the full story of the structure. This is why it is important to
observe many projects from different locations constructed by a variety of designers
and contractors.
Design Recommendations
As for design recommendations from this study, because bank erosion at the
structure is such a high risk, every measure must be taken to stabilize that
vulnerable region. It is essential that rock cross vanes run perpendicular to flow and
are placed in the straight section of the stream. If rapid lateral migration is a risk,
perhaps the rock cross vane is not the structure of choice. The functions of the rock
cross vane are dependent on the vane arms being tied into the banks. Rapid lateral
migration can shift the brunt of the flow towards the arms rather than the sill. Unless
arms are tightly sealed and tied into the banks and the banks are heavily armored
with vegetation, there is little hope of preventing bank blowout and subsequent side
cutting and head cutting.
Lowering the arm slopes prevents impact of the flow against the banks as in
structures with constriction of flow and allows more vegetative cover. If having mildly
sloped arms (around 3%) does not overly hinder flow contraction as shown in
Chapter 2 for mid to high drops, and does not lead to insufficient scour pool
development as shown in this chapter, perhaps, the design range for arm slopes
should be lowered.
167
Improving Aesthetics of the Rock Cross Vane
As mentioned previously, aesthetics are subjective. However, these three things

typically are related to viewer’s response to the structure: 1) how well the rock
blends in with the natural environment as related to coloring, shape, and visibility, 2)
how visible active signs of failure are, and 3) exposure of unnatural components
such as fabric matting. To improve aesthetics of the rock cross vane, several
changes could be incorporated:
1. When possible, opt for local stone rather than quarried stone. Local stone will
have a more natural shape and color.
2. Rounding off the angles between the arms and sill creates a more fluid and
natural feel to the structure.
3. Strong vegetation and grasses or mosses on top of the arms and banks helps
mask large boulders.
4. Lower sloped arms allow banks to grow more vegetation and allow more
inundation at base flow which helps mask the structure. This may also
prevent bank erosion, which is an eyesore.
5. Fabric matting should not be visible.
These recommendations are made based on observations of aesthetically pleasing

and unpleasing rock cross vanes.
168
Literature Cited
Bernhardt, E.S., Palmer, M.A., Allan, J.D., Alexander, G., Barnas, K., Brooks, S.,
Carr, J., Clayton, S., Dahm, C., Follstad-Shah, J., Galat, D., Gloss, S.,
Goodwin, P., Hart, D., Hassett, B., Jenkinson, R., Katz, S., Kondolf, G.M.,
Lake, P.S., Lave, R., Meyer, J.L., O’Donnell, T.K., Pagano, L., Powell, B.,
Sudduth, E., 2005. Synthesizing U.S. River Restoration Efforts. Science 308:
636–637
Bhuiyan, A.B.M. Faruquzzaman and Richard D. Hey. “Instream J-Vane for Bank
Protection and River Restoration” Available at: http://www.iahr.org. Accessed
29 July 2005.
“Calculated Risk: A Tool for Improving Design Decisions.” Stream Notes to Aid in
Securing Favorable Conditions of Water Flows. October 1998. Stream
Systems Technology Center. Available at: http://stream.fs.fed.us/. Accessed
28 July 2005.
Johnson, Peggy A (a). 1996. “Modeling Uncertainty in Prediction of Pier Scour.”

Journal of Hydraulic Engineering. 122(2):66-72.
Johnson, P.A. (b). 1996. “Uncertainty of Hydraulic Parameters.” Journal of Hydraulic

Engineering. 122(2):112-114
Johnson, P.A. 1999. "Fault Tree Analysis of Bridge Failure due to Scour and
Channel Instability." J. Infrastruct. Syst., 5(1): 35-41.
Johnson, P.A., and Niezgoda, S. L. 2004. "Risk-Based Method for Selecting Bridge
Scour Countermeasures." J. Hydr. Engrg., 130(2): 121-128.
NC Division of Water Quality. 2005. Identification Methods for the Origins of

Intermittent and Perennial streams, Version 3.1. North Carolina Department
of Environment and Natural Resources, Division of Water Quality. Raleigh,
NC.
Roper, Brett et al. 1998. “Durability of Pacific Northwest Instream Structures

Following Floods.” North American Journal of Fisheries Management 18:686-
693, 1998. American Fisheries Society, 18:686–693.
Rosgen, D. L. 2001. "The Cross-Vane, W-Weir and J-Hook Vane Structures: Their
Restoration." Wetlands. 22 pp.
169
Chapter 5:
Restoration and Cross Vane Impacts on Benthic

Macroinvertebrates
Abstract
As an engineering practice, stream restoration demands a set of performance

standards for success evaluation. A common tool used for measuring ecological
success in stream and river restorations is benthic indicators. Because the various
functions of a stream are all interconnected, benthic measures can reflect the overall
performance of these functions. This study addresses ecological standards for
stream restoration and reviews past research to highlight the main effects on benthic
communities, the expected return of benthic communities post-restoration and
address what impacts in-stream structures, the rock cross vane in particular, might
have on benthic macroinvertebrates. Recommendations are made that measures be
taken to incorporated leaf retention enhancements and stream mosses at the vane
to increase benthic presence and help in benthic recovery during disturbances.
Future studies of cross vane impacts on rural streams and studies on leaf retention
and mosses at the vane are also recommended.
Keywords: rock cross vane, stream restoration, benthic macroinvertebrates,

ecological standards
170
Introduction
For engineers, it can be very tempting to evaluate the success of a project or

installed technology by its structural integrity and the performance of its physical
functions. However, in the field of stream restoration there is a growing struggle to
evaluate a stream’s ecological functions in combination with evaluating the stream’s
hydraulic functions in determination of the overall success of a restoration design.
This has many researchers focused on determining the impacts of stream
restoration on benthic communities within the stream and developing ecological
standards for rehabilitated sites (Muotka et al. 2002; Giller 2005; Korsu 2004; Lepori
et al. 2005).
Benthic indicators are frequently used as an ecological measure of stream success.

Because the various functions of a stream are all interconnected, benthic measures
can reflect the overall performance of these functions. This study reviews past
research to highlight the main effects on benthic communities, the expected return of
benthic communities post-restoration and address what impacts in-stream
structures, the rock cross vane in particular, have on benthic macroinvertebrates.
Many of the structures and practices used are relatively new and a more complex
hydraulic understanding of how the structures work is yet to be explored, even more
so the basic impacts of these structures on benthic macroinvertebrates.
Stream flood plane creation and channel relocation, while arguably necessary to
alleviate the stress of high flows in incised and entrenched streams and stabilize
erosion, can be a very invasive process to the riparian ecology. Construction on the
stream creates disturbance which may temporarily harm the existing communities as
the abrupt change does not allow time for communities and vegetation to evolve.
Construction also wipes out much flood plane vegetation. Urban and rural streams
that are candidates for restoration are often already ecologically impaired. Therefore,
it can be argued that habitat enhancements that decrease sedimentation, increase
171
of woody debris and bed form features, and stabilize banks may in turn benefit
community development despite the invasive construction procedures.
The rock cross vane enhances fish habitat in conjunction with holding grade and
directing flow away from the banks. According to Rosgen (2001), professional
hydrologist, rock cross vanes creating a scour pool with secondary circulation
patterns, bank cover, safe areas in the deep pool during high flows, feeding lanes
and spawning. Benthic habitat enhancement is never mentioned as a design goal of
the rock cross vane; however, the cross vane could potentially indirectly enhance
benthic habitat by helping maintain stream capacity which prevents both
sedimentation and erosion, creating diversity of flow and bed material, creating
stream pattern, and circulating oxygen. This study addresses whether these benefits
would be significant or detectable based on related studies.
This study also seeks to answer the question: how might the construction, presence
or failure of a rock-cross vane impair the potential recovery of a benthic community?
In watersheds that are highly urbanized, there might not be a potential for strong
benthic structure return post restoration. However, in rural watersheds where there
are not as many limiting factors to benthic macroinvertebrates, upstream conditions
might show some promise. In this case, rock cross vane failures could have
significant impacts. The results of cross-vane failure are typically bank and bed
erosion. These can increase downstream sedimentation, cause a loss of bed form
and pattern, and take away any potential scour pool. All of these effects could
negatively impact benthic structure. This study addresses whether the negative
impacts would be detectable.
Stream restoration designers should be informed about the potential impacts various
types of cross vane failure to benthic communities, if said impacts indeed could
occur. In the case of poor community return in systems where either weak or great
return is expected, understanding the interconnections between structure presence
or failure and benthic failure might help identify causes of benthic weakness and
172
prescribe methods to remedy the problem. Designers should also be informed of the
factors which most influence benthic return so that these aspects can be maximized
during restoration. This can be achieved by future studies guided by the significant
findings discussed in more detail later in the paper.
Ecological Standards
As mentioned in the introduction, there is an increasing push to have ecological

standards for stream and river restoration. In a 2005 Issue of The Journal of Applied
Ecology, much discussion of ecological standards was generated. With any other
engineering practice, bridge design, machined part, etc, there are a set of standards
of performance based on the expected function. Stream restoration attempts to fix a
natural system that has been harmed through anthropogenic means, and as an
engineering practice should also be held to a certain set of standards based on its
expected function. While not all projects attempt to reach the same level of
performance, it is sill possible to create a list of criteria to which all restorations
should adhere given the name “restoration” - restoration of what? The restoration of
hydraulic, hydrologic and ecological functions is implied.
Giller (2005) begins his editor’s introduction to the topic of ecological standards in
stream restoration by addressing the increasing amount and degree of human
impacts on water systems through engineering, pollution, and urban development.
He then states the importance of restoration and humans seek to rectify the damage
inflicted on water systems. According to Giller, there are several approaches to
restoration: passive, active geared towards the stream channel, active geared
towards the pollution source, and active geared towards management of the
watershed as a whole. The most captivating sentence of the paper and perhaps the
point behind the entire series of papers is: “Despite the huge interest and activity in
river restoration, and the fact that ecologically effective restoration has been pursued
for more than three decades, practical criteria for judging ecological success do not
exist.” Following that, he introduces proposed criteria by Palmer et al. (2005).
173
Because of the rapidly pace of growth of stream restoration practices and the lack of
agreement on what makes a successful project, Palmer et al. (2005) set forth 5
criteria or ecological standards for success. They claim that many practices that are
labeled as restoration might not improve ecology and could in fact cause more harm.
Therefore, they seek to clarify what restoration truly is and present standards for that
restoration. The five standards are: 1) “A guiding image exists: a dynamic ecological
endpoint is identified a priori and used to guide the restoration” 2) “Ecosystems are
improved: the ecological conditions of the river are measurably enhanced” 3)
“Resilience is increased: the river ecosystem is more self-sustaining than prior to the
restoration” 4) “No lasting harm is done: implementing the restoration does not inflict
irreparable harm” 5) “Ecological assessment is completed: some level of both pre-
and post- project assessment is conducted and the information made available.”
Giller (2005) proposes a sixth criteria that the guiding image of the desired
ecological functions of a restoration should be detailed and used for evaluation,
which would aid in research of why restorations succeed or fail and provide future
recommendations. One goal of having standards is to eliminate ineffective
approaches and advance knowledge of the field. Monitoring specific sites would give
insight into what methods are ineffective in given conditions. Having standards
would help to carefully document failures and conduct a fault-tree analysis so that
those who are monitoring a site could have a tool available to help them get to the
root of a given problem and make more informed design choices.
In-stream structures are one component of an entire system. While a single structure
failure might not change the overall effectiveness of a restoration, a series of failures
might lead to overall failure to meet these criteria. Similarly, the installation of a
single vane might not cause permanent damage to the stream, but construction of
many vanes and channel realignment could cause permanent changes. Although
the aforementioned standards cannot practically be applied specifically to each rock
cross vane, measures at individual structures can be taken to improve the overall
ecology and prevent cumulative impacts of failure.
174
Benthic Macroinvertebrates as Indicators of River Health
Benthic macroinvertebrates indicators are an increasingly used measure of river

health. Although river water quality has been traditionally assessed by physical,
chemical and a few biological measurements, Norris and Thoms (1999) discuss how
aquatic biota might be better indicators of the structural and functional integrity of
river ecosystems. While empirical evidence shows that some ecological-level
indicators can asses various conditions of the river, there is no clear way of
quantifying river health. Norris and Thoms encourages further research into the
relationships between river variables that influence the biota. They raise a good
point that while cause and effect relationships have been observed, there needs to
be more understanding to the mechanisms of influence. While benthic
macroinvertebrates have great potential for serving as an indicator for river health,
as Norris and Thoms suggest, it is important to know what factors affect benthic
macroinvertebrates and affect them most strongly. The following section
summarizes several studies that explore various effects on benthic
macroinvertebrates that reflect river health. Table 5.1 gives a summary of the
effects, the studies and findings.
175
Table 5.1 Summary of Effects on Benthic Macroinvertebrates
Effect Study Findings
The study found that the greatest consistent impact on

benthic communities was the forest cover and
Watershed Potter et al. 2004 watershed shape. More forest cover resulted in higher
macroinvertebrate scores in general and rounder
watersheds resulted in lower macroinvertebrate scores.
Higher retention capacity indicated better habitat and a
Muotka and more successful ecological restoration. Restored
Leaf Retention streams had a higher retentive capacity than
Laasonen 2002
channelized streams but still about half as much as
natural stream.
Little thought had been given to benthic
Leaf Retention/ macroinvertebrates in the restored streams, yet there
Single-minded Muotka et al. 2005 was benthic recovery over time. Even small
considerations during restoration such as preserving
Restoration moss cover, increasing woody debris would not benefit
both benthic return and the fisheries salmonoids.
There was a correlation between boulder and woody
debris density and CPOM retentiveness. Velocity had
Lepori, Palm and
Leaf Retention the greatest effect on CPOM breakdown. Restored
Malmqvist 2005 streams performed better than the channelized streams
pre-restoration but still not as well as natural streams
that had not been channelized.
There was a link between bryophytes and benthic
recovery which is attributed to the stream mosses
Bryophyte Korsu 2004 providing refuge during stream disturbance. Stream
mosses play a significant habitat role for benthic
communities.
Increasing heterogeneity at small scales does not have
a significant impact on benthic structure or recovery, but
Habitat rather other site specific factors have a greater impact
Brooks 2002 on these. The effects on benthic structure by point
Enhancements
source treatments are likely to go undetected until there
is a great difference between the treatment and
reference sites.
There is a strong correlation between infilling of the sites
Disturbance Dernie et al. 2003 and benthic recovery. There is also a strong correlation
between the cleanness of the sand and the rate of
recovery.
Wohl and Carline
Wohl and Carline found results consistent with literature
1996 that sediment loads account for differences in benthic
Sediment communities. High loading and reduced substrate size
Wood and Armitage
might have played a part in lower macro benthic
1997 macroinvertebrate density.
Time Muotka et al. 2002 It takes 6-8 years for full recovery and for the population
to stabilize.
176
Watershed Effects
Potter uses linear regression analysis and multiple regression analysis to determine
the effects on watershed land used and physical factors on macroinvertebrate index
scores (Potter et al. 2004). Watershed factors were measured with USEPA stream
reach GIS data and NOOA National Weather Service data. NCBI and EPT indices
were taken from N.C. Division of Water Quality. The study found that the greatest
consistent impact on benthic communities was the forest cover and watershed
shape. More forest cover resulted in higher macroinvertebrate scores in general and
rounder watersheds resulted in lower macroinvertebrate scores. Forest impacts
varied among Appalachian, Piedmont and Coastal watersheds, and the study made
a distinction between watershed scale % forestland and riparian forest cover.
Rounder watersheds lead to a shorter time of concentration which gives the
watershed less time to “self-purify”
This paper was informative and relevant. There is a well-grounded basis for the
study, being that stream impairment is a problem and that benthic communities are
excellent indicators of water quality. There were also clear conclusions and
implications of the study. As watersheds are continually building up and losing forest
cover, it is important to understand the risks to benthic communities. It seemed that
most of the factors effects on the indices could be explained through the impacts of
development. The correlation matrix supports these assertions. Most of the factors
effects seemed intuitive, though not as intuitive was the level of each effect. An
excellent point was made that this tool could be incorporated into policy making and
watershed planning to understand which factors are most important to helping
benthic communities tolerate changing water quality.
Leaf Retention Effects

Because habitat degradation resulting in a loss of leaf retention is one of the biggest
outcomes of stream channelization, Muotka and Laasonen (2002) address the lack
of ecological standards in measuring the success of stream restoration by the
177
quantification of retention capacity. Higher retention capacity indicated better habitat
and a more successful ecological restoration. They study streams in 1993 before
restoration and again in 1996. They compare leaf retention via different means
(cobble, boulder, moss, stream bank, woody debris, etc) to diversity and density of
benthic macroinvertebrates. Restored streams had a higher retentive capacity than
channelized streams but still about half as much as natural stream. In restored
streams, moss cover is often lacking, therefore retention more frequently occurs
through habitat enhancement such as cobble unlike channelized and natural
streams in which moss accounts for a high percentage of the leaf retention. The
other main effect on leaf retention is flow rate. Higher flow rates result in less leaf
retention. The only species that had significant increase post-restoration were
scrapers. The lack in other species is attributed to missing moss cover in restored
streams.
The final assessment was that the benefits of increasing bed heterogeneity through
structures and restoration to improve habitat was compromised by the loss of moss
coverage during construction via the use of heavy equipment. A good part of the
study suffered through a lack of replications due to inadequate flows for sampling
and other problems while sampling. Thus, because of the holes in the study and
because of the projected results being that benthic macroinvertebrates suffered
despite measures to increase retentive capacity through bed heterogeneity, simply
measuring heterogeneity would not be a good indicator of the ecological success of
the restoration. A simple measure of retention capacity (not broken down by
methods of retention) would also be inadequate as benthic return as a whole did not
reflect the increase of retentive capacity.
In a follow-up study to Muotka and Laasonen (2002) that addressed leaf retention in
natural, channelized and restored streams, Muotka et al (2005) study streams
channelized in Finland that have been restored to their original state mainly for the
purpose of enhancing sport fisheries. They study benthic communities to observe
the impacts that the unnatural disturbance of habitat enhancements for sport fishing
178
have on the long-term benthic recovery. They compare habitat variables and benthic
counts in natural, channelized and previously channelized, restored streams. They
believe that one of the biggest consequences of channelization is the retentive
capacity of the stream to hold leaf and other organic matter. Little thought had been
given to benthic macroinvertebrates in the restored streams, yet according to the
study, benthic recovery over time was observed. Because of this, they believe that
even small considerations during restoration such as preserving moss cover,
increasing woody debris would benefit benthic return and the fisheries.
Another leaf retention study is conducted by Lepori, Palm and Malmqvist (2005).
What makes this study unique is that one stream where they study the effects of a
particular restoration practice- the placement of boulders; they also observe the
breakdown of course particulate organic matter (CPOM). They used artificial leaves
to measure retentiveness and a measurement of the loss of leaf packs as an
indicator of breakdown. They also took geomorphologic field measurements and
measurements of taxonomic richness, abundance, biomass and evenness for
shredders. There was a correlation between boulder and woody debris density and
CPOM retentiveness. Velocity had the greatest effect on CPOM breakdown.
According to their results, restored streams performed better than the channelized
streams pre-restoration but still not quite as well as natural streams that had not
been channelized.
While rock cross vanes are constructed of boulders and thus would indicate a higher
rate of retention, at certain flow depths they also increase velocity which would
speed up breakdown. Also, some cross vanes have no boulders in the pool region.
Adding more stones and woody debris in these pools might help reduce excessive
scour and improve habitat by increasing retentive capacity.
Bryophyte Effects
Korsu (2004) looked at the relationship between the presence of bryophytes and
benthic recovery in a boreal stream in Finland. According to his study, his hypothesis
179
was supported in that there was a link between bryophytes and benthic recovery
which he attributes to the stream mosses providing refuge during stream
disturbance. The stream under study received heavy disturbance by an excavator
but also retained patches of the original stream bed which may have aided in the
quick recovery of 2 weeks, lower than expected for the winter compared to other
studies. Sampling was done my collecting 30 stones along the reach and carefully
removing benthic macroinvertebrates and bryophytes. There was a correlation
before restoration and an even stronger correlation after restoration. The study does
30 replication which can give the data an expected normal distribution; however, the
study was conducted on one stream and it is unknown if the results would be similar
in a stream where there were no undisturbed patches of stream on the restoration
site. Overall, this study supports other researchers’ claims that stream mosses play
a significant habitat role for benthic communities. Related to rock cross vane
practices, boulders are usually ordered from an offsite quarry and are very unlikely
to have mosses growing on them. It would be a better practice to utilize local stones
that may already be covered in mosses.
Habitat Enhancement Effects

Brooks et al. (2002) from the Department of Biology, University of Maryland, test the
hypothesis that increased stream riffle heterogeneity leads to more rapid benthic
recovery and/or benthic species richness. They manipulate a series of riffles on a
stream to have either high, medium (reference) or low variability of streambed
particle size. The high and low treatments are conducted at four riffles each at least
50 m apart and by at least one pool > 50 cm deep. Their study finds that increasing
heterogeneity at small scales does not have a significant impact on benthic structure
or recovery, but rather other site specific factors have a greater impact on these.
They claim that the effects on benthic structure by point source treatments are likely
to go undetected until there is a great difference between the treatment and
reference sites.
180
The study is based on the argument that human activities have led to simplified
stream pattern and bed structure which has led to impaired habitat and benthic
structure. They seek to address the assumption that system restoration comes by
maximizing physical habitat diversity. They argue that most studies on system
restoration have been geared towards fish species and more attention needs to be
given towards benthic species which are better indicators of environmental
degradation and influence ecological processes. By increasing heterogeneity at
riffles, the authors test if the rate of recovery of benthic structure and/or the stream
invertebrate species richness actually increases.
The main weakness of this study is the lack of experimental data. There were four
riffles with each treatment, and one of the riffles was destroyed during the study and
could no longer be used for data collection. Also, one stream was studied. It is
difficult to look at one stream to identify how heterogeneity affects benthic
macroinvertebrates. The authors mention that other factors were more influential
than the treatments. Perhaps finding streams where these other factors were not
present or were less strong would help make the effects of the treatment more
observable. This study is related in that heterogeneity of habitat is also influenced by
the presence of the rock cross vane. While they were not able to observe significant
effects based solely on heterogeneity of bed material, the cross vane provides more
extreme changes in both heterogeneity and water quality which may be significant.
Disturbance Effects
Dernie et al. (2003) study benthic community recovery at sites along the Menai
Strait, North Wales, grouped by silt and clay content. The 1m by 4m sites were
strategically disturbed by scraping away a depth a 10cm of surface soil. The study
lasted a total of 213 days from the time of physical disturbance. Benthic sampling
was done 5 times during the 213 days on approximately the same days for all the
sites, depending on weather conditions. They test the hypothesis that clean sands
will have quicker marine benthic recovery after disturbance. According to their data,
there was a strong correlation between infilling of the sites and benthic recovery.
181
There was also a strong correlation between the cleanness of the sand and the rate
of recovery.
This study does not apply well to stream restoration because of the inherent
differences between estuary systems and river systems and because the degree of
disturbance is so much greater in stream restoration when a stream is subjected to
channel realignment and heavy equipment. In this study, removal of 10 cm of
surface soil probably will not result in a completely new soil type; however, in stream
restoration, the entire bed structure is changed. Also, the test sites are relatively
small with a chance of recovery coming from water influxes from all sides of the site.
Stream restoration is usually not limited to a small segment but very long reaches
where recovery comes from a sometimes far removed up and downstream reach.
Sediment Load Effects

Wohl and Carline (1996) study the effects of cattle grazing on sediment loads,
macroinvertebrates and fishes in three streams in central Pennsylvania. Wohl and
Carline found results consistent with literature that sediment loads account for
differences in benthic communities. High loading and reduced substrate size might
have played a part in lower macro benthic macroinvertebrate density.
Wood and Armitage (1997) find that anthropogenic introduction of fine sediment can
cause increased sedimentation in streams particularly around structures and
boulders. These structures often create dead zones where water circulates and
drops out fines. These are only flushed out in high flows with greater turbulence.
They cite sedimentation as affecting benthic macroinvertebrates in four ways:
(1) by altering substrate composition and changing the suitability of

the substrate for some taxa; (2) by increasing drift due to sediment deposition
or substrate instability; (3) by affecting respiration due to the deposition of silt
on respiration structures or low oxygen concentrations associated with silt
deposits and (4) by affecting feeding activities by impeding filter feeding due
182
to an increase in suspended sediment concentrations, reducing the food
value of periphyton and reducing the density of prey items (Wood and
Armitage 1997).
In their paper Wood and Armitage outline a holistic management framework for fine
sediment in streams and rivers. The management framework is meant to aid in
identifying sources of sediment, transport, and deposition processes. They
emphasize that the effects and recovery of flora and fauna are dependent on the
type and magnitude of disturbance (1997).
Potential Recovery and Time Effects

The term recovery is used broadly to define the return of benthic communities to
disturbed areas. Some studies will refer to recovery as a return to conditions before
restoration. An example of this would be a channelized stream that has been
restored where there was a dying off of benthic macroinvertebrates during
construction and a return to pre-disturbances state after a given amount of time.
Other studies talk of recovery as a return to natural conditions before both
restoration and channelization. The idea of stream restoration is to return conditions
close to their natural conditions as governed by the limitations of the watershed, yet
often there is no standard of what recovery truly is.
Perhaps the most accurate definition of recovery as it relates to stream restoration

would be a return to conditions of a reference reach in the same watershed or an
upstream, undisturbed segment of the restored stream. Restoration beyond what the
watershed can provide is unrealistic and immeasurable. Based on the study by
Brooks (2002) watershed effects outweighed effects of increased bed heterogeneity
and therefore restoration by increasing heterogeneity would not improve conditions
beyond the limitations of the watershed.
One study looks at the length of time required to gain full benthic recovery after
stream restoration (Muotka et al. 2002). It was interesting to read that it takes so
183
long (6 to 8 years) for full benthic recovery without fluctuation when another study
asserted that a two week recovery was observed. This might be related to the
different degrees of disturbance during restoration, but an eight year/ two week
difference is a bit of a stretch. Because the other study was not observed for more
than a short period of time, it is questionable if the results would be more similar had
there been a longer study.
Rock Cross Vane Impacts
There is concern over how hard structures such at the rock cross vane might impact
benthic macroinvertebrates. Stream naturally migrate over time and can be
subjected to extreme natural conditions. The rock cross vane is designed to be
permanent making it susceptible to failure in a system that is known to change. This
next section explores the possible effects and addresses these concerns. There is
concern over whether the installation of the vanes causes an irreparable
disturbance, whether the actual presence of the vane negatively impacts benthic
macroinvertebrate functions such as grazing, and what types of effects cross vane
failure might cause.
Impacts of Construction
As discussed in the review of earlier studies, disturbance caused by stream
restoration varies. Stream restoration can involve minimal construction or it can
involve construction of an entirely new channel and flood plane. Table 5.2 lists
restoration priorities and a description of each. Descriptions are taken from the
Stream Restoration Guidebook published by the Stream Restoration Institute of
North Carolina State University.
184
Table 5.2 Restoration Priority Levels
The objective of a Priority 1 project is to replace the incised channel
with a new, stable stream at a higher elevation. This is accomplished
Priority I by excavating a new channel with the appropriate dimension, pattern
and profile (based on reference-reach data) to fit the watershed and
valley type.
The objective of a Priority 2 project is to create a new, stable stream
and floodplain at the existing channel-bed elevation. This is
Priority II
accomplished by excavating a new floodplain and stream channel at
the elevation of the existing incised stream.
Priority 3 is similar to Priority 2 in its objective to widen the floodplain
at the existing channel elevation to reduce shear stress. This is
Priority III accomplished by excavating a floodplain bench on one or both sides
of the existing stream channel at the elevation of the existing bankfull
stage.
Priority 4 projects use various stabilization techniques to armor the
bank in place. These projects do not attempt to correct problems with
dimension, pattern or profile. Priority 4 projects often use typical
Priority IV
engineering practices to harden (armor) one or more stream banks.
Projects may use riprap, concrete, gabions, bioengineering or
combinations of structures to protect stream banks.
Priorities I and II both involve heavy construction of the stream bed and creation of a
new channel. Priority II may keep segments of the existing channel; however,
Priority I restoration cuts out an entirely new channel. In both these cases,
construction of a rock cross vane would not introduce any further disturbance than
was already being conducted. Therefore, cause of damage to benthic
macroinvertebrates could not be isolated to vane construction. In a Priority III
restoration, heavy equipment would be introduced for the installation of structures
and subsequently damage to benthic macroinvertebrates could be attributed to the
vane construction. However, the study by Korsu (2004) would point towards the
argument that disturbance would be minimized by introduction of bryophytes and
recovery would be relatively quick (several weeks). In reality, if there were still large
portions of undisturbed stream, a good size storm could reintroduce benthic
macroinvertebrates to the disturbed areas. Hence, it would seem that in the cases
where construction actually would cause permanent damage or change to the
185
benthic macroinvertebrates, damage is caused by the construction of the entire
systems and not just by rock cross vanes. The more relevant question would be:
how might the accepted practices used in the aforementioned priorities be improved
to reduce stress to the stream during construction? And because the size of rock
cross vanes is large and demands heavy equipment, a following question would be:
what other practice(s) could perform the desired functions of the cross vane but
minimize construction disturbance?
Impacts of Presence
Arguably, there are both positive and negative impacts caused by the presence of
the rock cross vane. The shear size of the structure might prevent grazing from
below the structure to above the structure. The rocks used are oftentimes non-
native. The aforementioned potential benefits, increased flow patterns and bed
heterogeneity, did not result in noticeable improvements to benthic
macroinvertebrates in related studies. The only stream habitat factors other than the
watershed attributes that made a statistical difference to benthic macroinvertebrates
were the presence of stream mosses and leaf retentiveness. It could be said then
that whatever positive impacts the structure provides are negligible and the structure
should only be credited for its hydraulic functions, which typically is the case.
In rural watersheds where the watershed could be in excellent condition and not
limiting benthic macroinvertebrates diversity and abundance, habitat enhancement
effects might be detectable. In this case, effects of rock cross vane presence and/or
failure could also potentially be observed.
Impacts of Structure Failure

Failure of said high durability structures is not uncommon. Frissel and Nawa (1992)
observed 161 fish habitat structures in 15 streams in Western Oregon and
Washington. They looked at rates and causes of physical failure that followed a 2-10
year flood. While there was failure in low order streams, there were much higher
rates of failure in higher-order streams and overall there was little correlation to
186
structure design. Due to their findings, they assert that habitat modifications by
structures may be “inappropriate” or “counter-productive” where there are high flows,
highly erodible banks or high sediment loads. They argue that a great amount of
funding has been directed towards fish habitat improvement despite an obvious lack
of scientific study on the durability and effectiveness of such structures. The paper
gives a brief history of the US Forest Services efforts to improve fish habitat and
refers to studies that have cited structure failures.
Rock cross vanes in North Carolina were observed using a Rock Cross Vane Rapid
Assessment Tool. Bank erosion was found in greater than 60% of the rock cross
vanes (Puckett 2006). Because of the high probability of at least a slight occurrence
of failure, it is likely that rock cross vanes could have a negative effect on benthic
macroinvertebrates. Relating structure failure to benthic macroinvertebrates, failures
of concern would be those that harm leaf retentiveness, increase sediment loads,
homogenize bed form, and destroy bank habitat. When a structure blows out, the
results can be sedimentation or deposition on the stream bed, erosion of banks, loss
of pattern, bed scour. As mentioned earlier, in cases where there are other major
limiting factors, these failures would potentially be of no significance to benthic
macroinvertebrates except in punctuated disturbances such as large storms where
greater amounts of erosion would result in peaking sediment loads and bed
disruption which would eventually stabilize over time such as when a head cut
migration reached the next upstream checkpoint. Even so, during these high flows,
in urban watersheds that are still under development, an increase in sediment in
runoff would be expected as well which might also be significantly greater than that
contributed by the failed vane. In the case of rural watersheds, effect of structural
failure might be more traceable due to a lack of other predominant effects.
Conclusions and Recommendations
It is very difficult to design a study which could observe the effects of rock cross
vane failure impacts on benthic macroinvertebrates. One possible way is to study a
187
stream with multiple cross vanes and compare the benthic structure at each vane.
Because it may be difficult to identify impacts of failure due to high variation in
benthic habitat along the stream, some variables could be controlled by limiting the
study to one stream and one watershed. Of course, to obtain statistically significant
data, multiple streams and watersheds would have to be studied and more factor
levels added. Studying a stream where strong benthic recovery is expected would
be able to indicate problems related to the stream design rather than the watershed.
This would involve finding streams where the other limiting factors on benthic
macroinvertebrates are not as loud as the expected impacts caused by rock cross
vane presence and/or failure. Perhaps more beneficial would be to study new design
recommendations for rock cross vanes that might improve durability and improve
habitat potential based on previous literature.
An assumption of no significant effects from structure failure on benthic

macroinvertebrates in urban watersheds and potential effects in rural watersheds
would be presumptuous without actual studies. However, looking at other studies on
benthic macroinvertebrates provides a good basis for addressing what aspects of in-
stream structures ought to be considered more carefully, which aspects are not as
significant, and what measures could be done in conjunction with the structures to
benefit benthic recovery in restored streams. As habitat enhancements such as bed
heterogeneity did little to aid in recovery in one study (Brooks 2002), other factors
that do not directly relate to the vane were crucial for benthic survival and recovery.
Leaf retentiveness was studied by several researchers who all found that increased
leaf retentiveness signaled higher levels of benthic macroinvertebrates. Adding
woody debris around the cross vane would help increase retentiveness. Rock cross
vanes typically have various size cobbles which are effective in increasing
retentiveness (Lepori et al. 2005). It is important though that measures be taken to
account for the increased velocities that occur over the vane as it often times can
cause flow contraction. Another major aid to benthic recovery is the presence of
stream mosses (Korsu 2004). The boulders from rock cross vanes are often from
188
quarries and as a result have no mosses. Using stones local to the watershed or
taking measures to introduce mosses to the structure would enhance ability of the
benthic macroinvertebrates to survive structure failures and other natural
disturbances.
189
Literature Cited
Brooks, S. S., Palmer, M. A., Cardinale, B. J., Swan, C. M., and Ribblett, S. (2002).
"Assessing Stream Ecosystem Rehabilitation: Limitations of Community
Structure Data." Restoration Ecology, 10(1): 156-168.
Dernie, K. M., Kaiser, M. J., and Warwick, R. M. (2003). "Recovery rates of benthic
communities following physical disturbance." The Journal of Animal Ecology,
72(6): 1043-1056.
Frissel, C. A., and Nawa, R. K. (1992). "Incidence and causes of physical failure of
artificial habitat structures in streams of western Oregon and Washington."
NAJFM, 12(1): 182-197.
Giller, P. S. (2005). "SPECIAL PROFILE: River restoration: seeking ecological

standards. Editor's introduction." The Journal of Applied Ecology, 42(2): 201-
207.
Korsu, K. (2004). "Response of Benthic Invertebrates to Disturbance from Stream

Restoration: The Importance of Bryophytes." Hydrobiologia, 523(1-3): 37-45.
Lepori, F., Palm, D., and Malmqvist, B. (2005). "Effects of stream restoration on
ecosystem functioning: detritus retentiveness and decomposition." The
Journal of Applied Ecology, 42(2): 228-238.
Muotka, T., and Laasonen, P. (2002). "Ecosystem recovery in restored headwater

streams: the role of enhanced leaf retention." The Journal of Applied Ecology,
39(1):145-156.
Muotka, T., Paavola, R., Haapala, A., Novikmec, M., and Laasonen, P. (2002).
"Long-term recovery of stream habitat structure and benthic invertebrate
communities from in-stream restoration." Biological Conservation, 105(2):
243-253.
Norris, R. H., and Thoms, M. C. (1999). "What is river health?" Freshwater Biology,
41(2): 197-209.
Palmer, M. A., Bernhadt, E. S., Allan, J. D., Lake, P. S., Alexander, G., Brooks, S.,
Carr, J., Clayton, S., Dahm, C. N., Shah, J. F., Galat, D. L., Loss, S. G.,
Goodwin, P., Hart, D. D., Hasset, B., Jenkinson, R., Kondolf, G. M., Lave, R.,
Meyer, J. L., O’Donnell, T. K., Pagano, L., and Sudduth, E. (2005).
"Standards for ecologically successful river restoration." The Journal of
Applied Ecology, 42(2): 208-217.
190
Rosgen, D L 2001. “The Cross-Vane, W-Weir and J-Hook Vane Structures...Their
Restoration.” Wildland Hydrology, Inc., 22 pp.
191
Chapter 6:
Summary Conclusions and Future Research
192
Flume Study
Summary of Results
The statistical model was cumbersome for design use, so independent and dependent
variables were lumped in ranges to simplify the effects on the velocity distribution of flow
coming out of the rock cross vane at cross section C.
1. For drops less than or equal to 25 % bankfull depth, the velocity ratio ranged
from 1.25 – 1.5. Increasing slope increased the velocity ratio.
2. For drops greater than 25% bankfull depth, the velocity ratio values ranged from
1.75 – 2.25. Increasing arm angle increased the velocity ratio.
Simplifying this by using only highs and lows for each parameter and high, low and
medium for the velocity ratio gave the following relationships:
1. Low Drop = Low velocity ratio
2. High Drop + Low Angle = Medium Velocity ratio
3. High Drop + High Angle = High Velocity ratio
Where,
low velocity ratio is less than 1.5
medium velocity ratio is 1.5-2.0
high velocity ratio is greater than 2.0
low drop is 25% bankfull depth or less
high drop is greater than 25% of bankfull depth
low angle is 25 degrees or less
high angle is greater than 25 degrees
Slope is not shown because varying slope never increased velocity ratio from a low to a
high or a high to a low velocity ratio.
The example calculations for area ratio at the estimated critical depth at the structure for
drop ratio of 0.75 predict conditions of overtopping constriction within the prescribed
193
range of arm slope (.03 – .07 ) and arm angle (20 – 30 degrees). This implies that area
calculations should be incorporated into rock cross vane design. Increasing flow area
(area ratio>1) helps slow down velocities across the structure. As the area ratio reduces
to less than one, velocities will increase over the structure increasing the risk to bank
failure at the structure, or the water surface level will increase and result in an increased
risk of bank erosion at the point of flow re-entry into the stream just downstream of the
arms.
Recommendations for Changes to Rock Cross Vane Design
There are many components to the success of a rock cross vane. It is recommended
that arm angle and arm slope be set to achieve an area ratio that ensures protection of
the physical structure at all drops and be adjusted to maximize the velocity ratio at low
drops where the velocity ratio is expected to be lower. At the lowest drop ratio (0.25),
arms were observed to be nearly fully inundated, and tail water depth caused backing
up over the vane causing arm slope to be an important factor for increasing center
velocity. High arm slopes could increase the velocity ratio at low drops, but designers
must ensure that the cross vane is wide enough and the arm angles are great enough
to prevent constriction of the flow over the vane. While scour pool formation is
important, sedimentation from eroded banks due to poor construction might negate
habitat benefits a pool provides. The qualitative assessment in Chapters 3 and 4 led to
the conclusion that higher arm slopes tend to be coupled with problems with piping and
side cutting. This supports the aforementioned example of flow area constriction or
overtopping of the bankfull depth in the test flume at the estimated flow depth for the
base discharge rate. Constriction would lead to higher velocities on the arms and
potentially scour around them. Constriction might lead to downstream bank erosion from
expansion after the constriction through the rock cross vane.
Downstream velocities are expected to be higher at greater drops, and precautions

should be made to protect the banks from high near-bank velocities. At high drops, less
194
tailwater was observed, and the velocity ratio was influenced more strongly by the angle
of the arms than the arm slope. Increasing arm angle was resulted in higher velocity
ratios with swift undular flow after the drop with nearly still water at the walls. This
observed relationship is consistent with field observations of deposition at the foot of the
arms and scour in the thalweg between the arms. At medium to high drop ratios (>0.25),
arm slope could be lowered to increase arm and bank stability with a negligible
decrease in velocity ratio. At high arm slopes, higher arm angles are essential to
prevent constriction and overtopping as higher angles increase cross-sectional area for
a fixed arm slope.
Future Research
While a mathematic model incorporating three dimensional effects of geometry on

velocity distribution was not generated, a model could be developed pending calibration
of coefficients of discharge and further model studies of rock cross vanes and eddies.
Some of this could be accomplished by flume studies. The difficulty of trying to achieve
constant conditions in the flume with variability from sources such as warping plywood,
changing flow, and unpredictable elements of the weather, could have been prevented.
This would require:
1. A concrete or pvc channel to increase stiffness and uniformity
2. A water source with a pump to recycle flow to maintain a constant head
3. An indoor facility to shield from influences of wind and extreme temperatures
The rock cross vane acts as a long-crested weir because the overflow surface is greater
than the width of the channel. It would be beneficial to calibrate a weir equation which
would be used for better modeling of flow depths and discharge around the structure.
195
Qualitative Assessment of NC Rock Cross Vanes
Usefulness of the Qualitative Assessment
The qualitative assessment was successful in providing a rapid assessment of the state
of the practice of installing rock cross vanes in North Carolina stream restoration
projects over the past five years. From this assessment, it was evident that cross vanes
have high durability. Although the occurrence of failures were high with 92% of the
structures incurring some level of failure, the severity of the various failures across the
data set was low to moderate on average. The frequency of two of the failures, bank
erosion at the structure and head cutting, showed that more attention should be paid to
constructing the vane so boulders are properly spaced, cross vanes are sealed with
backfill and fabric matting, and the structures are correctly located and aligned
perpendicular to the flow and not too close to upstream or downstream bends.
While the rapid assessment is beneficial, in order to understand more fully the
progression of failure, vanes should be documented from installation onward. Photo
documentation could reveal the speed of the process and onset of stabilization and
potentially reveal the interactions of failures with each other. For this study, photo
documentation would not have added pertinent information towards the current state of
the practice, but could have provided a better linear regression of structure failure over
time. The range of age of projects was limited and it was difficult to determine if age
truly had a significant effect on the failures indicators ratings. Also, the one-time
assessment gives no indications on potential effects of catastrophic events and human
and animal interactions with the structures. Watershed parameters cannot fully explain
the success of the various functions including durability, and it must be assumed that
other components affect the success and failure rates. This study attempts to identify
the causes by what is seen, the obvious inadequacies in design and construction.
Because there are class effects of project, it might also have been beneficial to observe
more stream projects so the data could be more all encompassing of representative
streams in North Carolina. It can be difficult to locate and visit enough projects to
196
conduct the study on as not all projects incorporated the rock cross vane or even this
particular version of rock cross vane design.
Use of the FMEA for North Carolina Streams
In this study, the FMEA shows that for the rock cross vane, bank erosion at and
downstream of the structure maintains the highest risk. While head cutting has a high
occurrence in the qualitative assessment, the economic cost of failure and repairs is low
in comparison to bank erosion and subsequently had a lesser risk. The greatest benefit
to habitat, scour pool development, has the lowest risk of failure because there is no
adequate way to economically quantify the cost of damaged habitat.
By gathering data on rock cross vane performance in North Carolina Stream projects,
the FMEA was demonstrated to be a relevant tool for use in North Carolina. Further
development of the FMEA through data collection of failure occurrences for other
structures would provide a useful tool for designers. The tool would provide a risk
calculation for commonly used structures from which a designer could choose the best
solution for their design problem and be fully aware of the associated risk. Standard
thresholds for risk could be enforced on stream projects. Likewise standard thresholds
of failure ratings should also be adopted for quality control of projects.
One stipulation is that designers update the FMEA for changes in the technology and
the site parameters of the structures studied so that the FMEA is applicable to their
given project design. It is also important to keep in mind that the qualitative assessment
for failure occurrence is a snapshot approach at capturing the state of the structure.
Future human interactions with the stream and watershed and catastrophic storm
events can bring about unpredicted changes that a one-time assessment can not
capture. Even observing the structure over a period of time might not be able to give the
full story of the structure. This is why it is important to observe many projects from
different locations constructed by a variety of designers and contractors.
197
Design Recommendations
As for design recommendations from this study, because bank erosion at the structure
is such a high risk, every measure must be taken to stabilize that vulnerable region. It is
essential that rock cross vanes run perpendicular to flow and are placed in the straight
section of the stream. If rapid lateral migration is a risk, perhaps the rock cross vane is
not the structure of choice. The functions of the rock cross vane are dependent on the
vane arms being tied into the banks. Rapid lateral migration can shift the brunt of the
flow towards the arms rather than the sill. Unless arms are tightly sealed and tied into
the banks and the banks are heavily armored with vegetation, there is little hope of
preventing bank blowout and subsequent side cutting and head cutting.
Lowering the arm slopes prevents impact of the flow against the banks as in structures
with constriction of flow and allows more vegetative cover. If having mildly sloped arms
(around 3%) does not overly hinder flow contraction as shown in Chapter 2 for mid to
high drops, and does not lead to insufficient scour pool development as shown in this
chapter, perhaps, the design range for arm slopes should be lowered.
Improving Aesthetics of the Rock Cross Vane
As mentioned previously, aesthetics are subjective. However, these three things

typically are related to viewer’s response to the structure: 1) how well the rock blends in
with the natural environment as related to coloring, shape, and visibility, 2) how visible
active signs of failure are, and 3) exposure of unnatural components such as fabric
matting. To improve aesthetics of the rock cross vane, several changes could be
incorporated:
1. When possible, opt for local stone rather than quarried stone. Local stone will
have a more natural shape and color.
198
2. Rounding off the angles between the arms and sill creates a more fluid and
natural feel to the structure.
3. Strong vegetation and grasses or mosses on top of the arms and banks helps
mask large boulders.
4. Lower sloped arms allow banks to grow more vegetation and allow more
inundation at base flow which helps mask the structure. This may also prevent
bank erosion, which is an eyesore.
5. Fabric matting should not be visible.
These recommendations are made based on observations of aesthetically pleasing and

unpleasing rock cross vanes.
Potential Effects of Rock Cross Vane Effects on Benthos

It is very difficult to design a study which could observe the effects of rock cross vane
failure impacts on benthic macroinvertebrates. One possible way is to study a stream
with multiple cross vanes and compare the benthic structure at each vane. Because it
may be difficult to identify impacts of failure due to high variation in benthic habitat along
the stream, some variables could be controlled by limiting the study to one stream and
one watershed. Of course, to obtain statistically significant data, multiple streams and
watersheds would have to be studied and more factor levels added. Studying a stream
where strong benthic recovery is expected would be able to indicate problems related to
the stream design rather than the watershed. This would involve finding streams where
the other limiting factors on benthic macroinvertebrates are not as loud as the expected
impacts caused by rock cross vane presence and/or failure. Perhaps more beneficial
would be to study new design recommendations for rock cross vanes that might
improve durability and improve habitat potential based on previous literature.
An assumption of no significant effects from structure failure on benthic

macroinvertebrates in urban watersheds and potential effects in rural watersheds would
199
be presumptuous without actual studies. However, looking at other studies on benthic
macroinvertebrates provides a good basis for addressing what aspects of in-stream
structures ought to be considered more carefully, which aspects are not as significant,
and what measures could be done in conjunction with the structures to benefit benthic
recovery in restored streams. As habitat enhancements such as bed heterogeneity did
little to aid in recovery in one study (Brooks 2002), other factors that do not directly
relate to the vane were crucial for benthic survival and recovery. Leaf retentiveness was
studied by several researchers who all found that increased leaf retentiveness signaled
higher levels of benthic macroinvertebrates. Adding woody debris around the cross
vane would help increase retentiveness. Rock cross vanes typically have various size
cobbles which are effective in increasing retentiveness (Lepori et al. 2005). It is
important though that measures be taken to account for the increased velocities that
occur over the vane as it often times can cause flow contraction. Another major aid to
benthic recovery is the presence of stream mosses (Korsu 2004). The boulders from
rock cross vanes are often from quarries and as a result have no mosses. Using stones
local to the watershed or taking measures to introduce mosses to the structure would
enhance ability of the benthic macroinvertebrates to survive structure failures and other
natural disturbances.
Summary Recommendations
Design recommendations for the rock cross vane were previously discussed: lowering
the slopes on the arms and ensuring there is no constriction on flow. Due to the high
levels of bank erosion from side cutting and improper alignment and location, stream
designers either need to be able to guarantee no rapid lateral migration will occur post
construction or they need to adapt the rock cross vane to be able to withstand stream
adjustments. If the rock cross vane were less angular and more parabolic shaped with
arms mildly sloping upwards from the center and deeply recessed into the banks, no
matter how the alignment shifted, water would always be turned back to the intended
downstream thalweg. Also, the low slope of the arms would leave a greater portion of
the banks soft (not armored by rock) allowing more vegetative protection. As the stream
200
laterally migrates, the cross vane would still be able to act as a sill and provide grade
control. A curve which fits this description and is commonly used in civil structure design
for arches is the Catenary curve used in archway design. Removing the angle from
between the sill and arms would give a more fluid design with better protection for all
boulders from drag and rolling as they are protected in multiple directions by boulders
on either side. This design would need to be tested and further developed before full
implementation into the practice.
Recommendations for Further Research

This study is a beginning point for critically analyzing the structures used in stream
restoration design. The summary recommendations need to be carried out and
reviewed to see if there are observed improvements to the functions and durability of
the rock cross vane. There are many other structures and variations on the rock cross
vane that should be assessed for their states of practice. Physical models of boulder
structures should be developed to help bring greater insight to the structures’ hydraulic
performance and durability. Similar qualitative assessments and failure guidebooks
need to be developed for these structures as well. Structures and structure
combinations should be assessed for occurrence of failure so that FMEAs can be used
as a measure for selecting the best practice for a given design problem. And finally, the
moral issues surrounding the practice of boulder structures need to be thoroughly
addressed: 1) Which circumstances are appropriate for hard structure use? 2) Is the risk
of failure too great? 3) How can habitat effects be incorporated into estimating cost and
risk? 4) Is there a way hard structures can be softened and better incorporate habitat
features while still maintaining high durability and functional capacity? As the practice
develops, the changes in methods and design should be documented so that future
generations have a clear history of stream restoration and bench marks from which to
assess goals and direct future study.
201
Literature Cited
Brooks, S. S., Palmer, M. A., Cardinale, B. J., Swan, C. M., and Ribblett, S. (2002).
Brooks, S. S., Palmer, M. A., Cardinale, B. J., Swan, C. M., and Ribblett, S.
(2002). "Assessing Stream Ecosystem Rehabilitation: Limitations of Community
Structure Data." Restoration Ecology, 10(1): 156-168.
Korsu, K. (2004). "Response of Benthic Invertebrates to Disturbance from Stream

Restoration: The Importance of Bryophytes." Hydrobiologia, 523(1-3): 37-45.
Lepori, F., Palm, D., and Malmqvist, B. (2005). "Effects of stream restoration on
ecosystem functioning: detritus retentiveness and decomposition." The Journal of
Applied Ecology, 42(2): 228-238.
202
Appendices
203
Appendix A:
Tables and Figures
204
Figure A-1 Profile View of Cross Vane Arm
Source: Rosgen 1999
Figure A-2 Plan View of Rock Cross Vane
Source: Rosgen 1999
205
Table A-1 Fractional Factorial Design Test Points
Level of Factor
Point Angle Slope Drop
1 1 1 1
2 1 1 2
3 1 1 3
4 1 2 1
5 1 2 2
6 1 2 3
7 1 3 1
8 1 3 2
9 1 3 3
10 2 1 1
11 2 1 2
12 2 1 3
13 2 2 1
14 2 2 2
15 2 2 3
16 2 3 1
17 2 3 2
18 2 3 3
19 3 1 1
20 3 1 2
21 3 1 3
22 3 2 1
23 3 2 2
24 3 2 3
25 3 3 1
26 3 3 2
27 3 3 3
206
Table A-2 Velocity (ft/s) at Given Point for Test Run
Left column describes the test run where D=depth ratio, A=arm angle, and S=arm slope.
Numbers 1, 2, 3, 4, and 5 are the factor level. There are three repetitions of each test run. The
header represents the test point location. A is for cross section A, B is for cross section B and
C is for cross section C. Points 1 – 5 are taken from left bank to right bank.
Test
Run A1 A2 A3 A4 A5 B1 B2 B3 B4 B5 C1 C2 C3 C4 C5
D1A1S1 2.41 2.53 3.06 2.05 2.25 0.69 2.79 2.46 2.62 1.08 3.06 3.61 3.26 3.66 2.88
D1A1S1 2.32 2.37 2.82 2.00 2.58 1.20 2.86 2.66 2.71 0.58 2.91 3.55 3.34 3.57 3.04
D1A1S1 2.47 2.43 2.95 2.09 2.46 1.16 2.85 2.48 2.71 1.01 3.39 3.85 3.54 3.54 2.96
D1A1S2 2.17 2.12 2.79 1.84 2.27 1.44 3.39 3.01 2.79 1.42 1.97 3.78 3.42 3.38 2.34
D1A1S2 2.10 2.15 2.69 1.89 2.22 0.83 3.73 3.10 2.77 1.66 2.47 3.68 3.49 3.36 2.56
D1A1S2 2.19 2.19 2.71 1.90 2.19 1.66 3.50 2.96 2.75 1.61 2.58 3.82 3.42 3.52 2.69
D1A1S3 1.93 1.49 2.53 1.49 1.89 1.47 2.27 3.22 2.67 1.31 2.31 3.82 3.45 3.85 2.69
D1A1S3 1.92 1.75 2.25 1.52 1.77 1.55 2.74 3.25 2.37 1.63 2.41 4.19 3.64 4.14 3.10
D1A1S3 1.86 1.91 2.35 1.42 1.81 1.51 2.56 3.30 2.77 1.66 2.43 3.88 3.63 4.06 3.01
D1A2S1 2.16 2.11 2.82 1.80 2.46 1.90 2.31 2.25 2.24 1.42 2.69 3.45 3.44 3.54 2.55
D1A2S1 2.16 2.22 2.66 2.09 2.58 1.70 2.30 2.47 2.62 1.59 2.85 3.58 3.34 3.39 2.50
D1A2S1 2.12 2.31 2.96 2.05 2.55 1.64 2.31 2.27 2.47 1.55 2.90 3.49 3.45 3.66 2.58
D1A2S2 2.08 2.10 2.65 2.00 2.41 1.52 2.91 2.40 2.47 1.24 2.48 3.28 3.41 3.39 2.41
D1A2S2 2.10 2.04 2.69 1.97 2.41 1.88 2.71 2.41 2.43 1.33 2.56 3.01 3.42 3.34 2.55
D1A2S2 2.13 2.23 2.77 2.03 2.37 1.71 2.93 2.47 2.37 1.21 2.90 3.47 3.30 3.49 2.54
D1A2S3 1.92 1.97 2.60 1.70 2.24 1.12 3.14 2.72 2.60 1.01 2.61 3.68 3.54 3.47 2.56
D1A2S3 2.14 2.02 2.74 1.82 2.25 1.08 3.20 2.80 2.77 1.11 2.34 3.61 3.57 3.58 2.53
D1A2S3 2.07 2.11 2.66 1.83 2.23 1.12 3.20 2.80 2.72 1.10 2.75 3.82 3.63 3.68 2.48
D1A3S1 1.85 2.08 2.49 1.76 1.45 1.44 2.50 2.11 2.23 1.55 1.73 1.79 3.42 3.42 2.79
D1A3S1 1.44 1.82 2.22 1.46 2.00 1.62 2.10 2.07 1.62 1.37 1.25 1.57 2.08 1.90 2.15
D1A3S1 2.11 2.22 2.65 1.87 2.10 1.66 2.42 1.99 2.03 1.47 2.93 3.57 3.33 3.30 2.49
D1A3S2 2.19 2.31 2.65 2.01 2.12 1.52 2.74 2.22 2.40 1.55 1.81 2.82 2.67 3.47 2.28
D1A3S2 2.02 1.95 2.34 1.71 1.59 1.46 2.99 2.19 2.42 1.39 2.91 3.74 3.20 3.34 2.17
D1A3S2 2.37 2.31 2.79 1.87 1.71 1.76 3.03 2.08 2.41 1.21 1.97 2.34 2.90 3.58 2.28
D1A3S3 2.07 2.07 2.62 1.70 1.55 1.54 3.20 2.41 2.43 0.85 2.37 3.36 3.34 3.49 2.09
D1A3S3 2.25 2.18 2.65 1.73 1.57 1.55 3.18 2.43 2.47 0.78 2.58 3.49 3.47 3.63 2.16
D1A3S3 1.99 2.13 2.56 1.81 2.36 1.55 3.03 2.79 2.66 1.03 2.75 3.68 3.41 3.57 2.13
D1BASE 2.59 2.52 3.10 2.03 2.46 1.98 2.24 2.21 2.25 1.97 3.23 3.68 3.30 3.38 3.14
D1BASE 2.55 2.55 2.82 2.17 2.55 1.98 2.18 2.23 2.17 1.88 2.58 3.30 3.15 3.34 2.91
D1BASE 2.49 2.67 2.99 2.10 2.62 1.93 2.14 2.19 2.15 2.04 3.06 3.47 3.39 3.44 3.01
D2A1S1 2.25 2.78 3.25 2.03 2.00 1.76 2.28 3.33 2.48 1.57 1.51 2.61 3.52 2.03 1.27
D2A1S1 2.27 2.75 3.20 2.10 1.95 1.67 2.17 3.03 2.72 1.76 1.34 2.38 3.20 1.89 1.46
D2A1S1 2.28 2.86 3.42 2.13 1.94 1.84 2.31 3.47 2.59 1.81 1.42 2.31 3.38 2.02 1.50
D2A1S2 2.30 2.80 3.25 2.16 1.80 1.03 1.72 3.92 2.08 0.59 1.57 2.19 3.97 2.48 1.80
D2A1S2 2.34 2.73 3.09 2.19 1.98 1.04 2.25 4.24 2.29 0.61 1.57 2.19 3.97 2.48 1.80
D2A1S2 2.38 2.91 3.45 2.21 2.03 0.96 1.83 4.06 2.08 0.56 1.42 1.89 3.45 2.24 1.39
D2A1S3 2.17 2.42 2.96 1.97 1.91 1.31 2.27 4.79 3.28 1.24 1.62 2.40 3.71 2.67 1.88
D2A1S3 2.13 2.38 3.01 2.05 1.81 1.32 2.36 4.74 2.90 0.67 1.60 2.48 3.49 2.61 1.95
D2A1S3 2.11 2.31 3.12 1.95 1.89 1.09 2.46 4.76 3.06 1.15 1.78 2.34 3.85 2.72 1.73
D2A2S1 2.16 2.62 3.01 2.09 2.08 1.23 2.96 4.52 2.31 0.94 1.66 3.03 3.92 2.43 1.56
207
Table A-2 (cont’)
Test
D2A2S1 2.34 2.66 3.18 2.19 2.10 1.04 2.96 4.21 2.02 0.83 1.53 3.12 4.26 2.74 1.73
D2A2S1 2.32 2.66 2.95 2.14 2.00 1.16 2.95 4.45 2.14 1.11 1.44 2.98 3.97 2.29 1.64
D2A2S2 2.13 2.30 2.71 1.97 1.65 1.57 3.20 4.88 3.22 1.36 1.76 3.14 3.54 3.03 1.49
D2A2S2 2.05 2.38 2.67 1.89 1.73 1.57 3.17 4.59 3.28 1.17 1.57 2.98 3.28 3.01 1.48
D2A2S2 2.05 2.28 3.01 1.83 1.60 1.54 3.15 4.16 3.17 1.16 1.52 2.80 3.42 3.14 1.38
D2A2S3 1.86 2.11 2.47 1.71 1.49 1.54 2.86 5.36 4.40 1.27 1.47 2.86 3.47 3.64 1.52
D2A2S3 1.89 2.13 2.56 1.66 1.49 1.55 3.01 5.27 3.92 1.28 1.54 2.86 3.50 3.49 1.55
D2A2S3 1.91 2.17 2.53 1.82 1.61 1.57 2.82 5.94 4.11 1.20 1.50 2.72 3.47 3.78 1.46
D2A3S1 2.23 2.38 3.26 2.13 1.96 0.83 3.22 4.30 2.88 1.18 1.44 2.96 4.31 3.58 1.78
D2A3S1 2.18 2.47 3.18 2.10 1.83 0.99 2.80 4.57 2.86 1.20 1.47 2.72 4.21 3.49 1.90
D2A3S1 2.23 2.48 3.25 2.03 1.78 0.87 2.86 4.33 2.77 1.06 1.30 2.67 4.22 3.68 1.85
D2A3S2 2.24 2.60 2.95 2.08 1.84 1.51 3.47 5.27 2.96 1.31 1.55 3.49 3.87 2.78 1.36
D2A3S2 2.24 2.58 3.34 2.13 1.90 1.31 3.36 5.12 3.42 1.30 1.57 3.49 4.00 2.91 1.37
D2A3S2 2.24 2.71 3.25 2.10 1.93 1.42 3.73 5.19 3.20 1.23 1.69 3.54 4.16 2.75 1.51
D2A3S3 1.95 2.07 2.86 1.98 1.78 1.51 4.19 4.88 4.11 1.48 1.66 4.11 3.85 3.78 1.70
D2A3S3 2.03 2.34 2.82 1.88 1.88 1.59 4.06 5.41 3.87 1.54 1.57 3.93 3.69 3.61 1.64
D2A3S3 1.98 2.18 2.62 2.09 1.76 1.61 4.21 5.33 4.24 1.47 1.48 3.81 3.63 3.41 1.73
D2BASE 2.35 2.68 3.33 2.24 2.18 3.52 4.09 4.21 4.07 4.19 2.46 3.33 3.52 3.26 2.24
D2BASE 2.37 2.90 3.58 2.28 2.14 4.02 4.09 4.47 3.85 4.09 2.43 3.42 3.66 3.39 2.30
D2BASE 2.41 2.85 3.58 2.25 2.18 4.06 4.11 4.24 3.95 4.21 2.50 3.39 3.39 3.50 2.43
D3A1S1 2.40 2.89 3.74 2.22 2.17 1.50 1.96 2.85 1.67 0.59 1.52 2.62 4.02 2.88 1.66
D3A1S1 2.44 3.06 3.74 2.34 2.29 1.46 2.05 2.82 1.74 0.93 1.60 2.75 4.02 3.04 1.95
D3A1S1 2.46 3.10 3.71 2.34 2.16 1.51 1.91 2.74 1.90 0.97 1.76 2.61 4.14 3.04 1.86
D3A1S2 2.23 2.79 3.47 2.29 2.08 1.14 2.00 3.07 2.31 0.66 1.86 2.60 4.14 3.06 2.17
D3A1S2 2.27 2.96 3.34 2.29 2.14 1.23 2.14 3.31 1.94 0.78 1.78 2.56 4.17 2.85 2.05
D3A1S2 2.23 2.85 3.34 2.28 2.05 1.26 2.05 3.18 2.11 1.12 1.89 2.63 4.19 2.92 2.00
D3A1S3 2.41 2.85 3.28 2.27 2.08 1.58 3.12 3.81 2.67 1.11 2.03 2.90 3.71 3.10 2.00
D3A1S3 2.38 2.96 3.44 2.29 2.11 1.89 3.34 3.95 2.58 1.22 1.95 2.82 3.90 3.34 1.99
D3A1S3 2.37 2.98 3.49 2.25 2.13 1.11 3.23 4.28 2.55 1.13 2.10 3.04 3.95 3.18 1.99
D3A2S1 2.47 3.07 3.49 2.34 2.17 1.06 2.44 4.74 1.92 0.64 1.78 3.10 4.55 3.10 1.89
D3A2S1 2.41 2.93 3.36 2.37 2.07 1.03 2.41 5.51 1.96 0.82 1.87 2.99 4.45 3.04 1.83
D3A2S1 2.47 2.88 3.58 2.38 2.21 1.05 2.43 4.98 2.00 0.78 1.86 3.04 4.45 2.86 1.73
D3A2S2 2.30 2.66 3.10 2.19 2.12 1.26 2.77 4.79 2.54 1.26 1.89 3.18 4.26 3.45 2.05
D3A2S2 2.24 2.82 3.28 2.19 2.07 1.20 2.95 5.41 2.34 1.30 1.95 3.34 4.00 3.30 2.04
D3A2S2 2.32 2.77 3.28 2.24 2.10 1.11 2.77 4.40 2.40 1.28 1.86 3.49 4.26 3.50 2.04
D3A2S3 2.38 2.67 3.04 2.03 2.07 1.54 3.42 7.91 3.95 1.30 1.79 3.15 3.87 3.79 1.81
D3A2S3 2.30 2.58 2.96 2.05 2.14 1.51 3.42 7.62 4.14 1.38 1.83 3.28 3.87 3.73 1.85
D3A2S3 2.38 2.69 3.06 2.25 2.01 1.49 3.57 7.76 3.95 1.32 1.86 3.23 3.85 3.90 1.82
D3A3S1 2.40 3.04 3.34 2.25 2.10 0.95 2.34 5.12 2.10 0.77 1.40 2.91 4.83 3.33 1.83
D3A3S1 2.30 2.90 3.36 2.18 2.10 0.74 2.22 5.19 2.07 0.72 1.52 2.72 4.57 3.39 1.78
D3A3S1 2.28 2.85 3.42 2.28 2.10 0.68 2.24 5.51 1.95 0.80 1.48 2.91 4.59 3.49 1.95
D3A3S2 2.35 2.84 3.39 2.25 2.08 1.03 2.72 5.99 1.96 0.84 1.47 2.99 4.79 3.22 1.66
D3A3S2 2.37 2.88 3.36 2.29 2.12 1.02 2.28 5.55 1.78 0.84 1.51 3.01 4.74 3.31 1.69
D3A3S2 2.43 2.77 3.44 2.29 2.13 1.14 2.34 5.72 1.81 0.94 1.54 3.09 4.98 3.10 1.67
208
Table A-2 (cont’)
Test
D3A3S3 2.31 2.82 3.20 2.22 2.08 1.42 3.55 8.24 4.02 1.37 1.78 3.57 4.00 3.90 1.73
D3A3S3 2.32 2.84 3.31 2.21 1.98 1.40 3.39 7.04 3.44 1.30 1.73 3.66 4.26 3.85 1.79
D3A3S3 2.28 2.82 3.26 2.24 2.01 1.42 3.33 7.38 2.93 1.21 1.78 3.63 4.09 3.73 1.76
D3BASE 2.42 3.09 3.54 2.38 2.14 2.74 5.33 3.81 4.67 3.20 2.17 4.35 3.42 3.63 2.24
D3BASE 2.43 3.12 3.54 2.38 2.19 2.31 4.02 4.64 4.40 3.36 2.09 3.61 3.97 3.66 2.19
D3BASE 2.38 3.12 3.73 2.35 2.17 2.34 3.92 4.86 4.79 3.25 2.05 3.73 3.93 3.58 2.43
209
Table A-3 Velocity Ratio for Cross Section by Test Run
Test Run VRA VRB VRC Test Run VRA VRB VRC
D1A1S1 1.09 2.96 1.18 D2A2S3 1.25 2.99 2.22
D1A1S1 0.98 3.08 1.17 D2A2S3 1.25 2.87 2.13
D1A1S1 1.01 2.47 1.15 D2A2S3 1.23 3.10 2.25
D1A1S2 1.01 2.14 1.64 D2A3S1 1.24 3.45 2.25
D1A1S2 1.04 2.57 1.40 D2A3S1 1.29 3.11 2.06
D1A1S2 1.04 1.88 1.36 D2A3S1 1.29 3.44 2.24
D1A1S3 0.96 1.96 1.48 D2A3S2 1.25 2.77 2.32
D1A1S3 1.00 1.75 1.45 D2A3S2 1.30 3.04 2.36
D1A1S3 1.03 1.81 1.42 D2A3S2 1.29 3.05 2.18
D1A2S1 0.97 1.37 1.33 D2A3S3 1.24 2.94 2.33
D1A2S1 0.98 1.50 1.28 D2A3S3 1.20 2.84 2.33
D1A2S1 1.04 1.47 1.29 D2A3S3 1.23 2.98 2.25
D1A2S2 1.00 1.88 1.37 D2BASE 1.21 1.07 1.43
D1A2S2 0.99 1.57 1.27 D2BASE 1.29 1.02 1.48
D1A2S2 1.04 1.77 1.26 D2BASE 1.26 0.99 1.39
D1A2S3 1.00 2.65 1.38 D3A1S1 1.29 2.07 2.00
D1A2S3 1.00 2.67 1.47 D3A1S1 1.29 1.84 1.84
D1A2S3 1.02 2.62 1.42 D3A1S1 1.32 1.76 1.80
D1A3S1 1.28 1.53 1.27 D3A1S2 1.32 2.73 1.62
D1A3S1 1.07 1.29 1.09 D3A1S2 1.30 2.45 1.67
D1A3S1 1.07 1.37 1.25 D3A1S2 1.32 2.06 1.67
D1A3S2 1.08 1.60 1.46 D3A1S3 1.25 2.38 1.61
D1A3S2 1.11 1.78 1.35 D3A1S3 1.29 2.12 1.70
D1A3S2 1.14 1.69 1.38 D3A1S3 1.29 2.99 1.66
D1A3S3 1.18 2.24 1.52 D3A2S1 1.28 3.57 1.95
D1A3S3 1.14 2.31 1.49 D3A2S1 1.29 3.56 1.89
D1A3S3 1.00 2.19 1.46 D3A2S1 1.26 3.43 1.92
D1BASE 1.01 1.13 1.08 D3A2S2 1.20 2.67 1.84
D1BASE 0.99 1.14 1.19 D3A2S2 1.28 2.85 1.78
D1BASE 1.01 1.09 1.13 D3A2S2 1.25 2.67 1.92
D2A1S1 1.26 1.62 1.96 D3A2S3 1.16 3.59 2.00
D2A1S1 1.27 1.54 1.78 D3A2S3 1.14 3.50 1.97
D2A1S1 1.33 1.53 1.76 D3A2S3 1.21 3.63 1.99
D2A1S2 1.33 3.18 1.71 D3A3S1 1.28 3.71 2.28
D2A1S2 1.24 3.55 1.71 D3A3S1 1.28 4.33 2.16
D2A1S2 1.30 3.50 1.80 D3A3S1 1.30 4.37 2.14
D2A1S3 1.20 2.70 1.67 D3A3S2 1.28 3.80 2.34
D2A1S3 1.26 3.35 1.61 D3A3S2 1.27 3.44 2.30
D2A1S3 1.23 3.06 1.69 D3A3S2 1.24 3.16 2.32
D2A2S1 1.21 3.01 1.94 D3A3S3 1.25 3.78 2.18
D2A2S1 1.21 3.28 2.07 D3A3S3 1.30 3.42 2.23
D2A2S1 1.20 2.80 2.00 D3A3S3 1.29 3.46 2.16
D2A2S2 1.23 2.57 1.99 D3BASE 1.32 1.55 1.72
D2A2S2 1.22 2.69 2.03 D3BASE 1.30 1.54 1.75
D2A2S2 1.30 2.59 2.15 D3BASE 1.35 1.62 1.67
210
Table A-4 Mean Depth (in) at Cross Section for Test Run
Left column describes the test run where D=depth ratio, A=arm angle, and S=arm slope.
Numbers 1, 2, 3, 4, and 5 are the factor level. There are three repetitions of each test run. The
header represents the test point location. A is for cross section A, B is for cross section B and
C is for cross section C.
TEST A B C TEST A B C TEST A B C
D1A1S1 5.49 4.87 3.01 D3A3S1 5.24 3.62 3.51 D5A3S5 5.36 3.88 3.64
D1A1S1 5.49 4.87 3.00 D3A3S1 5.24 3.62 3.50 D5A3S5 5.36 3.88 3.64
D1A1S1 5.49 4.87 3.00 D3A3S1 5.24 3.62 3.50 D5A3S5 5.36 3.88 3.64
D1A1S3 5.86 4.37 3.26 D3A3S3 5.49 3.75 3.51 D5A5S1 4.99 3.50 3.64
D1A1S3 5.86 4.37 3.30 D3A3S3 5.24 3.50 3.50 D5A5S1 4.99 3.50 3.64
D1A1S3 5.86 4.37 3.13 D3A3S3 5.24 3.50 3.51 D5A5S1 4.99 3.63 3.64
D1A1S5 6.49 4.50 3.01 D3A3S5 5.24 3.37 3.26 D5A5S3 5.11 3.75 3.64
D1A1S5 6.49 4.50 3.01 D3A3S5 5.11 3.25 3.26 D5A5S3 5.11 3.75 3.64
D1A1S5 6.49 4.50 3.01 D3A3S5 4.99 3.25 3.26 D5A5S3 5.11 3.63 3.64
D1A3S1 5.49 4.87 3.26 D3A4S3 4.99 3.62 3.51 D5A5S5 5.36 3.63 3.64
D1A3S1 5.49 4.87 3.25 D3A4S3 5.11 3.49 3.51 D5A5S5 5.36 3.63 3.64
D1A3S1 5.49 4.87 3.25 D3A4S3 5.11 3.49 3.51 D5A5S5 5.24 3.63 3.64
D1A3S3 5.49 4.75 3.13 D3A5S1 4.99 3.37 3.13 D5BASE 4.99 3.00 3.51
D1A3S3 5.49 4.87 3.20 D3A5S1 4.99 3.37 3.13 D5BASE 4.99 3.00 3.50
D1A3S3 5.49 4.87 3.13 D3A5S1 4.99 3.37 3.25 D5BASE 4.99 3.00 3.50
D1A3S5 5.99 4.75 3.13 D3A5S3 4.99 3.37 3.38
D1A3S5 5.86 4.75 3.26 D3A5S3 4.99 3.37 3.40
D1A3S5 5.86 4.75 3.26 D3A5S3 4.99 3.37 3.38
D1A5S1 5.49 4.75 3.51 D3A5S5 5.49 3.50 3.38
D1A5S1 5.49 4.75 3.50 D3A5S5 5.49 3.37 3.38
D1A5S1 4.99 4.50 3.00 D3A5S5 5.49 3.50 3.38
D1A5S3 5.24 4.75 3.51 D3BASE 4.99 3.06 3.51
D1A5S3 5.74 4.50 3.90 D3BASE 4.99 3.00 3.75
D1A5S3 5.49 4.75 3.76 D3BASE 4.99 3.00 3.75
D1A5S5 5.74 4.75 3.01 D5A1S1 4.93 4.50 4.01
D1A5S5 5.74 4.87 3.26 D5A1S1 4.93 4.50 4.01
D1A5S5 5.74 4.75 3.26 D5A1S1 4.93 4.50 4.01
D1BASE 5.49 5.00 3.01 D5A1S3 5.24 4.50 4.14
D1BASE 5.24 5.00 3.25 D5A1S3 5.24 4.38 4.14
D1BASE 5.36 5.00 3.13 D5A1S3 5.24 4.38 4.01
D3A1S1 4.86 3.75 3.51 D5A1S5 5.24 3.88 3.89
D3A1S1 4.74 3.75 3.50 D5A1S5 5.24 3.88 3.89
D3A1S1 4.74 3.75 3.50 D5A1S5 5.24 3.88 3.89
D3A1S3 4.74 3.75 3.26 D5A3S1 5.11 4.00 3.64
D3A1S3 4.74 3.75 3.50 D5A3S1 5.11 4.00 3.64
D3A1S3 4.81 3.75 3.51 D5A3S1 4.99 4.00 3.64
D3A1S5 4.99 3.75 3.51 D5A3S3 5.11 4.13 3.51
D3A1S5 4.99 3.75 3.51 D5A3S3 5.11 4.00 3.64
D3A1S5 4.99 3.75 3.51 D5A3S3 5.11 4.00 3.64
211
Appendix B:
Rock Cross Vane Qualitative Assessment Data
212
ID AC01
Date 6/21/2006
F3.P07.
Stream Avon Creek F3.P07.S01.
F3.P07.S12.
Vane # 1 F4.P01.
F4.P01.S17.
F1. 0 F4.P02.
F4.P02.S01.
F2. 0 F4.P07.
F4.P07.S01.
F3. 1 F4.P07.S12.
F6.P09.
F4. 3 F6.P09.S15.
F6.P09.S20.
F5. 0
F6. 5
ID AC02
Date 6/21/2006
Stream Avon Creek

F3.P05.
Vane # 2 F3.P05.S18.
F3.P07.
F1. 0 F3.P07.S01.
F6.P09.
F2. 0
F6.P09.S20.
F3. 1 F6.P06.
F6.P06.S13.
F4. 0
F5. 0
F6. 5
ID AC03
Date 6/21/2006
Stream Avon Creek F3.P05.

F3.P05.S18.
Vane # 3 F3.P07.
F3.P07.S10.
F1. 0 F3.P07.S12.
F4.P07.
F2. 0
F4.P07.S10.
F3. 1 F6.P09.
F6.P09.S15.
F4. 1 F6.P09.S20.
F5. 0
F6. 1
213
ID AC04
Date 6/21/2006
Stream Avon Creek
Vane # 4
F1. 0 F6.P09.
F6.P09.S15.
F2. 0 F6.P09.S20.
F3. 0
F4. 0
F5. 0
F6. 5
ID AC05
Date 6/21/2006
Stream Avon Creek
Vane # 5
F1. 0 F6.P09.
F6.P09.S15.
F2. 0 F6.P09.S20.
F3. 0
F4. 0
F5. 0
F6. 3
ID AT01
Date 6/21/2006
Avon Creek
Stream
Trib
Vane # 1
F1. 0
F3.P05.
F2. 0 F3.P05.S18.
F3. 1
F4. 0
F5. 0
F6. 0
214
ID AT02
Date 6/21/2006
Avon Creek
Stream
Trib
Vane # 2
F1. 0 F6.P09.
F6.P09.S20.
F2. 0
F3. 0
F4. 0
F5. 0
F6.
ID AT03
Date 6/21/2006
Avon Creek
Stream
Trib
Vane # 3 F3.P05.
F3.P05.S18.
F1. 0 F3.P07.
F3.P07.S12.
F2. 0
F6.P09.
F3. 1 F6.P09.S20.
F4. 0
F5. 0
F6. 1
ID BC01
Date 4/10/2006
Beaver
Stream
Creek
Vane # 1
F1. 0 Not a cross vane.
Data not used in
F2. 0 analysis.
F3. 0
F4. 0
F5. 0
F6. 0
215
ID BC02
Date 4/10/2006
F4.P07.
Stream Beaver Creek F4.P07.S01.
F4.P08.
Vane # 2 F4.P08.S18.
F5.P01.
F1. 0 F5.P01.S17.
F5.P02.
F2. 0
F5.P02.S03.
F3. 0 F5.P04.
F5.P04.S21.
F4. 5 F6.P09.
F6.P09.S15.
F5. 5
F6. 3
F1.P03.
ID BC03
F1.P03.S18.
F1.P03.S03.
Date 4/10/2006 F4.P07.
F4.P07.S01.
F4.P08.
Vane # 3 F4.P08.S01
F4.P08.S18.
F1. 1 F4.P08.S03.
F5.P02.
F2. 0 F5.P02.S01.
F5.P02.S03.
F5.P07.
F3. 0 F5.P07.S09.
F6.P09.
F4. 5 F6.P09.S15.
F6.P09.S20.
F5. 5 F6.P06.
F6.P06.S09
F6. 3
ID BC04
F1.P08.
F1.P08.S16.
Date 4/10/2006 F3.P08.
F3.P08.S16.
F4.P01.
Vane # 4 F4.P01.S09.
F4.P07.
F1. 1 F4.P07.S10.
F4.P07.S03.
F2. 0 F4.P07.S21.
F4.P08.
F3. 5 F4.P08.S16.
F4.P08.S21.
F4. 5 F5.P01.
F5.P01.S05.
F5. 1 F5.P01.S17.
F6.P06.
F6.P06.S09.
F6. 1
216
ID BC05
Date 4/10/2006
F1.P08.
F1.P08.S03.
Vane # 5 F3.P07.
F3.P07.S01.
F1. 1 F3.P07.S03.
F3.P07.S21.
F2. 0
F4.P07.
F3. 5 F4.P07.S01.
F4.P07.S03.
F4. 5 F6.P06.
F6.P06.S09.
F5. 0
F6. 1
ID BC06
F1.P03.
F1.P03.S01.
Date 4/10/2006
F1.P08.
F1.P08.S03.
Stream Beaver Creek
F3.P08.
F3.P08.S02.
Vane # 6
F3.P07.
F3.P07.S01.
F1. 1
F3.P07.S03.
F3.P07.S21.
F2. 0
F4.P07.
F4.P07.S01.
F3. 1
F5.P02.
F5.P02.S01.
F4. 5
F6.P09.
F6.P09.S15.
F5. 1 F6.P06.
F6.P06.S09.
F6. 3
ID BC07
Date 4/10/2006
Stream Beaver Creek

F3.P07.
Vane # 7 F3.P07.S01.
F4.P05.
F1. 0 F4.P05.S12.
F4.P07.
F2. 0 F4.P07.S01.
F4.P07.S03.
F3. 3 F4.P08.
F4.P08.S21.
F4. 5
F5. 0
F6. 0
217
ID BC08
Date 4/10/2006
Beaver
Stream
Creek
Vane # 8 F4.P07.
F4.P07.S01.
F1. 0 F4.P08.
F4.P08.S18.
F2. 0
F6.P09.
F3. 0 F6.P09.S15.
F4. 3
F5. 0
F6. 5
ID BC09
Date 4/10/2006
F1.P08.
F3.P07.
Vane # 9 F3.P07.S01.
F4.P07.
F1. 3
F4.P07.S21.
F2. 0 F4.P08.
F4.P08.S18.
F3. 3 F6.P09.
F6.P09.S15.
F4. 5 F6.P06.
F6.P06.S09.
F5. 0
F6. 3
*could be a j-hook.
ID BC10
Date 4/10/2006
Stream Beaver Creek F4.P07.

F4.P07.S21.
Vane # 10 F4.P08.
F4.P08.S18.
F1. 0 F4.P08.S21.
F5.P04.
F2. 0
F5.P04.S21.
F3. 0 F5.P04.S13.
F6.P06.
F4. 3 F6.P06.S13.
F5. 3
F6. 3
218
ID BC11
Date 4/10/2006
Stream Beaver Creek

F4.P07.
Vane # 11 F4.P07.S03.
F4.P07.S21.
F1. 0 F4.P08.
F4.P08.S18.
F2. 0
F4.P08.S21.
F3. 0 F6.P06.
F6.P06.S13.
F4. 3
F5. 0
F6. 1
ID BC12
Date 4/10/2006
Stream Beaver Creek

F3.P05.
Vane # 12 F3.P05.S18.
F3.P08.
F1. 0 F3.P08.S18.
F4.P01.
F2. 0
F4.P01.S12.
F3. 1 F5.P01.
F5.P01.S17.
F4. 3
F5. 3
F6. 0
ID BC13
Date 4/10/2006
Stream Beaver Creek
Vane # 13
F1.P03.
F1. 1 F1.P03.S14.
F4.P08.
F2. 0
F4.P08.S18.
F3. 0
F4. 1
F5. 0
F6. 0
219
ID BC14
Date 4/10/2006
Stream Beaver Creek
Vane # 14
F1. 0 F6.P09.
F6.P09.S04.
F2. 0
F3. 0
F4. 0
F5. 0
F6. 0
ID BC15
Date 4/10/2006
Stream Beaver Creek
Vane # 15
F4.P07.
F1. 0 F4.P07.S01.
F6.P09.
F2. 0
F6.P09.S15.
F3. 0
F4. 1
F5. 0
F6. 1
ID BC16
Date 4/10/2006
Stream Beaver Creek
Vane # 16
F4.P07.
F1. 0 F4.P07.S01.
F6.P09.
F2. 0
F6.P09.S15.
F3. 0
F4. 1
F5. 0
F6. 1
220
ID BC17
Date 4/10/2006
Stream Beaver Creek

F4.P07.
Vane # 17 F4.P07.S21.
F5.P01.
F1. 0 F5.P01.S17.
F5.P04.
F2. 0 F5.P04.S21.
F5.P04.S13.
F3. 0 F6.P06.
F6.P06.S13.
F4. 3
F5. 5
F6. 1
ID BC18
Date 4/10/2006
F3.P07.
F4.P01.
Vane # 18 F4.P01.S17.
F4.P08.
F1. 0 F4.P08.S18.
F4.P08.S21.
F2. 0 F5.P01.
F5.P01.S17.
F3. 1 F5.P04.
F5.P04.S21.
F4. 3 F6.P09.
F6.P09.S15.
F5. 5
F6. 1
ID BC19
F1.P03.
F1.P03.S01.
Date 4/10/2006 F1.P03.S14.
F1.P08.
F3.P07.
Vane # 19 F3.P07.S10.
F4.P01.
F1. 3 F4.P01.S17.
F4.P02.
F4.P02.S01.
F2. 0 F4.P07.
F4.P07.S01.
F3. 1 F5.P01.
F5.P01.S17.
F4. 5 F5.P02.
F5.P02.S01.
F5. 3 F6.P06.
F6.P06.S09.
F6.P06.S13.
F6. 3
221
ID BC20
Date 4/10/2006
Stream Beaver Creek F1.P03.

F1.P03.S05.
Vane # 20 F1.P03.S14.
F4.P01.
F1. 3 F4.P01.S05.
F4.P01.S09.
F2. 0 F4.P07.
F4.P08.
F3. 0 F4.P08.S21.
F5.P01.
F4. 3
F5.P01.S17.
F5. 1
F6. 0
*could be a j-hook with protected toe on opposite bank.
ID GrCr01
Date 8/8/2006
F3.P08.
Stream Greasy Creek F3.P08.S10.
F3.P07.
Vane # 1 F3.P07.S01.
F3.P07.S03.
F1. 0 F3.P07.S12.
F4.P02.
F2. 0 F4.P02.S01.
F4.P02.S03.
F3. 3 F4.P07.
F4.P07.S01.
F4. 3 F4.P07.S03.
F4.P07.S12.
F5. 0
F6. 0
ID GrCr02
Date 8/8/2006
Stream Greasy Creek
Vane # 2
F1. 0 F3.P08.
F2. 0 F3.P08.S02.
F3. 1
F4. 0
F5. 0
F6. 0
222
ID GrCr03
Date 8/8/2006
Stream Greasy Creek

F3.P08.
Vane # 3 F3.P08.S02.
F3.P07.
F1. 0 F3.P07.S01.
F4.P02.
F2. 0 F4.P02.S01.
F4.P02.S03.
F3. 1 F6.P09.
F6.P09.S04.
F4. 1
F5. 0
F6. 5
ID GrCr04
Date 8/8/2006
Stream Greasy Creek
Vane # 4
F1. 0 F6.P09.
F6.P09.S04.
F2. 0 F6.P09.S20.
F3. 0
F4. 0
F5. 0
F6. 5
ID GrCr05
Date 8/8/2006 F3.P05.

F3.P05.S18.
Stream Greasy Creek
F3.P08.
Vane # 5 F3.P08.S02.
F3.P08.S18.
F1. 0 F3.P07.
F3.P07.S03.
F2. 0 F4.P07.
F4.P07.S03.
F3. 1 F5.P02.
F5.P02.S03.
F4. 3
F5.P04.
F5. 1 F5.P04.S13.
F6. 0
223
ID HC01
Date 6/21/2006
Stream Hoppers Creek
Vane # 1 F3.P07.
F3.P07.S10.
F1. 0 F3.P07.S12.
F4.P01.
F2. 0 F4.P01.S17.
F4.P07.
F3. 3 F4.P07.S10.
F4. 3
F5. 0
F6. 0
ID LiBr01
Date 4/15/2006
Little Brasstown
Stream
Hyatt
Vane # 1
F1. 0
F5.P01.
F2. 0 F5.P01.S17.
F3. 0
F4. 0
F5. 1
F6. 0
ID LiBr02
Date 4/15/2006
Little Brasstown
Stream
Hyatt
Vane # 2
F1.P03.
F1. 1 F1.P03.S14.
F6.P06.
F2. 0
F6.P06.S11.
F3. 0
F4. 0
F5. 0
F6. 1
224
ID LiBr03
Date 4/15/2006
Little Brasstown
Stream
Oland
Vane # 3
F1. 0 F4.P08.
F4.P08.S08.
F2. 0 F4.P08.S18.
F3. 0
F4. 1
F5. 0
F6. 0
ID LiBr04
Date 8/9/2006
Little Brasstown
Stream F4.P01.
Folk School
F4.P01.S09.
Vane # 4 F4.P01.S17.
F1. 0 F4.P07.
F4.P07.S12.
F2. 0 F4.P07.S21.
F5.P04.
F3. 0
F5.P04.S13.
F4. 3 F5.P09.
F5.P09.S09.
F5. 1
F6. 0
ID LoBr01
Date 8/8/2006
Stream Long Branch
Vane # 1 F3.P05.
F1. 0 F3.P05.S12.
F3.P07.
F2. 0 F3.P07.S12.
F4.P02.
F3. 1 F4.P02.S01.
F4. 1
F5. 0
F6. 0
225
ID LoBr02
Date 8/8/2006
Stream Long Branch
Vane # 2
F1. 0
F6.P09.
F2. 0
F3. 0
F4. 0
F5. 0
F6. 1
ID LoBr03
Date 8/8/2006
Stream Long Branch
Vane # 3
F1. 0
F6.P09.
F2. 0
F3. 0
F4. 0
F5. 0
F6. 1
ID LoBr04
Date 8/8/2006
Stream Long Branch
Vane # 4
F1. 0
F6.P09.
F2. 0
F3. 0
F4. 0
F5. 0
F6. 5
226
ID LoBr05
Date 8/8/2006
Stream Long Branch
Vane # 5
F3.P05.
F1. 0
F3.P05.S12.
F2. 0 F3.P07.
F3.P07.S12.
F3. 1
F4. 0
F5. 0
F6. 0
ID LoBr06
Date 8/8/2006
Stream Long Branch
Vane # 6
F1. 0
F6.P06.
F2. 0 F6.P06.S13.
F3. 0
F4. 0
F5. 0
F6. 3
ID LoBr07
Date 8/8/2006
Stream Long Branch
Vane # 7
F1. 0
F4.P02.
F2. 0 F4.P02.S03.
F3. 0
F4. 1
F5. 0
F6. 0
227
ID LoBr08
Date 8/8/2006
Stream Long Branch
Vane # 8 F1.P03.
F1.P03.S03.
F1. 3 F1.P03.S01.
F1.P03.S14.
F2. 0 F4.P01.
F3. 0 F4.P01.S09.
F4.P07.
F4. 5
F5. 0
F6. 0
ID MS01
Date 1/0/1900
Stream Mason Stalcup
Vane # 1 F4.P01.
F1. 0 F4.P01.S12.
F4.P07.
F2. 0 F4.P07.S01.
F4.P08.
F3. 0 F4.P08.S21.
F4. 1
F5. 0
F6. 0
ID MS02
Date 4/14/2006
Mason Stalcup
Stream
Pinhook
Vane # 2
F4.P01.
F1. 0 F4.P01.S12.
F4.P08.
F2. 0 F4.P08.S08.
F4.P08.S19.
F3. 0
F4. 1
F5. 0
F6. 0
228
ID MS03
Date 4/15/2006
Vane # 3
F4.P07.
F1. 0
F4.P07.S01.
F2. 0 F4.P08.
F4.P08.S08.
F3. 0
F4. 1
F5. 0
F6. 0
ID MS04
Date 1/0/1900
F3.P05.
Vane # 4 F3.P05.S12.
F3.P07.
F1. 0
F3.P07.S10.
F2. 0 F4.P05.
F4.P05.S12.
F3. 1 F4.P07.
F4.P07.S10.
F4. 1
F5. 0
F6. 0
ID MS05
Date 1/0/1900
Mason Stalcup
Stream F3.P05.
trib
F3.P05.S12.
Vane # 5 F3.P08.
F1. 0 F3.P08.S18.
F4.P01.
F2. 0 F4.P01.S12.
F3. 1 F4.P07.
F4.P07.S01.
F4. 1 F4.P08.
F4.P08.S21.
F5. 0
F6. 0
229
ID RB01
Date 4/4/2006
F3.P05.
Stream Rocky Branch F3.P05.S18.
F3.P07.
Vane # 1 F3.P07.S10.
F1. 0 F4.P01.
F4.P01.S12.
F2. 0 F4.P07.
F4.P07.S01.
F3. 3 F4.P08.
F4.P08.S01
F4. 1 F6.P06.
F5. 0 F6.P06.S13.
F6. 1
ID RB02
Date 4/4/2006
F3.P05.
F3.P07.
Vane # 2 F3.P07.S10.
F4.P05.
F1. 0
F4.P05.S18.
F2. 0 F4.P07.
F4.P07.S21.
F3. 3 F4.P08.
F4.P08.S18.
F4. 1 F6.P06.
F5. 0 F6.P06.S13.
F6. 3
ID RB03
Date 4/4/2006
Stream Rocky Branch
Vane # 3
F1. 0 F3.P07.
F3.P07.S10.
F2. 0 F6.P06.
F3. 1
F4. 0
F5. 0
F6. 5
230
ID RB04
Date 4/4/2006
Stream Rocky Branch
Vane # 4 F3.P07.
F3.P07.S10.
F1. 0 F5.P01.
F5.P01.S17.
F2. 0 F5.P04.
F3. 1 F5.P04.S21.
F5.P04.S13.
F4. 0
F5. 3
F6. 0
ID RB05
Date 4/4/2006
Stream Rocky Branch F1.P08.
F1.P08.S16.
Vane # 5 F4.P07.
F1. 1 F4.P07.S21.
F4.P08.
F2. 0 F4.P08.S21.
F5.P01.
F3. 0 F5.P01.S06.
F5.P04.
F4. 1 F5.P04.S13.
F5. 5
F6. 0
ID RB06
Date 4/4/2006
F3.P07.
Stream Rocky Branch
F3.P07.S10.
Vane # 6 F4.P02.
F4.P08.
F1. 0 F4.P08.S21.
F5.P01.
F2. 0 F5.P01.S17.
F3. 1 F5.P02.
F5.P02.S01.
F4. 1 F5.P04.
F5.P04.S21.
F5. 5
F6. 0
231
ID RB07
F1.P08.
Date 4/4/2006 F1.P08.S21.
F3.P08.
F3.P08.S10.
Vane # 7 F3.P08.S21.
F1. 1 F3.P07.
F3.P07.S10.
F2. 0 F3.P07.S21.
F4.P02.
F3. 1 F4.P02.S01.
F4.P02.S03.
F4. 3 F4.P07.
F5. 1 F4.P07.S21.
F5.P04.
F6. 0 F5.P04.S13.
ID RB08
Date 4/4/2006
Stream Rocky Branch F4.P01.
F4.P01.S12.
Vane # 8 F4.P05.
F1. 0 F4.P05.S18.
F4.P07.
F2. 0 F4.P07.S01.
F4.P07.S21.
F3. 0 F4.P08.
F4.P08.S18.
F4. 5 F4.P08.S21.
F5. 0
F6. 0
ID RB09 F3.P07.
F3.P07.S10.
Date 4/4/2006 F4.P01.
F4.P01.S12.
Stream Rocky Branch
F4.P05.
Vane # 9 F4.P05.S18.
F4.P07.
F1. 0 F4.P07.S10.
F4.P07.S01.
F2. 0 F4.P08.
F4.P08.S18.
F3. 3
F4.P08.S03.
F4. 3 F4.P08.S21.
F5.P01.
F5. 3 F5.P01.S17.
F5.P02.
F6. 0 F5.P02.S01.
232
ID RB10
Date 4/4/2006
Stream Rocky Branch
Vane # 10 F4.P08.
F1. 0 F4.P08.S16.
F4.P08.S21.
F2. 0 F5.P04.
F5.P04.S21.
F3. 0 F5.P04.S13.
F4. 1
F5. 1
F6. 0
ID RB11
Date 4/4/2006
Stream Rocky Branch
Vane # 11
F3.P07.
F1. 0 F3.P07.S10.
F3.P07.S12.
F2. 0 F6.P09.
F3. 1 F6.P09.S15
F4. 1
F5. 3
F6. 3
ID RB12
Date 4/4/2006
Stream Rocky Branch
Vane # 12
F1. 0
F4.P01.
F2. 0 F4.P01.S12.
F3. 0
F4. 1
F5. 0
F6. 0
233
ID RB13
Date 4/4/2006
F3.P05.
F3.P07.
Vane # 13 F3.P07.S10.
F1. 0 F3.P07.S03.
F4.P02.
F2. 0 F4.P02.S03.
F4.P07.
F3. 3 F4.P07.S01.
F4.P07.S03.
F4. 3 F5.P04.
F5. 3 F5.P04.S21.
F6. 0
ID RB14
Date 4/6/2006
Stream Rocky Branch
Vane # 14
F1. 0
F5.P04.
F2. 0 F5.P04.S21.
F3. 0
F4. 0
F5. 1
F6. 0
ID RB15
Date 4/6/2006
Stream Rocky Branch
Vane # 15
F1. 0
F2. 0
F3. 0
F4. 0
F5. 0
F6. 0
234
ID RB16
Date 4/6/2006
Stream Rocky Branch
Vane # 16
F1. 0
F2. 0
F3. 0
F4. 0
F5. 0
F6. 0
ID RB17
Date 4/6/2006
Stream Rocky Branch
Vane # 17
F1. 0
F5.P04.
F2. 0 F5.P04.S21.
F3. 0
F4. 0
F5. 3
F6. 0
ID RB18
Date 4/6/2006
Stream Rocky Branch
Vane # 18 F4.P01.
F1. 0 F4.P01.S12.
F4.P05.
F2. 0 F4.P05.S18.
F4.P08.
F3. 0 F4.P08.S18.
F4. 1
F5. 0
F6. 0
235
ID RB19
Date 4/6/2006
Stream Rocky Branch
Vane # 19
F1. 0
F3.P07.
F2. 0 F3.P07.S10.
F3. 1
F4. 0
F5. 0
F6. 0
ID RB20 F3.P05.
F3.P05.S18.
Date 4/6/2006 F3.P08.
F3.P08.S10.
F3.P07.
Vane # 20 F3.P07.S10.
F4.P01.
F1. 0 F4.P01.S12.
F4.P02.
F4.P02.S01.
F2. 0 F4.P05.
F4.P05.S18.
F3. 5 F4.P07.
F4.P07.S10.
F4. 3 F4.P08.
F4.P08.S18.
F5. 1 F4.P08.S19.
F5.P04.
F6. 0 F5.P04.S13.
ID RB21
Date 4/6/2006
F1.P08.
Stream Rocky Branch
F1.P08.S19.
Vane # 21 F1.P08.S21.
F4.P01.
F1. 1 F4.P01.S12.
F4.P08.
F2. 0 F4.P08.S16.
F3. 0 F4.P08.S19.
F4.P08.S21.
F4. 3 F5.P04.
F5.P04.S21.
F5. 5
F6. 0
236
ID RB22
Date 4/6/2006
F3.P08.
F3.P07.
Vane # 22 F3.P07.S10.
F1. 0 F4.P01.
F4.P01.S12.
F2. 0 F4.P02.
F4.P02.S03.
F3. 3 F4.P08.
F4.P08.S08.
F4. 3 F4.P08.S16.
F5. 0 F4.P08.S21.
F6. 0
ID RB23
Date 4/6/2006
Stream Rocky Branch
Vane # 23 F4.P05.
F4.P05.S18.
F1. 0 F4.P07.
F4.P07.S01.
F2. 0 F4.P08.
F3. 0 F4.P08.S18.
F4.P08.S19.
F4. 5
F5. 0
F6. 0
ID RB24
Date 4/6/2006
Stream Rocky Branch
Vane # 24 F4.P07.
F1. 0 F4.P07.S01.
F5.P01.
F2. 0 F5.P01.S06.
F5.P04.
F3. 0 F5.P04.S13.
F4. 1
F5. 3
F6. 0
237
ID RB25
Date 4/6/2006
F3.P05.
Stream Rocky Branch
F3.P05.S12.
Vane # 25 F3.P08.
F3.P08.S10.
F1. 0 F3.P08.S18.
F3.P07.
F2. 0 F3.P07.S10.
F3. 3 F4.P07.
F4.P07.S10.
F4. 1 F4.P08.
F4.P08.S18.
F5. 0
F6. 0
ID RB26
Date 4/6/2006
Stream Rocky Branch
Vane # 26
F1. 0
F2. 0
F3. 0
F4. 0
F5. 0
F6. 0
ID RiBr01 F3.P05.
F3.P05.S18.
Date 8/9/2006 F3.P05.S12.
Ricks Branch F3.P08.
Stream F3.P08.S02.
(Shepherd)
F3.P08.S16.
Vane # 1
F3.P08.S18.
F1. 0 F3.P07.
F3.P07.S12.
F2. 0 F4.P07.
F4.P07.S01.
F3. 5 F4.P08.
F4.P08.S01
F4. 1
F5.P02.
F5. 5 F5.P02.S01.
F5.P04.
F6. 0 F5.P04.S08.
238
ID RiBr02
Date 8/9/2006
F2.P08.
Stream Ricks Branch F2.P08.S01.
F3.P08.
Vane # 2 F3.P08.S21.
F1. 0 F3.P07.S11.
F5.P01.
F2. 5 F5.P01.S06.
F5.P04.
F3. 5 F5.P04.S08.
F5.P04.S21.
F4. 0 F6.P06.
F5. 5 F6.P06.S11.
F6. 1
ID RiBr03
Date 8/9/2006
Stream Ricks Branch
Vane # 3
F3.P07.
F1. 0 F3.P07.S10.
F3.P07.S12.
F2. 0 F5.P02.
F3. 5 F5.P02.S03.
F4. 0
F5. 1
F6. 0
ID RiBr04
Date 8/9/2006
Stream Ricks Branch
F3.P08.
Vane # 4 F3.P08.S02.
F3.P08.S18.
F1. 0 F3.P07.
F3.P07.S03.
F2. 0 F4.P07.
F3. 3 F4.P07.S03.
F4.P08.
F4. 5 F4.P08.S08.
F5. 0
F6. 0
239
ID RiBr05
Date 8/9/2006
Stream Ricks Branch
Vane # 5 F3.P08.
F3.P08.S02.
F1. 0 F3.P07.
F3.P07.S03.
F2. 0 F3.P07.S12.
F3. 3 F5.P02.
F5.P02.S01.
F4. 1
F5. 5
F6. 0
ID RiBr06
Date 8/9/2006
Stream Ricks Branch
Vane # 6 F3.P08.
F3.P08.S02.
F1. 0 F5.P01.
F5.P01.S06.
F2. 0 F5.P04.
F3. 1 F5.P04.S08.
F5.P04.S13.
F4. 0
F5. 5
F6. 0
ID RiBr07
Date 8/9/2006
Stream Ricks Branch
Vane # 7 F3.P08.
F3.P08.S02.
F1. 0 F5.P01.
F5.P01.S06.
F2. 0 F5.P01.S17.
F3. 1 F5.P04.
F5.P04.S08.
F4. 0
F5. 5
F6. 0
240
ID RiBr08
Date 8/9/2006
F1.P03.
F1.P03.S01.
Vane # 8 F3.P08.
F1. 1 F3.P08.S02.
F3.P07.
F2. 0 F3.P07.S09.
F3.P07.S03.
F3. 5 F4.P01.
F4.P01.S12.
F4. 3 F4.P07.
F5. 0 F4.P07.S03.
F6. 0
ID RiBr09
Date 8/9/2006
Stream Ricks Branch
Vane # 9 F3.P07.
F1. 0 F3.P07.S12.
F5.P01.
F2. 0 F5.P01.S17.
F6.P06.
F3. 3 F6.P06.S13.
F4. 0
F5. 1
F6. 1
ID RiBr10
Date 8/9/2006
F3.P05.
F3.P05.S12.
Vane # 10 F3.P08.
F1. 0 F3.P08.S02.
F3.P08.S18.
F2. 0 F3.P07.
F3.P07.S01.
F3. 5 F3.P07.S12.
F4.P07.
F4. 3 F4.P07.S01.
F5. 0 F4.P07.S03.
F6. 0
241
ID RiBr11
Date 8/9/2006
Stream Ricks Branch
Vane # 11 F3.P08.
F1. 0 F3.P08.S02.
F3.P07.
F2. 0 F3.P07.S03.
F4.P07.
F3. 5 F4.P07.S03.
F4. 5
F5. 0
F6. 0
ID SC01
Date 6/26/2006
Stream Sharpe Creek
Vane # 1
F1. 0
F2. 0
F3. 0
F4. 0
F5. 0
F6. 0
ID SC02
Date 6/26/2006
Stream Sharpe Creek
Vane # 2
F1. 0
F4.P02.
F2. 0 F4.P02.S01.
F3. 0
F4. 1
F5. 0
F6. 0
242
ID SC03
Date 6/26/2006
Stream Sharpe Creek
Vane # 3
F1. 0
F2. 0
F3. 0
F4. 0
F5. 0
F6. 0
ID SC04
Date 6/26/2006
Stream Sharpe Creek
Vane # 4
F1. 0
F6.P06.
F2. 0 F6.P06.S13.
F3. 0
F4. 0
F5. 0
F6. 1
ID SC05
Date 6/26/2006
Stream Sharpe Creek
F3.P08.
Vane # 5 F3.P08.S02.
F1. 0 F3.P08.S10.
F3.P08.S18.
F2. 0 F4.P07.
F4.P07.S03.
F3. 1 F6.P06.
F6.P06.S13.
F4. 1
F5. 0
F6. 1
243
ID SC06
Date 6/26/2006
Stream Sharpe Creek
Vane # 6 F3.P08.
F1. 0 F3.P08.S16.
F4.P01.
F2. 0 F4.P01.S17.
F4.P02.
F3. 1 F4.P02.S01.
F4. 5
F5. 1
F6. 0
ID SC07
Date 6/26/2006
Stream Sharpe Creek
Vane # 7 F4.P02.
F1. 0 F4.P02.S01.
F5.P02.
F2. 0 F5.P02.S01.
F5.P04.
F3. 0 F5.P04.S21.
F4. 1
F5. 1
F6. 0
ID SC08
Date 6/26/2006
Stream Sharpe Creek

Vane # 8
F1. 0
F2. 0
F3. 0
F4. 0
F5. 0
F6. 0
244
ID UL01
Date 6/26/2006
Stream Upper Laurel
Vane # 1
F1. 0
F6.P09.
F2. 0 No picture.
F6.P09.S04.
F3. 0
F4. 0
F5. 0
F6. 3
ID UL02
F1.P03.
Date 6/26/2006 F1.P03.S14.
Stream Upper Laurel F2.P03.
F2.P03.S14.
Vane # 2 F3.P08.
F3.P07.S11.
F1. 5 No picture. F4.P01.
F2. 5 Washed out. F4.P01.S09.
F4.P07.
F3. 5 F6.P09.
F6.P09.S09.
F4. 5 F6.P09.S15.
F5. 0 F6.P06.
F6.P06.S11.
F6. 5
ID UL03
Date 6/26/2006
Stream Upper Laurel
F1.P03.
Vane # 3 F1.P03.S14.
F1. 5 F2.P03.
No picture.
F2. 5 F2.P03.S14.
Washed out.
F6.P06.
F3. 0 F6.P06.S09.
F4. 0 F6.P06.S11.
F5. 0
F6. 5
245
ID UL04
Date 6/26/2006 F1.P03.
Stream Upper Laurel F1.P03.S14.
F2.P03.
Vane # 4
F2.P03.S14.
F1. 5 No picture. F4.P01.
F2. 5 Washed out. F4.P01.S09.
F4.P07.
F3. 0
F6.P06.
F4. 1 F6.P06.S09.
F5. 0 F6.P06.S11.
F6. 3
ID YF01
Date 6/21/2006
Stream Young’s Fork
Vane # 1 F4.P05.
F4.P05.S12.
F1. 0 F4.P07.
F4.P07.S03.
F2. 0 F4.P07.S12.
F3. 0 F6.P09.
F6.P09.S04.
F4. 1
F5. 0
F6. 3
ID YM01
F1.P03.
Date 4/11/2006 F1.P03.S01.
F3.P08.
Stream Yates Mill F3.P08.S10.
F3.P08.S18.
Vane # 1 F3.P07.
F3.P07.S10.
F1. 1 F3.P07.S12.
F2. 0 F4.P01.S12.
F4.P07.
F3. 5 F4.P07.S10.
F4.P07.S12.
F4. 5 F4.P08.
F4.P08.S08.
F5. 0 F4.P08.S18.
F4.P08.S19.
F6. 0
246
ID YM02
Date 4/11/2006
Stream Yates Mill F3.P08.
F3.P08.S10.
Vane # 2 F3.P08.S18.
F3.P07.
F1. 0
F3.P07.S10.
F2. 0 F3.P07.S03.
F4.P02.
F3. 1 F4.P02.S01.
F6.P09.
F4. 1 F6.P09.S20.
F5. 0
F6. 5
ID YM03
Date 4/11/2006
Stream Yates Mill
F3.P08.
Vane # 3 F3.P08.S02.
F3.P08.S10.
F1. 0 F3.P08.S18.
F3.P07.
F2. 0 F3.P07.S10.
F3. 3 F3.P07.S03.
F4.P02.
F4. 1 F4.P02.S01.
F5. 0
F6. 0
ID YM04
Date 4/11/2006 F3.P07.

F3.P07.S10.
F3.P07.S03.
Vane # 4 F4.P01.
F4.P01.S12.
F1. 0 F4.P02.
F4.P02.S01.
F2. 0 F4.P07.
F4.P07.S03.
F3. 5 F4.P08.
F4.P08.S18.
F4. 3 F4.P08.S03.
F6.P06.
F5. 0 F6.P06.S13.
F6. 3
247
ID YM05
Date 4/11/2006
Stream Yates Mill
Vane # 5
F3.P08.
F1. 0
F3.P08.S10.
F2. 0 F3.P08.S18.
F3.P08.S19.
F3. 1
F4. 0
F5. 0
F6. 0
ID YM06
Date 4/11/2006
Stream Yates Mill
F3.P08.
Vane # 6 F3.P08.S18.
F4.P01.
F1. 0
F4.P01.S12.
F2. 0 F4.P02.
F4.P02.S01.
F3. 5 F4.P08.
F4.P08.S19.
F4. 3
F5. 0
F6. 0
ID YM07
Date 4/11/2006
F3.P08.
F3.P08.S10.
Vane # 7 F3.P08.S18.
F3.P07.
F1. 0
F3.P07.S01.
F2. 0 F3.P07.S03.
F4.P07.
F3. 5 F4.P07.S01.
F4.P07.S03.
F4. 1 F6.P09.
F5. 0 F6.P09.S04.
F6. 5
248
ID YM08
F3.P08.
Date 4/11/2006 F3.P08.S10.
F3.P07.
Vane # 8 F3.P07.S10.
F3.P07.S01.
F1. 0 F4.P07.
F4.P07.S10.
F2. 0 F4.P07.S03.
F4.P08.
F3. 3 F4.P08.S01
F4.P08.S19.
F4. 3 F4.P08.S03.
F6.P09.
F5. 0
F6.P09.S20.
F6. 3
ID YM09
Date 4/11/2006
Stream Yates Mill
Vane # 9
F1. 0
F6.P06.
F2. 0 F6.P06.S13.
F3. 0
F4. 0
F5. 0
F6. 1
ID YM10
F1.P03.S18.
Date 4/11/2006 F1.P03.S03.
F3.P08.
F3.P08.S16.
Vane # 10 F3.P08.S10.
F3.P08.S18.
F1. 1 F4.P01.
F2. 0 F4.P01.S12.
F4.P07.
F3. 3 F4.P07.S01.
F4.P07.S12.
F4. 1 F4.P08.
F4.P08.S18.
F5. 0 F6.P06.
F6.P06.S09.
F6. 5
249
ID YM11
Date 4/11/2006
F3.P08.
Vane # 11 F3.P08.S02.
F3.P08.S10.
F1. 1
F3.P08.S18.
F2. 0 F3.P08.S21.
F4.P07.
F3. 5 F4.P07.S01.
F6.P06.
F4. 1 F6.P06.S11.
F5. 0
F6. 1
ID YM12
Date 4/11/2006
Stream Yates Mill
Vane # 12
F1. 0
F2. 0
F3. 0
F4. 0
F5. 0
F6. 0
ID YM13
Date 4/11/2006 F3.P08.
F3.P08.S16.
Stream Yates Mill
F3.P08.S10.
Vane # 13 F3.P08.S18.
F3.P08.S19.
F1. 0 F3.P07.
F3.P07.S01.
F2. 0 F3.P07.S12.
F3. 1 F4.P07.
F4.P07.S01.
F4. 1 F4.P07.S12.
F4.P08.
F5. 0 F4.P08.S19.
F6. 0
250
ID YM14
Date 4/11/2006
Stream Yates Mill
Vane # 14 F3.P07.
F3.P07.S10.
F1. 0
F3.P07.S12.
F2. 0 F6.P09.
F6.P09.S15.
F3. 1 F6.P09.S20.
F4. 0
F5. 3
F6. 1
ID YM15
Date 4/11/2006
Stream Yates Mill
Vane # 15 F3.P07.
F3.P07.S10.
F1. 0
F3.P07.S12.
F2. 0 F6.P09.
F6.P09.S15.
F3. 1 F6.P09.S20.
F4. 0
F5. 3
F6. 1
ID YM16
Date 4/11/2006
F3.P07.
Stream Yates Mill
F3.P07.S01.
Vane # 16 F3.P07.S03.
F4.P07.
F1. 0 F4.P07.S01.
F4.P07.S03.
F2. 0 F5.P02.
F3. 1 F5.P02.S01.
F5.P02.S03.
F4. 5 F6.P09.
F6.P09.S20.
F5. 1
F6. 5
251
ID YM17
Date 4/11/2006
Stream Yates Mill
Vane # 17
F1. 0
F2. 0
F3. 0
F4. 0
F5. 0
F6. 0
ID YM18
Date 4/11/2006
Stream Yates Mill
Vane # 18
F5.P04.
F1. 0 F5.P04.S21.
F5.P04.S13.
F2. 0 F6.P06.
F3. 0 F6.P06.S13.
F4. 0
F5. 1
F6. 3
ID YM19
Date 4/11/2006
Stream Yates Mill
Vane # 19
F1. 0 F4.P07.
F4.P07.S03.
F2. 0 F4.P07.S12.
F3. 0
F4. 1
F5. 0
F6. 0
252

مؤشرات

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

مؤشرات

Uploaded by

Copyright:

Available Formats

ABSTRACT

PUCKETT, PAIGE ROLLINS. The Rock Cross Vane: A Comprehensive Study of an

A dissertation submitted to the Graduate Faculty of

BIOLOGICAL AND AGRICULTURAL ENGINEERING

Raleigh, North Carolina

Dr. Gregory D. Jennings

Dr. Michael D. Boyette

Paige Rollins Puckett was born in Chattanooga, Tennessee on December 4, 1980.

I would also like to recognize:

List of Tables.......................................................................................................... vii

List of Figures .......................................................................................................... x

History of the use of hydraulic structures related to the

Figure 1.1 Profile View of Cross Vane Arm

Source: Rosgen 2001

Source: Rosgen 2001

The fish weir

Source: “Fishing Weir” 2007

Source: “Fishing Weir” 2007”

Cultural stories can be a reflection of the deep-rooted conflicts about human

Lutins (1992) Fish Weir Town Creek Indian Mound None

Culture & fish weir Catawba River northwest none

Dujiangyan Irrigation Scheme of ancient China

Source: “River Dee, Wales” 2007

Evolution of the Rock Cross Vane

1980’s Rock Vortex Weirs and Rootwads

1988 The Bendway Weir

Figure 1.8 Bendway Weir Theory

Source: Adapted from USACE 2006

1. Excessive maintenance dredging costs via deposition at the toe of the

4. Damaging high-flow current patterns via increased uniformity of surface

5. Insufficient navigation channel widths in the crossing downstream of the bend

1987- 1989 Iowa Vanes and Submerged Concrete Vanes

Source: “How to Control Streambank Erosion” 2006

1990’s Rock Cross Vane

Source: Stormwater Manager’s Resource Center 2007

Figure 1.11 Plan Views of Three Types of Long Crested Weirs

Source: Williams et all 1993

Figure 1.12 Formation of a Waterfall

Source: Mann 2006

“Bendway Weirs.” Coastal and Hydraulics Laboratory - Engineer Research and

Gilmore, H. W. 1953 . “Hunting Habits of the Early Nevada Pauites.” American

“Fishing Weir.” Available: http://en.wikipedia.org/wiki/Fish-weir. Accessed 19 March

“How to Control Streambank Erosion.” Iowa Department of Natural Resources.

Lutins, A. H. 1992. “Prehistoric Fish Weirs in Eastern North America.” Unpublished

National Register of Historic Places (NRHP) 2005. Available:

Odgaard, A. J., and Spoljaric, A. 1986. “Sediment control by submerged vanes.” J.

“River Dee, Wales.” Available at: http://en.wikipedia.org/wiki/River_Dee%2C_Wales.

“Rockingham County River Country” NC Northstar. Available at:

Rosgen, D. L. 2001. “The Cross-Vane, W-Weir and J-Hook Vane Structures...Their

Simpkins, D. L. and Petherick , G. L. 1986. “Second Phase Investigations of Late

Stormwater Manager’s Resource Center. Available at:

“The Mississippi River and Tributaries Project.” USACE. Available at:

Tollefson, K. D. and Abbott, M. L. 1993. “From fish weir to waterfall. (Snoqualmie

Mann, J. C. “Formation of a Waterfall.” Wikipedia. Available at:

Effects of Rock Cross Vane Geometry on Velocity

Geometric Variables of the Rock Cross Vane

The Current Problem with Rock Cross Vane Design Recommendations

Hypothesis and Study Goals

There were two main goals of this study:

2. To apply the knowledge gained about the effects of geometry with

Effects of Geometric Change on Velocity Distribution

Potential Adverse Effects of the Rock Cross Vane on Flow

Figure 2.2 Diagram of a Step in a Channel