methods for inventory and development program environmental …€¦ · • review inventory and...
TRANSCRIPT
i
United StatesDepartment ofAgriculture
Forest Service
Technology &DevelopmentProgram
7700—Transportation Systems2500—Watershed and Air ManagementDecember 19989877 1809—SDTDC
FOREST SERVICE
DE P A RTMENT OF AGRICULT U RE
Methods forInventory and
Environmental RiskAssessment of
Road DrainageCrossings
ii
iii
Information contained in this document has been developed for the guidance ofemployees of the Forest Service, USDA, its contractors, and cooperating Federaland State agencies. The Department of Agriculture assumes no responsibility forthe interpretation or use of this information by other than its own employees. Theuse of trade, firm, or corporation names is for the information and convenience ofthe reader. Such use does not constitute an official evaluation, conclusion,recommendation, endorsement, or approval of any product or service to theexclusion of others that may be suitable.
The US Department of Agriculture (USDA) prohibits discrimination in all itsprograms and activities on the basis of race, color, national origin, gender, religion,age, disability, political beliefs, sexual orientation, and marital or family status.(Not all prohibited bases apply to all programs.) Persons with disabilities whorequire alternative means for communication of program information (Braille,large print, audiotape, etc.) should contact USDA's TARGET Center at 202-720-2600 (voice and TDD).
To file a complaint of discrimination, write USDA, Director, Office of Civil Rights,Room 326-W, Whitten Building, 14th and Independence Avenue, SW,Washington, DC 20250-9410 or call 202-720-5964 (voice or TDD). USDA is anequal opportunity provider and employer.
Sam A. Flanagan, Geologist
Michael J. Furniss, Hydrologist
Tyler S. Ledwith, Hydrologist
Stan Thiesen, Geologist
Michael Love, Engineer
Kemset Moore, Hydrologist
Jill Ory, Hydrologist
USDA Forest ServiceSix Rivers National ForestPacific Southwest Region
San Dimas Technology and Development CenterSan Dimas, California
December 1998
Methods for Inventoryand Environmental RiskAssessment of RoadDrainage Crossings
iv
v
ACKNOWLEDGMENTS
An early draft of this document benefited from reviews by Walt Megahan. Greg Bundros and Mary AnnMadej of Redwood National Park offered valuable comments and critique. Jan Werren and Jim Lugosi ofthe Six Rivers National Forest and Americorps Watershed Stewards program, respectively, providedG.I.S. support for the North Fork Eel assessment. The California Department of Forestry and FireProtection (CDF), through Pete Cafferata, provided funding for the early stages of this project. GordonGrant and Beverley Wemple provided many ideas and invaluable encouragement throughout the project.Finally, many thanks to Bill Trush for his extensive comments and constructive criticism through thedevelopment of this document.
vi
vii
CONTENTS
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
THE ENVIRONMENTAL RISK OF ROAD-STREAM CROSSINGSAND THE NEED FOR INVENTORY AND ASSESSMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
REVIEW OF EXISTING ROAD-STREAM CROSSING ASSESSMENT TECHNIQUES . . . . . . . . . . . . . . .
SUGGESTIONS FOR INVENTORY AND ASSESSMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DATA ANALYSIS AND INTERPRETATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
HYDROLOGIC CONNECTIVITY—ASSESSING CHRONIC EROSION POTENTIAL . . . . . . . . . . . . . . .
LITERATURE CITED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
APPENDIX A—AN EXAMPLE OF ROAD ENVIRONMENTAL RISK ASSESSMENT FORUSDA FOREST SERVICE LANDS IN THE UPPER NORTH FORK EEL RIVER WATERSHED,SIX RIVERS NATIONAL FOREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
APPENDIX B —SAMPLE DATA SHEET FOR ROAD DRAINAGE INVENTORY. . . . . . . . . . . . . . . . . . .
1
1
5
12
18
27
37
41
45
viii
1
INTRODUCTION
Road-stream crossings and ditch-relief culverts arecommonly sites of ongoing or potential erosion.Erosion from failures of these structures can be asource of significant impacts to aquatic and riparianresources far removed from the initial failure site.The inventory and assessment of culvertinstallations are necessary for locating sites withpotential impacts to aquatic and riparianenvironments for possible treatment. Withrestricted budgets for road maintenance and repair,locating those sites that have the greatestlikelihood of causing adverse impacts is necessaryto prioritize the expenditure of funds to reduce oreliminate risks.
This guidebook discusses procedures forassessing the erosional hazards and risks toaquatic and riparian ecosystems of road-streamcrossings, ditch-relief culverts, and other road-drainage features. Hazard assessment is theestimation of the potential physical consequences(e.g., volume of material eroded) for one or moresites. Incorporation of the potential impacts tovalued resources, or endpoints, is environmentalrisk assessment. This approach is not intended tobe a rigorous or inflexible procedure, but rather tosuggest appropriate data and assessmenttechniques that can be adapted to a wide range ofsettings.
The purposes of this guidebook are to
• Review inventory and analysis methods incurrent or recent use in wildland settings
• Suggest various inventory procedures for theassessment of existing culvert installations andother road-drainage features, both unfailedand failed
• Provide analytical methods for predicting thepotential hazards and environmental risks ofroad-drainage failure
• Provide background information on road-stream crossing performance.
The assessment methods discussed here areuseful for a screening approach where data on alarge number of drainage features are collected,
typically at the scale of a watershed less than 500km2. Culverts are the most common structureemployed for wildland road-stream crossings andditch drainage. Such culverts are the focus of thisdocument. However, the procedures discussed areadaptable to other road-drainage structures aswell.
The following road crossing terminology is used:
Road-stream crossing is where a road crosses anatural drainage channel or unchanneled swale.
Stream crossing is used synonymously with road-stream crossing.
Cross drain is a ditch-relief culvert or otherstructure such as a grade dip designed to captureand remove surface water from the traveled wayor other road surfaces.
Crossing is used here to cover both streamcrossings and cross drains.
Hydrologically connected road is a segment of roadthat is connected to the natural channel networkvia surface flowpaths.
THE ENVIRONMENTAL RISK OF ROAD-STREAM CROSSINGS AND THE NEEDFOR INVENTORY AND ASSESSMENT
Road-stream crossings are a common featurethroughout wildland roads and have enormouspotential erosional consequences. For federallymanaged lands in the Pacific Northwest, the ForestEcosystem Management Assessment Team(FEMAT) (1993) estimated that 250,000 road-stream crossings exist, but systematic analysis ofcumulative environmental risks are rare. Most ofthese structures have the potential to function asearthen dams, with a small hole or culvert at thebase. Such configurations are rare in naturalchannels. In the absence of maintenance andreplacement, all these structures will eventually failas they plug or the culvert invert deteriorates.Financial resources for maintaining or upgradingthe existing network are limited. Therefore, it is
2
necessary that those sites that pose the greatestrisk to aquatic and riparian resources be locatedand treated to reduce hazard and environmentalrisk.
Crossings can cause both chronic sedimentationimpacts during typical water years and catastrophiceffects when floods trigger crossing failure (figure1). When water overtops the road fill, it may divertdown the road or ditch and onto hillslopesunaccustomed to concentrated overland flow andproduce erosional consequences far removed fromthe crossing. Erosion from diverted streamcrossings in the 197-km2 lower Redwood Creekbasin in northwestern California accounted for 90percent of the total gully erosion (Weaver et al.1995). Best et al. (1995) found that road-streamcrossings accounted for 80 percent of all road-related fluvial erosion in a 10.8-km2 tributary basinto lower Redwood Creek. Erosion from failed orimproperly designed road-drainage structures
accounted for 31 percent of the total sedimentinputs (Best et al. 1995). Identifying those siteswith the greatest potential of erosionalconsequences (e.g., potential to divert stream flow)can direct restoration efforts.
Abandoned roads represent an unknown, butpotentially high, erosional hazard. Failures oftengo unseen in the absence of inspection ormaintenance. Where stream diversions haveoccurred, erosion is likely to continue for decadesas new stream channels are incised. Determiningthe extent of abandoned roads in an area isnecessary for assessing the entire road-relatederosional hazards.
Previous assessments of existing road-streamcrossings have examined the hydraulic capacityof the culverts (Piehl et al. 1988 and Pyles et al.1989). These results show that culverts wereinstalled without consistent design standards, and
Figure 1—Road-stream crossings can alter channel form and processes. They are sites of both chronic and catastrophic erosion.
R9801005
3
that design flow capacities were often belowstandards established by the Oregon ForestPractice Rules (Piehl et al. 1988). In a recentsurvey of road-stream crossings in RedwoodCreek, northwest California, 60 percent of theculverts could not pass a 50-year peak flow(Redwood National Park, unpublished data).
Nonstream crossing drainage structures are alsoof concern. These include ditch-relief culverts,waterbars, and rolling dips. An expandeddiscussion on rolling dips is provided by thecompanion document in this series, “DiversionPotential at Road-Stream Crossings” (Furniss etal. 1997). Detailed design considerations for dipsare provided by Hafterson (1973). Not only dothese structures present similar hazards andenvironmental risks as those discussed for road-stream crossings, they are also capable ofextending the natural drainage network if notproperly configured (Wemple 1994). Extension ofthe drainage network occurs when ditch flow androad surface runoff are conveyed to a streamchannel. Where roads are hydrologicallyconnected to the drainage network, road-producedsediment and runoff are delivered directly to thechannel network. Hydrologic connectivity ofteninvolves extensive gullies linking roads andchannels (Wemple 1994).
Assessing the hazards and environmental risks ofcrossings involves substantial inventory costs.Roads are not a direct reflection of the underlyinglandscape, and, as such, remote sensingtechniques cannot provide necessary informationon existing hazards. Simply locating the installedsystem using remote sensing techniques provideslittle information on potential erosionalconsequences. Several techniques are presentedfor inventorying roads and crossings at variouslevels of effort.
What are the Environmental Risks of Road-Drainage Features?
It is useful to think about the environmental risk ofroad-stream crossings, cross drains, and otherdrainage structures (e.g., waterbars and rollingdips) in four parts (figure 2). The four parts of theenvironmental risk assessment model used in thisguide are
• Inputs are the materials presented to thecrossing. These are water, organic debris,sediment, and fish. (This document does notdiscuss fish/amphibian passage.)
Figure 2—Road-stream crossing and cross drain environmental risk can be expressed as a four part model consisting of inputs—those materials delivered to the culvert; capacity—the ability of the structure to maintain the natural transport regime of the delivered
inputs; physical consequences—the erosional and/or depositional consequences occurring when capacity is exceeded; andendpoints—potentially affected aquatic and riparian resources, human uses, and other values.
1. Inputs 2. Capacity 3. PhysicalConsequences
4. Endpoints
DomesticWater
Supplies
Cold-WaterRefugia
SensitiveAquatic Species
(e.g., fish)
Woody Debris
Sediment
Streamflow
Fish
InputPassage
InputAccumulation
NoConsequence
PhysicalConsequence
WaterQualityImpacts
R9801001
4
• Capacity is the ability of the culvert to passthe inputs.
• Consequences are the physical effects ofcapacity exceedance, often expressedas a volume of material eroded (or deposited).
• Endpoints are the valued resources. Someendpoints may be:
- Fish or other species of concern
- Domestic water supply
- Aquatic species refugia value (e.g.,cool waters in a warm water basin,access to spawning areas).
The hazard of a crossing failure is a combinationof the inputs and the structure’s capacity toaccommodate them, which defines the probabilityof exceedence, and the potential physical
consequences of exceedence. Capacity must beconsidered separately from the potential erosionalconsequences. In the absence of maintenance orreplacement, all crossings will eventually fail.Therefore, it is imperative that potential erosionalconsequences not be ignored simply becausecapacity has been judged “sufficient.” Alternately,a site with a very limited capacity may possessrelatively little erosional hazard. Treatments toincrease capacity at such a site may be impracticaland ineffective in reducing the cumulativeenvironmental risk of crossings (figure 3).
Environmental risk of a crossing is the combinationof existing hazard and endpoints. Environmentalrisk assessment identifies the relative likelihoodof a site adversely affecting one or more endpoints.
Figure 3—The physical consequences of this road-stream crossing failure are low when compared to sites with much larger fills orwhere stream diversion occurs.
R9801006
5
REVIEW OF EXISTING ROAD-STREAM CROSSING ASSESSMENT TECHNIQUES
Eight road-stream crossing assessment techniques are reviewed here. These were collected as part of aroad-stream crossing questionnaire mailed to USDA Forest Service and USDI Bureau of Land Managementengineers and hydrologists throughout the U.S. In the descriptions below, minor modifications were madeto standardize the terminology with that used in this report. Techniques are summarized in table 1.
Table 1—Summary of existing drainage crossing assessment techniques reviewed in text.
Physicalconsequences
Endpoints Diversionpotential
CommentsExpertiserequired(H-M-L)
Diversion potential notexplicit but can beincorporated intophysical consequences
Drainages less than 0.4km2 are consideredlow effect
Determining failurepotential requiresengineering/geologyskills
H
H
H
M
M
L
M
L
X
X
X
X
X
X
X
X
X
X
X
X
Technique
Umpqua N.F.Shockey(1996)
Umpqua N.F.Hanek (1996)
Siskiyou N.F.Weinhold(1996)
Payette N.F.Inglis et al.(1995)
Mt. Baker-SnoqualmieN.F. (1997)
Huron-Manistee N.F.Stuber (1996)
Kennard(1994)
StanislausN.F. (1996)
Inputs Capacity
X
X
X
X
X
X
X
X
X
Intermittent streams areexcluded
A road segmentassessment withcrossings as onecomponent
Primarily assesses roadsurface erosion
Significant featurespotentially lost in matrix
A road segmentapproach with emphasison road location andconfiguration
6
Broda and Shockey (1996) of the Umpqua National Forest use a “risk rating table” (table 2) to rank road-stream crossings on the Umpqua National Forest.
Using this approach, each crossing is assigned a score from 1 to 5. Factors considered in this approach areshown in table 2.
Hazards• Hydrology/hydraulics—flows, culvert capacity
• Plugging potential—wood and debris (soil and rocks)
• Slope stability—channel slopes, road fill.
Physical consequences/endpoints• Sediment delivery (amount)
• Resource damage—trees, fish, habitat (cost)
• Capital investment losses (cost).
Table 2—Umpqua NF risk rating table.
Hazard
medium
2
3
4
high
3
4
5
low
1
2
3
Physical
consequences/
endpoints
low
medium
high
Hanek (1996), also of the Umpqua National Forest, uses a similar approach (table 3). Factors considered inthis approach in order of decreasing priority are:
Effects• High
- Large drainages (1.2 km2+) with diversion potential
- Medium drainages (0.4–1.2 km2) with diversion potential.
• Medium
- Large drainages without diversion potential but with larger fills (i.e., over 1.5 m from inlet invert tograde)
- Medium drainages without diversion potential but with larger fills.
• Low
- Everything else (i.e., large and medium drainages without diversion and small fills and all small drainages(<0.4 km2).
7
Hazards• Hydraulic capacity of the culvert and associated fill prior to overtopping or diversion (using 100-year
design storm)
• Culvert plug potential
• Existing culvert condition
• Embankment stability under no-pool, full-pool, and half-pool conditions
• Diversion potential.
Effects
medium
2
3
4
high
3
4
5
low
1
2
3Hazards
low
medium
high
Table 3—Umpqua NF hazards versus effects approach.
An approach used by Weinhold (1996) of the Siskiyou National Forest involving potential eroded fill volumeis shown in table 4.
Here, high failure potential is represented by an undersized culvert, evidence of past plugging by debris, andslope and channel having potential to generate debris flows.
High Failure Potential
Volume > 150 m3
High Hazard
Moderate Failure Potential
Volume > 150 m3
High Hazard
Low Failure Potential
Volume > 150 m3
Moderate Hazard
High Failure Potential
Volume 40 to 150 m3
High Hazard
Moderate Failure Potential
Volume 40 to 150 m3
Moderate Hazard
Low Failure Potential
Volume 40 to 150 m3
Low Hazard
High Failure Potential
Volume < 40 m3
Low Hazard
Moderate Failure Potential
Volume < 40 m3
Low Hazard
Low Failure Potential
Volume < 40 m3
Low Hazard
Table 4—Siskiyou NF hazard rating table.
Stuber et al. (1994), working in the Huron-Manistee National Forest in Michigan, assigns each crossing aseverity ranking based on a point system (table 5).
Scores over 30 points are placed in the severe category. Scores under 15 points are considered minor.
8
Table 5—Huron-Manistee NF point system.
Factors contributing toseverity
Road surface
Length of approaches(total)
Slope of approaches
Width of road, shoulders,and ditches
Extent of erosion
Embankment slope
Stream depth
Stream current
Vegetative cover ofshoulders and ditches
Paved 0Gravel 3Sand and Gravel 6Sand 9
0–10 m 110–300 m 3301–600 m 5> 600 m 7
0 % 01–5 % 36–10 % 6> 10 % 9
< 5 m 05–7 m 1> 7 m 2
Minor 1Moderate 3Extreme 5
Bridges 0> 2:1 slope 11.5–2:1 3Vertical or 1:1 slope 5
0–1 m 1> 1 m 2
Slow 1Moderate 2Fast 3
Heavy 1Partial 3None 5
PointsCondition
The Mt. Baker-Snoqualmie National Forest (1997) uses a road segment approach where crossings are onecomponent of the method (table 6). An effects of failure score (K) is a combination of sediment delivery andvalued resources. It functions as a multiplier to the failure potential score.
9
Percent Sideslopes< 20% 21 – 40% > 40%
Distance tostream (m)
Inta
StreamPerb
Stream
Bench or terracebetween road &stream, wetlands,infrastructure, orother valuableresource.
No bench orterrace presentbetween road &stream, wetlands,infrastructure, orother valuableresource.
Inta
StreamPerb
StreamInta
StreamPerb
Stream
N/A
< 1515 – 150150 – 300300 – 450> 450
1
11111
1
21111
1
22211
1
33211
1
33221
1
44332
a. Intermittent stream
b. Perennial stream
Determining K involves the use of a short dichotomous key (table 7). The score is based on the proximity ofthe road to streams, wetlands, infrastructure, or other valuable natural resources.
A = Snow Zone: Location of road segment and contributing upslope area. Washington State rain-on-snow zones. Rain on Snow =2; Rain or Snow Dominated = 1; Lowland = 0; Highland = 0.
B = Geology and soil stability: Percent of area occupied by road on unstable soils, highly eroded glacial, alluvial fan, or recessionaloutwash deposits and highly fractured and unstable base geology. Under 10 percent = 0; 10–30 percent = 2;31–50 percent = 3; Over 50 percent = 5.
C = History of road associated failures from sources which have not been corrected: None = 0; Some = 1; Repeated = 2.
D = Major stream crossings: Number of large (>900 mm) or deep (>1 m over top of pipe inlet) culverts. None = 0; One = 1; Morethan one = 2
E = Number of stream channel crossings / 150 m of road: 0–1 = 0; 2 =1; 3+ = 2.
F = Method of construction: Generally, if constructed before 1970 assume sidecast excavation, if constructed after 1970 assume layerplacement excavation. Full Bench = 0; Layer Placement = 1; Sidecast = 2.
G = Average sideslope where at road: Related to both potential and consequence, but used here to indicate potential for failure. Under40 percent = 0; 40–60 percent = 2; Over 60 percent = 3.
H = Vegetative cover: Percent of area above the road segment (basically the contributing area) having a stand of over 35 years. Over70 percent = 0; 50–70 percent = 1; 20–49 percent = 2; Under 20 percent = 3.
I = Road stacking: Road(s) upslope from this road? No road segments above = 0; the road is at a mid-slope location, or road segmentabove is on ridgetop = 1; One road segment above = 2; Two or more segments above = 3.
Potential for Failure (A-I)A B C D E F G H I
* Note that effects of failure is a multliplier
(SUM A-I) (Effects of Failure) Risk Rating
J (x) K* = (J * K)
Table 6—Mt. Baker-Snoqualmie NF road segment approach.
Table 7—K values for Mt. Baker-Snoqualmie NF road segment approach.
10
Inglis et al. (1995) of the Payette National Forest ranks the condition of stream crossings as either high,medium, or low. Crossings are rated high priority if three or more of the following criteria are met:
• The stream at the crossing is fish bearing;
• Road use is heavy;
• There is a chance of road loss at the crossing; or
• The culvert at the crossing is failing.
Medium priority crossings have to meet at least two of the following criteria:
• The stream at the crossing is nonfish bearing;
• Road use is low;
• There is little to no chance of road loss at the stream crossing, and
• The culvert at the crossing is not failing.
All other crossings needing work are listed as low priority.
The following matrix (table 8) is part of a larger road assessment package developed by Kennard (1994) for theWeyerhauser Corporation. A “weight of evidence” approach shown below is taken to assess the hazard fromcombinations of indicators.
Initiation hazard
PondingRust line
Hydrologic
factors
Landuse andlandscapefactors
Indicators
Culvert sizeCulvert blockagePotential blockage
Channel gradient (low)
Upstream fill heightFillslope surface gradient
Channel widthEntrenchmentChannel gradient (high)
Maximum fill heightFillslope surface gradient
Confinement
Low Medium High(0.5) < (HW/D) (0.5) < (HW/D) ≤ (1.0) (1.0) < (HW/D)(1/3) < (R/D) (1/3) ≤ (R/D) ≤ (1/2) (1/2) < (R/D)
(1) < (D/BMP)0%
clean out > 30 m
< 2 or ≥ 20 degrees
≤ 2 meters< 25 degrees
40 meters < (CW)(VD) < 2 (bfd)< 20 degrees
< 2 meter< 25 degrees
(VW/CW) > 2
(0.7) ≤ (D/BMP) ≤ (1)> 0% to < 20%
clean out 15 to 30 m
2 to < 20 degrees
2–5 meters25–35 degrees
40 ≤ (CW) ≤ 15(VD) ≥ 3 (bfd)
20 to ≤ 30 degrees
2–5 meter25–35 degrees
2 ≥ (VW/CW) > 1
(D/BMP) < (0.7)> 20%
clean out ≤ 15 m
*
> 5 meters> 35 degrees
(CW) > 15 meters*
> 30 degrees
> 5 meters>35 degrees
(VW/CW) = 1
The following are abbreviations used in the weight of evidence approach.
HW/D = Headwater depth (HW) to culvert diameter (D) ratio
R/D = Culvert rust line height (R) to culvert diameter (D) ratio
D/BMP = the ratio of the culvert diameter (D) to the estimated needed diameter (BMP). In this case, the estimated needed diameter isbased on the 50-year flood.
CW = channel width
VD = valley depth
bfd = bankfull discharge
VW/CW = the ratio of valley width (VW) to channel width (CW).
* Not documented in the approach.
Table 8—Weyerhauser Corporation weight of evidence approach.
11
The road rating system used by the Stanislaus National Forest (1996) (table 9) addresses individual roads.Eight factors are scored and totaled.
1. Location
2. Alignment
3. Design
4. Maintenance
5. Soils
6. Topography
7. Hydrology
8. Vegetation
Total Score: <8 = high hazard
8–16 = warrants professional review
>16 = OK.
0 = road within 30 m of base of slope
1 = bottom 1/3 of slope
2 = middle 1/3 of slope
3 = top of slope.
0 = 20 percent + grades
1 = grades 15–20 percent
2 = grades 10–15 percent
3 = grades < 10 percent.
0 = diversion potential exists
1 = insloped, including berms
2 = outsloped
3 = paved.
1 = level 1
2 = level 2
3 = level 3+
drop rating one point for plugged culverts or blocked ditches.
1 = granitic
2 = volcanic
3 = meta-sed.
1 = slope 40 percent+
2 = 30–40 percent slope
3 = slope <30 percent.
1 = rain on snow 900–1,500 m
2 = rain
3 = snow 1,500 m+.
0 = in burn area, below clear cut, or lavacap
1 = barren
2 = grasses / brush
3 = timber canopy.
ScoreFactor
Table 9—Stanislaus NF road rating system.
12
SUGGESTIONS FOR INVENTORY ANDASSESSMENTThe road-drainage network can be inventoried atvarious intensity levels depending on objectivesof the inventory and time and monetary constraints.A baseline inventory will, at a minimum, providethe locations of the installed drainage system. Atthe other end of the spectrum is a completecrossing environmental risk assessment thataddresses all the components displayed infigure 2. Four increasingly intensive levels ofinventory and assessment are presented here.These suggested inventory techniques aresynthesized from the review of existing techniquespresented in the previous section and other studiesexamining the performance of road-streamcrossings.
• Consequences inventory —This approach isdesigned to locate and document the installedsystem and identify the occurrence of diversionpotential over the area of inventory. This is thequickest inventory technique. Becauseremediation of diversion potential is ofteninexpensive and straightforward, it is meantto identify sites where large erosionalconsequences can be easily minimized.
• Connectivity/cross drain inventory —Thisinventory procedure determines the amountof road hydrologically connected to the naturalchannel network. It is meant to 1) identify areasof chronic erosion where sediment is beingdelivered directly to stream channels and 2)identify opportunities to reduce sedimentimpacts to aquatic ecosystems by“disconnecting” roads from streams.
• Hazard assessment —This approachaddresses the likelihood and potentialerosional volume of crossing failure based ona more extensive data collection effort.
• Environmental risk assessment —Buildingon the results produced from a hazardassessment, resources of concern (endpoints)are incorporated to locate sites with thegreatest probability of impacting thoseresources.
A summary of these data fields is presented intable 10 for the four levels of inventory. For crossdrains, a smaller data set is suggested (table 10).Note that these data sets can be tailored to suitspecific needs and satisfy time and financialconstraints.
Materials RequiredAll that is required for a consequences inventoryis a tape for measuring culvert diameter and ameans of locating the site (e.g., topographic mapor aerial photo).
For a more intensive hazard or environmental riskassessment, a clinometer or hand level and stadiarod is required for obtaining fill slopes and culvertslopes. A range-finder is useful for obtainingpotential diversion distances or for very large fills.
The drainage area is most easily determined if thesite is located on a topographic map. Locating thesite on aerial photos should also be done tofacilitate potential geographic information system(GIS) applications. Site location can also beaccomplished with a global positioning system(GPS), and this approach is discussed on page 18.
13
Tabl
e 10
—S
umm
ary
of s
ugge
sted
dat
a fie
lds
for
vario
us le
vels
of r
oad-
stre
am c
ross
ing
and
cros
s dr
ain
inve
ntor
y an
d as
sess
men
t.
Des
crip
tion/
exam
ple
entr
ies
Con
sequ
ence
sin
vent
ory
Env
ironm
enta
lris
k as
sess
men
tH
azar
das
sess
men
t
X X X X X X X
Str
eam
cro
ssin
gcr
oss
drai
not
her
drai
nage
feat
ure
Roa
d i.d
. num
ber
odom
eter
rea
ding
GP
S c
oord
inat
es
pere
nnia
lin
term
itten
t
The
use
r sh
ould
det
erm
ine
the
varia
bles
nee
ded
toca
lcul
ate
desi
gn fl
ows
for
the
area
. Typ
ical
ly, m
ean
annu
al p
reci
pita
tion
is u
sed
in r
egio
nal r
egre
ssio
neq
uatio
ns.
Not
all
site
s w
ill h
ave
a de
finab
le d
rain
age
area
.
Mea
sure
the
wid
th fr
om b
otto
m o
f ban
k to
bot
tom
of
bank
. Thr
ee m
easu
rem
ents
can
be
take
n to
get
an
aver
age
wid
th. T
he s
mal
lest
sw
ales
will
ofte
n ha
ve n
ode
finab
le c
hann
el.
Take
n ab
ove
the
influ
ence
of t
he c
ulve
rt in
let.
unch
anne
led
swal
ein
boar
d di
tch
X X X
X X X X X X X
Fea
ture
Site
num
ber
Cro
ssin
g Ty
pe
Cro
ssin
g in
puts
Cha
nnel
type
Mea
n an
nual
prec
ipita
tion
Dra
inag
e ar
ea
Cha
nnel
wid
th
Cha
nnel
slo
pe
Con
nect
ivity
/cro
ss d
rain
inve
ntor
y
X X
14
Tabl
e 10
—C
ontin
ued.
Fea
ture
Ups
lope
roa
ds
Con
trib
utin
g di
tch
leng
th(s
)
Ditc
h sl
ope
Cul
vert
cap
acity
Cul
vert
type
Des
crip
tion/
exam
ple
entr
ies
Con
sequ
ence
sin
vent
ory
Env
ironm
enta
lris
k as
sess
men
tH
azar
das
sess
men
t
Ent
ranc
e ty
pe
Cul
vert
diam
eter
Per
cent
dent
/cru
sh
X X X X X X
Thi
s is
use
ful f
or e
stim
atin
g ch
anne
l net
wor
k ex
tens
ion
(Wem
ple
1994
) in
the
inve
ntor
y ar
ea. C
ross
dra
ins
shou
ld h
ave
only
one
con
trib
utin
g di
tch.
Not
e th
athy
drol
ogic
con
nect
ivity
at c
ross
dra
ins
is a
ddre
ssed
inth
e co
nseq
uenc
es s
ectio
n be
low
.
Fai
lure
of u
pslo
pe r
oads
can
con
trib
ute
to fa
ilure
at
dow
nstr
eam
site
s.
corr
ugat
ed m
etal
conc
rete
pip
eco
ncre
te b
ox
proj
ectin
gm
itere
dflu
sh in
hea
dwal
l
Thi
s w
ill r
equi
re tw
o m
easu
rem
ents
for
box
culv
erts
and
pipe
arc
hes.
Per
cent
of i
nlet
cro
ss s
ectio
nal a
rea
redu
ced.
Con
nect
ivity
/cro
ss d
rain
inve
ntor
y
X X X
X X
beve
led
inle
tsi
de ta
pere
d in
let
drop
inle
t
pipe
arc
hbo
ttom
less
pip
e ar
chco
rrug
ated
pla
stic
Cou
pled
with
ditc
h le
ngth
, ditc
h sl
ope
prov
ides
an
estim
ate
of th
e tr
ansp
ort c
apac
ity o
f the
ditc
h.
X X X X
X X X X XX
15
Tabl
e 10
—C
ontin
ued.
Fea
ture
Cul
vert
slo
pe
Cul
vert
ske
wan
gle
Out
let d
rop
Pot
entia
l ero
sion
alco
nseq
uenc
es
Flo
od-p
rone
cha
nnel
wid
th (
Wc)
Des
crip
tion/
exam
ple
entr
ies
Con
sequ
ence
sin
vent
ory
Env
ironm
enta
lris
k as
sess
men
tH
azar
das
sess
men
t
Ups
trea
m a
nddo
wns
trea
m f
illsl
opes
(S
u an
d S
d)
Ups
trea
m a
nddo
wns
trea
m fi
llsl
ope
leng
ths
(Lu
and
L d)
X X X X X X
Ang
le a
t whi
ch c
ulve
rt is
orie
nted
to c
hann
el.
Not
e do
wns
pout
s, e
tc.
Rec
ord
the
wid
th o
f fill
mat
eria
l at t
he b
ase
of th
e fil
lco
rres
pond
ing
to th
e ex
pect
ed p
eak
flow
wid
th.
Ref
er to
figu
re 1
0.
Ref
er to
figu
re 1
0.
Ref
er to
figu
re 1
0.
Con
nect
ivity
/cro
ss d
rain
inve
ntor
y
X X
Thi
s is
for
site
s w
ith fi
sh p
assa
ge c
once
rns.
X X X X
X X X X
Out
let p
ool d
epth
Thi
s is
for
site
s w
ith fi
sh p
assa
ge c
once
rns.
Roa
d w
idth
(W
r)R
efer
to fi
gure
10.
Fill
wid
th (
Wr)
Thi
s is
an
aver
age
betw
een
the
upst
ream
and
dow
n-st
ream
edg
es o
f fill
(ta
ken
from
the
cent
erlin
e of
the
road
).R
efer
to fi
gure
10.
X
XX
XX
XX
X
16
Fea
ture
Div
ersi
ondi
stan
ce
Rec
eivi
ng fe
atur
e
Roa
d sl
ope
thro
ugh
cros
sing
Geo
logy
Des
crip
tion/
exam
ple
entr
ies
Con
sequ
ence
sin
vent
ory
Haz
ard
asse
ssm
ent
Thi
s is
the
poin
t whe
re th
e di
vert
ed w
ater
leav
es th
ero
ad s
urfa
ce o
r di
tch.
Exa
mpl
es a
re: a
djac
ent c
ross
ing
(rec
ord
site
num
ber)
and
hill
slop
e (n
ote
if w
ater
flow
son
to u
nsta
ble
slop
e).
The
dis
tanc
e th
e di
vert
ed fl
ows
will
trav
el b
efor
een
terin
g a
chan
nel.
Cro
ssin
gs th
at w
ould
not
div
ert
wat
er a
nd w
ould
onl
y ov
erto
p th
e fil
l are
giv
en a
val
ueof
zer
o. O
n lo
w g
radi
ent r
oads
, the
flow
pat
h(s)
may
be
diffi
cult
to d
isce
rn.
Thi
s fie
ld m
ay b
e co
mpl
eted
usi
ng e
xist
ing
map
s.W
here
pos
sibl
e, g
eolo
gy s
houl
d be
str
atifi
ed b
yer
osio
n ha
zard
.
Con
nect
ivity
/cro
ss d
rain
inve
ntor
y
Whe
re d
iver
sion
pot
entia
l exi
sts,
this
is a
use
ful
mea
sure
men
t for
dip
des
ign.
Hyd
rolo
gica
llyco
nnec
ted?
Det
erm
ine
whe
ther
the
cros
s dr
ain
or o
ther
non
stre
amcr
ossi
ng fe
atur
e de
liver
s m
ater
ial t
o a
chan
nel b
y ei
ther
a gu
lly o
r pl
ume.
XX
XX
X
Leng
th o
f cha
nnel
belo
w o
utle
t (fo
rcr
oss
drai
ns o
nly)
Thi
s is
use
ful f
or d
eter
min
ing
chan
nel n
etw
ork
exte
nsio
n du
e to
roa
ds (
Wem
ple
1994
). If
the
cros
sdr
ain
has
crea
ted
a ch
anne
l tha
t is
hydr
olog
ical
ly c
onne
cted
to th
e st
ream
net
wor
k,re
cord
the
leng
th. T
he a
vera
ge c
ross
sec
tiona
l are
am
ay a
lso
be o
btai
ned
to c
alcu
late
an
erod
ed v
olum
e.
XXXX
XX
XX
XX
XX
Tabl
e 10
—C
ontin
ued.
Env
ironm
enta
l ris
kas
sess
men
t
17
Fea
ture
Pot
entia
llyaf
fect
eden
dpoi
nts
Fis
h at
site
(or
othe
r aq
uatic
or r
ipar
ian
spec
ies
ofco
ncer
n)
Fis
h in
bas
in(o
r ot
her
aqua
ticor
rip
aria
n sp
ecie
s of
conc
ern)
Pla
nnin
g ar
ea
Des
crip
tion/
exam
ple
entr
ies
Con
sequ
ence
sin
vent
ory
Env
ironm
enta
lris
k as
sess
men
tH
azar
das
sess
men
t
Whe
re a
sen
sitiv
e fis
h sp
ecie
s oc
curs
at a
site
, a fi
shpa
ssag
e as
sess
men
t sho
uld
be c
ondu
cted
.
NO
TE
: End
poin
ts w
ill v
ary.
Inte
rdis
cipl
inar
y in
put i
sre
quire
d to
iden
tify
the
endp
oint
s of
con
cern
. The
follo
win
g ar
e su
gges
tions
.
If m
anag
emen
t or
rest
orat
ion
effo
rts
are
conc
entr
ated
in a
n ar
ea, a
ddin
g ex
tra
wei
ght t
o si
tes
with
in th
ese
area
s w
ould
hel
p to
gui
de fu
rthe
r ef
fort
s.
Con
nect
ivity
/cro
ss d
rain
inve
ntor
y
If a
part
icul
ar s
ubdr
aina
ge w
ithin
the
asse
ssm
ent a
rea
is ju
dged
to h
ave
rela
tivel
y hi
gher
aqu
atic
eco
syst
emva
lues
, tho
se c
ross
ings
are
not
ed.
Dom
estic
wat
ersu
pply
If do
mes
tic w
ater
sup
ply
exis
ts, t
hose
cro
ssin
gs th
atco
uld
impa
ir th
e su
pply
are
not
ed.
X
Ref
uge
area
If ar
ea h
as r
efug
e va
lue
for
aqua
tic a
nd/o
r rip
aria
nde
pend
ent s
peci
es (
e.g.
, col
d w
ater
), th
ose
cros
sing
are
note
d.XXXX
Tabl
e 10
—C
ontin
ued.
18
Time Required
The time required for the various levels of inventoryand analysis will vary depending on the access,frequency of drainage structures, and method ofdata collection (i.e., use of a global positioningsystem, which is discussed in the next section).Estimates presented in table 11 assume vehicularaccess, an automated data logger, and a GPS.
Using a Global Positioning System forCrossing Inventory
The advantages of using a GPS include:
• Accurate location of drainage structures onunmapped or poorly mapped roads
• Site information directly input to a data loggerwithout having to transfer the information later
• GPS points and associated data can gostraight into a GIS.
The disadvantages include:
• Cost of equipment
• Less than complete reliability of equipment
• Potential for operator error
• Subject to satellite availability which can beaffected by timing, canopy, and topography
• Need for specialized training of personnel infield use, data reduction, and analysis.
A crossing inventory can be conducted with orwithout the use of a GPS. The crossings must beaccurately located so that the true drainage areacan be calculated. This can be aided by using GPS
although it is still worthwhile to have the locationplotted on a topographic map and an air photo asbackup and confirmation. If a GPS is not available,the topographic map and air photo should besufficient to locate the crossing. Unfortunatelysome roads are not shown on topographic mapsor are only approximately located. Without a GPS,this situation is probably best addressed by plottingthe location on an air photo and transferring thatpoint onto an orthophoto or a digital orthoquad.
DATA ANALYSIS AND INTERPRETATION
Using a Geographic Information System inCrossing Assessment
The advantages of using a GIS for a crossingassessment include:
• The ability to combine crossing informationwith other coverages for analysis and display
• Ease of updating the information as upgrades,decommissions, and failures occur
• Calculation of the contributing basin area forthe crossing by using scanned contour linesto draw the area on-screen
• Sharing of information, which can beenhanced by the transfer of electronic data.
The primary disadvantage is that GIS experienceis required to manipulate the data.
Once the crossing locations (points), drainageareas (polygons), and associated information(attributes) are in a GIS system, they are much
Inventory type Time per crossing(minutes)
Analysis timerequired (days)
Production rate(miles of road/day)
Connectivityinventory
7 1510
Consequencesinventory
5 2012
Hazard assessment 15 256
Environmental risk assessment 15 266
Table 11—Examples of time requirements for various inventories. Production rate assumes five crossings per mile. Analysis timeassumes data for up to several hundred crossings have been collected.
19
easier to compare spatially. The crossing pointscan have additional attributes added from existingpolygon coverages such as: bedrock geology,geomorphology, hillslope gradient, slope position,soils, vegetation, and precipitation. Theseattributes may be useful in predicting thecharacteristics of crossings that have not yet beeninventoried. Crossings can be checked for spatialaccuracy on-screen using digital orthoquads or bycomparing the crossing locations to road-streamor road-crenulation (the declivity where a channelmay occur as expressed by contour lines)intersections. The crossing locations and attributessuch as the hazard rating can be plotted incombination with other coverages like roads,crenulation, and slope position to show where thecrossings are located in relation to these features.This can be useful for spatially analyzing theinformation. For instance, a plot may indicate thatmost of the higher hazard crossings are located inlower slope positions on certain geologic types ina specific portion of a watershed. On the otherhand, a higher slope position on a different geologictype may have very few crossings, with those beinglow hazard.
Calculating Hydraulic Capacity
The hydraulic capacity of a culvert is the designflow it can accommodate at a specified headwaterdepth (the depth of water at the inlet with respectto the base of the culvert inlet). Capacity can bedetermined from nomographs presented inNormann et al. (1985) for a given headwater depth.For dented inlets, the diameter should be adjustedaccordingly. Piehl et al. (1988a) used an equationto approximate the nomograph for inlet controlled,circular, and corrugated metal culverts.
With this equation, data can be taken from theinventory, transferred into a computer spreadsheet,and culvert capacities rapidly computed. Ifhydrologic data are collected for each site, flowsfor various recurrence intervals can be used toconstruct a flood frequency curve. This can beused to express culvert hydraulic capacity as anexceedence probability or a recurrence interval (T)(figures 4 and 5). This method is not applicable to
cross drains or small drainages where the drainagearea cannot be accurately delineated. For relativelylarge culverts in small drainages, calculation andextrapolation may produce unreasonably largerecurrence intervals (or, improbably smallexceedence probabilities). For convention,hydraulic capacity has a maximum value of T =250 years (p = 0.004).
Caution must be exercised when interpreting theresults. Discharges will vary depending on themethod used to estimate peak flows, and the errorfor individual design flow estimates can be large.However, if considered as a relative ranking for allthe road-stream crossings in an assessment area,the results can suggest possible high priority sitesbased on probability of exceedence. Becausehydraulic exceedence may not be the principalmechanism of crossing failure (figure 5), hydrauliccapacity assessment should be assumed torepresent a minimum screening level for hazardassessment. If the culvert cannot pass the designpeak flow, it is likely that associated debris andsediment cannot be passed either.
Increasing culvert hydraulic capacity typicallyrequires either increasing pipe size or addingculverts or end treatments such as side taperedinlets (refer to AISI 1994 for a listing of endtreatments).
Woody Debris Capacity
Plugging of culverts by organic debris is a commonfailure mechanism. Debris lodged at the culvertinlet reduces hydraulic capacity and promotesfurther plugging by organic debris and sediment.Sediment accumulation is often deemed the causeof failure. However, when excavated, one or morepieces of wood are often discovered to be theinitiating mechanism. Furthermore, plugging maynot depend on the transport of large debris. Thelength of pieces initiating plugging can be smalllimbs and twigs, readily transported by frequentlyoccurring storms (Flanagan in review). Piecesinitiating plugging are often not much longer thanthe culvert diameter (figure 6).
20
Flood Recurrence Interval (years)
Dis
char
ge
(cfs
)
100.00
10.00
1.00
1 10 100 1000
149 yr31 yr
51 cfs
31 cfs
Design flow to overtop culvert inlet Design flow to overtop road fill
Figure 5. Culvert hydraulic capacity can be expressed as a recurrence interval (T). In this example, two design discharges arecalculated for a 900-mm culvert. The discharge at HW/D = 1 is assigned a recurrence interval of 31 years (exceedence probability =
0.032). The discharge necessary to overtop the road (in this case, the fill height above the inlet invert is 1.5 m) is assigned arecurrence interval of 149 years (exceedence probability = 0.0067). This example flood frequency curve was generated using a
regional flood estimator for northwest California (Waananen and Crippen 1978).
R9801591
T,
Hydrologic Data
e.g., mean annual precipitation(data required will vary depending on region
and method used)
FloodEstimator
100-year DesignDischarge
50-year DesignDischarge
25-year DesignDischarge
10-year DesignDischarge
Flood FrequencyCurve
Culvert DesignCapacity
Culvert HydraulicCapacity for GivenHeadwater Depth
determined fromnomographs (e.g.,
Normann 1985) or inletcontrolled equation of
Piehl et al. (1988)
Figure 4. Determining the design storm capacity of existing culvert installations requires the use of a flood estimator and thehydraulic capacity of the culvert adjusted for any denting or crushing of the inlet. Using an equation presented by Piehl et al.(1988), this procedure is easily automated in a spreadsheet or similar application to assess a large number of culverts. Refer
also to figure 5.
R9801002
21
Figure 6—Woody debris lodged at culvert inlets is often only slightly longer than the culvert diameter. Note that wood length isexpressed as a ratio to culvert diameter (n=50).
Wood length / culvert diameter
# oc
curr
ing
14
12
10
8
6
4
2
0
0.5
0.75 1
1.25 1.
5
1.75 2
2.25 2.
5
2.75 3
3.25 3.
5
3.75 4
4.5 5
5.5 6
6.5
R9801590
Stream channel width influences the sizedistribution of transported woody debris (e.g.,Lienkaemper and Swanson 1987, Nakamura andSwanson 1994, Braudrick et al. 1997). In low-orderchannels of northwest California, 99 percent oftransported wood greater than 300 mm long wasless than the channel width (Flanagan in review).These findings suggest that culverts sized equalto the channel width will pass a significant portionof potentially pluggable wood. However, theremaining 1 percent of the pieces remain a hazard.Thus, wood plugging hazard can be reduced, butnot eliminated. The woody debris capacity of acrossing can be assessed by taking the ratio ofthe culvert diameter to the channel width (w*).Crossings with low values of w* are more prone todebris plugging. Using the northwest Californiacoast region as an example, sizing culverts equalto the channel width will, in most cases, satisfy a100-year design peak flow (figure 7). However, onwider channels (e.g., >2 m), the cost of employingthis strategy can be prohibitive.
The configuration of the inlet basin will alsoinfluence wood plugging. Inlet basin design shouldstrive to maintain the preexisting channel crosssection, planform, and stream gradient. Channelwidening upstream of the inlet is typicallyundesirable (figure 8). During ponded conditions(HW/D ≥ 1), debris in transport accumulates in the
eddies formed by the widened channel. Piecerotation in the eddies promotes a perpendicularalignment to the culvert inlet. Furthermore, whenthe inlet is fully submerged, wood accumulates inthe pond. When the inlet is reexposed, it is oftenpresented an enormous, often interlocking, raft ofdebris (figure 9).
Channel approach angle, or culvert skew, alsoinfluences debris lodgment (figure 8). Where thechannel enters the culvert at an angle, debrislodgment is increased (Weaver and Hagans 1994).During runoff events, wood in transport cannotrotate parallel to the culvert and pass through.Cross drains are susceptible to this because theyoften possess high approach angles (Garland 1983and Piehl et al. 1988b).
Sediment Capacity
Three types of sediment inputs are discussed here:fluvial transport, sediment “slugs,” and debris flows(a viscous, flowing mass of water, sediment, andwoody debris).
In general, culverts are efficient conveyors ofsediment because of their narrow and relativelysmooth, uniform cross section. Fluviallytransported sediments generally present little
22
100,000
10,000
1000
100
10
1
Pea
k F
low
Rec
urre
nce
Inte
rval
at H
W/D
≥ 1
(ye
ars)
0 50 100 150 200 250 300 350
Drainage Area (Acres)
100-year Event
Probable Maximum Flood
Figure 7—Hydraulic capacity (expressed as a recurrence interval) of 22 culverts with diameters equal to the channel width. Such astrategy to facilitate woody debris passage must be weighed with the costs of installing a large culvert on a wide channel. Costs areof less concern on smaller channels. Inlet modifications to promote passage can reduce the need for large culvert diameters. Data
are from northwestern California.
R9801589
hazard to stream crossing installations. Burial ofthe inlet by fluvially transported sediment is oftenthe result of woody debris lodged at the inlet. Thus,sediment deposition is often a consequence, nota cause, of the failure. In designing and assessingculverts, thought must be given to the maximumparticle sizes potentially mobilized during peakflows and ensuring sufficient diameter and slopeto pass the load. In order to pass sediment, culvertsshould be set at a grade related to the streamchannel (Weaver and Hagans 1994). An indexvalue for sediment plugging hazard is the ratio ofculvert slope to channel slope (s*). This assumesrelatively flat culverts on steep channels are moreprone to sediment accumulations than steeperculverts.
Rapid, more catastrophic inputs of sediment(“sediment slugs”) are typically responsible for
sediment-caused failure. Small failures from over-steepened cutslopes near the inlet may bury theentrances of cross drains and small streamcrossings. Small culverts with relatively small inletbasins adjacent to steep cutslopes are most proneto this. However, predicting where such failuresare likely to occur is difficult. If evidence suggeststhat failure of adjacent slopes is likely, the siteshould be noted for further inspection. During thedesign phase, inlets should be located away fromunstable cut slopes. Enlarging the inlet basin tocapture debris is an option. But, it should beimplemented where no other treatment is possiblebecause enlarged basins may alter streamhydraulics and promote woody debris andsediment deposition.
Assessment of debris flow hazard is difficult.Morphometric characteristics of debris flows and
23
Increasing plugging potential
R9801003
Figure 8—Inlet basin plan view. Inlet basins that maintain the natural channel configuration promote debris transport and passagethrough the culvert. Where the flow is allowed to spread laterally, debris can accumulate and increase the chance of plugging.
Furthermore, debris rotation is promoted in the turbulent eddies of the widening flow. Similarly, where the channel abruptly changesdirection, wood lodgment is enhanced. This is a common configuration for cross drains.
24
initiation areas are discussed by Costa and Jarret(1981) and Benda (1990). Debris flow hazard canbe estimated from aerial photographs, digitalterrain data, and evidence at the site (e.g., Chatwinet al. 1994). Debris flows interact with road-streamcrossing fill prisms either by removing the fill orimpounding against the fill. Fill volumes should bereduced to minimize replacement or excavationcosts. Vented fords can be an effective solutionfor handling debris flows by minimizing theintervention in natural stream processes that theroad fill presents. Another alternative is “hardening”the fill to minimize the chance of fill washout.
Erosional Consequences
When crossing capacity is exceeded, water, debris,and sediment accumulate in the inlet basin. Ifcapacity exceedence is of sufficient magnitude,overtopping of the fill will occur with associatederosional consequences. Estimating the potential
erosional consequences is relativelystraightforward. The path the overflowing watertakes during capacity exceedence will affect themagnitude of consequences. Erosion from flowsthat overtop the fill and reenter the channel nearthe outlet is constrained by the amount of road fillmaterial spanning the channel. Calculating fillvolume is discussed below. Where the road slopesaway from the crossing in at least one direction,overtopping flows can be diverted out of thechannel and away from the site via the road orditch (figure 10). Recognizing diversion potentialis important; large volumes of material may beeroded as flows enlarge the inboard ditch,overwhelm adjacent crossings, enlarge receivingchannels and/or flow onto hillslopes unaccustomedto concentrated overland flows. For a morecomplete discussion of diversion potential refer tothe companion document in this series, “DiversionPotential at Road-Stream Crossings” (Furniss etal. 1997).
Figure 9—In areas where woody debris is entrained in streamflow, crossings that pond water will pose a greater chance of pluggingthan crossings that do not pond water. Wood accumulates in the ponded area and plugging is likely when flows drop and expose the
inlet to the mass of debris.
R9801008
25
Estimating Fill VolumeFill volume calculations are required for hazardassessments and environmental riskassessments. The procedure described here forcalculating fill volume is to be used only to estimatethe quantity of potentially erodible material or theamount of material to be excavated for treatment.Crossing failure is assumed to remove the entirefill prism, and an approximation of this value is usedfor assessing erosional consequences. Themethod presented here should not be used forcontract specifications where more detailedsurveys are required. This volume is intended tobe an estimate to help with comparisons amongsites. This method does not include the potentiallyerodible aggraded stream reach upstream ofcrossings set above grade or the volumesassociated with offsite impacts should the crossingfail. Further, the single measurement of fill width(Wf) along the road centerline will often
overestimate the upstream portion of the fill (VC)
and underestimate the downstream portion (Vd).
A more accurate volume can be calculated if thefill width is taken at both the inboard and outboardedge of the road. The average of these twomeasurements is used for calculating the volumeunder the road surface (V
r).
The following measurements are taken to calculatefill volume (see figure 11):
Wf = Width of the fill along the road centerline and
perpendicular to the culvert axis
Wc = Width of flood prone channel
Lu = Fill slope length from inboard edge of road to
inlet invert
Figure 10—Diversion of streamflows at stream crossings can have large erosional consequences far removed from the initial failuresite. Here water diverted along the inboard ditch for several hundred meters initiated additional crossing failures along the way.
Photo courtesy of the Siskiyou National Forest.
R9801592
26
Wc
Wr
Su
Bu
HuHd
Sd
Ld
Bd
Lu
Wf
Figure 11—Crossing fill measurements—Solid lines are measured values, dashed lines are calculated. Note that Ld often extendsbelow the culvert outlet. The method presented here is intended for estimating fill volume. Some underestimation will occur on the
downstream side while the inlet portion will be overestimated.
R9801004
27
Su = Slope of L
u (in degrees). If field data are
collected in percent, conversion to degrees isaccomplished using the arc tangent function.
The subscript “d” in the following equations refersto the above measurements taken on thedownstream portion of the road fill.
(1) Upstream prism volume (Vu):
Vu= 0.25(W
f + W
c)( L
u cos S
u)( L
u sin S
u)
(2) Downstream prism volume (Vd):
Vd= 0.25(W
f + W
c)( L
d cos S
d)( L
d sin S
d)
(3) Volume under road surface (Vr):
Vr= (Hu
2+ Hd) (Wf
2+ Wc) W
r
where:
Hu= L
u sin S
u
and
Hd= L
d sin S
d
(4) Total fill volume (V):
V= Vu+V
d+V
r
HYDROLOGIC CONNECTIVITY—ASSESSINGCHRONIC EROSION POTENTIAL
Cross drains represent a special case for erosionalconsequences. In addition to being subjected tothe inputs discussed above, cross drainsconcentrate water draining from the road surfaceand intercept groundwater at the cut slope. Thiswater is often conveyed to unchanneled hillslopeswhere a gully may form. This newly formed channelmay connect to the natural channel network,extend it, and contribute road-derived sedimentand additional surface runoff to the aquaticecosystem. During the inventory process, crossdrains with gullies or sediment plumes extendingto a natural channel should be documented forpossible treatment.
Length of road segments connected to the naturalchannel network are summed, and the proportionof the road network hydrologically connected isdetermined. Individual roads with high connectivity
can be targeted for “disconnecting” by theappropriate treatment (e.g., grade dips oroutsloping).
Channel network extension because of roads canalso be calculated if the ditch length and road-caused gully/plume extent below the road isknown. The proportional extension of the drainagenetwork may be used to locate areas where thenatural hydrologic regime is most affected.
Assessment Procedure
Assessments may proceed once the inventory andthe calculations required for the level of inventoryare complete. Techniques for each level ofinventory are described below.
Connectivity/cross drain inventoryResults of a connectivity inventory will be a list ofroad segments hydrologically connected to thechannel network. This level of inventory does notprioritize segments. Treatments to “disconnect”individual road segments will be based onavailability of funds and transportation needs.
Consequences inventoryThe primary product of a consequences inventorywill be a list of sites with diversion potential.Treatments to eliminate diversion potential will bebased on availability of funds and transportationneeds. Further prioritization of sites can be basedon potential length of diversion and receivingfeature.
Hazard assessmentFor hazard assessment, sites are assigned ahazard score based on several hazard elements.If data have been compiled into a spreadsheet ordatabase structure, sorting on each of the elementsor using an if-then command, can be performedto easily assign scores. The scoring systemsuggested below is meant to provide a flexiblemeans of evaluating crossings. The user can adjustscores and add other factors to suit the needs ofthe inventory.
28
Hazard Score = Inputs + Capacity + Consequences
The inputs and capacity scoring elements are:
• T—an expression of hydraulic capacity (lowestT values are the greatest hazard)
- T < 10 years: 3
- T 10–100 years: 2
- T > 100 years: 1
- Site does not have definable drainage areaor is an unchanneled swale: 0
• w*—an expression of woody debris capacity—culvert diameter/width of channel (lowestvalues are greatest hazard)
- w* < 0.5: 3
- w* 0.5–1.0: 2
- w* > 1.0: 1
- No definable channel: 0
• Skew angle—additional factor for debriscapacity
- Site with definable channel and skew angle > 45 degrees: 1
- Site with no definable channel or skew angle≤ 45 degrees: 0
• s* (slope of culvert/slope of channel)—may beused to assess the ability of the culvert totransport sediment (lowest values are highesthazard)
- s* < 0.3: 2
- s* 0.3–0.6: 1
- s* > 0.6: 0
• R—if upslope roads are present
- Upslope roads are present: 1
- If no roads are present upslope: 0.
For consequence scores, crossings with streamdiversion potential take priority over nondiversionpotential culverts, because stream diversiontypically results in much greater eroded volumes.
• Diversion distance (highest values are greatesthazard)
- Distance > 300 m: 5
- Distance 100–300 m: 3
- Distance 50–100 m: 2
- Distance < 50 m: 1
- No diversion potential: 0
• Diversion onto unstable receiving feature(predetermined categories from existing data,e.g., earthflow, over-steepened sidecast)
- Diversion onto unstable landform (may besame as unstable geology describedbelow): 3
- Diversion onto sidecast fill: 2
- Diversion onto valley floor: 1
• Fill volume (largest volumes are greatesthazard)
- Volume > 1000 m3: 4
- Volume 500–1000 m3: 3
- Volume 100–500 m3: 2
- Volume < 100 m3: 1
• Unstable geology (predetermined categoriesfrom existing geologic data)
- Site is located in unstable terrain: 3
- Site is located in moderately unstable terrain(where applicable): 2
- Site is located in relatively stable terrain: 0
• Hydrologically connected
- Site is a road-stream crossing or a cross drain that is hydrologically connected to the drainage network: 1
- Site is a cross drain that is not connected to the drainage network: 0.
Sites are prioritized based on their relative scores.However, significant features can become lost inthe tallying of a hazard score. For example, a sitewith a high diversion potential score may have verylow scores for the other elements. In this instance,the user may wish to automatically rank the siteas high priority for treatment.
Environmental risk assessmentFor environmental risk assessment, the score fora crossing uses predetermined endpoints inconjunction with the previously described hazardscores. Environmental risk is expressed as:
29
Environmental risk = Hazard Endpoints
The endpoints value is the sum of individualendpoints values for a site.
A site is assigned a value of one if it affects anendpoint. Some endpoints may be:
• Fish (or other species of special concern) atsite—this should also generate a separate fishpassage assessment.
• Fish (or other species of special concern)within the basin being inventoried.
• Sub-watershed containing crossing has refugevalue (e.g., cold water tributary).
• Domestic water supply in basin.
• Special management area—special use areasmay warrant assigning extra weight to thosesites.
The number of endpoints is not fixed and shouldbe discussed and agreed on throughinterdisciplinary analysis. At least one endpointshould be identified for meaningful results.
Example of a preponderance table from anenvironmental risk assessment conducted on 383road-stream crossings within a 295-km2 area ispresented in table 12. An observed tendency isfor individual roads to have a majority of sitesscoring similarly. This is favorable for upgradingor decommissioning programs where it is mostpractical to treat a single road segment or specificarea.
LimitationsFinal scores require careful interpretation.Significant features (e.g., fish at site) can be lost ifno significant hazard elements exist at the site.Although the approach is designed to makeenvironmental risk sensitive to the endpoints, theuser must assess the reality of the environmentalrisk scenario generated by this approach. Thisapproach is meant to suggest sites for furtherinspection.
Users may wish to adjust the scoring system toreflect local settings. The scoring system aspresented is intended to assign higher hazardvalues to crossings with diversion potential.Elements not playing a large role in adding to thehazard of a site based on observations inmountainous regions (e.g., s*) add less to theoverall hazard score. Again, it is up to the user toassess the magnitude of the scores and determinewhether they represent the relative magnitude ofhazard elements in the area of inventory.
Evaluating consequences by roadsegmentAn alternative to the site-scale results presentedpreviously is generating results for individual roadsegments. This approach is often desirable fortransportation planning and decommissioningefforts where the unit of consideration is often theroad segment. Hydrologic connectivity is wellsuited to this approach. A similar approach can beused for a hazard inventory as shown in table 13.
Data appropriate for assessing road segments inaddition to those listed in table 13 are:
• Ditch length/km of road
• Mean ditch length/cross drain (or crossing)
• Mean hazard score
• Mean risk score.
Failed Crossing Assessments
Crossing failures provide a unique opportunity toassess design and installation procedures.Following large storm events, storm damagereports are often generated to assess themagnitude and extent of damage. Such effortsshould provide the opportunity for adaptive design.This section discusses features of failed culvertinstallations useful for incorporating into anadaptive design and installation process.
Mechanisms of Road-Stream CrossingFailureDetermining the mechanism of failure at a site isimportant because the site will likely experience a
30
Tabl
e 12
—E
xam
ple
of a
pre
pond
eran
ce ta
ble
from
an
envi
ronm
enta
l ris
k as
sess
men
t of 3
82 r
oad-
stre
am c
ross
ings
in th
e U
pper
Nor
th F
ork
Eel
Riv
er, N
orth
wes
t Cal
iforn
ia. N
ote
the
tend
ency
for
part
icul
ar r
oads
to h
ave
clus
ters
of e
ither
hig
h or
low
ris
k si
tes
(e.g
., 3S
09, 3
S10
, and
5S
30).
One
lim
itatio
n of
this
app
roac
h, h
owev
er, i
s th
at s
ites
with
div
ersi
on p
oten
tial
may
be
over
look
ed a
s is
the
case
for
seve
ral s
ites
with
‘low
’ ran
king
s. T
his
data
set
con
sist
s on
ly o
f roa
d-st
ream
cro
ssin
gs a
nd d
oes
not i
nclu
de c
ross
dra
ins
(thu
s, h
ydro
. con
n.al
way
s =
1).
Hyd
roco
nn.
Div
.di
st.
Fill
vol.
Tw
*s*
Geo
logy
Rec
eivi
ngge
olog
yF
ish
atsi
teF
ish
inba
sin
Pla
nnin
gar
eaC
old
wat
ersu
b-ba
sin
ab
Ris
ksc
ore
Roa
d #
-m
ilepo
st
c
Haz
ard
scor
esE
ndpo
int s
core
s
1331
1331
1000
1000
729
729
512
512
343
343
3S03
-0.6
1
3S10
-1.4
4
3S09
-0.2
8
3S33
-0.0
6
2S08
-0.3
5
3S09
-0.1
4
3S03
-0.9
2
3S10
-1.2
2
3S09
-0.1
7
3S10
-0.4
1
1 1 1 1 1 1 1 1 1 1
1 0 0 0 0 0 1 1 2 1
1 3 3 1 3 0 1 1 2 3
1 0 0 0 0 0 0 0 0 1
3 2 1 2 2 2 1 1 2 1
3 3 3 3 0 3 3 3 0 0
1 0 0 0 0 0 1 1 0 0
0 0 0 0 1 0 0 0 0 0
1 1 1 1 0 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
0 2 2 3 3 3 0 0 0 0
31
Tabl
e 12
—C
ontin
ued.
Ris
ksc
ore
Haz
ard
scor
esE
ndpo
int s
core
s
Hyd
roco
nn.
Roa
d #
-m
ilepo
stD
iv.
dist
.F
illvo
l.T
w*
s*G
eolo
gyR
ecei
ving
geol
ogy
Fis
h at
site
Fis
h in
basi
nP
lann
ing
area
Col
d w
ater
sub-
basi
n
3 3 2 2 2 2 2 2 2 2
5S30
-2.6
9
5S30
-4.5
6
5S30
A-0
.2
5S30
-3.5
1
5S30
-2.0
4
2S16
-0.6
4
5S30
C-0
.3
5S32
-0.0
1
2S16
-0.2
9
2S17
-2.9
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 0 0 0
0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 0 0 0
1 1 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 3 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
1 1 1 1 1 1 1 1 1 1
ab
c
a. H
ydro
con
n. =
hyd
rolo
gic
conn
ectiv
ity
b. D
iv. d
ist
= d
iver
sion
dis
tanc
e
c. F
ill v
ol.
= fi
ll vo
lum
e
32
Figure 12—Overtopping of the fill resulting in erosion of the roadway. Understanding the mechanisms that caused failure is usefulfor upgrading sites to avoid similar consequences in the future.
R9801007
Routenumber
Numberof stream
crossings/km
Total potentialdiversion
distance (m)
Proportion ofroad hydrologically
connected (%)
IS02e
26N11
27N40
2S05p
Numberof cross
drains/km
Crossing fillvolume/km
8.4 1,250151.3 987
2.2 34070.7 53
6.3 2,170391.1 557
1.1 032.1 37
Table 13—Inventory data can be expressed per road segment. Such an approach is often desirable to fit the needs of transportationplanning and estimating overall treatment costs.
similar occurrence again unless designs areimplemented to reduce future hazard. Table 14 liststhe mechanisms, the post-failure evidence for themechanism, and potential design criteria to reducethe hazards. In general, failure can be the resultof the following:
• Culvert capacity exceedence—Water, wood,and sediment in excess of capacity can cause
water to pond at the inlet. Most road-streamcrossing fills function as dams but are notdesigned as such. The increased saturationof the fill may initiate a fill failure. When waterovertops the road surface, some degree of fillerosion is likely (figure 12). However, pondedconditions or an inlet plugged with debris orsediment also promotes deposition within the
33
Figure 13—Burial of culvert inlets is often due to woody debris lodged at the inlet. However, determining this often requiresexcavation (a) or examination from the outlet (b).
inlet basin, which often requires costlyexcavation. Determining the initial mechanismmay be difficult when the inlet is buried or thefill has been completely removed leaving littleevidence (figure 13).
• Fill saturation—Saturation of the fill byponding, overtopping flows, or a rusted culvertinvert increases the hazard of fill failure.
(a) (b)
R9801009
34
Tabl
e 14
—Ty
pes
and
phys
ical
evi
denc
e of
cul
vert
cro
ssin
g fa
ilure
s an
d po
tent
ial d
esig
n cr
iteria
to r
educ
e th
e ha
zard
of s
uch
a m
echa
nism
occ
urrin
g ag
ain.
Fai
lure
mec
hani
sm
Sat
urat
ed fi
llfa
ilure
Vis
ible
evi
denc
eD
iffic
ulty
in d
isce
rnin
gD
esig
n cr
iteria
to r
educ
e fu
ture
haz
ard
•C
ulve
rt in
vert
rus
ted
thro
ugh
•M
ass
was
ting
of fi
ll w
ith o
r w
ithou
tov
erto
ppin
g flo
ws
•K
eep
all s
trea
mflo
w w
ithin
cul
vert
•E
nsur
e th
at fi
ll is
pro
perly
com
pact
edan
d dr
aine
dO
ften
diffi
cult
to d
istin
guis
h fr
omhe
adw
ard
eros
ion
from
ove
rtop
ping
flow
s.
•U
se s
mal
l fill
s to
red
uce
the
amou
ntof
mat
eria
l tha
t can
ent
er th
e ch
anne
l
•U
se c
ours
e m
ater
ial i
n fil
l to
redu
ceth
e am
ount
of f
ine
mat
eria
l ent
erin
gth
e ch
anne
l. H
owev
er, i
f too
cou
rse,
wat
er w
ill p
ipe
thro
ugh
fill a
nd c
ause
othe
r pr
oble
ms
Deb
ris fl
ow•
Cha
nnel
sco
ured
to b
edro
ck•
Min
imiz
e fil
l int
erfe
ring
with
deb
risflo
w p
ath
Eas
y. D
ebris
flow
evi
denc
e is
typi
cally
wel
l pre
serv
ed a
nd e
xten
sive
.
•P
oorly
sor
ted
depo
sits
, ofte
n m
ixed
with
larg
e w
oody
deb
ris in
run
-out
zon
es o
rw
here
roa
d fil
l im
poun
ded
port
ion
of fl
ow
•P
lace
div
ersi
on p
reve
ntio
n di
ps to
one
side
of c
hann
el to
avo
id b
eing
fille
d in
by
debr
is fl
ows
•U
nusu
ally
hig
h sc
our
mar
ks o
r hi
gh w
ater
mar
ks o
n ba
nks
and
vege
tatio
n
•C
onsi
der
low
-wat
er c
ross
ing
•R
efer
, als
o, to
Cos
ta a
nd J
arre
t (19
81)
•P
rono
unce
d sa
g in
ver
tical
cur
veat
cro
ssin
g
35
Tabl
e 14
—C
ontin
ued.
•R
etai
n ve
geta
tion
that
ofte
n m
aint
ains
chan
nel c
onfin
emen
t
•M
aint
ain
chan
nel p
lanf
orm
,lo
ngitu
dina
l pro
file
and
cros
s se
ctio
nto
the
inle
t
•A
void
wid
enin
g th
e in
let b
asin
•U
se c
hann
el w
idth
as
a fa
ctor
for
sizi
ng. W
here
feas
ible
, cul
vert
dia
met
ersh
ould
be
equa
l to
the
chan
nel
wid
th to
faci
litat
e de
bris
pas
sage
.
•A
void
ove
rste
epen
ed c
ut s
lope
sad
jace
nt to
the
inle
t
•E
nsur
e a
leas
t-co
nseq
uenc
e flo
wpa
thfo
r ov
erto
ppin
g flo
ws
•C
reat
e a
catc
h ba
sin
for
sedi
men
tsu
ch th
at it
doe
s no
t affe
ct w
ood
tran
spor
t as
disc
usse
d ab
ove
•C
onfig
ure
culv
ert i
nlet
s to
faci
litat
eal
ignm
ent o
f deb
ris w
ith fl
ow, s
uch
as w
ith m
etal
end
sec
tions
or
angl
edhe
adw
alls
•E
nsur
e a
leas
t-co
nseq
uenc
e flo
wpa
thfo
r ov
erto
ppin
g flo
ws
Fai
lure
mec
hani
sm
Woo
dy d
ebris
lodg
emen
t
Vis
ible
evi
denc
eD
iffic
ulty
in d
isce
rnin
gD
esig
n cr
iteria
to r
educ
e fu
ture
haz
ard
•D
epos
ition
of f
ine
sedi
men
ts (
up to
smal
l peb
bles
) in
inle
t bas
ins,
ofte
nm
oder
atel
y so
rted
and
thin
ly b
edde
d(<
2 cm
thic
k) c
omm
only
with
orga
nic-
rich
lens
es
•O
ne o
r m
ore
piec
es lo
dged
acr
oss
culv
ert i
nlet
Eas
y to
diff
icul
t (de
pend
ing
on m
agni
tude
of d
epos
ition
and
pos
t-ev
ent m
aint
enan
ce).
Ofte
n de
bris
plu
ggin
g is
follo
wed
by
prog
ress
ive
depo
sitio
n of
sed
imen
ts a
t the
inle
t. S
trat
ifica
tion,
sor
ting,
and
gra
in s
ize
dist
ribut
ion
are
usef
ul c
lues
.
Sed
imen
t“s
lug”
•R
apid
, ofte
n ca
tast
roph
ic d
eliv
ery
ofse
dim
ent t
o th
e in
let,
depo
sitio
n ab
ove
the
crow
n of
the
culv
ert
Eas
y to
diff
icul
t (de
pend
ing
on m
agni
tude
of d
epos
ition
and
pos
t-ev
ent e
xcav
atio
n).
Rap
id, c
atas
trop
hic
deliv
ery
of s
edim
ent
burie
s th
e in
let.
Alth
ough
the
part
icle
siz
esde
liver
ed to
the
inle
t may
be
capa
ble
offlu
vial
tran
spor
t thr
ough
the
culv
ert,
rapi
dde
liver
y ov
erw
helm
s th
e tr
ansp
ort c
apac
ity.
See
woo
dy d
ebris
not
es fo
r pr
oble
ms
dist
ingu
ishi
ng b
etw
een
woo
d an
d se
dim
ent.
•A
djac
ent h
illsl
ope
failu
re d
eliv
erin
gm
ater
ial a
sho
rt d
ista
nce
to th
e in
let
•U
nsor
ted
or p
oorly
sor
ted
depo
sits
with
in in
let b
asin
The
mec
hani
sm c
an b
e di
fficu
lt to
disc
rimin
ate
whe
re e
xcav
atio
n ha
soc
urre
d an
d st
ratig
raph
ic e
vide
nce
has
been
rem
oved
.
How
ever
, if t
he b
urie
d cu
lver
t is
appr
opria
tely
con
figur
ed, a
flas
hlig
htsh
one
in fr
om th
e ou
tlet c
an in
dica
teth
e m
echa
nism
. (fig
ure
13).
36
Fai
lure
mec
hani
sm
Hyd
raul
icex
ceed
ence
Vis
ible
evi
denc
eD
iffic
ulty
in d
isce
rnin
gD
esig
n cr
iteria
to r
educ
e fu
ture
haz
ard
•D
rapi
ng o
f fin
e se
dim
ents
with
in th
epo
nded
are
a
•H
igh
wat
er d
ebris
acc
umul
atio
n•
Siz
e cu
lver
t for
at l
east
a 1
00-y
ear
desi
gn fl
ood
with
hea
dwat
er le
ss th
anor
equ
al to
the
culv
ert d
iam
eter
.
Mod
erat
e to
diff
icul
t. H
ydra
ulic
exce
eden
ce r
equi
res
care
ful e
xam
inat
ion
of th
e in
let b
asin
. Deb
ris d
epos
its a
t the
high
wat
er li
ne a
nd fi
ne s
edim
ent d
epos
itsar
e of
ten
of li
mite
d ex
tent
and
rap
idly
cove
red
with
veg
etat
ion.
Whe
re fi
ll er
osio
n an
d di
vers
ion
evid
ence
exis
t, hy
drau
lic e
xcee
denc
e is
arr
ived
by
a pr
oces
s of
elim
inat
ion.
•C
onse
quen
ces
of o
vert
oppi
ng -
fill
eros
ion
or d
iver
sion
evi
denc
e
•In
let n
ot p
lugg
ed w
ith d
ebris
•E
nsur
e a
leas
t-co
nseq
uenc
e flo
wpa
thfo
r ov
erto
ppin
g flo
ws
Tabl
e 14
—C
ontin
ued.
37
LITERATURE CITED
American Iron and Steel Institute (AISI). 1994. Handbook of Steel Drainage & Highway Construction Products.Washington D.C. 518 pp.
Benda, L. 1990. “The Influence of Debris Flows on Channels and Valley Floors in the Oregon Coast Range.”U.S.A. Earth Surface Processes and Landforms, Vol. 15, pp. 457–466.
Best, D.W., H. M. Kelsey, D.K. Hagans, and M. Alpert. 1995. “Role of Fluvial Hillslope Erosion and RoadConstruction in the Sediment Budget of Garret Creek, Humboldt County, California.” In: Nolan, K.M., H.M. Kelsey, and D. C. Marron (eds.), Geomorphic Processes and Aquatic Habitat in the Redwood CreekBasin, Northwestern California. U.S.Geological Survey Professional. Paper 1454. pp. M1-M9.
Braudrick, C.A., G.E. Grant, Y. Ishikawa and H. Ikeda. 1997. “Dynamics of Wood Transport in Streams: AFlume Experiment.” Earth Surface Processes and Landforms, V.22, pp. 669–683.
Broda, K.M., and R. Shockey. 1996. Assessment and Proposed Upgrade Alternatives of Stream Crossings.North Umpqua Ranger District, Umpqua National Forest, Roseburg, Oregon.
Chatwin, S.C., D.E. Howes, J.W. Schwab, and D.N. Swanston. 1994. A Guide for Management of Landslide-prone Terrain in the Pacific Northwest. British Columbia Ministry of Forests, Victoria, British Columbia.220 pp.
Costa, J.E., and R.D. Jarret. 1981. “Debris Flows in Small Mountain Stream Channels of Colorado andTheir Hydrologic Implications.” Bulletin of the Association of Engineering Geologists. Vol. 18, no. 3, pp.309–322.
FEMAT (Forest Ecosystem Management Assessment Team). 1993. Forest Ecosystem Management: Anecological, economic, and social assessment. U.S. Government Printing Office. 1993-073–071.
Flanagan, S.A. “Woody Debris Transport Through Low-order Stream Channels; Implications for StreamCrossing Failure.” MS Thesis, in review.
Furniss, M.J., M. Love, and S.A. Flanagan. 1997. “Diversion Potential at Road-Stream Crossings.” USDAForest Service Technology and Development Program, 9777 1814-SDTDC. 12 pp.
Garland, J.J. 1983. “Designing Woodland Roads.” Oregon State University Extension Service Circular.1137, Oregon State University, Corvallis.
Hafterson, H.D. 1973. “Dip Design.” Field Notes. Washington, D.C: USDA Forest Service EngineeringTechnical Information System. Vol.10. 18 pp.
Hanek, G. 1996. Jackson Creek Restoration Projects—1996: Project Areas No. 1 and 2. Umpqua NationalForest, Roseburg, Oregon.
Inglis, S., L.W. Wasniewski, and R.J. Zuniga. 1995. “Stream Crossing Inventory within Anadromous Drainageson the Payette National Forest.” Working Draft. February 1995. Payette National Forest. 13 pp.
Kennard, P. 1994. Road Assessment Procedure (RAP)—“A Method to Assess and Rank Risks from ForestRoad Landslides.” Final Report to the Weyerhauser Company, Tacoma, Washington. 33 pp.
Lienkaemper, G.W., and F.J. Swanson. 1987. ”Dynamics of Large Woody Debris in Streams in Old-growthDouglas-fir Forests.” Canadian Journal of Forest Research. Vol. 17, pp. 150–156.
38
Mt. Baker-Snoqualmie National Forest. 1997. “Road Decommissioning and Closure Treatment on the Mt.Baker Snoqualmie National Forest.” White Paper. Mountlake Terrace, Washington. 16 pp.
Nakamura, F., and F.J. Swanson. 1994. “Distribution of Coarse Woody Debris in a Mountain Stream, WesternCascades Range, Oregon. Canadian Journal of Forest Research. Vol. 24, pp. 2395–2403.
Normann, J.L., R.J. Houghtalen, and W.J. Johnston. 1985. Hydraulic Design of Highway Culverts. U.S.Department of Transportation, Federal Highway Administration, Hydraulic Design Series No. 5, 272 pp.
Piehl, B.T., M.R. Pyles, and R.L. Beschta. 1988a. “Flow Capacity of Culverts on Oregon Coast Range ForestRoads.” Water Resources Bulletin. Vol. 24, No. 3, pp. 631–637.
Piehl, B.T., R.L. Beschta, and M.R. Pyles. 1988b. “Ditch-relief Culverts and Low-volume Forest Roads in theOregon Coast Range.” Northwest Science. Vol. 62, No. 3, pp. 91–98.
Pyles, M.R., A.E. Skaugset, and T. Warhol. 1989. Culvert Design and Performance on Forest Roads. Presentedat the 12th Annual Council on Forest Engineering Meeting, Coeur d’Alene, Idaho.
Stanislaus National Forest. 1996. Integrated Resource Reference Road Inventory. Stanislaus National Forest,Sonora, California.
Stuber, R.J., A. Beyer, and J.R. Haveman. 1994. “Aquatic Ecosystem Restoration on a Watershed Scale inNorthern Michigan: A Landscape Approach.” Presented at the 56th Midwest Fish and Wildlife Conference,Indianapolis, Indiana.
Waananen, A.O., and J.R. Crippen. 1977. Magnitude and Frequency of Floods in California. U.S. GeologicalSurvey. Water Resources Investigations 77-21. 96 pp.
Weaver, W.E., and D.K. Hagans. 1994. Handbook for Forest and Ranch Roads – A Guide for Planning,Designing, Constructing, Reconstructing, Maintaining and Closing Wildland Roads. Prepared for theMendocino County Resource Conservation District, Ukiah, California. 161 pp.
Weaver, W.E., D.K. Hagans, and J.H. Popenoe. 1995. “Magnitude and Causes of Gully Erosion in the LowerRedwood Creek Basin, Northwestern California.” In: Nolan K.M., H. M. Kelsey, and D. C. Marron (eds.),Geomorphic Processes and Aquatic Habitat in the Redwood Creek Basin, Northwestern California. U.S.Geological Survey Professional Paper 1454. pp. I1–I21.
Weinhold, M. 1996. Road-stream Crossing Risk Table. Siskiyou National Forest. Powers Ranger District.
Wemple, B.C. 1994. “Hydrologic Integration of Forest Roads with Stream Networks in Two Basins, WesternCascades, Oregon.” Master of Science Thesis. Oregon State University, Corvallis, Oregon.
39
APPENDICES
40
41
APPENDIX A
AN EXAMPLE OF ROAD ENVIRONMENTAL RISK ASSESSMENT FOR USDA FORESTSERVICE LANDS IN THE UPPER NORTH FORK EEL RIVER WATERSHED, SIX RIVERSNATIONAL FOREST
Following are the results from inventory and assessment of road-stream crossings and cross drains onUSDA Forest Service lands in the upper North Fork Eel River. Inventory was conducted on the 296-km2 (114mi2) portion of USDA Forest Service land. Culvert diameters at road-stream crossings (308) were typically450 mm or 600 mm. Drainage area was definable on a 7.5 minute topographic map for 207 (67 percent) ofthe road-stream crossings. For those sites, 63 percent are unable to pass a 100-year flood without submergingthe culvert inlet, and 43 percent overtop the road fill for a 100-year peak flow.
Potential physical consequences include fill erosion and stream diversion. Median fill volume is 141m3 percrossing with 15 percent having a volume greater than 500 m3. Forty-five percent of the road-stream crossingsand 67 percent of the cross drains have diversion potential. Influencing potential physical consequenceswas an unstable geologic unit. Inspection of roads and crossings within this unit revealed that past failureswere of much greater erosional consequences, and ongoing chronic erosion was higher than surroundingareas within relatively stable geologic settings.
The following endpoints were identified in the analysis area:
• Anadromous fish (this applied to all sites as all failures are assumed to have an impact)
• Fish at crossing (sites overlapping with the distribution of anadromous fish)
• Cold water refuges (the North Fork Eel watershed is thermally impaired. Three sub-watersheds wereidentified in the analysis area as important cold water tributaries and refuge areas).
An environmental risk score was assigned to each crossing. Maps on the following pages (figures A-1 andA-2) display the distribution of road-stream crossings, cross drains, road-stream crossings with high diversionpotential, and high risk road-stream crossings. In this example, sites were assigned into the high risk categoryif the risk score was > 100. This value should be adjusted to reflect watershed conditions and the number ofendpoints considered. Note the clustering of high risk sites along road segments in the eastern portion ofthe basin. This clustering effect allowed for efficient treatment efforts.
42
Scale 1:170,000
Ownership Boundary
WaterRoads
Cross Drains
Stream Road Crossings
Road - Stream Crossings and Cross Drainson Forest Service Lands -
Upper North Fork Eel River, Northwest CaliforniaSix Rivers National Forest
Map Compiled by GIS GroupWatershed Analysis Center
August 27, 1998
R9801012
Figure A-1.
43
Scale 1:170,000
Ownership Boundary
WaterRoads
High Risk Road Stream Crossings
High Risk Road - Stream Crossings onForest Service Lands -
Upper North Fork Eel River, Northwest CaliforniaSix Rivers National Forest
Map Compiled by GIS GroupWatershed Analysis Center
August 27, 1998
R9801011
Figure A-2.
44
45
y ( )
LocationRoad # m.p. lat lon
Quad: T R Sec 1/4 1/4 photo
Crossing type: Stream Swale X-drain Other:
Channel type: Intermittent Unchanneled Ditch
Channel widths:
Channel depths:
Channel slope (%):
Contributing ditch length(s): L R
Culvert type:CMP
arch
bottomless arch
plastic
concrete box
Diameter:
slope (%):
Fill volume:flood prone width:
Upstr. fillslope len:
Upstr. fillslope (%):
Road width:
Fill width:
Dnstr. fillslope len:
Dnstr. fillslope (%):
Diversion potential: YES NOpotential distance:
receiving feature:
adjacent cross drain site#
adjacent stream crossing site#
hillslope
other:
road slope through crossing (%):
Cross drains and other drainage features:Gully below outlet? YES NO
Gully Legnth:
Avg width:
Connected to channel? YES NO
Avg. depth
Entrance type:projecting
mitered
flush
side tapered
drop inlet
other:
% dent/crush:
skew angle (deg):
APPENDIX B
Sample Data Sheet For Road Drainage Inventory date:surveyor(s):
46
47
48