methods for inventory and development program environmental …€¦ · • review inventory and...

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

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

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.

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

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


Volume 40 to 150 m3

High Hazard


Volume 40 to 150 m3

Moderate Hazard


Volume 40 to 150 m3

Low Hazard


Volume < 40 m3

Low Hazard


Volume < 40 m3

Low Hazard


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




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

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.

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):

methods for inventory and development program environmental …€¦ · • review inventory and...

Documents