Drainage Chaptr
Drainage Chaptr
Drainage Chaptr
Provision for adequate drainage is of paramount importance in road design and cannot be
overemphasized. The presence of excess water or moisture within the roadway will
adversely affect the engineering properties of the materials with which it was constructed.
Cut or fill failures, road surface erosion, and weakened subgrades followed by a mass
failure are all products of inadequate or poorly designed drainage. As has been stated
previously, many drainage problems can be avoided in the location and design of the road:
Drainage design is most appropriately included in alignment and gradient planning.
Hydrologic factors to consider in locating roads are number of stream crossings, side slope,
and moisture regime. For example, at the lowest point on the slope, only one or two stream
crossings may be required. Likewise, side slopes generally are not as steep, thereby
reducing the amount of excavation. However, side cast fills and drainage requirements will
need careful attention since water collected from upper positions on the slope will
concentrate in the lower positions. In general, roads built on the upper one-third of a slope
have better soil moisture conditions and, therefore, tend to be more stable than roads built
on lower positions on the slope.
Increased risks of road failures are created at points A and B. At point A, water will pond
above the road fill or flow downslope through the roadside ditch to point B. Ponding at A
may cause weakening and/or erosion of the subgrade . If the culvert on Stream 1 plugs,
water and debris will flow to point A and from A to B. Hence, the culvert at B is handling
discharge from all three streams. If designed to minimum specifications, it is unlikely that
either the ditch or the culvert at B will be able to efficiently discharge flow and debris from
all three streams resulting in overflow and possible failure of the road at point B.
Figure 59. Slope shape and its impact on slope hydrology. Slope
shape determines whether water is dispersed or concentrated.
(US Forest Service, 1979).
A road drainage system must satisfy two main criteria if it is to be effective throughout its
design life:
The design of drainage structures is based on the sciences of hydrology and hydraulics-the
former deals with the occurrence and form of water in the natural environment
(precipitation, streamflow, soil moisture, etc.) while the latter deals with the engineering
properties of fluids in motion.
In addition to considering intensity and duration of a peak rainfall event, the frequency, or
how often the design maximum may be expected to occur, is also a consideration and is
most often based on the life of the road, traffic, and consequences of failure. Primary
highways often incorporate frequency periods of 50 to 100 years, secondary roads 25 years,
and low volume forest roads 10 to 25 years.
Of the water that reaches the ground in the form of rain, some will percolate into the soil to
be stored until it is taken up by plants or transported through pores as subsurface flow,
some will evaporate back into the atmosphere, and the rest will contribute to overland flow
or runoff. Streamflow consists of stored soil moisture which is supplied to the stream at a
more or less constant rate throughout the year in the form of subsurface or groundwater
flow plus water which is contributed to the channel more rapidly as the drainage net
expands into ephemeral channels to incorporate excess rainfall during a major storm event.
The proportion of rainfall that eventually becomes streamflow is dependent on the
following factors:
1. The size of the drainage area. The larger the area, the greater the volume of
runoff. An estimate of basin area is needed in order to use runoff formulas and
charts.
2. Topography. Runoff volume generally increases with steepness of slope. Average
slope, basin elevation, and aspect, although not often called for in most runoff
formulas and charts, may provide helpful clues in refining a design.
3. Soil. Runoff varies with soil characteristics, particularly permeability and
infiltration capacity. The infiltration rate of a dry soil, by nature of its intrinsic
permeability, will steadily decrease with time as it becomes wetted, given a constant
rainfall rate. If the rainfall rate is greater than the final infiltration rate of the soil
(infiltration capacity), that quantity of water which cannot be absorbed is stored in
depressions in the ground or runs off the surface. Any condition which adversely
affects the infiltration characteristics of the soil will increase the amount of runoff.
Such conditions may include hydrophobicity, compaction, and frozen earth.
A number of different methods are available to predict peak flows. Flood frequency
analysis is the most accurate method employed when sufficient hydrologic data is available.
For instance, the United States Geological Survey has published empirical equations
providing estimates of peak discharges from streams in many parts of the United States
based on regional data collected from "gaged" streams. In northwest Oregon, frequency
analysis has revealed that discharge for the flow event having a 25-year recurrence interval
is Most closely correlated with drainage area and precipitation intensity for the 2-year, 24-
hour storm event. This is, by far, the best means of estimating peak flows on an ungaged
stream since the recurrence interval associated with any given flow event can be identified
and used for evaluating the probability of failure.
The probability of occurrence of peak flows exceeding the design capacity of a proposed
stream crossing installation should be determined and used in the design procedure. To
incorporate this information into the design, the risk of failure over the design life must be
specified. By identifying an acceptable level of risk, the land manager is formally stating
the desired level of success (or failure) to be achieved with road drainage structures. Table
25 lists flood recurrence intervals for installations in relation to their design life and
probability of failure.
EXAMPLE: If a road culvert is to last 25 years with a 40% chance of failure during the design
life, it should be designed for a 49-year peak flow event (i.e., 49-year recurrence interval).
When streamflow records are not available, peak discharge can be estimated by the
"rational" method or formula and is recommended for use on channels draining less than 80
hectares (200 acres):
Q = 0.278 C i A
Q=CiA
Factor
Type of Surface
C
0.05-
Sandy soil, flat, 2%
0.10
0.10-
Sandy soil, average, 2-7%
0.15
0.15-
Sandy soil, steep, 7
0.20
0.13-
Heavy soil, flat, 2%
0.22
0.18-
Heavy soil, average, 2-7%
0.22
0.25-
Heavy soil, steep, 7%
0.35
0.80-
Asphaltic pavements
0.95
0.70-
Concrete pavements
0.95
Gravel or macadam 0.35-
pavements 0.70
Numerous assumptions are necessary for use of the rational formula: (1) the rate of runoff
must equal the rate of supply (rainfall excess) if train is greater than or equal to tc; (2) the
maximum discharge occurs when the entire area is contributing runoff simultaneously; (3)
at equilibrium, the duration of rainfall at intensity I is t = tc; (4) rainfall is uniformly
distributed over the basin; (5) recurrence interval of Q is the same as the frequency of
occurrence of rainfall intensity I; (6) the runoff coefficient is constant between storms and
during a given storm and is determined solely by basin surface conditions. The fact that
climate and watershed response are variable and dynamic explain much of the error
associated with the use of this method.
Manning's formula is perhaps the most widely used empirical equation for estimating
discharge since it relies solely on channel characteristics that are easily measured.
Manning's formula is:
Area and wetted perimeter are determined in the field by observing high water marks on the
adjacent stream banks (Figure 61). Look in the stream bed for scour effect and soil
discoloration. Scour and soil erosion found outside the stream channel on the floodplains
may be caused by the 10-year peak flood. Examining tree trunks and brush in the channel
and floodplain may reveal small floatable debris hung up in the vegetation. Log jams are
also a good indication of flood marks because their age can be estimated and old, high log
jams will show the high watermark on the logs. The difficulty in associating high water
marks with flow events of a specified recurrence interval makes values obtained by this
method subject to gross inaccuracy. If the 10-year flood can be determined, flow levels for
events with a higher recurrence interval can be determined roughly from Table 28.
Factor of flood
Peak flow return period
intensity
(years) (10-year peak flow = 1.00)
10 1.00
25 1.25
50 1.50
100 1.80
A key assumption in the use of Manning's equation is that uniform steady flow exists. It is
doubtful that high gradient forested streams ever exhibit this condition. (Campbell, et al.,
1982) When sufficient hydrologic data is lacking, however, Manning's equation, together
with observations of flow conditions in similar channels having flow and/or precipitation
records, provide the best estimate of stream discharge for purposes of designing stream
crossings. An example illustrating the use of Manning's equation to calculate peak
discharge is as follows:
EXAMPLE: A trapezoidal channel of straight alignment and uniform cross section has a
bottom width of 10 meters, side slopes 1:1, channel slope 0.003, and high water depth (25-
year event) of 5 meters. The channel has weeds and heavy brush along its banks.
2. The cross sectional area is equal to 1/2 x sum of parallel sides x perpendicular height =
0.5(10 + 20)(5) = 75 m².
3. The hydraulic radius is cross sectional area + wetted perimeter = 75 / 24.1 = 3.1 m.
Channel crossings require careful design and construction. Functionally, they must (1)
allow for passage of the maximum amount of water which can reasonably be expected to
occur within the lifetime of the structure and (2) not degrade water quality or endanger the
structure itself or any downstream structures. It should be pointed out that most road
failures are related to inadequate water passage structures and fill design and placement as
well as poor construction practices in such locations.
Accelerated erosion brought about by failure of channel crossing structures can be caused
by:
1. Inadequate design to handle peak flow and debris. Water will back up behind
structure, saturating the fill and creating added hydrostatic pressure. Water will
overflow the structure and the fill may be washed out.
2. Inadequate outlet design. By constricting flow through a small area, water velocity
(along with its erosive power) will increase. Outlets need to be properly designed in
order to withstand high flow velocities and thus avoid excessive downstream
erosion and eventual road failure.
3. Poor location of crossing. Crossings need to be located along relatively stable
stretches where stream bottoms and banks exhibit little signs of excessive erosion or
deposition. Meandering and/or multiple channels often indicate unstable conditions.
If there is no choice but to use a poor location, careful consideration of the type of
crossing selected, along with bank and stream bottom stabilization and protection
measures, should be given.
There are three generally accepted methods used to cross channels on low volume roads--
bridges, fords, and culverts. The selection is based on traffic volume and characteristics,
site conditions (hydrologic/hydraulic conditions of channel), and management needs such
as occasional closure, continuous use, safety considerations, resource impact (fish, wildlife,
sediment). Factors to consider when selecting a crossing type are listed as follows:
1. Bridges: high traffic volume, large and variable water volume, high debris-potential,
sensitive channel bottom and banks, significant fish resource, large elevation
difference between channel and road grade
2. Culvert: Medium to low water volume, medium to low debris potential, fish
resource not significant, elevation difference between channel and road grade less
than 10 meters, high traffic volume
3. Ford: low to intermittent water flow, high debris potential, no fish resource, road
grade can be brought down to channel bottom, low traffic volume
All three channel crossing types require a careful analysis of both vertical and horizontal
alignment. In particular, careful analysis of curve widening requirements is imperative in
relation to the specified critical vehicle. Channel crossings are fixed structures where the
road way width cannot be temporarily widened. Road width, curvature, approach, and exit
tangents govern the vehicle dimensions which can pass the crossing.
Except for bridge locations, roads should climb away from channel crossings in both
directions wherever practical so high water will not flow along the road surface. This is
particularly true for ford installations.
4.3.2 Fords
Fords are a convenient way to provide waterway crossing in areas subject to flash floods,
seasonal high storm runoff peaks, or frequent heavy passage of debris or avalanches. Debris
will simply wash over the road structure. After the incident, some clearing may be
necessary to allow for vehicle passage. Figure 62 shows a very simple ford construction
where rock-filled gabions are used to provide a road bed through the stream channel.
1. The ford should allow for passage of debris and water without diverting it onto the road
surface. The ford results in a stream bed gradient reduction. Therefore, debris has a
tendency to be deposited on top of a ford because of reduced flow velocity.
2. Fords should be designed with steep, short banks which help to confine and channel the
stream (Figure 63). The steepness and length of the adverse grade out of ford depends on
the anticipated debris and water handling capacity required as well as vehicle geometry
(See Chapter 3.1.3). Typically, the design vehicle should be able to pass the ford without
difficulty. Critical vehicles (vehicles which have to use the road, but only very infrequently)
may require a temporary fill to allow passage.
4.3.3. Culverts
Culverts are by far the most commonly used channel crossing structure used on forest
roads. Culvert types normally used, and the conditions under which they are used, areas
follows:
Corrugated metal pipe (CMP) ........................................ All conditions except those noted
below
Although more expensive than round culverts, pipe-arch or plate arch types are preferred
over ordinary round pipes. Pipe-arch culverts, beers having a more efficient opening per
unit area than round pipe for a given discharge, will collect bottom sediments over time
when it is installed slightly below the stream grade. They also require lower fills. However,
during periods of low flow, water in pipes with this shape may be spread so thin across the
bottom that fish passage is impossible. A plate-arch set in concrete footings is the most
desirable type from a fish passage standpoint since it has no bottom. The stream can remain
virtually untouched if care is exercised during its installation. (Yee and Roelofs, 1980)
Regardless of the type of culvert, they should all conform to proper design standards with
regards to alignment with the channel, capacity, debris control, and energy dissipation.
They should all perform the following functions:
1. The culvert with its appurtenant entrance and outlet structures should efficiently
discharge water, bedload, and floating debris at all stages of flow.
2. It should cause no direct or indirect property damage.
3. It should provide adequate transport of water, debris, and sediment without drastic
changes in flow patterns above or below the structure.
4. It should be designed so that future channel, and highway improvements can be
made without much difficulty.
5. It should be designed to function properly after fill has settled.
6. It should not cause objectionable stagnant pools in which mosquitoes could breed.
7. It should be designed to accommodate increased runoff occasioned by anticipated
land development.
8. It should be economical to build, hydraulically adequate to handle design discharge,
structurally durable, and easy to maintain.
9. It should be designed to avoid excessive ponding at the entrance which may cause
property damage, accumulation of sediment, culvert clogging, saturation of fills, or
detrimental upstream deposits of debris.
10. Entrance structures should be designed to screen out material which will not pass
through the culvert, reduce entrance losses to a minimum, make use of velocity of
approach insofar as practical, and by use of transitions and increased slopes, as
necessary, facilitate channel flow entering the culvert.
11. The outlet design should be effective in re-establishing tolerable non-erosive
channel flow within the right-of-way or within a reasonably short distance below
the culvert, and should resist undercutting and washout.
12. Energy dissipators should be simple, easy to build, economical and reasonably self-
cleaning during periods of low flow.
13. Alignment should be such that water enters and exits the culvert directly. Any
abrupt change in direction at either end will retard flow and cause ponding, erosion,
or a buildup of debris at the culvert entrance. All of these conditions could lead to
failure. (See Figure 65 for suggested culvert-channel alignment configurations and
Figure 66 for suggested culvert grades. In practice, culvert grade lines generally
coincide with the average streambed above and below the culvert.)
If there are existing roads in the watershed, examination of the performance of existing
culverts often serves as the best guide to determining the type, size, and accompanying
inlet/outlet improvements needed for the proposed stream crossing. For estimating
streamflow on many forest watersheds, existing culvert installations may be used as
"control sections". Flow can be calculated as the product of water velocity (V) and cross-
sectional area (A):
Q=A*V
where:S = slope
n = Manning's
roughness
coefficient
R = hydraulic
radius (meters)
(see
Figure
67)
Values for coefficient of roughness (n) for culverts are given in Table 29.
Annular
Culvert
corrugations n
diameter (ft)*
(in)*
corrugated
1 to 8 2-2/3 x 1/2 0.024
metal
3 to 8 3x1 0.027
concrete all diameters --- 0.012
* 1ft = 0.30 m, 1 in. = 2.54 cm
The types of flow conditions found in conventional circular pipes and pipe-arch culverts are
illustrated in Figure 68. Under inlet control, the cross-sectional area of the barrel, the inlet
configuration or geometry, and the amount of headwater or ponding are of primary
importance. Under outlet control, tailwater depth in the outlet channel and slope, roughness,
and length of the barrel are also considered. The flow capacity of most culverts installed in
forested areas is usually determined by the characteristics of the inlet since nearly any pipe
that has a bottom slope of 1.5% or greater will exhibit inlet control. At slopes of 3% or
greater, the culvert can become self-cleaning of sediment.
Once the design peak discharge has been determined by one of the methods discussed
above, the size of pipe required to handle the discharge can be determined from available
equations, charts, tables, nomographs, etc., such as the ones presented in Figures 70, 71, 72,
73 and 74. Figure 69 provides an example of a work sheet which can be used for diameter
and flow capacity calculations. If outlet control is indicated (for example, in a low gradient
reach where "backwater effects" may be created at the outlet end), the reader is referred to
Handbook of Steel Drainage and Highway Construction Products (1971) or Circular No. 5
published by the U. S. Department of Commerce (1963). Outlet control conditions are
shown in Figure 74 for a corrugated metal pipe.
It is important to keep in mind that in addition to discharge from areas upstream of the
installation, the culvert must be able to handle accumulated water from roadside ditches
recalling that roadside ditches on roads lower on the slope intercept more subsurface water
than those on roads higher on the slope. Sudden surges from rapid snowmelt (if applicable)
must also be allowed for. Organic debris and bedload sediments can plug a culvert and can
greatly reduce culvert efficiency. For these reasons, an "oversized" culvert design may be
indicated.
Inlet characteristics can greatly influence flow efficiency through the culvert. The end
either (1) projects beyond the fill, (2) is flush with a headwall, or (3) is supplemented with a
manufactured mitered steel end section. Inlets with headwalls are generally the most
efficient followed by culverts with mitered inlets and finally culverts with projecting
entrances. When headwater depths are 1 to 2 times greater than culvert diameter, culverts
with headwalls have an increase in flow capacity of approximately 11 and 15%,
respectively, over culverts with projecting entrances.
Note: Culvert design sheets, similar to Figure 69 should be used to record design data.
d. Type of culvert, including barrel material, barrel cross-sectional shape and entrance type.
e. Slope of culvert. (If grade is given in percent, convert to slope in meters per meter).
g. Convert metric units to english units for use with the nomographs.
Volume flow Q(m3/sec) to Q(cfs) : 1 m3/sec = 35.2 cfs (cubic ft/sec). Multiply Q(m3/sec)
by 35.2 to get Q(cfs)
Length, Diameter (meter) : 1 meter = 3.3 ft.; 1 cm = 0.4 inches. Multiply (cm) by 0.4 to get
(inches). Multiply (meter) by 3.3 to get (feet)
a. Refer to the inlet control nomograph for the culvert type selected.
b. Using an HW/D (Headwater depth/Diameter)) of approximately 1.5 and the scale for the
entrance type to be used, find a trial size culvert by following the instructions for use of the
nomographs. If a lesser or greater relative headwater depth should be needed, another value
of HW/D may be used.
c. If the trial size for the culvert is obviously too large because of limited height of
embankment or size availability, try different HW/D values or multiple culverts by dividing
the discharge equally for the number of culverts used. Raising the embankment height or
using a pipe arch and box culvert which allow for lower fill heights is more efficient
hydraulically than using the multiple culvert approach. Given equal end areas, a pipe arch
will handle a larger flow than two round culverts. Selection should be based on an
economic analysis.
Step 3: Find headwater (HW) depth for the trial size culvert:
a. Determine and record. HW depth by use of the appropriate inlet control nomograph.
Tailwater (TW) conditions are to be neglected in this determination. HW in this case is
found by simply multiplying HW/D (obtained from the nomograph) by D.
a. If inlet control governs, outlet velocity can be assumed to equal normal velocity in open-
channel flow as computed by Manning's equation for the barrel size, roughness, and slope
of culvert selected.
Note: In computing outlet velocities, charts and tables such as those provided by U.S. Army
Corp of Engineers, Bureau of Reclamation, and Department of Commerce are helpful (see
Literature Cited).
Step 5: Try a culvert of another type or shape and determine size and HW by the above
procedure.
Step 6: Record final selection of culvert with size, type, outlet velocity, required HW and
economic justification. A good historical record of culvert design, installation, and
performance observations can be a valuable tool in planning and designing future
installations.
a. Connect with a straight edge the given culvert diameter or height (D) and the discharge
Q, or Q/B for box culverts; mark intersection of straightedge on HW/D scale mark (1).
b. If HW/D scale mark (1) represents entrance type used, read HW/D on scale (1). If some
other entrance type is used, extend the point of intersection found in (a) horizontally to
scale (2) or (3) and read HW/D.
a. Given an HW/D value, locate HW/D on scale for appropriate entrance type. If scale (2)
or (3) is used, extend HW/D point horizontally to scale (1).
b. Connect point on HW/D scale (1) as found in (a) above to given discharge and read
diameter of culvert required.
a. Given HW and D, locate HW/D on scale for appropriate entrance type. Continue as in 2a.
b. Connect point on HW/D scale (1) as found in (a) above, the size of culvert on the left
scale, and read Q or Q/B on the discharge scale.
Good installation practices are essential for proper functioning of culverts, regardless of the
material used in the construction of the culvert (Figure 75). Flexible pipe such as
aluminum, steel, or polyethylene, requires good side support and compaction, particularly
in the larger sizes. It is recommended that the road be constructed to grade or at least a
meter above the top of the pipe, the fill left to settle and then excavated to form the required
trench.
The foundation dictates if bedding is needed or not. Proper foundation maintains the
conduit on a uniform grade. Most times, the culvert can be laid without bedding, however;
a few centimeters of bedding helps in installation of the culvert. When bedding is required,
the depth should be 8 cm if the foundation material is soil and 30 cm if it is rock.
Backfilling is the most important aspect of culvert installation. Ten percent of the loading is
taken by the pipe and 90 percent is taken by the material surrounding the pipe if backfilling
is done correctly. Backfill material should consist of earth, sand, gravel, rock or
combinations thereof, free of humus, organic matter, vegetative matter, frozen material,
clods, sticks and debris and containing no stones greater than 8 cm (3 in) in diameter. It
should be placed in layers of no greater than 15 cm (6 in) and compacted up to 95% of
Proctor density at or near the optimum moisture content for the material.
Figure 70. Nomograph for concrete pipes inlet control (U.S. Dept. of
Commerce, 1963).
Figure 73. Nomograph for box - culvert, inlet control. (U.S. Dept.
of Commerce, 1963).
Nomograph for corrugated metal pipe (CMP), outlet
Figure 74.
control. (U.S. Dept. of Commerce,1963).
A critical factor in the assessment of channel crossing design and structural capacity is its
allowance for handling or passing debris. Past experience has shown that channel crossings
have failed not because of inadequate design to handle unanticipated water flows, but
because of inadequate allowances for floatable debris which eventually blocked water
passage through the culvert. Therefore, each channel crossing has to be analyzed for its
debris handing capacity.
When upstream organic debris poses an immediate threat to the integrity of the culvert,
several alternatives may be considered.
1. Cleaning the stream of floatable debris is risky and expensive. Since many of the
hydraulic characteristics of the channel are influenced by the size and placement of
debris, its removal must be carried out only after a trained specialist, preferably a
hydrologist, has made a site-specific evaluation of channel stability factors.
2. Various types of mechanical structures (Figures 76, 77 and 78) can be placed above
the inlet to catch any debris that may become entrained.
3. A bridge may be substituted in place of a culvert.
4.3.5. Bridges
Bridges often represent the preferred channel crossing alternative in areas where aquatic
resources are extremely sensitive to disturbance. However, poor location of footings,
foundations, or abutments can cause channel scour and contribute to debris blockage.
Bridges have been designed using a variety of structural materials for substructure and
superstructure. Selection of a bridge type for a specific site should take into consideration
the functional requirements of the site, economics of construction at that site, live load
requirements, foundation conditions, maintenance evaluations, and expertise of project
engineer.
Some arbitrary rules for judging the minimum desirable horizontal and vertical stream
clearances in streams not subject to navigation may be established for a specific area based
on judgment and experience. In general, vertical clearances should be greater than or equal
to 1.5 meters (5 feet) above the 50-year flood level plus 0.02 times the horizontal distance
between piers. Horizontal clearance between piers or supports in forested lands or crossings
below forested lands should not be less than 85 percent of the anticipated tree height in the
forested lands or the lateral width of the 50-year flood. (US Environmental Protection
Agency, 1975)
Of course, longer bridge spans will require careful economic evaluations since higher
superstructure costs are often involved. Subaqueous foundations are expensive and involve
a high degree of skill in the construction of protective cofferdams, seal placement and
cofferdam dewatering. In addition to threats to water quality that can occur from a lost
cofferdam, time and money losses will be significant. Subaqueous foundations often limit
the season of construction relative to water level and relative to fish spawning activity.
Thus, construction timing must be rigidly controlled.
Another alternative is the use of locally available timber for log stringer bridges. An
excellent reference for the design and construction of single lane log bridges is Log Bridge
Construction Handbook, by M. M. Nagy, et al., and is published by the Forest Engineering
Research Institute of Canada. The reader is referred to this publication for more detailed
discussions of these topics.
4.4 Road Surface Drainage
4.4.1 Surface Sloping
Reducing the erosive power of water can achieved by reducing its velocity. If, for practical
reasons, water velocity cannot be reduced, surfaces must be hardened or protected as much
as possible to minimize erosion from high velocity flows. Road surface drainage attempts
to remove the surface water before it accelerates to erosive velocities and/or infiltrates into
the road prism destroying soil strength by increasing pore water pressures. This is
especially true for unpaved, gravel, or dirt roads.
Water moves across the road surface laterally or longitudinally. Lateral drainage is
achieved by crowning or by in- or out- sloping of road surfaces (Figure 79). Longitudinal
water movement is intercepted by dips or cross drains. These drainage features become
important on steep grades or on unpaved roads where ruts may channel water longitudinally
on the road surface.
Transverse Sediment
grade Delivery
tonnes/ha/year
conventional
970
0-2%
5% 400
9% 300
12 % 260
* 4 meter wide road surface
4 - 16 trucks/day
3900 mm annual precipitation
Sloping or crowning significantly reduces sediment delivery from road surfaces. A study by
Reid (1981) showed a reduction in sediment delivery by increasing the transverse road
surface grade. In this particular case the road surfaces insloped from 5 to 12 percent were
compared with conventionally constructed road surfaces at grades of 0 to 2 percent.
Sediment yield was reduced by a factor of 3.0 to 4.5 when compared to a conventionally
sloped road (Table 30).
Outsloping is achieved by grading the surface at 3 to 5 percent cross slope toward the
downhill side of the road. Outsloped roads are simple to build and to maintain.
Disadvantages of outsloping include traffic safety concerns and lack of water discharge
control. When surfaces become slippery (i.e., snow or ice cover, or when silty or clayey
surfaces become wet), vehicles may lose traction and slide toward the downhill edge.
Outsloping should only be used under conditions where run-off can be directed onto stable
areas. If terrain is less than 20 percent slope and the road gradient is less than 4 percent,
outsloping is not an effective way of water removal.
Temporary roads or roads with very light traffic can be outsloped where side slopes do not
exceed 40 percent. For safety reasons, when side slopes exceed 40 percent, traffic
restrictions should be in force during inclement weather. When outsloping is used for
surface drainage, cross drains or dips should be installed on the road surface (Figure 75).
Spacing will depend on soil type, road surface and road grade.
Insloping is used where a more reliable drainage system is required such as on permanent
roads, roads with high anticipated traffic volumes and/or loads, or in areas with sensitive
soils or severe climatic conditions. Insloping is achieved by grading the road surface
towards the uphill side of the road at a 3 to 5 percent grade. Water draining from insloped
road surfaces is collected and carried along the inside of the road either on the road surface
itself or more commonly in a ditch line. The ditch line can be omitted from the road
template, thereby reducing the overall road width. This may be desirable in steep terrain in
order to reduce excavation (see also Section 3.2). However, this option must be weighed
against potential drainage problems along the uphill side of the road. Dips, cross drains, or
culverts must be installed and maintained to remove water from the road prism.
Crowned surfaces provide the fastest water removal since the distance water has to travel
is cut in half. The crowned surface slopes at 3 to 10 percent from either side of the road
centerline. Crowned surfaces and any associated cross drains or dips are difficult to
maintain. Water has to be controlled on both sides of the road through a ditch line and
stable areas have to be provided for runoff water. Ballast thickness is typically the largest in
the center in order to achieve the correct crown shape.
Cross drains are often needed to intercept the longitudinal, or down-road, flow of water in
order to reduce and/or minimize surface erosion. In time, traffic will cause ruts to form,
channeling surface water longitudinally down the road. Longitudinal or down-road flow of
water becomes increasingly important with:
- increasing grades
- rutting frequency
- road surface protection
There are three types of cross drains used for intercepting road surface water: intercept-ing
or rolling dips, open top culverts, and cross ditches. Cross drains serve a dual purpose. First
they must intercept longitudinal road surface flow, and second they must carry ditch water
across the road prism at a frequency interval small enough to prevent concentration of flow.
Ditch relief is discussed in more detail in section 4.4.3 and 4.4.4.
Intercepting dips (Figures 80 and 81) when properly constructed, are cheaper to
maintain and more permanent than open-top culverts. However, their usefulness is limited
to road grades less than 10 percent. At steeper grades, they become difficult to construct
and maintain.
Dip locations are determined at the time the grade line is established on the ground or
during vertical alignment design. The total length of the two vertical curves comprising the
dip should be sufficient to allow the design vehicle to pass safely over them at the design
speed. The minimum vertical distance between the crest and sag of the curves should be at
least 30 cm (1 ft). It is important that the dip be constructed at a 30 degree or greater angle
downgrade and that the dips have an adverse slope on the downroad side. The downroad
side of the dip should slope gently downward from the toe of the road cut to the shoulder of
the fill. The discharge point of the dip should be armored with rock or equipped with a
down-drain to prevent erosion of the fill. Equipment operators performing routine
maintenance should be aware of the presence and function of the dips so that they are not
inadvertently destroyed.
Open top culverts are most effective on steeper road grades. Open top culverts (Figure
82) can be made of durable treated lumber or poles or they may be prefabricated from
corrugated, galvanized steel. The trough should be 7 to 10 cm (3 to 4 in) wide and from 10
to 20 cm (4 to 8 in) deep. The gradient required in order for open top culverts to be self
cleaning is 4 percent or greater and, as with dips, they should be angled 30 degrees
downslope. In order to maintain their functionality they should be inspected and cleaned on
a frequent and regular basis.
Cross ditches or water bars, are typically used on temporary roads. They are the easiest
and most inexpensive method for cross drain installation (Figure 83). However, they
impede traffic, wear out quickly, and are difficult to maintain and are, therefore, not
recommended except on very low standard roads. In order to be effective, the cross ditch
should be excavated into the mineral soil or subgrade and not just into the dirt or surface
layer. Water bars should be installed at a 30 degree angle to the centerline of the road, and
ditch and berm should be carefully extended to the cut bank in order to avoid ditch water
bypass. A berm should be placed in the cut bank ditch to divert water into the cross ditch.
Care should be taken that the berm and ditch is not beaten or trampled down by traffic or
livestock.
Spacing requirements for surface cross drains depend on road grade, surfacing material,
rain intensities, and slope and aspect. Spacing guides for surface cross drains are given in
Table 31.
Road Material
Grade Hard Basalt Granite Glacial Andesite Loess
(%) sediment silt
Cross drain spacing, m
2 51 47 42 41 32 29
4 46 42 38 37 27 24
6 44 40 35 34 25 22
8 42 38 33 32 23 20
10 39 35 29 29 20 17
12 36 32 27 27 17 15
14 33 29 24 23 14 11
[1]. In middle topographic position, reduce
spacing 5.5 m; in lower topographic position,
reduce spacings 11 m.
[2]. On south aspects, reduce spacings 4.6m.
[3]. For each 10% decrease in slope steepness
below 80%, reduce spacing 1.5m.
A Japanese open-top culvert spacing guide uses road grade as input (Figure 84) On steep
grades, spacing is similar to data given in Table 30. However, on gentler grades (2 - 8%),
the Japanese spacing guide provides for considerably wider spacings. This is a good
illustration of a case where local conditions take precedent over general guidelines
developed for large geographical areas.)
Equal attention must be given to location of cross drains in relation to road and topographic
features. Natural features such as slope breaks or ideal discharge locations which disperse
water should be identified and incorporated into the drainage plan as needed. Possible
locations for cross drains are shown in Figure 85.
Ditches and berms serve two primary functions on upland roads: (1) they intercept surface
run-off before it reaches erodible areas, such as fill slopes, and (2) they carry run-off and
sediment to properly designed settling basins during peak flow events (when circumstances
warrant the use of settling basins). Ditches and berms are commonly located at the top of
cut and fill slopes and adjacent to the roadway, although midslope berms may be useful in
controlling sediment on cut and fill slopes before erosion control cover has been
established.
The required depth and cross sectional area of a roadside ditch is determined by the slope of
the ditch, area to be drained, estimated intensity and volume of run-off, and the amount of
sediment that can be expected to be deposited in the ditch during periods of low flow.
Triangular or trapezoidal-shaped ditches may be utilized, whichever is appropriate. The
ditch cross section is designed so that it will produce the desired water velocity for a given
discharge. Minimum full capacity flow velocities should be 0.76 to 0.91 meters/second (2.5
to 3 feet/second) to permit sediment transport. It is best to remember that, in shaping a
ditch, given equal grade and capacity, a wide, shallow cross section will generate lower
water velocities with correspondingly lower erosion potential than will a narrow, deep cross
section. Maximum permissible velocities for unlined ditches of a given soil type are listed
in Table 32.
Maximum permissible
velocities (m/s)
Water
Water
Clear carrying
carrying
water sand and
fine silt
gravel
Fine sand
0.46 0.76 0.46
(noncolloidal)
Sandy loam
0.52 0.76 0.61
(noncolloidal)
Silt loam
0.61 0.91 0.61
(noncolloidal)
Ordinary firm loam 0.76 1.07 0.67
Volcanic ash 0.76 1.07 0.61
Fine gravel 1.76 1.52 1.13
Stiff clay (very
1.13 1.52 0.91
colloidal)
Graded, loam to
cobbles 1.13 1.52 1.52
(noncolloidal)
Graded, silt to
1.22 1.68 1.52
cobbles (colloidal)
Alluvial silts
0.61 1.07 0.61
(noncolloidal)
Alluvial silts
1.13 1.52 0.91
(colloidal)
Coarse gravel
1.22 1.83 1.98
(noncolloidal)
Cobbles and
1.52 1.68 1.98
shingles
Shales and
1.83 1.83 1.52
hardpans
Table 33. Manning's n for open ditches
Mannig's 11
Ditch lining V
n max
ft / meters /
1. Natural earth
sec sec
a. Without vegetation
1) Rock
0.035 -
a) Smooth and uniform 20 6.0
0.040
0.040 - 15 -
b) Jagged & irregular 4.5 - 5.4
0.045 18
2) Soils
Unified USDA
0.022 -
GW Gravel 6-7 1.8 - 2.1
0.024
0.023 -
GP Gravel 7-8 2.1 - 2.4
0.026
Gravel and 0.023 -
d 3-5 0.9 - 1.5
gravelly soils Loamy 0.025
GM
Gravel 0.022 -
u 2-4 0.6 - 1.2
0.020
Gravelly
Loam 0.024 -
Coarse GC 5-7 1.5 - 2.1
Gravelly 0.026
grained
Clay
0.020 -
SW Sand 1-2 0.3 - 0.6
0.024
0.022 -
SP Sand 1-2 0.3 - 0.6
0.024
Sand and 0.020 -
d 2-3 0.6 - 0.9
sandy soils Loamy 0.023
Sand 0.021 -
u 2-3 0.4 - 0.9
0.023
Sandy 0.023 -
SC 3-4 0.9 - 1.2
Loam 0.025
Clay Loam
Sandy 0.022 -
50 CL 2-3 0.6 - 0.9
Clay Loam 0.024
Fine Silty Clay
grained
Silt Loam
Silts and 0.023 -
LL ML Very Fine 3-4 0.9 - 1.2
clays 0.024
Sand Silt
Mucky 0.022 -
50 OL 2-3 0.6 - 0.9
Loam 0.024
0.022 -
CH Clay 2-3 0.6 - 0.9
0.023
0.023 -
MH Silty Clay 3-5 0.9 - 1.5
0.024
LL
Mucky 0.022 -
OH 2-3 0.6 - 0.9
Clay 0.024
0.022 -
Highly Organic PT Peat 2-3 0.6 - 0.9
0.025
2. With vegetation
a. Average turf
0.050 -
1) Erosion resistant soil 4-5 1.2 - 1.5
0.070
0.030 -
2) Easily eroded soil 3-4 0.9 - 1.2
0.050
b. Dense turf
0.070 -
1) Erosion resistant soil 6-8 1.8 - 2.4
0.090
0.040 -
2) Easily eroded soil 5-6 1.5 - 1.8
0.050
0.050 -
c. Clean bottom with bushes on sides 4-5 1.2 - 1.5
0.080
d. Channel with tree stumps
0.040 -
1) No sprouts 5-7 1.5 - 2.1
0.050
0.060 -
2) With sprouts 6-8 1.8 - 2.4
0.080
0.080 -
e. Dense woods 5-6 1.5 - 1.8
0.120
0.100 -
f. Dense brush 4-5 1.3 - 1.5
0.140
0.150 -
g. Dense willows 8-9 2.4 - 2.7
0.200
3. Paved (Construction)
Good
a. Concrete, w/all surfaces:
Poor
0.012 -
1) Trowel finish 20 6.0
0.014
0.013 -
2) Float finish 20 6.0
0.015
0.014 -
3) Formed, no finish 20 6.0
0.016
b. Concrete bottom, float finished, w/sides of:
0.015 - 18 -
1) Dressed stone in mortar 5.4 - 6.0
0.017 20
0.017 - 17 -
2) Ramdom stone in mortar 5.1 - 5.7
0.020 19
0.020 -
3) Dressed stone or smooth concrete rubble (riprap) 15 4.5
0.025
0.025 -
4) Rubble or random stone (riprap) 15 4.5
0.030
c. Gravel bottom, sides of:
0.017 -
1) Formed concrete 10 3.0
0.020
0.020 -
2) Random stone in mortar 8 - 10 2.4 - 3.0
0.023
0.023 -
3) Random stone or rubble (riprap) 8 - 10 2.4 - 3.0
0.033
0.014 -
d. Brick 10 3.0
0.017
3) Asphalt 0.013 - 18 -
5.4 - 6.0
0.016 20
Maximum
recommended
velocities
The procedure for calculating flow rates is the same as that discussed in Section 4.2. The
corresponding roughness factors (Manning's n) for open channels are given in Table 33.
Ditches in highly erodible soils may require riprap, rock rubble lining, jute matting, or grass
seeding. Riprap or rubble-lined ditches will tend to retard flow enough to allow water
movement while retaining the sediment load at low flow periods. Lining ditches can reduce
erosion by as much as 50 percent and may provide economical benefits by reducing the
required number of lateral cross drains when materials can be obtained at low cost.
Ditch water should not be allowed to concentrate, nor should it be allowed to discharge
directly into live streams. A cross drain such as a culvert should carry the ditch water across
and onto a protected surface (Figure 81). Spacing of ditch relief culverts is discussed in
Section 4.4.4 and 4.5.
The ditch grade will normally follow the roadway grade. However, the minimum grade for
an unpaved ditch should be 1 percent. Runoff intensity or discharge values needed to
calculate ditch size can be determined by calculations described below for culvert design.
However, allowances should be made for sedimentation, plus at least 0.3 m between the
bottom of the roadway subgrade and the full flow water surface. The suggested minimum
size of roadside ditches is shown in Figure 87.
Figure 87. Minimum ditch dimensions.
Velocity of the ditch water is a function of cross section, roughness and grade. For a typical
triangular cross section the velocity can be calculated from Manning's equation:
where V equals velocity in meters/second and the other values are as defined in Chapter
4.2. For a triangular channel with sideslopes of 1:1 and 2:1, flowing 0.3 meters deep, the
hydraulic radius, R, equals 0.12 m. Table 34 lists ditch velocities as a function of roughness
coefficients and grade, and Figure 88 provides a nomograph for the solution of Manning's
equation.
In most cases ditch lines should be protected to withstand the erosion. For channels with
grades steeper that 10 percent, a combination of cross section widening, surface protection
and increased surface roughness may be required.
Slope n
(%) 0.02 0.03 0.04
meters/sec
2 1.7 1.2 0.9
4 2.5 1.6 1.2
6 3.0 2.0 1.5
8 3.5 2.3 1.7
10 3.9 2.6 1.9
12 4.3 2.9 2.1
15 4.8 3.2 2.4
18 5.3 3.5 2.6
EXAMPLE:
Determine whether the water velocity for a road ditch will be below critical levels for
erosion. If velocities are too high, make and evaluate changes (see also U.S. Forest Service,
1980). Ditch dimension is a symmetrical, triangular channel, 0.39 m deep with 2.5:1 slopes
with sandy banks (SW) and a slope of 0.003 m/m.
Solution:
R = 0.60 ft
3. Obtain maximum allowable velocity 0.46 to 0.76 m/sec (Table 31). Convert to english
units by dividing by 0.3 m/ft.
4. From Figure 88, find the velocity for the specified ditch (2.9 ft/sec). Convert to metric by
multiplying by 0.3 m/ft.
5. Compare the calculated ditch velocity with the maximum recommended velocity for
sandy channels:
Specified maximum
ditch velocity
0.87 m/sec 0.46 - 0.76 m/sec
The ditch has too great a velocity given the conditions stated above. Therefore, measures
must be taken that will reduce the water velocity. Water velocity in ditches can be reduced
by protecting the channel with vegetation, rock, or by changing the channel shape. (With
vegetative protection, the friction factor (n) becomes 0.030 - 0.050 and the maximum
recommended velocity becomes 0.9 - 1.2 m/sec.)
6. Obtain velocity for specified ditch with vegetative protection by referring to Figure 88
(1.9 feet per second).
7. Compare the calculated ditch velocity with the maximum recommended velocity for
vegetation protected channels (average turf) with easily eroded soils:
Specified maximum
ditch velocity
0.57 m/sec 0.9 - 1.2 m/sec
8. If the specified ditch has a lower velocity than the recommended maximum velocities, it
should be stable as long as the vegetation remains intact.
Berms can be constructed of native material containing sufficient fines to make the berm
impervious and to allow it to be shaped and compacted to about 90 percent maximum
density. Berm dimensions are illustrated in Figure 89.
Water collected in the cutslope ditch line has to be drained across the road prism for
discharge at regular intervals. Cross drains should be installed at a frequency that does not
allow the ditch flow to approach maximum design water velocities. Intercepting dips or
open top culverts (Chapter 4.4.2) perform adequately up to a certain point. However, these
techniques are not adequate or appropriate when the following conditions are present either
in combination or alone:
Ditch relief culverts do not impact or impede traffic as dips and open-top culverts do.
Intercepting dips may become a safety hazard on steep slopes as well as being difficult to
construct. It is also undesirable to have large amounts of water running across the road
surface because of sediment generation and seepage into the subgrade.
The frequency, location and installation method of ditch relief culverts is much more
important than determining their capacity or size. Ditch relief culverts should be designed
so that the half-full velocities are 0.7 to 1.0 m/sec in order to transport sediment through the
culvert and should be at least 45 cm (18 inches) in diameter depending on debris problems.
Larger culverts are more easily cleaned out than narrow ones. Every subsequent relief
culvert should be one size larger than the one immediately upstream from it. This way, an
added safety factor is built in should one culvert become blocked.
As with dips, open top culverts, and water bars, ditch relief and lateral drain culverts should
cross the roadway at an angle greater than or equal to 30° downgrade. This helps insure that
water is diverted from the roadside ditch and that sediment will not accumulate at the inlet.
Accelerated ditch erosion may (1) erode the road prism making it unstable and unusable,
and (2) cause culverts to plug or fail, thereby degrading water quality.
Culvert outlets with no outlet protection are very often the cause of later road failures.
Normally, culvert outlets should extend approximately 30 - 50 cm beyond the toe of the fill.
Minimal protection is required below the outlet for shallow fills. However, on larger fill
slopes where the outlet may be a considerable distance above the toe of the fill, a
downspout anchored to the fill slope should be used (Figure 90). Culvert outlets should be
placed such that at least 50 meters is maintained between it and any live stream. If this is
not possible, the rock lining of the outlet should be extended to 6 meters to increase its
sediment trapping capacity (Figure 91). Coarse slash should be placed near the outlet to act
as a sediment barrier.
Where fills consist entirely of heavy rock fragments, it is safe to allow culverts to discharge
on to the slop. The size and weight of fragments must be sufficient to withstand the
expected velocity of the design discharge. Rock aprons (Figure 92) are the least costly and
easiest to install. A guide for selecting rock for use as riprap is illustrated in Figure 93.
The determination of culvert spacing for lateral drainage across the roadway is based on
soil type, road grade, and rainfall characteristics. These variables have been incorporated
into a maximum spacing guide for lateral drainage culverts developed by the Forest Soils
Committee of the Douglas-fir Region in 1957. The spacing estimates are designed for
sections of road 20 feet wide and include average cut bank and ditch one foot deep. Table 2
(Chapter 1.4.1) groups soils by standard soil textural classes into ten erosion classes having
erodibility indices from 10 to 100, respectively. (Class I contains the most erodible soils
and Class X the least erodible soils.) In order to arrive at an erosion class for a particular
soil mixture, multiply the estimated content of the various components by their respective
erosion index and add the results.
Example:
Total
Name of % Erosion
Erosion
component Content index
Index
rock 20 100 20
Fine Gravel 50 90 45
Silt Loam 30 70 21
86
86 = Erosion
Class VIII
The spacing of lateral-drainage culverts can then be obtained from Table 34. The summary
equation used to calculate values in Table 34, expressed in metric units, is:
Values in Table 34 are based on rainfall intensities of 2.5 to 5 cm per hour (1 to 2 in/hr)
falling in a fifteen minute period with an expected recurrence interval of 25 years. For areas
having greater rainfall intensities for the 25 year storm, divide the values in the table by the
following factors:
Roads having grades less than 2 percent have a need for water removal to prevent water
from soaking the subgrade or from overrunning the road surface. Thus, spacing for roads
with 0.5 percent grades is closer than for roads with 2 percent grades. Usually, local
experience will determine the spacing needed for road grades at these levels.
Field investigations carried out during the route reconnaissance and design stages may not
always reveal subdrainage problems. These less obvious problems can be effectively dealt
with during construction. Field investigations should be carried out during the wet season
and may include soil and/or geologic studies, borings or trenches to locate groundwater,
inspections of natural and cut slopes in the local area, and measurement of discharge when
possible. Sites with potential slope stability problems should be more thoroughly evaluated.
When groundwater tables approach the ground surface, such as in low, swampy areas, the
gradeline should be placed high enough to keep water from being drawn up into the fill by
capillary action. Whenever possible, well graded granular materials, such as coarse sand,
should be used for fill construction. For a detailed discussion of grading requirements for
filter materials the reader is referred to the Earth Manual published by the U.S. Department
of the Interior (1974).
(1) Pipe underdrains. This system consists of perforated pipe placed at the bottom of a
narrow trench and backfilled with a filter material such as coarse sand. It is generally used
along the toes of cut or fill slopes. The trench should be below the groundwater surface and
dug into a lower, more inpervious soil layer to intercept groundwater. The drains may be
made of metal, concrete, clay, asbestos-cement, or bituminous fiber and should be 15
centimeters (6 inches) in diameter or larger.
(2) Drilled drains. This system consists of perforated metal pipes placed in holes drilled
into cut or fill slopes after construction.
(3) French drains. This system consists of trenches backfilled with porous material, such
as very coarse sand or gravel. This type of drain is apt to become clogged with fines and is
not recommended.
A major difficulty in selecting a drainage system is the lack of adequate performance data
for various drainage methods. A good knowledge of seasonal groundwater fluctuations,
variation in lateral and vertical permeability, and the ratio of vertical to lateral permeability
are critical. Long term monitoring of drainage performance is important in determining
appropriate prescriptions for future installations. For example, perforated drains are
commonly prescribed but often will not function properly as a result of clogging of pores
with fines or from geochemical reactions leading to the formation of precipitates. Several
methods may be used to prevent plugging depending on soil characteristics and material
availability. The first is to enclose the perforated pipe with geotextile fabric. Second,
surround the pipe with an open graded aggregate material, which in turn is surrounded by a
fabric. The use of fabric eliminates the need for an inverted filter consisting of various sized
gravel and sand layers. Third, if fabric is not available, surround the pipe with a graded
aggregate filter. Although the cost of installing such a drainage system is high, it may
effectively reduce final road costs by decreasing the depth of base rock needed, thereby
reducing subgrade widths and associated costs for clearing, excavating, and maintenance.
LITERATURE CITED
Amimoto, P. Y. 1978. Erosion and sediment control handbook. California Division of
Mines and Geology, Department of Conservation. 197 p.
Beschta, R. L. 1981. Streamflow estimates in culverts. Oregon State University, Forest
Researph Laboratory, Res. Note. 67. 4 p.
Darrach, A. G., W. J. Sauerwein, and C. E. Halley. 1981. Building water pollution control
into small private forest and ranchland roads. U. S. Department of Agriculture, Forest
Service and Soil Conservation Service.
Forest Soils Committee of the Douglas Fir Region. 1957. An introduction to the forest soils
of the Douglasfir Region of the Pacific Northwest. University of Washington.
Highway Task Force. 1971. Handbook of steel drainage and highway construction products
(2nd Ed). American Iron and Steel Institute, 150 E 2nd Street; New York. 368 p.
Megahan, W.F. 1977. Reducing erosional impacts of roads. In: Guidelines for Watershed
Management. Food and Agriculture Organization, United Nations, Rome. p 237-261.
Minematsu, H. and Y. Minamikata, 1983. Optimum spacing for open - top culverts on
forest roads. University of Agriculture and Technique, Tokyo. Jour. of J.F.S. 65(12):465-
470.
Packer, P. 1967. Criteria for designing and locating logging roads to control sediments.
Vol.1, No. 13.
Reid, L.M. 1981. Sediment production from gravel-surfaced forest roads, Clearwater basin,
Washington. Publ. FRI-UW--8108, Univ. of Washington, Seattle. 247 p.
Searcy, J. K. 1967. Use of riprap for bank protection. Federal Highway Administration,
Washington D. C. 43 p.
________________. 1979. Technical guide, erosion prevention and control on timber sales
areas. Intermountain Region.
Yee, C. S. and T. D. Roelofs. 1980. Planning forest roads to protect salmonid habitat. U. S.
Department of Agriculture, Forest Service. General Technical Report PNW-109. 26 p.
U.S. Dept. of Commerce, Bureau of Public Roads. 1963. Hydraulic Charts for the Selection
of Highway Culverts, Hydraulic Engineering Circular No. 5.
U.S. Dept. of the Interior, Bureau of Reclamation. 1974. Earth Manual, a water resources
technical publication. Second edition. Government Printing Office, Washington D.C.