10.

HIGHWAY DRAINAGE
Provision of adequate drainage is an important factor in the location and geometric design of
highways. Drainage Facilities on any highway or street should adequately provide for the
flow of water away from the surface and subsurface of the pavement to properly designed
channels and then discharge to the natural waterways. Inadequate drainage will eventually
result in:

i) Serious damage to highway structure

 Reduce pavement strength
 Swelling heave
 Stripping of asphalt
 Cause pumping in rigid pavements
 Frost heave and reduction of bearing capacity when melting
ii) Traffic operation problems
 Slow traffic movement by accumulated water on the pavement
 Cause traffic accidents as a result of hydroplaning and loss of
visibility from the splash and spray

The importance of adequate drainage is well recognized in highway engineering and about 25
percent of highway construction cost is estimated to be spent for erosion control and drainage
structures, such as culverts, bridges, channels, and ditches.

There are two sources of water that the highway engineer is primarily concerned to provide
highway drainage facilities. The first source, surface water, is that which occurs as rain or
snow. Drainage for runoff water on the surface from this source is referred to as surface
drainage. The second source, ground water, is that which flows in underground streams. This
may become important in highway cuts or at locations where a high water table exists near the
pavement structure. Drainage for seepage water and underground streams is referred to as
subsurface drainage.

In this chapter, the fundamental design principles for surface and subsurface drainage
facilities are presented. The principles of hydrology necessary for understanding the design
concepts are also included, together with a brief discussion on erosion prevention.

10.1. Surface Drainage

Surface drainage encompasses all means by which surface water is removed from the
pavement and right of way of the highway or street. A properly designed highway surface
drainage system should effectively intercept all surface and watershed runoff and direct this
water into adequately designed channels and gutters for eventual discharge into the natural
waterways. Water seeping through cracks in the highway riding surface and shoulder areas
into underlying layers of the pavement may result in serious damage to the highway
pavement. The major source of water for this type of intrusion is surface runoff. An
adequately designed surface drainage system will therefore minimize this type of damage.
The surface drainage system includes:

1
i) Transverse Slopes: - These include crowning of pavements on either side of the
centreline or providing a slope in one direction across the pavement width and
crowning of shoulders away from the pavement to facilitate the removal of surface
water from the pavement surface in the shortest possible time.

ii) Longitudinal Slopes: - A minimum gradient in the longitudinal direction of the

highway is required to obtain adequate slope in the longitudinal channels,
particular at cut sections. Slopes in longitudinal channels should generally not be
less than 0.2 percent for highways in very flat terrain. Although zero percent
grades may be used on uncurbed pavements with adequate cross slopes, a
minimum of 0.5 percent is recommended for curbed pavements. This may be
reduced to 0.3 percent on suitably crowned high-type pavements constructed on
firm ground.

iii) Longitudinal Channels: - These are ditches constructed along the sides of the
highway to collect the surface water that runs off from the pavement surface,
subsurface drains, and other areas of the highway right of way. When the highway
pavement is located at a lower level than the adjacent ground, such as in cuts,
water is prevented from flowing onto the pavement by constructing a longitudinal
drain (intercepting drain) at the top of the cut to intercept the water. The water
collected by the longitudinal ditches is then transported to a drainage channel and
then to a natural waterway or retention pond.

iv) Curbs and Gutters: - Curbs and gutters are facilities more frequently in urban
areas, particularly in residential areas, where they are used in conjunction with
storm sewer systems to control street runoff in addition to other functions, which
include preventing the encroachment of vehicles on adjacent areas and delineating
pavement edges. When it is necessary to provide relatively long continuous
sections of curbs in urban areas, the inlets to the storm sewers must be adequately
designed for both size and spacing so that the impounding of large amounts of
water on the pavement surface is prevented.

v) Cross-Drainage Structures: - Cross-drainage structures are constructed to carry

traffic over natural waterways that flow below the right of way of the highway.
These structures also provide for the flow of water below the highway, along the
natural channel, without significant alteration or disturbance to its normal course.
One of the main concerns of the highway engineer is to provide an adequate size
structure, such that the waterway opening is sufficiently large to discharge the
expected flow of water. Inadequately sized structures can result in water
impounding, which may lead to failure of the adjacent sections of the highway due
to embankments being submerged in water for long periods.

Cross-drainage structures can be categorized into major and minor structures.

Major structures are those with clear spans greater than 6 m (20 feet), whereas
minor structures are those with clear spans of 6 m (20 feet) or less. Major
structures are usually large bridges, although multiple-span culverts may also be
included in this class. Minor structures include small bridges and culverts.

2
 vi) Sediment and Erosion Controls: - Continuous flow of surface water over
shoulders, side slopes, and unlined channels often results in soil erosion which can
lead to conditions that are detrimental to the pavement structures, embankments
and cut sections. The methods used to prevent erosion and control sediment
include:

 Intercepting drains at the top of a cut to collect and transport runoff to paved
spillways that are placed at strategic locations on the side of the cut and then to
the longitudinal ditches alongside the highway.
 Curbs and gutters can be used to prevent unsurfaced shoulders and
embankment slopes from erosion and then direct surface water into paved
spillways.
 Turf cover on unpaved shoulders, ditches, embankments, and cut slopes is an
efficient and economic method of preventing erosion when slopes are flatter
than 3 to 1. The turf cover is commonly developed by sowing suitable grasses
immediately after grading.
 Slope and channel linings are more effective prevention of erosion than any of
those already described. For example, when cut and embankment side slopes
are steep and are located in mountainous areas subjected to heavy rain or snow
a commonly used method is to line the slope surface with rip-rap or hand-
placed rock. Channel linings are also used to protect longitudinal channels
from eroding. Protective linings are of two types: flexible (dense-graded
bituminous mixtures and rock rip-rap) and rigid (cement concrete and soil
cement). Rigid linings are much more effective in preventing erosion under
severe conditions, but they are more expensive and, because of their
smoothness, tend to create high unacceptable velocities at the end of the
linings.
The design of surface drainage systems for a highway may be divided into three major phases:

(1) An estimate of the quantity of water that may be expected to reach any element of
the system;

(2) The hydraulic design of each element of the system; and

(3) The comparison of alternative systems, alternative materials, and other variables in
order to select the most economical system that can be devised. In the third phase,
attention must be given to selecting the system that has the lowest annual cost
when all variables are taken into consideration.

10.1.1. Flood Estimation

A variety of approaches have been used to estimate the quantity of runoff for drainage design.
When a drainage structure is to handle the flow of an existing stream such as in the case of
some culverts and most bridges, the flow used for hydraulic design may be based on available
records. For such “gauged" sites, statistical analyses can be performed on the recorded stream

3
flow to provide an estimated peak design flow for a given "return period." The term "return
period" refers to the estimated frequency of rare events such as floods. Selection of the
frequency of occurrence of the design storm is largely a matter of experience and judgment,
although the concerned road agency may have an established interval to be used for a given
situation. The return period is a statistical matter. For example, if the system is designed for a
return period of 25 years, the statistical assumption is that the system will accommodate the
most severe storm to occur once in 25 years. It is apparent that selection of a return period of
100 years instead of 25 would mean designing for a more severe storm and, in general, a more
costly system. Conversely, if the frequency is 10 years, the intensity of the design storm will
be less and in most cases a less costly drainage system will result, although economic losses
from use of the short time might offset the savings in construction costs.

The estimation of peak runoff for drainage design is accomplished by consideration of severe
storms that occur at intervals and during which the intensity of rainfall and runoff of surface
water are far greater than at other periods. Rainfall intensity during the design storm is a
function of occurrence, duration, and intensity. As shown in Figure 10- 1, the intensity of
rainfall for a particular return period varies greatly with the duration of rainfall. An accurate
estimate of the probable intensity, frequency, and duration of rainfall in a particular location
can be made only if sufficient data have been collected over a period of time. If such
information is available, standard curves may be developed to express rainfall-intensity
relationships with accuracy sufficient for drainage problems.

Figure 10- 1. Typical rainfall intensity-duration curves (TCDE)

4
The available methods of flood estimation may be categorized into three groups: analysis of
stream flow data, runoff modeling, and regionalized flood formula, but here rational method
which is under the category of runoff modeling will be presented

Rational method: The rational method is a method most widely used rainfall runoff relation
for ungauged areas. It is most suitable for small catchments of sizes up to five square
kilometres. The basic form of the equation is:

CIA
Q
360

where: Q = flood peak at catchment exit (m3/sec);

C = rational runoff coefficient;
I = average rainfall intensity over the whole catchment (mm/hr) for a duration
corresponding to the time of concentration; and
A = catchment area in hectares.

The time of concentration is defined as the time required for the surface runoff from the
remotest part of the drainage basin to reach the point where the drainage facility is located. It
depends on several factors including the size and shape of the drainage area, the type of the
surface cover, the slope of the drainage catchment, the rainfall intensity, and whether the flow
is entirely overloaded or partly channalized.

The time of concentration generally consists of one or more of the following three
components: time for overland flow, time for gutter or stream sewer flow (urban), or time for
channel flow. Travel time is the ratio of flow length to average flow velocity. Mathematically,

L
, and Tc  i 1 Ti
m
Ti 
V

in which Ti is the travel time for section i, with length L (km), V is average flow velocity
(km/hr) and Tc is time of concentration (in hr). Alternatively, the time of concentration can be
calculated by the Kirpich formula:

0.385
 0.87  L2 
Tc   
 1000  S 

where, Tc, = time of concentration in hours; L = length of main stream (km); S = average
slope of main stream (mm-1). Having determined the time of concentration the corresponding
rainfall intensity can then be obtained from the intensity-duration curve for the selected
recurrence interval (return period) Figure 10- 1. In absence of rainfall intensity duration
curves, simple empirical relationships of the form

5
 a
I
b  T n

in which I is rainfall intensity in mm/hr, T is rainfall duration in hrs, and a, b, and n are
constants to be determined.

The drainage catchment area (the land that contributes to the runoff) is normally determined
from a topographic map. The runoff coefficient, C, is an integrated value representing the
ratio of runoff to rainfall for the drainage area. It depends on the type of ground cover, the
slope of the drainage area, storm duration, and prior wetting. The runoff coefficient can be
estimated from Figure 10- 1. In case where the drainage area consists of different ground
characteristics with different runoff coefficients, a representative value C w is computed by
determining the weighted coefficients as:

C A i i
Cw  i 1
n

A
i 1
i

10.1.2. Hydraulic Design of Surface Drainage Structures

The objective of the hydraulic design for any highway drainage structure is to provide a
suitable structure size that will economically and efficiently dispose of the expected runoff
without detrimental erosion and sedimentation problem.

Design of Side Ditches

With the quantity of water expected to reach any given point in the drainage system known,
the design of side ditches, gutters, stream channels, and similar facilities is based on
established principles of flow in open channels. The principles also apply to flow in conduits
with a free water surface. In the design of open channels, an important design consideration is
the flow velocity. The flow velocity in the channel should not be as low as to cause deposits
of transported material, nor so high as to cause erosion of the channel. The most appropriate
channel gradient range to produce the required velocity is between 1 percent and 5 percent.

For most types of linings, sedimentation is usually a problem when slopes are less than 1
percent, and excessive erosion of the lining will occur when slopes are higher than 5 percent.
Consideration of recommended maximum velocities that can be found in different standards
and literatures for particular-lining materials is vital for preventing erosion problems.
Attention should also be paid to the point at which the channel discharges into the natural
waterway. For example, if the drainage channel at the point of discharge is at a much higher
elevation than the natural waterway, then the water should be discharged through a spillway
or chute to prevent erosion.

The hydraulic design of an open channel drainage facility for a given storm entails the
determination of the minimum cross-sectional area of the channel that will accommodate the

6
flow due to that storm and prevent water from overflowing the facility. The most commonly
used formula to give reliable capacity estimate is Manning's formula, which assumes uniform
steady flow in the channel:

R 2 / 3 S 1/ 2
Q  Av  A
n

where, Q = capacity (m3/sec); A = channel cross-sectional area (m2); v = mean velocity

(m/sec); n = Manning roughness coefficient; R = hydraulic radius (A/P) in m; S = slope
(m/m); and P = wetted perimeter (m).

The depth of flow in a uniform open channel depends on the shape of the cross-section, the
roughness, and the slope of the channel, and the amount of discharge. For example, water
flowing down a relatively flat slope in an open channel is in tranquil flow, while that falling
down a steep slope is rapid. In the same channel, with the same quantity of flow, the flow can
be changed from tranquil to rapid by an increase in slope. In such a case, the depth of flow
decreases from that existing in the section of tranquil flow to a lower value in the section of
rapid flow. The decrease in depth begins at some point ahead of the crest of the steep slope
and continues gradually over some distance. In the section of rapid flow, since the quantity of
flow is the same, the dimensions of the channel the same and the depth less, it follows from
the equation of continuity that the velocity is greater.

Critical depth is defined as the depth corresponding to the change from tranquil to rapid flow.
Critical velocity and critical slope are the velocity and slope that correspond to uniform flow
at critical depth. Critical depth is independent of channel slope and roughness, but the critical
slope is a function of the slope of the channel. Tranquil flow exists when the normal depth of
water in an open channel is greater than the critical depth; conversely, when the depth is less
than critical, the flow is rapid. Theoretically, in any channel in which the water is flowing at
the critical depth, the velocity head (V2/2g, where V is the velocity of the flow and g is the
acceleration due to gravity) is equal to one-half the mean depth of flow (mean depth is
defined as the water cross-sectional area, A, divided by the free-water surface width, T). In
other words, at critical flow,

Q 2 A3

g T

Design Procedure: Hydraulic design procedures are difficult to simplify because of the wide
variety of choices presented to the designer in a typical case and the various assumptions that
must be made. A simple design procedure that utilizes the Manning equation may, however,
be outlined as follows:

1. For the cover material in which the ditch will be running, select the Manning
roughness coefficient, n, the side slopes – normally controlled by the angle of
natural repose.
2. Using the Manning formula, calculate the maximum permissible hydraulic radius.

7
 3. Using the equation of continuity, calculate the minimum permissible cross-section
area required from the given discharge and permissible velocity. Calculate the
wetted perimeter for this area.
4. Solve the expressions obtained in the step (3) simultaneously for the bottom width
of the ditch and the depth of flow.
5. Check the depth of flow whether it is greater than the critical depth or not.
6. If the depth of flow is greater than the critical depth, add a suitable freeboard and
modify the section for practicality. In the case of side ditches the freeboard should
at least be equals the height above the bottom of the pavement. For other channels,
a value about 0.5 m is added as a freeboard.
7. If the depth is less than the critical depth consideration should be given to reduce
the slope or provision should be sought to protect from erosion.

Design of Culverts
Depending on the class of highway, the volume of the stream flow to be crossed, the site
conditions, and economic factors, the flowing cross-drainage structures may be considered:

 Fords
 Drifts
 Culverts, and
 Bridges

Fords are the simplest river crossings that utilize the existing riverbed and are appropriate for
shallow, slow moving watercourses with little probability of flood. The traffic volume may be
up to about 100 vehicles per day. Gravel or stones can be used to line the bottom of the ford to
provide a firm footing for vehicles. Drifts are crossing structures as fords, but here a concrete
slab is constructed where the bed of the river is not able to carry vehicles. They are suitable as
a crossing where fords cannot be used due to risk of flood.

Culverts are cross-drainage structures that convey water from streams and side channels
below the road. A culvert is usually, although not always, differentiated from a bridge by
virtue of the fact that the top of the culvert does not form part of the travelled roadway. More
frequently, culverts are differentiated from bridges on the basis of span length. On an arbitrary
basis, structures having a span of 6 m or less will be called culverts, whereas those having
spans of more than 6 m will be called bridges. Culverts also differ from bridges in that they
are usually designed to flow full under certain conditions, while bridges are designed to pass
floating debris or vessels. Culverts are to be found in three general locations: at the bottom of
depressions where no natural watercourse exists; where natural streams intersect the roadway;
and at locations required for passing surface drainage carried in side ditches beneath roads
and driveways to adjacent property.

Principles of Culvert Location. The majority of culverts are installed in natural watercourses
that cross the roadway, either at right angles or on a skew. In addition to selecting the proper
location for the culvert crossing with respect to the centreline of the road, the alignment and
grade of the culvert are of importance. The location of the centreline of the culvert on the
centerline of the road may be determined by inspection of the plans or in the field. This

8
location will generally be on the centreline of an existing watercourse or at the bottom of a
depression if no natural watercourse exists. The alignment of the culvert should generally
conform to the alignment of the natural stream, and the culvert should, if possible, cross the
roadway at right angles in the interests of economy. Skew culverts, located at an angle to the
centreline of the road, are needed in many instances.

Hydraulic Design of Culverts. Earlier in this chapter, the concepts and procedures related to
estimation of the quantity of runoff from a drainage basin was presented. In this section, we
will briefly discuss principles and techniques for the hydraulic design of culverts. A more
complete treatment of the subject is assumed to be dealt in hydraulics and related courses. The
purpose of hydraulic design is to provide a drainage facility or system that will adequately and
economically provide for the estimated flow throughout the design life without unreasonable
risks to the roadway structure or nearby property. Hydraulic design of culverts involves the
following general procedure:

1. Obtain all site data and plot a roadway cross section at the culvert site,
including a profile of the stream channel.
2. Establish the culvert invert elevations at the inlet and outlet and determine the
culvert length and slope.
3. Determine the allowable headwater depth and the probable depth of tail water
during the design flood.
4. Select a type and size of culvert that will accommodate the design flow under
the established conditions.
5. Examine the need for energy dissipaters, and, where needed, provide
appropriate protective devices to prevent destructive channel erosion.

Whenever a constriction such as a culvert is placed in a natural open channel, there is an

increase in the depth of water just upstream of the constriction. The allowable level of the
headwater upstream of the culvert entrance is generally the principal control on the culvert
size and inlet geometry. The allowable headwater depth depends on the topography and the
nature of land use in the culvert vicinity. In establishing the headwater depth the designer
should consider possible harmful effects that flooding may cause, such as damage to the
pavement, interruptions to traffic, and inundation of nearby property.

The type of flow in a culvert depends on the total energy available between the inlet and
outlet, primarily of the difference in the headwater and the tail-water elevations. The flow
characteristics and capacity of a culvert are determined by the location of the control section.
Laboratory tests and field studies have shown that highway culverts operate with two major
types of controls: inlet and outlet controls.

Under inlet control, the discharge capacity of a culvert depends primarily on the depth of
headwater at the entrance and the entrance geometry (barrel shape, cross-sectional area, and
type of inlet edge). Inlet control commonly occurs when the slope of the culvert is steep and
the outlet is not submerged. In addition to the factors mentioned in inlet control culverts, the
maximum flow in a culvert operating with outlet control depends on the tail-water at the
outlet and the slope, roughness, and length of the culverts. It occurs on flat slopes, especially
where downstream condition causes the tail-water depth to be greater than the critical depth.

9
On the basis of experimental work, analytical relationships have been developed for analysis
or design of culverts. These relationships are complex and the analysis or design of a culvert
is tedious even with the use of monographs and charts. Several computer and hand calculator
programmes are presently available that can be used to increase the accuracy of the results
and significantly reduce the time it takes for design or analysis.

Culvert Types and Materials. Materials most commonly used in the construction of culverts
are reinforced concrete and corrugated metal. Less frequently, culverts are made from timber,
cast-iron pipe, vitrified-clay pipe, and, occasionally, stone masonry. Reinforced-concrete pipe
intended for use in culverts is made in diameters of 300 to 3600 mm and in various lengths,
the usual length being 1.2 to 2.4 m. Reinforced-concrete culverts may be either circular or
elliptical. For special applications, concrete culvert pipe may be manufactured with a cross
section other than circular-elliptical and "arch" shapes being in quite common use. Concrete
box culverts are constructed in place with square or rectangular cross sections; single box
culverts vary in size from 0.6 to 3.6 m square, depending on the required area of waterway
opening. Both concrete pipe and concrete box culverts are built with more than one opening
where additional waterway area is required and when it is desired to avoid the use of
excessively large single pipes or boxes. Such installations are called "multiple culverts" and
may, for example, be "double" or "triple" concrete pipe or concrete box culverts. Concrete
arches are sometimes used in place of concrete box culverts, although difficulties attendant on
their proper design and construction have somewhat restricted their use.

Corrugated steel of various thicknesses is used in various forms in the construction of culverts
for use in highway drainage. Corrugated metal (galvanized steel) pipe is made in diameters of
200 to 2440 mm and in lengths of 6 to 12 m. In many culvert installations headroom is
limited, and a circular pipe that has sufficient hydraulic capacity is not suitable. In such cases,
corrugated metal "pipe arches" may be used. Pipe arches made of standard corrugated metal
are available in sizes varying from a span of 460 mm and a rise of 280 mm to a span of 1830
mm and a rise of 1120 mm. Other materials that may be used in the construction of culverts
include vitrified-clay pipe, cast-iron pipe, and timber. In addition, masonry arch culverts are
constructed in localities where suitable stone is cheaply available. Each of these types of
culvert has advantages and may be used in areas where the material is economically available
and for special-purpose structures.

The type of culvert selected for use in a given location is dependent on the hydraulic
requirements and the strength required to sustain the weight of a fill or moving wheel loads.
After the hydraulic and strength requirements are satisfied, the selection is largely a matter of
economics: durability and cost of the completed structure including cost of transport and
installation.

Highway Bridges
The term "bridge" is usually associated with structures that are required to carry the roadbed
over an established waterway; it may also be somewhat loosely applied to grade separation
structures and elevated highways in urban areas (viaducts). In some cases, bridges of
relatively short span are located to conform with the general location of the highway, which
has been previously determined. That is, the tentative location for the highway is established
after an analysis of all the economic and engineering factors involved, and the bridge engineer
is given the problem of providing an economical and adequate bridge design to conform to the
roadway location.

10
Generally, however, the location of a suitable stream crossing may be the most important
single factor influencing the location of the highway in a given section; such is usually the
case when long bridge spans are involved. The ideal location for a bridge crossing is, of
course, one in which the crossing is made at right angles to the centerline of the stream at the
narrowest point, where the alignment of the approach pavement is straight, where the
approach grade is slight, and where soil conditions are adequate for the installation of the
most economical foundation for the span involved. This ideal combination of circumstances is
encountered all too infrequently, except in structures of short span, and many bridges have
been located on skew crossings, vertical curves, or with curving alignment. In such cases,
considerations related to the general roadway location may still be regarded as controlling
factors, and the required adjustments in the location of the bridge are made with these
requirements in mind.

Many times, alternative locations of a proposed bridge may seem to offer somewhat similar
advantages. A careful comparison must then be made of the several possible locations. The
final decision should be based on a complete analysis, including factors related to traffic
safety and operating conditions, fulfillment of the purpose of the road (e.g., the direct
connection of population centres), and economics. Any complete analysis must include both
the bridge and the approaches to it. A comparison of this type will generally result in the
selection of one of the possible sites as the most desirable.

Although the preceding paragraphs have dealt largely with waterway crossings, similar
factors govern the design of grade separation and the bridge portions of complex
interchanges. Obviously, a grade separation structure designed to carry the roadway over an
existing railroad presents the location engineer and designer with similar problems, as does
the design of a highway overpass. Similarly, the general location of an elevated highway in an
urban area is usually determined on the basis of maximum serviceability to traffic (and
availability of right-of-way) so that the engineer must prepare a design suitable to conditions
in a rather limited area.

In many locations the natural stream channel is somewhat constricted by the bridge structure
and roadway approaches. In the interests of economy, the road way is frequently placed on an
embankment on either side of the bridge span; the distance between abutments is reduced as
much as possible, and piers may be placed in the stream channel. All these things serve in
many cases to reduce severely the area through which the water must pass, particularly when
the stream is at flood stage. Two results may immediately be noticed: during flood stage the
velocity of the water through the bridge opening may be considerably increased, with
resultant danger to the bridge structure through scour at abutments and piers, and the elevation
of the water on the upstream side may be increased, with the result that the area subjected to
flooding above the bridge site is increased and adjacent property is subjected to overflow
beyond the limits of the normal floodplain. It thus seems axiomatic that the bridge must be
designed to pass the flow occurring at flood stage without excessive velocity and without
damage to property located above the bridge crossing. Bridge openings are also normally
designed to pass floating debris carried in the channel at normal and flood states. On
navigable streams, requirements of navigation must be evaluated and provided for. Generally
speaking, navigable streams are spanned by high-level crossings or movable bridges.

Standards related to bridge clearances, both vertical and horizontal, are an important part of
the design of highway bridges. AASHTO specifies that the roadway width at bridges shall be

11
equal to the full shoulder width of the approach section. Along curbed roadways, the full
width of the approach section should similarly be crossed the structure. For low-speed, low-
volume roads, a minimum horizontal clearance of the width of the approach travelled way
plus 2.4 m is recommended. The vertical clearance of a highway bridge should be at least 4.3
m over the entire roadway. Along main highways, a 4.9 m vertical clearance is usually
provided.

10.2. Subsurface Drainage

Subsurface drainage systems are provided within the pavement structure to drain water that:

 has percolated through cracks and joints in the pavement to the underlying
strata
 has moved upward through the underlying soil strata as a result of capillary
action
 exists in the natural ground below the water table, usually referred to as ground
water.
The design of subsurface drainage should be carried out as an integral part of the complete
design of the highway, since inadequate subsurface drainage have detrimental effects on the
stability of slopes and pavement performance. The procedure usually adopted for subsurface
drainage design is first to determine the geometric and structural requirements of the highway
based on standard design practice, and then to subject these to a subsurface drainage analysis
to determine the subsurface drainage requirements. It is extremely difficult, if not impossible,
to develop standard solutions for solving subsurface drainage problems because of the many
different situations that engineers come across in practice. The experience gained from field
and laboratory observations for a particular location, coupled with good engineering
judgment, should always be used in conjunction with the design tools provided.

Subsurface drainage systems are usually classified into five general categories:

 Longitudinal drains
 Transverse drains
 Horizontal drains
 Drainage blankets
 Well systems

Longitudinal drains. Subsurface longitudinal drains usually consist of pipes laid in trenches,
within the pavement structure and parallel to the centerline of the highway. These drains can
be used to lower the water table below the pavement structure, as shown in Figure 10- 2, or to
remove any water that is seeping into the pavement structure, as shown in Figure 10- 3.

Transverse drains. Transverse drains are placed transversely below the pavement, usually in
a direction perpendicular to the centreline, although they may be skewed to form a
herringbone configuration. An example of the use of transverse drain is shown in Figure 10-4
where they are used to drain ground water that has infiltrated through the joints of the
pavement.

12
Figure 10- 2. Symmetrical longitudinal drains used to lower the ground water level

Figure 10- 3. Longitudinal collector drains used to remove seeping water

Figure 10- 4. Transverse drains used on superelevated curves

13
Horizontal Drains. Horizontal drains are used to relieve pore pressures at slopes of cuts and
embankments on the highway. They usually consist of small-diameter, perforated pipes
inserted into the slopes of the cut or fill. The subsurface water is collected by the pipes and is
then discharged at the face of the slope through paved spillways to longitudinal ditches.

Drainage Blankets. A drainage blanket is a layer of material that has a very high coefficient
of permeability, usually greater than 10 m/day, and is laid beneath or within the pavement
structure such that its width and length in the flow direction are much greater than its
thickness. The coefficient of permeability is the constant of proportionality of the relationship
between the flow velocity and the hydraulic gradient between two points in the material.
Drainage blankets can be used to facilitate the flow of subsurface water away from the
pavement, as well as to facilitate the flow of ground water that has seeped through cracks into
the pavement structure or subsurface water from artesian sources. A drainage blanket can also
be used in conjunction with longitudinal drains to improve the stability of cut slopes by
controlling the flow of water on the slopes, thereby preventing the formation of a slip surface.
However, drainage blankets must be properly designed to be effective. Figure 10- 5 shows
two drainage blanket systems.

Figure 10- 5. Application of horizontal drainage blankets

Well Systems. A well system consists of a series of vertical wells, drilled into the ground, into
which ground water flows, thereby reducing the water table and releasing the pore pressure.

14
When used as a temporary measure for construction, the water collected in the wells is
continuously pumped out, or else it may be left to overflow.

10.2.1. Design of Subsurface Drainage

The design procedure for subsurface drainage involves the following:

1. Summarize the available data.

2. Determine the quantity of water for which the subsurface drainage system is being
designed.
3. Determine the drainage system required.
4. Determine the capacity and spacing of longitudinal and transverse drains and select
filter material, if necessary.
5. Evaluate the design with respect to economic feasibility and long-term performance.

Summarize Available Data

The data that should be identified and summarized can be divided into the four following
classes:

 The flow geometry  the flow geometry is given by the existing subsurface
characteristics of the area in which the highway is located and by the geometric
characteristics of the highway.
 The materials' properties  the material permeability that indicates the extent
to which water will flow through the material.
 The hydrologic and climatic characteristics  indicate precipitation rates, the
sources of subsurface water, and the possibility of frost.
 Miscellaneous information  all other information that will aid in the design
of an effective and economic subsurface drainage system, including any impact
the subsurface drainage system may have on future construction.

Determination of Discharge Quantity

The net amount of water to be discharged consists of the following components:

 Water due to infiltration

 Ground water
 Water due to melting of ice, but not a significant problem in the tropics
 Water evaporating vertically from the pavement structure

(i) Water due to infiltration, (qi). This is the amount of surface water that infiltrates into the
pavement structure through cracks in the pavement surface. It is extremely difficult to
calculate this amount of water exactly, since the rate of infiltration depends on the intensity of
the design storm, the frequency and size of the cracks and/or joints in the pavement, the
moisture conditions of the atmosphere, and the permeability characteristics of the materials
below the pavement surface. The Federal Highway Administration recommends the use of the
following empirical relationship to estimate the infiltration rate:

15
 N Wc 
qi  I c  c    kp
 W W Cs 

where, qi = design infiltration rate (ft3/day/ft2 of drainage layer).

Ic = crack infiltration rate (ft3/day/ft of crack); Ic = 2.4 ft3/day/ft is recommended for
most designs, but with local experience this value may be increased or decreased
as necessary.
Nc = number of contributing longitudinal cracks or joints; N + 1 for new pavements,
where N is the number of traffic lanes.
Wc = length of contributing transverse cracks (ft).
W = width of granular base or subbase subjected to infiltration (ft).
Cs = spacing of the transverse cracks or joints (ft); a value of 40 has been suggested
for new bituminous concrete pavements, but local experience should be used.
Kp = rate of infiltration (ft3/day/ft2), coefficient of permeability through the uncracked
pavement surface, the value for cement concrete and well compacted dense
graded asphalt concrete pavements is very low and can therefore be zero, but it
should be determined from laboratory tests.

(ii) Ground Water. When it is not possible to intercept the flow of ground water or lower the
water table sufficiently before the water reaches the pavement, it is necessary to determine the
amount of ground water seepage that will occur. Figure 10- 2 and Figure 10- 5(a) demonstrate
the two possible sources of ground water of interest in this case. A simple procedure to
estimate the ground water flow rate due to gravity drainage is to use the chart shown in Figure
10- 6. In this case the radius of influence Li is first determined as:

Li = 3.8 (H – H0)

Where, H0 = thickness of subgrade below the drainage pipe (ft)

H = thickness of subgrade below the natural water table (ft)
H – H0 = amount of draw down (ft)

The chart shown in Figure 10- 6 is then used to determine the volume of lateral flow per unit
length of the longitudinal drain (qL = q1 + q2), where q1 is the inflow above the bottom of the
drainage layer, and q2 is the total upward flow into one-half of the drainage layer. The average
inflow rate to the drainage layer is determined as:

q2
qg 
0 .5 W

where, qg = design inflow rate for gravity drainage (ft3/day/ft2of drainage layer)
q2 = total upward flow into half of the drainage blanket (ft3/day/linear ft of roadway)

16
 W = width of drainage layer (ft).

If the pavement is sloped to one side and the collector is installed only on one side, then
q1  2q 2
q L  2(q1  q2 ) , and qg 
W
For the case of artesian flow, the average inflow rate is simply estimated using darcy’s law as:
H
qa  K
H0
where, qa = design inflow rate from artesian flow (ft3/day/ft2)
H = excess hydraulic head (ft),
H0 = subgrade soil thickness between the drainage layer and the artesian aquifer (ft)
K = coefficient of permeability (ft2/day)

Figure 10- 6. Chart for determining flow rate in horizontal blanket

(iii) Vertical outflow, or evaporation (qv). In some cases, the total amount of water
accumulated within the pavement structure can be reduced because of significant evaporation
of some of the accumulated water through the subgrade. When this occurs, it is necessary to
estimate the amount of this outflow in order to determine the net inflow for which the
subsurface drainage system is to be provided. The procedure for estimating this flow involves
the use of flow net diagrams, which is beyond the scope of this material.

17
Net design flow. The net design inflow is the sum of inflow rates from all sources less any
amount attributed to vertical outflow through the underlying soil. A set of relationships for
estimating the net design inflow rate (q d) has been developed, taking into consideration the
different flows that occur concurrently. In the absence of inflow due to melting of ice lenses,
the design net inflow can be given as:

1. If there is no frost, no ground water, and no artesian flow,

qd = qi - qv

2. If there is no frost and artesian flows, and the inflows are only due to infiltration and
ground water,

qd = qi + qg - qv

3. If the inflows are only due to infiltration and artesian, then

qd = qi + qa- qv

Design of Drainage Layer

The design of the drainage layer involves either the determination of the maximum depth of
flow Hm when the permeability of the material K is known, or the determination of the
required permeability of the drainage material when the maximum flow depth is stipulated.
The drainage materials should consist of sound, clean, and open-graded and must have a high
permeability to accommodate the free passage of water and be protected from clogging by
means of filter. The amount of fine particles in drainage layer affects permeability and the
elimination of fines significantly increases permeability. A number of approximate
relationships have been suggested between permeability and grain sizes. The most frequently
used approximations are by Hazen for filter sands and by Moulton for granular drainage and
filter materials.

According to Hazen, permeability of filter sands, K, can be estimated as:

K  C k D102

In which, K is the permeability in mm/sec, D10 is the effective size, or grain size
corresponding to 10% passing, and Ck is Hazen’s coefficient as provided in Table 10- 1 which
is dependent the nature of the soil.

Table 10- 1. Hazen’s coefficient (Ck) for estimating permeability for filter sands

Type of sand D10 range (mm) Ck 1/mm-sec.

Uniform sand 0.06-3.0 8-12
Well graded and silty sand 0.003- 0.6 5-8

18
Another approximation devised for granular drainage and filter materials by Moulton is an
empirical equation of the form:

6.214  10 5 ( D10 )1.478 (n) 6.654

K
( P200 ) 0.597

where, K is permeability in (ft/day), D10 is as defined above (mm), P200 is percent fines
d
passing number 200 sieve (%), n is porosity ( 1  ), d is dry unit weight, and Gs
62.4G s
specific gravity of the material.

Figure 10- 7. Chart for estimating maximum depth of flow

The thickness of the drainage layer can be determined using a chart shown in Figure 10- 7 for
a given values of slope of drainage layer, and the length of the flow path based on steady
inflow and uniformly distributed across the surface of the pavement section. This condition
does not normally occur in practice, but a conservative result is obtained.

Design of Longitudinal Collector Pipes

Circular pipes are generally used for longitudinal collectors and are usually constructed of
either porous concrete, perforated corrugated metal, or vitrified clay. The pipes are laid in
trenches located at depths that will allow the drainage of the subsurface water from the
pavement structure. The trenches are then backfilled with porous granular material to
facilitate free flow of the subsurface water into the drains. Design of the longitudinal
collectors involves the determination of pipe diameter and the identification of a suitable
backfill material.

The diameter, D, of the collector pipe depends on the gradient of the pipe drain, the amount of
water per running linear length (q) that should be transmitted through the pipe, Manning’s

19
roughness coefficient of the pipe material, n, and the distance between the outlets L. When the
flow in the pipe is full, using the Manning equation, the relationship among n, q, L, S, and D,
becomes

nqL = 0.311685 D8/3 S1/2

Given any of the above values, the other one can be computed. To determine the diameter of
the pipe, the equation can be written as:

0.375
 3.208 Lqn 
D 0 .5 
 S 

Alternatively, the chart shown in Figure 10- 8 can also be used either to determine the
minimum pipe diameter when the flow depth, distance between the outlets, and the gradient
are specified, or to determine the maximum spacing between the outlets for different
combinations of gradient and pipe diameters.

In using the chart, it is first necessary to determine the amount of inflow q as:

q = qd L

in which q is the flow rate in the drain pipe (m3/day/m), qd is the design inflow (m3/day/m2),
and L is the length of the flow path (m).

Figure 10- 8. Chart relating pipe size with flow rate, outlet spacing, and pipe gradient
The material selected to backfill the pipe trench should be coarse enough to prevent the flow
of water into the pipe and also fine enough to prevent the infiltration of the drainage
aggregates into the pipe. The following can be used to select suitable filter material.

20
 D85 filter
For slotted pipes:  1.2
Slot width

D85 filter
For circular holes: 1
Hole diameter

Filter Criteria
Fine graded soil particles from the subgrade soil adjacent to the drainage layer may be
transmitted and clog the voids in the drainage material and eventually reduce the permeability
significantly and make less effective. This intrusion of fines can be minimized if coarse
material satisfies certain filter criteria. In case where these criteria are not satisfied by the
drainage material, protective filter must be provided between the subgrade and the drainage
layer to prevent clogging.

Any aggregate used for drainage must, therefore, satisfy the following filter criteria:

i) Clogging criteria – to prevent the adjacent fine materials from piping or migrating
into the filter material,
D15 filter
5
D85 protected soil
ii) Permeability criteria – to carry water without any significant resistance,
D15 filter
5
D15 protected soil
iii) Additional criteria – to make the grain size curves of filters and protected soils
somewhat parallel,
D50 filter
 25
D50 protected soil
– to minimize segregation, the filter material must have the coefficient of
uniformity,
D60
 25
D10
– the amount of fines in the filter material,
D5 filter  0.074 mm
– if the protected soil contains a large %age of gravels the filters should
be designed on the bases of the material finer than 25.4 mm.

21