; BOTSWANA ROAD DESIGN MANUAL pa9

Date August I%a2

CHAPTER 4

HYDROLOGY AND DRAINAGE

4-000 GENERAL

4-100 BASIC POLICY ON DRAINAGE DESIGN

HYDROLOGICAL ANALYSIS

4300 HYDRAULIC CULVERTS

4400 DRAINAGE OF ROAD PRISM

0-500 SUBSURFACE DRAINAGE

48DO DRAINAGE THROUGH WATERWAY BRIDGES

4700 REFERENCES
-
BOTWAHA ROAD DESIGN MANUAL %&?
Date augun t982

4-000 GENERAL

4-DO1.l Swpe

8001.2 Purpose of the Drainaga Chapter

4-W1.3 Application

4-002 . ABBREVIATIONS
 BOTSWANA ROAD DESIGN MANUAL bee
Date August 4982 '

4-100 BASIC POLICY ON DRAINAGE DESIGN 4-10111

4-101 CRITERIA 4-10111

4102 ECONOMICS OF DRAINAGE DESIGN

4 103 HYDRO-ECONOMIC OPTlMlSATlON

4-103.1 General

4-103.2 Design Life

6103.3 Repair Costs

4103.4 Risk of Failure

6103.5 Optimum Design

-
 BOTSWANA ROAD DESIGN MANUAL %
' P
Daw August 71982

4-200 HYDROLOGICAL aNatvsls

4202 FACTORS IMFLUEMCING DESIGN FLOODS 4-202/1

4202.1 General

4-202.2 Rainfall

4-202.2.1 Characterinir;s

4-202.2.2 Selection of design storm

4-202.2.3 Mean annual precipitation

-.
4-202.2.4 Fload return period

4-202.2.5 tntensityduration relationship

4-2022.6 Areal reduction fa-

62023 Runoff

4202.3.1 General

4-202.3.2 Time of mncenmtbn

4-202.3.3 Climatic factars

4-202.3.4 Physiographic factws

4-202.3.5 Runoff ccaffieiant

BOTSWANA ROAD DESIGN MANUAL Pace
Date August 9982 (

4203 METHODS OF DESIGN FLOOD DETERMlNATlON

4-203.1 General

4-203.2 The Rational Method

4203.2.1 Assumptions

4203.2.2 Runoff coefficient

4203.3 The SCS Method

4-203.3.1 Assumptions

4-203.3.2 Curve number

4-203.4 Alternative Methods

 BOTSWANA ROAD DESIGN MANUAL P W
Date August 1982

4300 HYDRAULIC CULVERTS

4-301 lMTRODClCflON

4-301.1 General

4301.2 Rural Road Drainage 4-30111

4-309.3 Urban Road Drainage 430112

4-302 CULVERT TYPE AND CLASS 430211

- 4302.4 Hydraulic Performance

.. 4302.2 Culvert Types

4-302.3 Culven Loading

4-303 DESIGN CRITERIA

4-303.1 Headwater - Depth Relationship

4-303.2 Inlet and Outlet Control

4-303.3 Freeboard Requirements

4303.4 Velocity

4303.5 Culvert Alignment

$303.6 Culvert Gradients

4-303.7 Minimum Size of Culvert

4-303.3.8 End Structures

4-303.8.1 General

4-303.8.2 TYpB
BOTSWANA ROAD DESIGN MANUAL pap
Date August P982 (

DRAINAGE OF ROAD PRISM

GENERAL

DRAINAGE CHANNELS

4-402.4 Hydraulics of Drainage Channels

4-402.2 Erosion Proteetion

4-402.3 Alignment and Grade

DRAINAGE OF ROAD PAVEMENT

URBAN GUTTER DESIGN

CHANNEL FREEBOARD
 BOTSWANA ROAD DESIGN MAMUAL pass
Date August I982
!

4-500 SUBSURFACE DRAINAGE al-rjol/l

4-501 GENERAL

4501.1 Ground Water

4502 SUBSURFACE DRAIN TYPES 4-50211

4-502.1 Pipe Subsurface Drains 4-50211

4-502.2 Stabilization Trenches

DESIGN CRITERIA

.. 4-504 MATERIALS

4504.1 Filter Materials

4-504.2 Filter Fabric

4-504.3 Subsurface Pipar

BOTSWANA ROAD DESIGN MANUAL pa@
Data August 1982

4-600 DRAINAGE THROUGH WATERWAY BRIDGES 4-604/1

4-601 GENERAL 4-60711

4602 HYDRAULIC REQUIREMENTS 4-602f1

4-602.1 Return Period 460211

4602.2 Design Flood 460211

4602.3 Flood Velocity 4-60211

4602.4 Backwater 4-60211

4-602.5 Freeboard to Bridge Soffits 4-60211

 BOTSWANA ROAD DESIGN MANUAL m e
i Date August 1982
BOTSWANA ROAD DESIGN MANUAL P W 4-00111
Date August 1982 [

4-000 GENERAL

4-001 INTRODUCTION

4-001.1 Scope

This aspect o f the manual coven the principles and guidelines t o be adopted i n the hydrologi-
cal analyds, including rainfall, runoff and methods of flood determination, together with sug-
gested parameters for the planning and design of a l l drainage structures and associated works,
such as scour protection. It also coven drainage of the road prism, subsurface drainage and
drainage through bridge structures. .*A -, .
4-001.2 Purpose of the Drainage Chapter

In Botswana the hydraulic and drainage structures associated with roadworks generally ac-
count for between 10% and 20% of the total cost of the roadworks. This is a significant
amount and consequenrly the factors associated with drainage design must receive careful and
detailed attention.

This chapter is intended t o highlight certain design methods.and factors which should be taken
into consideration in arriving at an optimum solution for road drainage.

Users o f this manual should appreciate that the vagaries of nature affecting runoff, streamflow
and erosion preclude the use of fixed methods o f design for hydraulic structures, particularly
larger structures, where a thorough study of local factors must be made to arrive at the opti-
mum solution. Certain unusual situations encountered in drainage design will therefore be ,
beyond the scope of this manual.
'

4001.3 Application

This chapter of the manual is intended as a guide to sound drainage design and practice. Devia-
tions from the recommended procedures herein should be discussed and authorized bv the
Chief Roads Engineer. It is not intended however t o be a substitute for a proper engineering
and hydrological appraisal of drainage structures, particularly the larger ones, where an ade-
quate and proper study of all relevant factors must be made.
 BOTSWANA ROAD DESIGN MANUAL Page ~-M)~/I
Date August 1982

4-002 ABBREVIATIONS

The following abbreviations are used:

Symbol Description Unit

A Catchment area km2

A Channel cross-sectionarea of flow, at depth d,

perpendicular t o direction of flow

d AASHTO American Association of State Highway and

Transportation Officials

AMC Antecedent moisture content

ARF Areal reduction factor

C Runoff coefficient (Rational Method)

CK Component of runoff coefficient for

soil permeability

cs Component of runoff coefficient for

average catchment slope

CV Component of runoff coefficient for

vegetal aver and land use

CN Runoff curve number (SCS Method)

d Point precipitation depth

Natural unrestricted depth of flow

D Culvert internal height

BOTSWANA ROAD DESIGN MANUAL PW 4-002/a
Date August 1982 (

Symbol Description Unit

d~ Average deprh of precipitation mm

9 Gravitational acceleration 19,8 m/s2 )

He Entrance head 10%

Hw Headwater depth at culvert inlet

Average rainfall intensity

Absolute barrel roughness

Ke Culvert entrance loss coefficient

L Hydraulic length of catchment

m Degree o f meandering

MAP Mean annual precipitation

MC Moisture condition

n Manning's roughness coefficient s/m7 l3

N Design life years

BOTSWANA ROAD DESIGN MANUAL Page 4-80213
Date kugust 4982

Symbol Description

P Risk of failure

P Wetted perimeter

Q Flow rate (discharge)

Peak discharge (flood peak)

QP

R Hydraulic radius

S Channel gradient mlm

SCS Soil Conservation Service

T Flood return period (recurrence interval)

Tc Time of concentration

Ts Storm duration (duration of excess rain)

V Velocity a t position of head loss

Channel velocity

v1 Velocity immediately before head loss

"2 Velocity immediately after head loss

BOTSWANA ROAD DESIGN MANUAL page 4-1QlIl
Date August 798.2 , (

4100 BASIC POLICY ON DRAINAGE DESIGN

4-10! CRITERIA

In drainage design of rural roads the main object is t o protect the road and adjacent lands
against potential damage from storm and subsurface waters. Roads are vulnerable t o damage
b y floods even i n semi-arid regions where the characteristic feature of the climate is that poten-
tial evaporation from the soil surface and from vegetation exceeds the annual rainfall.

The basic policy followed in the design of road drainage is as follows:

(a I n rural areas, soil and physical conditions permitting. the road shall be drained
directly into the road reserve.

(b) Where natural watercourses and drainage channels exist the road reserve shall be
drained directly into them.

(c) Where conditions necessitate drainage beyond the road reserve, additional land
shall be acquired for the necessary drainage channels, but this should be the exception rather
than the rule.

Various types o f drainage facility are employed t o protect the road against damage by surface
and subsurface water. Drainage facilities must be designed as simply as possible t o convey the
stormwater along or away from the road i n the most economic, efficient and practical manner
without damaging the road or adjacent properties. The options available t o the designer are
largely dependent upon the types of watercourses encountered, which in turn are functions of
the local terrain and climate. Watercourses in semi-arid areas are typically dry for most of the
year and flow only during and immediately following rainstorms.

i
 BOTSWANA ROAD DESIGN MANUAL page 4-10UP
Date August 1982

4-102 EGONOWIGS OF DRAINAGE DESlGN

The following factors are important considerations i n the economic analysis of drainage policy.

(a) The basic cost o f construction, incorporating the use of local materials wherever
possible.

(b) The useful life and cost o f replacement or extension.

(c) The cost t o the travelling public of any delays or extra travel distance due to road
closures.

The cost of repair resulting from damage t o the road structure or adjacent proper-
i (d)
ties caused by overtopping of the road or concentrations of water in adjacent properties.

(e) The cost o f cleaning and other maintenance for the structure or drainage facility.

Drainage design should strictly be regarded as an exercise in economics where costs of con
strutting a scheme are compared with the benefits derived from i t s performance.
BOTSWANA ROAD DESIGN MANUAL pase 4-10311
Date August 1982 [

4-103 HYDRO-ECONOMICOPTlMlSATlOLU

4-103.1 General

An optimisation technique for the selection of an economic structure is justified only when:

la) the structure is large enough to warrant it

(b) reasonable assessments of the potential damage associated with various return
period floods can be made

(c) the design return period is not necessarily rigid.

Where the flood return frequency is specified, the designer should consider the consequences
of functional failure and adjust the design return interval accordingly.

All structures are subject t o some risk of failure associated with a natural event, which has a
statistical probability of occurrence.

4-103.2 Design Life

The design life of a structure is not necessarily the length of time that the structure will func-
tion satisfactorily. Rather, it is the time allowed in which t o recover the capital cost of the
structure, at an annual rate, with a statistically related degree of assurance against having t o
replace the structure entirely during its design life.

4-103.3 Repair Costs

In addition t o the annual recove;y cost, an amount should be set aside for damage repair costs
associated with the probability of the design return period flood being exceeded i n any one
year.

4-103.4 Risk of Failure

The design standard may be defined as the value o f return period which is the most appropri-
ate for a given scheme in a particular locality. The selection o f this value is based upon a
knowledge of the design life o f the works and a safety criterion known as the permissible risk
of failure.
 BOTSWANA ROAD DESIGN MANUAL pa!% 4-1W2
i Date August 1982

The relationship between return period, T (years), design life, N [years) and the probability of
risk of failure, P, for the N-year period is given by

Using this relationship. which is sumrnarised in Fig. 4-103.4, values of any of the three vari-
ables may be determined, given values of the other two. For example, for a design life of
50 years and design return period of I00 years, the risk of failure is 39%. However, for design
purposes, the most frequent application of Fig.4-103.4 involves the determination of the
design return period, T, from a knowledge of P and N. P is best determined from a knowledge
of the site and the consequences of failure of the drainage structure.

-.
4103.5 Optimum Design

Adopting the above proeedure, analyses for a range of return periods should be carried out in
order to select a structure size and assess its total maximum probable annual cost for each
design Return periad discharge.

The structure with the least total annual cost is then the optimum structure. If, however, the
difference En cost between the optimum and any other structure varjes by say only 5%, then
the structure accommodating the higher return period flood should be selected. Fig. 4-i03.5
gives a typical illustration.
BOTSWANA ROAD DESIGN MANUAL
page 4103/3
Date August 1982
BOTSWANA ROAD DESIGN MANUAL Page 4-10314
Date Augusl1982
BOTSWANA WOAD DESIGN MANUAL kse 4-201/1
Date August 9982 j

4-200 HYDROLOGICAL ANALYSIS

4.201 INTRODUCTION

An appropriate hydrological analysis of the area to be drained is an essential element i n the

design of road drainage. This type o f study supplies the information on runoff and stream
flow characteristics, which is used as a basis for the hydraulic design.

The design flow is established by selecting the proper combinations of rainfall and runoff cha-
racteristics that can reasonably be expected to occur. This is usually restricted further b y
establishing an interval of time or frequency period as a basis for design. The design criterion
would then be the maximum flow carried by the drainage structure with no flooding or a
limited amount of flooding t o be exceeded on the average of once during the design period. -
The accuracy with which flood estimates can be made depends on the amount and quality of
relevant information available. In hydrological calculations the consequences o f failure should
dictate the effort required t o establish an acceptable estimate. Practical experience under
local conditions and the application of sound judgement are particularly important in deter-
mining the extent of the investigation necessary.

Consistency of results should not be expected between the different methods of flood dis-
charge estimation used and estimates must be viewed i n relation t o the applicability of the
method being employed. It should be noted that the methods presented i n this manual are
generally appiiwble t o rural catchments, but may be applied t o urban catchments with cau-
tion. The often complex nature of urban catchments requires the application of methods
which can accommodate the subdivision o f a catchment into various hydrologic elements, in
order t o obtain realistic flood estimates. A river or stream passing through a town or village
would be a typical example. This type o f problem should be anelysed by computer methods,
if possible.

For practical purposes it should be assumed that a flood with the same return period as the
design storm will occur when average catchment conditions, consistent with that return period,
prevail. It is therefore necessary t o assess the probable average condition of a catchment, com-
patible with the design return period, i n order t o achieve consistency in flood estimates. The
various factors used i n the design methods should in turn be selected to reasonably represent
the assessed catchment condition.
B O n W A N A ROAD DESIGN MANUAL page 4-202/l
Date August 1982

4-2W FACTORS lNFLUEMCllNG DESIGN FLOODS

4-202.1 General

When rain falls on a drainage catchment, part of the water may be prevented from reaching the
catchment exit, while some may be delayed en route. Losses in precipitation arise from infil-
tration, evaporation (see Table 4-202.1), storage in surface depressions and interception by
vegetal cover.

The excess precipitation travels b y the hydraulically shortest route t o the catchment exit. The
determination of the volume o f this runoff and the rate at which it arrives at the catchment
exit should be the prime objective o f the designer.

The factors affecting flood peaks and volumes may be conveniently grouped as those affecting
rainfall and runoff. The principal factor used to link rainfall and runoff is the time taken for
the catchment t o respond t o the rainfall input. Time of concentration (Tc) should be adopted
as the measure of the catchment response time. This is the time for the surface runoffafrom
the hydraulically most remote part of the catchment area t o reach the point being considered. .
This remotest point is not necessarily the most distant point in the drainage area.

4-202.2 Rainfall

4-202.2.1 Characteristics

Storm types are important when considering both the duration and intensity of rainfall.

Botswana has a distinct rainy season: the first rains have usually fallen by @e end o f Sept-
ember and continue until April o r May of the fol!owing year.

The country receives most of its rainfall from convection processes, such as instability showers
and thunderstorms, which are the result o f the convective rising and subsequent condensation
o f water vapour in warm air pockets heated by the surface of the earth. Most floods from
small catchments i n Botswana are the result of convective thunderstorms, which generally
yield high intensity - short duration rainfall. The incidence of rainfall i n Botswana is highly
variable, both from place to place, and from time t o time. See Fig. 4-202.2.1.

An observation of rainfall records from the long existing gauging stations indicates that much
of the rainfall i n Botswana occurs i n spells lasting 2-4days, with fewer occasions of heavy
downpours. This accounts for the bulk of the seasonal rainfall. On average, precipitation
amounts t o 5 m m or less for approximately half the days where convective developments
occur.
 BOTSWANA ROAD DESIGN MANUAL paw 4-2Oa2
Date August1982 ;

, ~ TABLE 4-202.1 MEAN MONTHLY EVAPORATION IN BOTSWANA

. ,

.~-.

..

$"_. '

I , .

ii
3
 BOTSWANA ROAD DESIGN MANUAL page 420214
Date August 9982

The variation in monthly and annual rainfall from year t o year is considerable and there
appears t o be no regularity i n the occurrence o f wet and dry years i n Botswana.

The following points are noteworthy:

(a) The annual rainfall, i n extreme cases, may be as high as or greater than twice the
normal annual rainfall or as low as or less than half the normal annual rainfall.

(b) The variability of monthly rainfall is much higher than the variability of annual
rainfall, especially i n the dry season.

(c) The maximum daily rainfall may considerably exceed the mean monthly rainfall
in all months in the dry winter season b y large amounts.

.- (d) Defining a rainday as a day on which 1 mm or more of rainfall is recorded, it is

seen that on 50 t o 55% o f such days amounts of 5 rnm or more are recorded and only on
25 to 35%of such days are amounts of 10 mrn or more recorded.

Tables 4-202.2A t o K give the annual and mean monthly rainfall for various towns and vil-
lages in Botswana. Monthly and daily maximum rainfall figures are given below the latter.

Because of the .tremendous variability of rainfall in Botswana it is important that the designer
be aware o f this and exercise the appropriate degree of caution in estimating floods and run-
offs.

A brief study of the climate of Botswana is fecommended, as it could influence many deci-
sions made. A useful reference is "The Guide t o Botswana" by Alex Campbell.

Botswana has at present only one autographic and other recorders station, that being at Maun
3 from which much information i s available. Data from this station has been used extensively by
the University of the Witwatersrand t o assist in their nydrologicai rsseaich work in Southern
Africa.

4-202.2.2 Selection of design storm

The response of a catchment t o rainfall input is dependent on i t s physiographic features as well

as the size, duration, direction and speed o f the storm.

Storms in Botswana vary from those covering an area o f a few square kiiometres t o others
covering an area o f hundreds of square kilometres.

Storm activityjs seldom uniform over a catchment area resulting in an uneven distribution of
i precipitation, Generally, the larger the catchment area the greater the variation i n depth of
precipitation over the catchment. This is discussed further under Section 4202.2.6 Areal
Reduction Factor.
1 I
South East District
 - BOTSWANA ROAD DESIGN MANUAL Page 4202/9
Date August 1982 (

. .

m mco
-
m a m t.~

i
i
/
i
i
' 1.e

. - -
BOBWANA ROAD DESIGN MANUAL Pa, Q2a3O
Date Augst 1982
- ~
BOTSWANA ROAD DESIGN MANUAL Page 4202/9 4
Date August 1982 .
(

, .

, .

1
i
1

.. .. .
BOTSWANA RQAD DESIGN MANUAL eaSe 8202182
Bete August 1982
Ghanzi District

/
BOTSWAMA ROAD DESlGlsl MANUAL
 BOTSWANA ROAD DESlON MANUAL b 4--16
Dam August 1982

Far the purpose of this manual the following basic assumptions should be made:

(a tire flood peak will occur when a stationary storm is of sufficient duration to en-
sure that runoff from each portion of the catchment will contribute t o the flood peak simul-
taneously. This means t h a t the critical storm duration IT,) should be at least ewal to the
time of concentration (Tc).

(b) runoff reaching the catchment exit after the time of concentration ITc) wit! he
too late to contribute to the flood peak.

it will be evident from depth, duration, frequency studies that storms lasting longer than
time of concentration result in lower prolonged flood peaks with increased flood

Intensity-duration curves are based on a stationary storm. In many instances the prevailing
storm direction over a catchment is known. It is recommendedthat in cases where the prevail-
ing storm are known t o travel in the downstream direction that the time of concentration be
reduced by up to 20%. This figure is arbitrary due to lack of published research data in this
regard.

4202.2.3 Mean annual precipitation

Mean annual precipitation has a significant influence on the depth of precipitation for a par-
ticular storm duration Ts and return period T, and a commensurate effect on peak flood dis-
charges.

Mean annual precipitation for a particular location may be interpolated from Fig. 4-202.2.3or
be obtained fmm the Meteorological Department of the Ministry of Works, if available.

J )
r 4202.2.4 Flood return period

I f an went has a return period o f T years. it means that the probability of mat event occurring
in any specific year is 11T. In flood estimation this event is commonly defined as that equaling
or exceeding a specific peak flow rate.

The degree and cost of repairing damage caused by a flood exceeding the capacity of a
hydraulic structure, combined with the hazards. inwnvmience to the public, and classification
of the road, control the determination of the design frequency and therefore the design dis-
charge.

The annual probability of occurrence tends to an average for a number of drainage structures
scattered over an extensive region, For this reason the design return period is prescribed for
various types of installations and need not always be established for each design case by hydro-
economic optimisation, as described under Section 4-103.
BOTSWANA W O A D '
BOTSWANA ROAD DESIGN MANUAL

FIG. 4-202.2.3
BOTSWANA ROAD DESIGN MANUAL page 4-202148
Date August 1982

Design return period is divided into four categories:

(a) Large catchments (Bridges)

The design flood for bridges should be based on the minimum flood return periods, shown in
Table 4-202.2.4A.

TABLE 4-202.2.4A DESIGN FLOOD RETURN PERIOD FOR BRIDGES

A major bridge is defined as a structure which, when designed for a 1 in 25 year flood, has a
waterway opening greater than 600 mZ (including freeboard) or a waterway opening greater
than 250 mZ and an overall length, measured at right angles t o the river centre line, greater
than 60 m.

A minor bridge is a hydraulic structure which, when designed for a 1 in 25 year flood, has an
opening greater than 6 m measured from abutment t o abutment along the centre line of the
road and whose catchment exceeds 25 km2.

A longer return period may be adopted, if the cost of the expected damage due t o a higher

9 frequency flood is greater than the allowable cost o f the additional capital investment t o pro-
vide a bridge opening for the greater inundation.

(b) Intermediate catchments in rural areas

Intermediate catchments are those ranging in area from 10 t o 25 km2, and with a discharge
greater than 10 m3Is.

The return period for culverts falling into this category should receive special attention. Criti-
cal culverts, where flood damage might have serious consequences or which control the road
grade line over substantial distances. musr be checked b y an optimisation technique in order t o
establish the design return period. Other f a m s which could influence the appropriate return
period include:

Shape o f catchment
Slope o f catchment
Cross-section of catchment at exit
Road type
Road grading (whether culvert k a grade limiting structure or not)
Effmof overtopping (traffic delays, maintenance)
Availabiliw o f funds
BOTSWANA ROAD DESIGN MANUAL page 4-202119
Date August 19k2 (

The final choice of design return period should never be less than that given in
Table 4-202.2.48.

DESIGN FLOOD RETURN PERIODS FOR

TABLE 4-202.2.4B CULVERTS I N INTERMEDIATE CATCHMENTS

Road Type
I Design Frequency

1 i n 25 years
1 in 20 years
1 in 15 years

4 1i n 10 years
5 1 i n 10 years
6A 1 in 5 years
60 and 6C -

A grade limiting structure i s one where the level of the culvert is the ruling factor for the grade
o f the road over it.

(c) Small catchments in rural areas

Small catchments are those with an area less than 10 km2, or with a discharge less than
10 m3/s.
'
Small catchments are considered as those contributing t o the small culverts as well as channels
and side drains. The flood return period should be taicen as that shown in Table 4-202.2.4C. '-

DESIGN FLOOD RETURN PERIODS FOR

TABLE 4-202.2.4C CULVERTS IN MINOR CATCHMENTS

1 in 20 years
1 in 20 years
1 in 10 years

1 in 10years
1 in 5 years

Any deviation from the values given in the table should be confirmed with the Chief Roads
Engineer in writing.
 BOTSWANA ROAD DESIGN MANUAL %Je 4-202120
8 Dam August 1982

(d) Catchments within urban areas (usually sman catchments)

The following recurrence intervals should be used in urban areas for the design of minor storm
drainage.

Two years through residential areas where short term flooding is not of major
consequence (in very flat areas, where the capital cost would be out of proportion
to the possible damage, a shorter recurrence interval may be permitted).

Five years for areas such as shopping, commercial and public building areas, where
exceptional storms are more likely to cause serious flooding and extensive dam-
age.

Ten years where the mad passes through high value business areas.

Minor storm drainage consists of closed and open conduits carefully designed in a manner
which will permit the unimpeded use of the road during the more frequent storms.

The flood return period for major waterways passing through urban areas should be as given
under Section 4-202.2.4(a).

4-202.2.5 Intensity-Durationrelationship

Intensity-duration cuwcs for Gaborone, Francistown, Maun and Ghanzi are included in this
manual. See Figs. 4-202.2.5A t o 4-2022.5C respectively. These curves are based on "Design
Flood Determination in South Africa" by D.C. Midgley. Curves appropriate t o other areas
may be derived from charts prepared by the Meteorological Office in Gaborone or from
iitemture such as that presented by the Hydrological Research Unit of the University of the
W~twatersrand.

Most of the smaller towns have only 20 years of available rainfall data. Designers are warned
to take cognizance of this fact when deriving intensityduration relationshim.

4-202.2.6 Areal reduction factor

Methods for the computation of design floods require an estimate of equivalent uniform
precipitation depth over the entire catchment, a quantity generally referred to as areal rainfall.
B(XB'UANA ROAD DESIGN MANUAL
 pa@ 420212.2
BOTSWANA WOAD DESIGN MANUAL Date August 4982
BOTSWANA ROAD DESIGN M A N U A L
Page 4-202123
Date August 1982 ,
 BOTSWANA ROAD DESlSM MANUAL We 42DZl24
Date Auwat 1982

As the catchment area increases so does the variation i n expected depth of precipitation at
different points throughout the catchment.

Such non-uniform rainfall depth must therefore be averaged out for design purposes. Average
depth of precipitation d~ is obtained by multiplying point precipitation depth d b y an areal
reduction factor ARF.

"A
- ARF x d

Experience has shown that ARF-values vary with rainfall duration as well as with both
3 the area and shape of the catchments. However any possible changes of ARF with
9 return period are unknown a t this stage. Untii more comprehensive data become avail-
able, the use of Fig. 4-202.2.6 is recommended-

Consideration should also be given to modifying the ARFs for travelling storms, as storms
driven in the direction o f a catchment exit will frequently lead t o more severe floods than
- stationary storms.

4-202.3 Runoff

4-202.3.1 General

The numerous factors affecting runoff may vary considerably within the same catchment over
a period of time as a result of changing land use. In addition, a catchment may display differ-
ent runoff characteristics when responding t o storms of different types. The relationship of
rainfall t o runoff should therefore be made considering the probabie average hydrologic condi-
tion of the catchment consistent with the design return period.

Runoff is affected b y the following factors:

4-202.3.2 Time of concentration

Time of c o n ~ n t r a t j o nKc is used to define the design storm duration, whilst being a parameter
primarily related t o the physiographicfeatures of a catchment.

Where a basin, such as a lake, swamp or dam, lies on the longest collector it will have the effect
of attenuating the flood peak even i f the basin happened t o be full a t the commencement of
the sxorm. The flood routing method used is left t o the designer.
 BOTSWANA ROAD DESIGN MANUAL Page 4-202126
r Date August 1982

4-202.3.3 Climatic factors

The type of precipitation, temperature, solar radiation, wind, humidity and antecedent rnois-
lure condition a l l influence the hydrological response of a catchment. Engineering judgement
should be relied upon toassess the possible effects of most of these factors, which are difficult
t o quantify. Only if it is f e l t that these factors may have a significant effect on a particular
portion of the design should they be specifically considered.

4-202.3.4 Physiographic factom

The important factors relating to the geometric and physical aspects of a catchment may be
listed as follows:

(a) Geometric factors

Area

Runoff is directly related t o area of catchment, although not linearly. A small catchment of
less than 10 km2 is, because of overland flow, generally sensitive t o high intensity, short dura-
tion rainfall and to land, use. However, on. larger catchments the effects of channel flow be-
come pronounced and the sensitivities tend t o diminish.

Shape and orjentation

The shape and orientation of a mtchment play an important part i n catchment response.
Catchments with irregular shapes, such as those with two or more main collectors meeting up-
stream o f a culvert, should be analysed using applicable methods.

Average catchment slope is important in establishing the velocity of overland flow in a catcn-
ment and in determining rainfall-runoff coefficients for use in the Rational formula.

(b) Physical factors

Soil permeability is one of the most important factors which influence storm precipitation
losses.

I Changes i n soil permeability within a large catchment should be taken into account. in deter-
mining the runoff.

1
 BOTSWANA ROAD DESIGN MANUAL pa9 4-202~27
Dare August 4 982

Land cover, use and treatment

Cover is taken to mean any material or plants covering the soil and providing protection from
the impact of rainfall.

The designer should attempt t o establish the average conditions, for example, interception in
the form of retention, detention and infiltration, which are likely t o prevail i n the catchment
during the design return period. Information about future urbanisation should be obtained for
catchments on urban boundaries.

Lakes, swamps and dams

Swamps in Botswana tend to be perennial and can therefore be considered full at the corn- .,
,P
mencement of a design storm. The chances of lakes or dams being full at the onset of the *-- .
storm are, however, remote. However, for the purposes of assessingthe runoff they should be
considered t o be full.

Local flood observation

A helpful and necessary method in determining the peak runoff is that of flood level obse~a-
-- tion by the local population in the vicinity of the site or flood marks at the site. The road
maintenance authorities should also be consulted in this regard. Especially when the duration
and intensity of the relevant rainstorm are known values, storm losses can be calculatedand a
more accurate discharge computed.

4202.3.5 Runoff coefficients

Runoff coefficients used i n the Rational Method are discussed in Section 4-203.2.2.
<*"-
2
 BOTSWANA WOAD DESIGN MANUAL page Q203/9
Date Auaugusp 1982

4203 METHODS OF DESIGN FLOOD DETERMINATION

4-203.1 Generai

Various methods for the determination of the design flood can be used. but, since very few of
the catchments are gauged, the number of suitable methods is limited.

Rainfall-runoff data are seldom available on small catchments. In cases where records over
reasonably long periods do exist a statistical analysis for flood estimation could be applied.

j .. Empirical methods, which may have been calibrated for local conditions, are known to provide
1, ' I reasonable results and can be u ~ todcompare with the results of other methods.
t -8
The practice of weighting the flood estimates obtained from a number of meshods to deter-
mine the resultant peak value is not recommended.

Two methods are recommended for design flood determination in ungauged catchments.
They are the Rational Method and the United States Soil Conservation Service (SGS) Method.
It must be emphasized, however, that both these methods have limitations.

4-203.2 The Rational Method

This method has been one of t h e most widely used methods for predicting peak discharges
on ungauged catchments. The criterion in the use of this method is the subjectiveness in
selecting the runoff coefficient C. It is recommended that direct application of the Rational
Formula be limited to catchments of minor importance and less than 25 km2 in area. Appli-
cation of the method could be extended to larger catchments, but with caution.

The basic form of the equation is

0 -
- -
CIA
3.6
where Q = flood peak a t catchment exit (m31s)
C = the Rational runoff coefficient
I = the average rainfall intensity over the whole catchment (mm/h)
A = catchment area (km2)

4-203.2.1 Assumptions

1 The main assumplions inherent to the method are :

\

(a) The design storm produces a uniform rainfall intensity over the entire catchment.
1 1
 BOTSWANA ROAD DESIGN MANUAL page 4-203/2
Date Auwst 1982

(b) The relationship between rainfall intensity and rate of runoff is a constant for a
particular catchment.

(ct The flood peak a t the catchment exit occurs at the time of concentration

(d) The coefficient C is constant and independent of rainfall intensity.

(4 Time of concentration Tc is the time from the hydraulically most remote point.

4-203.2.2 Runoff coefficient

8
The runoff coefficient C in the Rational Formula represents the ratio of rate of runoff to
rainfall intensity for a particular catchment. Suggested coefiicients are given in --
Tabtc 4-203.2.2. They should be compared with sound engineering judgement.

4-203.3 The SCS Method

This method was developed by the United States Soil Conservation Service for use on small
- agricultural catchments. It has since been improved and extended for use on much larger
catchments and has been used since 1962 as an alternative t o the Rational Method in the
Republic of South Africa. Details of this method are covered by numerous hydrological
textbooks, one being "Stormwater Hydrology and Drainage" b y D. Stephenson (1981) -
Elsevier.

4-203.3.1 Assumptions .

The main assumptions inherent in the SCS Method are :

(a) Precipitation losses arise from an initial abstraction followed by a time-dependent

\ rate of infiltration.

(b) Runoff or excess rainfall may be determined b y using a curve number (CN),
which is dependent on hydrological soil group, ground cover and antecedent moisture con-
dition (AMC).

(c) The flood hydrograph may be developed using a typical unitgraph.

6203.3.2 Curve number

i
Curve number (CN), which is used in the basic formula, can be determined by applying the i
i
suggested values in Tables 4-203.3.2Aand B. An adjustment t o the CN value for moisture 8
condition using Table 4-203.3.26 must also be made. This adjusted CN value is the equiva- ( @
i:
lent of the runoff coefficient C used in the Rational Method. 8
 Page 4-203/3
BOTSWANA ROAD DESlGN MANUAL
Date ~ u g u s 1982
t

RA-~-IONAL METHOD RUNOFF CoEFI1ClfN-i- C

BOTSWANA ROAD DESIGN MANUAL page 4-203~
Date August 1982

TABLE 4-203.3.2A HYDROLOGlCSOIL GROUP AND VEGETATIVE CONDITION

HYDROLOGIC
DESCRIPTION
SOIL GROUP

.".
.:

4:: .-
than 1/2 of area.
Regularly burned.

i
i I
I
i

2
BOTSWANA ROAD DESIGN MANUAL Page 4-2-2839
Date August 4982

TABLE 4-203.3.28 CN-VALUES FOR SOILCOVER COMPLEXES (MC= 2)

DESCRIPTION

Contnursd, partial cover

Grwnd cover. terraces

dustrial district$
Reiidential 500 sq.m

Paved parking 10s. roofs,en

Streetslroads : tarred, with storm sewers, wrb
BOTSWANA ROAD DESIGN MANUAL Page 4-20316
Date August 1982 (

4-203.4 Alternative Methods

A helpful method in determining the peak runoff is that of flood level observation by local
inhabitants in the vicinity of the site or flood marks at the site. This is particularly useful
for large catchment floods.
BOTSWANA ROAD DESIGN MANUAL page 4-203/7
Date August 1982

Another useful technique is the unit hydrograph method. Methods of deriving unit hydro-
graphs from records of storm input and flood output are expiained i n most hydrology text-
books. In Southern Africa such data are generally inadequate for conventional derivation
of unitgraphs. The Hydrological Research Unit of the University of the Witwatersrand,
however, has devised a novel method for the compilation of unitgraphs for catchments where
input data were poor. Reference t o this method can be found i n "Design Flood Determina-
tion in South Africa" Report No. 1172 by Midgiey.

Other methods, such as the B.D. Richards Method, can also be used provided they are appli-
cable to the conditions existant i n Botswana.
 BOTSWANA ROAD DESiGN MANUAL Page 4301/4
Date August 1982 (

4-300 HYDRAULIC CULVERTS

4-301 INTRODUCTION

4-301.1 General

This section considers the hydraulics of conventional culverts and the requirements for select-
ing a culvert size for a given set of conditions. Design procedure and the corresponding use
of recognised design curves and tables are left t o the designer.

4-301.2 Rural Road Drainage

A road is considered as a dyke which divides the countryside in two parts, and therefore
disturbs the water flow characteristics of the area, whether it be sheet overland flow or channel
*low i n defined watercourses. The importance of continuity in surface drainage is illustrated
by the fact that most road embankment failures are caused by water problems. Drainage
design will influence the road geometry on the higher standard road t o a large extent. An
efficient drainage system is an essential contribution t o the overall design of the road. A vital
factor in rural areas is t o ensure continuity of flow from one side of the road t o the other,
whatever class of road is involved.

The road, which functions as a barrier, will cause storm runoff t o flow parallel to the road
embankment until a discharge point is reached, whether it be a relieving culvert or a culvert
at a low point.

Between two watersheds there is generally one low point i n the road. The distance from a -
watershed to this low point can be several kilometres, dependent on the nature of the terrain. ( .
1..
Provision should be made t o discharge the runoff across the road at regular intervals t o avoid
such a concentration o f runoff at the low point and t o maintain a balance of runoff. These
intervals depend upon the locality and type of road and c-nuld be in of 500m in flat
!
semi-arid terrain.

Wherever the situation arises that a road is aligned parallel to and close upstream or down-
stream of an existing railway line, the siting and sizing of the road drainage structures may
need artificial adjustment t o match that of the railway structures. The designer i s encouraged
t o liaise closely with the relevant authorities in all such cases.

To prevent the collection of water and ponding on the road surface, roads should be cambered
as follows :

Surfaczd roads 2.5% camber

Gravel roads 3,5% camber
Earth roads 4 camber
 BOTSWANA ROAD DESIGN MANUAL b@ 4-309/2
I Date August 'I982

4-301.3 Urban Road Drainage

Roads passing through urban areas should generally be constructed high enough not to act as
collector channels for surrounding runoff during a storm. The designer should at the same
time take into account the damage that might occur as a result of a major storm.

The type of drainage required should be interactive with the type of urban road. It might
be considered that a road passing through a minor vOlage or small town would have a drainage
design typical of that for a rural road. Open side-channel drainage combined with normal
cross culverts a t selected points should be considered for roads passing through the larger
town. For major towns in Botswana with a developed infrastructure, a more sophisticated
drainage system could be designed with kerb inlets and underground pipe systems. An under-
ground drainage sysism mus? not, however. be incorporated at the expense of a satisfactory
3
surface drainage layout.
 BOTSWANA ROAD DESIGN MANUAL page 4-3ozf1
Date August 1982 (

4-302 CULVERT TYPE A N D CLASS

4-302.1 Hydraulic Performance

Hydraulic performance is one o f several factors which influence the selection of type of culvert
for a particular location. The shape and type of culvert t o be used should n o t be determined
by hydraulic performance alone. The economics and availability of different culvert types
in the various regions should also be considered.

; 4-302.2 Culvert Types

.... .. .-
Transport and purchase costs play a major role in selecting the c u l v ? ~type f o r a particular
region.

The advantages of corrugated metal nestable pipe sections should be considered for small
t o medium sized culverts and the corrugated metal multi-plate arch for larger culverts. Metal
corrosion is not normally a problem in Botswana.

Prefabricated concrete pipe and portal units are encouraged, wherever they can be economi-
cally and practically justified.

A newly developed plastic pipe with radial ribs may also be considered for the smaller range
o f culvert sizes.

Where the aforementioned culvert types are not economical, due to a remote location, insitu
concrete culverts of the b o x type may be specified, if the required aggregates can be obtained
locally.

4-302.3 Culvert Loading

Culverts o f whatever t y p e will be subjected t o loads. The primary loads which should be
considered are as follows : self mas, water mass, mass o f backfill, traffic loads, temporary
handling and construction. These factors, individually or collectively, influence the class of
culvert t o be installed.
 BOTSWANA ROAD DEStGN MANUAL paw 4-30317
Date August 1982

4-303 DESIGN CRlTERlA

4303.7 Headwater - Dep* Relationship

AI! culvefls should be designed t o carry the design frequency flood with a headwater depth
that does not materially increase the size of the flooded upmeam area. This is particularly
critical in urban areas. In rural areas allowable headwater depths should be determined b y
field conditions and are not associated with design criteria. These may vary considerably
depending on circumstances. A headwater-culvert depth ratio (Hw/D) equal t o 1.2 i s recom-
mended for cases where insufficient data is avaitable t o predict the flooding effect from
headwater depth.

Allowable headwater depth is determined b y the maximum permissible e!evation of the

headwater pool a t the culvert for the design discharge.

- 42303.2 Inlet and Outlet Control

A culvert may flow with either inlet or outlet control over its full design dischargerange.. .:
Alternatively flow through the culvert may vary with discharge from inlet t o outlet'$on,trol. i
This criteria should be carefully analysed by the designer.
- .
Culverts flowing with inlet control usuaily lie on relatively steep gradients and flow only
partly full. Guidance for the sizing o f culverts for inlet control are obtainable in publications
by the U.S. Bureau o f Public Roads.

)% 4-303.3 Freeboard Requirements

,
'3
1~

9:aina:e structures should be sized such that the headwater elevation resulting from the design
discharge is somewhat lower than the road sh~Iilder.hrPakpoint at the low point i n the road.
This difference in level is referred t o as the freeboard.
(

The minimum freeboard should be :

(a) 250mm on road types 1.2A and 28,

(b) 100mm on all other road types except road type 6.

Ic) I f justified on economic grounds, freeboard on road types 5A and 58 may be

omitted altogether. Funhermore, the design flood c ~ u l dbe permitted t o discharge over the
road up t o a depth of 200mm, but only i f approvedby the Chief Roads Engineer.
BOTSWANA ROAD DESIGN MANUAL Page 430312
Date August 1982

Such low points should be localised and the road graded level over them so as to minimise
scour damage caused by overtopping. Wherever possible dips in the road gradings on these
low class roads should be located at watercourse crossings where the deck slabs of concrete
box culverts could be used as the riding surface of the roadway. Scour damage would then be
localised and rninimised.

(d) On roads type 6 no freeboard is required. The carriageway should be aligned to

follcw existing ground profile and cross dry watercourses a t bed level. All flow will pass over
the road, which will therefore become temporarily impassable to traffic during floods.

4303.4 Velocity
_ ..~
A culvert, because of its hydraulic characteristics, generally increases the veiocity of flow;
above that i n the natural channel. High velocities are critical immediately downstream of
the culvert outlet and the scour potential from the resulting energy is a factor t o be
considered in culvert design.

Culverts should be laid t o grades that produce a nonsilting or a nonerosive velocity, ideally
between 1 and 3,5m/s. This is particularly important in the sandy, semi-arid regions which
experience sporadic high intensity cloudbursts.

Fig. 4-303.4 gives suggested scour velocities for different streambed materials.

4303.5 Culvert Alignment

Generally, a culvert should be aligned along the natural watercouse. However, the culvert
itself should be straight between the inlet and outlet. This may require some modification of
the natural conditions.
.,
(.
Where natural conditions call for skewed culvert alignments, a reduced degree of skew angle
should be investigated in crder t o shcrte~the ciilveri, and reduce costs. Sound judgement
should be used in any associated channel relocations as these could adversely influence the
hydraulic performance of the culvert. Unless downstream scour would present serious
problems, the inlet o f the culvert should be placed i n the natural watercourse.

4-303.6 Culvert Gradients

Culverts should generally be placed on the natural gradient, where this is not excessive.

Velocity is associated with gradient and therefore the following minimum and maximum
gradients should be used t o ensure an acceptable range of velocities.
 B
[)OMINANT STREAMBED MATERIAL 2
P
C)
z
(LI
0
c
c
X e,
?J G8
m
V) 8a
BOTSWANA ROAD DESIGN MANUAL page 4-30314
Date August 1982 1'

(a) Minimum gradients

Gradients less than 1 percent should be avoided, if possible. Gradients less than 0.5 percent
should not be used as they create maintenance problems resulting from siltation of the culvert,
particularly where the catchment consists of erodible material.

(b) Steep gradients

Culverts on gradients i n excess of 10 percent should be provided with anchors in the case of
concrete pipes. For gradients in excess of 20 percent, special anchoring devices should be
incorporated for all types of pipe culverts.
I
-
Where the natural gradient is steep and outlet scour is likely, the introduction of drop inlets .
and stilling basins at the outlets, acting as energy dissipaters, should be considered. k- :

4-303.7 Minimum Size of Culvert

The minimum diameter of a pipe culvert under a road should be 600mm or i t s equivalent in
corrugated metal pipe arches. Wherever culverts are laid on grades producing self-scouring
velocities and where maintenance is not likely to be a problem, the minimum size may be
decreased t o 450mm diameter.

Similarly, the minimum size of precast concrete portal-type culvert under a road should be
750mm x 450mm. This may be reduced to 4M)mm x 300rnm wherever self-scouring veloci-
ties through the culvert can be achieved, resulting in a maintenance free structure.

4303.8 End Structures

4-303.8.1 General

Inlet and outlet structures are required to prevent scouring of the roadway embankment, to
provide a transition from the channel t o the culvert, and to improve the hydraulic performance
of the culvert.

In the higher class roads the design and type of inlet and outlet structuresshould be aimed a t
providing a balance between hydraulic performance and cost. In the lower class roads the
emphasis should be on t h e cost saving aspect. Unless hydraulic or other considerations require
special end structures, the least expensive type of end structure should be adopted throughout
the project for continuity.
 BOTSWANA ROAD DESiGN MANUAL Page 4-303/5
!
Date August 1982

4.303.8.2 Types

The type of end structure selected wilt depend on :

Hydraulic performance (see Table 4-303.8.21.

cost.
Availability o f suitable materials.

On minor rural roads the use o f concrete may be totally uneconomic. On the other hand,
suitable rock for pitching, as an alternative to concrete, may net be available.

Coefficient k, to apply t o velociry head Vz t o determine loss o f head at entrance to a

structure, such as a culvert or conduit, operating full o r partly full with control at the outlet.

Entrance head loss He =, , ,k VZ where V2 = V - V2

1
29

TABLE 4-303.8.2 CULVERT ENTRANCE LOSS COEFFICIENTS

.
I Tvoe o f Structure and Design of Entrance

I Projecting from fill, socket end

Projecting from fill, sq. cut end

Headwall or headwall and wingwails

Socket end of pipe
Square-edge
Rounded (radius = 1/120)

Mitred t o conform t o fili slope

End-section conforming t o fill slope

Pips, or Pipe-Arch - Corrugated M a u l

Projecting from f i l l (no headwall)

Headwall or headwall and wingwalis

Square-edge

Mitred to conform t o fill slope

End-Section conforming t o fili slope

Box - Rainforced Concrete

Headwall parallel t o embankment (no wingwallst
Square-edge on 3 edges
Rounded on 3 edges zo radius of 1/12 barrel

Wingwalls a t 30' Po 75' to barrel

Square-edged at crown
Crown-edge rounded t o radius of 1/12 barrel

Wingwalls at 10' t o 30° t o barrel

Square-edged at crown

W i n u w a i ! ~oarallel (extension of sides)

BOTSWANA ROAD DESIGN MANUAL pa* 830316
Date Auyst1982 (

End structuresshould consist of the following types :

(a) Composite headwaltlwingwall/apron structure

This is the most common type of structure, but generzlly the most expensive. It protects
the embankment and provides a transition between the culvert and channel. It is normally
consmicted in concrete, but cement mortared block or stone masonry wall, or gabion con-
struction should also be considered. The apron slab may be omitted where the culvert is
founded on rock material.

6.,.,;.i.
This is a headwall without the wingwalis, and similar construction to t h e headwalllwingwall
structure should be considered.

(c) Mitred structure

This is the endwail structure, but mitred t o match the slope of the embankment. if this
structure is used it is normally constructed in cementinortared stovework or gabion mattres-
ses. Mitring improves the hydraulics of flow at the entrance and is therefore preferred t o the
endwall structure. Mitred ends are recommended on shallow fill slopes. The ends should
be fixed at least 3m from the shoulder break point, wherever the fill slope falls within the
recovery area.

id) Projecting end -

6-
The projecting end is the simplest and most economical end structure. Hydraulically it is the \*
most unsound and some form of scour po reo
tcint may therefore be required at the outlet.

Cutoff walls are recommended in all caws where the culvert is not founded on rock.

4303.9 Wingwalls

Inlet and outlet wingwails should be flared at an angle of 45'to the culvert centre line on all
pipe culverts and minor box culverts up t o 1,2m span. The wingwalls on major structures
should preferably be flared a t an angle of 30°, as any increase in this flare angle will only
aggravate the potential-scoura t the outlet, particularly for large culverts.
 BOTSWANA ROAD DESIGN MANUAL page 4-30317
Date August 7982

4-303.10 Scour

4-303.10.1 Culverts

Unchecked erosion is a prime cause of culvert failure. The greatest scour potential i s at the
culvert outlet where high velocities may necessitate scour protection or energy dissipation.
Road embankments that are designed t o be overtopped i n a flood should also be protected
against scour, if this is deemed desirable.

4.303.10.2 Bridges

The need for scour proteciion can be minimized by locating bridges on stable tangentiel
reaches of rivers and by plac~ngfoundations on non-erodible materials. However, such a
solution is not always practicable, economic or desirable from the road alignment standpoint.

In such cases the designer is reminded t o check for local scour a t bridge sites, which is caused
by macro-turbulence resulting from the concentration of energy. Potential scour around pien
and abutments should also be checked and allowed for, i f necessary. Where scour aroqnd the.
.-
abutments of major bridges is likely t o be a serious problem, particularly where thezbdilge is ..
sited in a wide flood plain, consideration should be given t o the allowance of gui'd@>banKs
'"

or spur dykes.
~~~

Before finally fixing the level of the pier and abutment foundation footings, consideration
should also be given t o the possible shifting of the river channel during a flood.

4-303.10.3 Protection
:.#
Scour can be eliminated by the effective use of :

(a) gabion mattresses

hand placed or dumped rip-rap

stone pitching (plain or grouted)

(d) concrete (in-situ or interlocking blocks)

(el cutoff walls at the culvert exit.

Embankments and training works should always be protected against. undermining.

The choice of protection to be used should be determined by material availability and ax%
) atzhe site.
BOTSWANA ROAD DESIGN MANUAL Page 440111
Date August 1982 I

4400 DRAIPdAGE OF ROAD PRISM

4-401 GENERAL

Previous sections of this chapter refer t o conveying stormwater from one side of the road
t o the other, as efficienrly as possible. This section concentrates on the control of stormwater
runoff from the road pavement, and on diverting stormwater from outside the road prism
away from the road. The control of stormwater from the road surface should be by means
of channels and gutters while catchwater drains, mitre banks and berms are used outside the
road prism.
 BOTSWAMA ROAD DESIGN MANUAL pa@ 44021'8
Date August 1982

4402 DRAINAGE CHANNELS

Drainage channels include gutters formed by kerbs and asphalt berms, chues, side drain
ditches, toc of fill channels, channels formed by catchwater berms and mitre banks. Details
of dzinage channels are covered under Section 3-206.5 in the "Geometric Design Parameters".

The design of the drainage channel to carry a given discharge i s accomplished in two parts.
The first part involves deciding on a cross-section rhat will carry the design discharge on a
given slope. The second part of the design is the detsrrnination of the degree of protection
requiredto prevent or rninirnise erosion,

442.1 Hydraulics of Dainega Channels

P
In a drainage channel the discharge (Q), the depth of flow (dl and the velocity of flow (V)
depend on the channel shape, roughness factor (nl and slope IS). The relationship between
these is expressed in the Manning equation :

where V = channel velocity im/s) '4

n = Manning roughness factor (dm )
R = hydraulic radius = A/P (m)
A = channel cross-section area at depth d frnZI
P = wetted perimeter lm)
S = channel slop (MI.

The channel capacity is determined from the equation :

where Q = channel discharge (m3/sf

Fig. 4-402.1 gives selected values of n for use in the Manning equation.

Capacities of drains should be calculated allowingfor a maximum velocity of O,G&/s for earth
channels. The velocity should be reduwd when the soil conditions are extremely erodible.
This information should be obtainable from the road centre-line soil survey.
 -
BOTSWANA ROAD DESIGN MANUAL page 440212
Date August 1982

SUGGESTED MANNING'S ROUGHNESS COEFFICIENT

FIG. 4-402.7 FOR CULVERTS AND NATURAL STREAMS

Channel Material

Degree o f Irregularity

o f Streambed and

Variations o f
C h a n n e l Cross-section

R elatiue E f f e c t o f
0 brtruetions
(Debris, boulders, exposed roots etc.)

Vegetation

I -.
, = meander lsngth
..

-...

-
Y
..
83
W
Z
I
(3
I>
0
LIC
W
I-
3
-I
0
V)
Concrate : Centrifuged.

aded area generally applicable for

andsrd box cuhrenn.
mailer R for smaller ruiven and
 BOTSWANA ROAD DESIGN MANUAL hge -213
DBte August 7982

4402.2 Erosion Protection

The amount of erosion control and maintenance can be minirnised largely by the use of :

(a) Plat side slopes, rounded and blended with the natural terrain

(bl mitre banks and berms

I
(c) protective coverings such as concrete lining, stone pitching, grass sodding, and
hydroseeding.

Where t h e are not possible or economic, every attempt shouid be made to maintain a suf-

i ficiently flat channel grade. Alternatively rock cutoffs acting as weirs could be insttiled in
the channel at regular intervals.

The velocity should be checked against the maximum safe values for unprotected earth.
Tables 4-402.2A and B list suggested permissible velocities for various erodible linings and
for grass lined channels for maximum water depths of 1metre.

44023 align me^ and Grade

The road resewe width usually allows little choice i n the alignment or the grade of the channet,
.-
but abrupt changes i n alignment should be avoided.

A drainage channel should be graded t o produce velocities that neither erode nor cause depd-
tion i n the channel. Critical velocities shwld be avoided, if possible. In the flat semisrid
areas channels may have to be widened due to flat grades and wind blown deposition.

3
BOTSWANA ROAD DESIGN MANUAL Page 4-40214
Date August 1982 7

TABLE 4-402.2A PERMLSS1BLE VELOCITIES IN EXCAVATED CHANNELS

Permissible Velocity mlsec.

Soil Type or Lining
(no vegetation) Water carrying Water carrying
Clear water
fine silts sand and gravel

Fine sand (non-colloidal)

Sandy loam (non-colloidal)
Silt loam (non-colloidal)
Ordinary firm loam
Volcanic ash

Fine graver
Stiff clay (very colloidal)
Graded, loam to cobbles (non-colloidal)
Graded, silt t o cobbles (colloidal)
Alluvial silts (non-colloidal)

Alluvial silts (colloidal)

Coarse gravel (non-colloidal)
Cobbles and shingles
Shales and hard pans

Rock

PERMISSIBLE VELOCITIES FOR CHANNELS

TABLE 4402.28 LINED WITH VARIOUS GRASS COVERS
Permissible
Type o f Lining
Velocity mls

Well established grass on any good soil 1,8

Meadow type grass e.g. bluegrass

Bunch grasses, exposed soil between plants

Grains, stiff stemmed grasses that do not bend over under shallow flow 03
I
BOTSWANA ROAD DESIGN MANUAL Pap 4-403/1
Dane August f 982

4-403 DRA!NAGE OF ROAD PAVEMENT

It is important To ensure rapid and efficient discharge of stormwater runoff from a road
pavement.

Where roads are graded close t o the general ground ievel, sioiinwater runoff should be allowed
t o spill off the edge of the pavement.

Roads passing through cuttings should be provided with side drain channels for the purpose
of conveying runoff out of the cutting. Low points in cutting rnus be avoided, unless
trapped water can be satisfactorily led away.

High fills should be provided with asphalt type berms or concrete kero and channel coii;bins-
tions to prevent erosion of the fill slope. This is particularly necessary where the vegetation
is sparse and not lush enough to bind the soil adequately. Waxer collecting along such gutters
should be discharged at regular intervals into open chutes down the fill slope. Piped chutes
are not favoured.

The chute interval will depend on the gutter capacity. Capacity depends on the cross-section,
grade and roughness of the gutter. The method for calculating the channel capacity dan be
found in most textbooks on open channel flow, especially "Open Channel Hydraulics" by
Ven Te Chow 11964, Mc Graw-Hill).
BOTSWANA ROAD DESIGN MANUAL Page 4-40r1/9
Date August1982

4404 URBAN GUTTER DESIGN

A gutter flowing a t capacity should not flood more than 1.5rn into the carriageway.

The location and spacing of kerb inlets depend on the following factors :

(a) amount of runoff

(b) grade profile

(c) geometries of intersections

(d) width of flow limitations

(e) Inlet capacity.

In general, wherever kerb inlets are deemed desirable, they must be placed at all low points
in the gutter grade and at intersections t o prevent the gutter flow from crossing traffic lanes
of the intersecting road.

Where a kerbed roadway crones a bridge, the gutter flow should be intercepted and not
permitted to flow over the bridge.

On the continuous grades inlets should be spaced so as t o limit the spread of water on the road
pavement, that is when the gutter i s flowing at full capacity. The use of "side opening" kerb
inlets is recommended.

The exit of such water from the roadway is normally led onto natural ground, into side drain
channels or an underground drainage system. The latter will be the exception rather than
the rule.
L,. .
The inlet capacity should he calculated b y a remgnised method based nn free discharge under
weir or orifice conditions. For the length of kerb inlet J.A. Zwamborn's method described
in "Stormwater lnlet Design" and published in the Annual Journal of the Institution of
Municipal Engineers of Southem Africa, Vol. 1 No. 4, May 1966, is recommended. The
percentage flow past the stormwater inlet should be limited to 20% of the total approach flow
under design conditions. The apron slope adjacent t o the inlet should be increased for greater
hydraulic efficiency. For practical reasons it is recommended that kerb inlet lengths be
restricted t o a minimum of 1,Zm and a maximum of 9,Om.
 BOTSWANA ROAD DESIGN MANUAL Page U05il
Date August P982

4405 CHANNEL FREEBOARD

Freeboard in drainage channels is the difference in height between the flow surface level and
the top of the channel o r gutter.

In drainage channels alongside roadways the following freeboards shall apply :

(a) on side drains - lOOmm

(b) where mountable kerbs are used no freeboard should be allowed.

- (c) where barrier kerbs are used a freeboard c?f 50mm should be allowed.
BOTSWANA ROAD DESIGN MANUAL page 4-50111
Date August 1982 (

4-500 SUBSURFACE DRAINAGE

4-501 GENERAL

The aim of subsurface drainage is the removal of detrimental quantities of ground water,
to ensure stable roadbed and side slope conditions.

A satisfactory solution t o subsurface drainage problems requires a knowledge of geology and

an insight into soil mechanics.

Many variables and uncertainties exist regarding the actual subsurface conditions. In most
cases the need forthe installation of subsurface drz-llnagecan on!y be established on siie during
e.
the construction stage. , :.

4501.1 Graund Water

Ground water, as distinguished from capillary water, is free water occurring in a zone of
saturation below the ground surface. The rate at which ground water can be removed depends
on the effective hydraulic head and or. the permeability, depth, slope, thickness and extent
of the water;bearing formation. Field exploration is the most satisfactory method of deter-
mining the discharge capacity. Generally, exploration should be undertaken during the rainy
season when subsurface problems are most likely t o be evident.
 BOT-ANA ROAD DESIGN MANUAL paw 4-502/'1
Data Augue 1982

4-502 SUBSURFACE ORAsM TYPES

Two types of subsurface drain may be necessary i n road construction.

4502.1 Pipe Subsurface Drains

The standard pipe subsurface drain consiS of a pipe at the bottom of a narrow trench which
is backfilled around the pipe with filter material. It may be necessary for the filter material t o
be wrapped i n a filter fabric if the filter material does not meet the grading requirements. The
pipes may be perforated, slotted, porous o r open jointed.

This type should be used as a single subsurface drain along the tw of a cut t o intercept seep-
age, along the toe o f an embankment on the side from which ground water originates and
across the roadway at the downhill end of a cut.

4-502.2 Stabilization Trenches

These should usually be wider trenches with sloping sides and have a blanket of filter material
up t o 900mm thick on the bottom and side slopes of the trench. Filter fabric may be requir-
ed t o line the trench. At teast one subsurface pipe o f 200mm diameter should normalJy be
laid at the bottom of the trench.

Stabilization trenches may be required under side cut fills t o stabilize waterlogged fill founda-
tion areas that are well defined. They are usually linked t o herringbone configurations and t o
filter blankets.

3 French drains, which consist of a trench filled with unspecified crushed rock, have been found
to hi. i;nrt'H&le and are not recommended.
BOTSWANA ROAD DESIGN MANUAL page 4503/1
Date August 1982 (

4-503 DESIGN CRITERIA

The minimum inside pipe diameter for a pipe subsurface drain should normally not be less
than 1OOmm.

A 75mm diameter pipe could be considered under special circumstances. See Figs. 4-503A
and 5.

Surface drainage should not be permitted to discharge into a subsurface drain. The discharge
from a subsurface drain into a stormwater drain or culvert is permissible, provided the possibil-
ity of water backing up i n t ~ t h subsurface
e drain i s avoided.

!n general, the ssbsiirface drain grade shouid not be flatter than 0,5%. If this slope cannot be
achieved, an absolute minimum grade of 0,2596may be accepted. t...
The choice of depth and spacing of the subsurface drain is flexible and depends on the permea-
bility of the soil, the level of the water tafjie and the amount of drawdown needed to ensure
stability. Wherever practical a subsurface drain pipe should be set in the impervious layer
below the saturated layer. Table 4-503 gives suggested depths and spacing of subsurface
drains according to soil type. It is merely a guide t o supplement field observations and local
experience.

PROPOSED DEPTH AND SPACING OF LATERALS

TABLE 4-503 FOR VARIOUS TYPES OF SOIL
- -- -

- TEXTURE OF SO11 SPACING OF LATERALS (rn)

*

- I
I
I
%SHt I
I
%Clay 0.9m D W I
I
l z r n Deep 1
I
1,5m Deep I 1,Em
I
0-p
1
I Clean sand I 1:
Sandy loam 36.45
Loam
c1eyeq toam
Sandy day
Silt/clay
Clav
12-25
8.15
6-12
4- 9
6 max.
I1 15-30
9.18
8-75
6-12
8 max.
18.36
12-25
s . 18
8-16
12 max.
I
The start and outlet of pipe subsurface drains should be indicated on the surface by means of
markers. These are necessary for the maintenance personnel.

Design of subsurface drains is associated with soil type and seepage flow rate. The interela-
tionship between the permeable filter material, the filter fabric and the subsurface drain
pipe selected is important in the design of an effective suhurface drain. Cost could have a
major bearing on the type of subsurface drain used. All possibilities should therefore be
investigated.
 Page 4-503/2
BOTSWANA ROAD DESIGN MANUAL ~ ~ t AeU ~ U 1982
S~
BOTSWANA ROAD DESIGN MANUAL Page 4-503/3
Date August 1982 (

FIG. 4.5036 DETERMINATION OF PERFORATED PIPE DlAMETER AND OUTLET SPACING

NO- FINES CONCRETE

's
t%>.d

A = DESIGN INFILTRATION RATE lrnrn/h)

0 = WIDTH OF ROAD SVRFACE(m)
C = PIVOT LINE
D = DISTANCE BETWEEN OUTLETS l m )
E = PIVOT LlNE
F = PERFORATED PIPE DlA.(rnrn)
G PIPE GRADIENT (rn/m)
H = APPLICATION LIMIT OF PIPES 1
BOTSWANA ROAD DESIGN MANUAL Page Q5Mf'1
Date August 1982

4-504 MATERIALS

4-504.1 Filter Materials

Permeable filter materials for subsurface drains should consist of sand and/or crushed stone
plus gravel of suitable gradings. Grading specifications may have t o be varied t o suit avail-
abilityof material.

4-504.2 Filter Fabric

Synthetic fibre filter fabrics should also be considered where there is a definite engineering
and cost advantage, but preference should be given t o the use of local natural materials.

4504.3 Subsurface Pipes

Ccst, ease of handling and ease of laying should be the main considerations when specifying
subsurface pipes, which may consist of perforated, slotted or open jointed concrete, clay,
pitch fibre or plastic pipes.
BOTSWANA ROAD DESIGN MANUAL page 46Q111
Date August 9982 (

4-600 DRAINAGE THROUGH WATERWAY BRIDGES

4601 GENERAL

For the definitions of bridges refer t o Section 4-202.2.4(a).

In determining the size of bridge openings it i s important t o ensure that no damage be caused
t o adjoining upstream properties, and that erosion be avoided downstream. This will apply
especially to river crossings close to urban development, but a l l cases should be investigated.
Allowance should also be made for potential silting at bridge crossings.
 BOTSWANA ROAD DSSIGRI MARIUAL PW 660Z9
Date WuguSa 8982
J

4-602 HYDRAULIC REQUIREMENTS

4.6Q2.1 Return Pericd

To determine flood return periods for bridges see Section 4-202.2.4la).

4-602.2 Design Flood

The method used t o determine the runoff, rise i n water level. and velocity should be as set
our i n standard tex:book:.

Except in special cases, the attenuation effect of dams o n discharge and any possible flooding
caused b y failure o f damwalls should not be taken into account.

4-602.3 F l w d Velocity

The ratio of: velocity in the restricted waterway under the bridge t o the natural velocity
should preferably not exceed 1,7.

4602.4 Backwater

Backwater is the rise i n flood level caused by the constriction of the bridge strutture acmss
the river. The water elevation resulting from backwater should not exceed the following
limits :

3 (a) Bridges on Road Types 1.2 and 3 : a water-surface elevation higher ffian 300mm
below the road shoulder break point ar the low pairit in ?he road.

(b) Bridges on Road Types 4 and 5 : a watersurface elevation higher than l0Ornm
below the road shoulder break point at the low point in the road.

(c) In urban areas and i n special agricultural situations

- -- - - : as little as possible.
- - - - -

A method for determining backwater is described in "Hydraulics of Bridge Waterways" by

J.N. Bradley (19701.

4602.5 Freeboard t o B r i d s Soffits

Bridge freeboard is considered as the diffeerence i n level between the design high flood level
(DHFL)and the bridge soffit level at the lowest poinr on the bridge soffit on the upstream
side.
!
BOTSWANA RQAD DESIGN MANUAL PW 460212
Date August 1982 (

Fig. 4602.5 gives the minimum freeboard for rivers discharging up to 1 000 rn3/s and carrying
large debris, such as trees.

For floods exceeding 1 000 m3/s and carrying large debris, the following formula should be
applied :

Freeboard = 0,6m + d
15
(with a minimum of lrn)

where d is the natural unrestricted depth of flow at the bridge crossing.

If the river carries only small debris at least 300mrn freeboard should be provided in all cases.
 Page 4602/3
BOTSWANA ROAD DESIGN MANUAL Date A u ~ 1982
U ~
BOTSWANA ROAD DESIGN MANUAL Page 4-70011
Date August 1982 (

4-700 REFERENCES

(a "Highway Manual Part 1. Design". Federal Republic of Nigeria. 1973.

(b) Hall Dr. M. "Course on Road Design in Developing Countries [Road Drainage!".
Planning and Transport Research and Computation (International) Co. Ltd.
(PTRC). 1980.

(c) , Currie J. "Courie on Road Design in Developing Countries (Highway Design

Case Study)". PTRC. 1980.

(d) "Design Manual for Standard Box Culverts. Book 1 - Design". P?a:ioiial 1
Transport Commission in collaboration with Natal Provincial Administration.
R.S.A. 1981. *.

(4 Campbell A. "The Guide t o Botswana". Botswana. 1980.

(f) Thiel P.E. "Introduction of Storrnwater Management for Urban Drainage

Systems". Urban Stormwater Management Seminar C.S.I.R. R.S.A. 1981.

(9) "Guidelines on the Planning and Design of Township Roads and Stormwater
Drainage". The South African Institution of Civil Engineers. August 1981.

(hl "Drainage Manual". Natal Roads Department. R.S.A.

(i) Midgley D.C. and Pitman W.V. "A Depth - Duration Frequency Diagram for
Point Rainfall in Southern Africa". Report No. 2/78. Hydrological Research
Unit of the University o f the Witwatersrand. 1978. -
I
(j) Richardson B.F.C. and Midgley D.C. "Analysis of S.W.A. - Namibia Rainfall i,,.,,
Data". Report No. 3/79, Hydrological Research Unit of the University of the
Witwatersrand. 1979.

[k) Midgley D.C. "Design Flood Determination in South Africa". Report No. 1/72
(Amended). Hydrological Research Unit of the University of the Witwatersrand.
1979.

(1) Stephenson D. "Stormwater Hydrology and Drainage". R.S.A. 1981.

(m) Neill C.R. "Guide t o Bridge Hydraulics". Canada. 1973.

(n) KovSa Z.P. "Scour and Protective Measures a t Drainage Lnlorks - Guide for
Design". R.S.A. 1974.

(0) Ven Te Chow. "Open Channel Hydraulics". . 1964.

 BOTSWANA ROAD DESlGN MANUAL kga rl=100/2
I b)a@ August 1982

[P) Bradley J.N. "Hydraulics of Bridge Waterways". U.S. Bureau of Public Roads.
1970.

(4) Zwambom J.A. "Stormwater Inlet Design". Annual Journal of the Institution
of Municipai Engineers of Southern Africa. Yol. 1, No, 4, R.S.A. May 1966,

(rl Transvaal Provincial Administration. "Code of Procedure -Structures". R.S.A.

(5) Orange Free State Provincial Administration. "Code of Procedure". R.S.A.

(t) Forbes H.J.C. "Capacity of Lateral Stormwater Inlets". The Civil Engineer
in South Africa. Vol. 18, No.19, 1976.